From 96bc550eb916184e8ae95239c4151ab3fe68b904 Mon Sep 17 00:00:00 2001 From: Matthias Date: Tue, 20 Aug 2024 19:05:46 +0100 Subject: [PATCH] Deployed 8364fc1bd to 2023.4 in en with MkDocs 1.4.2 and mike 2.1.3 --- en/2023.4/404.html | 1124 +++ en/2023.4/advanced-backtesting/index.html | 1402 ++++ en/2023.4/advanced-hyperopt/index.html | 1522 ++++ en/2023.4/advanced-setup/index.html | 1540 ++++ .../assets/TokenBot-Freqtrade-banner.png | Bin 0 -> 61805 bytes en/2023.4/assets/binance_futures_settings.png | Bin 0 -> 81500 bytes en/2023.4/assets/ccxt-logo.svg | 3 + en/2023.4/assets/discord_notification.png | Bin 0 -> 48861 bytes en/2023.4/assets/freqai_DI.jpg | Bin 0 -> 314012 bytes en/2023.4/assets/freqai_algo.jpg | Bin 0 -> 353611 bytes en/2023.4/assets/freqai_algorithm-diagram.jpg | Bin 0 -> 501262 bytes en/2023.4/assets/freqai_dbscan.jpg | Bin 0 -> 67965 bytes en/2023.4/assets/freqai_doc_logo.svg | 304 + en/2023.4/assets/freqai_inlier-metric.jpg | Bin 0 -> 468744 bytes en/2023.4/assets/freqai_moving-window.jpg | Bin 0 -> 276574 bytes en/2023.4/assets/freqai_pytorch-diagram.png | Bin 0 -> 18611 bytes en/2023.4/assets/freqai_weight-factor.jpg | Bin 0 -> 189828 bytes en/2023.4/assets/freqtrade-screenshot.png | Bin 0 -> 145638 bytes en/2023.4/assets/freqtrade_poweredby.svg | 44 + en/2023.4/assets/frequi_url.png | Bin 0 -> 11391 bytes en/2023.4/assets/images/favicon.png | Bin 0 -> 1870 bytes .../assets/javascripts/bundle.51198bba.min.js | 29 + .../javascripts/bundle.51198bba.min.js.map | 8 + .../javascripts/lunr/min/lunr.ar.min.js | 1 + .../javascripts/lunr/min/lunr.da.min.js | 18 + .../javascripts/lunr/min/lunr.de.min.js | 18 + .../javascripts/lunr/min/lunr.du.min.js | 18 + .../javascripts/lunr/min/lunr.es.min.js | 18 + .../javascripts/lunr/min/lunr.fi.min.js | 18 + .../javascripts/lunr/min/lunr.fr.min.js | 18 + .../javascripts/lunr/min/lunr.hi.min.js | 1 + .../javascripts/lunr/min/lunr.hu.min.js | 18 + .../javascripts/lunr/min/lunr.it.min.js | 18 + .../javascripts/lunr/min/lunr.ja.min.js | 1 + .../javascripts/lunr/min/lunr.jp.min.js | 1 + .../javascripts/lunr/min/lunr.ko.min.js | 1 + .../javascripts/lunr/min/lunr.multi.min.js | 1 + .../javascripts/lunr/min/lunr.nl.min.js | 18 + .../javascripts/lunr/min/lunr.no.min.js | 18 + .../javascripts/lunr/min/lunr.pt.min.js | 18 + .../javascripts/lunr/min/lunr.ro.min.js | 18 + .../javascripts/lunr/min/lunr.ru.min.js | 18 + .../lunr/min/lunr.stemmer.support.min.js | 1 + .../javascripts/lunr/min/lunr.sv.min.js | 18 + .../javascripts/lunr/min/lunr.ta.min.js | 1 + .../javascripts/lunr/min/lunr.th.min.js | 1 + .../javascripts/lunr/min/lunr.tr.min.js | 18 + .../javascripts/lunr/min/lunr.vi.min.js | 1 + .../javascripts/lunr/min/lunr.zh.min.js | 1 + en/2023.4/assets/javascripts/lunr/tinyseg.js | 206 + en/2023.4/assets/javascripts/lunr/wordcut.js | 6708 +++++++++++++++++ .../workers/search.208ed371.min.js | 42 + .../workers/search.208ed371.min.js.map | 8 + en/2023.4/assets/plot-dataframe.png | Bin 0 -> 216176 bytes en/2023.4/assets/plot-dataframe2.png | Bin 0 -> 195009 bytes en/2023.4/assets/plot-profit.png | Bin 0 -> 145983 bytes .../assets/stylesheets/main.ded33207.min.css | 1 + .../stylesheets/main.ded33207.min.css.map | 1 + .../stylesheets/palette.a0c5b2b5.min.css | 1 + .../stylesheets/palette.a0c5b2b5.min.css.map | 1 + en/2023.4/assets/telegram_forcebuy.png | Bin 0 -> 17895 bytes en/2023.4/assets/tensorboard.jpg | Bin 0 -> 370209 bytes en/2023.4/assets/windows_install.png | Bin 0 -> 94230 bytes en/2023.4/backtesting/index.html | 2070 +++++ en/2023.4/bot-basics/index.html | 1386 ++++ en/2023.4/bot-usage/index.html | 1572 ++++ en/2023.4/configuration/index.html | 3082 ++++++++ en/2023.4/data-analysis/index.html | 1502 ++++ en/2023.4/data-download/index.html | 2021 +++++ en/2023.4/deprecated/index.html | 1523 ++++ en/2023.4/developer/index.html | 2196 ++++++ en/2023.4/docker_quickstart/index.html | 1700 +++++ en/2023.4/edge/index.html | 1979 +++++ en/2023.4/exchanges/index.html | 1910 +++++ en/2023.4/faq/index.html | 1902 +++++ en/2023.4/freqai-configuration/index.html | 1872 +++++ en/2023.4/freqai-developers/index.html | 1399 ++++ .../freqai-feature-engineering/index.html | 1694 +++++ en/2023.4/freqai-parameter-table/index.html | 1753 +++++ .../freqai-reinforcement-learning/index.html | 1632 ++++ en/2023.4/freqai-running/index.html | 1569 ++++ en/2023.4/freqai/index.html | 1496 ++++ en/2023.4/hyperopt/index.html | 2694 +++++++ en/2023.4/images/logo.png | Bin 0 -> 11025 bytes en/2023.4/includes/pairlists/index.html | 1653 ++++ en/2023.4/includes/pricing/index.html | 1488 ++++ en/2023.4/includes/protections/index.html | 1436 ++++ en/2023.4/index.html | 1490 ++++ en/2023.4/installation/index.html | 2187 ++++++ en/2023.4/javascripts/config.js | 12 + en/2023.4/leverage/index.html | 1634 ++++ en/2023.4/overrides/main.html | 44 + en/2023.4/plotting/index.html | 1666 ++++ en/2023.4/plugins/index.html | 2209 ++++++ en/2023.4/producer-consumer/index.html | 1507 ++++ en/2023.4/requirements-docs.txt | 6 + en/2023.4/rest-api/index.html | 2023 +++++ en/2023.4/sandbox-testing/index.html | 1505 ++++ en/2023.4/search/search_index.json | 1 + en/2023.4/sitemap.xml | 238 + en/2023.4/sitemap.xml.gz | Bin 0 -> 594 bytes en/2023.4/sql_cheatsheet/index.html | 1585 ++++ en/2023.4/stoploss/index.html | 1714 +++++ en/2023.4/strategy-advanced/index.html | 1570 ++++ en/2023.4/strategy-callbacks/index.html | 2455 ++++++ en/2023.4/strategy-customization/index.html | 2904 +++++++ .../strategy_analysis_example/index.html | 1673 ++++ en/2023.4/strategy_migration/index.html | 2424 ++++++ en/2023.4/stylesheets/ft.extra.css | 13 + en/2023.4/telegram-usage/index.html | 2191 ++++++ en/2023.4/trade-object/index.html | 1688 +++++ en/2023.4/updating/index.html | 1347 ++++ en/2023.4/utils/index.html | 2553 +++++++ en/2023.4/webhook-config/index.html | 1652 ++++ en/2023.4/windows_installation/index.html | 1355 ++++ en/versions.json | 5 + 116 files changed, 94405 insertions(+) create mode 100644 en/2023.4/404.html create mode 100644 en/2023.4/advanced-backtesting/index.html create mode 100644 en/2023.4/advanced-hyperopt/index.html create mode 100644 en/2023.4/advanced-setup/index.html create mode 100644 en/2023.4/assets/TokenBot-Freqtrade-banner.png create mode 100644 en/2023.4/assets/binance_futures_settings.png create mode 100644 en/2023.4/assets/ccxt-logo.svg create mode 100644 en/2023.4/assets/discord_notification.png create mode 100644 en/2023.4/assets/freqai_DI.jpg create mode 100644 en/2023.4/assets/freqai_algo.jpg create mode 100644 en/2023.4/assets/freqai_algorithm-diagram.jpg create mode 100644 en/2023.4/assets/freqai_dbscan.jpg create mode 100644 en/2023.4/assets/freqai_doc_logo.svg create mode 100644 en/2023.4/assets/freqai_inlier-metric.jpg create mode 100644 en/2023.4/assets/freqai_moving-window.jpg create mode 100644 en/2023.4/assets/freqai_pytorch-diagram.png create mode 100644 en/2023.4/assets/freqai_weight-factor.jpg create mode 100644 en/2023.4/assets/freqtrade-screenshot.png create mode 100644 en/2023.4/assets/freqtrade_poweredby.svg create mode 100644 en/2023.4/assets/frequi_url.png create mode 100644 en/2023.4/assets/images/favicon.png create mode 100644 en/2023.4/assets/javascripts/bundle.51198bba.min.js create mode 100644 en/2023.4/assets/javascripts/bundle.51198bba.min.js.map create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.ar.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.da.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.de.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.du.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.es.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.fi.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.fr.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.hi.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.hu.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.it.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.ja.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.jp.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.ko.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.multi.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.nl.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.no.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.pt.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.ro.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.ru.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.stemmer.support.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.sv.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.ta.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.th.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.tr.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.vi.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/min/lunr.zh.min.js create mode 100644 en/2023.4/assets/javascripts/lunr/tinyseg.js create mode 100644 en/2023.4/assets/javascripts/lunr/wordcut.js create mode 100644 en/2023.4/assets/javascripts/workers/search.208ed371.min.js create mode 100644 en/2023.4/assets/javascripts/workers/search.208ed371.min.js.map create mode 100644 en/2023.4/assets/plot-dataframe.png create mode 100644 en/2023.4/assets/plot-dataframe2.png create mode 100644 en/2023.4/assets/plot-profit.png create mode 100644 en/2023.4/assets/stylesheets/main.ded33207.min.css create mode 100644 en/2023.4/assets/stylesheets/main.ded33207.min.css.map create mode 100644 en/2023.4/assets/stylesheets/palette.a0c5b2b5.min.css create mode 100644 en/2023.4/assets/stylesheets/palette.a0c5b2b5.min.css.map create mode 100644 en/2023.4/assets/telegram_forcebuy.png create mode 100644 en/2023.4/assets/tensorboard.jpg create mode 100644 en/2023.4/assets/windows_install.png create mode 100644 en/2023.4/backtesting/index.html create mode 100644 en/2023.4/bot-basics/index.html create mode 100644 en/2023.4/bot-usage/index.html create mode 100644 en/2023.4/configuration/index.html create mode 100644 en/2023.4/data-analysis/index.html create mode 100644 en/2023.4/data-download/index.html create mode 100644 en/2023.4/deprecated/index.html create mode 100644 en/2023.4/developer/index.html create mode 100644 en/2023.4/docker_quickstart/index.html create mode 100644 en/2023.4/edge/index.html create mode 100644 en/2023.4/exchanges/index.html create mode 100644 en/2023.4/faq/index.html create mode 100644 en/2023.4/freqai-configuration/index.html create mode 100644 en/2023.4/freqai-developers/index.html create mode 100644 en/2023.4/freqai-feature-engineering/index.html create mode 100644 en/2023.4/freqai-parameter-table/index.html create mode 100644 en/2023.4/freqai-reinforcement-learning/index.html create mode 100644 en/2023.4/freqai-running/index.html create mode 100644 en/2023.4/freqai/index.html create mode 100644 en/2023.4/hyperopt/index.html create mode 100644 en/2023.4/images/logo.png create mode 100644 en/2023.4/includes/pairlists/index.html create mode 100644 en/2023.4/includes/pricing/index.html create mode 100644 en/2023.4/includes/protections/index.html create mode 100644 en/2023.4/index.html create mode 100644 en/2023.4/installation/index.html create mode 100644 en/2023.4/javascripts/config.js create mode 100644 en/2023.4/leverage/index.html create mode 100644 en/2023.4/overrides/main.html create mode 100644 en/2023.4/plotting/index.html create mode 100644 en/2023.4/plugins/index.html create mode 100644 en/2023.4/producer-consumer/index.html create mode 100644 en/2023.4/requirements-docs.txt create mode 100644 en/2023.4/rest-api/index.html create mode 100644 en/2023.4/sandbox-testing/index.html create mode 100644 en/2023.4/search/search_index.json create mode 100644 en/2023.4/sitemap.xml create mode 100644 en/2023.4/sitemap.xml.gz create mode 100644 en/2023.4/sql_cheatsheet/index.html create mode 100644 en/2023.4/stoploss/index.html create mode 100644 en/2023.4/strategy-advanced/index.html create mode 100644 en/2023.4/strategy-callbacks/index.html create mode 100644 en/2023.4/strategy-customization/index.html create mode 100644 en/2023.4/strategy_analysis_example/index.html create mode 100644 en/2023.4/strategy_migration/index.html create mode 100644 en/2023.4/stylesheets/ft.extra.css create mode 100644 en/2023.4/telegram-usage/index.html create mode 100644 en/2023.4/trade-object/index.html create mode 100644 en/2023.4/updating/index.html create mode 100644 en/2023.4/utils/index.html create mode 100644 en/2023.4/webhook-config/index.html create mode 100644 en/2023.4/windows_installation/index.html diff --git a/en/2023.4/404.html b/en/2023.4/404.html new file mode 100644 index 000000000..60e334627 --- /dev/null +++ b/en/2023.4/404.html @@ -0,0 +1,1124 @@ + + + + + + + + + + + + + + + + + + Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + +
+
+ +

404 - Not found

+ +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/advanced-backtesting/index.html b/en/2023.4/advanced-backtesting/index.html new file mode 100644 index 000000000..65ee457d1 --- /dev/null +++ b/en/2023.4/advanced-backtesting/index.html @@ -0,0 +1,1402 @@ + + + + + + + + + + + + + + + + + + + + + + + + Backtest analysis - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ + + + Skip to content + + +
+
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Advanced Backtesting Analysis

+

Analyze the buy/entry and sell/exit tags

+

It can be helpful to understand how a strategy behaves according to the buy/entry tags used to +mark up different buy conditions. You might want to see more complex statistics about each buy and +sell condition above those provided by the default backtesting output. You may also want to +determine indicator values on the signal candle that resulted in a trade opening.

+
+

Note

+

The following buy reason analysis is only available for backtesting, not hyperopt.

+
+

We need to run backtesting with the --export option set to signals to enable the exporting of +signals and trades:

+
freqtrade backtesting -c <config.json> --timeframe <tf> --strategy <strategy_name> --timerange=<timerange> --export=signals
+
+

This will tell freqtrade to output a pickled dictionary of strategy, pairs and corresponding +DataFrame of the candles that resulted in buy signals. Depending on how many buys your strategy +makes, this file may get quite large, so periodically check your user_data/backtest_results +folder to delete old exports.

+

Before running your next backtest, make sure you either delete your old backtest results or run +backtesting with the --cache none option to make sure no cached results are used.

+

If all goes well, you should now see a backtest-result-{timestamp}_signals.pkl file in the +user_data/backtest_results folder.

+

To analyze the entry/exit tags, we now need to use the freqtrade backtesting-analysis command +with --analysis-groups option provided with space-separated arguments (default 0 1 2):

+
freqtrade backtesting-analysis -c <config.json> --analysis-groups 0 1 2 3 4 5
+
+

This command will read from the last backtesting results. The --analysis-groups option is +used to specify the various tabular outputs showing the profit fo each group or trade, +ranging from the simplest (0) to the most detailed per pair, per buy and per sell tag (4):

+
    +
  • 1: profit summaries grouped by enter_tag
  • +
  • 2: profit summaries grouped by enter_tag and exit_tag
  • +
  • 3: profit summaries grouped by pair and enter_tag
  • +
  • 4: profit summaries grouped by pair, enter_ and exit_tag (this can get quite large)
  • +
  • 5: profit summaries grouped by exit_tag
  • +
+

More options are available by running with the -h option.

+

Using export-filename

+

Normally, backtesting-analysis uses the latest backtest results, but if you wanted to go +back to a previous backtest output, you need to supply the --export-filename option. +You can supply the same parameter to backtest-analysis with the name of the final backtest +output file. This allows you to keep historical versions of backtest results and re-analyse +them at a later date:

+
freqtrade backtesting -c <config.json> --timeframe <tf> --strategy <strategy_name> --timerange=<timerange> --export=signals --export-filename=/tmp/mystrat_backtest.json
+
+

You should see some output similar to below in the logs with the name of the timestamped +filename that was exported:

+
2022-06-14 16:28:32,698 - freqtrade.misc - INFO - dumping json to "/tmp/mystrat_backtest-2022-06-14_16-28-32.json"
+
+

You can then use that filename in backtesting-analysis:

+
freqtrade backtesting-analysis -c <config.json> --export-filename=/tmp/mystrat_backtest-2022-06-14_16-28-32.json
+
+

Tuning the buy tags and sell tags to display

+

To show only certain buy and sell tags in the displayed output, use the following two options:

+
--enter-reason-list : Space-separated list of enter signals to analyse. Default: "all"
+--exit-reason-list : Space-separated list of exit signals to analyse. Default: "all"
+
+

For example:

+
freqtrade backtesting-analysis -c <config.json> --analysis-groups 0 2 --enter-reason-list enter_tag_a enter_tag_b --exit-reason-list roi custom_exit_tag_a stop_loss
+
+

Outputting signal candle indicators

+

The real power of freqtrade backtesting-analysis comes from the ability to print out the indicator +values present on signal candles to allow fine-grained investigation and tuning of buy signal +indicators. To print out a column for a given set of indicators, use the --indicator-list +option:

+
freqtrade backtesting-analysis -c <config.json> --analysis-groups 0 2 --enter-reason-list enter_tag_a enter_tag_b --exit-reason-list roi custom_exit_tag_a stop_loss --indicator-list rsi rsi_1h bb_lowerband ema_9 macd macdsignal
+
+

The indicators have to be present in your strategy's main DataFrame (either for your main +timeframe or for informative timeframes) otherwise they will simply be ignored in the script +output.

+

Filtering the trade output by date

+

To show only trades between dates within your backtested timerange, supply the usual timerange option in YYYYMMDD-[YYYYMMDD] format:

+
--timerange : Timerange to filter output trades, start date inclusive, end date exclusive. e.g. 20220101-20221231
+
+

For example, if your backtest timerange was 20220101-20221231 but you only want to output trades in January:

+
freqtrade backtesting-analysis -c <config.json> --timerange 20220101-20220201
+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/advanced-hyperopt/index.html b/en/2023.4/advanced-hyperopt/index.html new file mode 100644 index 000000000..e5af2283a --- /dev/null +++ b/en/2023.4/advanced-hyperopt/index.html @@ -0,0 +1,1522 @@ + + + + + + + + + + + + + + + + + + + + + + + + Advanced Hyperopt - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ + + + Skip to content + + +
+
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Advanced Hyperopt

+

This page explains some advanced Hyperopt topics that may require higher +coding skills and Python knowledge than creation of an ordinal hyperoptimization +class.

+

Creating and using a custom loss function

+

To use a custom loss function class, make sure that the function hyperopt_loss_function is defined in your custom hyperopt loss class. +For the sample below, you then need to add the command line parameter --hyperopt-loss SuperDuperHyperOptLoss to your hyperopt call so this function is being used.

+

A sample of this can be found below, which is identical to the Default Hyperopt loss implementation. A full sample can be found in userdata/hyperopts.

+
from datetime import datetime
+from typing import Any, Dict
+
+from pandas import DataFrame
+
+from freqtrade.constants import Config
+from freqtrade.optimize.hyperopt import IHyperOptLoss
+
+TARGET_TRADES = 600
+EXPECTED_MAX_PROFIT = 3.0
+MAX_ACCEPTED_TRADE_DURATION = 300
+
+class SuperDuperHyperOptLoss(IHyperOptLoss):
+    """
+    Defines the default loss function for hyperopt
+    """
+
+    @staticmethod
+    def hyperopt_loss_function(results: DataFrame, trade_count: int,
+                               min_date: datetime, max_date: datetime,
+                               config: Config, processed: Dict[str, DataFrame],
+                               backtest_stats: Dict[str, Any],
+                               *args, **kwargs) -> float:
+        """
+        Objective function, returns smaller number for better results
+        This is the legacy algorithm (used until now in freqtrade).
+        Weights are distributed as follows:
+        * 0.4 to trade duration
+        * 0.25: Avoiding trade loss
+        * 1.0 to total profit, compared to the expected value (`EXPECTED_MAX_PROFIT`) defined above
+        """
+        total_profit = results['profit_ratio'].sum()
+        trade_duration = results['trade_duration'].mean()
+
+        trade_loss = 1 - 0.25 * exp(-(trade_count - TARGET_TRADES) ** 2 / 10 ** 5.8)
+        profit_loss = max(0, 1 - total_profit / EXPECTED_MAX_PROFIT)
+        duration_loss = 0.4 * min(trade_duration / MAX_ACCEPTED_TRADE_DURATION, 1)
+        result = trade_loss + profit_loss + duration_loss
+        return result
+
+

Currently, the arguments are:

+
    +
  • results: DataFrame containing the resulting trades. + The following columns are available in results (corresponds to the output-file of backtesting when used with --export trades):
    +pair, profit_ratio, profit_abs, open_date, open_rate, fee_open, close_date, close_rate, fee_close, amount, trade_duration, is_open, exit_reason, stake_amount, min_rate, max_rate, stop_loss_ratio, stop_loss_abs
  • +
  • trade_count: Amount of trades (identical to len(results))
  • +
  • min_date: Start date of the timerange used
  • +
  • min_date: End date of the timerange used
  • +
  • config: Config object used (Note: Not all strategy-related parameters will be updated here if they are part of a hyperopt space).
  • +
  • processed: Dict of Dataframes with the pair as keys containing the data used for backtesting.
  • +
  • backtest_stats: Backtesting statistics using the same format as the backtesting file "strategy" substructure. Available fields can be seen in generate_strategy_stats() in optimize_reports.py.
  • +
+

This function needs to return a floating point number (float). Smaller numbers will be interpreted as better results. The parameters and balancing for this is up to you.

+
+

Note

+

This function is called once per epoch - so please make sure to have this as optimized as possible to not slow hyperopt down unnecessarily.

+
+
+

*args and **kwargs

+

Please keep the arguments *args and **kwargs in the interface to allow us to extend this interface in the future.

+
+

Overriding pre-defined spaces

+

To override a pre-defined space (roi_space, generate_roi_table, stoploss_space, trailing_space, max_open_trades_space), define a nested class called Hyperopt and define the required spaces as follows:

+
from freqtrade.optimize.space import Categorical, Dimension, Integer, SKDecimal
+
+class MyAwesomeStrategy(IStrategy):
+    class HyperOpt:
+        # Define a custom stoploss space.
+        def stoploss_space():
+            return [SKDecimal(-0.05, -0.01, decimals=3, name='stoploss')]
+
+        # Define custom ROI space
+        def roi_space() -> List[Dimension]:
+            return [
+                Integer(10, 120, name='roi_t1'),
+                Integer(10, 60, name='roi_t2'),
+                Integer(10, 40, name='roi_t3'),
+                SKDecimal(0.01, 0.04, decimals=3, name='roi_p1'),
+                SKDecimal(0.01, 0.07, decimals=3, name='roi_p2'),
+                SKDecimal(0.01, 0.20, decimals=3, name='roi_p3'),
+            ]
+
+        def generate_roi_table(params: Dict) -> Dict[int, float]:
+
+            roi_table = {}
+            roi_table[0] = params['roi_p1'] + params['roi_p2'] + params['roi_p3']
+            roi_table[params['roi_t3']] = params['roi_p1'] + params['roi_p2']
+            roi_table[params['roi_t3'] + params['roi_t2']] = params['roi_p1']
+            roi_table[params['roi_t3'] + params['roi_t2'] + params['roi_t1']] = 0
+
+            return roi_table
+
+        def trailing_space() -> List[Dimension]:
+            # All parameters here are mandatory, you can only modify their type or the range.
+            return [
+                # Fixed to true, if optimizing trailing_stop we assume to use trailing stop at all times.
+                Categorical([True], name='trailing_stop'),
+
+                SKDecimal(0.01, 0.35, decimals=3, name='trailing_stop_positive'),
+                # 'trailing_stop_positive_offset' should be greater than 'trailing_stop_positive',
+                # so this intermediate parameter is used as the value of the difference between
+                # them. The value of the 'trailing_stop_positive_offset' is constructed in the
+                # generate_trailing_params() method.
+                # This is similar to the hyperspace dimensions used for constructing the ROI tables.
+                SKDecimal(0.001, 0.1, decimals=3, name='trailing_stop_positive_offset_p1'),
+
+                Categorical([True, False], name='trailing_only_offset_is_reached'),
+        ]
+
+        # Define a custom max_open_trades space
+        def max_open_trades_space(self) -> List[Dimension]:
+            return [
+                Integer(-1, 10, name='max_open_trades'),
+            ]
+
+
+

Note

+

All overrides are optional and can be mixed/matched as necessary.

+
+

Dynamic parameters

+

Parameters can also be defined dynamically, but must be available to the instance once the * bot_start() callback has been called.

+
class MyAwesomeStrategy(IStrategy):
+
+    def bot_start(self, **kwargs) -> None:
+        self.buy_adx = IntParameter(20, 30, default=30, optimize=True)
+
+    # ...
+
+
+

Warning

+

Parameters created this way will not show up in the list-strategies parameter count.

+
+

Overriding Base estimator

+

You can define your own estimator for Hyperopt by implementing generate_estimator() in the Hyperopt subclass.

+
class MyAwesomeStrategy(IStrategy):
+    class HyperOpt:
+        def generate_estimator(dimensions: List['Dimension'], **kwargs):
+            return "RF"
+
+

Possible values are either one of "GP", "RF", "ET", "GBRT" (Details can be found in the scikit-optimize documentation), or "an instance of a class that inherits from RegressorMixin (from sklearn) and where the predict method has an optional return_std argument, which returns std(Y | x) along with E[Y | x]".

+

Some research will be necessary to find additional Regressors.

+

Example for ExtraTreesRegressor ("ET") with additional parameters:

+
class MyAwesomeStrategy(IStrategy):
+    class HyperOpt:
+        def generate_estimator(dimensions: List['Dimension'], **kwargs):
+            from skopt.learning import ExtraTreesRegressor
+            # Corresponds to "ET" - but allows additional parameters.
+            return ExtraTreesRegressor(n_estimators=100)
+
+

The dimensions parameter is the list of skopt.space.Dimension objects corresponding to the parameters to be optimized. It can be used to create isotropic kernels for the skopt.learning.GaussianProcessRegressor estimator. Here's an example:

+
class MyAwesomeStrategy(IStrategy):
+    class HyperOpt:
+        def generate_estimator(dimensions: List['Dimension'], **kwargs):
+            from skopt.utils import cook_estimator
+            from skopt.learning.gaussian_process.kernels import (Matern, ConstantKernel)
+            kernel_bounds = (0.0001, 10000)
+            kernel = (
+                ConstantKernel(1.0, kernel_bounds) * 
+                Matern(length_scale=np.ones(len(dimensions)), length_scale_bounds=[kernel_bounds for d in dimensions], nu=2.5)
+            )
+            kernel += (
+                ConstantKernel(1.0, kernel_bounds) * 
+                Matern(length_scale=np.ones(len(dimensions)), length_scale_bounds=[kernel_bounds for d in dimensions], nu=1.5)
+            )
+
+            return cook_estimator("GP", space=dimensions, kernel=kernel, n_restarts_optimizer=2)
+
+
+

Note

+

While custom estimators can be provided, it's up to you as User to do research on possible parameters and analyze / understand which ones should be used. +If you're unsure about this, best use one of the Defaults ("ET" has proven to be the most versatile) without further parameters.

+
+

Space options

+

For the additional spaces, scikit-optimize (in combination with Freqtrade) provides the following space types:

+
    +
  • Categorical - Pick from a list of categories (e.g. Categorical(['a', 'b', 'c'], name="cat"))
  • +
  • Integer - Pick from a range of whole numbers (e.g. Integer(1, 10, name='rsi'))
  • +
  • SKDecimal - Pick from a range of decimal numbers with limited precision (e.g. SKDecimal(0.1, 0.5, decimals=3, name='adx')). Available only with freqtrade.
  • +
  • Real - Pick from a range of decimal numbers with full precision (e.g. Real(0.1, 0.5, name='adx')
  • +
+

You can import all of these from freqtrade.optimize.space, although Categorical, Integer and Real are only aliases for their corresponding scikit-optimize Spaces. SKDecimal is provided by freqtrade for faster optimizations.

+
from freqtrade.optimize.space import Categorical, Dimension, Integer, SKDecimal, Real  # noqa
+
+
+

SKDecimal vs. Real

+

We recommend to use SKDecimal instead of the Real space in almost all cases. While the Real space provides full accuracy (up to ~16 decimal places) - this precision is rarely needed, and leads to unnecessary long hyperopt times.

+

Assuming the definition of a rather small space (SKDecimal(0.10, 0.15, decimals=2, name='xxx')) - SKDecimal will have 5 possibilities ([0.10, 0.11, 0.12, 0.13, 0.14, 0.15]).

+

A corresponding real space Real(0.10, 0.15 name='xxx') on the other hand has an almost unlimited number of possibilities ([0.10, 0.010000000001, 0.010000000002, ... 0.014999999999, 0.01500000000]).

+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/advanced-setup/index.html b/en/2023.4/advanced-setup/index.html new file mode 100644 index 000000000..a9342bc63 --- /dev/null +++ b/en/2023.4/advanced-setup/index.html @@ -0,0 +1,1540 @@ + + + + + + + + + + + + + + + + + + + + + + + + Advanced Post-installation Tasks - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ + + + Skip to content + + +
+
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Advanced Post-installation Tasks

+

This page explains some advanced tasks and configuration options that can be performed after the bot installation and may be uselful in some environments.

+

If you do not know what things mentioned here mean, you probably do not need it.

+

Running multiple instances of Freqtrade

+

This section will show you how to run multiple bots at the same time, on the same machine.

+

Things to consider

+
    +
  • Use different database files.
  • +
  • Use different Telegram bots (requires multiple different configuration files; applies only when Telegram is enabled).
  • +
  • Use different ports (applies only when Freqtrade REST API webserver is enabled).
  • +
+

Different database files

+

In order to keep track of your trades, profits, etc., freqtrade is using a SQLite database where it stores various types of information such as the trades you performed in the past and the current position(s) you are holding at any time. This allows you to keep track of your profits, but most importantly, keep track of ongoing activity if the bot process would be restarted or would be terminated unexpectedly.

+

Freqtrade will, by default, use separate database files for dry-run and live bots (this assumes no database-url is given in either configuration nor via command line argument). +For live trading mode, the default database will be tradesv3.sqlite and for dry-run it will be tradesv3.dryrun.sqlite.

+

The optional argument to the trade command used to specify the path of these files is --db-url, which requires a valid SQLAlchemy url. +So when you are starting a bot with only the config and strategy arguments in dry-run mode, the following 2 commands would have the same outcome.

+
freqtrade trade -c MyConfig.json -s MyStrategy
+# is equivalent to
+freqtrade trade -c MyConfig.json -s MyStrategy --db-url sqlite:///tradesv3.dryrun.sqlite
+
+

It means that if you are running the trade command in two different terminals, for example to test your strategy both for trades in USDT and in another instance for trades in BTC, you will have to run them with different databases.

+

If you specify the URL of a database which does not exist, freqtrade will create one with the name you specified. So to test your custom strategy with BTC and USDT stake currencies, you could use the following commands (in 2 separate terminals):

+
# Terminal 1:
+freqtrade trade -c MyConfigBTC.json -s MyCustomStrategy --db-url sqlite:///user_data/tradesBTC.dryrun.sqlite
+# Terminal 2:
+freqtrade trade -c MyConfigUSDT.json -s MyCustomStrategy --db-url sqlite:///user_data/tradesUSDT.dryrun.sqlite
+
+

Conversely, if you wish to do the same thing in production mode, you will also have to create at least one new database (in addition to the default one) and specify the path to the "live" databases, for example:

+
# Terminal 1:
+freqtrade trade -c MyConfigBTC.json -s MyCustomStrategy --db-url sqlite:///user_data/tradesBTC.live.sqlite
+# Terminal 2:
+freqtrade trade -c MyConfigUSDT.json -s MyCustomStrategy --db-url sqlite:///user_data/tradesUSDT.live.sqlite
+
+

For more information regarding usage of the sqlite databases, for example to manually enter or remove trades, please refer to the SQL Cheatsheet.

+

Multiple instances using docker

+

To run multiple instances of freqtrade using docker you will need to edit the docker-compose.yml file and add all the instances you want as separate services. Remember, you can separate your configuration into multiple files, so it's a good idea to think about making them modular, then if you need to edit something common to all bots, you can do that in a single config file. +

---
+version: '3'
+services:
+  freqtrade1:
+    image: freqtradeorg/freqtrade:stable
+    # image: freqtradeorg/freqtrade:develop
+    # Use plotting image
+    # image: freqtradeorg/freqtrade:develop_plot
+    # Build step - only needed when additional dependencies are needed
+    # build:
+    #   context: .
+    #   dockerfile: "./docker/Dockerfile.custom"
+    restart: always
+    container_name: freqtrade1
+    volumes:
+      - "./user_data:/freqtrade/user_data"
+    # Expose api on port 8080 (localhost only)
+    # Please read the https://www.freqtrade.io/en/latest/rest-api/ documentation
+    # before enabling this.
+     ports:
+     - "127.0.0.1:8080:8080"
+    # Default command used when running `docker compose up`
+    command: >
+      trade
+      --logfile /freqtrade/user_data/logs/freqtrade1.log
+      --db-url sqlite:////freqtrade/user_data/tradesv3_freqtrade1.sqlite
+      --config /freqtrade/user_data/config.json
+      --config /freqtrade/user_data/config.freqtrade1.json
+      --strategy SampleStrategy
+
+  freqtrade2:
+    image: freqtradeorg/freqtrade:stable
+    # image: freqtradeorg/freqtrade:develop
+    # Use plotting image
+    # image: freqtradeorg/freqtrade:develop_plot
+    # Build step - only needed when additional dependencies are needed
+    # build:
+    #   context: .
+    #   dockerfile: "./docker/Dockerfile.custom"
+    restart: always
+    container_name: freqtrade2
+    volumes:
+      - "./user_data:/freqtrade/user_data"
+    # Expose api on port 8080 (localhost only)
+    # Please read the https://www.freqtrade.io/en/latest/rest-api/ documentation
+    # before enabling this.
+    ports:
+      - "127.0.0.1:8081:8080"
+    # Default command used when running `docker compose up`
+    command: >
+      trade
+      --logfile /freqtrade/user_data/logs/freqtrade2.log
+      --db-url sqlite:////freqtrade/user_data/tradesv3_freqtrade2.sqlite
+      --config /freqtrade/user_data/config.json
+      --config /freqtrade/user_data/config.freqtrade2.json
+      --strategy SampleStrategy
+
+You can use whatever naming convention you want, freqtrade1 and 2 are arbitrary. Note, that you will need to use different database files, port mappings and telegram configurations for each instance, as mentioned above.

+

Configure the bot running as a systemd service

+

Copy the freqtrade.service file to your systemd user directory (usually ~/.config/systemd/user) and update WorkingDirectory and ExecStart to match your setup.

+
+

Note

+

Certain systems (like Raspbian) don't load service unit files from the user directory. In this case, copy freqtrade.service into /etc/systemd/user/ (requires superuser permissions).

+
+

After that you can start the daemon with:

+
systemctl --user start freqtrade
+
+

For this to be persistent (run when user is logged out) you'll need to enable linger for your freqtrade user.

+
sudo loginctl enable-linger "$USER"
+
+

If you run the bot as a service, you can use systemd service manager as a software watchdog monitoring freqtrade bot +state and restarting it in the case of failures. If the internals.sd_notify parameter is set to true in the +configuration or the --sd-notify command line option is used, the bot will send keep-alive ping messages to systemd +using the sd_notify (systemd notifications) protocol and will also tell systemd its current state (Running or Stopped) +when it changes.

+

The freqtrade.service.watchdog file contains an example of the service unit configuration file which uses systemd +as the watchdog.

+
+

Note

+

The sd_notify communication between the bot and the systemd service manager will not work if the bot runs in a Docker container.

+
+

Advanced Logging

+

On many Linux systems the bot can be configured to send its log messages to syslog or journald system services. Logging to a remote syslog server is also available on Windows. The special values for the --logfile command line option can be used for this.

+

Logging to syslog

+

To send Freqtrade log messages to a local or remote syslog service use the --logfile command line option with the value in the following format:

+
    +
  • --logfile syslog:<syslog_address> -- send log messages to syslog service using the <syslog_address> as the syslog address.
  • +
+

The syslog address can be either a Unix domain socket (socket filename) or a UDP socket specification, consisting of IP address and UDP port, separated by the : character.

+

So, the following are the examples of possible usages:

+
    +
  • --logfile syslog:/dev/log -- log to syslog (rsyslog) using the /dev/log socket, suitable for most systems.
  • +
  • --logfile syslog -- same as above, the shortcut for /dev/log.
  • +
  • --logfile syslog:/var/run/syslog -- log to syslog (rsyslog) using the /var/run/syslog socket. Use this on MacOS.
  • +
  • --logfile syslog:localhost:514 -- log to local syslog using UDP socket, if it listens on port 514.
  • +
  • --logfile syslog:<ip>:514 -- log to remote syslog at IP address and port 514. This may be used on Windows for remote logging to an external syslog server.
  • +
+

Log messages are send to syslog with the user facility. So you can see them with the following commands:

+
    +
  • tail -f /var/log/user, or
  • +
  • install a comprehensive graphical viewer (for instance, 'Log File Viewer' for Ubuntu).
  • +
+

On many systems syslog (rsyslog) fetches data from journald (and vice versa), so both --logfile syslog or --logfile journald can be used and the messages be viewed with both journalctl and a syslog viewer utility. You can combine this in any way which suites you better.

+

For rsyslog the messages from the bot can be redirected into a separate dedicated log file. To achieve this, add

+
if $programname startswith "freqtrade" then -/var/log/freqtrade.log
+
+

to one of the rsyslog configuration files, for example at the end of the /etc/rsyslog.d/50-default.conf.

+

For syslog (rsyslog), the reduction mode can be switched on. This will reduce the number of repeating messages. For instance, multiple bot Heartbeat messages will be reduced to a single message when nothing else happens with the bot. To achieve this, set in /etc/rsyslog.conf:

+
# Filter duplicated messages
+$RepeatedMsgReduction on
+
+

Logging to journald

+

This needs the cysystemd python package installed as dependency (pip install cysystemd), which is not available on Windows. Hence, the whole journald logging functionality is not available for a bot running on Windows.

+

To send Freqtrade log messages to journald system service use the --logfile command line option with the value in the following format:

+
    +
  • --logfile journald -- send log messages to journald.
  • +
+

Log messages are send to journald with the user facility. So you can see them with the following commands:

+
    +
  • journalctl -f -- shows Freqtrade log messages sent to journald along with other log messages fetched by journald.
  • +
  • journalctl -f -u freqtrade.service -- this command can be used when the bot is run as a systemd service.
  • +
+

There are many other options in the journalctl utility to filter the messages, see manual pages for this utility.

+

On many systems syslog (rsyslog) fetches data from journald (and vice versa), so both --logfile syslog or --logfile journald can be used and the messages be viewed with both journalctl and a syslog viewer utility. You can combine this in any way which suites you better.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/assets/TokenBot-Freqtrade-banner.png b/en/2023.4/assets/TokenBot-Freqtrade-banner.png new file mode 100644 index 0000000000000000000000000000000000000000..6087056b6e8faa9abe64188a8d22d1b006fc1895 GIT binary patch literal 61805 zcmcG!Rajix5;aN;2oMMf9wfLo?v~&#jY~ryxO?Ly!QI_GxN9T99RdV*ZQNb&+W*Nu z=f4m4@ve`D)y-OSR?k^;)TmJvq$n@(2893x0RiESl%yya0pV390>U$N~P)fvzh>%oeaP+s{I>=LuVa@dquMYC=( z0Fm}jp+L>~){1q4PFwtz1q9PzkLAgz!i{;PlX}tD#wXUQP&8KuLVHg5cNWExW2Pp- zP_!WDu%)oVIq_(TVCWbm*Y|ra%y9Z=k(8=>R8hPO?pFfHp4i-b=dfVzqEn=sns-MA z1Nk31rCNr|l!)cyyeOZD4W6NY<>jJM8x_-OHpi6GNwcSQK*vs_4vFVeRkbqMAa+zz z9^k1%+KczrJ*dU- zH7wl=8e^)P8oNb5$_pInn_mR3v}8RYAuq%fD4%*4<#e9?$INf`WBcR>QYUgZDN;~P zLvjDhm)QI7kH8fX^Sg_GDk0D$zGRvMDQmQ6o`myn`WJ3-;oFGN8jIMrC}F*|bsA%7 zuK+MrG-*%>UxhI%9K1we-4nR#f87%8H!hTuJ1v><2`8;qMi6^Hn||@Zks$V2&b~< zUfQ#g#+NAhoEt)#1g5fkhJ}Rl2A4+Pi44M^WgU~(SJ=`wSLS5r_MdqW6m`MS5BOuH ztd-*R+T9WBUTE*?dDsnqAV@5kM}7Pe=Ej%6HL3Y1Zav`tvx9sA1#1+;NJc7KfO5QS zcCgeN(@1u;?yu3m1Xd?Hj6ldiZMC{V4hdUjA{g)X)Tnhp^4_=*ygj^B8UrcJ%~qNI zk7%z!m(}@1*lD~5Z&CGU63-g;^T{AO%OPS+#C>ny#JV(K|22;vKA-F{A8dm{U|71k z+f|JI>KyCn-S3$BU0F>(zT|6!Is31+R1vu=m}^m3O|{6|94y$`8654<>(^+uP{CL;G|6h`v6}9xC5mGp{AH8jWEqA}6$#RX^6 zlr{xYKxv6gm;l+ckZd-`%0@|abZc4c!b{q+mnAQhdDBW6T0O+q#4fC7jYWT~Sg{Ulhlb}DH^ekL`|x$B858o!#tY#Vrlm34ta zljhZV&at{^$MZ|52#-tyvi~E2)I)z8pI@8Y=~~4nsc)Sn`M!@T74wzl?2$!fAKO3% z0gTxxIW$neuT3gJxT?2tKM+I0mL%(!a)(}pL-1a!-urjx;*dNYL^M=Yg}+v$RMCGY zCTHiq_1Ds4$w>Kh{-itapy+$v=tr_0n^_n51ZIe-SF8Th@%M4G-@(t~#G4d_>U-_hy) z-!r;Gs>tfo*#nkLepraei;S(Msle_3I>7x+Ok$2acd1`hMXARjNFY`fvZYh2|0Y+U zzton+$ld|+oS0fwu6|QDFvK4#ZL?`+tLYlL9v#ShC^Po=C;ZCk5d8Ib8iZp6dp?_g z58TSk&oDXEA(8m`8@j9doBV>QBZ!O!<1A zBBH`dBwFe3Tr^pH3IHS=RzY&~+>4sRbDyn2(CTUQ5zHdyIo!Ciii)L_l$t|$GTz;S z+XZ3bs&9&#qE(!_fLP(^z`j{lVWHBc7$?+U5U262GghR>&>)Ngl~k-m!a`%noRz{Z z3)k!|m!r6@Gp-CN{=5F{O=2Q3OMJV`HIUdsiv~&qawJ(ig{~3w-}M>QWOuBkHoFY5 zMDhH-)=F-WdiT55E4J-(uFNN!qB=E(ieKP0nb->-N!(>;thOMERx*gSw`Ac&1I%GX zHKWLeT+?^OTI0PDVWLs8oCu zZ<y->&l)uyfl$`}K`xu|X9NqT4 z>>@LJ0tCa_|0m2Y^$&j`jlV0On*&!&7~%fAhvgmQAnNxTe{x|Tr#64OEVR${HcbW{ zK8Lc~l{lfvqe&Sy9JI#duV&Wxk)TgItl=3fL`VGD)d zv1sT7CkvM85$5+x|8sSIx0(ua~cO#tA1^#2BLf=`ut zF=Qi^i!he%Av+~gRpCki%=5oi7g0?g7Rd||M$H6}g~OmH1~q7^*Wd18yqtrbg=;$2 zUJ^S3H%^W`d0oZC{T-rnjUZ5PAKjXyrfTkD{R`Z_x0I-4<^G#{9g>azU5xVUn%rGJ zT1zx8GvVwFV{Mh|z_t1&T~guPBC*MO^@j~5+(j-V3tmdCBl+#za2DjZ~9JrRwmAU?) zaw3ypg8p~OBSNET@y0(E)kJzE&knCg`Y8>`qHG^z7h`mLNjyf(4p&6pD*r*k;+Jkl z&B+X|rZq)iv4^euw#j2YIq>xSEW~0BMjefHRGPI{g+at)xY9cl97gQ@%==*-OdK&vlzkDw((Vb5 zV+{dyW=d~zd}T@E;XI6!y-%SU$r%5Rem;Z5{r4V$_!Fqg_g`S<=l;LJ_`m%3<1Tu% z)=1Peck9Ka@wIBa?~Gy?W* zuZVx9BG&L1ZDNUKmuNKo@rE3pjiLHzi@(lnG!dq+QW2uzAcaeNb*9cklG)EtXi7zkwl)oE*dT#Bs@?k2(np<7Z4cAwjY{I=IT{SA2jWfn(;iSMJEb-SMv~Q`!^P zal$%SlSx=?eRA|N6qE7`?G85BrXKieSVv}@9t5Lk8zrh{Z_-ao%i4t323+gVFz#Q^Y*2KG z{lnMIwRtb^uT6H)L>~(SH_IGKTNkr+G9H>{|KV!{VB5&nflHKH)l?zUt(f^&J9x{Z zNiY$`Iff^L!EGFpKOfU5=}W79uiH4GEk1p#{7AJ|g7Tssk;)StvNy=sXeZVZR(6rK zu(`0J@q+7u(%~88K>l_Q1T&4W(@)r6?sG{AImEMZXwy6Bul-BHezkyI`Zf)pI|yQ@ zIgNEs8lZ+(1AG+)E7cXn(P2Q>Do zu-yG5b}l8>3QPvFv~BG6^a!`h>`~1`VHL*G9!Y3Y;RpT{Ws9AgjKE&~b171VjBHX( z8MN{H9%Vz_KUJ2NeCN};>G@)h`>bLKhz{h?q>v;6oPir<59LSRlUZ6F4a4_4@5Sn{ z_CGFCo{X7ZHX^@z=pN?8%oAw0`U|$-+&hZX2j##dU^PZPTy^a@)yAB-;*gT(>Sq7R z7))*3dmn3z>^z2E3)_!&srL)Xh#74HzV$5N7xgxWX8Hdfocc|Eu!9uw>P%U1L9kwsib1WaZy4m47(OA5L8J4S%X6uQoWeDt zlugWq78##zn5Z0S29Ep-^xW;8+@B1{w%2vo7{0azO4`lVr8XZDARE|HnzLfKl>t5< zlo4DfaregOjBv}utu1g<{*0mgf zb{k4veT`54p=(;rkfHq*V?&j#aQsVFsE2s_qWbTcg@;?H)2tY=*yI$ri*Er1SNwC; z7i)QmS`IA!uDTY(bg|!x1BnW^7U2*0Jbq!MK^ZeVQ0fe0gH3G;tV{m>j;@hL9xgSD z8ypGJ97#-a=aCaE{zUuuU`rI@q5o~3Kk1RN$_kFKm>+V17jsWj0Vx%(4o0JM@Gs14P@6>MOA;2 znM9dsMb>1nsg#bD*Lvd+mGYMRU|A4cajDgE@HO8WghF@W^<7=HyD1(IZc$+Rp@AT> zNaY!AX27CIXJ147%nhg@^6w)4BKo?qV1VvQ0X@(Ycl0?TTQ^qiBX4|)$6us5B-;!h zb$U(9*Q6xd!6>uZ)hfvw)Sbnby{q7;GcQG z;sFbPzV-_iq#AC8X!6(gs3P5{XhhN(u10iy~(Ddr?J7_EL|U18pxroV0wj4 z5mawN{cqI3;PE$_y%RtLjBGH$0qJntZ}i;>{G$SW^G+$gPy1^;0tVX4e}l`)_kjP4 z=kpI*{PXlbR|IlMJc_g8qcN+IUG3-RcaaKvtSC7ZZe<2ik3DO1AD^}KmA_CoorK`N z$0p8hzA$X)j}v{p{n;xbPF3DFuOlb6KK^{VfuQmc(%(ic^V&}VyTrf- z@EC>;FLLT&kSJSa zao+ql#3~BA|FDm@r~gb?87mkNYDNY@3yQr7b4eJoX($~DHCq@Hg>5iw_N)x04t9%* zekIvG&`Esp>|eB}QL9MkBWvwXl@X)vZrWcr1_v4p10W(s4S-0sM(>Jf#wUM@X~^Zr z|7K*AB^Cl*mBv+AK~A)9%^ce%!5gepYvyP;ErWUVlvh9gyK@?}#jpj66Ed+9Nm!9T z>_xD(YL~qP{WRXjD#aW7#XQ9S{fLgX543c_HmBCIx^KkhJm;fWb(=$2)8}`Giby92%faJOmI>Dg9V-x{q}E)|MEwACY`ta1ZD0s9SUI_ z`R^}iyyy#e8}|Q^#!BmGBXX#<8_4g0A> z!ZpaFqEIPV?8#n9wMY?&j(>{wi{|&gQ|gLlO{ZwRy8Bm$CH{hJ8;dK0!V>jf?wYu`@TnGX8!Q*@)0H=A@B#kU9s)h zEj(r#cnn9uuxp!VHR5IvwoB}3cjGkjsm0I_G~WDJDOZvnQkR>S7s>;q6F($ks!*cN z4@dvuB8Ome^B#wsWtiOq4rZEw!_Lb0ughLSrLH0VSME9pP!JoN$xd|by&MV2QC8=j zo4fns^yhW~+mnu3qWP;cuAFT047@+N3~j86_%rKyE>7rVxvWC(UMr+|agzvJObTDcQX>Sc)S|u`u7~ZpwW!I=WzAo_|a#cCx6O z&;Gz^Ub|JVr^arkr9ju+Hg=pjFfjDEJE(V^f*GX`umzEC>!h|PYaFY3b*^Xr80O!; z&i2GmkIZ_R;8Agc2%e!=eE!xM%n^0TG>|-+H~)okic0-6>eRv1=^@p^a;MZs(UR|{ zYIYVT_jgL$^tnIRPFjH3`LogYpY9fB2g~YM(bhNfPESry-k^pDB8LQ`y+Ju;?}tj; zCyU64i^q?d3Ik{Ckw_i1%#4Rtw~X&!i1%Uh=*Ybd>D862zP{4hur`j&iKTavtX7LS zc95ujTki#l9|BU;DzA~J*v*F^#+oG@de086r(RXD$|d6iICN%)ASs%bB#+BV>9zQN zBGf~}0yN%CnLFEGs_~j2$lxDk)p~!<)pX2;esCd7ACH8jJTfw~>B8J`f1yYjKK1OG z*C9-$Shqimk`ttr6}i0V@(h9hV(LRoj7{F0SAryko(E6i;8&}otIU=3M#qy~nb8O? zywJLo&!3-ZLmAXuvk?3mB=~>nZ=FJoM@SvQFY>zxtl!Q@RbqG`@3oyeFcF4yoQ7!Q zwd*I1%YF!5e2R0dJp7nCx_VkMb;p^f+#?GvFX*G)SS)IW|3e{!83xVSwZgU#`3Ft@{YhLhJE4+Q0y~B@r=oBFT46j)KN$pj*pG4X>{lq&Ni$9D+=6Mquj z9Kg1iD(iWAZ=pZ?yp~F+?NjMgZh^^;liTVDmkmYRGmt`p~hH2KkbQj0wQ&9QLg zSg`0{;}G(w_tm>Bnf{QkkfN7oZaJtH=GUST&w+<#x#J-#o~0;BaVTh2&APpm88Xh)7s>Fp-mdZoE9u!pdcwJ-IAtT?ll}3_`9n2;4&0| zTzN^!p;7wCfm}??EF~q(F_x$f+r(K*76J&mzl?_svlfH#W4(@n!q|M+>yvF7QJ=02tl=O6`gMs1wR2>6aeliL&Rt6AAobG{Z>xlrZ=L45D zC1t)T3W_F~9?4O~apd)GoQJ39;fg6h7H*~vQn;ee&QL?CV(RPFMMX_(uceK)5J6{+ z?JZ;|A<@y`@$ssvY6n$1BSS+|j|)2MqL}T5B)vQZvmnuI#fG6p@~*0q3(x$T@(Y37 zeW8H%FBKXLb>uVkADnp8D#SqQE~i%r+oW7^cj)sjSe&ij=wjX%f>*ZX5#1G3)qSaF8(T`B;<(!s(O zZJO2`y?_1iVhr$bVlRFD_VTtH*$V_yY9w(2c-) z>w?^Q>w(QvO7D?k-Il5|mb2487HU?DRj#8q4(hU0)!ha_m03TS$#-|Moaesqaj~gB zrhnkt#h4indc<4fG3)=ncXY?mjSWhH+R6OxHF=!0r>7^cll4VcW_noiA8d%?x&gynDUEaxbAXJ>B@hg4dVHBE|yw)cUkSD@$PNn0{g5 zAG}aA0ZY*~^OHCoNIfBs$LS_6)iKxC_3UYz+oV90KgmWi(@(9YazMoDw|qQ-CqW5U z`uh8|PvH5r7IlFgAqYkBy?muV1||p{4dgZPl$Y1{@h-;qfY=se?H?W^!`a z^`vNtZ1I)?2`8o=BbGNa*YJd1czc+&>O#flzp5Li|FXS$TPIW-7 zlpw;z?Ndm6SgM|mP4negRGa)MS7(QaTKU~l%=WJxW!wy+3n4)^Q$@vJt z3GH7zc5Y6n!!xQ|z`htG4OfB=G|B>JA+Hxnj#jozU0ru6L#9i#VGnBHqFJ|*DWR}G zz0-Fmz4I7EOXbdKE(%oTNyA&H@bbq-$IdcHz^VPDadp4eLab4VH$hxNkyMPEPr&8j z@5)Uf6?5rwF3b5Qo1JJU6L~%Mq+y)H3-En7CN}n%_ou6N5%&OJL{}FO@V{G!*;@e^ zdhz6|sQ^DA+`@$VBF$<<0Me}%^QDsP#9NjsT-OiZ>_OAHz0?XO9296n4oYRo5(eqi z%eW0SGFl&CH#f6>H_%_x{bC_Pt-Kv>|;FlY^I@9swV_yijf=j>KpD!<*~Mvy~@mo|szR>LkwQKFnNtbO|& z9RsFo^Y-uydomj0daw|e7t0<}ZPTyEW^OL)h2*pD2*ADhzE1hZml8DII42v;0$f-Q zkB^O%bj<9haZJC*#XwjD^(f?eZ=i5u$!e@>EKQJN?T`6z=x8gX)}Z>qBV+BmCIe2+ znB(@!i3n1}(|PB3<@=r*>xr)6#tnF3+v9G1cZeDv$#4hJkXy}*UUVYv%LA?uDg{rN zYB5#lOeozL-2bzUMxA442|lVXsZaEDVZSN6)pDITJlrS3$%1dkhl@_s5A|wix_s|2 zE#Nw(zaD&bbO$pjmp>#Sj5h+PAXxF8SS={i0^8(C*kN*F^>ybuJS{H0q<_EpgAl@D zBa5|hN3Cdpb4Yh|LQ30t#D&GQrT4?INjfo);^rpkJ~l0l&*Jt60Bs=;SytWz!|99O z2a663QWI`5LS8VpFuE(}?&S;LT&efQ51!{WBCrD@fR7G6+Kpwau2Y5XOu|C$u5H}h zrb5lmxU3eeIQaw}yW!C`3ytf)%F>XuYpqXLH@J#b!W^9F3gjWz`x7?ugnWC={g=Y6 zE+=yv7;pydrqy+YC9|2I;Va=9V{PTNi-$LbU|J-T+jKSVaMSXR_9v?~!V-iay?E zpnv53q4E~>;3R#0^W7!&$k-bb7?jBEtTrM#Mmdd3lOdil<&ytt|E@P8!zo;Xn>l4z zoAU@?;M5yHig%0Ojb@!J>LZ9R4p-acV_n%px~m*|+fBqDZQz$LI-Hr9R59-M>KFN* z-aUwd44LujZ5Ej<&F1Sg4Gr2H8(SXT-#%_$B@kXGTla{tr5}K z`1qSG3z9`oN0EXF@2i8Oqh^jto~G56{Un(ji5B;bW2q|`&sWV$Ah(lEW1S{#TK&k# z;V?A#p38b>(;O+LzsBGuKS3{Xg>cc<_RRP+F@d^NySk-W#A$sX?gNv6(BWv_UK*2I z8l)-qcrFp+QKM>PW9`TNeIq?n3-e>Wp1O;D4M40~PS?7QS|R&yOKrspCKT?@_slYe z9NL6%%NMGSy5T+Pf=-`jmRi3K_Mzk{e65~sw(FD+QEr&{{Q2|hRO#Jy3X0!-5ViYb zN|0ALiO}7E%zL8j59qcM`MgcS_#`$ue}0%uFN zPJD;@V)N_0BZ6GF0`P5rpOT@}pmpPsw1=3Kw$q?>c>0HnGA9?%3lUC+L1JE){aKTx z#X6(%?GHO+`Lswimmw-Gqo{)FXu*w?(1XN0u{B})T*{sD!W+mmy!%?pFnd0 zPpdE)sa=IC+vwitD&E4~s3K0+5=$ zHRv#>@KW*|Y>z3YldNt8on?6ND_o})iINrSwnztrGs5_%@~syM(3NXP(;M0E&Pc8| zKQu4G9trN2Ow073 zQlkp-g8&~As=ZmcJyosUH3-EMqtY){bD6->`Asg{M>z#s!w+IEg-mhMVoaW>9CMlz zxPksI4X>XU*UR15%3E8vZntow)qX0yaXC#$V2UHsIu*6IpGztCas;*&{lLDFpP!$C zfni3U((gdWST?22HKy;%LA`9j9@}#n* zTJL8BWRwR0SK;eY)5$AqaHL($!|kC)(bWNXaf=%i^U&dbW>&ji^hcbq`xW8Hz|g9V zTB+Jj`e>#dIUC#Edc+nAdEGmDxWKUSV1^7=i5^VkoO`-#(POhaXr$aT4~&VAkAsPy zugIHEE~PM<%x%2|fYnjAtF3-(1eIr$<&&ul#k;ZT5M7&V>-HlD!6&72MO-G`4`<_D zgYsjlAPgD5n7vJC7!w zYdwYQE6}Q66;(fZpptz%xp>Dbz6P3iGKN%B!iJ<%qFW5C6w#}Q`0u5l-_H{o#WAPs z>BvNC=exU2tvD=7E!2kyK(*!@1sl_Q=p;bV{F{Wvu-0TUXX(>))*42+akcB!NE`9d z*jP*}!{Fof0||lL4${=mpMeVbEz=EV5DR(ko=h&CW*4j7*hct>?$*_8mD%V7ITEM2 zPJ&JUz`%p;IBrKdA-=!aVqO0PdB)G>%zWW206uSAcZX9h8Sb5ooSi{!^QGE_$^}Dd zZ7a(9&$|ICC;#Nw%Ix34ESuBgB+f68sdgmhE?!Ap|u#OwI898Y7+ePvBfD zeB>}!0T{Y*2qlon%*VHc;FgUH(6z8OJD$vi?ktZaax4wcJ=Xjg+8pFW9-Zs$)y@OQ zauNM%f8qUfduAVC1NHQ5xw}02hXn;!DgoH$EPbmNAkXmR;KL|17`7)R#GnB#bDyog zt2mHubw0QTHcp zD=n7LJ9kdjvb*#BfRDXgku(*L>cY9FwIu*Rg0y2wW7Z_WhmE}vh!)Zu zwK**VLo)A8vC7>w1jzB3n%PT?273fsm%8ku^4p#StGDap3KSqKPwc#Dk3K;e)1}&- z`jrZ4WZ+G!l4A|!iF$~?OocSs&4ukgDiH3hX|Yam7(bjoL!Ym81Qex+2=ye- zm}qO%wpO*m*t0Efowm%;)I2*d-p%j{s>DTn~>IIL*oc6R)*L^=8RP9`K}YTnR9%I}f9at5Rqy z>tLF^+84Y(O)P-KRWc{ZqKD8TqV139jbWQPt@p~~=EIRuP!do3`68~ix0&m%ucdtg z&Vi(kisaSD>^FPPHI}oceHDUbkA!uB-*iAn3wamE1-D%0I=zEd^3Rl#l6XkG&yDo6 z;%%ZNDN5%#lUOW-U0jftHL`?VV)8H2z3sMBN4<4`6VU#}kt3i3JKkv1e)fzB-TffU z#22?^Yz*^&uLh63VVcjzD>XT}y`E&|32I;<4S23c$gCwXz!7GD9KiuST)W(#ks5g! z4&E}jmiLYGLv4gSduUEbE zOnP;8Q=zZ!UzN6hxpo;y;IL2+!R6;hKX>z~{qico%5e++KAF9=*9Ij5xm3CS_P4$# zSIIJu)9FHL>Ui&2-)amXU~bd`o#cxp#%bP_HlvvHa`8q0G37C}Rfv;r#u~furC7c5 z$;deRMNJGni$q~T>ol2BnxSqCE!4_p3V`L$(Hqq-#nmB0y+1OTnrv21co_+W7yc3wJRxi-= zx^)1;z;&QD08tHB1yA?$4Dd$&O#^qk{>>@?iqfxF_d2F!EG<>$oriV3*OM@)yjmT4 zp9zW@G2`XSKUS$7nCr;)I)iYWmlpJV0U<@@_Wb8}AkV=>XZrZt8_-kk&OI&L@4J_W z?!q!oV^ULjZ#sF-wiIfIZ^z6->a6Fq-Dr{=_kGn7F?NRQf2s@&#LV6(7f)uWY`JE1L(o`q0a;2`I=T`*AxWPXe|l5%W`KZu})EFi!sTe1pc;0 z!Cl?i*sCKQ;;$>kVjoGd5ARHZ+kpelKja%j+klOfG<%6#T+(NwmV`gsLB9%!VQcGR zC?zUIgcF6z^!oqkNjHfdm2IPhH5vE!3#cLzVZ;<#p`5I&mLuvU{{Dz(j8rjZHyIyp zY@Vjd7fI2)0o@7^3&u#B+RskGa%q=`F(Ug^-yeaDkvYlW?s8b9Q~K!8O?bQ;6{KF6 z%IyWZaNgl)1diK8IlZ^TUVKXeD85Wri4L1Y-nCd_R6hGFDk?_OMYm|ce(th&ebyAVwxSaK3et#N-v;@*JsXwUz_jo*8FH=IpMxEppUhj zOqq_ApQmsZ3%Ty{E&WI)7ZYo;y6l&8KDr$wRS@kIck;S+&1Y_EkLfoY734D9t${)y zVF@APIS)4hMjf>jaw#0w8`W1H-S+K8Dg&=$bR7~ME3Ai|1^+l5><-7#bE4Ke@Bgu% z+7HU+(4%%v2}fS{vKq8g2Kot1 z^L?Y~iZmNQZSFnjgxh^9I^`T^Z`ul1zHEmpGezNsK4K{kc-eS3ohH=!Sh({?TOu1R z=M-pt2S}KF*($)wY>zs1O4k90^X0`Njy}{eQhF|CGZr?X?>LEtoVT{4>=|wy#OUH$ zFMHU4byYsUc=wKg{c4Za^Umo#Clz%;sgCY#bCWCEEa|i1m8_O!lB;0TV==0j7T|@^ zHg4G&m*Zb7kJts%_`WB-zYw0LN>jK=dIx1V95wXxLuv!B1p=7`tB3~>kYpO$l7^>cMP!hb_)E){~)6Z%D^K(R(nKSeX% z7YO`ppD%MBP*kro>JF06khJTaQ*;a9S0!MMsa54#p8-{a3{QQ@2gKTIl92A~k!cO; z7z;qwq9YW8T@DAJ?X6Q4W~tD5o*kH2OqC4OCk}4`yLnbDnf;Ya8>I?SgCQOuunoLp zj44+rFWy?saboA=KAerd=|}^zy&g}@;BTLv0GF3?5r~ORG@wu2k%sqb0dnwk%=4yf zKy%tv)%IglE$Rs;66%?s-*hDQ*H3qKRs?8Y{8ejnOG{5P@3`FT$|aFjjinJT7PsA6 ziS6o8hRif-J>tu2&-?x-Q6?i9e`x`CCj!oFJ=cKUeE@K!q-gKu;u3aCq3I6FY^r(S z*4tY)KpkJrr~!(|o71HwI+aolg7nP4pK@qzLUe`C0QGRj`wLWmFh2}NH2#9`_~ay= zd543`FCg&e=gzq&_aiZVWO5FU^NFb) z=#TXYF>GjTyGfrrrZathSFD(1kT6&_+6mC|6+6D_T@ncJiy0)N_icIHGQILnR?5jJ zQLg9pRt6N=ITs9RGRNI-k3lnuV- z0P6wep(Q9F;U1RXlYaA-c3hNHjUTdCOe;#V%z?58*5G3k{ao~pbJ#=d! z=4~v~^^Wm+66(ubI?Ncr%to^QKnvGwC_--G{y-`AI1+#(T{_APrSUJ!C~Z)uvq=wA z(~A0fTp`uRfz3T6jQhn}Z9cd_3PEQ6DuRw<8l#`iY8C+loVC7Iq%=9~;7{*I>L zlmc*}<@8er?K^Oi<4mBJk-Q=Dn>WM&BcLYn?ygT2bRrdCtTwQWJvel`^0XY>61Xq8 zf$*V+6Fmi=ad?0JZoIri-VslYc(LL=+Se!OJq3Dz2NobT(Bhg;R45$)q9-z2;{LpQ z4bX;=%{ORR_4uM`aC398ffOiM8px>VcQOw}W_f1kkS|<;X|R|gT-}b+J4YJ7F8+j2 zw}OqJxbv(@H8Rt6Fz7?Ko{)*rM&dBeJcq8D+TqRwuHW6&{$itXX%PrrO_nSw)YWQZ zoN|x9gnuoV8NZYxSf`mS4$KZe#|Jtp zv}%hR=XQgtO+c>;Hm>jP+!C$p+`Y)!isjb5xKM}=j?tZEaF2(9~-s8_VKqKH+8ou`w1#gyUnp}Nf9M8o2Uy# zj*^SXcedf1)ojGa564a$$OB|>;?~x=Itwa~YF40mC9%A&LQ*3@k}9Uodg4#a3tk3O zX_>(q#xLiWm*!&~*5>o~<+$>AeCuldB+4eWITj~RP@{K}~ES)}rMzp+Anrt*4RZg$bfG`;QxtweJWr%1lxbXN+ zgNuYOR=PXIMu2Z?Q_6!1Jx%^;GNH|*{)^PZEk_D_>2!&!4*&@-k&tMx)?%92SgZWR zG=R#dujcLG-W)b88VB&5XWt*w9YceIk>8-=Da_>^=h?Hn`cC{qhuUDoKPy-GcQ5$Bjdjx zAra6f(peo2?5kiccsBzu;M+ImK|P*2m~zakzebnU8y50Em9D-Zr0p+66fyhNC-iId z?DcctENj#or@%#=zOdsFp*XuGo`F@&&Hb4@R^ae;FYMP5`iXA9##g)W@;~_j;d5__ zkuTeDqJkv>Au)qHCMsQ3%c3kc{^DY4Z)2I;#iS+LIy4UH>S$x`tFl1N2&T`*Fn={r?Agps>}1u)3>jn-h=$_MUTSWTf-S*C_ zGh8Gjq@L-Y3)y0w&Ic*eWklr_daY$@7QU$0fbw;8ZJ`@)m|qM+c(>-Umgvr8?VU+T3aOkn1?ZYMa9LrZ>66GW5UDby?c}Sxy{-gz0scTRQ z{ZGT3S?ejTZHw*9q2e*N#%eXU9`C~eKvVa+Gx7tvH)_q}lGzKVvy=zjF3hiQV6xyB>F$9b zYX|TAj)%Nrwp4X<=CWzKW3IeMEWvac^4pAH>*EbYZT;T(={KtQQb5zz01iy6GSmFG46uV zf&PBJ)9L>5*89z!RhyRMg$5vxMkOccQ2nrA^33l({A{$6f&N~4SuKA2>=6Be$?@_< zu}))l#gwd<*LUVKGR!nyS3$7~!94lJ-Hxujy2iD1jGi~zb(l(3FnY26H{+)D6@6;- z(;{AvYk3Y&q~bn{iY?8mp}lcB_eY1qgS;=6L5EPTI88#`=Sxd;qcXv~hQ(uhI)c3$ zC`oW16`GBXkAU=nqYHmO_7yVk@B8wE3-!+{*h{|vN8X2?i@fgW7cU;%FAn!tAx{O{ zM`IRhR51hm=vqOi(3JST?;=$gIUH=^ofG@v{oHw@qvBzEk9DS(^H1(s=C#~#u(Th+ z0nm)#^=LMGmb2__ew|XRX6fug)=uQ2VQh;<1dq0h?({&sN^P}{5>xim%q{Z9P;0n% zz{7qeATyw)Gprs{jGs6Xfj|$In*vMj#Lzm^vOv?v;B8w#t+~AAwEcnDFHp>Ts1!_j z`4|rHna*mVr}km=b?R0(y?hDbq^c@=T@Py=H?4f6yLb$BBS_)#rJA4bOF-UGDo4DzaE+KwP| z13c>S`=n!4ZHhgEV^IrmYi0>(wnQZOGlpj44f3uGTGeqZ=?@>fxos; zxJM<7Zr(D41ML$SXT>Nc9~Pg*N$)5o6#OpRRlIST&5->3Ija-3_f5e26X&!Za5*;q2X}qIH3q^7< z1FCdxF5*wLy1G|eWIqPiJfKWZmh~5QgoY?vBgYCA?zQs1GEC+qw15`ix;a@OA;`wO zs_BI@CXF*OJrulhLKX& z*D1%phc;R8LhGSV`asiGY z;O4wP)#9m>hc{~S`~`mr(D#6JMe@F`#%g*WU5(dPX77>`_C}XzzlQmea_u?$69OeyS;t9q{fr` zk&W($)8`j7T}jDr_S9aZnksvYcMTcEl1-5XZWF_z7&nOW6@!MOL;~-F->SdB0$mtp zMDQCL6I0W}Yq&R*)CFW~Gx9*Q6->)zjr{fG&O9K%*Rdr!`zAX*;q=}(`_Gc=y~eV! zKJ(ugfF=>aaZOFLgjgv;K|LqQzBx(|^#p1vRKY(WJJjj)Ug5B^*twALjt1&3RN~3; zgt>J;dH=3zZH7Iy6nBr(&Iiv(rpmf7GR4n*E93*`Pw&g8_v??K37T+Rg#1^gN5Ct2(R30#x7TB~35lT`gl;_e zj4iC7r|{rDMZ}O<*Mnfl9qki|A%VM(!3TA~J{#Y#Q#FbP@rE|4-*JDl@e5A1&*Nww zsOI_jE%m2OQBNFac)Fo&E`P0fCvz>8!4&gQ0Km=U{PNObsY(ODkE@=0Er23UcbYJ+A3PG)9q){jdG6@_zE z6iD@mKy$5a&qnpTODSb`IK`%BWdyi>=f8J3lwO#VKrZJ5izublWR3Oq?fygu5LuQg z588L6<(H3F z?A89`cS;|^$oAwEWtvj$Lc^2use56waov;Y@%$;uF&brr+w02>JtzPe0CunGbRuu$ zt3>`j04#oDbFw=(Eav5Zl2s|yyz$lo;)&sx^9_l=hdhrCdh6 zjiH9?;Hed$_^=&t1`CKCUMu}rPsAo26szB9*M5=mb#F@lsaT3B>@O+>8ANTp35~r{ z6BjDq=X`NY>Kgdd2}^+D=Sqb-w25C(SuMo6bn&#^QtN4m*}zvnzlUS26ct1WBHMk7 z)U`1^GHZ*;GKES_w3N@$zwL+;hIfEW^40p6`UcrCU^bSf(p1BdaIo3Q@-lhf)S1lS z{;`Y#%1dNF>wqATMj)M%)3aaRojaL){bsT-n(m66;3fg$h^wY2HBg7$?vKb-HL1_r zXx%JymSUR{e3NSN%ow{>tXE~ma}^=I0;(G)RYmgyQ@sE<2pHF^Ebp{DB>1^4v<8MP>+;uFu^sT_N<>$Jy^VkV4vD z{ZD?DzUOT^X8dFBcFIRqA;IRXPCTv@`84-?=K|CS=bSPW^-PTiUDVY3O_Xw)CzaM) zwi|(^IXMto{=G{}RiO<%`t6Tu!l}5S7Jq(>9kmdPAwzYRHnU>d6+KP6J)?Vdo7t;3 zB(lb;zi`t$$#0xN2GypI$@47)p3)|>#<2=LXB(OWOm`kl-7Yc{J5VRneI2+#M;8XrCgPfo@GaqD_ik|tepNLQ8l zTxZLpodBcuk5g;Sepw(=$Ss7@4dbrGX+V$Nw)fa03DlkAko^kqkx3BrUir?y-Na*MIT_&U6o(POI zSLik8f2x+H*W_@V{*q6f@!bFBi6`GLOtJLuL}`ECn(8?{Adoe?Y(I6qkWP|kO`4|( z*mUlFX3+LHs^$2nXz%g>GNiH0FwdKV?V{lu8+gF1D5?({0_vdAdPb5V#pPh?VR>Q7 ze6C2Q4ItJ(9t+X&WHvy_j3h_KMOI@%>(=s1HjAsv^|}e9^KA1V!rOKb=@>{iw!Jk7#a!nCC`gvdO_tdF6 zuh?dLZGITcRjRyLs#IQ?etVJ5QML9@<9zZ(CX>t878v@IOse$5pK0hV8u29-f{ycp zu%~ie(ztfP6B7fza5FEhU9|-OEH5oPHcff(qtSk!_d9;`r0)p~)tuwoiMB;Qx?>rU zp_)WrS%bI~w+8Q=3}#lb2q8KA0&*Ry`jNOu^pe-F*c)&;>MZAje6}vEz7uf2n5u&frzR z^8qv~nT=^``^Bo&jk?@Z8>OTCoHg))-n1ETb%e;Nigsc08FipU1~4 zLbpbN{vzAwGysBQ5P6L436iZzxitFeW<4xbIB`?0fo%> z9SXPaK-uGXBZt-!@JaBoKym-XsB6;v$#C}e6I!-tJZ3$7JT_heRi`5?}UG@&c$d61tEQ@d0H_0 zJX5vYsnO3M*Y)S)bV&8W@zRY~om7=H2%Qhsi=4J)V#UQKov$Zs2HcjZ(md|=j2@p4 zR&EHjeyK!mG_SU&^6f0vwS4aD-~T#D zQz-dnt>CAK*4l@NeDEHprlNAbe0lT6(Z4rdj(bhRufD5>vX*$@(3|@Hy6_ukxJUP1 z&Ag=dk4BoI)*TO+Bb>7DmsGcud10@uun;HTn`-_h#CBo&USA2Z;zol$6INExL+(cuYJEbd0pT|J}YYE_)1x7m(&7{&tyDr-@I{hV$XP?qs_px zyuxsDa?%S2%>jop564O|(s6iT6Mt%ftfTsQ+E!G;SB_YHDj{LoX^L%+|Mc<2tny(2 zmF(W4WLxCo4hT8$1)Mb}F6EDui%rUZO`Cn4mV*!9>e<*)>FMvsXfs>d6&1;?%7Y0c zNnu#sxP)Eg=OY?8e;hJUu^Z3r^)eb>>$v>mcWz2Z4>3BmmOc?X{h~pG{f>WUY2N-* zEMblc)zNZ`OZjX}ZMNI^$y+P6*D(GzxDb*16Pvr=xzn{TAJJh@-_Mq3RmaxGE|-46 zC#sfwbG^FKO`8&l?Ow`enf;;@O}O;qa|eS?)X_)B7VrML<6=48eqREcB_cu3`_l|| zbvq|~t&@3&plB6c%?9(C5~mc7d430ni2Tq|cU<%W))qyYjrs4v)NgTg7zD^)43w+1 z9Y*FM=lQu8Jp5wy>|c&#j@abh$|YuGG#PLF(&7EOsooY2G^0hMxM7*@loUlzto?#4 zq!I(({(0t7QV&P(X$1`wbAw3kx=QmKHL7%UGAU%<2Rt8iAU<t=?D^{xH(HzD)0(Ih)t9FZ-}QqXFTdDqV*KlE_6FW@0*7HSw+e;M3-rs%VT>mY`*hq*u}ojX(CtB1$xgx2FQnJwygw7+sfF zS$;_Vm-G>3@)kvRkZ4Yj6OQ61dT8OuTy_vFrjQq0P=%-;+TWaomcp2SmXwr?&XK25 ztDe`W^fb}ia!Ox=D#Uzo8%vv?7XziSkUUkKq)YOlvm(*AM8S*8`*}(oELdZfR7tYW zva-mi_h$lhEEt#bWvRC8R+IW$yCFPN>1$}5zNEBPs<%YGZEi%vpJt7_ zZR|K2vuH?2xONBSP{!)&qf(*6#Nf_U!%Dk%tbBXatvRV*f z7lI^ld81I8cpp_HChqfDT5|RX1F^qc?WEo1QUQFKMoTp#4dy~j4bVAzoE{6VG3weQ zSFtpqIQj|FF&zcY=cMcH$B#zwUD!KrzlVl0c}za9cnTaZa$a7QbsMuN#!riistE9g zK;;&BpyJOWB2r(L_!7Qwx?H7mZYQ%o?x*m~KaCu$*86yM?SB1wlO|2u(_`Rf%Rd|U z-D0^wcg?pS#lUC)70C<#{O<$&HEG8{76R@HJ3Jk|;L|u- zwdkr|tNkahls2D32X%GhYF)x~j%qtIGw6?!y0X6V@_&OK$DN&R@Yt((t=+RRIUrzq ziblQHff}NO zwRWL-_#N(k9HofNQ>$L~dgN#~W9<4R<85>7UV}L{IOvCsdvVrtIkGt?3jF4yK}9nb zO`+PU>>KF7#I^DCtYr4%?81as&N7FDRt^sB5RueuwKJym`#4dY8n{wG8117Zw}nMj zX37}kO17C@m+6dWhk1IaOyJI`)BT#*@Hrz)d?E1Ff(~PWrfN5sZ&l!%2|dxGf|xjb zs?rcY&uxL0eGz(~gHU+o+PL!X;V;2lCx*@7dqgp(V+q)=728L{Y4fw4zZc)Jq zOH+vE{`8-AEFPCw;Atx%mo*iQ^XMzpqZunZni4h z&L-yeI*g>mZB15qT$FuUGgI|?e+f}9X|T7~Wir0>(zJS(A-9+5;zs6fiE{(=osH!T znjmHFXpDY_M?I-}yP@c+d~x-?>&y@NMTfEvg#hKF!6|8jB?6E0!E&ol*skDecTfGs z17o_twcktXd+`KjEO8?vOcVx+^1LsWYS|n{)`O@7SV%jkpL+$xe+l^AXg-c&A|Q=t zwm5*OGioetNp)FLoH7tOWqyWlr71#efQJ!IVaSyIE10uXo^tBUCqso* z^6FouJ`ff^QD~KJ^ohZ|u*d+{nSMW*Gf9x-n>07%FL7j z$meJ1u+TLQvBEPXh`kN+bcg^$uLeJPF8@m;f5XdqL7|2264ZGaIEXY`3N)e9H_y#N z=;&P##tJa9q9P&}pDw9cZhl?+XRjb(#$$d+p+!Ot`JMv!M*Id_(8BAm*5mReWxQde zH8<73cBxUClCngaLJr_MQ8Lo9vU9bz^Yf0JHX7O*a`^=kqcopl?PkXn!brl^>6pz| z{w|dF{-TdUc^GfDT01(c9nJdg?}SPfDMhB}a|nkJsMFTW>*5NJ5u(|JMvrQWZSU$} z9TD-BSpx9MH-y^7TLqO7$~aMEaFzI5zuxFJ7%#?pcmC5&z62c}p_*vU6njL$4XchA z;|}sE&i=W)=5{tVYCK3nOoT0)E2%Si#H6LoXT~w@*{N$j*5)uaXUu#CXD?(4nngXe z)b+ZlEVVS}I8o@iUXP0b{bjSw)W1y}gQmv-UQ8s=)M7Q~Du$v5=w;nnOO7~uM0Dl; ztRul6XlwJuxuV6YH#eG?qHMZ9Ez5f1{u>OgGiV!GC zrWiCE12UdRv%w*U5J|3to(z4Y?S|J11Hw&jxD4es;=H#EX zNF#)O<=&u~P{_M$J>!hML&3XyTL6Xj0#R9Ul+p!3(tZ5qw*X`ar;F=bs3d6EV#wm8 zuIh})%FLLTjwF(v>eYQd!-Jf`6WfKU3A>J2nG3D;t*w|D+>hbP9nKl)>3ZEb;r)uF zQekOpcS@AiR|iUTRo{aP_=F1;fAGn6cnXMy-dHbna(50uVgUMnTzkI5b*3t3FPDq@ z(45V3diVM@QxLXXwFDMM80`Om9O&=eU%xKDb?~hxL2z_(%i^(-1|}U%EgzUe6SQ^z z#cQ^D>8<^1s;?$`<|2jMdv1GqH>BFCb^D68YjeMcZI#UD<0a046tan?N`=xTe+ zGxNIS!^ip&pEh2xJq{w>hTq;k)`Iny1DAxKwzf8*+p&hn<~*QP#}PXh=0{K|yTk5;G0%cc7L%&H@zzOW_REcW!+{#+CN6^$P(GwUDJ$X;&uTBR=Y z9v$ZK>Ue}U<|+@bMEM{v3%H|I5Q${?+=u`<98GO930CE zMh|j#RwYS|0pWsCwav{d9d0vKgrDSyx~z{vCGsbMhQHprU;1gCgwiaG|8b6B*$)s- zv>q+q8RLd_cC4;X>yt+CxWm1ho15<2^F{ROZ*)7n-%vu*gg~j3>w4TQM;u2}UCb{m z5%9LS4CalVu7*|xOe1RsMFZBwA0kOQ01C?cmSt_rOci>}X9ncH2>-NgC3}le|!=&0DJC zFJJAvcfWk-y)}qyY4I*qerMH_GyR>esFJ?gAfM;>Srd>fSd<;G5G51=!1oXcS`7tZ{xv4={w&SC?6m#zu9=EJR(8}1XQ?)cV_q4nCWen z3bBB9r`zLadwX?P*DQ511uCpEBGZZEBbzdn%L5>B0RhF{w8pxL-E~<<=L>g-ENqfo zPw2Xu0jC|GH}MTMBc~osk}!Iosg{FBhH54L8{GL0prJ{V8?E0k#8)bmBG0U?QlBK$ z>h75wGlu@_tSOUSwixj?NSM?0`OH0oekr;A&7BB+ZQ zJ0M_4|5RvP|4izLCGKeVK9@BViIk_pj3Nhl3plQe)5(0P%J?ebnTh z)!&((z^09EW9Ho#UW@MT6S=H*q}MZUUS6Cw+g$ddd8{ski0ukU$cf?dz*z>kfE;E@ z9GawuxjuumT#q{vDspli67(SxM|7Bd6rM1n<6OM7T@b@)TUbSQXx_a;3TX<#OJR1Y z8CmOibZ&Lrx7F8|gw!L!iZsAAWqR-0)V}Cac6+ka0+v~TL(v4TX*;cz#9UF{U`zKzW+ z6j41J>Y(eSK2y&Lgp@PFKpIs?L2v6V06tXfc1Ars73K)#G?qKISdVh1h`I2iYga#p z2n7p0Uyhr-|TKApJ{?FBUr{`Apsnz-q2Y&rAX+DGS*PF`O&&YBYQQwkM zSmTAI>sg%o4I{Lx^jY@7Q#@Fi;|& zMFZqBk9wmK6*cmSNl72Ezfhz1?KoK5{OGE9Euv|(G5sn}1tmXYyc9aHk!I0dhl?c+ z8pj-Gl1XC3L(O_!>}%?+Jwr*T1YfmmsF0Xaowc;GI)dmsljKDwE+-uF_slu@`T3!; zGJ3H64Ze@#U;!0AvN#}@%z7Oh8;UV67bgtHMpWh}#A7)XL-ZfL}AH#3gu^h?KfAlz2dX@cNen1(D_beCf_wHA4u-V8)arCaz7I*n& z+|+c?fl~Dxgj86O9Ub(-ats7h({%vridbRwF0klXTq8sXBp^H{Ky?3*grPc1M{9*0 zmmy4%D^Eg6=|#HXLXkUZjtzsA@LhZXpG%&Z1QDVy7LrVEN8Z1TkX8su$OW>Z6*_v? z)UuJ(|4f&80H%B1nUL^_6_Uz<4$2;Pj@)bvsuX!h5R5~uAaF6AWbuA^ZvqS*_ z8L6J5pZ=Z$%>uzcGIDg>

Xsp72QbApE=F8&^*MUF`}vlfUM<#ps3!t6s! z5M0xDR6%75Z=XIba!E>sYct>=i34MP@>Ut5GP*zbECmY}v|6a8^~B=XB4VuV8W8;7 z+P<%!sMd?R;0JEo`R|{D;->K6f(#|o4>b1J^6t1oNHa>>C^@TZ_Shs`5c70ULQn@9 zn`UC7^bLh1wEn#Q_HDyB%4XfI2~zhn+UtBOiIGXAkJYA$3Q+kK;gG=tkS7gmYW__} z=uzaKBTC=DU(h&k);s8f&q;ch&|B;3>fr3)P_ez=qd)P*H*bw0w?>F0`s)McdyKb) zgxP8FSt)7&uuGJ}RW9Bynku=u;W#;AP$^TrC{?Kh^-)t(d-l90mN?jLR@OS;&uH{G zgFtzrqfO}7NZ{UAsxa*@RfvZ9dEeb03kv~rQwjKa@FOl#D@~Cg?Hw9*JzgE_v}OtY_U)TF)kkP}C`nsrdo+bN7$%q`$WY&p zWoBp;$&)4?oGr|(NUCVUUCd_qUSmV1;n7!>HPUKyjB5rOlHo_S;z^X?&{#id+(3To zfW?goZ|?900EHWfRC@s*vPnVHj5tdR!EA!o9M#6Tdj98UzPoTpfdJ=hptV*qV_52b!+}vk+pA9W+<%dNF5&!ezV`I$-(G|tj zR@Mq-bO-Of-^X;e>zTDTy^`zG>&3Bp9lp;3oRoU(Uu$R<&|!p?X+nC8mCLqQ8#JC7 zbe10eT)cE>D^^yb(6-_fi)B-N5a0&q2f(;b#FwexpGhQZ*SXM@3%3ke-hY+70rlMe zRHIB(a`2y(8h}S@FwAvC>f>S{1ao~t49OFPAKRSM9-PiMDpu{9s{tqvz@sa&-|uE|zO8bJtp|@w z9_#n;aF^r$$>3Wx;8mju$+04xoG#WnV&)&$1{WL4dc09ls+#qzu}vcweoZyamLA!y6h2F|n(w(qYX;TQ;-Rf;0gY zzh?~xNqbyhNTOJyDiYE_k`3Z~Pz_+;vEepSnbS~G+Rcp_fk`05yap5DaX$3*HUF&d z+qXpa7Vi7o6Ie^mQe~;v0G-l5btyGuAMWd^)4!RLDOHXtJi-1w?m2-+Z~RH6Tj+n} zqo61a(Kj+zv(K8F$4^qA7zX=#_lS_+#3T7haFEKF>$RlGsPOcxtkFJErW!Hji#{%p zm+t=EU_D*-v`%#qjS?f}_qjMI*2t)%QsqNmK^Yey7kECbT(Ld2m5RYio-p`^kc-#%Xc;`e8m21~@#*PZCR+iB zMz5%^xGM3UfD-|0DBRzFPa~Ckaa1aa4{Y$0l9m`iv*ab<&?Y%$V(b`?pRj)(g++Ej zKqPI>l#W0qY8)(ZUg?A=Tm9}H8|}%1!tzRjhes5GtXMi6>dUiCLAq)yuu?cRke0lV zb8pmS8KUxkiO7q`jBq|K)1gf{{yj*|{x~*Q@MnQlm1y4WmvP1y4Dw4{U`qJ2$Ni!4 zc)=n->dpp+i0Pt>{kHCNm{x;+Z>rLq)M^^FL5s~~pO%_i17aI$-gYy1@V>G6dAlw{G$6fIN6M;-jbqkVEg-)Lq zhR@7Mp-?D-D5?1l509{3Gz0@7u_DORL4{2#C}dx&?iK*2^xgk^v%C;dUnBMA7Z=O4=LYf$_-x1LSULIMk(=St+sQiL2)O}xKj#lY0F)N!Ogw$>MvNu-dlH&=X ztR|G2yGkXHpzPso~HVJS#9Eb6Q-dZv%m$4Iq0Dh-$G1*AFn6~P^_jg zjMF7!UT)BCwt|){zJ1k3!zQ$*m<_g)g>{YPZfk9i1=8163RoSkJGuaOR(g4_5|3yP zCSv`eewkM`AQz)ZsMc#PjxYAwqPUhS{0V5%xoV4Dq7c$>?W*-lSVY9ZI1Fatagi=x zaxBm3O0PEHtTCulv-OzTtSj*Gj9Kmvl6WB@aGA^2s*UJL|4S13OBEZZB30^#OB^`{ zJOR)Ie!;}VBp5Q4C_uvO8L^jF<}dI6{VlKC4F~0jNE-`r?HSqm35v}XwURhM)hGmZ z0+HYQ6+?*5U-!B#K(q@##+i-=KG8w~6-?b7+Cjz=cRS|uyp&il)e=JAKI?Ikl|{?{ z^*x@hw&7K<{k37)uUyfB)I}0}r`f946p4@9Ujp;~Mi`|q=(Q^q{pTErmi?N5&jsPB z_6H`egw$kfZmvJ$e2wjRZhktzNXV;g_L}lvkNIp*6qrzGN8dD;ZttIuU)74CNF3Z% ztS|z%c#xe+&ay$J0>aWIUrXU6rt(_~;qY7Oh>0BK!30 zB1^ckWUWdI{WtF*C?;=+2whnz^vmjiuPRVO&RZ_YH&iLSMEu?kS7#wCNS)nz@>HPn zT&vjDW$mF+s^j8MVsywQ=k^}XIu4j!c+nbz9Ny*@+7b0+H-{KVEiR1SP~Uuhb!G2d zXQrh!`QEu!k`yySNr?qBL9vf?+7&)dvfb%XK~Iku?RE~y@A>Y-!P7$eTgo?#Er$V` zN4NaLzBYx4mUsZxeF=lkDX{&%OT=Mpu- zkxB;fH5>Bh*Zg~A2%taZ(rAkunhU|SzA4RCMDe6|nUx#yRxGAOb**4oaE~eE9}>)- z$_O$s*{RZKj9OphLI?zK8Jnf@+;n+h3@SWC&r^Hgj~{*C)C?@v`gzadq>NvpII{(s zqpZB4?0N>?B8h{)2Y!*Q<2bpx&NsL?6iWSI8OKeOatH2HjDS>~&a~@A$Ls3ivTE1F zq`zQ#UsDt!kVLT*)ZSQ62P->KZh?9s8{>I`zIV=-zeKGYY%COJmZsiSJx#s2D?8$7~*~^WOp4cI$(t& zrq%)sSgz8tvs2(5O-t>hE z^-h|6i6A{Z!9Tg*P_;@C<2CvnZ7YTa?!RQ^-x`d~9T8DhcCx_Acgf3~w{$YbN_x?q zy8DBJ0V$cg-@@z~I9j3;_{JZi1VzZ8aU1~^VxPREY`YtRRN=Y&!CzVBZ;Z(zr!r-d z8ghl|dABRsSRy(M1Xu4}+kdw~(F#8@5B@gu^XFgZe)n(yf1WR$(Mt%ODKjC-LlT|< ziT=s+>Gl#QvCN`WUjAk*Q703aNH--VN|Hhgn&dB^7Ob8OwHdqyPTsNmv`5T&?{6NT zotZzeDAP24lI!BWq-bsBeoqG+chbAzp~#tHyFCFPUO^ND&3Z>OsmMuMY~hJ%%Icew zrHIicidAKTiVBStue%t0!4l2uAs)LPrl*bPJ^nVib_7qOsY+5x%Gb^NrGEySq?$ws zPU5meF`PW^Ou62cyC3Xg<$9=veYN4Z03;vp1?sBP(*qhNSW=%(GK<5|G9qF=GnFP^ z&@tVe98*5SyY`;&`PioTehv|tY@9YHs9Y;s%~Go(W?FCuzUsdfuNOwV(~}$FH-Sgu zysO8v&w3@?N8_PDI?Z8&p$Ut;h?-+4gbnw5wS*66LL|urx(?EO2GZ2YUv=k`+?YxX zB5XmAKMRIXFijg=AFR0wq$#q$?znY5@^3gQK8V4&rC2Yl3@dHP!iG=xdrFo>1P-+k zbJhGz8YB6EhlZ~fzJ!{%!zK{I`amm!$64z3auk!v`}N`yma8$B$`kwrHq*mX!}c@Sb2s3N!2$PK2Vw|0!zDg?#vTB?5`+TQV)|H+ z!>dwIP;9vT_WDN?HP7M#T}CoP_GE*GPtYzIQpgm5TA#ya_1f*NdKdXca&fzwfD9t$ z)e50!yFxGouvfC#+?{~V@4NQQ5AGzIs)NG#+5f4v!RPjEUbNVCw;dv$&~a2=7K+~*Ok`K*aUZk& zwyzvUKwQMkbK70irlo87PPYHkj8G4Q(1W7My8k^y1}c>yKKB`+Mv;jltYMUOJY5TvS)vA6~QZ9b2UsKWa0 zbr?m0s=KoNN{5*W5XefKS{6)OeEf8~RyVq~+9MVkb#&BzI0q#~zS5KMi4oB5!FHiJ zljcsZnSUC!wCXK=&o@U?xbis=Vyu1@4E9n!u?PQf5fueQp;|>&(AxrNi0(@K*paen z5}<|Ub2k>P-$gykG@-&WKl_$7tTxuv)bz8OiJF<6g|uiJm6lf?Mn-OUh?Fi4nRY{o zvZ}iVu{j@1vpHF<#~f(Q4H-6h<@LMxNv5*l8~P`{{s_g!#TqW`oq8s0-K84p!gS!6B+c(JCii;5f3 z-AV!!-HltEhon*bM0=s~0vC$nLB6(b_}7>CZ~h4xDkhb8W{bG>KY7?1z%1ZD=5IA? zKnE&@#FQ5L(Zk3{hKJLvsmaF9Zy4d4*KXT?*E}~!O38bAykwsNrA*^8)*jKfMdIvF zV!ob3G;5tt^!Di~F#L>am}ZV`+;9I1X}2IbnB?ziM6b7H>TiV$p6!nQ$_M70lA#Er zme<)*r6(DRa8z$RI?RogOmntH(pQUjG~f3u;e{vu);O@=E@PAc7A? z2I?*JcTfG^HFvV@=!XLXky-T z{tir2Nsik$;_EoOidDe!!9_Q&r7<(emEs^ja3b)nz44`gN|7;R8X4<)ytdNRq{(8C zKIDjzS%f+()ddL(jk`!LIU<||n0Ol|TRFlXIa*7^Tt^JEq4NadsbTG(&Cq8ibQq5- zCi7o{!*|~K?+vycTLzaK0K0NHF&S@6H9K3O;q+?ibjH%3M}=qq+5(%D@L#(BFRPnP zObrlC_}a6M+&$tI_0Ks;MsG)S%huNtN0~?%ndH*gA#8@*rtfhYVAs9z{~6pW#!MCf zAuTDP3|o>UOJ|EqgMF7@PprDY6|Jl`l&{SoPh;TigN_aphIVRWV>3n`*)yBW=`P9m z#6m3St%?^7$vkt*gmOCXdIzeMyZ(dO%36a<5n9?W0wQ+Q#FDQhS@99RMHlQ!YA$o&j8$y2TmMdC+pTvbEiG@*zYj}qW z!5p>3=zLdosaav9(CZ;IWSlq*4f?EkknCQL4i5e_>D0$(ByFt;x1M>GrQYx&q$9%_~Nbs_Pdi6X~3JS=s?p$ayIf5>0jU7!SSoaN6y)HZz zxc-6sz%Ti~OyRLD*6kO7XwM77_{#QU((gZmN_1-7sTbqF!FYvKFD#Ugm=SF3>>pY9 zc-XFt#M~POM{@9RB6ypLIBgDi?d{(@y8%qy$>}Wc=-_^~OkvvDP{9ZQpiW4P1N5QmOT6?=QTE|f%rq3Wd2 zmlzWuiby3b{dy&&!ox@>Xc+STz=4^4lzw*l7b-!a;LuLMPZ5kxUnvzkIZ8?=k&-KW zC%2oM8(?+6jV~G+q>1F_^tz4L1d&mCnr*pw7VK7=3gU~~Jm_D(DweYC7YJz=_ZrsK zU()eT3>4#a;=l=jz`TLWWz6Cs3|TwR_K}O5cfGA@UA{}W63{|GVRpx;7famc40Lj^ zRe0q*XyUa-kCVxj_-(B``|a1)R(PG+vI!K|be9B^n_6pOqqRRVWU(XqQd3)Yckf94 zbEn|rzVd#9h6WB+3^JhyIoyCU+_@L8cAr(3#c;1-+&E&OP+SPh@GY|sVbK2=> z7u%wWRI&wYSB;GBXnDI^^z0$iy;VG~JXMOnZ2kKlswRhu=6->j3)BOZi!&Uyy3RWf zkf4r|P9B#W=W#kndoeLOyXc+zQ5u0fPlkr(dY7td>?p2Ep_`}ue;MG#r}w2L{%bp| zH5#~)f({cm7`;>R>_CD*nzBW`j@LW3T12+D zb3T6uTkxZ`T`(TJ5`lobvqkjN^u98q?hhSxs;LDHeZJQc3e=d+=`dJ#_jM1mZYdLk z51mZd|7&)iJ@PB~xf(WeusV{SPN&!NWVS9DziCRNseU>VeWlO;NgSGnbUyOx?P^La z0`m>gf0xp>0Noa)EK6YDEv5FWN(a?1#`r0s9mfE%$Yz8$fh4hJ1~*7_d_*{FsA4qK zCVCK9;`J=`=UZsfYJnsddIo5*c?gECJh(_9KYLV8m=L2(e%7OKWGQS8^lqqEubkO| zR>t$Fo!xDC{x>rV)IA+f-=#sVLFIvLV{>l1VrjL=XkYvP3umFBk?w%%iro6QY~&$dJ_P4LG+3GvK|!(o`~Y2chm(2s)+j&__k(V} zEIBA{cA%bMV$fldnXdiecw!>$)|m(b6#_dyp>KETP=cYeHnk9yt&sYNy-rfFf0UWN z3JakZ$PZCjOmg}*60`%=tJ9ubT^)--ipRy@!Z9N9lV*R*g9CDepokP9EOC{4r7ynN z|7Rp?;Vg;@kf`ahC}Y*((yc;D6dMnt!3!wg_+EA` zg-e)kATsTBAyk_6y-f2cc3nPLOO0HP#M{<2QX9?tpow0DqHEtL@ zUg|&4@Gxt|a+%_K5)q)ZD+y-$P2+MH5eKUyB*_G7PDfUjrwT{6{UjW@DDqZc z)B$)tI|`pI=p~b{=ykn+HP@FLt+!qr26)DHw9k^h*d`4mrDHAnD1RoLar$T6{kd8P zaBHmn#UB+S-p3R^mRZAGNE#M?==SvI{wzHjtol};4YEL34Bu!5+21-s^=rwr=wDN` z{?d$8ga6e6oT|40e#fL*-(j!Aea@A9jrtct`aGe$T6G=c-_|IX$eLXtji<2~rY1yu*sI(uV zO=8N{3r`_Q#J%@$cN9m5NsUG%y+8Ihxa;m@1#@Q7WYD3)P)~KH2OiCu%SRMwHj;*$ z^jf^->OJ&dQyUsaMwgQ1srnM^-)(8=OLjl+Esg>z3WaE~J`wLRA#V6xeNnYLHg4X+ za_6G{Sh!WCiliy|JeTHR$i@sSGRmuHf5>h>S2^nyK_Jx zu_3C1pC;R`M#^Ob$wesd&8!nY z1E=r8mT$^+*yaPdWWnR#{}TLHf}cyu2W&imvF>0=KN?E;CXZ)t(V2&S-kS0kUg;?J zAmnx?qMQpABFSL2nu#+?w|Y<}UtB%~EIThepH~LOGg%{Bp1XaCc8g)^tlpSCTcKq% ziKIa;y4AY91M@w7c-5X3sUR1O@aEFxuPSGABrBAflw1D_8b!PQZt>vXYeqaw4t zw~SQ3E($2a2>4yQvn|&r&T*K2lrBPV;&EAbs;PU6RApA-aHEvGhB!egmyRt9qm@^) z>awn@BNoHA#G0r3JHiD$-yf(MC{Tp{M$Xh*&DCm0BqJ!7g;-_zQHb?fxpdClJy1YE zw1YMnXZO>1<4wuWvg3U!0fnr8Ws9U)$;i5pLqy`#wMW?B34hDT;5(|+EmiLHd7GOQ z9!OH-nu*u7VqLbqz4ML=tN#i89g>N!ZqN8}B&aHWhvV8jP~yeRR(L#T&lAPh?asBM zvHv!?a62V!lq*ArnYpYSG)T{wMu@i+kd~WVJ9NE+4$k*Fhhg(mQdZy(L*oH zlKsr_A&=@?e8yyzk5l~t8MTu113>`EZ>lWk{Jdh`sQ$i5MZhZ4P@yl*YgSQVVG>Ya zMmivIaW5<_#1RpY?ZB9s!Si~hM_=W!cr5`WXynvs}{T@d|f%H9lwJ550r~c~E z3^(^0X^{rq?jY-u)eMK)j(_m>Twyd8YWF57o+aueX$0y}Gr#7w{7Pw^ua7E(v5q#v zY`9Q`>R+rernVU@wQf^L{btOXgWRdlxm?i7Z~P~PcBnRL2x|^rf&hX~g4Ruh#$~OP znYkc6L)02e!xPF0TeDVVnG0C78OK-zA=v&_1PGL?T-gCdg6MQIDKFqE*zizaKZ#Hz zWb-+^)3b(!hYzpI$%Dt9UoaY+3IPwL`HCZjU~=A9aT@wj!l6V&IZD? zEnfNMwp%PgKbpZU3#bxV(x69E-X_3cCf}@5oYJwrX z^iRq^a!RFJ$g#vJK2VEm>tq(m;+Lbs7}b7%9m^i62eX946ighx#fR%YHz}7YT5N7_ zXjWTTS<$}ku==*XWsl55f_1g0^M)h2*SY(*yCsVo$X_BrtdjKQAEFy9VV*|(8V)}; zT{IL6-s*;XkC@;KVNV00zWxFa)o@TY#|UNoM%@M4-?mWxT|6|2$TVT`r9eMou|pvr zNH@@7z(`KN;{%@zmRP93)lNjasXudp4@P-i>jSY65EBKZ)%~e^Y(HALZPr@6M@lsA zJP1Ssd!KOZR%nh$35RCKwuG ziOcvsx*|?FA3SU_SEWU`HvYG-c9`GG&Do{Dr5Bfc{pC_F+wgoCOrOQ=f(Q`_uzunS6h3D~fO<-d*5X($z6No&s z@>i=+06;=uFoDC5BatnqzRn`2+&XgYit`PbT!01^piJ9L%$>HdSz>2@%HsZi zJY7{(lwI3aK@dS&KuJ+Ty1S*jyOHj0P*Pezx^w96M!G?|mF}(~X8!#ie7|cBJaD;Y z_VesJukbaPh8(n9U+#`-Q(*tI#9r+rw6Gr0Y4M?q(lk@#^(I_Lst1YgRREAiJzN+n zr;t(+QFd~2zL>9A^L>~ziKnohDa)-zM(jo>)?RDE`?gv9D>k;>?%F@Ck)A_iiV|wQ zbW4i3ol}^B?<)r`BlJOssL!1c@CI9eG*Ixi8_FN|3H;ERcH(5x>1j~;Ac}0 z9{vp4f+&iAph^CTD7Wz9*;@?rN&kP4to&!V_^V{NCg-11lgLSg&NIKknGA*AL6$9% z9;z1PQ)%P}sQ)_#L=8^!_@VGkW1OT1`-DyTKQ8~Rib9_3*ZvmtKYNx9n3oWZmx1;v zhu;@EzC<*`yn>=81Cm92;?k@w$7Mh){Y2H?jW1I@w>ozJ-#;48eqG@S0k32T+ zq+m}|FiV0xc`=Sg=$lNhO|NHIjk6&k*u!j=JJu`?0tWJZ)dpQmP7ZJ_0O@mkBqMJj zDA(q^DbfAF(QoniB0P6@d<|3PQ^XTwaUnSgko6S>eBZYLnE%$+RnT4KvKvEf;{Z+} zy~5!3Yw~a^p6qAgWEMF>$?55MI2mULv&)k|F?>3{8HxgGj{Ryw7|c&c2e9Du+)nwz zRiczpP60Sn_~&PMq%n&e)!X5*p)6MG>j*L&ZegJCw#SU{gRMj4Wt->+X$qgmy*^60 zgQ<6>W$D-E-^hw2KZQq!!?BZkcToX~ur*W$%+nM2_^TZE0ZlIQ{C(b`sRuJYuCSP~ z6d>m4+K^`%bQv;%T82i|;Tli?yCTa($@_LVa;GdQ`_KU*D6AKxt_@?HyGM}%XUA6oza6rPgVc+P@0U66h-^G~3#~02~85yYX3T zRe;&=?(8A7&go~5nZOqf8uZ?bXlX7>+JubLG&{rpOa|up&^9*zeB{s&G7;>>&3ql*?;EYt=SYP{5#!_>L1* z&|HUk_mF_`f`daWON2bOUvvOakhyr4>Ma%};SL4Z;}g3QzOcCXjHLK@jDV6;$=Q;G zwo*?m9i0TG*JyB_-cE!cqH#q?+mc|vT(m;KLWnDy+mBX`zUOl4@ z?Wx$>S8De%C8zKHo@MaO_r1AunPVm=?rYQM28wlWUht8u@vA5*{?IGga#%$Bh=In0 z24ZbVMSl?Pb^Cju-~N`iS))>aaV|Qv^9TYT88n>aUiW{8g*Y@g7#$`*I64ZH05VKj z_yKQ;q6REr_loWweFPGuT>-&CajIN8L0Xh#!cTaZT2!nyv$C={hqUKge0dI6oivS< zs;W7-vzV&r)SFTW1DwAsY4&y7NakYnpFa1E07^IEmvtrG!>Q~(qFHLRZifB_u9Gg~4^;}aW(pRwg<(VQq7Qbt z4bE2l`#T8|`>3PL!kx5~WP>s0gfQTVk1k zF9=AbeD>Y>2eY(<4^j9iec!)-(BUAyv*K%#k^$QyUn{I( zB!lSzLPX*83D~37UjOw2EJXka0XsvG!bIN_rX{EUT52~(W_c4JWk>*f!`>yU)&~Oj_E9%&hz%W?oxUOmtf-u7a!coJ7k zf1BlBP}0(>gM&fBIGim~(tF)^Hq&Q_iiP-ecYX<;2C1OeXF6P?5$EP3yVYe3&`EyEmz8%c4RnYZsOA~IKb_)-JD^+;ZN+2Mcr{u zDSV%{o&bZ~pv@z1nr|%iJ8&Pl%>)Yi?vBp2qHuB!$`1D!l>3Lvi@UM3tG*LQXZH@* z+hj@y*Nvc;+PBP@h@7^1D-e700&G9P8~=x}yVf2dq*MpS2pO$)b6MheXD!G5c$4f^P zKPVN?aM-4ygI@^rX#36F`-Lz+F@Jz=1TWv7^2@-iy4WZB`l>;C2CO#lCkRc?gPR-r z1?iCy@IY#%Ak!7Pf{MUExLVWc`l221;rjb1Up!-Sa`7id2-T!(_UAmiuSW)*&4G91 zg-5y_|56M9Yqq8)^_ihR9!@S0lkxK(4(43mo+1@O%#Ez`rY*tNfQTbk0dROwkX{}F z;>g1_jF(=Ga}98Iq6qmV%y*-fL#^*U2uoj!O)11yQ$_3fT`@`SR*dgxx*x1~foc-W zn#hfJsOfOQ4EzGD^y9wst?6V~73f0W_qw zE?orV@LZJhA)uSk>3;iJy-3FS^1MXXsH@8k2*!6q=>cdSV3%deGjeKGo+_)Uq|}5E zPi$j1a)*1r9!MN9^h`*RCSiMf zM>yKx@ixYTIfmliG_dFq(7cELCX z_;CRUc#HuZ^9RqJtGBZy-u1WEQ=q>9>~la|h##6PgA8q71iUZ;BZQ=;-Xz9#04M#K z{4-bOKd#8j%Q@S%hF)ETpX*rH7?8{o7cIcwkcd99x53`~Yr$zb{9~5ZEwyY{OwnTD z?~fgN!h2|aky33*260Sx_D5SR8DhS zk#nu5nCHTi)*7ewx}Ct($Kw+oxLBKBT|J?~@su$HoJLH-f`Dl1dC@xycOzj<=8-1D z%oKFLyeXFejBO(3lFEi8{m?dpHbWER+><9=RtK=ELx&T%!u9{wB(S!(%eZs{t-^Ah zr***Bi+V#5kUceZ4nCluO%`zW0X3*}O`IdEaD$gXmAKgc}9FaGef$(qLExcz}wby$6GE78VDG&dTG%PSZ9DvI&+bb1aR9KWEWorVG}^9YR|lt4B%%Lbz>o?hgEF-&^>46+CVopw{MO%a z{T)e6&!*!Z8c&A2ip_wH$ee28 z;qX0*iP8FkdG+^u?ZZv_@6LB??N1&+sN+uFIaO9XjpIl`UjAWyW=eFNAW$1udOJ=m z?PSHzxJ<}K_II6sBGx(b=*Tws?m8$vA0@6?Id^`)0MR3h*PFe=D?=m){6Z@)?YYEb zn7X0d>3^1Ei$ze_2BG$xyzVtT=gsdf@$DnjolH=@N#~c(f$M`iyLyfhdc4ZA1xR z82}vyS8Lcu<8MMrY%JtE{_%JUQ1cnk^cErFS^oF!IKAsTP(L_|S{1Z`dLz|m`2K>z zayY;5kdvY6PFFCq2HFzf_H6Kx;iL4?%bMrU{O_MXze;^S-FKS=gLHizMg&Xl622N) zfxoB0R`H&_59uyyaRyIUUPTmgn`xr zAJ_9%s6Vj&%S{jc`i*RGXt8^>M+}Zm7b^)!4UMML!tu3SDY_lLBmc5YSARrFX?B`n+(<|i z!@QI#ZC#*U}wt1-oy0BMD|}OqnsDB(?#_F@R@3NEW{LklY(g1p4HU? zpb_>_{Q7XFxn3TyKV&;>FT0a1_+~GaKTvV{L;u(=5RxGZBN2a6LiByy+Xjej|HpzN zW=!zj-pY2Y@!GPmFqkk6_2qd}_;p~~CeHW_p8_SH-l7Bo|(N%{WPL1E}${@Y!Cg}&Yh-+ej&u5PWzb$snDA>&gjF>OF-=yKg%HGB} zfx~Hcd+Jnc=a);os@SPFy8Nv*t?_q+Fe%+?mezi;qj%+;m@j`994Xz+8JgFP*TGeF z{;x8?l>!j?8)z-&k^HLwK>2^q)g%Gh1wf5|hASsEG?Gj>mc)&sIS@Q^tk%1c@b;|w4Jy;S+WitFF(6RwVdbAt{mxgJ2a z=2({AoHP$kBUaH%_B{FR&xD4G_#5jB#N6Efp&5*rdEq+yhdMSnzctZJ>Uazl6R->9 zua)v18f|6=r-n#`Cx89A0y1Kme16-9^SLM=&_04&WCff!Gcw-D_kdE>D)WJmqm18J zwG-HU0|g4|1P5hf>8`+CDh1u{_X&WZGQ7=OR;HPnC8MJQB#HQeJgDOAa=bhQ1u@uu zT{NBF_ZV8Mc(i#33LBuK!ECl7^vLi>m?pMdyZYnj>OCA`%q^7>?>}2u$B(Jjp)5yw zTL8#}02Y}OMXJA-8%gXYtI*v#f@0k3hr4s;XOVumJDZiyrFtJ7*_^BhM$@Axkp+H6 zQpDx=JxJXV0tk#9&C-fD4fPO6=*<)->qqaYfHxN#4J7CTfCr}zK+om+dEw>sjDnSa z0Z38?-%il)N|`Q^I$=0tIflkD>LsxCxgOTzsm@}14`fbw0Fb(PJ-Z=3zFm_}PFAwR zk^k>pltilaDY*G}hZEo(S#YO3vt)9EM3hc9odap#lOn7}ky*~rLyglcZdZE3_{R&e z$u#i4kEgPBTeztOU-VYXq$<7#AhaD##|0Z}j*-4urExm*jsIZ*RQ`^a4=5=a)NQB8 z_W$mKqY`Ox6v~jn+SH!q9YWgYw0QG?(5C1Q|mksnO-i} z2|P1_5RgWmc=Z)6-^UM3@2&9>g|B^lL6Vuz75n0itb-`rL0Q1>Mc{JL`_?8Jqgl4Y z?fN9-jrHM7VrpEqN4L0*4|Rt8;bisf!g(OTiOnRt_8MEBmAV5#ON=+rZ1$9}zBoYN zNav~_%aVtJv*q-&>@BqLYu1Fvc%`s3U2W)s8W|E(SiP|4s8S1Af=UzErmBxnWY|`m=AHoGfXD-u67nFNf(`kLZV0RSRE*XTKtU^`ho$ z6Y>u(B}cfJ80OazlA4^znpPcYjX!3qaA+Jy`=_L-m$+-|I37zf9j!8{uBYwm4>bvy zHP?S9&4u?JO>Tn5nDcG*lmwJ;c4;eD47i_B`|mdTRIyoLw)<=!X;rqRs97_k8wj2P zE)=_+vEEz}hl`YS`uRaG1*+hT1yKfDs}FQ(bYDi`Ztu;2SlE_bO`nWG3deO+PpJHi zTV!CNWni~radiG5iF<+Z`5%qAmImkZ8z9`2()hRF<>i_y(fiWqy}ch8g8G`l1X+zS zyFplNe&-P7_N0p8e!Zokl+l|=E#le2y@GF7YhO>eI{5Q4p zu~ecdu$lv(dyKuC4b)48OJ#&nwUpl{u(eJW{|S&gdHR$j0iAi_3Uz; zR7UZVkU}L3taM=aBZV1lwZm+3f*eh^E1oLQX#!xxJlkSsTL*vFmSr8pUyFUouGniN|RhZeOSIm%g47ud#NuBcppav`xD8em?1{F2f_T8 zM>q(i`^=^nK7OIsGdlao7$m2wkle^%d5F$c|HF=A+2`LZ_T4gHLaw(5%D0>>8&8gT z-7RYm^6Uk)p30l=N3X3L*SvQxIW^i2yuPF`NafR&LVY*Sx^a*KyE()g{Nc4NR}p~@Y_Y|rmN@FDdrn@`Gux7}~L_VpcD`3bGd)pO>_BDt6-hsNbb9di(DGB-@V z#1-Onv{azx#AL5Zmo;+EDf62P@Z$X}az073`X>D|>-s+vQ}5V17o+3C&Kh8BQG+ol zf55_51RD6uIbLVlw>EBiia+7Qjl00DqJcx*4y>54VlW5k-@D_n&bqDl!|N5UpLnT9P-A;7%d zZe86mUqWdHcgLmA$?JZ~Uo#`#mb>>s1M< z>2N(>jJ1|S)-Z6Ht8XHy-$T_^eXtb?P9nV@J4=XlZQis^$3%&ud_bgD$UccguAW6Uwf{WFKeAN6MbR9N+xu*bq%)5-a`vS17a{($N3md zw6NXka8(9ck#h0ayLY%^FJG#q$MU;wq}$%c<4}zkkF*-IqD?2<^1CeydMglB7X9pE z8oi8{i1IfJG87g6gdLT2b|=Fav+IAa*O_>K86z)!=jrWT0X-Vqr4%H_!O<$-LmF`G zgw=eeFDIW0$u&Hi?2uVY-=|Pz)1$)^j4YGB?XdQ-m90R47Y4?JvX$XR%yv+NBLC5>3iS>0Qpca4x6|V8hQ%fGS3SN zc`4_=tu^3z2e3HrS@KE!hvw+vJ7;mnw*2>+_9mMX-jklmMtfls{QKFQ(jaQ!D4=gJ zHpe;Tb^U8&O*939gU5fP+Wi9eU_z}D0;Z_Ewu61{XI+jk zHt*OU85~dVkdLXv$lW&bAuYE@3(Yb*(*?eviSC}{Ev>D_{i9)iZsFFjc29h`Zh$IH zO6OU@PfUsHYZK#-=Hp)Dj!Prm2m@p78gijNAY7~!JL@1DaIUPbCbn@;JDL<^iHeZZmOA2M*dn?&)`LLA-!OrxkO;=Dz7cv$avJS0hB zalmNiA|l)N-a3?Q5#qQv}*7u({mjTCJbJ^d~MdF=lWwh??YO7m9aEjs9*IlD0zPm?dkEI z+4~@$Xa6DRdr!!%yF=hKByLQlG13_pm>z;Jms0hZ|Lv%3W$W^-C9gV;E?Adu8MfrA zF$c4kcD|)kvrARJJ<5;WiD)N^Pu=(jrhaQfqzuK@yfEGb{mNa1t0rDhW%c!Q-|t8% zsWzVARc`$LO@0}utEuU7HE0PtCM8^vI4O%x$jWk0V!)e-4@8c(u6NU+T$Vmq8YlY2 zeDjst&Na|PLLn_rFY5ik&2dsMBS#m5vb}J_Mmx?9)9ooS{V(*y4bm z4W^%96Dnx|oe-Cz?Kf#Ms{Zf_(F&y-G7tUsk`JF$OGU_bUqe3- z&YK8|JD$5#B$AaWZVE?|GNypD(Fq|c+S0>^%fDF(Q$l7e9%frc#^hiZktl+X7w0B{ z-Wtu{o0a!rgV~wu@T;j&t#}$j+Tv~&eCYz?zU9ib|A!c$h-n_8{yI*w7vRIaPkW$Q{`EDW9B)necVU=qxzy7G+NnEGzO2I? zx&?XXI|cl$K*MxBLI3dZK&Qj~P-?%npFT6!t)+rY9e#0p`;OIa;nd$(U?c21xzgCz ziwCG@rFN~$zqtMDgWy)bpP%s}!1FUz-?=+5GMsh(z`;;r$tn@5Y-YPm5yU-`?(WB4CI{d@)#z@5pS1!{C@ zNsH{(S|UzTQU!l_c%UpO9?n1YE-Xk=MYTT@2|(cch1d7_J)4tqr0yr}zjVsB%T+NCGSbUVgXX;83w~@DWW!B$}%>bV0CdU>TD8!cobny17eK zUS8izI}u7#torAVUq7C~^bX$Q>-oxcWsO1-Q4i5qaN5KF3`Mj)5e*Rkef!s9J;xMe zibpLP_0r+?s2x{IQlDm%TD-gDgMr#cl0^Z?31=ir(gTF z{aCXP^dNDp)%}F)XPi@e)1c>raZx`ije2=#G;b9Df{Qo zF3zORmakKxI}`NY=)`;iE*^p!^0lqaLjzHy0ydhRzyj;B-Sc)|WtWxJNw-pv2D=(7 zHpdnR2aL;B3T8EJi`^osGgM-kW1yH;U~RS|E7_ zz9`#|lSPO#uz9xH{Q6!_71{oUbo|l>&RRTaMxq#_`2n! zrK{WSY|efp}|?> z<;!LINxxnB=bgd{qU2#wQBIEaZwUJ(KN)v(gRvqdKYv7Lum0<>@-8y?HzC1p778Q; z_sy2Rv@dpcYgDBjuBP>#ZA(0b(K$O=7A`$&c*T0`0+fFZvu`ZMyKFt1t^LHEm)rG#$s$EHV4cUG63AF> z{cN&K(As!=4>);#@M9O~?8OX?twK?+7 z5?ug`h;7blX^W|#+#+C`NGRxhv$`310G= z7uRNIcvkr@#y^=akNwFij47I1B)nsBNk3aBT2s!*&IfJstH4}ZqUFr_u;Wk?Yx zRP+CW`x2NdgAXVaV-r(7KR}05Sn=&vWhchR)rzI1Wn{jy`tc1Gbabrc7zsxa4zOF` z;$UNwX;MwN0?douYGH98q2#9o@775BrN1Mt_Pph+5=y^RwZqCd+zU2hh53=ujxtkH z;+>%PLpZx}6A+-jvb(kQb_+jFLApG#(6U;SEY;J9S$Od)*mnxuJU&Y}y_S0^cO^y|9gvI^xfk&w)8Li2gi3#!X5cMPrX?C_RGM|h&IX#8G*=u?@fS!c8I`+ed@AY~< zQzIjr)gOFGJnJ;v-0aN6Z=wh-fgd}?0k*RFn#%s}E-~)b>E_|~@v-Oah|nC0S9|vX zZ)3A?9xaWli9qyb8TM-{t5z`nAgX*88dI5M7%L|H_zQUxdh^1&@s~73MP);OnBHJe zu4VZ4_8hk_r;};MC&B>t%BU#Abir2~o;npB%q6N;<>R?@zov`DZ;8Z|&@YxzQN!=D zQ0?rTAkHAQpdfCrUmyFrdQKl{0^7&7AColD34Rk*KXk z-jr^t<>zRigErFkw04NP+u7(kLgG9Oy1+t)@>Vwz%rH$s|Ez_DB0K*7Ocx4yxJ?r_ zj5%^=$C{)wPRdaj+AD@MUjM^)GRo;=ZD`W02Q9qXLR1{KoT+z~>&c&+Qbs@KV6oSz z5##C27U&N7x*drLYIrQD=qOBM<%g3wv-kF#$$0vmiK(%frhaeFaClNVZ2wM_#>D#% z8T(i5d9+hbSwl%wGZ1~ zPsV-}LoVt3X8Y>ui^JZyrO-9bw>6JbPYeky1I?00e4NcnRnG~{;WxeGyI%x;p1@{U znsxDT2;bolF8Uh+Y0T`|gDWq7&yk&Plm?d;)tah-Dm7Y(MvLz&;?%q`-pfr33&X$% zywa4myKLG~;=sle4Jn$HKyu@`T=Xcg-_|5I)@ry%-mO2sJ7Y#u*~U!RJZ*AK{hjft zfxubke*YS&T)lLjFw)K!v0F~J@_!Jfuu{dZE2evT--bza)H@5(N$ zgEl2kdzKa|)HH0Gbr0@w{*;V`7tWk~mI$esYvbE`ogqIEqsEO+yDn()5X)GKt${Ye z1)?@pyc7Yi0n?RNyBgo>wFXWP(@h5f0oOQX-3O@X;7@nS@s>brF zlV9lUZm=PQ-n%Sw7M-N@K(wcPjV(W62(h?8-=jVU9f?cwv5NiI(Imc^=0 z_}Y_s-Vy0zre~G9jm&7at{(9zN?|;3EiJXO6bY4IX{MM}g$>qk9|l`9?3tRp=4Pzw z>sNpLt7u&|k13U+Nzf)HXcB1OhJ9ssR8}i!SLM7@Q1Z%uyRl!#6r9Uc74pq_I1f-Ab! z_hB3iCP5oe9HY)yL-Qh_r2pXfSD|IFh`bt1HnX!+!Iaj)(}M;FM<*G92%Nd>oiXDsMl&R|!lN$qF3>W*^v1iD~2yt`Suz~SVTGNb7!@@-j8~&fKT-yB{ETZFD zQ63){e*IHEf2GtftDOFHV*?F?F~x~S#1gGHR@H|Jlq`hDhTIXM+dXIltj&~a2ZPaFH{RQ#uhpR;jeVYH+w z8i9hM;AZK=G+R$4zZ6Xzn4$ijVP>dEnW#}mr84a@{HVXoi*|^1nj<2TM@6nJn;^d^IrR{a_QXPHuH48L-(M-*L-aglOUExUN zJ_1ADek!yBZ~YZJiZn(*H&=|LgFG*B`#1U_4-~_5kA4i9Zb_Yf~UrD z^pbzrAB&y*IT?5?MI{G14VCf&Sx%exMSAoA~P&36F)6cc;zm6DeegUFPw%@sj}#qVa1W}{^fGQP6990uEh$N& zf%zF^_8^*=dp2A63VNAzx|Og^BqkaX5~7Ysx8!O8okBK9%S2k#YSvJ;>z%E1-|KR` zd@KiM#zS=8!@XL5Xh08ASE~76woHB+5oAT5)kZ_M)_LG|DWZs*H#^EmIV|V z4A!?!UO>Bwztloaua40YBC>eH&t9!%*FgaZ*>E+ac51ul2~+uQ#ZMrcMJvMZa(w-R zT~Cae>{w@4eh@sTmda0c7QYo7nw%kqhl&1rpJN{;68*;v*h;gtL_Wffl7Ii+02a6{ zHLGCSolHBr_&VlQ^|1VUont8NL#^+!fL}$2WN(9%&B<_AOWujEQPlxa+v)m8byG&$ zeV2P{4f{gf4!e#8-fx``;2M;Rc%>7+?&9LYVzDH54wmzU>dGt7G*C0*uc4}HGIY$KKsiew!Dm-q_ySsk*GD(vB2`nN zy-;No84AE>i5te(6&c97i;bO}r!*V0y8z~1R`AUOlvto2d>^$F^F~Jxn$72Qw^fhU zi8=m1EI_FnVe5+hfz7u-+0{dYUm|xuR)KyPf~rBhvCg{Q8talv9Tty~cA>-g+;o(_%v=;YD zoDM7n??-d3hI}U5Jm^g$vY?y}PdQs$C&g6yN3mBy(2J-0^_TXav8%(5=|uQA?OS-H zAklx@HCFn{!$oc8?T(?sL8Kk3C@}ziq zLNBfv>9b}*P6`c0)_}pdcTXzz7KU<*_rN~rWT9c=ylA=cv`n*M>H1#fR#y+*LGks^ zYuXX%zYwYgGX_lvjbgra*#Yw3pjO#g{&YOm^#EAF^=h3jCiRH*nBeqllY4PJKY{=5 z5ys$ljnT*Z%Tn&0UHlFf-=V0(;+v3_wiE&Ln_(MXm0P_}3DV+x%s>CVeW0eJ^Rk`0 zkhy`h143IThqWg-%T4wep93Yp3<0m>G}_0snE-7SZ@r2~B=2{2Joo)3?&KtgD|&$E}$tRt)fgek4s1KtE_J?#=^J#bT& z)CF)PDC3pe!U%x6Jo~IX*k*93OK1tgo6gfJ{n_h#>I$Y0{O({er}OX=t};x_y=8wx zcUassgpp3`$|y5~h<3b-Sir?;D&Urw^u=fX6;k4r0Uw_c$11aH_Z=4GhAXe9gP8fJ zyPhna#v^7uaJ;10R!Y}h*XAF@lmKSn{#y0Yba}4*rX@;ycDuUlUa7Gftw*6|`d1=2 zF&-kN`1>oFjim2QaZV=CKWgau==gD!blP{{p(s_ZBbjY&K8*G;CARG{^P~bsM=EEs)ce?P-`xM2~ z6AF+`O!MZhWN`U^J94dl_x~m?cNsJqV(iCCHj(Q#+`592NZZrU|Jg=w9TT5f#{}GC z{1IrV*sN2$KyK#U@=L*6^vCRy-;?-*k}PDdBl7A)P&y$~m3LYL$|D7SoJv0*zwc$C zo|qV(9*HW=^*AwdKOg(?R+w|)$Dao_jS>^ldM%n;Xb{~F0lTY6LHL5YA@oVPoB`@S z@vl0sFin1)d#wWOJAdI1Ty)xzz)we=adwFW{pL!jD)jBSJU3Rhq?{Y=7JCf-13CCT zwW{w{a8np-!5VFymM&v#(L2-x>k2#AqTxc@J3De|_x;hw59Ku^pe}UTUy8Jh7|-go z(U(-30Z@|C$-3p`f%x@!HOMsl;*zq|FG_9m-#-|rd9bC#Y$7(BOl36MnNQnkKXcv< z5D*QlHf6SaJ(rNd^8F8Do1tlexCz_;{*7lrZq7puQVj0E zthZLr_A*F(2Zb1@(-ZnF8{emxzVj>k#n@pPX!H^3%;04je2M$GFBs%^P5x=N_s47g zF_DD)n{}ql+29}4S?WYwzSF6TB>j5y$|Ny0mD%$55osrUnkJz5^DGE$#v6B9%do}6 zs{-Fc_5LPi)XqKDzbz2ltYtn zGk?A*>4`xIvwK7@D0DXp$KS;4sCSrBqD;oMEf0hkV~V_54P9B;s+`E;aq7kOJ2@Xq zvt;b7V`5+^ky{S>F^KuPh>%kmuwGKq(~WtS%M_bH17)$k_EBxELY6Y)dg&5SmIg=5 zBn>$-NF$=dWMyT4{6I6++}Vkm0qsb^t^E%62jocvA8vP6 zcPOY7(!I?_X~8kd9LzB=b;WEpft}uJD?0G*Dyf&s{kLzY&0@Avb#{haH`p@ro;^B)$95$ym^2l`#%@j0`#czGi=F0v zI>pSz^{ucNAc7=fGMAc7{3}e82Hc3X1&lon|K^^ZVJa|U_(}Ae$_T90IZAklR((G0 z@A<;L52b=SzMj9{TeRG(>R9(}?QRpTVW3AnH;@ z#yviarp8F9dFJ96_IqDaP3gG$Vla62qoQN$1;wwiH(c48KW~tSd4^tp2*NC>S!gQa zzAPL~sSn8VE}K3W&uDi!DeLNvvBVCQhg8ac3;Y#aiqCNxl~mZ?q^P1j^+E+QZ6GPn z>?>8jdIC10U^+;I%$2V^WbF(A#ltvUL|0L<4x9t9GLQ9SCVGV@aCUjC)pqr^H2;1{ zNJ+I5w(9rb8Ub}st(>RIQuwDKxIf=Xh{Q!5{vI$aQdIE$9@f`)a3VpA3UF~twTC~e zvzQB}QSxuaXHj*U_O@x*AZ!_@#CMXn6TqIiO|+bW+vTju(@oq#va88lT$R47o9QZ2 zgDQjDIhJ&dMJ^5QeOUo2b{}r0NX*c(e$RhD2!3YcjfBdTW3QcVz8innXZ$E)biib;zcCV zr7bo|`R1s#(2z~u1{dcgbv7@x<&$-uMNIhV!jF-{R7IZuO0=3ez=q9%!pLaY!reVM zLv4fe(!W7h}lupP#9uyg27xbXBL8Yn0*t$?50fYj(2x+IdHJt2uzkNOPJl49gzDfxBypkWQMr`+dt)T2a<%lC1aVgGi~d%!V}(ME zvr-w=(e~^p{L=jS@@w{Vd;eTkd5%~Z0cHRGQZS8&-buh;0-Kc};iQb1O z(R+)i(R+`G(Mu#5B1jP>gy_8!-G~IydzTPGh!Fj5=iHq0zV{Ef^Mmo(Yxe5v`;_$* zdBDbzcaGX)Mx6EHH`+q(m(#mJs#n$Q$(q1YN!fp?7JX?20>W4@&SxWwR0p{4g#FY} zad3VnakAabGGd&O2%Wr*r@^TPnn?N_f{)Rc%?i1{mi^7^sBMIccQ08jbOL+oF%H;ZgkAnr1M<;T7HM_+ zyfP`S4U?fxQ0Zc2{k>eII^tT;iBE7m%Y8>hO)WZ_=si22L-3$B$Z;uBD}VgBex$+x4)n#vq5ntN1(1451 z^l#%NK`=ES3gtD~;F$yBtdg+N)T0MM@haQECHR%a&u(~-&WrKo=cU&MfZQhimvO%i)I5-c4w3vVPOnNoKi~QOLX6?scSU~9Ch^WHlEM0F?$NmUTz zbE}y*PJQJQt@8~UuO1fXyW9h;GFb1oRQNoV3^n#OPxF{)`OM>%4rLk}?HM9uwp2!?5lR&!%FtR6MjmX_s5!mXdm?g_qz6tg!0S5 z({f$FD(?j^_y|#Hc5re=3vBw$)m7u!*{UP8y~D%%Tl0;hcb>z(2~YG2YjxeM+>a?Veaz(sb94@X_^J^o(1uApa!A>#?q0=T!F845<)U&}=?JUXuy556867Nm%cR3|s{wR-nUQ8J>TXbR=XL~ zdU|mNyf5K36SDHY+Qp%nFIjzUb#-%hcedKJ00_K_Phc!f_|kTxFDp3Dba#;O@Uq_T z^kN{r%<=Bs;_>{qZ`8uWvo6bm`!z;pX>MfIiK= zi3*i)o59>T40klI155+Y7?_~!LV)3&WsQ4P_+n*g_-k!<&b9~XdEMQC0;wO0hO~UnWJ|VX&>Dd30@nyZeEc9Q;m%97PcOy;WuCa) z_zWzJbyMa`C%#&}K#yYzFMof@aD^=i@7FP$@alvFf!y3qt@qNu->B!O5L2qjixz$@ zyG;mWZVLWBQ#v=-0uSKr67_vPrqC`SE?#AtB?v4{?Y;VUZ-H%;DgFGT548&w029-p zj|d5l>bGM@r2sn4IHcG$E>Jw2O{?}Od5cbyxx_?`}w6s?yZuQk`Fm5W@?l!l%fyQN9Xc?|ei`BXMFm;uTet-i0G zo}$17@FOPDH_iRmH6moCYAu$qSq7SqE9ZAs?q@N)o!laUBx;W36GSQTJ;P#0tu4-- zxchtncDJX;HB(23=Anb=K3`dKSH_5ew!W>;yPNjIjAm)i)KzvgW4`-WFgOWgLi0@0 zH)^C4-*($o8u9t~JzXyTEht7usv6V77m%P+@P^kyWkF~Cu!`>+yD#BGz}?@XfI}{+ zn?3oZvJQj#qQp=QAS85l+C937GORJ;P9AW)ks|7+D~qIOS`)U$t?E;%WguOz9cAY6 z2J)i|jiB$6YzQSMFlud_`~UC)%~RX2+_X?#yIB!fbPEZqWh=GaeIpYOkUkthq%0b? zt3JVg&g>dP%Y@*c-I=HIHN9J4NR_XLb#v*(X?lrFY%3*oJ2(R;=FZJi z*XsKivQsQReZ%|WZK|effandqN7*hGo6BV-pS2#HLK>rvj>NS2o*BLG?hjkRfp}xS zK9<1Vq~nh@Dc@_pfE3E(>8qtIDFQCS*HMDc!`{e1UcKAUseJ=?=Tv6}HwaZ5T>&%S zXV-wUw6tpj7LKk&E0T{FMLuU3vEILYWsCweKA0_*5)dJ=c^_u&VagqK!PrrA#=1$9 z>zMJVyAm5iQ4J&&dy2;S{AOgrYf(FN&WE_^UUQa3xr`-CAE_Rcn}r&hITXE zmBEUb0co)LaCG<6dvRwvF|Uq?I=Uo_Ezch;cT!)ZaPN(~swQ0ZDRe!8#2d3BE&zh$ zIB-l21T1~v36ad~vctxv+=;lLRq;?7gA_Jr#JXZR&fRMJu;gzwt~-YIM2x#|6;K-g zzJ-*)$Ppj3x@aZ=)s3_03avL*(*`tE(L{Y6y8Ugu*f6~;0UAq38hhu#2|tg_WJ~wt zkc@nr()ZnKS9&g2>)Qga*I@X$gKRr|W5VOXIZ3pklQV~AHyo5Uv$F4zYf3~QTGJTc zGh*?cr(`Mpl6S5H2Nk4*ZvAElZ&J2P?gszWu+%R`mAF{r8opuIfbvh)J_lId(E6{E zo&#Uz?}Y7%y8#PVKuj+)y}97-_ACW}V0`p?Us*f9OeJJa5JWVNg0*llyR{@LS)8XLs(SO>pBM zbwEY*>{bPCbQKPy7H0uU_J!six5&V?^_zw0{|m$La@sV7P-I;bi}+{I4iIu2qR0DZBQI<@8>ub z`s*^U#s;(9!{IG*3JbPv5jYLXz6gWuC%zRHKR>zu1mgF&oHlFnDoRt)m88c4pppXv z+%%?X`Dp}$y-w@G0573u2!QJxpB@6DKE>H)%q%JaqbC|qBb0Af`j&)lkkQ-8o_7D3 zQ^878G595$?=6}r=x#n<@64gcy3p5h6YpkYtJ08fj~ri`eIU$x0%nF>tu%^57`uD^ zWMQg$YpEcsM8WSc_pK2}uvj8)g33r4Y9RgXBd;lBY?aQ_tBs+@{8JcDGO1cY#iRdY z@`!@^G>KS&hn+{bw`@gKae$9azKPpw=Pnhr15tv{)om*|eRBNwg=+hw*UV3c?kv3D zodWEj%Jb%sf_xL32iLU$Rc7HmTi_lfXlF#6TYTU8f6rzw3ce)F{6hlw9Z#w!2=%RX z;rYx>oLSuNFbxlVpxzC{M7FI~RIJF`tY7y~E*iZ2c7R#jTerX3d>zNrZmU8Vcs<0h$NWlW~zevYl3;(;T`<4DchM?4{~Z#QR#Ny)TMhRcVFurHg`*# z3E=y~;zL2q9^vJGw+KKB;WUbA31rBp7HVUsk%S7TQ8cZ~SXHA$j;KB6Cl^Wm6;i8} zT|p$DdTA6$Ie0JOhwHnl@4jia(2j9~7Hxx_c5d8J+^vKwAF;}j3alAUPp1heB)t?#c z1jSJ^Y?xtGqLvvp9%C*W4Ibvwb(UHVIHeQ8-Pzs7m&X@efn4Uk(p2rej68Y4rh9^a zdag{)PU=0C8&O8Mf)sW@FEecjf#&KS>k|1}_@rqpYx>b_)sX+(WzUX^a8X_nN^K?y z!uiwgy#^a}7mK|cT$(5&usuITIBpr@jUN=mN|y~rKSns}@hY6z|M{`<;gU$TSSnMY z8Qc*Kx7X=2Uga{BEa)1PrLM?)y0VdGm(8KB4U>i=byM0R{wx$CPPl!U9in)bR9b~T zTmT2f(7!G007@r$#h^%%Eg>Eg&;Z@aqF+_TtAQiKvm?<^5pK8P+IvoaD##%uC6Fw- zsrk)cQ4XHTmcXbA!f&Srn-b{}SrMaOP$BU%D;Xay&<)un*^-Fp8Nm5-96*}cb7%w5 zx>DK|YF@1`UZj|`Q-bc1>_$~yjIJRj&25v$tR`%lhnI!~0a@E^=x5-060 zY+GRUJGdt@j<@J^iC%y&3?tyPv!rYHq3`_Ele(e z6u-5jmsB16CUBtV9t}TbsCICoAbDgRZb&SHOhUs3NsbLA)fxB%7cRNSgy-q;id3X9 z+naICOOOOV1d3OGs3=8R>z%LHz_@?O2Ru3;JQ7sYU?NlcdsVLZcJF)YcSz~^)j^$F z>0}bG32)+9>k>Wra6xC_lw5iDM8<)*2Yq>c8Ih@V(*y*9EorDI8HFgM1o$>3a>d%I z16hJOgDa(YyOcX<8pL0yAovbVNX&sy+hohEV6Gg=$NbH?YqE_ld-`xH5!-XNVENy2 zP7>;G8VM$O6qg4Rn>!j%OIk!30xc42KK-0Wz|E6`f|j$y4`#xg-Kj@D8zjqLsC>VY zajm2BmTus7%5%QFqL#3srhu)Rj~D{Vk8rOxSb$-Fh~y#xsVc6ihF$iuh^R<+q9E;{ zN|gK^8=_SlZ3FbLO6xKq^iP5Q>y(x?xC|`ZytF2yd<|oiiH9U)$;#V{*&7Bdjl*9% z#tU+OtE>8VQROw>pxl>$B=NWjMQ%OCzG8aFrj5uGh@2~rj`L<{I2K(yH=&pp#&@Zb zkNj949S2gJvBKUqrI=?!X<}1q8Uy40-PKi)I$|{m^~5f1J!VQX>jp!~CM9LO<24J* zB6|mqNK#b?R`+MtxWDLSP?~I)Ue>hFW);xi6mn+YskME#&=ep(L&9DU zhy8xy9?y#6Wd%RqjTA@O#D9_Q51?R5ITvZ9`Rlk{{kfr>?s>$ggG#x97fZxEAk$`& zerv+yUDzKQUP9Xvyxh1CE_lVP; z(i~C3icG;v8?eOG65Gl4%Qv(I7*0cU7N-#+W4;M zIIJbQq_N#sl?T%eAy4j3ar*;w2qQ6Pq710*Zi+c4=SQVv$rw(8+;DtH*5;@dHKjG$ zwcxC1sU%dltwTcn`Jm?qd56rDjC};@wB1y*!`2}~V^aWn>JgiVLsAl5obYQ{3rIQb zMJN9k@k*vVp~wTSv;cXew;RdlORzp>xJe3of#>%<7iSvte{- zAud{7do8?>y`ji1SoIbfm?e92^I^_!;gA+Oa&I?O;Kx#C;*P5W6c1PzUu~6pl6$yJ z+$Z;k-{KZ<#a+R~-ouG!CeIcZ)dNC1X4Rn#a7 zURLGUDWQf$V6<}G!tGZCGbiq;HZ`obEB`uN-a_eZzL95q$;dAyUgGDkqJZMRH>N`u zcNF;Ve}QfkmigEcK>n0P51*6$-L1c#6#aB}{g;MqAWR8oyY0~bzPv6euy!AJ^Oj;d z_*I0zCbgRKl+2O{2wf$nLLRb(2!J||_772kbJ>W@ajiMNOIWP4g>%f^r@ua#gmShC zLpVc!>-dg2*9Nm%h02dGiXu29YjOmf8UZyQ4}er5_jZFN;a64D>7qIY1%YcZztXR+}wA~QzSYjUbn-kQQpQO4T+$g~NHmnS!HQreM*(~H>3BD@LkK2|YKiWIVi;W(5Mp!AO~ z_}ed~h%7)gR>4GnIvUMad>wzVzKWQguAc!Fg~IhSKEU_pj&%^8vOCZ2ceF{J7$EDx z%+ZWF;l~QU9{y1%;@9b_PHeS99wUPnI}C|GC7go~biqz=8^me$>y1<-)Um4sR^d=c z&=vZdc71y1eX915?U5j%Cj*KlQc^&}yx-aq0Q#40V3(h4PNhF!Agmdb-pf=)0{iwM zHYI8Ozrp{4;<0Cj9c$4gUK;HV{IMqnrX=1tk@iiYN?Zvbv8+~$O}r|Y_s=I)*9Uj? zYSeG{$86b(ficN(fAv|d(P*S?5mA&tRBX%3n1KV25k?=N=J=CM~R$3q>IB zBB^=V099M)zbJj}J6W?Sy`~bsKk1H5e5+gSa|e%S47orU46zJn&mXu<_%BO%rZT8m z2|Uu;_ag-+svSRS3M%$<hVNN}p zoK-h-*1zF@r|{{+%JjsLM9tUpv-F7?=6aUUSZ}UQePT7#L1zaYA?p9&J|=}GvfGx; zi??E~wRrw+p3V4nIpJOaTYcLcWm?aM(d~4 z64%_!@1NlE!9_Vs;kD}QOS+4L7jz>n*#X=?rW+PrZZ6o?fMm_=Z+J*^Qd}`mIq&g60-T}5_%NT z-R>g}g&!KMUO)M&tpnX4fP z5doV$tfrQE6E`=c@siqukJxV*BAU#72F8fjy%!?Do)VAHqBg1FU$W=bk(~bc4FTT% zStJ*+W0N_Y^pZF9L^0xMjB57w>UPIJM|l_(LAxaU`7XmMQ9kDiyW56rpmHZqy;2wr zDW5@|(M)_^yd3GR}? zb!Qr+9rcnCP1k(w%?)i@UXd?rYK+1?q^G&eTPnjn=>J2i!D^A$PohrX$X0?^y>x2M zq&fVGpL;G~(EZD3U^))qbYqGGif}>%Nrycf$T3LovuDGN>Jol_E&>fIyjFq*4jj5> z)ELilFO^;vD2zhyn#(1FB=dcqLT?M-=1#fxcZwsh)YeyOR+4h5ui`pAIE^?G z5Gb=dJg9b1KCfN{wZ9bGV?0M)7`{}~&|Acdmk#mb&*pF(gcWMYpg^lJAj{DH}B^?oNo?D0e2jwqD&>@_8n zqENRzUq1U&4O6G`FcZ-WLdBoBOikAsMCkV)a2@rjzyX0EqN;v=BkC`RAEa&H{E!xp z%eD>v!Qd)!VuOr4h{61#ApVw7u>fl!I^&*Z?tG^FWfoIOdJqH1Ok8+v5S-D76e99} zg%?K`&-UXqPUa3ESi+csVkeCWX^%H|+DH0IE=VMIIy$hL-T)U+HhybphfUBjuz-CU ze);Nvd7}UpI|7jeUkQwCQ4P`R&?BAfRbCLH{!9i+aq{`)f}9Z&UZ4tUkRq#^G*OWN zsUz2CB(aMQPVvx@MRZa)LY((-U$k9BVEI+0BRg-avY*kjgFoC2Oj^h6 zgW`;zLFL2by_F+KJ!0n(1kZYs;9{BQrQR0sClqu5HShDRsfk?Cj2uGVh)U24u zC`cCN{tq$HV2l_Fj?m$tLfvbCuxwUezg zqT9`tlc^xGtYWo6yo=x=wQT}`y+{9|G~4-1P8ah7R*yFlp=x6+hpv6v%$|S~5(%*S z@c3vJC|quzXYwC~P~lCzYth#T-a8328NqK^`yiwCVwhxRH`_;bSv!D#sgdda zN?)bZ`&2?jvyQQ9YG<3e905CI(db;-Ej=O{85pPgjr5`%a9KJD#-0y0ITuP&*@;J$ zB(xVvG3T{Gp6{B|k~y}p1#&dh(#|l09ZI5Wr1FZ8S3H@57B3gB@B8SjA&Uc7k@-h| zvt+P;w*pxKLe7e$E>e(@GkMMV_KYG@ElNR!@S9kkS;61>4Q+N}D8vITqN0~ViXV>J zypdv>v#q+%f~;JX5p+F;r^IO0bj-RMz%b|mK!rXd3_`UfRoqE-J*2%rVS|1?B2 jj!(Cw_y_FuGLOO^npjla3ev~`0Ur%j9hEX=o5=qG;zOek literal 0 HcmV?d00001 diff --git a/en/2023.4/assets/binance_futures_settings.png b/en/2023.4/assets/binance_futures_settings.png new file mode 100644 index 0000000000000000000000000000000000000000..a3f7a2c7032d54c80b2206bad0a38cd5c7729d2a GIT binary patch literal 81500 zcma%iWl&s8xb4O@Ft}^5;KAKp1HnCbaCZw5971ppE`tOQF2UX1-QD$0&aL}?zp9y< zsgds9-Jh37J4$N&Jmla-NB1ppX20D#6sgob=l4EQWSUXUDQw4DI}t^3~- zDv<$=2mr_dSqV`!4};?scOO$tU+&klMl-Lu;ux=FjivNfYrU#Yka5Jx4}b!n^&yVu z$1TbZ#ZB@p)mS{N7#$K9yv=fFfUVx|=w|m;$z~nbXeoWdk*LB3?=#+Z;n88sU8{Wc zm6_A5JCxZx=TYpDs9nF^+Q=w#23FL+tjUBy6F-%xrwdU3+Q@69i(^Xy>(iW&FCdTM z!rT%Rt#f1i?_co0HyZuFcOw7)@0y~>bzjsl%9t8%{(> z9ULsbTozN+?_?d`kt~`bQlLSrs$5P-$0ClEiXRAMvR%MxD~J%NaWKl4u5fr!3@CCEq^;g24PwMT%5Uq`w}RikaAhYb5k%n{(B0q5 z0X9^qUpaJIC(7s(CO-hIj33t4eLtc)e{B>|`Fc~v)l41+fY9IyTF~<_EeL>=Vcf^} zeMq%babbNX6M8i_8=g5OjZAOxnsGaE8&HyQS30%{ot8E>cHD%DhsXhkXLLx9D_<-{N@l+UH9E$dWq z*g$dQ#Kj}d>nYk0reov%+yq&))MpcCZ{@piy`bnV9yC*JcwqdX?z(PvzBI$bZ&4Tz z)p*LT87%;gWmirbQI^IeJ?opPY_7icABTAi2IQkn;}qMlEzitVsxV4G=0J^&lv32gr0&!ZC6ET@0Mt>DJb0bE!x~9x?b$;QV{k+fdC_@#XO>3|5fV;08NM z@s`;wXG8%S=vX#7)2l!Q0Ls&>_|q;*^Yo81?)6F50FO%FR(?M4qhC$=tR>&>52Nl; zTf}+EtG3BXrxZR9JFW(|R%(5dbCKDpX{GQ~^T>iOx@NHQ^#if3cgD9V!7HLuLbrs3 z!*_?_d*L1e>)Dz(cFg6YM%DBCJ6at1Qv}@B!zxAGbc9Smn6YSVGM&|p6aZw6`E(n8 z^{BDl8%;{X#0uTY(Y{NyQ#Sg0MrN)!%-MYB-b74Dy-Rc?G$Wi2JD%uzdSpe<>44PU zT=aCD$blkR#BFtRh)xEKTe@=i?Z{Ja4s?H~fh#|`>e=bOTdu!1F7{gv!?e)0V~uk= zh6IIiOg<`isC?Cdg@X@3ak10H=U5&JKK zHJ-UytMR13Zw`I^%>C0$1OGQ`m)vG_jgO<$JV*S8-4D*JhtmRxuF?PM3tQLPKqdumG#v0J4| zi9qK)b?>7eb1P@E$hT?VE5Nt;S^gI0H-?lc(QI5N3n!D0f)+5wiq2PUqa8K0F9n&+ zOD7-3)`0}B`r`xC(RTJugP-M9_3N%*7-Og~14i)G^$I#)N%7MIo!hc7bev_Vf^t*b$L^rG2(OnYJa9>O~N)m2L9BNyATB2+-H%F?pY z#xIW^DW689uR#hpn2rhx zWV_TKSB+$oq8g8Do}IBNXC9}ki_qDwfwKjK+Dv5?I)sIJHwMNG;<2*IPfwu(EO`(C zRX)WA?_bU*$#(EnGy|prnHtI+pG=VeF%%(n-riJu7EQ#e(yCq%sn>&qy$sa30&Lff z5A6G+QKA@)5J4z;MalfVgDoNHrBgvgC3Y4#i@u=~+ANR1g4PB1V4`5Pl0>2LLD1-BpZ7^vDhOgMA@(W*7%AKjVD`HOgPEC$`So zfmy$`X7!4)$#{5EJ2F=5nq6eMnQ0OFq@ZD468Y~mv|IBieycU41g;29$%4RqZgUGo zkGRzg7F6xr8v47SmaFP@x{uO6FK}d?O^3a5o@ILy#4v2!XEr6#uFCLOuDCF;d3@=K z@f>JG!bh~Yk<8nMg?T#3Xf~qs86rxV2n!fz%QhTJH2mn1K0-y)ay2o$M9a%Pi4OE5cO1wa58j_QiiCkzJ%7)zu6ZBY z3{KB4xH(8~hn%Wy#Yj2-zMF}6%mdd?7w^Q0>O)VO5uk8oZjyOzxdH%)k(x{B0EHSV z;4ktE+Aq2Qs6PpN%b!MWGWpYJMjbL;Ee}23c2b)Fk+{i-1b9H~_vmjPB7WR=Q1XU8r-c-U& z!wgJkh6XJSEWoNWQX=9H4Ff238%wl)!|p$zf&ye{ei}_U0a_ZYw1|iS=fqtO;pIF(bwcOp&xXRX|GBW$)vls$YgQF?oQC<>oLyDIntJ zid4h3m?r2=1a~$-kqzB5>(?Vt6spHvPuzsusptN7y-WMUmNrpA<|oypdbBDZce1%N z#A)x~ z{w?BTMKxPz5=#B8s+N@ln^{Yi3E?@8uvr`h&9|FK4>K!a-16!Fg}}33PV8>Zt z(W@nIt0N29bN}Qt{JwfiK2Bv!Z)mkN!NSzby0@;LtgOG}m`jy-EHW1!ppxxMZ)li? zBwe%E7)KkpneD_K1tpJZNz@ub;yw)CXsCdxqw88m+!#v|Gp{ojXlJiHjLrv5Hmjjh zPIg+a-3b3~Iw<@6!m_j}>0M+P+yG2eDlk3qVt_DQOJmzAsY@(^tN>_6O>-e+T;eAc zOjXk*vI63FmmV*ms1%;bJ{27-(ELIA; z!*n-fZ`&9Ed!b{29Sx0_o56*WexG7phe`8UV2^WdsHTf=usJLEQ`%@~c~ z*66a~*zjp~9`VO{^axd8T*9Y?mgW^iGo0_WEch_LxsGk7l`?&LY}+68ZAl$C-F3TQ0zAGqt7w1cOGFJj zH*k@10}2%Ig#KcS6&_~Zl(dmPl4!ZO9lJ2Q+>O}{aD<|a8gcB6HB$zUKNxr_a=mM< z)S}AoX9Fy6yvj#K_B^ytM*megR6qb8Y~DiCDX#&19(ATFOQTL|jY}Pwmu^JEpC-5Z z86MOYO!Uz|V_7$_*w%CJr0HAiTc>M&0e~gLqPsh1LJAylDhZ$GBFPUA-IJIV+e9=$ zTFF-N6$|)h57~jFmNe_lbe4G<$8>=WdY_%uJ7OoM+s;R?Y|oX<>Hlg)7sC7Qz`f;? zQ-X?%{t2amyMhK9L2hiUmp62fVBp|`Jhv_1;pWNJJum@bkuSXhWj zf&g&}l{*O|qStR9brnp&_LW>$$Ql7a0C5^2BhzR z5%Msv*E-lWsM0dEHlK2aulYu`HwgfNbR)N!#5BN<=mld>>&#j2)9W$`Sb6n2(Z658 zx~Si31lu^s^}jxydhLZ%G^x`#!^t8DT5$PCuibufIZP>@u#r5c_cC;8uPCq}|FD?y zB?W~hHH>LQ)hmzJf60rHAo0#+% z2Jl1?VqnRa;i;weWO&ny4(Gjd2nc`|iIL6=x$@?TpGni&PecUzH@>2i>$XZ-%r>|u zzyRb>Smp5lTDwaM`2!R;t3#YqYu`*=)W3R010LAGel@=KPt2+-7A2j68Td*R8f?8! zcsp_V*OM30<{05)+usT27*9#~YF3ZfQ2&7QRwfMy)+=F0OxW(JR8Bpu&h` zJjq@CV~f|4f&wiqp2OGtIpa<)m|bmF_x)Dq@4}{GHg-}P8O=pC6s({t8zA^g3TIbJ zzY|}k9R0nVUU%LH0Tc^Ml=?k0HMOw^ct_$s#|*}SNw&;cqQ1a69Jw?i;;9Dr;}RHv zm+vDVEv?kk^#~LHos)<|7x$)lN#uXSb#G#5%dyzK>ts#FE(6G^5OWdJm4X>E%7heYA;vL#09!Hbu1-40!wxCGe@V(oV_VCi%a-f)XNY6P zkyj%N?|WU)rbldH#DGpcuG4-ofbP4%XE-$8{n=wMo>;ZF*TQ^ORGeGwq+kj4&FW0ht1JXmB?x{M*RIfAHHbTLO_P6;Oxo z&z5@cL3ygz1Fv6DJ$ywxc7;e09AI<$DX%Hu-}TIzPsPZh)@XyofM6Jx_=GkF41ag} zy2niUxBPlYI*!@;w2HW6{-rY#a_Oo)7hvH8n{PdTBxO2dtsRC&xaVforiKn`IfOo5 z^b*+@YWg}E2X#6ed9Rr?EM$2-CbQSDA6o53NW4$R+KMgunXS-mome-LiHOuSq`}Y$ z3qZD*7cJmZdUa;9Nd10`{;*`%e^(#cm|?VI+?IV6h^uXQm^*rVeSLci6(E%-=yl!S z=Ig!Y^-*_T^o_`LZ+_UDZ2aH3Tx<4U~Da*~|YET_=De7Zh#2mVk*Zhu77 zM-2z_h}NbPuivCB|BlgEu4pRG5*q+=UtV6C-A>gwwV=kXQhrQFU=^e?Ac*b%E!pON zzi}hUeE6X`6BPKd7r7VKg0E@T-`nbcZHlPIvyUnYsm~E`nA!MC+!r1=K0kjRpDjdy z8gh7ah)z0h&q3%b^p-ao9 zJxyfany5enwJzSU!X>b2XH#60!o==eQ`vG!NOCcKbS)igpnp=DWL8q?zrH>288%EK zkJo8i##T|0K9i7;P{{R=!@TuiDIoPnr9GpoqE+1rykHWJLRc;xK2{(*zg%&81s=+O z4pUEl?j!D-Id@bbJgcUe&Ffs>&@PM_&3AkYuH<{ZtTenpcy_I;JM$f5iWPkA*0paX zK?~JaS8w;apq7k_tk22(^1wZ_y1H9h>UDTvfjKb!+Ujlh=d{Tp%QtB?Z%X@eda9i& zc-iSnvDy7Zz;6>4(DOdwEKz%;74}ud?m4RP)}+HGBqeCv-JdXJyXnu1Map$psXJ3m zsF+%@sBTAab|eqht1es967yO-C5Qc;YZUdd*ToIb<`DHMlJ8^uKY)4jWTocrbN2I! zvhSt4@I!MpYpMdrDbO)}t+poe_WazEOULha%(LX|X6{Clsf7Lxodq_?)zuYQ16#5) zSf)A{>BFM=f#S*H!cA|wJ-m318vVm;l?e_Gj(JTg*OiArH6?=<0Wl#nOCbyl4en$waa(^D z78>=K>crBQ6)yzr*;3cu+U_REpn!lB#X=b>XaL`WPhRyPK?4j0fM+mcX-%oworOtb zVFE_XXRJ8M-sbz*xPcBw;!qy0MUROAz~8RQkKs=T*ji&Jqm#98#z?j|#VG3A_)HX$ zu0VCL!}|Dy1m(uf=&PZgfOuh{b!=|2&L50uVuzEt;B;u{2(emt*#hDHPW<C;?|_7{$M4GDCxi)_e-4S#56r~E5|pI7a>B!V)>hQ@ zHMRAxMs+Lo&1&0lP?!wjzoalI(SAARFJUp%TrkBuspM=39S2MD}&V^w+R={Mm* zjv=m(R5O`vsUx26MK5*tOsrGCB%$flC(DV)a$EIxAksoE6;o7GD2bAVxJjf2I1`^p zea@ab_>-$`Yi@r!P75mjLJ(N9E*%+XH=rKi7-`J4|)yqF1P2_cYVhymg~{L>hd0RjMcWZ7F<)h)@JN@k_u-;J8J ztsh)}FWlt{nn+Nf5;*OYp}NwB0KMR12Z~@Hp|kmY)|biVmf42KC@3@ZROdOYglmUS|+;N?JlRrhl~*HZeBgL zwhBV3-gaJrdEK4Ts>S2R($mjH+O4mYtGQSe8jwF!G>GUU*Fm~@-b}~E*VKKm@eXH(vI}a_c2Y4QX-}^ zrW((zJywc{K%9`%_o9dk!#7cC<}sQagk1R1E#&<_O<4)S)`W~23}SSokukB2?>B}U zIQS58kU|;g!L@uZ<3{$D6b|R5?$O@Me66b+4o(7iXrTh98_T~i#fqy-n`+C%@@59A zx9>)jV`7b?JPmjp?&*MOPL4TF4w#Pp{s4D7x3gF9^Xr4-n6cpngxZLXMjIPdfC^Yq zQ8n0L!;i`!&&NcnbnoToZ|$U&v}gUJ=SmZ)j@{$Jhl31H(<_yDEY+YBU0>KXVO%vq zOG_9TgLiwDsS;J203{ez7Rr{Oult8v%Znmg&!ph2;2p+-f z9ehv|8UPTZSwLJ?3M@~9xn=Tx(}uE)p1?uBK!gXT31 z5txYgy@Ky~y!p!MHKSYe)hef8L?|)RGw@O1@GNIMD7OQoM|g+ahY4pNuV{XD6aDg247;lk&O$Y#$K zIN;QLMj0%ZVFu%CVI~_PlM^tetD{4T6slkzsgakNc5W24XI@B8ukR)uZodlo>LWM5 z;#%m`=dFsg$Af{39XknM3(eAwt}cekHm~)LS9g794iKWi)ALnr!`E_K;8(l9NabAn zQz4OuK}uGZ-)o3ri7vL+PU2hzuh+IXS&QP&{MT>aR@;OCkCU6ZSYqDG&b8+N#v$>) ze`%Gf?SC%8Q*d~58t*JQ*BmbHymXzCr;PVRA!|Ua@QQtR2$rQ>LmII&!kCqi-~Cjv z+vGG(l#!ueC-Ee+*?TN2qG$iq&8UW7LX=%#nx$xp=P5<6&b4MLex)cDrDc468InWi zQa|?{bXHb1j;N^e336f}g}T7|+A0QVfFa(hj{R^Q-f(v;EWrD5_i7Zn!^y<$p%GG! zg$90GNc|9nlQ*IES(gQQb32%}{bqAVY-oqQTq9e7`Z>!}`=^VJzL~47ZfI3QQTi-b zgS%505edo9$*(GTRXQiceIbD;e}i5WnQS33DO`dWnCx+iwGzwH0MiEdU(ZnmZ#_JI z-*t@T&kCbl|8#=3rpCZNBP*>VLe)ThWpUN@aeK*{5S}{toLE}f zrC^E(vNa`Y^q8Xs4a2^Z{4~MS+zm?6-U4~z@IfVwb>eY5;TI~4YI!pwV$hqiHZ=9f{W#>eBZ@N(Rc#yT$%b3vgjd z`K4|VvTD%gu6fZu4hurh+S#zD+T3h%qB?SJiUJM9(QQ6{U8_Pt>^6=IznD+7=!X*UmILUFN4DWvIz z!0r&X`BmQ=Eh$OU%c4lpTt2+UW|&>9s8+g5l$QxY$>A&RdN9(_16vt^*M67Mx*#;dwcva_=p>MyzKj{EZ5kMhW&l;YM|7#PWi12 zRf0lE^X4%c>&>6SU zlnAbq+>Qs!n4kmH=y~VYf{?VIIcs9$+c>Ujru?$;fUX?4wP!uquiDJYvC(K2_V8Uf zzoft{;9}&^puN_L1!GqpdD`CP#s)UgBPQ1i-QfJsl!ZUGM)u*_JIky?Uub{w@}%$cSj=k{T_*qaVF+KITdtn};I zyleeard~$iai38|x}FMl))rL2-k)VF9^eypYW+5k&>co81lv1@pzG1Bs5Mgg#PNN2 z(CT0%ec(KIhEN>Qq>K*I@0tXH79$*`0MPAedjx3n&QtL83t(EWwP#**CB)d@6VUvcS2`|v{;*t@dF&vVLetbK(;%VCXwDbOu&RqjO6=NU2=(+ zNTv@1br_P0iQr!2ny;Kq3R)szdk-z4`E7Rf*f4-TL20QNVxp#6VnUL2-Qx0v6nLt~ zL`5vYgHc0UKN5t8#g>{WcgB!JwF} zVe_HoVYSyHE;Ei)PlNLj7NLR?Q#R7bG{sqMHaN(h51YV17xzyMUHItxPHo>?niH7p z;%VDR0l6EWm_O0P%Sy&cs;I3%QX#p>FP9|eJ!CclTONpxL5>;&I$BE8KzpX%S@Z1R zb1Ia!n9Qm(BCzp^)%|HeLJXU8LBj}AVd&Mt$7{d00ssm~ajD^~AK3#BIEMsHTt8H4 zom_Y@-qFAYz_quRwa9^|aM}i~@0(Y?&rTzy{dIhPtnet&BCNms5Rpq4v9`yl%b2J> zoV_qPB`EOwY#k$hKrBx<1P2aM(-u{Q?|DA@%64k4<24*_c#0-r0s&QB7DJw* z03fN7)QkavBJyG5X~j_9$;V^?tl$YX#+pATADUs}hwDp_HTG3hRxINY1uc|lFdZMc zSt1Bywjeol2S>dTw=o7(x*tHUe;Rj}myh8JkRwt8$AS*7eh}(Eu$W%^UGOLCJ~UAf zZJ%H*%$j-4+$^`FKp^GVxz)@pCnQWLMaA|~e$iHe72Q8x{F|o$O>``*tFCXTuK$^q ztKjp<dQ4LBkbtXka`RDeMoWf=N}rw~*nnN_qhn^56(G!^xKqX-j-pYna4+OqoH6 zN<}TJl<`Sc5Oh%TcEtNYS~zoV3`Ky}v61DX6c5`(uNKAhOD6*GNhPNO00Oia(-g)* zSZK_!x1=qL>iS2dq@OtRwp%J`U=gFIgF)p|OI0TB9E1e!=tYws#AIAuGe0&jQPeHF zdWwF+rf#>Ol*X3KTyZ@k=W-Os1BQ6vmnS|82H=3me*N0<{f)RL?M|$E{jqXre|AinK8wB zGR}fk9k2Y5v@|@Fh&^jqE)8jMCuK5CpvxK>)+vUUI~_By!pIME+Q|M_LoE*VN@cRMOIlL|zsd%el*76b$I z@(?4ZG2LENO%(Le2MmKIRoyHOAobHz`Dp$Wk@;RB6gfVsn))0eDh(f;bO%ljj%(MBX-P-`xJ}(+7?xN)&OC4#2z#L zUM{>|9!#`Eo*^1qfl@q5LZlB7#?@LrZLW3#A3vUEWf}Oaki<-0MwG_oyxN>iCxIkX zg1W<6cJ0rh`>@2Elmp4b%)x>+A{#zpIU%KoP1To!1(^vs>tY4qB%_QauB0!BP~ef% z(uyq)NU(UCB@-Zz?>1uK@+^i22w$1xe-&qrA>njazr zK61&?nqi84e;niFYS7S_p~4F9e09E#X!ER8Dmpt}3@ zQvOc;J5pKx9A;y@(^xw6EIHA)xdj!HV+Q$icSW0DOi@&4k{L|q?0d!k$stO^IABJlfiwUqeR>}^Hk!Us_h*qR6e)7abL0=aR& z?foipN@k2Wr2bNfiR4W!9LH2mVDQb+riLxaXT(|J5Ah_*4OU~Hk3l*#{)~bRqj5`) znQ5-=Q)<#!xdRt<*Jy835L9Vt#l%r$X#Ng9s+6~^c`ptQDyoQ3C(SO+BVXA**I zTpagEAwgq$apP`;7RwQOozBsT7!QpJ5TT4C4^Gff%jNAzp~9@nRdHRR>0Kt#_jWs8 zJxnbPO%)C1r}OJJBgDpy2lFACq)!46>L@oUYoF`L6H+8 zRlW~gCrK5R^7=M<&1X}K8I1p^4?P6-LFE+?s(Su)*Q#A63H z@bNu5*R=@A-&hi6b_|VP3?#plp<;W(Kzu9fba(;gN>N!+!Gb+Lm+pO!((MnK?lGJr znXC-_^o&pK$dGVWQ@*W=Cq{?Y2r}Jc_zmm!kvzB!GgAMv&Fk!6rM-4djSu(?)r<@g zj0|r^bq!SUq1CiSUd7!Dc{B9YW#bztL7KjX~Ezlt3yK1Z^=}=xX~GR z9Q^E4@jB)8RVC3|4>DBSj_7-;Vcs5)B04Ea9fE5G{4ixBK!$hHZSa-~Jv>ZbT6A|J zCL|$9yBi(#wt!P9E3w5+57weOBqg1%vS#tcW$g`m5p;d0*nCIYrKOE{{}pyXDB`MZ zMONEWQrojLI6uB`?efpRq4=6_0Wpl9_=jgF$BT-0-+tim#)TzoszvCKfHK1R;Gewr zE6l5@qCKVP!KIm_Y{9>truijrFDO&rSn3o;Hd#QClHT4*5VX(whzLW>g^Wdq^DI-b zNHwt{?R%j{FnJ~fjO$$XBeRxtdc#h@Nh;eDP!Im22aH3LV-Rcp^t~APl!&N~v)rf+ z-2p|+r33X#-CEDdJrbQ97lH=dG?!X4lv+(5Og}FxP~z5Ye@{skiuIg>3xHJUgC2D? zwJo(Rm<)0pgoW&TtQub4O3hY>gY^1@DtUBHY;o=G9n`zAaz|}K?H<_4{?_Oa@r{?L z$NjbED_e^E44037!-n{x0$)I7cdk1P7C`>~TGo4jK!xfWJ2sfKiruzx>4 zfRD=o1^&7PTA#2+qp73{HZ{{g1m#dQrr3FSbp+Ap0VQoA6V{e5TqfE{rrJqgPrEoe zY*rY;6mXZM>@v6hz=XznE&omK+l~KIlX0YuzCpsK72Cs-D)hFju!;h>;~uk1WH*U_ z4?L-l?GXptNU_-5)c8_lXQS+9O)lCGTX4<60S`)qEo{NkN^75W(hYyEd6W!!tD8uM zSl>D=l#dv37%ppvN3Qj=y-^cJC!rM?B!R^Iu_9l;rSQsP3 zyr;sX5aC6L`-+?`yoX(#ckO&noc799?mYB+Bfghr=<)o*Q`-nIn4c8s_(!+tUhn(u z{j|lfm!Ywfp&9+kvBs5_J$;}Gi@SxVtw*SN?rZ;E+P)A!cwF4;ub}69;mQ;y!_ta4 z#2iO@Ci!YA27Ih2C@78nLCG#>Vt^8s0a@dB;752dp#Jvhfd;m++{a_Mz~ZYw!3VdU zb?*X|8x@GQSXRrfJ(c9(ntSrK^1SO#VV7+Gh9Y<$H{-_%6a&7$8VQFs+7jxU4cis5 zyM=%h$ZOL4sq!*{3PR3Auoo;kAuYQEo;wMxRS$4)wbVo zbncC*u76G3uSs3Oilvq;27|As;jriFFjA{ zCwBgfxX=MXQb!kBG)5(r<*%O3dc%9P%D1|64?m1!-DGnfAr#Mm&ki#kZ|oJzuwbSf)Ix{V>n=^19V6jq$Lcp^dbkUhk@}iQ&_PeyPX0tyeP1UP z3X6HJ)Ib|1rpDA@8 z2$;$k7ppBd+?++9oHZ`;%cFWLb7vi`!H%NY*^E24E6*k?57efK=y~Ug;N`Y-+`)!3 zc;L0_RG6+D4uC#?nA8F8X-L}iE}M!^k6PS)SK*VZafA_=KTzF`9a223aMOG)Qmix| z3T_7gyS1>UMIUVI1X<)wuh+IKqGP$s3+W#E!v7ZeDO$-sb1Qx_{8DLSyxLGB?U{D8us+kHlp+3x2hi9ZO&4Y%ma*@wP=L=V^&8axBi^{L5QjF zwbyM>diz@)k-bDL+b!yLI5&CRa|OLgPK+875&bRO(E6NVIepk#b7A&miY>F@^6}OoTR4zX3?ghZ8fScyFDh5=P`Zo zSN8Cu{nOzrAjUyv>SaYI1~Bks%4ylw_kgRm&JMRHaVU6dl&XIFkX(B~s`&mFOEz5y zD(ZQ2q{880Mn(z6C#Cag01|0fHSxX9X<(lN9l9)4H7k3He4GVh|JbbS5TYRVy0AaD zP~x_0gn$SeKqtQjU;p@3q%9t%H9cd7I_J@b;|lFKMCTh<_TidB0}kUxwo?R(#IzV} zOa-GK275`_|J|6P$Bak;=^U#__0F^*_k)Or?fp^s`PcN|`n{Nl%J%b5)HJ%3s|BO^ zL2k~EkXEunf7k&2#Ta4!(Lor+n?dqKUCPVck}Xc@nR2^zPWYz3%f@QP;=TF5JsBb* z!G|RuOGg>?sMlcOlXLo9&wkaRX+MPtS3?g0(@0r(WT$&;SnxEqc~p*DL=4BYd2td7 zIC@peBqC_+{<~4;^2f1KF?;TO&8l5&c0!_W(V%1~f#zwlwp(SpSwt$zds4%ddWIn^ zAzf7_tjowB+@D2$XaNn zHPFpNJ_OGQK*K^|ADDbGiNRCuXqZ@O2`>8E=B1Ln*0>|vQ&f;ofK<*0qjl{WP4JH# zA52UIDX(is-YD;yPtOc>0RGdB5K>ZXZt;UwPJEkYjV>RQh@`Vd6`MI&lJZMC%aTf+ zK?ZH+`QU*rqf&svSfPZBf{OW$nmc8#-ROU3Txw$!sG`E2KUD%w#=wQ3XWs!xMGa?y z8WHGd-Us8eDr-P>7b^!D(V4TMzMMXn63@eK;Y+b<~|yE!?UASaLv%r1-^-Pj;;bm^(6<3Yj)D31!SRl z7(tz3AUUc{sO+(`?6m|uk>V+WyY~VE7|{!XfSWPLcNwwusGrq_8_M7%_J7KfM3jj* zVd$|`rtG^4dJm`ohQ!Ca&R6AzT1iBR?mNt?s3NyoTmEH@`dvynD_hFR8lQc`whm(G z=A;O{p8&1`Rv@T4Et4B1S0j)|YD$%O_(<I zpXMO5mD!KlOa}uKvIg>_^C>;5GdtmDw3ia}aZA6gd2r5@B$T@UT42UaxOADbeA(;$ z&kB7I0mow1q`OisPiP?EY<9X$YmVYr+afB(*FgB9`m21@sPqy^|rm z4cV_>TeZUmIc5v@O8J!WN8@1~yQ#9rkJKLZV^<4^LE%e9pL)x&dXCzt%t2ifDBfvm z2LCt&xq4I1q30z#=+56*HS0bb5Y8at_~!Fh40vJpA|f-9d>tE;phm z=iC&-tldBDc2-I-Mp5l5JCBm#?CnX9$H&9=2+zc!9x>9XLJ5gobDqCpZVJAa>ze-a zE6!h^t3d%`{r(B+yP7ZZ`Y&M`kw_}oj12cI)`zRVv3c9alWBYHVx%nF@sbyMvKX;i zak;h3Fpe%etZSRh3>2XOuIBjv6~m8V7;G0B+ zE3BJ=Ph>ds=w;zF5&nF{7Wl$7X34RTdK32J7`{Bxxy4CS3)O5vsailDIVF0jiPxat zvOaE;_AiaduZj|;Or;X*|1LM84+83K@^m{Bk2xb!ERqx&MLC*-)zNZP8HIT`_H~?C z-(7Xl>MVYjc-ISyrGR*HCl(Bfmf{yY7?9U}=wx_O{(@1{=wTpJRL!zr4$ETkmACZ= zft7&e@O8*3w`n2r=?q>f`#mYQemYDK2?nUctN@0 z9L*#u`e`LlGOm4#v{Vf{W@xl}EQ*(zw{0(49)~2?n3*Cqr}ygztRr(ay#BHbXdpf4 z9VbMO#;Sm^5F8wkb({u}m{N)O58Dzzp$NE7n)>-Wh|=jzn5!m$^Wut5(86^yoD=~M zgRtNr&j;|YYKQ1mn6Ve`?U49P4RnW&XmTP=R3dpQ;PBhEk#-Ypj?0r}jO~Kk=6tTRIa13(Jsw0O-f@DEfoGMyeLPh`4(8YKP-;rzO=~fR!u_YH-Wf zdI%IMu^wVc)$Hv1o!$fIfu$zAp4kHl2mooOw6c}>F8KX+Epby9emoC)ppM8@7#kM2 z=vF56+Mam6bOlLXu)NJ_LTEU(qAA}UYyXEyc^0tlzcU*y<&~4Of9a6!6?k9S`rPu5 zz+82GVXC`{9(#Vi2=FHboO}%$MWrH4@#rUm=zH(e&2|b@AKltn=Dt7XyU#L2tg$#^ zc0mDq#k#PnIPzcX#|7jNyHr_ci>B z@7kN_m%nw52fajVzEg^KzHbNqyGY@th0ehD}RZDFPk7z6|4fg+vT_Vi$R>Vqa~lNf(4&JCQNA||dFae+{5 z=-}N-GrQDA>=MleySh{MmA9Mnvu!_Y*bRicm8bAn-{U@EKc3VjR+AHY`axj5NZ5Pw zC{uIppG0q12O>dnj7^sU7c4L&HvvY=TZJV@E&G-@y|5q|;>W4Q$EW4L{3v5p(ZK5$ z8;G((p}_H89|N4h$? z)2`MkHt!(VCd53az%0Vt0`t-=4Yh6t&Mza^PJV}kq{rSjJA3;(SGg+&KmerI>IQt< z+P><6KXYOsJ;YAL)vJlE!)5L9R_s&uQESD{7v#yrYU)tX(yDGysj#ZemRNJ^Gm^pp zkprl}9E5jed3(x!eSUjUm6aW!6?)}QOiYxnY|Y~G9`m+)*w1w5;F_lWS+dm^`!?6m z&d$z`0z^l@Lw{cPTihL>B}E8WCaxSE9UWVBog2;M_sBXU!@ocE_dlCHn5v})mhW2M z-~-z4FEw9}m0wNY+FH?tuFvmIThW1x(Vt_!w>8o+dS+ayG&F^z-p}FL0uQFAI(N?- zd;WWQWWHlh!sm@+KD&kKcH>qYeKBNjb8k=fS67jFH}F9l6>Cr1(cH`HE=l(=-;(`b z2HoGD{4v8}%V#V(+Mh0BTkixL5)u>PaB&ShHZDy&20Yt<#pR_ZHq>F(mw1hJGV)?;$EZpX+eimeq&Fv@~5iKX<;&wyPL^ z0h6rGoDIX)drZLRu|oM}0I!{um-A+SD^^)SVcBIX`pkdr$-w8Lz>=}h2ov=?BF)T9 z#Z%i`(42vP-jjCn`CVlpAvzae1Fj(>N*M7ntOZHC0 zcUkH7^;(QDK|!=inS>-HObiS!tM2Vfvx}Ekh5|Q7nugBvC+)6#B)Q~m&YQ5Uvm4?F z?ryHU@^Tpk)P1W$NlLQPv7`&F-g9IA7udGs5u4oY7dJ%qe_mbgQZ)^|nX+Ws-e&Je z1sv87EV3wq&js%6-}Vew-63*HpQG}kEI*o58J~M`vc+bLJCjgEN*n~{swGK%)jx`T2odH@Nh^+L3Hv`>c!@JYQ?sLY+^lfQzX4Q&y zArMF*gQBn-cg!!V_54-eD;~bJ?g!H? zK0IeXUR%7MpaTS+`%Ug%@7&v;3)=6;Pu%;f73>=uOAc7tt`p~0J>EO_0fsGmq@jN= zQk#Y6gjoRqUGTYI6Y_IqpJ(?h?+woeZ;rXLz2!z)zst{F53%j%((bE{Pp`gDmORZ{ zw@dCWZqFxgcXwmYm)eW#5bJ)PV)4tVs2DvVeOb3{AenD}TWf#42_^H4n$0>Dx~i-6 z+o@lB7CD87RK6)E@7vcR7ab0v#SqB>=-?lunXi#I(QAYx8z?oJ&V0U=V^91AI}rE~ zBuN28>IT|UL9ycD?XAlZazt|)+lGA7?n>6;0h1oo$=rzv~%#2e$i=Dn8A02P2|;Pb&f6~3&u7yfa}+rDvuohwToLac%z{*$wG z_N(!Pd-bhtp|P=#)SpuG38Hh%%*^zCd$Joz{Ff}2v<)qqDzk}8{^czSVOYinL>rhx{w z06YH++1mYBD5)P+dvpR`7x1gDfRi*9AxO05!;~TdFvZR;073D&G6a|6=eY9f>fnIh z-oC3!G5)_gI_Ya3M@%|D>6ZyA8zIbMETJm5mp}51kzMJa#f;qz>F*%3v18KgryZQIU7T zMK2_!Z*_)}KKA;*CO5fA(!@^~u?pM`70uuPyT@TA5fi4c0rtLEM2E;g%|IFRAX};e zaQ@`J_5yk8*Ot-+vahrQzhC@ffqJA)8SR2IIUL*FxT}ssdB~!86Qta;V)TxHnXvbQ|dIYe)1)!+;SNl2c44 zK;Hg)TyC*~cARE(jk{rKB^o-2@s`$Vo{d?ukYM|t+XUp60wnHDFexyl8r2OfmpyDc zZ29?X!qV!NMbp_yqmkQCo+GZJ`MDt`g2$c)njbXY{P^&qc+7Hw5CFa@(i2JvI-RWF zi(kbxK70Gd0QMIY4g-S$IC+KcOcZYXOL4llz>D{YitIv-#6=2&G{bv>rsy@V-;Ctp zxX6t(T-pN$krVA$bw7gs>`&zm-F_wMP2E)|H9!rP2==+$-+}K9EV+E#NcY9S2nVQ1 zo@I)-y=nE`u)%4Pd4a}EQnsUa&CzXN*Hl}pta(r_jEsf!zIXJfe9YM368;#u5qCQm z*z8aekPR5VjRi;m4iG?8rQEPvR#RbLaem93-G5kd#TTOZNgw(=fUPb91oQ#-+5y)OcZjX0WNhBIS5BED0hjrQ|8|Hj^KN^? z#m_R09z4`uoRSiKkv0WwV}H1qXX?O4KOsm57QM=$nw*?i=QZf78WA@#()IzZc>?oH zmCY1Etf;f`gBWntrLJ-~J?7PMKDA-Zs63=JE@I0GDD zjW-~Wvfb|Pu(ErY zQBkUeT>P2G!%^(R&B(*m8ixNBNDhw(=@4#oCvVs-1eI4ZUW;_>_R_(3EuiHzooXRq z>6E>hgIic1Uf2QxP8sR2PZpnGN=e5c$9u_j4aewqSKGcG;a%rbU0HY%wpU8w)YLR` z2DGVU`kvhEmEPaB-p+x((9Y4Z*NpFxawUMR!dvTHbL3SAYsTcG#v*q@nN&Ga~CWewX`Rh1$mALP6 z5S8drTfY-)CSC5~^!>*OaE z*lmK_h&RpFii-3;=uc~1LJ7JlKo{m(x(%@brE%LO_OOzdNVUPl=vw?LX0F9KS$LSA zS0OgEv^FWj=l^}<>wVUUiyKxSsVPaJ#$-$|`-PmPL-R4IG-{8a1-zv9+uk2J{-vPl zrIG?|;?%vc1!zMB!u$MqHI8oOr$5=WQ!@z=R&iP9;Mgzmc*YxV<5ben_gOPs_aQvm z;XDqF&sMJUP?gcPzg0`1>QKljs5t!_eNB~_kO^#3U2Aa z6pu7pk0%wot-Lh&Ckx@`isMZ=dAYhJm@F=n6;}O>cGO%$MqOPP&fonwoptgp1Mu^h zS!F0CS`xUZz&qSF1QN9^A9LM2;VQb}e`o&!Ey8OP3w&dc-!DzxgZ1B?(XXHyPQ(89 zTDqN-{lAl>hF}2A>6-IqkMCZ&LgPL|YUX+V_Xb6AY2ph6`uJVC0DG?oDY-K8o}3x&hsG4p?VdndlXsL&A+5cp=K^LuKKVX`@w z-pWoFQ1>|zd$_8M4Ad6Ne z{lwulcJt#E%uP($pueh`8PswTTsytKeNJxBY#HZbVrlv1LqTzA0TS4X(QLKmEf(ug zp_eu=|Axf#zq@wV)a4!O#b#q;^9~O38|34P>@tb6&_XpF_b8QlT&qTuZJ+sA7OrwrxXdUKrE zz_+ziRztC%w?Nx)95^YbO?u8Ub%+Vxe*^^PQS5Q=i<3A%UWlbuK~b55Wx0xh-OIAk zl*!+3!>ED4mdIJU&p)@{Z~bIQHTI+%dFlwOAig;M-%_=iaTRuV*B6Xe=-HXXLG9wQ zWr1?tGA`wm&j1jaZLhIRh)>$5^xwfalT)IcbcBr`yrG6{2^F!BPBy@pu2ZLiXM?d+;5W>#L_S7Pr9$d%>g_u=sL z*QRo9uTlOdCAu*mNoK8_6h*aEb$%wM(BMa4U|?W9jY{Kk;QTwZFwW<)byMCesIwNz z_QVF58r=N7Q2t_)dRZ#=h^C$^iT_EbQhC86U+(=jeS6#g z3Hbl_;?<6vM9kwKF<*IFT7CbS>P}%CP5ih}xKr(?=~FKLJx>3V4@{?mbAB^#ng4|B zRfs+Q|DPA~)GtDxNDeGCbHDy)Mtn&fB&(p$%SQ%Wom?Ro^ru=jg*WPKc^^y z6_V_4Z7vw#N6bOU4o39qzy4-S=&RpB$n3Z`hN0bKbla)N?L0wEZVOmA}c$ z>o(>q_W$iqa^4Dq;3KRbLYfSO2#dvs{bx9=$%x6S6}kSF6z2b99$A|E^*$0 z@qr$Tnu{~n#23V**h2PM1Gp~<_BP=92H*KbL@M}mOif>JJ@si60*rSeYg)Kkcd^Wwc62{Qoy_W`l5dQ25v3@GEeO}9d_@N= z?W|=Ll?fxeyh|nr<~-gFf1omI82eLJ@TRP&1+d*)-0T@phzgeYD+={Rzwv&067y5B zyK6v8M!E;NEn&oOsp2@H=ecWbjbnzNq@e4_W)HJGcsPv+6ncB>9jQ#AGBEfg>dQZC zo@<8Zv5X+#WfWjxK#A7E*y5X0tYxeaTX|#EpQ%mN4Hd)IHFUXV zlF~yv`<8b0f1;x@52s^d2OY*LVSZ4^HN&&m0_nsugRRZYVT&hde+6CEgy-~o(u5!p zU2KMnVsKU^+X*z_-fuzX*}1* zF)C^({jdr~VB;(ngrWp+Jh=)2*YA4#|MP1{8+N8!#?x>+~4lzdK<#z}b!5$UtOd zGBc6Fv;FH2Ibc4~qu|r(UG)W@kEX7vpQelt79c6vRsL$M+(ea^j}&=yW)c|y_b|c^ zoCQ*)kedOF>KKi~#%9rQx%uDaYgWOmKnV}|tr1UWHCuZ-pi-iR1ytAYNK1BIU1Pif zc|wiC`0?sqFJk?OvQ(F+{e3?~D(4Hj-BBxXmkgah7rjw5cz@EciVT$hWPQCA3A9<{ zOqDi`4Cqfzjqm9EXP&Ewo==m}lY~)Yh5?^n0hk`YyxZjs&KK~KDl!Z|m0_0XX=wzmG(F z%rlu(&H4nsjFv9Q=fAz8$rEd`^BuG0nl0wJoi^jbM{Os@RFamk)ajhRy0*7Jpp!xZ zV#V=Hl@#8blai^A@t^=&MVdu0SyAXT%bsVZNUe#AfMLXc`k)&lgT1#IN!LLX_4KWP z$yeg0r=g?bdFVxip1w!xrH+Etq`Y6J#06#P?6EtxyVptvcm$@63>@bxENcyHyXDiR zQ2;i!3t>K8)OQ@%+~?Ax$p-cf%x#{xy+o|~5PgW6=Pstw_@($6YF#f6n}-bBZSQ{q z6%_|8zyC0>0Hr#8I!*UOjtRjzRBml;{_h6|DRzfM@JifRj2g26WTE3PEg^~@QI(am zZn<6wdhqa^em7fW2#w6-PwUQ(1&I=^B8{9kus^Bc8fB?@FUQ)AM?n+Pkz*1<<5S1b zVAPg}L3w=itx#G5u>s$`Ke$iC5?^>!WYYntQ53|9l%gFABs-B*vb;xL$oE?m_xDCqL4d_57wpQxE3; z&5TLoN?Ll-$sNlND>a&qWi?c=aSHk!pYXofC=YP_LJO^Jq+y})-hhw^>!tdcS8fX) z-+DNZHEsvKFn^t|UA$Nh`pQP(IVMU%BFYk&b+w+Ig1Voxb@#~+5B^DQAMa~aRMbo% z|1V7^I33TLNJ%MSv;PFx#b!)66+}h5Od{HURp}M8;ZJ?({h4tfcVBJ&W?;(B+i@qd z$-AlPFHN>=q_)0adjs$Y5A224%T0_Of2nXz=acNKGv0)~@RJC>e=7Cx&BWN!GEo1n z66frke5TdZRq`bsFM|k|x2KMiT2S<`(Qk(<@2L!MN06NgUjN{puCo z*2oJ1@5v=plzUOlAGr}sFH;YEB=g5Lj2LmAMVy@+gY(V8tg~dB={!pIu5eO|SF+mg z2*O20{fO3C2qO$)s1`x!k&g9KlA!>AU64rvu0}As^#h2GvP1@K(P>W=#FF zZ50Jk{U|Sp*!&jY7iei*+giUE-KQ~3?%1~mxh%$iNXf*k;FbD+(RkeSSFTc?blSdWEW>=H|2+`cMlD*G6)Y$O-)&HO-@e2U{ff8 z-`G9ZzG z+1WV(G8uIaz_?2AK)%`Je<-mksZ*2dkLtw4$?2{3QvMHxm!t`dP9`v=+hgAr%#I-NW)bs9|PCA|F3rZGK%6O@St}mcgH5 z5%;^F8gGUmEe82hIL{1Crt`}1!jMQZuPvG*qaNdkZ5zzy|AE{Tw`M5(^mO+ez1H2@ z7^JgsczQc;od_F!_@J#_HboW|{=@J7o-5gpo&;cIWj@?H9Jk~VTNLZ)=m5nkZ)bQ| z*(jME2ni35{rZ)Mk2iU6^Id2*c>C_dX%TRLc(N{iLZ7Wgz@SA{mq|U747O zh)+ZW8_3A`_v_cM9xA`@tYC~vwvt8m^uJibatzL-6a4P|SW8MSZD%cW?1E8(rSeD0cUW0jL-SQs*rM!%M>IqD5`%A0B?6!UM2O_`nL?3q}S^@yK4NvrP|UtyExIZnczRL zm+z#3*s{f-MJSP5VMmX5b(3bJ3?h`f>Pfjy70XOr4?zl}lM^#C z=6R_oonSzZ0u)SiO!b6kGsdsGFJGLTAOqUk5lyL0on4`Hn1H2&194=?j+Ip@zP5P7 zQvZ(z=WEFa4%gtL;t(j3&ikm2uUWy!Qc1svfe@ZL!kM6SzJ4(~Jj3sYwrf@DAFn_H zKEJx6>zR~vDV18xQ?$=Gdxwh^E*}1BY8KMc-qN$%y_UxfRRf7E6yRisZ7N5$rKzbp zYVOlTA3h{75r7_itl{tPV6-%r)8kxOwYCN}c|*Lsq*PSOPW85>spl6&tWuByWletu zahyHD5Tupi**$AF{>-DC;>%%FRCb9&o0m-LauI50!-;;&pK9vIIx9weBt z>sJ4%tElqSU*AxQd4Jb`7M;Vm*vT_J@8Zu(Pv1u$M2GpAMwU4B=PT-2 zTuv2nhWQt}%1v_H5n8fD(hwtNHlF*OAhAd~#(31vw@tC2fpTBj*~Ahi{m%>u*YupCpt4+5GpCU#0L-<>GlvdklY4G!O?3Y1U1?Hln}C zb@flijHiCYK*%5ps*g`11C^^$=?N+_V884PS4Tgd5aC4%`9fsvsmEVL)~=2J)>9vf zNqzQx(e%T|d>d9Io{ni-KC+=@-m7OFVKW2+7fHlM=Ab<^VnF3|;&nYMV<4gIt3oz# z+wW38<|*Zi2rWN*if$IdJ4m94Y(*Ch`Aa!)B})Kc1Dx0W3LMy`IWJT4UV^q837I79 zRcXP1cq6RaNYqD|3owv={Sq|uXEpC0JFv8=Jn^teuEnXWpXBFaq#BBD&odTb}U z0AL7sXJFQe#B+;h99zM#An^`Yd1(02rcW)@oTaYrczL^%moayVF_XDfJOKLVq4uVz z9N6!^T?x^Em(KrJa~RfZqE8Y;`5YS95QdAp;63Qsb8Qo`@tD}-Nl-vF>~fgm#^!jd zV<$PxyfSa=v)|?ZP+2FEJ(nc9yGlMd+9p4p zAYpv0-MH&^Xn~Zk%c|v?ADV46!@uaicwBl4M+OcV%fzd`&6TDP_PJo;^BVxpBg`-sC_>fjn@ixdOAr}*q zMZ|(bEVu zd4%IoId_D3{-|QV2}3882$txG%=&06D#^d3!Uv9K^el5Ytr#1x_MWv|J+~7>d9(O> z5Bl$eI^y;)fMu|cQZ0suhX)M<1A23!AdVP&-mqcMj2~fkx`{y^^+HNCOj8R~q1*aN zF+)sY`6pfZS(gx?fVr-g2K0pu{)e1p{r;1anvmQ|6Wpd$9wu z3JC*dGBK3qIcxw4rB=)7un}O~5$0ywFOImdS3?x6|NdQ)>7V9tx1EVN!k}_Ej9~qp z4_1T&h}S8^%N;JUmhPWrj(C_k*?L{AtY&|%ulL$ttr`)57-*FHUURuSf%gLL)M-Kj zl5xk4TV}wG9HP%uxGvy$o_nNxZoc}Xgp*o(Az{PXT<~G*5tKXh+BtbJZq;%72#;TT z3ge)a7+{R_y{nM1vDs^E@<|)`otxKuNp+jidKXMAl^rqd3Z7QDbw0ymb(*a8x(38C zqi5|gvUaO#*gVGs_ZI&NeRMWE{^&@lRz+;ouE9SO?BRyt)zvlNS7l+p zU`1Z(utm<+P>qZ%@HW0PK}Dp+V+oPd6T`0a?Rb}&y=tdVPVVq-<#KUxkt(3ECj0Jh z$8*uERjehK;x08cH6-B4@rCSbT2>r@>)-fvS4fCw)eP9nUuSl_O7*ZEE|hNHGraKT zd>TA6VuC;gcx(8sWynIm(UnP`5=dmzHPF}R|D!7KfxQ3C8&Xm-s>Nnq7NWAjA)IGV zZ?48i(y-|0>F=)7#c!_rF5%)6)1$r*ewT=Yp{beTiSOSZzaqmhBR@*Hsa`h2XVYqn zFyOY^=5oPDMd~^yE+QBCLuVGzhEr+uJzA*ar zJ_o`0$;k;S@WXmgaLl@9x0;kMN(#5r#-UkIq!=}Dd;3q?K1(Ex9pBAPeg^d=GKi$MCf4n<*r`JCD9ZD+!ml4yUcF%lu$|Iw&aUYqcvMKc`q$UhS|6 zCvm3e?ZFbF1s(;*?!ZL-3~lmYgH&EEJ=`Duv@R}g7Gbstl4aycQ9RuoI!zd;^E>WZ zXGlJ352w0E)H#ScKKR*am08={GqJq!Ml^h)Xc!SWr&Ai=RXa}kf4Kl7{-@Un1QOu6 zkxMo2aVT2A#ALp)@m#d}C1B`zi~)o30!P!dvA(zZOJ-B@%8G-%y#kI$qk9f5Xh8nS zsrdDFM-QdfIvg}36_0*s>cAb^ft9uQ*CJ=;4;IiU)WjHh^4%LBHSRmf(&Azk$p=hS zYCm@@gsTjYGL3KZ}w8YNH zAxjenqE<}}gU3-zch|M-)9wETEl%6o0_$R-=BJ%~B5|ref7Aega}p~^E;U~JxgEJy}>guU8?L+0eEz(Nx4!2J(0s>ih%s6#*eL^xu+?fGiA|kaI!=}M22BoTJ zD){%hBp?9VDxYTXNYlcKkShThq{rJsoAk_5XmS#5%-dBv`MaJt>5l}Y*}Pzv8+CdL z{;qod*DR58-)+spF{#aP}@LQH{^%PA_E$*5jR+#wc zG706axA$!S{Dy5L%-Vb}qEsfBXg_uY zBPRJX^{Y90=H^o23%Rh_ST^z2<#4d0-FBw3S^at-k*%n(@Fi*Ev}o#s-^0DfqJL9C zLqqwWUcvMQmxR7;Sx%B@V0q=>*&{%2LHNrga(j4q0G@07t?tTnUf)k?Ko@7zlanmx zXNlWbBcIxdwY4+Tw%022JJ|($c%k{L#R>R&e=XGM<{!NeE}y2I5-UIl#(4{MOK~LP z4wO=HZs%?U8+oQYq^v|PDk@84t@kA^H_L6*z};$6>rEGXz((3LU}J+2z|F+XOo=!6 zpjSAxx3{}H{y|;j_kk@2JK%X7dT6K9dSLW$tzu=BJY#!*Ugm$Ww>D5JmZz2U(nmQqIW>G=G{Srf_fVM~bRR&fl0d%K}1`qgJ00SF=3-P1!#7(Vh zNCCvjfl-l-4MhZY^yLI~$f#SGQnaP}Ou(YsDGjd_7ofVt>bnVfxU-HJQzRWS^RHP? z_PC!}0g>uVvn=%C?st}CdOABRg$AQmnO=?Au?b&S&)jWpQp>t7N4D*kZR46z0S9`w^Fm=5!BU*uUSjX9N|`H=hrffAH;>gTehUC9({d%Vp9BnGq4TDz(?2c zvmzHgOFz3%!nJXDqbKzTh#HUrZ|qkT4*MEH0sm%Quf}2Xym!q&n{`bKqC54`u($f1 zp2x%}*XU6Y5U}D$Tu=#k3I==nU5!0F8cwmB<@>uk19f$&|s;WmF+J<_yu5w+WSp6?aCDPqom8l4!m+j#XS9?tl_fz7h%IPPkr=Uh` z`O1NPkb;2iT&*Yr^WK8^!;b5iRYa#IebrPQQd?+YVf?>;;gs{_r1-OoV?<or|@F#yE($BC5Bk5aoA#3M3R4pgR;Z%<5NXyx} z{7aK6Vu+5p*0sFsqVr96D)EBzPtS3x&JD?5^xf@eeFWbR!NumA`@o@6nLY?%sZYb6 zx1K%S-DhWKh?R74=y}tU%YagTbN1>66+$&jwb;mGs`9nMa2k{U!Q4^?)2<0PkxTuB zrY6LWCmHyaS37CR0}jV?KSB;{Mch+?L9xi`M1lhp(b#V6c|TNlhQ20B#$VZBYVBra zB?yp0)*WsdIcZ_07aJK;*!%^5A*rOLz&t!xU ziq|U{tNJ0|>b*D!79>e7sCt~2o{<8UVdXPl(zH8Lci}lSLn=eYB;B?Zv$j09l?12) zmweYp_Yd8}$sJ;RsY*CFN)D~3$_fg5k`V((y94Z9-NRltGY-gYpjB+WJ}*mTG5W}t zDyXXs_gT8XNDQZ(ZLqnV3-IlUy9yBZc3ZnS5AeDyo2w1j(YP?F^*?a}=le02A-<$K z`Osie2i9NN;Jg)p+EKspHdHu$Eu;)7$h4&#PR?I4+&0wM`}>RokG_s2Eq|$5)(=#L zc(p0`wub{WE+M_je_S*tCud#+jAbI;Qt$8C*ReYuIIL>9DV47NeOk5`l08lT?*9^G z30BViudl9V^&RYu)pOo+R{I=M0=Q44Qu&G>7E&EV>Nm!iQ|3AaDNoMRc_QmvCVBN* z4-Qns=7cyv8Y2=C(%wO1ls@vpQwRG?Ic)CIfzg)xL26+`?bvwjSvthAOl4Fx$70^< z<`AUZxc&3sSCGy5wGq_hXYME{Wh*OL&&4Y*?Z>R$mO$?y@&5DN?5vARQo!lVr@Qr& z;o(w`hl|Vje;jU+B^Zlqu3f3`;gytGTknlJW@x_i17Tpl)!5UX1@S|zA~ zcL_ap_H!3gt@j;qXuz1-SXuMg>U@KXw6tcX!`wyJMHxWN!n1xlB6dIGbGTt{ovR`uAs=*4ylIPu6IKQ|o?oru8 z^A;mE0ER<`i1w(83eGn_yl<0?{2JYgT-X|EG%2!$*}GFF!Ees=UL^zFV8Fwac!L)h zVIEMAUmkggGx{f><%yy=LgqrwY9Oi9nV~%(2PbL!wA z#c!_G1_#f+e}Az>*m`$Tl)!KbEQ1c|yj>zZy-Jd1EOXIIpM~h*dZ9H#Q2lNdi@mrv z{pYFu^*{4WKg4y@A%c~b_Wrns%1v)1=HoWFf8JBGyzTGaeaw2rtX+K9X@#7ae8ERc zyLP>mvA27GSg#tXswDD^jSXQVih4NfsZ+8zWe+&L2?`ExIfZYfGb~G3(i_32ldC*; zf}e{!=MEh|CbZa_WutQ4OfHO&Nr2CB<2<d>CAnA~%%LBB)GWUIsRvR(m|!KDvpL@AD0lp zv7(h5JNcQb>KY0*3LjTGzkJuH^mm+vLOd=Zrq<_J?1(n75pJ#*Vl?9uTmq@;@90=# z5Wi2BE*csT{*jjExdgqW6!ll~%l(hD)3c1!HZgqO#<7n>7~`znJ~g$K3?gPAs%w69 z^z0S{mDV+3;h`RG9$=*zAW_N-3Tppy!~rlevUu-w9Xi~vuib=&fhZ4zq6~I67Aqe= z5=DhZHv+$)P}8+Ax0r8Zcx?q5rYkEe`=-yLy6!SB7eOcyfDC|i;>c6Ak?J}m1Y!ZQ zus>wOqqOIGaop3<@mDThq}+kONmi}7qC(CE?J-BOz4sLZ^{W5g*KY$ksMmz|`Pn&R z_>15dgq^HjUAw!x8U6=e%BzYsmPYPynU-f;g_(>J8%G5(cRt+X5-eV#GDr7Hq8BQH+ ztVVaLMJQFMah;*pDg;kLr4u6Ph zCORB$FQ1~(N!8TUptRksujg%AY~lRh-ET2p?j6D>cd7y78olDlC>%U5c&$I8>-ORP zlC120bG{+od>|1M^^U#e?kV7ZJ4uGzvRp(V{3QNM%-8vffknM7W3Cjy_`1>@oUjU6 zSco~32#*!H|IdHT9zcDvX>z~}MF3*JNTt|)Mwyt{CBb5G?jGY==^zAgkolmYs`LFD zE<66y>67qSa=(k6eE)~Z)VbE#m4nTTURwP{pNHU)T7NOC98h;~ZyIUPe6z;>Y|$@q zfDu6J70iK#hQ{!w{ZceHgT3`8U?0$H^@qEKnrw#>x7_ZJt5Asg-+{`vUhH`?QIN5i zVT8ELpV7ri8;_P=dv53rGv;W%`imtvGxO+arUZ`as**1*?ygTk;$y&jiO*C%>nq%! zB7e|c6mmfgAO10mlT&E=UCu4(wc0P_XWHGbR?2*odd!JY#P%B07q`~=c*-IHF~jHf z2bAw#t+??LmwK;zyWXa}rk$b%y$w-}LBPO5ecWDORy)!|n{h-ZA>b*}qGNkl`Y0Z+ zLQ2^w%j)z$RY`@&d_c|`GEuv9v#CZWPN4}ps%S)7w#0l~hxbz$BW?9D=dnyTyr>bdl5!6s@BIRYJ=aFaoVdF$uIh`!AXXL9vqn*2Hk%s^;0wqc zAjWlF-4?U&F2=(r5)|h>J&E0xA5dZDws``iF%NH9%p*$AMlIePZor$|@9#7H&QySD z&!xovlT!*fYZ_NoWK^fy8dU#cf4*?aN>Dd>1_apiFAq3rcwV9ae5ui?zB8U&UTu=w z77v+DcSzeVY|*23zqcnUVkS4y1Qk5&A39AMiiM?Q^UeDc`e*xeq zXWF8uzg%@rYBDqz#$CZrE$V!0UyS-F5eAPwf8A zAsi`iU#h`0hqi{7FsMCiCpl_+XjsSR=vht0~#w^7H0ndyHzMr!Wvc>+GW&s?7VwlM1%jMzH%5`|E7<)0#> zNaTs8f%hjOrb6c&cY6)0!e4cn&!0Os2NizF7M|VhB<6Ydpl@|CFyxy7Zy`gWSQ}e2 zo!`p8RDTV%KRBT6==e0Uj>Cjbv5}urd!$%aEHS)l5Ztb6=o|)79O@;)4C#{cia|D9 zY?+Bl1dHO^=d)PmTY9Nvz%r<6L}?;_{i~wC#-wxe2Z>%pU)ovOSYczuIwyI@6i8#K zP*A8xBLhB^A;@v_;-l6zX`(KLO;r@r{L^=5)Af>KM$>cSY6d?)acs}P!R9nc>8{4c z#vf^{2JuZqK!)$ZEwARr>rYtu3$H)X$*~#MzTpw^R^;W{84_7=5p;2iW6^8;a6l!H z_d!#8Hkjl;#f_|V0a-k@!VcOp`h#f#q}!)mq3LVzSf?jUP7!?~Ep67ETU%h6EA7mr zp}}XJMQ?@qlA9E#Msou+nrSW{H`V#aQ$=XFEd0F_C0F^f$=|T`t3P8#MP8v*5qX%A zd3gHt*+v@Y)}}_mgs{&T7gciRyRsri;enPX?9K-Kf4g$nCVRInz4dm3B*KG`vSdjJ zC`$NA>O)J1J`bXO8cC-jrPPC3=Niyck)&M}*K2|Sy0nZGe4#Zn)hQLMI)oA+WeI<+ ztJ@z%P=K9-I#4lA!J7T%`w#u#stOuyBDIVbSe!POA2-pA3+Qt=Xxauly7k*->Nc_ z74)tnqR(Nd?>ZTB6Vt^-d3ctoaGPy68rj8fe-%}U`W!zcCXO&HLkTQD>b;n>7)}I- z-B!W0elWbwKfiQ;J*dJe9x!=bo4hYIAQwMhx91Ys1M(4Bto*w}Ph3PO9j^P`lyG>5 zl-xjhE7%ciJTB`MjB)`TVbYw~`YFhtDwQwK)r_9P6-adtPfv9PnQa|3+0XYEl*e7) ze<++6KcWd=V?Gskc1#-k1)kY=BPALX3XpJ@rFhV4u>kcGh1hyUqzcFO%+=ju735!L zzh;(^{tk-4hl0z$xkbY-(o~<}prePE7*y>P*-;H?^mxVn9dD_82Df7c4LW&{wx!NQ za(2QnnlsUsduxy3$0v+blWg=<-?nATFK{%C)|d>uIxv&Nbe)u4bOL zif)(PL;B(fe!H9%F84lh#^1l|m*3BLZArBBx(J%F59hwLSSn?gjaQVF-&n7lpP#?I z^?uVoe1m~rqEqwH$0)YX?Cd#qq)J+V@ko+4l#i90!QI}(;qtf5R&)J&>G0%Ta|eEe z&B5+DZAWBK1p3P+E0u+62j{G}Kda0(j%j@~Xlb+X)bP|?zIW`egO5Oh+YP00Z@yKV zh^!$?A+vs}Oz}e=O2h^!fLWs7_!x(+IU8O!Ha)cXuxn`b`9+neW_)8^0)H*`pJ zm{bUeuoOd~w1S)!Em^PZ7e)967KTJ^-sS1A)Xv!Enw#IFpa8gdlwtw?TuS35C3Z2< zsAy>P4D?9g{CH7yUw6e4cQdd`FY$p+N)04btLGdQT!|pS^wYtCy-1|>;dXIhso~Fr z@uj7u*-3QZbaR+6LUMzUFyhIRSFhUJa7m+YT%9Ul)=--v(ymce9UV(MJIJR`Q-9NJ zBZ^Q-{*)G2S%LLBkWKz`n9uRxhL9qeB}y+VBlawhf${Rps91{hqv1N9$Tyc(7_Sej+8t|1>6O{PuVR(L6HF7QsdS)HpawYgX4$H z9)=G}-E-PU7iZ0?bVNj;EJQ8=4j!^PU}pm=h*xID_>@35eqN)QVYpcEQV<&ll3o(y z8qvwh!&13D!TG~lL$Fvt-IGhqN$ef&t1LWkXl?QOS8qkw+yabqQG5^v8xp6!nokzV zY5%1b`y(kWnVY}8g7?$-LAeVOafBo{Umm5p74Pn9cJpv;ViftXaX{I-GE4Zh$q`C6rN4tC3Ta4z?R<{ zDcYv_J>>wl zZh9xobaE)dLO;Y!cgbw_r-Tu)Ka^f68sBK7Qr7C}n1G`|xGKf9(G>2h|8Lq!WGOcUV} z;wmQBQY`syvg0bm_W^Y5V>@n(i!H$N8Z8y&2g@4klsu40>bG%`<1V0`W~A-X{%Z>7 zInKqU`@s4*Pm+qL8#lC1%7jyCJTk~X_}!AKzNWML2bbvy?oC`GJUgCkW@hr`e36^OaEj5%NlsZ&%y+#P5sWdwt^rnVrYq32 zq|D8gwI*j_MMXt6d7WnCvI3I-|B_>XmkE-%( z*MWi1fw-r!Vo643+Tdo&&?bejJB*(Pcq^wnQ@Y@D0bk0D%TQMP9>^}Mq96(;0T%CX z=Vs@?3@|u}`E}IDi_C?9qveXVo~kOMFOu2E(4R0iwKX@v=^f#hVXwZ(5isP;#XvtN zG$_NTtMDb}ma@U|v5f18L_HRw#dT&jaP+P@MOm+t0aTG5ETQ-!Em^+;S<;7k0aV=n zlO$nIE~iJwKiKu1h*ST~{?l%VQD5Olc!_{~*J@sw?oPf5&4D}QLFpXplS>Fj5$uCLBrZN9Jz(|QX<1fy zev)CCi-shqwpYkmlB>H1uQX!|$h&N{eD(^%^Yh;_`d%Gx%}61WeEw~51Ius^r)U(K zjsL(kY47=VY-w(MyifWKH$}`^r`{m_))rUtSLCd3euIYex~7G34Neij@~d;o?aXbx z-yfnuE+|D5IWd#S`48Tg3~b%l9jkoB=feyL2+-#5i<~!bSEI&_M|~ID z3Ptm6PusSv(PJfwYHDl(d1vbJ2`ZUPEwaI;U@0U&>64=C3jj0p{d14l#zybw&-sZf zbQoX!1JqsY9)}(QY}v(wBr$FcP~{1pL{Ns$&Mc}eV_%> zPc3Re)=bq5OPp+TvxrM&<+NLgRx0@8B+nnKyPblnjr{u?gwsce1!iNvg1j#5XJ>ilZ0lGd?<+vh&pW!b zM}^Ds!@Dw?2{h?~QQG8!b=)(wM-uWiVh0S)TyI=Nh5V2x5No z{C!LyELrT(+5JY(y>DhiWXgPZ4-P(VV%qf~hsTa=druVN_*XEA`U^;6*U7(WN^QNjo05`W)`}5WJ@KMO4+1ej0a6cJsQ?zDf$R8x=X-~f z*3}&y{hNtI5m8ZNmLk4qj2;`hhxjOeETBUC@UAaksX{)r;tvm@gi#0HqrClwdYdf0 zJOaO!00a!*ZxMU{`$thymBT9JPAR42`zo0@i6@)fMA%blz@o~U*P6=6#|8Txu0L3} zHLmz6^Ic;9`CJ(9Z<6oc-q#qY6>j2wq6-M5cVb+v?fyQ?xs>V^@2~;Y@6gpDTRxES ziyA5&5;ihD+*|omN#^;%rok%>+8j1Ee1xZZLHNXg3E$4SUX9+y!O9M!^f5eKgii>} z4$_mCRS-u(L_~fA)aPgTO|s-0A2%u7spbA!RMph+rZck%7#ZN=Q$Epl7ElKtj@Dxe zs;Y%8HXtqYo$_Ah{X&0s<4=e5YEE@$WNO$|O*Xs8YKvI6k0K-=1}^mhp`>H=FQ zEG)a{dWeBR`UK0n7QrokLg487!ZeF9qbc0gw)rBC{U_t*&PL8l^%+~OXp6y)afDxg zO_u2Y%LNFelk@N>uy*~h^kAjbq(JN;IKrDpNf0UdKJF)EUBksNO=FiS;Whx#7a(@K zZNw0Bz`e~7?MXw_^ref9Rq5Oo`a*DnqASIF*TOpi1>}rDp%CK!VG7}DVLBcY>vd69 z*q;x(cVfp!?xjF#D2s){*v|Mc6ZmuzTfPA^F2E5l3Mv9cN5=r4do`%NN#l+LP*M3n z93I}bfvb+8{Z3-omR^9_4P zV>!7S=GYyUZbDTve@j796v#RGJJVb2hStiRP|?nh^p*g%-T&H5g2R=J7PD5__+QVs z9ht|mr%3I1?px|=ox!Bj^qPV!_u=`gn!Up&RsR@@9%eSqjp;1WdFKR}#*a@dI<{QN zJSi0+Aup2W1C|ch>q7y{S-wK@7zJ5->py6LEsLdBzcs{S`9*jDpm0o$pWmZ)%}de@ zJGSyrq=O@!g7t|UlU9_Q>pFfgU2>l5o2eb1bczx%Dl29b?6?pvEhsc58zwSiS5O+U zw=Z^abHl=kn%HxF9324q!$cN+mjhcQfTdSgNQ6hn{lN7idwwxvxc`{!$x62HAdzfJ z=FS>3yP!Te03P|XL{GOqA^U@)KnRVRuZjF8nWBktVNSvH683_U!g`x=l#Y&I`A`~0 zI>iHGRnN$K^g&$IK4Q#t2ZQv6qcf7vS+bEmYq}*`DYb4MwXD5{U?0F=s|t`~emVbO zsk@V08pFrV?vN#0tQHGG*I(kCOq!y6^2b+kYA$Ko5gFVnIh5SVN^+;%)ucFc!gsWS)DrNw14}SCd9?HK7X;*mf;oj*0M^MDUu(~ z2KlErvLQf-4lS+AcR8-q%-vEZK*b|V>Rjj?6oL98wZUM~wPRYTFw0TFOt~C22QJbL zf20Kn{6j&V;$}XqP%eJ`OgZ&pLg|+N^7MJXZkkWya_5&1k15JL(hXlsflSV=V-65k zZ{?f69#R3b_Gd?D@;@`(cdxP{U>3OrR1RMGCo{w6Ja)qBzKM3h%*@RCP3|>A7B*HD zm#DBE*Zh$*-3l6F5{_i0d=+A0hu`3{)5iX$`S9CS&J@$;6oNpr#IWY1cg&PDx?7E& zhv&>cS+rQ(|4!0MSJz!@5-j~9BZ0A+cQ7TOVQLCi3`*3=*SOu?^xu&MOmX@*cFT=93qM`vs6>+cK?=!akj;9t;Fq0a++uIujd4)&DeBZFV zl2VpcUUPO@Rh;hL8d(UGZ zkPF;a(9bpT+2@r^UesmTfAX;b1=Y%dwVAvtlmTpR&<{B5;7cuO zs3@&m6#;ZgCKF^X%~=g8lLT6uigK!;FpuVd%bm{zx>!jL^;@rsx@P|w7^#^&Q1dCb zWbM#^lK!WZMZu{2s1)bR8uDU%~8b{`AvR zfv!Odbu~oA#9$hjkrYbbR9ty-!8dqPxeE6GmCc=fbJl5pR=KQ`3$0<|6`j*Q5HBR} z7k3(sTO41&BY9EjIb$9CE?~Fwqq{Jek};|9ye@R}ZTVG}&;8K(QO4^NX|#nR?@2Gg zNT#^Ds=hks|B1Tq8)f1&ood&O8z0L5hpn%Gsw(QXzJP$VgtUOPbazQg2oll_f^>IE zHzLyAA|)kV0@B@`m+o%n)?RzAIpbOx|gVH_C#W0rkoUU1K(QZ4DVKFzwx0ECk^HdZ#Kq*LRQ6&Up%?k4r5 z!c2mVe}Xn4_VT>0uB(xPKb2nb zaJ-pDLqk%e#koZ%hCoO<8*UFk3Qm$k7k8>zgYu_y;`)L<=)TJ?2`xO;h%059OB{At z)F`s=yOz=LD>;R4D9WRslUcQjxrp9Bd4qI(e=GSPY4m#;(J73#FMz&nFUFtWX64ti zSrSU$phK5@BVXl;ZE2Lsj@MMsZH?DWM&DkJAa<9lTIyL+KD5q=LGw32sl#6>Mr8l! zSa;M*S{e~+sDqktjTJVWK7aP7+1ew3qYmxQ2z>{p?R#sBdNHkZfxOYZVus&_-(Vy; zfAZ;Ry6XtVn20J0RGQs|6I`b1p^@tG-ufa!*y>tpqUC6DG#EC3*r#v-Z*NHAg-_hI zmjbJA`P5VZVD7(s(aoZ~G}mD@4{>YAqLta$v|&rJwn|iTFAYg?{Akj%q76?HO^S;- zF*0anDxNUZ*C)j*(4m^Gj6;PLUOgL%OCP{wX#Q$?KO7O_Db!Na(5^p>B9VO*f*(Fp ztRi1;+Ge+M-#T~c)65o@OB+TejfJnDcx)CioEBj4+z$df{UhAqvx(=4ykl%9O&&FM zLE0OEo`G7yh34Sc2KTh_vDaSPbq+i^Jq*VsK5qmhr(C5n zP-y82KvkZR!qxYMWFE8KIr%y1Jbk1(=L+^U1Q3{o?A;uU7)|Ao3KODFf7uro<{)MA z&r$Jj{F8QTTVjQ3kuwOGOCKcvy$P+mwtmTMv*bZ~CwU&I1nKVA#lj(;oLlH&bVRf6 z;KBq99Esyxl2UVLrJd~?*X!GBWt<^nHrN*r4-Wz)pE!~szix+|#J;VI-#G~0xYoVf zQ?{yTm%i$`ze^#qC{d+<`_(dfN4k^7Xpj9yrx>+f`GxJAstX9yL_5}A?*UKEF$z#1j_=6ISp(|WxpuJon{T#kEph%`sa80`o3k+7-C}v23dd{+uSTf zMfK^x`alDPO~Meg%uW>%t@tXG?hBr>cjQ87fA!y~8Cw{uZNv)4Ut9@KakDGjDfGZSoA}z5Q)k zrhaz(GdvBjvo&GVCcz@KW+T!f+@?m3N^EpXzD0Bmo^PN+|$sxTicuGhl^(ulh85onGF+> zxbG1_)N8iiLVp<^4i2SB2_(ogRpK$8##|Ge)rVD7g4`O?Dx_wxI1+HZZ5UHq-dqnr zL!$)1mH*HgC}TZZ8`g`$gq4uKxL7(U5BVcOz72kDZ9Vh-B%6Sf!C&x!wuAh5ma_5S!#{akSH#r{{sXH=90yuFnbc$=fBl{Qm& z64za#qBPXoYtNCw0o+;9xQ;5JSXac&!98hnyWW)Tw?;4%{+Q8p8x_49Dc4+CSKz6a$X{daC$Zs@Z8+ zIP4e3lD?ttqQ4L@t<;EWdbB9&wBP=-T{h%(;Ko50SHGDA2Rb$s2;crFd@!;0me&73 zOVFp!+H!X{olp?fMNNqqEa|!zOx7J0n-cxV}kDkX1@@DN*nRyW>{>?5pJ%G2IE+&_yr zA77m1;b;Feyi1<5#`%-#`a{`*wU%}w(avG0f(*QG$M=XYi=~C3tg$`A3p38!f6Qv; z_8oa_Thrtb%UikLu<;F=v${A{5un|eb&4x=?T<%TvN@{=g#bTiV->k5L<9tne-~@S zmJ|T{`h}4PiSdUTL}L93%mhPdylc3q+%fliC*Ry1IBZQh6VQ~ zRfhPV-@|{6M2VIswi!1&pXrbj!x86vNTAUI1%#&rUDJw-zI^+m;EA|oXveqVqwcAt z8<=wGHWouVqnHl@+$CgVDRRw|zp+E9G)t1_$9GhE%#wgACUm+)m&%u{TfTkGF0_Df zfI7A%BSXGu&AtwmZ-GAz5$G6e=HOsqLrNt>)#v9&SJR%m&4Xv~mTSYK7@+I?PA6iA z!-(u`0@EKn2KdhwT_58mhT6!wv7dyMmRxWm3VM2q1_s)e^Lu;63*`b((8ObKuovg( zx~a*FE{=|dUyJO8qs!%d<3};{QX6yYa?J#vhu8dwKV4rxuho|Ke@lZC9w7<%tF$;@ zL~c2(itChc&Bl>6%gGIe|CAy2KfH?8H2qt z`8oHN6KJelH1<4zoK2hZn|XDLiCmE)JbW4Y4i-{gI-34-KY_O(Y=+{wH?5pN={t5y z53CL43%|61UvD9hRs^t}MhKb8e z_8X9UVU6Fye);h$SF}Wz{Dy$r7_{A>{@Y7;fzpYUrC!ls?qg{fW(pfV!@(|C*O0Br za2)&Y9$T4Bz=)fJm8;*l+}`=+sjH{Phwk)y=4a0kU|RL{)C|ofU5z;s`V{2*<}c^N zfs7RR?G=m>aLguq2Lt+3_&`s0FW^sPt;CpV)PZrNBQ-t`%wApa_^re&{^Hr&-|#p7 z;b4gE{!Nx#4HY+Td}*xwf*KC@-(13~3C+v!;#Ejow?7uN^OBQpeYVid$gJ`_L}70F zIq{>OZ|={ZoJK5HyE<((=5%_7!gue7*a#wD)uem&(++o{#bx zvVH1YcRjCUrlamE)8jpK`)YwV^fFG>cw;?9E}x$f?bKn+c+Z`fNmz&#>g)i-DN7P_=@Cr++>h)suuXrEF1~?~Tbu2< zs+42ttC?-VLF_sOp!>DfDtvTJxV}F+O~7O(aZsZOGGJ~xN%PBQI}D+TeMjlNAL){9 zwAw@g?gscF(N$afB32eNMpmaS4;?jhXn$WVQ^)lZ;yVX*Ub8+NX3QDcwrRWq3)I}xKZzkV~ZZ$Gi9?TxTr z%7VIPTf3O0C6+THuqifb#+i|RW zbvzUa!})#h?YNv8tsPKAPLi8U8j5bZ`J2&COTBBY0|Nn@e7|h_(D5MPvT_N%Vdu|ihnkJ|#{VS* z%fi(JHw>8|1-Br@<@9aKhMgO8lYQ00->uyv7L5szFM!c!`0dbOvgwv<)Yx4|9IFgkm!4j2 zeKloBYb###@fB^GP2R=DGYNM&1dG<#*4d7!t|PAwCpU9XFrxjbY_?Gio1eK8P-jk- zb_UK|&kzts>fZ0Y)0p$9nhg44g%We|OH!!z*bd@1Kp;PrK1FWsdDM=$7*TQt#V`Utbw>ws#> z72LNnsr{%~G=g(jYFvJ3#mV%axyY6EPHylwc;w=SJ+oK0XD_rGd9-y+%%_jc9lcUkssxbxz^>jJj}R*pUidwO2}OChYQbw=SDArs zt|Zi&Qm=jIwVabLYAvVt->drL^EDy1)32-;&mqh9bs26Czr+be&SIZl;^!4&FRGcE zbfaO8;6?f3sBvT+lM$^h-D_W#J(;B2)@jp04VCZF?0pkvj5T9%AP`U$oZDw3JnPt4 z?|2Rg=}(o`i?+ zRm;SG->GX}w6A!-s7QuI)C2ZC4kgd^JxywWVP58MhYO0@OZ@-zjx| z{wS&KpmeUoA+WhP4XHC2ZE3`UJdSbN{Ki9w<=-oek@U8Vrhu(_#>%}(o(yNi%rQni z+sn6pisIE86nJuHfB3O3@}oAu7{Nbl;s0&=UsnjzDBanlDY1@%`A&--~3 zxIW%roa)y!L@XyKbH3d80=={~@`zw`W`)MmRrZhNW2MdEbmK- z9v;42OF$h7_jmFfRKwvRHy(U`Vie%2-J5}IGy&JVR1Yis9-LK!VX#_uGvY6F#o|A~ z$2Ovx;#U%Cz6dQWq+qOk#rBj@)Pj?I4pK4^0w#9kux4a5;E;foxUapDC89?k`I%@d zOYoSTkN>VquV&=Gqmc1q>ya6=XviVN;*)tVRWm7I92ZLdn`vKZzK?C^j%>kx@%{YH zu{IS$cNcbiHSj6;Kd(raFS6rHn98-|OPk~5gfTxB{kN3O_{tY8K1fuIDYFxEHQDQy z8>?52ll_}*-Q1}I<_n9VgKzHJ+s8EJ2yyp)9%H}!n+$$rE>d>0j6qbL-vX(yDw-%T z|1*MpF6m)PGL21q6ts%S$s8@DNl4u{>2c$nw>6N3W7- zcWh`pe#W{zZ!vY6%d zg(jUv0IPtH3#bB>fig1uv1Y-TPNSnVB^G?IG4vl}haxA(B9=k{waWO#(dP0lCK4x3p|)Y= zf6u#OP3}GOB|B_u>$1MHiK?~y@43j`^P9=vT;u}$^z@93FX8%E9JOnYIFm?bdY)2i z@K~LR_Te37PhBKy{eWy{l@!bbYo`1KVXHIp>MaY}|F>Y$@0 z12r&csly7@1KT6X65==v*TW;0`-`c^^zYw)H5f*_bfAmmDeYQV|E{WffIePNE!sUD z3EjV*FeZ(biV?A$)Y4hl<*m;hOtMGl%yob_@epGGJuePTMVu;of_QbJsws& zX|5C}ggIY`%}*+(?m%R-)ay z1+V1EcQK!>n6yCV7ok+%IVawSA=Q%5d&5vtJ+hN1*MJ`?=9-%3w)(?|k5>=N%U@lp z6yyuo=|PJC3jb+TG^7Qsk5EUxsr;umEuHIg1YSkuvu?i)Qyik-PHunfJiR3`$xz%| zzg#SNL!k*6q~GAXH}G(UguGYhuTjKnGgNNYTCSPh>gvqncJzGC57Qey(0=ER_TH>B zO-DjS%~dKy$SE%^^;Ty>1f1%7Q(@Pd?Cgl4H8hOqyWKMHiLtTWI_ARPznO%6rd*c1 zPnjQ1${@=FIZJjtd)4P?6COXiIlva>@cudjy0Na0?|dtN+ww41W@2Iz@?(7bd&@)7 zBLy9sp|SDVX3N##I#a-aJo3}r*`p<90vVbEj{tgb*T&Hdu~3R^<3>5J_Tl>Tldnh; zO4XX%uj(qLO$#_^X4aR61A6dK} z>7s#!h4sE}c`()2C#%b$iJfr-a)8G9c`*3a*4F#mpF`;l`op_Dpg-m#_v1s+A3wk3 z+4__kbL5j#@6%m9BtYy~K6QOO)q9|NE!13DY2j>rVB>SQHnDai@CLjm>*YB~@6#=; zRy%wSMrVD=Fwnv~snZP_>BBq4<=9umWMnEfuiVQ&S6h7RG;Y5BOQek4+2C_QQ84DV zcNMCf;dMi^bbk;@#Ep%O#m3H-oUHk911y9a7POfzD`UHQUY(xU1J|CRN3*)L!m27B z5nKo)X?wfDdH-f-H~evF=n)fQJuyyvyL`{B>vNy+*!4(udG5H{(bLZS>Sj^PW{9!B zZlN?CM!3oIoDee3&Gm3+Dtte=(%$mePO9hn;6@UCOWJb3 zAk`&=;Jd!QE_`>u-JvTaA11&C0@ewBR;lE=8Fkz{=Y)9IU|5KVwkL7QfVGuX&Iejl z1UQKMt$SZz@0EbAwl*l11^OqX~C^7Mj}pWu9p`{f{? zt<9*oQj$UUQX2lT6IWmvr+;fq2?>yjkR1!a6fFpy^m2w|P%kYAep z=_|QQ4+obqD>1^4GYX*t7g13@j{`POj6+~Le0N5Z9{Psk=VpF_5ow0&_VQcgK=-}1 z=%(zD9B^Voz$Ak8^#zBH+q1Gg5R&g@VOu45iu_lqX?QcULRoi#2GobSkcDqEh<}OvD!^6c3wQbVHfIQyx`Rs1*QedxH>)l^~ z&h!9d*yc!>kZkL!bi{@JRMa>bocd$D%W?pJ!kCqI?)fTQj1;USHM z4rF*@2G~#}Meu0q>G^;qcYPcuNneac-qU{X3hD5JZxn!1r{fZ2Epc(z;abx!(1BwJ z0le?exYwtlNXH#0M{Y{Sgy20jroWk!+FP+RXue@QKm^H+Gp@gGks`g(>Nlw%wYkX2 z@KD*wA8{CL2>}x({LXh*{7a7yw=IwNX&27I!e$a~*VRNIK+p(-ReERdfQh01aXI>t zmya(#HPy0m`bvPLJV(ButoX;(=Hr8cWDZihT~pie!kEX}#=zjg!SA^maJ%U-S1{MG zjY301XEZb%XN5BaeZ2Si%d$>gzm1Fmn{te%t5F|tQc!HigIs(({PY-&On5tNy!Z_i9hBasgt%iG@_#vEgR0<xkw$KsUI=gXs+>;t$wDS1*^Frlgb=9KxWw7w5nG+EgUT93d`l&f@$K$a()0Ijgo*6 z;`N}TB6*yJ8^^i|9Q0s!Jst&L4y6jBh)0`#SPu5NYv0=1!g$yl0;%6~zj-_af8HW| zy~U#&wEBu_tlA629Ry*unA^8cb-dcmpPgO>yVUde#SzDR-YrI;X7s58WqcxJ?*jk@ z{^dC`ay^tAyA_&@d?n_eGdJJ5e0e+V3EQU%+ABHmxl8c zcD_-Wbg|_pzbv&X9Xyy1!vk26_wIR}nS_H88YCbfAVF27l6SsOFKq=SK%)+RQ$CF) zrG{f{Bd9SddQSwQ>4J5#oGVx5=R-!zZBGq4srw=GLu%g&U)tcz%nXJQ-$h$+SL46k zu0n;a(dNL#^=&GLp#w(Is?+kYg;I;JES~((0A0ve081E_Yi!EC;Sov2NfI@{&2{Epz<4HSDpF*>5z|u^&GuVUmzQ|14nVv;)*b zzTh;DVhL*N206OmKh$Y(lxO99*dg*^%SkM5aDvX2`9w#MW7x0eEX@|Ha-}G%Yhq#% zuGrU0Lg7jVoOq&80jDz8q^w`ug33 zjWA9&94WL=46cbOfu5QvRI&gK?F?3=bJ5h_PZU6z-(^vxd$ujnP2J5!3nt#(-|usI zh~1y(1;Q7M#U#fV4e#btOFC&r`iNw9nwuB5@Cs48)Mz%H)+No;A*=_Vg55 zx&17+ygbPM;}>vswce4N+aVq)X~a=NK>3OoM@3VKyyddGQ`FDjOofhB_O$Z$62Poz zYC&GkkKX=m%jfwi$9AbwL84j3lRk}ociu1F3qwRuWO}kAT>2AXSPWdvGtKlrK);FS z!SmrUs&Es}V0$}0nEF(}(u7%>viVub?J;2n$7zF_yzuK)|U_?-`m9GWKRkuoX7 zr_D!WVWA9K&0|o?>?_aLm`qr*eqw)iElPWXZ(#K8tR?HJAP}b-COAm>AG^<=#q~dY zb}>zfUhDcxQ$90aHt20VybBg(HF*sO%9A*0Gt0&;7 zyFE*VE=5bTNsP30f6gVuJ&u7a)6!+zR4?@;A}u4MC2Pbr$WqbRcpX$4o^14qYOH6h z=2+k-oL`xm&Q=wTNZN!r}r{=HRNONO4Ld%aR2PN=vFT{U12nbZUa26S}% zog5#Bh=__dinQ8TG+^W4B&Gn)pL0V~c^&AK>iWkiPNs7cg%!*VJ@%#a~Wr%e47yZqJw2k!9~-pAF2C z#r)u4@3b`M==%ER`1!ytrGp%lWnjX4{;(967n)gnr+Z!k-ozJ?56TUE)-*h{w9MRy z5KAr(Bjw7)Oq!@Q$9kEt@DL8H(Dj{ai?HqO0sE6@Gr6mQ{FZZ$Os`y$jaISP?UB2` zBAOMR`q?-#AO~tTJxauxAMZdN&>KoThbkA}OwN4{IzzMKN#J|Ruml9;5ONv{5W9|x z`}=Earx%+eX^IE?yFXi+D-}){mQPo^MT+C%|Clge-~lb~8}7#o*?E?~0PQ?*lgQ4+ z>X~uaT&U#-d$m5O3>B78w9iUW84MXa1qD5O<^F{g`SCu#OD*x>cTrX$)^blr2N@kpFT};a}ct%oON462Q!UIL%`;EKGVglxNodG3J=IRD}$it3b$bq(iBhV$}WCcxKF< z6dVaTA)w;4`oUmDk*@@SMZZGb=;U>^c~MRjm}r5bP?P}ioP-bL6K0&BUkE>|A{bc-gV(=?Fa2mKR!_eQQ3rczA-UMKF-( zNJs=3iIf|fJ9|J@(jVa{T)6bGV-Cq2`Rwwl03>c(jK4 zk^P~A7@VMCsGhCubcqF?;3DEv*e4P8%mZ_V`!mP(-X36&0nfYZ>+7|(HPA3Fy?h^x{1T&&)DfM&pl@bmM&lQK8E{-y~V@bh>ncB#3(L(-epoQD=S;7 z)39R82hne^KXFwY9~)!RaMv_la*Ql_O2xEtC{hRw`@oUKei9SrXCOJe9g9_c_MWm`ET( zq5{Bz&E@#0`2*Q{pt18|MtrjUsVW*oTv8GNzBgG@6NU_cSbTJ&#%yLBn!pYgL@^?w z*!1`vTRt5oH7~Cm5Iiu{e}isQd&O$uQ($?*Hd=up`#?RXZ+kzlZDds0x8M6tAv!eb zubp9EJS$=mal*wV0t10!29m6$X|lP|O&jy>tiRa{h)QWZ zUalD$jg5_^q8Ic*fhot-_PMwrdT?b!)%U%^y@PE6BGSv4*nNq-QPqs%`KWEmZixy- zu;)bh2j5jh;-hX98-Ci!-cRg8sZ+x!Tb0ltiUP6eDLnl*#P|FpBr^=}WSnm4w14if zQR)r_9qBS^@twKkVDl!1TadVT7l}LchiQlnR^j6==swDSrkiWB-grOg!D<(p{-wuE z0~qt@x*A8AX4-t5V9w1oaSJ}6ybk;XeipM*8m7(^zTRb5++`yQaw7O&gS2Y@NIJd7 z8?z?;o9)q23|b4>n?eonO$a2o&jS=PyFY2YShoYH-+JLkpWU5-B0Y`%{58@$6KmC# zS55wP+JjXQ+V+THQhpW2ma&b1VkA> zrYs^!Flb-%Bc}`;qt{DSj>WErna=r%W*r>mmq`sl?VX*f>a?})VPex`Zf~>>dtv=S z{dK*~vI2ms0+Hw>LO_~^^YUdM-t)~((;aIahxtCS2K=B;#)mR;@&z9X!4cn|bBF2; z5u?{sd6qHIiWf{=I9A7OU&9(Gn&q{$^c$9MRV6zVMBVa)&~P6d93FsHvC48$M)>jZ zF>|vjDJgoQ)GtDFCGuW+!+gQ5EL@X6?LIp3^hc65OX9tD5mcMYJBBQ|oEOaq z;xJ%aYoZ!;Jl0PB+%Z5pI;f~A^Srzc93W)&F9}}KzX)~Ge7{ga-;qoJrWFTYWRHH2 zpJ=LlmyUxym_u1@i;2_XB+*Ng+G?ujsF0`G*|Y!@uueoKp;ZP9BL$PAD((6#dOo`g zH(F|Am8a?q7sDcl!y*c1Zq@H7Rhbwk<{JUg%ckXg<t zRJgbowKcUNz*}m4>#Cy&Kqm{V`gu1L2y+}%y0089+#PuuGiydfXEo%LYV@W9#ySdozfI_{_p z;ZrlOfr+D#2}SBW=mg4XC5_O{ezmOWJsd27Tzs^@D*@pi=5z5cZ?j?)?c@u@SgU6- zB=VFl1swEE+)vI-NjWE$`UjYiP=R1f*k=Z8<3LS+D4oxMW$XSfVd`a;QsKv2^zL8; zd6HW%ul;Hri3aXlOogiQG=8s>y963|SVc|A-)eqm#R9F%;jyIJ4>u9wt6ghIU0sr( zUrWz3%$V{u`b+upj>BDt3?Jt^3~eTx$J5|?K1~IZ+oAH9w$YO&;~QX=P*_;b+;Uah zQc&5rqrsJe0yz#XR4hN{F7PV#PAUHGWHmXq=7QOu#eWNZ2vGL*?@ z%=k;2TJyt6M{vtWxu`e~9B+Nwt+ctw4B}fIvHkOkFfW$@X-7FxJr_{U+x*+L4sQ9X zeI1~Y?edpI+SV{!M^O#qM~r27JFfKiWm{O_DQ-wVJ~sQ5;;6%)*Xew&vFD|$?ch2*n7F|G~ycI4<+}f^9OO8PUPxB3+1Ld5& z_16V0nq?WnkRdT6$AlB#Jbz4C*Dx~mqn%#@Zv=OeuN+{X{-Y^>$^y-OVR}p7x4*&r z@5)IW3W{pDM2Tj-`*E>uSz1bxVdacy`7?;PA(DTOc4EqFECu-}zBKvZb|E1?OE$vL zxR1dI9-UaiglY1*Nok65QIYs(_05Y>bp#UHo>-9F|4HJr8SC8U(O%q}uNp7`U(juG zo%E`Ka<8Ajq$~ujxw>px0W7tK4cT0kRies!U{LP8g2EyvjsLw4h}q^7pf1N$00&7$ z51!liYxvaT!eC%&{G!V)O}>GJyGUE}*PNAgjh4_n=`ZQX8I-6s`CvH2`EAD+VsKvC zGv`as0k<1aheut<%*(+@FMj_G$mM{%8|Wt9IZVQjB|^PvdpHQZp1JD4IpNv878(mJ zj|usKZ%Ae&uF@Z2I9k_8-T(7iBhb)XyKKsN-ueXNdG3cF+DyXaZ25rwQlgrp#KbkZ z0O(XF$rDCshBI6;;%{_B2n z*R9W%hK>lS6$q0p`iLSfnjPYp46XYKE?c+kWq6+PVQxZObLq}KVnP1j@8KwGIqmC# z)2K=G#9i+tL6Jj=X%NoWv>G^3%yZHhAD3?O%QhP!6-jLV`qqcUZ|S_mC?|APTd~`M z`Tvk9Sy8dWa!(xu*P}E`G%4f1rGQ7O9A`hWZHv(Dhr3})kz*K);4Wx{R2nDBl`?=I zG(zB*rEbWTLX6IjSvvM}8%wNeE7NK=|8>E>5!$#4D4jgD13OBE@uGjo zq3!qh@F+-;iOKjn4cI3GqQ&p4tP!-xsFj)yCqICK!dK2zrUKQX*W7{MIBVj^GJGt# z2n5AoxorY|G`U^Z`Ik`^e9~mt_#KY`U4@O4@sC8Wcpcmh$XH-~m_jo+ejxhA^ra?d zr7dXIVlitM2PdIn{(xQ2bI*}ybv4!q`&%Pc{~(>b&#@iQm9{W!i44X7%5IhtlYB%^ zPHtKjJyUhnq*4sc$$2Ol$l3T9**E)5Ny}__X8P-e^ZS0? z_9FGT#CvuLQMxo$=%OMl09vp+d}+%yD5x%+%hCLm0>`V;tZwmH23!ik94>zYqQ;yd z?fI0_i+Hc0=Cl1n&!H7ydDVUE&ey0+wnWJ^)q7a9pIDb3P)~?yEDP525u;_CAOrC@ z*@7un4F>A)_W@&{i-b+sDo<@{#(nM{d{X(Yvpj#`aS3;u02ih%so+M~YQng8RrOKD z?`WwAThH7}z+V`c;+j*jOC;YnQ$XJ3J~;BhK>%5kW7a-_^ew}A27WBImEw13hP%b^ zXO7Ad9*$uxkXv8N_SscHD7M`ERgUxWZjS@^4aedph$wp?n;$IFBf9fA}cl9sfXnPDfUSsm2Af<>wic@aj5-Y{UP zd&(|ZHiaBO$`rH|#_ zq~J1KUhc)i2im$e>cHo@!RJBrSyN2%XV89AUOEz)rb zbej;PH+baHgn)I!w49ef!_Bjqj6;Bpk-ny;=oKU=*nJI48Nfwuo^Q29E7WT}qf_yT zOe)lQ8^YWKni*(~haT`?Vf|T9Jr0)XbuUFrB;Uls!$ADoB0?fMS;9;Ou^<zZEUdt+k9E(I>khfFWa{Wgw@EXMT}yvi=Mq_-Zf|Z}p;IpQTV` z$3%ntDr$keVWD*5hXdgB!v6tzFNA;>-Hey0qu^B+pz*vNFFADoBuEe$1mb-~*u7}E zIBoo1=`3kVA}qvf;T2Mk8Rkg2Ksp&KuMRH0fWS23Q)q|H0wlaRaLU|I*iwsoa)QjO zqpPrzS)e(ct6=|@e~)ouDn8?zC7`>_^M(QWOVh$D2RI1aqr_YBzW)4!V+> z2C+IN2|xz?e<;VMII;LaD?HYfngY@DliGTky7yqe05n{`H|iJSZXHg-w31alFu*wQ za1vVXX0-tETWFlSobAfe(~^>!sj#{{Uan4pJ*)>f095qeJ!UMO zKb@aC@)kvvho6c zYIs=B@vhq7(s{4D*^ee&btSjuJhe3R%@EJboDw8RmF7o4Z>zhe@)Xi5QPQn*oV2ci z@$IxU|Iw75qV`?u!|38H009k7+0W}kl-DT0{8A>t*kAmpym0Qf8it47v|(SjE_M$b zq{OeP46~~mll>Ie+wZwcQ<>*=K*1v~F*%W#bm2z40&H=INUl_s%&E%@6^odBy#Ctj zO+A0hhE~>!;o8YEa?ovqm!O!AD1#l-RSAIb}7Dts$x>s z_wqT=mr(+ip*`DeSq5M`0)2JN^MQ8UKR8;u$uZt_8qmb!Kkn-#rfnb8@28!u7_lT) zhQ;0dn0htk(Ks?Y$6C*8gH0DFDS-OyS9AX%;;}yeyi_>eF^u|Pd(;C_xPUP|X-C@p z);az$JUMRofPO`)Jm8a~Qlqrf#=l}G$eD|@Q7Z|T%#o7pl8;z&5T-am>(XcWlVd0u zbkj&dg3T96{UU7IGHPVxuKmBc06m{u?Na|{vL`X*nEBI0VXnWO7wg>hFo$~d!VgMT z4w5Z#m>@%!U|}C=&%5l1Ju;Ud3XGlWiE_AS^KELtQ)+G>(81B1rsVP`f)_IgKi;O} zh2|wNryfU%yvWeyzbP<>j;>!Nx^x&BR?a2w*x(#4wr`@?w!Wtx+p>L}?*7{+K&HTr zs^8udS<@-#$mxu6nWc%h`i7SgyQ>ZAEY(GccIAJNZORGF~yE+R74nHJ}X&QDV zD4D7@B|1O~8l;au-xmx626b%kDK%Ox5T)CFI)AR)OY4oqA^DOjI8Mn?ZHnRYU%P}p zsQua&(`}@v${)3w1m*|ydsNWl(-^v*A?Wmn-_SH9B5a_0vD)AQ09B(Ci!y`X&O@Yg zABGx_PL44i+W8IB(gaY%ycK~wiNk#}vK{(Lz;S(V8(T}GdOK^@omZGLUGM8an8mJ= zQYEavSec$i<%xp1q`a;-((x?R{!Icd;Z7;`=|S`H0x8(KEY|+!i|GKOgmdQw0-{$J zrwM{CM{}eP4++NO?f-?$JYjt8p|KI&Fzv z>g;Ofn&vwLV4~8kJ?cxe=b*YKkRwY+Sk*%JT76q#g-Twim^;lY}&LU!v0*op$^!@tr6~G5SeH?x&{^!+;w{qbLe%H(b8e(SxKL+r!JZG;j^Q=irjpi zGL0jjmE=V!*t#Oe@$m}+P&M}R912o+cONsTnYpLsSJydPo4Ccj#?sIRT9Mbdv*dCS zJvO?x$^R2=EYQTo*H+O4_M;{RFJ)z2^uB$$>enBK7JxoK^+gv#q*z!&4$*yX<Do3=Yp3dA6Ph4M=`F&tT{(V%;(x-B9O?9R1 zO;Y{wUH*frh-kqbq^8E?O3y0ml7)*XiG!|ohOygXR~W^mKahF)kIb#4q6)MX%FFa{ zn;FA2l&jdvIfiRC+Sk42EVWRCKemX9~JJ=nC?) zJDsm4cs3u+z?HEiGN@lq$gdv5>v?-e$mGPS3B|^$iK9Bv(Us(Wpxqqu6wMn!k)>m+ zaCZ7#1AMk5B)hzv4%lnv(&A$%P>_>|weEM9M%8mBeuA=osb1Z(a|{Lx93(`1^xMhK z-`U-td4WLY+jOgUx4Mqlq5mDN@;x$I^fO`=E8}eg`pz@ zGDek+n1g9TikjsLIZ787Rs65e5&Wekue&#c@~RepCsbmj=p27N>P{b=gS%fq@PGaK z4I>Zb-Rd=Sv8sitytl>#yIM1!zkxa^-=s}~TJUw|@~^O_As_F%X!PJqY^{L$6g5D0 zfC17%x0OyC?535@hN5EC-A$S$Ai1N0PMDR>A4G8xMyl3bUaz$20u_@US$7XT7l9XA zN;?Y!HCNl`@z`DN+SI}H(mCGiPEb_6JPR+aYMeExeA070aUXtF1)|NA zFbDm+uk7%FIAAwU;8t!w_y zRkaV@Yd2Bzban^4GO0SALwjCe7MDU~y|NxD-#XBCd(Aa*Z!EM2(RAc$P#GCjcE^zi!fQ}y#HoV9k^;ai(ySO_YMp;D;L z-wzo!RnV3=j5YSFqyJVWFGc@1AV&u!FP9FOv*dcp80!Jh^#%_>Wz~e$B4poW?hdQC zzDiN7(8p2cjELln%xIT>pkz&XZ>$A6dgrqzg4%<*Ap?cS6gy&bH4ittpkN&fccnW{ zMgaNt-TaO#bBQV=mTc%0kx7bN&;N}rs>mASBv}EefpNtX*B7wODhPblC+5OEf7Mjn z@Cql%)%{}OWz8OA#LMriyURDJOzI9A`*KZdwJGDiBv521nWIaF)3%XUuCeHOdS9rn zCqp!QVSiyzNq?Zpi*6@5)NG5dcb&cZgqTJmcsW{n%4l*W^ueM2FXG-ZtgGnj7Tq9% zBBe-3BS?2QC{ohh-AH$b2uPQ7DcwkSiGc8SB?%$!;}$vC4Bo0byF#!bRc|)UdYPt zv2<0}NVUnYM}`;@U}kKt)t2080};IR(~ZcKUF+$~0?$MYLZ_wT(>t5yylbYp-M!-p zIB3(VXVxU%M&^zd1|L?4(>+Ix^t>-2##nxhLUvP4T#Xt;V)g3V9FFeOP1*yvx%J&3 z5?R{C*4Gv`I^F;>0ptVh(E5^`WovXDeF);xn?e-pwELyjXDV>Z#YV>>gsoP6c@27T zPrS4jwmB+V3|&I`S-@x~L+x!J4lyrb1Zc-ieI-A9eQD;bVs{f7+41}Bz43b_`v_03 zgBO`!DnvAU&nTGMY}@otM_k*Q+!`i5tv$E}>i}H2$F3Nt_LhF3dGkco6dPv)TGA9- zxCqmZ?8aPMPtdyAL1ETG1p__*aKcjQYGs)O#iDW7Aa}q;l$IWhh>!+j95|Qm5iWjD zpfD74jxn9n2>0y2=uMN9i?^Y_;t$n~iimW$GeL{`%%MINl+#q$Z^GT7A97Uie&R*y zta$U1Ce9E0@i+37J-BDBI5(?uwv#P300$K=UsR9;_4 zk^M^|Fz51R&yp)pehY;HI&-c*q@j8d_4GMH=I^Kl&b>MEP3s z4jw*Kk~Rp*!;XM62_+UY>)pDC$lG!-WSn1BQQP>W`#7xlu5f2l=@c>f@zTRu!n(IWX?qxVpo9uqimn8PZT ztdBtJC_nB=$ykjyGTI3FXz&QB5BIvb z8CWLv7_9HnY#d)z{D|Ve)qVC1eA$=L@%xg7|nfx`|Lm;_JIVxBm}9 z06W!PqKE+3jZMKcJW0mKX%=mv&Fv~G)LpW73+XB6wSvhuvyW>tQkA+c4}E4xkQ5kO z93Mu8__bX?%O{=xh#onBg*6Mv2b}PsL3Dln>uXre4-<{1vS}omc#u$$^7vvya*FQn z&C_w3^*Bk>TCP@SO$RS?w{uuoi*b44Z1S$N>cLmnDVRJa6r>+^OpK<^0kSG@RNta_lCX z`l%>FV1)`J70ncevjzfSpTwT}y{u_Y&Fd0-VoVx0J~&-G0diP?Dd{}-{NJHX2&%Ry zD9YNO_AjeT8#rcmbLrmi-N6yK-l|R4#W;C3%xnk9Khar-7%9D-*QdJPGjblyzp_u- zB+GaR2VP!7$8gR(Y5xn52J$WPNC#PvK+Mm`-S>qWJq!5Kd^a<@w(+ltBETx|vuh!5 zzV6pihB+#Ma^|$m)w0YT(|E1UOpVD91Z`t`_Mr@$Rm@Y5nei+Dy zMf6@8_M&U~IUk4Jis0J*I~Iy|>)L-F)Tro=n11TqjkW}}yr~Qc_^73iqUw(lwvF>> zd9-wVCYF|oSL6`iF^oOrf1-4AC;tPbTVFYm65;hsil&?&(Ao|D&~AScGOMxkkJm%0 z$CNmVnME51U-mfrqj2eE&3n2E%qZFJtWr^g<1Ugfk6lFtE zi!WBquonIM+oL|LjCrNDOsv^Ln&cXNdRtDF?29Yl^7I(JA0+e)=rjP?S(?hIoJ+qE zi)$XRL4P0R@s&pB*$Qg+F+dE7`|&~N|9?C058k4DXXD-WyliBgZostJ$VM#55_-G#a~KR{q@ejbJK;T8OI^~2lS zRaDZi2zK(8gPX_85(6IgC-wuO*Bhy)>8EkVH1-*66O(eY@4$@f-NTQcqO>eHWvK$R z9V}SoCR>-N|3+V*o#=Tqz3R&O*iBJW8;nkT)p;_-FP_fTKR$kc?#(<>k$fTr_&{e`i)DO1x%EV{7wza=b_&m zx0;n>ceh^SSC|4Hy{}a{(bH5E zbu^UJY}3el+LnIReLY6g>MoTLSvaR1tKg^O@c7WHq}wX=g&cpUuuO-iY;u=K^C2v) z`3b~#znL(Y$~88oX8Mqe@IGH@p80q+zk~?1M0l|5Dbp*dh4pOJJXa8q7*rWzwOf-e!><(@G_?nndBqcL;_wm)#qC8j`Ct$E6kZ5Go6ZKzw4kRVe1 zEbF+}mXF80Fwa_6=es{!Qy*btye_RkUb)5FZ@ZF}UeQO9aOC1dVbx>`I}bQI&XHPi z@_P+|XA=8T&HNv6kw`C6RY=ZE(F>^cvrptmB_Av;O^uHS>*G!at15g~)mBDSma|m2YfJ4hbUQZQPo&t50Bjma>y&fC> z!X@W6&B|D{nFg2!m|df`S`(z^#bt1b?+NiRF&F1c_V472O);EvJk*&hvp>K~VV!Ebx5=zjwARR-TREfY4TGCL`coq$q5qB=uHzX?}P?MO%4+;ics4J8kXW zsyC6}M1Torep8iN*(C3&7Nj>5CYG-}1SGPU`r3!>mP9SF!2o*uc~OC)3DmZ2h}1VJ zw|T7f8kxKLI8>+Xa7W=JRU7h!IVt=#9%ET2m$2|~`LNF22oQ3|b%!xUg6;wp)Ga{P{8iC1qjWg_WbuqwU#oY5xFx4Xf}^~$ zG9$Kpsy}U!G|nhPFeeHYs5bE3yb^!sa}muah=f7E2}llJ{p45jONuO+J08W3A-{hj z45HN3EU)Fu)?rf0C6*as#DOFL8Pr@;Kh`6}8 zkPbfwjIEy>tR_tAm;hV*Kg6!}~bNP4_X2QL4nw zZFcsFB=j9rOD#w|nJGy$uTtHzCC|mB_GD#R(Z2pQR9sVB@0X*lMU(rp%g^*F}+Li{a3(8G;bUN6L-l`JpF@$a#{NYrBt+k8!gHUgyn42=BA!sT+E>rtbH2b zc8auEHk3V>yCL4OH(*MXR28Uo*+f6^9IX4Aog3KjOFlZ=ugRNUvuKr03kt9wUAoiF3L<!%IcRJ)5nP^TrQS(YiBFQUi3s6w} z6e|JRsYVyG@Dn#PNQripwPo|@&<>ngtX z@VPE8ttnJ20B6DSK*{4fl#den3v1?;H7>~RS1 zzea*Ni3?tDR~uPbS*;HvYgN;E5fA_FGg#5IH+%nG=QXxG4upUZu%_LZy)7@+8IcwP z=6xktzuSMzG->gbeX5hb&{A4@fP$!)H|l=G-?S zt;>D!xJ5CxOd3jTKhG~y+U|je;l7-bl79N%h@);BIxQp#E>|~CP11Ni;GS7jTPS7Q zIS4V@=tAHja&of&0tHUD%*3!v)^P6NKt2s)z3ap4$julYW6=Z!Y74An(G*`kw+@`&PXl?$i2|~*)}2BY7jyykpCP+{ ziVCUvxqWWh7Oco6fK)pKr5LJeCAu8+l{=_+a}jvp;HcJ1A%d|VF@}Vmb(ZD@-%&F} z|Mr#e;=hL4~d@|DTs<+5F`K!diFXi{AV9C-*KCKN-K#ICyv zG`w)or*}fLjA+)lPeFZ7>S7@!Whr8A=VcYAFs-hmL5^MhRpes~x%tp8oKV^5%dhXt zCNaq+Fc7f#rl)zUs*u@}(>HPhN4wD#GWrcOSO4WBQ7|_CozJQa;IZ=B5}v#B10_>^ zTwEtOA!D~-wXwGeFb3_uidq-jQ;fEu-j~%Jji>x@z5{ww)4Xgi@DX$tfH}p{>spQ{ zZD6SbgRe~NpDJJn)5OBE!LeV1oF|i;TbN2Nq3oinD4Zs=H0}`rL5Ti-Oc>2etV#nqgK)k-x!sJ(Hz^g{^g{;EbHI6Z*nBEI62xF-k#2 zLp&y1C$BL&g^+{hRR@!YN2!mH?ZEY@eZADof;0E+Ug8!PO2^*bejKt$-iL*py`~t4 z0fPtM&g~@PikZ4GQu$E1aiD<`XZ-Gf2%aF%@Bg~&lvU$vB>xa9NSGf>x3u9y1O#>I zIxY1HgXj6kZiWi1b;gza;e~|s%(Shr&g&W* zHOVdDd}nr$QT%hecw)!_W~k_uhO}5&D^AtdN|7M>&T0iVZCa2_Y!FNOjKz{jJTf7) zs4)w(tnMgqnpK%Fry{cB`Gc_du*6N)gZ82vhrf*c9Vt9Fr zgt+U(Q?He~4HVhgK@8Qe&~VH_&gCkqX;|SK$-# z>E?hgi}6kBMsAny4~$gNR%K62sF3E9|CbIEfr=mHEnWVHBGP-1FStI)YR$JJAtHWBaMIJTn|7=-oWrI06u4KW+4;t8h zyIh;Y0Y)}1)DkgTrK3w}W$w+#4#^}ay#q3ol<8740^jYkUzIoWCjgnoBmUVxB_P+d zv1!c>`rljtG`1Nc>`-a` znh+B4;JWeo4#lyVzVb>&M%wfk4tyy8fdg9NZv4CMWq5BN;OMASyj>8yP1xA*-I2%N z(B|-ugM2}QQ?vgSvKx|r4_NPJ#bx2a#GJf_cAB3n<)(z5CI~Z&uO@$?2^aO($Mx?S zI>SY*Dc-JTGBX#>KD?L7S8!2`+gsZMa)a^w$9!=SaKX!LD7&*CS}GMuUxsebyEo+& z=+IePLX?>$@)JCvG&A)F)&T3Pt0|$irsr&dyyaevmI74k8rL zOBILRq^ z%qXx%XKGJJ$E$p1tdgINfTR zQrHM%xlC)c;wG2)L~g)|?w6M!3m)@ZYa*;S;j4AAm7&{23$ST*Y{leTO^;l+ zQQtn&6M4c#W!F`!GwkY>b6cWy&(BwqL~59<%M`tS0+GOV6dRFA$Q3EK9xQCE;xD!1 zWcMd(M!8h)sr)n+_TGU1OI!O>fYiqNYLFpYL6vq+Lv6{lMBOPk3f9`)N&Hm7qri4>h0|MPSoIN-Ms*qs z51|0X49*7Xp2UuP(O^2aMf5W${JfgCc>d1k-hyMCR&T&Sv$149rkhix{fa=q z;+x-O8b|4PJUe0Y=t$rP(pYXTTXjngCo!}j!)_xMEacVsESJUExC5ER+2SBXnQ&mz*i#HaGjw;j|iFGvCwv4O-msv(Q|*j)DoM|o&bN?xsv zgCrSL#la~W77O~;}A+?_jR0lDc@G;~-vtwjr^Tc(4j$;-!>9fQS!_G6c7YqMf zWGsD<&-TkpWuu|tJ6LeBc?0&~FWNgNednOfcEQFIS=wf`Sw{BDLyzq_d+|&K6jZv3 zUfR{Qvc8cvZQ2Mr%;mv+>6ML0g7vi*#yhuG3ix);>%y)7X7m7oX9X=f6_M z>hgHd43H*O6r1mB73Gg}sxr+GjALSWXFTGVq-1~d=O|m$+?+iCJwhw#X-d< zmw?d6aH!|7mw&wtmUBU@>FaqmXGDCIK-O;bce1H?DN$qd^JxhlJv;Yy(&0|vJe!&2 zWn&*p=4mZ6G7Xnay4m_-3^W1@Gpl|r;X$u)Ho^%rcqk2-LA$(P`sw{;xG!9u>8IAL zuIlj$3Ku23DFNCS2@EFE;(!BCuT(awsv4($=bWjcuWqit^ZgA$gm#e9hmsXI2xJYi zkw6x$;LtNwp2{+jgSDe+!-_pg>8*a{^eGPgV&y6lr%OwQ&r!c;isUKTGkaov=jK^F zc!-@0`rA^C1^=a^S_2g*zpELmzl_68V2U7@D3P&Y!a$ISArIDP`ob#^b!)N|Cxr1; z#^&h!MQKS%b}52y%XY_K;{G&iEq^t3ryGLu`~Zy=glj_s7WLnO$t~3GxaV0u=takG zCN44Fw9n05k3C@<0|H+x3QJsF29?d8cY#)dpAg;$PVWcd6$qffln}N@%Y0rqxsd+V zcDOuBz2C|VsaR=0^cKNbTXwo_12Sty-{`F0iXMDd8#`I`X|0h0|Koyy%Kr8eXNN%K zADosVxAb4ydn{7_^Q!~lr2u{0WSRR$PFqbtH?TA8{LbXJ31mT zWut$80ihoA65BPpD-I!_zKbp%OUPT6sSnpMEB!xuZS|jB-nhQG6|_Zp2E}}QvH5VI zkBdkS1$Ghq6XRPun3rBsjxwF&K)cQoj=lX4z$b(9?&xnGW*HkLg78X&+!Q1glO<38@1me@OdEqJg_b*wv<2p(DjtZ!pVJyJ-edd z*ZYg1Zc^drO;l5yL|tK$>V4%ojjCOrYsE?g*LfN}W!Se1E zBjec9cPEWD$|2Pd7c@Mtq(sdus#Du;A5M@!un)_D!*%~N?lxJk)uJ~^X&&AWj^9(d2!sc&;sz>W zdJQ>d2oj83`S0w$6_43VxfV&*rvC@1vV`q@pvVcI*dSa~B%2)OYAN5wq!gu%z%t%1 z1#00_ZjmP;7pY~e5H!>jYx>7Qvhso`&dfXlj(U_bB95?b%!r9`U{Y`@xSN$<`R_Wg z12O47T-blYxD{R|{_5Em1QMuQ!s4Fs%gy)ZdsF)4X#SIVk-m=sI*4DuIfgLWU-IuR z>~)te1mvUPK|+W1dqg9kM4nSeE!$%@oS)sE5W*dkcI!r;dDB4e0OYy*8To`X-S*-? zzJLM;ZbE}n%o@Bw^$fyCy_Q&N;1fungxsnG`^~{j6g*=L0!_nPW||9lk}xjXjqw59r>PcTNtFov~J4#B|_F*c4Y}r@B>e zM1=i%@k6Y{!lLjL{TZTfZf*lnb@<4^9Jm6qL<|BPZeVotXjA^VqP*OBw?7;rxdgm< zSeslV#lI)~fk`-zPU`5#@K$d{EYfLCd{B^7K~!xGn>;#Q*{*hE` zH->NFpnP7uVG&<2GJQ9JLTLXkWmEhKHQY?$DQ}UPPS1<(QMyVv2;)=E$daiI($MU7 zXN{qSvc_lsLk67(5@ckBvBn(nN!$^Z!)8pqD(f3+n-h0NxBzpFbaK1qzP zCraeqJsG7MXRtR?@fM>3e5r4j~##ZeOR)=xa*DGw~vcRpxccpo}7!3)$AG z6m9cHkW7628c|wGQw<^sxC(M?&nu2G-|Y+b)ejr@K{#Q9zgmL+KHJzB$RnCNmLmg? z`rk@&`9DL;Ut3Asgj6PmBzIdeL1JgK{Fn9UL;rs2uh-W=JRhI`d<}l%udT2j&VPRQ zFVp+~-@p2)N%5bLaiW>E%+|&U7E~7up7=bNR z$XVWJ=!Jc68DCrRfBYYx>HotgDSzcEA}RuO_0@CV{{3}ydbwC<{^I-)IMXkBJa*)7 z;liY9SO2Vf-5s~4y7IxEKz4R^MUzdRlWtOO_D&g_y*%j^i4li~%h9V?9`81Rew}m{ zQJK!EwU(CF<0^W{^(|Uj3G13I7e1%evMpCU5|K$LHobb4S+Fq;G*ya(@SZB(4=1#9 z!{aF6H$|u;!W$q<^ssvG@^JjX`+2g(BQy^oP>?{N{jZK|^Cfo=x+; zjF2A5xf+$_Ku^n50@fQ~kR3WQYO0=#R+hRK`a!?acn2e~_Th=KcOco%E~5Mx&Cc3S zAqCy?hMA8i2?^3c2MswGYv{l2;glbJ+v4L*fXYO41WAVkO`HO{y*)$jU@8^?L41Z! zvrl~cmSaOKWrTdBIa>o^_kg4LTk}})87bN5+22fIyKVaiI~9CPP4@N&6k0WA_ei~t zfn?d>k|z}kqlQSOS<}&tmedH>t6Mm(!`qI;#EY~E4xXN#Ml9+sHbHP-?gGWXWDnax z7|)vG=`|0CG4kd9MiTPJq<94MMnyCGe6Xhp?QBw3sQ7#t!C;IN)Nu_S{i5e87LYj@ z4wQvk3g=V-R9tWO7pcP-N?m!`v+^0Vq7azAbQ=g>>lG_(J^!hBg z06``t)hA@x9u&N%dR2Xvh5;!*)RYMJr?Sts`*~1P^73x}#+2ow$9RT^kLPo@{g{ug zt+~F+)#7?QVz(Oe50cxPH`cMYwDjz1y5=+8%2x<30hK#|#zKl% zODQuUVKOF%_U`VEH+x?!l86wOQV6^BB+m_UFmcN8`>()r*WjQtPOS*)a%TihxSxeR=#-t zeALY2VsBbmLF;4l9}Co33rB~4x3St)T{q>5-XcWz%4+_`UyTlNc5`Wewb61L*~W&r z{)02P&)rD}IKAb}chE1J!_pj>N_MUcyl;3)jpcMxL0$G=C3L;j5WF>Hf~#6fK6B`o zS0H%7KZk9xL$Qc6u6nw6Mczhlzq8~p?TIlSW-fPcKG+$jMMB@BdJB{zxL`C-4!T~z7-9hsQ|)hEq4iJ zk&HkJd8E0i35QXq!lC=3wajgIBxest*L~W+0I<(B&hI3zuic2fj~>cyFBx=OMmjsO zI*zKgw}a3Qd>-%F^JLOoFQzv^5gWJI-(ONuxH9jeYBPH=+Dn;UQCrE0Wo%=5Ie7kA zaORiw-PPJR`7F3L6;z)JeXsZL+lR@5N<_687)qrOPb&un7Hal4sTFfY#JUwkO57iE zid0LRDmVJ~XKNX>s<^1Q{@f|o3``F$`m|?uBBAZwnsx9LRLujs@;haJeAjmjJi)?V&CytCWm0#&;zM%w~^&-#S6-qG6~+EWrZ(#ViCM%ao8XjB`u1J;;~csc$Ht) za&_lleMlXrEz5sTsrc~oF2k2Zco@fZ8K_lb$m!~Sc7FZd`w3Ts*Y$WVW9!V>_+-k% z&BrFKU-S1O7uQ|{(ObQZX!vMkZ@!r^LONg&(tvKJBgK) zvQ{c;Y89HeYd20ujS`_HiS>k@wQ=P*NKRe;r`%sJ&)eK1-A34F%<` zfkp4fQ`Kpn&9${L7&rH184x<0Wz_98w`%s63Gh?>-8NCH+)?YL8$)jK`Ex;I#BVZD z?cRFam(0iz7ePd5C_qEa(x)xLjnla2Cny+z+%{d4gmN9XW&$$wJuXfmjd>l*80djX zNUZ7);SpsIOWT)eD;Fq{OQiC+&H3#DF4?o1!S>+pqT<3m6{m#wg_*m3PS6r*qH5MD zo5UA?QZQ`xZGkq)U^gTar3oIXMo(8A(z`{0`_WX)M!9Jq?3H%S_r^3`Qc7yC#!4{c zV`gq{S6Zv3Y>EJ3M_7jY1T87O*X2Q)K{-I-KR>qc>;Kd*Ewxtq?}g>EJ;gmWx*axE zhd?5Mmp0(!NcR`U3JWIb_)79OlC$^;^jpaBFVC4fKjezcC$gfVNEBV?)H`Ml$!V!^ zeV#~F)$ypzZ78fdQ;~LxqaIwsMF}iQNl_!uO~*q)1di)RXO@z-D{$Eo z&1301Vdf>K`2469N!YcAl7)prbjb;j^PQWU%a2rmNF+WiHg-1W<*ELj0%vy5kMxnP zs+~gb)(9kB67$`Z9~7}Ja_FxxyJb>1I)b}9Y{Pb!+KiYA9EELqpdTs)TqeL#E>^Wb z+4;g*uQ3oQZ`?vUvU6oZB9`KbZ`&vug85e zq(8m>xV>?DMS(#lE#u?L@BQ9Cy>?q(ekpi=)mnFj;DLmstJ`|hz!U1*L6bJ1CZCp; zrWBAe)u2SjAOb}>JGvh6>;2b@{Ug5K zCP^RQAx9eC$+AZhdq*Wby|#=DkB6ypGMwW6;Wve|KE4YnvvI8sw;UE|{X=eUZuGk4 z_N$z_;}a85ThOjRMIZA=4vVXZsOGYs{^kb5+ldFH+qqf?4%R=k%M2EY3X0n7f)1o9 z{dTrXHPyjIMYJ8ie?Rqn?-U;Q>`>sB$~Bi>Zlc~19uW^gM{x^|?=gW)I+u!q0>Mk7 zmzbEmWL_ILPQ+%_DmMZpbse-8-{_|Xe>_F=^nsX^e{i3EQmeHt`M_}Kg8Tk}7Pd}K z<;GcBTU)CMvT~ce>Ku_a&WImyzP^XkZ!O>;g6=aOFYkwQuiJ~g*;Yy(ikd@qf51A~ z-+$H8b~XM*G^@U`G4!LC?$9AvauP97pJt!J!^K6c;-*>OJ&exGteA2%Qjv2Z>^=?; zQ((Kfx+bQdTzjH)fD?0eW+wj`A|#bFPp!$$$3+op>`sT_`icA%gi&3S+~o_kb$3s) z?S~ocroM~efLl~!1+3CHrmabe1xi{BlV{)C*_;ni*FXJI?htlcyt#4rYh$IOYiekC z1@*#U;E@3ZpQ55$>7#}hH4Iog)G1bXPvP1|VJSN&O;)Jr5uF@0z{kM$^2jW<3e_GB zg(6FR>U*4%`)XI2_?NS44rlZ6G#z=cl1NO=Li0s{bhlCAo0hi9JTZDzTz{IPxFOQG zWRoj>T$gA{cj5^EGR={ZPwFBfSWWxwoSX;=xlOwI)YVO8p0t4sS+&IdJqgAX zFIN~Imif7$K@7>X6i4hdℜPfmEGeOXY2CBrYzI?vOku=IV<o9Ii1%>ViNqZ-!6eiN?(LKZ2 z?4LgeI{Y(IV|I2nIILz+uwl3fQw^{BG#6D+(bv6D5xpcUZD*>gPEJ^$d@4qC^vI+0 zV=J@#%03>maCPHd1vuYdL*J&RTsd7XL2C5#r~YKAb^z(v(GrD}_;Y0BjTd|gWv-c7 z!ZUKZNE`!x9m{vC_f8TLm|O%h;Fg3VU$ z9*FJM*Rw_#5Wq?7b4Tqbw6wchYq=d#%xWq=!ewAq{c6A{O`~q9d?`vUz4G+S(@+*1 zx*b*UKG9vS*KywmCh3$d-tBi8!{P^%PQPiDayc%kS|A=LvF5R`;+`x2o85n4827OxCWm_GoQ()@?;0>XLDyn$zc0YRAc%Lmy(>%^i)2XV~ zZt7lzYUJD)Poj-FF2%=hSZVsWIxc*{qD?o$=CE358eVaBFHxynZwwD5){JMiX{pLG z98A|u4v>ayPp*z9b$m3-MmN^i)A;P}>Ot*rKT)*7NPw)#jsk&0l2)ZSmB;BO#inWG zQG{f7Yiszd>3jnZrF$@AkcN)VcOf&LK|4L3Lwkah*ZIuWe4GAM^?AUDqRqJ>xyRjy z=A@(-Nm3>gNvz@F_Y@_HuaP?=_BB~qHU=|VkB*j~Icx<#+W*z*umP(`di9Hm*+Za~U8Ks8!Dp_ufNrVuLFPh4Ossn|-7CG_NJ;6t z)pW)5cjEU4IDG@{p8LEpH~M%8=+DrIcWebtkg#9zG&B_2yca>JKkQQtx>;qSPcU{{ zBD!P#Jo5W*83IM@JIj2yt^JRG@`>*~D#HI8BeEu-5)%ggAz_XLXh_}Wc;UwvY=H0@ zg!2zfd9y9>u%>HK-D{Ah@b8;1@hmPyF*|*CO$Mu8xp7#ge(`2wK_-}BPGNH{%{7nkZNVj9}RgBo}&hut?_4_sQ&Iwy z+{Y5Pu3uo6+#Y%D6a@JA99L^G=w2>sjEVy5gH?YgVker{Ph=k!bv>$Nwv{Ge`abDr!Xo6e;nC#M&E@MpC|{eB@Zaxy$MJnA1#R8&Ya zkOr&Zv#srWQKgLjx8rx{aIHEe7%yO^yZnLA^V*g48`cMX2e3+DO0yu|Xg3-4fLd*0 zV&dldnqKP+_x&pz&o9sZn+q^$Uk2X>e5?%&4AJizT61!y1xfTC?g9vS2ZC3-A-&l_ zL7cTsdW|N_`vxB~Hm}_pzGUXd#(sQhS)@&%fQpIlQAv{F&ZGYUIjD0R0p_Jmwdm8D zQ{dvk`$Ru%+|q4#-^KZGL0#9>nDIAKzhnA=HIb~~ZK~XqV_s<3rjhPz3AYPVmtGfU0iSAMRPof_Pa_N{HDh+(*_5QaHSf-5_RRVX`o?SNB8>r1GFaq9RUi z^$H-lg{WK+rhO@ouz`K0>~T6v&{{@-w}F;k@F^-#T3>B-OiZXi7|1C=Z&&y6#$INi zlfgJaC2!oUN(fjo&M0Xp8v}W4A8}jp^P!0s{r%0L^0K5k`fbN&<$kkH%MFUhc)(Z^ znx3YP`$Vr;@^g60nSd8Iwrs#7TL5&hwb(4qws_H+?8EF>1NF6#tX&f9B*oHXTuweN z270x6C_wE($Hw+>Y%{miJ&G2quQ_hpXJLzkN1(|h0NC*(8oSd+*oa76OXgXJFn4^bt z%f(-4q+QG*VZn5mxIN}K`}Jq-2~Q{4VCnQ4iehFwJ)e9CT)hH${`CfanOyq;363VvqXxVE{j1x{ z{bsk(sh?&Q?){B!lVxDcuCCEzY)t8Js)gO59oT1oV=$fske|XP#L+$#Ly&3aRs3?@ zf6WYmFv9f1+&0W)_(mz<`)eG^ylf1_l{P;wnre5j4}rrYXQ}RNqYI}*4AWM|!l2+| zRJuUQ?oyNk_3hEmC(tj)=@4)W&@a2uc={Ahtc2IDxM+e zL<&BP4W!ND;qQovMYt^ow3ce$GMvBJURmi~=Xuy{wCS{Zix_*nxv5j<$jy7%Ypm*0 z8x|&|9u!8Se|-`M3X#u`$1W71)hEGurduZId~T^cuL&_v!5)28fB>n`>8i^`)n3oq z+&zJ(kWB)RF%X(KOApBTu@ohp(h>tR0+?*$%5mJ|%^l`x77>wWy_CuMI(*9y?YHIP zmNqu0OZW*)H9H0u*C5Z5A7<_p35f4T9NKjqT~14SQZXqx;waSqyA*P zxm$K`wU6|82TtaaqOxi}K3}&^jgm)Zz}8v8Y`L1uL;R8`K*mo_-p9tl!$nENNXNJG zCpZiGWqG~IC3Cws8VNLeW8T3`2T~pGu3qx;He%6c+*zNB%1eH>)bjkWq85t2C(BW+ zd%ea=-~YSPRQ59)9-wpZu>I}a02^3c*$#yHUq?sEMGp$f>cb@!1Zig{$N#WB+tiDl zBAdH&n=~*CAA9$g@tsCwDXOo32I|_Z?Ntc=~J9|_N``_?fY<{tF_h0XEAE!cfkNPiX;~N zMEFUZz~f+Vn?r&Z%9cq5noK)Ii~@%{JLw?M<)uH2=GqL&QLt3}XvB+-erM&b0fO_F zJH7&?WUu27z}m^)p2jI305)4ge#yx@492RCN?}^0yq;S^DcL2(RaHMO73Vsq-_g-| z4yLDl%2AkE@V*-I`yL!C6+kTD_@gNkP#w_#Lq^9~*|@f_vFdDM!=|+yk0VJ*S^2Lq z3k1$yLJSO{tn=Hh-v;H5?#*r?&tp5s@afRE&v`gftfI{T!8eu9jg5>9vm@{f$Nl*C zQ%1%^w^M8k1QY@C%i088AOj^m)(RG3e2_;pIRItENve5DWIS%yEu^qM7d@o8@&$h! zp)imAH0yF-v05Xau+0Yw?z)y1hRYqS{}+75773~Jy-j{e!M2jKPLu1yyQx*+%6SmA{rHOHxxq;VX!SmgY5bePo$Ut12*E(qEQNQ_0)j?ip+(P3ZZPsRDuh z*$soGbQ<{WpTx#0GH$MEwZ_w0*|d66=T(nS`QZ^~0E&_0`X``12q+OfJuQjJEW>-K ze&oS6Ef1O5*=hw!8=Lb(=GE|in{jvgp z*OSATJCHq`IsZ&9L9=!evyvILBnOz4fD+}$76pj8B!6dX4BvY%fb<)b!sZ^g;rNpP z9+VEMU`%JEq5`!7T-#aI*?gl*NaxDILFxVN#n*tq`i6${xGyAR5J(0^vHjYLKw_&B zB0fiRbe2ib+7Hya=|<8zGr6rj(Q(!oL8I%MI*#?yZ`ma3AKa~U}TNXDjTXJ=~U z8g;J6D`V!P-Qkif8Hf)D9*-E$!m5v#XD(;lO@EqKCs&$Q5_vf;fJ%UoeajFA&IfjO zc7@2ewB84#!}J0Jt8pJY{bf)#>$vUrC+`%FjA*itD|<241oFtnkn{5L7BEn?s$9S zD1b@_;b(r_(2Knd2yZaZyp5S9n2j4Zva`l zv!5{YpIm0;XA&_xMC|N;UqBXdy~8~(Ejjtmzi-_eWGNKw?Ug8PSj`Sc?{fLH|6W?c z08pe6OMtX=u5j*RLc3mFT|>*s4t}m6{C~qxiY5rx{p#N2B!6vS_Bd(~E?W$4esl9Z z0I@;C(e~s|c}&d8x1(=~iBoU3StrYGmr5`F6RK*tdASX7tu;p5KayjMd6((V617M7rCylS9P&mv`CtP!C;@SL*npTuM2Bx5Fr7r%V^^WUY9&+)@Mc1 zf*)(Gx;rE3$MI2!?@xupMOi`Kb9Ic_7n!VCYrX$hofhFeBHoA)B4FsnPnfg_>ABM^ zd4zoQVjYzxkMSPOPLx_z;9OCO4h;=$_J49d7QMT*tnbBv<=+E#gt2kjLftWK42f4s z_&Zd2FYS_rAHYU{IVUVQMCbaCA|$8g%1TZS>;py}smR#?<+0AGXzIHGton03Vl$}# zL;H6Sem-M!$UjPHqkT^&I?t8XMGw2P2EzSz^9fwP|cxjM?!^y*4A<~W|S9iPP8ci z4IC7FVRWXI6$*-!on4)cb#*eSY}{O2S10QPFewO#_VZCLWW70^pXdMGL1tjEl7EJ# zE+((P94jl(f#Z7eG+kEUKJR%C^}Q}K)V6F3I_C}9+0O~kD9O403@o%X&)Xy44Q-#_ zWQn2E>mILteO@3SAOL!wWu7YqAt&|_MyLkkx>4)dq$ec;@MKA5X6x?X-a+@rv}P{^tcZL9i$~mOplPftc#& zUxEPVh&b=5!EgKxXI8sX`Qd-X!D4N~V5g@xQ?s+CeK1Em=jUk8wa@QD+wXVP$t|25 z931TJ;eEe;{YVkpV^Yq9QVFnFO;oFD^B+FD=m90U#0c+SaiMP7a&bk>*AQxIY65do zI3T0)=g%J?pLpKO9qVMV{H6Pg@CUbnffv|xy;{HivLFElPqB}UP}E|G_*jM2t}JMK z*EKY7yhv%CZyNRtr^PIhxm|p;)M@bStTG>!{Pf8h8yn#YNg>F2Z>qs)JEV3|Cu@8Q zv{1P1uk;=h;Gz8So&8zwfy9h@kiM~~cY|d1HS#WP1lsVmgDEPa@J(ORzyQOmV57m( zG){NbgsDo?JT1};UKgUsP4c);kpu!}c)f0S!Ib%Ny8x$=v{-jn^!pAv;Nhs}DwMeM zjK~27-4`VQ#n~Lxb-g-*J5?qEJ~a;~lL?`rA^G%h-VfV8|5CWL*47M{RZ-)d{*;a4 ziO2s3xD9GsJ&uW01Vy+F)#)YNQbO7+k!HnSN$ukB_PJ`AmVny{E~KQYY@*9mYrn1- zp(Up8FbYTr8S60Ulcc0du^PA2&Kc*TYbT%Af4x(0Vc~mF`J0$97_lU&n+~vX;F->Y z<=nq$t<~V@rZMBF9t3K_j+`8TB>=%Y%1C0L{qtI2B&DOX0F2WzG|yLNZJI5LgdA${ zW-LbAd?$#JL7e-RyptRjcAeq%upWrIYcm@GNC{pm!Fd4WTrAkTb9!9JOZVm=Gi+~Z zZltSNH_~-z^-Y%xPX2wFVS%M(*h7 zq+DQ3^M$09eUwKh9w{n1yvOJ{z6JCvPKJmpP#?m@#R+ML>tJeg-{d!d5~u5mJNrHq zJvp2hIY5>cf~EQf+iX>)-0-(NVYBSjN_L1{mJ{HfN?Ruz?uy{8^TLJZ=ED6C^Sc29 zbL}#?V|j^7I;*~)zUtXX_wb9z$hN0iZBRpwkx0SM%np{pp?c9_)2Gq(-6-P?w_5hZ zc=0?b8%jljDl?PjY!&J0gpM_`c^q?ixtToNFyLAori8o>5@=EkVd+@q`SKXq`Nx~-Ob&ozq{7`zI)GN@elBF&g^+-o_VHt zp_GIMujqNdbBFuS#TPF%iuu%) zExPk4V;e6r@C}lLxLovjzu;5?PZd}IpcuY7J2)qwiv3JY8s~jK?|D!v<)wb2jEEj? z?%Glh_Q$IEB!<8s-^0{&A5%-JBPnr8K_?U&_GD8xKJ5o)@CyK(uwjU5msc~F z`z(<~hwoyg{fvr$dHwEsR4o^8P)!34;q{kEUiNXrBg~)WH}6#S!c-8Re56}v=r5#S z9V)vGF_gX%Vd{%SOnk~ail=n??bs&+k^?4e|I$FuM4jC_&ul(0L`^Me$4nX z5~FN<^mg_CVJQLp?qQ;aLE>YZE=}i{dsB^KCYvRq)O83p8CxWYS=5By;ByUpymQAd zXIQP(8?iw@a*b9_5Doe=#^DwP+dg+9Frv-Tt`r2|S-Z-k~r)rLR$* zW%^pPurHmnuy6hCq13H=aHv{s$24$;C(?i0r`2?j5{bc!I6UH)L~er8;-aoslF)U+ zCj|WNo>vB*(fvOxnW=?@yc!ajQQAmWz%(CeR7-nHDhbo-l9%2^^8L-H)tF6es$e1> z*Y=*Cts+l1QL=-p?Osn+#icg~>1H9(@iwDCxD(WP4 zP$IOAvVo>g$3d#2W)+tPy`RnpT1BM*-9?4PN|AS6zg-OXx(LUiB*pz857i&%#GnY_ zzdd=86kawqyeIW&E_Y#7W6gWigi+H`XCw0>tZG> zybNyId=74nF9-2iFmd$i=H}%1D}KAH(NkCru(+{*KMviOyR6WVk#Xq1yRCyl891j- z4i9A{378I!z!hT^nUr9dm3qK|;&yi(P_$-yS&KdqcgAb<8lN=(I!7@Tf8Fu2zZ!8d zUAP!or&L{d+Z2BRCEdJR40H2wnzjTA>07Ci7e-!}2zr{|g(9T3A;^!n{Yd6rDVV#MY;d3w3HqMI=CTa}L3NU9Ixxo2 z+!aL+4jbY`If8hRb)%2p5rYk(R+8K8U`F4r?Hz2D<@u>FQ5U#-bZ2?lbx(6V?1eAT zc5{0}>ar@It5oZGZ}kO#C)(c8@%#6n%y|R=zlz# zhfrs>M`EQ3ApwRSg^++$1aEy0#L(~G6Z91x0SakCLIKEWKK7eXQTGtsr+uj1%TbMJ zwDFWPdnabJY^!i2YXK;xDG^xFC>FEjlW#{W7r$#?aFL{%k|XVd6U)H6%}a#;^bHVG z`w1*Q`22cS6dmIqL3X8XMRR3t$u0>RxxHEsLi)w|~~!Gu(JiD-Xr)(2jx zu8^q6C^#?imp=;%tjx_>i8C2rf;6#?&VI8)LM*@AL?Q8Lst*bt9gaVr+g0rzEgc>B z1(snN88c2-S%=WLr_9cR&*XO&91~I3NLIE^;}2{iBK_EXr~HEPP-XL2WB`%Uh2foN zKZWwU^~A9aa0#0L#}1L4-{}w3)yuE0BW0Dsx0S7pZ^tgbnHgZrW}keSUsfpiJtm+O z0H5BB^4vz>V1;$Pe}kKclbvI4szH$&=he%X{I`8}f5=3GE;ODuZ!)vqrzI)84f_9h z&4p^A)9Q@FkUp2up&kipiz~1QaNT&RZr{Cm;}}9_puG-KZKz!~W;h`|hK09_$>u6e zQzPk!02Rr8!})lj$2Dq$w!p!stKYbb4d()@$CqGE229cET=mjEZ8PqVW3?GJM4Krr z&o`X-O`Q1aJ)j1b1qOX9Q_H`&XqV;JX2Lo>Edgs4{~u;GH8=tM6u;f=lgi2higQ1J z#VtWWsQX1)Rc;$-BgQjDO^0VwWhUSPZ}7@+?q8i&d&7fSYLHkWo~bfsRKp>`9Ln68 zt~~7mSz}s(pl=UoA1v)KGTBJts{_GZ&dFv>49NY|*3>-lX$4C+$0sHfw_Z6qI`*pO z6X7z0-77jZu3HDGI!ODf#YWqs<=i2>KlkMK7wR_F`#)_8tZx6$N<~bw#U4$u&YQR*P~meEB3+ta3sLIvw?$wL0fk6C&Rm}|E(1D`38z?Jcq${ z`3L%E(Dbd#!|eNX*WNz$0!}mWk{oap9k&nJcaTX@Xv%#&Fm*|~Uj%Kqa`?#^-C=Bf zaZy&Wf16kvO081#T(JlHM|iCCI*5=T$h$S2hV?l@&ccq4PPYRD{V-G5IBC0nuRP(g z-n+a=L=LlsJPyMNNs)XdGHV|N(N=uu=Nkdd-G22uL^#5X0|K1~l1VmTW$ zLf3Xz!^&RTqCID5%_tKju)%DRva<4B2^`aE3s{iOQVRYilXm@T7wJ6}4u_a%SN+9s zFO1*0DTS=*b_A8F>1aA`%zFD2@#$9%8)$ou{Qdduv&8pO{|9->J_dn6zr^S-Ds`SeiY0O^f4iocfxDg@tJ!5oqt|h@P*_ zoPZT1g}7{tQxT2}oc1LXY{LeU1yBGQ8gP+2cGLa-vrGkUu(U;SfcF|K?aZ)MLEP=|HM-WMq6cB@40+-<}dX4)C4dZOM9Cd}uu2_VM!a;s*EJPhYWD zD&l_FN@uLQ@dxReg#53$TU*lgU`%JAe?`pSJ|6dHJKq&|@=IVD|pn7aVF> zw12l-AQ5hrz_ZWUdv`;W7akFF6XIz#Uwhgc)a>?g>QSAV zQBBQE<1&WV|7}!pvENTxOPkhrSKi;ieaCCP(OH4g zu})oymt7%=)}xkZ{jq#ZJlBg?4tWME%bnT+{B$rk-?uX=;_pdf`L|0W>9p;+60xmg z*qjdr6L#oji%?IE2B)X5y4#v`j&_n4T&{z{MGNmKe1#r=ohWo)lh&FWjnC?j<+1MQ z^@Mn(wtbU{d2Q3-j|YPd^i{8oGaGc4Ry_Y64SPmH;=y}A8O4x{PRwDk)Ve^Bv=v)h z3$h{^88}%yJrX^XqM|~Hd9nQ z&gkKscbK@r<0aKNvLBk&)|)aLp(0JLRDKyG?Xpq?o0-;tEKZbd#7VD5Yik-;T=ge# zyB)3OLaRJdQSHbb_pXe8C%EWgF4fdbfpRHhMYPJz%|$~+WsPS*^L0Pj9D2TBVRA?V z&ai2?!j0d>A-}OStYcvj3u)XYy`RiyA2+C~+MT@H?+Lf&bv)_LkO&vJ-(kIL8ATp; zSp-S?!j5Vi?8ZA+sK;GNR#w)d>+>mE+Tr2h#P?L!dG(o@9ZSXFDq7#RJ^MFM4UL5N zZ_Gy5*48o?JmRNDN9~tI^w#&ZChqi~0-8>9)rmZ;9^L%~Jc`#vmQd9q4G9`cOY_Xp zy^oVx=)P;~hS^zJW%jPqISC0w2d_yhl+B@eZ9=l|Iz9R`va(2fcxI=6B^z5=y@#4ZrR?m+TlNP=#}X@* zf##%!SZ3y1*OSebZ6ZF`msC6;A5Hth;JU=%9Vo{qAlyj^jjf^Mf$7S~^zUi$prD`> z)vRl%=wJ~IXK$|deEat8sC4)grsC9G3oL53)qrv^`(`0T#Ao(4O_7PpUx+PV2qI@ZluImsaz-K(BGehemYW{Q`rIpPb-%U9b% z_`xfbhC`;qRqVN7UR1C|@W6mTtgUx9p$GDPy8u z)}a}iTvVk0lJ?CT8k)<@Xw{PZ9z$A2S~#!nZtLW*B`4=~CygS*Gbu06 z=H|~fzQ^N&JR1fT8f~p@vNIR2g4>pIe*X0L4>oplDe$+>)%3W;UFllDAi@%s&QO&L zO+NjS+ztWcl$F)(k5iz9kpNG}r<1T2zbH0R4FruG}d2&eFXv^4!EM-?6x*>q`X28L0GyYqdQD0@Kx zAxv*vSU5SeONXn2)=>Hio}-!!Tp^kI7UI`Ay4x+JE!;Y9`YxxX<$qlu9B=dJ>u)&h z%thzFG1U|D(4Jpv{qpJHX4(`F&1Q>$knl0{>B+v!-E7V9@DEs*EOLk+m!fKP6#q{( ziwVhkP0m+O0<${BCQRbO-4C+jv()LI=~KKAd+Bl1^X}V|ByCo+0rm=Z_ZL^c@;^c3 zBw^dzS7*CWO~YA2Bj;LEtD<5miuCkt7;J5N_XAXPAnDFsSlA3!evg{C`10mk44g3a z7)Nf%>KbuRFF3@0Yb&}xj=iq-r-qg5&)2wC4q@vIIv$|9vRY1`7#o|_?Kc?d>TuhnVE{>P9PNaf^Y^~87)JBK;?PIU00YC63f z7ua|*oH_7{lCrA{-Bo{kS(#3?+s+`dyRxW~=*`K>fnJ-U;)qoYo6FYmz^_wYfNSQX zM;JG>A#84UK(fv%1!bU(K0E}s>BM1YF=%z|6d?rxDzHm9(G_sRTooD>S3>1ly5w!+nQLn_K( zX=%yd{Y%iNyK%B0S?u7TY=}cT3eEu%5t+2;q8Cgm6VMrfrOvbA%Xrs4s*Iq zCrk`a%aU^QX2NUS+OGFWc%5G})AY!tDi&X_u`^#3>7@tg(I#{YjCh zbL;)d7I!D%5uor_7DTF z=LMgXPe}6zZV1q zNdn|YfY;95g97}QCyV57Z*OlEhlfF}>5oG7_xJakVzat&p3s@ZGwdErrJ?x%qNtV5 zAX0vd_Ro*50o|^%-3e_sEsd+3!NFqYi69wq@yR+)Y`{w{yl1v7TL7Tfu)hz4TEXUc zxOgcj7#Wj@$)rePS&V+M%>j6yo0{%fWd#IWPsZ-%l1`U_?Z=lV{cnN z2pQdm?y}!%dwWKL#Y%E|^n|pa?jkBSm9LL4h;L*Q*j!}St){lO^E{bN<%A8DPIkDR zM#snf+{MgFVo>X5k{IJ==<=5{-iJ@E?(bW(gw?WE)YNd>ZOlK%#bpO?tW9PZtNdPrzQAf-IT;&J1RMCxpK;nxo8+=F;zn|7}E z%hw8v|GcWJtHD!I2_g=PEN{d}jEpi(wqaJ)dbUvLOaK$(b?=ctQ}NDa;sQAB2!!_e zg@uev6Ag4z*g-+lCM#s(8^gh#+Aq2(7e{lCQ1_ecjss`b`)bnb^X+f1*tenMbNW>? z6Gc!~^YNX>*ssRN#w{<8j_d2&068V4a+i?>&&ysN1SH>xsHl9Mrnm-XTBiNM1a6S- zioe(-C1JBy*3b|}R*H!36`J8m=5upA-ySo+YZblYy9puUyQ!*LU>hG)R8gt4gpK=> zy3EdxyB;jISn60NHXSr}k-Fu8Kscp%v1TKh@AxDf%+Q;~vD0@8G`g8|*1I^J?Tyvt z*Pk>vE&Om_H#X7YG2Q`D<1fDlFZEtpvi7*Jj_@}}{qb461#XuyPsP%$lXcB$B`CO~ z(jy1hL~(!0@bYkTb2Am*{nnNMUyaSmb=3mDd5@h`e0==*?p%0cA}UZ)QUYr2uIr0< zJkVP3Pae#lK&@_1cWN9QD!>q^rMUa#bmG4DbXHUQLJ5 ziVxIjlVm6&P+(a=_2h}R`emu>7dtTQS@5{d2eHpwiGiE?hsT+(br+itSExoOCae{# z_I$|h8N>xcU3+uAnwih67!n|Hv(&iiDymC4bh#+WhvL-v#(9)_R$0 z$J|a3J3F;qI+TZ)nE^SQ?a|NsXOTE=P`5p^i_24LI!1W1<1g27$m9V0MFE4fATosF z_YfwKM&$LHh|%7SrB&<+lra?9a#&C8e7@ zGL169bm|t^(l)N$;5s&@y=Yn1XuyIAc3CgIz~Z#Lt1EeW81!R#IXUVB`5Y1s6a1xm z$qf_TuV0dOc6CG3ot{2mRKT?|t^?91{rZ()AhCC_^j6!dcS?JC1)IeAAdD}zj%h91 z{)HAck{mXpyaV43f0M)%?yQ-}8wmIf1Aa?nRb}NA(o*8VXOmZwXOmXp(pz)JVB1?D z_6eMb45lFLEC`~YJW>+|y8p(nQA zn+wD;%$Z+voRxGXDZRKp_spADA^ekXt4(nx*Fwt<=dUv?>Eoz!u1l2&A>q3_TE6GM zXtoYQct|G=p9&rd*HRo|KV}jyn%@rEe3$}|A!Sf`b^5v!M_9Ox7|oa2pUhRbjl#@4 zFK`GAlL*VUpWQ%HK;31Y*r~I<9S3Lik)U`oKU0zOT3%6QAc4cwTXa4%me{O$d%KV% zz7UcgJ7|zX+>ak;0#0*au<-2m>@I439b`yiVq)CZ7ap7)T~YPXY|Sw)ZZ<$hMz+Z5 zG|z7GoA@Diz{iR3yrG^KL=p;>3YoIthU#QDK*NfWKB=zA#&|W8MK$9xtx`#YXmjmI zK;X!?i_~*RukIG^UKkk_6*iW44l63)(fR;lXO#WhIaO#?eWizty?x)@IX$nUj#FB( z?$IiZcPOP)(l#*>i~AJ=*l4Y+9N8KCM-b_ML>T@(-_@v~uTbMsL0&#gC4bTW%>%iR z(gEXs-C1LY3U+PFMl~M&%KOzT)pW}NVz&h9zS#!Yv`!$myc))TN3x*R0V7%1^oxfK zmFQP64C79CUwh-TVKjbztUdq1(lT-@XLBMb^8yxBqo?D%R3^dNgb9=Tl*c={; zQptkFkxD8_Q8Ewj4$O)16DmwSJ$0!Y*=?lRSK6IM&>|;a+@nTzz#Bc7765>yWQ~H9 z{T~XnO+iL*=kgGLykXpuDT_`ylQPfy#vsyqYh)50NBakC{o}{V3KAc)!{N_JI!Q@? zREGx*r0WYRFJG3$!yu5?B2_pIz)17Mguo$|4XXl{uZSdbd2E)yWsyET;oHK{dxAp^ zUiBYvjo{CK=7y=91W25ybc^wvL%K=7RUIV4Fc!w4hi!MCBUYv-F~J2L-f_1E(@lx zG?uVmPXlrHx90-HAtQ z#F~e~9t00~|K-!zCSCcJ%j5fV&wC=fIkjg+;I{WET5VUCv2_{Gf$qt@+&Dv6uDp4B;VBurmmT1hBmas%jqkOKypdSiX)fJM?}KzDZ)-|_;93C&!IUzf@A7}T zQatxIOnyX=X3dVu;ySE&tq@HZ?#t-4til-?PoNb3__XWNc>wub!XMXi94P-_P2T>s z{-Yk30SniXxa#8NtDa91$-lFai;&{dFqkJ4{ZKVCr4aG^UVJAiW8>xTB~6Ye#g87j50w-ACvX4;LGGSm>lO4Qe|4qw$5&H_5K0Z*L>3=#JnQHWkkxpW&>j z;3Y&yR|HP>UOJ+#nNpB*VAGhVSPIkHje5Euss5S50>AYsN4t?47%KOMi!_&}XE-~( z;3rK5G=02wN#KeVgv+sX{a?;W^&dg`4usnIU*=JMdz}xJ=+Uq67T>zUlmFHEEQBAl zFAu43zX3TQZQ@mb5j6r?z4#br!B4nz(j*m6jIiMw$*(NYSib8i(kb-Dy8| zWD?7zq>t$l_vJQYF0kwj9o!NBc^xLXl-4XSxTZt62@FPd_zN0mvKwa}YZ}*&heU9* zCipnWsiQqnWqvr6GWE#hH!au^3Wv=+ue=p5#l*gnVcJ;f_=qYU>Z2d^*y0-aXRDkN zgTDE}Q`M(CF+~}hwYLtZB2z|qjS_81C_yP!Ana{2!GJ0eC&$}Mq!auV(7?zk`<%!i z12dm8r0eSKGaV?vJu;aNho)|GDF1n6g0X#amc60Kbc#QVtCx+jl-qe{Q0T~-XHjrG z6(tS#wb}f>a6HY+qE9I~^R7uAN06e!)YGb`Vrp{%8Igy#6N(}{c1jnP^-GTlwSpF% znk7qw31?8azG2&8w&En{jjk||VxE20Oeqye`KbQ=yH4T1TDK|?BpbZk5l0OGJVlg5 z$UyZ{O};#{--g~}46?#X(e^J<-x!OoqDs<2K)3RsD$9=|NR=nDJ8VO_+TX8!ZSOaW zCRvi}^nX(&Z;b{4>Lagyi?3K^F{TN`>}xz1+{TlF9Tf(zPfv5o{Z*7hR0nhU7!*7L z%YNiQYy6WT`Ptl1xS3T}v7o|CWeT#5PRSGY)3~lD4s>Q3VX%}=y&2F4Jm3?ud3hpH zJEmXDt<~izLXv?nwcVED)iu;~GG>J1$!^hNL4mkHEwTmefBz{A<8txQurP9OcgO2w zw}U5kRKZdq)3mr=ROCNs*W{ZS*{T23rwsQGRnEOBq(a=?CR$1rVWpM4+(3YAM&->C zs%EG53HB4V{YNwB2qzSGi3JyWIxeHA46}cpA>-Qlc0uQ0R>+8H-20MEE7S`j%@YHR z!}WjPL_(+8Rp#Hrc&I`qrPwo;sgte!5s^QBvz(M(>h~MPOYPyR8gRBh@ z>Rz_SpdMD>%sf+zfBN^?hFT#*TZV>rvhhd{Az3ttc|)bcv4-4w=zH%X&!4O~rbAV* z-)F^OqMJJK_J1`IDzFeq%&goN^@78og%6BP8fpl1;rl0`+d*ic3FHtHPS^;UnJc)j zIQb33AqS%(Y0U~U5XT@v^=NTqLMo0UgX2qpvkO520FLFlc#4F z5v7AH?SJPLEFPuF=0Rjwc%OR*m;+uPp9}jD9x9wjL1VGVOVX*Ig7b|xjh%(nG&vs) zRndO9lKD0jY(?6?Mi|&eR1EV_1uI*cSoJC2|mMFC) z+h0aDrvAK)njIeS1`s5ve`5HOX|Y&|^N3 zd^DI(EE)J!*JY>w zgTr%>EmfBOK3`L`v?4fQkG?gih+5&Lyd*NY|0YJNd>R}z)f&W$2A!NtC7PO4F_8QU zo@X#`hn^xh`CyNV@XAV#ys;<$L@o2tfI!+*Fvp)! zd{X|EymFg+gacWVNGKB9hha8J6*wT%%l+BQ%w~a;Ff6!Q8ChX3h1L+Qt=c6@-nMto zxP_5P#iRrJ9|0Go&*ES5{AN;2sdM-6<#yV$kdw2{8}t8bvF8X9#WCd!#Ucq%&*f2D zam03RFz3JI;iXzgNHS?}6SdnubD~aTu&VjW^fA{gL;a<7HXrGD`j2On)Ap025J8yo zZsXS{>%}v$fl)ZG9<@K$y?X$iX%7wnexb!V=13q7IklF?6?MNvNZI-H%uUHi)UjZu zctxT@`a>(}NjMh!1vkCc^8lFKt33_XkQj=Yy9`MYv`*s}8yUt-qdZpZ6rcqWck(J_ z;AQBYuu=1a1B!2X0F3y+Pd1X~X{}`xl2hEyuV}oBFuz-?wyTi84JVra1}FE88Xf~X zBVSW^E!9;!T*fD}eB+&(GC;c!K&`T>Mrb+B+dF;8fBDn}f~!M6hxYH<^Q6S#AvAt{ zM_p(6XeKe;)+t1;P**{S?Sp}fRuoIPzStvS6}8-T-O}$U3ia(N^b(?v;}I~+A{hXn zeLjWO+M5td&t!KBJ9|2cGM?uA^$8NcacDeh9BeBN%yPEClF2$-E>n8qX4< zL6LuAAp&+&89H|536U7`O!Y&!p(&=u!H23GO{(2|q+o2~cQL zz#3tm&-OA z(8oN?Kj`QEa)QIi`Ra5@fjm3_`W8%mm;D26-HxvNa9g#t>NRmGhR1%SJrC}9>Fxu~qL zp<5`FXO??d3!nF837NnE{=+E7H{QuQAtSg(pnvhYLgnpLxPq&4rh9uwr$#WqNv)m* z-MtsF{Uh-=Az7Wr^r>&6=_4;Fbi||K2L7TcP@H$9 + + \ No newline at end of file diff --git a/en/2023.4/assets/discord_notification.png b/en/2023.4/assets/discord_notification.png new file mode 100644 index 0000000000000000000000000000000000000000..05a7705d7f599de625de31782344f98d93ccb17b GIT binary patch literal 48861 zcmbrl1#I2i+vV#pGcz>}Q^QORLmOslnDK-;4MUrxX_(V6r(tGhW@cti=DhDWpGNo2 zXzpLSlC5J|hO_Ox?PvYga`;z8DHKFPL;wI#WTZbU0{}!Wc)o;(1h4s0_1OUYgWw>o zBrSs7;mC;PjW^(gDLL5BSp04<8%d>3a68Au+sDjJI_ zivE1>5Ibh#1nK|2(rlg@Kt6zUF8p4l;NjY0lgirQbzhk|W3;qP2_)?Z&2kZm{=_2< z7V`)f^EjlAuD09uQr4J;R zHh6y=_dCbn4I5CysQc>`qTGJ&`bOgjquQ-o1;z!K9C&halFROeoK^!1bnz_ zBnduf(G#)=7x=&hVeCV6=qf+jDGK;+?_cx8(R<#qr9;WXUd8`yL?|VAk2|#gWCyQu zN|JG5*ZckcIV8(Ef^lTT{?X-x@k#|rn+=sfW30OVNFD%C)ZaYj{?h~RxU39+wII-g z=S+&ZN08~pXp%LLt%IE$**%S2H|Q6})n}~ek5^Jh2*tW+wnnyBzn%K4aYV?nSY=Zw zC8-*e0p+t2)X)H$qxqIevYt0$dW0yH(Oi@79FDDT;;KJ#{56s1xVz@_dboeH;cB)! z%WBa;Fq{^GsZ0Q}g8pZVZ&-`Tr%lDI3{m!&{q42cf1&(03EMvM4a`T6cgVv=S1YCY z?1&tKvciSh(>7*tMrGpl3oMF?1_To{O4v|D%kHu+AdM;aVy1C?owBOCxvs}Mphb1& zB%Tl(P@7s3vV$vpGU}q%Jxx4D=x~lvlp3V7X3`;nCH(V=U`o4qdW?=bF^nBHf+lG= zHx$oXTS62-zfVgZnejt5l{{dWy{U*#+cfr0Gnl1B*z~((E*@8`V)B?308*&1`a3=! zgrfogR&*}`E#w;78BSvl(tM71lK%lFkajK^#UCaKF83pzcUZBSoZ$n&D618 z$$YlP+DtMnPh_{_HHz!?ZDvBPk%aa&`W$`mF7NVsf05`_C6N&PQdH$aWBJ2q%82po z`my2npWRM|SN)Y1S=CGtzJRt)_Eqv_(>>pZyz_QPWE9Z005kYRgbCZ#cdnWI!?8lbl^E?!& zk|1sYUj?>U22J23+}l@gj#ly_v}JnDDRhrt41O?|CUT}n?V+j8G7meMAn0=db1NUT zd=#SqXB}4_ZU!1#gz-AsK;+f}jKflVQnu&1(57mOiVa3Yp*0PjxttcyqL!-&)P0X& z`|vwwAnT&bjiz@5cj7Nff*OhyrstUi&*Qn`GWLx|&Q>H8!M8wRH?^ZZ;TXtLe}zVa zT^!#84R?;EYqV57J5W8^X1_LT@S5B zx_aWTAV{)PoLc#wCp{Gn_1qHx@Nw39d^``sNT>F4ovwUJQOBD2=<&z{0bH27U(!u@ z7RmEY;GMgaasE;81Xx#XXqjS5Y_BnP!U>z!j>*1NOFLS9c2bNd1wIK~$kp71v;KIf z2^HgQxg`FHiWjWdDVji5&(l~AWzm)@dlz218O4HlZ5%iH%k<8kfWX-eAMBYm{iAQgWevZ!p@~Oq)i(7-)DS;rmAeDJ>)W2AyuzU+-ff*$P*tJR zTsdo*J|5@zER$i8j>tkN-gNz`z^uSR(hJEJ9-re*#l=YnW}S@Q`GVPxOeE^jHrmU8r@GXV%XH znXUA<+zMX<-8Xpl!hV2<$X4RIturi__C_ES>t{=4B{xE$l+XZFphN6$ofQE1QYirU zBAx~m=4uJl>?H>1V_m3q}9@8?C%8M2$={DQttbVfI1s%J91E6AkJqx$;`FJ6@ zm{zDt+V)KLMNPazT&Q-=iyHzxarh6cQC9PO8ldYRtHID)s1qT8m$Am9$FhXI72xvQA+dfPNLAYj z+pPOz^Ai9#mVAz43d;-Upw38Y7oz7Tb+rQ8++qaCGW1i+@!|(fAU6J?;B91vao=QT0FgR|lI{Xp&OQA2_ z?0a-1;bGlg*om_6i@&TS_{z(9X-h8wZQ(N^w6)z$+&tUu;tkfkt<%6z={9Iu-6IA` zkpUM27OSrV+&je;me|-JM4WL$SdjL78?*{059IL^$2$XTn6c`$o6ee|aQ|f7R11CS zcLFNWNBdp>=#$#uSCEONY2;jK9W+r=IZsgs$Mjv^+pb_0*^TgVgy8Soh$ zSgO>piO(pOvYBn^V4c?z5@^tJIhjGi+GCx+U@{VdMEinTIKHQ$)lE$Gpt;!Oi$;T) zK(ybJAhzy;7Xp|lFb^X_TLi(tH7W;;j;c3G1QA-+iOU62`+Fsnm-*ZpQvmhMT2FjRw#&(68 z`MKC0TT{Kk%NeSzfO{4e9wuYX_bD1Aw>g0x$Iyku>_``ffSWk5bb zSKL746*KCENlZg3O7Lxp5l_EqgvNwb#UUx&w=Kn5p0qD30Q+O6w;rHBbp50ueUv_dmmAou}sPSQ8M8z-beiA+2&(KcG zNGb`PG~zYaX!JmWvAo)Q!#Fkbl{aj&rMwFC*xGhUny*t%Hu+L%Fp?b?DI>}=O|y5s zvW2MCS}Nn&#!6j!9bcaNx{bE4CT&_RN0al>`YZY&AE`m5s1IE3J5U z_TNvOhIZ`1EO+i?dyQ?ChstZP_Qb>jTSz9=?2~r*{_z?+8EJ9!GAgzbgOh4PdfmF(!?ww9^y)(^cQLd4~r#Yx=CQK0{Ektz=YH;K}cub$cA ztSN1(%eL)@<>A2fMz|FCb>7jcajs8E{$AF2EV9LuU(PfnEjMln+s4Xg#*9VVGLW7s zUa-L(F@=_nz5Y;UO}|+@?;LqyD^nRPWg`N(c89i;C9Ui&+IH`x(0jA84FX(a?Sk?ByCEk?2Ulq zda0)4w#N^j&PvT*0$GKnOiT7+?>4fCc{1caXYioU=#Ml@f^SHROTm9gGuR z2!J&cpctn7vg%i?40SXQ<%EZ19xZ3@W7IZsyMM5=dn*j-FQc#lK(8>%4&IIHwQ@KT z7s?4Kme7czvzmXB)Lp}=&6kD)z$Hy301Pzw#FFh^MpQJ_;OSCrMDiRzTmI1s1e7vPiDd z9a(>sHFre3ZQtQ9^T<3offxLzrs97}=^hg$5jVZb$K<1oCpHXN5}@N;523;OA?B2! z#W7uXEy4(BwOad(i3^Z#x&g@kGljJuE>^wMB92y9`;-TW<`eRzmV=;_=ka8gZ`{Ct zjj==t^Iosp4p(B30DDiL=L_8wD7zH{7jF|nf^1x-kfNZB;bITM{}3;*5*XW@5Hae~ zq>|DcK05mwFT!QcC|CYh$nF~j!mCKAlV8KJSL~ZE`LXOUoG<#O#8x$HPk#V8Uy`#> z)If=X{p=wEQe_{4<^;|OY$uu)$#JB{$M1bOhkCzvA6nNQ@yA6tJTm`0=atDa-ohbe zGs-Cn6u@=7y^s=r-X{rXMym-Kh}@-<;=>Qv1}=*G(}L~nNAC9_X(2X3L-Lo6L~G_v zzqXt?w7xkNoNe(g%b##{pdkE?~nP- zX=(22#ySIQ(dqScldFUZ@(Sj0*T^{w)a6ZEKejnB2-u9OC``%rBIK}**xE`y-64D+ z9O(_B*1Nv+NNxOwTJ=L%P|?7PYXvqUK?8ym`NH05$L_i+$*$zh#iH4i@_^qrG@;T^ zZSkGpE0;5A7xvjb^Dwo)G89#jg1)>OGml{+Qlr^D+f%7Yh9M56XExe0=I)ftAN%uv z9ruNO+v4tJ#&A04s9$@RCOyUYD}v1-?te&#cweE3nfRUq(u45drAJ1-K>KGx_jG#a zXZ_eUtsI}MHtp@?mw|Y#OS~D%@M*h#PxV-9IuH|oh~6tuEWc>I44AUld0=Quiu`ZNPW?S@#oPqHr{3#p8$E?nCsfz0(+%+K^B+`sIihTpR9p$E2E z*|s7T)(knIfFfJEC(70vIQFlriRT730_cgWRyr@d?BETU#Gdr3uPFpUmvtX(FW%ti z69FK1y2vay7TVwI$a%TW_ceaM$+&9??Ou9hby}sMLhimSat%~Og8%>v0Yu*x4eSTG zMwVMMGFs2_a-QwCX6rg-h z$7^rYTxOv5u;c`QS?5s2&~YVNd$%iiEF^vYgdCT$OgEXf#MeKvR^k<%$`r`f#4 zzn?Z<{t-rXg6S57`e@0p?PZVG?sJF>pOO`iA{bBTBOse^)wt`uYO6Of0QQEfc0s})yhau@tK7O zqKYClzBl|Jl!UobJYCFPT(=51dl|ceU)SjL9;Y1_Oj^AV@QH~!B|CcZ(QjGeY9-6P z@-y?+rR*R`&TdBL31M4 zqLGd?Rv%(Y1Z^#hS>}5aw?9j z;YSs&U$HTRuEc%h**wPPzwek61UbK{=-!+z1ZkA&(IB!oeAIKwa-0Z)_HdbxGQ+EV}J+zb%Y5HjPU`hN(&qf2!}e6i}@-Z_@}zIptvAtCYG)(??@?_Szp*iJv%lfSFtwP)!UE<2#cLvaCG zhFo1jA&EJO1zP!{C37byH2*#xBnaz|u?MsHOsa}^qDq|4+G(ukJn_84?5uXXB9ms8 zGuK2f_syNpCcy>qece^28M`J@O7y7pL%qsF*=^NuHZJ-_ihqmihYtY8R)yoH~sXw)^48^q!%ZTdT?pHPKXj{v?aa5Mm9}F74hKW^qQRF zEJ9)>m$6mxmCd-V9wZ%a*8^3LQ)3fACXv%pJCkbqbNp2!7Sw#YCVPkQZ)v%hEB+EA z&rP(v%skJzMry2+724#m9^N}VPn!0QDU)*k!f;;*9aU>!;;2z7fG$GKWcXIZ{60*v z(!XGxok@ar^m3`Qx;ZOl{oXT0tE&WmE^GGNRtH5SInjmo!?($pAZu3E_>j5cT>k5_ z+&Gbaq(sIDhaNz>8|CoQXE!|^>!fyK{L_9iM8~(5RdtP&zY?Dog5UjszVe{5DwO0A}MSn*k-6!`UYR@xuZ6)8Yx-9%X;eUY=chf+ncL}_AyY?@bd z^i?J@0+B_XAz-`4K~M7Tu48fz8S7_$`MB&0U6lXAuhS|@-+79+0rtP_*!l-+-E>M~ z5E9f(42gA+2Kc|-Pp%z0r{P`e_>b8L@L1)*dhH73qN?reC7d9KNi3vqqCInLBCWIXu zn}B*x1ikRk`}R1uFxm>_cV=vbnzX&WBhG%j3Pz8PANZ6kY1#JEpqc}Qdvh5S|1UU- z=77RY<4MGcCOoP?Oulr=SQ%1UGjqXThg23=Vxy5x0huy+^@K#+-ARuk*Q4Ihu<#_klcx(op|M(??Q?lh{HC;<8r!nbOoA}@fqW* zBqv({z$&t9tiLR48s{^aoysnk_eo1_t*$XGW7i*~&db)d(Zjy)tkK+c+!-!)O`iw$ zLBjKR*oFL|Q0w@SSoIBGvg739^4B7LHIw>RlELUG-c`O7cE!1_(mB+%%)wpEELfGY zWN5K&LfFYoFS-dtNY$@zO&&2>`zN3I2%oM~7?? zz2CPA>Gg|$v=Flj0-hg4kXk7622<`H-?{7**PG15z5MyE^ zhE%kQw7zk6b8>tfxfbT)M{?GwqoMd-F}6yOIOjjEpNP4AN%Q3YckoRpS5!~WwN{^T z7C75HpSKR(YBS2Npl|10&gzB{J?3UHp>N~W$ptS*O6Yd12(84U@4?;{OexNxGk1o2 zX?U#(FD()CMs=bI?<8~Fw9dDArn~z5x5&i`JkZsU*N6iZ7WRGU^WL+&N4Us&9mr%S zpkh9c&{oNV0ucRSvP4LqvK_uR?e)-AA!O~Vf=873r+ofI;zVWRz#RS)`>o&{t*E1B zz%Z6f`_3^2_SzBgA}uo75Lh|OIDz>rd+$eVkOdc#q0=&G9syU6m^s~k#R9e-i)*SF z&^UB-tCG7M;P&8;nGXUm=JOu*{yPg0J;;{1Q|}7+3)oTjVYAY2mz0*At<^_=wLv`gtyE}`>3YHFemYF)FUn zDVV?A)?5ywaj3K;LFkLN|4Rx8{#~xvcwn$|hI0RV+`rA8U$31t)x{8n39GQ`cPnI& zRmbb(pr*LqjporOxG23(jz12;YWa&B;=!keF(Qr~oDuj(%KPI~7Km9uW*MQ1IB&g- zm&oJ7xXj};jdl4HoDTEBjVdI7`}^0mnNIAF@i1VAB)+JI!Tg+&{!a=tsd)iJ<0H2GSFO)mX`?&^Op*N}o1Yk&sfCT)!;O_wHIl@Y)CCW&^ z#9!2*E9O81b#f&OzwkrHa9S)QAS*4I01vRj0UHjM_l?M?(Dxa~!_fX0j56*TDH}|F zzGseqZbO4V0DyEQ@S+%p3m`6f?6iz*R(GhDC~1QaDvJUE`7sjqKHP<6EY`Rn!z9nx zlUau^9-%4H6em5ipVr=AbF$j0# z!h>TX3D|emG*mu(Rxmq<4{r##xJ!BT{hrF$dL7fV6{e*qq8iC?>oZ*aP_b!fHCyN# z2tmKTwN!uey6OB)9nJ~6C^B)s@M0oYjJtoW5g&|6n)pwFkQ&EJ3AfTgT(Rq)_m~a_ zeq^u*DMJkJw&|E;!qe-gpv@J>{%O8>ABT49vjhLWO|Z>q1ueph8>l_*mq)0n0GlHb)h5n0>{Ii=EB<)Lpvg;qbL|1)}+0>WIoDtx2|~8f6q# zn6(-UQ#seL{|zBH#)wL#0GXJ-(L?ckWBqE)QB85ygL|P%*8;9AA>Aw z3;$slGx70aIaE+QDb+g*Jy*{p)ptWVLyd56KcPrb9_NY*iT&8Tn6gi7yTDK<^JR9T{t$f!LDUG9U}oaSG{yJ_6&zt9{ub!hX3!-4$7g*$Ec3)AKn@hLWxG}cn2nOuUqy* z8R3)i#xzst*`QeWsPkZ=xm?eXuKKrZnM4W!l=18LivHbM5 zX5R(gLpY`=zM!pMuU3YiDi?jzojhZ0ANfPSo!t}h2KD$9}Ro1h@$x@D9Nko%x$A8fgdH{HcyuYf=&Mn8wl))U%luydN zSPRz~O3W{pRZcz}JMZZ$O>L*y)ICq^V>f$Ut=aC%ut=VhH~nHz>oF=ed&GiLm%4Jw zV$YNuuxceGIwq98R+9c{&h(7$2^8J5-vi2;IwyR7llfkEvym~f)wH~7YZt;!W*y53 z4z4jSX(ta^A}21Xv~d@gw9|MLX&OP`b5S^*LY}0i6HT)W)P5;^RRq01m##;krE7po z`-MAD2!5*IcR-d{G=f*|My&30A3tN!4C2=BV~T>8T&v&; zOZ7Whred!M!!W-~)0iQ|-3!yOK~da~8%NWLQ2vzN#iB)xXrRNF{oCWZ$AX~0>nsY= z{H~pcw`ck(MT7wUkpth#Ywq=eLs@hb3c8aH!;BfC*Kf4?+vU0d2I709lb1T5 zx8*vmFdz4>*ciUYpakEi+-1}{f})j<+sxAB0$H8{n5?D;#e`D@g|Ep)qzq_bEnw1b zXb7K+J2lK@Uq)S+8(42cs7Cq@gXsLj5Px;j_R{=0@tZ2^o2J{@u)yH>Yw>grUn{r0 z$IIP~e;pAn*9`HBm*vKNd6oGHu~(Tp83Yhz8Bb7O&MRe$KyoZykW#Dh=6Lm2b=7|E z2i|O~jD7psF-hld{`zHtaz^?>SnW@<9&f~F?PFXhzu^46+TOWpIjWWbfn~-XeSyU_ zhS^oOz^9c)E-D~t_K*27ached1kiD6k0h%$=NmG|v*O4D<3DP8jO_#ojMSGGNvGM7 zKJZA8hCN!XmA87$mFw5qc)Vo8-sL>pKi%Dy;>Sao#aIn5yX8F$-D}^Se2u*Zg8jRY z(qs~Uai&1{cbUF(uUh6QIl1>QEi}}^9W5Xk;E*Y(6OY^BSIS6b z64&|rUN)S1C1E^YnNX3WcQ%-_uADv<&DKqmkl3!JS{(A}&f3cUZe~eD6mh$C;(Bkk z?bycl`1)P@G6;k{T39+z(!g(2Jz5|D6q07kHB`5TG7tD!(7ig1_aO(J!y!rDdR0-+ zzMfAnm5h~?wvX&J9}O3<4BVOltR=mVoMswm`W0K^ohxg0E++S$j{K&1@wuI>9kdjy zx34IjzXDqqqoGKJXJoP?C1u&^Gc4wbJgzcq&Sx!Hl~)P6fl@e<)Kd!^d$JSLn!Y9$ zByVLQg%sc6i8t}iwHo?m$bn2r2*O^Qv$h=<#~r85{;RXM;!Dre-)yn3m@7v$Kk(b$ z{M1xUK)5x9_Afn2Tp6fSXVPN9UawSR#;&;%yAN|jQGrP5C2UhZTa|U}!naJ&s&9b{ z1Cvubd28cCH z1L3q5H>s)PoJAD?5W7JE?}g{4i0r_!;MW8BdR#&(WD~B0yBp}}TD%m=;Lq~!NH{>D zuF-|+c0K2-*=s{G(HjMUkbSUlA|l|DK`Sf$J7M^_oZf40h%O?UY(sGU~ zV?{&YflkyhFKBX7(B&1mE>~iio?%RgLjBGto*Ua$hJ>uu^S+^u8qBDRTOIzje+)4nu;)@sWYv`hdA&{Mn|u`vS-ra#?89KhW0?sI+qq8FQ%H_ zoX#bZ8Ho|6l3Znttg}XWe}F2A4JvrYO3}0Ts!cq{GW* z3t~}Tk-D;ygT*o?+Hc#~{t;70<+I7tZ06O>l==Y2b#z!*jcP}EOrc1&LZyt=K<{F z?_YxdD_7vV!9oZLa9Lj7{d#N~aqFF{&=!u8O1>L+4wqhP{Vd{3!(0k;##1k2M~wV7 zGiD0ilV3_`P`l|}J7kg?y6_`A$|cEKRTn=qpKcYCz#tzEFhs03rG^Ptmq<{fnG@QG z4HJkX>TTW{$E+(F7tG&Jg-nsiz-Xtq{sc zJzWBA41BIt$0{kFj(ZwTSPtIb?KnoPb&BMj+|}e|GxF3adEIE$7kK~nrPOikj(BLb za>had7oILgtsl%m;5#4&rYla&bgC2e&PU|%vIC4oaU8SidBH9|tK*$iPG8Jq67mis z{sl1Gy*Un)HBOZM5DtMqW@V>7;{#{7)pP}_Q>qMfgKO9WM5W`_G@3@q3b!7YMw6BX zY6C~qI8uq8e1>+8G*h^&0hD1cvg)xL5=7p3#5!9z8F3np98t@>?Y`4@72wGGh-qhj z+CTY-ye$)D@0EPx#p4&^CXJ&nPkwI5JC+ zE$rc4NUvWl;&%P0xU2F%e@7H`epJ=Rg2;+BBN=GSV?${}`Y3T4o*rT8OIbOlO=(+j zYOoLKKTn0ZEin?ZQy8}6sf$mzCCR}ME>^@4W@@{9tbALsH5P47w~sE%*V&JY0Y z%e`x9>fl5Xib(DTeo>)7LP83)I=im8ahJgRw642+X`TOJbIxBMiL#I2hGu+6Z{u?_ z*VTp+^3s!yf+4HPqg6$ zt2bRE2;y@-dkDZe^>`RcLFwB|K-$SoiopH_U+%XS|7EVm0^Z3=)Gd1zwAgjJ6T zkQ5KG7s$4=KSIN6;d-pSsU+b2C`c%>fZQx~6lAZ1;`8vZlmSm#43m1sc^uTAn5hMC zi5d}DiTgpfrg20cG{feqRd~P6`nW$ z8P&qlHZ`$aACs2J*~Vo)v3?K`l3S5=5)hv=HL<@kM)z?m2SMOZCbjIOUo_Y>%J>S{ z?T$s!_mZS5m%$l2HoX5SX8c(KAq^T0SCIgH*O{;(J@wQ&s=69lxe?%X2!2yx9T`;+ zp-LIYpCt#~-OI`vSkn9c+bLcnBIxx8M0Oj&D zuJfc1x~zBpc|L&z*(%MBKC6$3fE3qe4mxaG;4hio<^h{=tgAqX(EX0HWz@Utn zIl9GD1&#eNBi`|FoHQdhpisS#3HYoa!L-TUy=&)-i1&J_6(Oq~0wcZR^9 zH&p)o{w5({$X}lf31m$Rt}q`6p8A9AiWR>wIxUkVP(=dqZ?@qY#QXOdKw3e zz<05!!tnIIepqCE!3$ot5-8q?nciV%Rrdp>>{R1(U1VUxZ8hB$=w57i$l;$eLCU7+ zaQVb*zO)ZcMTkL!onK*1^*{pv1r9fQO{-7h<2A_*4xZBdZ8F#QSX5~bN!hoXUmG1H zyv)u-+6aVx5s66gs(T*`X5F|BI|dnE#KlfGFt9r~^n~@VNQ$O^&-yitnGafV>X&hB zzR-!;z4kkfa@Cwg+BB@lLnBIr!(HyJLkqPL1A@SW1F=7sQy}Xv>ZE27Lj`E0%i#*Z z+h|7;>;CvA;!~C4LAwJ8UX>OeZjFfjLD>7wanFql$vx(OMpZQaiK;NzR`bWUqeFjv z2w*$T+ObJ_$aWRYv}AxSb&g_x1EG2+z8}U?%H)cfw-yQ@f-P2KMQj~MwQOa#4f_Yg ztXxrTxrGy5nM_NEw3yD}2X}j=hj*M1#Ut4_h5tpSl>U#rNI9g#lp~)}M)A2g|BN4H za#ob0xEBPQomm05!d(2BsmXkKB4!$6c2H?@F!{6LuDi3xlJb~}j>;w` zE@SPtD@9^36iMxTRv*i9vb?fWu4%4PK@A18A1pqFHY5?JDwh*cMTDzonBc2 zIMfuy!|_gjhFN1zt-ZW(W@qDu63TnG;E?ZB6)r(A#E75-(o|5aJ2+M^QvY^zs6ypDI7yX= zP4)IC!b(dbiZV8wD#4zGQzsqKSpMy$zABWO(iQ3B2Dhy7aWyEq*Aas>{VY5aQhiYQfL?3LJA2y|1p!$N*4?tnohKX7qUffCE}<0fu+ z*;54dP8zlX^M2{T!P#)Uu%qol6Jtr420dtI#-r2kKA;-qId6^{ErgB9sv8pxt~dO_ zXKD>}sTzHW?5(CoPh{amHq;YF<+zgb#r5Vy0nvUi<9-Rt4f}O-{km%Zbd4-;6|eP+ z>pNmz^OILxVoyXf0GxQ}VA9BL$MvAyqk1&e(9%;*BnKFYyfKbS3wZRxvqo|I!0PF* zU1imHE?l8~s|;FEL@~|FZHhEAN<`~9f0~3xz8@%K$h9k9p<#y#zE@mPZsP!{#SO{3 z`OocvUlJe1L9b)6DiJZo?#bj^>q_q0aSc@wBtQc`=5 zUmp2wo`t)QM#GJi!0%GOjmu@`c8gu++apkb&wtTMUam~9ayRWj>Q)i+QJ}@d77hVNNmchZeylys#~`8c zS17Uj90;#rzzZH$A3f_q(5`?DH*GERylG#ZpN@L3KdVBHnrQpZ&n_PlqG6E|1Sbi= z&wUImDD!E1wZHhJ7qQ!+Oz-_qWFxM2v|bf^O(P^)ab(c|5YjgU3-!*)418UE7do!r z65m$?vy(cXbc#N+apym%>T6#;9mXw_=a2!jvb-D^qJ4pmwoaeV8AmUrR%HYo-+pEa zX+0mU?_U6b;llTUTBJ&QS3CK3@_AHzlD~i8?oDv#SIAlZy>kK`3*jwyzgsfI z_Y}8ln2RNu>eVTVTN;1KxS@kdj6oildSc) z_m8pa^?O9kG}10~hjkz0dulUf_H{?@ji^C1T;MUf}9~(N%m!b>(w)S{>*5= zYV;t;?omR&Jpr>zP6L~8KNo@a`k1u%QgWsHq%=4VsNV1L3nt)2W3BH@8yF$a^4ZL}Zzf6?IbX3o#)2keiDO?t(W1$({cV}^;iE+qKfPj0H5x=b1SD(B&o3Zs$bmk!Uv^T`0l#73p7*&6S$Ude`gapMlN3zyucuov5 z)BD{8iM%N_++TnC^x^h*GQDPlslwm6rCPU}Fil7r)P9ZiPbxGOVYkg*-pfalLBV60 zY!Hlqao`-I$uGOe_=BQA34ZnOf6D~Rs`b!;ujzPYSMRqybFEal6{ktU?g`|#BN0-8BfqpU-v_uhCp_CxbV(X15H2 zuV$52_s(xRU6PPG)U7l+TE{CZrj^w-mK`t^Gg(47CKl2es7*hDAjn$`FIpDTs2>DAh#U`ER!P0PC8Ixo+$o zSu}~is5i%YKAIfpK0Qmwt$Y+7FasSf@=39rE6p*kfODF)pTB`kIybdbxeQ;w(~KV1 zYdH2J1u}VcLt7bwiangyoiqM$MAm`3HMNAZNSy@Da*%#X!+!2q%3+BcwZBA3Low&J zO+e_rCUMA*gKIMre^IN{|EB@N1!eI^V~IxoodM`5v^+Brzb$KXt_G=Sn~(^*oy!UMQvb?JO*b}BcI>^HVeMlEhaT6{;DMa_A0uqlnLs|N$F&<9PhDw2G>gvhH@ z7InL22j%DX@cGy$Lv|csBw&5-Q(39OHEYubBb4<0-T?nxj(aCc46;O_1Ocb6c+HMqM&aCZ;x1ZdpdUFPI}&&)gTGjl&R zH8WMz{?MF;(_OuL@4c?IewU9L?P$fB&RVBmUHow34U@a^X@M%r@4C@AmO89-vs1P8 z(dT>7MJuL_k+=h6d?u&uPi`4JC>b4R#lPt-kzszptlUg&A4lCZhYP5OvHzA)dC_gC=ZmnUW(LhnjRj|13zdqzS_@9*QT<{(vrrbl5U) zIxLVmPNZXww@lHF5!ab%Pf@yZBn}PeI)t?h1Gl|(^Y0dFA#=%yE<16<{8=CI)zBv2^`UhV*B~Pv*apJTFk2ZVD4DLN;aOp^_HWl_KQ7CDc zbsDlb<{xLXu`F5qNs~pdqFLOh*_t}?QeG9|q*Ko9vA=~<)>mY;*^YL}Kd=4n3*jE6 zMLR3?%eB@*-FkZYUO0B#hK4l*5}iYOHYtPx?Hl*SROS{`;UipSbWYx&_{@S}6SHw| zErG-m#Z>RxZ-qG=GET>)+}1$oU7t*t*~Z@ZMIs#5F{)Rq7c6eT*JR5T1(A#z7yDPN z{Clf$ZBS;@lJ6hBYFoa%uV-9VXhe(9)8Ipf0e~OT(iNwC7>^H2%YH~(X>uEBMG5Mq z-()-yJ3jM`$E%_oy|lw*NWp!pc={+>=O;ITHttTj@?XneZe{Y#oi+#h`yDbJm-=kD z3s~WcF+HAj#y;>%Buuljh2ME0e;S%MW_WtvJ-7jGXIT{x3{rn~?e9*z^|TgsIhCyt zXux7+Z)sr}J>k)4`B=vDnXR7h}?AV(x#WLhAjV@YN`* zib}TgktO=AE|j2pSMi{Xlscnpluo}av)R%L$r?UgHWvLE56L5xiL{l8v9|qa5`#el zkN*$f$~zO(?ZHfFl&V<~&)3IuPKK81v3NdI$*(z<>ps7fp(?FlzM5(iOL7_-cPi7} z5wESCmd!GSkYR?IJrJnX4jEHjz%iHRs02DIN@hOq)@`)H>u)D}4{E1XcV&Em1GrQO zKYft+3V!Tnozu_u_6KPn6l9#o_O6YhaAs`3cYCH40hmqi9I|fcU%1?1fUr3J=}08R zjB9+b))xC;IUO)doWk7)$Vk3f7_D3$TYYxcK!Qu+->dA z@g@@6oe7lZ8rC^6te#X^fAXSqoNh$ocWQd4gS%DNld-jxs25XvXF7X@J1pr_Xmofu zpfCr_^6uG`RH5aMBw3^p*{n*YY;7$Sgn~<|A`4gOc9UAl_tA_q?fx(Kjw@jkcrd-0 zjnY}`!PqTQ6=kxEbOb*IlRcMq*G96yFRBzn6%PBI#z`oj{X_P-Ayqj?WS}o7pSk$= zH}{(R3cEjLh+fuH4^N=q?2Zp&Cgki`qJf39G41MAAqVBd$4Gg4)f75AIP1j#`7X7eK)oh@T`jE-ew zc;#mx?->-8_9 zi(~wU3oB`FyEc9#2rTeJb(f) zhF_)KTf64rW6*0+>f(z~9HQ1K*Vk-XOVZE6ge9#`$}b$K|5QA_jJnaNaFCw43T>#N z4sX*po19w{ShjBWdR{HBLWtBXLoq^NPUBar{32BAbBA;H?CIJ5Y+gDN9~Ocn!+ncr z_CB1J-%uM~7Uz&3qwO)2H?$K4+pmhIVz%xNXen6wSzTe?JDCfECh)!S$j&T&sqQPz z1=d~r3i?=7i$lo^@nN!3yP)BaO#h z6DA|Dq>ykpSlG(0OAl6eUaA#rkNe;d^kjP-sm5M?!|q5++s{9C%=zI_fpQ5?p@{$3 z^*bZNnCCDQxL1mBu7FfZ8{vA-n4$z$TF$l?es`v*drCa_H#o-gk3n;NRhGB+O~MkH zgkWy`sHAlu6rsL#-hKytOw;}Gmxjw_0^fEsIe>(VA7S`4T;M8bzEWu1&25gIa(8?e zOE60Bv@~4%f&0nt-79?^*?+N@{x8PBWQTlri^UJ12=?4^IVn*mEVC#Z8xgW+hNiwK z4?fDL?d4k7dESF^h1&X9HTdVQ#_|&gkFut3y__8S>1EAv&VtIi2Xt&*d6*=8_~$9o zm`Gb#R99OG8I8H_LDaS%UVY~&AkLQg@$S`m>SytP5K;eTbZ^LYA8|kjp50&Cj-bJR zU$EBx@>F3d=dE2yi;L^@@~l5H;#yw|oW}o1IMN3%pg74((Du4B`aM+9B8gYJ!Zy!D zK=N!igl`Y#{l-FGo^RTE&wuoKtT4^1+82t8HhX(p9njb1{r#=a%W}JQPT5%Lja6L@fAF>saWAKPWMe@%*XCV7%wI;=kN>CRjpj$9 zIey~@l^?Ugrgg~ac$js~6+ZpHXuZ~C8@z_rW9sS8y4+qO-pLGpSz_FIucv+H`s4F_ zT`!<3XH2U}Cbxji|JpKarxYMj-uj5rVc2YAleiLpGG)x{9#<4e8!|1D(BmrShNbal zhi&9ml4|MPbHO6NTzBNK42!5jY)xcyQ#kTR@g|;x%-dEH*i={BhjmYf{tw$HMjzvI zVVP=N*Rlq$4Eyyve|H@+q>JOfEZ?s?0fD*``-Pc%$2{=HOr}bXlc_N7Ar;V4k2AeO z8KW!|Qf_HgMcIcvl`UV(%y89f{dC2Tk8PZ^_SYSS@mJwRevA&k7lH{F{6#SNI!F%87n|QfS4UBKxjK&K@QCQr@pdI5~CNHw! z11;aH6LD3=K8q%T1ek6u23fxFvgJ%AJT4uT912>0?UFe?&{Ir79^o52j&TG4E4J!;$>ETm1r98%%b*6P%{pFb+p zsb*34aH)H<_sm>1R}sa6qEIra%EJa@5WdX>#ex|>C+VFf5pHpcooU0bS|`L2VM;l` zF_j6QI%MBxyJ~m(gYWhG0ULt0Tv7rwpkA@raJr{H2Z50`u)hlFw+Tm^5dqdp+lr2TJP@bPt8 z?sT(C<+B=O6PrH{6M=M{5ox(;s{MA2J+^7c6{B8g&d!C|{)fSO^F?7l7dVo=u)jns zRN*`&_{A;I)U;qXcwh6D7r21D^Nv4S_mxZRGWtX@I=EkAWVGUm#T<(YG*~YsNmNyR zSv$rBzWmfIE#yylI{>@zQ;Acw8fmIqHR9Q9=LW74iZcBS2Q#E~J>7B{^+JX_qc#Xx z;Ehm*{D9N#6k+cmMU@@spO+neKbBq)?VyY@wyPzc1&w_~&k_2)ol}ctHoM6qtENTy zhrK7ZW!SkrY^Dv4rA%=8U_tHd&y0$RsBurau3qnEviG|cLVxDmnet$*!HwA%Nh*lJ z^2b;KF<1#IE)3?o7@MV>+`s954wmxoL<*f-QUj}9ryoCS>UdE$n+HLt%W6vFfiRNj zJI~|mLIJ@OP9cpRS2I+QU1OiYGF>0uNdG`ZJBee>l|;w+=Ww^$RN`w5KuQ(4HT#yu z%jS}&e6C;4m(KsOBVFG!vi~WojH*bs7RIM7EBEqHrYDFPxmm47X@^P?7T8T0i;LI# z7{Zlu@-bg+Jaz^gQpW`^2g zBGW01(ZjbI8(|b^An*#cdCogJ2kvWl_w7=UbC1{XWJ}XyPY)8p=c39>k1{kC_~wnC zLgCeISfzT)cyyrK#$uc(i3FSXhh_0tRO&$zo?lg={TxWV3JFamwLw>1Cw^UbsoA$( z){k!?W9QKHJ8ke?U+UB;G(BRo%_$Z z$%;O%Dyb{?N$IOL4bDG_^%sdL_2`lLT(K>mtTS&6%!VaKw18B!1OD6Zp4{ujEBo^TZRvC zzkBt_U||*%#BOhNf3lD%WcI?j?GKWzoYw0*kCzv_P1^|2I$ld`LZJL3xAkQW2pboDKM9~w{e|ma04(VtVbu=aPUn9H0W|zW4xFbGLW+8rqgU<`qhs2`7Pih;{G{2 z=w0{m-qoHPLy{yp5J73qn@_a&{;gBGGpkbkPslYoK2F_DMw>cv?9Qa;=j^*_@w*m| zERIhJpUNho4=U0>L)mV23sbeZc(w#NUCnRVhmUy5`(Z@cbTwBiT#EV`uGO8tcZ!DZ zNG7Uv;mPMrvfV5md0WNvU=w%qwZgVDW=g+Hw+oje99p+9iZJFaz@jpdZPb~XmvK;| zo`$$_I?oY>`f<`&Vn;1uV*N6QZ}w!Wk`RhUs5{GY4+kNc?Z7W;J53IG2U*$BSgI*= zdSYI?jjwe7iWy%i>s!}Srd*Y$lu}3ZQg8OIjOS)BR%>!+c5@DeFKrOdPwEyI4yOW# zfr^lKV}&N!G6meR(XMh54KAQ8hK8dMJLY_>CM0~H&%4FHCr-h(tWBzxpf!V~&3>_A zAdh^2IE!h!+u|dRa`XoVFMypZ0nNX2+qRL2pJz6RKj|WZFIvLW#*c-%xxKNyq@Pc# z>61$F^lKkx+jljnJ81oZ-th&X@bqN&>BLajFTO*o!HNgllpxf(B6#rlj2gaKsYiV#H;u#?-{-T|n$0n`>z^#t1v1>rYg9bmG~}PYr^6V1)OFW>gf)9~ zKVYr9`Ftq#tBQd;OeKYDBi19kKTR(yZG6e1l!`OSu_k*ZrGu}Zus7xcv5UAqn^l}$ zDw39H(rAMrV+35P4LJyF7+@f;&L$cveh+C4!vR1{Ls>DGJ;#$z&5LCi0nMOhZ}aJ& z$1*JR$|KUqvD1%4OJ>EC&OK&8DI%Yhq`ZPw&OK2_cGC|So}^mO>il2o2R_x1{9D+v z^Q7%Y3a)F_3mE#Uhb1!=6)i<|>mttzM9Ac+_>>~fp!KB7@`v`*+{&L@7tZY0_PLe7 zquz%HRKd+)+`oIW2+I&&c)(@%xymJ9+H19Xv;0Ao|4sgO_Rq3B&?oBBC)TFENl?4i zEfH!rU3Zo49GG0%GhMtaC1KwaXtU+8$v7^?93!)<=b(B#7)c z?}{Ek)I#rre0ggZYIGG>#mtH5yF}bw*NJC-E&i3or?H*7>gA3PiT}P((w6Jf0Gv1R7@kN|UF_CQ|QR%SiPf0NC)0Mcg?&_GMz(p3IhTN)g-WE0Z08 z4@F2w`Uor%yZ*sWmcr?QFy2GN*NT_rJIaaQmoALe?@65u>VB)GHn1*P4pbVMn<_>F z7Hrl7=&n}3{Lq!C_P=3W6|65Y0-;Ph2X9foV7}QSX`&e}D`R;QfLmM7NNi$?LZ8dK zGod%L#Zm5eTB*xwyl zruKvvNi)w~Hi*7*6=ocDy1DcSE&JF4N2;aTagoGGOgdW7cl`5h2a7AbyA(+M!+oc( zhCk&@LOu80*rOvjARBLkDLG?<+n$&F$h=We9Pkl0qJ5W(rzStK-6(#B`8mCzc?l@J z(|;dpG2g0h7J&`;Knt8f0S!riH+sLxf;f|E2ckK@Lyx;~Tq}h2-lVRh;qx{${^G3Y za0CGPbnG`8LgPJ;=*#DwwuF5R?R(J0%oIc3@>8SJLee)An!}VW`)#fi2~Jn4_gUbY zMY_ZLf*bwvQ;XH^fk@oFj9a4}H0@cbkdaWn7i02tg$t)mT<CPNBU461Bmx4}hXFREEH}^K+nAD0z5z@y9rU$mWj6t%M!lFC3Za85u zjPS4Z#VRo%ED$DOHLO#BGuJbS&3CLDs8Q$U88a;$gy=KN6!5!Ov3$lHevdyyzYiji zy0W2ckBZ9;8sRz?>Mbdb69FH8egMQLUU3;OeyJ$WZwV{x)bra zSZxjEm6dqi_b;2+TS?Y(!@P1bpZ2&WmXShHZ~U`atKU+Koisq}&;f~Fq_2=yEx^za zc^Ipme{K1(Gy$UFm+h))yMs9?ko$aM&Ag9{b%w@&u0W3hXq^O+;n}wM!cHRER#z8-VDFPs`{b45fai;>wm+9l$ zfad|xaxMLw;gjJ-7{t_S{VKy9derM%6?bIt9$h|p2B8PPHyD$=vrS0GH%-(*2 z6)HOoIH#rFJUL$Eu{JWa^%qi?Z&iF3w&WZ{VfpO0!>XRt(3Evkt@<~#s&dNi*&F(- zg2^@LWNVFU!jmC9ZFt1vk>4a}#JeBG+AD84?qNI3AJhfUJ0kop_4B64QRT)&2`YXB z%(#g3>wv1;>YX#FrZ2BuL()g_g6Z;qKS~xQvwb2n&f2jq2;q_*t5-Rc;G<~|K3jPc zK9lw{buWlwUDI4ry~rwzgNyIIab=KwL6+-jnQ4v5IFIc(w}+&PLW6tn`n(4>VH30u z=^S+j|M>mJP(L0Yb@tO)Muhk%CGy99{O8plbXfK^cflD(KUZ?4l8LanVjoh}XU3e{ z0HDSslJJ|ULHmQ5Vw8lz+Fd6)3QIC+c;9K*d|Sn7s-*3xTY5Bj0C>q32oLSr*x2}B zaxqB06*>`lJnpftniaD;riFD!S5^vm%G|1*#mp_9%*Ue&oNWY@udJM(?pMOPI-#!M z9H_4U#5kBYVb70`rWrMTzHIFD-KbGR#t|Rq{IK~ci}SI!u5e1>c|GC(f&U@`_Sy^xt>@noOhJPyd&{-l8?dodX+Q8$~zj_30UvB@LU4^WDPk{Od;&qVRSixEAI>qmO z7Q#(dtl9KswpG4xYMm)xMcA@gw>mu)X0D&(#oi!7Z}i&nGPNP+ceytxrQjR}eTO-S z)Pc!t4k7J^(Za)p->f(v-pr{L@d;kjp$ICxDSh#?Xt=UhvK90_48-eDOHCvkh5vL+)apV@3n>d+?aNW>?mc;6_3>XwOwE%X54CVVkV>-_APz8eBLP!-TV@Q^?o z{qF3CO<{qTQu(ovU-g{wZk2q@TTJY{^mCClEZ~+a7STT2hjAEXg26 zP+C0^*5-hC4tyrYo3N|wT)n&(0|5JP2H{N^*COejigT;3VWAXsyrx$Zt}igZNe)mz z>sP(DS<{R&Mbq4Z%k}CuO59C9GTfI@HFwrhPp-bp*K&)&x9bEjguFHtKH_CtXS|U? z1Ekqp?!mm7XTRV9A5IEJ-!sc4xve4Mp?bHXQ9O#ak(`%+v*PE=a;-@XY#VI4u*Z&X zG!am^6&bt@7mA^Hgi$?a7+x$^z(X)fo>&PaA!Yj? z2`=Yx8{UAX+`Z5Hqgon2VNYM8u{mQETt9ce7N*ZAU9NqvWd_?+QUR|2j#|GaUyh@t zO7_DXB4z4CN94$_@!MKuf3bg$sRk1OQBo^CM{%gj7&+2mJW*$n>2MB`YBL+wOW!(Q zSU0^V^qa9C^sIPd(n+rNaNg>HC7d(wnOd~lM7B8i=RVu-$v955P$<@ujA?0vxMZoL zOUV$o$bBAGp7(6~PQ%OU2ma6j1h;frbp=)(SpF#D8e5(7PYF;qBAiW>dSBl^j?d7A>>6&{Td0`)mUqR*3=J7ti~e7!Yf473#|u2&Lpyb#HS5;ct1c&I=v;D>lLg3iC|$}psU_`bh5wcT6yze}K_ay|G;#zKhbiIf^E(bX`Hgoqga7 z#=SCXSadxH9Wwh8{R;a8IWY-ij%CAEd)nh8W|#=-SeA9Y^oeVdT&?%mY@!{2QY0kv zf^9o&4~(A1rJyrJX!@ ziau?mWcxjwUZllUbT?Y(`go)Epkq~^lb7L~7c5Eti{(X?boIT)K*)FaejUH2nuiZ9 z)eKqaawue5@ZMBe5k75l*%%g}lG)@>;v6vJ?2fbxe`lBufLe0<+QAdNkNm_gP4a;V z$iU(9?sJdDYbEyMf!@z`wXeMB$7;QKWv_Jxmsm?vV?eZ4zFx>#p?y2oJ_2;Se%0zS zyV$W&uK#9dJC3i~T@hXf^SReK^28AcePHbuwsgI-aKm8 zde3x<-qo=MkfN>48XGNM&j5<9pB_GU;aL1>#)n}xGv)KP`PQ~@kK~YUdiLFh0>y_# zGl67TX@OCoYx&g^)3*pbfTnMeR;l+Bff0=7-J02jf}U=JC0Osj;+8URn8<^@YAnyk#6kKqn4-(9><}iHGhALmdOpo@}R@aBa4> zSKX=Z=Uo>R5&A)z!6D6U@#-?(7Mt>d4AspDIw(j6cnc*?wJn8(r%NBPeg#}?FxkgI zi)!E``mi`IE3Zd0KeiqK1w?wvD|3DFp@{*up0xu}V~!I~Qcp z=|t5;bbN2P?Z#3QW>n(SI2+_|n^fhiCPSfbhcSg6XFoWNQNY@q1uw^`{7Yvb2L=A| zGpd5v4_&d+sj)LE=_BD+uFL3NHD@3A={jVDe~daVdr8rc+?P+lT4W8R89xp5equbh zPHTLpka}Px8GjQ~V?9PyY*eN2pf6+rBD&-PW~!EyI?o=EjpC606#X36_LE(7>`5VV z-d(J&`YP;MTY6|J|C%HxIOca}seWzPyM3Bz8TpLV$!W&Wg?Oq)7HhY1(4{ zO!1#<-Yl!y>-D`UN~egB(4p8@9kF;BXuRF}9;;+2c@uT2M#oEh#e?5%0JW?5)iRQ? zxm>siMmK!h2{ta2ayo&A-%VcnSAYaYb}OUJ8w#t)xAJs(6^rVHJM>>AUm0nX=pQ3@ zQ|#rr?);oLl`|B}Fv~A@G|Y;Q;8=>`W(F;rHYCQc_C>bg zsZ`Yc^V&HFuIe;DJE{|fxrdvkdP`zK0x_qK397SvSAik!1%I zR&VaGcJ;adX=fxC7_T#_y1`OP%uTd$iec#Z;pE5u4Pank9|8cI4-v9gtQ!3sO4wnyRwn=sHpU;v4YHluHLwTa^&JHD~8eTUL@ zhppqrHrg&`RWr}jGD;J_xJmwPV*6CUAm+M|&htr-l-J}4^%R5gVjln;OzP;QOBpJ= zTcB?}TE-$KY9j!Ejh42R*vF_Hk& zMAC81`Iif>SYarLQeJtE-AVtou_yeIRPA398023hF#NwHuvX`zodwh~G@tH`>Q7k3 zqilMO=V|~~ZPP{O=htB5C{|ij73emz(47D<`fz^FLdlwP=#ryFDuXF|CR7(P=v{45 zpaU^TA_D6|P)SSTlAC(w#LZA4^T3nuhQ>TR5SGq4h8u0cx5=?In$-54;vA?L<-ZiN zV7tT)ei$>#b@}xHx@+CDdmZ(6SK5KX!jExH$k7<42t_Zw1Bq=;7Pyy#638eLe6W7` zaY+^%u~%y`*ep4oM*Zoafz^m#vi`~+PV!?)cDtOK#K6iGzA>&I)GP-dGFNwb$`KbE zWcGljW82m&lND;zpt>V>_Qo}l2%K~BCY%d;(y~ zM4J~zg6zzsYAPZbc)qG8P_?510tW8Vb{2*ER&J0qZrHG$a3fz$zyZ6bV*UG9E4HTGVNpY%zZv9veUK7O|3|HATFWq?0xdsjN z3sJAdIyo_TS{oYnjLjB~<+FEmr}~uk@QhS1!wO9y;*Hpe9xHe?F})hOCO`Eirc{iN z*N18nQz#_b+@#9?D<%v#6vr#gyYT_<>u)clP_!Tk(vakr^#@;el|OOMb3!{V&iy)vd-qMozJjn*$r+zXORZ^zWA|DQ`@@4;UxP zMP!<$nwd>r>9_cOT)AUmDe|`0Z4TerkDyETE-pd(L?69x96}n4Au@odotgNBsyc)~ z?fs{iJ^z*}xeIPoFqa#<2X@zEnR`Wu;4{Z-GUDw0t(+f00AyW1i@X+CgA?iROp1-M zzg~GpH@XF94>hKJ=jXTTagRb1Srr!qTr!GRW7)$YD=Z)Qs7&=zNdl&e@PEq`awE+N z!?_~RLw{CE^M*i=Ye&162I6evaqw?SyUICFs!bD(y?*=g@ym*A%>CA=e74No7~ksz zdD0w3>i9s(i`Q_IVrj+srw82+($kl?Qfme#f$od~&ei134|STXz$S&2Zg0igA0iLq zl5R>4!Kh{a_K+_j0_@yqKjxTS& zr>b05^*ddfopgXyP~2n;ABl;0UlQi~JT;4(@ZY|H|LPw2Z${LEdP!X0;7=_I8OIgO zi*$m-$i>wi{JSS7$Z2#InvR;N!1qF4Td+yRedH4MM&zG^YV%=-@{^?~SEyK)B#B8Q zV?g)S_7Csfy|7lvyk&)i9b+6;g651#=Eoti3YAY<-cx)0Cf+|k9H|DN%ypoDe{9(I z)Hv6kFVT0Xv{da&6%rpeH4s&ZW|y{Jc4~@S+rJT%{S`EqNL>5EB&&0QgqXrU66Y)D zGhL5K!Z%-i){Htm3KK_)&QvP?1BHJ10U1U_GkWF}nj5Lm0pX%oi@Lh}vQX&ep20#t zMYEi6&NhvvFH9!BeWHez?(Ry6s5h^$@-`@Vt+gNP>J=A7W%bz+M#pahFI^Z zYQynHaM=#BQ=T&9bF*4TV89!P!=0A6MX%jATBE)saJGteNg)QUk*L>P8^-~cab{(I zGgmS(DuKnyM(o$eh$#Qu-xuGQW6pdkF&Q|UbP5!QU)Z`12=)(<<+v~Y zqasHr-t!uqgnq2SK7l|9T%}-|OK}-M(1p{6P--22i$ycTkLL&@nD6-hj(+d9F22)& z8isPSF;RqDLv?;lRKcX$0U9>9yg64mbsU-4w{v$i*`R$c7T1PWcJyzc7N+ND z3G3bS&hV5>dC|k9of(!r$6vf4k7`HjBxHpg3u0jWra6{Y>cOkF%%JzedrmxB&K(^L zEPWBOq*$M@Vtch;;?QuDIeP-CSdY%$%Sv}=+J1As3$B@VhK1=ilX4W^k3|{JtJ|%# zB2*MM5R;OfLJXN&Xjw?UjMAZ`jABYrddBgahxDdLKi~HusWS9;9ts}?iLhE2^%Tt> z=P0+~g>5|?VNDjAY6L{7mYQ>TG7?;y@=nO2*19^iNDXGiua~SC$3;!sAePWMF$wwBwk`fGQI4z00v%~dork0o60?u|iw_C>_ak0YKi={ndDIRLu z$+O*rJn(ooShfKqB_iAzy$F4^Hbv`t5@`^$~%cky}HpN72CAI(r)c*lsFSA~BiZ zT(;*|BPOY_wAZ`IT@*Otk>xahdGB7e!j#2~dxIKSJ~OqjAmVJd&oyFVw3oU&UQ?(_ zJjJO9G;vdS<3^C(1fiR{Npo4LR; z9372@)g!*z4o=NS^ zvhFfWNy)o3oL1a68Y1-fvc9AwEKEb75)C{YBu|^r&5ba0kj#ML;O53#T(~FVg7T@{ zPl2{S#rd>|VE{c0$1PKAN&U_*N4|4uN54yPbT`_vemad8RiZ@R(#gqbUkL%wTuG6R zPIz2Wld5P7ID?%Pzmnb*phklNkj5#EFIGM2B$|U^&BwHCGM$JE3Gu2ZQXIC)|6q+S zxt2QsnOqMeS4_%^kgBrj0Ipd`tyFqKBR}tzm=d*m1^VXU8hf0S^Jc$FLD7yLG#dF#C@7u&=;dDSeWQade`9x*B@2@soy<< zeoSxmUpKs#UXb(9YwV)oocKV?N$vd8|FJ z`ZHW!UK~ENY@EQ8ukcgwwHUa230vF8?Y z$cs@VVuQ4^H_Y>6cxS~#=jY{Cxwv*_88i4yyQl33_GcGyN9&G%UP`vf4d2@qxKGrC z5f-CXOwi*xDX(AO?X4uUSyJwDJ4U>F1%2wVEc$)F$D!rnfaqIB&uxi_0Ygpg)lt^p zC34;-Qkx{H@%mP;{&bh#bJ$r5z+_+{9Q>ILE{2ds4bn4tMMi?=#Q z1#~_4!+-n-vHpKLo}b#(En>>lN0dl^c*@2h zuhU{V#P!hgnU9S86TkT8v267pbAp;@UTLZB%64x$%_3D~(qh+7420Ov5O{N?u#0#~ z2Hp9UK4vHbisPfWOQ=>h<@G6nBka~VWV`G#f3-&^B2)m*cdc#=RkC1 zfT6P#C_n1b6S>fespEY=?U~5dSb+ZT^pM{$VQ_tuIw-pOLuSc%?r1)8j`w*9m;Tw!hSZ*3OS&)iH>M0JRHz`CZDdP6OH1l$oQe0gJSz5)_guB-0%mtN6z+!l73s~s1E?{q-b*MLLC zuay5$->7bctY1X3X>az!-4^YRb9#zKYbRf6ogfw#ZIF_;z&0EG`x}R(Y<2R7;}X+@ z{{9B=ZyVVOdm5>!2`aY)Ap-xOz|=gaSTAbry~IoQzFt*X=M#5q?CSwfK?9eD*a+Tl zffQw}`HwACnhUzwHc!y*lG?Hjf)bvRIOk(jQiLZT2P@n1CGJe($Rn)hf3)E?+*0IM zpcodnjqyD`ImYWaeAORVY6`Dyh*lR`~Laei~oqn|Se{AkPUjzb^TOfEQ z<&*GUjXQ(^LmRrfaxY6P^YH;PkrJUe7nhTyKcuzuT}3R!gI8hrn5jgD=dkma?W2y8OG&};%=I8b@%BI@G za@E2~jpkBr+v%w8d0a?~->5P^o@aEu+v0RT&&_*ii{H0#=Kq$FZ4Kkpxnv@<^SzFe z(@VxmT4ed7r6{Lbbeo6N@A`!D0|$ImC`|0B-f7JbmgK;~%+?Aq^>Z-p=9V^)Xj2yi zx^=@=$Oo4&nrp>21~Pa>!3F5Aax%rbX{8qH5#etORSlaco!nX>rcTL8Zf$}D_PUCzlQ%8W1I76yP z)^VF-C~gdTyl5uPQt&?)O&G*r+PVxGp^|InEaNx63v0wE->x_H^%McS}f&77jb^LWu37Tu=!XT!3y$z^D2&D^0O>>bPmI zAjZz|^n0{}UO-{lSmS8^I-fq)$L}oyJ(yITELe z%9u+#9T#6C*I6U}fW0k4p^Zz+n|`i&^dgutVOHqj{#AF5zhe+Mb-G$_U=l6in3$_D9zWzc(s9W&l;!-&wG;;`}b56W>rPWNa*Q=*XnI76l?110~z zt*o?0vEmQ}2}Eu{u@W+DLpKhH{8fbM&*X5317%c*fKNO9z*sDIv zlT9JI8s%&|YjRjYfl*ChNivi2R}LvA?+^OSkNIayh{pC4 zD-%`iFZD?GAZQb}ss(c-=knkswFv{3fXFmF;%dcVa&tR&t05&A(+doC!uXWitO?tc zoGV{@>=|6jg=DvX&BGGZ)VpqSM}J7~P`Q>Q z2;Z_KJ&DoZq`sSJ{6@NNpF09Onz7Ejg{r9n$}d^=Z^8KDtG)EZv^*+ffu70N^q|SC zLHisd(Ta+dbctUTzn+tCc;wwb3Jg{CI{Oxz&sb0!bH?5cB0k-;qv1s zWUR%~8$XQcHlqk0qK~Z6*9oTuR>-gem|z-?$F`eO<_C{a?)GG5sk%{|BNe`&wl3w~ zq1qS0zEuJ)-}#$s_#A1uJ++@^8i!8}A|i-zjg1?|gqCV;w)s$7Q$* zOY2FZ-BaGaNZZ%~4CZviD=9f3KSES04TEQFWzmBR#LTz0Ez8`frr#%J8c$Un#;5nv zQx#t)gS^frn&%z@WJN5L08!?WJ1b-3Qc+8Ihy8-ow-JCAVU2g{Df#h>$- ztbszJj5KtDym&PKCbL^ow@I4BH1{Di^#5^5@rZgMk~Z_{A?Hh#hxF0nsFq z4%xCr^aODx)yzZFyo=F{W~2O{g}z5SWOhJGCR2x7)OkB#=B_fBTRwujdG$$91;2Rp9y1!LB8y6^3a>>}QU;z6Y@Yb87gJhD%(4yMky>4`1r?HVDSPGKk--ru z&7@~PjWh3ClI(BWF)jF}4H*;Dr=dyj=iP{QmbFW z&2TF+;LoyyR_O~mYkJJlpkup%aHYZUBdr&1GGYCg7|VdToY14-+6pUoBOg*`P7)1+ zMy27V$rAMhua(C0+x*B#=h;l@pj)Pzvi$bM17~};4urkD6_s|@Q{xryCogB;Hnp9{ z<$Y81II`yrV^YU3Q%8&8?{(q*8l@*uqZv5W>SDiO<-+II{MX%7%axj4(uXqHWF88tqx6j#JgPy(u_$eS0^u+#xaZtt_c#D-Z?!-P=O* zDN=)aiVlyYru-#WkEoRkVR(y;m{Yz40kA)a{KKh25hUDNe>5aofAk;-wqeB8_)2YU zd&}c9?yNcByk?lABT&>^d$ZLr$f2?Z4QKQMdo zx1N+Ppa6aG-o{@mlk8!qp$*bL9285CVc z7Gi$&yZeXghFK$&6D&rr`@`B`(V-O7r`|8$&X$SP3QDY;X{zt4Q;Og5UZs%WH`7!5 zPY?|sIlC+GZTX(0^3A+hw~GxuEE?V~rV`%@y5X&Y9qSJtDppSD#u2>Z+hchiT&mvy z!|b_9FXkz1bD`kEIr=gLrCK4P> z=d7fTJ5%fn=y8Pqm)71gD6VLY7Cb<33k27Mgy0t3A-KDHaM#8)Sn%M%rEv`ojeBr+ zcX#*K_s);0H}mSvOx1MtzdpPA*xvQ+wbr*5GqGQ0t2)$GJOTqdh9A^hU&dk8T z#C;kJ&?moL&pzYkhI%@!*VO;P0a6MpH1^mU>EJ|x-uG1BtNPkgl?pk1ZOqTF@0|cq zK$Uc$(a=hb6W&Q~2jBF47sM^Sr8hG$B2=%$HB`Vmf5z)}o7QPdA5RQGVbDj1!Jiex zZa2-`FDaOlH-{q7o8r+BcnmxGK`*DOo;LxVU9Q37L%%XK3;O3;fA}3fEpOkSm=77z zgkb_v8+g8kZ&a{LbZ{Vh`Q1{6hP87{@4)|*a+l7nQ|?e}81PROr!8-G$ zN>IlHLx0fAJlfuV;=!?KZT*dAvRTH)>6@by?7ka)=ljtsyJWj&GaWm2E4%ycB=7}; znfr<>7A(+lfv+&0ko8~Jp6mKj+m#4?!a3@@sW9qa|)v5V`z2@TTlucAA-Tt!cvL%!(1^H5|UV*!;8)vWv;;Ht#5}ZcRgFgVJ+H{bpS8r$&l;fwRGi5pfE4DP+(VveEQk zG>7RU3}(QR@n=||QKE+^cH=(jGJuJH*1G7M(2G({?&p5hlf~s92-#15uIuC2 zI_rbg0H$}6Afs>kAXmr>=<&i@JlnaKh2?0s-j1vH*LZ_T+Gd>zP6MM%fE;l=P|!S{h?mT+*cqu z(yC#w^L|aseq>F?6>6#RYW1$9=B_`hJc6xaMaPtlEm6j6u4W;9zrwdh*B)Q;^6l;R zl)WO%`V9ZxKFu7P!V%&hGcSzKXIldW#+Bi6V^GpLemQ_5{Bi+Jg_?6GwN%Xh1|TH!_Jna zn9OQlW!251u}Uw+p0x{lE+(Xogwle>*Jv4R40hB<&7Ga8+pHzGTgJ=}K47S`u1tfT zh}u=`#UW;LtZskw+Uw0X;1b{}{c0x5Y9_>)?H*ubh!4_Go9$Ge*JGa9`)6rAbNuYk zZPEobS4QsJkT%Nn-QZa!(kk)G6Q%jp<546lWAIvG7{+V(+6raEK{YCY-~ zSkGz7lWfNdFDy*<_B*WG!0?95hxYiS^i8JPB85q~!q} z4%RW7i@@D157*mrOk4yYb&S=buD#Qcnax1ZYrC2Vh}7=A$W3)2=ah!>Q9ffK4;tmw zH=J?{8@V0j0u&`JleyQt>{LqoEp7?LSv#dLMOK4M3I@&IsYlO|hZxo{0D;_11EKi} zM+}HpU2lL-Nl&n5HN9|ZX4c|um&QuaTG@_l)XlG?{`v5Bl|`n1Rq&5hs_jXap&)rc zxrO-VYpa_Je^-(VHQ1k2kjs4R>ajZoV!sln1-rjUYzhUeQGi@-!6NrA^fwOX6p(8rPgr2xhLo_uuW4YxF=_eB1SC z;#q|_E+Os8XXx)!uw}8c=I)(6M`~tABJ?@^sB(7BML@QY(|bY=DQ?dZ15<9dO}(UKh}vk`4u@@-b_ zM@b==>V6UINv`QMkh$Dy4S(rSg!Gpzm3%%oouTHSiq;06Cxe9K0NO5Z+`%?Y)a zZ;B)jj1b!`RBPLNCy3QhxO-j8J2_J5Hh0`S@b)DX$mY|#mJ!-UWMl)RKSq15$v^ru zk##Ug4ZK}MAUWs2_`9Nd`HiEeY+@pV=mXB0$;v;4V@JMH7rmKEMPaQsA+#M?EjCr8 zegNM~F`%Ger^<)d%1EBCovN#`0wByBS+t6JRL_t}qjM*`2`@K;|Nhu{2D|iHzP}aJz^1+M;=&-ooSfA8O ze(NBuHA|Q)Bd7x8t|qk&j!iAvTFe(-tq6{yZMbo;NR#oTYoZ>P_U>$Wi%$jK0)H!* zK^CkHbn1*tXcjcRC|K&oFGahkZwoW;sh!jw%U+q;-dtRlD(ce?T=JKc_u;e39=>9S ztdXq>w43WVDj=SyFn6&S8fgbHbo5mol3%K@ZhQK>dT}oAA!=!nHZR!=%+FNS?+00a zvG~r77}Q$Xq+|uj0$VFIe11L^XiYO)1HCSt2`+p(udEqp-v6V(;oEJxrE$~yx!86? zOr_B_#@1zfmcO3)+lU~p?a)3Ak07ozmz~Ex36bH}hOO3UQo1%m2c3Y52u)O;_SuA~ zbII)X*}Rw-<9xa}9J5v8-c)=98*se5_NI969)3x1v&2J16BxNto6CXMlw?%bHOGLsaf?Z#k zx@pnWLVWwdZ0y|C=AO|~>=75v(1L}?@LdO)fU}(RW;%d4g!@=wPH%)nHHPH2_3C0_ z+V#ma0*jddt>cVt1CKx(EvXXSM&X%X$Jdda_DisNILa6fZf5NB^vUONzp=KrOi-A5 z>yrxz7^ihsjvGQhj>|in6nTdAnfHKKIHy3v3$^YJk*DEzJAlC4d!9BMq~W<-*hRIS zE=UvDR4^d+6{MnKFT?WWf3BOs<4-;i^^m@&C?IWQgKPL-izg+oIv;3jVyYjn8Rqy?PHD9Q-cY4v{@XpR z`tduikXNY*b;7#iEl>yP%Fu3G0(So-~$ zAzs?KJ^ZEyu><)*TAyZ(c-H9MN|MEz;}FbNy!;RoiS9a-*v@o54o_4q>W&;P)o{W# z6D@oLYx;_5x0pvaFf3K8!XU13(?}@%=hPo~w|eA(b9I+h+=JC$N;N7@;djH#tx72S zEfuYP#~S_e&@22`IViuEKkTpv3%K<}CWsW~n<^g+3`fb}+b-9z+KCpb$dKQrQ;=%l z;Jjuc^+8gJNP)jWHM4y&JM1NO4oJTgq>!*Uq%cnTeWSM6-A)Rca{Dg*j^o(@Ok9tp zyA*w>f{2+-moME}4Uz3%^<*^@Tg&lEfZ(`F5;GAa$!Oc;JGMk}`5{JIo1=H%k|>|; zZ-I)6=jn~IsP$H>*y1+-(6YwE!AIFXicwVOw!Gv)9+0f zivsU_eZQ;6UwiisY38gV_x^WY_)o|k9VEj{J&tox0uV>yPn-pE2XH~!HgrWcnfkf8VvR@1C1<29*U_G*_z`yg=P}Qt1V$K=$r=is~jh-e0ZROi9fv5QL;M}5f zRi%w$R(8nkXqj-*H#f_{)9EuD21Dm2U8o2Ad&H=9LKDPEr!Qi>?Xhf5D{#H#_Ds8M zQrH4$SsAM1uRm;thyp4cwtnU>DI}~}TD2@?!+B+c;SicJK|zO`7jfitPV#E|z)kOp zhhSU6&Vr1mU*>5}bzb)Oa_h%~%akjPG;Js#KCZ{Brg-(`rJn)Fa>m(K4cUe3*H)I517$Y+jg2(`%9fpW^1n{p<#8Kro ztLhe@l5FAMn@uUYxq?NVc6Ruf?XzT0*%am`C%3M_fnYd9o3|jo11JC9jh`zN@105j zAw6}6SGzHIqciemDCP6y&pS7C=DYxOh=ZkH-mltwAkWLeVq(pWXDi>d#0nJNea=u% zsbMH@S^8n`eJ&K*5LC%4Gz*Nb1;OGGr4e~qd@a{GboUdriw-_#Xnq*M-hc=b(6GU0ghd^(8h^jF8AVr>lgo2rx6%|+Ls6~t_Bn!mzH>k7ff(yj= zBAc<})TQi3xx8XnBZ7?Y+yt0smvSu&v$3eYyK1WsoU8rE&4XCk-rjGkgp;8Rv$7_K zknOy3AWgZW?P%zae~O>blE=@pej`=N;rXY{duRqcK#3Qj16p<)NTJK>d6J-4ca~9O z(A_|)S|2Qwsnn^MD(Hp~Vm9dlm7a2cc3*F|Kf+|g0`=<*v4Is7&O>b~8Wp9f`m{z^otRbK!wdN5Lv|Rs%ukH z2&{v{2gDdm6f$xgKK`t1b@SXLIvAmQ>~7toEzZghP0<k2pZhfnKq9^m-=EIxz1-6n zKK(NP@@tq6_b8#^{LCQf)}`nJzcbQ;i%rwiLC^M&z`%8OrH<|5PJ0fM(e#Yf{=UB? zY3mH^gDn$(_QBrwl`9Dshb7JnbrfcXK5FCFxJ;mrl=w^}|ln%Z$+Y(y)Z5SG& zbH{LXV+t*1{4hp=JDOM(#O2lmh^yPu`$%^V1TA%^f|d03a&G4)8$KIly=0rnCB6BR z_JSKRlWXbsL5cfD*N=M!|1FSg|MwNq*~5y01n-j^1}dqbZUzgL+F)(I(@T;~=FaWD z0kb91j?V5`k0_6xhr0w<_ypab=yRuh)8hU&w8j3kaz{Is&=}!b$!x~GJ$W=q2L-jap zRcTeZeiN&npw_S-hHKmZR$qxU}7D0ZOOM*l!&EYKNyI- zWWqE&wQiFx5tD-}HnZ)@<2*(;Ik8GUN-#?wPqT*ODmj;E7eQnD7J;A8I3ZSH!CQ6^ zTT{TpSLElSL$kwS-K)yu_OhuwV$aub;5cM_jXaRAJuB2EcQV25m7v|x2Tl8=GyYi; zFK3;WzlXbEAplBPfSoEaWdXC%Q{c<#S#X>Zr`W^hYX>=80ICx-9{O^0Z(Ct zfKn-6wbLyYupU~W)yrGURXS^`z1#D-Zr3rcUhJV_ zp44>L(mPjASA!&iGM;=Dt*I}3Dy1g3RsC{r9mhs>mzBSCEbKktd zs2zn724E2*gbC`~Bu#ABW_(~pC%}6q(=*U5r@S*x^Uw!t_;~C-@~`PD-??lWBY$n_ zr=znvYIoqL5ctdQtT?-z8E{0tSuPN(R$>Hz)g8AL7fR0*91tQwlip#IIKIB+@%nS~ ztzC%O^rLjm#*ad+xn;Nb0&5jMDgdxZwV&Y?ht1xB-SV1+>DQTVd$80Gd^g{Ox{e7TQNt&8Q!mfO*?XFDd$6~$*(9v}L{KnseIGn3Fm9LQsMApB zlaX2Pv(h321DwBnj#pOs4O=%fCAh-E(`v!uu;eS$xF%XF6^;itu-0$m~ga|5wG__@S6-d ziM8|jddpuqakMqjwxxSCM*3}pR;k-e3~76*nr#=;&1C5cO^)cQ``8yeI~t8>0dr1O zK2N|AX+EW!RCOvOx8!jfb&if4FUmKKf2Nu7MEwrvAclJE=JS@ya;FTorjaI>NiMNV zDJfi$XUfX6GVC+H46tRa)TgzZtQMTOnZWm7N!x*lDeOwwTwh^PNc&P0qS(jn{MC&G zqKf7$7AjhNg}ht}EC*M;q^dcW5D;uL_&ARwgNqBLV)Wy%ce=qhGl(2^v{NvKID!XM zPtD48AUlDsN9#}$sveZVy9{4eK7oJ1uWF#(^_=QxXMdlFHt#trGQX8<@P(H2XZKAH z77^>YO389_>WSJO%8AsVZDj_-JzTXqkQ*3PN99l zwedRZ5I#8KAekh3Ah)^qepZy)jj9<)nC+8WC7or_sSA<#WAWuM)9z4%K0-y$Aya^5sw*nJf^%)LSDjzO&!@*^}#}AY0B+}Z?V!P-(4j$5zCVQE)KB*bDH>#92 zxrwCne;}T{xeqK60!x)nVY&~UhEuSuE;`r8DdbGNMHu&NjK48~#a#Y}cKCAXNOYUg z1;44Y?wlD3rWXdCt~{>e2hB}LUSz%XGgQx)Y!A{7D>9Ktwe zC^wh&co%N(4DGoepG!(w>C9FVsO8gGeCMx(H-wfc(8b8Bm2k9gbQt%njUc4r5jMCm z7Y_><+rDnlmSYLkK*DVMPjY>#F0$S4K24dj`2;`TrU{(Q1M5-O>uQ6Uz7+TQGig2$ zUXLNOU9GRKB5eA@ZRu7Jyid=D1^eQ&hddGV%ocSvUII~vw-vEqxXOp&(ecl*+k}IFMgGKZ?f2-h zXX<4PY%cMyMQbL8k;ktbCw{Qm)ZyrTWzKyTI8=Akv|lO9Q8BsNoAOE*j)VgrM&6VK zL7}77LF8<2!!x;eP>=q@EgD$Yz$}*KjgBpfO)o5 zd#z#ayX*%gP6NL>S=>Y3bXpZpuh*E84-LaLT1~5wE()W4yz??T)!OI){T#rQsn@pzrkz-G>c+5;aHZdjuFs9*olz;MhYx9R+6_syhh8Ddq!`rdCva=a9%B|JuvbGuUU{TPIY$) zl2jxO);SPzzUbkn!UMHedByrrsN2jrwN^s(ciXx<0~|B$4wJp-FY0w%jR~f@!HQ8D zpSasg)4HYYk2&ao^!@9X3DO1Ah`V56FRcPsR8)4ovh20>n|ENU zKNUWpF5Sdc7i46s4DJndVtlT=rwuhO#FxanJ1e5=#Hfa&O`o+BtAg>dQtof)QT$)^AHo>8 zQ4kCvQk>noL)xu^e6&2KC8Px4jn}%GC1v?~eT^J7f++M;&CdIxYYp%Zd>?WYknjsMY z-yL|!YI#5@2ju7TQ8a{H%=-&#(2YpH%+jZ#2KMJrx8B6L#cy{6?-wzBKu7Rd=p>Cf z3cvoexXDlW?19P=LG$^KSG68*bK`+B-9pWf(aMy1xuc8A~WM zkST|tw_cO;@{9b>eVzOG56N%r)(b3GDy@n88&c_h+F8*be=#=7YTbk>^besw4i48z z$XHSw{l7-ZTOY_@>pvarX*@ItPLNo{dSm^qE!t7~PP|uq zKo}8>uP&?OCe{Dp;VGP}gq3m-L*d2MAU&RR!{)IyzxZ_Vp}Tkq)ru_D{2-#yPDN0M zquXG+Cs^L6hQj;_e?bFnzW~ka60P*wanw1QG=^;Q0IRbE?n;G=PPvcs^w~uCR&FR? zA^e1o(s2Z~YFyaa`?j|BC{NzE0lKb!#XK5hqLssxf2)IZ{xG8Ze13)(hy`%@IxEkH zSK21Hc>F6KLNRgU{)q*C?pDuVIxP4iO;2vLQh45zmEVW>R+savEBrxY|8#>r0gyCLjppG9sBss{`1>vgdV7O{hNnnk`n?`{wLl%p%AdZ9cyoMiHI+@ulqh-4ln{ zpKIT|DRq;tl*wu20`wfArWPy;O`Qo>8_ zl-@4k9%B50n#Y^AQ@eb`1>d(KGd5MK8Ul+IPxED0G zCr5r%xXWE=vmlER)L=0nD?z!UuBWZ4wq#yT&%^zkrEwyok(M-{z@t3&v;m9O`C z&;4;%%M48AeZMd0*aZ3q&n3q$`h#e0c6(|7mE&^n z$N4HBZnveNW1r<}sO0rB5~*4#l6UP=H*E&(7W}zkO6Bw4Il|oadT3yZp@xd>>BXK# z5wP=87P^n21(eS%kE9fi%2cWk?(8{F8cT4m4+s>M@|^H`3Qls%t>}Dh*QRMtB|jTQ zIK|7`o834etjhTW&BIYlP4pZ*-YMoB%fqME0cM(3iu>t!Vq|UI<-0gne&UO_9U-mU zgj*xjKRu5~8GXO00?AC#HJW9JHkbLVt+*mZt|Z{h5=}S4u;%Kk;F86&E%GRYAsuk+zBc*t zrDqz<{WxRiq%6GOFVL)FMYK4mM$7mfhb@v`3zF+=cWAN!X$gxzo*o~)4+k4lE$UQs zIc2umjp(}Rt%f9~jn+l}Y2U56Ycz}r*J=HiMJW~j%(hjaavu5KC|()#0+%kqdE0jVObc^WD}`bhtr`i=IRP)JT#PiHMcQ%B711a3nJ)pv<6>}qBe{{f zVyM|k$1)@3%uu6iem*Pbbj;?B%;=Z5dALT%rAnwu!&&@6`t#@9!m!NsKuFnCTi%^S zCFYR{c0Pts1PWl-m|6F7?PVaR)rJB!{68Vpp9pSFZfnNI#;5R9bfkE4-xET-dsy(e zf*~NPB#~d#9RGK#eV)m>x4SXjD#k7zY^bZ4&SlY4kFs#VBrvdp0u4&#;9w(K4uT^k zQS3+P0AX|0;^3W*7Ei%AybryQn{s%Di!J zjqM5IUw?c@NM5nYN|~b6{PKbdRpk3IxeKSYw)J>8Jkx)!h#gi+WtGC_gP-3k z;+vD_mgLcS3>pHWF_L;2!!y_W?D!JvYSxt_f=&b-qu%b7ND>{2KC2c+CqHZUbE{$D zGE4h7x@C4TZ~s_XmaGH}ZavDg8UlqqWGDAl708FXv<;iUY2fPhQLg+L2;?y?ee>T{ukwUTp|Cl@S(=+tdRm!|hDEHM3ffm2kCDmx zWV_s^`mq^vo(J6p&@>|469N9(!f+vTy^oAqYh)9BHcT{Hp{>BSx_-NOcFwd^? z1O0HaEC}}rEgFNwIQ}sj10bVSyMxIL^e=zsj~6(K)Zy=!*82WQ%tSGn?6vyZy?{%= zN9Nbv_QyN0biOHBM-d-LH3tsxdYflQW?eyCZqoP78baVXBkP}74S)~5NQ>xY077k1 zqrveiHUC$dfDo8YkDnGDCfwMgvbx45NaWrdXo<^Ds^>ny^Iw`I($NMO{onQ>Fb0-- zH&`I8&WC0MU>!1|-)t5@-k_9IWK=A@dj7TZ7a}p3EAfyuvR*HKLd#goU}IZ?Z5q1){Fbqq3Kr z1Pfwwj;0>mG|(F(I)^`Z{q1#xS&qv)&xe;2*)>OO+c8W5(}ge|T~k^?Ll4d*crxdf z^Vn5n+c_nbCuI!YOq7^4VtL;FIt&%iR8160&kYSB8F)b*kn`iub`-A=Z0au)qNVMnb|1`by9yrCUz|NsDUIC7L6s+};j}H=J zBN^gow^w_@K5f@fOK>zgo-$VdA^7u>xgg6^dj$+^pd5Y8J5eM~WstDyOuV0iESbGS zfBTMcPgXQe^gQXGutsc?=~^+}oSK}i9y0iHT5M;zNa-~?+ue7Q)b^NX^|f zmt`xnu23r)G9-D*U`@@s@E7Jb_*s2HU7MC<$JTR407idO_lC#F%3s#AbXDJqlM?c5 z9=GB?B!YzAxa5*+eWVaZB8R=2=9JX`nE>LyeH$idE}IhA{=btk9%k)Z>hv6G#?03l ztdm)zpG>BzcH|2`#7CfG{G1boi7!NGU--@K>KLrHx|W@L$u>_<4Q-g8C-cfwsqHy1uF;WQEgC^ z?-wHI1|??RU4$MOT`xXq{713!YV*HmW-8r2liu{@sV%wrWw43OkGz5S?V(s_MZ&L7 z|Jj&!3V28CFcIrJOm?2cRwQU)vzztKq4Eu5Lqt3n=Y4Fh5`>cuGUmBrFr0Q$d^wpq z3Wd?}`paY zC;wc)Tb0+4&g{GqO8&s5s0Tgled?JHmNNSHH=@%78^LrGDxF z%ZKUyxkhYe&ZCinyRKr0+=VBbBg4TI9`@?Kicg3?4F%~CqDc05?+4l+rl0{kwRh0q z*}8zTLnVYeQINgGub!5CIZR`^?R!fy1x0c#`<@r= zok2NaEa2ngmRTE8SEnHkshSQZ+C!fSAbe!<^>-R*{bwo+EM$7)@JE_T;b~sqSE~cg0=^;EkC{@`W zzB51;{#@5=0w)KrnGIu;9tfs@3ufyL5znl-l5Cnqdfo2%)1Ik;cI~z`@p3}Y8!!zw zcYR+jpPz!uwUB=)YU`!zkd{cW(_0v7yZBmRGq=c>e>I$UQ754$fMrZ`piwM{WG?=) z@&sZe9$yn^HJ9-{c^~oIH_DDN-xx&y=fE2A+W63#-BrYc-AkT_5ksw9U#AEj@Ei^= zCyNtX8e~8@#O%yUClDF3dRT>KOth2YaEID8%c8gxEV#N{9gq(6GC6*(fo&x?P;Ihy zTPO@JYST0*LmZ;qoKhswX>ReWtZz=+n|a4E^q;grc|ke*fLx`574o%}05=f#eevR zXXOpjVw`HGcLrcDESgItDp{Eaw-2QP^j-3Y*|J?y?3CbB|0gheMqss&Pf(v=zSjIE zTAQB|Nx$vKKprH|8(jz;`8?KMO0LA1x-!`Elkdaas8!VAmL=k6yx#nW*(C2|XWarb zptrM*uqC0RBypg6_RQx0gyhFL)Ffs}l`UbsE=tfGe z6h?oiQqPEqEmfVs^p8Y7Yq}X)p@ckoy;5-hIE-aaB;&R}zi?w=Q1XS&C}L0G{ln8M z{Q}bz0Du|e)sAT9Cc397AL`Kj3ofw}Ybwp>K9>uFy)8T{YLNjjeG(y`PY}J3M%!DX z?{jV1(H5`q>LVrqa`mu4iMeA$vV^Nu< zeF$r3tQyWs=7twKZ=u03a2Pp1=JedMf?ExEkFw-_k2u2erwJ zH`jfvM%`RM`u6n;Jf7%9VvS!DVl%&}ZPUCEHpx9NVsz?6Esvv(W@` z2oc~fwscc^iV2JlM&L|oU~p&UANTIQd^fam`c^XACkyo;J*)%i?AD z#-}l`OjxEZm4pd?`}TDhzJwSBx7%d~_pqezW>5z+Py#Z^S%O}9|8V){anTOr|DaUx zE1fOQtzL7RwMhLou&&`Fdj7iX-4c$VgxP0hOz3h5QU|H1;&D?B zb;Cw=lO?C4<|;VNMS+0l4rwqm(G9~RyxTI`nwm;~n}ijYHW@o?fZS&yD2dO0X~FGC z(V?;fLzF*ktTm=}=h$CQP6wmt8ti75-?a1|KWi3Wqu%Rye zp5;uXUHej*1uTD}46onW3!u82E-z1{etyC<-Ddq1yEzM^-MmJ?F>>VT%tTXg`){>G z__d<^rKn;k-PgRV{{=O!o*B`@8vc^n%`k#obU~o{2|%C`Pl+}H{>|5_1rYW|(ufsvz##kq-O)BG#& zXU;qoX6DoqktKOoKigP2&aJdbbCrv~HJYAlN zlr{!sCktz=`1!)FJZ=tNnXmWvdj{-XWuMM+LA+mLQnV?ID$JR@pq1p+$$54g$WC6T za+1y3^2Hti+6=NqW+t*EiT9fyWwWCH`X(?_fBO*$ARvhQDl64 zUK}FMp<{cMz@+dM>8HE?P_viu1ve{!?Om13iMcg!0%u#W~Q=m4@J;256q0kke3!(+~O!~LMvQ%sT&-4=)f|#|Cf4eDD-^x2FKZ* V+b%kpPX|ChQeyI=72gei{tsZlJYoO< literal 0 HcmV?d00001 diff --git a/en/2023.4/assets/freqai_DI.jpg b/en/2023.4/assets/freqai_DI.jpg new file mode 100644 index 0000000000000000000000000000000000000000..5e2aead34884cdc2c9ca20e798885c0405ba5666 GIT binary patch literal 314012 zcmc$G1z1&E)9^kVN)Qo5Kw3&Ar5lb&Nh6|2D=N|{9S2cSBn3peq@}w->Fy4tySx1N zIe>A!_r34;e9!m%<8tq{XT{8_Su?ZN;d|frufTa}iCYo?1OfmM@E`Df5V#3IQBV*8 z1XM7fVWOd-qM~78V4!23#X5Ty8w(p7=N$fdoO5{Ru(8h*p2xdDaPi{Bv$#aWgapL+ z1Q!Vql|Z0i8C0}0XlQ2$aIkR*{>SkBGk}MQ#)yLfg-`=1cn~Na_Er}-#bju|* ztnLnDW+`J*kO-Irkn+eL?$83j{`_=*&sksxAI;?sz)IV15 zz5}!AgyAkO&4aSdFH)u-N&x`0b73Bx%DuH5kirV6Zul|(p!GAyb%D*&zkRU^Zq8%0 z%0}qyc0xPSCie7JZ`yi>38_e0EnS^&9JaBf*c5t8qE-{#TQ@&Hr0~) zd%f0F0MPRt5D<~uF2T}Y_E0*QZc(Rvezx}5P$aQ+M@0|4!@i=SZIRk zkp*R}Uz6(v;xqe!MIXlwJFpgM{NpKJ9xRJU~(v-Uv|Anx)@!ob)fKxcxCC zr;J`{;pIg#xzPiTUOfDnU;v=hN;q!-U?2tqa7w*7XF@4F{s904DY_i#$D50uq(_P< zua8EGn-iaO7z$btZ*4H~+&x0I{XIet^l?nOcv2&5M0Ho=|u!#lV z)N)=y&O$?*aGPZNU7IX<;<@3V7d~w)67${|0>xs(;}(R`2{x6L0AriFQ627~{lA|U06^GFYi@0F>y3c3h-DvVp|*HG;l||N zb*XdsOC$LF&hnm*zEuv7SJpl?;`J<)#_mn?QX@Mg8G|OJ8K=psFPeZe`iZJNzq( zpzn$vy|v5q=nUTcdgO-ky5k{j^(3!tz_a^hG9nz`0a+2=?|>xE&7eK^_wrrnaj2h% zdn{U_`-e1aQdzInK`7pLKfU!4`^?t#)Pm%!?%jbWFv#}wM`8cO^8jw7Gbpd1iw-Wg z(pJaG=Ix><15GI|Q2Ea|0Td;BKc@+LVQpz+ONn$w>HE7KK@wleb^52G78Q`>Qc$1yT&tt`FIovnSrfjR+wdzAUYBaRrWZ%vxUXJ+Z#4MjQ3psyT4 z{k**fFTV+cZyvl7oULo*s;k+J-r8U?5Rz%slr}e2w&+zP^@-Hjdlc&!q(m8^>4e%+ z6%X(h$J#YU)$_mrNj{_uR87YUF>wJ3($y) zaE>ac!YX%-1+}s^0Pw=rh%0$#8KBu;9#T8osoe*8%ZHIuh-F7BdG*;P+kc-gIBH!( z13E)JK|24E_xkRO@mEc72^cWe(uvpQaqjA(xktkbxw_Gqo8GbIZGLeY#<9Nq2fVF` z)*0XA=L!8jCG*9`eQ^``(q|^=;C}6y>2&4e_sivaSep?%5$Dk+hZ_bcU8E|j$?ntGt&9O2Cb8@(;fg|q>Jd1!mCc()|)c*0j5XA zyPfeWxX|r|LxDN2DSBk>@EEU0g3+rtr9Q1LZ7#9OUt8N(ed+)lhX<4q>b?5jRDt|V zy0*{_F>`CiLU(Sf50;N8_t5#?-|=&wmW}KiU*AuihyutojxWFvUK{%RCEKXMRXEeH)s>jJ;Y zIy)*zj3W+{KJ)}J2i=SMk>$UMKtqb|m~*RVB!($;ticc^AS?T0;hcMGvT(!~0Br2@ zSHc058@N86WhDoC+utsszyOSP5$6L0!;YKN3%VA>a4`B00I3~DD~JD*h{3<12|khKn-^Ri8ea(@`sIHN5V7r>V-_dHFh(P$!R-jLD&lOn)7snm zy6zH~f*VDPG3wSi4X(!B?FP2uv#|d$mD2(T#fPE?rH7(Iklp^hD;4f{>)TL|ED0&u z-1)muH3Ps24|&{*AjD zpxU|n+=8AFG|&7nPwcI6O6EVjq2eC$zLrkyZOvD#(UPqGG9VbO?`Jo^2FJNonJev~ z*`zF<)P!7yH`W_jwFQF@cO+oGAReO)9{_|8C?bX7kO>JMQ{fMAWNuu{EieJ#f|?qH zsse3tyU$x*aUW~}0NZ^Q58^pmd@onP=_^{)z)cSiS>_<*Y7> zJ}_r{o1p2!zT1bn^-ZddHDvROH?1>UWrK4}GeUU*roy4wdz{|yEv07lz?b1Z0DZY1 zyKcL52BS6Br-;W7-(osrObdDebu{qkG!xD~D)RIyTCF_A>y==KuvbKhuw5xRh6~EPsBK z-yydT-=_)o>yM^Frr~M}w@^AeuK8#S(_EJVd|?0z%n8(lz&TfrkasQBdx^UIM1cb`5JqqAGIy4A zCW3Qjoc@N;wUSX3cIvZYNgJWg)$h*KUf32kc`^KE(4zWSN@3t^)8`o}ZQ)w;j(=R@ zK%Z|FBCmPj#Zp^V^Ht#!WuFt3^@Ucu&^N-^X`F zYY}T2-P%zlDO~0YdFyVUoKn!kMR(R`JkzW*?MbWMCb@hP>85Rr>z>WE;@aUUZVZyE z-sfvZtkWzhi9n0 zD2~Y40|gj8-z-P0lUrLTs2{mlU(2yPn5>bsol`e5d3uPjMJ&oXQLS#FH7N&8Rd%yo zk=OIW(m$9wyaeD?QdE6dn&pe(nr+b)$O9sG+U7K#QP{D``@X~EP?}i?-i^FIM}8pL zy`Abfk)qWbBZ;3rEjd$2@U)-Us9}Ghj9mb0iK1VQ;Y{2x2ENtN@@saoz%?tr+Gx56 zI)P{MQ79z-w!D#gWWAm1=|#4Y8)who;O&|8PC#_auW)l6t?=BSyzpy6iGKDU3FBG~ zxk45S+NyYA`C#u&4|PQu|497L5xzmqzTD2_it*s|_bpz*a~-@K_O-~@qE_~7PpWTwwxs6m_#9Gp0j1o6TLSI&bvtnTdk zJgFl~l{{LeeXU7;radjj1?$!ePk4(stS6VH_0zIPmYSN_lkyH9MTXN5_7?j zzy7B0y4W%*Nvr%ik$hytXtW6U|BA>M`%Y|}ooN!E(U2Z5zd3J*GhA<6f1IUxoTfzA zj+&WDz8U=By^?F2l6Q7{S^+fal=iM|i12*}B#;k)thdAWzP`v>`5lr3ip^1K^1+($ zpc;2cb&X~B=~t$uR$nMDuZTF3K}Sa;zXMSyFP$0b3-+CY6k2dQz5^{T?azTPPt31K zuT3kheJeC{pi{mivSxZ8TDdz+KG$wtkFLdO zhK1|O8%e2-`pm{sF)TDQWY_)egOKR!-{Sv3LNF^ddw8d6b1`V2u}CqeNs?Hj^^J*; zm+@-nSd7b#iL|X&A9`8wQa)O|QT>ip5(97R>+S~c+x(ka<7Ja%9R7=Jq9mahXGUy* z=VnfHCgvA?tmq%izK|j)6i9t>&n?s`%!N_*r~+L23Zw8d002X>0G&A*cmT4#8~~*( zn>~Cv-ue*$;(e?f4?diMLvXd6XqHDKRpBWz^9-@%QJ`>l3}M+J3#XI))_iz*WxDL; z?#9=&vAR9#I>`)Sil{fE-gXdLeWA#?$Nm!;U!5CWJr1wo3}3*J+7cyOY^U+x zzV$_%4cm^%he3v#K{|$Ep8+$8;q`F{sw$%9e^<(WJ&DGa6DzcoLoq44^ZLcLNNh?Q z&Q`8#wk&hHf|o)%6jeN+5nTKsniTnrDf31n8>H~TAyZJO#lxV!W&FMj+(HEehnyi#WDunq{u{|yM*EbG++01n%t06W%9RL?EEQ0!s zeSo4&Dv!lqGXl5-CKCLwbPTr!r8+)+ynN|?BQ>>5%yD3>UnU+WaxvC1Ci0u~yB6ce zNFDho-iJMY=*y%J?2oGcGeTJ1A%PE(1+sG7ViW*44^H!6I{?}`>M2S^yFEU~;J6d6 z1J3;O&`}?w8F7mV^*3eR1YL=UrXUWmAW>nd2wEDUo@8r7xe{ zs$n8l(3~ZP$(26i(#k)V{3C_naPZ1kaA${HsQ{aU0CXzg3@W&_{~TFRY&-2#od+n& zb~pJe0bkI;zug-ti)a{Dz)@B*+rV+je|tjk>+gXNpY9qIR?!$k-Om$l)K@AN zaqC&a(tj-QYa-T_NK|m9`hP`*AEPl5FUSTe9Pvg1hHnJ`*sn=u@`G=LQ@Dr9eKZmV zKu3$XboJUV!9nc517?TJmq?4%g%|&pmE?US%j*WYn(9YQK98vWEjJ<^JOl~n_drea z5X8^LOy!e){YmHfPQmgrISYe?E+q`xEn@V9V-gs=M>6@nl@aT9#t zmX$b!L0*GyLO)Bk-)M{O||>us`BTdWSDBN-}WeD{zOU+pw#r&-m6i~gYdw-L~H z5T#>7qn#Z%_WYjUTbueie8*Eg2P}X%;-bReU!MV&6^4J8liLWMYpt~3eFjuGm;wMk zm%+Yv+NhoGFay7!eL4WRoB)Syi5#O-5qX~|Q=VPrI>}Hv0e%6Z%pZU*SSOZ;Uk~q| zw2ca^Kqiad{-n(KTSjb@Bl_A#bf_wV(cewrSM!F2J#M{=4qxJ9C$_vQ z&c5`@)XU)kK>URc|9;~W=$*eS`+HrS**YF;7D<~ewT}D^BA4e3H=qsa1gb>J|~1u1=(Dm0RRT& zjfSY&r^t*nH+4EK_gB|vsNZN^r9<@IAAoY5V=vFe9uUVbQI!6QNJiriX-{Y<68=p# zVpjEWF&*sy9&|st??+W{muX9Q>G;7LhDT70ClmEn z_9#EYrB(f2pxO!zs;aasNd(@-fvQdHa1zd6X0Gg+~r*~sy z9o!dBBk5wyonx5{s?CxWRr{NiLKi+E%h|6l0#yK|vlIBmDC|^ghUh{xCR%GpzibuQ zFU>~g{`o*F!@Y6sR-uay@U!nw=mTQ1RF3OC&WZ907j~`wz6bo5TtDhCu>VF|yG6>D zXy2q3^1?m)3tbaO#-lnnx);6-+X75#-{7dq)S`A{BT5Z>0E9q#^#&vSN8vvs=&A5D zoyViF+YCwxYZmBIX?Hx@PnXC^ZcOm)gb&~-L!3R&W^SOsAQv0_)aYQoh;#^OB}FP^ ze?ra}01(>U05>WC>=%e~H~)u4bQl2aIQgsyM5ot-(VwkGl>G^~7zN8{r60}^002Vn zpr#>`&@vN;^!>rwP61j!Cs=7pk0Jrq{+EV9CN{dcMDKQPk~j63PB7_Dy6!nvO>A|Z zaWcd|Yu)U@nVcm^ddZ;k1x|0ELzZfO1wn&f4X*8F`S;+7Bk<1yc0w8*wfW|EfNmT+ zqhsaX`ikwYnOyn6*z`u zXR2?A3;znH{6jB7B{){!Ybi3P!G7%@RLgArSDm?Fu@itwa$}aWUVWbr0};QiJe0ap z14^8fdp*`S$zS+#yb8`8QVr@mw4u^nv*AU_-7?AOPPIbumSMJfn9*GD8AguL7-BEq zHEEKhn!D5q~VtK-~{xH1cMXwBFO{JTn9ig)xviN7|JN zQ7?44H9JjkLjKd6Sc;SjuMsFfIJWH}_yTtWaJ?#}Tsp+dK8LbDV1W zN^Z!2YX1}xhrk2a#J=@L#pqIq z_%HPmsOkH0H6Jn6=80ZfeX@*7msEv~xFYiu$h*wPx0*%c{wP zp3Z-o83di>hq`(WWvoE)SRsPC)PoBDC#`vNe z)MjwCIKDDwY13Yr&Lz?=;`NX1jEmfbR);PFS@3k4L7v=w%lzaz5{G!8C!~*lIAaQq zI0&bi0KgzrGhS^MTc_95!CAnf`RSmhRzSXfWSylDVM^{90*@Rn=}FS?DB%Yy?w{gK z`?$vDq4R`FO^$Vhm-d~2=z`s3!~G{jRF;p-KlC$f)7jOk4lJ5&I>jnCliZboUW0?g z{$YJ0_yDDvpuVq%IeSv7DZ*MuD=#9_Yx`y10|ptjQu!Uziwz1@68EjK1>*Yvh_qKf zx50)Ut1A`Q`Jh(vg!{GxhQ;nCl`mK!@IxF^;2K-I=H#lyH}qQ2M>h`E)YL|&rahdn zShlx_RB0)XMi-%CW~s%p4!XR>DYx?hTo-%0o3Fu25$g#C+(~%xIPA5s%+BIxBLotJ zAF~4>DUoTD>)M@r(mF@HFj!N3oDTZizwn)R2x<~GJ4T##G_LPt&xTze{4^WMYUpbr znJFT{W5ka=8=VK$c&|CHNn%e@jHmNuoTEQkZTMX$+pe+*1LMdWFFE*4o9jML1NwBs zcda@(0f@uHW!v?9<<}^;SQ`!3&{E@h<9#q2z5#pPG@rSo|)N05IW`N zl#5^A1k!n3vDrciMz=4^M}Y}~4mm_t$)_?%btlY%njM*+v$^H{C*uk*9xJgvW_B>NvVDj+I1{EWbL4E{gdKnV5I;}(!g+_R z6;H4LE*oA%bmRc+hKf=h3x+!?e3#@^d{c_S~jXj)5nY zmVxx9W>u*-^ZiIz;9wF|wZA%WDYy_3zt%+<$uc8|;7qx0uT%|Y=70u&ei znK&3clR1Ht2~QIqdPk`K6bY~qpT+YCGO0L4Ox17unyk?ZxydxBPB(0uYi!;u{4~YB z$Xx+IRq`2?1{n$FfQsN!fu4y6>i`tnP%*bcZQ293xPKk$^qMt`NO5I&B_#MU)noI? zrSADVSIvDi*5tI7)RW9fAA1WAuy-ddhD0D|%g?R$l=$fzi%v!xI^}Op5$YL9a7Ocw zpf$prPiI&c+Rhl``ZJh2ZW8oQ^&8>kE-rBva;`fDwk}K0M8_cw8<#-3Cub5mM>bwI!p`w;3ydoFP!k7%Mn}V@fB-y$VDG5Fcf`HDqJ^ zQM&G<^hk_L();$C!6Vo5&%FG~C3So( zg810OhB5W*hG#VixYPsNVS#Pr-)IL6h{Phzk>jEjK&m*J-M)iJYM1=1+Vyau@M^7j z&6Je2c~6t{##r+me;(DCrh%~!+v%^DqrioAyt{q70Zt4)fC_r1@tej8!a z#wC=kJcwlYr`2}K|M45LUY)e`u26ot&C6yO+ldthqM@yW9Fz7oI$#tz#TBY8m^OvR z$DHjuxZmx7SO7>+)Zc%V8^j6+1r(^Zi=_c^ckZX9CWyeDc7$Jm*dMF(Tk^Q|BNP~j zKy|th1-?KN%U-gr3K0W5;#8AWq-bV9t4mJF`zsQ6BH*viI<&>A%k(w9Mqmv?iL*<( zxK|Ky#8!M_m?J)s`$1XU%RJVPRETdRfCCoui(P)w`rgdy!&kM=_qE_qO^%S4P6gGE zVGvK2oPE!4cR9C-hwRdZv+b$JiE8m`k!uzascSXuI|Q`PY#@2--YR+ zr{;9cK8GPk2Q=0xuYS`z+bohBdKFz;6b6z2Q13B_KMlma*1N^5_XI|ZKncG=8eoW} zg|Oy80KFP23?K18UjW(zZ)DGchs`}^P1X3HR4z4qMO>mm3$N_IWb4tWG-GDdJi^p> zf%C9;3yVC;vjVCR7ca@ZYx%w#b3rQTy-grL+!v|xywL!lq%Rxq_GOmU>L4itgpYfL zjCj}7T99Os55%BJe?9$%_~eIi;EAOY)L0Z_ykY-PawJ>(2zFLO(yNDGkbK?dE%#ZA zVQxl^MT#~F7>-Us+p}TzQ3l=IcIyufQkk*EDzjK;GBYxVUV3eC-wS%Gi|K9cHo+cB zXYttc+FS35G()eUXJ)TE?`c~UX?<~VzdW9qN8rFQ9cs_DMze*icZo{eG|P ziDVzA=aSmw@REQZpR@1vA%JW!&D76buWrc!Jvie*vEBzRe#chv0#Baf$t?X9bRbdm zYs-~(a-T~G8~#{l^DLp;*Ldkjd49kNl%c3_kcDIO=mXPIMp!q<8wiwQGpOJTS89(K zj`%(q;Paq`gobcAqQXHO^u}nRR;vhIcqcb@i9R-^S6=KuxVXbtp!Sg@jdFOQPwEt{ zvD>Y|zJ}0OvH2Ew2Bfho-Rd_mS*Tu(x<7M zQn$G2e0_ehk=bY4mCDH=IRGlhLu~!2#Bf0Ncf;|w1&V5Sc`Y!)T3aaIh{$#gJ=i}# zb|5G#4&(}h`P(#w(}JB)txhL*UyI71(K4S_09&732e#*CPnM-mc+jl{JLcNUNlOC@ zjr-g*tj)_g27@L7hQ#ZN^>g@rova*>^s;9xD&2<6_`VTJI~EtkN#)!1v0BiK)V|vC z&@^^8{xZd>@o!5Rjb<_SYr&=}*rZcIi7Ud4TXxUA0gh;9u)U%3sANdqS@PC<%|=_H zIffSmFp*m>==WQg1U?#9RU4SPGoSXj9o=UlMSoLf|H~lDmu)rxx8V+)J3|c3|FZ-eBjJ(XQ#4S(0^z z0q9WSum}CPBA?{>jL>Chyc>H!l?`HerlmQzklML$lXvlp94HkLdk38vDhG#7-n~g2 zce=;j>wNVhDnYZ>j95-OAQ@adSBK?==Z0f~uo$%7jlF1{yz{PyICx7laQ8@lD0t40 zr)Q64WTNuyLa_g%?Rj}pHzjI2b@`FzFyb9TjsCoQ+G zz968K3eLMW7+7MT=D0^-V4ueV14$ffdEM5jiv3BNj&jLye02yT8sQ^JBnEr6T!>}{ z;TZt|RfVDYcJd7X6#92Ir+ZF@7h0QvRA?&G#o-;OX1DdFOKmTQ#w;&@=N5_F+)a&Y&>{ zNFw@^7U76$|IP}BR<5z-pS@Ot##m{LyXEN zk_zxBsLbgh<}o)Y41o`mRH)#f)+4mPx=bEbhs2-yKeePk}w&Ev1_7dVusXG-4 zTQ9My(J!b#6^7Ve;hK6*R2)x(H+K>eI>EsLfuPFx`EYA!)Sr{~x7GBs4^e`ZWnzm6 za}w18xO?*KljeB=u7{15S#?kX{lRnLjAvO4WjXGf+M6#=4JU+@WD`h_D-GseFadC6 zjJoAM0R|3F(lY~LAnCs|J8<0_S+&qw--7&rfDoBB?1YiR*gh(7BnB+yDb7|!8wLQj z+tvn`F{#|nMI-w5-IHdt;yjq zYR?H9p@<^{1+v39oBY8S`)Mdw`I?b7OuiBe#G^UD2oylFnDM~|g&$x`;?+qbmvE@|4@)HIC$A&G< zI0|@>qVW*tHKJXAv!+lw;xdw^Huez7fUp|g$ZU*h-0)~R42b|mjC-%&+|@wpcS_L? zuqcdlo=N^Fs77B+Xn|8_b9Jgw+&L$|D47GHPJsj?0YfA8SE-qn`+55{kq^?U+}3>j zz1K%F$w!Hb+G~bXW*YF(gSA6vtjqINbgw?w@BqC9hy@?21NM{GX0%O=Nmpyjx#UUt zGWv|OkVwn#01VL{oNUAHEjWP29*0GtprwLkFmmDB^0(>b(bxOi6)9Uj{WU4INwu?^ zl+K=~Y-1*?c*Na#4nVurRCIMDKms8Alu?i*E)EeK;LjnNYw|@Ol&#MP&5}IU*n2qp9BGz7WRsGo;f;dv_BV&j!1VC$xX)4X`BqBdnr@HnhHN0qzagPl_h`gWWT5F}_YHyT3H`$8B-rLn#P zSPY|B3eI}O@!*_LJP%MBDyEa7a3H(ka&*c zNX*eOu}uz#h`BeNH+u@p+|9EZ_r4f!#^>je+>MJEvhL}e7h56+)`=S>Q)7FSj_erjWwQYfM7!|PP5wLQ-OBM~)ts7SAo}n2j3}xZeoNMaH zE}DMcaPQ6Is2>91ivi#40N-7)hVd+^0h)ev_Whb+B_cn`ho1{$lwA`dL}&{oSxNn@ zx09B{Vgn#6t9l!(&JJKJt|=9-HT2InP3(gju>J0=TW!|WoxW4{V?AuGD3)P>) z8wBmQJj9#2{`_bj>b(%{*p%%nX%LDHU*E1ioBAKEUceHhm_oM(E^$^iy4?)Y+DASq-IU{15@nC5axAlg37 zWn^51(I^7cm%0vS9XMrW<<)Vb5k2we?!e60@3CR(69pzxDC9*7;%b|;MpR(-I<@6@ zYyiT_c?@mPMT?h)J_$ZL07~%f!1CvD*CB+`<`1s9j`z9dEG279R4TLFh#BixOzw%> zHX`nVVjxkO?{^rJd(>0lK3AsUS7)&$^q}C>5!lJ*P}w zvcr#6V*ouR_XcUwoZT>o&aB;HhxBV3X(y*%P^VS+1e~*-F^>{GBoAhKaQ_|66@Zyx z;B5%fsRcO3YN$D-%j-^$YU|y5sgM*mm&~3p-@%?Bx)5f+=Q=gk`px(L&_%F?$aRQc z0G)1^X?F2{Nm3?6)8_?8B%EvWe<58Q!;7jAoVW4ZUHRWgvo6 z3YJaS0K+m3! zM1Ip-dz4gzZ8w|>@M;R7$t8~4HlSBS?Q5zHzIU#qIe%}AxS zZF))@_f|q~S z1=3g8sZ*vOJqbc3Gi@x#hvmU-?Rq@V9pCZK_jb4X;?^hY@3kR24NJusaefK%<_rVBn1 z>LTc4I{~yL%a1%31MG++cIdd6@X~j{2@a;=*mmZJbISk_J-hgz;w!^as|B`OKxSE& z)PN0Wlp{3*(x-sv1t;dfoq3(njBo|MG2PEncLZ)APsjgI6qrqn(5g#UXTZAGm`21^ zz?}m^vRp*Cp*7&RgY939=vHjy=Ig>Y9T*#J%;=koK2^_ShorV)*8jNHfkn7|BtujV zU9Pmd{dr|K*?3%`v~S^UPZakS<#ofp;iiY14cqjTc-gvMHkPjTk6vB@FX=Ho2bWP6 zM-<7|Pt6fYw_}`A*FYxFF7#xZ79w@G+YPBS-J1+iAOm3U^vrb?fh`5W_M-;V0nz&n z_@1M~3Z)5_@X|}Hh$;mpam2kZh(t1?6QRwRxW@PMF9SvnhO(YqSi|J5Ni@uPR|jI* z4J)+uns4QK$@KJGy$7Ik)^o1DtJTY*kIq#=X)5k0GePIUma%2ymwytS|NNXczqVgV zBx7D~xZ!Bx>5J&VSVNak<+l=KP&SA$?o-{8yZL_=nGX`hkiZ zez_0-t#t>aA`Y$M@*njq08nZ#=oaGeuXAd`(PMMwpTr(3PBLs?n~-$=R5<|@uT<&B zG7VgCaGAc>O7e(g#f<_`P}dX^v0a$*afs|#h0|J*rkve1b*CNN_)rvugTc;qhlnq6 z3#6P2{scE@EQ<#(+}1D5t0ukkQ~^iIIQ2Lfa%&=|V0D4Q30sZttp4qBZ;U4eJuXa>wh^jeF~VL>k7N}X zfWrPF;he+VZg?jsW6x#`(mm}_nYMA0%Q?BM#P!xvzdLzRBKlnm93aVb@-AzoALy8zAr@t*? zND`@`x1=ubbbPZs!bNu67R;5lX&|qp-2%928^!L?BgMhsXp~ zZ~ao^EC~>#7Vbae`}~P6!zi74LDlqPk!tA|q5w}gsM~Qd7aXexCwks-<$tw80LOjV zoSiLAglINg>a*XqwFjf9#?tZ=F%*mBgr+oJ%VK^q15sdcl%BCF$TNwR_2xs`VPzj{ zlNbAXk#7C7iCaQWE1vV!YqDyx($4@OI(_H96cTzLarq(XB;UF%va z@hliYRNe`pBkQYMxHH*12nk(JhE@FZAqMQiwA8T4{AWD(H3S z*27CTwNFw5f8gDj*B!bwBPmc8_VU@=QHE#C$z2jKWW#opUQbG8yk|k&t3rZey4b3v zzL~?Wu~fQbopATn{h}@P9_0BgWzpa};s zysc)Y21;BDHiQfb%nGw2RcUCEb0MR*xg5i@`GZKYZ8Rt}FdBsvH?y_vNyIgyJijHi zM%b6>f!1K-Lhdq$pc3PH@wfPa&qvISl0waFERE9zr5A$TX}<%vdVLUHy>9(Ho0Aq^0_+$`{w72@ZrWkNk_&viCIC zmFJlS=h=e7txvB^;%IcyGE3Y9C5}6N9W-T%N61aKkG)2`aW;!bpfsGsxXj6eqch|F zmtDh>Hv%rN9!rNy&4Clq=`QTS!dFtVT@gpOJn|64@>Nd6Lu^ieN-yB*z1*uku@?6b ze!eWhiynb+03gm{c>Tambs>mCR>D~vsSg6oSY3%8pq8iwyTjrH|ESozMmG5HMRbtI zde7CS?hW#zAnl3!*Q!651hO(E2N_t$ZRQD_H)2-*mdmAmbr6qBD1@%Z$0^-;T3ez- zesKNHrDgub!kh~B6k{!`{>is*1@}|^Rc_BOmTl~RXnhKgOMF$0!+M$W@$xE9Q{J1) z%JVvwwr@xhl*Qs`_o()aCriQ17InVpvgV$!9)F} zdy%GD-Tv`fZNJUhVCCFWI1F50@e_r4z1+9^s7oT|{F_xR9b)@dFDWR=3)mRH0F%8X z#0fN)lIS0>72@3eWSzxspknb-ePz9M7=7@b@0Pqz#d4f)V<3O@CGZ#W-$*s;26Z~N z&iO#C{2tmx_!!E7>VQ^|4yG1knB8*fWWJEQOdmb%8w&8SsSM7ghUE2k9Red#$**Zx zra=G#@;++`LGB;N`EVF7x7;m!_g;;=aVdqyWPosA65%S2CC>so0^+IbFHRJiT<3SO z%0IbBRsseYk0`pFmwtJf{EU9t`?u8d>df+q67mmc`xtuJh5D>Y-C`?-X>H#d#16Di zf^$Ev4+YON{sV40$R;=?@#v)d4qTfzE+-?U+aEh5z_iSyz+)YhkO2r27?&LgX>}%p zdYBy3@`>eb2!R(?qpeA;>d1tr z{99h`dy_&9_NH-eCFR*kot*_YH*8)6+f=Jc`1yncBUY1oKndf;nEKO;){rSzb5Zv% zy}k_lS2YK;;m1P(K2H>Tj7Tmt5f>PUz$=0DlmvFJ7o&q#Cuy>&#ep&Hx)2tq5qj|; zwvltDv*$RAYX%WJNN)3RoYY;_*nwulXU;t{-A$lTN$MP(4+mBkUYlZzFtxSY#ERD; z&S)I1A#0a{2N~B{mf7-E@(UO?Zcu^L{#5vXwKO2aH^?8h2mEr8gCKRSe$sdYCVUIM)CHoPwuU@sZ!y*!$4QO)pRzWlm#K2)s6e zFx4RPt)s4LfvX#*=e5d7ve_teJqszumIg{FM_;Ib+@kpVU$c{Hi2G<=ij#T@b5>)P z`{I?~UbuF+t6GG{D6@azD~sKleL`M3yunUXNva31@8VO3 zEzo%Twa`f4*X9(>jJ{?|uJ@fdl16P;ZiCAmBU1G9rO;^Icc7b6L{Q-CNCn8*SZlA_ z;wFF0=}lOOw5z456mCMb{bZ@j1L;)!-pk-$koW<=b~Et#NvG$H4AaHjl0DABQ#Ato zA!Pv1oeA>ivga!bJSOuSjY>Ub&l{Ph zskvoNztCvxRBi6srF=MSX|q#st)`K#bqV1CLyrV+M8UyY5IO}ymp*~N`mq5f)&&Kg z9-fyli0I(x^SDWp3UC{FX?WqAVyf!tA*wK?{!Ww~I7&0AhR+O?FjD48zXD9Nqx}&` zUt@=I>AE7s)QoMe3kDEjB!hm6v7vk@*jw1zA#bt`BDP|chZmP|Rb^S8eqM4uaUuek zRXKGIzUS^&Zj4U9*|V;haCPH&jeih&eximsSq*ChRDVx;(or~3k%jrdb!L5QEK{_+ z4Zz|`BOzyx51S^vj1zmZgYL#%6cTJ2 z)LuwvjXD`=oGaF33544#HIK?(XM2=O>XZ3(ga)-&9cETtd0J(ua2Y?WJ<51GYh{1t zc*mLc?`<>?x2J^6Wt!fdZkJ938^@exjCCIy9OY9u_I{myA!ilhw*nA}c<(#IaPVMD zPTjOS4QQuWPmCFNspwDmWf9>3?k;8FNa~rikr}?G9Tb{XXQS`PH&+!Ck3{G zYm;{`3ak<{*39Qt+#k2>kay4=>D+}z-bJu##hda;Wk}nsfPc-Q`$2eO|JF{1bbi}m zJ@Wn?Ja9MS2A!4SZvp#h$57CmSl}5g+@`m$Tg0Ml17aHAsU)2cQlje%lH0RFqZ5rB z42jD%fFm1sUu(S8CNCd`{6IYZI#1tOZ*ngAI3q*F0M%JUo2FUL^0>uh#ZkFjqa)7|z=J z9f{1-HAyTldD?;>P5Fn2>Y(P}8`YnTfChl7W0Gv?wy7CYx{(6UdmB*U00w2UQ>*yde1cS#RGH+i`9Yq9Ey((ikKd2*q4k>K4J(ZP`8nVM+6W4WU>0q6V9~PA$vNVPNVH;$R#QN3VR*$3umPp1^M-2q6eX z$JQCD0DqF=@-b+i7tAB*@L67P_a3rx)4>h$SNReK`+9aQ%P24~^|#8!;wwG+j;}|8 zNyk8#uZ7K!^GdW~X(-l4-|8*^Fr|I+$afa<5ix3wP*EUqE`y(*>VOuH=0L!WrLBey zpjK@1s6VnCNYL}mXKo@Jw{(4#)oZ&huSGt@I5L!*q8TdJY{>TxSDO`A2RNr({p=1` z*SdXAwI%(xd>S*$;>LaKjYE)yANqkUVK0Gmq!UJDpfja`;*N%6Wp&L>t41v%Y?rua zlyNE|l2J&GQT(=Am%o6D((xw}_iO(%oAKW;zB&n-`kbF~C$9U}I4@Jkw~BRCn|NF4 z=YiRGH!M^VRH}wWc;c6c)|`v^aK+E{A*%Vc7A7-(yT^==%HW8EQ}`{z3W>fR!^9CB z;MB;SSXrrY-X;+r{>EpKpumkZsO&cQ2!oWX$rX4KmhNS}3g!6@NFQGV{dG)qS4g5C zFio6NIC-O`lI~Nd7aM+V7#m5l27tWPshLEz2>mn#9tJu!>vHUd4|}iN%4P&s!Fdm$ zrFLh+D3*;(Uqz-*Hap)3_xTrWJQ9Pmm6sN83U#lOz8a;J3fSwwUM_R7jx5^rp{M{C zh^e`O67=!|XRw0P^}J%AxB8XBVIIf%O3ikUM^>$AZ-=bPVjO4rm_1Ov1V0H@8b zcQ;K~4)0@s2MWFeaeIxkwmC$GB6N!aH}3k+u8nu=fEAv?&z6Gc-DElw7qorh;9scl zA5wYvf+Ise6hpar$ze#B>Gp!jXLqC1QorrIN{kY=`eYTRQ9o0`*e@dL+C^PiP7lNV z9ni9Io*{l$D@+9*y1OM!+X!v3zMt9Rxh$FUv+OSgnHyh!!K3?tGxIa`ICeXkZLi7<3&;_Ug3?D9WTq4od=9Stk3Ym-3pW7{0@-?uZwZ?RIPsfaaO6X5G9X}A$ z)^uPT>@SVDkOf(vdCjHu=Kizw-=4?&; zBhN|X{)i*9<3XZ*L1r0nf2f18ziZF_6BgGeg)1GUwXJdCB3e>Yx4b5vBb;&=ui*`y zgY473PU$AB!u{CYzNPp7`H@Tk-fm-Qh=c)zBa1NMA$82Ld+t-)6m7=zB|NhrMvi%UxkZB^qAb~vl1!w?0 zPJV!V^i)&Rh$-yX0zdPv6A==r$J?*`ntB}nF*LSSHFU}aPt(gk3R8F!%ZH; z|5|{;Xt{(ipL?V+N^x_#q6o4QYdl|97;O*!*d_Q0=9ak6BWkb*q1fqTo`U6OptGCx z61K(geBIA_aC1#?FTwyIH&*S<*3`g3#;(&7@!L&u*MiYf+fZ88f}db75T?t1Zrp?e zyp7op*CUH51NM*t^JgU|u^$c8A2@o{I|=_I!%ss=pBLvv?<8mnlWg5#5+nCRX@&ik z`&8EAF#bR8-U2GFY)u#~La;!9pa~WTlHl%6aCa#*xVsh_5+Jy{1=qqQxVr^+4Xz39 zd53i0neOSnbML>_%=_PZyJ}UPv(LBX>~DX%4JMuSO8HOEe!KZcOX0D>nlNDafMA1e zloFphlxDdR{}uX&h5O3Qxt~z1H!DE0>gi}Akd+LbiA+zH7qCLhQeQ!gn?k@^eF|6` zV2U(a%@Q#Ef;SBpTZ$`QfBv8qm@W$}^=+mX<@J`XQqA@Ny|kfej^BM4DtW?IhYY(h zS(l7S}4*cjthDqx5+P23f`CWqnXySQ&252!J0mxa>5%;CS*>>tECCB6R^tjqW! z;;$gT>~g?$1D|iQM%`!IZ3KX{*&psTM-l#!4g7~Ng;ix9#h$3wuW0?7ynd(Ar@|wh zwPUkky*_|?f_Fh@sbp5~^H*>*7COr(u~xv4&4PDD4oF+R2?Ur5U6M|pfHA}e-|ZuM zHbN!t2RZrOvl5v68By@N_aWR&wz; z--v6!(3%3wY@X`Zi0a=~@mEWAW9oK4cZ00_EyQmfkM$NV!!Lf<=WjrMwZn5luijs@ z|2F`?iQ#Wrt%?lxcLIM0@E>uS@&(emvrCas!WtSj&wCn-RMXt`s76tiK*f84#s~Y= zf}fXpqPFwza^ZaiX$NhEb*uRQFi!qbjL@@um@*lmRxL*zX!vR#7~PFNKp+8VxN80^uY9f8_;;-RKg$2d z#dDN5+!z%5GI z1cZ~EUx(U0aC*|7(pNbQu#*SCRMx)D(eHq6GQNAU`uVu!o!wQv$yBIPRB}vvQdq z3a^?x_<)33^Pcoeo*>tXUX7Tx^XvBQ;{StEf%&`P94A$#S`oAX!mA;FP)Yt}CZhJn zG#iK2dGZ?tL|R=hfLi{$LFr~60Cg*`s{x&PTG@lH&KEKOEIyf5ohJG2>2V2hBT5s? z8_D5zZwpt>1%a?j$}f68Q>r|_SoVGJ^?`WYs0yAG3Kd2aX8(2(U-|WK3rFL70E?9e z9xDE4l1BHx1^%Cnq@{!$!??+V)E6IQBrYz_1st+PFwjb#mvlu2M{+f4=YGVXM{CQo z5D+b;);vf@IYaBwaH^!rB#CcM|66==9x=Xkwm zIIkSh!8X?x4GjM^&0p^+W3dW-evf|4Y`}#`%QnIAFWAsua{(%_UJkn)lrsa$umXPq zZtd)BY&GW>G=aR4i!{dmsI_(U@F2h!2?7a=PG_5PRF<>?*Z(x25$(k3sby%@ z0jzsh2>=6@BOM~UyC0D!eb*=zi8|Gw_lhYB7{IF%0NS~}irRnK>)*8@w1Lw#ig!i*$X&7#hq^t287aQ?W^CTVii|y zRvN$%IsqWMvNW8|$;&l0bo#?)D9hf8dsELckq_WUGiAtNZ^y@45pcKz1DJQ~aDW+g zBd}pfx^Otrl>~A`G&dSWrX)Qsb>tT(sO15+gNBXWRmyiFMn}2BOWB%RPU6q$I8i$- zI#E7atW}%`N&~@XTYVvNABbb+!X?(w@d1gZ?2LMBJ3lo1KLSiB6F8jHk+a-(A;8v2 z9f&3s|4{gmoYmAeanjl=4H;O~npQh3_EvM@_pI-`mn0mBPeq#(lUNa`(egJ{| zA2wOxdf;0BJme|08NxxZ2O?lhWOEn*|Igam(#o4Lx+&iSQ@W{ishrZWt%1>Z$H#*} z9(ARnWslnLTQ7gPr+=aO7w<&~_d+J7gL>lP+zWu{_vuIS+RV0kfu;C9$jR>_6bnI* zW(9!cFh@NVfCQ(+X)K!LPZ84`Qok)9PE1fLN4{mV=xbm6)^!)B3p9|rvvMF3j`=5J zjMgVrh4FgNqeHlA{k+8k2%YJ-u4cFa86|*1Vf}SW_{%F6;9Dns`=jWK#&vI@f?IP(PP$wlzcQ8!L^^FDa{MIHUgs`R3X^>O<05uHd z9M8P^dY}hf{v)6=?t2e?s9paF;tNxV<8*I8ZgMcF;&5pAGG8G80I^)+X6n7V$>gle zMlO3Q@QrA3)#htt$70*nt@C|12xe#^=ZH)+{;_C00oyKobBzz5${F@cyc&^Yga|xaIGdm!nlC zxO>MK*$b6LToqSGtjMNtwp21Cu7E(nYeXfN0@1ROD5-*~Fli{C@RoWx)avWF)Y!c^ z(=$#!8$LG%%xn<)X-6gjESJ-nJzJCj=&_FQ84Gn*^;eMsp6?$6IIwFs65XzfrU*J+ zRObiB1C!SSi&9WCK2kJI{73?vwTk&0oKMVw&FH784^=!UP->wnZK z)os1KkLxC%e_8vwxzxB>U>l!xr5iI;Wbu*f(?)Y|*9L%n4*4{_)HEl~V|RcO#Eua@ zgmQC7k)N;+RjljD9H-C&eeB`rh6%Lcs}Ge-hV1mJfR9oLcWE^=W&h(dL)k{$*K!+E zGhHuq2TFQ2I!^h?hoj$ZrLccd-$A6$I2XEK0pV%XTc8uQ>y=*PZFd{Cf*ynQGwYKw z7R3Ltf6hPY0?25wW@s(&g5PG~xOx2Ee+NWbS65S;jg6f(kb>^uNCY|=2>6fBHtXy` zlwh^bfFP92|J4D2=R~*ezabt@(nUn~Gl0U;o9~bl_4;3GPUOzI1~8_jnkC-|$ITsV z`(#>q7sx+8+zX+FXy7Xd4j&xTc3M}Ami`v-t)wCKgmJT>u@~UcL}XVoJ;=7F{I<`F^Zx+xa0$%$=BTW?x=gzqs9Irq zyJT<$aPUJRfV(_Ad&%eXaPg18aRMcw{JvI_GFD?R20}a+0yaeX*FkfBX%?h4UC{&r zFS^qB04l4NrdCYBLa1@nU=?Ownz01(6G2Q(>}0mX6pzQM_+$2Qsk|Sjcqpw zkxoE*ivSwYEslUTco7>I*W+u%5qKyYXr|y5uwsGqODP~djt=kTuLg%5DY9R9GT|9t zBqXl|Of~ibVdOK=PM7L1s;7@?TB~=E8*Fp597QoT+2a@CP_T}lwh#zZ-!axPCJ)2JUL2hs6?=E3fA7`dgsL_AdO4@d?9wp&n4x8QFx`EaVNt$gU4(}y3kwd?nm zaA%=yz-AkKw=}?BXaC&2@ITrPCq~}Zn(tlAPFvhgK{a*H&mI=X{R*cZfWZM{qPPH~ zn!2+MTmkCA1Fjy{)2>uJ%%}l~^HFGtQ0Fn4OB_+~243FRhg1FLN2zXVztF*K_K=t= zA{5P|i5v%1WTrB$`JugD_IZ#7a8xyZ#B{Vr;R1lelrIDVCE4rkrJO*Bo*qd@2Ksxx z4o|ooz!WMv#lEUN3Af_8K3oF9$JYxuI$A{qkUN!xyn5P_WYmvGk+H)?{7|ER8DLd{ zGy0qQPBvoI2OtZsB_i=+!CL4bP`9DJr>53kac19#Ph4}Qkmh6)WP@}YG>wLDa7J_l9$=^zu;{#eK#Dh}9-w?>K=%H!mR|fy{vcQ3 zCZVH%S_s=y{_gnv>oKr}Bfx1i7h+paw$(ivssURWA!}ed@xPPV0_|!dl}K=1DKoE56(}2y@IJHKT@n`F1`t!HJQN-;=AxQu5A#Uf7wAqb$x$f zl;u=P?gX_m;S!%_7cWH@z1_-t4cVO9JvE{@gQoP=%a1bGnPjmWCE@-XoMzhjO#m{_ z^dv#iLN)Kx%cc#-L!Kj9tehR-G`SDt!337`(GU@Cy-FZ(^L21m%g^U`7Gv3IO zu4BI=>aHfMK3*%euQt8iOqCrw6@0U@g7yeWcV2DwKS5H5TML=$H-w99Thdk1yX+^w zTI}EY!LmSstS!JUZUkgr0oKA36%i=+YPjwdO}y`r+jhl;EA1Wxz(ahn!%^!?x!Pt= zp!slOj=FqrAF>B9q1S?g`{1Dm&hQd>!k$Ic@?E2(EAwR5M0Mq1?Z=OOuwt(V4Q>lE zzgk)_By1>b?iKeD{ZN$;yOe-z_u~?s}>3);I0E)SaoGUs4N( z%tt}y;`2vW-7;}|Y%J|?-~G_(*CuDQmldOREzqHgvn}_1zxByQyUyBpxYW&dR{FoK z)Sn*0HMRDXSoz=re(LfsfaLfUP6VJ<(jeZK^nBNU0{+ttGCi<7WEssfm%8pP9QbP(=-S@C)%?% zs;({IJucDT+&QA}CseaA!IxoJAK%)R>X>nA{c{zx2ImM@Qi>R(aBPETCKn&a-&EwE zMkyB3@abw|`5+XYx8F?h&q8X>Rb5YkZ9HD`5yjkMPsI$>vj~8IRZT9s0{=9VKj?dZ zf_X_fDL^ip0Y{Z_va7=zk$1F>q~EwP_v2}z(#09*m$dv7{9kXX8iaTFT>QEB=?AMG zcM;IFhU`!!eC&$NefQ$ys${;$4|y1iz?o;MO>G>&^HhE2!5FPws^2sa`TYIk|K(T&VB-HkH!za|P>2AZ@z#93U#fY>e_P@I^Z0K+LF2D1G0}d4WRDhoJ(gQC z6@Zgb1=)~Vu3hDlFV6b;?+8d69^Z!`c>e^Ej;PN}CA-f+x0-^M{_}4QFGt6=R&Z~E zG^uC)6}h&9gk)8K4X$BGlG><*d3GhYSXOy$oPB!_Ii#@4tU$HywTh~9_M*{`p%Stm zYBwR{pK5UZ2(B3jmt)IyGk1?NoT%_1Vg|AOlZ`~#!s>=igy-FxP`EsH+_>|5U=vL2 z4{?W>&H>>5IiqoInx@U{MgJ^Kqov6gv!lB7VFzW#0Zj{AYX0F3!OM4k7KLoJOc$>) z)7IDQf9ML8^DB<^?!mb%q{y<3=GT-!a80K%&a8MdhXv<2tC^l2_%`KKBSF?nGgIyb zHyM5WB*v)270qcm^D|!zdFeCBP^zyIUKvK}tY4!drMd;=yt!8b^GqECCmm1LRqe~y zv7cK}o%Cc)(Z4bE9(ADF!Frqv8$Y(da_yx!q-|dG9Mwqs^LrnGwh@B>vRg9uuDuhQ zwAaKVua}_tXsfG90_8tJ)BcYsrj^8G4J^2sCT0KVj59}R$meQ~1sm=+(6UR!;c^a%#3p&Z zrgZ)|#5ECsS!(a)2DN6GrnVYgi<8Qsm2cFd70G{XFFMedj~4=Ntt^}0rvooSofuuO zIJx6=X{jbMl)NFe-=q)xl5L8{1Jx_D>-Np9Gbg-W?WlX2OKpkKVUW_ z9UJq(ZyZ@OnMc*!YE2SVQ~Cgv=4)c1n%ZO!OD$U3R85Q5M+0u6A}J}(5W9S&s97&n zuZ9djy@^W|N!78XR@+*|_bgXh&mv16WSB6cMsWqF4~g@Jmgz)>kI!1Xy^_GyX(NKJ z&0A|rcAa4;IVsGWG}fm(#Fxn8tEN%Eq@{*P$7-uZhB^mz+Q#WPsnt)5a-F>+v7EY= zN(!yZsq@!8s*H%s4ZsGQV+AJ^hQ3u&eMfYtbE8KsBUCBC3mwif44Ba1CV<4nk7Hn< zSh9B}nyo{V1TXlqTQ3g4PhP90PtDJL?wl{FJ7(NtNmztJ&Ypx>X;8+HCaI zrSxAPC{B8lZ5s~>o}}VE<-i-y?c=pUR0wP^uI1zCmCuR_%^wjZXMCDv|AejbSb~?v z(8-I{y{Uwcw@OLu>A17+ka+uIxIwZ6CboERwk*};I31Dls2r0ei-l72z=XbiHSbG^ z^VBbX7qw`nkvTQDHtFWcka`rx`y&?m}HZ_XI!@N`K5#ZEyKGHFX|O}cO@V7 zSxtqrjy9$jOg8@nm24kG&lqn0qOT#jII!2EdP!}uuu)fpLg|!vS&GjXrC?a~I*HIu zA)fU2ujdO@`4-L@*^%;6*40A<8QE2*=%GXAA=V7nFHhB{Sc}LVz*(58_`>Og>cvwn z%oB}8LD?!}#zTnbSH+Bhg!!hh3u`;42^)b8( zrjlbt+Ny!)wH9V)3ka^}rD@MS;R~*CrB=q1t2S#y$xNFtw7UI=JUF_tHx@S12Xm+f!Gr9_ciw6Rc!PaRK&RtA0}x(EImNA$_tmW@YhXc|U;;y%KZ-=%LW!@oFTPXOkfY7(joqh`eg!FO&RnXoZr{kB z7g4)j@N=haZP{KkGL^7OD_B`sLsIW2+0=LA0!6-8H$+Uzh^58TD8q+!1>ew9lp%(8 zXDh}*lSVgW(5`99cgc9I#JI)|`r`(yg6vXHagmQ zB9ut^>rYVaW~Ez2?Z%;dXtntG_^u(8Wy$hp+;}DZNnc62+nr?cAIy1iXNW4HJL@MA`?IXbyrMav%ac(ssnq^WNxUZaKp58qW zppr6GHP4ip$ZVUL2p{7u*ev#yh5kCvVf+-lx5Dng5@0!8RUH+&-L{=}LHIZKLFc%Xljw&=M&P7fa?Vxa9-q4v>v3%H5ks-}c9}@J9amfpx>hSok!GFXWNem3n$Pk~35Dgnb1N$A%vH+P>>L zaVh^Sgpcd$v&^cL)?_1a`AU-$qQ)n7GUkp|mczW}YJ;&ts*&7lMLFeM4QiGa*M8D& zaIE-+aY7V3e*K4GNkYBMwFl+u&!Azs2JCjZ^hQ;edFb2h%k6vQy{Ycu6o(Z{(oc>s9n};XZ+31+p1) zrxNiRDzcR>(*=hwoA%O}F?#n@<{ucwm7(-fyy`X5m7@vMc72=Wswv`GT1~R;ELkL$ z-iFPp@RKn3%$%Z-u+Lg~QtxuqH4E~d#+gX313gQyLxZ}UsXouZSu!lScP5q1GMNoH zp}xUPa_5Upe$A&xRPM4FrJ|W63N}gc2(OOUJ0BNUQ&x;nCUIh&^NF!K1qYm^;)gjA z^{|FjZ5*oEME^k9mVwIdYg*PQTcB*Rf))B%6MM>3P2GfKM2!m9{FG@?m%zr`F6>S? zvF_$LnVZFZ?r?miT#YQ`66>$ZD_9fiT+*j?lrjtnE}t}|Jt{TvDVphvdN$-QQ_^4VB=r#OEQWoI{F-+UFB?cspI^r-;jy$Z^-9#*;QJPdGHFmTbM_(y zVx5&npdcEWNSL%)HcGuklDi~1HJfo0Sz|@9LTKDlN@<^*fuOF0Uo2x;ZFe7btw{{@DB!JHHkPgUm(#T@Hyq`Baxw8K z8DKKvXuH)6Y0G!+J2Nl-Yx}o*1IGs~Rq$%GY|$?yoh`ygw_PT$gRx7vN)}Tydzhlw zqi%2_8D1c}h&BKWX5y~KV$e!YF|C?Yvd>xY3}2Yk1^N?|1sr4getEA0F!liFlb3dV z{>zJ0Y=%%x)}HNC?%i|QT^13y1J%vpC4v!mZv(7qQg5RI&RKI=XK-|CT=v#)Hv%6& z2_I_~_fC?WoW0?HpMoFjHIc+<|l8akk*uaXqjV-=N zS>&!dWM-c12Bir-ytcUyUT%0ifBvT68P(LA+{?*1`s8b%hSx?erfglb41H*;WPFV+ zbc>?)s3rAEL^Ni7;76)~m&W>&&h#!uTj z@^9;t%D47SGex)eYo(~os=La_w5-h?Dy5g6vrxnHPeZ3V%~6^z#I~BZZ^giC(x%)g zydRyp>_x2-%!N^sm)kxawh`V~PMqU3HVCnn%$WaJ9M`$s@gfTLRs_#Kd%XE-5}T)% zv|$|B%+Hf=@rnkF%yh5uskh;TlB*7&E25P>Xqlvu&BNgxwes=Dw#0RN93H; zF2B86U!wB6SF4$eTE$k$xkEL-p>%($4@PC58Wb2{=V~=KT`WUWTt`%#MhB;bSg=u?vitCokSTtKQ}AG3{my^GHcXZj5(QVUc973{9-4_{d^4 z+0bVdz^O=)5;3Le+-HOmqLW&Fev?L1ATJpaS)Am@$+FZ!c;~Lk(~|Wygo3$^f#XEB zs8qsX92dL7tR|~~xI~(jdpl_=PP%jCxXEjd6&@pGL+Q+z?acYJ^qs zk*wr+?%?!0*;~RIsd{QKGyVY^f7=)PXyL{DCk@u+1SHWom!HVC@^Y&ABx9v*xi%Pid(q3V>iTF3>CvPKHTE2x~1MA#J{>kR3PCM+A5cp>d6@|zqQqc|*d;=qTr`MW25ATY_U z-pn)ZYP83v;c|ar9C5*~JK}J5>pT9b@H8cQth;BYTA{jaqq?d&^{KTU4Z4*syXSUZ z(V%5WvcmfFECjHDF4=H;O+=huK^*73xuAbXJ`16TXv2l1aYJ%ifKo-lf8RJSct^4fBSpC*NL43G@F|^OLlC?-wO@DdU@I7(Xr?Fc#Z2RqX zmr6Bn9^G&)nU*9SjpyRa?&z~e`uM#2Z%vvKEUoLbGHL;*&Ju3RX`ZO@U6cXe^SS_e zR!mJ0GE%0O=0Htuzv{m96z)^Hu|r}DtUEacXSp-cp4XA}T`Z%n8qO!92|q!~8{oMU zMN1oMbX&ZfS(RmTqk6+%WFSE?RTUFapg~@DmW2D7@k)g4!2W;7uY*%qJlSK zeM`+YKZ|P`*Vf%I6mA0%ntE8`bW{dD4P2ipjRI z0n23f3V92&de&@JyU$D}+>PF=ShJg7jvhe|wl}k9mQ;yXSfwh8nkNR7%05nM;Otd7 zMsLnKPpng$$L%>LVpS8(!b9oiq&`|)9IVX_nJ*M)Rccv(O%$8JIsdvbf&YUlU%?EP z7$ahEQ++|JnRK_<#!-#!nf@F8kV!L`cQRufq>d)HQ=)&dwc(7jPeU6xt z5wlGw_D^`@r)ZDnnKTrIrW9(BZ5iXya<+jq_jQxvs-@_syNqL z0k+|4;9|VTq*>(v*zByc2;H?1=?iU|=9j%|Gpr@!o~4iq%kt=QFx}}vE^9`8cBK+s zIg{-Q_xJAgOQT^=P0u(SS`NTuM?y2NoXl#-ZMb6)b4EPcnvkP7U>|sT9tt?*>R(46 zKE=w_eZehr!z-3@$dwx;f%Z zPTKp`X0@B9#GaX~%w~BTduYt)?yfzZtWds;SSh)7rF`m(4X%NWA@Y+FbM~kzxj~J> zdzZ2LIYyO>kcbE?w$67ZZ&LV|RWCYS!A?0~k$8#HCW(Z_=zD<^NlmF*Z~Py1 zst{xkjUICmb&zhD;bGMphvaKgj7#T7b{M9SB0uhLURK{1bD|c^%Tf#Ee3C(@kOotq zo1r3S*_k9iYKutg+;Gc=2{4&tR*C|mHF{yvmetFy`RpC07(^{e&O+qtkKoK1y8-wS zThNmi#TeaYy2WolqM)67S{~o~dpPe2k7`Tqpm1gt{X+X$yf%Z4fyRBZ`nRXBp2$n2GE-w!WoJ4vHt?796(U0BKs(?DgvK4rY z)4#Nh*nNmOV8To3imMnNe@UBlVj14^E{`_vA|1PS-o=ZtnwK@~2tTY7kbCd%KrN^V z$pF5$=@ws0z*&>>us%z}`!w=Y_?_fqx<%6yZOp#w0fd8ydVba=>?VVF+>=RHJs+Q` zqe|2CxD%DzBi!>v`D40IG2C0M@%>=JQ#Z!j;H1e7EA%%m))w*m+p7ZA1XCjnW~EUy z&01r+T$@>L)WWx2Nu^tm9f8=}1_bOuSX)E>v^v%%K}0BycuuiVW&fPm77vg}{x;P) zkJKj2MWe{JpdB zE~)a7YXdFOL})H9d2XU0YjP;rDLIDgM6!@MwxT{WO-di0Mk1LscZE}fm;#<$&o|oo zkeT5E2vwq}4P7+Z2k7>?chxK_yhI0wKJLf#wAa!y6TspKz_-VfPS(aLe#aNdD!L(T z>h5&G+OR+OdZTgDo4f0-R+%R_liy{?=r}IGljwKa6y}>w+PLB3-az>il5)S0LR15< zf?0KkzrWNU$V+v8d$kv>Zh&$Q&2*rA&$L;GKXDp9>;_HK%edn9-itnaK7en3n*@j>;K`sTW@qz08qqL`5pIBlwMe?G>Z$KNE<{{=~wQrEJ zX`XQyz##WQQ3k7Rg!^?-H!J1HLR5px9#dD)VMR_zPx_58+QlXb)yye=G3ea>wq>1a zc`$y!e->8M&X?ln?^CaKQaxd*!!BQiUBCtG|L zz(>{01|O5m+b{Hr7EyivBl^r<`F`?**X1_jMO2oHKi%|6}*_9HnXE<0{;3)CP(xluN+ zjE1IO(?w0=x;oayw(A`lCvnEel zZnHRKhN#lK*bk={Z%uZ%Ta$2@|2Oh;MJqQNA8sJIJ?q`6u+h}m9b?gOy|aiEL=vHa zynmVPd(+_AXV=*ugw?@-m8 zr@pQwM;Wq`DcA7XR}D=aGC!{Qjd34)wR3dk0I6 zzO<|&G>f%387yrmrbT#eE&)CxKKNufe-INYq|XNv)0`+#U=tCK1Z70{j#am%^4#1J zKi9t#JF0w{pSR`Kc#et_XNw{TtnF`qkt#kkWLh;5NUK_>n z)ysFD@y1i+iE{&nh?HlRrW5OnGYYx1Bz&>hNE*wJx`t$mq)TTgUaC+zYe; z%|Tr`o{lMMpURumT$z3Sx%54&d3aOgT^;url~#NcyU-$un;D@=Vk6VXwDbms5(y2u z$TOAVxb5p)*|~3fp*dEJU&}`vMu=)B1FtL%`X?76fa2 za0%03f@q1p?1Nu}2--ZDe++VN3^RMBT2{U<;{(D6;J*7Wp>HZM6$%%Xi&Lf$;{y=& zL58>BwjM^lo)CO-iZSwFn6{^@t6w4&SUv@Eyw&r+1ud2dAh27{ZNf9Ig{s%+YX@bE z^nw=zQ5E06&9RC)44V2R`V-WlE|7*kCOj;uZTxOV3#8cxD}*%VXw5t|(1!UF2ZCM4n-?pk*0Z^O`N#!JVlRH+r_W>BoUzqP2ZxdE`{$ zPD0S>Mw%o=7KyC`p(K7;o~W4iRGQ_0d9mcQ^bvp^+?JAHH5UrrmK6 z$*ak+9QoXrcU150Q{Gx{JX`d&y8nc5eKytz&3uxfMGHo?e$jYqsLcg=#YIKiNaIb} ztAga3`?RSkY#nUTO1OFji-ihr9vF$Z`FZlWGi@S+Pk4ygQ}xhNa;9DnfT9x)N$7cCMyW=VgW#p877jlPd~MIvdFI929a@ zCC_eOPIVB^K>KawTds*a8m@g%4>G+aVefK;l{CqUFb+!wp`-Sgy6Gwv)axPCNR}zZ z-LuFB`HYoS6=S32LUKxR5e2<%(xj5gHP3gcl$fR8l*hwOv5DNu z%q}@$j8J;LXmGN}4~H3xkCvyt7&3&kciSB;H}c*Vjf`=gu*e6rq%FCzudsf^CiJSw zv#3^I%{k?dOSD41l{0HF^^+lK>P`^eYW@$R6|rLDVU9{PRp-!UFg(1VU#>NDfzv>jiY+C zse*BE>--z)Ptd;L1loaf0wTFqXMaudyE|>DDr3kX8NrRFr}O1-5A$7d$6W}2!=OlQ zQfOY|`*WK;${u#>CUDJeQV!qd(#rgeV1Ua7`k5#PF1NmkrSw5yYB_Tag>&cdOyWG| z|NW?hCs-ZZ)Y8gI3C`BS0) z=QsBD)kr$#?XM*1T&|t`CzcS6sE5~~MEe$;huxKF|G)c-d$OTOD0wD6>0bNUVo+E-OQB3VyFr4|7ZVj#|yI*U1bwqEz1>J|6WzQrC2or_KL14Y6qQv zfi1lf75%AV75E~T`vQJTGX3w>%&{9M7w3qFBjKBQR$iT(7zO*ohy0M+a!;PvoWbP& zGmR=#jmEvJ>2ni@u)Qh;Z*h^9x_2{w3MctXRwW{Z%rsb+~I{(Ck_L5>ix*t968V$6M*;m8Et+2*lH~s`&F3LF?X3f)v#&-9wOD?Y;)K|s1 z!RZL@sq07pWO3dHQcJ=>SoPqIVm>Y3~{hH zo1VgQmcrEXi|5ht>h{(?lsccegr~d?w17{ue#U-hjql%Cm~kJ zmTdlCVF*f>0J| zw5`L*M+P51U?|ZYJkV6Cg5l3kHhqk%(?ullDapx@%68c|p}H_@?oQ#2pWClyi9ezF zCVgx*^O}+qXB?j;t%5D_Id~Lprdm7db3=kF!jTM78^`8tTrbir=TKBl&QfEkiu?y6 z3HtT$o#VK8v^2HCp_QHpnfTJ>+fKj4!G`)fqbejfYktUmSYfsxxzRdZp-@|8Y2`+x z+w}1v0of>KulSPOGn6r;c$$Z*?a7l*KNHUVDSRCr(MUvuoq|lK0d%4lM$y*E zdmUuVwTVe>dMT9RQAcS)DlgiSye8auVyzKW!ITs=-~q$QCI?;8*Xv5mQ4D(YWH$am zfo6y!aZ*>L*A{uE<<@lTHXm@Y#<0^yABL?e&Tf8!;viRun~nDDZQ)@+|490FQ96hK zwsuPOOp-swWmOda^ctA7QM!?5iLwXPY~!mb5yED20J@h`yf#~J=NN{S| z=qHOFDR42LWMc10#zRd_tkSx7pXBYATx06?YFkjzyvA3dE|^^P z^f)VfUiYCO*u@hGXi<2{MNAyH3Tf$g7O1-G9-184>3!89uPrEPL;c_np9@76jxmXg zWKa`toNPVC66tu;8snXhhA3)^S!N?=vOl4tF`)PPkg3wp#B*E1o6K7(rRK$JTWqEt zjyYg{cIYxTZ=(yo_wi=@Ta|d)mw*0;U*k{f^Is!3Xp;JkI7WLDE!N#;) zC6yrOrjSi^F`_pQ!hS{1cE46^hzwabLacd}@%4e+qcZ(OSaWq3Vn4n&(J}GmuJl5D z?%?z5Q#m}s&C?WZf0t&9`s)*BeQQ2IF-uOiPL~(pXZC0$=K&qC-$tWjEKmmVdzVGB zSk&2e9%zu#%|F62s9p5b_#nzSK+@t0xkxYdqsD=K%~D`fjOPWsh#rYSKPYs=h8!!?+sH7{T8-~lMQV`i)GO* zRS=gm9$4-xW2+QcmbGlU{#QX(QP?2nd$Cqxc!ZG=kGp%%+4dK0slXds-_k6caNP*M z)76}mVak?G?K#p3652w-sfWbtXpDTbWQ+T*@j}(^)ky)+3D+%gs@}4igf8rxarg>3=+r$I&%X7YJc-;8*fP;2p@jdrv)pHHC%kF0>cIP}E zOT1_{5wYF!GIO*fZ`rTmIn&XoEB;8fe^$Y49Ex07-Qz9rU}E9kJM zUR%5;7uOBtN)$onv&k@#X)@9>LWR28+~)A8Bvih^&Cre`uOQxLppeI#__Q)-T<6!Sc)+8!-+pI+EX<@7*45`}CZUkD&lP8IGL z&k`4I67WtWkGVh2ARTVdA37}adqee8S)$jmI@%AAU1A#EFJnZQgok;Z-O zv+(rV5Q{ES@7~MJQ5mjrfJJey4`vgvs26Ue7d`OertHug(kxBksy#V?;0r5q1YeK! zOEwBhi%47W%`sY^d7W@)>@`~95~Aon!=7?+48AqM8e@(eh|$PeDkW#kenLqf#T5Dt zY4LFL2TOmD>)GzJ1Hby{=X&7pMrzDi!ODxFK}KBr%1X>r8Pg;y8G4Bn^wrO{EW74X zgTZ-~stFB8+~eRm#9E(OazlSlO2vHEe3qRD9w!_bWr6;+jnuPST0w87;3V(IRR#aQ$MA(?((|2#f0+XfjKGjMmS&uYeD6F3EMqZZVE=?0Ujf>Im2A z14k!+yny9=E{t>Lj7%+gOjBe-gW62w>ppU)sqis&UAa@a-C90GQ6!%u&3J&@%ne1d zX1Z!!+i{+J8P)Jd)gyr#4~FACi6kz<=O>m?F_Mh0SXVIE!RuK8k^?l`ORdSq2N|TR zqr4?{={JEc;bDb0QBVPGpP{7&qEYaW>G<#jJ)V<+-4xT2&G!#D=%b}er|T_I{24P= zR-Yx?D>gM&m0F)JY~4MH7rMH+Q^G>k;5Z{$k5uBelX%m`+kzoiE*0&3i(!1Q6xL=c zgs#6k(CD@e=m>ke*Q8bfiJkCBMqom)5z!ncoruaSq>Hg* z^^Y(pC4DapNnslZ23HZ42Pz~rXR)YgR}{?Kx)NPbV^g)~Kk7Xf?!I}XY7#UoA&%f& z7DcoZ6mFapRk^%DYZk*3+aROgdO?+s7EqD5QHM?NER;kdP+TYbf#@(P8_2Dsn%r~> zNm2D9kVlp6_~X-nCNtaJb6Z!-m1O3o>Ws=l!U1=d3P||kH$4tb9n7k~dsl@Hp$=cF zpDsf1-%C=#T&gK{Tq&OWI-dkN8$2kKOEy=gairO21_cG4<#YjWQp8N+po#&P>Xx1f z=v`1w9Sg3gDhc>42dPNz=fZe-FL!aZ+49q5g}%GGs@l>tVF3(VK+Kkw5~1Ktt6*(O z7bG@Lrb1FuTsy}9qX}qW^cPvKwGUw%5fjc1=D^%IJs;OWN)XE;2pdl+h;G4(=q%dM zkn!dA85=cW0sw=&GA!Fwm>4p7*dI^9DYpTsft^dAQ+=t z?6N}DlbM3T^D(%=;Zs!ug|cB+lTjgPvgGXy=^Cm2E(VHqss_NsM-o=FoPV(8BiA91 zX1)UpG4DKg4SAUdbhvweRUWe0x4kIAvY(aS`BH(B|G&kW%g+hk$l`2v%qB%)&n zt_@*4nSmZbuEPLTlxl)eF3XJf4XzwNLGM0qbPuhHo8k?sjFG|GIvE0#%0=3Ml6#w4 zR#jSSrtN8_^^wYGhrs30yT(Q2t)pv7{?lVZ#@CXN?y=@zuDu52h^K8kbo`J_PvYcf zl4QqZn^WDQ-^LDKKHe;!!EkJjEr)lUkY&b;c~G4874d?3;!>x_Oe`akKE;V=3g_y6 z)RIjV>K*KlmeYGO7s<5-Kfk+mARG+F$oIB9t+C;{exdu_$dlzCx~W$jpd6*`7MJT0 z3+|5c<$BGiU!cb_ifrXP2u+T#72e45k^SSb=iN_ZASvzKKI8XO`mMp^AeguBLLSpY zI1+ldKa#rPsE3X%R42bKrrir&p$^Jj@8SL2Nuo$vjt}f!(WdWQd|U2U{8$;ERY3^6 zFKk3r{is5*@BXcohTH$c+E)O@)n)5KZ~_E(ch}&7KyY_=cXtR9f;)|CaCf&raECyE z;4XpS7Ci4XnSW;P%zbs|)_Zljit28<&*`=I+WTAI`qny&Z=FA6zr2hDK^w>>g4P5HuvH$BQk)NLCPDVG?bo+M@(t{QCZWrKXU4!RQjPKy3P>>d~uj zsB3|G%NJbv_ogta@sZv(p{uQ?`jLv!G1v0JAqbyi=o{C(-_wuD&j*ec!M9bwcrdmG?&kI>0NwQa(8ax)O)vFiyMidsruMVmVafaBaJW2xbvWb_G3#5*fzEEfS z_?}wPLY}EvdX<2eI2jEQ;*Dv(0A3U{ut)z?3B&)%_tQa+^C5U_(=NQgauz4a=?az= zlSos4kx)nj@4Jq-)ydm(o?!|K##yJuC%sH_pZ5t5r#A$88Nx?QLoa&sB6axJrQ0@BciIE%SxiLVd579VgZgz zt*7&djLhQd0^#c_wkQLHC%g3Biw@?7Io}OeP&2z(3F~X>uUtq0kC;D!t<}#`t~Z1y z1AsJQRBKCf9|J7(t!8+eAw@hd-B6@Bi6~1_l~gDwgUK#2qOo}K+NJLGa}xaE?H9}X z>_MRFV{4Ij93L$9k4b#CA|x$-$GDJ7y!3Lv%DEt}2CTgolSDNC2^0ps@R^bx zZ3r?~SLW{JvQD6W$_8d>@{ZfPY!D}II6}1&9pf}snzrPcSh=WX^X2t|vg87|AcMlR zyVEl&#-)Ax(-wS)Qx;YQw37a|wgiT3NrKgc>JYnpkux7%hmjX$p-MCZZXwEMw;7+c zkcRR{Nkki|0B-1xxEgu-*A$Ab`_ytd_~pi_^_r-aq9PimjN?XOACjsm%?MJzD1VQF z4zq8~MnYhO@jxHuuD>VqmqGzg*Qe5;Sg`B3EX8|LD z_q{Qt58Ju$ol-P8Ek>j$;{p52|W@(D$rin~8sa1NoH=vj=DS3zRv$|KXo?f7VM@H0X|E zDOxb+rU}dE{rZni@ZX#{KCG^{bh+t&x-{UE$<>(O*K-(be^#D~_#t?BXT+{v*U~LZ z`SB~wY4L3#3pin~b)Rdk-@Z9@&GzI|ctb?A$E`>J) zG}*}i`5W4$ljxcHwX}VQ<;n^$^cg>j{r(r9{&T)aHA>vEoBzeBZ35$>?1=Cesfx>i zFm=JB{&C5eDez)J~ z1JV)lV;1rrTX^-`0IhlF2hou-#|S*oX94B&w`Ac5%p~3m%)bB}J70|7d$t&V&1`sc zr^B7)&KOQe^&29)KAc_x!L0&W)dhEa+9BP|?bW`o>GyMNxmMQ$@|;B}?UT|OgXX*8 zleZTFS@yH`C;kjk3uMHh^Z*6LV(Y~le-^ztz2mq+9p=uU1+&z2W=F($Y;=IU_+7Kf zvSH+&im>eQoB1$@LxMeKy_}v-(bt4`?D9eoxi6oFmo@J3WNTXaHAAAqhRY`TClq9~ z?&Uy$Ts@8}<+Aar0Dz6uCa3E$kHHG|1zuDsQ&-R)M$vtg%ukg-mLSXlibKDXF*}!Z z?W;V8nrdo?v~aPEYDeq{)fy1EaQ2?)P+}exz9lIoG2VS82k=AWwLg35dxR0FA1zr< zzaJcsOy3PxZ&R#NyfQ19QZDVd(vG|aC=!9aCy8&-X-tR8>@C9|J-59Is1q#fSV!Jy zq`wn`A2`ZSNzfOU!fkkU6uX@um1QrGPgSUEn*!ojjEtt#(QT4P(&6jFeJ61s0^*QH zPA|zfR6QtWuNfR*x6FVAl%fyA_olIK>S@LEjnn47Nlw2Zo^-IkJtcR7_s;F8qRAM{ zj8;yHVaCqb5)&sQnLbj2w&uyT9Usg*c?vTX5R#TLx71Oi+C)@+E;fhx7h=*M0@;>B z`wIy?HQosbG0CM~BpF*F%qKEV_j3b=Zd6@h6I>-b@SS&szikkgK^HOoc$u%mTi5p9 zmALFhkLgALe{Z&~Z=l?4VV#8Z7&Uh6v?mdh$en>YATILrp`6&&bT^?mxWYMls`Ht= zCmC-tEUuRtHNoATDmjXe7iX%Eq$nP=zf#2J5J$Q$u21dZ>@e1yPCyt4WDh4j>i@Rz z;YGXWHe`}DX97$C;UOqqo~=UEISfV6JJCh@8>e|e-T954+&0ZU4szGRr0vJw5Ok}R z9qWPbi;brnNme*7{V%lLTYp1DmsbMZ;VlM<-xn5a%#Js0@!QD;65n`RYM}fe97CnV zB;#dB=eK|i&MqjN;>?KI$?1IpI?`PbukZq1-_9pEq*ISG^|rfkWerg6CKk^bVDWxW&(L9 zl=P!ux`ez)gk`aLuS@EBJIbAzfV7B9Q<90$N12F7ZxHSgzChi8Dkd9d z&}27GM(^BGLytC^^&7&Jt@3l`0Ep7r2{{p@V*Imc7Iy|S!L3fuWzqbrFxSxNkcSP! z8-I7M8d?jOBzH74CD|A3Z>bHCY*pOwzvsK5*zx*cV1NF}5!ZEVVCR3Q`HVnf(N;pZ zv3n9CA-G-stTmDYn_rM`lus}Bio%iRPv*SK5KZ5ISVCT3FG~el3w-C-J%WPpIP@nD z%J50$i%JJytFUaf|FFGJ*(TR(fWre0EZjkV;uv63f~4-8qB2J+Dck5x%!2X>FBYfK zPr8v;BQ}0@WH}-m<8i}A{f2m+tk^zq?pU1tBBVtUT^uHwV{ol5U9s<^J{Oet&4VE> znhHY>`6q#RB?1g8ez+k3nZ!|VO5~UL%M2T)!n+I3cp&Bg_l}lkJAN!B5$q zqlCi$DOJ7AqE*u6aD^QSKWJ=&Fcz3PAUe)&s4ugEE7lmrBGAggT?d9rq(vG5KExX6 z48zqVgh3`g4E?Y@-Q>+sofM+^OIcz$a(!!t{tLvmP2*86+B2l@B(@ZOVc?uhNwQqb zao@|JjP_O6!SnKsNQ)0e6*9rZni-xfPL=zQkl0r5bIt**|ASL>Zy^6QYQ`P@^DO@Q zZ2k9+g&POD#1YrfVAH9NA})KHai2f1mN_gqS9=_K@oRqL)GfoQx!x3_FW_yxLT4Rr z1zTNAHqhz`lP*NSF9fi3f)Zgkx-c*l!m=ScE*&=;k?miX{jL&5J}5nefakP)2M$L` zDL@^T#$s+YTrwwhKA7vzu^y$syd;dM#*TGWNAWs26SpH@e}EQ~?~#Mr7E!VueYS6j zJY0t&Fa)S7q=~;FXsUVbBdc2U3&}t40UHNS1$es%_XylI13MiIn;O5&r+53huiFs9 z;)dwrD`&my#rw_ItbnN5wlPkpSt<`@TitJSO%Gne__T7U8w0U~#7~S9JL+7_WV9Vj zcuF7r@y86ix~#8TGX2|5wYn>bkvBI*-Yu9T*0DgyHeWnUD5^bDtePxhXQ zL6&N4PXRLSdwk3xahGw)W6plUoD%1<~8`!T@Ko-qm z98nw%NJk=^LihUuRqm~ed)B-qOemYYz|RMY0@O!N@g($e_N+2J`9pam{(ivm!0~W= z1V7+-B)CyVV^`14V%WRlW@2kyLVLezFQg0hfa4@%A8Gl|b@X2b#2V7|5OA^C^L8VG ztmDeq)G}@^O_w<0v_58A$SHdVC1;nYptn+L{@^=jl9`@)Tev-=xkZGjhWw*6tTd@X z)caTX;Hjf9{NiFC1R?=7kTo~Yd+eIg(#GfW{bM}UV1`g}amb*YLYZjU@pSTWfvT0f z?pEzeBko=-_lEOmyu9v?`;gLFA|CIs<`QWqk?$BWgeh@q91;?RxZg}c#Z42G@u-pm zFOpI7Nn64&X}%cn_&!ZIn11Y$U~3SjA0YeNi9}D|C;W!U%X%Il!UujyxfLeKa~YQW zUA7iS2|oyjo|Mo^lK^?&gqA%e2J#}&e&z=UrLZLX3pb4LGO>hR+7o0D8&TU--P26j z59%FMHSM8*6BOV3I;-08r~*)&r5Y^0NtBDaX$rL0G9MqTgEg$lzrRPH9 z&1D#+w_(|L^@UwH=v<({iu%+HlE@`Pcj1Xjra1E)#)BQtu0-9+GQ1`rvh>0B1$Iz< zsT8|KHUovVhQ7SD#3{~D8ARInHa(HdQZ#~>1WSklZsX>!KtSjM-$on%XSf{s!YJu#8eih_J1qTP%#5_ggO_1I|L z%4>7EpEZTSvBR}^I1mtupIAGH#I&g4zc0~d2msdJ&gykt$z>389~RA9gUqu*QO7fl zK;dQd&wKWh{>=NPxQsUM=apy#sGA&w;NFnw{ehCEMn}F-Wa5|a_4uHSG@fL|s%Q|z zy1gx|KP7N^7q=AxcIFTcE467$zDLP*#o^|M$7aic3;iS(Li*wW*~l|^WU@w zjO7JSrS^WM3%b3F2!L~YEQJ8C9>0xg(?NZUpLWzYe@PF>VAjl{vYw zHX7S;n6Mq<7{F44_EUxnAH__w+>TPz-B1&X(0fhxfFSG-3;~1Q{Czc)U3NO2X@U1q zF6M{osXA0tzMT3c`+WMnKXOUka$?Wd( zW39*Z-=WK`N-&O(xkj+t%@RVa1+%wmFK`sDXCwMb|R z9m!z|?2ojwnXSW8N|u%_IYuKlTYQH>(18E3n(u3Dj3V|@ZS&VgiTb;<^NysE!dnO% zqPgRIdGX0C#U@mi&6w37so*)T9| z7U>}jGbJGMY%LY;6?j5UNsnJ=20FjDROH^jUbmI9=FqOxYn12C2Z-RGoPPp`M=VTwx)bg3 zusrd7sSywlMP5w$Egil2fmlB_PX}l+2}_ernzL${7jtt#jUw^=mWjbH3_RYaZikR} z{xu%Hk1HRZ>^{Xw1knRs74)|oQex`j??v0{0P-vpZ3B|%9$volIZphu>0?|T;CKR- z%-qLv_K+!wGJ5@S$7?%JC2oLykRjajlP&JBfiWUjPP;W3x;U>^_*wz~0sULh-wcp^jv`eye z)qCwMQE5sU7bSV!pTuE9A|GSibvlE@{8W% zV)yS3>0c4<`|Iy`hF3CbOm^r?8$Xa$hAOnZEI^zmg=D{mD?Y=MqQ zvp}Vr?55lK*rsrut*T;4v4QL;sHe3%-xYrL<;gKnO+b*72s-e4dfyQj)9*@R2mO^&vLzs-_lMmTYT7>zk`n@zLdV zr>Eh;S#v!^#KdK>C4X-WT!4uY%m>6;;yY}Hu^{bQ{9}{Sf10`v=gQ&Y)FVu4%cenj zAx?PMiN;=3)Wr^XxS7WML;}L`Z=`3$q;@SE!%fQ~?V3l^*xwDHBXT~J?ii2)5hANA z(`-8Ng#pEHhy~f5U%#A3a@z|5mB9OChuhj6cloJG$HzjSXDgdz`1RmKem&s*HNEhN z3;RIFC*@nP15`T}Vt zsaJqD@fm>5fz!I_K!!UhI{aFUZ8>lng<@+e$^pbKU9!o=N!!3BBMG95s6*`3gJHwi z)>*Tp9@ZikV=ooYny$~dPkTm`tW}60H}^ekmst4`L4kM*O~ArVgSSokS_TjGgtIIR zSKtgJVr~;7K3d z+5eIY4Tj-?X~fM?q4C!BdQ=PZTBKK&ob9?t&Xd@dX|Ae$3h~By$7DSDlPJPb?g#<4 zz&>>$;#a4K1@$Bq`=Hw@mEpr1@5E7Xv2vuf#Lb?GfjUP`FR|zHK%eAIaE_wj|1clOz+L+onm`DeLVm|;afPR8S zs()WS7>30%Tce|-*%;Lpza6}`jKscq60<-t$&Zh+YgW=gC;gMGL8;oyz(Q-4$IMYq z%`u)3rGudajFK9uBG(16jPM(xzQ(rm=~$y{7{y?<@j~rQ*DN@|C5!GMfY(R;i3`MF z{q;(~ttq8V?%02xtYqa_*kqOL;qIYc0Gvm>**AOnM~lyJVXX^Y+%Zll&qvG~>!?7Y z%qT&Hs-y_WKMW4ttXF|N5C`f(&`9WLmLp9r#m69iX_Yh9Sv0xapgG)~ zAX_x0>S5|pPMM*HuLdep8XJpA=(J^bxpZZXiOaJMJwB~WZ{o-{qcwnGSQ6$-&@C1` zDpY{rwnM)Fh9Ki2a^^r{{<9gt_ry|W)Iij8u{QZdb|xM0Oy(*HK1k){2~?Xl`f1d7 zoDH&-53<}zrgO>By|!4SJ!VG3c40It^<8y$bwf0lRf?6NqEtJ|&j&@?rH5XAMvJ%B z_{xqgc2qMPws`v)`mc}~U|7#36_RdK$7)Oi*_pnxpeMhVA}7xl^sE2QNHG;a89?86oWoby18zja(D#1uws)Ic- zv_TTHV2WwDGrb+_=x|xj3r4fm58m|fkR*IgCDXE+xwuoVk~?@eNemaR`(J7`+_egP z^12E^jD}+U4^w@{w+PfJET6k6jq`_Gj^`n}5ao%Dz_{d-yW_c&1&)Nf;~t9R?y=rN zZzgX$C&i(D-T0nf{d6}<@8ZOcs%jtBN_@3IDRbx`O-blcBpPXWz6}e}2L0oFIF>eo zxXLhdN0d-S7g_<4*a*55*Qb5kY5&_|cj@7*+Pp0U+?a8|ZJbbAWw&BS zW0(f@r5S6fnjqD-n!jD>><9LM$)nDv!`Rz|%-QPO`rS&us5>x-`KY*rs89A2Pp*La zDE&DSo5FxFEDt1qvrLhdx@VRZXJL-_1+Rs}eQEDxvz_G_hAZq7stV$E7zoed9K3ie zv~xRG!_1ehderNf44Gi{y&u-C*l3gg%I%=j%l*bGwNvWCIU~{%wMHS%S{a&R7DoaI z&Bk>fvxA}@Y^6PjB+nSa+~p|pn`~HF)fFTThX+2>eYxiZNi5jltEo8&M%!Z2H8J7J!QragQb{CG9|%xU za)zWrRd8;zK$V?Dqw+QUs9=Dis*wdneOX4R^LD0IW?&VgT$WQ%MR5%_JF2}j$yYw`5CpQG zhV-Na@f8a`ej=pfaX)xYx$vYy!3Yy8(#E{dL z5LOzV5k^KD;UPH7+Ka!S#iNP>HKo@kAAkI-=9YwYRdj(2e^rje-VG**SB-osQltd> zD1=Xpj#9d4FdE(Od!5#%wID`u=^lNey!GTkZ`x(M^+~HNjyY*H1!}xJ z5%LY`7%MXC8@cW^Uf`YypYCQhvybiqN5{^~tn0(pO`?TO_uOoTs4q=DCwQr$rJpOB z3>fQ_sf~!w#6wLxISf#R(VT_>LU!nE6YWTUjKQu5(HD_2cz-d`$2D(US{rb=O~zUL(bx=@)h z4VR(Fj%UgcA4Tmn;=bzuG$SL|BeSI&jPEpFR6vFlB_nrk0duVLZ9+%x_;%gk|!1li70Q=R1}H^{d|)i zV-_Jbz}HXy;aJN=hVZJbgM4^^!vNw$UVqg!ZI!)$;?7g|>=lZBaR#r$=1@T2;nYiG zhYY50*~U2K%!=)Amh=~hOY$~BQ~67Ow4mZ`*NJ>C21EMwN`7lUnV4g@hK4}3L_nqM zbV_cEZJ81}%xhVK(kt!PVjt}@XL614RE@{%E#y^97a{9{Z0q{$qrEWFnyGVINhMu0 z{m&=bI8`xgxR49*`D$@YQ-0* z#<7HbfUb8x4v&H9W{nlt`*jzaw@2h%!Q)0m%d85y5-j+TENViy7M37gVETb_!FkqiaJ0ncRde1m-!D@BB}R;gI_#fa#|eTEv)=tEzzGQ|;L45R@Xb>5dI-)m zAj-CZG>K`r z0JJ?o+{eMKR?Gjz8|su&eQNQ+k22dz5=*Qj!c?(MksWxE7{HC!75e(qy_77FG7|99 zF}O`evG@&fYpdIACebGervK@ZE>5`Rfjsjm4bfJ3sG>wDwC5k&XFS{fI!*2&7{EoY zQNY4_hb{J{EPsbN_ZH`&jygIj-mW5>I!NMTH;=t>`i?dawEFMWU4GKEfE78 zmhD=m7|tn%=TO|WwGzWz9PesDidC#-=J`#>=`*%rP5z^*=O@g$19-iZ&*_elJ91mu zxB3zJt18?qpVwm!(N=jNTLd5l1r%ZsJ*iOV;ZVOQ!WH)6l~SW@9EkwwCkcMx3AXWF z!VNA*l=q$BP$gqnUFmky-P*!ym?bM}@phR;1EWHW+8MPp1;;T#6_EXFt(-fY1u@Wm zlaemPR|8!eS}*IE;G!=8{Y*;-Yp)xfFf$f5(GvC9^x8r%n{?_o=7-r1bDGSG=)+UD}EjgIHK+Vg29b%&xAOTkrO1m zdjFqUN^nq1V0Who2KBp5FP@p}2&>l6S-K9rKH$G+7N2SwIfT1nId_vE@AEm-19APz`PjcXB}VuL`ZrEY%yM|~rC#NGsMmTIyRkU2RKkHQ&y#;g$X8&SBd3hia&KWN^ zqL)#&3-rO9`(+avglo-#K3Pw$OZE-uHA4k-V*U+5u7;=hT27WAR{1aI5J{n97b2)M zL4}|NnBZ&p4NGHH#!P66Q_-@Y;?&u2NOSKP=SFpR6ZN`=FF z*jlP0{4B+eS||b3d*1Cp~#BcGs}cbPq-!y2iTgM%P2!yq;^)!OzIZrg^q*$3v;S7 zS+9%FYfPCOk$cWOIxi}5K4W6o%X`#ErP3XNsxNx2TKhoc2!}UDpQ`yx<1L_whwJK5 z3Qb?)_FJAwb&iY0EPG$Zx{dBcxbYo|q;}6BHmZgb;{bSV?k}ud0Pul$?uK;C=^JE} zLB`1W>Tym@K9e^tN$e2#(}obVZYa)0qVP6B-#NaA4>kmv=ewt&Jq<`Cb=@|}oE%@M zb?v>{2jX@C&zSXj#bqKiVpO!XnxM^nATzB^5F8itSV(=GJKkj(vz}a+wK+t9-@7h*I=$2RU3UWl?t!>NU-dqOb#Pti)^veL%7 zo$;lTgRk(ojPW@!DurTL?vex_g3vrkuScXKRr&#HGS3i+Jp-JmWnj#3#%kS51cJr| z@E(2qO0>GKi$zu*GO)`1W5%r~oRvIb8vuO6LQZA9nF-O#&^bnB7$Z^_j7SFsuM3vv zDey`9tje<&SUY_%+is15St2up#Ot}GE)z6o?svW|A8RN$CAw(N+!DVL`{-}(WZaFp zn6tp4nbPOKJPzFLJBy{xkN-0{+lSe@Z&p`s8w|rms!~+wi7~pL_|~ry{j!cJ^Fd!Tg5;dO3KxYdW0y2ppKmJ#gGa*v0-xr^#GE+TNuC><6IR+e-Bnx8} z$!}ko#f5ZVzoc^`^Pns(Y9EmaN|Kt!hHuQ(dnZf9MwDQ_9uma@Z9lAKWa+*w4^1#x z=M&6I$nDTmJmp2b5|&TgP)zmhMDyJViFg&w`6Y}{x1x*RlS0f3ldNP%2s`wE^ zY{DaZWFeKcMK62O3_XcXu6w3xXzXqo9E>B`^5eZ@|Hd|bfKWIR)}Xj^$n;%j~CwkT%=?$-cAuVZzaP>y*1>T;f*Lpzx6fwr%=XnNSn&Z zFKZ2iiOGD3;$q1>srjOrxD3efYUFFw=^;tv1%CI9;;!-r=&#q+Cka1b)v${RWl0`T zQD05G#KZZ~7-K)KFB5krw;>WFfZvkBnh8W;J1@f4sRD#hqCii>BqwbOwM3PDZ(6@*n`)-5<~4*2F!A{_*y8#`U@{Vm9XhanrAoJ8905(){Ax~3IWI#r z)-HY9`-}@&UzUj;2y1ADzIV=Ke!Gy67+_^IK{2ST2M9|qi?`?PmtZ0A2Z&`Ns0_a^ z$eWNYyo2&Kf%N-FMwq+ObAJ8G75VDnGYV97ckeJl%W_VLN>kk7Ap-}M^yyqx2Ljw; z5eN>Uw+J^0w+LaJulc|4szJ}etdnDRSBUL}Dk2@54QKb$%D#{baj!3!I7_1TSPC?p z{|Z4MYURXFDV>*4pfUO}*9JB)HKHY=rGK8hhF%H0wIdz5G!D+ej>+mMlAxOr9?4Ie z#aL6`F_$ru{bJlPIKSs+Lw7#XUL`9_7V>??O&^pfV+BoUnwb`-%uGo!gkVSw`9qaU zSMV~;-$1gunj1SCTW@y+q>&bFj#Hs2*#S`X9iE=FIIYCw2gN?bvA zz5zP;QeLt3Sr3yq4zN2uOeqCNivTcBuexYKJY#8xWH0o?9BH5(=m-2Ul%Ufab4Ju? zhnS={_mNZGiR{m$fSH~?&HLCzv z=}cT5K!SNL9BPO!Dwa}5Q0|>G9qS+MO_)yb+6p7TAMW|g3P!QHVz0{kjDZDUb(NT| z%s;cLad|zb1g8%!yZ%nH)ALmHgVwH)1>V(7|Kteno3@5y{kZ74u9pYL0j>w^G>+$y zK3&9s6m%Sl&g4nlgYJOflXK}%u9GN+CRSBHOR{u3<;FAwp=3y3SSN?=?Q|!hV%mzm z$^=fSh397(YI~G`MNwnqgX#Ry8u0Uf)VHe<|O5Wt80u`)ZncM6X>nA#9Y?!p0Rx2^m zK|;r5q#<8afI#)d!y`scdLU#wP%5GkyCgxiX45gT@A(V2zaff7YHNmK>7xcV8i2tD zCz29fXUk+W)-=~wn#RQr?w9C2>PxVcDOCaVcvnV!k0QBSqjwQ8hfkma)0n0J(h3e2`& z0fJW;Msj6xPuX`T&Z&yt309aP;I4BnA=*rA9mLus*{#!M(|=_|@YP|Sc6 zeZ`h(ysSK^qZ$o03yb%xh`1=lB|+)tsBOXLT8Sr#&fH1h6Z&M{m} z#Bg40qW1>ICe6?wdBjq)ZO79SHGh;DqAbU5fA6?NaY(Hg&KtFt+4NzPnBqo9Z2R&3 zk0_MYaA@!5GNMlUC0FLLb*Rxt8L%N8avajJ35l5{w_vu*KW4o*bQZarvsf^r0YlKGFdPR zS*!2KR75H@+BUFlQbX=DA#`a&GI2C=>-2huW)sZ3`8IUha&rX0BGMv&W7^$7+C&9n zxJJTN>K>fXLIH{uC8-J;c0{8!DCPc+L_K5m#mserKp?`7-1{*|T3PEBwdy-@MU}}T z&P5r2Z}un45oI^D;kwvdmE)Sv>xM$QN&APbu!E@=5V_N99;z1vuopNlS(!sm_w2rc zxq~Ak-?gpA+XVcP4~RW&6zh`b#F9L~ZfvWpH;WF+GZL1iw<@dPuNwDFSoQxgB$W+DZy1xQS4b5!0L8fAZOkZyeIocI z5Kx76I_dS`JPD|vqclW$0%l;B&FA5{;GYROQXO_u2 z3Vmf%8cB3{WfHWBIqwcq;@2d(qGdWLO>0MaZW4T64lifrBB;VQ58<^Ef zes>ems+%IDwl(f8O!A2{r428fG>&P&qNBafiRI*`1S zp)Nd~xoCSAtOlJ1%J#gs-p?zfBuY5Gcc<3to>pACj%)w;M=%e(=ry&Uh7cmCNf7QncX+P+UjUli^Jnil4vU z+-^*hykWq#mG8ox=_gY`&PrWR6p3eh^rEE37e~qGjq_44ft?MEkS! zxdQ{0rfqDJ0bVj;Z&g?mhv?XHW3A;e_BDhPMLk9q4} zWc#@{D-%exNSoLl;6?hI*#L99^uqR4OmE1>yg`P^@NOmcA(iP_kc{S+SqsZT`Qz^@ zp6Ov+Z_+)l(_|ODmD2^YXLEq6X0VaCaYg#20(&67Ig3UKooGn0N(lzt+*=0N zqa^+MZEMw_ea5usyKrKzVm41a}zh`U0SLF}r+UeCu9(zS%|W2Ib|&FWikr z1WcyB-mCey)%5%E(=3(-7Y;I>uu(zpSXrSw&l%2kG`4C488nr!_!}RRIgD2Y(1^ZR zx$Q)kp#B(C_NK>Sn!{Rd{S>O+#U`0`hUHLT;)H_Q{?Tmo#FtDF3VQQRPBx?vFV?NE zd7*FU91z4z=|L|STqtCu9+d=&kj(#-HQq{sY2EIcNIL+q5xA7)kFp)?Wy6umH8vtR zmYjK~cv0SH!>)$EAts%^bv~U`faA&Dvz~{Jh?pMO3)e)hnb8)mKY1>?C%w}5Yz7%d ze)cLUju(J-YAYB8burhQF2#esW8^{JxkxD+Clxi~EvAxOn-&FgkUg|Y8dy~TXdW)@ z)f+WSRm*5Jtt#SlrGAY4s*}D{p?!Y5P{~Mw+X@6dwrxEaeiDKW!H7!7k^fUUxNKjt{h>DHw zTqQ2MiME1RP|yF=ERdE4^53H07eJq??E z>e}#o)68tn0Q-oM2luzB;sWn_t7#39mjRKRpN&{+k`zLvk?RJ&Z*}jrbziM_Ieu&8;6u(u__#O+36UIauCGCrvj(WUL}C`=I;Zie#wTN?^mb5 zFj&CFAOEOFhwAK3d@o3{{_7(_ zoq9b3Cthj%f4ovE+lo|)zP8w?*pt+U|K!mzT(2en@AmtDeq~Vpt4GOd%Wnu~xF7tg zx3zCm@{Cq_(mdekfO~jF@b5=G#R~GMUT0h}PiKVk3-zL$0b6UafU$+1t#a&k?FN(plMw5@|xy2m4NPY}Fw?^#POaj(EotP}twvMor%ZTve>~}~Nl@tcZ(jc|{{%dVEfoHo^pB7I>)Zd$JENb)i3Pd> z$`{t@uRZ^|W!C?%Cka!tbIjJ5u(Lb9;If?A$)%UX8%bz1q5paz#NW!s;bD0TjR z+phoN8t9R?e?$DQE?c*Z_3-~@-~Yf+|28K7{Fyp;zRua%cY$*a4qxItQj9unX-2>D z95Z9`geDHI)hoS8p8PQ)2XyOB+-UI{y>KmN3S3!6mc?!TNneVp&;>LH;18G#emxS- z#nAXLOu!h#w!$!!>X+FD>}Bbnn5j+|U5=kC7SA140;}M|w#N&e79dp^t-k;~qX0Qj zSV|}}2M{5?KuZ&!zqbILIO|D}*J+l^Bcjvj#v?-NGcn;?1dvF`ICL+Q7C|ut#sI@AlYgNq^6GVeo~P zAUYgn>Sn`FKk#FR#_TA->dJtH&4Kp>+25d#p; zC@Y7%tT7Jwv1Bj%Cjlt4Uw+#6q!%i92mwwPTRsONn`Ky-u~cNP%GfYTzMqOz{dGz0 zS<1nHI~?4E^d67kv>s?eYVQRX`POwjo$aWNWm_>?^txY_qjZ*%l7L@s;n^~I>Usqf z?Dg@TFV#!x9DX(e0{=W^wE_nAM(|5vwGQL2&-4x#DZ8ts-$}vu#oDT=u<#UQCRm8N z$~OY&Mzmv;hgviyz9S2Ezft0p9&SIFsJF(C#~>wB!qEuI^XR!4aylWB6PH^V&zFQZ zvOiE5#rR?Q0z-CP3!Pp3)KyWgPETSw+C*9c9b|6xVxew1*{l9-JeV4^-5$!yRA?w2 zKII}=() zU%%EJdq4os--~?K^Sy{yvG?hm-w#cN=(%!6nS7@ zFwMGO9#89y3Cc9jaOnQfTj*IdFzs!GIaIqkXaQjX@7gFc^vLP>p~G>or}z^R7KdjB zBpc9baroZIME9NQSjMDcZWARj7_aRY{%H-C3oC%2MTMcx&dD?INT>HW|EHZwMC#y= zyEect(*NEzXZjy-MJy147rX5)hymCP1nho22B9_dq;s!!z*3EYUvJFyzb4^Y7+(=tn_C z!`OR}$~5L(0P|oCqKGJm3bs{-;FcSY!eH^DXMp`PT9@3PX3R+2jy^Tc1G;PP+YY@1xN}6GIPPDn(@^EevlRQb<8w+y+NS|Jvl;0)Wl11Q7ta)R`vrFW03e*F z-^Mv$e|wC+(PVYN{(d@uyit?ifqAoOfMOD}xdG*92v$agBFZyDtPDSgq`sVjcEPtz zLy}DqR=pJd1B?U%oZGxyvco|#uWa~}lmy9(DCpPGjzigT7wQv*`{|;VH39o>*vGv;bOEQ6p@cK5YLwz+;dYgf%g%q{^HxGdDC&q1sNz^&h<(D3q zRY*qpXAOOa^gL~I20EVoy{_n_?jhSw#tY}wz?5HR3KRd3MS-(#0q=_eXiWk0Z8Xi= z-aO!pGAXk(t9A0FFpec76m{(l5Hky$TgZ1ns6C#OhEQr3E)Jw5c~#EN;&<= z&Sw~^GD|LZz2Y2ruZYuk%H)?k`08x1A<3xsQtoN}YHs_TI z!HIvUab;*FOG`2}eibOMf9;|E?HaOOq#X)PKYaRG=j%uV*?-9y01Vq&LI|EK`G^^XseE$bX=j2lN1F<}tX$Eo9w|kVsmBaFWc+R2R&zt#%@y1!2&F z0|;4HgyaDT76!Jrg^g}G)H67OiC&RfU*qZ2Ea%XK7TTB``p7u!lWB{@j}orOG^5k6 zg9v*@6hMOw2HJ5x!A1%giMTc-U+Q7qKZkZUcpcPKh*XN1KjvkE4u~RsgQJnr8Xu$7 zkAIPCzvvStWyTn*6iP==I-$u*E7Ezsj-lQw#$Gbh6b_Y&Y;eJyC(uv5zZlMEr=htq zs0z8H&~RVCRn>U>i{!ABkHf*BP;>$7U+YOgnp2J|&D`vOaK1u6tuDZpuo%F&xnU8? z3W^Z;Gx@L9_Ov>`F$hHJNe?cU!MwKwNmX@RkOUBJ=Q8TCxqGQ(2l=>N2ZnaoKQ>!ce>9;29bnWboPP9g)b00u)Nx&Wp#O%wn&hQ?=%?EysroYL0O3E*2r}{IbNTKHOGv5fBFS zhNPR9;S*R-FD3O!G0H&|%dz0-->%J6YJ0}75NRwD9)qv?lmp42|DOx|S6>58Qs@h? z)iGmk_s#j~JRDmH*%#%Cc^ zT3!GzzR9PorabE&h@87Y{3KgWv#x{VTftghw^&@+?^~;30>=+mtl~MYHtG7_G~($~ z;5q7wAhR@Yka>`uGNR&wusyCzBoLGI%z@I+jxuP~9+eAtM>VE*PBN%~sKN`9R6_chfwt?xmWy3?hNS4N*MiPc@!n1W?I zt?+v7+;nt|HTtx#_;0ZAq=pE~`&T@Y)Z7wH^-pu+cMRI(KiZcOxoI{Wla(ZO#Oq~u zs1bT*E4qR0>%rbW!r;4~8^M09Nf`8c3^2&@+6K7<(@W7;#gA^`BxTvSq5Rf?Q6k+J9poqp}= z8vdD$#Uh7w2Rkrh4>YZjEO#PHc@kaj0lSi3eBjg?0{LM%N!#wsUC%s2xbnqV(aS4g zovK{Y>D)Hf>Wv*A1h&{bgE|U$*W}_cJqJF^pK=P**fdo-9mJn6OnYE?BWgCiS2#%{ zFFKL7Dggm~ROFHuXu77gln5P~nwCpzb~-HRrZXuL+{IDt?u6O3QpDN9X7Xk z0riQ0m>zbat=o7Jv$@m^n+fK})~7W<8vIW*=D)R%4QHV_lD zoz_23NPJHZfm|_8z_h}XkqLo`KdfdM+ld)~{<*JU_TYeC`+!03<3#{Wp;19YQ(e$h^X0;G#v%Xa?sxD%FN(y{p?$DdC#p)z_#N>D`3OcA==fK$gi##F&^}WoE z11r%XviS{&5W~&^9MTpReBXOJz^oUV00%-g%M|F@jDzR9^-`_I;ThF`I%E4A2Ahm1%?yK_{ zv!&{p<&9`I){BoIQL9L8RxjV6$6xN+>PM9tu}1MgFCAbax0Udw#R-G|fFf&jAMCk` z1~8p|jShg;G4Qn|i!^4sM?*Wtt$6jgoPpiQ8!th?@>>GP)O)UnwOTCCG^c`HAdKF* zxCm8BD+D1sx>EM&7=VQVX2DW%E`Vxp#Awm|p{9cMY4q1G=kjLj80CL3e>*Z$Q_e+P zXn`14R}u3U-4KtzHax;6sbu&e$u$%-!^S0IWrx$$*rUdVj z;er9@16v)c=m8Y5e}eVXDp|^FrLk_B>`{wiEesznaa})1c+*L>WYfIDA_^mc6X$7nhF8xMpl$?4pWBLs@u!7xltbl@&92u~6?k8n!TH)Igd^6Xey4{X z?Mv^p8KITQ)rgh&|4CzRns&I7{Rfy>_A4sCN?k*kI)aKzot+o@J0U^8vFfq1?3 zJH@U~0TI*v=jL~xUHIPf1{9AadA#KhRoipuzG~WVTRE@2b=7v~W8e=g9oq}|_phy| zP|A+1-P4O3=Ewi8l8&9;RW@!KNb}EsTTfl}UhkTYlQxd<-~axf7k;Xn`XOkgar}J{ zaF=le1=E0nn(mqfi`;$F|9t9y{_V9nJs3?JO~?2n`@o59o!cZHJz;4J7Zlrj1Y->+w58ey4VSS8U3kq$<94z}k5!w1Iw6MP&D*%i#;|EnMSc}y1XwOPgd zY(}(74|d$3-EDev^rW~R)7l2Uz1UV<_KOZ! z1*pbG#E%IKa9c#^L-00Yv_xi!Cbrk(xm%ljzb7tBZQHmld19-q(xhi%uW5M zT6%MbUDr5%(tqzva~X92Y?u%?ok(3(AD%hD{h8#+38V@73CqyUvX}nH zK-sNAL|$3Ta>9@Y?w=BF5U%jjd|F=6oIg~$`4$J`1)bcTL($FufoNmEcE!3V>fm{H z#$97!wVDBa0kvvJ;}JlQQNWaAAy3E6w9i8AD(+Zuh2R3?p(g681uXusrzlZEOl$ni zq{z6f{Fwc=m$U>u7KV-MPiuuQDay!d64?o>8!0`Ufi{SnKv~1{XXO_=g%QN3>-^blmTTLX#qUtdc-EgKi1Z0CBeO6|22 z@%?h5iQf;%GW-qJ+@2-L>wdZ`!@n|U+VcIKf21t7AphtNQ`qQR4LN&dXS7nICPvs@YFLzSxs8@>aSY>wY!4`x zU^W1(7kD@32R8^1f6pqwG*u+i&sw^ z`>|R+dKQ)QNuJ`Rr&-BHo?)BFXJ(#E+c^!)$0)rdTa4oAOQ#ivE?af@*JyMP^eEB; z3h&xJ-j<^F?0H3o5{7f*@`blmtJ6^Xd&M>1loRIbL&F1@>ym#i1CP;5vjGvGm&W7R zVpDUdyV1{@Rt?Ws59C^ZsEVGfOY!i-rDinz{LFUJPT!b@eB0J7@nF}k3n2J}K7FCH z(#Zd?@=;}XA7f`p5UJ>C85qCwK->SN0E1?*#_MDuvTSXu^gN3cf^smMm}UU!&zIts z;E4biI9nJ%i`pLh@kJNuIW1~|GuE6R|Cii}nmanSt7k$7+l)uiJ|*RnXoJZG&Pmfd zOnp=-?b=)XHswhhb}t;o)buZRlT~|j)JEo(@~cvGFt-$qmwe9+sHA1 z)Gs9S*%`xQK%R&U&J8XL)rm%7hk#RzRt%*(|Ne zzXq4Jt9X>vG>jIhrR-mGFx|1uCB{)YQh$>O59rCWZ066G5`m+;pSmnJ>UhMRdc6`_5^V>+p^dsZ(G2o1>{fJ@f&4+mQ?OJd1=dl# z?gv`657c7@;f;zc9mN;Q-6HYA*Ya7)q#sFtSf*w;&9Z58F426^pc{mr*3?z=jqi;o zKe%RZIm4&fIcZ+Tg)zvU_KjWr`%d2KrHU_Mos_F0mLFQN{Zi22`wk!$*}kL}4A(KL zJgs8Z!+mh@Vf}A~2@Jw~v?DjMC)$e_;mTTN_Zph0YYExeFD-KeCYS^MYZ?$@1h^Jk zj{JdmO7uF0f&uFUMZX3{5GyhTzG*aOttC8xfwx23rxz?qq+m*8YoFh0x*Hc?n#s``E$nkj6>DaU4ze;WL34*rHP^hG%YdECyDq0I zyO!px4m;4@c;i(Q-9Q(N7%*2&Kz72c(luJ{%^3JHwkoMmmpk_{j2C-YKiNV+^-z8( z`zWNMkag}yHd_*oZn(Qy-_t|vb{h7T#!n+7o!)B7-0~7RiD@zl?X^ku@W^2C8!G^G zgqBXVPDuf(q%<|HB`z&z)UdLyv`?5lc2?`n3>o~?lZ#U>n?2_u{WFbR?3X1Iu1WMN zl_@3*VN01>7Hj|ohJLe5~r(`Fpinzjp4yewzNg}&nmtGtAD;fHebzy`A=@ou?^8oen z0a~R{KP6MJ@bEdQPe(dSn5Fi^nzdE*akgV5>mt3@dt!5OjFEIQTZwCAOR*QmQqXnd z`ej=lni7{X7K^pQ?)I#XP`;_TqH)FCl@6u5c_^lhld2o_4gwt3SzChpnQ{8#qG|iB z*58vzM3st4NNJ|#cz&elPLY7r>-vV#D1-KXTYzfpU%cBe062unQn)v# z#7@E_8~K?A%`m3kNHY#dupvP4n_279TXNc_S5$D&KxDwgVJ zR`s+53ZxQ6KwscS#4dK_{bv5Ev+n7f!J7|3fBen-g|wU#Dk)7A64$x-nfvOq+YlV$owfoxeQ+Gvk_nC{Gg+e>=x zJ&joF>>bn`^gff63|2UhlaC#4aL*=!04-M5bvr!S97APj>A?XT_6}0t!ZG#XqFMP= zZCE^n*}8t6l-VhQEAs4Zqz8-O%W`3AUrs0P8dSyJSIhgK+Apry9qCXNJ2%u3Cn@AV zbE)G3CQZY(88f=N?2la5*c67BovKCAgC>)doVbD16Z7KFEk+v81?ms>HL?Z-1BWSt zE@(^Y59Z72gVol19uVP;Zqv0P` z+gY+BBYKE7xz`k=f);~*)y$fgjYh(}A`2H} z5VXVnz1d0`!`dC9o86#bCguebMZ8U4(Ul2OALfS{=u4$j8UIqc{_@}-2~lip@d?77 zGhxd|KMoPPThyKe*~8?r3a(mCdA#FK)I*ZclaTd1r7q>e+)B<-CU9ig22t6MD7Te@ zqo(I}Q`_~ZM&UGq#7ri&v#3^QR!K}GmkI@98q#0T*Z{luQJ{>i*|~@j=BwbdNPVCH9zB$r44E|dEOFj+PQT_YtnTlQVOyAi+=6-x~` zFv?P$PMOK|v0#nGdBGp0Sm^C*aUntYm=T(6z8XI7KADl%yU;ny%TIfrTt3+yp-$m@ zDWguJWF0Vb#~#aWelj!j!oXR}$$jQ5AG>5^m0Ju;y~r34xp@fpyeV$K+lDN!mt^Cn zP-QKd#KrryX&tj@a_dma@W+o50B!hVue)}oB!9(9K~}b94g8qIVO*k%ipALW=Fg%8 z_{ZI-&$xqhla-TX$4M4H%0tbFbfUgE&d}#>SL+Z73Dhl?VoQd`T4IUotg&T%2;8En zs8iKfKY_CdM=$)vdenT5I5VyDEfPj?w?=2XLw$TM(iDo9h;O=i@zI4L^aMi-Lc7c* zR>+)ydWdwDy2v5bC+iK|>38k6U6FDLw)y&NT4ku|MJ~s=`o#y8_`n3#%5ZbcDB&-s zm&EeBoAw)2Vrhv5wZGyXac11=gudi8UvE|^Mip}mYUD>o+7-r{O5Wh@px}6<8Fjok{&p)}i_t$B|B(AcSwe$`sK>KY`H{SYN`23ynHWQgfV#w_uv#KZ(B2J&%08`f znW8A2JilE+aXY-?dgJaGSql2Aro))xcle>P%(DR#bVdL9m)_$k4)J#bKf;y4mNial!i1}LoJ0b>IRKyZ zjCo8zNODPx-T1?zCI&m>xf z$_$PlZmzSgxN%#S;sw2jp1IfbdwxpXX%6w)* zPhuD^+1P-?AIObA2V=wF@9wMsEl1{u6~GJnVu>UZU>KJhS^fVB?>1UnBzYqSJ^@lt( zF}Vz}H32-ngOe`E_2wJNbZ*AzZ?LY=AaX$c3TuNBq}jniuyc^F3n*q91^5Dxx&Zm( zl^+0ZZIfuB>m&gB{)hG;^M=ino^nbF0sxYJkeoqe60Y`yEf0R@-A@ZPiMg*)2{7b- z14%Z-Y63PNUKtSNWW&Z)V=rErhr~Zq*kdIW<*1>C5HPU3LZJM|0qW|2(i#!wi_HrX zpoyL$u)R!8BQFY6F{29-c!~a_>5lGc%m&`i`3Kl6aJ?)!jKeK`_%N*S9fc^eooXAQX699f12m}^}(91Q544e z0}!D9c)^g5v{4IV>vOitk_xzo6!M1Ekpdbgq)gtR?2OR~=No7^3Zrn_0{+mEG;gvq zDQSWJ@&B>kF=>^Pa5%oWY2nu()@HvQ-XfRbn?I&+7g`KLI zXKFt-zg%azfw>p}0IISXd2cQZWO_!jnxqNIM*ey2lT5ae6GRK}&^GH0UVxD{^F$_k zqtV&tv{QqthxY@ZgYM?bdu`unEdVki4ZUSedd3KVi2!ArX7m-6v6Z`y)Vu%Az)R0s z;^A<9{Zgv9EW>=*7?xy-v4IM<*2?gEkkxe;Dn%KS73Ft7C&=9t2c_awD9Xib`y|AV zRi1vd>FP7P#f50rST1DG$6ejL(jYYO^^wnZpWM^10)&tE-?o|TU{o_Wx&Y-4;DF8!alMrE(#G}z(v^HpeLp-}2yn96kH5igOAr$s zdtgHX12EN5-5Knt5(9=DGlx2_xR_!ZqfjT1X-ap(j`2h<8-7jW1arfK2=A^!T}jF2 zAbSA=(3r#dWpt|uUn1kuNOPGTh zh25Q*sKxH525uKL=28&4c}rqjY4w=Nt}>t4#*x@Gw;FZMvykzQ0Tuc`PiFWK=n0?%;qeqc(EI_yk3wlLZjO>@T>MVo3T3Co$*%a*Ds1h zZ@O0%>BE~OWO&0=aw24KLdZ&Z@Coi;_CBw0ofBN%!vM0~s^L=$fN; zj(G(943*lK<6T9wDa!ZbwNEur`l0>~=!>-HxLk*gqLJ(p2X|(N$7^u%;?^Jq{OsUj zlhg+-z{Zb-UE9;gra6}@$C90Kv#R*OCvwZtPITtja>%fzGo%nd`&>HxzGes$i6GhQ zyPVFOul7G~5wqQNR69U4rzP~N=+R$tV^_I@=zrwVzVpwvDULL*mfU=fhH>u0!j=E&yTQ|Y_-3&L&clw`( zXyWT+Za2M-4A5W+ql^ueGF+CKM?sy}*RVJ-;Klm5aiOghzYJrOioe(~;UIc3D!yhB z`3)xVx@~g)S0f=d94_b7y}JI}yI1-ivee6Wde1_LNi8YE9!f0$wC};@=!&`PEf9bY zs!vluQ%`w6-D`jkMvdRA;UK*1(GZvMcA)XCI9|TKX$&&$)AS^Gnem@#SzqL}qOhvl z0AsB#dNMaXh~Dhm1Hif;m{ey0?R?+%8_b29+h@5F6a4x!Z>8$)A42v_`iE0l51w(1 zGzrN!jzgJXN5Uw=Ty0t{#{)?YYuLBOMElP;MxC%5g_pj_<~kbJZeYaM^o0moa;mu@ z=c<(plTn>bg<+M-riFAHbW5>?hexJn+p_>**7+`-<$!S_K|6?kBJu`bsY`I+o0DSB z_is(4Bh`pT*(<$G6?!v%pH4yA=Ps~B)ejaS)BdOskiFWKm#DWv=Tbtyj<>fD#kDa- zY^RiY*dr80cda2qR%CaE65OzKxZU4eH^*o2 zIsj$@tWG8rO#rnf^C~U{;MA{UU)PQo#a4i108xOfKT%o?=T9G4%mC^hYL&eni%Hp3 zc*SEyqCnY#@t|8822@uubK4Sl$0XGsInX0Ab*d`13MfC$UZOg>?5Bv=83!_&~UY0PZjzD3!Az7 z`~-Mu(|b>izwU*InKhSi5aeB-CnLxnxPyh|=+6Rb`6D$BBZoC$h7d~{b zpLxDY$d%L-K0uWFZ?GvHAKM+U{XlS$-(Ud#`tM#aJ~S|Jh&K>W&`{ve&=6q22OR7T z1SA?d88Z|rDU%`wIm-tjTTE7^Pj+7^gblK)Cb1~ld}6wyJ1>4Ofx&`9fP)KyAG!SE zSRU+XO880r*x!`9?(d$UP&;18%JUM6hoIg@n(h?ewBX|u&zCr;-o3CwIEqsFP(<#; zd*1W`VbmATcyh3T7F68;w=0cw0i{4+lu}zGQBVa5W29&XO8>G_=7#+izJc`7LjfxR z3#4^>olZ!Wd_4U_>Q|Ony47;dzXV7aKX?sq0Q4k7^5^meXs5Yl` zTYRL;D=MSrzc8Oy0N+HcvR`m~yU$&ao~2(yE_~i$9uJ?saoakDT{E<`Rg$D^YA%k+ zG{ApubkYCjb&$Y}gy2elVym;ra`al* zDIEwU)Ua5|rUPFrXSsRE@&fc|icywpyF zo6{Axmfx)Sj{KAGDyaSQFw>X9l}yhct@(q-+YT?G8n;pJA8xdl9C*cvge}2_QD3|l zReH6)UvggdE6{Q~#r4@DIKhR3E#q9$!-jv}E1OfnJKjCSz#H>8+AJ@8~?@OKu)EK!oNO2HjdI$!;7b9}7QWVR@!ha;g!@M1@!mytj0s<$0DM z5ZBl&ET$Ij+s9OnPPRihS5Vcqt(lZiJX1xw+Q!43{;Ws}AOkeWT&Swy@KrTKVdhe& z3O`|}rFy1X@c8AI=!tPz)o`%a>4$z=FHkK>sfyfE?O$gp+J0WcZ#*ASm=$+W;zW4o zO3ygB>O>I7+4;U-iN+Z1i{)^{kgp7Zo5YA=S%4$m?VrTscRpXu8-;hZRpxP<4}Fe5}? z{gyOMWaS&K<@@BMe5-^5B>`xcc*QN+8E!BPy(?5p{l~FVimjKd)F@Db&zmA#YkaMk zYgdr*incQPa4)Qmc=#w6(Sb&u-<`a(55|S|9r}dMDpbs*#Y8bCa^hT|YTU&^40G>Sa}&o%76nn(Pfx>i zV42X3pLs^;a-$q~M-1U?f!vP`-3;3>sn|4iMsz$|Qvk6&(lH4p`|riz#_BOn1N-2W z1+wVF*&2-Zru#TE<~uFsHU&b$b5{96X|gJ2(XbR1UU>8iF}Ee}#JsXj-caexbt zAdP#}PL8Jdh9+~>*nDq)EGf}7*sRi2#|kh|=VRFA?(##^$KNA947GA2#y`~HUoEvD zo=DLWA3>_Wr`t;53?M}Nu|Xs>h6o~WXCLFH&FhcZD9>|9Tk6X;9IS+=cDAm+H{X1% zhZxq)^Q=h?qXjG_(8i8zsd*JGssrKUsdM3JZ-o=^yzLG3l{I6tKUHKw8-wnZoaiU| z?%%S#$c9+^kjlRzk2q@ol!YGRq_=G9UelqEBvEb4~ct@;F|tROrmy`l4!Nm^z2w8h{IbtBhDr!SYH_k}A( z5sRbgS(Ps;^eawqxm+JXg&=A#thePW{M3yxO34vJDI5;K?YO_>t8pEE?PcUy72~xa zFr~Ko1(6bqi^~)V`uFWm9V)}(EU0Mqv!h!m+Rc0QC+jCVWO zTZ)JUakOuG%z#wakftTc4IE;rdQwzbwg%1$i;m^dij7sqipqiK5MGZI5^axW`0s9m znN{x2e^fQevClsTB`6^<$I<87&5AMdMLvX<4{YG)tMMHIR4TGSO=V#LN!nQCrI>Sz z(6i>f5z_U~^t5=Snph4u^U=XFL|pO1xkO>B97s>^4V9G^Pcv_8$9{Sn7=@Ce9njUbps6Z z9S@bRRPawwG54y+P{iGjyAULtx0^XUq#`wh7}M36T7xNkY=1hjT8poX)DJAK#MYCW z25X^?;)tP)u9@3dNq3`br6%Qdq;c$t=xQby;P>x~%ij)RLmCGqFg|00?z4)Ad5Rfx!HFZd1C&hPUkrC1vvM=ZAI@Ac_S<9^l*_dM8wOBa%-mPAC5B^C1E-Y%@R&rn372# z$aQMlsdm!i1;A@9SqG?e8{#;(xs4)Y+grb5;6&Og@=;}lwKKRGsSNlSJd_~kd)7Sv_6AevlE#q_bu$iO8Phj@p4;@9O(5YVm^wTu&in*KQ?F>d#zArzs)d>7Pn%Uft* z*B;6Rtl%$Oz~SEkv3c={tAmQ_0l}qe4hP8p;&0o*tmcs+E|IILVy#PEcx$!1EVx^+ z*aKrJ58{%fQ0)roWXO?1M1v9MS7RXJi(6Q)sEQ|HD&R!`DHa*v2UWyA<*P^u?qS&&#WdEm^ayQ^sFX^ z88?djB|rCD@f|;=<2M0AltQ)I)1oj4j7%rK6TrK|cWIr}%ey($SL^eMPIe@x^f4k{ z#d!e=K_`9oH)+Flr3smL{FU;Hm(>(+FK>x+cCk2z>Eqr=sp+x5A5=|8&$=&ma8pXy zT)?8WZJV*~d!()^n{a%E}iRd-DeT4FohK6gVsd_$$-@1`Gn=`5~DV zp-?eMnS=nY{}U^llC2%3uz}Cl=$OoNOe~7dKOFy?SC0QT*jLy>(#WPqg0M{_Bht7< z@bP2e$|`%RB&FC}PBZ0x?26jFoeu}=x1@yj{;}i7p76*(URT=Km&k%ArHpF`7`2XO z80#J4y)%s4=K?r-T%mdFi0a9bXP&GF<_{2?QThk-P+&+~^ABI<&jM1LH&%?qgNib; zRs51%R>Y4iY=wecAHVdPj5Q~+SQF2AS5587`CH=M-}E+3e3Gcj$S9NFZcy)fq>wsl zR+w6_%~xhhQxCmKP3qQcTt5a8u^WdQy!&V~cjq|XYY6wL?>Cr~+cKIsG*20J=-7G2 zX+pa*iJ_8BP*p(N#XC=3P?!gQIKx?pp~Zv5T^YnL@s5yTr;l#Mh2}Gcr2e0_26cG? zbMUiH_I(r>^Za&;GjWt6eY%;2B?L{PqU$tK^d>Drk z#IGDTw-8_)cD9pO9pmp1yIlcY72GX2x>x2@si}efW4TZlB-eDV; z-Ipw4d|dg+LmarAJ@L=wBp%qwNO_MC->+GCf3M3}U?WlE`!d3)W<7BZ26^^2KDr6L zZ46jAR)}Y?xEF6rnGm1o%Fi>1Ph=u+f`50u>}u| zpo(11LN&&X943}dV1taj`LXkuFEeHc$7I)|$dHD1p7je>N8nO^ok6rQP^%>4LMJzh zT&Z50-SC(t*Wr=@T>RpQcvU}@ULwW}o-<1G_K_f{!dl7Ssq1VpP}B}Zuw;BPBPZ*F zt~gU9lY8K_2_D+*;WbTH@t@Ov*4J~qqo=*Ek#^Gxi9xlrQEPBjZUAK1M$kz>xn|0?*1hAihXZI@7`9#+_rmIt0V!!svuZb)4SWH~}*=+mUE> zo^e>KXR(hk6cp%?=Mzfm)*3z#W=`s#GHhWLNN^u75`Xvt+vn5= zJpy6oNmmb7KW6_1gX?B|w8-LC8E`-1ZeSisAK4-rywh!uVijojH0Z5oGHmklio#QB z{+6x0hKHBRYe5_VkM}E_DhCwKmC%Q1%INYSjiYjgDTfvrVXyv(~a7 zp@i<&$w%*zxcm)|%W*x>{2MZg;}dsY&r1OHQxL9<+e6))9{vY^!&}}Bp7r!u@mSYFh+iM9V6}SLKXI&i0&c9}NKW?^S zlevT>*LccfWzT;a_fo@9{a@|Sv||k5z>L3MfN38$P#2<6#q~geXxvWM~R* zFVFqmniv#Rh1$L8I}<3!KJ@91lz@-l4B^bKB~b5rO&FTaBrL|%IZ^ww&qv@7!l??p zIpLcIfq|YR6xl*cs^4pS2dueG+?x-QD3_H^U=dCnlhp!@sJ5fNCUKzrkct^O}V7+@2!~M~AQ|%IR6{10)~ci^qJ$6{&i} z6mG;t3(_G-(jb5xzD33RJK6i+^ShKFa&jAoay#uP_rGzar^x5LDnKn|TG>gcxc!n` zWcZa23vHU_vZN@W#D}8B)G$v5w?(&2STrbye)pxEMi&`mQ2x z^PAg+A~S3TOi|8Od)-B!=p;tDqRcWVb9RH`a&{BEC^u2a&_^lojomI-Lu|&eu-v?Q zZi8R)I)>~z63iDN_5^!RZ2Cxgw_vXIPk<+;+^536)I9H**`C$|c}gkuW(wpLxVtY< zdM&7I`3t}VOMZiSWsQBBT1@BZNl7-`=!|C)zsoBtiL2N~ub)`%TTa*q|S;# z{|Sw#!(QL5b(c|%kCV=mAMnh3u^z<;b2kj|0RRL%zy-X4hCxGuhJ=KE0|o<}f|(l9DsA*u_9&QgD0_R{m`7@HH;0s=8|ui%ryBb!uH7|4c ztb`oZ?&SQcB3r9u-ljs9p$Ts;Up>avpU|%m0L*w#ttmPOMw_Kk^d&7(8B%jvPHI#i zvhmV_aeu(+L-P0Zgrvl>sd^o$mi{KxncCWJ^~d}})C9EfOCjf#j6~;;IT%QpdhlCd zR5A9;E?U;en`6E5|LzbQ-qd_20nffez*79M~I7i)OvcSw!XrjSsg20$4&ceY|-g z#AlPqZ)v70xzkyBwedL|JP|xw;N<0Q4*a-c7=8AziOv(XI~erX2&H*og#}zcAAW=F z5-3kIHc#m|z#^!YOPg**1&-MlERrCl?kH3f#w6-Qa>P+-q~+Q5ran%q$Tgo+NeRhu ziJHy+t%2{x8fz?9`0kBLN>}UnkThpNk`pNv)|-~Be5&0&qe_+o=`KAi_@OijP>4CS z?OD)NOM|bS;z3A58JxU}wbiOrH+J=TE6&q6XTU-kIc;SmH{>w8qGlw`PWdjpZ}a*@xb}NtFDw`xYU% zeHA~a)u^d-&7|s7^%K7UnwpGpUVE_|tL|FV`aVGDG_+I`t*ryH1F$Ltf51M02e}jZvh}3N1@YdD6jbapA-@Zdm=l!tMcw-MZEE!yW@E$iMT+=zH@ki*fHfl;D zomi0M#+PJZ{e}+^{mU#&6_d$I_3}u?Oj*m`JXjvAlZ9FnO||}H154o|ZB()YOJPgoA>+88@)p8>%yFoGB)OgZ=S=S)`h9XqfKj zI=ne;s{-~Xt}PUvr$@L7B+bd=uO`|qYD}3j3c1S9tZ0?n$4iAvc97vEJdj*rRLCqt zUj{7@S}mEDPTsu@r``UZ6RfdByr6yTCR*JsH|-;-L};w^uH>T}y_kky;vBsA_B86e zTm`ojWSd!xio5T`d^x7kZow{226zZXEULeI3;FiQcec0Mv=U3WO$H8D^N>%cXq?p( z6$Y$Um5D1lSy>*N^5mA=;qRJK5KG8RGbs|W3^RzmoET{6d;72`!d3gZ#{3)1a@DA~ ziHH~2hB0VUOl+l#>R_aEg~f&p`&}@(T9A;-jT*#$e(z4H!z*->XqEpB_Hc1euB9-K ztSzN0#;MwmxOj<30e>Vk!tzUNAd?ujJARyurrI4$Fd>^@uRovhEx8F!p!Js$!t;eL zfr#Oed0)?z-f1|wy=7dJmaLOLJ|^OLk6|t$tENmsarBV}QRQDC0sCekG0Gc8=`r2m z##9gE@~==J+A>r6IEzobjF$_xP6KR__G!&D%JlKu-BF;~sBD@kRV_gYSu7c~{80bi z>L8sc8lT?MN7u@b)q*N@JB;TIuaFRA8h(h%rolRpIAx}>b}g)qx(Ob@);Q2KHS-$x z^EWna*QXk7R1Ok;&PPmLQXb9rpLghTG{?UYQK>iTd$ZM`yF#TIg*&|$lJ4cm%hgj$ z-weIIJ-D$DA5)#t^&ZOK#Wr5Oi6uf$)6OB%z!c9hPdXC|Ph8;)yEnWn`OaV`w{fEp zD~xn_BZMwRY=v$5pc})H2JPqetX$5t3L|uqIcHoByjXgT>CjvV?5`ho8Afn{qL0DD z5_un=k(@Yto6$^^bNF$% zT?XAT9m+}qz6-lr$rLlp-GF{bt-Oj# zWC{tI+0C3sA04dWGDcq#5n~|}-eERuiNcg6UjZ4QZzSg+G5F#Xvrbrsd{0S|MUb}N zG1rEl!dy*Sj^JF7%AvvkIf6qI8puQKeA)C~*vR^a0Y3E}AFYyg`m*|xp5AzJCtH(k zbt=AEry@GKw%ET=xRo8rf_-LxA=;VNYvar^cl)n;aGxSiNPfO;dW%{V7B*o$gz4W z48mE7khVWN`PyYZa!_w??=dOw!#1a)BdhwQio8{4rvpouOizw;tYQjFR8L*54X!g` zaLXHOPQ^_ynkwckzGIQ#8y+4R8h(q9#u%2=y2~g#!m}Z?Iy^@6NRFFfq>GF+g`!=$ zN$J^TeG>q+Vasx-i#cMRj@n|)NNvS|9(}~cYG1$(-Iyp^!y2f?kjv>TisqbDwHlJU zTsA@$SC8OoZiUTrfwcK~(~_BCa{1htFg|M0q-mg?3BkQeY8W;yPX1kq?T($c!c&`F zPYfJ=Ky!^+=^C&2e6ID{KiiJ_$hf=BWR$bD`28Mi&R6r1Pq+QDkhzM8J|fj;6tKfZ zwA%7^u;vL<(~_9dsI<_%eQ=!U56aKbv57;7S~ZOLja#tvJaWuGs;)Koalb#{x3O7y zPv^`djjldtJ2>Z=w#-H82h!RFXMEPr$}6bgly4{}b`^dk(@Z^GzV6l0-+$o;=Hi3D zRCV0sJSvgnfzr#t&mdek@V->JS;{>$>Ls-^)6Zu-~_|xGef{b$()*jddDnXrZVXbeznGiCz@A~fu z42ag|&>8p}2}3d@Mki`(W=Cf!iZW_dxSO~`C+{QV_n0ReV89BWhsgM5-=+oBZwgOT zIZ1KDhO#LfJ2Z$_;wW$ceUV_bv)8UR;Y82@>2Lin|84Vxa^pE^UFf zNC{9h1PJZ~mq3w1ix+~sTX8D|3I$4g^S;0Locq7`jC;bA4p6t7t3|PK4Qlj0vEr(@tmqKZcZa*`5%YQ{FaMo%7t1TC`V>krwLQ$L^9LnEAbm@gN4LZW%c8xmtCP$g}OPDM5m;0}=RBFifKT?O) zH->yUsq!xVA;VZFRQsrtzPvuT!-T9_y%?x(ghpBi+Oz%+0w~7h>-#$2lhi%+6ljkB zYT2%lLf~XoB0Mb0_TY^t;A%QRUzEvA4wxKf4F0rPc^1sKlA3Qm2ilLEw$&LGpV))K z)&i$t--i0@XG^&jysceUt*KH3!@DZ{jC(`rFM9agxIzx855Pd6BKQ#AP(%kUN& zb2ik~*ZAmA`zs`E_;og=dGrK}>H=jA?~{BBxB#9ysa*g>e2*;k0WR z>6u&bp(Uq)rRX7w?8xc}&TX`;FA>g4=c{ua;;^3NhCqlohuf5aO-FFYl>8+4n(uP9 zNo;NT(rFYToeB*aywm8(@pWvO0+#YjU=w*8fAqplY?Fie<-4%@9|wIsr9C(|A-}H? z7;BP#lAJ*C!XK3R%W3F4VZ8Hx?Mo4{ONp0m3xBc7HEa1JiazciTjx%NVG{>G9Hq)# z4$qADxMv+~aKBzRiXyRL{fh@&!gJsT#Q0`Gw;DPj z?2aV*lD{uvdA-WgL*Du4Qiyo8B1EGcY`ANU(L+Z|OCS947?uEV8G&1RngsT5ZS?y3&;0jZhb*Lk;fT@#@LsH<}oxIW) z{}0>@@qrKHSteU-=yH6Oc^Nia2GlsY-3BMA4uJeg#pMo_)-~d#{{JyFZVMN;jpzUt zVG41zh}eIxU@AYTCFFWYnl!p$Hj`^3{CT9iJ~^W#s0aBcyt z0#(oSf&+yY@Z{t3Yi*b(B5&04(|k*1?b9CKop_d0g8+=3PaGzo828UsAX@aIG7~{> z>0p{D-y>*srQ?xJP}KPPfv|ac?pR}RPaQO{z(P`tcAW89k`rhjayd_{G_iDv*KP6v zYd||%LXZh%OTW~Yw0osY8<@CG^REd8SZUsond~cyeGHH?-i#N!H$3F?;Js8nsO7Ky zc_c?h112pfPi*i=Y&nXE<1d`sq_bBhofc9M{5+hSt%6{zTP@@NDHH`>=Dtxg{2{f^ z1?X6MJ?Vy-m-lM@Wf4P$`Ho4Y<&(D7xj$K)>*ptpFA&aW#lPCogC7i3^ccR#ihB zo|N&&wsIm>{k$#=A`{YQv?WIvcDd^2OsEiacKuQl!QZtLj%aaxwH%(H9US#ij`nrD zJ1_XWw9TOXgh#%6A8MgAQxrYXE4iho@R*xwzJ=LsbJ#X|nYT3u-3tkTfmGppWa>9SO9aXH`V zNhP5bhMG5E2SpbkqsUUhbbOmJJ3Gs|{4G2%PA5jIMWJ<6$pe1|rwK|)@ z%?H?c1$)ub%{XlAr8-}6WQ6A=*YX6pW83VawUL^+J4V^Dqb5MZj3LS+ce^%d^b?+u zYJ1MZ2}$l8&oz~DGZnGZD}kgUx&6#6ZRuH_$_UXQqpWM#TM3as%?3j!`3c69G)P?F zkiq_soH7B48dEK}|0Gm#XZ-0&GMaa<%EZ}Uh(D*~K(-LAG5JivMofA&;Ad{dLXh3e z8e|5>1Ht10Q0v%b_`S7vrcCRN+LCfsiQ(A!^U%-RxUmiFYywLEJni(xG^d9OVp#Mo za?Q%zxcbEM-_C)Q%Yb76ud*h;T$)bjTBbZV)N#4lglp*vC~#y$`yxsBA4~r^g3jnR z3axw6bo2UbMU0B{?fF9~ePxX_C>-4maxJ`ave^ zH0BT|ZPO$4!Sdc0;Ku2l>5o^t^Q_{$KcW9L?z~=i%i~UsFq%k4dei?nL3JM8oPtV~ zdZdA^X*nRjmr4)ok~!3Km`S*PIemEMsXesLboQ~gkhwx9`ZY$Z_Jt-`i<$h&Q?*Ge z$T2(b>qZfV@)r}n<6bB6A0V}g*;zieeySBEX>c+WwaW7Fzb#3OLbTlBD$ccGf|m86 znt(iN(^1kIK6r-T)fi2@N(_{ucyyQsU?B5MLKPKCQJgrn_@}dO8I>_{mHgvBl`3kg z%O~;6TAS+Ki3Catot0~QCB@r&(7HWc8qQgYplxW@Y#jYZDTUz(6Gw0H?6F}!66IWl z;E&x@w6;CHPA<`nhR=(84;O;mAz!9Rh>#j2eaUU}M=KncC|6Gu z-B1zJe~O%@ziVF{i(H~^`Qkxc#Unusw?GAeI;;c{37jvIJQ>n1&O6D z^H20&Z4he@cev#+@9Qh>sgZvd9pX3u?H(Cc=wVPsQ>LA4hthB;kI4kNmd$oMmdbo( zpG4)X(oXu7HyR@^LLGy^u}fTBli#NWUyXcbNq2_^O?!^5yD(!BQm2v3)Ea}7cpI<7h8G%BgT|^V^WM|ELJ2t*{ zDF%v`RVFp1))3IA_Tx`c=#!1f4oEtWOT0w+l8s;~u#!oiU^O@EJ$ReznOh-aOP0xd zJkF^ClY(Jvd`W?~V_Qv)19z&8Q~BT3q{* zizc#SqOC=p+(KlZSs8!JstUrFAFG7lD7M+gi^#WMmmk1O!#Typgh8fKyADivVx~jRqHLJ1fxnx7V<5L+iOqur)T--Pc3`s%)Xh zV((BVhC0x$D7PO0ha-=dYUdgeJvAZg83rb=I0t14gR>3Tgv8@kfA&~q{l)4%?(H=v z!%N#3t`m1aH0{v|j;zjfyfGJ#-l>sfY%XCCl8#ylD{G?;!wUK79e_M<_CMFb<`;cgwh(}%{n+z@V=EB|c2Za`1e0sK{P4qNnOwE%) zx#5<5l?EDJ37-f>Z_8wOUdOn;BzaX?n`kI6nc=W_v$pIfpsmMtI{dwcv@T)ySk)6g zxAf2_7l)Wrx>LfprX%r>cSLxO`=Zb!IJH&2>6|_W6Z-P#@sjE%fT)@IJ91<2!nkDd zc^zkE9yZA`)l?c)?SmKV+O;BjMl(lVvOP>SoHV^%Ryd-mEDz?hPG{sEt$N^U;%?idlNDve@GIiYcLmNV*vli7$V+?Uvh+oq zZ#&DP+bCFH@9V>q&@xz|&G00xjd!8GG@Oz;I*y|8FIL+*h)k!|r-nCXxYj~a0FN^_ zGa2Dnz*M;c)_&;Ye+)En_N%4xPN*=?v`O8Z`FOXheg&8K)9?;L`AaX|E7fc-W>OHnVnW*~ALuc+gy;?<_7U}_6~%E@3q*q0lxVvuh0fiZVt zAfE>sYUuy`p<{2*1ravDwAM=nuw#AQVxtQ=BdIzUW5UM+5F=$09G?cJMz&@@e^WSk zC=A%K^oAT~EN1}hzg}BU3K=XV$-_1n9;?76cFH3;t)1M>4xCDhGF^u~Vh^BD&CY{M z+S$$cgC5+(3pW&xu^u#dYk6=Q;CgXUxM zwxE5kY-w}tV^oM^pHx(u)n@&<6!7fjH+&9Gps(X&xT3!?iQbs-@8g+K(@*4ruxm`Z z==6y|1}MD3S4wL?ZnPP8`{Z(-vSl`=Nwg0goChcD{Il|h*aztxnPS!FLqc7LwpMwT z=QX6wux69!WRqjR;1y%U2th@aatJTj{)|cBaF4 zZ~%YX-_|;;U}WBme}b>h6fDdf7wBowA@qDyagt=xtFiQ9EB%%&g$kpA5RM_LS2t|d zVILz3@?vu)-74MnZyd@}u|pdX-{$Mlej{?idkUPZ5L}H6vmyaD3I59l5VQbub!)zX z&s|r%JW{J$o*ekJM)^FGl9HQvVS`Qo8Shq7pulS$I7r76`eCedSj_pM8}xl`YzmRh9;$tm*vP%dt|IeSs4tGYS<(1JMxkqW}zVnF;;bavt2SWwN^9H zesfUeHnDbJ3MF|Tg2?-g_ncfp8o%#E3 ze^~sSoW>mIb0kb=L^6MsJ-cz<%8J-Yio=}s1g8+x9yi<;c{SKlmO?Ea8G?=d5}Dl-~Ip)ir6j<*&Z`-h8;Ch?39VPEvy5ip-4zpvpKoVzeqenVdS6 zIxYD_USM9%7o`B;DamuXH<|;ncg=Sk-_*jE?vtT1yeslOhblh0^b#qvwm;>saN|~M z2nRPl442uRGF%NR4!0b6J>x0<@-RO3yMAeDRjy|VVpUcYlSe<@-3&30{KX{Tjg?wd z2_70`Op*NVGmq-~bi`}T5GFOnL~W*&WQ(ePmm=I0I^+ilInbVdU>_DO8@0uBJ6U(} z{Jfvyxi0PzP#8Wp9Sf_01mk?Yx^dJd>K*OSIXD(5NLjL`Eo0ZDf<@j{g6?!ws!mr;nqZTI{!krB4er|s{Qr6%Q~)9X+8fv62@{jA!miKu)9%g%k%vb(W|Lg`&;MR{9N(`GHDmQ+Wt1i8+j+MzP8 z;2HW)ySUgA$bM^I^~`>rYI0v&_tlrzSwlOo&E{_m8x0Q4fj&RmB8}F2SR9=Ds`Se-Qt?3SQ-4 zMCX$~BJ^@gT}Q&Vz{+BAXuGUGZMAKe!XudEX592WcV1EY7x3e`E;gng91d(q&+5fY#&&yh7}ZPsvQ0g}wd1V34?Q5B-)*P}#m4E$8*<%R1 zN&Aac)|U6HPM)N+uy6fxnu$4n{ZzjT9~pw7rePu>e2uR1Ge8rSp*^i`qy~(cX7GJ}*Et&g&StaMhXeeI(-K6~}reh8^Qd^RvibtmfmSa12z^W0B?e z;Gg+Mg`Y|HbuJ>ONYC+!%|}+ETjyT3+5SNXD~AM50I2jRk{h4q!N*~@YQJ-b^}U4Azbu1prlqtop?x@Qgcgg z7G5H3W9c+eXug}|Q=w)#>EmmK7ip~TUD)1{6pBC#7f6TcMo9`ScA1U8(24x0OL8)s z8=W1Im^2Yv{c0$b3ecvu?Sb8#ZLvI6+78>P;bAa^FjG#ay=V{Pm?sRkn^7G8B`E1m z_8YZ7E;%1mz#i9J@^*5Z@ai{OY@KHBEvlJTjGyDi1iumhD^%ALFnswwnIxXO!+(T* z?v(K&H)Hvz%vMzue!UshrEv(EmD$)xk}?6B0rx3Y@$kLoSuNi2UuhM`YtX!QAH4O{ z(u#z_%I!~k5DH-wjLSg`B2J8Y_HhFmZp$E%7AxF$Ay zYGLa^e(&)}*YLV-qQm|=O_=eIfN`Mn)iAKJ76rZH4wMwbc2@xCYkcM(IfR<+eRtSB2S z)lprJ{w)jGD+dYw&_7E}OS>a)DdDCN-l7UAc_0!JREz!$k$jIZn=LzRV9^3iQ9W=h zsoXTeabDo@gTVN~IyqQhOsJ{T5-9S$m2(^@GTUX_-Ca+9VkEM;PcvthetC?`d3Kr#ZW&@EMRGW5;B2r8zXCI$6`Sm{| z*d!)@Q|wuE&GHsI<$IbS9gBYhO=UWX@RYaslnP}omZz1%a-JjYN1yea<3%_|dzE0c zmv+I$JXLH>} zL`;Uq#nqf-mnL2Ms!q6;Y2-%v&fw`oaz@fdk;26ibmF$R$9A=X=;>w*BcGW#HsysE zE?)jqU#Wf)VrzezqIz}0Z|2*fTN8O4jJ0@aeJOvj&_e`}sgGqj*KJk-&jP~<)U(C7 zC0s{ag(RURZzGbAkseYe7}i&}7MY>|Z|v4le~X0_L(-A0$sS!kj#U0WCaC|T-$mC| zj^6x}1a#0@;t^>+iwwz5tKXN=W25%noNv{LDpDcwzrr;So<$pb_+J)-EahoNfF3&Q zNpS$hePx~Mf#s8|N`mc3CD-=0EOGDjXRkf%8)_-Q*0uJpx11d9BHU|UQ8kWE6I?v< zk{B{Us6~A{gm;VXTi3RF80fb!cwOxvYJ}RWim119ra@ZTg&;m6tlxcPgl$>F7&%9_ z!`eF%MxmHmq~bUonDM{|A?(3YXFlZ@8o|E!gz0cZ#_?b_dp770=3g;$TjYQ6e8Ql= zbU5iUh6IMXPzY9s{3?{Uvw_^xkE``6 z9E5u%@R#OGh|Xa@xYI+b%4>^=qEo+fwwbT$&yp;TUI2dY*F%N$6g!5y&ol~W z^hRZQdD>8;rB=0Tm2H@VM`oC|sGeI5?Pa?^r_xJGJDDvY$k6*DZC>9@gnoGd^`MWu zv#U6_!g5u2FD7YRDE2A=a2P>(pZIp2T=b=!u=C$B^s8{4s7pnKMgRm}F`EZ`8>*z; zG)CmTV!TK^TTv{mWNLBUGdg11k$;0&m&KlJyrp!zTz_gvu^GyTrqru2-ov9GJ9t)# z9|_hnsx;+Mli-$W#dde$-y_01INMstZ8#3_pP9qeL0F55u8Jh?Z6K3&m)XN&$ck_A5+X7o;`-;D&DHkziP-> z-dv+wpQOT|Wq)&2%^aN;!qd{dfp0X(R?US%W&G3&rI)(*x$#grX69^r?{~khpvB_b z&4nbsZ|*9_=4%^|%vQdI+v>LM8E!!}VO8Ve!h+Y@_dKuVM1xFdKJ>m1j&(AGe@bz( zr+x{(5dKW$8_G+SudDbj+zs_7VXw$B@A9+c#k;+Xdiz6ODd8Ac4@>oM*9(UMhdZt5KC zGh#*Y~=9VhVp=eFxn*oo5C;2aVnFVrQ8%X1E1@DBEA$8|v3{6pS)% zo^U$T(9Z9?=inS)`;)wKSybjzfrHijF2@PM?)u+ileB49spj9@9x|8(2Nh=wVN9_j z+4-!@6zj7P!e&Q7gn5o-f{r!zV|h^^s9;<83VWvGtq*Evtyz`lP%tR(IY9!_sIP?i z$(MBaE+P=gxuqLSX^Quz;+Y(qPUwjDh7MsWCG}jI4CRLU;`j%$w~{{OOQ=Ju z84SlkrD4O*bAhtjc-oJg@@PJh* z{ZH0)0erulW?e$ipZ|Q(~hslQR$~S60Um-O-vre4mq(q4Wv`V+;jGfZiPSQYtsoij~6@n4K_x`s`9nnu?MR`OqxPD z9*N%Z`m0{x>YP6$bo^PnYx^j!YZ@cW$8~)y((!>bIi3cR**XxUi{6KxXGU6&n^dQM zkBWbz6hP3wZ8Pbb9O@J4&d&1l#}mm|$^w=%6LSrtR(m(GzjW2oXgHT+QqrVvXCy^y ztl2z=Git-PY*)M`h%vtOWB3haTpW zk9-UB(-Gu@y<&qNfXwY^24N6<$7K=Gb8yt?2Igd#ml?=KPZ0@rBWBw#p6mL)cE&$d z=b~#%+9bw_AaE;cz9N&bP=^s+9siPABV$2x-qpt{UK-1Y40%9Hc9mfzeJBtEbfSm? zmyMZTX+BL1A@)z`a@X|^H6w6c@M^9v){|2?OsP&dBapynqmUC_!1ZkWNJ*1gV9t<| zf38N33p*6Bau!1E>&!& zO38h)Xg?P8s|?vQX8HoqzGFp3K1l+A8p+BM6o-8ALoR=#$kS_{m#-t0Xv-_ z9VZmW@T>KCnTD_5DI1@blum7q=`h>B1asdfRI%t7+kj8hr*taol*|tl9I|IBc3~Zd z1Q_DSv!2%Cz%u1RtLSWrM6-gm(8RIDK1nl*0PJ2 zJ3^7{Zw|{Iy54%k9VS(>q67i~L$x6wy~eMF_B-73aN<5xxvb6(F6E9>?qpQFcG>RE zIH7*o;b0|0COCL8kj>3#jI^g@dm^jtM-OsEYOv(+-8M4I<|P2^p0XWvachdYqDsA1 zP@xme3e@Foc9=_b=hGWtMj(7ddsQl}i*t`PLRWDn@$sWDMkm=QGUZ4qkh;{1Hyc%y)G&Em!z*wEDF3tM-wVC4*&FvAVW&Ux$}m~sJ-@r+@D=(4$S4Rlezp~ z3T)NvYaR3L?!~B-TZ>K8f5VJQD#a?jxbG|D?Lyz#u)q{nSjDYAA4cdKp-Lvz$caUN zpA#7^f()Dn5=`H;$qcw^-gqFl7g@D6Gy5>i8W9AG*2zmo<}Y|kd=fjO^gu({(f+$Bzp&+K0W#u+fF$0OM!|siu3Qrd_`WCWk^EWuxqr-EE3; zL|eV7;^&+ z`P?Ef!SoYit!qpaTjSqWnalNCJSST_19JVQJ_8Z7F|l6b3-R*@Q-O~WCT5!X`y@Yd zx^((Zsxh31J4+W;v$PQriU|f`{I`S)Ij1(4I6A>4H$~I`AZ7q-bfww2d!Q2Hl@b-B zs3HQ`$%nfE%!^YkQUMNSUK@4oO0jEj4NP5;`$B2t6j6-snf;@@bhEu#<)aA_Fid zCbI2LFU(8vh=8S=%pANU1cgkkKdT2%hzP8$Tm}i)Un#IreO6aY4dQ+UzsDvq8c1BDT z^jT=bIAt|*ca4cHz&yuj{N=N3C(ZrOP-9HT)B2Hh;t?7$ABQ{A`lRs|*dMN1v>l)}RzDJH)-x^>6xr~29Db~=l}M$3 zdRJX}{uhf{>z?ndQlRcKf@R8l%*%`t$>*1w+T<6yG1AJ@HYgm~_V|4roaK}Z!N2}f zsjMd2V5k+qo!(oqA~RUff9ea`aItNyyAl23b|S!9Juy>G;y zGTs$Yy3Fs9wBUb(yg_Z5Fo^3IysKbGH%5=2FM>)!{6g8rgBX1~6wf+COSqYRJ>s3) zJbrjuuini*bo^Hm+3G@8&-S&pMb)}w+U3-0M2^}b)$e_zU+x&uy@#E%+Uo#$o4;7F z|9tshx65Bl3=2AV@7x)WH9@5t3j5BI16A6laor>`3oU5AI&xp3#ph|<$?J@x8$=%l zMY30Rq(F{irl!NV;<|0|)lbH2IaQ*KM(q``OtZNu?!80p&86)=j>DjY2KA+h0-OnQ(Ar20^Ny{q0p-&t$n#gW*(e!A=h6jBn0^bmhQWNp)g zW^L6F1ZnNlY6u8olJW*ka#5oT9tDBk9jQk+Y8`Y$uu`CFwVuxBG`WJ%4KKhg#< zdmP;iZquz^-eCk}l zjH#Y7Dyqcg!6Tp%sp6puRn9d1|G1t^4?%bGr^kFMx0^?2y#~*iek~AYzK}N0NSzjYEc0ZhhbbE( zHMiuu%CMmYbVQ9&J3i$6#$mzLx5V9}S4*6}HX`2FIYvqn{0n(fHZ`r{7lr#53;0&i zE#JRZuM#uA~!dT@`M*Jjz(#nAKRpo zxcpb(KlovV4Yj`FwX54l&wI&+4VV}ReBYxUfm;cHCNky9x!UFG&(9z5?_6YX6N$eY+dv^Vg^6-TOZ{M`= zhf75eN6ZM#HgH@%>#JzIOMmGeZ?Ru{iwEdf&P(hsd_B=h8TA1ZzgjQK*OsH|M2EkL zm)jq`<#3w5#UOtESx2h-G@^C5dD~lxzEpG8vzkQpCm*CM>Vy}XXSUucr~5v>giVC% ziLGVqE~MNFZJAmrYkO`1W)L#~A8r#lYyv#T}=wsZ_m`z?4j0hrAD>Z%l6QI`2*7bbWgN zdQwbceniRef##tTIu+th{I$NV{GEYG@Do;U!M^m&Nu(E#C+R9ijc!6Ayg>;whpCM4 zv=>p8p)sx5%`$~1Cl*T(11K%LM0geNDW1I;EkZSfPx7s{mqhE*C0BckPl(#49K$zE zOxomw7P=DLH)79g;26-`wKuG%>j(+o*RBx-e@psRx8p7BeX%nQb4*>hh}CWB8@72N zh#P6fB8Fg0hk^Np6JiZ|k&lP8q4dZOg#o~1`1!1!O*BC$tPamsS?igzG5pb-bNuV@ zD}8k3Svkfn-_4h=Jk*{_bLpVuR=S%S0;0;Skl+Grm-f?xr8GKyrPxy+R?&FoeRU=; z=H^up5bKtr2NInd_3GyN?lP*pdsmi7!Yk`d<|2^s#hOXIKVhKFp^a;#_qfy|D+`X3 zGN>2xP_M|>u6+l;1VpwZ8DE)|OHG1)*431hOn@5sU9=Mk_1-zY>}+rw(On#C?@cu0 zqaJL`s`Jm?3y>n@nl3G8e z{4ODTp;^=Ikh0l_BmH|gwe32ZWF^+%L(#IY6b9n0Ez?K8^03KV>`g)Y0#b)G*9j}t zW-g(pRrLJ0@@9uZM8YdY@GzSTP2+2MaY`n3%b9eS79>78 ze|d&X1I`<-q z@3yAC4_$a#{`UU!7^9b;Rs9!5ONOp`cSuAQrNU%mDbr@g+;h(3vdyl@ZBC`ZN>_}r z=SME0vv?9Xym%;H#e(I17svXo53)%+#`#+%5L;Wq@Du>QD@2b=47AcW-7@_Gw`Tc7U^A#7;f?F# zPxaaEc8x8j$?s2pmbl29Ua0!j)vw&?$5)tqn0;GR5B7W|_n_2?&xMXwUIvHIDnA0U zk^Ze#_V}Ofoc?!r-cNsyQ97hMjAnxnsYDaVY%J00*aqnVZ*^@Y$(n5OcEOq1MmMUa zJU<5Iv{2AMA#%!^^j4>Er`d`iFKtUw!Hdi)lR8eI*>q&6T$$Z5&2e-=K}$3f#PSom zeel|{pmC8jNPXHs>q|?zf1&i$M0ca~SO|pgNi`=f$YAnD2DS#?UNwTkS5w2M8(D84 zYHKEtPBJiF=lS~cwc?Ccm~^8axMJby$fp?Q2-QmS!KPHUNrD&6a)^mI5&cxKe8KC|6VNhKLrIM(<9=e6nUCyChk#^;2mh7M{*TOwQKp2ERWHiMDlkd9A6ml*9Jg(oVD~w$@uuEoUX) z@b)GfFMQi{LQ~?6;hGPs?tK_0v+6*{3&z5O>>4|~V|%zZ6z2GovtadR-Vg?gfI|UAQW!p*A#a)bGp02NTZ;-#I@hSt|mHFoin3 z^qxqeY$B%etzSxe)FYL@#VDFFFxJr;kg)yIFr_G*RD2!G0 zywhN~62v`D&c;GZKQn&#$T*ncg<5B93k&GG9d_T;ekUbdcciNgnw?>~K&QTg@sd(D zg0#X6MzBn*wKxRC`#Ss!x5^Je@ARFX<)b6XUheeG0~8tre;>}ptoOxNRO^oGx7)zO zGiZErit78zMMNSo%$%wj)E*(GP7mti&4#iYAC=DQ(ByLg%@t<%V0Za=vvyAlU}&sDm+_|mF~voz@k+PmyIT&mozoiAdzSj z|47u1fHaMPmNa2Zewi&p4-!Z6kcmTJ)|`i$a3*4>tv#8qz|&Zphyg>Pa{fX_($CB9 z1-+z-YpGS1*2smUeMqMOn}A@AfGF(PWnri(?fH{sY5_Jk8IwXIXM)u%uI1cROD#pO zglu|A48OROb5{$^Ui|so>+1Fnt}GqHwCDH|f-}Q%bjj564OS4DQ2#$5Wks(zrj&@4 zHop4|xw0-d0xv=pP3sh8da3LS+JNGt;~YnzNRYi9GGHHqEe9ZF6MV9x6Rq}4KZ7&< zO2tm%;whF9aqlaXYn3`%TENvH`FSR)=7A`zK5a)ShR{Tak2Jo#A%n#Qs?bktlBV$o$zHfppdjQE7sKl!{!5tPRHnAB zhxHJL&PRJ?$}~fcoRMK=64^rQoYkkMPCtvo~@~qe2aH#$g9ngBdjH_KHlE@hDuAi~?;teKsPIze#%$Nmw z{h%S0;abH^Od7b-X={ewjX~tWsYsf?;HrUA>CwD~OOdoj z*pAEFZI;1G!y05nvwB5caRuS0)64;VKg-8gevB)t;=k(%s_cX}D0pUlIporP|Ijbv z1+dM7Y#EsM2<}q>eH~KCu^OqUFpj)O+~XfJMqOcSp@)tVuf&cPL>NJu!^2EFQ^ zm=&7E6|ZBGIek6l>J#uefI{a-&oV5>73_7Ex^$?+#Qg1|xT4~s@-#!BjS*7xvl>9c&vkyReXlT{Tp zsc1#3L@u}RTC1e`l4H`WsAStrdg^edkV@57=I)&9Il53nuCkw^L@aH_#d{1qxFmRw z0420ZL$K3OHj^H!6#iN&X%U}aY>OITdRMchbqETl)QnpA5>;LK*zs^(MQp1?aW!9P zZ2wh^5YlW%L}ZN6+UNSky4`Vlc^=m@nD5V82V+a{!y#oVc)Y)$d|R@z&X;~ObMBgA zx%YcH+$!BBCp#7bPOb}nAu_-NE8+*f7M#FI+j~zx!{hFsLDY;nS&l+%Q~X46Ww)Ew2w2nKv}P9Jb(es^OUMztsI4bK;$Jd<39-ZJ!gSO6_TD4UQZs(!V7#g9-$7ZB7Gt=U_i!A}G0$e@V5dD*qSGLJ6Mo~MjQNgf56S$;3o zhT*xqIs=?l=JMY#{zwRs_As#w@7M|SIgoY=^wfjOJfMo{gin|zOrYXt&wGA@A+7d( zDhX*#w<6518!IkxMNKCm3e}~s3i`{q)jvuiyGZeiHpS;>sfJp1WM&2Iy3@st%&gD1 zp#xN$mU>WHzI>!B3(3k5v$r`FUb4=i@GP}S=O!7&$i9rKq81NYya=Z1Y z+8_FrmP_M{YonL^1r*$DSE;u?8!Dao!O4?L!A6fq6vO~LS;2YKR@%~AelzYEJaQHh zj+~yEnV#X;*!X7kbse&W{4f@$igx=Dx+sI>r-5Z3LUvH}`Xcu8sh&r9WS03wYlZ9d~$gRjT_@6tz*6|BP1P7KXy##Svvo(6IJs zfJH~ST;k~7;ciK9qP-fwE3~~V)Twm$@13iHM|lD~Kc|%v1eWA1#Blz_ zdc>&TrZFk+{s;qn-FYvtGoC@-o3+*+j<>WCat-vq4G%$j5!{1)D+?5p?o{h~5M%3)iE zaC+(I7k;04vcm=wM9%Q|8M#5x3Ndy>PCuE4qe~bpEdHeAry_9mWTi~eWz5O-xsTj! z7^Z}n+Q)43dkwpKNyLQDR@-6D=5Yt{_AjSUBO?t}tyTI!D}&6?0n5N>QMZyhu1eOM zgM_z(OycK*B=QZthWtL$=>NYLoQxD%3K8i|_AL9gt>$M@Ihtv^IWDBqo8@q6$W9_d zsB=u7<^LxO&;t*0^jE!36M~iaLEip+TgmU@X$L*Hpnw;=9HQ>h8k9$dYy&gEbbuM& z|FS8RZ!rvfkLa`JecZmn65OpO@Io5ab3t&kds`IwU3=RU=}nXd;W8rSU>ccMb5sLh*ulIJ7;_=sIZi@f+h5xs@go0bh`QpCvd#5()A{~R>CXXj? zH)2Jo9)a}DzuDwy(+v3ieAMJ+@X)i>%<8pptw@r8|Ah*`$_q4Cy!Ml8p5RCp$5Iu+ zt#-Iad1Kq;Qre;dGJ;M0@6r69&f}uHMglKc)4O~GPqyM!{kMFCEOTCJfMd?@$sZ1P zdrIQ3EfsIZM~S@JBy{<=13#J>4{OECj{U{@aqa5!o15}mj!n+1VpdwRF8>S7{>bnD zx99#B7gStaj$Z@_1c#j(f6VExalK-rb+CTt*%S@C~FI!{Y|1EIc zrslcuTX!47M=G;N5rIO;hx1IMVV>i4dz;U#M(&C|>Np5DJ}&d`Qk8@T+oF!$MEmT4#$q%_T{B;$G^ z6XjEg8>ADB1c0a#zrFN2_Lh$LG&%CwS7QJ5$}##A-hjSr!Kd@qv@E#? zJhhJ(y*#{tHX##5BIR)ck~P|DAD={O%!mNEC8p<1BSKq#3h(5P>r4T@ETM~I{mH!fF_iT&wU=O^v{KQ0 zcsJx1OEzn?@CvF5L$>3jL(wp@7n*tbGRA2oBVUY6+9I^QBr2`{f-gPPu@&~MGGZYE~d5DYuqGiXZ=A*fEbGO|AV!+jEk!M+J+Go85Bgiy=aE+1_`Nw8DQw4q#1hX5HJW) zx?|{Wfgwc%q+3!_x=Rp{P~Qz+_y7L%{N5+-@8@hbd#`=2wT|OBj|F)*Z068nb1W3I zm0ek-Cc?uxs&9rJCWk{`1sKX4EHMwlBoiqVb=Y~-?r{R^_<(h@o?yTkAyt(#OOs6| zi(2w;K-B^(DXqkY{N>f|*^I8GnXNrdx83ck-K5HX@O}n?1TK0PYkJ*c>r?1YF6JX6 z{{gjrLH3nG8N&JOL0WF61@nHDQpX)rOY_%+)BirHy<28LOe|Y#mR-lh;XBn-{v@(E zUCz~1O!s5U^1{Etbr%f5RVHN@y90xYh!zj|U#5rg{2{<`MeWiDHoXcgv_|{%t3u03ll-HhMwgNBm1Fk?@m!8-Bd#x{kT?BRuh8lm2P?E zdao%D8hhVHla+8Cx51(;RdaET5p&t_M)kC39Oa4vwHVe7*25zL6M4sS%aT1iw835B zH%{9Hg$o@6X`+)g+f++rnSv12V^G(?TmN;_V-Ay1z$@@uJDc5h!QQ=(>rk}4>3jPi z1o{hDfEe605DP4t@!0c8yO_>29ac_6*1~3-F2~c17|S*Z-@zA90&XkRY7c#q_7&YK z?FYKY{7u^w`>G6xPb-Zti~a4Ti?pv+x<+c}mIlQpw`B%|oBw{LQtW;24b}E?W4hIw z`|UNyO2~yTOEPPX&!)Rwzq88B9E&YjOVMQo3>yuLANMqn@wdi2G@A&5kFU+vR#dP* z?Oqgnjhz86El@U1+H(Q*j`&^!LX z^fg425mFrt*2JrjO|tSd71&gpJM7r)`?zixz9F978(gkc7q{?B^bCXud0A?I6XAEK zpAXXuXoYqBnE7{E_joM~!Q=O1WljiXOI0^F_{$F1Fq1eJY~+l6L^Cq=nybze?(4Is zqmOE1tnF*7gIA)SIk&y!t88)7gbe+UMuQwuqV35WjYQd zc?kUCJ8oa_>A;NM)vis&`lrSk+%MkI(|DoTeIk-&AJo^&Pl=Q)5r?JM{|$2-LxtR- zeF(8`>&L_tcR6aMyI4%m+v`mbz_)e1>kt+=b}v zhX~@wRGvJg=(R`PuSdsb=&6>$Q1YzQaEMkgg|7zR=LkXm)ebbqi&Ot(YXbnKSU3NUEgX*y~; zb1m&tXEI=Gz=prCkwsgcIIjL`PvKt^d5`g2bJd!nKz74N4Hb+~HiWO{-lA(bY2K#w z><;DGQY{fxtL2AiCEHB8@0DF|>ooQ2G{CN;qiT@#$W^V;TAz{>*wmo9 z#pKdR3@*BUZ0HCql2Q#vp+4koU!P%qDTCwq>E?ZYo7^c?h5mDWv)FX&F9p^ed4j>< z5m;5c);f*kN_+*AFxuW_Q#lAa+si-c4O}G_n8&S7!i=jdN5n?NkN@fKJLSIH8_qe2 zM52P+f5nhQ*JxT*2bhfL?x(l*crWoFPUAasHV4G1l*poLl#3h-Xx#{Lp&RK}kZ{KS zGct`-mMk&B{k%eod*@LkV2IrAxs9mYFY5w{l@yshC2reF`S^G%NfqT+L)#{5g!lFA z@`+=Q+=|wn+sg?tuyS-)DI@d;19dW@w)bk zBv;3JfhRXCJ+9IE7mmBv@0W_4Kc`SHtl0B=!`8?hktc(MxMnQJWN_FL)!c44)?KUt zkw_cqTAQikYTo+vl>grP32cj;(gqBQg^(3H)w;(wA8TuGbd2SU@CP(V_hQ~&EXc3e zp_%Kx{-Jtr6;LHLA2ZJbquAA+@9jG#4>t)n9r0(}H?AJCJ%f$=%~KnjvRB+TF0~Lo zzRA{JXPjbTzdZu1=w(6r_ZgawRzva#*GcjczwZDv%2xtIscg{i05#8(o&ux)`{(<6Y+#R;+2!G!eh$)bkNF@Y;v}AnNTQAv58FB)=>A;7F%rN@< z(d+I(m^hD19l0WBxCA5K(*};~LqFnnD}^c&Q~bu`!#}!*f+6%^ie9m7*XP5fw^)GH z2J+;@zOV)^NUnRClVr2K%a-OsBl)wm`aHmi9Y)*;=Khj+g39=j}BBmxHG~(J`6;_vFmGY%wEH7+$ zHQDuS#l@!u&BS!t=!o&mbqdnltsFni7}J;u%1Ls(ozZ&dn0sWWa}vmwYtc@Qs|8w^j)yr9E!U{+ehb~byLXwy8DQ{ZVx=Rpe$Q1&vI(+oG`Eyq8r9_Y zD6^HJCMnM{Qbl*-3;BGJ<-QUpeSsmc4P94qgdv6w14ADU;G2CO>7Xeb6U`z2+9_=@ z{spG@=uPJ;D!mQSSDgLAtAihTBgQausv?DO4=Ywq#*|qx9gmgZj8rd5U02=bg*x|) z@iNQ=eXV1UKwkVX1&%1{x4#(r6J#r;DlVP-)T$KDg=w;BooYlT?gOqXe&=f4%oZNOyF7p zI?%uFpl5&gje|c}beQz{!7L}#%kX7V$_H-f^qpWUE7c|9iT>1dAI@s@hIDqp?YJ%A zLe)7ijHdLk%33}yQAQ*V((BQKS=5*1g@2Z`5S zH1-xtkmg&GeipCbeARMShAk!_@DG-NBWDmQB=Bbe7KSBB7_cxPhzx4Iyv;MiLNvye zT`(4dOLE)dKL3${BBW?d{7yj6RO9F$tYwzCFArT8oNIgU{u=L??yxo$K})w4X~^a8 z=CPPP`{KrNmiUQpT#`viQoI&ARHUiJy|%2pqnQ2P(&0z`*1|`~SshM7fYEP#-~K8i z?T-FTf?D!S9YxvB4uP0`kJw0W%v@8#2eqKX=i+^ z&%);pKM#YAm;_92P1Yps*%GE$`VH!^zx8BXy1$xhVZbN2jC}d*S=#OL+iu&epUE5& zHLlv>KFFQ0&tG3u$Px4v6q*1f5mFOC7*$WJ_2?~h%convIPG{%<0A8z>qgk(|IxTb zWZmYvovzpb{so=blk+)`P>8cV-;X)wh}_ zQt`?5uZw~iIYUS6Fo*&cHI*>I_kpR&fs5a+p9dS=X_Vzqo50PRTz zQEmEdBZi6kIkmZ;UE@ta_5a2O_**xdR|&B5)N_Bc+CR;Hc6bu5dPQtE&(p%@%NXBohy?wu&u*K!PTCzhWX=2fv8mfduunz% z(Cdl_{K2Z3?s-x>Bj$`=q-#ijGHsk@Wt{YY<;TEsRR~VpH0L^9i7dt%u0Iz$VaA+9JAwYmU_P6U2Th_sfX1|oDRK#AuQZ8|-S?64t3irlQ96|!DiINHrt?YJzURT2 z-B96}_rT{kos+grLqLC+@!IOf7rH)xpo=VD1SY6nF>qPr=ybZWni0XvspTfU7S^cx zyUMYN?PepxcGSyc&Win&={~;Txl^nkXUvBbvqR7vb1#?75ZT456zhmDaBW|=lWm}w zc-tAWnwB%RNk==eAeM=GI^xMfeQAg({9Xm0BkXS+o*mT|(ccF1DaIChtUhXTR6gol zrT$#nnQ2qAA5qcrpoyQc&R~FXUc%tOoy#J|=EYFnLz7=ONmO%iXvko%SWkN;67O}{ zau}9^M(<~i-PNP9^@*DfN&tD&K>^LH+yZ<;%=Y8KSd`1uE>->hO;|?!!NR@Re=wuw z&ccT@hsfbuPT_R@1V&BlbxtAKo2qVNTbz<0hsC!7Js%Ac5(3r>Z9nX z2#s`hX~4Q^L+THa%Y%x|0fFRAHS^J2+DA?HDlhV=KU-23Cni_O)LKZ&O9r%Aw;cZ3 z9hOl_^7#GrSIL(R_ke8$k~Kt$4;W;ajpK5ty=IcEN2cSdv|DYiQuz&b*o+@#KuBua zyjM_GZ_+U}A8u58<5cv*p?E7)kc|YpsQQV4LQB;dnk2#;NJX;}V(uHA3D@74b?V2A&%&+msLa zdOp8c@yu>lvl1NXQuD-o-6tbtwckA1*CXPz35e%?skia{*=Th&CMW%ga+?$JSuTde zr_~371kyfUXC-2tu1L`z@}49yUAYFqYqzB&DZ*Bice3A!0a+k)I^Hnl;A6r+-(7~P zlTi-rZ#jgYr%=ZGpMR^q_1JPN%*R@xE1yz(@Jq6As_>I^<<*e55@Nvs4e5hdMRL!F z6bjP$E22)5p_v1Rq#$hbQTWwJ}pv%ss7gri40v+gat+T(t_yd)|M* zNz^^=(=p$VZ+elxIZhaeg5n!r!-wJ5@FD2i`8CFefBx4CBeX?-I(jaeqP93$%ctF!um9;lz_oCmu7a+? zy0Bl%A`kvxiS?h_Kdu{g`w-X=J$KZAIx-G+tHHG`4QgpLvH{P4H|FJmK?akyci@9b zZPkG*ETeJ(ef*_8^a27LMpuhHKe{A7_PMgQy^lQ#Lwpw=Ge4RW$9gxJw0jiF^xAI8 ztJ7v*kz4dXhn*a_&O>iQS=qzgCJMwRhnziAl<+xVjo!Wr2{Tc0Y`6+^)x6&2hN7h!PdG3r`$XN^I7-Yo|evD1xz8}3go=n->*%Fr0pQcA&EFC$Kt&=UI}RX4<0mF~aT?U!JTFa=_J@R$l$JWKQ?RKQndx z5&6*jn@%6`^nE90UFA$t<&vw2oQ2rSAgOy(*;K@t;NYI=o!0l$jeNq9^%y0`N)Z}9 zO9K^Et@|*e>>wMLoiBMm&6i;W#_iSvUcd*3Tevn z*8$b}D&h~;!((JvsY3Zon&i?yaWSws1)d~wB2v73?YUg0mA>kG=uAodI`hG`zJBrr zabtsdAJk8CgfSPpnUi?Nt4N|w*QcE|j^T_2R$fdmOQB$iw>uK4-t;P?U^3Y;wy7*( z%IA<0D_EnKoXoHm(dp0VR#?M?-Lh@AF(d5!8V6l`D&nCfFm#Qz+Z&wVy8u`_Df8J_ zL<-4@Rn&k5-H}(MbxWPi)F%@3v*plyy~*_6QFOPNDu|nq!`&vtg}Rv)hwx<#L%MdD zJ3WB1wj8qzgN%G}8Wf71%LTM@so}k6DFO88oo#Mw>9pXRd!7`_Q!n_dixrX?AEiajtx}Z=M%Q6@)w7 z%#(@i30dQhlc7l#)c}IC_L!xsHdC8g&S%ATX5VGqXy+O(C~CYx=tMRsO>cF?XD|}< zHH#>VxkpIg=PL$VhfU!+SoY;+c_7M4@$-fefowGg@QX=*jmF+2Xtuq?L0sh8vbwqW zyOg10zRd%enwlt3e3-x|q7{-HRA$!rq2Uqe*BG$_`|VBxT}HWp@tujIn+p_A*Z;kM z)B1Q!vIqI!sF!U2Nx+o#Id&LNbh?nPp(FyQVUsK2v~R9}DD<1AGkf&VqE-~=vo(Fk zAX`Vlnc{A;#1XU=L=c;Mn*XO|V0fVG@N4lairRM5Ob^}gc2r)IIF&@aC5?~AZk{M) z!ENEIg9&Pgp1d3u%#|uzYcBQO*ozD%3Q`6scVG=xryxDgxBc6+c?o2v+1Ayb*dgM; zMx6^`{LQkA?_~yrta!s_q4|6auYD22CR*7}J9}<&YfNY^$UpPRXhp4j{GHLVIUh=M zgBQBKB$yID6TP$WWJJ0N6>7!*>;q$|#sK{X0u}vIU$a4y?$r+OKE>pO>ayQKm%l){ z>7bpt!-rtPj-VH9Eql!Qp%(9g7fGdY9TSyybs@aUlUo$SFQrr7Nz0!dgwpI&&p+mB zTY6{Eb@w=^HQ(4y4khDb_i3nl+rx=NW$t1?g>%h^!tHeM50*(MT7vAA58u zTxJaC2yrS6%}77XDKRQoaZ83~Yo)gZaO4MSjQb?q1Mj0#8kfSRg8?IB${`QAbs9vW zBf0-63OL?^e_Qtix{*>@o^o=qC^oX7e(MWxLb-}JvbR>Il!89rT5R60Zi5{-1z^+% z95Rinolvb2%Yc|B!&nP-4)}jh*abKGXaR0F|1zcH*FeI=rvbIBQ8Rx~#^%PMg)}ZP z!TuNa*zc_3)eoItn{CW33cjjE;U@abPbp}Rp)DR~q8p9f)QYHaV9$81DH@#H1Ft>m$_9B7JqH0bTMJ_qd2k z;oi2nun4@hgc?6IhH27*pwbaS*`YLZ%#%&*MEEKi-RcDD5BOILw(6s4?u+S|6dP7@ zaS6WE;j(^#t3?J+viUlZn;1rd+Jj!2@mH&bq*1yTZLvPl{9D+p7c!~q_xktmQTM`8 zy6L{|Z_E`D_0$yOu02^$lWYvf!X`+IK-WVygCL`xv#ErL#sDeyJ+t1>Ie2`{2M@zS zQ`55TkaATsa&GF;x&ZB(Sq{k6NTZKJaj6u}&X}?wH_fK(acP{+GEqk@bSt#vF~urS}0k4943st3iDC3{Y~Xc2I$ z=AD$S|8^p0d6QA6jb?HPHMx(sC!aeyo|rAv*439`GgB$5ogn zuPM!FQpxzXdsYPL+{(CU_DA-4CBf6Ir802h?>>}|&V%ZV`?k1r`0u23osbWjk0XU= zbc*fhaM88sTC&~JTwSwzK$L^4=swR4b#<;oX)=0k>}zkE&@^L^HtBEXRY`4VJI3Ov|02bKy~v>5 zAQ8vQRZ39E6MS1a{&=_cANg@tYfx6YG$BtEOq6_kFW6e8O; zNui_Ms-1_pNzAn~iDUAM)8l{W8qjHs?w3=NpvT2u)3bBF<$?~^j&!*wk3_+$U}>#& zYBf-brNrfjY%Xt(ByuTbp8 zIo5q9sFgZ@D)%3>0n3B+FP^NaZczA}c7V}%%e$MtIkRSh!lxSyD6d3s*rBxgSSyD^ zX8|jwwaJ5Oq97)E5D5q-JB}XmN{LePsoBLCR1D-;1ij}=J*7fsuw=$3$P)~t4UueY zS9qgcH$|ggM%arJ(F^D>j;5-*6K^zmh4)R+sXfLfn{8j3Xwkm7x6BnF&r`%d1PP3-|w z#;5kTW;GAoX(&EFH8!JB9zHF_KuCS2L#6hOv-QY3q-Zv(p1YJ0ZnbNS1$pFZyLQIhXxp?Sp+u^=(~ z!G8)25{%9Ak#qxvb^CXOEE!)dwnJ(0J)hVvFMq^MOb}sNPk3P&Sp#Y5DGmcGq(Y`Q z@2lc=)f@NBU=5@HU~y7wrJ7Ffp5F5Kz{N@=vx6S>P<&8$S3o)oPti)Ww8Q(Z(kl`- zdOizC%!f3e1P$myb7{ zp9A(dx8vBE?E!a|u6lwO^u*004<3|2*B$cGdnVW8O0ePBa6T;Yq?=T8LdC`R*0@hV zSdgH07)QQn`x~FL3Xciuj8A}b1kkR?J$f>T&z-kmKJZs4;hP3D0l#TKu<5n5lu^rt z1ZxB24ZkAMIQiilhZ0COR2Z{T)LCD{iZ!PRCC*x-Yl!}kSYfi#;>lcApcjbc5~amo z^Y96$Do;f+VqT9+vQ@qWr_2Ce|76^BYyIz3ixSPAJJj6$z%B#fYr5bAtBfrTFsA<+ zlI!GasCG{G$`-!w`EJ|J*a9oBEVzh0NiNb?jws($gM@vIIL>Aj0YASZ-DnUD5q{)x zo_E+P7FP!N5#4By|+_YXfHYvrS<5vZbpzrh1wv4Q7+Z{c<7{~ht-~1Asg==Zr z?-o0Xv+kMHoWU=mrTn3P+mmr!;KAzW4n?6jUIeDT-t!QxDO>mdQb!(ZOi8_$=!Sa+ zo4g_E*t{Q{mWR&)zq{7dc!dtW!CtSKUjjxVS1`@hra2&P3F)*&Um_>uhNA(7&TQyG z^}h(1s__j1X3El4&|O*Fc++{W)-#22u_`h!E5Jpk+Gwi2tW(z@p0IL!a66xP7wv6< zH!D&gFd**S5b&kDnQXUC?3cwO-+Fy@s^ed&t*#S+nUNZ;u-I@MV{b^J zyJR|qU4F7{+JOb)`;usvT887O#)CM5v>~4Gi)u$QfF*X!x=HaOdTV8o-LF=MADc&k zm;1&Z)P|(#f8b%8ihCnlTka9|PSb+iymch=U{Xcivfd&b;$kjvDe7H{V5*!alKldjjJR7X0jQ!A2vZ>ycrYtL9gZ4485paGFB}GMm&g3cpE; zrnE6pRcct``BlvLne6F{p0u(1bWbmqQl7f38_P4cn}3d%$)QHf7v$Ax{RG@`Y+Bkd z2t_8&$>k`V!_b-a2$1i^;J7$;yrSQ9aSp5I*P%~aaXa6x8Q|l6t|&Lo0{@ZPsNTri z;*)N&7)Nq|N}WnLS#^fvn2lduC#4Ful^)qIMyRLFapN409l>=s$ABfQq;m?HgAs`8 zhGB$vmwabsF@BPTAnP9N0}xz5sb9R+4OHr4pHtn0^TIzDU-SPSBzGjNK)SR!_(Jon%X;6zhTsQvKi;$`z2YTv zMDIoRRdIVh*F@5m2!2*G_(ou+Fu+(yy?R(l%R3S8lac8ak|_?~-`oPqMxzwbdihJD@Z)jTHh`~M}K&dGGnCD^s6XacXm9or~UOmjlAn3 z1;$+~g$COJH|Xq}CKEh8`$#sL?7b?FI~^qvbL_ZyyD*EzkjMH>C)`aLofbr)Y*NYq zH^&!i=4v!dLsA0h;jI3nhhw(y+SvzmR{2WB=QsT5+%(NaV)@-V?f6K-6Fn)0=Yt5m zw1Poox*Ng+{H5VP7qZ$%5~aCOY2v?0EV)e)#izKHo3?v&yFg%l7+S+=x~n++V%{ zymg%MtOqP-xL?a*Qu!l=E`rreYcn@KU6q>e+-{rluK>EC>waj4z6pF@D79};b}hvc z%@VUGNjxf0s2S<@^gUjkAfXfvJ`bLg>CB7EdazC`fMmpJhpt<~oTdj^b!D;PP?}?; zSQuBhmwjx@FmLyLjN?7A_@3DgZ-)0NoM3!4jQgYoLknt?WzO$%URy+gQP5l&*-ho&zpjad$kwR<3&a?7g#Hl zf_!8`3Mh5f)1j*w*;M;}7{uhf!O^CpjZj`n$kKGy2bCw_AaJvH&bO)gO< z6mk0{I|+FWXQ8_GzgH`j&c4;#x|Z^6O|9G(K(c?ZYSV{##`d#akEgt+t+M!&zpdl$ zR)uj|B`0%6uM&~@UK;maUZD5FTIGglBgf-9SeWh~EEMC8%gQPUfQ3PhkR`DMyA{nY zuLM3Dj3YQ`S@*#g4BfQPJu-n39>f?0Fitke#&tMU{7SU;#;Q#86`T!cq}UFm+@;dP zu$MAo{0`u^kgk3W!@T&lk?v< zvW2zGh_w$=7)($HtB^Li_KNSIvT-LoSEW6+G%&1ky*J9fi~SGzaMWftQIu>gn6A zGag+$J!POnd%yckY%Aq^jO#qF#*`jsL?7R?J37Q)?~Zp;#b+Co9Y#iEwIUjv+u-&!-XbDca9VELF<6wj&jNTb;cEadG^{OqJwK99kRht#28*3D@p9T13Y?= zn4DHBOYl3px^YaG_CSrl@^W{i*nBN`qFABZ#1zztF5%rR>jzBL$hG6}605na!^4E+ zwZa#;zCXhLQddas03*OfIAJ20P9W4x0Do5{0Px&p7a6 zfoA@?(RfL0=AjylqY@aHmds-tUJWxNUx;Kr#}63RWH6|kkU)Ms;CDDoDBA=ixRrtg zEKc4Qb_UYNTpa@G{yKj@q#Bq(S%J9+Xihjc^Ky`Ad;Jx(`cP-pj5tEOCkNeKqR~!R zJgaix6K4}rAqBADjJnYRq5T+D4Uk=>Xet+nP6`AvW*w%?;EJ;K+x(F};b z0;a}Ja0ZnCj>jNtRN2nBQ1U}WpXv*Xr*Q#CkKW;ha&HgkKLr5~1!3TDef+qziE zg$Nd*C`2uTpw(r>o0N%CN^Ij^tJ6oEH%BKpBkOOda8Blc4fR(&O_*dpqcHcNIiVgsXh;$dv)Z8YSpPUZPbzObZ4& zEZ90cxt(YuP^@}TNIbTEmJ&EnWcU(tZD~Fh1!#}Qg*;4e1RXO1)?$J|ux9qe<*~F> z%#`qWdp(;}phLM&l?P;Z$OfJg7QgK4VFE}+bc+n@6GWs>MGEP}uj>u_#ed{>E!#=YIs(ORzHkjvNc+O`CGv6^?TvLHak5-rV2+1ML zF#Mw^W`_qT7gL69c+6yE zC*xD*^E_~=5X&)tKMD8g4wWEllJ>u{G1ptTm`oM5YH6IlPbD7cnR|4dT~T|^=1#Bj z#dXwi3u0x0rzf%DjT2Y=qqLed54VaHms2M?`n5bE?!N1R z3~-^RgC&)IQrLzcUq_fCO`aC8IO+s;q?OrONzxhF9Y%yu?=YqC(CzufpJSh!_eDQu zKld?Geq07->#E%+5}c%RUR|TO-5TGC4t6=jW#4~Od<3`_sXB0s(v|kUXDCl>asmABuGtePotU0AjHMZQhd%J~23&HZox7uO4Qa`%t5jVIS>h>&x+{BI^q zU~K=l|Jcnl4NF;blbJ16_~JnnrqU(TzZA@1%`v0v-dlt}NoV7%^g6!5Nyc$W%vyr`=HD>xF683!&Ew3$7QAT`xrb&n&E78?%2R zB=97Nd_F6`<2AWya+CZU5)*g10^JD$<+9 zog_(yv?WI4N#fbj|I|NCq~eweIxn!^piV$l%XzW)$3Eg<>(e)*B+4jTcWWN-1^DMM zg@NL=`H`4Hy7Wh^&SLDCfw}t)iyyR-i^Y|+`+58KRgVW&9Jm})sY}YA@qq^)tbSlD za|B`kPd)zs_ZQlk&OTpqH9oLy1h^-NVpn;bht?0LV^TC8x3N_M)fz)bJvf6;kd@@%?bl{2bhLAw`a_O+Ka__IzaL)M; z(sLgd>ku9c-wPDjte5}$3o@2CwG0p@ieF z@xaOHqP*D&xH}VzbZ`blZ^EXc{i*xhx$ukm%(c$0r_2~8W)KOMH8ru@#&5n8E~(bJ z+@92*W{xM-u3OG~v{MzFJ`TmeS*nh3lw?d8to|j9{oKVOllg73!7K@|nA`8Oe`|d^ zFVK%5zJ;;%ZM=gUcBO(16O4aHN#2odUxD61O*jc*qFus0#HA(dY$%K>Plq9C>$#Kf zP`!?h%)TDF#-@?CLAVxs^frN;hqu;acm1&N)pIJZ-rvH2k6d~M_jTjkQl4Z)P(6kw zo8Yy5xAEHafdKa`^gWR%zKd)ha_RhIRDr;jhw}}8(bYHO0YcTj1+$ahQ8E>>jKZwe zuPepd4C{B?4VPi= zFU&DeF=CHOtk=oaTnTeG+?22RXW^E4N!H@RUQrpiQ{C&3_ zMeO?X!IUAQc0EdWzXYW+3W!U8>Fo4nDzZb%O3P+xfDj#u~%>OMnR;Ftuf-dAJ4v@zlHaACHXEfSeWvj13bb7+eckGXj zi^QscExEz6pAFyQuK)1vm#mBJsvW?#$IvJuixZtGV*=f zC>x6j<&E&lU!e*4dFVXJGglOMNZF5EECK5Erq1irP9{3^z31W3e^#klQAink_WYdWzT@6L36ZrAJYv59s0q%-&@mE;FoG9l)Nf z*pcZI!`R+m^gV;?WAPm8C}FpsdRnR2jc{#RtIjTV{L5Ka!h!P{;-TY4aKUy;vNLh@ zzRRripdWA`5*Ur1mkdszV7X+B?{wK&dqFo((4(LTU0*VJG}K9%yP$oKhAg^7Ww@vu z>nd&D<_tnvSZsw~7#sejg6ah7e1U2A zQzmEiUBQ>GQA}L5^zdUxzug)%)1~6aVCnBUijFKI?5!dxY^R+rrIjAHc&K}KYdRIb z)7x4)WO#U=mM&(lXI(1dQQkeyGk?_dObedtYFp89;CT~}W$c<+lq!)nNj9pW&~`l( zZMwI-F6SW^Go+67(CdC<>vliS4vXYQDR3eeP-jN>vilEA~gw^;`{v zFUM<)N^h5sKo^1Whgu!syjl#u;~b$y5=NZ&>SI=El)oR${ftPP)0n zXsh>?Jj4#!`r`nKmXI&2n^r7Rpq&xQ^3D9~P+(26tPq`8m@ErtM$_xZ;(;0mr7-+VgH7}s8E5!My zgg-ndmz)6=$zPLe-j?V{uckWV`cvl*Rx&E+$U82&!iK6>nFrgTCWq~?kNgvhw7AYopxd%48TGa(_ z-*QB7z_d&GXcWbHL1@dxOyyn7cbXR3?ZAjn+kgjFXAmnSTtmXN+9JD*ozH_+;4$<{ z;!d25Ko{YAX0*SZkFDxgO}J~YegE<6$7(jC+ZAuOHQa69-U|(SM2eFmb|XO(-gE(9 zhv<0!b0$B&kES_jS%Y*&p+iFazf9x$0tqArDlS<4nkt?2dS;`DHj{LE@g ziH_K8b^Tw+W)qTUN}rhjvgg^Eb+^G*x5(gQ9+IWS_YfC2p^hUaU{0hakFT{PJoZ#O zO4cwDEOj$hr6SN@hd_H12Fp`?f|m5+_kHXBrri?pqw8B_URok~8*VOk9mTk+PHYus zMj+4NGtxym@6G-*cKxe3Ch&1rS=gaoBp!1L{z6x70z_P;@41ANSH;qcyFzFiq~>bqnbE!DF*$(x_dah0o}j&W7BFhbTCd&Z zJme*XRnc1-S6=v%O~930&7xgj4YK-bY6dN)K*9kiu~5JM7GS4gstnTbPcJJK8%+91_BQPrBKx5$hVHGZ@^6O3C(kbV9Y zKluQ~jER!k%a3)>{&7y^Z}=e`j79N!Fw=&`H;$4ZMlUXBH(A6zgUxUHE0g zmB=_-gJgx(wL9iuaoGVg4D3UwTpEQ~DmV)F?)|I$QKoY{-|<&e1o84EZB8#AM9(tP_;a77h(V* z{OWj1#Bec^2TgHdA)TaMVEp=0uG9YOo8}zt-1zkHT1EG6eqeenJK$Kca9;qoB|Mi; zWt6~zpk8>(p992}qLIG@ajISJ$9Nc2;V=818XoTW+vnnCNWMLB-EHjtsQ(Gke{q`M z7BS&uNVV9o+&5X4w~LMcz+$P8M`P!m$uMRYO%f|Y(JAMrHLY#YNptW|tJR(H4fYkj z8*ui&;xmT5ZF8Fx{^1cROWuYAX&1&uCensCOeZF8G(8{w%`T(YM0Pct7?f4M239OV)Wo_aXCgn2 zebowIm6^InyH-k-=vu;4QOFNRmx$j16*rMzbR7#yH7)s14(`^eHGQ{DaS0&F(CVya z&hOpm4VT*hfqOM8X>n?bT4g~Sp=^Z^S9}mvi2CnSa#2*DN^C!EO$ynkd-u+BQ$Ce{ zJ?ZEyD~qEJ^H&QG-J>dP|A&?HBbtZSnazAvHyQmz>BF1%V ztyowJo@V>k<7PX`VW_fs%9Hdrtivdt&Wl9y2=&A#v1xoYft%>I5bn^@mj<0zAFbgf)W>}F`>y$Z?> zlK5h^!bh@8A|SXnFwgo;>wZaD62W#;Hb2=LP_oxT!M$61h=kHOZ)<->m+aRjxoMXZ z7z7$BKaU(rcI|bKdr7_)?|bTwDr7y{TCn*!`#Z<8TJ59+9{-%=bdOMQc^h_<&6!g_ zhulBLbuh8OBb^x&R{JDc^`pz5yEt8QU`vUGtB_y5$`#Nm`VjC(7bk|^IobMhmmTsB z+yVE7{*JyGwfzPklh!A#OlH5gsEevzvrGChW>UT+ut^KrzDtYuI8XRzUtcBNh9p6k z+2eint@W?Uy3D2SUZvza27xA%niUukzxdy2a)O0L%*pA=_)X_o$^#ONS@ObtTnZqj z+=ZibTgr;`jGKc;6LqGOYCgl-L!y-nF^d!-V<}GK(1)k^YXU^eHENL$tV&F@#>ONliY@&bn+~z=db#%!7t&zk?n!FZ z&z7Bc#rNYR_spVBTDFTGUod*smK+|xdfTiwsT6hlpFa0nWuFtr7mL|9Rr)w$bK$5w z|J*4D5in;dV84qq^WoMMPpc#*B!3z@nW_wV>Lz8=kVDF6TOTIov??7hey%QWMG=j@ zAd(yjYIas@P^~PoLFo_IwAJud#v$p0Xa$!7n{wMQ837HcNrvcGO!=Q&`yTZOCsJWx zfDh*T{eMHx&MWP|QnZM+XiL?P;KIS}X*|c+-%=wg-B5&i8>$fd(|1p}))XtZW}0_F z=UHmg!GU2%5l@d~J!-OV;)% zJv?s^oYeX zytkWQKdC7{iWTLH?&8k6p`gu&99fXy`C^4UcgA-NLyo5ZhH~m1Cu3bwR%jq%cdq6C zW9u!$qFlT8e=NGBq!sDz?uMZoh7yqOlm?}{yJP5Ph7P4`=#WwxX#@cg#orC==lSmE zfBfI<7al8P5l$UtdV?;RU~js#{A_gYk3gbp3eaojQ?wZ+6f!(KbvSv$X zBxpNyhhQyT??s|4l25%qEbZdRd^etfWQ8?+>-OR!{$N9MtTOV~Ic^IDlr&aDhS4Ch z6wdw(e!ImM{-*xlsOMxz+TM3KhGn7qw!YHtz7s3Y5+_NiM9!s*vmVQFDaQr{@8>CdeG;yC~~zFloyAmrc^*Y z(uYK*O?0?~*z2P1#(8pT8_kt{%&2;nu7IJrt3DT_pzTg(X^z807eVi;)sidCIG+5c zaXwtqzL)}vRxUU66i{nN6a{Ttgnt^OB2`19*fz+*v^y~eTV^z*4+BLfgkP&|Az>xlciUtd1*FG3%4t6I%X zg#7uD9mFa}wy@wI`;tRvw$I#uTc`@Qw<}(R%dXMD2wg$=Ul+^li{2LL?GxFn`c&8|>=+h(xkht^ZtQ zNp3e;ZXJ3(@AtUkO?FX({6W&6bbaLl7Or|}3W@yimS4#iN%LJ*Hfgh+^hb@BGx&r9 zXS0X-RX>`zO0UWfvV{DdC|W~#m0B1WpYc=e&Z#Yl=3RGe8kjG={pV1$^HV5QKax`| zKoX)Krpm~0=yRGm3RN$rXTV^6PCWBChG6+{p6(S_9Lh{%6n3OE=3j)dZlM0?(FXhS zMEEMNdq1bvW7%6ckGijq;}e}6i%1--=ALn)(^EDQPW!Sw7FV03V1ZLLeSPDSz9V}_ z%!=mgWaeQ<$#8WKD4VsRyd3bGOndpAqx@S?o8#G;41wO3&=*kg*6h1nzXaXXJP3*q zQkZG|!3c#tGa}fz)-Tma^MF{UMIRR#ldB`YhpPHnnx0-et3XXV)Y!gOl>BsQqGJhT zE#>QT^<2#_w`!O}j3KYr%?{0FZ$R(b@1Ko1-_vL|*8)f&L#h0MlZbl#+E-sSVe!eunn{_=@@O1h!|$e9!eq9tB(<%ivS+^v z)ijs!G4GQY{m%;%?a80|=V(GLs*#DFtQ+jhh7+&phOE$N{VCr!W5g`Of5UR}YtL%A zdd&Qi7_g-j94>WKDzUGmmm*(5xUkG=DwRE$#;6k0=L3r19#kkJV7> z{%MW%1twtlbUBm9S);#U8`$KjURX`uTf`^G0D+rho~MLl{nZV(SOQc%)e@_-TIBNQ zbdFzbE~?!q!458mK(I39c6z+9cz$0ys{oZ3Zo6B+9eik z1SkHeb;y|xSkn33z8ZB=)m>{<`jtz<72RBx?FoOJo8^prA>9nWR2$bjS^oW}6bc~i zE>!t}8oI2c^mvp8)ocYKgaa)fVK*O&3wcmMh)CtA>PUc+zQeyyh1Ldwv)w(ZjpU5b zyvr*W*pHtMSSiX7ScAh`XzZUwzMyZl^PnRN%OyA;y~im5yD(3Z=rnkA2O`W(HU2dG zpuKxyf+-M9?GF`4OBp%=+)8UXBqBfZ(PP_g!gebXE^NDIy9rN5D_`Rrl7>0XlLrMfjcRD8Hp z7aQAKFB|A(`wYYd_?^Ozbe`e!Dq;)UQobLJF}C5)idT0`Sz_j^&pfh@i+f42q%4|( z6H}r`Hnm!N`;}&1i9aL7y>|fe3X7wuOvj9ysrJtg$4dZja$RdQ-g!2AjJt#UN z_0Rqu_XC(@=89d@MnGg-y{c)00_R!x`b(oD;hu>MgF=GTS=BS)?_tsl&j)x@>Tayv zA(8Z`gwg6I3YiV(ilh$gt9vT;qa}|g^NY(gpU3y?*jW=IgXt6Z=>O(+>T}!`PDbxu&KW ziS{+J2O*fWCbUVfeh za}_^oAX9D1kub0~jRb{=W2fe+=QIGQ7;R40rCQh|P+}h)(ev~hHRj?@mrs9i)c6%1 zx@z~u9srk5^FFi&qe)m*Jfo%+YQakyL(Ho`xhlHkJU~YiNi@0nsk^DG&W@GXVsOB5 zN8sKDFFdd-c=S%yCHcs%jJJh0)GrG4l6X~(QR}dy@!BrjJcRX==Z|d+u3$8 zdjIT%h_ZE%SWlFyPMhF*M6%iCe=(K1=xof?s^lf_Au{js%=P(_j>OVfK&i-La>c*l zcjD)rccdH~L!voWxiNsRjeBzumwYVaU`Qf0!c{bmu6;I>$F37btZ_LW{;&c<_4bJg zPOCXpxclwmN||g>$nSP^J-WXL5vw9H7P4q*IKH$}np1Rgr$DV!QZGHe%&{*fY%9^i zKHPRAYd&bet1kgcc0;xvlzy2`WAZ7#>PzVd=&8NEtf#lBIOp{}-kpF=e& zo;rRG15vbw!^&7S|1aGyHmxrpNNSg)lVMXa&4^r8X2j|;*^}Gx>nlxci50@GhH;C- z{qO7_uG)+_s)dr>M$ei*W6eNyGC-<`l7A4T>iX#YL3z@DijxPPN=#dc`Ftx%Z_1vp zPJ2Ou5g`P^8f^LdZK1r%l8ok01`1Y%wy;cX>86UHCP0LP(y2lu^85ACG8-D3c$F32 z2`9RnV-ti>T#_ljW`{_y&yA8OYtSFb7WKzt+h9uYGy%H|dOkTu$2y-xC*EC0E9&hs zB|)H9jcu~=GoSt$qX1g~7)`(!47l7#ax=uGvMhd7wKe`l$Sf!ogAXS8tIq>_fSTp9 z*db%*9KT*V%cNxu6lRAWx5O)TGxhZyXzGzt_^6OS(8~TZS$PoLd<|gP9i(ka3#FkB zbDgwC;*TQS48#VDlZvx<*kb3TWe)6LX*nKILL3B|Dz_iX!2}e;Iq+i|JD;Eua{r8K7&HZz$T9Qv`a42UG!OYa-8Nvji zyPKmmIPkK`B&SU3Q=6XKd9$)>)Amm}jmfFBSQrdf%@>Tu_Gz&Miy+9(($%6RpWJ1P z^Taf&0_R87FKZ%#_`hNuspt$fI58ZI(p-ZVZ_iIi&)DUZ>8QVFl)60oP8#;92Kxu> z>0v`%l@9-SQdzX>?5I{tef+lkIA%TUz8PI!5`a(--sFDC%krdyjyRUR(j4F^)fkxk zJ(Xzp7ok|Vt}nTrGhHXe4QDccbFiKzqfeya0^~$jqk`^aF-Ri0jHlpQxnbpFQksaD z+!9|$Kqu{X#GeWnKEv>(Ny=%$@pV2>JJy)k(3Nw_x1B5z`#b+zX6dLfRqJ0ow;3=r zdwo;9rISlXCACzc=yrc-KTE)ySIfd6S?F~Nb zSLS-VgBZ44{ukkQm(u>dpbWgMPl#x6etjZPusCPV`z=r>5b((4+00;;yu{aZD)@+uR+As4|A+nDOJ{K+LncyUb;}QfFmfM}G z3UYrV_XH<5vU45tmf)vY&T?S@Vg{xBLWn|!B=RBfKX0%_cs#?-;Tz^3UTA0yd!)rq zjWPlwjUaVuw{VEk4^E9WOE0rI(5QTA*wES-tFD-uev~}~`^hH?uV;N*N;Pn7ra?(} zBNMUxISYo#)#k+UXOEV3NlOL)^N!{|4Q3v6FXi<3QJ}wrUN@x!E$2kF-o?o?STQqd9}v!D+NALw z;2*wE*WNs2WbmwyrDLLF06V{~NDaXGIgn80i0ni+3WuF&FrOwI5$DT;Q50}HO_NR> z4`jDTcY_t{x;%1i1Ah{=7*p{DuW{gHUFMs1O5SNU_xeh)Thy7PM`GV?bOU85Ff z+17JO)TS^=r)&q_^zYGO7g0l8NqO3A{z`ulZjGF49}499C`dbX%0q)^F1{Hv!vCoyzgrsu>;ag-!RmQ|PlOKDY zLZ#OOJOHLB*H$TDt~*=eiFTY2e|^!Wz|U#S<8ZHYB%@O~HBc#Nxc{HS10ZTA?=5z3rAXH=+W z(qfb^qqMdfZ4jV*&y!pRK(88DOM9gP_(Ls-c{AM~SqnrPdui54MVAQRTG>vtiN^uDhT92R7FGTKSD~67FR*91> zpvj5)R_W^r6G;Sed51&v&DVP$>HhyzG21F*d?fb#f52nZH{uOr`MjA4{6FX?Mzi23 zDJhP*_DIjv=bWiL6!?<>Co=zz*H^!E)iVr~xjL%iMDje{YOH6)m$q}2LX2)k>26=V zE>5#`uCjRMc%O%6x~-TbEL1LLr-QSNi0$vd0n|Q-Vm$9dhp+gfD~JXpYGE;yurAkd zr9VSMp)+I{0CfF(x^t6)wU_BP*SdwZ(DW3*x=%qtOLw=cSnwJm%NABmo2Y)chYC{G zDb%4ZRKNIu$G>$4Rwo>nwaM(lnqC*Ps*4z-EFIL zrfatE_rCgu^n4mK$0?Q8csD{LqIH^1rc{(inNGX|?AJNNiE)}G`o(IOek-(8tc)Ib zOQ>!*?X{=1CW&~}xw3anFgYYFz`R+sWA22MZpkfl1qpJh7eUW(@pgw@gS<0-Fos9_ z$UhY1If4prt`cs)E_j#p5@cC_bCuZ4`Ft72rKE?+Na9CS!_wk1lS-DI<%&3JMs$a3 zxoG%;sYJL%g&&Uz*Z03`M@?J~%*%s2p!v}A&#$Vy-F|QV+wI|}ap_Y6PuJ#t)32|0 zlo0=YLtx{mnl(`6XVW{ypRPr{Q3llZz*Rqhm!#_9D70p08|-uL*km(1D)x-W-G_q1 zbpZ_E^w;{renQC?CeX6t#Xa;ae)J{!@m%Z?{X8qUe)puBNt4d5-l{a@^$xz_31!|9 z)z7EOLrJa4ZYw9CRoX;|s0FESPQTUr0zrD+Kp9DBBP7e=cO)|n*nR&rEzTvE$z8PS zAAoLlpd;jt(Iv^SqV-fSMh{2A!f?)E2%i=v9A4sSGv}?wkSigEs%`J%0KwFu z=e4c9PJ1HoSm$^Y7cYt8%D^5UYVS}_dU*}N9qqn-nysDG=(YP836OJpYvMN#T=rEN z#0kd>ihHW*XIHb_3?+?Q9P2ly{j$U@MzvhACYfGMuJV15`B0S^s0M$bsS&*1v4;0_ zW(<*{4B7c&i`In;Be~4e6g(osB|K#zI0V0&&Mf0<43cnMq?p}qv|8l>Gm)?Jhu)V2 zy%w?XGKE*s3viTE)V!zE$S+hA#JKGB%&yKAo7Z|C`vxB6DkneRied+wgGG->E*?}( zB&$1l>EnO=MA77FpD;{Msw**35^zBv0$OiJ5& zTQFK55coSU`>2!T8N&-8g3R>5#8unr`gzQNPwbFKMd1l1_;v>nWptzvk%a@)r} zAfV=JhGWSsVI%c~08VSvI<7?VD*lQ`uUG%M!u{y-pYGs~ie4NE|4cOGaaGmj&_yI+ zyRE``K9Hom*zsvo*l}q)1fEC`$dE}+5Y8%8IU=I;VP?21C=-{}s;Zqap>^B6E&KOM zUF!`$p~6jh1bf1lMS&N)XFbMghIdUa@|Xu+NYwD%xYd9v4SS|ZKX{j3ULW-q{8I~N z*-FCLSUj(9O;}QrG#`+fy9^&BKts>xn(@ePQxD#HLjLQ=FUU#;a=$Zq=VPYp52aMv z_&(Iu8Indt=%4~yFp-#;MrZ}sOt2IkA)yo}d)U}du>wqv`W<%7+3n2b>zg}_lSY^(ZWtPkG)AE15 z5}7D0rPm?jn_|hKH3}ZUXe$TtX1jaNZnVGlFi1BNcuTDKBG+&`Co6KjbhM#)mbLOj z=zvz)mR!g(t}OCqacW9Qt4@E0aM@9`%IavKah8>KqdN-F0R)NsEbiOJmz?K8To+3J zIcuVe6(S(X^klJ;md5xgJHyvU#jE8b?uV|t7wzPo5f%-F@3$|by_?N4QnqU0ZY0%O+cYtl$|0HF_GIA3a<0K{w~b-$BklI{UzR~; zU5=?U-i9{U+MVS?zQiUT3ZXH@Ce$e7U(GIk`vp0F+XF?OpbxNaw~HfsN&FSql6a!n z#7b{7^i**iRJX~Y!9qXa7-@t*sDfd_F{)*p$Va*2-Zgpy%Bg7h9UR)X4$J}k?04d9^+O)RK9RpOjjF)9dZtPz z?#j++COtDk`X%nwH|ZHco3Lqu>*_t#0Q>K} zGEjiOl?-viiUkdY-X)W*=R=Qz^wM_JM z9nA!ez$IVl43___fx|-ZR^6dpAi8oo^^eWgY$%}}*Yq=-tS-30@ww~F2@@z%cO$&1P$l5yLjjqyUVHX6D|n~OiyB-Cd> zdg=_Vre8E2SLq~=Vsqo+^i*ciEoL5Txb*x~sy{+fw5Ei#Bq#*`<1koj>Jji;W2>kW zQ1-d~=RXH*Z*`Y-|3#QzR8Jc9u5JlMJ8Sp_Y}_P@;{kV?%2fTI`w~KF~TGlX})H zub{2wg@2|TJqV*rCUmgTePfSe{zc+N3T8CO9vsWQUd79{DrTfv$)N0(^Je4oLzlcv ztGIWYW#$70khymt>2PGGX1`nubDef0Nv>*A)#YjUQxZX9_4f%K!CrA6h$QA@XnsBZ zIq~%D`m!=(ohu@w(`b(LW$(jbP{Fa74xJ8u+Lp=TI!EkA*2J70z;$m{_~$$<{jhv2 zKa~AF0#oA?Nr>6xyTVP=$~0}9OkgtO-NXuN4HOR5e2M7z)Pz7Ohs|HpA;UH91`*qy zDnc_liAYZk-UeU)Y_OV&c=!dFeNn7hA8Kv!|M&zwOnI}b(c*8oI*yZL7Q_L4$@igY zKO}$E8M2{oGifr0P&KOQm3u^!@9E-cY9rl;Wd0$z7k`Ar)h*MRWb?3>^n}w_|t|(eoAChO18~x0a#(mas&zP`iIUv z4)e?FBl#|(&bE?F>B9a3h6M8>mylvLN^!Cj8C^dYj&gHHx)esy7X&p0_ikNbee*wW z2eU66?|CyYrcp)Lj$CPYQE;b{zrT{4sa{Cahzr82GQMZelnyl`UzKOQ6d3qNpVCub zwnmna#z~Gu;K8?C9BJG4yS1e=Q=RVAYdk<_%cL@++GOA#<{zIG`dqO+=X=F}nXU|b zGFq3#qWgl}`~AlxNy!A_TC)u~wP2WPv+~{o0%Y5V^HDxVHu|VZ-_A5&VZ+RuQpuuu z3P<)RW{)XaUPQD`y+p1{}Ld!*?K3jms3(kHV{!YCe-wV?oTZbh@*nHcd z&AGxD%vB28nnHGmgn=J)t4VM}8Kb&}hk5ev%76_1j@QraM$1?xGC*GY#~}BbpB!Wc zsxST`Jh5VkoD4gE3EO?5+;?vYSN<|1U}yA+9Ut5Q%i?^?$ZiMatZPgebLrSTTD=Q!~;9^#% z5$d0mkH!jzY)x{`*s5gd-r3u<|A}+3$2-=+Zxvg3Hb?+it7zXy zY)xmYxvQ%LeHnApKE<%?@p)A;;)9UXo9byo=z4h3g(-(3dyA`as+VG%=-T4C^|O2V zG1vq(A%~yPr}h{>!3Y^Q(ZRJkKS7b2mj$!mjPG3BJ&@_fEBJ>D=~_pOU!of$FI~v7 z0w)GZtwQbHAaFNuT;?RTd&Z;EF~OwoPXU3TaD<0h!VA%3yJkkZ_>26j&ipO0ahONZ z=u;z>j`gndgLS@YT8sM^ufvQBNEMUL+&oewE}z|`^2oP}dsi0=Spj`|HEBFEoK<#D z;)v5;4A5iW71qD!Y1maSXr}`DF|ZEt^aVkR!r}oh*jhhj>_9u%X2i8$x&Ni=F4xvu z--08UO|&o7h9FveonG7EIN<<~Ey7m5#wkOsP@da*W~p}8%0r9a zl8>4e3;p@zm*c^WQPKmp#iZQFkdR#L>7WIy!)bz1IAlJRu9!}fa5V+4XWSuJO8xVO z>R?2u&a&#$1WwKIYO;7=&1$E2<9DE}H+l}UTykyu$fLPm7U;sUHm%q>@U*G@K@BH|DU+ z$o0ruh!Rd1E~Jl~isa69`qZQ7Wo|sYr0GOO32k9+o|5gkHDpTrdU0f-mEBs7ZmxLM z>xqone$*oqC29$CU~<&9>afw_Ce_hKEcpoe3F6mh0@mkA|89(mll-jYwc(l)GD##U zkCEZsce(F;k(vzX^sUqsOr#bd6W-6Kma>amfirtXSg%a&NB%5d2pQ%?sW4VC`K+eSu4R z<#mzxP{AeNRlSGzD8dW)mW_nR1GM3o?D_rIvWD?=gW8xoox>~Fl!Xm*C12kTO1NHY z0VX)Mz>!q*609LbTExaaOk>0!CAs!2ca{Ou{C>NO^V#N~^dXyMJ%40}QSEs%Gh{LK z*4_Jl3Y!rOVEiyDsL;f?n~SkWHNEHL&?`;kl{+c=j)n1x85I@KYl+b5Z@3BNU2G0R zRnWXKtHLx%)CW{q=ze&aJ4hB=2uBVAme^pM{Nh!8TRaV$Nh!0>E`~phtv|U-;9DpP zd(0+(n{Je?CA0a9fEwEAv#BchST?JgEtNyQ73Q?IM2=}qL6{x~v3 zi&#Ob2K4rxtA8kiI7H`yRX`ch9FHE~q!9LV+36+HkHNRTmuvohO_=H!IBIP#B&JTw zJ~bR07~?*=Riy%E#KS7%`HJIm{X1a{+E!EXdmE&g`GNCyZ!Ehz+P@D^5pFpCM`G4i z(XVM1n-OT@BPpoi`ZbFU6>@zHBQ66wO3YUc8apu~r6k9+xhfkmeSZq$e!(gyNTMvJ zWJ6$HQzTg~Z>|Q2Mc#(5ER9um$uEr6zeH;UmbXg4qQ`_+?fer>Q1587#egDxe&B8LmFhk0S#4_2>ehtGMAvEnD{W^TWxz7m3cH`tdN!jD%` z&U>;rr&S6hLI^eK2tifDPzt4+W$BUNcc>)Ap6oX7&6ni4J3Uk0YoUK0pd1uWqM~TF zIgpxS*(FXnTIA$KB+Y--ntL2meeWHy*%})yHxKRCguQ*v1tgTXN^`zSF2%=>q8m&s zqEZ-g6(i1d=jy<38*zentgp)Z>*d2+ z&6%6!D~jb}&C`qg;!eSNYLUy2lX398;d)kq_XY&(zOuAewN?i%l}w;gryol$2NK)| z@WF&8pv4;jTD+aFxaI)Jy=GD%-erN(pP2ZYoHj5k`|5PO>XO{mv3mn=N^3;qNs`Z~ z*&^%lWgv|Z#7XrWn9M&ObVt@0-ulVp*_kuY3OX}i4zAW3RON80vU zE9s+@{{rO%@}D0OwG5adpih#l5M}ggvXPmAq-aO_UP_N>D|f{GNq0sr|LGL4dO)|e zFcH}_5?YKUj`0nYe3-UII<=^dhTjkrZjWy>K}(gYfE$1;4wIqVt1;{2;|xoVE{ zRd-J;3^ekNqG2?Dol7+91X=UdA?P=qw>hZfQ)u$?rsh@3qqV_72gH}Eo2VGdb&hrG zV{T-!6mcF|;`SontrbYiN(~3{7*z1U+Irk(1)VgtXQvSrB5*I{V zwZ*=0oi+@gZj2?;f-ISq*js#TD3Pz_=@_AA%a$DJPJU1dMXa!>H4 zXZ|AC4eSrfpJ<3(e~eiT6CMa@sJ%&OZ)C;`DDau0VM++fv?`o7x5S{^XGh(oA0#Q? zSw2aT&E+MiO8i-ZdlVbSHhq&4gqt{xtJuxYk~3|qnv_{@-3TX5N9QX{4nb>apy!&Z z;r}8XKdsEo$dK%w(?-3-Yve;ulms2@Ri^Cv|I7*Jw@w|EDVl!~&xvK0GIVsrH)lKD z1xO{K0YmEofhNu62*9U`s_e<4_lGQRHR-><4XX4e>+2x*g_#sf0ks1Fh~I8fqW?l2 z_?gh4XzntbN=XVMmzDzWL%>N?EeCVVa#zW4IDwjAGcW<-x02 z+1f;Xxs8DDO4X%j)meu`j6prt1XD#QF(YEmBvM(kv`XbY*Xe@oSV#GKt2J~tkl3HO zGl%tNZ@znZN)&Msr<&oQ7PTyT_;Z#Xgv^O4fI+tM4Q?-C>^&AexB5R1=JLEq;eF(o z&o^B4qfW&TNn+1ftE!N~_U_i~Z{+~8H@TmgM0h^zWV?;iUuhwcEwK?}j8eg-nm0rV z#8-|olQMtC+O90^b@9<=!J4CZLvfnDt{V&NPOecG=Ez`V-|drOi!*aX9Q_(R2Lgj* zNU~%*+ACq&MQdy!MjD*VW;P~WB$kU6#1>W&XeNY^QW7$Z??p>NYcr@R8IC!n%lP`- zeX?(pas+8+&3TvT+xdn_xjcNW&*GknQdJ1DpuTK0s(!~CNLQ0z1#T&54SpQ;IXML2 z;Ji+%B5OnfWs9C@FDZ3Fc~sJVENRHS?=VTBWGe0U)KfY(3?qG(fATEz9%$#2P)5DV4xHU&&a@FHlM>u#>Y+_crUl&I%9b3zIB~K5+yO^ zcv}woQAviml1H3Mq_Y+SulMN!6S=(bFp5Q32;>=V$a9{W+8Wax%?--fRjBry~Fm{N0Rp1 zl9%}G3h%J#?9ysr)GZ@p1IZ~5HN=0sGeV#4t?&BF6Z#R6sr>C_y+%Vb?LKIcKWe#b zisp@=4}I-w!CtIba=?n9<)b0&Onb78uS|D1PHOMW|N3X3{m{YAjV~OsoB=?-<~}-O z^lAX-5l$G=65-~h-mD6b2*^#%*fS5Ts6LhA4G2W+pS)^!l_}PXk%*mHnyRA2%LGkR z5Q?*XR&)EI+Ti zvp(z1`az{&(-}#AdpaF|D$NClcc-h+3W)Z8vbMs z9wqgtb*~Je`@g^x5qnh~;Rcy=%F*8IcP`sx3zW$58>f1{6!9&jdLQ`I|$+kF~0& z^|jw5EU+3=Q!E?JsjL1J=e5-w<^3Gvt*8v2D6r28G?Xt8*XAlZEZp04>gyY8Kjtf$ z@uU!w?RwsiQ`@T9t}yK0C=KnnQltUTLGT=4LJ2Y}vo-xjl36m-_oE5|_v|SC!R7Q< zcXJBMHbiHP<&_@?AZbh#qFL^5CH6;I;`IiEpIZCVkI*Xm42Jvv_ob;Q#r_Z>U12mf z)C%3uSjqwtIjn=(dLNdl;G}m6c%idR(@`E`ea#x-W0&Pe+`GXAD8O71pesM>Dj7~C z)M^oP7fx~ET_35LwZ%@AsI3j&3}cvg$!ToFVmMHus+QGcF*WD4q?rqK^ANyY!z01v zsp#H-6U~lzlD5oUHr0-`p4jEL)V!k1pb|XUhh;c_`B^!llbq8r-N;V2B*dT|@k67@ zV#-eQQIM;45VVI*P1)?!G4;+ZFsl9i|JWYP`V05ug)iS7QM~|yxQU)&vNn1ydx8FOBR8!420@bmjMUx zflp4tNyA{n+goxa>`_iu>dP-bxz0RUJxc41YHyop2but2NP0SMV9Po4TB&-1vS-?X zv5H-uLz!m#0j83|6r=97)I|;EH!oXQLu`fxfWsb6HZR~7WJ8ujqUJ33#zoOA+AuNn z#%~?0)HLL9(0*Jpjv^e$&4&Hno$}R}Hf;OF{#kaG=ErOmlN3xLEgj}s?mdms3rr7Y zVWr65Q)y;21BR)0r-+C*(`sMjAw>)(pC1+Km(qIEl;UR)MCuQ>m>4_wxE6N1{dq+m zG!nzwrCqx%_gihM+Lxf5->qn-`z5rKm;Y;tj>`5^EGmCEH7ko-(`O_*a7quD64a;sRg2J$y_jHWJ}Uo$#7!hB#mX;{>`$$yq}&* zDv-_K*~um=o$2I|QULHgg2M@tf3XMxDK1R@;sQ6Wc{0kJEbGP5iC+88VCxvM$V^WLR|E^duzjMA28WWP4zkOTQ}giI3OYN@HpcNf|)Xfm?||n)w0(D*e)KU ze-==71asj64d1tdMS9EKQzAo(H$wApqB2=3_T{+5fK}Ox{FxKBupJVc^)krs$dUx) z&MATfh5gYkJf>$UD?d{m*bvz4@#2 zecXdAcGS6?W@p;`NF~nK`JLd)j<#2Ktm@eNyuY`JR*0VY6-*|{Tbu#Eg;|&i=pWG& z1`;(-uNCGQtrycx;@i%0ROOZmI?s0E{;}1UM`)9_X-tc7K^^*Gb0$44rq9m!qBNqSBi| z9DTX-FU$03a$r0<^}D~C+tb_al@%E}T{&W>4^NU5Fw5K{)=rx44qa^fW%bT~UnkgL zm(7xz3I(mWrWOM#aqRL#xJOIV(j_Wj+9fSlEo1kd@H${OjBfoF;PFmG?%OmRtmNTR zinx15H}hwbuxE`Xpm&U_X_sPjZRxxRuUv1Ejz+hyhH?=@hGnTb8tdl{lH`rqoeAuz zKyueev%IQM2$ff#x>JqE1QAVOWz4&e9I>B0-FlO>{ZY0Q(Z)xUCM^KYUByxG8=BOH z9<89B?uY@c3@q=o`0L%f{aKUN2=(W}tZmk^uc7(-BE6Y}nI@=gi|RO!cJ*VXghI7iF?%&;ruXbt!YB3?Pa1(eA@F>cXEuNfL>(#Gf5Nt; z<5dyNK|{i5K_tuhT}6 zx^ahmw|@xHdHtiWSCLks5+ycpGpSq3+#yyoQugpt@XzdMxSiMVnG7!#RhUgmSHM&M z8QXFGC73(VZ9;WM5n35&QcCQlH*bM|a3x5w2OS>C0LoykB55=i>Wb)Fv$#db^H*O_ zES0wXmi%m<>&uuN$X#L)bN>!{!Tg2-N)4rbeaDl9S+YK%O&Go%)d2fcWjj?Erls29 zSK9~yj^*c8%CDM6*ntJ$YM{a}Y^qvu+T7-oc2fO30?U?H3GRzY1*ij-xw{elu`RT2 z-On5-!w*Nti_O0&TzGF7o>@gL=&WS7W2v;v#1(Xu+a~h=8g%z~+xO|)*}{-@KYDX# z>-m0aW69hEu8SBFNrv6C{8*II^#1!HM&= z3A2(LH^EI!M*uwkio}p=%Ld~to}AKq@vKg>5GNiGvGG;u>O-ZnWo+Z1KUx5mEy;dr zA)t{Zx{AAh^y)A8Bpj-m%e|p?q0M0EIF5%-XD+@t3=_PIu~C&d`I znq&*q-&9l6C8JbOI-*pFmD!;h^9qG(QgiD2_$L0C&;ylJTtf!`p0*Q_ zGR^)(^+mOgt;|1r<1C-R*HSD;Z>wcA+^yUN^Rd9tEie{Mcy_~%e&M@k2o=o;U{3}i zu7Fm4%=D+&>dqfHdy^tMLSIO-sinX$(Kf-Zd!(w{N$gV0_!?qw=MyRk>5nwI$TCu_ z?2W7%X9^9wrR<+GGW>H~G}SUMm4@bd z@~A}fbceEbTa(wFeR|({HiO5_{!V2|y;9nCOb@#j@J*4PGW81t9O4ZZ@9n zxz_bOu95qb3G#MqO`{{NY@;mDs~U~r)F8aZ!O9h2E}3naVq}g>F*mZ>F)7_CDSc9r z-WlSO^Jds1IcjW3p};TQ&S?9~30#qv1z14=eV4s`Y1sn3y>-klN(%OR-4v6*8S$sD zRNOJy;VwDKeA$G**Batc>2LCJMt%e$y$5B`hQsNcnUAmYJv!Q6mmDbZ zu;g)6V5?0L?NB|(!i<#29i@F$7<%)rZQCywi%1iUyK0Ihz}c5ebZcj9W?^cSqNk*` zik^jLo{=&NUmNAqWC&p(5>~AgnDWD2f1CIrv-n90-`Q6rkH z1-xFg#7$Taq0MM#W9Gq2TOvc5iX)u#(b(SqAB&wj-%UwOzGnY+R<(O|bP(`ElBsTO z9oafw)bk5M7FayvRp+nAZ8(v%R3AqmYM5DPuSJWh)v`LC?zq(?tD0iGs|bhSZ>XXu=*vGzTsr4Y;*;E*<5T2)Z?KtOsnu*Hbg! zlwik-&)zFi-P%_VMc)b*53s$JB^&8X?$|0%GCHzn5l={?7#Cpg}c+YlKzy^;`!{*hPn?*-f!MdV}xv_&u z*Dy@SL~)EOEmK>QNmJS!uTn-epjW6SGIrgP6{pp-47`|~b`ZLLiFLKc&dqF9N9IKT zEMh*PRr{McnQ6J9krg`p3FlkBcPtb_Ps%KgUC5j3g9WCi&oNOU)&t0)I*d`z+EcU( z)WN4_lpxRyIvn;B+>zDp}?@4MSKL~=USo5tq=jT_1XlJr^IpvuL_Lvt7AG9fWTRjFLQg8YMEZb;hXnQ@*0vbHx3{P;pQkST>>Gw{g zM4dtCa_L%Ubox^i=MWrUT0+`4+wM8PetX_SYtNBpp%GJpdt&Eotz$ndSKdd7zK2{- z%&l~2@S~a*dj8$Kgt$ZOqm8&u$qxPpyP5Y8%W%z};5OSphfR5Y?Zyb93K8A3l&qakFCF?u7CG+L_E26f2(wHe zT724$YRO?=7F>ZzwR5@3YBJaG_RJw8ie~aer?>)5o0D zyP1+_{%jUIhk;#@+cVK*uW);3NofPoNRCC(m%~KY@}?;3Bb^WlF0P(#QdJ{(;Vf({ z?{U*mA}Rf@2J59VH$Z)0ytB)FzBtOU{4KGZcIzJe>sRjGyjt-2`1x0lJornbyc?*$ z)lZ?n2&+ZsQX|F7euQYvb;D1X-Ea$;A7*UtcRnx7a=v(`xk|b#k+3rpNQ@%>OG@z0bN4wX~-EKjB^DI z&<`VdZfob=go=RZoB@OlW?4%&@Cw!tDX#4V z^sy;81h5UhY=k#k@!qt4at;_z*ab!$Y2I+*te)xszWz};g_vp<3O&I0Y{IC^JVyCt zhnUwu_W8xzh(#}GiG8AeJI|M`m?cl^i&*g&RX3@y=x>8Gn)p(G5mH@aXMfTsz7Tzj zzr^!8*=s{#+kD(t_GREtx2VJRkW@D<7RunD^!-(^wnabQXL9tmHM9m4EP4q+^bEB(w4V&;6scFlkM-vt|41IA9WquAq#gNp={a-ls)V`#X z=jOJI*0UUSB-+%BG*&2*xQu^_M-qv+K%uTn2cc{pwlR>gSKvHdW#m{48K0%X!?z$w%4)W5yF4vSNNHyAk)09CbUQGsvmS<&B~1CKnvR06i=x3$TXRk&3?KX&SF2ZKc$$F`kZLbiUkx*8v^*}8{@JT8i zAjH&d#-Poh7TM)rj5Ff+){QkN9TGbO?=WyLk}v;mB@n$Ly`&aB`Dc^KmMnJM;d%pv zSIeiO@xXX@j9k8T_c2lCw}1vy8L3CkiSfTt(B+^=MXj?d;I6na)0Jv;(;sv@@67B> zu)exusjgsg+3pj6da6k}*}Sc4n8bI*H|p~Uqqtg@Et)d@i_VP)$AIo2>3x!&qbvz% z(2kJ|XDrF-a*TG7BtmS$OFdzSeR9WgX(Pw&dX5b9GnC}CsaqbAm* zv(%r2{Nq2WQ2uwOuQfkuOEnNa4sNZ1nEZ6hwz1H4q?Qb8)`~G%Pt_&QmVhF7?0#HPUbOAghtyXE0fVf3)P&RG6Lm3iiq4+}93=8$$cRGp;y$d3~Wy zCI-!U2V+s|f{17-HH%vW@kZ(+8Vd(wnPE!1{uP!$$73z#2D|wJz@KlS;6ttSC4ir% z4?>&Py`Q0eC9q{RTEmR@HQ#-|B(JWuZ`84_9C`9iCy3+UqKD`uFb3eWw_bglU zTuOz^sw#rXHj3sgFL4w1HE`yb@(aeI0d{(iPc!|D(#EX+jYqW5`|JIzj3yOg z9MglS^2e1aW|}!!GFpD_)55$69AstLqUOX6xK-aFAjxN^{6fImWL|rE{>=p_+j_ey zV<5$=VjhCctP;QM%D0WT{OrXvyryKY01||r_x(ODi*IU`(r#?Yy|hz94NsPem3=e8 zvK&74IOWg239M`-maM`+Lv2&)?T)VOinB#B|9%GMlrBOHXrIBPvt1Wd+O)kR>uyB! zuy2biDN?~AQQ0Dx_Gk|3Xi0H>%xfw<=LnE$Nl23TM_b0D_(BLUw*b9Vv5dchA{RK} zMe8$~O8G$#BpZP=c!cz4C_ue~u}q9ikr+pHo@EnX;#c9%@>WyM!d_b=H(oMqACf{q z>`-`rn$zuWC#~!ajR$R$=@^^CP|#ipJ`Bqd?!@gXHeAjd}!r(Ju7sK zJQhs59-A9!ICevA@ao)eaJ+s#4q`E?#DHg>$H76)G3Us8Rw~#qRo7MD11(eT9&P&q zI0UJ5ZQK4=d zy@Os^Q+vu!QRQAKRn-_KaLl^0OhTvBPmG2Smfl~b#Ljk4-D7RV4H=l1W<|L9BLbNV8VT z^aQymzUs1-sLMk&_B5fc*^%fRaUtZ{wPLX{IcZN4JoT{E&?eEwtHSIgx&?9U@o$T= zh_?be;yJ90$=%=a7C)KS2E)|p)@B+iaJ+JhZ(}QJt^6F#VNj6s4)!$l%f616O9=$A z@e&w;0SIS%+VDb0d3COM5-RJ**XM6?BQ)K+A~2 zhORK5&>mH*Z%*uRFiAwQg!NIZe(*izaF>Xw7M!Nb>LybE}_t05!j!(p&Wjpv?^CzT2FS=1z&8U>7U2N=Drylyq1mJ|P4s z&jVsom0zb&)1h&R1+n3Eu{TKWd3~z!Of_25572eDGIQpDffAg=Z!*V*p5QhquP`V~968)8NK2J;QWA%U z{>-07K$irqb3Jy)Z#mCP(Beb@Nn(Ri!67sUkb#~A%WnU=Hnsg-O_9Yz^C$Yx!niC6udd76q+h@Q(;ac3Amy1?F zpyNvtf8QJvHdwt|$yeG}g2B^}jlpRu%Qbs{JfCehi@55D;B~{xI@*AoK0@^QM*HfN z$=d#|zt_L7X?)+^U9O8M(t)Lvn>15aX;9YO7B0EUAcl4G1_lv!8f7F9+Z1qXq$RC+ z3r>HVl{%KA`o8uks}G1ju*&Hl=qR7PP)YbPl9{e|T_tVq9uU#10`sIK&fOu*pQLmo z7f{c=|Gk@m#_LDiL%y{~!9pvlh(Fg2t7rtr27S!Bj52H4-gQCP_=|^rVjnxbTg%Y= zg|HgwCq?@++w@2h>4lKF$@B9ULfd6O=}IHDjknc8;(;ON)L{SH-*Yja6CCHAT z=T#BYj(hKyss;JeNcydQN-gNP_+CYC+OW>n0gN@ulmkf?R*EgF(T!RzMkJIgme2cl z&hR^FiC3TD&(Dn6*f%QjEjJ$^D%`ho-?TVP0?{mSL6Yg6a-BXgXf1^K=SgsAW4MOW zQ^InI8sqIb8QYP+?xJ~^_xPE2Ypjo^O07Q?nKIN4<#|4n=8{cin!i8VGP}Blz|(Tz z{1!nqpvy(lfVNzM9X%BL(nM3$f6F&foE6{Paek*DjYGCP3vWt{n}LSV`Rqt3P)wlI zc^g9I>nBR-gxMOj1=g*qJ{)UL?ZV zVN1U^_t6;*b4j{*^3evxuG3TdHNH|*=SGUTV(u8|02hkTAn7jzARo4eycPKx*k8{P zPdodVx?1Y{Yb6M->6_&_S#Ih@>cg5(PZ0y}<95F)mMG!rbuB^FsMeV26gI zY?u3DNZX5d4oR2_^!?H!5npF3?{r3>tcReB+;S;%mofA;v7>MDq2Y{juCX@Jzr^>? zk&a7_Pf;Ihu&v`A!U|~0PY&+!K8L_?YAp!s#QX!$Pr772muN|vR4NVKNhq;j1jA_p zaSJWhV5#MGF|{NJBKmH6Nh!+7PG+FNWjs=h8jVoHY)Yhkb~Aq2emtF|?b}wFEcYr+ z>hSL_jpre5N2>4a&z{{DZ}`z07hF{|AZ9Vrln+l490_}X)g2DeBRJXoGi9Ftbp3@e z{cs4Fpy%)j3N|T+@Dwkrdo^uh>r%>jDc?GL>_fCmJalWUDi#VI3QXK_OGu4bl;!$N z8SqS;N4r1?dIxWWe;hPJl_)3v_+5Oo!F!fKs0K@|Nu^l^_E;mqN~6n;h(ocqM%fNu z?_x%hyC&G2+Jevj$tJFgjk%UI;5wyViaU94;G>D`>SKrN?ecf_jm+LgEIBvyNymo`v(O4I8*nqJF&={eWwza9_VEPtX4rDYg?KUe$6%)08G z+dU1RuW8Z05Z+qr)LwKw6mv*=rfn`I;cgXoV7RkECVgmfAIwNu?f4|-E>TZD*o9`j zSkq?7w3*;}27X+|qc02#0KXocg9x#;W>l3sh@ksZRoRAaHT5L=`j^i&(}h<*l4Uffh@zBl3H!1TLhV~p8G3!-&@FKKdLZaO(%`#0gNEcU@+8lS zLPeQ;_r<|oOfk$sa8LqDi=KOj_taE7pkFYnGDvo8t!~@t$4V+iUNlTkya$?m=^xRcpQa`zew5zD&YeTmf^iqi zav0gJu#>batOLm`>A$Cdb^SgeT*o%Hu!_Ew8uLMnWI+GYI~^VY?;e3>catk>mmm^NX*&1$SUx-LMiOyj*PYS?=rtd`Rz&KWuBN=Oj!ZCytF2X z^r#Xi4kP7hZ1{$C5siCMy)v~m1Cq6HJ4eY^WF#g@F!yAr%onuirgukjThhsVBeCNn zpPk((j@zfX6A^4Z5H1HD_1b0~G2KUgy%QHH5m4X0eaetK-NH_=(P z#j>!k>UIReEvw&RSUYznK*{uc^^IqHgX|^7tuVzN`t5_iy=vYSk|({jNxCYd$^L z^yzX~L+uh0NP9=}X6nawVBY%Nr|Kp8N#{eS>53lsi}WINlk4V18XuAYVwxNk;dYJY z{tIBQrgQfV!*`0v+B3H;!HWs;c5mPDi@6!prAUxE@5xo^Ho^SEHXqkN2}%P-e!?Ru zJd9*epP9Z(@FTnui8FJv>zup)CYtq9lldTjv&d z&ZfW6`%Og5HNDU9p1U*_rJILA5tVh=Br5NGH4446OVV)q?4xy_|5Fnp0XS`$d!FM7 z?l9SkOUaCk3Me;9glH%;@|A6_Y5J%-C-+Z6-Ip(es*Xna*mL@gE>y1SBMDC)A{@0h zE=0vM1g0x=1c*G<#Z{}jI4Y}?0R5S+>;DVQHcKy z2}@i@EW7MO@I%ivKMDzB+FBy_;MDqY{{!V2hH_F|sl)_O87(wAEdVx zaM>nSc^L5HD6AnQRy8Wiy_ z)M77Guj(D$jB*6}By!}`oFqLtt-DXi<$AfAWnLN5CoUJ>YHUJj-ebHDh@lo5*h2Zn z+>NFk$FgMT3#4(L)*IhTI01Qo8V<6b$)st3rj=x-Xj`JP9hXdu;C|*r*O@T-h0s%G zM+!B$Px2Dn&K`@)#+=#*(MpUt8ZcBbOKCyESlol0fAjFWsrVx~RG1?hkoz-HP(H^F zVt6UGOje1CfgI*1!=s2$?obizCoylOxmDuU>?cg50u+aoRS(dVjAZHeaCdnxxr0#| zi!9H3yzrxhQpu*5tX`Hwet7Hrg$w_dvCYFCigeM!#|k{CQd#(Ku;Ijet=z`i^joPGrZ_5sVlEPRcx3{o zbH%~2@uo94bv5S8Uy)y9K5IdPH+KT_;7xG z>k6hXwk>euVH;{2X~k9IVKB}z>QLULB+fV^>Uga-O(*FqK<$pyfC+vE8nH7%b;Xrl zl~4eu``!m89KV;cxP&0VZY;C*8SdvNSJ+zJ07l zrm)u0UL!BhCPth?5g>}I$`(=3Ui~76_)`#xW8&4CFE=lTbu(RRY0I=|Iq?kJyG*t5 zX1V+TniY%2-B7Z}r#A)&bLsIM=?P!$pO@i$?XV2cy#jbM!R$A@nT&tq%}jJxl}oK* z2jT3KplN`<-B)}|Xprq(5Q7z+g>eoq4E;kO><=dKKRDSE@~Kg_w#)lXFN`Yej@`F0 zWZCHE43kdMP4zE11?{1O5bA5>9EQs(@9+zF*73|5z^hFyXC&OAzu#vV(~&KH9h-#I zAK`tEUIOmu>L+X`U}T&ZeHG*xLdJRkfG!^Z+0+@$gwlnff4t{)~|R&oCH{)>RyU)LwdgiO6VRel5CWjIxS zrZF>Ch++vGsXgi1j{i+(2c2~x?I*hFC-i>KLu)E3P47Q%$BN2R*1t<(7X!y_Tj05u z;(JK*4yQZI@2A^V+dlW;7aRBQ=0;|=XI{K;Mqx&-ahYDZfFE1Dpf;`CaHEyJ!Xu{7@QyFxk%PU8(aOj<%|))0`q_~*9{s| zOXVW-Ss+78J5ag!tk_);rjbOQ~TmNUq{+>(Ai@Ob8#?O%%dZxy0e4|D{@ zJs86E0|?15hkL=YsvUPSx(t9N8CJXSYSnGx&B)g=8GC@u{OA6PPWdlhV=vT%S?U94+LKO5{st*Tl= zlx)(VYg$D(K)tJN?vXS?7wK+$nr<5V%DV8T&d~K!t!)Ysdi{`d>1ag?^?(`&-`d@- zn2DI&_T=Ow*?P~o!e@chre!??^H&ya0B&U<+9|j7iKepV#hZ!23GH~y@~KN;xi^+& zda8c&<@Sf5;xNp=&0)#{SVvfGP%ZbxN9?0*+g3R*KAl5VGR%al@t%(Swlmg$@Z%+m zj@lYZHJHDo3HdpLr2V!%>}=^LrWgd>UkFu1iJZ)4ijWx@jm^lc>sof-H4G7qY~^ zyp;cGeLIG@Ihjwhl$(AT09Th9N z&8$+kAKgkj9F)ipDXot5<6nDl-t#H_IN-`?8TPIHN7a$3mx0gpH}*D`4_00TOZYGA z1kn$}p3RXVxpqI<|F=(gJCRAwYTm+1jNW%-zNTls#h3elSM3+VM*f1{NOz8fS9FN_ z2vAe=!Rq16KN$u6REmyumb$E3-$zm{;x)Rjd5PcZ_4F?I{bM%-qK{mq$Aa{Q82|0D z$`AIvUnP4Xj_NMc_ntlJf9;9?r72>S{nNw~*ZfP9o%ucB$^alQpZ6Y+>*x61A3-AW zK2*cG&Doyj6i3Rgh;xB%Lx>knDpV#HmP#{RyKvDcacZeO@Dz3`k!_(^x&Bi;dOFWz zZ8h0P>RZ9f7(x2P^~ey&CFP%g2*b-y=}I?Xp(A}0b1IougD~nc&~LkdpheMK5*1t$GL73*h($+ec={k0pd{VH$p>W6Ulnr z|G_*x)~UglYsmraBcpa1E50$hlKwXy@jp$%d-ug5NEv5Azk(hy72)_l9RB}wMgO&j zBi+>2Aa2dIEM0#y&spA)!n;3L36@>q@pVQSk3OZJCJ_z2&Gs9?JsO`)ZN{scqkHb8 z+YrwySrdsLI#~y7czZp*QbX@}tun7=joPgSX!q{NgNi@>q=p6doI*5bke-2SKj*<)lh+w{&k7gLxls9N!-};EkK#gj*bU?48fStGFp6_NDqGI)6Z>Y%`#D z8g0uRFSQa}LMj1GxA%>+NvvwvIWF(Wm)zvDmQd`svOoW*m{%qODBr0lZ@E<@R$ZM2YSOr7fd&gg1ag&g zzIaMwKj%5A@UfH>(gvJStH6=CqBRv*jE^&XSydW~dty3JfXaCN_8G6CO-sr#(U*ov zHu~Cm)g6YxB1q&dpus4;u1NoZ$Y?@j8?R(1^+p$LKl8h9Y_%jOOF zCQdw8z-j**!{sB%Q>7nCEq>kHSMU(0T7`x4`uje|wC{mQn2OZ>T%Y=a`9@~aEr&Tp z@C~5V0J(%e+MS8>Tamd63wz*Tj}3SK`!Hn2enb$&QnEnOHk&UST^5?4Eu5p|E6lYo zB?H=9A9OWZSDtz{Q})h6FK*tv%9+x$?W9;Ym_W{nCdU7KdO)VH$-g*ID%s<}!w#8Q zsOCwO7ZJd#0Y?|{@(L{kWpdm-&1**0(#Rt?;zU zTVa(3NxAzD?~0Zh;UBIVg^FIj;%vp$c&fJBM3#uPj{7)l_9e~mMY6Rtq4V{q3(~Rt z)}OhsdZ9#4()}F8SK3P+v6jyWp(ygDHchnULa|n!18sQ^^}iFUgx1hfHB(##eH1uu zqUbdeFeOB5vQ6xR{P#ytCZnk9RKlp|-m3B;XO?hw8)x7{%Vu}X#jOEdB-^m3?F0-kS)@ZV zF1C9bWB|T=oj$6(Hx9#2YWmC2j6rbAvEK`3CW)(vO8?cGqC-blH?2owu@*y)AK>Dm z-_u=8*^6Zl3vtJGE|Oml6zTZNq+|e%ZjnYhLF1fWUAKu;zP&b6rR3naRS#$VQ$V_+ zNT)uY8|aP$4g>%2(>V3-H~m}z&>L~C#2w!ta>a#}S5G!)YxFFLZd7Uj3 zX!og&%rlsKm8wIB2AjjT>+Ks>*3P`fJraH=L_dVrB)GCCs;M^mjA?KsA9zxW9Jyws zxDt&itl5v1?SBV-aioHT$%9f>-z;bZSQQ65Rd|>gGWI{1^307E#3id8aEujNS%>Ym zZ*3G3>n1>$5W~hmwXvD*PO4dacHx;CeOSKg3{>Q8owm1O4*qk-=n7Vp9w^Z2pYeoz zU>g-evwL0=pcF2(7eTs{S8&a40%71i|#JQx&jJHP`NlahoQ;ivT!g^c~k6a^>z z?8Ssg4Ps@n^agR81ZU~|UL+lqfa=Rk*nwn0U_}>8 zur`-U2g>KT>z>>70d3Og$ua_1K|2fEMYj=hDwKX5DUUf-B}btlANRhUjhf$K!LA~r zmN$3tK7WzLP@J-fjVTCd(+%5Q8{kZW8$h6x)Y&$GN{v`dG>ndzRAXX>78nq%t`kaD z?}yuvZN7tv_4@mE)?$mu`0B?aUgJ>u+|4T8Z5U zkW&FiV-G7Iymsn*ZCUiE%5=iBQwFwa7N33dKG}CYD%X~tk+tdvSzph+kBj(n#qbDP z&@1^(8(g>Ke-yt?n9am8N?^{5KtP=POjKvA~{Ed^6M0ZI88_&FQDHR=RPVj_gG+E*+a|4mqQ z44c;hd{v`(*>ZFN%xIVaIb!1u&*A1T zPukSWi3*t>@+X*u#4sKB>Sqca%_B9kpdOrJ5*k zyQ21*&xY?<2-hg}|5nC`|am=S$H>Lra zL7(ECb9l1UXJh8Zoy~afImb88+mJQO<+u1{?!k)OGYWpdopp~_)OlN(XnyX|29JJ^ zoAgx?;+_}k`^dZm50yN6&b19?a`is39+-_`f0$ebm)FeYQV;o@S?h7QAiJx5CQ`Vi z&lO&=g=r3}S3xpp=`Sq*cvmQCA7$U;!)+{_u&nex6gDeFw+-wE7JSYkn@i5B=cR2| zJO}zzU&x-T8j*H0mH$q#6{$nCm-J^CsTvrTWr_F?Hg94aSEd@hyvx84d1z+%Br9}Nh3+SElD=M=v#ry0F+#0_vPlL z{X*a#^DH>Ch#0wL)SLUh=-vq(y{}fA0-VJwTT><%p{Nq^p|}-MmqtT6s-J!&E}d54 zTjAbnoF;*RR?u7K68XNJ9)y9f_sq)GQLUj1Qah||@G_ILyPGs5zm1|=BV?Mysavuj z$a1<+_qSD?ln-m><(_}d%6h}mo%JXG4Ao~+W}Ol6&v*!#7HqBnkzoJ2*&A0AfxGLu zX*pwTSasV&KQUE=6!UhJ2WeVD8@Y1`cZAuUz2UaIhFe<94q(k!x$N^p~fNTqNTT}M9H9FK7UUh znr3PDMao}6XZr@449w6AYC5GWVfp*#Q@la<+x&jWu~@Zt3rrTrHkaOy$#$2?L&JcR z?qc)WqNAtCZ(8N0>Drj{W!*!F3ln~rwA-?AD`C25StqlJssl5$(5nK>7((FIWzHQY zz>f7v&r`xJls^&MNL-?IoRFV=r+)cj|Fa+n4vrN&OV*ph2l1T>FG4jlsHeF6>)G<| zu3|^iN?gU7C$M$vrDd|+ECkaJ9F=m=aBVse$7$K&^{1JrC&8U8)QUb-`5yKBH4of% zxi~ic%zglT$4J@)*mI1(J*McFG~Ia*b@9eSLezv~Y76ImvIeWNk0^bNZ)#4IyYVz7LzPw#SFY)e#yo8)Z?e! zZB^H|8mn&ak=m)~xFhg>f+Na>!no%Gk={#?8-`1H*a+KrO}D;*6c4xRWGRPdrsW{d zj(({y|0Iu4lhq@-1*u0=A5HY#H;wU*9g)(4&0hPwOrvknG)d*01;b(D&1!hVs%I`F zCN+C$-%i%1jU0sMT{Jzeu5PJ#0FGj}X5E8?RRQw_XCD72|6HoFX6?LX6~4A=CWtlz z_xKCShg%93v7ZhRNle+B2(mbWAcl6s}r+B^=n*kn+wL)orwz>}K&WbS)=m zH(na!-TJidz5apKoi~^zCQ+*q3ZE zvK>A5MWIL(XP-Jji7xJ`5(Y4aOvVtzZ0$%!Ic-&vhZ&%^vB|)DWkO|@M|R!e_c^l* z$01L7w?WdgCUIu@G6CKRxAcUQ=S0hZK&_~ijih=Hee{wvg{0mvT?wG6lZM0rqrgMl zVushNv@u&aFZa`r#ib+4AzR}$noK*FP&x3kmJcGTktYxVdZd7_m_;NVIGi%TI+C(R z&_YqgYur$lPEyI%s?1<4d+?qcECmo%yq+}#OJ{l-y!Wkq341hqr%ErM%d)oC^Cl~B zXntxdlalVnHar$DDNL~Z+&sUZud`eys+{<>P6|?yC}Ukh3OqI-2$}i--#)r9znYlp zeafX!oniz)K8@jwOrBMgEvE@$lOr520~K)5Vt_5hfYw$qG(3#@LdjnI^V3EsNJR5Z zL25zi>ZMS>f|Pk+>}|IG;qB_2`_q9;oag0wYXHxyMweo2Gc12gH9Yh=&y3M2!PVYoVivdZk(ZWBdk6AM^W@gAx@;(D)?lX#;7>UF%=OEfdnp zv!0+md^i+#;|Y1_#lPZ8?mZ5k8UltY$vg2N*||#*Z0otVHcwe(M#gt*f%0d(VEOF0 z(w$#-Zd|N#rFvVC&EH+Dm3g7I73DVvyO*{iTL~%8`AQ;|jnA$J;h^eJ7WjHZOiadt zw`)BI>Y+OUZTTe*BRBrpO64@C*DhT2=c14iW_nl8nI{waC_I6unObBrrZ{=ZzLn>k zn^ZE8hydD`cg`4&8cTRRE{cS*TCJz^!Q?#GGur$1Iw> zqbq%5i-+gNBR;qLJxTK(uo-BwSPRl#gp_GuY!yi91S{w?2Kis+D9EH^)0d{tbo+SS z#PD?57rlRbT&2$ywCf&=!$uyOzr!-q%Wg`kJlp?*Gn@Yr&g>|s z@X&1<4S5{)v?Yd9EIin>(Ld2#jhT|r*17NKp)M!$s10;f1c&3!DJOi(ZDG0$Qt79A zq%vuCPdAQvCILxYZr#J!r3tqJn}G&Emd`DGTy8hs_Dz!v06YRPHyrl3f79)3hBTEd zfF;;m;sy=-AKu?AM}^d}#v@BG^GGWk48#p3)nyL&I2Cj`9{wH4RKu@ANtjSW#sJ__ zuZM$??MY(LTzUh4JEe&1o>PE4m+1Al!e|i0Q8kh}5IbHbbSQ|i84;xY`Rnk0Gk{?? zW&y_%uuGlX$&I^@A!Z34+bCjPFesvOEXM<*6y*1(blg!hdZ(T;5tW&6UXyu#j@HpC zu0GcCj%*5S{y~I~eSp8zE8WT2n74CF#``dR4?f9G#V5KqsV>g&8yMOwn!<+!Kr;x{ zMBQ0sHS3Faw(P1!wGZxJJP5cSI!?Qf0-ggGm`u_YPSS@86SSGaqh~JLx}O-F0iwD( z&|0A|*8DF74v}4Y?xgA%SCq*u<5viD$go>S-6o6)*tt6gN&&37gSC)5Nb<;O(QrVz zRPI7AroN=!vKXcEzRHcTyURAy4BA#tYoVz|fS+nD3rr2Ulqz5DJ`wp8xWVzhBk|9;fOsW1B3-Gm zevvDn3m8iwXVN6-GaVuB=t^-)QjF7EZ&jA+Wc)blbVf%P%ApnD)%+1k#(;;f< zNf=Y*BCM4FhTC%Is-OiW*cY41M7+*`M@98VY+9hV;0u9VPFqGV(?}x%Or_jY%f0X4 z*ULs|1sR27wvSr%HPGLI9xJA;v>VlMKv)+`y9TdWu&bOnGoRQ~)n{4`no3>}1d51p zz4m@t>RvGlC*X(tJ$Od3p6oFAz8YjnSwjg0V!YJGGqQ|@EN6R4e|r!9Nt}KEKg8Le z-@XrW1H1LO(e(J^HCgCK$*Ja`m}=jai3H4Tj|IUZS?eqywE<8-AqYC(ek8!*4IIXZ z723l{ZgO4k<3Ga_9r)949j=G`gk;;;Bc)0A9`ukN$VL7!?X^0xNzxn2Ixoz_LKf_Px+svQ?l5GH> zi;l9UA8+IWkPl$ki=3_5Y;kXOsqXZ~KjLfo{rO&SP^>q}Z!+ekw>vk^;G^DsV_YyA z^bO81W<{a!A7a_ge~D#Z8U6r#f@wMbSelXA`DK~<$z32b4o>KAAFlOSTIk;qKD$1I zzb<<;67sQT;Df9AT=&`=@S|hBB4lyqx1N6nfS=nUG|#)Rn&bTg1jOLrkGo7p*dx}=3~lJyrC%#S}2Z?-I@ zjnsjSiDF>KjSnwy19seg*UVHPJZx)Z-_dh=7NVf&taaAD&YPBH4v*sM*sAbTqRAqZ zZ?QJB9dMKINnftQz0`o2YnaL^fFZ=u>L&ACWCD<0xEKuB$~6Oqx90SMJw!R{ar3>{ zmH#@ic$74qT+`ox_3iGrM+JQCorxR@tp1$JI7KB9HCo$%^6n!sr*WW;HiysRb(b;; zm(XfaM9PxTSK!agV>L;6#cw5)1wGg5*keXNLGQM!;3rI8{zAzA%Xyse^WbOL{-ZO1 zMhxHf=5EFs&^LIh za5})87zI&{2y8M+U!QHlF zpK<-Q>(_{6re6poKPU@iU-!J0+KyQ|UK;&>|MP!n4#-i`>a2q&RkF(PIw#gMsVMaI z^8U!?o_>x^v7elKfT=|E=Z!jc+Y^6~E=|={~uE8$ohu zub`ADI@ce`cA;?l6bVYb6n*06?&nvJXFqhKdU)?4%5aMu(2KO)dEhdXEZ2u+Q}QOn z4YMQmVFMwg%3W4Nm@++&=5l36zllhny!YB_)*3a#g&D1&yvT3mE;c94$x8$6$LRs` z3T??QxFILmpG*f%VGsR9Ef5?cS5K>5YL^(sF?$Ddu4Ff)N6F>d9#kvX6-SS|h-w8E zK;|*4ms{e|kF2&ie=Z&^(0RppzYhfu+n(o#RFAgFne%;*Nf!@cdyUh~z%lI{BtAja zu5VE()f|-;kz4ireV{4iiFe|CiE;^|_f{dT%uihHPx`NO&))Qzt&b%wc~%P!t(j64 z#5QVhltDzG?;Iq&9!WL0I(?Qf#U;KBl;oRlH?VAME+C4IzR$*#-sM2eJFmt`O_xG0 zKuuh#5qqug15qoziVw5qPe%*hWrn2DWj6#6OKU(dBpz}+ZeY%rZ5!3SRCkfoypC{@ zvE~#xp@X~_r5JlZ#_n8S1v;?LGb{4tWWi|`|9TL|x;!==x%RgG%X}Ye^@X8#R0YJY zIl{g1ZR?aqqGir|0&(JGeeT7*$eM2-_je~?F!^Ul{f2qpb6ngJ*!oeLM%80gc+Nh) zw^j(?84`uac6p>5O!B2<_7J7;3Vz9D@}VrDWzuOeI0yhmBKPlm;_E42fF5y)b6VwT5$yGkXV{$&9ny-yxO;Y`LUoDNJ#r9 z#TuNgx2zdR{LT|#yp2c8TYdV`KAGM{m*DZ7FxfAJDo*ZL=#BIw5CBg6g`kT3%9%+h zY$@3?MMp#)lQ57_r6M=fnS>jh39B6Z>tDnN*H@DvhmQ<0IW?agLp#1~7K|8R@&Kad zwaSlpk#_iOYK78Yru_9Uf_M+>nJuDn6Veq10-+P)Gwf-yt}|s+!sUd?M1-zJ&YhBS zbK(8)r}StKq+zkj*U#09s4Q9CYx6s|r7P8)Byl3XjNXkAiE+UGg+S5$(*^lhk4rah zD>2^|dxU_4l}4FweI%(hcqB|EW4a@{DaBd#2xF+|(47)cAX^VhHAz<6>rArNY`U{z z-@}_@uKEf}4ARPeRPl^TJ1koK@%xZ9joqp7B=-mI{k*BZJa~DdO%9z+FYEQa>p#(f zA4|wJv{2AgI0|2vI$YS&%Ye!)_aag=e~Mn_R9}x5PkeF0dTtlEKO+W{jTPQUXzk2f z-M|iradgQbV(5fg363ee(gmZbEbq#*efoH+mS4(Z-{fhq?P@LQSI$_M9ZJVBf^!<- zC;^H3E;1UG+MOZwaY(Ub)aO!zKqLw3%&Yr#&YEmtvQ$8XCLfnE&{(P0zS{c~Cx&%% z=oAEzn>%^M{9_XXb=)8jo2#*f!A;>EYnW3oD1_;In-{M@+)lzX#eqJg(njM|b*IjO zw`A<(&CS(x_95R_MHI;)arwGVJsUfc1yqeWF5&^BdkxY#_V(@61x)peZJ}Sl#5|os zEE#5*@#gvS-7>+fV~)QNKELMBkKasdtD0<8-uLf66{Vmd5~4$p77(+Ma~uer7CZaQ z6X%x|9vnzrfS$W6k9D+*nWUV!nbp!{DEltoi)T4+Hu5g3N#M)ZDXPr6QhJ>^XAHgQ z{V)ZBOaIBQ?E%k>cAN=%p~y%QShhS~*k;)d`&G=J(abO(0epwZ8M*w3#X(BOHyL;d z;1dGYIV?JM>!+&%E90n6)~NmWPH7pRjl8VmG14yXJ;twn_zJHSwsl1X*m*%Ml$hat z#qt}QF&2ph{!LPQFPdLczoJ6B^A0n=>hPwx*uM}u)$M;1KaPOAF(ZYqb*{{t_}e2r*@-6&B~q8(rV+Y^nnAdBMo z*_!IXIR9mQb==2s{TNK!d3nlPQ*PnIM`A<66Wh6)rqhzd*Y&Uc@*i*>Fz?jgPubLLE6e3*~>LFSr%#zDb8YnNezu)9G8 ztS7@7(X|Q7JY-~j&gemu1BJz=Sq->$xvy%r(hjz!^1#?=B|pa-Es>?TOl6oCnqTqD z=171<4`tZ$@g3iG@-BBW6m3$YMKCfjm6|pvM=Etl&{-|25;DVBiK9HKgDhu!t5siopgK7 zf+k~PlY568lcBN9KUZosKY5Om7n_@O$Z;{}qYl>HL@Nwwd;%-w$=F_{)k1Ju@eM+V z(=n-gow)Sc{FH>TIt~sh%2gHYe<56p*;+Vn_T!rVBu^7N-VW*IKj=!zjLqM{+@d=Y zws5F3hRN``k`3

ad4hjey6oqT^+F{9m6E*N>?xo7jCp=N1%rlwCoW?eUq;Vaewaj!GsUu`*NSd86VoM2a%Q6x*W6~Z4P z!Sbv%p%vuRBaQJIP8dJr8OV%D;y@18=B~fm^>M4^9~Id8CbzHf)y3iOUtDp}00G|z zV#6g}?~TA&D}9xTD3lB40dQW~r^8SJ8K=^jI|=Jxr!$z~n_7khlGN%zOpy_5heKJ! zM(g(UzOme|!M*LQEEAu)bH9^(Eif^;{`dSjYWR9Hp5PN$ey*W&6^;M*uto z8cXK+O!KZXtsM-3LT4Sw@4m5KNKe)4)*;1lLa`IrSuf*K*sg==A%a^*V4tfu-aBQt zO)Co+cTR*es#%4KArr_%wnUYWN|Z$l2}p3KV_;H6{?wO0Z~^b@p|*1_gk=n2e#CKv zQCFSFE>K2&6fdQA`gyKNdy;8|>5yT*;d20ed6MSo`Q(S~r^h``Q>(M!H|MAVA=oAY ze7y;m&#dOibk44E+ftn>OFmYvNW-AaOM+aYpkFm)VX~ zHU$}=N()x^-PUI26I6vPgezu$lrvjD6`Z5*yWFR zFcy@$lROUHzN8Ot|L_uN5q5|?8oZS6gSSui+^;;7>A5|G(bN<9db-t zDJ^&J<7pVLD9cT1CC-Vb)Te-?OpJI293qAPit5UbCedArd0Q*P5xc(t*R2kbte+L; z6Tc{pqc;6I_L%tE(#lt3(n#25JO2WbmFEr(hFxboCln_%f2xObu=vS4H)ZAvb#Wz;leflk(Ey40$YZ?#wVCyyEy`8x&K zt#j9#dRj(mYbbj3Z50)ZyPNWE&V)&M_D>{_sV&t8*#*j{79)r`XYPR>C{|3wp!>L( z1kcWs4NA5=kQ)LQu^rerHZoh+trF&>o*#a@xrTPOz(mB zgEl&5L>xo{_#M%222cN!7c9&|RVy#&8z}v7(O^MYKC?#ohA;-!@{I0CLN6fW;osvM zt(AAqnZI7AkWu-`R1TMr^}^r=|9MfHi(FapSL!)&?r;StlU_)gZE_rEoRahpvo&SB zDVn^9t%3{d?q%w_h!1VF!R*PmJA!|L?lw(l-w|Im(&u5h{Ss#8s8)q}sF!8UQMTV0 zA?UH?Mq$?+MK-~%EcA$z7Dw48|9Dwnzyz)IKVYVo&Oe5h$w@e~wq_MbTD}{V8KEnk zE-5+Jl7Jr%d+SZhS#8>kvE8QeODgkQVhbFKSY=bsalK5F0xA89Lr0U%hHQQ5x;`?&&!k82K!sb%mbb1P|S zq@VT}iJ^nV*ddzPMht3e6>9_hgc(LSZtn@jBz%K5{mtf~;<2u5b3ifQk zN^0NOTA!3=;s*1VyCO7TRkCyW#~6QJE6HsrhK>-w&oLKu?uk-t*Z9WtOAuz7L3HRV zt%B876qK!KVig#AJh8y(#)3q+>UPt_O-v#)k7@%Z4J}fkDd8_lXZaSZXyLK|4aR?T zw8nolw69=3+Vo!ln3)ECO=LEe&fKU39X&Xv2=UFK)m=tuE43V9lyBY^bj#lcs{E;V zw>Fcy0BEJk%qL}J@71|H*xvd-F&&pA*jX^_iz)ghxgHZKkuJSjuxC>4=lwrf%9gUM(`mMoeA&1W_*N+wVxtDW-BZ7k|4^A7t_sG-roLA?wWA=rQ~WQtR*MuR{9ymq3pXpnu#AY6A(#ColJf`5ibeH`PJ3n%3GCtC6YX zS#DA>k|+cg_N;i!aECw!+w^7Sv0@jDI#*9p_?dwbDFC^4orgP!8ciGqE zeeP;2tj7uID&TewbhBM(#CeVg7Qkc(_#d^ZCKK3RG0a{+WrJBDSV2U~K%l1rR52diy^cW`BHL^5vLNkW|0a@fx|WB6U+Q&i;L?{@Mujw++JmF^w= zZ{m>JDw1gm_u6|+S{DGFG&CfnjGVn9b4O!;0S5Qd&3oLh9okPb;MSe7`l>A;VmsN% zxuqbSX%l~~+?bRtOu^i6_@_%3XYJ;HACh=Tvps@l?A;8-w3(5V_LiLBNS#}MF%Y*p zl0}lbV2@rsw6oTBXjf%OYPVN&+VioMpMI4IYBU4cjbU=KQo=sX7LQemQ#x0&37v9$ z);GcN9hbyO^Vo}J7p0Y`K|N;xi=tH2TF;*JJnVkq?MQ;>h9f8X=SjYT!TE#R-}WuS zDmi_8H9-y6-4L6eg^+|Car3jgnU0L)4Xi5K z((@#D4x+1DQ;c#@xk*jgDJiKQH}eW}Bu%G{wRxt?+tIl{6Iv zYI-OY4^)Gihthc)EL7h2TMfVHrLw^kOnuzSch|v=e3}EN~EYM6EYcN58hoQVi z)Y1;rzqskrER%X3VX3!=eV1oq-FI$nF~GQh*?%eG6#X{6*fO0~AYlGDR&LA4hFtch zb*9KqZCXud*Xs4IWd4Z$2^+QBSk5e(5wwzBD2Lj0GdgA+nFW_5yx5@Qk$Za%_%_&E=$D450aelow*iSo7TFRO6 z96gdy-_S{-T3tzeyJWZ3DwEgx8-sxoVj9%25|=F-|DFbx1}#;s`W1_gOd=D5^}Y)K z^sZ_~>vByRCViNR6Bb>^Vz)Q3ZAW%eNExwdXbjbP((+Ez&2eCPS8m_^T@n|{&B8Xk z=2Ke<`oo1&JcP)mf@4QUuCf#uyLX9g0jz)%%l{|&#>C1wR(_Invg!B zBtbLWU09IgYg+zCpsE|7H3MlTpSQ$p4S|k=C0T#Zs^J zw(jZUE@;!OjY$(Zt6?b1U;3KgXWErDeuG#k$hJEq1M6_+1Qt7)USKSftEKb&WP9LH zZp_o0|6!P_&a1o*|5YG-?cw@3B9ohLbb9-Meto#%J4kbM0M+y_V3hMe!+#pjeU}$C zF@K`-yD)=)f7MRgaB@gsiWznBM4gbYWe|#-?+4;x^pa$ z!rdZX<%NBMj)T2C5VJ+^qchdqd$|0}Sa+CeWTs?z9Ix)yKhX*&>DU+BC97j^TQ7&G zich+^Sw;iBC4LVjd!TmiipvGg4npOc&n3aN1WNuxQK|pd-ZaMh?|jxhb;)hZv; zQMMh{Yb*vk&GaNo#B0tGzs<*a>??{Oo1E@o&&?WVRnU8PD*P5=Giatp$)|Gj82b1N z$Rzfdt9z`MFO3Vajd`1;t}swwrlmlqgS}TT_M=h_1zt2W-JHgqs+KI)B!y-@n8(O@ z%$0M21>2iy_tmgCwU0uvkQ^(N)nKbK);(+AkJ`$ePWUM!5PA+Bt7&fCH4-qK1*SJ! zsTK`V6iUL!h#NHC=lYu$PA9Ck_Yhobs~DE;64=z;=M;F0%U1mS)e3AXZI<`w-o^H` zq(9+1TWMpTv=B*;@(}3vwM9o;JqZeP?bHryF;`QWS6ck<5KKE$_WolfkiL@gy#c+~ zXmsl_XqCqul?u(!^c2Hy?bGq~VdMK!)5kQv;c|}B!At-+MdY$mF5(=3DR8Ni{eWg9GaWM*LUVZ_z zshsk(!r^4Bu{QUqUJ05ymSmI~x$@vgPZ=t!Auq>AE}A0mpqEof*{k=rj@h7F27H-I zo1xo~1OUcH{?6;|mJ*Bk{J=Mo$^ho~Hwwyk(elJbZ+dt%j zP_>Dw@x^a-H}U;E=?m?dt?0h^`~mu!_9~fE5VfVSFgxAVsO>at>KB1O_6S&|4+l@Kj8T`BFX?k$|@S>>J5DZe65~E`EiHTn1cF5A4e;;cBYh@ zx*>YlZrgw+W(PATf{Yb+W22Zf4pa?|blLL=CtgyU-4jDz6qhAom$0`N>iQk}vz#_U zdIqOoyYAT>?ow^f9o3#Cs{5I#8B`Tu#B_;!VIfVuf%LkO+SkybymDE>_Rqa&+ZhCz zfzd%l-Fo5Jd({n;&&e)VSBaZhqJEPP?l6uljnf#sIL0 z=xvMimx$oq*BsPW{~*Ww``7+lr$_C{-UgUyZGMzZT@$x zEpN=nNLZE^Qz^Bk8FMG!mVPsYz6%QERNw@ASg4}vZp*0x%&ffD`{oO=Q#ZFj8lf?Z z`hcAGwR9#?`AMqb*D-0<*-Cg3C8Ba!$&4%$61WCl_f`bLsz6ut)k%fOIv;nR4HJ)` zZP3~%B~D&#kL$NGPM$?_RK&%z2p0|$x9uge6)K*U7LXOA*3RL~>f;xZjZb51X;&+2 zk^VEyv5o?@;67TfVRNPn1*fr2t*O(g;MUV>jUWK z4Q97DdLuQX2O5?-$ z-?$T98Xb(Qg?0U-0#+#Z7}uJ39`kSBj${WCkGiXqM?kg-7f0yw+%)KKB!o8J%5~rN z`!jbcrNxGMyy7EpTXhUB6a*$Nm3nIkB;;fAtnsVsv5WYesh9baSvW1 z1C63=*WYCJzp~5y1U);#JE>219VZy|D~1rVDB*Pa8iRt$$(pT7aTV#6_-eN`FUBu>_70ObGfxEkf(r^hhpz|zC$NV{9oPXOAT1mrQBl#2BZZ>lfQ&>5f zIeMiHD$tfsH7rhAH$s=T0106>!iOuGmssnxx-3OGz+_~WN8o4Q;Eh6&rGgz$skG`w zQ%a>Zn*lTtF)*2vBL?GCQbfXT>`$yq^-iJhG>F-W6^EQ{2R)n?xxB0_MctO;KD?V< z?A6I)44FPk?p~J%-9IGhO7x8Xc!S`l7e^$vemh#IiA&=z$Scp-1v4+#_Br7z&uHP! zk68r-bw!4^g_(f4TUZ+@(8$H9p5Vg$}0f+zj7X3;Ng(Blhidc69qvF>TX4ppE0yx@WP;QyEL?2 z#BBqbiON4wIB{~ah$L;2F|C6u>T%a=d_VKQumOw5`i{Rn+>j(cDJFV0Y7a&mzh7-S z?h)S~x$yOUsIiqJR=1>6UF<bsqhP%m1+2l8_+W;!gI3 zJlZW5zA5J-jAI&-NUR8wM!i_8n0k~1T2iGm+Np|8f>>D^F)HhO%+s*dCQ)fzYBHrF zGKUt|nHSTdJ3I(IDUC`MBKOr_wpSm+bOY=6cXQcj{sHPxE>8Qg26S-eTivHVo--vJH!`3*c~3$2$Z(ZHAg zXLODft%nMQc(N5f9@Y)H*oO<=k8eW|^2yFvNgW~P6^Jb*QRvp>B-B+xX#RC%gid+2 zKn^n>Vk@C11vp1!e{wOM6S~t1X=V}_L&DPWsZQC2{#Y^7o%{vd(Z27GCSvweA{}jJ zfBy>zliKhHk6Xj7abT?M#r2b;{O_-dWBKFfhtVP*i9-5dSUT!gS+jQe?0Ih?WTH{p z1!6|Y?+YEQvkJ2!J?c-feE|+Y+FazI!zJ8&}quB0X_j;sS@t)!a=zf2yo+wJSah`up5ge(g**cO_ zPCD2!S^r{z-R3v>SGh05=2v|tL+6$OD|83)^q@&vDQbB6e-)1fNCJIzkm}u4cC!S5 z9WR*64?B9K%E}P5Xs;@WDW}_0v{=PXbGqVwsC4T_?TID5FJyparw>xb(epj>8csnk zr-G_TkEGsLI@`r&D|PK>&R#_mi9+FO>sp+QlopGjg`97r+`(695d6eAe{vcY8W)U3 z2fpl4GmtMwPaZ*kY^CxwY;J0aGtgaGR#dm=ia0Y6w(9`{0%=GwIDhr{QCz{!+N49m zKA`zAl@k?sDT5(s3lV|8W5c{>WAYjI zuNI%c|1QKmuv}u>)kq&@8Cj^Ds*L$a!|*i>-5uXsZ#CKTPQ{WF-#~pv8O6F&JLfoB ze(lgu(gC`we;U$bWvGO9N35@%tqHIYYc9_y59*YO)Q^+v!UUn0!;s*V$&`(NF{4bDoS{HHPFr5C}NZJ zv?5m_p_qkMVoXYB_fr37+SvMH8m)vhMU8~~I*;=#+s-9Ho*cV&S^18D`XUi`SI%X{ z{3*GRr1kY2`Rawf1I|wYojVOYeJ2!J2`r^J@V%&@{{8r$&fg* zosi3RDKsmGBa7=!8ti2~Xn=C3*Mr&pe>o~z$HgGmr zG`X6eyK0!{n8EAwdxS$2BOmXa>&{4%wS@Z!sY)|}#Q^mnf zxKyeOd(6pooan|%YcZv@kmfc#$E?~`OqhS@Vxx`?ao*6%I3rLf^st;R8v3DwIi~`> zjSUs%n~B-xD13d+&~*DyQelokV_3OJ5E5p~jC5Sq=LYaZ?U**%Sw_Y3cnc8Sx=Fej zM%zpWRSZn=vq29G(!RhgD6DAFUxaSX1O(!tW6qnArZMJ*Fp_K->}H$?ow!>=TD4i-rhib8M z{%G65egs!(lzKXB1 zD{&^Vr_RT#@6}|>zKdd9B~rhfM92{d5w$w`Y)$2@&+?lDcs|HDjs?aR3*$&ZfDG3F)B&+NR&vTqwS$g-7ABD8*) zFWJJ<26${uEWSXCFO~RT@uv4C8p{GJs{8R+BboGpA|;J#@a{do{L|dGmqtyrJXT!! z&@!u>bij9zs)g_=SGjGekdx5TDqLD+ajucJ?-#o7A{o75I_}r>_$ZSkDVK77a(8nW zXfJzJ=GIw7b_v(!b2qCFV~epJb%=pURTn{ z8*bEUIvL}U<<{ z1yo*fbJUBd9gk6jH?7D_6QZbKrJ@y$!P2&2(jrs1$hCy&Q7>W|4KE_YSj!pO`L)X& zho}i1?#{bOSR3^gqEve#9LvckLfao%D(3Tf=?i<`C~A)}vJje4=BJhf%lcxWsvtij zO4qZ@t=~59!7a+1?^jus@h|_NSMoLc6^%=DEHmqJs%T>^cGVurE$pv89(Dnd>oiNK zP-(|k7;XRHMKtMITIQlgRxiNwR5>gNh1Y#Nvum*AsStI;CSvPe4~c}*9p?dRndXBc zFWMM&!2Jt=>KEiK0jLilsw*1;CIk2YlBzrktWpT=5SvQeFhN9lXxPH(5gQsa`s~AJ z$pCJlPsp44?C+=AK`TkN7k&MDRhSOMB45&xu(a^gt|&H+*IFa^nvujj%y(8(Z-c+k zj2SUQk?nq8zSs~t1X{Fk2s6mq*oIl<+Zd6sL7hm2oywg!0pMqlSmoko1@RTfv5RM* zryA_oWv=@o+;Jxt5n#MX|d<(bYZFfoIotl+mL!H!tt4;R0ndC z;!|DG`lim1RkU*%9Bfsrdoo@b8#=@s0(WoJS=i&q`dFW4vR{>DAE(21y^!Ci3;3wK z;EsnH%uVxT?4VlQsLc6l4nxA%O&-C1-%%BzS=e zNiv(H!iy2IGZHW$_iZ(S=#5ic&6Dve4^|Gt=7WHjMKSeqq5)jK>QEQ*kH89ji8y1s zbF(S$p{wi*h)&7X?CnxHQ$ctkx#$8$)|bnL;L$g{B?N1C!rIDt;0MqTx|42W+$8Wb z?ArRPwpQ4P@7Sn^S&oWtaZ$o3SpVPynu!6(QQQ~BVPeF>VO1r#?N3cg zg}$(?G<4?wdD6zsb{h;8Nx5J@CBe6N>(If zh~jjjU73wIUR=soGssjf#lfz^TtfD0mll(@H8dah0;i!TB` zm18p_!4#%1B{E$hQ@zmf!08Dwy3cf~`xSV;5c76dbZXiLqCfylcGZK_+9t9Vz?v$3 zYG1CTp-Y^Xn9@UPMHncB_H%+mGeA{AH) z(=$(6F|0P2D6ce>pHdZI+s$t7;Vb0O3581&09PUt@KYtRDJ&0g7r^B6sexae1rnc z=I={KpY2gFcjAUU_Xkv*Ct<{b?BF5mYj_C!TUV+1(b%+jEen08$9)E^TL&wP zRX)*>3*3S*s~;9zNj-SF< z0$f(%WR{>n8Fz?uJBoCcCSW3<4c3v=ARHM-R5#lt2-pX>t~0ATaI`P=r4_z&`l{4w<2T<`l1pw(~97ISB1ql;&;0atq`SA;4 zyReIOqJj{^4TFcd^aWSEwMTA8<1s7YG#!iU=h&7$+E^LOD>Y{@>0aAUHaD0Z3idJl zNJY5IODnT<294r7Vt3RsBZ8i!v7K~c^C^UsBbrwDXOG6zouF^aEx2J%uEwQ9A~4O~ z_cdxap;h;nnsaZ`x(lkI!$HL8yHamD#kYdN<4jY%A{Ixk`GdGimHNjrcWS~maEI*- z7L$h1KqaKr&|qj`bIPIWp@h?~$_~Z#Sodsmv;Cx?C?6uR<{_Uc$3_i<)1NP3jLBT} z5?{lfSyJ~@bL*RD+PSVV&2PLqM+jpKi(#sBrET=Z4xoglF%+%JFv8(De{g;IYGCGhLHR+2nxHuvLSfdAW<{(!Ihv3PPn z-7NnCm}iC)_ax$*9Unerrq7BiG+?NY%ChFh&y!uVM?mjVBu0fJ#uq?X#_GmcFPr+u z^bD&RV(G;lg@8=j>o;Ls=45myEg5GHbxP(x;3Pu#&}pci4*BrX?}x=p0XK&*SJnRTu?w)@ z_Yc$!0#Rs9xA*JII2*Om)foAUGUUfr9(#0XREQaldt5e_nSz)Xx3=weY>Vu4Cqv8W$m37F6_$>mE) zIU#;ESclkSol_3>7v^XZa^$$y)AC|}dY z!Ib&DA`Aki##u9#d9!vbzZy)7_XLmYor_C78>dB7hq=pBmpED|Xp24y;6mOGUX?8B zaTwS+w|CWU}jxm8<6)&JLtAFAi)178bPK{B3tCQuo!^_KEM(TL;i!CxpiDBJ-in1gT z6cY@AMtK1z0xz;ht)JlNSF3mJRUw590FZH2jk}UkjL~c2{!l;AB$br{yxbJHIo90? zO+2gLR$X7rWAShQPm!d?}d1eWVLpmrK8}<^C_<-8h%TPo%EW6u*&mjMd)rYjW zI7eetFZOmP-jZ})%8g76={guWiR6uDf@K}vSB6_OwR9L92A*6jVpzUp&Rcskf;o4? zOc%NXx$Kl!4xR;p7{er|U^Q=)m?Yl=wiXV2;_vRWq0y6Ig6n;u@ncidZJbYP|UBgf{l2-_r zge%H86)*QH8n2t8lY$q@uLVX$;?D!Y7E)Q_ES;5N9@|)ETfvjtku2Xs0L;_I*or2i zHf>+M%s?Tpg7lbFa)(Ay{;Pmr(&Bn^pD<^Ah_`9Zx&Z({SO*M?`?V!h_>u+?}9~UEvQn(&W{oTkSd77 zSCwWlq$|-@9d$-!s?ECf7)~DGz8*th9H^&s(?4xX;%T)%-&c>>lx8`NXMdKxO5gcy ztzbgTlB0D^{eaEwh*a~IKP|=6fgSJ45IM4@BA0)I&0p_ucjUt}>UC1c{_~dNrOVMr zrS;Dpn2T`4CPEEBhuWcWrfcx)UqAo^rmRdi&7-JTb50?yNY;0n!fNi=zkq~~^{XEr zegedaUxdz>#?CG_gA=rOV?H31v`+JArA!7W^zE69#4X*i4e0K2i|2$4EaiPI=Wu*2TFQ3(N2{&s$ zG#^d+D=f(1VLpi{I;r_)7Rc>6@E*pP^f$8O*y zNa29$Hp9`RlNHh2-kIV=rf{hCmHP{9-Ksa0WKCFcNpGY6g}YrG9*4p2iqEhK0x`ffohZ#%HHkPlVH zSLT;y;+EI8cd>%hG+SYIPM>sf=i1x12c_zyD=lmX|Mk@jomFXNBohm2y}SYixBS)= z)HGtrjU-;{CYH!Q%iF-qe3KR=5KQ1No0_#p%6a;=j86Agm|fSw+0tPbhu$j4iP zQRk4Mdj|BB4S<-N^vau6qKSIdVl&}pA@P0p543>Pc_GH71#D5$^TSd~NRIT&2N(a> z&VHcbmPn4~eeM9R3c@km;*n{&MX_{>Ql9yZ$Yh{r%!hN+{V6A9OP*I)7&)1BUo(BJJ96s9o33h{Zx%r`Bbqn8Qc9Fl_WJzo82DYbdCXgjv}(4v zLV2NXw_lyZumk?|Q0KhFUMOdkc1zqym{SgKYCGyk*D0ym9R(&)x|R%zRW)hgemi-5 z8kqSvL=VYW6wW{R>ww(s(XOq3B1~X!yCU5CEDM{hR-c|*IqcEs2vS;e;SD47_W2BiE8X?i=>{-7=&D*3`6+u@REI)93;9#%_ z&PdDoidaB#bdg_h41!!fJw{U-Hd(HRV7AHAI-&p2Y&mKZg9ST_iC*GRwy!iueTU10 zBaS!}8hWdIW7FHf=aF)%xAdL9J|J<+TbGnD7gNdSQG6%!6qFS&!PS9ouS3j&rhm>N zTwbuiSyr2uT}5~o_{9PdGu+lY`rsR~%J>%9Xh9&qY;{cbP=O3Z<@M1wPpvSZMk)_) zCs`I87Ik=A4(x|KB*cfE6&dR8w@{n4l+%12jz$CxkhX}qHf%Jf`%}+YMwZ?Lmo2(` z*>$Adfl3uz*` z;?sU83%L|(E;r61kK@?$`JzKk^bH7J4hOwRlaN_7|264!J?2n5wqtOW65r>R+LuS9D<(R~be z?-|U1Cp%w{afCIcVGz9(cn4ZlU6LYJ+WuG_3=s4egc9OQ8V;JyJ#hS}Sf)El5oe~z zT(`dyN-+s*=_NpE~2KYLi=x^lM_OD)3aNp0zH>C-W?C8{Zir8Dn8hE0|eJMCg~ z=Mr7`C2K8Oevnoj!hfzx;)kBYu2Vx+%ucK7Yf4E+M7KMm8rzpb+2i|6aIH}{P*!S` z&{Q_|aIgYlRc$E7u?~4I{X-Uy0HaKuls5`#cBV^@_lQu%L7ITfC|w|Qz4H_-2O}(y zlyj@hLCi72sxm5YkK1+4LN=pmJxB`>Dl6~-pM=rRtwu@9P+#w;F*!>MgH8uw5w zow$z`2h&;!haO#rKgd1Ql1*(Js*oxSiPb{XD|sslWTv&QhjfeKBe*flKXCqU`iKes9{F&w>CJjwcpFhn<}d%0S@W` zeQuHS;S{SyL3vNso?|TALDmltnf4ihW1l*)*jb(yA~)J#k+fSx+VJ$nRFdHJxy(v| zs%Tbrxey}jLSc1DhD5a*9`X+deT5|pNYzC__ZmO|(Z$VgZ$Rm4jL8E!sXNOhoO;kY zm00rO{N4LD-U($rs8H9iEdI$}im2aHx=L(U1#6zgl2MV_Rd`>21!oIt4XQ}k{_fYg z5|hRdo&TAKYGJxZp@VKl#q8SV>kg}*P$y#rB=;zvJrDL9t$(HrVm@BH@{?9i7=_cfiOA_;dj5vA^)GkA!N_%0d+Y^7DUm>2QW99xA zEls&?S8_Q9xhGPK1bw?>64N(S22d4|8kI0MLkBtOW}SjS^+LS2kM&Cono5xw?Rx#G z1dMZq%^wM~-Y;bCv>$2Te7t@$p8M#OHO{1iV}`xapt^Z#fxgTVAqce_C_F$u#BhmE z;loCZeWswBCS)R!4pmGhbY48NCW^>3OV&*tpnHK`kboQ|!Oh9$<0^`Z6vochWx5|o zNO_b?XTLc#d7#RH43Rugs)SuoKz&!`Q9{lgclJm8UkM-u!`dj~4L^X+$R$GHx?Iha*meCOM`dKm|ZYors;{08Qmr<$oMt+!OW6Nh8EP zFlg?0%RRU=JSB?JAsu#5!3a~tgGfDTewSmct!Ey_K{Y+kpMs=_rlF?1o&EwK=EG-^ zhRfHNS((t!mzQR0n1u5z%F#aR@GkQqh(H*Z z%$|Nl3)>sN9T6jIrCu+J0^dIj>jkOhXX~<;+K$L^L!Y6WITooog#s8U)nrHlBGik3 znp10FOu}TFO2{jg)W{OMHV1c>in5pW06Gt%;0#P|*4oA>89KXwo;><8!E8@tP8>wo zjPH$qAtKBI9N8Ab*t-=u8p0)Ew2uoMNn9+Z9zO9p8euk$qm6qqTKK>7KA-Mg?*jr- zCF{IJI(t7E%E;yWpY2g4f<8h3Y>*hSu>6rJ!msQ@NmO?NU#5(i!)UrefEYfXJivb( zoFp7A1^_T>yU)I?`pa7L-DIOw83EQ!`DwniT>K|u5AS*pIlm)4Azaoo;AH{DZ>#lv z%XoBvx$)GO(o-efaf!+c=xbR+!t(BL`BXq{|2+w3P_RFgR)EZhNcF2)gefX8d1&S? znE%Eu%(?Izwp9?0h28Q@jNP2&*2_k=Q#(~R`+NHvcbP4+qR=jF$zs}(h;HC7w%q)z zMLS=t>^@@i=VmK1ZUk6&pDrQQ%r^gokHSvjNZLM@8tKnd2|U@sJ=gI$YcB+?)+5Vo z=&%!Zf_K6+{BV_ZwOP@`d(RdJ<+sAy9U}aymg32^9ht^eLuW^*PJv;q}_?EaO;47McFH(2+K;sKSbO_s|P?p1RXx~ej?Y)xTR8= zjw#`HrxZ33neX_`(U^8Fv*sX=RbY-reyZ*3l<~UtSa08K_t>vFWxkhvD6CA1&>IJ& z!n$i<7<=9Q(9Y?lx*Cw}iC(QQ^N@Od*tgB(T>A4GV7<2R*e-=qbdK<*swpcXWm9{; zDgAwI(dDudXQSKj>NP8u28{tSm~Ixmx9){fZ}nSL4S^xb4Nye{-I2A8G_H2dot%8M z?LN6n;i}9XVisgqMQyrQjHys~OoP$Xw}SSoQ0_)n2|FB*7UL+;b~^s^y7nKK3+{eb z41llB{9U8G5%@2$)?l4qvc=L*h}_4{g2o`cpax*XSQ7!gCZfd}|Dl`M(CjZyRw19e zvUpjDh*tqu7cr#!TdmHQ7-F&-F2kKFKI)nAi1+YWWw;*?P42)*TxJkrbqqih&&Ksg z^|%r|#FxI;!XgbAlCDw-8&UL9XZOcjQuPiyAf{|Y7%R78GYTHy6w^fStQKq;lRQ^9 zyN^wS=-5dT0FUm(wz z%*|zm;Z4~KiRLW|?4sE1~O+*?S@Su;#)wmPI3Xb|052G9n9l-AUAzG>y@t8^;LYBlIPqP|6wlA$^D) zfOCRJf{%T*Vx{aWXY2=U!;SI@xVPfMd>Np?iMEtjFJeZWxvef_C!1fGj;Cd2EVKMT zmvDd{R{SG>6`H5WsI|h;-2N{>1;Ai_Ioc(eDLNddO9;OOE34+9RaCHbhEuZm=z)6l z`RoN-cpB!i8}(qq2Aji&{u?jeGrjxUP{-b6v2eocUVk2YoFzHK zi7>&UtTgwf*aC&e86b6;t~==z!fOSi_;Vuc|F;pz1yWM!dhrQ;$kh>@UUP>R5jz2OgK}nb#||rVJ;NRS0i8)| zE;Ldwg<_=@?@7cVHk#WH;zK+w@39DAdmt0N>E_SR7R0tl1~KQ?|K^uP}q8 z^Qt;Zo-#V9De;PL*h^SOP-EL{>k21h3EDPjIh^m#f&NXOgBBA$&GGA$(2I(XN%Lk= zSVhvls{7pA7ly2bG=S#n+ZcLytrl4aGR>Gaj#j2)u=nvAma}n7l98=pQ4F?M+4{ax z4Wy)t^lKe_T=0mateT-zvC;7*q|n%KA-R}0%3^Wmm?V^%ez3|MHwpPSXyTCrgKe;b zchWgrMAAKebh4A9Dpf7NhO*Db`>7%0)3+e+?#D^IC90%k5XvQDxq<;^Wf{|sn7ZZI zGM&MYyH&(EMv#fGIObDW6c;7l=HU^{gQEo<)v8ZbrPKLI0DR&QXgh>7{Uc{0{Za-s z#b%CnTi>e)8p5F<7&Hb)<`l3>`)OD8M{oN7V(u?vBk6*!L9oos%*@zkW@ct)cAJ@@ z&CJZq*k)#CXfrdr&F-z|)hF#}W_D+HeymnXsY<$8d9yNa+=w`FA|i=fD-ZV^nsrNY zr-;|-;Hhc1gI9(SC~BqP$hu89V5=bjo}bFp206&cNgBumh|pM2wU7YUeRpGiE=Zy0 zf_}stmcZu%IFzMuC4LzIG4XyVDhOYSFB&{QtTaiFG06O0kx~y=-6T~x-lI>smEW`w zqPgCg{O5%V1Flje;DO5=Y`Gv9?q>#7-{Gx=-Q`5mPeNrlYAN)Q;1{MZDN1A50;NPW`K>4`}eau)FHPRyBAz zwKk!HY9v&=xR!l^(`Y@GF=B9sa+=P7PphzBC8lBr%VJBg7{zi2%NuH@<;WAHO_2$x zgXH1Qprufu6G%E#?n!cwgla}aq(8J_GQ3n|GY7Sn)z+exe*FBoh(a91#Y&IYiUtq@ z%g9QRgVD1hF-)$z+X=f@K^(FyHRY)*5`u?JDaV1$#)k#_9a%#t zg%k}-8DJjeYZj3cxRxp;QPNKlOUFh|PBRG!V>rh4DlVW}s-wLa`Mx)?a7cp z2-VDm;;b^XP7Da*Gk1K!f4#PP+ThTbvNt|86J6*zb;I)^N^9BUEu_qLss+@})bcBsyz#s(Y=+cf=;W6OGrL^!fG)u?g_Owa~vop}M!-8@M zXPo-G@hXu|&Y6nj976D3cG^-`x2$a9McNpp(F7;Vnhuu3!;iy_%P75pOuc$8uyUYE z{sRQkQ>K*n<@fNn+7O~w=a*p@=0B@?kwJ%6;86mZB2AFcW&0W1MZJ&Tq*G$}@nJME zfkZS27q}}Kat)K%NVHZBQ;iZ@KN2#Z?kiegPmksY9^D z(6ynJ^qn*w*rd{!9kOQ7s7T9@aY%N*wr=#FzV5Gczk8N`eQ61@8#41gpZ3V?N7b2w z6-4?P<=@YfS_Q+)*2)HMmSboH{q7avwEFJ{J5Yhao+K8OV%Lc{4_s?YlCPr|&!1*=q-0 z5!+VqBEHMKilip(6v>|XKyUN*%_O*Br9x!#7EN1)z0_X1|%vAi_uZA}>(*ks}?dqIhsTxVQ zOkv-}atude-n0?;gMWaF)5h<9mVhdBGq`AEvSpg{I}1vWZCRp|V3J^HN~II4f^#X` zCB>9WvC^`cB~y!0wr60LL5$^R|4TUUn&*23@6B2L0o`lVH)&tQ|6hD5injBEp(!35 zj>Hx{{F{OI|LgmsF5+IW~4C+i=vl$f^;PaL1QUJRLz0$H~GKYvVFp>2Xwurav+ zw@Kdx%43RQIIDcY#1GB!uj(WJWry-Pnb)~iU(3y2)@wtYzx6R6nxlTzrvfXF`8IFB zfz!=imuo}rzx6pEnzR4+2mVjyI22g9&$s!%J5b8$-xp!!tY^AJ)6&84H@O!_ps0Vn zQ~w`;X6}caX@o+uN>e%$r2jvFNEj?SwNl~#mmh*CCk7tb0dD3V4$+fFscS%pFVd!uza0v=H$+ODPQj*(6i1hFt3os|$_dY+=BPUq{O1kj5mjKf*O{?BMnuSG+X zUbKJ*qj3=0b6G`I>Hw^VAJlI!T33#QC5E?DC1wQ(n^V(tzXK|~*h-LiFOQ%0FL%Y9 zwtjhA_F?;{O(7wxfJ2fU52^b3_75NDS6w{%e;PH(#XwyjX*3&sA6__e;4XoHe0bfx z>gS6bAY;z#cdqwsSeO$Besu%8?0&g;tmUkW$YFxzb7l`6a(s2@#W@)jq=VXN@UHSM z1#-K?UyrlY#Cc>Mi_%_zJ9*HHbQ>U%$;eyMBpi-yqK0WAbFsS4lfPGd;)J{ zzQEcSTIP;nUer7ZGsj*91EB#8xPGsLkl0^Q2uQKjV<}8b(_waejipkcsiV>~JWPN*wF1)785q(J zNcM$llrx{*dMeqjWtwcwW{}d#xE8C4ImMhVQjw=&WKr`|hr!*uVMpalId0Y+tcvFr zr&oDk(at3I#O(n1%qWZ$+g_? z$WK=N_iy!^-86KLs;Ireej@(iB;cq7bwzxrgHD@Zrl;*J8#gg#q6f)3NA zq+j{VEMymrN8si7Do?rY#U%F+keeU#=Wh)QtpNKggnn7;z$0TNyNM+%OrQm5;14)P zl70JfRWfvU%FRm&xCn5}pMfyWpc&v!!4h}F1u0DF@-kpsrTPP-+z6W|Np?LSKW;p@ z8!->o{=}6!Ie3)x7}Fj*cRGf1+i(2OpX6N}_l+ZLNOco5WZVru&bXpi`}^ail8|<6 zj@%F4ty|xn;ULSI-^ZZ#Nwb+q#PkeKQ0OC6i@xe0>l5t;=W$B8Il+a%<#rS~F~VVT zxeE8yUi{TN_Bv(c83vd;f7c}tOtpk>R&AyR?sh8Cf?l&u`~f!t0%+BUX4HF?ADFvF zc*2jqWJwoT8@2!7YL8^gSuLK55wKU$A3Y*HB5al23qE;FcyJS^S9xj^u$-AJ3{x+; zbqckqVX5e#`X>+?FcTL3V|39 zaMUZ(MJ50mAOY8(7Q0O&H3`)moq0cWkqhdR41Z`)97+2mY0kJV_&!UoM;h{k`8Vzl&D^bbUIx+GSH#@4u8b>#Z_br6R17Jj1m^g>;Y0zmt*m_ zuFm&DhHt@Om(!lB^#K}P%b6xCQ5DEZk$R*C!vz7jf1?H5SGkzF=Njs%uR{JE^yYe$ z;TR1hMx-@YHQDuMxmBS83SK>(|Bry=J0Md1=l8Q?Q8(DxT+vVulkJ1Y@VvJ!=iyj2hT0Jc$xC zPCg*m=%YHK)pMK#+Oq5~`!FH4R~gCTA3jWVb|{B~aK+dny>FHqOrb1!z@%zh+SG~- zBuLdA$?J=gI(rU_jRK+}q9(ay;hMlaD$zsIpndE0{Uf$^LSO&Z`a7(CYv_{~t|{{e z%W@f6W_bt#7tF9Jd)ip~vul#jhtB$I{U|TW!Bc5KC$@W|Z`_|}{Pa(X(^?1L^?;!c z;Ik3?%1w9Q2T<|Pm*dqT$ZNK9YoTCLLU^mf9Ry2Xq7$VBXuLb+_uD}IQ^zryN=!SarzdIj z1ltk=-e*Bt>NWyD_{S4IUmj`7^U7@Ugd*oBA%W@J_v zp1EjLKI)fjK*vu^#j7u^zDt7*K9+W!%V1udX;^R+Vl>e(fJkI1rDNy87Ljv^)*X#w z$8k)h@ha!FTl6J;9(TEbY*+p@x?BKGzN4A(7p;X1Fxjc0qto`V@)j-k>mvvxrp=J1 z71W(M5v{NuG2Z14Xskzh_QrtGo-0aji>1|1>nO-<2x`WGC1ZG*t*VAC);XzcRXVEd zBptFHcjz->k#=A(+8HC z(>soowI(l6yD>J$bD%6881{Eo^2iT@Wkom_NJhN>37{+hj6u3}IQe_~H>Iz7CnPkV zKv#9DdSmK%Wtu1RJI!&Qyly+AfTYKLwtvlVcl0uY^m^a)w@fv>%MgtxX!HUM4`j^q zg;JUs7@9p6EjdC#Zw8k^*C1xUxffQAr2g3_8e(Tm=*FY*7|oH&5Zlvcl4%WbyV3Bv z?srN1J{%i29HMPQABm(H7S;khM1NdE>$c$@Egc%B+)vN~I!db!iFOz8JW*?tj@Kq> zj|tovq{Tc=xgzmjlo0a9ka(2Q0_uti9KcG%*)&RzqUm_p+33x zNwb@uPnuI64;(JC7c4!jYu^1|pQBGJTmKn=0dU6IHZJ_-J+Jz?mnzM2aYHlKj;JdA z%|FH0nXKL~VL>mC9!^{RL1orzdVpum^eJULzKH6uWIXy(u}Xx#{F+T4V*@i0e|GE_ zooxMF$TMZlw;iJjVW666^LzCr=Ll7*@zh&5UA2%BUU$1s* z&;KkDE90qfL|?I9EgNh88^obKIw*{j#+94jWf!>9{Pp@{!Ypl-8??F+CM+FqDyTEV z$vW=`wtDX)S03SwWbkpw@p4bQI^G@CBwTiVN9{zE@l+Jmnd7R@jb#^IYE&`Sc)C~3 zmM!rv#S?YP^)lwN1k*<7oeWGehjtTA7d9ZQnscUiNgA(qW z_d?6LO9YXcsDz6bb&6^=2u)9%0IT*#6o&1`(MUo{f^J=?2=|-3iPq2;?%`GVNv;nwU4a+z?CoJ%>S^q%s2z%*A^!a8|r!9Aml#m;q8RoyYE7Byl_5xHzS zd)5XWzi~h`!06=qVrc^49#})NcMVNKZ^eKAb$oa6sfWR1W%`q9RD)in-8xxORC6uk zQW9SUKAIwgE_;)_gJeh;`y(aex&8I_OZV)%7m(j$d28=FePJjg+ z;F`OIhN9yxPG%W`oov|pE%DmQhx{I&)0)SDnU6ZFR{XEG^?V?=#JzR59ACh4xc0ZD zZ@k15<>auGZIN1Z7^b2)nwg|`bTE{BHXuC8Gg)(C8?WUu33zC*5-)R@xzWPHbJ*YP z*%X=GScGN2%rj9xdSqAREQ^yUCru z(Oa=UL%{&n{b+^_VCwx?R#*tmaOERSMi%mwl(&V*%?cZ^WGn{!W~f)i9802km2dg%8`|QQ^%w>-_Vd+IZ@V<@9_>$I zmXyCr;yP8_otddc-Nd8YYCNzW931WCPp0oMQAq36Ej;7cyok)~3Ajh?`^_ju)2iQp~u>TRbY13xC#P zM73u%j(1Y*cR7Mk5d0p2TMUn)p0>O=zJr)RXfUWX?8Uo3_I`8_CN=88SuvS>y)}d4 z1e5OEKUT)SEu$xOI?BB;MlRwdyoWsyCY9p(z2e zuK)ltE|Ktkw$F7X5wv2-D(#4jvjVb7sh}CRTbTg_VXqDTI`{IQVkMXKxU2{bCHEWS z706)3{?}KW7fp<+u~R(4Mm3*RX)l4FE{s5~2!ag({>tvJgV;?P1$T@E`v##uk(T$< z6doXMu&(S>7@4k3okA1wdvsGF2VS7C&RZEBpI3YZM?V3-pII{areF(6jn~w*b<1nT z4T0SjU}gy0x^yw;?)bcw=p&6G7Ql}raMh&gs$g#~Dsix;^>!p8eya)yxsBlPk)bYx z%*6R|*jMtM7#e{4{tS)v!J`RzmMsuA4!hU2p5{LVN+g0%+RAZY)#@{0%SN#3a9*`T zTbMgK<(R#+$kcf|v9XjJHICc1qe-LEb&lmR?Nfm`ZY&|VZ*Gd2#+1zXJ1_I-W$v4T zk47=4gJMn3ud^QAGZvNRNPAO#^zJ|2bFTDxbwSnjsglDc`qmAZxHbdYDI$-yi*?9? z1O;ZG7w0Nu7Wqjf2(KS4Q&U21a?>*$~7^%Moh#s#X5w+otM3v^_tQPT;8q^lF zimA08aM2Cd(E*?X&l0)X8eft?5EdFdT9N83|M9FB%hHYfz8ivYi{HP{9_oSNIpb~F z{yk6j8wGE5?R7n1uIP29=cQN|2E*mE4Hr?`J1%KOT&tt_!u*a-$z1pZ%Raf(hG$WJ zJEnVQ9F+PerzS=X<+;n3Io(&0$iwIGJ%i3fVf*P%n^byJ@2E>FH$*TUM`I+}jl;b?o-|#Fd}9e=o*+DtA^%% zm))d&Sn7Sl=hjr(?HCXyFjVZ`-s^+dI#!V9QyfAda*`W5JQ3c1p(=(csmMa00|i z;q;j>E*h8jC~juiU01#4MNT#hM5>z$_dOW zIgmI%Vt~PYgCRj?vS0NbUJ8*s!_76&5&q{hyvR*I<_C^ZP*QciB!4Lo{|}IauUqZ& zh{^lAl=wXjf2Eah82Assd@+!`Jq|l#W(+yYH z<}NrB^EOK6cTL7Gs32%0#c#)9;7);~hVcZ#CdCJDH$AzclRbPXDAxn7j@q5P7LG=^Ul?C4(1Cd0(lADGVK{GdCxE6+!~lZc`bI`(PP} zzgLS(gSs0|nr!0EwseviA131gG%EEizRg~}y+#4Sb1T~?UsTl|a9e??P_%tR;IpMB z@Mu0QW!A+^$X>}&0j`51eX_o?=x5?1v&t5EEjL)0B0)@7&ZRZC(%PCl=TV;65DxVT z@$Bd@;J4P7$3LI@$*&&P9-mkrJt`?FZ*mlB8c5Tld?0UHrvt)^rCzn~436HFJMDPS z+$O>4k08h9!Um20$L-G;4Mfga8@e-o#;GZGComNS>>zN|m!}S&0i?hHepJBa;EJP4 z8)8w?LoI6Nnb=rZgWIR}S1x+}ITop!j01b>{MZZnE(WMSpYn0qyD-xR(f{J;MddQy zxEVkG%EcV_=u{aRSGQw9S%R_x!&L^Tt*BCY**qbd70Q}F@Ki46O$Y&?5xA-!!c{nx z*kzc2)=NhKe#IYLKp-z7yf)XV)54Hgi?T_<={CE}w#`L*!>GAz%XZh4>`*efS;joh zXj#F?8p4lp8a-smbj&{-C*icAUZ|}gAQGs9F0UDdgdBh5<54_yX@7)57I~X4gXx%w zDp>6ZrJFn0a4|6^vRu0on$rQ;;uA*)gLp#vOTsAiBIknmxB$w`j?Di6!XY|iRX@fh zNG^+ogoLyZKlV94?r^N>8%i#ezk7jNIM>G1yiawx*Z~umOXYEzq$b^iv_MggPl8(e9-| zt$zqD55+cRk9;IST=QHbB(-o{m4)I8bW1`T;4K4K>d&V88LL%nvdMl>$rI8|KCz*5 zsz{copRd^n5l9K9f|E#8bSd*Y=BkN~)CS0vX1Zqj!Q8X_@Q`k? ztl?$wcF%@Q$hPUdwC-Kg{L_Kk0wAhiU&k~b(qTFm zQm;yoB-(e7Xf%RcjB1T{FXK9d4#dsIGGj|o_zt35Wsu`&hjX}tG^T+V9AuanR|HVH zKk&8wNtkMDT!yr_E;P>z)1PPp^wK>g(0?swyO~T93d4pi4i&}{*G!Dx{e{H*a$!S` zk)2#=*YmMNJ^C9jiNV#0)wGJN?!KwvcI0_J4#e`YCLYzy1UZ6rIiX+SNl5Wq~rch6>92# z)V>7~?Ny?-X?=ddfkU&9K+mI?E=H5p5E#7#&E{uZ|I(gZe{1!IHpUCcJ^u#~$Y1M{ zPgotJww7N8`lpyJ zo9y>;E(5E7OYjB7=#uCNNL{Z5D=Gw zN2OP#B{Kru{m7>vz#=l_bB2|o?>~UngY_%=SZ)5;(kJ8W)@rx;;sh0O3ngh%mU*xD zM`By2pAZLI5uhm3yY1ZjH=jhR__sY%nLzF!2Y)S#_!*J54LSO}J1+C70EZ`pWT;+V z$<(M4rf!ZvLLYA~WMu;sSzCkpIONv-+<_9cUBu<70E@Ho2R0X&X~PQg$Yt?#prgn( z*{S1Jac&kU028e7M<^VPQMyFGeg(XV2pO`j-ec@wQ^N1=U}@cwi7+KG1c*rJPF=Z; z2Cd50oSAtX{BCIM1P|#Xb%Z&CRR24V-emD5Ul;4$X4#dJWD0L=qHlLMLFUCv0iUz zCqk(V7v!VX_n{PPHMsK3F0?rNS6|M&F9EcGo`fD}FdJup24J%Fw2Tp=dTDYhU&CkF z_U_g@S2TAeES^1_z)?V*8pcg;mDcCd_Vdq~IX&p|r(2ssD&E;}bMGAEQ$w><{YV)s zyJnHO`4|>C+lD{8A_UWEdm5MVCfW2J_oYn$*78>Ba(CY=h$;Ok!i|{JCvz< zyLYk{)IOkZB7*P8B(P&&pQLMi00rNhK}Kd8$j9Zi3Ok_6t3_padB*rb?cC4j>+$60 zP*xjLCoaV<{A)px;k~rddH={ya`5U}xO6BedPg?sc7jL6^UIrn&Q(`OJ zYTOnL@MAKTQ|GP{%+ibK*6z_wxMtI2v5k3XPKp@{N4S0^t6Z|PZH{MsbWi|2WC}-y zEM)JzzE7dz1(Pw(5i>4;28e9ol=yYZ8a=Z1K2`)yd}gqpj0Ufo)=ZQHCG8o72#XXv z2sBxFoz^-RWBooBLoa}drR`T=@OrCyf(o-`Rk&F`P9ToxP(VdJceZvP_3JN;go$jP zi&umj5oF!{1U?q;@Esd%^J;=Ajbs|9c)Y~UianvVCF-LaBV^NAUP>=az<(*CiA@{tua!?DlY3g8#X_tb3fZmjr zzjV?f`uv@XQgGHCDT*o16)^M~^90=;?sG+|-}tc?{MKP*9{tY-ZE@UROcgcbi@D~E zJ#-L2IHJ7)<91kK2SId-&#DkST_AkEr1K5G6{_!8`y?nv+oU9Ahr-l!b=PYyQ>(`x3GG@S#=zD!36 zKX0bZ1ggJBx?fbQ`|4RA1BE#6pm+oJRpw=~@OHTXtfoF&Wijt#DBFrKI6-nx4AQHs zC14uKD58{YUI($JX`ylu8;cRQCm9Cq)Ra>`ELFEsylBZfnOVoyz?QfmsvQPIrh#7> zuwCKH##qv4C`6)52|lk_h}V;gVRF#lNH$vj~b$4DifLj;#evh;0x<0je<@d;x66a?b>UEQ{jK|InCS^#c~@8Fy5IbrZX@1mU7I9UP zFr7iGo&QO>&kNjRDMceOr6f%es#1#pbwJA*ehv z&Pa5@t-;p!VNX=K2(F0s^KtE^twTROAKwTyRT!M1a_*m&MCo8E#nh_~`XK3${FD4t z#}37LKY&KT4xQeuOr;eDgtTE|sS}>D4aV~IpRIH`%`v~r!`|onUGLZ19$$~0Ufx?n zf|vTZkG22ffQ0R*TV-xQT8n+i;s4dSTJId|x$_5nUMkSxfe$;iCM0A^bGn^ga+tG{ z(%}OCZ)W*_IX!ToqN40;HB~~!AW#RvRIOWU5pg231%>`Mrz3z@HI?%p8Gd^FKd*wo-=xNyPI4)EOZ*MQ z75@RW@4C27DM+Wg+hKm`S0~UeBZUS07AVa!XC*$LtoK_q*x7u19|*@b{@)+CTu5t) zVX?A?sRC@Pmt_bFDa#})1G;R3pSM=PlkH|*Y9^s-j@fJcHnV*ImxdA4NKpR`nD{fBn(qMyeBIXtXFP%gKu~P6ma#IGz)kTs+IzcyPuO;FLKG}d%nwK(KTh5+<4?5NyR}?q4nhW$6|_};>0xmH&lL)!S>*c*S__KSr6VNRM8v>> ze6=Xiqnrh>OBG0TFwyt{Y^3%8;Hq5)Z~2?X#U$}0vW4?`GW&XBO=*MI&Aa`+GtvDfZ>uhq zUb2ctx^!Qp4%WtIUor_S03$0pd@nmF$vON{VffriBFuJoo=uoY88dvO>b|80a0Y{5 z-zt2t;)S>F?zY%i-?~Fh?hL+`hVa`4I|FF#!>ZPD1=R$t$+LkxmO^6^2wSftU9{(5 z00!#let4om2OeXwGFTf@>~kIuhKTWNZAW6-@#7AS4rM_13E-7b6;V~IVv6VrzsjWB zg+8zL;Uf!LET{~1U!>t$izo-nkSiP~hj8wcL+EFRQZ`J2 zXy-+uO|(8qB4d>fo-V$(qXe!mfcJOrVEnm^`I43JZy2crpcn`9BSq+&j*>d7iqCIc zQo-{Hv6i4e=u0BS$)kgGs*M#$-UKE{9C(26vs^*T&#&NOnfK`19tWOlidUw0y~%$! zAtO)+QfEW4?Lcn4rOQ~&H`Lp)^w$D6~dUz zQ909bm30BY3h^0@iJ%^J!NW`Ha$6$@nl3)5p6Nz}QAwK#PsD)W6kfWI7DH(#z9QI- z0b1ptcv*9yuZNu5M&S;oZiJvw_lQcIG&6vuCku-k#>u8nCDNcg$%4xeANigdu+2d% zfH(RJujrf9|0dbA*3rK%yBM)HDyM*V_&Mv?m@ntyk(Y@0kq{UO8gzWY2-X>+|7mUm z;wZ2rn6t56gSuQr790oGb;Kl8Sg5YLq5K^z%Qz+e zR9c#vep3VPUjp%KzY!WrdlQ8LQ|slvf86yq`dPk%z#uY)mcTne!+BD@!L{4dw7*zQ(}6EiFp4CGpWId z%EeV7V7ngVXAgiiO}vA_A37mDf)q9JNcG&GPC)n)faPM;6>nqXcuWf$7D{?W(6k|P z_$4~{D0haBde2>13)=Wq9()@<6EoG^Q-*#o`sJ3lmo)J5;U(JP)uF_M@*@ye+k^MB(Gg`pcg8- z#f7H^$A#5od#a2ua5*!O3D>90ZzP34dK!Slk?pM^pcX*2Wf8-e<}};MX~2$X(fUC9 z8oq6D5Xyd!)L;ZRsM;}l?Q%@LmO_J@x$z#RFY)4^qvvT4{W7Xii;afSnDadqUz56h zGHC5*p1^Tv5P49#FUWvCKm&8E6oZ!Wjh@3ci zOzV<+p$WuFf&e`3vR!8N)<1xb+G}EPF0swBCMZn0mBaHLX6Z)WEXQ|Io>$TzUj0Fs zW{r}d9{k1;+=PB)5`_I`pV~DfGSKfM!JZ#*AYK(xdr%s0W1<#)prEt&b7zM(G^EFA zZW;!t9U7&%96`dJb?JYCePWSolW+A_$U4>&UYmXHf$w<-^jq!507pxMH#~s5xdViO z=UAwMmQ&#V)VSBO3iz5DRi>m4Q}Sym!s5oOR%;YFAyeQq`sfMuF2qPt9-4BRl-h8_ zV)4Ey`%_bF*^;x|ky5)ayXbz|GnKoDm_0MdVMH~mSHhAcXSk%*&#_t>fi0X z1u+^C^jcfJ? zL>ae0ie5YRsZ{CyyTUop)VYD5lIpPkpSS%SC&C%n8-iNUGmX;4_I-}|cO@>}l)y0t000t;6V1P8chYgAJ+$#W@suh!CWsy_YecI` zP|Ulocl$%h6f&*36A2gZlzb-lWH_f>1qiTXBMQPb-|XTer*Ge#P3!}y6-E`h$Z=)- zWStis^;~nY+M-L|5YFkA{2o^@vuVJN5MH*T23+>BJ^7FS<@nRws8V7YymJ}JgF6>1 zez@sdqbg_4A~fa#QfbfFnM3{ol=a(E`Z@jiDMN|v;__|%iW)R%2V@Vns-BRJH{KBVU;iU?+xJs9O%2g8UIW+Gb>L|$Z(5?T z1oz7V^6{9Gmtd7H=Qf`&f=Ri6m=o2pBFTA6x`c7hQH^9oMIPPa3749*=bi(lk5&a| zl&(NNQjZ%fq|ajfRZIq$R4y9nc#65#FJAZAS-mI_;?T(YpD`IVf3%etC5a`*cM@$Y zO&lmA#hxVJ3D^_LRXX_4cor6+DEPG|Jd%kLft$mj zpCGL2^6D7FC+$t6+th9#j*XogIzgCFM~uhm;n}ryT_{KrzZq5H`_d)u$P?wF9uex% zya_|}fU$he9)*<DnCN~;qdPx&=j*iWAf zI!89%648F3D9!wi=V}H^D_JB-%87cYyD65-;l4j*`iD4l$q6e9L2h4h))?tg>Lq}t z&yw4>CzPkYbQRS4RiD*3Ge;5iTnY8hKLC%UPI&>_eGw=$;`4ao8Axim{DV2ZZEY8l zG!3%iuFmzMqn{o@{(l=-NMS~yBt9RZA_G{bAH4=jid;B3i zI)Ud@#CK+_3^E)rn_OnSr=I@tgl%M^E=8*WG7Ln37#v3y7Fe6<4Bx@I?hId%W$m`_ zLLWpIJ$Gv(4++{$w~iVcfBY@F?zh@xgB>w`w=>M-ou+|S8DGGpasmjzsKiQ2e19r+ z;9xPu6`y|P0oDfTqzDP|vi9}Q3*x78AzPaXYJGN-6gScRv8u5>kj3th_vg&|P&_8o z_0788@|f@C9_S7?gBQ(MG#T8@opQKGO)sc3kOc?2?4IEkS^Qk1gSb6N>9+d~gS^?9 z&ae~RQ>S|5R#f?LMT>K9Ed0 z$e$y=xl&vX&_mO}<5LQ03fC++be)5iHzTauhWdwNlt~i&|)n zB8paV^>9PAo-NSHsZ-6)-_nJU>eQjPh*hrYOiumNk4BPgNfn!L0~egPzC<>uOg7F7 zYRtDC=zM%&=>i~CHXGL2D>Mx3+H${Q8vVcB%20)55sWHWMFSGG_hLLtp`QLk{}{KT zcL&A^x4$h)Zo9~S0cl+6g_tmkBsQI>n~f4;zTC625sIXdnALL8g?Z{Pa7fVP zbNa!H`Y1a$;gARvoINYA+CP7>WPb`PW^8Ni4!P@!@{yehq5egv0-P*2^U&&HEJs!j zjML-M<3iSrSa@I>Bl8t;Pt@Xrjtk)}3dvEN4y8`)EtY-#luF1W=fyk_UwSe=xccM) zdAjOj7zEPMVEQ)iWGtGsnY=9@vw45$GQ5(uVuzg9;8^|A!`j2`*G&3+#_Il2>4<-K z)OmRV5+Ly2#0Y!LN`&q;)Hu3kJR#{6ctHMmR zL|$tYDw1tfiqWk1pK(`RSQf2gTj|xcCv;@B$TsMqV!I!i3_-7)J?(JvS|`{rItbU1kzwDVQKX z$J1#tQ^2Jg5!!en&9OP`w9#}7OgNdu)?}QOSft~YlEGMx{8S6kDP;6TB!R}}>))WF zIghkfZ>b%PXw%6d)!jUyXit^*RYHsP85}(+yRf!lyZEUsU=mS}FObJ`ZYP{jBbpn0 z)6hc+k(*~c&xQ$)uIrq-ybgM{4p&RH#MHDTYl`R-Md2H2o*96%Mgwb(+!A(|Kzfd>;?c00l=?=(rr8ZwR40XY^e?Z$lw1vWMRE zON>;2Jh2ZOx1;?nS>rp^)lTu{l4ZoJemkB)0kF1Lnk9@IQ zrIu+#;_Y6#NJ~&0&1UjSZ=wVq$=kwwDYY*PEXtzkn1p|!Mwp#UdQm-`1VREEfn05H zsI(*`7ob})?+$=n`mIem4nEO=(^o5>EDb(A%fhvLoEAb0;b^qCJ>q(8k*4I+BgmF8G-)$TDA=ftCppXPzZDv$649bR|+ znFa1{Ez1Jo?|W$xHsmnjV1tBJ^~S9^q!gJivdyrpMuyB>Ac0(6A|OgFXc9eTn>FNI zs1YEA!Sh^}RLKH-pPiuRnuez&0hvF6J|7DU1AwJvPz|4zM}?MCBITTHjk@}dUsS{h z93J0MSoq-SB~qCncO!FQ!~N6ZugCugp@~arN;fOc?FJBJ1T`-|;@HodCk{C`<638) z!D~uf%^MtpVzfNgV0mNcK=c%((v;?fg7 z+s#O&uf4!>8mG@Fg=#-J2EotGbA3_Ygr+79bCIn$Q}Zm91T?e=qdDCum@uv}aBC3X z7bW(c)~`7+t5>cr=d+TDnq&>K2{FtoJ0j0E*WltG2DxN_&`-u*^2!+{aWVN`5-vo9 z_Am6Au31CX?L{!NTg_62clcm05F{3TO|RD1G4Cc#QVGr~h}9AZ@@;?L6%sKwJ1Z<< zi-ff51uTchC&BfEaUx_%$3ou7>W}430Zx}h26=KrgMd{95KWQ7poHt;$>u#2Q+5jf z3#>mG=79^XySK0EJA4lGglf1_={WUaUe zuJ#WSyxz>2Y;Aa~iPlPr*&Zqs6x^!ATlNvHpi!5#r?jo>dy`DXs2QNBxJbfS-ltKI zlc9&A1A?;EmhE*ZnMmvi*pdw{8_%04f1k3%zoyuUx7!1`<8G}fi$(=v-v2gzfCN=k zkQ#X8JBXV)FJAluP(b__R{WiLHTam)e<=2+U|B9?s)X~og#g>E1iHVJ(;YWnx1>N8 zf4r4&_|?Mk(lp{JJI`I;eV_mtpoREq zzZZ?_Di8KoQ7eXII1*8AXHALx6hOgJ&)Ee`nDl-FGHP>%GW9h9jIn`T64)D_Pjr>x zm7WMEFpYoHvx9X^h3R3fyn(UIs=E;R&s@`qq{b?LM_|`eBTP(QQlE6C52GWH^3ed( z6kIG~ys7m_E4AZf-Z2a6(+DyB^?8mu8OhE{d=lu1(3?onpS2X|1-t&8OPi_zN(PxH`@rpKXFPtYJPW zH_Q+^(N+CAK!eMVz2a=lc<8N3Hyhw!;3bhGNDD@iK`~7J$;mpo9vlXZnsSu3pr2+2 zhbl1BX_AU>$@JH-n7viyP1zf1)km8UFN6(1q1TQrBKC)X>fH6}0I@URgZBqt@Q}oYzgNw*@ zrDoVZmT2KUK+E#4=m<-VBZz*j3OZF^Lehu1W@|&W%qko7lB1O58NMXM&O!Inh;Lt4-lPH@! z7KBI^#R@f95Rp+<&J-}N+v48jn=sg^;bhs1SoFAVJ54J+@_Ve%QFgEKIEM7{+4VeZ zm+%(I@l?#W8Q_X2Ejcb+!^?gr$Y35uBYqcMeyb0zIu=ru&^%BMbL8~CDDxS89M zl!6&Md`&v&w&s|x$LW+tlekFGn?p=P=kPHi+^%nkdOh7vsew8Yz4hEYaG0#Uqw=?d zmJk&MK>(;QsM?y%9+{@?m33&qv>Rw(4I1uL26tf$b+GkaxW4URAPo>kNJ2#9hD2Mu z1h(XxT|ixD6E{u#H!@Hn>C{ctAn$^qJD|EC1IY1aBy(Iq0GF?DE^|07iYKe%)*8X zLurybavsoyAR30K0^83~MGHMhIl%aeQmBcr#D}?8cfRx~v!LfpB6^2{yiK4N&`Ms) z3)-WsR9h_N^9Scpk+PCSpUc=3{#3ow$&{?4xS5I+Mda_gZj=R8bayIA=8XOYDJ^A&3@aVH#pD zGVoMvkezHp%t1m?G)xs-cu`VE``8olJ-(m;w~uM6~I@B#)qc^CZg3xYU~#NJ$V{ zxK;BMU(>lx_4e3;>;CCI_$gfZa=>LZORkp?3WwwsW`>-Kme4Z3h%hK+sc=!X?a`ft zYcV3D@BlN_2;x9>p~W`yw%Tl}_4KGBz2Z_O$q(Ipz0=}2A+nb!mK(!=iZ+w(aL0S_ zpC5~cmu4h_OSo;^pP64M_eY|Q1}2-B{>&L;xS&%WMV)^d`C%VTCY|(MntxE4Xyc_|>!fjU*r9ml7|2oK=`*oP!h_m~{|(4^_M{ z-$owqq-F5B3J%}6|lRgu1v(psZ$6B==UhMPrQYu9xnYh-8(*e|*p!GcVbv4j&1?C^ z&zm3ria0E|iV`HPaDUB|{BlFIEu*vw!HUAcX;otP)@O1(%*BN>XRtIkmI8|Wxe-m&ayZmZ&C2$RCw}-fRLNQ zCz76nYt$621Co9`1+mC6N$!>JqV&W4fe^=q(W2@W13Le_@efDzMIE_noo z0I+Z8yX?IL7z9&qxdjg+o*#I=9*KB)kXdOgE0@|U+xO8SmE)_^S10>L7JRr3bv zTqwaPzc4M9HepHf;z&7I&n0cuQqm7N8gRDFRTsU652j|AoL@fDT+F7TrgaHB&BGNT z#s5G%3!`cEIE3<*;a+h^BnZx_yNY+_(D=ZV$|A(V$>0YQvbQJax-$Zz#;ose>FRK+ zb+wYw?*kEIzxiGV!A4N_j)i-0K*++h-JjV+q-9Tc!<<2S^_E-=Cq;*mVmI<1xLNuq z*;((t{99^-q4+i4LiY#W;Hb{4W^b6e5g}4UulB>KucF%A!D^#r+VNE@UCj4+(wz)X z{G6m>^$p464ZRwxI7m`L8!<(-EcemT8l3X!BGPEbp7E)cKp1*=zydFC`;3s=0L)3f zsA)#qu}NHq%$4&e}T}{xBbxgQ&+*0LS<)v)4CU zWY#rz`+02K;hl}(7H?~Pq@(F0ewz0{vmySuTESUk%uAi?;u10Ss!ZN3C|<-G)jB!M zReMe4pNGIjhmVOA5jH5=Vg^%J;vKp*#Y7{>YPXO65-#8Xhl{LzHu{NeA>w=}n5Sdz zFhU2znlPbpbD)3C=mJgKwX$&B?aJ2apcVIr*vu*UXMA$!@c6X|s6@Dlj>YJ7<@+XE zNSz;blk3q$(E6x)6wU-u{DI@$i9X0jqOxUoqsJKB5MN+ves)+a+aNCEAm=4bTmdpr z)ZZTm`ak>Z7L?R2Hc-bOYzBSzDhbVL;q^E!wNiwvcV!d1KM?(?mPj%GIUVokt=YIj zRffn7+P5+N`+R7ccXd`~}JnbyLtCxgZ#DcXu z@5W}~s9T0$ypX}s;DMHy)Q~h#_+%|DmhI*v)W?lpI&+x2`4AD-Sl;j70ebUNSr;livp9{iE2$ap%9ww!mOkiJ)QYe+E9_D*{W>br=4Ek4OSJMYO(FncofXdx<7~fMJW)-?L_e0B{pdSk_vzY^uV9F{6*}dJ(6U1JHPMxs0p)Q1MBNduNl)hAy71$1G_L=`jv9LdGj{10K-gTKyOdhx&DLg zTEN7~_dm&1fSG>WsheQ7w+iQiFkx%jX*%9lFQlN6LoH6Hau^O(Xvm9K#DIs38lSWq zCA`o42<$cyM#jXlJh9PB(Wb0S10T%1S*0Z)xDxI&uEwAojQ@htt@~B+Bf;%D;L6Ba z{7@vNL9fdCV%Wpc2KvQI5_fT)N(pNaRr|T1^U0 z&rq$-7CrV*K*!#R{Bwp$!h^&~MhWY;ss&Hf4RS%md=g42BQc}gl{E=(npc21qBtq0 z!KPlX8@wWdWYYZhQF~ae;aB~RppuUjKLKQ&cL8FXcDRT#Y0-hnk|SZA0wb;k8r#r2 z=rK6JjeWPgux_nSlp;&B`?^n%;6O-|Z>&R^r z3$Xvv6|5jS8hcWg($z58lril2khO;$t2r+7W0-`4g98aNwhvaA=7G1M?1)>VqNhL2bK7+WR3;>5+T#%fN0?f8^$@f`^*LtR(MWxf-Da)#%o z(AnlY!5Fzod)F}_vfq+{@_i$?o@zVql4o!r$FH#&f#3fsYq*#^!$W>+V%OC1W&c#= zQnq4)wKznCzqs1uTlHj`_puAfS}NpThM?R^?l+HFes!~OHk@ZDGQ%T!jX;At{qLqq zXZ_Y4KDhg(c3<(d(E}L@o{B5vHu%4w8f9x<%GsY!2~rZJwxH1HAa7#S#DdRnA4C&g z=yZ^XHzb25saYk>OYO3&5c!AAGWTgc`Af}4mh{>q2jP7koVm;<$gejy{({!@wwiH&NUPPh4>QUvwwtLx z`{*#LktThn_tHDfmfx}G2ulxgF`~Y4Vp|q(#lQ07Q3KQNhPD=g8qw@V`x2yS!(}gI zGc>6O25rlYMA2ay6r7PXIE~LR?p`8LC@`+K?|xarR;_b8oCWB);mPH`2$42y{sa2Q z_^2vMcm_)9!9z>yammSv?=jTCXq%ZM%W6=VM~_kUFXa_&zwSm=9P3$lTWxpr;*7*0 z->P4MZ5u$+=5th*=fQF{+7a1Z5QoZ%GBRDbflCG6Q&xd0@jyvc`EpA{a7^|a8O<2H zluz!jrQ8{-2L!kbL@hVAm@HVPo0I8nA3_$&;;%n-08uu$yom-X1zrBb6zIh)oggd* z!mOSv#ah2ZMMS|-fChZPL!fuH;m6tu!$|zr#SN1MPHt}(M)2ND1({dM5Tt# zZpSgv#-{~5={-8I3gW(cdJ&|V2oTpFlt8>=Tq#I;-6Uus1m`sVg zmB%mSayx8KXx?J0TIgaAvLs`@Y*FD@V;O5aadAU62dM{k$#+_2$oC$(@hqZqQ!U1&+L2S+hiHyN zJr+B2vAF^aa|`AAA1)4LHx+30V+ih!^qY?EV#y*jg)LZ9nNf)tbx>eo66rd!pBkPe zY-eM!rSF$wo2;&hG}wm9&Gv;b5?}ccF{(wiZJP}1Fz-%n1aFjMX?(xxtmgm5Yd|&q zwrFim34zB#;)fh=sF7}((6U#AAVfLkS@59?4=F~j0I3~V?+~ncd~qdQ>H9TQKa%KT zQ)^of|G+flvF)Z?z~Aqeu>p15n7&LebrBl4%_C1dXRoG(@1EStPND6hll8U^v@=vU z0faYr2Pf~x9AMWa3@&y(uR0ue3zskfgS$k0OIYmp&6MG!G3;|0Vz}{O0Dq6X^+zzJ zSPhdOo*(R>+d!gbO(&!m!watL!CKx55_}!kYpdSbeY#rN7mvy<6ewYDRSezoN(Z*@ zYirmXA)v2+V2bfr8E8B;L~i{zx)?wg9H>2~_u*BBh5c3BlF^JFft3JVWM~3-^f;)u z(5Gs#RA=#X9HQ9GevAp4@-FpyfY|;c#mBVtY2Y#Vc~Es4PKOW5?i$ z5a5JgtGtMU(XpGF0gGbpqJLUSoOXz2V=r>m3f0?%;HpqAgC1fUG+2s+K? zu&IF`X0sJGWMKzUX=G35G#nFY8f9`(cL1t8)P`7G5!hz9KV=_D;P zQj?nvaFNdH7vU=gj8!X$0`q!e0n-$wcjdMpf zAcA;(bLb~w zpaeUtlQE#NJ1;jHeWhfkQ%k`cL3R;#}i@zT6Lm^8Tn*$#{IMMscR4Y z9@hf~(A#eQM-#1gaYma(b%?DbSOEWg2`!h2&J0B3>ca0d*~o%>*O_1#+jt*obS)B; zrNEsxQMV%SO6V#Kg&~Dm0F0mf>9toD_yvRWdFo;6-#DcTk`u={UIGB2s}du9Bye5| zknq2SS6l_qkPxoGS+93cDU-rD0cCzCc1F5`7A(!%ywj~VZOY;X!encmCKsm04>7mC zd3v#Y`cot2vaHB9yA%feNot~s!87zbqHnMiNJYM?0dX1w=L1Ne>~k@ngE(ajum_#SJ`KZnrH$${(w<-g(JO1~Y;6GYr^!lgPbdHP{I<%oW5`XuMrkGEzsp;K0P zr>fU@p8s0rr@=lHXk{xZvY6|dn+j1I7cp{sPmT#9PcKvXSjtZV??V~9fGYPoNlVCVW|um~dyc=qOFf;0(M*g!(99W3xL4-n zjFg`A7bIWli{pFNZ)`NNIGpLPR|ScxeJKPm6T0+jPoS_;D6+uzEf*{TK@0>y*$kZl zP7^#$Vq6GzEjtj{>(M|DdvoxHp>5>BTSfTGY8|0jMrF+?g@SD>!I0Cc3I2WQ;)KYK^36?n~S-)mp?LoK|sDI&gZn`A+vYslVuF zb_Zi*PYDwj32uEagu%QQ!dw_=NP^q6sY=`lB1`dn2OE@k{gAIY8P5--OMM`a#JWK~ zsR0CI3W?j(rClPoRam@cnc?9cPh?iR%wk2qCL)=73-FGdXhAeP6Y=({CtT}OS+Tsb_okm^=gUd z-#+u);`-6YIo3m6X_={andxH}Zw8Fag3TiHmCvpk6eTwq5>EEcpi*ihoYty4gCl#I zF?L)phgZ*?=~Qq+5=*4r(H>WVrtK;@QFCpoAjAomqr`{8&^)9kJXFD}Z_bD;l-hta zPQ3{A-WINk(vW>(Y4A8@+pEo1^m{LStn9?PZ?*2x{3=wTWN+95!Mx9^RMH8{e zy-wyYMThZ;2kM|3{z#lBciS164$e+Z6N;2+*RJ(^PJ3%LV9~a?!adbtSbXe~HUF0h zDx$?43KwleYO$PV%7k)g8=tA9jbe-FUW4lv_rPpUud-iphyj~6Pd*h*>4B0 zXM)BWS$|TUJ~%&>vVc&bk?&{gJi(HzqP{um)92L*wxx&VpF<;UL@>c;n-Rf1S9w}F zb!oRbYCN>Ggk9#|KVsh!gdNRGzgNYKa#{LuQ z04~Tt*Vnq>@a{lm=ngKXywEPHuHZ|-dgzH3<8KIB`bu}UV)ADsliWT}m%*_HJ7(K6 zSl0?2S|Ul$aEx?kWV=rL(UC*d37A*#7sel|RY5c_?a?88Y3$Ro-J6j0VRy?J+y@6} z4kTpC2n>nAp9?j==?ra(0tI_aLOE-TwkPS&trMLZW0Wfh(sZnq1}L0$-X;|Icp`kH zwJu2$jmcO%7CgTkzJJBV0ykKobDkw{x2li`7ZPe$Om=ELeIp`DJnG=2J<)^7qOWu< z;{8;M5M(oPg~-`(IW3I|-pQZVS;HtRjc*_9;a;b@$Uv@U&hG4}T3E+=?C`9)TiblX zd!#$>+>K&eCHa*Ib%rcPt#U<)19GWlBi${kbHzjRgG}w>z9xSr|1#>c%4R*{#0Dj? z?1jAYPjH2eO-FN{*S7T-+g!_V*_KAVKPa36GP||5Bj8~bG9jOv7hh_6et1KK36RD1 z(4DttWRNa)VdGTXz`l|XCE&@9#Ec~v9;_#K6fW20*RQkX)g+6o(29LGcEG1a#J$3gpeJxYJTJN%)Czo3p0#Lz7%-u%u-y zVHD&h%`s3#&AMDb69v5`@-t<#EsXP1%*h-VFNlRLuM)jd%A;w2q>NGk9r<|}hgJZk zb?g39Zgq*hkw4ka;)6zZ==XddKW<<@1Pj&;8>96rU|f#8PXiqJIAF0rHYar(tyibU zGbGIIoGU?MzvhG!5y%u3hG0Xks~hc-#$JVDT_HL7Q%ubBLNfUf_IJZ_Fn*k$jZX1r z;f+$pLWZGe%xhEL3TOI?@XiuOnePt~;kY;zd5P&i(O*g!#FPq{iIW z#M4n)JzS6v_r?{8!X7obY%YD&IqXC(4o{VLlJuuH2fO}&t25M#XWao)lpBY@V&p;& zc50M(ZookF(zg`tRSc3}5-9w66jrhb3)rn|t*T)Y*dl$CYbtY`Na}Tgqbr|VnUItH zsMKB7G1zUem}Ljcbl6v5z^w&J@fxijo?c#}JG#>7J8A9^Zz>BBnxvl2!aAOMHR9`JqbBIWT7cK=-t}3KgOk}L zurUGlwfp!+M?b&bnv&Da{!q_LowZeLgXmJ@s*9BE3e&R~OI$Db;3-JTF8b3B#Rau^ za#8dt*Ez-m4_7W=BX=MLofWfsK;VZXK4C=v_OPByW8w=;mfU;CsEe)XIQ+;L;SwOX z>Lncyf!k`;rmT+dFH{yoUQ1AkVXROA)lywcu@ss~-T2T@1{HWjHZjy+!x>kkdMc#+ zXo>(7x`;d@nzpLnV=B~k-43+&u!#`04hBj2@h@VWt{8P)W zx`h_hceKqAmXAVF0r{;cQVLuz4eCF#2UlkPWH6{O`9Xf4hC}toVnlOPc+SqK7+vAbbiu8bs2#_|U zEd{eI1DNfiIjIJ9|MqAfEsKe>~;UrJWZ{$+$fVO2~TG;Q} z36YTg=CqZnH*AK*>(Az;{&aI-*-)b#)n@SQg1O=ed0KAK5Sq2zZ(sZ z`dZ28j0pY(p(~5RI62438apAmV>{FMF?@A}MiF_GoGDW12##^|nl(jRME46#|O znhP~hIh0h;?i>HJeZTZc7i0cs#*h-)BFf1Inc{mwh+LRrvqg3_e~s4L)fFQnQ~|u! zKE~YBr+iTQHT$-TA%>z0pC1;4M)im=eB#eW@%b)gDiyfGVl}j_Si?SIeir;Dd&e?F ze)TYOs&m2{Pd*lK4J+u(V2h(V^9QX=bPwax)$E?6yO<@6k`f_2C$R*fd>k;Ni}*xxpcB?f!ON7$Z}SfLdlinLq$G5gN2SMwgq`c=0J z?A2KSUYAN^oKy=A0u)uHW~iVovFHaR+eLb=-n$=sbKO9Ln21%cC5dwr*>x{X$jfYN z9Q)<+JIc8vC*pb-k<;T@3@!m$33BUQ<=ts#3R-dgV98v7zs|559`c=bdMce1fn3UH z>BEw(SI(FC@mAJ|vyYA{k~Yrd%+uTxWXGjUOh%!ee*&yOd>hLt2GUA( zvJTINGa~55htn3lC(ixmM*C=x2(jF?AU{DGEjLwvKHhy{V7jHV)8S?krGejPn|r15wktnUf?PzsA;tZVq-cxLxFTt zGY;-2O5B-ROhknGbDXaDt&+o&>YszB<^8?oYCcv--McBLKv zXFa6fcuIJYC?>DvmVS$juv4Ho@jN>Pe{i~2nv{JjUSF+Ux}og2Ci9k1X7v^s(Yfv^ ztt2Zm+f&oKvSxA1J9@$bF}M7zX>OAgsV$ZG34h<&x{#lJ!94eJw++hq;$(_y5$K+O z_`c`4RB&7nT#fg-)@ZPKL55|1vygINCBvcyH&2w?(6Oow=;TA6Os>=9L4bpiQ{yoF zS_5#%Pzn-s<8)DZD=_?+kUI#@b`%H8K7D0kmYu8+$?g_8&cB0AvZ4rC(;FF)t5r$0 z`nq^$g+~B;P-e~MoEN2=Ie7;Dh~@daT4A;du*aIYNS)&nr?2B%vgXOD`IMS@CNNHM z_mjD7reMg(g?ps6bL&)ipIaZyJ=D?%DJv&dM1%tptlx)-$B}8MO=n!>Z0?-a3?w?1 zllTM>kyKsciqmtzjo=aXLrlnAy z?E$BVkxV@vDM`VbP2`e~rsd7cIgn+s3PT=-8Kh5DFJ18+`Sgl`12@WfTH{wrDN23T zxTDnChB(^ltf4|<@$irzf8A}Yz6PSQRbq4|m=M;sI?EilF!OCP}7cndKct}gKH zK*b#_%u<>bs;YBMiuY_HNil@*ggey3z|5pPV6oU2_Y3iE#W}*~Mg>s(Fm)eWBgPSV zWq})h0bQ#h2zCYzE!oC~SF~XHR)v^tP*tR_McB~wn_bc3%Xv7OAnPNbyP}8U^@poT z+^XxmikuX_xP907sSel)Da`?On^Hk?N|3`+p+keN|9sFxa%GaF0C${c+nZL;q>0}z zpN{ek7>ZZb0b^RX8E7prW(iM2wPRFIMNTGS+?GzG5`Re9CP)9J@#xK;;{Uvig`O5~ z3Z&%zSV`csrf+2v2=P6siV_l9U$J6kPEFG>O8`ilamf?GgJZ_$UAy^r$g8dqC1vg% zh%l@=>Cth&){5qlO-4RN-fqj8h-gyD+s6mrp!uS2{1L1#G0OHD7q>qc&|9x*(^fIA zleA+Ip0V_BN7G&FJMTMJy2UgoQg0^36GjS%bwgTsGsJq6UL{=`X8-zTvgS{hAPj4- zH{Gur8lRf0a^7O^E9}jBtK&_=HT{Tr<$bJiGAlFZAA5E7u?=Nlx!&xmREKxV)+o}E z(O2Oa^OkwP+O>4ExK)hBU*ic@n5hvV|Fs-5rSMCN=0KZ2c$ZQZyfyb1?2y4Q-&1}1 zp9(|)tVgWMTgcWb4y($%-B1-e6K*K#TZywmvVnU~O;Mk297m->cT}b-$kw2T&|U0g zX`#g}>(3VVy^^ZfK2rD0)i$BIIS!&ROr}a?lljt?W=L2tkmF~}X+E7oB?TNV5O&E` zj6mG>F*;0)Ys*g+CM<~B6S#~Xs^}+B1_ljGyPgs-Bz@Amvxs1K`(!x8(R`oTX?DDi z!mFbk1Uglj#z8kBAmS=hC{-jK3_aXPNwAGmuJd+*EdWy)w~bTHI>qL0t4ECWM%fBw z#u_oOg<=j&g6SvJru+5b32CSG31hb&t$CZ!)D;q>jl|!8(r9I|LYv(e{(6ebZX%5N z&4+{RFQ^V%xl;fdix4Y3uWrdu5XDh+!ljK1qjTtc%BkEzI|Zl7jO;VU)k5Z0R?Vz1 z1Ki{HF@Uhut2+ZfVtt$CTl^MjWPw=v7nITqfYNm;iw+9Qudi>+Kr53ljN+_7(wdZ^ z2o>x z_1!wS?Hr@qar7^USvE#tm3R=P@Cd8{hU|}Ic~Y9h+>lN==Gpq7^Ql)z6BmS~u@~_t z_B;Hc5`9nnB|y7Pn=ipGE`Ioyu{mCRAlHX2^L};8KP?S5{9Rs*LblG_;o1>{9^XN4O_c&6e(^X4K@35efnyY9 zaHeC3HgTuFvIRwQ-~)3d#eCpaA)McWU-B0uSuAZ)|ARosoX_Ca8xw_;aNspq0(Bi6 zO@QnVY~0?LM_(wgk9*YWi;f$nHR(>L3f`A9!Bq6tyK{%r#_jsozwyV+wU}Bt5Mh5U zKJBl+78zG+MC8zb9c`u(P`z=c`YJU6#R>i*axtbmm*qg-qyyWQBYJ{>KSBksAZ-x8 zUOLWm361+%fbk7dx*09^&P;dkkvQaK!?onEV`(LSV%l!gfUpC@iIE&Cak^L*ir&jAdVbbkf|7yJF?R@>N|58O_Xc*=YWEYI`i)xft z5}Bh%rkX9}<&3FSg+HI;Ct%xHZ9p*`&dp1{JgU?ub6$z1loqrVo`W&28Z=nAd9;!m zhLrVY!!EzCVfEtZX8m=+bQa@J^KB}M09?CT!5n!&WvihL!4WqliAaVjM9VnUp^7X{ z4U#5ou0Pkg%)1WpC9a!la){3Gj>J%k0tNp;Rob@DX{{kNqxq~ z{7Aq!FN6cSx~zPw9f;a}+|8)B(K9ao%=w=F9of9+nPYCe zL2a6PP9^^SPS%bU?u;e6UTr#?w%Xf@(S@0`lv4qOQm&k1@m%1g!Ku~cZ%BFLPZyOz0{@?ZEbcYgf{#< zRi}RF_Q;sEsa%7hWh{xi++)qTr(SuDsv%O3S@|rk>uD z6|9aQ9{^`alqSx#`GzqY+Hu}Z=0&qV91LiSkBVHd$YkC=Sd{5mhqyTChIvA65w-KS zTo9T?6$wTa0Wkx91WuoiU;RE|u;0+I&TAch|BzcQnp+;LXrpyZ^ z`{Q2Pxl6tYSsCzpcRx>z8?=%K^FTkq7%<%EAW(SzG3lM8Y`FnQKsWYub!=d1NO-(2 zy_YNwv?JYJu;}qTuU@xYJyCP&pbNdtnJ1J6ZIpWfpUze2_ck+keVLZM7Ng75F^d2< zyE){T+Xfy#zY-s>C`dD=KAw@q^%hI>(gEuGZM$dgbqy~_Sha@e_JI3k3t$nI^n{-O z^s7@fAc8tX8MHz(oBs=HGzXC&zITUm-Ny)rbGuhGptouHo@Jf6ICd%eS|ei9WVVZZ zki4;Cm!tN_8w3vC0pC5`Dxd$LB0p<7AfgEKA#GpQBRKDO?o`2{wNqA1iBU;7jL z(930^fVYj6tL~~emWUl_&UwiPEy7ft_(6`|!{n(TD?%`b3ux>Ci|?VUgohTM)j!4_4p?`RnVb&2F4n|G@2mpPCfP zYNkJ=NJ>pv$u=kGnCr~qgB23knv9x_RqQ75umPld`7+AbFw&Z z0C=61Z&96|#wWuI?rrGn+Gs(QS<%$7F5vo|neFm|G$@tTpu;&dU%3g015YdGy=sEM zTk?cZJ~4$_ob9*XFH=;f&}AajX=#tlFih&GNMIn3Po&j1WKZna)D?E4Qip1uSXU{x zI$vA4EN)5nnj>=}CuW-J)V%Ffuhd=R z3+@@L@h>%q zM>BZM7oe~1=kO=BArAKX{=>u%I&BHFc3#?J*}J$FKNZ-8D6%6UMlSu;zEu4?g?)sA zEl9LyrHjT^9v)OCUsi5mpm8b4CEDW#^VYTarO)yHkVk28CEn9A1fp{(+D9JPFd!;t zw9d6RtPb~If4}h4j?@P-OGlp3Kt1Ul51=q6mnDh4AMHc)MLltH{)y z=$Jq#R(8c)*5}{uG^eu(lOi!P6=13xl3P;59cPEYrgA8!Wg z=E}ECspm3OjH_Fbv6a8E?sOBoXJwE*kJP?S8WVM6mJ(1j=JUd6K|u^KHuw-dLe2TLLbfNK)N+SKR$gz9@$E&k3;o9o^_H|L0c4?@(_)l9O{ z)<&HlcU%O5QnP?&pCFqKj}0n5H)^V-RCxt;tJK~^3^!RHwV9CPqmi?=yUppU;nWu} z22A{B2wg%}tX=fTxdxlVEdM|SaQ|CDEUirL?0MM{~DVY(GyPAMMEs~AJ6jdANk);k6zA2W0?m)Bfw1m z)o1>H94PGl#8*b_{fscBUb*>dgAton3vvF)_!-@f;Mb2UY& zE(SM?*fX8n5LIdL%;g=n|MPbJH^YFNqzhTkbdMC{_WwM`fAKW_`_qd_xYGKWolSHh z!_*dC300r^?M$FpF)+B&Wpo?U0be9Ij5W@0a z;JgKxEFveSYJ>GA>@SEb(}T2`#!%F;SC_*rkKG#@(ht6uGo5|=cYpsalmDlhw*mxI z86j%s7eL(kyXO2a$n7oXS9inP^23+5Khuv!KmVMVf__>66=KuUt|GKLGug8wBs&<(iH&57=yHrd9 zkvAFXJGySKjG8WTB3>;cn#82sEklIE0&xdZ0DxA^8Zw803MXkjkeXUJUMCz{cCA{T%wJ*M)87v+|}?zg02ulDo>S zP~1UEjoF8EDJYzyZWCNqJ=w8FLDhy&M~hF=p=-}yhW~=-KObnLpOEN3Auw-bLyKT* z#$6IQ|AE=po1aSfZ7{IV8}~;iFP1aNaUjr`zg2qWdG@)-bFty^-ZA!5>=$66RDW+! zuvza1>qvY?q`;l<+PAfLNsRQ~#Wm*O<#*oc-h?%x~lfWFK9&Ix}8(>mYp6$ro( zt;O_5QTs)^d$biD(N^Uz{XlXNz`57~zqHABFxlZC^B!lJx7tPjz$N{NbILaJpFa*; z$BsCnhdIr&DKc!&#+npg7e9@9HC_xUyNDl=50Em`*U9)j=~T)2PlYN$kmy87PLP+N{J7~7F?MfeKQI)Nd5MX8e~{2|)m>~xnk(64Y-k`>H0N9$Qa zDybTK#N`2E*55x^z0Vp9Af!KGOQD0xo?`99tbqsBC89|`s3W2>E; zLZ|C7vuz0(Kh3Hqi~ED{J3wDz8AGIAYXbdg;8u^rU9=~C@kUu{HnTqnaVIp1bBYF% zfOl`KRdg&gkRwLBQ0)*O=*U`3N7y)gJ9Y!@hs6LgJ6XU=OwONf>u9+mK5HP$P}4o|r)%7M}v-4~D`x%Xf@O>*gbY4A29uf-V43G_|qHb=vV z8%Az(x;L1sb#n8kqAYQ~g0ojPJo-o5MGUu87rufx4+j z8<9H6qFwQmFnnwXl`#ArC?8Y+5@s|);h+TLZ?;6sq$Im>??G)I3A1ZvG_?5R+}I+X zYWm_#VISR1S!KfoOLWDH6QIx}ncV>%xS{>}{V&Lh%Gsiu`Sd=gG<})~*Zptq%3e3J5#n0IbceE39cX)qWA;tTZ6Kc@y44s3zQYWB6)rWRi5~3;pr`U2r z>3f;n)?7w5sf%OZs3?q$;b4Cw<111}3jmR(k&|FR8G?K41)oguy5(FKkq_mL%~oLB zw5I3|L%cH{+sZ>Mg{p-)M$V3k%!t(xa$reHL4KoWbQb-SWjHeC_7Xrfwl+!N%!NK< zO+W7FCW{`b`O{MsKFj(lqy0IyY|z&>!MZ50O>j8!_>u;4@w$b6pb8=_z0_XIrrxA|_52~CF`bOb{RQt%W%>iooAgj#B)3yPD?930!lG3Txjunb z|NUD>rAf*^G$xy*ja!9YiD1z{KO`3rQp%IpygX)G+Y&qrD&;sdbiac=-^ildrN|mz z->*h%@dAA!wj8o>DQCWZy*#6U(z?SNKrEU8AG@|C#<1PK-Jdw4kb1HMw* zATc4u&Z|++KXnC{!LoY8qW+;G3yVtiNXt;v-F>3+Gx-ayPcDiNG+{@U8`koKT>00D z&lq0Bab?%0UmZ+O+TLPrHh89za-UW^?8Vq$C%h$v=ArxhxYLWX*D;43b`-OnwCOyU z(6&ZoIZ|B2eYW=-^5ATs^8b1G$uqM>B)3}nOno)G)%NW%A zR=a@R8}<^jlP9EPHWSR-!k_trD9Alq*^=+F_&%g^P2+0i6)EZ}a4uz5Nub06?PuY= zhbd?2z0>;tMckLbL)pCzkLtyITy_ z%G|l<__I>|?oeIL73PD&mfoIvWjo2oXXOu{1Sg+xr&LKPjC78A{`G7P(XrmuQ}Cxh z`Q*J*hkC3E>Di*`&5xYBzCU3xZlUL>xaraGyy6>$To`rY2_Jp8Z<@zspEK^VzaZP^ zSJ!b4d;x0eM-iL7JGfKxYs-#1xOzt=c8Oi}N)Ho> zQIUAlW;gJth9w<*jN6@H1D=Uha1{#gKUqiQ2-Er63yw~)Ux1!}#ir@ig-cxw9*%rs ze)N0^jeIsgmB78HOi7VS9!W>Q2X8M$N=#00!*2-3Sr%Ol^__KW?%nz6OV5-;xH%_8 z;i}lf^x&o8rRQ|a4P_yC$1(P(m-c0TIE%?qL0!q;y&v< zKM)K8qq1xiQE{ECl;?6Y7ll`wKc3pQ0?_Y%cPN*eXbXFT*LoD`K4ItVAM2>2!8T;F zIsdTUWA)bjZmU^gL~N)YMP!siGS$$YWLncVTdzGmp>6KmoMrz`D*D5lO=%Gi@SO9! za(HI0g?adA`n(W0TOORgrEmS+D_wFpHmm{}kwJ)`U3wa-B}h)-)*nF`$%Kf-V-4Z`z3lu^Uhmu zza{uS_f2_wXnXsEYZAN84V=O=@21v)_!(F|JR*s`m%~iK9GIbKZpC!`X^v~vUT_{= zP0}Ipw)BsYELkc=PS&`)&9&by_*N=jMoNE;2XTH1RY8wbY-9*t>7Ak-zpM64dLKo~ z;~~C*!SGbJlz_@X9!GacU58II*r`dLZ;_?%yS|usse1lEq$w5*eRQq%82W**0Wa1x zo*sH=x3|F0F3&qF{L%u>u;{eM^YbDUCz4#~m~=Vi(@O_=GqnmeA)Nv9LWTN47M(lg z3mjzwlBKED@z?{8ub4c4T$6gTP z=X@UOVsq)@Le3rD4|=AW_A++H7Aj+pdaAyaQk%x=p>ao}bMyHd%$+ma1BM2cI+4P@ z?=(Q8q^ew>V)Uv`uFu@@+Ms$ha?@egIZC$&SJ|hF+m-bUuXQSMAi_nrT$%~u6_BOO zY3ZhZ!IhZpYO0*3XBxg69HDe7wr|^J1m`zSz5Z-1;xxCsk4Z|DhyD{qp&311jfKlE zxo);gJeU3$(NI4tI=X-C%5q$-WmoIlePt7?mm06>NziJI6`pmnKCMuCIbPCM^_oH# z=c1JVJfFuYu|N%O*e{iZjcLS9r`D&P%TH6*JBEsJl^9VPG+arL95@clz3)g07zV0BJk zr){M*Z_8}n%L+dM3$PY;EtA=M1aU70SALn_&T+AvO{lGY7KH8J6+8G8(~p14`6lps zWwY`zw>nY2zR>qMXN)BOQM{Vy>O(!j82i42LAN?1|pvSbRZ zc$@Ox*|PH$_2TWFC8szv<1J3qiXrcPQ;8FKZuRtPP#>j@Hs!k`3c2{|-t*iBkZriU zp!VUc5v|^W{gYEJJ7m0#O>^`ntzMikROz3-hB&m+)_L+PcrufDyVL2_VI3XMR=xwv zyI)1ozPM<6_Zt`tw}WU_b2p<}5v9Q$=Mwh8*fVV>9(O9=_ZFgirW3CHTu#j)hovvM zLFeleDEO^&wo@z8OEja~qM}ASpYm=|17^m1xN4)|16$sI zevDAE6pQq@xrzgfPVc7x$8r~z&|m!~Za0-=0i~)>ytg?b9I=_>sg1yEvCzy8g=Q^w zek*VLN7J3q`S$+v`jhYXIUoxz)(k%m%pvG(DnRsIZ2C4t2q z106x#M9H#sq&DlF<~OBL-iv%8?ymb4Ckc<|CYg8ckGth3Cr7WkQ0z*F$l$)c^YIC! z4$~InTRR4^9N&77OwKbe``x^jJG8zIG>TMjLX^&rO;89d>Q*TYD&A<;bMc=7zuTJ7 z7p$e=0nZtELif_h(Rg<;b|=v2dki1^;)mku^Q_`rjPZsjpk45FeR@};^86b zRlXOq?qWX>iIw07<_pgwn~bhIaEQ-@-lP1u-z5#N6&fk@`t+45qs+`M3tit|njTGd zm@=NwbJuMOy6I<9-$u|=woKmQBB$nT`6)y>EH`jypslgA&SqA%;pvTU_ZDasCqJEh zqWCVqmT2#!S~-$f$>%bKH)0d)Q4Y^T9=*O((a(WzDGI^*$#ofg-n8prJ?T59`m@(WV9 zDwTb0Y7SGkzi4YNO?&Z_HQFA0&?n-iK;SHQlgX{HnptPH@>cXo@K%68yOV&BKvV{s z(S%cgA^76Yr=g_9V@s*vds4Q>RWX#Rr?_dJ8d6NKvVL^1%~s@P1g~&KP^75(mmMBJ z-tM1%uG<+KCDR3IA z7&wh}#XX~+m6qnR-QacHCs1(s{=MzIhb&`5r~M`TUSBOOaY6_sg=K6l?L0Wta>eiP z!^wU0V@7qiQ6Zm#c^OOs-)+)2+-*U**(*ku78$T@JFn@7X03DHIoZ=-QP;8f71Qd$CEG8lW|MTte!5q9#(c@EsE1=>Pq%I+eHNmaBn^R2sX zi`Ptr=hMTM+iJd#eOa9)xHujx#t-Yi{DDvyv2+_VMa-wWlK7ID{UrZ$L57la%#}Lg zM(0p~$E&wgByYSnsVAQ+rXwSy={3F|DNlXL)|vW{kZT zrBUyt(=HxGUHdU-VoXtek^^;i22#CVSO-d~pfYkEL_6zcJpdDHrYQj1i!4!V`JKt z(&yDIVo|pme0$V*?iTkU6MwDw_Rg<;e!HZHQuUh9O?#;&OJ6QM$bOQF^3)lRe#9&H z1Mw#MgXdsg)vH>8cLAMuFWZJ5p!0ZK`lai&zTbyp>-}qypnL9#;mVT2gKhDv z{mTNp%a;3}k3QaPQ-K=bZt^WWt=VsS{mc95w(oASJCN7dXgSlS#s$OcqAdDSKXHj7 z3u_)Bl=X#Z3fm=JKb)qPyK4Wmt2zCE_R$8PYNKAm?gNqxgI#srJZX`0QpFQ&UusS{ zNf~yO_J3f~n(>I3dnPJ$Z6fyhy{$?e{`DUpDoJ_=HKWpnu1OEoPg>j@Y4>c3Gq70( zZ=goWE9}htbWqi?+o3tviH9#Mj6__8QZDI`o+(nB9i&#d&|i zm06{Bb>OD`(f4L(nEO>TZ+#daft+G(Yx9D|cGcCnMp1kSV6Ay+Alr8iN0qOaE}e5J zjj{C7<@+|35%muGyL7*vdlclPuxI=TPjQD*=)S}~LMi&^J(VbZw^j#mH8e@X-jeNw zYe*6)Be2RKst~T4^laua3MJsz)sjwVc8=q^|ObpPr_g;K$h?kge9Q4f0Z z-dvfpkhgx#Wa$lga`*1csuVPn{&BXcGGA$LY z@^g>4gfjvZmR`Jg@i1)CA*eRFMoIm}i)`9edfTRUpW#p7d%(Hq36)YmoNd{ar}yLo zV#3dy@fOaEDz0mrwBkZvc4Ht+aJEH|mMI{8pZiXGVkV=|(5`<#$$Pq{(wRtrXs!-8Cm$?S*{O zujTgnLqAVT=c;7e0e)l|-!opN;k<>fe(!pHa4n<$_viU)UGrkzUuo=#z*O08_ ztu!g6;XPjMaAL=xv3G>c-?InLYIHyCE<1d!^wo+9lKOtxEe^=2Z|A43+<^;c_MV|x zWr)ld$vC>>6IjV^vK z{4AdPOiGZcz2I_z-unz6x4l;}AJe^rOS-suX&2SE3tyIMK`ME8dT8r(u3AiA9OqnS zd%~T7+2wwkPCf&k&C6BH!oAO zAlvw7-dm%lCTW#~-3KKoc+Dz#k4vv!-Nxm_b%)+)t5eJyE%6XZlW#_&8n_n}GwCM)6atUYq6q+|RN7}Y;W7NxbtaP|-Joq_dTT=8KI`BI5 zVp>sA(CHyX4!>f`raG-8)eFX%MbYzvT!jpLGLEdrA7XHt+}W1a9ExH$mSS(8mzYrB zy;{1bAAAW&htXCopYCX6C-KEfYTqI3lWZGiPAB=j0|^p_K8X(o9)9e4(=e#B=-FH( zQ+L!ay70pyYIw)N0vxZxr@Q_)lpowGR_V(-XZnbRE~T#g^+MQG{;|TX)+QQ>Z%LM8 z<=Hc2nxkYDA9yyb_U^r|v7IS|f}Se&{l1$0Tt1D8ZRsz~ z?-d(5tdsS6WgFc`sW0d+2*ZRKm7_Pe=*tq0S1eYLmEa_~}W?#^Pgy|I)|z5TMo z^j)N5(*-VcXV}|?+b4<)#iY&51nN~KCaI33$!mt#b`MLdL|;tpwyK&--OSK+`73e$ z@@0mDD*Dq+Q#bl|OsB@vJJ&CHUHQ~=z-3eHjt@?3mRr897>>@Sm1~819KGQab#S_f z?UwUu_x#aGL~wPeL-T5Gx9xC?BQne?33_eM{F=5~S2-7)8LD>Aq|)@{u_iwamRn)j zVZq7cpyw>HN4=u<`fS|&Q8M~)omTaY&i;(7p4>;>mQVeuMY&z+!fS4Rb$p}Oxh-Mp zt;((kK_?RR9FR%^wVAB`X%s`T(Edxl>eq?M%en2;P9s=q35DEgS1xJaSBo~!%Tpb& zQA{HvpY;6ePCQCr9su9J3s4AG6%@)?z}&o@j8e(88_##7-E>HBx8~6=(5Sk*g-(yARC z;X-d=`-pnM*=!q@H<>jK44T>8GkZQHpjo&_FW9RMmw=BLr;j~UqS;YFznBS5ji6L*F-M%s{aCB}oytH}iHovYVNL&%W9#w`#7{o0Cjk zK1JD1>byR&)w~*?T^*dM@h$#+2TtI>^)YLvP4c&BzuIz&hU;z2Ug@A?J54Nae<<;8 zsB&#t8V{8%u8j@<-rO~+@M%orAagA`==tK)lhCo|&xDV(Pm{2NmZh334np z6bs2og?p2-O^p@`MEcITj#}7fYY8SE0~_6m=L$}3=J3*d`^H)|k&11IDdfq0Md9%u zh~v)^0);~taJBIw)secDyjA`<{M@lglunz&x~kQRC{a_ye+rm;qjCAQet44F zsvSk7S0$_4b?eKTlhe39eUYpHUX98euB&WOLyqqh;Rz~J0sJI0YcF*%^5{;5DK;PJWrm(A*V%ptW@N6hht2 zwYOUHwxB?R>QB* z`Q*Es_Pm&h0y6e|kVBDitcRPjguCCOT=dKx-p{*IP6)#_@Y%D3(l#wq2`*oFwD00g zh=bM(0ZbTa9+d|VHd?I&i{%wgs7SQ5ST7v;KwP%(+BSSY3g0-n<*G-oEXU4Q z3)gb5Z$ot_oiu6gouU=!*Xxil3Z$(5fe`m+(Q02g6IU_QN<=;prf2@5JidQoze#Cl z@ejlcXM)#by&8j%_ z>ubDV2%5sRGo3s4!sBD{<^ctNH$$YI-BMZ9&TIN^V^Rz}zHj871&GFYe5Tr+e)D?S z!A=u`oL|AFdzars$J35Gr&vur_q$`#`>7ZHs0wKW&+V!1Q;WDd&s?zmUZg$Mn7*3C z0NQ~zwlMYftCWw3Vs_go#oL2J`Z;vP0W;TLUEM5xb<2H&kj7VUw)h-}y1U1aR6T4a zx1{e&-7|VN)0D1fDItR}(-y?$?5XjQyQPvJkMJirTFHv%HI_D9zm*zTF?fG*M1CeP zkmEs|kjfSAJ)vU`$>JW-aoNYizLgx<9e+^U)B6mIbkXNloyjBbyaLJ>=C}I#5;}V- z-pYljc~UnxaNIBu;b0&th@IWqKX%&WPQs(@>LuUV)dfp#J)D{CJrf{(?(E22FO~g{ zd-+!myzHj!q~8-i*cIDaQ@h8#gSuf_f}1%F7Em6+Wftji9VOrI`z?uaT>gG>aeLc-@6VjZ<6?Xl zQd6^Z>;=&sMo7;{tgPfRHd&0JwDa)X8An%^L!*+LTKvn5yPmt<;#Ss55aePr!^&#EW#2=!m!Bvh(SgKdPKF42W;NXqB(8CHA*9g2G1?Z$NAJG>Zr7|CEgi6E`8yvWB>6=bB%}_ zw|m=6x*jEc(uc~M700L^&*z+xBrf{dc1p?V^@;Ory{#VHCaeI_gfR^6;laL$bQKV0 z@$;Q}((Unj2Su4e*vewRiq&M4?}<6B_pF;am_|h(8oa*l$Hgt9dDCj;YWQKnn!0i7 zN2)?6OYJ)Xbdq*HbYeK8rQN1Hwa2Fg-?-m-nx?c}mhI5xEv1&_*FFsH+XWrfGA(&I z+pl%_#-ME`g-rN~G?y%EpFSEu8IlrGU_ zwwQIr9at*J^ydheG|oL9cM>0L`>a3l@s{&uWzW=eL&RS0?J@T)IPzep5C*gBifs+8 zo~R^Rw7nrp{H+^I%2$u=<;F)04vuxA$G|>kQRY|DO6qsqF1ZWUBUvWbsbp z4+`GbZ?T+`I~rmy{fzmGG+tfhJRYe=tK5yIjTLo|#AWZK<~{XZj9p9MVekTxu0Ni) z-ao3PDJYGhJEolell6yoF1x+o#Z0!kmUbTD$*lknN9=(TpI9Wj0>*9MNq8yb% z`h_QQ5HHX4x_XAiwXZBuMty!hSw=aUdXw=0XJw7Bp|95-L94)9W)XKIn(PM^Iu^8{ z$4d8fkH?F-S`?eRUy>@O+PXdZ`nBb1<)PyX`qz!NhR!{;-Q+wtYw)R4hgyi9x$gR0 zyfZ5JrDdZwW(Rl=sf6AGVy%wLZ2w@ER2B93;H-BbVnv^aM^A$0V|EfBI!DKyu7pqe z^`HmZlc|IHp{ySKfo{IxLyG0J=LG$_Q8<1LE>qP?Iy-n%1mHR=W!)nxxPPXKz>cmJXFL^~)M(O!{fY zjCs>+TG65GTDo?2&n|%jkGf84_~m+K z(hhv*#IPE&bTbk%a&n?U7|L!mdWS|6-u13u%6?R+BhfJ%qZRV z_r*4834Uyww^i2c1nOpy$4BcsuF5kaT7|x0Z_QOR`+49xtZ?mUcOAz+nwPzu@VaoD zkyB9cr$Pt!JudSvW`7gyw9ZJ9IK`^`Y6_YVY3O^nZcwpG)_ zHtqe~V{N9vCZ5y_;=A%^0vW!<$fv!TyLcofAeOn&q%=6*$Z_7+TDSmyKZ5yV;VX}? z;bBgT+ZsDyCANEH1Z|(Y(HiQVzt_rYY8Is+>|kSMcY)hT>OAB4AZbmIZhc!dBV&^A zeP<}2GDU}VI}W^6aLKOVnn$FQn=x`4MSK13_d#6Tuh z51wWb5!sl>IpW#vqu*bWn`YvYJcsVW$vJU_MK$;EV9xCP``5=Wd_yS&C@sjHeCnNf z*2XTryAezT^xpdO#ki9fjv7h{w-TZr6=96#9Mlc9J zXjwY)m`?j`XS3%g{)s)Jn_a-qeCM`kse<@=Q86MSnYv*4loW4?j-hV%S{sTA9uLcy zXdVCT;ST9t^kY#BA3qgof3f{$8-Q|$;N;;G(CRwtXu5o)-%T@@YwWfr?I=EDgq9hp zKIgH34aTq61wLQ%kA+0Uws?LFPkeh*O+x7iebJRS%3rx|I>fq+8d)43-5-*Af+qjH z`6Fcc{al?&gD(a*f)R@%_LUYPop-XejwTTDH&5N1Fbq}U>XV@dr=~tr5v%&K#OP@BieOK6NTKQ9ZLM>H}4(dPKbX9N0mFcO3&T*bF{>u3$Jyz|Kf^sQ0 za%rPzjyAM5uc{qOgMDF39*j4k&SG9Jk^eI)&4FJP~~f;ar(E+e~?>#c9hPs^AUZT$piuau2- z3q~c{_%+Gfs{mXtd9%pdY4$ebX;dGV%UoGczj+$kpmQlHTmMtCN8xGQK8Z(>;6Ciy zb9xe^wR^Zc<$5#c)z6DvxeBtT!f~F%0UpO!sCkYFoX9YtcUE)_wkSm|+*#RuA91=S ztRbOTQeUD+Y%$I3n#wjRNzu2)_AOiZ^8(F0Wb52|3Ep`(aeACRw zq^m5UYaJyK3waF8pS);_>@AcI@6_j%*D z(vjLlXoZn-i-~v6ikc#y3q!{h{gAH~-vW7~cZOMb?GQP7zua}@eJ9htgD2`O(#!og zT%NE#)8OZerIx#tUuOSl>`XLR>QLWnuvP%mdj9zEYmxY-X|z<2=68DbuAV4 zIr6qmlk5<6lj1eofTIKodm!tIuxGeWU+;;xV&4(PEDyR&C9+rD_RxO_NG)Z!p!sD} zjV812;V6wgQHgav%kHmcS&Z%99lf7`sMFo4#4qd>%pgs?(0|=jNJ^6BbdpM~V%nx> zDocf<=ZoNBsU(EZaW&8zE2uCMo)$yNq{h2=$>0NoL9vD(WT&-$yzR(a**5T%4cq&X ztGX@cKc2@o1>KO(%KmiGoWgR>Z*$_&t5nzs)CDH7#LefQp*_*ZKVPGHn5u$E>M)D) zRG`V$vDg|fVVo&??U~4n1Fj|cQ!%z@)gqYC%3K$O7RB1VzBk@fj(k`x>*BYb@IHvJR z=X9&%)0t%3tu#VPk4A-rRvvTTs^HJMjrM&ithuvtOg}P>nlme~FX zoay+jF_|SkYQmy%Vtb$N5OeVmy|LBjvpSWh)`+3P&G9ad5TveyhiIv$yNVIFp{D1( zZH_{e1g#4Kibq5|Gj<_T-(y&(@Y_~s3T?f;L)2@+UTGC9+e=g?CFcRsl;oR>|hY{L6D*|HVB}ruy zzEgVdhLrXDoUNa}>hw^|Y`2OFU1Xx>E5C8bXX>Cmdp!l^$zBd-nwVXF(PL7W0r6*3 zv4|~#H_h$7RS#6N2#hITTID_=p2_kYahC3DYgb3e>eqUif??|vy&REYq{Ob)6fyJd zt=ytKSGkeJVupw3O7%=!3mbcuxDHk}T*4;barbNoM;Fpv3am(#yThp~+7;oeR3+XC z2e1^0KM=fE6puYpycI70KHAWW=gQ&igC)V=1&1h4Qs!-Kb)<3`oc!?_!AMDg*o2@& zAP}rXJoImq(VKsZ`R{rD&tPKm@cqj!|GFaoGeZBJxJC`|KFt8j(8~Y6w!`39$hyjq zbr@KyKe`2L`UCL;QO&4SRQLBbY5Y?K8_7H{Ex~^({X3KY3Nl~{5%4IIUgEav$OIO^ z9CM>b3D$0!Zr>{SW$y|JXRQVxKZzG1L0~l~@sPk8EzG0^5A8Vu!~fsYN}dbMgoQIv zlQj_wftKsj5ou8uM+Zp~+Q4q+55yEBx@~z9Y+^uFPRdd1GP7EpmgVp{Cz))|vl? zwgD)fNX8$Ajr!}a9F3#f|MUl9vS``dosR$Q28w^;v<3$i--tY20tp=Y1+fKQ;00y> zjoW&ay) z^#K1@-2m061ybK#*K-_Ql><(I0xFVPvfrJ7mW;j^-~!VP00V9uZi0Se25Yo|J!Q2F zWd(-VfW~2=>3VThtayLzznaD}fd`fUV)bz4`?`MQ{}Gn(e^Re5iUS6pOrt-mK$0xn zf`y2+2m-7H<1bxGehpbL(C-&-aGT(bObkS3g9fk zLwXHJB2J`tGgWf~RCX8<8Yn~b{)0KKuxn~h*U7Z}rp#~E`CVIng9<3j9XXBX?8Q?V zVHeZaRGh092KvG3g3Bt&*sz*;1+WltDhhv111J#(O;L#$)Ing-ncfWGD0eUMX&{<^QcN|9}aMT_-LL9JmT(AZ(k<+!^4L-^1v3!0om%WjC)7@#m>n zBLg5A0&qqZ#>^v_fhtB^83yQ~r$ZHn7nfEM+3Sl5!=s1|2%B<*Z5>3cL-$hBZS@l= zZ3N>OL}Y^2r`dl{=%4G;*YnH3f?$-1`ss(oJwH`7bvb~ipkggMY=!IiXI4L34Gcdv zEPvCaPJB-tBf1P_(D3hVz0L%?uH+~`(9xabg~g9181M^~{9{JjEyMO#-6Oomcq{oeVI)Y}f`H!S(W+VStfxw1IJ|g zi|%hSHxxA@|5GVkMMk#&q6TDcGPuKgKKkMqGfAmf2@EtY^I(80a3=SaudK>cyrAja z#nCuMnSynxGN38{l-44(7FdH{spSvI5KO)SXCwYvzELI-FPBDYHB#GwhARd!VbStL zk}#?LmNJy7kZ`+lqG1u2*VthM{>Zlxxi;Bje`y8+POIRJw{ZU1P=PcS?Q)jl$RkZD6$9LKbXGVt@hksjJIlAC^b&k}R7~b)}3{N+34M*y5ozYm=lmQ_^ME^AzI+{{r8@ht$p$ ztY^o9QU<^o#rK)lRg7F7d9$1dx7u^E*SV4#j?^VGa4ew9f_dS0d24(F{|JZsE+LpC z53U$;OCzH#6SDUQ0xI2DKvMiS1zkk|*&3w^Io+ruS#M)<|J=7i zRD~sxsb;BO|71u|`w7QL|CB`@Sjb*5N8QVW8t|tBUne7myw1e_gLjZ<3AK#Gk~IWE z!t87}Y+-t(M@||fcUyD169N40B(*{32R*iuNJ(l1vbf-Tu${@=NDWGIIxzdL0oQ$# zWbIlxR0}4v0qXS@u?qA&2;=5O-8DbK(H=<8C-XtfH-B2B6HP}a2oIW!dvFX=>*8YI z41wflAOi=icsK!ZVWbRvi(3N{8gCAO1_v%w2Q}*oO4D!Mra>YPUPB<$u3a1XOR)fs znFjoCogo3mTVS3y>|*oRV!Wled>5eQ=q3-WW&pn*tQ;D*S5;35M zeK8DXL`tv>QmSTY1(^m^uQ$}Jj@*lrAzH#6W!>uQXk_F@1Z*WnbP0T|8vB(;_b0SC z6~he&2D*1}3l@|6!%LEG`wi+zO$ctJRY_E2K22n{Byny(^PVIc7)`pN06jp3CNj!A zaYy4c1QX9f1~A&6e6E%=1Pg21%K{_3YanY5{12MON`zh`={A|NM2Z6zNPU=-EJqNx zc}=v)bUjQ3!cr0ynFF|r{BpR8pFr5yA$VAyDJ@J^5bKo63ORcDgU|Vj+mZ<1!@4N| zgroc7K;LizV-FD&*xcurrvKr|O$`8YWNKQbag7a&s0c=0OgafcwiocYF5^!Ae zc!-Ik0Kf^8Jj+UO+bumNw;-$={hEQw7|~rMCgMiv5U~{e*c=a?Vj_{@%E;XLgV)87 znu-ieRWc^xXizy>q`RweOyr{_$>ODcj%iNHvx!JMsb z?+?VTg5U{SjE7r}v?Qk!Av+h2P6lU6s^!;ZD^RnmOwjZvKo8iR)_~zMdp5`;+YZ*x z)dO6M@E8M~z|saz3YMMCYnz>a^SUZQfC&gQNXijwo%t&ugGLcEUkob~3cwTVAP-oO zNcMNwuuJ#B#tYf1I7WV)lqD#Fh*cs}!0YN*ad>euHyDTqoHvO3<#i~sclFgkA+<<7 zHOQ8v03i9R*SW#Efrv%lncb08;bi*oXd)@BBF}rt*W{m`h z!;9P2`P>44?53V&KP|`A8e4r20LbA&qPt*WRh$7yIzuT)3us&kZp>b_L#Q`cDof=+3;`U1e zm@}!NHi-M>axhP-zEbD|i%b?YX{`o}7J&I|@BwgJP2MQPkO(mZRS@z)E#GRhaN=LR z+6~5Xn}Ev!v?@ko0Dns(Y8tP%*uEH@1-e5uU1pgKKaM4xTqzOaJ6<3n-dNDpbOz?O zHd<~F_sivg8^1L0ho;5!v+GdOAld*Tg5m3LF4u?(1Mo(193X6zk;aQ>G>UuZzTB;X z$_D$m{zzA9l3lhtTQjh3Y2*aRnXgjmCEi^;&NCQ0yqQMHtIRCKrMI~J?Haa z*d#DP#2#lDPAVngY;d_id2BjN)d!bS5o?jex_#hwqYnUUl4--hC{vFDn>}Q=w3$R@ zW1&nODp4H`<~zbdc{X000!yB5l^WU97xCf)Fk2+C){rC@uvGv*WrV$a3TyQON97TO z?pF)`i#&s$iq@mv>?i=GKs*~B?yJPQ5_xdl-~^x)hBqR!!N6370W%6vz28rKPZj3? zo*##Ynl%J%vA{4>OaRSiZSQ@pdbF4pX2v~aldQ7^(?m!;|4M<(78s%lug*4z`{i>b zbVMDv;if?fg!-@`uZT_u*9}g9he824o(jy(VDkuQNC;NafRzE8#ZV>|!fA7sfiS0$ zKO`88CL+x+GN>a8TnDtFfd(XKU(zn1TzW9*^DtaGfXb{aoD}1bkT;53_qik&=foPg zbQ+kd9^ShpI~;(J4GA)EDe&4xuq_6LY_KSTh=G`*&no}{G#U{v(4zxA2}ETbL9rc{ zX=xw>0bInO5gk%5g0914(t=xzEpS9im@RHIJPIdaB|92WWd7j#Q``W{n$Jz$WM3=w zhPq!Pi4_}sF&HphPw-O*@g~qvUZhT217&rX5CfrJEXuYDiqJV5QzWP-W}HdF9|&Bd zn%0JzNk$=havA{F#6r`9(S%?)7$mA;cAG-8E8!XDb7U#MT)alcfGoG}qMbI0RRcs2 zVkbV;GpJu!Q)z<_5bJ;iWnj1`1`w81q55iRy!hw_>i5hGqYai!2AW^GRtB0fku*QN zrW+0`6GoPT$LKe3!bZW1H@^l~xrr!2tLE>7mP&)NFKdg5pWMIaB>EH}b zc{ov*PK~Aeu*Pts%YiP3If?~rQK0$*STwMa;HRfqN1%4qr_&%Ds1cWCVsare0%Zu5 zbbuj9N9##56DdCpXi1?r1$8nk&S?eiz;(upSc%(k*M*Tb$XoY0g5WV=Y(j7#K)~S@ zq$?nqd1n&xFDC%E{lO$p0(&Gdp$O<&{bj?%U<@I3#-k3J!Xp!b_&Sm{zNAqn05w2M z9z{r^*TsRgE9_ft0&B|rgm3UUSZ`Qcz!0(t2EP8@3FNT=vJQt8CLsVgNOjOO%m;`A z;~bO-ia-x>gCC>;ATda0lMW!nxB;M!3?L-XV%~$TbASdTfb~|ETemRE2ml0n zR+6k=B-tcw*O*tp?mAruxm80%&|;uXz$Q4%Rwg9KrmRB%eI!;COQdg zvJPJ#2iHo)vzdW|A#uSZght4q2IxQoKG%Y_#q(2>2LqV&2!BZ}dVpoEPGH~#EXiq& zG3>l1^Dt1llvK0^-{77Af87g2!tjb5mH=5i@h!|}9WL0i&K3aSL=F39W-ZkY_!cv* z3H#vU8F(%uzzYBpp)O-qB?Kl2YaIM$ff<%WNFrZj3wynUde7hBfhbtS&qTI1$vEFw z3jPYm!#j7sfN);W8Y@1ai)9Qucq!}J?HUQm+KupdL%<)|P(+6F8<6;HfI~aH!BcL*X>v#nCcB^Ahs2N# z$4LaMS`Rq3*+{jXC$S~h;x_~moCP}nyEp@YQwN*oFA0V>pa}3T*YBy46GSG>+OX3R z{GpBR6~GzH=xFi{4J+rn2;UvCf(@rQLLhiX5${mcNiV4vvc?avH=PXlf>K;ni{U~A6} zw?B&jycPwPk-s66OY6kXRYRe`wW>fo3H4*iY;=Ru+oF317!Y5w~g?4mj5W?FWLr!$>-nVPkRf(T?UWI1|gmYa5jj@A`VRa?>P`l z{?OO{nH<^j@0qjk+v4Fz5Vk-!7S9$?my zT^An1`&rh2M26(XZ4&o-*{Yba4 zp$7smEsZjIBnUWB9>9t9g4wWHWio;x0t?89WMkoL%bIA^#$5=YfGS)H!5VtO)$e7L z0S~gsk6km?wT0RKIblF$ZOPML06c&YvGCg?q%JH1<}X4JBOE4UNs|Q#km^LvF#>ZV zz@A+%*wd~@mufST$l=|CepuXZ&fuX0IW=Mk3=ww`3MqK#RaOsgv*!#%SC5W%1{F5ECc*SKwStGjNp~PehtVP ztT1ql19>EjQO(4#0!SE`)@;u=X$Mq;29~w}9|}B{hIn0xc$vBZceIq`fosF6kkxbE3_6lr`Q`46elzz73}>tMcCMS({q z0EYm)G!RN;f~G*eC+%Mcv5FplxV;|9s)K-U2NPpE)( zNUQ@m;5BR%$v-hf78r+EaTk2a>TVs&&1F;4{eLv=>mGC!@2lb^#pOHp|K^o4vT|TN z1y7l$!5HKw4kBh-QtOaA2^k?9i;)K7V1uBoHe|at1Ce_g`&?Aff)H{0*nh1X;j{CQ z6tsCb`2|s70{dU%uj>YkDsD^0Zxkmt+|Q!lGrtY5phz{{cED<)?o7e&Xnzu2_o6@X zdr3X_GqKE5>3>4_$7aXCBM~#E83g*O|A`Qo`8PGjc$cb*2=lzZ0sn%te!GM(upMTc z?r_d*asD^>KXImm*MzoUf0w?_nT+{gjOM>1|78CEp0S42k}Ov zgZ7(AR96lP^E1#r3v7RBMRVZ1`7y&y(bk(H=(dS82lD!Yx0cVnr)$=K6RhtMLr00_ z-IDt$rWz%L+r56sNm<-T8>O-Hns>EK6K zGdFJ}CErz#bmmJ;eKNQ781EENz1Lvgd4R%N<)f>0+#3q{Z5SPnr10_W2c=I=vCyk3 zT~)nScJ1QkQxrjrR^6K_`^vLf5jlkE}g+e3~Hv>>yK9?m%nxNXP#ir6V^4IjLd*y;s_RNW37?~k@Jw!odwuF%(L-MXKnL*igE1G|Fk6E%&1#wrxY zc0(~^%iLbf-6z_8B4Mq{2ZZPy(_i1tzZmhY-rbykn&wOEtlDA#BZIzpW3lWKQ-!JD4@%r2N67`om+}cE=cviS%ZZ;&df8O8X&oN=m)G zipC}-rKzjDt;JD~(nP|fy)hgr55rqoi>%|??YAPOb#d+k9*^C7T&Z-7K0X;c7`k#m zTX0rk_pvjQ&IZ))8tV^fjXLUqkD7qHGgC3&F$+%k1D^G!r2%@4FsaoPhkNhm%Oexh z=kL(p={C>ea2P#!aZ>AWa==(#vYC{&Rdqa7G!4IZqmSlP^mWWwsg?CHO>8Xj;x?%3 zsw}s(+=T20mHBUz29u$#lqOVXnJC3{4}0uTbgDTd=}xb7(xq8 zLujG*Dq!dxLJv*p(xpo`K&S%Jt0=v9kdD&3(yMeqktU!hcw2rv@Au80vu4fAn(M4H zd!Or^y}Rb^#L@oi;U?SZqs16M?Uw3H zT6B)roPD)AZ!OCY=oRZ+Ur70{IlT99?)vX_Iho~kZvT-^b8FGeg^_)<%8zBIpHfz; z!*}ofXIQ*DcduX!zxd7R+rDy;V8ivFV5}qamzOq|MYGopFBY=gJBt(5^!~i1-S#SQ zp)Z_&bF9rB91(QXj^56I4NmN!-~cs$@C+xdMEkc`ncE(6hUl$>N24 zsSTS-e`{EmC(k5_`O*!E-@Nn{8`~hC+||CHmq7mWuQXW=#d9&bF#D5(*_mw-xkbPD zy+_-In9F$NWTp4Y_B$d#*j|ptv5fyh1X*K0kox%dY?|hHuvF~K z(}IKpee$$1gEY-|VumjAqQQ2N`7axY?@w=Aygb1$DFEzVL>lMeDw$@egr$r7&Jpel zKOBpmYb84_PLtw{6^k(aFR22V-SMnenc$d;meq=~kovRaeB--Y2{vu5DTlxK@Q(|4 z{J(momPHZz8|2bF{VS~9f?v>)bKh3i>Kg6z{qKE20xMNOzskTJ|DEv8!KWB|#+Wpriu+e6lG;$6HgXp8Sat z!PmD4r~dQBu>k|kFb8w(-7d8uToUon>FbZOrB}Ckbc-Q3{S~V#Gwnaep@NiUUnjX6Jo?x z#EZA@SecngTa}mTB4WSwq7ZL4pVpO}+zkqD;r%vKG2v!rc5f^V6bBV| zNb>|v*oqqWr%&AXKw>DLS=)x*KonaMUCBfn@b6pWO#-J49MlE=2PR)EsJep&u(-JF z-NZEAK^Yw`tz6L{^^+1Hu%!X_#_WCwV}i{Z>i~2gc;T*0x14*_r(WC8^PjsT{>5WA z@Z0j?mr?!Qr-OGs8hrlq0U}qRmC0*RNZ9{1Lw)!(^q6w-F;Y19`}3c6Uy7xnlCFFY z)E42}A6UzU4}*-)&021}eE^E%Y-MqmlAuGWM7-87_5Ygu4w6AoN>ztr$;xT66@f(A zOgXa~mp^g<>XVC>y4ssSd?6UQskGuF&s5;JKI~R>GM!GFoa8vT_yfO{nWKHGAv8#lN#yh+iOGhdX4g~8(YUEDuQXq_j{|>eA1tvJNnfQ`~#@g zb=L{U;0nwFKXr+xWFmuo-;QF$wOXS*0@^TkM_eIxO0`fE9u{8nM}B+ZrTkCme2 zx5{@uDPKKqbf%QqQuq4B_zw^u0IZmG-p|$v$7nDdem)4C3eEZRpRvI|K+v93XwsXT zArg@r{ThT0m^ndgu1rgu?IcX5a#wDoY0>JbU19g9zu}c(PKh6yA5H%3LdE4toE3e= z)w)j{boExI+Dr-*g^VbSgg#Cn8!}Yo_Hgk+wEPNP?xCMBort7=aKk1$?@08khC>pk?iQpQeD|L4x;yQzxj_#)Numawk|8>{=yIPEy z%WeskNYc0YW2DS_montmyO4*&^6lgG9r*9lKQkeeR)UnUFFM1b)W*uwG-}C_KMnWy zOYSCxg%4(*hZ_9$WB&ct@9C|R;=kO7y_fWN-gZU(xYAOI5>HcBrB~?bvltK!HcfrR z6kqEa-hCr2@aV7pYlr!_{vTc|sFPmjF(wVB?a6+$z)_GhMrr;mqO7a7_A5y1tnzL=usLq6&MF(dQ5t*t|FaPg5J(0lBLUou_TL;p05zM) z4H;3y8DPD*f4K;|?gvq_5Txbv-3^A{W zd5Z>}^Ex%Yw==Q^wdGd;g@Vi60~1*VAL>S1Ua+_{Q~lb)6w-ew}eBc7I@*%~r5nc|GIXQ1waPo&0v8#A{otknG*V-({2y zknE32J2Vr*E8kOkhZ-=h%&}^eF$Go_(J7 z@w2? z9F;Nv>M2=@c&`cJp9AoXRG8^?g^30@k@k0o^SvC@9ZdeR-QU8RGirs0g?_3WWzc-( z?`~!laxOdIx)4CdHsZHy+ZBAuTrsd(ro2_1=FE)f-xeCompQI=7a8%}$=j3Us9Jje zJ4w>C2VE>Pq42(%fV3}re)6X;9Q+m*sIG_P5FXJ}ulh4!>~mz7GRP=%r{X86!Q?>j zt$%>t2Wt(`aO)L4<)FUP@6UP5=GISOY*1Y}SYfX4CG|7P9iW>8$l4Pilc%4lqV^K7 zL`_{$0>eK>QkWAoZCMA@r83H#(582Q)*=||4cudAC%6HOr^_i|dH=5tJ=Sl%>xa9W z>sGJK04-wgNA~w%wMl zv;(A#ES?U@4~If?hRgrh&H2>agB?bVu?R;_qH?u*f!dp4R!^pKO`fcC2Q&cN=_h+ z7W<7A>!kR}R;&B3ko1=;S%M+%!8+{tIt6swgW&=?MbU5BgTqaDyJwmN^5M2$O$DR& z701Q_s$}WB6EeYP7)l$ppu7^@r6lVz)p|DDc~8oZt>js9x~Du@`A1TuoG-!ql@l_S z5^DKvyWOIhGk^ZfIZ3|4mp2`fN#yGDITg+aY??Viopb2(^(*3Qga2bjxXATj=#dG= z=U;6*O4lq0tj`nOA?;Bm(CeWIi7yu6fA+)SWeckMs>(Ll_8m3BV&337w$6s_v#3Hz zj*9jOMu+m@9y^v7hBF<#XBG`RfA4#LOD1g)%0BrApca%yy(9RX%=$$X`RweFsdG7= z`J9{?rR+n2`#+ZjNn(2b0W@lT?;2f7V@NUvd5)4hy9i0Bu;B311=B%lqchPu8#b5- zox&|@Wh`+(Q)&F<$J}KzPa)q2>*r`<+X(t=p!E&z~z+zofsLK{twHm7P z`TOS|({m4DZ(IIuX#(p{he!k4>LUO*w_Q42W7SX-eLf5jeBB!lY>LBX)?7xtoafsI z*|Aqtn77K6Q9Z6z_vGn2J;s@$jEfU#(1FKdhaBt-#x;=qJiB}SoX=J8Mid&f867&y39Ng z^KZhv{39PW<#D~bjH=*zamAvI%=@woZi=seMLKhhTNlQMk_$jM3zfYQ8AnP{JtSjn zoGRsSfJzHC3Qn#+2TWr(DKncq+Uuj194&m_F^48BcR|V?8{Mj%X#M2;VO#ydfnDZX zpzy&;s@ZlDfd1E(KXPUE@pi9q@bKST)=9rc>Mkcs6q=5nUV$?y$(O!zo^odJosO=} z))ejqM6O*lC*poLRBzvNoIC%K6svcToN8z``sOn`^U4(GZj_on+n@I4u&PtViy5<}e}GCwD*KIu z)sw5~^=q0_`9yDCNJ*P;@CU|xRhPf&%vuXNpL-0?>VDDpIsduUT{A!2@8E7BZ_`lp z``u&x!*Knf`R^~MmcQryeYLRkD;#ZwQHr7_@v@Yck8{!adBH0F`ZvT_Z!1m0=j{UJ zTkdW|nRB?XWO$}-U$8Vcwpn}Y&gWhXuy%t*^H3)qA{f7{8r?}Cc!Ju61 zXsc_t3AOXk8+I^K0VjB$Cz$+y&r%qRsKox-vk-X_d%xo2SOP17Nrg4Q#09hL6e@M# zUh<$gyY`dO6#*&9wvgHU%)hI=bpEg7{}6%z{D1XI5cvPEAIw z{pED+->3m-+eejg!TepGJ?F^w9(_Qqefsp5KHVe1 zFQ(qHdXL%08^GXk70rf!fKti~T4}MxiuwmHK8`<34k@9!_Yd&v9{-kIZRDNrWxNd! z0h|eOGI~1oH=-ujEIA(LPQ~R_A9BBao_+;WVgi#7by=U;<(ATdx*|U7Q_Gd1)Cg~^ zPJj3D@F;e1XKHA?Cup*{C{G_OJgDhtq% zmAUV7Ta*BeJ&kP{7yj}f-{7^{);x2w^+1$yVg+lTPk-EEv@1C?m&}YofWWI#tEKZm z1A;wlEnTkBP$)4cZuzKed=jn(c0!~Vs&de;Dm^*x#}L&`nbPu4(hZnhF3W*&j2iJ7 zv?1O@(PD5mEYNh`a3+ATy$(X-MO6<`KA&{7vc(~IWB>;ueqgK(FaDr6ITFn#&6q!Q z{}05_tMEquRyRIn_s@Ud>Oqk5h~Y}HH_X|z^b~yl+kbh3pa^VZyLwc*bjFLgb!!oP z!FR~FYeoCF3U_?Bs9+g7Ta1lq z%Q$qWekCaI&^vbv>?MC(%I2|fB=*{wnV1KBptZV4X)~M#QBEiU-}Ob2p&b8!0D#+M zB)OyFxl$ov3ohl#Y#Bz*;vQUCDOKM-TjvNmEFW!Zd&ut9X;Y)(Ih+@+npNVmRF=QA zeP_v+DRhsr=mF7G_Yq6&Cf{S%@-@Nmp`>5pekfazicWk(gj|#1OwU<&q?>kVY<8NL z-=yWD$8f$AbDmf~e#%(Zpvsv z30%wStj+kuML?=DHl1GX3;m^6$%A8M_qnn&LX9YNS$L3#s97qY0`FmOy@dq=?$9zU z7mUj3D&JK}O!=x9k@?|JM`_DDAXRe|`Lm8&dc#surSVinFz%hxq=PgALq_H2w|07t#W!GsbB1y+WTK)FB@6jw6;${+@V9^f9nZ)8XeQWX?92H=R~) zu_^L1B6{K`Re)d9UkJws0`;*SK4KJmM>^^+MmSu*W0w^xuV$K=9SGL+0aYVihtv^? zV0MJ9E6*S}W7N7rTj{KIv6op0xFc}=J=q+w3@ zeOB3=4Fb7-R;4Hz$*Lg?Gay5>2QwLN($tSocAkX+=``8Uh_#j&qO3Tjibm+W^5%s- zTr~*N=*(QN0>#VDW{IB8BeG*KP()J51gqMjf1r~Jz!J%Zricb!^hIg2Ev5n|s;^|9 z6TT&gIwf$!t~zjE@ubJ4HusjwFJ7xOZ~Q9oyp)KXJf-rLRB0}YOz)x#nU`f0_-^9s z-J%`%Ly|p?v>yUs5T0X#fn`=wW;ABT(Ubj773~!9W7o${WB~oKnMF?x z_bOwIH$Zl_d=`~Y1S0HgX+p%s1celhdCI_9I^HjSWLK|_^#N}B+;zlApV%?_Ilt*n zi#e$C8w?`WI^&><&V*T2)W@Hm<@9KNU-}G3wKkd}>W*|*&D)>i{UV^E56nwD=N*JQ zlGWyhs}vK*q*YkKgETRPe#7mpq_P8D^p$!O8Vt?PJGzY(@+FZdRXe4NLIRW&!PV1p z`ibdGA`xcE;OtMyZ$uT5zp71o(C(VwdM@>?wJ=TurUDgxU*t2TTQf81ie_IXM04(( z#5y_HvFVb__eHK}P$HYkuwKRF6LZS#E{Fx{`a2r7%of7L0aTpf>3ZFUaHe(Z+TvqA zycRtLgB7lsII&=pEh$lSKZ)8GY}Bz$3VP28&^u7p@Ki@plAQ{L7pb^1ImK8+SXj&h zW1uLml+W(1R@B-{F32v&)^5jsDa*P}W+G@gS!>H)s{O#=mS+j6v^@vvC+-D#_keaNV&3QP2HFKs9&~!5ds}2Q9X|I&ko_5;(bqljhuuDmd+kN>f zE?DTR0?LW*#WfvJ=K`%lN=sHT9Ek!nG8+|-%_4S?s{AvrZL!-Zs!<@Fjg9@2$7be# zDL09#pU0*C0R|o$HMDP6U&1$z4exqgP^L;?br7_zIQ_!=Y@VaAZpk^2F9!G`I>g-m`AXX(~F-JK7bLaa)+@FV~|FB ztmlU;?PE^|##ld4UB*=<&~ZvpXD|z!qzd$R`Q#dy9=3{%HO3h@xKgM0x!2b+Y#VSN z&E7*LsPyb%AOIAM78@js)4dUwY<=t_YK&Otc)}nKrN-Q;P?l&BHjhq5@J~LAD1KM| zjKa);XzlnD&D!39;j}5?cib2j;RzxQf%kL1>`TFuq4=ab`?7$h%Un2qUCUqNI_a<7 zFGHEsD_02p|NaBKF&TzOdAVhA^Y12X@E4VSv`p)Y#U59S9=yn6fy&xgp#4kD_-M1}D z*d9GVw9wdWGZ_LAYGVr$@fRv1fF>I<3|5q=C>J-)YvvgRs#%)aidcLkx*H?T`87MJ zs+!Y6)3Wf+3xRMkU7?%+lLVg&lpjEb4NHgim`T}ZWeY^pKWiK`540+-eL9?~b4%M; z^rLis(GD~W!q0RI$SSIcEKunmu5D%+Sy9OQDk$=J50|L+Igv(x`?gs{KT&m9$5lUfk@_>n-nU7-IHHKB7Z51;-#}ttDti) z9-wkQun}g|Zi1HMOPEx}bAOjsZ8K7q6lFnLP--WQFH94o_DA4m`Ey);z9_Z>fZhXNafpRxXAS$ zj`Vx@Bo}jf@_Ab(0CELV5ihlGeyMmPFqxg>VRP?;TL=_29{_05q_LaB#sug-e1@J+ zIlf~L6@SiJo;7R8SJa~RCgf4tjgS~|vmACi&*W#2U(e)e;5n1$wE{r{%G@)iDanrp!_c5{$oHU zn^r%kS0A5Z+2gN@+vI-wuFH3Osa6eD`P?l!7ScN3fs_|~qUiIF>YOE=f$dP!=#LXb zzRw;|h^q>eT6^XOT03;oRu?|Z?0vc|37^gc-CAo^63j^SB?^3H0kR2}O;%n%R-xhg zLh?i0KF3ny`QvU2`|s&-1?5o}7I8k!j*{7V7qW>GQv#2QKc59IQVR#T4}bwaE0IcL z_yj4&EmLg_g}_{GH2=>hjmnZXPpVgP^yKMoo3HLObs-VA{=PoEfpIeP3*`2$6b$+4 z{NiAnA~luGcG`2{B#B}Ca_EoqLE+l?=u_-Wd`gqKdU1*Edu)hsE}EVZ98E>mhv#V@ zM06}oedP@OG0kx!8wPzV-jFGYWRh*ku^!zcdQAQ6Vw(4{mdCml+ikb8RhWPYY<90OEwNc#_E) z21+46W`KWnGv*@2THE*9})c^6BmCoh4ZblUUUaO9#_&ZW+4lbhc zXIO`xaIrn7yhreiY(WN-7{?gGQ3QSe7!a;uq}?ZD#|+Rem69eJvX02>|3_<^ZfKJ| z_P~nnHdsw7uSlYWoTIuZ`kfw*AEwMJMh20Ou*=_sXAQ2D*gN>k4A8q`bT_osfE_u3 z)+qy|%?`kWE+Sh(9&zKfQ0h`yp9!SGW$}I27ngj+Bbk&Ps8TDC;0Pz31J~(y!9C+& ze`^S1JQjojS8kIo*cS7mO-jjSQ^PCOv2??9uds2!m?yiVE$q48ckUW zCrD`&q{5wIrH|!OS7B12MKfOi;$@0k$T7RPO~dg%TRXUQ^|L`-fz!aT!Iwo1?NQ@D zN`trA{eVj7Vwgel6?{E2HS6g-b4L&b6aRi zcie;4aWXu&kIF?T@{zm>Dp;LkEr~QS808QtCH$O$A`1@)R(p4%-%$o}mZGM<18}=m zgqo^oJCl?k&;xt%0OXkj^HCUvmaDBX5-M1nt zV|So{{ED0sA_2i=>oIuO!=gF!#*JbrF8!x13o0(2o22D1)UF4TPqvjjqqri}_c!WM=fxm1!0otgjEE)-i%hY@QK?Y> z$I>IysdP+3X3;pz!#ALrt*=YN&hph?%NUJ~Alu89JlCmJo=RX5;)oPyIx^c@xG1t*r!+!v{ z{R9t(*Rs(Ay%;#PIy41$%WR4i3DQhDGOAIjvC&D6GlXAUQl>&vpCD=1Iw()}^ylZw zyStj=w}(nKIDEM3o)vm1?b$J(S^gGp9609SLT}@`_>OS9jKyyY2<0CRHTZDsB z1!EdQoY>iz?kYIESl!PePiqf%9*;Qm__{(az{-=M$d6Y6vBDmi9IKf4XN=m3#Jx00&TNzELWXQ&2& zm#!~*(-T)b7fGZemEAOy)(w@sRnOzc2T6Re)%4^5a)Fm(hRQhVn6Zpf_zWO!hiMQm zQAnr6Pft1*t$H$t29j=Xvo3W)-Lb?Wt;fo3R?Z9Hn|8GvYbE(H9?*JYm3(M`OR7>< zpDD;$#Nt{;!yX37>kItyaNg*b z>r-8|&8i3vs*j)l7Lj9kVjq_#^bNtfk0)pYnK%Xj{6HJ&yN+X(jD)m!6&;mOViRRo z3hVj=7&6EpW9vr~zpui9bjSnpk!o^Rm6@7azySG8-woyE)YECx>jZ|Y%o=#qY5Jub zMt`J`69lb)dtu1{$tu(cb+t%{$OIHi6ChoTfLXl)xA))*v56wS8F*m+OAIGK_vB-x z8Rab8o&CP6h=D|Xl#}5j(b@W%O0ra64U2&n-P&y0{z=J6u5&<&5$@x229Z2(QYB&X z;baTr8`D^BU`;D`FNr91@GiZ&Z{qzSrT4jFd^&3c4!zIDq{FYhn+-~>$f$eQ!f<+A zZ+a>*IgNbOdXOHe%>C{gRICj+Luk^3r_~rM8|}K zP=fAp+Bch&_IfPu7F7jPOn}WaXYo&K^&$*+1$h25nuu%jYG+aD1ZPEq8HF4~8BDdy z@a;47MvaEeJ#SJt3!{d-`?)QQ5^~QXo!JKlntuEPFbi6BFKUrdCzr_;)oG$Y@_&cy zORfytm9t_Pw^Q4IdNyb^_r53ylu-=}@`1WMWsyiDT@0r&U399-Cp+1!Pj;}ek#pvg?D#sZ#1MzcPQS^~9tCw+*R3ZAMLDAT56{RC ziXbA-1WcSGvI5~s=UZWZpC!_alz(ePRO`B7`X3-hhn9B8{t;M7^5+gIzu0IotkTJc zbQrA$IF7HF!}@@Gyy-&=Q$$Q-{BvxDMiSs1EXgE7rQ)rQ&}r> zt_g}`Hq_3qjCmFx+%`6D!+))e=uvbCC*1u!QE#>8V+((-AR{VTOMGjn$r_X2T7=&q ziii$99_Hq7bMG^yCS)c1{mcXcPMdxx3WbqsWdHR>ymRl8){COii`~#h#zF9dLGLV7 zXNJP=1ZG$8>b164%n+C4L_YcUYk&E!4elthKK0nudil(TMv3KJ_1dDsk!T!!PI|bB zonGWN?mfhu67V z7<>G%-yFQ!^_m&8xzq_x@Oi|8Ha@|y>>(O!gfyjWWQeNHr~7O$Pjfz`=nL`(15fG{ zs0_9G1H+aI-LI!0hdO8;#;G_T#^611jEYK8e4%$FAtX{L-1YKkxt^U?(w8t3LEwjL zv*`1sihpPOWz1B003gV9p!~9Tu6QhR@a{87YAB{gV?lIb$VV8AXR$UY(^i;Ov3P3N z?C3Zt2%nOP#SXBJh@4u+Z$EhurLV+F+rT*Xw z4|Ju_clu>O2b zwTIe9)~0OWv~ae_RfR^6hi9dR&0QB5IvWXBeHsCWt84ZB?-5`x+`D@`dJl_lBcJSf zk2t&kq2g5oe(lTqDg<|nRK^=wRvnaQz7!PzV>4UFEq@vpSUx}9jifODmeGaaEWl`B zOi~9RacSK5Qfqr{6sf7j#AD!KAIoFrH{LyLP#KOiz19IF85-tAxlk~#zck{tO2Th!fP<4akdg%GhDS+IQy0dJUP;g z+sx-g1u6Xb8O@x%Ko}A}tCe8IBos%9_^L#sql4e&~XH{sD9^`se~!hj%z<&BF8C^o>_Dp%YpC zkm&FkV9Wa%0qmr|(Qp1rf(E__;4!esn@MQ&@R}`wH{qpBB$@|MM!GE`tf#9@x#5i&?EQauXd7ndT!sFMnlDCYv+bf6F8LE9dvp3=yu)H zcKzqcHVT7!t{CK;QDiRx*OY%-$~HDW7)}mpanMNN?|4vpXv_Q}smmmNq|+!9A*Xif zS}>S!8J@`e4xKQ!uytSUE>>ieN~-5;A)d2IBmZtTOj`9Y(B5!m6`$5?fH*PL%Q7dB zZR9u5EoWk!)P1Kar+Mc%=3KEf>ZcCOZ&;sZ45+QNs;Do4ZyOdlL7x{DT9Q)>Ll0L3 ztxTWz`~3qf2bpH|ifrcphaBj21kFM)25*w&G$y(S>h*5SW5vt|nRre1bz&6??@VQe z7u>bR1MyuO%gJ?7N1Pn_WqB~6%jo+#rAOEW6;Lf^gO0k%H69QqyYr>;}|Tn@laP8x-w%@2%aELu97}9 zy_dg9%{Nq|`}k0gO~6nrxGA&1Gkbf0ltJYAN>+H@bOOIs=qckcfzCUPj^kZJPQeV8GP zb8O;`x;%gYTE}}Dl{kzb22Q~3!5kq-c9q?^M)m9850+CSYT(xH5AWUUp6ii5!_Z>r z!XV;WQu`@XQ+=$;5qZ!kRg@MMRLDe{hxPG3f*p1kC(B+6Mj1Gr`1ZL8FTtSWkrc&b@K457hsm8HEP;!h5&GGf z#Bk5HeD;<}tT8MW1iFu`IrN*E}MUQI^Zw6f+bdmaR+R|ku$CPhuX#9?(;IpFScZ)cl^;~7DhbzBht5m=0HsMpak1k z)B^pU#cNkER%>ysS=gKssb|neUCd49;TzyLEtXCs%sTv#X0*}DAqj@)RQBL5%hvppX2y8j?T42bi&1v5ASqIg-&wHbfaaDo$CmI%XN`oC^T2VH86abcDBdRsKXN632g-2c zwiF-_MzL!M!J#siH=kImEM8>DL7k)}^hbjd?jm&M5Ou(m z-kWNN1GQ1GPGME^FiFlY(B6N%^g4w^Ym_=Ii7$-f(!WrF@J+ND#V7XU*$zE}PGTa~ zoZjhs*|_BtF>=RMk7sQ3_D;LoxTYNnATq2CS$DLv zf2gY|`^DC#m?7L@?YSH~{n1%7;|UB^&-GX(g)0|YVI42xp}i1i^PX%3$4VY=LZTU> zCpnHw((~GN+*cIrt4xTHyDlK4vAq0miRc(+jfvC-0V?^&A>Ju#ri23nQw??I_ifrT zB5H`WK~gCWZ7FH8nRummIwT?`KW1uYV1$vPq;k$dxVJ#lNrZ28nDS}@YZC)4rKxu* z=ub)k`r!ao0ON{NrAv7H{-@`Tk>|X=={(I1xXoiC_Vo!DU={SDaGmn=o`jOvC%Puu z7z%T-5D276bdtOlp3kp{@WXME2lZH9t~bhZ8M^8Jre@(@Aa7)}V9rDF#gpl`RdTJZ zJFIwS^eOhU1ohuZTOw@7>H&m!at4IOgYKg-DS1^9JoJi65McGdjqx!_!+b6IcCo^& z5BfcLr}u-A%w!P!p3dCPqa;Vb1c|GIapJ>^V94|H2z6*(PmWpR;SgE1G2=@-kh3yw z=;7xDlaKQpjeWUDPoH+rjjl3BVx1<% z5{z8@fz@{G;Rei9-OnS7^g+ijWiShPrpXV;UqL*%3jNs%GGNJG&v_Uh%(&L5;BIjt z;WvyO(e&mkuZ*ttX@;$as6BrigJ?Q>2RHac6?Os(o-r0F`KB`Az-*Mvc8k7X_?FuD z{Mx*W1Npf&O)@MqlL6Y(te0v%27ssm7#m4w6BIl1bNn)GYbk+3i(A#(kBmMgwhz0` zIO?d+JnH5R&td;UVfa!+vqW4}{~5}-d?aF$1Z}COm4f3c#&z3#>NgWpD zbm|v#>ygD5lu1Bv-)nqS1s3iP=T_`$NsRd3GM$mndi|#`?`O#`d0JYU5_xwab(aw` z0tx^?mbe4!Y0Xg+=rQ=myG|Z`?m8v@Te0Kr9|<8T6ex>6<;9npn4A7F8_br7cob>8 zv17hJZXI_*L$6)Hkw#m-Dsq}qJeH*vAHi@i%~;H2sNv;eFj)h;Vss_&DA`_&9 z-x&n>K*&dkEZ%=FU(c_;=x(u1s-pFxMK{(qf%?kr$KtON%HuSR!BB(ob8OrVO$7VB#HUx{uY&)iSi=Z`{B627YyL4$XZD9eHuOPg}LAkS0cP%E95- z49osue%`Nq)-)N>>6jGCOf?M-L$|nt8J2zOR_SNskA$T`MZZ_-<;e1U z7$o4jtSIeB@8YTUJb^aR6h7d%m_pfmae8GoGrt@r54Ul=eQrAOk{1~*6ZeG}?FwGu z!bCfjL9n_0 zOx@P?yb!9@-YTCUxJ^K+=oH1`3Zv$JcQPd^q_|9A?3`_wIey|>8`7H0l>I|V zUrKbkDfIn!r4)rI`|t|Jm<3NY9vaZQZ2BM;S{=#2yAu0`I??p0 z>wtixojmceZVAu%IFZfH-D$!4SEFHOw>PML-n>%1CxR#?1t~(QQ~v>wa^n8dUc;ng z?jC<@7+&uZd%>NLO!vK=8;g>PSBnbY(YKh}*ufC=%veCcZc$O2copEWX~5N8m8?Rp zid({{Nl_`SPyCJ*%;#mc_N8*)lpNwE(Z9>?XQVcIx20Z`Mv=ql0NMT|)uz?cTU6(` zS<}cl+&f~d7-B2|mZJ~=X!PTeO^=~+1TcUsUL>Ua$Ne$3?_?W-(njC%36T>nalbc=k@vlo>3ESt++wd&)txJ9_|M_ABRk#WZ1>;;+k2u9Wo`lkRYL zV;^iRH{m7%7t6Xd4Wc^Pi|J-a`gnSW77k0l+xF1hLXtREMEo!C6x1K_H{IuKNxuDZ zYua&LH8I*1nv>JPe}OvoW%!A1I`;BXpCh_Q;r!$ly+Dl z=vU<7McA`XPf8Ook(q;m7lDkcIXiZdF?2{l3E%cV=`RaUH^FJU&D(dXhTql_btL5u z)QlJVGG-lybD~8gk*q9)q8Z*4n@=`SR&DR|Zz0#{7^-TkGBS7Mfh^U7crxbZB=z}t zjmC^Jwlv%LDG8GIq{+TG#-d&WVQ(m)q+j&yGOOTwCvkQl5?V8 zCK++Cv%!hk>+ioa{)y?Jnb`Xnjly_jd}bKIkb{jlZp>VA3NLA`rk-KHmFB(i%EPUU zl6+>O(63@CV3J1hZ-M*svbgNzr1L}B@DF5jN=X|$-Zyc?g$vTV-F@S?Xk^0#O=J}@6c)5r4#GMmlB z84Oq#CJ%^vRvW?riRMYi3E9V`obh68ta%S<)Bbv(`70^X#3>Gh2N|I94oBMK-^ntt zQkVhQd{~NAil#{2Te8RZ#GEyKWN}KXfV03^2-fpsnH=0owSUCRvtY=8TD9CkZ|N;5 zMbznT`2!6aDaO;)UOoo&bcF@dJc}0)%HlvTz>nJ>8xu$I0Dz4kazbqfCx3PCr+uA1 z3(>freOVPoUbM7|IVm8|RGJ*(I~P?)XNs&y5d$D4Dyal!T-GJV7!4cHCFZ$V3b1nS zg2*{2e1QF9TBXeBmwqK98cMLAHfl-<523b8V4w)^C0Qz8nV6(q*)#A?QSo;Ny{J-d zP*PO^Q{aj;TN>=}4b1ti3H2qToO*DEOKg7*riN$bOxWP1A}+(%H|ct_2PO2ae8uk_3Aut$z6sD#vOzX)2rYy}& zGUl5jKG!6DdC^QHg-wi^jA663;cFtkii_}UC9&shUbIM8Mdt@Zcb>-DzNYzQb#`%< z{3*Os_$m(@pR0-{8Hb&68m>cw#NQe0920l&y%^9Ei1ln1OlIuSD-vp9c?7Go7BZ37 zFr{CT0~+IADAJ`xB=WFik6Xb1A9e2;)>QMpi>`zO5}HZqO+pC0BOSyL2pyyo zssRB3ktRsj5Q=o^RWwxTQl%qZnn)KBlp=^=LB$5@yYv3`-sif`|I7LB^XZ&VvnE&8 zthHv%dS>Q+?&p50O4jG?g>)?%TDkF~9mJ6a=WMdmpgx^EPJGKwSCCklZ3?WxdJ=9s@E6JSz5c6 zVMN@S0!+Q-o}pRzEy?%Hmq{HC*Ud6dYbJhC^$1^SiWA*;zwxA9QW|ILtbCLmlxw>H zNN;RWmrIr)@@#<3W(7~VPI#qd;9#;_NHL1IeLXhwkZ}^VMycf)r~e$65Mb<-fMFWp zm=$w$%4T~ym#y^s;K~R+{v&GAT;HsSW8qxzYSrDZh&EGbfLHU9MpU`?XhG0Q8D@l0 zL2-8Wfyv^jl8~t)%cC~k&S@|z8^hd|LOfs_1tntwg)dr7aAz7uxXMI&M)JPU5t$p^ z;~jEu)N8erryWfmSbbt+A0_}6Gl>^|`NldymAbgKxOB6JWLBSn2WJ??!oAUPXTetn zwUcV@)Lx&^T$*-?7=QWl-9^QMhjFn6l25CcvoGg$2cTF$05gO5rcbMPxT?JHZCAAf zDus!C6P|^wvMnHOaf-Pq%f-wHQx>$U-L&Z9u7kZplxKwFSvnDgHCY1jfp$*H^yz>+ zgkMyM`RyKraR_Pj-Hb}2%aU*i2viH*Npb`Lw2~Uy%v0|(05H#p6-!Tz4^qVRlkVeh z^$2T+B%m#kRiPDREsp%BI#e`>tj4d5Hap1e>%nhOKF2emPZQZzqzrC-A|1gPV=?Ne z66Z%u=D#R!rRrpAvA2T&O?VY%06(*)SQ-n$K ztF!PVkrJ|1 z+pX@|uXd=KukWYXS+0ML1M*syZf)5&VlLL*@R^@u>MWfbw`Z#H*XkEW<>7Ev5mzr0 zE0O&Z-d_;E1FLBG}f+JRu~epX+ZrqCWj zvhg!(ROvoxC+Tn2H#NN*+T7LOsZ+z}{xD*THiB+^e_$LL=P44;70}3v{@vl(v|N}X zf3YD;c^COx&XKbqa-WZ(Bu0|q&`o4v+qF(En1GnYbulm8?hM zba$bfLZV}rO$=HDEZc23*usHk8OvEb)K6O|U!B_a*5uOyn<+R6o@2md-h{$iU>Xr4905afK3|3h}Ujc|D8T=%k|D@np3K(uPfrM=bvobSr;N?DR~PPj%@P) z*a@AiO2&C!m?ec8igk3$m5xbdzxQlCaD~l$M`v}e$bnHLYDekjF1tdmjFoE&J=8rR z4pvIZQ+p7d<&RCszIS&AKv#o@xfox`rv32+09IOzhn*DRs$>)tgxB6jRn(YU4H^Hrcyt+h-sgWIFj~p`?EiZStfQ zl5b*-6_6hQR*3NCrHelkz*Q+Jw3T#dGRtvCw0M%dftK}Zc%6y z=W9P1>9Sf?>hL7;mLVs1QO00<9@p!j-_m^b$jEc$Z2s7+j=Ut31+}h@yW>~&ux=IV zJF~HBMr_f9DeZm7YQ`r@mR>=oL&u?fA`a7OPReTM&n=)0ac+*BX74PzCt6cn$*I+8 zhM@ccvZY}+z?XdgFW}v5RX4q3rY(PhIO$X0dtWU$RhlQ4;<9~RIGYUrTVqWvBOtjU z#;lGE8XzYVeK1ypL_hKUtjQFIGCC`W9x~l6P-Y@1!keELW(4W$!&n(&h0FU zz8z$kc-O?hn1Bu{36h3Chbl0_jWn?)Ta5uj*Cpa(q{iOuP&psDUH0x8wb<^M#dsvT z!xfU-MQueBvKp)gK!u4H*47$deAt$@`V%j9cwLJ@jB8|U>jnS)gI5+eSi#ep*r=zX zl@U0$!z|PUQt#FzefZ}=2dkt`e0&~*xYq*&SO}StZJTN+4I{_;t$n-U+-rp1A))BM zmpPx)tM|Xs3PmkINJyJZUc>Z%fYp+Rn5+W1_!g7z!pe6Vrw6T0#Gh!G>4GjS`^ISc zGU$I18D(RKOfWEMydRqlu+CQ>93i`z@OdR?l95wR^>eg7sg?UKDy5+eA}(u~V4#7D%5nnK2PM z=gW-{0^!g6x~(GJnkUHx!sf`(CGBHF-nhSUg_Q}|(tod^8Y&8_1B$T{!nIfaIDe42 z1ai;6IVI@2d!)}(mVJ?p75|JZKWTaKq$bjZm`0D(<+W%LrQ5zJd``FW{@y(R)ty7n z5NppFeFxNBy{T_RsStk3;{z zS3z^|2}iwZP>g$UO$1Oy`bJq(5|fPC4NYs-6u8d%0K=mHk5Axur_1ESU&ryJ4}~9l zITfkm>l@JS!9GM~Hmr{f09PnOI{NmzG)=9!J8PC0y<`;DFzde*xL7c|IHC;ypCY`N zc8tcwqG*KyKXV9szQr6iS5l`Y-B=$=;i11+9{AHursm0QQgGoBla1~-dW6tkHHO>G z1q{hFUzZTEYnElFTzX8?&G=vP8UH86^Qw2*?xPMN2{5FJwKCfTS>sya2~s^}6$$Y1 z?J@nKe#I)`ow!@yF~ZP28@#npCSAJn8-6hRCEwz93JT6(~+Mub5>a~D28V5qL z1)2cNUae{!qR4LH5pgTX?b4_tr!n@I)>PVfVorQ7y}mjWWKejAIUZ< zqW*QitDk)d_-!=ETwi9)D1ccJ_RDt|wvCA1ZK&-H)d5MN>My}N0mF^CN`lF_&y~YF zTxU_7Frj42X6SC^Qe*seq}Rq^j@kU3-l^98dQQ}!$;e)3jDu_t@iQdjvNkha2XoHx z9f9#4%0V+NB#@)>=U#1+Wssvs`d)5CGE%mwckG4BrN@3B1|tS}^amxWOgi3|zi$}r ztx2>R?K(EqY#7(oJ)B>m>}BXl0|=vyQ<<&|sX1l+K@JP3&!Q-I|L4G~+53d3s{2QG zD?&LuZBNO0m$^^mkBfe>ckAFb(ZXtf6 zy;Zf?fc8*;r-N-zUESxOW;OV8t@k|TpeYWgZd=4piqqqC*EylHlqJ*MBfskWR;2-B zt)EKk;GrjPI%7Dswv<_;FJHiwbnbsLNbevApeR(Vs6|xnrI62O^(PGzYL1iGfkDBW zFK)ca^P|gmE@v8T?u}5#5>q7!+lDkx1c;=a(JVIai7FrTa=bes!(bxffPSrV%5x!H zpb!-7a6_Xx@oj&GGmit;Bs&OzfZ{%ETlX=8LhYgIuqyhAwnPtRueIV-_y#uDqq0gD zHye1h`mQ{mi50iL%bw^eaT0X7Ug|N`@Zo}?&Z~gEU3uv%+3`3&r|GE(0dIq8L-~?8 zAwV7@>90a;oXjQ_fn+Yf`M0Z7oD(Nz(0n^W2prxjO3TJy*W!aB*l}zkNCF>SQ)io> zpLcpg3z68mT@>GC?;@lFm9c6oz^Ld@yt?pZU!v{)0m?`B*l?fMx}VXDe>%3v9kwaK zH5YRd-3XYaFVZI4>{O7^5Z$MjvJA&|@`*#trW0wEBjpp!zxtTyGnoMG>_>L(sNJzG z=$moxX!%u$3Xp9t*m8HJ*wC$A^mC?gaIM0o*O#9j#0lIg!^`+PJNKp+dvGQ5@l((n zuJ*Wb5%=83ufsWQyvvN1bu}_gS0nWSeR`#DG_cRvDnI$7@2)qS4ZuJoNd4~t7>mA#=b)(T?eF5baZ)Lo1y*WbkP z0(QvBFC`ZyE+0-Zrl6P=l*f%(^e%?7DGfoqL8J_{C+R_H`da&dYo5bt$o!@>YPF#w zqPEZ|oB6i9IE|`uBlLwyq!5~7b5V6H;6z4?wq-HaCO=VoMxov~p|kmHm`es9EI}*X zTG+$5kl~pe@}*h1lN(`vB!KQyFY}&@z8^^6PjQ$u*D=QSLc-w@bl>Hp`ymwvI_{`Y z1SZL5?p_ONwTc(1%uJ1O@6k8>TOB*lMLb{hjV1V|j*u&9$ z3|X@sv2`XfYv)ssLk*rOF>*0z?wYVlYp9S&t6yvUf+y1D=i%tC-ZZByWs!Me&)Zn z?)=8ms}i=WUd@cI+g2+#t#l=({7JcXEExg0FaD}mHE^HQ&neb$`3Mrk6WK+NEG7E0 zuAhw664p^EpC&zLhfDJ`+1oD3b-tiw1RYm@@6RE7+tv?m3)odKdvEx$4At$;dL_!ZO`cFp)UQ zqwt-FbRGzF@MHlK@W(@cL7vZUGZIO;48`z#vNX*ddrB&hK9G1JR68%+$VWPWOl#P+ zn~#Z zAI6^xHr@Y>yzHd|uY60>egHUu#rwM|x6-?hTbAg1ayh_8o)C2Bkw9u`0RkL?CV>HF z&;vFR9TUvTas89s@2Amk!|kUW$o4%Vt7A(X-2VV{;M8-NU#K`V1R$#%X-nOjzd|_R z|HGYobP;N-6758fRHXBFS7j8XhXhoGeYQ*d;?Vl=@gtY$PJ2!}b-o1AI5Y1Rj&p&a zDMzWUGm9&Cf7=jerUiyErV9}QDfc6m=12L)?x8xMcu zN%Vn^E}BXD30w`4bPl(3PnU2^xQqF%$7yN%#*`@l)IwqhXQy|v_?bIcxTd8KMbt3n zk#qCvs6i9Fn=;g4bcP5K->|N`#Oe0nx5qcx$Y~Qa*pMMlSU{sYpVL+5L?0v^*x-w~ z=X8K?n#SfJw9e3m^Erl3P9Jzji{hlXEXX?kbvMK$(BfV4n^fbW70UF>u3Ab#0j+;; zv3f&)DJ7!Tm`9v-!^@}cC>hNm-J$2akM5*w9GjMk)4fwXGW{vJ^C0o|OdlxJg)y3N zNLaZ;$UX-P*!-ky8|GglZQ^5l`f-H%#Y}`ihTCOcx0HA2YV$LVFL?doT&U*@J82Pe z!GV-mZa%jhmB~0**bhjKo{_~4bpIP#S?dRisL<=HEh_BX!x~k8DWBQ3dS{yoCyCr| zOO8Y!=p?%&8p;uN{?cnw4a(8YB5$NkZd>Aa{{eUe{Dr&PbcF>j&opr^FaV_`b*u0j z8A=kb6<-%UuN+dX=RvotN>@$jh6qt<((GC9W~Sz*^jpq|fq)Q+A3tFbWL9mGA=H^b zY=VO_v1rSW(YBplwdF%u?21bUbnmgkBCROfzb3A#u1gY3bvqE=t%tUS$99RUhm*I; z2W(WvCbULnBm2mNar#_oz}C--N28Tq@V4424`1%eF&D2Wp_feMEZd7RZ+?<2nbb%< z(zd?xle$WG+ZX|GII(vR6w#d-$+Uqk;A`I;T6t{P?wbNmiDy1mDIGS}@|?By)etKU z?RR$5klYtfXFlT(IgG5`KUuhNLW2$?&Pc`wS$+>*r&!izNdusIb8CFAyL8o9ysHrP z+q84l!eQ?m8aNt|3Y+IXu?{G=(2@Wyx6${G#kw$t?}^N!b6lCz-mO)mn9#W*o4@Piw(nZ-bt z8y>DuL=&ntLEVWUzkQ{~A`+jfq66VMt9lI-;4Z$^DL1dQEU0m4mOA~L;zjj&5Js!J(E-hzfEV48VR%EZ_C~zyf0)|LF${TS|i`S7z zgfvEjlPT2HFE-b8#u7c_8CwOobe=%(=2LP2+qy3-U0`S`GQ3{tjwX?Mk6;(eN#*pP zZC%YeU;huF{M;Nn#&-0`q%@-*s90YV<4KnBaF>x!0bg*ygUIv=80qTfG@P*FS`9~# z1cc5_m|j#%koM_`xR{>Jf3C?r4U`j!ZhY9p3EElnXZkr{FhEIiNNkZKK61Bqvf<0o zX}BP~Q=I|3@awqNrL@?by>c!hS4OPO0UKd+fpl)->!TD4|AP=w8#%LELz%69YMDBu zDRz`-U^|FrDm#-uI}Ie&PJG*G!*E>|;rj=;OwE32_;@}ZxpkH<;|E2cwaJ_W%HV+| zK}1X)IG|mUw|z2ED`y`2VQw*U<_|E*W>P6K?Y;Sm+E09nGLv;riWf39zyizhn|Xf9 zK5~5?MwvRRWKQk&*`?guaF|6CIX2ffN}MCh3c{v5EF$Qc+-_MU?1PWAaM!_&{~SJk z3g$d&y4MPgCw#((YC1C2o0&A=%|QmOXeQenAkUA< zg*7N@bA3T&hJDgR)PN*sTbvr%{*zBBp>V)R zNHor2Es`+vdQWV!2IazOOG&AoB||C5@0G&Ea>J3DV`UUdGw(i4gwQ?TDLHtUMXpCo zQDd?_H`C-9XpZ7m2*#J8y*QU9A$3PAx#pZ%f(*3TEU85_i|!@z22=&h&TNcbCzSa} zU<-)Y82csj+8|;;>J&{adi5=mV_w1jyVl<%<8ZHN zI&DaZ(VUaSg#74e+UnWo#$QB3Ghn7D(gmlMZDG+GNpHc9Uvmx zQ=L?vss3poKKrbBUTtdWiR+j5;NymG7OLeV^sU(E9Cw3>On;x=Tw)cqyym~YFI=Wq zukOHFEdZ=gYvZ)^kTAUY=^V(7t$`*d!f&K?#RL*7ODE;tZ|tCIC!HP(^5xKoe>)gk zdNKHzYT}-=!a!&zS5+}x$>!3>TPu(ac(<@NLCr-z_(5=eiiaLbI-k>YRpU!5YoAsZ zBZxBstcz!!c_-mQ-_m{EC>A^ORV@qgs>}9)c8{_c74W~Oto&WPUmiu?~{A> z*Y#ym`i0)wBR^cujp;sfKN6zwPVR>l*SCHyzw4yTaYBX75wBvx zq_c`fYwbTkvKlBaL@JsDUY*ZJ$$P;(u!WG`O5;w4XgQj>;CAA%1pF5wDk@h;xsy?< zyt1spTX~<@#QnSu>O=DCmQPz3C@^*Qfq#JebXzK)r%Zc%DuSd-H1a!byQ-6RWW(t9 z?w{H(!8tGiL~#KL8_fBem1Ck;5wJXa;HoL|K=lu2`wI+qG2+32Z&ObmewZlxF#WUB z-=s0)ogZ0=-VW{QUz&eK-Pe=T7TQ!7=q7HCrPuY7nCqZxy1cdx zKEgdiZ+4RUhcc(`4KDrtV;0-&zZ+^_6;K~!)vmTT^dRSaib+S0fH z>CWXW7TaV}#?cYot7>l9lhSE>(le`_^a;%z#4CiVwUUoDv005D-gb~LSRT6PC|a;g z1W8gE7h5ElE-}T{ZOyei(t_OUJinBe)PF0-wT9M2h&}dmTl-++3w7}2D{VD zm{TjvlhUP&OS<)(wBC?tl$TDq%;h&c=dXMCrgtN<;07sWgNw+qKR%74vH9nQA5>OG z4fR`Ve8B^J0!z>kM&m2iy&NzDU;JvVNFmt=Gh0E)jL^>sbFoABfS0nH9%bg_*LWdY z@N7d=03fo0b4~MYV?}}?kM= z?mUkSdU0BwRIaRA0R1hObJ;#w^$?&xsBSmuE?5`{pclxca>|Y=*PC zrS~4}NgGug?kD{a!Shs|eVjfS zKF@NJ4U(-XWQA)^W=Npz6o+31y0cNaqJesnj{|5qRu#Ke;!>+pRGK=%$DVOiwjlZ`P~S^O@*wk} zRZ`OAuFQtb3QxL@>Y!jlLiK&(%&}9Fz0oApbrCV*WXFYfX-IQp)EJhn4V9m^nk8KTKT11x3Xp`4)(BEx)Ypy`e34BDdue=>k$NvR8w9d>f7*UaD;Fq zI&J1A9sQ*(RaFh??BcGXt159c%mzZkZ02UeJ9*fcXZVS=;O~x07IF8&>Ih+!vkb%6 zb{N*eJ%rAS-_|$sG769{92U(RsOA!*gS=|*5>4AbKWNyiQ#5)m#n;264;&FR;`rCv z6ff{M&3RV04aqEDLn2S{z29A0GoLdtj8 z-;Rxs(R-U!II<~tx=%^ab#7qa&svbnH{HIS^7(A!X%8==f+06K;jj48{Uu#)GD2IE zc7px~YOr^pbnMRC+`Ye5~#X<%YOc4@==$P=#+5-G`cBrU%l|;3S zSPhJ>&9QtQw6|pa8+k_lSM182Tm$+yde8J`Q@VvI#dB54S$3bKeySJj0`!L_nGR^z znH%oC;=UmmFTR<@+9s)T&doPm_=TZXleyu7ik{)9aaj^oldn1xJO}}e2{cBp%g^T% z>j@U0Ro{~*rQ8RXcA+_1z|)DLuq&dXLwJ_6z8nwpuemtGL`^$s>V#G=v@`CvC5U7c zxjW!5v@r3oie-rIxUp`%#sxUmrtdzej^r;B05n6CQJVI9`mC<$0?E`qQqIuK<{6M5oatlGm zxxbS7Wp;nl4(@CCvAtUM#T9wVP zFUu*BBh<7jrsvK1#H!dNZq>XIpKqb@4z&?XUj_Kjc;Y?&0lI30f}|3> z3U+IgpEEmwNBeJ9COut3H30sL^T8qtoUYqLGx(a-{G%|FQhOYk?%)stwIDOs8$b3@ z-3I5M4?<_E6iDUvNTxz5;3vy){ru5tfVW+)u$vjhPuZ43%m-dp?pT-Y?HU9aRw1)G zm*&pvt2#mk3cie%6~)a9n<07dp)eQ3p*}^W=LYO+B(HvKLcitBBo#Npcs=}rKXb@D_V^q!>J<)R{9SxHm2H#*nGIu0s8knb zt=BUg!Xg9L?RoVvQZ%kZHNO&re*3kYH%r&BMu|42>24UtO~N%ZTb}0-ZsuHcE+Azh zVEL3P;W_jL>O4&h7B#Zh(7Dk%q3yEYHiV0z;32xA;2QIXwIY%7o{1l*7HsXQl5F>r zA3((4BIGSjPbcU~$T*EE>WZ|uSMbd{;Rs&*R9sGERe_k2sx&yr$8y*}k}1h-$I4ZE;dO0ZT6*u|Ev&Qg}!%eQaPKfM1^t}9Q4K< z*{FtU&`zs){jJPPYf}drM~mlI8=Qjx0QkxC#6>=j^k!6_eTWk?Sf_L&K@#dIXA7JoJmb1!GOWDs#AWa$Db)xn)!{+kDm8| zn)2Yy*;YnLKP8qpIKLr=*$hv)NlAb_v|=QShgf4^(kB@Ia(~Tn=<1^C$6?Y>63M8| zG4-ionb3=)fR<}5tCWEY2YZhGLAcCG8c<%TXM{nsa-1RgrnUQHv za&sFm?I|HyXo`>qj0b%?_wMJ)PO|oMhAn2Bj%Yp>%U6>Lo|)Y^P>NH8XmDH=d+J5T z@4b28Zc`e=AQzQuMFebv*k$}3WI-@fNRcO`^3Vf4y{gc^Imcw9I9IOkNl=l45Rde! zqEKdIxl##|>g`Z8>xXj-$-)at_6|k(q!;djQ&mpuVt`{qdz!8LDBe$#O_r1|b&>zE zOkT4Cd=Zf{WU3=f7`6Fs0_5i#VH0o8`8^j@5z&Y{6}&jRsd}H&4$gzg0hwvQ-YpSD zq&YSWR#uvyYt`n8d!xFD?)6?CN%~HTbcOOUHV@i@VsjXT8O00bB~%Ok0fO=slTp24 z&q3(IDFl7pqBhCyh1nA|6?Otl!|lGu#{+R678-OEm~)`66v^=v#BZmcOUM9&eOcGaSx~=(~;pF@m8Mb*J@kvSsm;sN6i! zdxZa>{hXMG=QQwaY$}MS9lU~{831|u1lBBNkqG(K4_Qh0bDpyDJf`%aUh(8%ircmR zdRBYg{2*~5o`=~AQ3Hpir8+J5d|wauM$d?<9~ zzGRRfqK56c<3Q=`jeCXzB|qTp&JCwHMNFHR^izL0eLl?ewA6?iFv??8a?Q=7GlR^l~ zfwbv$L{>6iS;EZvxm4}7!~+ES+!8Chfm!JWi?8k5QKJjECh$MN&>BKK5No#C3SVRV za!7ofmQ1?Q{14DW6f?g0JPfIFBkir?AJ8aT@?VF!LA8cfn*Ix{!-(xuVv1tpwk=EY zvhDP{B>bRxyPRok?0fBlku>_q%BVRoEL(Iy-7kdgZKc?fojRStGc`fO?VAXU4NY=z zKJn&+db%7gS!qMQ@NT_q9~OL+2ov|9I3l}bhRKqO6M4q&CN>UIw-1cg^uQvv z&GhZnQOl3omG&{9ewn^ydlzGBtTd|rm}X(^p)i~fEJ85P9h_WHY|sy*C~DkwS#o|f zd`aJz4>W)KF9NJU2(PSo4-{i-XtE5dd_{pFP7k;Z2vj?Ieog0^JU;l(${a96Gy3GJMAo_&46iCvn~QzJXx4lh^_JXkOaOovJ^o}~ld z4T-}fF&hRZQj?W%ATYn|>(WBJKqv21$mo*-s%`q|Yv9CG9Lf={OwDL_75fdImM;2H#J{}6c-Z6PDF{Y)HYoU7xCV@_r^eH zDmHyZLaE?<4FDe6+#vPGUA9Y;4HaqNmW?C+|FZX>t9#br+DMU3*)_bW_yM1v943f zy45Vk-xRa$@a^P`Py*|7;N^q;!6f`w7To#>(4eVw97N(1uaZuefp(F`2068R+WR^x zgHt?Y8U2&JpT&V7b{_J?vXd;HKbi7;T_lb*G84C z1BH2UQqqnLPT3YoPjObYqzM4v<4^7ZelhD&lOk=swjGLIAHP-b; zQ2)cfU@!M4LE0Z1Qz4ZP{QXF1H&8 z)EM;~3$KWBbte1hk|e8QtPRp2%qCb*!i%(W&|dk1V58`kCLm&ulX!}samEG6zFkb= z&b(%90bWR(QaisnK?j`$dJH+BgWk}}AezF_>640Gm6%;xiiH$=M`27;qvGr2?k(96 z0XvY+?<+RbTj7Y2J|Ufo5fqv5UeycpeV`8{DOqzUBj?e>jZ30nSkCvGo9hwlhOv!P55qI5 zim^yZCJ{ToIUER-;U*q~Y7|%Zp=S&9K+{7AX2p}Yg0c&}a3;iGf_L6I{S4l}6kIwC z6|vXYusi5vUz5J)58hrX#7!@a>M~1FaoL>Bql0er7Nnvk!sxAh4F$F%u!#NPo1V)DRFYC;Co}AAxdJ*MpCaUb<62>F|U;NSbleiERrv{lp_OJ*3`k9cIlV!%Uy+vrh60>?< zvo{ar%5l4*HUL0wX8-Cfes?^BGd~&rnXOG?u3}D>rAd34y)-EF#-ZHT`h!n@KG$kT zM^j(i1_Yg894zB|kK2P&Aj(>J}u)|xvrjt z=Yf$+H4?2!=?3$3IMFG$GuCs?4Cc)i}rcZ z9x+rvFC17-B1PCa%Jm)UMo7P~MdJnQe#)$8&S_qt!xcGXn`vC2cY{G+GNi=7pYJ=A zUj`3LFn(4ZCm7hi7t_ldq$mLm!Z}fgEfVWWCr8OaLWsq9R8-C zsL;CvU}M`S^5K$$*tkht?C6=A*pL)quX)DLXp#5K8N8Fjl&44WN!aq6_0%0@;a$_c zoaCVcyX~e8qc@0yLJ0x##^43#zcG_%T6%*tG{)>zuKdk4#7;C7*`?!w_rW=j<|6MLxV`jr7Ho zYz;U6v~{h-p%9D3lwhC8P2CDXuu?~Kc`XTKJ808i*_)ZMlhsre7f6uC^tm4-N7dzD z`jU8v<(ZcX?)QuF;!!Ah)M(WX!s(MJ&W%+xkI~v?;F(?Iq;7&TH6x64UP5!}A~eCj z8*(e+y8KGXn*9SE{hT`a-zleP=cPvV}TE2r_XyJnR3uNak;QN|*`wbu`K&m?&aL#T)ktY;@1Wa18~# z3DcbT;w1w2kh#Nsf}m)Vv!uq(LfrUrf}a*gtxL$sGG!X)r9FRk!A5oF`@_VW^;`Gr zeqK4qG^C}P-MWPMV?TPq_VbzLXx(rBFDth1SdLV7>HRVL62GG(P!NsOi{`@1l{(AD zn5IHajF<61LY=nI$-|_Z_1m-$k)&IEfS3Rb8B;!O0qT;$yH*W#G~@qQ-Hi+NP4s#BM7TTP!o%g$=xMQL7{w{GjJSOh2FwAh z>&iI$2VD9rFC7OrRP_2sm*ozJP$+Y9sG02V+T1I|%Z)6}M}xS?pm#XUQ?@5Ga&S*Q zQPQCK;K+AnsZLz3Zuf-Ak0*luc&q8u#;|GY^s_dbuds2CoKaJ?=hr#gF~agBn}rTU z9EtFCZ8c;51|qM!<+J^!ao0^SSC6r^4a_(dJxhUip*T3nZqt+N&fIezqIXLC?T3p4 zS8t(2o%p}1lTFHlvI7&hjWLQy}{ss&ZG1N_@_HA6Vx7bqpvBG7!E z%Zyilm~x_l-5U=4vX<`YW>mt$Bo9AlZp-ctoII(=VFfNTeA3`OGfHV4kzt9)Acp9e z-iXpAivJOrliKc(b#Um-;9gq`^HR_#05{*09miex=n^hESW$NF?;5{1eJQn>53n7T zoLINlkzcMF<0*gwBo;Tg@uwYQti)kFHYyY72A#Y`41q`_A$Z0!=|(}y$gHoB%LEC@X%;AChgn^GR+ zN!j_72uaFljUPnX@knc8t#Hbe@5GYy?n}Zo10`2W&5D5)VCLqiuKZrh3p5j-wHa)Z9WzxTJ zy#n8y70O5_!$(yjnOt#S>#asz9i*1x7j61h)G`I0msDZKW8{Y%Yx1V*A2MC6Jm&IO zwRq={tdt_1S^L0PY`uNX8qxEBBwdwvnJ@NZ){QsOZ+<_#@b45kQ~BkvHP5)UXP#_3 zS7R8Hvwy9bCcxLe@;=Z;=d!&oJ>sm~@U@vZDpd^t7lO)xB*xd-ZAb~Et7fIb#Zi{Q zj(>os-qs4>n60dLOe+bE=V12=8_aT9YnjBnqvkG2z>!_|7$w!@pBAxIJ~L*OnR}IV zK`9d?d)-@)UUlr)b_ZV`vd-qQ~yEMyfr;V$?b7* z(R9LU4aJsRY7)}uSZVE_B-59?cixN15WLQ7uf_NJsCmEpI(uD8vcUIXIfCD~LrFfW z`8zboaa1DZeZN0IY_ai6}x~631JW`d+A@`ZlWoc!T6K%lzI>m^dt!>5eT53y&_twK{0W2?mAUw#+4+=t;w>UVL=W0lKQ_pEbhMqQL znmM^uD})M2XliG0k8W!mS)6z<1%%qa9ckC zUs1~&7G-4KSniV@hFj&o8vAZ9Lat)-(|6>*l((jdM#kzdR(fcN#kMY#BQrf(W38EB ztov)CW=r7rOXE2`Aitusf!AUVBb*Li*)Tv{-kiS(OzE7b3RH$^IAY<1r2%4IS5?!n zMLINoR_*#p)CF9bg}?E|F3*gvL#xTVre|KY1Ww6EsDHdTKduHj+w>8TtpQgV`ky4W zkijS;AzawyC2lf_$^nvm{c#V9W1|E_p0(j6Bt5hcPmXWW+do1*^nfUc>V0jdeA7Dd z)>94Ff0Jyf_whjFmKXozp15ilh`o1e!|uzyZt>QctljI)*9VYvk>t2AWIMZ1Z2B1? zXd_)}G5E&`U_7j42=mFXM>&OWbI76X{f zV)+viqS(Tub4El5sG3y1u8&?~Y=v&^z!^#5M=h#m<<>Rzk~*JR(K+TxAUBH?{GeWe z(YG=0xmx!vF=oA%1X8u%C}O&%d-+J<-QwSW0LzO-7FqKNdqKFmbX&CHICSYLsq3oR z{fIAHF_-FU1*N}4WNO9z&jtJt@6fOMsS$o=ljX1vI11f=j1!B`rgUHcZrHBObLA4V z=bHont_`tl*m_w(L6wdvVUYHufM3UpNLhIK^WN_nD{T=wR+ zIciys<2L>Wv1AFZql+6%&la=TU#E>e_NT_23py9%Kni%)H1VpQ#$|uWNKF1>v9euB zNtyoY0yepO$3E92EnYsxR<2Ao_lfgtx=P`X;CiQnOIM`gBbCKBb(jc)7cQLYyRPx3 zojJe%Z0>jdM_>z!*t?;}^vF8-;B%|(2q)j@#ksTM1^a)m29QZ7sFdarMY3w53UsZy zN2)Q;snK6DCi|;fciT#O87*(d=6oq5xxlcw-slq0m__wi>8)s=CP*>xa zB)#52>d)&aJ?gdM%&jR`>~aTeBjakVnl!){rX7t;>KG}N5 z|3TANhc)@WZ$H}@Y?Rv=EjUJv?zFZsYSidPQt6TqL>(|X99@c}w1jjhASf*@A|NOr zs3T6=XHi#_`u$O^4bz{vkyY6Ky@_+!|NFTG3C6+oqd^3 zLQj|vN+nXRL@;s2nRcz+Z+kh&gSBNP)R8XQkL+?080I`Uka0M$8S= z6WBXh>TZVpw(x{OF$l_A>t46`(%J>hXp%VkUt1&iudVT~jAP7~xsD$dl9!8ayN}+a z%7inBM;68t5^nBMHPd&WU;8gVIIqV7mSnPqt#!Z*e(`~s+~wgi3Fc1VBiN7rG?YbR zc_6iOpp$21HK;(b!z;%4ato)k!`6^K7;wTqMLNig%_fuea$~L1p*Su0k9i!ChWBKqbn-ygQ)3QgN6(d4Xal z^-4A$sbbnBZn~w)@fQgD-9&KnlH4PTC;(b6lXz`dW{2to7r*%_x4vn5KRkx`ys7uC!!hLsS<`gK9?n75>sk+dp!3DllNo;>1 zY7k&#Tyd00T8fH2gE)W|E0|xkU|O#d$o|V@6!XHkmjJ_FBUjNVH^q9%=D?66>n5u+ zGV`F-mIVi<^3-M!@8Q#P$-jusnrlrz?HG%~gViqX=(lUubIlP>51uwDTio`%@Za{0 zZP53wOh&~8CU|XD`N$0O2~YlaWV2#0c3$nYzf#-3YKy%0`CFYZ*uVU8ce=ku6C?5u zxq8)}N=VC=5Rbjcp6g@R%)`6^O*_`ue2&u{^TEZes@B8^j*CsFN&4j}mG+!ej?Eqf zf#dQ#m~gQ4cqS?=S@k+F=nMx%-1p{(l#Y6aH5{g6oe1(4y7HR%cox_f_WtTT+Q=H~naaTFK;H(# zIQt6=n0~`w!LR4+c784i`Fe^-rNvE94ha2C#~5P?z?0M#$T}=8vJg3TdIWA^M?(|g zqv`c0>Jn?>;na8xA2-6=VBahWN_IjBJrB!M^At0Ju6|gTtB`SVQ@;8Si>{`E#;vU6tB zvaPj?5hajX2(iY2t*Fvy?dyI*IY%s<{Hi4$3q{E zqP~7H^uq`ml}%g45L8=@>1|#JHo&+LA?dq?JZ|>w3z*aB4qA+Y42KDdgO&h$`exbq zc=J{BiINroG~_3Wb(471o)T`f-OJ3!U<{E>zG9Hl_kiX2%&ldM`JJYny_CXV|1_ce zp>_$`kc0;b8z3XpyDm8}Yu?N?Tk>cZ$}H9M)_;cUdJtx{d(*J^!6! zr%wkABzP5q{aWRjH%XA?^~PFOz1VANxuaqC?`rOVz#Z1j6w(&HfEup+)k$MSe9ToAkFT_(?=2IVSI)mO#cv6j z0q=yS^&LXsQx1Fqtz}ZO5Z)<;5JWHY4%FWw+RY|xrGxHbQ=`?_52X&&K0d--8<bs(_s3v>4AM9ZYldT&W{d!8Puo9xa6Nf%+@ip(Xm??b)BK$&n%Y#zNtX(G66!>ic!mMIUkN|-iU!P{og&} z5`WUL6jbdi^b|-kzy<*U@?GF#1ytUj`MT5Qn^PzyOSXAeLte>` z*6B#rEctO>cct*qK&78!uWr0M%`;w5npc?E~rl)Eu|LAOa*{ZsgK&a;?P!Yx)ZyA z4J?uO5y3G)nCSG1pc3xxW>V={&UL1Y2<_&TDycmbyTAqd59pqXBb2mH06L?fZH!Ls+MN zK!hl?mFKzNUBG8Fccz?EFcAp4+?F~SqE&kYLU661BG>$v+20umE)Pc*C^Ub(`7!tT z*Fi1b4y<+FQW=E-b;A_?oYD=5rN0j_McFziqy9*%J70>Nbr<$}$x+PgG`+B4X!U+% zX!QY$?`|%Jr+&;wRC;;}TY2k7SRwzEZZ2;X;ycvH_^tZ(Kut~#0X5=!cF~@no095p z>1F;XVIj|Nc@yCSMl#@=4zBxNXHxgZ)+DHbHcSS~Bu`WOg>m=JM}AV)dHH!XXI6!r zRFrl>o8$zu#C?q31Lj#=u&=*QuArA#zhFV19sYCeO`g7QqtWHwR^?`s9egkjPF5Z* zmt5w{b@yM#*nB{?Z?oinWky}9=3efz>_&Whlce%u|GHn0(j(qEG5T5!>ba==r^z-n zt5YJA=vP;b8tlr)U4Q6SaK~Pc-+0_Du516%73x2u~BE)gLw8mV1pW`G`%1hf~)@C~{y$ zev1(%<#^I>P|l+eQl%EVF|SXmEFTFt^`uT*%-DFAcwcF7_S{dqVGv4R)G>!$`RE4- zbGZuj2*gi&uUUFU@0PCb2(7+F8Vc#LLrTj+S(a-uCxSN}zS8UhG8p<)Tyx9Zu!9a3 zW~mDNs&Z8p>@rzy2pJ~tTw`mXid%swjUNN;8crf4@%|e8H&I*xvK+M*yN}>0y$wse z=*K46L)tV~Hiix(`K*eiq|Io~1%o#tU~wZ?wrPPM!We{>UKg8yNGDZ5e$CfE)ahq2 z&>h|95WkQUDd0b}f98Ais;%2ow@(Zv_v+xb%VNEC-f}D+)l0e6@b9jb>VNy3T;Jso zx7E<0XZqnUKw7yH!?`BWyOWx+{B50}U7~hEb7P|%@3sIiS;SFrZ#w1JccJg@PK1oF zf7h~ok`f;KoGBDjAhmTYURnBj;S@sbOh~8&Gz-b4cEYSv4oeC5FSY03wymCIINc}X zkxZrlx2t^nyA_QxE2ci~kdPO0ftg1D9@J`w2_@08U)L5O49Ic+b;}-uNp34*FC}K2 zAE`VakD$g(j%gqlKQ+{(`5$o$7;Bb!d?Spp-xE`9yyx-foW^M2X8v~`-_a_6j3yH7 zBd>5Q!gNhxI^d9%F{fI$v>Up`L=c5{-lE?HX_{TRrwcu$dHVovEKwP9f9BSrp@;&L zn-U0F5k<%r!6$&2s8Z&xUQEP%*sy9VaWeh9>|Ai`JdPLT2U~Y3UT&(aqTPB`j z3UijaU~dGjudhs22FTqwX6JATAvh;TH?Qae<#MUrUfn@kzG3S&su2J%izbv56$Oh1 z56yY&13Zof&a8z7N3wOd0uUA-23hEq8pV|SE-@XZg-kRVOn%R1?CZY6Pg(*IZyf;reohPDsNpHWl#_# z*J{f02r`lCDPxfi2L|Qm3a5{L1QoxSx|Dcg*`ItQi%cH#@mE%`yYc3m0Hs{VH%$U@ z+v3dQoyW{HStQASKF=EaT>R(Z(N2hdcTr|7+L!`c4(qI5FX@o@aqBODHemQx6@R8K z{(1hp`Vz~nl3!OXL_gTx|AdBqbP};@0qSR6-d+A|iwto|g**`m#yC(IzrD(ViH?fU zfnHg(SZ&({vl{6Jqz&L`j=Xj9#0-naQo0qP+>HXqYEexZ0%7}N)vpM!f0!*>c+ZDAo z3T-ZQ<*JlDa}$o0hU;$kQa?VWvV=>8KxdBO_T%Fz<1V#<3Hy{p!+kL*M@e zew$zv5}XaJujDd_(C%=9NZ*h0y!O0T?myL_Iw_`LrNMsU_MA@dA8%fwNJbOIj4;VX zr36jEl{d6rY*WG6U!^`aJlbX)Iw{}qDHbc>F9z>Lq$E{`m7u`4LGpKF^et4fI6N?q zkNXiVfo7J&be9r(mx5IqgwCQWZalG36jKMbva0~rrc~#rUpm+OC`zic=Z*;oRR9LQ zc^eHQ=Px{lb>-=&B490D@5KHaq6Y7=%5S@NSyNjwyvlic-RU6L=3t`0S@!573HsFN)XNOU zJTd!vOru6lW*B#R@Aq$xPTu2XDM_yvfjle$=ijy-E%*K5K0JK|eKmWWUPdy`fDNdN zP3YauDs#D;HLW_uc7DQF{ZJ^OB|To8&Pun#Uwx&+t${G=l2$6qw$ae@O5r!iSdfBG z_f%lue3%9-Zsd~XrQCe~PR`i+Ug&D(8-*)o z8ffzpE$hoUCK9$hh4gbT`&RGCdbuW8KQj0lBft)gEjbI}2@6>u%6EhoXp5%mF&VH) zVtvH|<^U;(ri#c4bm8R`?$^9t-1zr=uwfu2tigh_F)24{vGUgRc17uu_Gy;Cz~yGS zJe^f^F2ZmTo}7h8@PPqEmh=R%loXUVSA8cR@g~8C`;9sPsmP9N5^+hCPi+YI`U?oz z4HhC0YN|XG5RNKC#%-Z3IlQRW<(Zb~Z7wY(EPI-4|MCnrj?$KVHyWy11;)KE!>qZ|Ntv=sz4hfI8X zaqtKiHl3!WZqX_PMV~EZK2-*p%=o`Fyt-YQlB{H{{Od`n_gZ&U@1A+9-UwpiI+*CV z@3(Q{!@XZG|I7LLiSf)npSFOgDcUIp4Ew8ZzT6AAeSBq)l3wlXKmy5>>kF=$oT%G( z|LReS8Y;dY`7@Zdmg%Gp(K+kQ?vYvC&T>#St3`a}&gaSnq3@!E#7n~cDUTVBnm1)& z$*>H+_w^S=8PNo^AsfglHJjTL%_SG_t#V#$EY)1h>3)O3gu!RUt%@GQbt31m^33NN z@VVrcSN}O$u%&8+yJuyklnoWDo&vQ=R^(o+yF5Jn#>7ut?ghFK+{V3oZ$$7pvP!ki`j~mdN0$S4rvh&!+Yg z0jV`9Q&G8$k+gmJVL95cXfTO_BOx_A#QgGTfdIo$O*!Twt5>^Z!vMRqi5<`Keq{af z+b*ll#=(`BUrv23`KG>RsD;gs@i`w%F$`A~S(ncRG3<2|g=Hg1S8c#U(p2Ae;fRfU za?Q-1fk>kl73pS%M`*C1&aZF}ur=lB@Ki2Fz9N`5AJ-m>$cE?g1*X-254{7lQB1We zjoa2@G-DzF9ux)ip|zqCLtyeFR@_Fu}`dF#)e7w~5glPIP(@$J{t-bSel{+7^}o^&te z?+BY8g|~EMi5yv~4G%AfZY$absd!7jcmGdD%y_NG4OV;hj5(ac9UhjEYM~q*FqEXi z!}-I)MLfzm>PyoOf7qF~hQn%&Kg@eq4J6J*=uq`c)u_2vSibC6B;%B3=B)x^(QKMl zvyIS;{+Hr!;s(>;94<5_#T@(vRs%`zjP*SHl-JJY+bI^1kBNHxh`$d?)+sov7UGqUdOwTf*Z`mu4POdseU!{w-7~yYrtTe%O(*T7l*OH^-0l2QR1JB56~Mxt z3;B%YiaxV8yk~xVrM6qtTNTnv-E8{xD;EvLiuI@$kZ+XxtFF3rm4y3Uqr4lnB4m|^ z;gC$Es>YUND_Shs&X~ul?!EL25hmtxCRIq{ktHmciH1p*1bsg=Fi5PBGCi!#mz$#B z!xzoG`iHacsDID?sXI`aqC$$oJX%7&*5>zF!~&-z6;cgjhjk{tLeRviS#UQoorZ91 zWj-d>p_1wO;KPW&z$v@I+@*qi5;CPDH9eU_^IiLPesZ_96S|lK6PUqgdRa?UFZ#+t zRDoo;%nG=(Y0fR$uwPr(Mt)ZRwiE+q0S)dc-{DAVj&yaFCb(GNFZcH5?yd~L!{_*g zU%^)^{{mxQ5?v};ik_tm4?~}^k>UPgG@XWYSX>~>^s!q`!svL&ZNCu5x9gUMV;CuY zSDyIokfUR#p}eE@?zuJ8&Pwl6Bg`%A*Qs&O$TxLO z$3ZgT6f_|^4Zhc2d`qxPrPvta&~7R^=qf~qiX^bAyiT*=u@`Hzqe}&B2-qZ_zHj8K zuU7eXT*$;{Z{1upqa;xTiBH#b6)DftU^*#j&#|aNnp1xro)q)inA-bqBHpy!U%d9v zEk>Tkr9m!bYtCdjGw>@kFbnq`|oAzTmkdrWxZIDyDM-ZyuImp z?%qsENwF6qRThGcm-B?lRWlK)nsB^KtD}fv1(%()M2FCRw2DD^HNUVi#IdH@RTz2U zEdz+rBAXfY9+w1uCrmC`TQf`{8-~3#HPg%eo?XP{T~n8estI3r1^`YugxuU$ z<&2#mnJ6P2m}BZ+;HfDq)Q+veG?NAkF! zuKyB4(kDX@2*wxHEwk zPj*7Mmo0*0#9O#A^W7i^@0gV6Uw}0`EIYO=1zC-AHgj+$hedc841%)dWY@f;_;%dB zT<3gCa;(2&H*Ea8%$|v;441)S+~+K0#WJdVl@W@uSP9Z{RYIj*NEO38R9D)WY?njS z4>_Al7MXiMKc!mRsKs5D7UC&4shGUvXUZ|6GVj*%oDr^&oFqzVn|k%*!gmg~lv#Mi zZ=oBWlb!Cd)ERDThJ#ix8u2sQF2$2aheGWzT@e5vEWYd3bb%E@fS9u8juOV*1$1Qg z^yX+_<<_!ngIfpy-1_3w{w|{0qVZlk>!6V#G;wN{;zUq?;%k+1GC9g!Qk6}8a!}EW9PQ|!vmJ@j`x~XyTvQJA5F(FP~MC2qWB2TReQ~H zQp$3T(2G0ol*K;uCYIf_NdS6ZyPJv-3yko(-LXnWv{VT@6f9x1ruPbxewfSSxE@-4UxU za=Kb?dE#cMS!-FaWxScWO<8QAlP&v@*XSFyXqQ3Y1cy%7G2^ImWX;a(q ztBgI%oE*8$URfp{$G^bMMppwUg{`65BTPR$Q6)roOMZ{FKZo1iV%aL!+S)-<)7zbF zQU%A!9hLtC*=-OB@H^86K8{QQ+s6;Int!ndXZQB=ND63Bz_2rpt_;#wY%$y*n$VNBTGvJi^Q~bd3K}Be6s;4q15M>=$B~07YcJ z)-wUysEa(fng55C{NA+}lf2!LS~4LMq=Bxn2J<2?(8MJV*y-Fr7&)V#@>P7_oKO9w z;oZl(_pgyeY;Q#{61=L?Kzl#aVP(^W1(>?jSsq zHRE}DM=!7%=W=e)5ImOa#uTpk)DJJk{Dt(qm>xupV>&zHn}ZB~+_D8NyBtz$JJkFx zV@v#YY&XMc57togr-d_()pKMN66p;LEY{aqTU+~S;Y8B*s=hQn`MS@PuFfoxa9X8k zs0yjsSW6on|DH6{zUrgO#&1VQTeueGS}-xy^XB`r`SQ7Sql7CMe!oa2vTFnA>NK8k z4tye@3Elg7$mUmHu(I-;<@zpL=+F7C$8V$_kun#X_#10)G0U92lN|4xIBHTkUoc1hV9|az<7}<|MAsm$=U9+EzB`y<&*U7ppu60){aXO! z+7VP&(rpg=V9}o(!E%bx1w3G*B4~!F1Tj$y`KLKoGSF03-zSyHrzG;@zn>YCDV$|- z0P&3EXGtL(f-K`HZTcdEBZAx^g!A9=pnRb~_#IzdYqWQnr3_fP8cVy_qDZL-Nh;LK zW#5{gF_86d*Y#B8XA2U8JO&msD)#+#4|XflMm718#d=Yrqx0=I#xR9UG63*?5b}s3 zZh(aXkye~dGL9CnPs@Dv1ljph$R4M0OoJ7K8*~O7@>{MIMFRvngb5Zta#@pGnEb|- z6-Mt&W&0QE2ObTqWv@EA%UAP7d)5l>~m2BS1aPXnig0xF`)jlkk9Evt9&oD3%_KoxToQR zWZby*DiE6TN?1RuhLyPa+nswP#OBPO}>EZN3hcy?#VFvUwp zN5}T9aIVHk?iN3*MVl)b)FJY@!PvUamWnOowI!d3GmcwOBw4KFiAIyOvkbQvR*`-<^1v#9 z7LH?2KZFKEw0mpM7TL5w9;9B%n0(XUS5Jk^(@#=+PWv+qQ0%JF`0FOdk77awabZ>^5GsVs3*|1ex5&88D znN5^7879b7b4$L`4vjXbl1tBy%Gzbam4syE*Y_-;t{`0ld zvzZeI-cf1&DpMDl-6}Um+{HzwWBwGCTv=w2Z9G($sPF!NPeqbwXihD&rY=?AZVLRx z&LG@)_LW@8Utw!%EpB&w>0nKs>>Dqzye7r>gKgRBf=zPk%H_>JZ->uj7k%JuN?S?I zLd$LBIm#^OT{*QAE32ZfXFAwx^R}$ql?TN;F7yOp2XJOEB z=nxxV^$5(_^y;|V(Vi$)aE2vXZ)nWm_FL3IuA@%{-`tRS2u-q)EwDQfgonXSh;US9 z8cXTR4oD>Pf$sNv#y#N`$JisYRE(9F$VRcU_33poM7gj%SPUPC2jXp;>_n7o*6c(g ziIY6nRdmD(qav(C6}oDq;jjC;!)fqs zxfCoLg^hY--|TgrLzjE~Xr>#Riu5GF6;6V5N)tu|`CxjBVUb zd!FXAZX9Hu$~65APRECP7KP$Qt}dicV?Z)-H$!#~KWMI#hl4c@j&>k$Ne4;XzJ0}u zTTE5v`=pjg+|SIF>q!i^yQzYWl-#rXA{ke5x3B39;rhK~?+}D9kvD8Um*44BBGL5_ zG4*;Y(*{;S^wbj5)_M8+p|ijguBMjST^zUn}rJo|B^$e;?-afO|DtK zWk#gp(A9^pTHRqEj!s>w`Ih>aDG_rwBMWE3o0Lx>vKCUxfW}sKbH3V{>K)3xL6z;!pVk!?{Fa1G%g0w02CPKRl!t(u!mCQ%<(WH>F0n6G+sTw|J#3DN{B7h= zRhTHMP{Zw11g+T8KYKS&@cRknyG!}!ngV8Lw?16|l%qeNtrHrtuTl|Bgebdc2=D&| zbVc4wbbQw37T*L*aC|)w40g2>%bT^=8UQ3J-fiT3t>vE-bM>#)BMC5rNTN+na@##N zLPa(pr)+eVWuu(<{V1It1>0vbBf4YdfrVwlk{>b7XYIoqb+8BHDGB*<9G5sAlZ0=I z*QuN?v1oYl2GNzb0t&PH3%nLFSSC}G?S$aS+mah2L+uiQS4-Vi5F@1qUL1t$d&@b7 zOLBQm%Rg?H@gjRnOE530z89bCUG~Os`g|I0DUnZSYT=_*LSdG>Hf&j-UO9G{`o$Bxc^umen#&P$%F z90(kcA(kHZ^vxe!y6c0aB_{IAE@eM)T1roRbMvqy%A`m2ncaV{7^i7iwH%5+nOtQV zDsrxu!kk^m>k`MLAR0`<5l@gcyeP|p7+B0W>7Vi<1if^qKcxm&wsLX&gPHN7wdHvPt!}+ZS zzH|K3EMX8<5ejbhSJ@DIlBg}Bw;(n3!kCpvo2p~Jw8;oW&lu)dSv@xX4Y}-%=PpVX zAR^f~JynN!h8_U_IWe3Dkb!(+x1C!N&olD4G1iU7AJLAhhUc6-XRW!K5?7baUBRV( zNYA%EZHq160|yoI2%bKy{@!sK+CA=h1wApQ&3C_isLfo^_YNvhQ6o_tIyAhVsbr~tIDsGEtKt_$+L`6krwly5A)a$fVMgQu<8#){2+=rb1` z;N0)T=tfa>IjoM(L4u(lSobMdT`ciFYiybwo3&5|nHh}gRd&nu1YtbMSYNpOJ99gX zZx!rJ`j)?P!4)nE!mbD_+c*kyGQt397|P3U$JE61%}Ni}Ezb%Ugu1n)+WE4y>@BVb zdnTU~7(*2#DrB8NgL;fgR67L2F!MuG;p`Q2ukc5OqQB2ORYox^BCG`LE)h?S*;=0h z4O`VZR$pzLMF4m{Y3m<-ZMkcSt_wjxbF44jK|W8^v%DVn@W zm_M_(Oor!q3rP+Du9+TkXTz%$R+_X%zhZ31!I$-4Ja z#oy-!P3-4GLciLel9ZGAH}Ov$zzE72gNqjIN|>s}y3>3MBfS}{6cFOx6x7}PHxH(><$Nx1thjM37-nI}~&OjX- z!u9oBorg{p{F3wF$b^Jr2fq+^rSwd9?G841{Q&?p)?(3h2ro}h)$l8&pK6W6FvuV* z$tCY8SZhbXk8<_26m09onTLt@8A}BOi-rNE_rh5&+eS@amZF;?)b&%+Ph_x916-~_ znij76DGzEu)-eZDe$&G*RVgq3xg{VkUfrh=aYo8e{r+s*Gbjl3~yM+7KqvJfCzN zb+z|l{k%i7HfK_sjZ%2#75vzEi~aocxb9D- zJBSaX?NPIH-Ts;O1jMY6^x!IuPd0^f&|xR>;ZE~gef$@qSw`Qkny_>A- z1~69V>^p}b9+XH}_@Cw}s}6S)u~bbDEDM_;jBnjcYp$j~2f2xMD%d{AuSr>12^4I5 zx>qvZ{m*CQrPXWaz&f4?YUxj#)tfh9A5GD{gXk7>Y_dLsdc-$txox7w)RZc!?X~`c z&$&Wa))jK`UbcFO@gt$IEkEIkCYh?&&xRAzDR6{D;@MJ!8eGZ&789hKR&Z+OeH2$92xd{F#%RhI5*m%)SW7e$qCwD%o zbkh7u*l`tFxhD`a;4S-#Lw(S0(`5!)(s)Aso@yK9J&R+@bari6+W$e2HH@uC}O4|NLG zm?dFzjlS@+;Tl8fV0p>ZfWM%~W$q7WYtkRFO=j!ayK6M$LH7J)h{1zwREc0?{kpYR z6sZUx7lQM+?cvn9Tg1d2N%NiG&#m3ABlO_NjQDjkCi{rPT z_^v{vV~a~Fq;Pr%0ne9m12UxXg6)lb<`MW?Y-9R)Qg0I7L>ui%-2rOkuR@}C!UEEk zBR^l}`R_?dLbQ>~I%DB5v1S>Hmr1!36xh|9Nt8vE(ej5&#VlcX5*>$yG}jvhf6nRR%tFHLwwE|hr!X3 zw}dURw^F`S0@%Itt_(Qsseyz1tnQlcVYg=m3RlauGFZ~7X&_+~<-c&Gm^wnR%+b?g zv`8#l^lCz!7g{-1w{n^egSDt)gX^=dWpqkCL*&k+lAh&XzS|+69@IG6Ie7krNLvwA zZdRWuu1b+RBq+i8a|PK5P>9B9j{IQ$V}+2#>9>uYc~(y7*InTjYEZZVWkU-*sT&G@ zZCIRWV{fcs6Tv1h&ERw*INU#8Mlu2>#h5gjN@zzX%x}MaGGs{vS zY3Ad+&kA<0TP*gXI;f_@ZRd2*)!|nUxeoms$zM9oD}*NHz1IzNIk+ltGO+=U9>`ex zaXaqQ=3wAGsplk^+-24%sHA;NI7``L#AWL3NBh-l_R|obn~K`H1{D9;k81Q=pQ|RVLLjzO&J9zFgGcsAOl_hwl zT)$wBkJinFHRA^TaZRqB%L~+|oyrG0M6eLXTJ{k%liC;TWB)nMm^eG72YJuXN~2L3 za>?wms(y_>G*&&ELbZ}Bz_i_tYe?k3{9^cMi>6{aO^wp4Ovq6N2^aE=#~Qt387T7v zc~iyYJVsa~!YvT~$s=l_{@2=+JX||Y@F}T%?Aol7CB->QuyjIt@OTj5%Fry?J<7pU zbl5h6^6-5+<1D)TDgsX_ZH5dNIVOOQ(_B6_lL}|x<-I2mWr6vTBbvx-W)L#^A1cL? z7W1BR`MmKT`ow9w#>f|Uc(r~;o5prQm|^r|<34e|*kEx+ln_(s^M&T6FD3^Y zF>LaE^~XhDo+`a;yI9UVmP@xr9WYkWo%LzPNR^l1uNEV&dM0SbKa45^vkCuu0641C zV!j#N%|CO77N$`xC37&u1Nw*u@UDzAa=XiJ-C)_d^VRG=a}$V(If=$z$liX)5SDJ6 zvJY!*-*;w>`{n&D?Oy*wcvmu>#&t2yNN%MDNU0p7rb|?_^RORp_MzR_Q?Ow^4-&E^ z@o#LaWm6ZhxAYp$IGY6V<@u$gns_t^&w1|1+0o7p>IL1Rx*^O$x{rM}qSLEc^3@ET zs8N8YpN9kT5Ipn%?8XDkqo}k_=ru*x2H}u44xnX>!&HLo1t!{+OD8GfWqUou01?}Z zu>PXALX{We%=_69vdPKEufwr_f#+)euUDVsEw-E6$=ix;Z!Wp>$3EusVXwvCI%ImX zPu8=sl`~H8IPo(6;u0L;3kQ75o7mJWp?#v9qk^mqmV-|NXgQC(?vwu7wweLO1yuyI zKQE^nUwH>Tuz!g?K2s9nk^TuD5dB!;wB`9d^K02l87HpJu>-RQhdHuxI)ZU#G~r!< zV`)?=x%J$jyHj4m4cg`~ zuADdEm?bsgAq~ogIslRCxe+=SNpb&=Li9+Ds2)hG}4BO>$Tl7f%Y%(!-U6UJY&Dj^io!jl;|#w1-Qw zFefjnKmGJph7Xz`2?_VND15|q0KPg;H-21mS@O2=V%^cj*jTDHI5uu4^O2p zV!SB}7XNc&J1ZP3Dx&x_eozRq&!FY3 zz;DB=szW#8>}0}-zkf*z69z=hoBs(V|hWE{0c6E zdmA1kdZXHpLW%xmd0wKz@8YIhho8KnzkN4boSFjnGHfJct&{_ML+cz@5Y2cCdq=|N_cyU7 z&si=H>p|ag7u(bMNz)f_GEt_lcKOe|#lW01AAPy2QE34%5W{oUFwr~lJwl}L1|2D= z@LpK#a>t^F%e=nb*VHw<74qw>pH)@DJE%r&fIv@B9TU?0V!>7 z-t3*>=W_1!%>~x!wcepeiycu=L(vCZ=@7$}9(VP8)6^^`Qk%9pg_4r9%2a+;W`kbb zeO_X%p^FDshi>r%00^;$jm8mRCJ80pCVYeQqb} z2>syI=eVW*yD#s<@|a9H=?Dn3Z5pSy*etgQs+!8}h%}OXT9y}tacYvGV;GwVmh^%a z^&hVey_M*-ZLL&${suc)3?5QaeQNEa=a_$i+~DxN2S}a!>+hG75#O&2I@`g+qb`ei zkus@r7x8?S%@K2NT7;7BHZPK@?TfBQKV4u@C{{$AEe`i5OSy8X)qJYzn{Dd+k!1iz z@rrTcAa^b5a0HV#q!)T!Sp03@Y^Y{S-ji*^FO}nre$NW|-m?%M+fjS>W>BYf5&(*BqiP zd$vQ7VKUMz4SF+2a(2Ug>v$%nlNy$n)h?UAaL|QiPKzeLAa|sEtOl~UE+-8K98M-_ z`~}*90>M(qi?o}+xLvFnBJ7}ArIw4ZU8AXq5}V>LtGQ>KZ=1!GAV5pL=E&IC^=ohK z?lsHmet37OLI)^Z^Rb~)S{gp=Y2t4HcV;jX!7g8qM_9e#||P|j0u6Cl7V*@=-@{1mUm^!F*_a`W!D2h zGwkQeyms_iEzEhsB$9c7KH~02MA`D$4i=e5*LC;K4`htVIZY@=;{p{Fl+YaM4C=7f z`>y8M`GLDy3J|KX58_hX6Z6cI756vl;e$$+Q*l<*`wg5s)gDny=1@jvH|0O zu+_`Ho*O#0Azwd?B=+_{R~O-`oqM}TG&M@`D+BqSRM86+kG0dL>elErUG-9?=NKmw z)1lCWCwjSZ{2VORlH)lB!E{#%a@)7XbRRIZ9mlQ{X;Pb>Z>k8t;}eKkvE!_?*+u*5VFzk@vOM= zNSy3_J#hG6AZPeKU3nI?(MW0(0bA_rS=@%%rGU4Tx-G*$^%KMmVA`zu5XUWC9LRnV zJ^DpZ%2SU%VONo4C_aZ%o+!z6=7GCa(RKS8)ca0#(Zbqc0wC$p4P9_2t1}AHz9$~9 z#$OltkZ~d%4w#~Bq&HMn^XaE-@^YPBjQa#Tr6PV<26y#R5s0uuiQ>xPCEkR1^tk)x zOJ%V(LWZ)fEsPY^3T@F`dBh5V84@M!w(a_Cff% z#Q0|4we1h{JNvk2zdU-D%J&KX-D1NlIQ23p7Qn)w6$tUY^J}3 zH?KQa=6lw)>I+vM=?kY){dAGFihD}wwh+iT8tA{zM>uU%Z=o=*df_uKX1N z^6|^_sKM0jBF`U|puH+G*A8#CGXvRRq@GOmF~^;KG`9_U<=M8kAR+ZlhRF<3W%=aM z#~N=#hV$QS-7UxMhP!BHecb1V;%<|fb)co{epWwk4!eoIJ;DD++@iS#H8$tJN-T5 z+Fzf!R<}6nP{`sLT4uSV*ts5-@1CNdb7GTkDKZ%@d@{Sw2eNZc)kwzsab4AVrd56Z zNBl^TKqm>;#{DKuw2wb5M^(E3j>R8VFROk|HsXsgTL-7vZ3s0V{P=7^o)Rq!W#QwX zo|tQkcS@}2_Fr|A4qEt)pz-0X+ZT9iw&;=7%G|K(N%zMT8#*s2jE&-@$5XV#jp7@rq- zJL8W!`8Q2Ye>CZA-xPRUE} zE(!)s!AYX7lMeTh8rGxDGgB}orwME^D?U5f{+&U)oU45;RRjd;Sk(W}E;hU|zV1qR zWHSb&(p0(N_?055 zX5S@q`V4DGiW+>x?qkbj34~)9u=}~)#Zx2sb5hcQnsi_Z$Scr7ZMPz8iL7vU^oyhZ zdaQRtpO*0b4coHpBwrY?aJe|MB#jmHU)0v?pt6T>#wx==NFcxcm>>7M;}vntL^(J)r05# zg(3B8k)G}taF4c$C-up#jO47+aJ&fX%5P&&_#(MR;v?JexAr@+lr)cHP3VM{(u%t* z2v$GZe}(bP(mv7wh;kV#U;l)x#`?=MwybB369KcK5_LfK6Zmt2ryA+W20@Ss8Y%*` zQ(~fb^Ae`hC`x+FYdgP${#+Z?>rl4zkS$c}>+$_Bl^(LMHlSZP%+wVG2te~;rjR;V zy||V1`J0J{Gp)P78UC2-P`pLY^ZQxed+02|B}a|m9`bZIOkcgbF3a6rJG-7~E!Bri z<^+{srf5F|EX90b`tqf0F$wlV!BX}3ACSb%;hIF*E~e67(YZ>eAWBdP&XI^3tv<@+ zUCf<4*StTbT&Vw;zw8L$l^Cy}h3H?-@=EoT9qSIa{kD=(ZXcMhP)$%^s|}2nOx9wI z3FhLm9^+9ejvEjo#^kFa9b(uS^s*$>AQ3fR8$>7gA~K=fr|3*mi}suj0LQkCwD!`mBYz=5XFsHgSRVp?SO=HA;!>ipK3uA(2v z-Bns#2YknNhRE1cjqs68MZ^Tp9fm5Hy*YW@818v~ra7?rN=GgxxjHD>vb?prT1v*q z^be@XIpjNbo3K3-hp2`af=mP~Hp7I=e0GJtiY+(*n)4u1kd@4>0V-TFHM(1@wQ8_w zd-Le1e$mt{+z5#8AYZ0*nD!P|PUkL5em8Ew-+uUu%~WK#kvshOX_-h>!Yl4Z?Br@a zdXSDkoqj5N-YmD6DsDU7xA9Nb7q*?r;(%`-AqxpKjGXgtKUp>*YTJPItkS!z2Ew07 zGz-U9N25O~IL4&vU}w>Xw*l9v!Fau?Rc(j?vU^XqrONz>>`92K-;RCsG1_~ z20%{DgzrW%M^;k$xQPX?<2b@Su+TrUGh;cChpCkA#$$#BnI9#d8QLU&HRT|O)-{0M zMkGXkT;>z$b$aUcFYz*MUgD@8i(Cz{XPN zVceHOYtWqZC&Khbx?Wgo0wu}}kTEkR6w-Q@usVlL!oFK?Bq}e9@!1adH2z?>hFm0L5VZwL_?Tfz@APnEiLymF5 zUNgA+;(SB}bQn3(o!5tTNUO1VA&LA+o6_!bhe!Q=6w8W8K5r^}`1=M;cub-Wsvuiy zy!ygk_rc~iSv-On<0lVpN<;9zRydb>;7@wWUF_7_sNfHwUa4v_{Z0Okpq3CFgZLfa zHfI;*aCxiTdbaj#woZYctJngW-L_|~v6UJUmG8r+ADCQKOm8WlQZ%-Z65&YgCpm2y za(kz&92B$jS|mfG<>ojWrczl763B5PXG~x9R0IaKv}|`9d8NjyAQB=e@x>HbUVrvzHCslLHM({hY5PUZx5tc>N1=A@p1NN<>O zZW%(Pm?CPVKGz+=qg-?GFo-XmI0ZA8TZ><_omh=$RO3;&3o6_uXQCAnla}QiL}}eJ zOUi># zQ2{9sj}^p$kSfROo$j6m%%WR&#d<@<2J*6M%veOlGx3s|xzr>3xsympX4FJ$dFlB>$4b_~Tc`Nw-{w+FKjbi}u*H7W0jUdsWzp?mA zaYYzExWD0l-Fg-I66dPe9)Qmt#_KN&B@o>aBPg{Oy z&3+_|v5eJ>pjM|mdLHXWOn?REwklm^Vr(p%YT4&@z8%kcu1GieI}4Yo^m0g@3rB0sOl0Hln7n$( zAAK@?$tNiV7u&~t%WktMl^rzFHSc_ss+B=&HXU0McJGrU0I?>ai|#p6ka-^Fk7Kpr z`W$Q~TX-~w8|i;jE;)Z&o;HhF-z6A7DXi4iD9I@IT%Wtj(!1eHm{!-bMM{8*jG(th zt}#wR$Md4)J+*?NE6y`ZF`WHhT@~!#7a2aUr!Vh*?vLlYYb5Z1eqhawc+J<)q8yX9 zptzTBnRJ_9qqhn%4Fctm4%X(4Ni9Xh3pH+jv;HA@Z+66t%+82B#XY=eQj-JA2aV!o z#BQWn&?n^Y-s{q)>!t&eTnJ-WjEZw z4pS9tvaDofvdch2313Lh;xd6h4JusLM=g)@CU`qEF#M?k2&?0>D-AR62cT)`RNA95 zO%$}LNG0m@El~%)h>st&NT3?Kw}ShC#21<>!gKLE!(21&r(|wONz-uIYf?pfy{X;V zLGvP$=?~KSaGGJ_eD!2|C(r5#7mW&2n_BCdj-UIqqiNWwg>ru+dg{I;ii;;(iC~DZ zGa`jB`c(ZldJ_hkxwvB4a=59|@HnZ;U>E-b2pgZWdFhh8?c*?N>=Y}5fT?J4&WwHb zCxD2iwI%(bjP*{1PL(8;j=-qztfn;T*z7oB07RkVp8YMfS<$qRWSHi6Y|%c=7Fe3( zGNTPZlysO-Q-H#=`n-F~31{tdxGB>92wY1TExcCkR|-DmKl$b5j$B7K$#Wwo@zljn z!NSNJ8l~jpQKQ2Ijji5O2{YOS$umlBsg)u$Q5~4rbftOG0JWp;w(d&!TRsK2bDRcJ zQ}=sk@~!m!@#iX|G{O%PzxjV1tTEV)wRmQJ@kIhk*r`CqcN6yjD`n_Q3>{UW*>h@3 zld8EeD2d3i<)h+y_snfEyuG#J#b+}84!aL8P(H>_5m$rx;mqEG0Mw+~l43nnseg;<1reL9)LMpL_aLz=>D3(N%kh>4R zOjjh)L^s_^Vq|OO67=~Y6|Uf-;Gt%*e#@gx@V7=$=rX;&7Aw1aoq;Gl7(6gi?pUmA zYm{Xv*lEBlv1VxDReyUcElSr7N{|EsphBVFb!FArD$>=fc`e==S`CVPn6PSWUC4g@ zT-3>9TySehu`!H<{cG~;N2`k;&MzHZE*N=7c^mO&X0C&-6Hp0n)w(gnErOL# z-x4axlM}Gj6N3C~y}L|%4T2}k#SuxaykZ zEb^|xs!*qpKDUA~b-zL+Uw%C7RX4x{2}cBB@8G2j3^C=izzhUgo>QFd*E}rPlzHyD^jk zdFF?@=Y>>>r10Cc{DXb2dXW$dit7#RzvUB1mbS??p5^diTmlA zvQKXsWuZ*m00(V@e(mJ3hIQ+NAew>ByWL%f3Ni@YWI#B_<#vsfl#xRe&)yG=y4HqC zuIb5pHH>h_bTdk!s(6x952inJPw|9=75!G*9V_@5mK?iz|X^a+It53*SS75BMh<-}c}aFYI@MoRJsqeam-X$ zQMHj!1pO6~U8(PO%PR%tRyg@E2dpVA>f2$^EQlz&Hq?FSjb}h`sM-0Ae3kY#z#1j5 zfXDl@opIfMC8v|E7SWTs@4EfE;{yySCWtO^eYYe9aiWMCrU%0-zE2NX4Sr_Y9t z|N53R#kn4 zGGmm|@a9sMWW(fhGqs$>g(@BpF!2S|nm;oguUdB!y@~|rc0a!&rzUY_FA_LeRQ?-vl5uD z2E{5M<>@ey+8<kk@C(+x$zphTAt5=+$OYk}tGPlQ;W zD)qN69W;T#VbV*#c|TMZUQMR$$vsn3gEO1USiV~5nAIDNlIzoYaZj3=;z9%l5((~M zOK^GT6;UBqg?dw!&rqV;HT2 zB_=>C&R#!iK;87>>cu#}I4;k{(sGN1w{(k-4$D8UMo>vKk%!U%KnlNaz2zM4z7wi= z(+-Wh&HYr`PY>V0Cx-m0<@aS?kDZ`{?CJR4h{=$ucwLg8LU!Wu#>B1?pLc-L479qn zGN1%^-t44bjY;1GGv~zCzKcw(4O^g6N>O=G$wHGla`ee}ojuR{!A66;quHitL>=tw zhnYf*gp$_YQ}=1r=^E4S)-Y8)kh3kx{i;4ZxP)q*AcJ01f_jt*)E5CWAIwPD>pzOl z)lk!jV&lf5qQelyAjT{nP79|jKFfVes29AY@gn>wIH1giFfzLFk2-t8`W7Fe(G>qt z$29iM9{r{9S-DdEP`aA_^-v>9fe=lxx0D-cNA-|k zw&X^LjlIP)5ex2c1~26|;WPP>j5(_Vutp&g`7_6we?T_Cvy+1Uxrfiib0EQ7Ye3?1{qW9;;XRmEi$C(c-}zB)bOd)^+H4RLoom8ZF}C1V9X8S z@4n=Le+s-{<0`}L!=L^E$v>wl9^MQHjz760|8DdB;r#Q>z*?NvW-(2W;5)qaUFPqC ztN;b@?(6;U>+Y5_@$XC;Sz7=6$moerySXDos9dI15b2d3Khu!*ya+>3UdI26&j4ews%I@LdFk4Y|DRM>~Q@hP**!BWyTI$VTzO~d=Ew*++^ z?)TctJyT^hVPDwBqeJT6nnx$SKl%)O20_p&mNsGd*>-wwzWXaG)w%Ru6FTSv!8`O@ zksXV`$Z0*XxrnL*u*w-=%&RB2Ej1AwXVc#m_+Bm+`4jx@1a6%)Lav4AU423edpyc- zblEG*40%v`rv6*(EYtf!O9g7cc#g=#NeUxXs@V4 zg21Gv>;N8Am*P#8*v2qB#l`7eZjawD3mxbJ#OKO*dZs!AG%5J!jH_02fV>9@>*YV7 z2A|cRyCQ}U;?EPp_v)EJMNg#d=hKB{VOEeOb#(`>l8WUSD7eXP_DAKkp$eiDOIoY- zqgL6T`Nk}GVz6`oPhw#+ti%iUL3G2=O5Mx`8PsAHV6KhMny11HLRlnWWu(h^oBG3B zyx|)a@>Q)CLC`V}e)9=ZXM#>TX3qO`Tf{u-)^UTgCWMq`qOysMQBe_TRUGL_B$0`A z+zv-i+#pn|d}YW(-!2wg*$V^4#pO~e-1yB=@BISMdMWBr#zM=)uJ}7G#Ai9iDjY>2 zLDS+U_Gf?8>Q!<)LzSh!a4qOsNw11yCLYq1!EI-FfnN9;Ui2t+){JuNg7<{z=OSGa zJmI$BN*OnL_0d$C!%Nd_FYTiBPm+wHodh4{n!*-Z0PyvbPK8cg*7R31C2L@`b1{JZ zqs0dF3;?`rjBNH@$SUD=xKaaV-DIR@Gs3~?GtA)1?(h-Ksd2EluJ0x2E^U(f$23{} zk+JKmVol1~OamO5)}6zX5nwn&}&RLayV_bbl7oFbJ$Hl(-g(Q zdvQEg`LwE&bVVb)hSapdaqFl=acR7!uy4$%KM?^1isO!Tn(g=iWlF-It$Au>>98!givffmm>J%0nl2KXdc>YgZ(!d zOyE|9w~%^8+X&;D!HY?%QQVQPvTSu@wF>8Blaiy_=WUh%pYbK~eDGf5XSFlmL$SY9 zVvUG7uG*nNi>40)?L#_pE*X1E%j3vHgQ-_P?iR8qkFUtzjaofq@A#@oFJRm?{{UQd zOPf4iHn{s-IH7cZy>K^vn+R>*E)%L(57U%q5}*PH++%N=@=q0lj>4Fs*02?Ihe0gz zr@$BQ2sitNPMT4|uF}p|ov9CSS=u-pFK)0gpqxb~V4Bpw$RbnS!fKfVVvom;%0L@P zguZ3xZ+G~v>n1H0s6Te6t?B8t>HRv_)|)`>7W8g3PYx1AKK}1L@(j97CMsYJva>SC z=Fpb}Yzt_KqGB8G<=TSo3h=%j<2deLDg!IPeGP54li!rQZgy+wz3_O=z^ zw)D5tZjpHEX||7CTn=ujlI^WVMAj2Wfo}{*62@)x`q?&rhL5cK3*0_!ErwZid`#EZ z(+^{MbmQ^UQNYJKnZ}*cIFT#RRkPIDd=iUk1Nqj#^D2YBRqORWjvtUO%WQ{nPrpPr zmQ5{0OrXO>!h5`1!IvhgnGI4PATuQ~5M+=ST&6>*K zU^ZXD--wub)Bl)8bTX>vmV`3ghVVx6l5N*cmWE1qU0{eyRxlT?J|W)Vu~mQBw<0B? zRNiAcs^dGB7x`gr?Mf2+8W59PHNh-_r>h;_3tH3o7Y^spmTX@tO%L1mU~tpczYXg> zfo%^|CG*`gJ={Pba!Tgwqk+onV$p6vdy|Sg(;@7?RsYlS-vZsYS1qD7GW_i2rNfCc zZaM>ELh(ZJy7|5Q(fCLU=eSZlQMpxOR$hW|bkDZe|DL?vz>Q3`V1NNm7n4*m;vE=5 z|CZ%{FAMy;n|Rm;qQqTG^YuSH`S11qx6h~V^WmJu(B(&ep4c&L&AQx^e*~lSroQ*$mGWFsxpWf>i(*G^ABj9F;q;6ihwB1|6v<2H-e*VF$!sim) zZ!vr*iRll!{6^)IyJmh=uv3{aQoC_;jW=o;EPTX67z=m^5uJcZ-p9NW;C!5xen(cL zoL!zDNB(8eLEcSxKG3_oG1)J98WD$Mr37A1csahy!+!Oc9o}(_^3|NscKFxdnbT5S z>tskXQFoM=Hbh!L~yCUn)mn~+C_YEo8ko&jC z|90@Y*2CN7g8T66|D`YfH+hQ^6VW%4_aZssUTyBg!nn5%k>vaIi0YH^ zhA;KL;$N~1fP+Cq+LnU8-iX-k)V|W%4SKLxAC&^fOAWB5rS>rce^${0j#%VI&< zfOjDaz~9}M2B2Y!HHr#gJ{(*trEnA=sL)5Wi46xnqdbQLiKzz9l_c?c}|)U(Q? zwqAziaJjMC?JA5eXl#V|O`*0{4{G<$e7`@V(D3-$V?dQzzwl*(S0kWEh^r?7SUhWC zLZkW`BZk~8xkbw9%h`mqUykxo89Tr_p6aN{9bZ1vbtV3G^4N9A>t_+}8ZPN^*K!8y zfZ~q7A!*0sb5CG{$6z_PyY?+V%xi|PmBs=+X#BNk0!;?qYN#jClf^w&rPvRfH=yCW ze}|a*`qSz5=%#bhSauO6&j-9FZ2C^Cj*17D6?SGnfgXU?bP`wvMjyS+WA;p^BLBb@ zGueG)3WZ}c?c9!kuQ7zwdR7`La*wGo89KJWitd>cz1PyL*S_?5?odS)OGNa*g|i$n8E!Jnpo z1pa0|(dsY%-wG2$QBqw&9CtXb#4O@CqC|7t^A4y>bu#>NN)=nRFYth2j! zYQs#U!V6oelu}C?iaIt$Hwt+fo_Q~_e-RQOxmDYLgI|uRRwU-k{gu{w(FYN#i~$0H z87$R=D3v4U_ARP#b8bX_jqtB_K&9+*rUBab9sN}^tYu5;k8XGthb-b8N9Rc{`jDsF zD4RYJ@DFxIaJ72W)lAmZyo6MJW?{oP8EJDVg&@+4YcrKiY3WF6UeaoH5v^y3GMCHU z3?h+krg(R*{`j>ao=%wcg+N*&a+QaV_3iQvO<}aunlFg4Y1KZ^OTpk3)q#oJxNtCh6d0@ovw1Ng9B@|P7+7`Du?ZbYPc*(9;%f)nrz^LdFT2i&LL=T@V z-g(gc0wX0n^AG6oMr8LYJul$tTRFEa&Sp;R2v64---IBMAS5Zfo&c6NQ>X5o#p;sc z^@CQa&-vj_%ex7TPU3{T#4J)jryO3uZTC0T9^>Hv=j@QvZSh;2@V$4Ix{L2AiGR~6 zgTz{0G_@W5z4sPF0KvC#v3NqvKB3a$Mn0a8UizV zUUO~cA-B5Y+XuoCRL;H2cN0>ruDWoZ*kZ34{t{4aL}b;hvEHOTnjBs8efVg}&-MU> zR+_8GN_S|cmenh5FdeA7+uZluZ)4X2!I;aVJ&?2SEYZ9vau4;Q zdGJP<4d=N17U)=>7d9EpA|w)6+YQBB(u!P)IM2k|9*l@?36efTq-uU4x!TDulqBI_g7_q z%Vc%nFyAi43wx?^RhAQyc!SWEaqrml*5YD6A$78C0tf;b7x^C1@P;VRJ|_!+xqj}D zmCC+!4y<@Xa^~@)`O_8A{JRu}I6mQU?>vlV^*zjT&^>MCpm=vu#c%gS^g}>MgfeUHml|6|(c#Fu(%TU8FVl{0r>}$9*fO zSj%!Y2aI-Y!zX`p^Mc75n{3y{OCCuQ=-V^J7vC$(at~RX$f%4?VpvZ?<1~cV1I&(U z1-EZihI3M-`r(T!Z94R(nYl`YKX;hYMUzK(zd56*XO9TSi9Al35@8D{Vp1CrC?c8T zw^;p#`{G5r%Dr80ZsJ{TYHEJ+$<=V9$)f&~_P46_6p}YTBxxrsp)yi`>B;QWO<*l|L%+5=zsA(mbT3c8@%IIu zQ`E2M5^PJS+xt-9ZzRY91!Tf9$^z`grNSdj@!zd*JtMoz3M-wupA33M=Sv9U(*T(r zR9)}KKTNAoW`fxvo*MgY8{Ub;sL(3gRlVqe!v^7r$vT~&!#HX2zlWXgJJ&d$)FTY+hh^V*k9+I*Y^T_ubiEg5jLTXhRov{4M3av|fxERxnzbh>gFt6XT(iyl z^rfGX1_U6t2VJA(1XNzwV=mf}C}%UXjs>PpdrgS_+t4=D$7<)#L7sq&mmn=-y|Ui! z4K;sopvJQ_g$-4)=9c||$Xer(!8&y|K~>3bST`xgK3s}XVeKcGofI3*7%L*e^_;&UT_ z7_G0eA~0S8*TnJ4IEjLxm4KdMe&mFRPE=N7@NJr=I$8{{cm6z0ENLJ$&D+ z(y~d4lx>T1)mC~hKE^2g@#8TN_nqxoBv~&JfirS8RmJdbFSucOxCw&N8HE#yLxG1y&5AJcyk6Q%*{gE= zlrwUr(*n^J5X%#-kV)XK9?=0xh~E6aO@_}@v&GivEy=>kG$xYDhKwi3e)N&T(nvvy z3%c=@ObN>LYFaGv>{_Z^_)->{=ha>{yg_UX`)fUIc`f|2Em0L$l^$wF4KEng0%oi< z8o&1tJc%UfF6He$KpN zYPyz&Im+??6B|q|H?e>HMhe|u7w2mEF>q3o>u6;6gC8!rF|Gn|Av zyCkgF>m$_T&Cz7{9`Je~${B6tro_o5y6+n!l1r{%hNdO}fkDA;&mt$JFn-3HdB2-S z_*US}^aSdRsrkU@e-0TqD(vK+kw2i!kpbf*6hC0SICPOPOJz@lmeE-ksdhfgat=o@ zV6|dEhyxL@`#hyvk#aNPg8P-YKD;B(9TFwqOsd_+{pGuXW+t}W(pguH!j8O?k7Sv@ zHHozokYB>qdP(^>l-^+6EFJw6Yol$%nT&6zeKbHBXpj%ZKh-r^5X|P4me`Ey3Vl?W z1{nZ>6Y{bbR*y|c8kyU~^&5Vt!Ru2ANb4)2iY&pi7lLo{i$~r*l`aB7VFeuq+3yF; zy(5Y=A1ZO09O->xs};s@608B9N#rCf9}Mbp5ve*I#Z+qQ#*r&OnA_+U=v`*+sdC&R zbWVKWw{des1^H;ndZzs9%F5bsrBbI_?(oy!Dmv!V_u)@Fmt+D`n}M)0+pw8~o+zo% zNSA%d@rl#Ex}_8z4}bXUd#Tm=SJmVcHL4Qq@C`GH;A4aXwqeCo_UpD7_pF#CV@RTp zFBReEdw^z^456sv=PHex;?xqAqZ3Hont0m}T{*3%&yiC5T)rh(EVZ}g@u(HcE0@Wa zf146K5*#MyU>aouhEb)P?*P%>@{%6hV3>t%Q>T-Mriuc`ZE&M<>2dQG{gjBiI-0Pq zt}6JA^6BxPMhJH@zHoGdxv59jXXv9x?|MNCcKLI8X85(RWkM9CW(OnoEO19GTaQSA zQQ>9y4n$qmpmmlQ%vT8~G^4-UsBr z=!QU<7LZu_!T|@^Am<8woW4IHla-M#*TX{s?kp_)vV1PBBjOfsAN0p300DnL6SeKE)b3c*T=>kx9 zLp1kEK#jI~$t$KP>3mn_$({}qdY;yrKUc#=pLKauad5MIyjza=eV6r3{3*>7^Gwp_ zs>TdxFZRP!VxwpMM>&$tkt5B!(?#XU5hk3mZ6DQYGlYKI>b{Uud|yhV*BcumoZTBB z=oRy)OS%SG3tt8_Ro5+2TfU7@TadP<97k5dZVD>PKx6_AeRI7-b#Zdm6?lM-Y=*Bi z*|1!GKQf{D=p-UNQ+34*g&5G~vzYyGf~{h(db#&jui!-WrEOPk2&%`t(DqFGewM{> zKXvLrlokl2)dsZ!=}42TFi;1lYb_b|4F*}SH;I6WxF6hW#{2{Nz?bZ1$g!*EyrNRy zR4X}Qz;P2pQ!akPUxT#lnQiI@Fu})ry1RHX>3}K*T6r|NPbYiH#caWKp8gR>U(Jp( zO-;jo7gdKf(SUklz_wMLhP?KeeQ?Ics}u{-OQO6kQ|sQe^|Mzr0QRaxvY5KEZcGDF^tRdq+3Y+dLa+ib_*M6s3izmocds1PS$Y?c-iebA zst%Vym3?;OUa8@wPGD5`WiY)ujChl>=q06ZmEpbBQf;g0I6C89H%i5fJn!>}4_Mqy zX6iZ}6Ls+#oGT9gVSKUAVD47Io&12F$ZVj9k=Q;rfFHvBBP028tP%tY++vZFGZE1X zk3Hc9n=7YUeh;QlYTO)icE}I*cArqbKPa!4?eP|stp*3j+jAf2YI9lo39&_1`5C_p1Ix@6Uj7`o=GfB~)xDDsvYxWqZ zWQ+*m?{;Rgh7uhHx2UnwKO>`6gMJS{WWEy^GzY)S<%rp4jbkoY2>a`6#B+`ao4bUF zLlR=n`N7){xK09dnTRtKy>4s_&paZNVDNCI=HL{brC-;h{}nSVKC_}a%mh| zCnmVdet>aYrN-;yrZ>sI7MVbL=={H|REeG)eHW4XP3@xQMr_a91}|;nqImP$*Sb=> z%@uG>p^$=De2E+3po>V4P*#J8z|u%?kAJ9!%W`{88@xau>;^*)&IqsL2N)oJ0QJf? z>jL>t&74|BNDM(<@*GvFoWVGTM{$jn$PHd@C)~n(P`dcE2yis^PFNED5Pu=X+Bx3qniMo31;~Vx)jLHhpvotjkPh|&G6gT~4$HH_5%4T? zSwIC;BIdkNp50ryweo>g^JxgIEugUA_MBZty#NKGrTR2PRhjGE0Pm$hUfLRQsLW30 z>6p7K#hQYcoB>DU`AD-~thZqXVHtsCh*4mN-u8QW9=tfYd0{`<fQ&dTREMT5V#f#e>f!d(nlo#}I1+$bMeR=&&AslAo8dxp)hhJk*4 zj*{JZZ26yU`+U3c!h+DFk{}AEazZrPTlSp6IUtnCsLP0uii4lNO?BKj4re7Mhs)6j z8DC~cv3rY*!%$wP#~^0|;W-b7wM&%lW^JO~gvv7yPTL@8b9aAO^I|>0fp^6~rU9Y5 zYCTo=mOQziEpdI%-b%M$@`7pi3)!2{N^F__cYo)M()x6bG71%OzRNL=KW{@VuJNgtMWybDnR}a7mpNBzq+O9mb^-9nO!at6Xdg%2$IqoN<1`QVr`sHOMD?3 zg;^+$lIvEa5*t#{d7~b)aQr?-Q*?kGg{xr=2RV6eE#}b@Y-0JXdA2?elMB^=$kUtJs^1RJ#G%z$G-ys7q%6<#4 z?35%>MwV?#-K`-Nh&NeX;;t452LUceAW>Vz9Z7w&j~8N`q=`o4cxtVi{wfd@qTx}% zNxu(MOa(?>V z^1nN2dIEij$BW`xV^c433b#?&*EXfqmt08LYb^n5nX1wouNj`P@Hyc>Muwk!>Xi!Z ze7C@xvXFG^$Bt9hyffHyFFVbEWrqWV#+XzUq_gBA-#bCu-V!!-`i`ojP~QMm3_!jx z91)tb+=zDap-YOEAO7%roeGf#>|)RNs7XXDBN9fuoD=yfA`bAZ90^pk1!RgB4x=zP z-BK$WQD>uw2C_ZntGt5S`q9XVVE|3Wi=PX;H5 ziOeMN)+0G%T*6X|+%qy`}_n4cve zqM+;r%GcU7qR9N^A3hPN|Gk?vWg2mL@-qrqHUQ5Wz69&tgg8{BWq9tv)9mA2g` zej<<-DUjq02#0Mp}w8`6e~KXnbxolE~b%uv9RC0{Qu>CaME*gF#tXZ#6-{QxAxm zI%POQ)bT?SSvMt5P`AjoMdZ1_~7-Q2uCQhY7>>wmES^RZ79ef=e`ikG0dCe7M zFvDwd3mrSn*c>V|Hw-P3*u=D6JL-*p#~=_|-L^&dDcZ|E{Nz6#Yzi!z4)~D8x-;IR zx#Ja1cc}55a76`;H|XQt;^6DL-xFugM=(yq1F`^tL?@bAQ-0Ub`%72y3r@sA%#{-u zt`34p;1<&N{H414&jki<+UvQddsr*h^p~Tzv+BO|v4buvc2D}SrmJ_s0~!bOdiujt zfelx6bKzzLX3rw}E=+%O_1%6Mz*JIJNYJ~}$H})M<1eyf_mONE3k3X3;5->rr=#_U zW|AxS!Bo*DK*SH5%nz)B@HEX^eN1w5D#KW!?{}hBhwdng8d;MyACgcZQ&C4MoV%nt zNf9>BKV<0aI}ZO(Mu9ptx5ZmvW>!>25hkmE-)Qpmh>4} z31Bzl{Vw;ea%A2r@R zPN3#?+Se4gkbT|d>Y}2)oz>S$p-dcKtsehF=7y*A=B~f*i|~zHSw zUE)^f#&w>qC{4teyNl&_AQUllY{gvEZCEu9-qzFiKb$qOpAL;2Plc)ankiB+MDJf? zAyhR8%)*Pz4h(s01HvNNV-IhMTmusDi{vTwF7h1zHp+b3%)3Ah z8u{O{W4zECxcBh zM5rSB7a{y_b+@NrJf;!-WtEp1fAP=fjDKo9(bPZW|HE_V|8bS}mPlH-6ZrtBm^aZf zI=Ru~I#zCeE-F=s#Ker2E%(8HDVLiPq??Vk{lD+*|L=~m5ph1FDtbj2!6BhBCnQUl z{Hyk_b~VYxOHVs-&FB;`iWck9Gidd_XW&)9PGG;`l+%l={VbaU=7lC}rN{w^f`PQ> zvc5Nl(iCK3l&}zUx94ETpaniqQc9%&U~sIZEgk&CrjaO_qaKZ0Q7%1U#>B%*r(}as z!PM=;6bmC9zdu77ZR_5A`J3SXB?}VI8%eJtDzb1@5t9bO3{wr0l9>dyT=3kz=w8mO z^(~d6Ei{TMF(EDfy9~iy7$8TU@NT7QL zbr9x*U@42pZ{_#TIQdG3qco9I2$p=YM!tHsF8$7=xHXItv#1bxq+$3OjkiaA3CXsv zQ1ZPsd0IhXEf`+DMpSWONAwkQpZ9&)Kgtl4SUJiTn>hPNvnre&8%KE` zhF0578<>R_6}k`XQ)!WTwjT6-60b6D%Mv!hYM5BFHt<2OCyvJItKpho0GIw;3aeUnvXsreGgsIDnV`0l9j!8A4r~vr7chF*oyD4gDv# z_J?r@5-3t4VMPu}Nca!PhYp}**^xHIYG@~zpiia#PhH;u)nw4MnS=nLhu#rFLT@6y zXhQEL(tB?L0t!fz7P@o@pn!n%4pO8yktRqN1f@%nCLKlG`0f7x?w&n+Cg)@(CwX)7 z&Ye5Ccb@w^qEmyTA1bnk-nt{o+GKobeQ)~(r=JafbclZz!78}{)i*Jnw%j5lJ#-&)bB*-HAuAmGc8SVR5K6q2cpjiRrEMu$0!jUmlS5w5dcJ z14$GiQ4`e7nzGOyOGzT~;ee5FHDc8R3mPfB?S&kR%<9J^Jp5ihjC@8tPQsFfJ<1XQ zd<~}g3ph7EH@Mr;fmzV;X#=4Wi3k-UZ0&s@d3UM=csEU_tcrG|}93~%N_X{`s1PYv!fPn13bD5XwSjpWNb1I1!B?)pjt z35jSb9qi+w!{Y!__=&uePUhzj{@P#0mS?pcyO#db?MEfPD;ht#|KFE|G6{q)n}P55 z|IhaVN#eMp=<5D|pHZ8)^9;S>8flc489lUWXgOH@tr*i8OoyOXDK;1DIHvKvD0XzP zJw`BV=C%rgPkdcHbq8*q{RL1LUUnx5Y;`og8SD4CdXP8#JdhPPMtk z`~~!Fvc^^ajP%Vnx!k|9np}%{eZF=>9uuVq$*FCj9$9GJ43pe^erIFoDvzSew&Pv4 zy@Rbr$*e_zOp?@2r80vefBW4A07t!iAs-*rNFb*#cb{XZUAIHA`b`oN#jkOd)(x+s zg>mT^%fzwheN#S=Ek^)kL2CM84b3ISpj2lQ$nz8#Ynf70HBfZif}^ScwW-qR2-=5W z5Zw>VWNg6lHrk&RF9}T9QoKwQp=sCN5W7$$7L1)WdQq(vt3TF5qYu*fpmGlh7sN!0 z$lwUr7aSuvQxH3z6_e6>1P~GjtOLO}Cq~g4_@|ll(g}}!jhCNe;TgilP58x7FQNhh z<=-4l&jV^$nY~N6phPU{#Kc^r(j}=jxt|_9r~oQg^9}n5F!oPgDp3ur5d+d?8LEPd zmZJj7w$GxQ8lGHfXaIErKZHR|g%lKaWmxJjp&~+DKN7s$ue>=E81 zZzlhsZoZT&^>huN8W@LO>-lO@*(WeW?DWSl^9QU@Zv zs zU6Cgmp$%Ydf5Lo6lT{vURYlNy`h(p-pkTnxW5KJ)}vUVANU4g=@kOx z4oujd8AxfL_y$ZRuYgsFxEF}3K=)Y{fRxN}t2fV=tGS6%q6W^x^(7h3^j(lkUEEV= z*tKd?h>wnEq7HWhUNy)++cXHS{)5iAbl71d!(F4@ANXK;9&19hh&B#~hl@A6Bh z{ks$+Iz?}&Vfw@yM!F8429s_Xw5J&Rs8klH1OPOa38INR%}(Z?eg;c~_A&}gKeaR! zCO9qN_^IM@Tu%02M@>Sbf)DBZv*|9mv$+viI`L$IDy^kN4qf}T*E2vyZXml?Ns7M7 zQxy;N)NIG2>Uw=UN|}5K)d4S=E(WNj(OG(iss2<4_n)gDjfz6-0sKV(gc2Trw`Dhn zwFQ`TS3Ig`DzbZl+7}wW{NKw(s^uac7F&+Ao-2I>AGcYFQXWgbC?sYIV_duwcZx7 zKAH?Tj7730vEi0_ma9X=4{fSyHaAdQ;}Ou21B?I*5fS6{?+nKg8a-?>*thzXPaF}| zkdr6Cq+#`I$0qRXnYSzXba{)CYFH*Y`_vvDcR>48((oxlfT>dA&HH(toz4Za{U4ux z{`RRK0=Ull&nImci(C%}{Gm{H{l=~Tv?nr-V6$k1qsF%l=7|=1$x=vVKJ}@XK**?g zAe=7eXC=X88F zd|hZI?kisAY~Gl_c2P3yA_sVCB_miz1QoR%Y{B3-t6*LseY}87+cy-KJmPlF3)O_{ z&?Hs!t~jVZg?f!_tf8Bvf1k0Efg7lC^${bC*K#xkS?i##~qV;b``!~WR)}Ut10Gs5^rbVzAoO6G^Kuc(;wDDaXNB9ZXK}Z zxb^3%l@?nf4bdg}cJli&x%qj(1B!nZmvA8M_V{d2Uf^uz)#AH1Y7fL_1BAgiY(ND% zI+3|;p4e56>F}rgWfuM3~EF6!SAN@6h*U29#TlJ>az+anQb#TFO`v-bv zQcc58Z&|Pls#eL7g1Jj7xPu&%=4qEXsQ;*_6;V!w{_@%4G@;XwCkrseTulkhz%22w zo*Qj;Z}=iFwM~BTSxY*8q!C($_v0{Sis=gvl8ol-BbQYiWfd4F*=EA*{91vl68GxT zbl&`Mx#6XX-6Ndb;#6Q+hnka}*S)%}!qwYXUia*EKmLUqRffhK=xqsXQKh#q*)M0LyaY%YJXA~~%S8<54Y%iJlaQ7)v48;*nnC*r0Z%9;2?)|f9x~e?8_(~W!RFT5&KjXTf)W)hx1%D#hm0%( zef@8xcwntXN`_aU{dD$woi)i&{uRM$3@@Eh^g$_`_WGel^5r@Bc25Ts?&mE5ZiLBI zPtNQ-%tmrHxhk`ul-}ppYeE61;Kk&_b`jZ;`GjDlH2*M}x)O&E-ky7jZ~BI>osufJH`>*SZD-laMEBF1B_v2} z4XBHUUx@RiZ-Yxk^AcM`<^seYP# z4*=6k@e_f9I0s6q;lBW^9yEXGD0OjnlymInnQ1_1u<(M`N1TJ=uh^+AE9wSyEBtHi zR`F!%Q%)rx&t}&`$?WTNA&QSjFPD`r{?`^bc#)ehO)c z{{7F8*qr<5qxIiMpoP&SGH)R#_FIflxDhv2DG1KzyJg2M6COGk z9Uz2XUEl%E#ahkQcU5P}51*%fvQmtxwF0UVS!HP@An-Xl0l<*gL%pd0;9Wd*1YSXm zcl|wA6b?3X5a%==Qt?&Twn&kvdZh7;>*<>_FC2e~JBx7TNcL6wxS4tmlzT_Y^6nzE zWT$c_fham%UN^1*x2%grffdwyCv^hge6Mm4{s)dxhbc;LGG1(glNbQ4>{gKmE){*BZCzo{YLJ{%Z~TuO@S@ zwVum1x`f;QoSn3|0#LS!m>>9n?Km0%)T7Uw0iv&m(`nD`0%i|F8$dU$iS9{qz50Qf zibMjvo7)d%lM!GlRdt5xuQubuVw5V0sbLNo3B3dcF)Ly<2wklwTlY|^)y5W$Vzhky zR&R)nObMzq8t2{x4l8AR#N6eIzb{vE{% zvvhK>B_NX?-QGDmCvUGE!qg zSMfXheu3UGDKhqxdUz)1q38_=%?4QWJ+x9lDd!55Gvl@K`}z@p9X2-Ak$_a#@rnXcn^ON{ zwCFlQxHEq5FCbfw2G3hsHe$K^>OYQK^z^n32SbR%Dv`;t2dgmwWA2&N^V^ zK-y7VKg(yoVE=+ePcvNuamIs7*Pt8feQ002U_M>ao+yKV8NJwC_{jS1du)Vp>x0gX zqqgx}@HA}8arF<{bkaCdC!>#tuHKd=>dRSwBHXB=TxEoc1R|-dx)#Bb;748<7-OUzz`)eSR*Q!? z5~>OPV!1;2#lX>`@Fff`25@U535rT{%YdarqfYK+zm|hYI>P#!AaFy}pP$(ib z9Z+EkWWeBDGNQIeoQ~yZi-p5BX+!P<#CR1ao6rG;Sp0Q%gD+|oc9fzo-VSIcsDV%b zL(TGqu$WGeECjD%HOq=-FIw29b;`|%Q{p#sFKF0vJbE9r@H7>u?k+s=ryww@)wwLa ziw4BC)D{(}#!b)1o3B$SmS&}FQx6M|EiItN*Taxak4rYxa~8#Mqquz%yk5LeFrz|F zX0P;bN+2l-xQMZQ=TT6LVw}3I?x1w*b;3l^d*ts&CK@U4%%}~-8+P9^FAIJy6;iP% z8ZTP?u!=jb1`|dQoyY$APH_w4_15pz(RW;8&80=y_eN93l{zf5lF;)oAyaYsqND4D z!yVXq#{}~sBFA)l?Tnf<)Jr_6nS*#D`+_g$d}7YCi8?hZz66$9AraalV`3mfOj|EL ztqiBU19xT(d6R^jONGO*GpdEeC|KaEEv_CRqGsA>9J7H*fD5>Iwcfe&)a68+ML@!R zN|ZV+O|*dZdx8m>4T_y3#V|+5T!}E#nQrK=8qn3D56K0U5xf_pd#z9ke#`C6BmT`> z(sU#b`orpVd58jV;TX2)MD~xmu-Ji8Ek{T$RvpIkq?fVG#%4$q$s^W`PEcD=B5m#~ ziYw}>oW^&;d#c|oZuJtzbW*`GY&`&to#y2NWKiPS^HF-hJH*3`P342BUyyH(az62D zwJfjH%gA&kL;_?gh+kZ?BPkP7$(`nIr|v1Axj#lQ$N-oTsSYz5d8(k-*l$ zs*h*l^W;-MDz(%EPk>!thLWZ?!Md4jKWU*l5s{jErNSw9YeG~J-8_RPV>xA~NCqG0 z2dySUz&*29sfdK4HADQ5{1Q;Me3FI*ZIQBEW?WfD+@ZRIR+~s$fYp(Gmc;;KGT{ws zz&@K81s=P7P&n$nV=pG^<17A+&dPd)wDT{Zj+}P66HeAih|V{6+S%+{kjBLh>%S*- zU5Cew$oG`D!4qQ zPkhH7nJe(kElz_19LGC9Trk=POSyH{ARXM{o4oJP#` zp#AErK{|+6AKFRO$hP?Pt46t>U+cEP(AW?Kjmv1{$D6Xd!P((?}f!-&Um|Jk}OFeor}Yls(l~k#1Kva5S`_ zuc!O1r$q$qKLR5;AZcQ_nhe)o-*ZfGic@5_1tr9kGS>A&20u&_c6z^ZfW3~snx-?= zeq78I5Jt1yo5mH6Tvd){2hnX?&i&%p@bu8`LYB{FXuR)aB%aV_S>(a zo;|JBOLYgIFMj-fng$r0xNno}dH>gmZ{{bwZ{C;7r?`B-0co`!z=nN`w3|cOgzQ&u znibJ*aFRp?V1y&0|L{x{AcBO$<1H7xw^p4FM)vOveOad?0TEmg+L2IEqj#)T*tuw8 zkr9KIcW4E|Q|`GmolYki!-PC*Few!lFs)x16TnJwQ_|w<9F(nZ4JT1XM~Y--tteh(RhBAR-P)qTMqsc>NguN;bD)2ydnD#)|yC(5G-*T4?2C zf#@F~jSY^_z>;htCX&JD>(-auR&K2%Tq4{@(S5&ez3Qs+maixNzdOZnZm-RYe1w*S zHOkrbgW5hJ%Pg>gjj(g$u&s4ja3#&RxAnP;XUT*bYA~O`b2-27CiF^_uv~paaO&+? zb3)z?{iDM5_`ZhLKtFYhZm`sve^zk4kMh%;fYAS>Y!VLSH1xmxqHtT^9=-9VUY5HP zFtqZ-wxbi9KrP9?`M(>lqlGsuQCiqno#3x+@{w160mq&ri5)(Vx!7)6?pAECD$q|! z7Sj`SD`4!8SL=o5PG}6;Y+CUQ!?4{7QJIH><{xCAX{A=5^ycY4*J^bZUb~eR@%tH4 z7H2h+X6E9=Ut3$*zW1Z5wzj>!^4}9~gf;ykH)S- zleAbntzFmk!{aW&i!hU-*s6?ZV<+tyIjLFS4oKcN4%U#R7_;OE|@Sj~-o(RiWBYr}xBv2@#uZze4sK;lzfm^!ztHr$rq-PgLW4zdgp8 zkvLt&K}PjEo`&FR<|cIeFJ4Adi@G}@Bkqo> zSqSX@Go^`QG%rhTpvt$mV+YbWYdSVm5<#Zp88-F>i5b{6J_H~&7Q~P7MN+EqrRe}H zQgduJHEH3}cd4{z+E5m4OG>1ibXa?Zg66{JFwD9GYzdb)8Tv}()aSf{+dAXq@6m$i zcT=<*(@Ebi4*h<|P`#8w#=(PQy#6GQvqbu zr@$%Gf!k&x|K+uD^Dl@8QyJ&0b~*IX%=L7c%cu*^JW%`Dn>TJ_cxH%+nJY$D9v)xB zz<)8tz}r52PlR{23hDJ+xyb$|qc|v%O!truxkNN11W0NLuMA{pS&lq++P<@0b-jmO zTpWH89@Knu9xAWa|B6Cn6vIRlcNyY321E9XjUyDgtqA`D?)BiUM`h;{3Bh5eRM=;) zk(*Dzd9U}J6&4f+M#4H5Vx2VYK5#F)Q~S<=45-JVc4vu_x|NX)G(9e&2rWR((dTrb)XxBcaV-NDzn}!ENvq?aE z3%CPvKhij6vAu@W<;_?jX~_at&W{FNnbM9;S|Q|aqS}dOI4&3II>T?XjVB7qKkLH( zYe8=44TeYOg1K`0Y}uZcKHMJ9%1Zfcr0SF!LCu|T_%2-vr%LA8y1_Epv96k1=c&FO zel*10cXNbSH}0_-dWwTiR8fwqY4_Rd(Mm8qC4H_pT3ch?Z_?=U_^ehqwn9ja?ezQP zY9d#f2~4Iu=|{c!LmW&9kR-y0h|0K519Mm{pe4rf7YK0eO62`jO!g8=GQsp^}1ss_4XkEh7 zPVye{W~S)uOehhKCRC?BWeQ{%Tg}Y1j^H$p``iv;=CG`=4bxywrCcb&1!)ciZzKt+ zwC3r5P-0|bkN*O``>$p=M~W5lZEH@36;yu|1s&q2eJRk6BhIh@2+T;o*TFIxrAf*1 z1$u5K7GBd;=L>PSvHQjY1(Z>65{MQZN-cKvk)J#Sp%K>JbINCkaQ-$WuP2i-B6 zy3EStt9^+c^)))WJ_w56qTCC9_jHdCEKklJ22+rkM^G&4&N(h{8q6g}>041k3yRWA zN@`l0SesSz_&bQxdif$I5OD?V86Dv1ZugA&R8HUWu^Qww3B3xSLH$85F$>g2f3A=C z)tln#!XM^;0lvo~_z_zD63+1#+f^8X3YexhGx=FTILloX{P!b|a=8xH#QW(9QGTNI zh10UAV4C2Q@&Q|pPON|YT`{ycC@v>%f}3yX)u~vrs`tCaHv*dJ5WUPwYP;)VaSmNW zj0E@HHeG$prXcAXbrwdW@|SnTM#u*n*jDy820XQ|J@~69M@J?i6mY?No=EE1BbxRh z86NyVqp1L%Cz~^rNx!au{C@#aFvHBz6RK0z@;;Me6EL=Kx+l+iqHf!W{2LFN$nJj* z9%@pT|NK`?V64t8Ro$UIcDniPpyJiQ^sAp8{<-{|$Z^Yo8YLBGxwvnolB4=zz(T9j zH4CI34!FAr%_?1UxvNNrpYR(Gi zk*4o6YCr&0aCtl=7BKf!5?7e6cH!*xqA9F9J_oWn7)hbRAz`M9SSn!qgr_SfP|uLw z!0ihccl>M&PCAqW zjY7FtI;lAuv*girubcoEn^r3uEvt{G9K|-zn9zN5DK+`MPicaq(1pyV#%K&BmC76G z_kg=3_+d7Y8v5QRgL*snK3OyNfcLV&agjOnP1p`M%ugsN30!U|y!h|DuSd=oOafaiY!KTS6YsG6acK1UZv&lOs@ z9m_oAeiZzgYWFw0IFr-TUq-jfVR7OsGbW*m>A1fiE02Dhnj5!yDLH@$vuIZEKgaxY zwv0Vw&FJ0JaOx}GcvUVDoL&Zb=N)MVDIg?q-BlOK6EbHKWvA8z=YG2Cl0Xa+CMNXb zZnzC6K}@As>z%IhY1;P3Ejyg7kE?Sl3){VpBlv3SDx|TYKQu;rc;o(>vm-0qI(%ic zt7pig_p7lYlWXRx>ay!~{*;|8WrV{E^$G z5Jk&#snRQ<&i}HI!jyFz^ggldOY9O`7LCkkiu+rF6s_17$Go8_Xv`=|(vIxWgPp_& zN<`5plAXFt(i|12{vs11*;~zj5Vg>_@pR51k&YAo1Ob$#$Z!W%DGO#WH%dl*Y14cg z!i<5WE);)U3s$9S`Q*B_6wNMzaHfzJhmHLO zD~@ks@cf>s(b(SEp#OEUoZPGfr>XNVtovfnSB<`zc7BNhnHm8|t8{9SAv_*}oYd@n zB09TWa>+cEDlgsIX)99j9WD#h8m8C`hyj_s2ARR&MvT|DEhMCX3Y9t)t$oydPXfhZ z;129H$Z6vxScl0Z-s3zrVXh9E{c4B=f95B92j}92d5>eonI7py^8W{@rsYsp8gw}kcm4J3ggh&K^h>J2L*-fq3j;Zt`$lqMQ$U9pJ zQ>N4?{fJ+mlfl7LoThS`w4p!jsZBGDIG-AFwKL2}{af7lIj@MgQJ-4QFf=Ze&pW&+ zb)cO*PLor%A<9*v!=MW_!ma|SDv%?YqM?puZ2-{Hsb>nI5@{v5s7S!|yNasA2_bq^ zCrKIE0PFUqjKTT&vwIMm;Z|V8bgAB#-0ud8j!)C>@AJu*aX`H%{LB8uiWK+g zBRzOtq2{y6cI)B`H0@eHKdS6k-1{ds6HpM8bVkbl;QjB3`@au_02r9s=Z7-7daJk% zeR!HMh+Q|^o70{*>ia2mAJ&vKIgI!MP9(+aR{AkuL3!dri6UZ5=080S}e6i|P{uiTa=B)dPEckwWhi_WU=pXt&*vRCuME7J_ zj}>FUvkz_pUPg9hkh(UDYI$1x)bBO6sm91o-+XeXQ3CAZxRivSDI{ih-cV5vLh7wr O3!nh^(CeVTi~j@Iqd_wO literal 0 HcmV?d00001 diff --git a/en/2023.4/assets/freqai_algo.jpg b/en/2023.4/assets/freqai_algo.jpg new file mode 100644 index 0000000000000000000000000000000000000000..44600e71a7b969372323cd090425bcbc8453a61d GIT binary patch literal 353611 zcmcG01z1&E)9^+FluH%jf;nS zMFb8WY;zOgHUh$JTr?Ck-2X9rZvn7w-F$v~6CREdxPb)+j|KO=4Il%V1X;V{_a`7C z-9)$ne+%w5SWE&I{9X#c-Mj&hfQa;c5j-P6;_3;Vu*k z3H`GR){ZoyRTx1T>+`f6$#I^Ys6nLMPG%X1g5=b&&GuzFDo;-j^z(IGhQ-%&iPJv+ zmLZH}6C3pKpsh`6%~TVlPl3*YVaGj`!vD6--xapz9}QLF=C%=5|4=BAVneAL!gH18W)GeQaxCr6~se>Z&pB5&kmzg_5AeP+qCk#O2` z0re~o8~P831ouW0oDf1q#Fu{JQ+54Z_L zyh|3`s@-%$KCJr$X5Oeuw@NPbGTHIi3)eep>P+Yqk<%?O=f4FE##tCr9xwl+TOY;} z&$B*iSd6iMxi4fM_#u$566ky=_e3xTA|&=g;Rtbt+i|Osg!@p^$lGCJiCzbQ`*sQd zK9HQvfVqbo+c^WIx6piLH91f6HM5n2BR7$B?$0Q;_7r%mZ?dJ*dnPc%BIc3ym6i0z z?RpHSTc7wPt>l^egfxS!x{eKKJ8rn1wK`_Z*w;M@_zGkgmC#yyS-Y8kgQ>e@)~7x(|<16Yg3lh2NdU#Iwp#=0 z8|K`Sb<~#MK24R$dApsHS6?R%?=<9M+0$KS7Sv&d{?3f8Lrv?+!y1c8Mr3>C^Gmnx zw~Cgt^!x3MMc&q4!L|Y&F-t)iOtt3AV+{%#ES?&R4TF;%yL1|?|jn?%jY+Q8{T9j75 zISf6thUnF6PVJ9n+|$a;5f1>)Qh3*|#G>mV@;&Z5#)D~33) zLk4rovvI%vGF|(|xH^q#rThd3Qx~Tl;=RUoCYK6f9auwuZ%}MtxCVriJZ2pNS)-c@ ziw>X@f)srryX-Byy2^79clsPGK^oc}ba(~8f0YD9OmxTg#GtG$TmoNfnwwf; z`26$~Snhnw*QWa`jemtBMGL5a!i8~nWsdqNH)yLy^{MHRWiq)Lc*=@~c^g5u_-)S8 zjm5)Six{9V6$GIdRzn|hYywEkwP4?raXeeK#|1n+hGJb^6Hi(SPTko?ObXHR%2chC zl`WJ{q395+B6;bh*In6jeK?^ehuu$$%F&du%$}hoNHdJ2)Yi(;n&igqR+Ctm=bb5~ zP4hi5X6tB}*Uuunc|K$;fKs9HP~;)!8~v~d@2Z5j1PI-pHl>Zpa#l@Bqa|$YjIif%K}EEljoHvBy8A_jc5Qf1M?Z@-i8f%V zGo6svFfh1t7vEy)mYMS5BRs=2PeppGCFizI;Yo@RGeNcKH|kyL#VVkZLHc>fvx z?4F3RT&*Yf3NAF=${M#FB4v}e7FT&#zk29Y#_#bB!BuZ8#`W}vIzblD{aeE1jxd#E zw-m2$GiJ`3JNS_Hj6Rjz4jazO)TVhcNSXnGYP^@zY7IegMt78?xNqIK1uilMSR~ zPrOy_tkilgb*(YJazbkWSy=-+W_4vX9pedUr0?oDk(iP8ehZS8lO745JESp;E*U-G z3uCIlN*Bswj{E(1F!9!A%1T9o=b{3)pUoN?-!_$ygnXd2`1rhPETNk#J?VA!4l?v( zU(opJWA;O7N}6ii(QoSk+m+O{ryo}`Y=)Rdafs`vdpIyH)af)ReliB$`3u%}Sx((e zl2HXH0Jx)ANK_zmlGhzCthjri;}UU47#s8(j&|>4_1j{F$uc!|TsK}j*?+OKIy7#F z7559RUr;o-CUd6_P6Ffi@5W-tEY%){&PcAUde)iISKMuS4_W{UBVLLJIwB!CRt0!6 zepFR9P*3RwdPfBk4!3*uXH?Iv2JWV6tG~$mDTA-BCZi1wZX*>57>uV;5Yen>;l&)- z;}JCu7bW{7D9KLvAy3ENqtA$Ig*fVpUZ&n>p(iX$mzP_zXe6V$uh>w6MwxUdFK?r` z*UdG&d;)M*H^i)Cz}8GkE2sm+ zJoEQhxe;8m71YOQ-(^{dIWZ7rYuVnr_|@2J6AWuGAjr)_uDp*dfp` zKEVLyBH`>xK%!osaYH>$TdQ_etoa^AT-EvqP{#L8X+ZXoyme1PEz>`P?t`;gnImrJ zwCBT9=ICyf*W0u0J7iNE{Fb3Uc8j%rAs%4)LPz`PNA~c>e1VeneOAQ?BJNBc6`m+L92Znu8Lq{eCuYB zDZP}2qCB+5w9sMnIM9OO(Q>@zN(zUk%%(lcmmGIJG9QB_^_*P{T$$3v|$vJzU`-x$|M<*+0m3~O)+DKh@eqH z8rl907Tw^x6fTE}R4AA+*xJ^Iu zG;s^N?ajhmgJIJVQ*AUyer?r}5jr;b+mJn&P8x z7T*gn-0&%WvL93yBm$}LzF0l`IF;QnZ%O~ewn3$TfA9TPZCg|G70=|#u%kFnw6b1n zOnia(qtEvS4C=*vgn=8tO`lty*tf6h0C1#U=i>aOC2Z);J^NDg2it|7N=szxVZ( z%q8wc>q_rD$&^hkN?O3oBB4rCvh|Ik#mcY3&S!Cnlx zN@cfFYE%qFA7#z6DD>kp`YRIP$?7}AMfC9mAndAW+^2O{!mr6Is2s3=)*?44e)8o0 zQq0TlE@leXCP4*OiZhqcFvV1AA!&bC>leJ6=obS8>ovLZ-F`Ou;<1!U(`sL@IstAt z&OO@6SH_6+2vSOBJ*7>)KivdIs=<2zh=2n;Z{PF*tAS2|=W+PU;IabLF&%xR)?M^Q zobWlnW*~4bRjzxt_0D^z15cw#*2(L-zh2;!^ww=n-ARewS^3u?wq-H#%?fvXnRptz zvAwbCYV3vRv5dG~&ym;|u5(dMpMP#0a5QWMEr0Fgq~4jSZZS)^0+55+60|q_7jNBX z%HO`Z)2$5wCkphS(K&l0IuYEl5Nivo^ZwkjzPgHxO>PL7BD}|V;MN2Y9pxB!YhhA+ z*Ky-m;+6xO{iHfE38fEbtxzS;kp}9i@zcDZ8moEZ4aoxUs_^(2Ew#2uGXb*IS2h!< zYn-FFwS+1Pf9f1q~sa-Xd=nIx|8nHz~vK*FGpe?-?^IdR6=-$RLpi z27w@rJIsAd*vKco2P>2bcll|0CYFs`*UV(+Dz-W((P1fOKUcnaZF7VO+#MVLF=TfH zeSRIgKXJHa8<~^fp*sFGy>%b`Wx_E|&eUi8ClO(<0Ut?>_dk{$00-U0y{V7AJQLm< zXd}l005|S<*-7+wt^vyvLlY!Cw|;Z}u8DtOwJUMe5ku3ujn^ro=_#9{JSuHvXuUE- zrbe7A=*%VbtRcR#eeQ8VezIQK@OGY60;hsKX)Boz@RtXWFcCVkrs5uI+0~1>p}(3* z%}|h^R=0X+Y3TMCC}QB<|LIiUmB_!5RwbszK8GAdyoBqqVog`TF@1MnO3GDVi zSjErJ_ctG~*!z1ImWydM#Uu0CTD}=-Y{T_h7{?81@t`F-uYbPpE~>)GLi?-`TScKJ zQ%|eCa6oHZ)89M(A+A_!C2k+W#@**A{rb!BeT{FRMGBBvk^-j9Q8g@U>q5&@!Z|p^d7}0C zA&ERhvAAsDvw*1^I$z+b$NL&uK%H6mIzOudm@;Klz zg1|h-X({~~92~wd91M=)k6Me}!I;A6nvyFts`mg8aiv-_fPvl_Ss&@%^dDmVNe}># zaR}%MqG8WR`Xlw1na`WVqNBQ1-qn7Ms-_VcvNV!Ydk$c3FFkIXQG9r{5PuqbL~>J+ zwGFrMyTf+c$ud~&hlmW;lF?5fTdmH#9;H%amSVpe%@(dFyx|(-$c_C-?_^bIL*C8; zH`V1qa&Y-OX_*MIbJ#IHl-TMl_4cg$z|98jtgKXtr9a;I*X_6Ufk+{2;<9I!{t(LpQ}82b!~Fp6{e?D6;EwByRcxL)&uf*I&Z#FXB;G zEc@#G+77j#Yrg{#uUu1OH52kOZ+nIs_vPIkrS@pf zeVFt?^aw-81M>}po#Ey5^Z1K1VGhBsUS&OE_Op~pX40D>N8d#@>gJDPERUjcw;Z>bIQ?3*V5 zI9+URb(*sMLHOtGhF@C^FP$4*P@u!2K1xUTj4JdYzDLZLFb%B|>ovU<0Qo!MK2uu> zGb;s~{HcXTIGN7SX`VbNp)YX%=m5HfRs(C1kgbdlkv%w22^Ez4 zt`0ofF4R*svyRj;P}()anQ?UdsQOxpqE1n^ z=~e7oGS%4AS5;4L1eSNjI3mR25Fy2E+x04IiG76cQ>Hb1@Z^i0>ronu+b4})>4fxA z0uOC?K;W6S$1wh6c0WPsT}kp3S7^Up$4ADolr1Vfik#(4OS-%=6j?{%U6j}4!j-g} z&u7Gq@k)#>v-B2HpBi9atR`Pz{<@-G0zOZ9C>KVuXC4`qrE~2ETR|lI+h2O|E>Q&F}V-MVAVPI0=7w99$I}(qmjFg#InE= zwqJt^OP2qXy=VU|YgL(nIK!ee{ewok?|`462Axl0sNxiA(MPrE*Rj(%w8$_9bVS{6 z6^l!3BjL`GP4z}?2oJGPHX6pTtS_ds2@qbo!K}Dqk?TunQ=cubt{CVv!uK9E zg>P4KRF#|)ZR`6&g57%dQVZ+rHmUk-^Rxp3%RO?|=D2#klwW?_r(@g|JQIN+R4Jdb zcjc5D+7)rc=Ck41;6gq-reYaL=U8Owt=^Gv{X&q8VS8d0aytIfgo+Z)b!0zQWzy2c zvfv3!P*M4Q&^^D(#JExH1DlYG#3g0En)cRZyY+mw?I?)KGW~`yq0e$#%6!<@pThrn zht7TAsJnMV7yu~P*HF(Kb%zYWqFW}Z0#^HTd_CTtBv31Fo<%p6dZ=>-UvyzWun$~D zQ9o#jY;GS=pqT|6j4&dgOVbi3>H;989^Y1;0#)_5v^5_J#T6&g%=!4OPrHoLtZF_#T3W7 zGLJ?gvK7&tRz6VrmUM>AHluoJP(sb<(+2Q)?Jb9a;og4+?)_02R@Z5Hm6V7v-Jiv9 zBa;uLnKK+NbOdA=@!fKl)9C*Hr*zndmBZ0i>1;RMm7}}Y-;z(Wx$C*(xKV08rch&jG<(SLN{LfXDiJ@XU;7L#K)gv$>~ z@T&u3V-YeW%!MW&AiflU?T`wo2X%4s@Md&~yiN>^Xu&9Ei46{R@2?jN1{GdEcrHUX zF@?c65om*u*74RE7zA?i?B(Yl5T`)^yZU@7bhpGa$4z5RQMW?qWv3mlEZAE+`MBTk3>A*in!ztv}l8~$qVm%z!Oe_aLQ2)@A&*x7H zNL`6~&@R0wT|4Lf&dr{Adw!F-sHpGn!oud*H^C*PuTpvdX}&*U96qf8qVHBj|1>-CiH} z=hu1Jx<3T_ayc%ieQ8*if(C2^9X^4Pvo|uByVUV7rd1+$T?{$l^@gjuQ9uuO)SHNLF^|63b zvbnLwwx%G|^sMGjw4vTkI^>k{r6BWWJ%x$4*+c+iN2bt|y(+TX#EfzcB!-BI_L&95 zwB^)R%J5kc0R0q7D0jhDZo_~ zNJ#=*ByR!vvhxSFI+LSZwvtl3t42Dhf)M^iL>TVtx2`l}m?GO;>jGv-9kB{+h-ZD7 z;YueK{dkWU%bNe7{$JeShV9$uKQBg2d^1azG18cQxJli(S4e&SY|hK-Irx|{ zO!@8@s5Kr_e+$#*yXZ)@wKk}|2q=mo(Q;~)e^ST6!^^B zb@}Ed(hWSSF&oLnx7;@~njFu2tp)in6BSw*yYY<+zZC5S`8jEkH3`)AY8HlwXP(X| zPppsY8;k~x!)jVxF5bSd4U?o~lS&@wZL`s|KB%dAKGegyNMce}YJ|&4jQ<@#BEKw5 zlp?QJJ8;4`o9r%xzq{Pk!OA>p#ZjT&u%w+Yvss5t-)KlUD1den-t%E90qfq7wZn|r zfb<7;^I|f7&{!7KliMq)!dl0W5pnQ=mzbD;yE7)^?5|iM_P2U*0Qi z-}{tyLOW#nsDJ_WJd=nmK$_Lx#?X8}ZWO8)6{0*?fLoM`)t9qNGW7mUpIxy?r93}F zzpEFd=^l<_$v9^H_MF}*E&lSH(*>Pg$U`xX2C(Jd7FAs`o z^sf~Jgh^hSzR4U*FYH4XNFM|voZerCx`X}x>icgPqELXq+*j!AZ<)V5AfnDEzm)Gh z_2Rwgf!3FDYJ{cim=xXgf%8|GSkmw~zvwzjg0bb5^IWDEo34Wmd5;%Tz4p7gp#X3P zjl538SeBpay&x(V)z0%PvbfeD(_w=hH1@M_MW%FJdD9_!Jntzu=L>UEG z&T1D|L1Isc_7*YE+`?z#D(g1@dq0vc#_89a+w8CpmgFtFs0wPJ03bQ|KJ2xS>T2QZ z`sjYn9rb4oYqr}NrVJJTMXKLORCIf^A|$O`_t~`iiUtPKS?j0OE2jSzXuG{3$wGJ9 zoIGyfOtA$waO+4QlPAFqcBF^cEp7;MlF;YV9oeR5xP)^YGy`x0P*SIhE|3!54z$qq zYvf4jYiorh<^c7fgXT*jVNvEY}~*QjT`ZKv%gtLcMbjV9xl`ww+{*HFH# zg*pj8%Tg;FA^nPDE?Q$4hTs{lqhsQwBVIpML7b+V!3gi9OSD10%xQWlSNSRQsWd~B z#Oo&u=XSj@){Wd!lccSNK084aAdgpIR>f;RkQUX!ne;i5h5d{NQ5S&EOD`>5UNjHO z1ixO~b`aOpq?nh~*YHAYx?r$=w;W?l$7dH*A>wEz#-+N9 z+0`~d2H->C*N}HVEN_08 zKKXo_JTjxQx_8n0Y9121^@*u5_dDRtd3%H*==4rZKv%)#tBJhvI1Ii6UBJMzlWZ#j z3>*}67aC_8y%PTvG6x}9c?7+-H{r%6M)iDCteKv-b+56$jG7w&*BJc!!oyP{+1OV1 ziHHND1L!G0dKV&rQBa>k;8iO@+H}2NjHGZV6_&AnA>>cr*nLRSr(A(*wM)#_bC;Xv z7@maYW~BY(CQ3qe$)hgk-e0rY4!=QUAR%-eCk>JJr6r6%)5 z@6QmUX3Fq_Hm<=?)qRfTZ?&D-Jplha(`!(hkek5jAHNs4^L){eh{HL}+(5!TMFiDQ90>>s<7o3-#?q(N7M0kA;;Q?W_Lk2cc*=3buEq(@1USB|P z5>*N3E+EBEz2-Q&@C9)dKNu{lwc4krd^caerlNgW~ zw`aa_RX1Nz14fc9<(OclTIaQT0Jm5X?d=W0Wqs6oYYICycVPg~H-h1nr_s~xA3sF` zM^8{inG@j_bfv3v_}I$HgVG4;F#d}(c#fh5h3y6zE8h9k1M`PCdU=bL#D(Yc7!j86 zI1C00TRy~>YOU#ElB4sx0o3HvaA{M{Vf6w&qb7&wg+%b_sk)1u#f~RUgQ>{u78Zs+ zAd6Sl|7{xQ}CWrPjuvcu4LWVv!t`LlC6-#Z=#D!G*Up(NS0#sQsT_D z?V21W&ni)-Z{<5W?Zi8^IGHI$f8Yp_`4cn z(N*PA`!wxkn}h+t@nCdF$#8#%Op|qCY*w>0Jjv(BYcS$m)3tZkkjz?{WBb7@Gk zBZNOC03hCa)1~~h`LS^`NR4pv+y%=901!M>Dehp@&V+Es!{hmWn@*CEfKLo$0c-r! zmx+&!M|62z%}%!UJ&t40{d|y=xEx^nB_PK>zinG^;Xuj6dR|PhaYBvhoBx>Vcw1@m z$_GrbuU_?!`xUZ|k^6g#;xJ2lyD^#ROJ|tC8yzw2*H=}40Gh`FQMcC?5U~Ad#^KD1;Rmq)`Q9eQ2QwuwaL2^#M6%`#5rW75vu#~Y-~k0wkBAMmq|pnleMOE+S4HzIEg* z5)E*gGa0vQiwXvZn;}8(U@-*daR-N^s2Znaj;F%3-7BPG4S1_XkHhb~jX^pN@33vK z(tAH1Z4Y{M%8C1gE=WqTmi~4K#hZn0|4>MJITYsW{B5?yh*-a;_}Kn>5Px%MqU5S9 zLX}5p=?hWbB54WO3~3vn2tQjQ^iL-cB#Jod*F z0^ft=dolm3)rOm2N^2i)u(sKdsb1>f9wg(ocSfT~o<|;ld+PH;xK~#nS^7E;o*e}p zje=x^F=?xBFy1p_ENp5X>;u2TPrrL}@E$G#1G`*0@jto2tmF~@G!UphibnkwFX z2ej(Za9gUTLY&cthdCnzf(wF`=iO)EhhXpNO6B>!b00tHMa|>hX^~8xS6e=>yZt$Mqr(NaS8*u$h)Ts+4L&2>WiCeW zoa4nyMd622vfR2W@1(wn_`Ximpi!9eZL!!g{IZ1g>w4optor2fH=+eE^T$57*>AnJ zXxzH#zUAZEC}t)6+4;R~sqobR!DUXjKIkOMx086yi+$7CK9b-AcBGf^dS;=zMfbd8 zxLh3I+MtPRYL zL;+`TykExBGUEY6-0<^})W>aVI69HTh*{g96%ofsxLJ!JY>xXibWRId)h^Ri11ndI zT*I%L)8VTOpB$0v#yQksGrhH%>Wo;u=&_c5C!KVHTLYHxG>$)`t#BpJroS~)AMep8 z6M@_#5w!OJ&?c@bc}YTHCx@ZA6xN5cRC$W<_X@}3Cu~m8|s^A)3#&RtZpT*rm`KX zG~&(p_Z?{F@!Uz+TQS>FEpKAN8*W?8o4x38qFK?I1DWt>=d(Tbt-9Cwc?kdW*+ESQ}VWTH{+cc|n3ynK%LWp1AHp0(3g;5lH3l4WGD^QD zN!=DTw|fM2z_-Pj=*WL;t22)G9iSpR58$J$dT?)tHTT{?4k-us1QAdDmh-{Z&TV67 z^_||G0Ca7{!y%P!g__|(8A|Lk1c4Vr)5p~k>3^QFfcyF#aB<;3^0|`ijm8fG&MQ0N z=&(80!?gmhyl_cp<&!A&SvBM#^KZ}{j{yBInbrpu{p&#i{hhiuh+=9I7-^pbt1cIg z2Yy-SKQNGDpbPYTmMyn=P45ovGyOLk#}S*yFQh-w<}fz#9o(a{cxI9-$oCy!y`f{} zy4X*u<2_e6-a>nd4EYFax_dips^v|Y>t&j~|6F+MDTn8+u6EcQ1==wFedm&afjT<~ z$3a!#8sAoWTSX26g?!~dlB-<~mUBUp+K?d;2 zG){*}=;(r7C%`dJu3={mrGZ>sqk0GJh24@>^hsC{Crl*nL2gCsf=S3n4@O}rK`(hf^YljpS66Z@5u42|2 zCd-ZyW7@tt%X@MG@YcD7jX~Z6aCp&T84oF#yYD1(baia}EY}b*1;0WnjYN2NNT5>vnub8F~8Os&52#d%f)<(mX+(m7}b7?h7rvB>eMCzaf9I;1zzIumw{bg{;IKa=m%I|)MuD!aFYc4 zL|6{H^M>fzebfb>uE$+e7xD}|5^hTk$33#|di`1jr7;6M5^m{C#06S8tJ~`h;e378 zf4N{3nK?Z*N^&*ypg&<4>fZF;aM6?E?Stv4?)eAg=?~QvKu!D7OBnvn9lc>kBVBuT zQtsr!urlU6TqnoFB{I)uu;DxebK~=Rub>Zz1B54JNH$E3&ET&`z+bu5?sW>1*2x6u z=mkwg@Ye2&BN?kant~w3$82tam3jI0z~I0Q@C@~b57?|;Zf1Eg3H!$ygaXNe>4FW^ z0%7J=nCn4*IICCG{|0_@F#LK$DgyYxTErW9%PF1>S}H>g3C?9MqO%Qa2Q~t}Hd*Bp zvb>IJLss_cRsCo5@r_ByZl;xlpk)LKid_Fij)9=heE)vAbZAzhWkCgLktKM1zE}eq z3a+u$b>+T}lJ1T#Jc3J9j9OOKo!j5ciEd8HSw!aaerZ`gLnj({9jMAJFTaa_tQ78; zMPIj?_lzc$+@^3b@F43$H~T6DsC3T)Y!|k8jKCKP-m!K6_7b+$eF7U;&#Zd$HN+NFU7?he2=U1Xq!xcS6{wBb&6}m@IH*c7ukuzTK(w(G>;&YrE@Y<4sawN#22i z=tFriDth5l*BWenhfv3!7m{3^gt{tW3j-WI&zg6@?+1f%4Y%H*Ktb(VJ{H@u=`Xn` zm@Y87QCiYB+y94}{XzMkuD@HlN(y8ZAAOY~wN`W6-aO)nd9bv*XInWopryqjLsGh0 zH!et`FT^ZQW%sz^Ye@4?V%%f+vSia2%Cprms~QgUPQpbdwH_oMM5;NnaN4R$y^VN* zGkk10=1YCG|avo?mMsNvMEwDJ9n z(G)N+3uzfQQ6n`maELM8U$bPWiHmj}i&l{eE~3$qC(w|nM@4q2Cr)UtaZ#>_T+c9w zQ+*oaWYeFMqxFRz<>QV#&RfgCugUD=R_1Q-$sBfXo^HQKxllFjF38VD>8cyu4u-qw zPv8sBProf^ApzADEz`qJ!w`1Zv{)RZD;@ehOyluj3er7VaQU#ZXf3*Jd1R1;fWZl~ z!LuoCS)uV{ET(Zf$XD$BU*DI(uEQ@o>z_}FJgG0bmuPO&u07VZ(pgmSZHqwGQKp}* zUw_?X^jIq`%ee@Hu6jzM@oB_%CZ4-&s%>9*yJZti6PP0W~fi#@&Y@It7e zmU?wN=xkm*NoNy}mm!@=2 zDM#qXED^U9+lp`QGF`EEUf5Pq;OnVuLGy-$tOyblwsI<5tP2UH#-YP+9$|3Gb@Nh- zS9(s_4L~)|P^;J65B+$wOJ3*XY?tvqZ89N`j~)pMFML?altW7~JIM9oJ@ctgs4B~h zt#A3HL@+VN-4!tbG{yC2pYkW(QD?SK;TC+jRoSbF56ON;k3nUwMDCwc5b=w*0tLg^zoD0Hj+Prm^(BGol}$yYj+Ui@#53x#zii+R?JWk#&6+1Y~K~bpNbaq^P!N)rj zs58ei$7^Bje)71|Iw9wo1nS+%xUmlvDp0Y&j1rbHWpyQQGjpXV4S2_hx$96Ow?dg) zddF9Cm7BSAbAo&eVx;7D{JZtDX8_MlD32b$(2nZ&sHY)KYze46Non%n zp1pXP`r5X4Y&F$oe}k3pluhOv03XP&kX4n(y#&{`JTTOJwm0eMe6%m9XQG#^3)$=> zg`0~C1vj@VaA-Hi%q1~SyL=DmFY70uZsQNdgWt?>%8<7ftT;$KP|q1Q*{&+0hh3%< z2q5!H-Iu${8(wzR7jLP*S36DMuRDyYHr+lf`VR28 zJ%fA!NUN0U{3MH8Uk*!h}u>6yxRp8$I6#JOVx! zig_2>E6hk&xd|iM39Rn{DED92r-eM^dt;uY9vS~ps3NA@mNYK)(XZwm-nPHC`hsMo z+e`rd-J0VQr{hwzfCOfh65aH1!tJ<+lQ=009E~O2krl*lGiF{*H)tw9uF@z)7xoRH zb|o1-aF&|qGCd&Nj-Mw~EhsS09Ia|Y77V#&^fw`0=ltUZJl~0x$nW;TP3WCZ^@jUg z>=s|=ojG2<<2WAGhD;q<=SQKjK`~A~fS~S@LGpx<2p;{>Alf_0|CexZvnF%4gf?XWjGn)^<gYB{L zBzH1(r|q2?32eDm5h6z?qm|XHC!8%|8#T;b4y}8cjMKey zyBfVZ1GQ89-7Sq%lFxSm6}ux~7< z`z2C4jbs7T+SE^G0VH^Ws&wvChY%I26$uj-kOVrJ8ksl+-~Ls!2~>+*sK?-%elnBL zr+k`woX5y}66loui>9+NA)k{=x*idq28t^dIzPAbIPHO6ut7L!YLyq%UGuqemuNAl_y&Ispuo$wj|;v=!Bpu z_;m*psTUOWmA7R?pdcmV6$FlHkEn4_584*e2OFrM&_sk-&{!U}tVA#S4e_t{I_s_< zOuvpnAHEV3Vbs7+F=2ht#>no&(HB{zq4*;_cEv1!SY?>(&>J3&wsvCn2EEgA9tp>f zR0h+Q`%&9JqEed;`BjXX_mG5!5c!&~TmR;7UFZCAL98;8QCmE8oaCb?opiLN;&rq| z{@Y|l3 zr_a+)+tZJ+8xEdDc6Rn_y`!1!P)YSf_J@6tZ+!RQw^aL}=xzO_a%H%7lk9YKgG;@= zAfY$AAfaU;jY!o?=q^T}4b#gg2&#$Q*aiEzW=JL1(A|E%g}3}tFGeLDPoZW=1z95- z6Di)ay9OjAkQhIp(PP;|qS9*ehM z$4KaO%!>MK*@N$h-uH#7u#&`5t|7mcj~350w6Xw4Z=FpApZ{4jHu#IP{)kDvz0r}d z-W5(6YZjg@^`xg^0{k@`6M?3n^$}`~@_)nMH--32A|GDHU2mJRV_8*pa zxHSin($gy>^?1xu9tya*Vc#<-wT)cgbC?*7NHSntV4Miso>j7E>Q>Q)inXvu+}&qz zTD|KUZ#H*e=C)|~Hm`hR`+SO7NTW1o64m+OfDclwd4d{?lSPV}l3v7z^k4@uExg#( zRb$^Oz%Re97bV&*dH`E6k^aZhhdc3?F`Xq_f_@wwqAXzM~p0Yjm_SU_gK`GBt*{{Q|_c4T?BRZy?CuK7yt{|pMTWL*9GNB5oeMPUwh&dJ+c z?}tobQAX-hu25?QgWdK^B zvq4soQ|7WUVAUZS{8ZGSQb5A`^ci8)M6+4de2vJFo84sE+lB~JUdNu;u&}109SHO`U*t?oQ5Aj>?HJ8#=q*yfVvhz1_*X{ySnC7Omf+M{9TytwubY~s}Lu!h%UK;aO+dj z*i@mRjc!d;?@-|+S-eUV?_v;slX3#H)ri5O@k+D5YGm&dVzfim7_uiu!=V9HhC4>; z(uXYC52f@@)U~jTDc!Z(78IQWs=~vd#V}Xc)u-YWmRj4()&_Cc#N+z$#Xm#PSFgRP z!pMf2n>h-RH2+s>;HugWxRFs=>kOrOYrZLK5otq*Yhpu(^>32y7)dY+qYj&wqrIoG z3TP5Ro5ERMu%7x>F$c1*AQ}@wTKa&3(h|G+-ic$_aPH1d@@h=ORQQKPwN?%Chz{7Q z;n=8^inEHx>}`{lj}c~0Ni!F<@r^1XQuuxOmfPcvp2M(mr_qyI3=XGh@N~V=h5AXo zj(gFA!l4!8oUW*jdlzpuoEGx3w-Zzj3p@+msMXQxDO_En%&|J{##>eu0+L^=HH8DDBcJ!Zb*3__A62s` z+v@Q$u1ad>DtybPGwn{ZONch>uO;L}O!K5ln+e*Q&b>GDig zJZ9|n0a;lt{`MT><;w+us%9caGog_>TSglv68DVi$;h_j#(3DrxSYk)JsOHcs&_DD z%Q6HKwu9W|Iqn+>xY6Dkw6yaR(bA>Wjp;;cnn? zuq@$<>V?7OgTwcc( zi4Hv0%xdn*)7vsg?hlj#u@aMheG~ZGHxXl&iImAlxu9eG-@r|E1^jnAtB5JrbH`DC z^_+!BOG2TndNBgfdx6?G%m4DVO3h?%YJ8m#dBXJ!x*|JlweW7yHX0j}yBnc$h&agn zAipBNfs*TXrggZhYjDn&Ap6mK8iE*p>LOJ!fks&4w&~76IbVa~J4@baD=Etyk1yC* zd}&lDB@4R#CHhd$EKFrKWmj7Ccc$U-;P~>rh3NOU%wUUV4emTVnvT)0=kBEIrl0)lvcK+29B0C{0SK#i9 zw%M4~L&tQhjNr3qQKsg<5bR(Y`&4lG1`|Eetes4bXi&F{Otc25NJge^tx51)5l_Vh!BFZeoMlPsVM3 z6OX^D27yTtB6Tlvf@9(>AF6FkrpZB6iNf_dY=qm?jL+rpQ>hQ)T^w?ICQM5t*Q zZM~7&7wZv?19RmW-k^cZhD4+JmJxPQTI_RO(?;9Dz4NN48mf#J4)6d`-0j8T=mCdI zX0h-~VTW5g@oK8h9cx{@Cb3?{8%@wU) z1@{Z1+wg~shzhQzayQ%S81p;7NEJ7|sr3YKQFpAO(hm;Fs*bMzFzt6-2b&EBG1RtC`4s;mZCk%smdyk0n)VeWt^jB3NZQebVmIb`*UUfd~)& zO=-~Dj-h&s<^7LJfH*>NL7}>Q2|vo95%@$NkZpm6sg1vBE&b8E_Y_v^OwGgPpYj1_Gs`&!!zWl{!WTV1_@;}~r z@pZspJUe_Zww#N)qi^J(JniYeud3ML!lKZJJb&LO+-)g+dN`I{=(+)FuK$HyOWpo9 zgZ-(C5XGhb*b(K>ciQjTuLVwZ9B9&o7gwU|;eg@9Z69TV+oew(Tl zqii|kmGB?t)!l>U7O+m5#S({MEoiLh!wpPaL=`izKn2*p6Ey2pjMT zH{cy#LKufJy0yBWR_DJ<0f!{(#5f7>+PFl3Yh%aQ2>>M~#wY06zV{jGwX?--atvFK zXjM(h9cLsD_euDU<&xg_m1s(=wI7*Vsy$LRgMo>Ok!j$ArW79Q{$7%wj44k;Sogm9 zUB&{#5&=l^(qF+e{?Q^H=IR5$Y4?W-qL+dj-^lh{1-CU6jad~>bZ;sBmn7#az#ml| z0O&;2MvsFxdGbH7F0dv)8G{wGL3OS97Q(lkt6&EV9+$I6b~_iUId^nygoPY-`?Ij@ z$lRvM9Ij)v&b;Lg$#a(GlzrjR*D*HL>P4kgq1~fW8fHZl4^YC-5y8Em$oL0bYary<`*88Xq|}y>{woA zil(0W<(EM6rEiA+7$wEZaClR9wGpG4v{#fxJ~NdASNuFM)jw#QS+rdots~@P=XkO8+^D@VHwzR4<8+cOd1Hf0cRrgvC8U z31a4Ax99vxd$z%#a;(IgnN7b+s|x1?SB{3ItbNNvFV1jNCvCa3ZD?q)W0f;?O<+(B z&PCwhc?R9@h}7LX6P#`Xcn3ieKE&OQ9GU6m7Ipzxw78yVb+u*qbZy|131{6bT9Ahm z;P+J*Cl|okc!m5zfLMpl%SQxgJ{danm93DnQi_NQK?IRx)3( zO>~;rZ4Uwm_fKECe-0V|fYY907u>_P;yqG8%K85H{BNfp4(@~jrgw4vln1ZbtkpC2 z2C(G6)%TC|F1)pHC5@U4$SE%zg16|xk*T}cGd8SH{flez7b%&!CZLJ-=2&3ycf4vn;?~)6S!|ve9xz|Y*Yn~ATuK;LLyfF+Cv)}-^|~0 zd~%HM1+wT^4bCh2p{J#y2JGRPS{eKVqPX>D|#l~m+bNXxn(EmoMa=`R24Bu)jjlJ*omDtS<$`FOOeBEyt zeUxu156|VXTvKy`M{Fc`e|78TGy6c+EHB{!Tk1nbjO#Bz?hR90!HEAr<8bnn=E%_Q zh)3mZhHb4l!mu{K4xVGKM=Po(`z<~_rQ5gvp+5K@%Z2ASuU?u6na@?T#ASH{4vaqZ zmU!{`OG`KS^b96xJi>mvr-6dECQJT=Jb;CT_{60(vc2o*DDqiN znhSQov#({jlevs|cR#fU?)ikLBD1<@#`S~84ZPCD!&Tpeh)8>W`Nta`jd5m-@$!uh zqjpI?40;|s7gzR902k*Kr1rhVeR71XiZz=oowxm1-|*Y2@&f=Cb#|KSa*3JmQ%~&C zO%N(-O6@mOIJ^N4TX(ncXK6iE7oHyYUf>Ah<27eq9Uu1>mckrB&zwd00dQVDfCnJ7 zI67_NsY|aqrI}d{;$GznU+y7FVd0!cgxs9C-#l`y1azD*FH46pMX%K|GY~P+O4T}j zYN{+gmDq?!X?@|{E+K5^2k{%P2R>i}pAYQOe>r#u4o{}>0uQYy-_JX#LSSdyPNJbjW@1xKh~HR;KzFa42BkZ?noCUA zwATe}5Xp|_-a3CIs6E6A5a)lgqc%YU4-w|1YEk~Ah*+9Bqpi$3p}hFpsp+qN{A*Ro z-O>F_bTFjT?;EaL{1y|u&M1WzYeOv#`ehI_DKoG?>SQ~tBY zrbb2=085AO^)_5)hHLmy?oncIre`-GU9q^sz5p)pqfcxYaW%y$W`FuZtQJx_SVDIw zf^eqJX=YRNSO)?YdbYIAO@ZxFaB~Ir>nm5;HSTl8npwC*HX3CE*YNsa*=@c?IPZ22 zp)o(GC|ImTYdt#REpUtdclrx#8~5GpD5(JqCE3&Fci!)4B>3OTf{Iyh@OopzJpjNy z*!=!;F!CfYjLGlpf082LO$c~>B(_WR7Oyq4Y&FxWIHwA{6-vVo7hyji6R&o{n_uYK zO%lRA1}M{Q(txE!JuEL}P!sd8!M!>sS+Vwfs)s+)cJ$1kd*V6ebI=RfY%qn26W8EG zlsDGK;np=tvOmN!V{eh*m_y|s~vkq+g`I_W(2M@JxP1-e#c0Amx zXKLr;+J-RK8gGX@^5Dgpp(n=vveiJ`KRc=eWR9Z3-MPJ#vaD zczY|K$nsAOas*|r0h*1$JM_Fgz&k;GLOZdoCTQifR2>scQu@0d5ghXk7LZX8DmNJe zf1VZYQ-r?svFt)>NmuGWUqH0QMb3 zdPHA-Ju)e^kJR?ex!^Lh=wocQ(!vq~#e~3o0%gDShrom|`M7lqG{#H}9)|09Z#A9_ z#RQD!VYXzh$NTZ!7pNc+1d+SBXL*azeH_K5lfto0FY_aA{h34W;}LlUe(MopM@31v zfTPNQ^oAFZuVJv~G#L>~kDOOD@IhJ*+AeOxsJH0(^e}Ws&Y!&9@HI9*CnqdqR@C5I ziS-}|S#h@z(-y69*m?td``-%>T2kxMFk46O6;bh>`^9Shajt7K@QXpaiqs*~zE+~5 z8&X2ZeD|PHDLGK;nK29+$stHnls?tTP_8_tX8B-wlbGsB;=1)Si@rK9O_*(u6*dJ- zhs{dTCA}%-XI$Ssj@vH)3{yH^3aNU<+RGBhN)wRt##L=gxK%~fHoQdAW^q9 zWC-mW)2eIj#t+Yqc0eQP$RK(5f0(>=8t8xJn3PfkZY4H<>hM%1ZlO`#LYRt^!wl=} z5bg(DX`r(QgM9$6s9FI0Ntu`4BFM#a#n~ZX5hOC&*A~J{R!8qzu(B%;Q(OTR7#|Nl zIRgrNml#_yCmK!Sn+mdpw}`x4VC(kqBr^$nO8KM~VG(>@ox92?8W z-Q*b8%()b-N8v5$-F{lAgF@Tved|vl&va`mh-K*G#;!fP9b}ZDzW^$=yslqo8E3}t zc~lMsGW@lWLB|m%)IErnYGm`?wLM^S)}NkHxXTM(3JF-vfbjMv0j#V-iQ&WxtWS=xtVNH0ceFr|fb2}ULI6o|(YRE&5raE(z1GrO@ zD;ILcGGea3KI)>pXo9iC}k zC*dsWpv zuJGb<>U?YaKNGYyFm&*ZNAIeA8wX~rH&WOi-3@r_o5ldhlAAtKnz8AE@>?=@{{9Yq zQ5(haLvRQD?dyL@r$FpmQ3?D3sd)o$-WZMLt3h?itb^8NOL@eSS}D`PmrtLAmrqgZ z6+jv?-ov0&Q}w!TOl!N`aO|g^eQO`ARkqSj)KW9+MzFrUL-Dx=`W9|6yZ$i3`wki? zBi6ND`JfB56o+*Mjxa}uuZ;pH8du>QfCa(cTcu3}<5oR*owV39x3@iLu>#_g0qh45 zf=<9Y*nsGZePBl#3GFEHXZiH*mc`hL@Wlp8ODKH#YW^e8QySM5mbWbs<$V&vv+ME4 z{%si9KsA!1NvGs=WSK3D+9+0*$JY5BAkMo!Sf~fpNgZ|6w*+si#J=kUwgmJOBESi` zAJc2Cacgl~djI&=@PDQJ-73m2qNCDVS3&z+GY{9Za$I1Ew@x=7R2m4{!zof%#_Z(J zXFsn@kH$WB3usNSN=hd0ac@rnBRYJD^q+a2zL#guySxey{E)r7ry~tIJW%xdDCy!R zxXbd6hi7BgmC|nB#kKcITEHUo6(U&cLu}+BK_2jV^MZ!m20sCxximnxq4&RwwQ;^?Uu4gJ6r6f?ef(!PnQeao4d3kt&@2_nQSYDY93 z+)tQt4e>bk{)WT)(%!yZ~H^gde1V2cBZiCxn6+()hC}Xf+BwfZQg1=`N6h(nhC8+uHbZFi?x;V(@ z&qRm(z${Pgm^I5YEDXxUzXYh?{{TUlva9`$EM2!f74!cIoT_HnfLdiwfFIW~R3_0u5_)~xG#XXaUmn?dagD)!@(q6UlsKYtxY zLSclA8&@BWt6T%SvQ$1TWyLhyA7BSYZl7^EpE;mANN7=wEv53c8E&_D#&(>)GghGG zwcl7Pbv9+&bM(~(!YxYoho_&fB3GVRu^(L940$}zF_`uFCF3c=*@-yPp^)2CwyyLtCy*fTc1BW3Y?e!YBcinxFBzYW^L?W zV|l!Ng4s4yY1@QLyoOYq^}LOY^(>Y21?tYzdA4B!dVIbO9&>I08EM{3ytHwku^@o& z>*U*i;c+Lwb)D@a;IYB&I=68T=C<4GpD89g?tArI>DB5!feYH!h;kUP zIJFNM%}B%+U@u_S2AqaXPsvpuXh)I>AR=5xCSdf8}jb?Ny=1{D0>XGU_k! ziTLrSM=Z%4Vp0EMT|pF=y90DcK9B4GuN3T~--H-=h{9CQzV7k!sK=W}^PdD?KFfj` zLuwi!)?OphD1nXS@Ll-I`OtBI~8B`VFl|Y*U{&J^bY{`S06+;)t0>nD= zSBo@6hXTz&EYyC?QB^!Iff$+x3#`Gw;+`6an96}@iIM2k+O&!k&b0Y+AI2^juE7=A zjTVKe@U#z3^vf4fJ2q)c&Fs+t{y)9p10L$djR?e_@7_UzLjFG53;6a9cq$n<00aVf z_82k>D)140GzL~eMkeNuZ2X@|bcu*b$!y4lK8c8W15f>e1ReY7pQd+OdSXy{Hof%waETPn*=yS+=D*V^ z?G&273&cNGzf~Ggzx@TccD4ssjP1jv8u_N-rU=o3I&a%s@_E|4X9m+kp*Nl4!T4@Q zB5xQD^m)jaw5qTRdvVhIo(UHeu%Z|#4+k`W2#G${bg#C6{W89o_r=--{ z)Qa`c0i`_(TjVjXYfU0%eGQS4Kg~bYsxXGaS!PxIOu+fGnT(x6fS{>HJUdwgEsMNX zO(wc3S#^n>qFwya=d?+Mt1K5LCgo-|eZNlxv4xCG2Fr7WM7*^YsUsUar8NHU> z@oE_tbuNf&jrloGCaPF_J$X2el!?=B=Nq)J%Zg*^4gkG(#N9<2mP^;bT%@pi36YfJ zYthEDpx0PZiicPnob)}vgiz|kTmM{WGA|j7$X3t0+F4HL|TE;Kwqa1qrp{S~$81?s9j*}M34YZrW-TsfwikSq(1%X0+ztYk|sxOcu z5m5`;?ImVT#wa(`Zi;r>6F&amWFz9D^ztMiq0$E1?1^qiLl>CnHUH1f-r$Pe)k#ir zB8)a3-7owC^x>++%?I6v?pSF2pp>L1sht-$D~*{^Bh9atlr~IA8KPy0QMdR%@^*%< z6NnTH7+RBE^{&zINPHqSow8QT}(=9>H!t9OON{Sx6o2$)wPtd)wuC9TiBCjSBqoVEJE<$dTWJRfZgp ztf{mYH~s>kvb`SD*>HX$itL}RB%5(pRZj1N7Se<}kSW8Hl^BCG8iM|W(NhsEiS`~@ zTem@lLE|t*D*57^gX_r8h=)ctL{r;=N*>H?Nd5H*eVl6WE3fMY78jc~i6Z{YGCoPL zYBxqQQ3QCn{F?g?s5dj0F)VUIg9D_PrEWvAcK za6Y0Dp1647Xcg$P=q|5|wAb=>WMKnB`GxfqWoJ{!ROWEQov@+1!HuwO4Z-{kDN8n# zT#6otdmxLq;VM2c5a%G!qUeA%G+mx&1G&LgNm5^wXm z?f2J_40uv3A;mIXldEW>@cJV5ad=2wJujEEmUaghiagbLqG z#`!+T*wW8G42L{e#Pb9{jH_cLD8V0T5dqAFu^mxb6q6-@VLoiiX&RZMk79?{;=R)!etkbRLFwmHed3~j0giv^Jfl_?yYm>@SC zE`|3CrLsu`Zm&cHwzJma35;u9X?MdJXT$Q_-1fe=4IqlP*dFUz&9KY++;dsa8lyzt>z)4lAktU}Ig8YW*sv8(%%F98lXSGMP3 z$`}+ZT%lBQk&AAv4fxkNb0XM9Rct!;lVYvI#Sy%WEV_twahgT$*F<1N@~PW2Mfz}Nl2B^ahG(PU zqYV8dbcIMnq;1>~$7C6)!ETN=sD(+jeUmjoo%};Fc-ndb90$XxdB>4?y8`faCeK>M6j%vh z%-Pro8c@irkXdnoJDvDM)pO~FyHFyzo)f^rF@ntywFemi=Ev*94k!-aK1&WhS>79Fi12<^n>^VoOuC;-8P zy~sttyE!vaZVusTQ3btxBR-94SVGJ)V`zpV5$F;_Zm*Q91j2j%lYrW17>an_W}GF!k%=cIOG84=C6h@KM-R2*rlWE%(U&FcMe0V(7> zc#@w&sMYuIgx30>+!+}>HDfTmk#H}wttEwmNCwtU;xX!!ZrMa9MaVpDQDuOYydz!_ zyt`&^Pt8STSUb|g`4x-IS*tvxaaZyD=qw2z&I4xiDp9bk1mP#ejN+fqI!adaEgK^c z9nU1>bIZ8CK%?0JBOzRNSxD?`WTjrlI+<)5xXaJ}#L z3efXJC5s8B%j$%(c`dwSIm9ACVW4l2$@k>frx43U7QtiIWi^N+ix|zQNF=*Z^pz)D zgp#4grA~sSgysk%UFzOzekU{(X!IsCbB`!-H1>3tnoTpq!B^81c$(Dcj2Bl;ptP&{ zawgnIe1rub&sgZ%##0@07nOTuNMK~4El>dy#Ts&XuTsbR4}?|A(B(7`m6&jP+@<}Pubf@rlI{Rq|14MQ06^#DqJ{y9=(8?!NP1y^;ih&;alON6}TlfM86Jddh`GQ#hD(iNn|p20Cay#iS_twg3nP7f6V zu1K-TJwxH!Mb8V;FJo&Qw}d~v@VtnV7k|aq+49?Xi9tC!MZkz&@*GWCp28p}Cm8%) z!?%vRbhvgyxw9Fr6II76KF_~_;QUCON8usm7l7a|jBJ1v3($gT=oRwwvZHvpVWK6)Y1b@{*Yyo?*2ecOgLe#ux33~}i9a2>w%QcXLRHi&9 zCn|X6rU=*}V8lwQ^?#8F{D3n13*#qC&8TYRKGeZk3b@f1DYr`*5NkN~&YwpzatYpXFP7cQ+OKDb>dzCQob&y7| z#!@9zP?kOPjXFn<#3drfBhgs9ySQ_H`C#X0Q){e9ey;{LIvWunwL#q9!R=z`3AS`5 zumY2uOw4ry#iOu){CaljDGjGR-$90QbmSsmMa6`t{fYGQ5luLXLv37ik)h{!Jxyfm zYiJNB0qOJ4xifL~*g8J8c8IHg?4B&acfzu6^q$HF zUdrWad^R>QRs8Ee-RLIn6R~+H3aU2obI6v{Om5$GfUp#?eM^!$<05;IMoYrx*(6Da*nYfNJ#ZAH1m3 zdm;ZQ5gQc2)5gbc>J-h%Dm}tLmdNZra@k75(?9WKlMxX`On8?qH9iJ9#dMQ_?S`M_ zE0Q3hnO-D0!^K3bko1~D)8xg;05++&qVUb=vKJUO zTqI-ZNq6?+^WKAdTmaJ&b12nUS=Y2Ap-jimv+^l}Y~$%EZMuTH!aD@bIcUUhU)rYj5YUC}EuE2a&ggRN zSQW@jy_t*kLp!%ik`|Ha=v&PhDL}e6La?3c zNHb(*uZrG0BH8h|H9*ZbksbArk|&wBr?E|7ssh`8+h@BLFPJ>*R_7Jc6)$k&tDMGp z?IasA1F%fR%#A3ktx2*awW!|$#S1l%P#7ZsamXC_x;=>CXFL)cNJOlfb?WtDoR6rj z(6GX^LGLRIPt+`}1_UiW>yTmUpRimusl3KmK|hC5#+I7Ih^b&M_CdLP{sqA79&3rJ zETU&L*lkS^f)}Y4l+24vK73p@LM(#uk*Vb7;fB(sP#Gd;!St{L?c*zOkdH)&8dM@q z;;4x1j;i2(NE61XpNiL)tZq z_%OTLxAz6=5c7{fr6W!>-eKyi#s>7l^9I96oO3^H zi3PHtGuc>`#v*hLYN}dKHK zi|mmX)x>d7k&>`<5(-4-K>@2T^kCX}741ie4jpoejtv+MkuVFb ztT$U5VR*9U68-tB>{og5qs;lhja}`h;`@c?y(W48i8cl6F7y-B z-IeSYpG)(~R_yl>l$2ez=4&|%6mW7Y5o_3|)^B@?QrZl)o=6OZMt?Rzz@SEu1Nihnh7@_yEP?`-2f|xIhP#{8oX$arRX*tU?|-wF*i~to!`Gzug8L_2)S!FI7wx3l~U&=B!7kT+qu z5R0t(!-wvG#@(nLm06bT4h!r9b-2;jXKSHgL3>vLLI}ZBTe`yDlcWAJI-=%i>7-C# z00OIbC`a4jP+;FR6=Fo^(a34}g|aB8=JDT?i7pZbDK$=uIEDohL6ws}=s#Wb-EM{+ z_h2jp&=P}7F5ZpAZl;T*njjQ?>=eSU3%@Ahe(NC0;VJT8VTC`Ol?*fPK<&rYe?vZ$+UP7v|klBaf4RI5$ zq2N}Qb&>nvqKm+MaLw8u!m22&w^9Z{fnbF?Xy^m{^q*uH!58 z(rnM{p=|Vp8}*^DPzEx^Vm{7fKeIHX(#8RTcKko4muVB(#3ef9*!l!U6zpqK!)w<~%_H2Bq^r;h5d zDvOwd1?twW_D(q-^zqWrI~g5{o>4&oHm+dvH-G3!Ybeq>jy3llT@vByNIjuH44@c8!N;YXWSh-|Yv)HrUi8ZrMvw1$ ze30jWMA76=!8i=cSeES%7q-SCh9|UjK~$r{a6R=`8e8z2byX?mq{Ww5u2%~rvuE*YS2P=&53*W!Av7r~Y-QKPgZ`8d(oU_~H#KyEI#|FtpF)+Ty z7s-|_CmZy^vJ0Uob^W5cU}DwUZcnZ6w1)l*LVtugL`!OI&QEMxTA$^~b1g=HPdjSs zOpq{j3gTS?A3dwYe%7$l*%C|L?egbq(`Rt3L2W#(n@w{+MNR7>7~8_|T-Fc4-sO3F4|`gCB}*>T7(6 zRD9)h$l&y!Hg(hd7(pNY-6W!1Jjf*(Pm9zrb|5sCu_nT1IRgvD(J1b~R}`HAi$qtJ z%uF%^YcM>|NhTAXn40L6o-sajUw$tqV>?XB`8ddn-86B69ApIMkZMjnonF|5E2Phh%jwBBnl#9?+lA?oy@`@U7GRO`2lj-QQEHZ66#^01Tta2&lSnEn8e=GR z7dJA}=w_!-JWwO}{1Y+7RCftJ5Dm;Vmk{AlvP#RK%$bpn#|D#_J_=+fbldz;ri0cj zdzE+4mMQA9qUAh{?%E+A#*_mFTf+}Y;$s}lv(qOEhqJbKG~DF4S3XakGkN^hXi2m> zAqL$yo)t-tx>kgua9M5IVwBs{i&_$2Ta|5-FcKO$tO?~FEvw8hN{kgWHT!a_=)_@= z9~nML>03}^SZ%cJu39e%T`Cquts7C-EN(=tJg6KSi~=PePzS*zH5rQ(i$P_|`0`}@ z9F)hCEgpo-B1J)u5*vj&&op!c6yEgUyDc}@j|1Jy@y;5ZoBzrd@T&s4{)o74@uZY` zaKo^XjDUStKY}?59oQ>5Dq6Flit&Kz{rO?V+=g*&GDC3|P#e z_j>TWUr*cQv)Ny7N_LZld4c(sN&9u3s{4!mmvyrPaa4|iFR{`sBf6_BDf|vb#Uw{k zG3n1TNN#NKJ{F^wlCAm~B`wP6_br|~_z&AG5~NUDgDS;|6;@s6qE((NeY!~$ikj0A z;S-|m>0wcLVatrpQ>9wb#hXfDM$Ac@!R8kSeZBfi=;DwfpA~X#p9dQ~&>3O%XfT!Z ziqK(U0S5hy8vfCuaziVt+jfvl=xp>@CR2X{`6&hiEB=q;h3B1&qGLx(+9g!i!5+V7 zE?4x96BTD^#)uaQPe;i&67ma>w>>v|u%jS>`K;ICCH1C!n$>0@H;Q4+(ZO7AZuw0o zJi?)b*UD`3(*LLtICE_4%4(oy`IjV|$i6ey^czU3@5V09gk}#IA{Pk?QGUO}?1g46 zXV+}zyy6gYKyYUen{NCeE77D($I-v50GhtUcIi?=8Fm??KMk3DLOV^!K79pN1dGk< zq3279Cf|L}7*Z6?E=f!xv5b>UW)uw;fQQ67Wq-<*&@&Pf*kybqNc;_Iizm^j+%$`^ zhNv|1^Fr9-^muN4wjTZEb3Y+3V)K*&E!$M+dblKd6i5mof;6%?Wk5V3>m?LUeBLGQ z=VD06w_VQ8J-Wjoh*tUG%iETMrhG|IoJ_562o0hNg3kd z`f`fwRxlV{fGwy{_%54I`pZE{OMcRTR~~P30Z|;tkL&)V>TrguC1`G0HyLXgw_*BU z0Ey>a0U+HNke6|BW6VnL<|7216MNI^xcKR*)`C=~q={s|p!}8?F64xj$TD9Q+u2f< zrpPyuUcvr00~8Bje5!901U1W5!&z~4a31<5>POb2$r=@W>Wg)ctw4{$6fb*q_b!d3 zbC{y){T6?U6{8YzO=*aBGBQ?8ZW*qC`HvsZB&s*U?1Q-%$1S*^+YgcljbT{|&(%Ho z^~xKJu_f_C2%eRznYoi$ZQ+YIE?>+T7V{%Z3~UYXp}#~BDKN$%VaL;=^mgn2=%R^c zm)b7ES{1mMd%DRex&nv6guEzUZn$@P@jW}tXWYn8K+pOG=uwbI%|kr>ElS`qIfEi+ zTUgT;LBzl5on^oV_857ruh3hTq_%!5K@K+o^96{Oufb)*<0OAnRoLi_dDacvwDw{v zlIV!*Nq!@WvVyfu>0vt-P&Sf}vEvUzOU;)=-(NE(KN?~d2moDJ*W+D)4a;uo*kTIT zg@)28#Rv--T|~mgBqU&j0;)gB*yGWte zU=O^a^CJ94`Z7aNUf_^!%HvIJ4sH61h*OcEh9zW$yks<iK#~0Khj$mpuD-VsYk;?s65ei^9Xi+PzddyW*47Z5Uqp|iAjT6CSDo*JL{BoF*|4D%&u(8C;?VDKc0nUsTHh3dR__rZ$ z&HINAe^h%8#3mHD?8!=oLWkh+Zaq?ymSbEToWtFKk}GKtj0>~~ zX#9{THk|6l+bB4jsXd28+X~E5UxPCAu4ftCgCatJ>21zC9nr+T=t>?_5FDwAi;-wd zwB*AL2ThG}MZsz1X|Q~>G)DViV`~!&rGWbZwRCq%b*8#c)t@t}iCKCPD;*ooOl&(C zwcp^I_U9ntbn%=^!JOmP{}k3KA|J2LXI3t)o_2rJP>%a2*)=MwMWQvotby0Gf>Gd{ zeHs&=u<%PzD(uq*%nMvK!jUq_0#C;~$g`Z4EH3YLlI(^gao2 z!uXkBxXjbQuX<5Dp$kOyxd`I<)p1u<_IwsiGh3NoAb z%Ag0ivBuJQEhl`U136e!>y!&-LR@rde2B7!ZTt!!DVD79V``6_Z4s2D@}O-&IR8zN z^OqD}YjO9g-SKgb6q&wH^G9+OhcVdJ&6>HpoJoRHd`yDVbyPF4Jsf7KZ}2()P`6enWbf9IqN3M=H8 z0P9UdgU)?ZTX%e~cqm_g0wQ$Q?BB)!M}{8oG0KwV?rKbwm3IN7yezPJS5#YNV>D=8 z7P<43X7(CjR&&tH43#PHT4mIEaYM|HbXKZht{V9uGBjpQqse0NK=byh?INt3PKc|0vlcFm|+YgWZR||Dd(R` zAI0E40}cJ99{r&(tryL2s7AN%kc(v*MT(kDry)H$gVohkcL?Qn+fWE_LkD{%R4{$+ zuE`5#<#sZZNuDzEKM{#zi3{l3whjy@Fr*WkTZxgtO}R26=VCS-)XR(&!0C_KF*fR1 zC=K@nGbMv6j@{b#JX_;3u#uVivoJKzrcsfIFgAl6=19mYKFWVw)gY$Ipe-o=6iRI7 zNpL|)EN+)5xs@1q7-U#C%$+71%`WnL>kv+3~@HngafLNZX@iBNH1E1(g>YSd3nu*IL7 z{5itH0%)Xu211vKDnB;+&fEZ5IR}gFP)UTerk(kAtTJJm+T5FwMCbs00n^MkML5J$ zN(=AGMzUJLzxs||B2OTqESWDEOSiV`aScX4>~@SKMv_<>!=lyKSlz)6S##Uc zzM+ZBQpU8d@AOr1Hwi~cGRW3FE@Oczepoy-yLo95DpWjnSVL|CFR=rI(aRv3w?sX- z&TXhgg8g`tiPZymg3`{BA75Wr>xETfBIEwx2q`JZh~&ZP z-rRuJ0$H-?G2A?|mc3DaPbIvFx(pWSC^e#m+wv-!wkNbRl|A=DQz%~D}nt2f#NW-4jR4g=luP*oKW#__7SFLZm}sed_G zWu}KI^h58x*&DIEpA&=&z%;Ee_MxGmY9^8C(l-=E$6jUY4y-^4JET7UqPEot)gvzO z$f?fDxme^J8?IFrS^XHgT^13-z2eo*b&N;l>9|1iDw+;`CwVBu*nb@NNlQtCf>+Ur zwJfjyH#^a+4IqK=)?@g?T;0!%KTMB%?^V3B`tPggiR7RR;^xi|ClWJvU*ecg!Dnn9r04NZ}IpS;Eq4iw_hRR zot=#fF^&_=$}3I5S-!tW0B!IPW|(si3s4i}Mfb#5Q1fFT4=9j?q0jdDr!R=Wn6bOU zAPZ2#vSC_MMsswVd9ae*R5dT!^GwIi@k;mt3Vk}b7>lX+8$-OKh@}tu20)6C$+G1> zev(C%@Kg{_jk)hdEYg3s9THF57syyh+t^4Fe>Tv_qZc{^#JuaIUrcS0JOuQ&6p>AL zg0Bn*pMDe|ENQ1J^kJu~$9C{GlJ~Ak;`plmWISgOC6z_ji`o3 z4DtwLN|45!TV8yoUQM1?f6abRY(RxXJ|o5MR6HzEvN%fgC?E z=2czq0{dx8nA|$LewYv>jC&6p_miiMRl=;B>J$0Ephp?DCS!bIOudYenshQQr~E~9 zk!6a|If}k|cVhlV1yJuJ)9KVa3p7`n?KLhA*0v!}Yoa*lgC3gU*v!?6uRNvl?a=)e zEdOaUy$AABL&1zA`V?{bsXXuK(L_-{eT<;-u+LzkjOt&S4u=trGI@%9e6fu?X{GB!{~I!o6dB+Fe$ zi%YOfLO*D?a0aEPU=q0D6mKhpgvlZMCNUfaZVZ1~d4cgS0FpSV%_T)Wp2t6J z-0nBTb!T8zDuEFA3psbg9(jYEZFrN|81F)>NX>fLD!@YZ4CwuZkK9JKt30sWHVF7D z5^8eDX#9!8nwIv_E_Ee95qlaiw?uH&V+ovL?Ptc%`l_~uV!;Js05i5!ZlxHEbcpTQ zmia{}?wiv(EpnzLx7%Ak%u|E2yv3(bZCrbvDU;{>CyV$lBbu|enss`gkJscFf zVc&_k=maTvS>0X8Bbiy16IPKP7dtkJ-4A}^-sDp}5kt3O1g|%gQr<>)P=k3DU<44Y zz+3k81&X8_h{xWp-1gPylAP0#zIHw?)HgSSEmR;#YJW%wl@jMwT!M_v_iOe64LlcN zNPvjf|LQ3v9ShXkGz4o1Azjqozvd9S)5gQHKqj<4nlCz6YX=U_c!a%2MTVYSfHgJJ ze#H9{O>W(T4!%7S%z}E8^}o1#52z@*ZC$j9k~4?|36gV;qU4-2O%f0Unw$kuGBi17 zl$@F-cM}_w9F^3BCPx7Q1qq6XxAp(`d;7h!&mHHSd+&JnzCn+oYObnP-K%P?Ip;UO zxt6r+$gbF{7fwuS6Y6==E|f>uE6!%oX&290MOf{$v>MSNSYNT{9bpy1PJ<~_XG8Jw zX?b9xQ2vL*%RAWAQPXFW@+)~wk7C=mUnG{|W7;z=k@ZyJQVjQje1WQ&pTBT`>9V_J z?OcVE`IhLr1}vIxu5H3%mE9(s%MKWU8FDNxE}A3Y_dJW|o%#0eZXi$h0Ec+E>Y%jL zcZPi_1I9+jMc{mR4w1F7tcncuM;hT1p|D?oyRvNRJx{k*Y=yGT{!jZD zMTHMupYCJG*HuR}mlBM{@$xYL(=KFSdGJ{~pdd+)E5H*T7*Ka|Y_W=J>4H5AbbWt< zELw%gcA5VFNL+1X=50Z*lnGW+(%>7FiQ>iV_tqxpn&CA|E^S~`v5|@Ieoh>8#&bGn zOm1!}zR807`(Vdpdjv9^rf)NiXUy3ue>a7t;l7H^L-e`SWbb^vX%!pYx2vIl1b;5x z5i3ISlo1KBMGmJfq_F-uIR;%ZyN_<8S~b7$%ZI;@g3jnpcf_|xTzIx0qd(k_ai2lFg%-?oeVXXI*}! zX5-4BYdXS1I>K7?gh3d+ZN8#@%hZB_j5oX5`=$xJFjF(vx_YFAb!~H(iDN+R*s)0# z3i7|DIK8b>X^&RnnHgq&I~DS6Wj&6#7dIqa$BMvp!{~p$FhovNzv_;OwvpCa_E2!b~<7BFAeC5Ek+ z<6@6J_oEL@3N-b{p{tO@BA7NQ(`>MCRB_Ds#8+L{SlZkdKYnM%-YqwpybL?Q+`z(4 z#W9*Xn@Zo|j=RTk7uA#FzfU6I<>M+pEyMJQa4Sx14E=MG>7VnNQ)r6}&kJ3x{vT5D zCSr;l<@>A!d|dxjLuHJ&WbVmAAoD+6?-)%rwGc2R+X5b#s4s2qj2}DX<*lx(FU>Ew zkrsuJ&DZ=x@TZI$!HJpHCu5&wVXflmZ!EPnHJt>)FI_QZr|^y&*nFw{08=ROadTRN zV|b!0QSZ$o@5S?#F{C##9OQ?!{w*J34;S3Exg^dW(O}JNud(95y}*C3pITwZ&cgi* zP-?kex`9`1ifV3uH#ntjyy(iULh;;H;Rn7s+##uC$>SGb^Xua?B{LC3X>rbMFz=ZM zLA$3%#mPW%a~n0u0~JOglx9>^z)x5&=GtvW9KZZsRZJcjQk=^@ji&69;uk=NpWsJ#Gx772Yqv_MXa~+Xc!OK7N&=u4~^<4e2 zHM^$kyjXFE%d2nBu^|0uZ^N)tBlRL-@bPSh3&>4%~+t`S?cwFwKJ#tqLLE z@-Y=A^s)th#a)CfHB7Ptfq!bmU>7^h>ufqnx>dRv zUwhXwJGej$&b~w(*VhwUOClm2J$M*1Dbd&}L$!&cB7}$9Lcew_+SHh*xx^%kpzeQb zp$zvemo=3*7wi%`cQSDoP-gCHSRS8XuD1Kw)-eJ5z0i(-qH1Cr6Sw=^*3;q4Q^bS} zc~Hc{Gvy=baXp(X%VCNp+346Wz=pl3iOuys`zMH{39rFZu-)j1k)TCR>G9FcC>H#S zHs6m=2krjIV~BRxkMwm0bqcay0M&H1rV)o)fsjDUWtAM7xgCi?0IklQoskgSJ3omX znEbtLw1BKl_jXa!F6ptzv!!PBVjM6jab~zTiSvwoyv~lMQ>k%_Viuv$kjDbJ=}HFB z=PWFvxVd@i61j!#0q-y$CZ7`Zilzq zV?`E}qUYf^Dk}gJnDL}dWvRbUuM+s zlEHnoILe{tT4pq1#Uw+RQM>?VTJ5%UH1U!>A1iXM!!eaZGc=3y%trR9+5Qcf(nO8n z%HbyOZq=8^ZSz7uY&a&+t3NylDsb+ky~*p;E~32n9#%z`X#NI&+%#YIPD<+cM^`Su zXjw4HqViC;RM1AJFz=wC;@ePqne&&#O=L=fJ%u0Z__3a+v7~NE-4c->Q#w{k4s3Xw z9Jz@Mrr)zv{wXpTM0N820+`y) zKY>l#-Sj6D_$1z|horG#QmqD(bY@x_IwddZ{j8`+5*c`q6h`ChEpzBsEAWxSS<#ik z3ZYhN_GLdc&zH|4fUB4JQQb_1ykTmZx+&*D=1)w$ii%OG9q4|yoJ924QM8D|QN?4; zIXEy6let(A$$IS0*rINASrq(N`y`+d(Rp_4I)R0zSE%SYW(%ZRqU9E&_2*pNN3-Btg#BjuEpsr*1jmeH} zzmjlqX0DBBJ@3(m8~|k<)n&CFMA*~ohN@fVrxw~T;1psiYX%KKR)eV3@;ca|{R0{8 z*NNMT3i6@Gb;E-TChqu;U;oh2jqnm5w4N&SH1dyo|FSa>M>J zf~T5NHLSasn&^dPWW_qaMZeLpurJW_B_L;N*T#7-U>rVi!mC`hPhi>s3|R0Tf(^>2 zv$`3xI}=(V7M}*y@GzaU`6LVmsc1%kg;hH{%&WxJsSrHEbYxu+AoU$C^_iJs*X0G7 z`}RG=E1jyM+B!;%@qoKhvW2Jq11v7bRhc{0Uq1jx?28T#(|_FaNGD@pcAlS~O3GqweJaIi5KZ1c@ie(@SltJt{w1@U zCn?)U`6)9?v((B&mDU`0-p$iZs-QB1QkD+k9h&mYMQ;XmpD1qR15Xd_nan5n#c+mU zKL2fgGnp)fTvMoKY7|?iup}ZJet&GrL$J{46kA3lSX(byEe};z`I_V2%ZYm1pL0q` zvz|`8fZBIUhp&1QI#DR-42#?HR=azP=lH{oDt5#uFxHmG6)2~*5W`q`agJTshbNVkIkMhTxF^%$FZ&tDox-zDe(Kck5p2@Fx}%W{%v z%Y61wvpy)x>XfHr%fxq}_9NKj0|n$YyV?J8u`cb~a3puVo8^$k!{=$r1r+qVE)G&L zNNV2y_}z_(y!E`*yfI#DK0dM+MTcJIn+%){2Mu#Yu02Cd5tsFV!onvE=>&By)92Irzq!daCxtN+YYNT zIzFkaq=vN2a58VN`?4}pVhr|~Mk)|aYqp;ptstSAuQ z@fQ5R+`dfrI=#}v=FQ+T{kj$(XHhU$tMuDT-1eql7+B`s}_0xIv%2<5+RH++gpEp$1k0fz8uhflc8ec6!2 zPAZw7pWmP{=xzrOijdmlRINgPV!er@cf9*Fz{{U9f&;L^?{~V)+KX25Y+D7}u5E;! zYmE4gZhUQS`sosOyqK)dL3H)NAxL&p?QWfG$N#|x)hj|tbofJb1+;=h91iKsx9>s@ zO9(;o3XqWy>a|sO5bo;To?0?^b>mClvt*XhCP}>5v8+&DHMg=6%=B(zVw^lOL6Erw zE#DKjzL-ABU~enC#y&-#nJ^dpd*>22^slh*|aA$6JV>;Z|BnP9x4O7_Pw&xI! zOxmqXcqOl22y^*zr06mrxbGyWYeG|((fzKLtrKIn;3J-;cnI1$bB$np!*Vc1OneG& zDps5R-5Rc8#4leWA+BrQL+0uF)I(C=eHtpz_KQz(qDy|77hb14N(1!j$zN~G7~Z1A z{l_nskV&gK_&+Is;__Z8T@wsOm$1A$P~0?Ws4ZZ_EhkrvI=eX#g;Vzxt^huNz~9Y00r+OF7%^e^0>WHN=uOVt69)tao*lz`c61Kvf86`qvr*tIp=8Uv=d-%C4;F~mEg^DFm%)=q7X?e7{%c~5 za{dIIZ^E{nOjEWC*1)y6z8_I*Z5IosZKVP|KWx;8ivbI}>eRUiO|j@=opQw`$olWr z^BRQIH9hxX+pAW%r)uxh8345mfbK# zoJL|OoylM0;ovx-hw|IoZ1qUxc;#%NaEeAjweNq%az`lNpp8KV_SWnJcG}>kuIA0h zrLVBRzRk?(p!;q(mup|$_uVph*ZosgTfOO8$N$)F#iDa##I_Goo2QSL=+O(alh|`C z!DjP@R)LtNBdH6<(Ddlg54I4#Qn;xZ+IyHp)0YG0rhXziA zK#-Wred3GdQNy}SJH9D8>j{`A$m62^q82V_*)@OIX3_<716D}D(pv+@sOy-=gH+R* zZz#v#hXebDD)13#X&GW3vq~0~N6~i>6A{W6US^JXj_KCej`(`+Ri)ML5ma z&6Z}}2jwg3^Q20YsBavB%VS0}c*mkcUQ_2EZ!5XTnvA(-wj%N`K`wapB{Ir)wDXLL zT|(urwIb6^%SWr4E*x9bm&+w?%#xR=uO*Eg7+K75MsKpPI$V|*NGac#N(NYQp6iAu zvWKn61J4={9=@xZ9LVa^osQLE)}E`8H{{^(%TvcMXe0V$)7O@IeMw_{9REh!;^UT) zx+EEOpXL3pTfKgdiwTw$3{6HJmt-P7y)g(!VHB?xObzLN|4`>J1pu8D-c4Uj17nP5 zxsY8r1H`{?fAhm#Hk%SU?1@0EX;%loPr3Yrc?F@i>@uyd;*2^DcQE$_8~z>#GI`BS zGnC~0)cMpWWR*~kINodasx)8}ILuf3> zQu8{6=4MI@=?y#6Vuk029|!{?X5@=g{gR1J>WJ&B?45J#W;)cf`WmaIXXtQ&3_ukuvtGrB3X^U*Q`pzR)E*hv? z=W*1ow_iltdE`x;NK~$dayi%)#R^G0&8pLDrXn9mqD`JU>DWD>$7DS)aPdxgiG~&u})pQme7O{i0#5&Q|FinTTm# z3+G*8UECUDO;-KU{FRhx&LGvNNrdswSdruT`qjGS6;B(o#U=7lM`Jk;t*X{+pdh~nnB}%Km z>Uk4NPA{fD)kB;5mWbny#v!3LFqU1AL%N9=UygilyqvT8&s&i45Ff0nzRT+uHwn12 zaqr7M?|Shn`*RMP^7TiOp5KccXtfCIYGZ_HIa*ifU;Oq?o*C22M>w(K!) zeW^1+r|pl>Hik@p3;Bfe_&Iy!uA4%Jw#7TfzQ?Mj4}y8f4;c8bkMUlfp)-cq9>w$v zD#?ypkV}op4@ef;&*QRDsy;L6WIi+9UyfQD$#Ij8zSi`sPW=4PDYK%C z6Xm>A@BYU(3Y)eisiAuVFgyizyU=L3G^B}z2zJD4${WmzL90GDw6*10r=24$n9=Hq*CjUOec$g%Kc%YQeacVW=3e61 z3V8>UY0a2vK4zMOOcD1EV%LT#QW&--YMG%JO;-2ZahEmYCMSSD8- zAqRes(!(0bny5B@f-(`D#!*8Hce&`Kw_odF2*I9n}IaYrkRdvquJ2U)byCOADiwe2H1i;L$Oi7OV^ zPz!U`LsJhg;E?tlKzCD;$R2sKNAMPqO@LOiv9MwQ@*69>%lXPA5oR@R^11$+u*j4G6H29_i)8HwpMC?J2@>iL-!|6>d5fE2;^~djrzH$o@aGZ4b5=u zwLO# z5qpF5j;05B?nnGkR^!y@0p@3lNcTHJ@ET(~oI-sCO{Np`2@>e1i* zo1%TwFx!;7IWAmNSf_4m;4NJ((OdfPcxAAkz0nH;-TKACB8!@py{BD6UnU|c6~RGN z0zo@bzKfA=KeZ^#XSY5s&vV_7eb9b<$z>JY(E)yOfZ^DG&sa=Ch}%VMK+}Ntjno+Z za?V3q)Yg}a?&Oj8g{?G(4jv_Png|09m#$e^DXH^AsTenCsK{mIKI=%ycZ{x zAB4zb!FKUX0XjlU!RZ>9_SK3@Xwwhs?nicXdoLXMvUu6`Rs&um`5`%=2R7ChEuI^+ zBdn~*Z!!U^)t4)uWNf)uyXF^WTR(mM8=pF_`US}RI3GCr>?G8Db;;|^w2Z2`4!*{i zo~J0KA>Po6oEK&#Ic`5`bZKb@#Tx}^+y_Z7o23DoV*F`N4KsKbRk$$Lv=_EnnGr?k zvAb!feUwuzqK=owoS<{WetR6FtTb>{evVn`{qCB8D&u1Ks`5 z?mQ}j5%Qxug9u;#UaU4LbHq%B;{DpIDJ3+r_hHKhwXd*r0E0~_wd{^5HJ3`s=^%jU zzVl2?D=!$s4(Et)tpc$0+oF5;u-s0O+}v)un*`~JVRtSi#p$Y^rvRe-K4y%;8W8K( zQ2k;~SA!w4M6=X@EX#m5?k^KSPN4eRTvCG6PLb-eh4wqds?6V38cr~4p+Sx2?wtS) zaE-F92E3)3#s)jvk^>Sx{flRVL20doF$2=4s!AD3JexFryt`7T>1nYeanLij)-prJ>bzH*~LEI!{cCx2ND52_7t=npI`mtdI1YY zA>3_XpJwfBkZr+hHf=f3-A^IGYin(7(Br>$(2@gCp0k@~YAvrWWUVa_isXb9GIp}t zqTh0^x1NOAZ{wZgOQRmhe=m_-yY~O_A-;jve0yx$C(y6mp9}a&?*qxqbM1qq)ns=1 z#?Pc9x~}hHeW)RWoGJXQ_HRdOk_#fX!J3u=o1__r7Zvc5XD+snSw(6<3~77!D}KIT zF~)RwoWbRwYI@gvCehOvN!Y)CUBhS3;JSaattd|FVI;l)ryEFB(wBXZkQ*|?f#*$F z`3#Pe%!23QpWD00>No2J=(&iBFVn4vl zFRmcnfJ=YTjuxC~F3ts~M zT5{P`sPe98JiBYs=7-WoP?PG*Pupfyr+21PZB;j_D$8F53M;c$GcG5l<`WmC{MAhl z9^8fuYpk=s>9}wRbnJVjd`nqype`Uyu`Re=PyL_xyr(HokHlmp#Z2!NKwHXZ-YoT7 zdQ5&T{mPT)Ztnc%FLUQA37HYPmVAf(I02^;k3Xa{`6gv0M8YsLf=Y3a<=U^a^}IPF)D${u<>X~$hbc)PG*Suj<`v_iR>WjBXG0>s-7xk%s+lD& zSg{`ISM~d3TEa3a$PI*5Bo)g~42P)994jGm;Y?S-`tJv#^iA7r4&mi z)Ma|>`;^_qO$!$sxQjNAUn-XZjr|@-=$yZI~WQYlOm5cL5rqU zU$6M*m`>&j zOi;B)Qaygni#%fsF&GZwO^r85{iKlAk&bU}6s)SYp<#wyUWZ?w*HwAB-~{?Ibk2BY z@#7yLK{Q0eLp|?SLsDpOY1}py*i0f>YVqnz3hp~si9YfdghgZy;U@saZM*N7)L|4< zG}ZO7DHjiKtZ&TU9ljEsIDP^mz9|NaFoGkknohfBb((TL4oAh4 z-8&TE@9#~)!;?-A{D?p9eX60G?Bg(gL3c-TgRR_cSc=v{fsKxFWQ)Fx^E3UV7#j`Y zJysild81KA5Cdhi6frmI!np7gy$_cC=oEUIB|zIsPoSs{+FA^f3M)r6^;G{cmCyN|6CvXlz&~zpMrn4^k07=RgKIyIulOSve5Bw z$#QpIWWW6QH*i`;OGD@4zOU(~Vi@~n$UNhcn6W`E_Oj8)^0v#mW~a^T4Wx zZ{8jB5iUTYp{x`HXeXeR z1rABU)JIB}6GlW%sJ<7!RNUy^nu{DH5LNmdbkmMDNTjwCbS*<_)%2Yy&EgSu{cRL0 z2y=bbxaJr|`)}i*_@HJac?Z=2`J^5{sfJDZdnC?8x16w9q|H2am^lG@y~wB5dK-1- zKvvU{Om4DFI@!;$-o&RyVV1OGN>d-%nlPm&4ddB0YED9Gqbu5|G&IDWlZ)=x$XN(f^~uB*jMvh}N)#BBs^K`E0j-!)N4pSix;eKyz8|iP>Z~pg3bu35 zRL@8k-p7-GtnS#L!>SAAXr2M?{28`W?bZ1qG-3Yaq6@7Y(MaL%9XA`8>mh#rHT$ba z(x_`R&ww8lWrM)DKMB0P#s(y}`#@rO{>`SkPqTD3$4c6yKkHv@RH;lLUe*}G0c|8E zUb|-hosD?}{NGgmj~mvfLXa~N$aaA`%)>#-;LWjmhT4DLKu5*9IN+Kj5ow{TUwS4Q`Gm+i9|ZQlR}dWpkn4O9g>6 zah?!8T8DJ4_W>z(?w^Ci9xRymctQXTPpP|zf8Wdm_8s22_qIJvPM?LsiF4ybr79a`OMnG}#Y+rhBuv z_#z#{PKU|b{V@o5O6th$lDBg@Ou;0`mer05e!Jh^CW&vAm5(`Jk}dN3*;s z=ECROSP*hu7oqit_7}iqDye;jafpx>8*H_Flm*l9h+!7e;~^!mrw98s+qDb2QzrQ5 zMB1%g^2;5iJqSrR%;vQ8jaq>{&%gh(;c*W2kCKlzsG~561npIN6Geu6#tZ~2no(xM zZcner@PRPk)Rc21eonyRwe}PHCWqA8Ge=JXbJVtgnoP?~g$gTkQ2HX}EH}R$VG@2& zW|Oq$%x2bb-%G5Lg6!Me(Kf8)*zLgjA`F_5xcgt?v+`;@asm!UEe;^fdFkPl89P!t{ulnc@ zqon&s%t8s)vtle{$>bt@{DCP>&3VtOPQ)r5Y$_ZguYDFJXA<(M7;9Vzh~;kjHSv;g ze4(~28+vSIX-xy~T(xO4g(f^#f&s6O=exdrpwsCo&1(pe`TG&w5{0uPnUg!QQ<%>b zi_>SMN{E+D1nvTN2aDNT(>i7TsK+;8e#;$$j{tRAGZqspUk`K5BL%VILSyNq!K9B9 zWNxy=!9H_c7COtcSw;NDAIUv{>#+h#Y@@T>g!$7F>TfJ55lzZzt=a`P#?McpJqxZK zaNOISU-0a9oTFiOUE(EbZjNo0SsGqD#-Fe!$JQ2s+odwN&N-jP?%#q#}4j1afF&N zelBJ|JR1xxFtEG9a(;Bul{^nAKuf2aLeRmM07#C_VN(vwJfJt1)bV;feoc~7@}=q@ z8_mb$9kV^NqhwDfeutE$b(oRiF93zwUjWeKr~`)fAD$|HMby2JB+B<$#_O zeBjTsc-_Qxmo0*aUTJ9{Ttj=xZ{mH=v(6%699E<4GZhaA=l&3CFZu}sc>{@2!H+{z z{j&5D>>M`rJZc#kiMMo9s3$9FvL7}cwpn9BWni;df!I09Sv?=`PBVFS!hlp{)NFj} zVrmEKnIKZOuyQRGVEN<~-$`75=!p#Ng?^;n;^HB-UnLp+th8nZ`ym=8x0E#j_ z!EkGdM1Nu#;)dOuy63r$LK+~U^0{)uDJ}_E=Q0^>x%1f`lbVM&$g8zjn*NUep`CC< zFrW3|J6QL-1nXn>z^j)St+xHszg<1otA)yUXv@*`WYrAb*{1t3@bAw#{bb~30TIQl0nw})0wzU3k@hN+BQzpiq}3%XTqlFvkX&j2-3^QH#NmFr1IEDj2 z2j3Bp`d3^|@0=EGElqg8&JBN?_^A18QAdV_d*CUn@i0jM7KBk^WTvwt`C$ju9GY;X zTUcK4QYzdo+LbboO0u11Lr*=@&Z6@6{@1y0kQk5TiWSNIyRD{|>q`B|K8_emHW(D- z{HICKltN;-`wpI^S@9jv*n(?2y5I^4twrI=5l~Rnu|A;yeXAY1R!7KeL2C(VD6j0n_)az3l0IOMO{S?{51==bIKyZ3nEu zrr>S}T0y=feg{I1BWn8)wL{2;hVobVHYYHq{Z~`J!CVg>k0b&K`yhZMfCx;oMX~=^ zT;(3NSp*B_>Ftd!7P=NstD z3+LR8lc3I#%zq!^o94lGK9bPs7=6z{>mZfnaDj%9hM4`v_~Z)+4K_^KDxzXclzs68cBmARRy+ zNSG__#T0+FoadWQJ>jv=#5%58zNW9oX3Ch)mhqkeee`4Ak|h2K6u#O2%?OCP5YWF% zgCPYbDLhIa+j{30;`=(6p|8|OJSL6OHvo8vx*Jd@>C?{bBQK~%l3bgT@%;)G{S0>- znBwNkKTscAzGC#kUH5x0dXj6W1nIA5my4o}2!#YCgK>9lE}II*_r!mjYJSf^FR zo=L<{J2=Eyf`$NIx!N3^ZR!wTG3d=5}Jkcwd7Qf zXq8c?-E-<$zb;7(uW(H{@!&ykb3YoD zB1!mhe!mc=f){PDTxo82)@oU~_EYlG8oAJ4L(s*q(-MrVMK6{#ITjgeb~eNqOi0r! zWWRt$ywOnZR}V)kvz`LntC|{!ev2yOLu<$d;k$0-3At@Y#G&r8ZkX_>A--Z@4Uh)su*sr?8%&*v;H~5k zZ|mq8As^vb96g{&!oVXM#NMV99aFo!w&{);D4fE1n>H1aWUn-TA;}n9J!LCCc2VBJ>7Fifo3GvL(*S(7To}4IKcZ!1*IR*r z&6Zij>A3o1fq_q>S7jo!n+9FdNDfVD2&rr$J*!DbR0YLL>fmRcL_HDxM`uLnA;9%k zzK5R%4k6u~B2CbELlZGrthhAm>3%NgE{;$N3D;%Q$ElwaKAd4nP`~At{E%`v^+Ig& z`0E5o?UhMG{hCGuvbgc^tKc8(?k{zXLfz%WY)y5~{cK$c~G_ zL94T|QgPN@>aU(7+}6b2_iYLMtdT=Py^`r#*lqq%pA-bl+hj&LgVARsrfZRg*ZEa{ z^FEb0Su(d67sf!M<9Rj@L|lv;@m_U41{<{#?ry&sTci+}1*kSMymUSxsKD?Z+-wT^ zkoY={b7!0`K6o&6O47^PyM|BdUb2;4I9sSisO&CkQ)cpE>E0-fU@-y@C;5J8&RuC>X zar0L!CN>V)%~lv0z1KCy{)*)1+0Uw*9SS^@cYoCUew2Gs=BF&tK|_Q0%gYD(2OR3T z26AiYT-a2Y!tH*v#ffamN+H=|xLozaH;;K)ogw*c0rQKgmbSkD#)mUvm91Uj0|-+5 zpjPq<%N-dsE6k-WPR_J?}>m;JsafsetL@{jX@8#OvTflfjO_-ZBX^r7q%nR zm+qg7?^xeJ;^!fG(@$(JSPV<$gwVkvqr{Gy#vM}U*P2D_^q_p|FfX<1n8&Ao^r9z@ z)!raGb;eJa&~8%h8`)c`gHp3XjvSGCs9qAu17t*ru(U(Qz@sOhNg=g zGDLzJ260Iea5i`1u?LCdw6_LmUhezsMaOJq*~dGO%~@m$?-6PG(s9l5KDn1~yQGoY zDk(F@dXT^JR8faCxt4iriGj6Q8jEhXWpq~f!?g2wX1R)Uc%y)m%`RPFa&J3x#(N0) zQBpINVaHJW+^SOK+Xv(BU`pEbSe2KYGR8DB#+;rtJE28AVv&@XfUFt_;r0j-=ESLJRs32FciGW z+`4S6+l_Ued%YeDrhHAdDDl&wMSfJcWJ3TjwT@oo>4(zHYwAv^0M-i@_s``h zUED|a8v-%}k4m9C$=9D%Vn^GLU3=lz{3oKDQ)a!zPmptD$LPpnvBFYW{{}C5_PrUh zW52x9KQ>Y1ra@@>J^q>&6~vld_NBgO{i;0;+ORM2O*tVKt(me%0M(nBX_{K%DKA&5=T!GogZQ^{-Ljac1|n5v#6NA>uLz_l!vM zYo&qRC}s4Tv%2}VfF|4H=-)kPlJ12`0_keELJkEGK09p9_vJ%e(nFB4@bX}8d75Fl z5+DHM#~mj-fIBIR4PAPcj}G(leI50Q&S;gp_6f9o<}`fL7fNUkYA0p0w3Dd!y9u;* z)u9Js00Ttq!PjgEZU!sY=R zMGTFT|I7~8gOPhjS<@o1>`AUD&PQH@PjHszJa7*j11UuLL(!C!BXh>0nC?}+6jtrW zuYgPr$lus@etJ-Z$WmrU>*Xu~$vfa*-Rc#|RwL*I!8&H!Tj&?wf7)5227L~<+%Ww{ z-Gk%#Z+6V>?rx)_1kXo$&x~TR<-Jv6;pWk3GUQnrv+j!7(dtq|>-(>c@YEV+E^x51 zVu&0RHY0uLgE^yGq?xfppkKm|A^yt$?x{;Jd;{L}9IK~m3XT7x=kFx@9Qx=s;}IiH z2pXGK)X%^x!}Jbyx-s;8!*pD6BOr%pY1?n>niN@^dSU&51`5Ef??RhggCPfvg26dG z(m}1E@82JzGsO5*->5hTd9CnQfW8vdxBU7q$Z>p?4d(O_;T z(|}p;hb+`}r~SGblJ779w9%Py^%MW~Os(OG&IflB`^2Q4`#LBxcdx(4?^n2*wFgtG z^pRfAW#1lupW)c3)}k`=#yVw1JJx06Bt7d|_m(eLIbN7}x>tXQiNEsCd*zOcf>v&! z*)t$@h59=BJL<>PKt5hnz3jS%%`ZUTwX`1kXU>+yal@?V%NswtTJeJERW9ID?)A6c zZ-^~d(e^)R-E{-EKH;|l0$$8R{p~-S;z0GF*?}%j9@35~?DC-;4Y#xZmkssm!VT8D8cG3 z_8yORZWF~DsOPkbZ0wc_nvl20L_<^Jodo?sN{=m_4e~KsWvCIFWJp7L*HCq| z<8%la$uiL647l7Qv=XtVVgT2V^J%%Tc>7?o4`((l4JxWp-*P0TwI8)ss(0Ao^#$Qs zo_p(Ve@jCE!&Y1Onv0d&~!C`(7Z#n1r$M9K3Gpy7uR=e2`(J06~ybJlc7Ex)%vNY|>Z$>?6 zDuBjHT#=A^ui33c^OwiN&0`Ln?1p(qq(0Xk;QOoGpG2-t=s~?oH!#_I$+PT-wbJz3 z1GbsluPB4{C$e-NHd955DzeRMo`-29yW8BwK1`=-Y%%1q7w%Zh0dzdfj2KI# zUwH-JL_Yi8arQnwoqNi#VrU;XO@-6WO(wv8pQ^M_-;2+Efq3fG&zc{fo-%DglpPM( ztSc5Jsr>Q!>?LJC+gf$~SOB?laf;$cozIQ6gQlU^QdQnJ>S!eAfaLUeKZ&^yM}yv& zpVCO4USUSol^XZ@$g?xyTX1w}pa4o$4yb6G)8b0&4szyQpLc}Wj-|JGX_YE_z79_M z_6K&5r?--&sbo66vq7%{jN;F4^pLs4t#@>bGqJCoNA89k^n6!!S5Ya@@IAxkFVK2g z)o1}z$u6&e%)+bm9y#k*mX_hSXrxnQ4;MS-x}nn;Dn}`XGFD8h6oaD3%i7`uK?Gw2 z^q6NP1Q9XJQM_%gcO|Sc$1!>Fcum5|iED6A>MQQJRVeCiGRoo}zb~PVxnF5}Ci17- zkLt`YD^ICgMyy%qrZWz(|6lDnWwSF(+?1plzh3F0$gz`uZ+QTfXD_lqMn&R?Bf~gj zNJmrJX=+*NWwYvE1yKjiU@mUx>z8kgVX`^S#lF+ik4p!*qf9AeHpydqP`xPN4rQV z*2@w?n)cw)UJ^HrwhB5%(*N?_SQZCnvk@OX=9@irBIPwr%msLQHw|6;_iVBqt#Lkf z%%a}xtCCGl+8CLJml?Hk$4Y-jh#E8jOEq_^0MF6f^~wJY554nNb37n}N+($eRwt9j zmo-qU+3+NHn=l5|%A6YeZ&iO9G+6i#iv1jBSo@AnBn_I|f_(IT>)xM-7rT<_A*kIw zulEb})Tnkd*G+3kbijXfm$CMMk>5(R!mGYBY!3lIF^qfP(P+EOamS*09~OybfZBNf z*3!4~?_fJ&bojc!kc)eNSrzZXkW5DVGzvr#4&vSsWveY}ZT}e1H-!rIi zUVc%lOG^~gT~|)~VlI`fguRPQW&q9GP&nmu=jn}Qt>WBB^Af5C;&Kw^gp#-p9YMA3 zjc~)h@d)ZX9uGmUz87pt8jzMs5AP%kh*-Zb&U5ltRP;@s)uK|EHrUq2NTxXtnfbo1 z(%(@O!d+yceK_{a^gkk$2a{Tch@h!WL%&=9Q(AvOfMQtX*()-;1y_M{D7UMUJ6uhE zt*=ex7hvoX4#BgOxSGOfkWYmA*IyQs+G=&31r#s|TRmxU$Z{IynM^h*V5mK7qn~&7 zs+-z0&5}a>QvpOm+SJXbZgJP5X$&p4G2RZIB%#h{;1!cH{d2xWNgtHk8}^Ti{$J;# z#oZM{=QX>UC=NxKbO1Ww`0rQ#Z|X41l3I)8wgt6~D>^42z5LuzaDH?G$YGlqI;h_8 zz~iDD@kfWFq4vydV4u4e^)n+qmJY1^yxc;spY8KEQUPRR1GsO<9HYB}AT+FPNv!ZH zfQ;m070So8#`$Aut*2cKLjy(@%p13X`edXX8mfxYbb98E<3Ic2BDIW-am;uCS1ya|oTV(}XT+tzeHwzUG ziir&WYbI{_FHQ!7-@$y$k{B+n|IvWp)&^qJ763(hKmvo8daZwEa*LW11Nxpg4bshA zram!$6X=|I_@+OoZv2LdlHA#5nAtoo7JCQ1+(7pr(}k^HdpA|Ya0PHj+0TyNBd7EQ zzned^Hat)2h>BSiR!-c2BM1*iv!}}g-Th~U?_8gpm-qSlM0(_MMEwZ~2DTq9Mo6S- zN}8V;>m0j*r`J`*Q(1<+!p%wJUdf|aX<>Z4A?-+}iR*D`z-0@Ho zkW;EDGfW#m~jWgWBTGKJ}k4rLtGTV;#QSFW}nRkMf zKI+ho&3QFjs>6wNesUqz?u5OQXWGIUSeQ# zOlh|^!orT4TyIAwa?*PGym?G$06Fsk&Rg3#wh0IMVG%Ay$Mkui{qRy+DC$#bh%nb%Sz0uErz23kbS%R;`)=!kfhqKN zRvP_8}m62sjPkaxTa7tgjH4;ysL$32E@z#d;jJR+p52sVxyHa=7iQ z{<@>r+Vo_*mUDWX37d+du6`4(<;Rz9qQBhr2LvYU1tPr8_V+9>FgpAa6LV*s z`%KLSws!V6&R)og;O`yts=agyn@zMd98!?)G*bfPQFWrdb2vkacP{-c0U*Vze~m< zGUkW&VqMkMF{ksa^A7LQ_#ebKh0_>S+H+Aj6q;|NNk638+-X;GrN(GqL1$wTViC;~ zu9XrNiGBC}B8RGrt!+%DUcjFD9=0*3V8~Ggw^J+_NrIz5+G_8OjXWxZs(~w5VThzE z6-ob^w?p0y^)ZC(66Db457}F74FlG%KQHhTqy@CokxupD6)l_aQ@XOkBpM`)>b`xX z`Kkqk0~gQY+px%kfyuFkXPqB8$+9UBF;$?6Kbn&GCK;`PRyB84irIR->*%9!Btu{p ziyRD}LX%RU?ame`vtiyfAWO@?Vm{~TwrnV1;Ec>O$H1U5zP^~N!WK=$x=kUoZ=Tzv z{ac1Kg||qq0$jqF&R2%@VOLghew-3pW}J9|znRBlxJi>f;V(okE<)x6WHBQ&8t@WU00o0r!7!N-7sHpw_mC4{w;)*#+&z zwV66TK_K=wo*$0&lp<87?C9?AKVjd=jF~lqKp7;6I!P({dC{t1p?PWS-3lkfxe`c9 z?l-M2`(WX)CY3ZSTPrsfglL_l7C~qtWTvc8Q@gbcL$`kK~Q&gS?YktBuo3DLK6)=BBKW!t$|vLz#Rut5UG2v;4JV zfe7(!WfE|c5J})5Iqo^FK1r>lc+OA3ZVhtX{JyS~laHMkh$3Nx^7w zI#DCU>bYjE;uJ8yJu$#3UsS(cbya^0P=e?g3{okPb3RKPzBI@Y7BiPXV2?w5}AGNTadO7ct zMY7C`cXA_>-n`qc>NKG?j53#oq}8?6 z837R%2@Q01K9#=3C%FZG^)&?=Ow0Ny`dkC5o=izT|0e-EmTiJHYNtYZB0J&2hP0A$ zyqyUgLzru9%W)2d4GACZK;*1JR%<}U^se$+w6f!j1VzCy-6JyHkPAg#o+lHm%vgl2Kyk)NIBl<90TDz!hnMvfrBYEvTG&NLqz5Dh@sU^T2>L z3=Re@-zbt?=~5PJnf03>2I`9J@o=wbp}~pml$k}`uyYg&PZfFJ6zyR#*FtGA7nZS~ zTZytEkt1=as|48tGa6mGMk%jaya&V#Hn%ZxzcW-780RoBeR!La3t4Gx%~2v#ECKS| zYQ{urktL~uE3<@Q38f|pI{gJU%AAVYM;2cL;asUYmTrCiEKS^Cnk4Gk>Fq!*y1r-B z4AK>Kq+`3ISd)iNJpcFNf%qbb=_dD^9EyKH)nfI{7M!dsb~&*nV&Nqr>-JkZLDam* zm;q6Pkn(1U9li9FoIUEWmS+0v&lR8SW4~~HEgA$s9w9Vw@V-NGJ{Tf) zMk$X$Zua{Y9`><%R?;SP5)UjF_-lJUUhC3FLYc5l78XfkYsq(T;;*(u7cuu~s|0UL zq7FODxXLb8KnlkTbfBH}9}vxUu&-Tpy6xsxE1Ni*j;{Y2EQ>qmqppeVQ)cTn29|;g zPg>VYk%v`miB6#+C_!Dli-uvYVHE4@yt6#Rl5jeuqcum5YQ{Mv{(X=~J=Jk{Q|o=27bl-YrDDRh`Hi|}FfZ&6JbkSYv@tE?xkhC$8LR+#0r z?~#SOU~WfqeCyaA4Kg$x3x_`<7Ip<1N=ccfrf#gdrc^PSSid>seKe(7c|56$O)+$) zE5J^if5k4@g-{f?8x3u`=5Hv{S7YiN5%euyKGzAq$OD0k15pgN3;J!rO1;2c>LlKr zw0%+~ap+ljyfJb2H>HPa3{n*eY94<2Al^ZG;%1G4h9Q$h^4$L+4I7D&Qo2Fk)QsiGo;^UF>X+jnxDiwGh6)jPS5oJ%a$}ql zD!0jx0Z)PS@5;^Hapu`+XQYbPt~JF@ow=u&X^2*(PGooTOfNde^0HVmJ?{Z|=U|fL z)ERsUV8gb`v|otX!0Jol66u2FqY`T~A9gL^1n{Y2jE71$H(JAXIK@g>&_c9*Px@

z$>n4_HAc~XF4Y(Re=6l|F>7qFMh^bqCL51-4f|3^+M;#QZFw$^SKzq|n5GOJ< z&8UeouVlKcFCPO(@=m?zjJ8sfr^!5lK5M5u9+PVE>wv0h5I5qo7xX);tlDYy-1fdK z|6IzEJRubIq7^w!`aOHwkiF!=1gy}n<>7Nf!=#V{rP{5(76}jN-dxa&nFY(Ps)!Di z$mGD`nkx%*6_)qlIlVuTLBhw0nz2v~_I88@?s@AA>D~NcdAKCu+b-Gpv&+E+N!}l0 zuVw_FI%{c1p**BxAbTp2hJDJ%frdW0%O}tJ)3;@gIVpG@iVv{A59|H$0hmYE^h*NX zTID!N9y&m`3~KefGCDT@#SnO9#lz5te98@baRsnmwU}uzJTqu;cKH5)>|>s&7P@j& zd4{71Jv>e6d9_A*7YL4@`wnXr`p5hVUo=?!pz`93K&F(Qg|H< zdRbc?>>Te6KD|Rs_mxMl5mQt`kD3R0i3w#;f-uqA?QIKVF@-aSal*Qqp{VTpKrYmyfSPBQxRvhwA5cp#5Xz+M+pyC+ zt&?s~Fa64g9l~4g6IsaijQ0}h8M~^Cd@JolE%6n}E$D{n4@k8lmQfjBDvLr67j zLZbcd_(2D>MNX*yPTtx@LVYQE0zku)vIYlKt!)1}}*7sh{Otc$dY z6$r0vU&i0HE_m%{yQLUPw=)&zVz#c3`$`E zz2w1PlGZU0jy&S`kG^SWEXefD<*)u*O}~3A?|dfDZZbDd*t(WYJZpN-_5Sxw4gqh* zjs47P%k{458&i9rg)?edTRY_a3RXsJUpHlha_e;+xjqAppZ=E1)!>7>K7Na+c{zXP zyZy80dVkprOMg>r*EpI&hk5{|asMzS-({17S4AdL)|8&3w1nl`)eNw2`Q?V~(}44@ zGDk{@1!Pke=e|-#1@(Euxm;o&xuQwT2gewGO%(ZiZNT-4;d5ekV|Ff%I+GqP2B!a9 zC%BtpmGs$aWNkVgoJ(ZK%e!pvP_w$O*Vx)@?(p9sSu&Fk_KPc(Q5Ew9HT28eT!js? ztp`-F+hJ~uB%wx+bfM@o1B`Fw_@W((oeh*rfa&Y9{ZhJwlV^~nW8`a+8|8T`aatv+ zpp_EWJ+v;rg6>Bo}fE-v1upZ$}a{d3^n@;G~^0@ndC~RTto_KE>KPQT=2z-O} z2FRKxzRG~#DA+3#jF zlER`Ib>qb9a`-D_=xyn!GA_qk)Nn91L8(+wSaVJ`${wy>9~v>a%~S8ocHiLjA<6gd zQ~*r#X>u#Gp#;)4F{B*Q4jwzChC}%QwA&l)jbg zYkfxHTiolcU6GDmaae*T5)m72fNe{c6$q$3ZC61cB(^_{o%~nBWN@sy5knQBA-<;% zQBuph-ffu*K}!}&r!L1X!XkE5Pp_>70JbFbk6x)gYx3TyoBOS;eemW%1~0Bd1%~NR zF2ZBEnd0%`_QYs*jE~4b($_P!lT>;!HK3sDPAn zkW_D|l9EhQG7B~w(2HawxlaYEFcEiHdP;xV%%=0~Fo(Oa%J)hC(@2T~lVwU<31jL3 zX(qTKFC>@*1KN^FyYh6T0MgIe!%Jx*Gdp98`%Rgzva~hc@G|J6vi>T>@+sQr4``DF zfOy~y&a*{IrSm>R`4TX!=Psfk`~kJ{KDq}c1evHtkTlRU?vu@tQTZJdy*xwQ`=!3Z zX`jtbjKU{Y8fp*z4lYS%T3w0;mZIAgv)OWLHyfHyW~phPq*GF`xEUVpMIXsAAv@XZ zjQdf79dJ=<%Qp|?)d+jCBNT!e;RH0^BcVD9_f%i&hThBt^zS2UGqdz}rF>a1*)AYM z!Dk9l}vgucutor{W$b8?R?k84|FlAc`g2ewoIIt74~uQ6(Y1pvZ5H`{d@zl!tg z>d%hw28*@A3w_YpC_21upJKDCey?pU)z>Zg9{*+jtZRjX7e283mEm~ZN7G#c-#m~W zcYwQl9JX4|4c3<;?UtsN{7;?yVz-U1bw&3|_*QJNv_@yqH_R*@F0LQh&S%LY$6t+L z(Pd&)G=`R)rkblp%)*bKeudj1j;$}WMP8=ymFdW@t5f$KUc#LeJ%^4tv#Z#~GSfg^ zuiPEN4Hijb@UtFKsVK58$2nMIu+e-5JLb|Um`AXn>k>Jiy}ex~TF?eY61V{L=MM39BzSj+XoLH&E~if?hPg3r3iG=*?;rT1 zRlmIb15*4u6QQ4w z4gmgc&DZ}sVwbAlc82;_+W_>g^TQ?m?m@OwK=YxYgnq2oMo}`g6sMWdcB-<1(_$I; ztXIlnl?KjHU;Yn=mdbkRJXV^E>CXf==T|C!Eti>iOF{IFpEe3_;-o6xmtQjk$|aw? zjc-%;gswWjM|GrbA$N44Sc255T!E|nki+}=Un`hkUS3TUOZO?v&d}1SV$cnue67*Q|dI>xo}8Er(j_IGi&^esY(4!}*l~>*XEnF7ZOfx6-h8 z_)2Knl$u-%HwWSa!Az@gV79+cm(8|97$?VMWd3&&~C631H%GSeiEx zcHJdXXBe?45dNqcy;*vWYoMypy?HES&PpI8u;>B1PQqS`5;$dOmt7s__8wBZC>~*$ z{;dsvE^gc;09 zGn_9I-n<9A!>PLORk!USw&ZlB_`cXkZsAF9udf6K1U+$T-qQF$Ir^}Xw=j(J-9mws zKsj362t7A~JJ-9fW`{Z~Wr;&o5i}4V*eq)IuJ}a-!9y_Z4+syp49#w_M^GAXIBDwR zNI5D&@r>*03qcnf_ravj40Sz*Rzu4NJ(!J#o%sA?W2qb881JcG($eJ8{6Y_M&Yh+o z(21`2StWKB$}593S)qqxbj1zcYLiq!LR zS(CdB_y-IwVddPs^Lep(@pH(()ut#MW-CT8f1wyJJWT#HO#Mts{_JMbpauk6FnT<5 z9K=^dNT3ryfwGZn_T_i;agQ|5bT$2q2;jfKLL7D(`1X1k_ZJ)DM`MKELuPv-YE z-la9(;C`i&ep8TlEfjnW`{65B2u9^t zKc(COzaF`h{&bu0b-89TB|Y3`@56IysJafj^Gq#wjzUsD;6iNqyBN+FA6+Omvgk{? zTK8Ki(_xl{!J4TyG_e!C#iEvta3QnMK@)xt!_WNTjhH7z8RO3_J7W}%NXrzRJ4N&! zg2=77sOK-n7eC0eckLjQqFm7CoSkI8wUrt#Ct^?|ASuO#t0Z9BluiohrN?IeCR00T=4%sLMGedQ9rkfwj%u-5^PC9v-5LBtysg&R?(hk zAPAI3Hkg`cZ|h`t(j1#(l~%eND>vgWp_wBb;kcqbcMxaF%;uFhv+&j{>*`Wz)$jI+ zR*AvZp-7H}LsQ5OHBNDTj}Zr)y~S#o!!7A3G}4GWa5)jA>zOp>Q8TPvQ9A1NGkDm} z+X)pF=}U%_Xb!*naKR=4Q`d>!ADK2?XR`GON!Y_i7+on;mngRcdzbK^^IJOYCr~J7cZSVP>^@1a_P#fG!E+_oR7plleZNGg6}3 zNToX-X^Wx?2tTTYgT7Tob2FFfT7^x#bJAQ)R2~N6)PR?bz+0atw!p*FtAIJO?e9=Kg|CMXW!u1}frPxDbAx#anMM4?1iV zXT%frlipU%$jO*-m|=-_yhu`1&!l|Y5N@=b5&gLpTrS`k?fgNZgOhSdLF|dM@8jIM z-~sHtmawj~b)E+{>BR8q;FzG}`rc+WR6tl?5WeXhWF z2x2q>K$Ku_KfPyK*P6mtfP!cHtPvWCpnST)UhN3L@L+D%TF<2c^gPA3@zN8}5EWN# zLv2I@{MmtyK?b@lv9HaN28Ljo>^Rn!0^rbyaDz*M0kL=v0OITHFBIlK902^0MSx{0 zLdNa6%z_4RpAOe@L7RlN1=3;e)r#Ib?>q0HXQ0Wc<&*8#?rODxrK%R<-a^5?(1+5? z&EGZkc4<_8qcva;gOT9q8FAlcZj8G|KfgniYtLZhPAXx@O<)|OrNjFyQBF@0h&@v} z{Q$+0PdA}RsrQ2-jj!X)p(KOY?#(^^UyJ?6sezsNm=ZJcjF_igj$FLCj4@HPi-ll& zzcB}MDn^||OF6bzt=WV?wzoR0wAZ`&g75r;etsebKf{q-m|70^^5RnklX8WaN;%(a zip@+MS!6G@e^y$~;&IL7hD&8Nu2?v5L{JHvQDagT?_G*eU4>OEc+=*zTaE9&aB~;! z;Gpg7vJZrJNal+5vXM7jiu9%brpU+ux!m*C=%I?IASw7B8Zm5IHZPHqL@%Mkm=L-@nAVqIk{WT4|M?$^y z**jaus9d?PqsW2N;7DoUH$ea>GADBgk)PR8WIynVJ z8!j4{B}QV(mf3~oiClvQ2D4GPWvTDlhCx_zgR*W@NF54gLtkd;J0uHwThNk0)k&3 zkFG{9)ziY6KnnUR7LFFI$GlscFx%?~xg}k{{RW*%s(tOYWjTsO9&AV~OWZS(m-S^O z| zt2v&+zD>eeMukq9B?W@TKcR9kCsGRTxXu4&F-8K)^9tl2mb;|mlp-HTl8nJq%du?t zl918eh?5NFq_oRS$5hwAt~@)FHA^$2%wa3W(u_D~AW!Ba+TC{`aA;hGk`zC0XyKG4 z2o=Zo+YyQ<_q7T7toMI0|N^*hA-Ru2&nx z8ayaB* zI@f!I3>R)})acDeVxh!OY18pw>8e6q1flp-;Z~E%X=1L8yD3;V9Jh(}lHlGwnBTWR z?dUE`tZmiXmn2ZLwkEF|wFLP*o>|(h_(Qce9Kp8&BT$oy2N2OcFTh!kUTF8{fa8)xJqGQAcM9}XW5s)Vv#A=sN-P|&t{ zp))WfWbP0yYW`xZEKGyrh}DX<^UKZqpb!y%&u#+Ek<%LBX;h34xzv#(>#Y;hfKMSR zi49efR9rC9qq8{KH`IRe1v*N?jF2u~pRKMD47oBUr%x>@vNvH-+r7mYn)w|ghpn5Y zAtOCzR0h(*d}klq$!#i@k|5K3P6J#qgf}JbPRzU%YRN7-E58M$(qaxtWbDAv8(;18 ziNGknxD**GA6mG>L#K^6C3V~*D2v+412@St5juLL4=0A;%2PkZGxcG-py^CA14Q+Y zR@3%%SQ@KFiN84h`Z2zwaeM8N5(r(!oySd>&al&jv!zVx{(_=*fy`)i1fy1BlFc!E z(C@gCM-oUTg32N#cy!6iSBW#37*NOwY(}1Y9rlJos2g7}rNP|&P2irYmk+D`&9fK| z1{mW6HP~s%m0A*PCf~w(bGNNldV5e$V56ys3Nm9$UX8@3##dY6PFJHZgomU95D3C6sv8X5>F>$&L9Fu^N#C2ww zey;l(Y{e2(t*^n~cIIa~!+%ZQbOoGZf35h%XF{hGx&sN}tZThaR>f8#E};&Z!4YZ& z9SWLpB5eZpmMZ6jLN^z~sB&xv2b*kF5wUCwMEt^2j>(LHu=0E~wTKRsoQu=&LXB8g zPYY#qUS1xwq!Dj|ouJXr9|6Lx4Dl#d{)J6^5*0BaIBZ)~5jyCcHiAKb(B}PBhafI?ocYS{ztKVMO-K_=2b-+-NBPze*7Bi5dGcRyGwBn_npsNli($<6< zccUIvG!Zojt1nA66zss;TLnxjbwtK>T84aW#sH01OBc6I>%1zAr7j3gOSIJ`j8o7V zZtd~(KlJJwcwU?YGbIkkym_5C=<1Z3bP|2=Qfv?KP31wb4$SGoJL7BgRY4_g7~RqR z!am&HvTph%?0iVKPACWQDxWNK=w z-kbII`fKhz&|An6=b4~s(WH%;35cQCg?!&+McKoGqn3WtYML9gfU5eLfJj5ge=TQq zg*lkiKKw<}deJ#+mo-k4&Ks>i)m@;P$+_+i=x4uMpibQ83IY_EDJ|Af8SjM`XU9}% zJR&&swXcK-2gp>d0uXS(?j5ey>JEO5N#YNJ&4yC>W{*ZhR?q{U`#H9#6G|*?_u;KE zhgn6r8|<7_ic@;kO6_K~*&L!PWbCqQ=19PuU z>JQ^;nvw2fxjb7aPd_0|ZtubkghP=^FIA60>cl_~|B6*= zRE4Hzl!n9A~_rT$88ZCUXRvc#Cs2;xSB}ryo*tzrpC34DQNRpi90UEzy;@ zA}&*^MO3e*arOVlHou=(N(<(a?0i4udKbz6d*7J<dr~_CYG?)@?oY-Wi{X>6m5}ah6+#^d{Q($~3zo5!utXzM6 z39W6liOIK^}#dv}Xbx62Ae72M&JocN+SA{NtU3B-%xf;y5#AOw>kPp=2q z1rX$O!G>e%S$Wk+ggqF679J(&K4lZ#Jh$`X4AahDx&5%X(~bnqOpAE z5#E6lUaMgqg}JWr0-_ABF57Mi^H3uT4<`Xt8x3X?cIlz0AXufjuo>HQ9wRgP{g5pZ zq=4CxdHWDMmMCKw-HHQCH5q*QDjXIUNsjb7EMn;quS$cbc{3Q*7^qQHewu+#CkSF4b{jvr`8f$t2ubhxBsd z1m6i>4I-y0>FhKkw%}XcmjWG)s&ZS!cs7S2I;&wOYIQ?|b2+*hJ`lBdX4d0-9+#`Ur*$j_+K0+6LAm{EY{NweCO?|w9_*O$Z_bC!-Dj6w@jfxgSSD=z8g1vXiW_F^)lX&jUlh}oyM)ioc4>-+2!;3VDd5udJUkvl| zpFSW>Z%&mPY8yxt#>dYJ#GZst-Uy$ zSJgQfKR!yJ7)gSZAuXhe!6oDONilA-OHP(-Yf*n~{5JSw^4Xc>Q%Pd2e3>Z45V!X5m@h*{pD>b)T1k|~a1h_&^`94KJ&zBYNRB)UMQoMYxiUiV4HKAu z$b~wA-`Dje8j=}Y$3KXowr$v3YA@bu7#SZmQE5H zW_Y@5bm^0rxIIK%f>yX!dr=Dh&WV(6&c0@{lrn0j)UDQ6xQGEUt@2`Xk@iYe+|pmw zL>2YO?Hg}cf6AG}mS_F;3ySC#eP=B^T}d?|a|nfGw6>PQMJ+ZF-6&bjh&OcFz>Ax@ zL6j<3Y!lJcm_r+8{y{ab)S^6CcD}OgHq3#$w3KkzhnNUcFliI(y$v@Ir>JH2yOgaB zmaWV>@kX*ly4;wtxMf(%LV;4;C~S*kDelmkx)_Ki_yd%Aa^(VcQjuD@lIb|SIwP;h zviB`G)H0slP)Z)Aob6x33YU)Sa;j0sPME{Y)^$2=gk@I^H^oBrAexd^aGqMKjphmW z)3<3C=h|pZsOKc^f2PsC&pE@N_P{$BLhWt3u`4>Q^;AZG0|BfAQZBEEvrV~J()crq zl{O4d3e+RwP-=ENIniQ}|=3?(BRgK5u-z}}8Z-`q;A+j(~{{j60F$Vq_0x4mDARwV2;ZR_qq2M8)p+F!A5F`{d z1||gz7B&utxCS~ox0sr`OK@^wcG0oq@81*y>?)eOGez4Un9pgxoQWt3jeq!^m@;!^ z__DYE-#O5ZbvW>m^{fx#iPqeNO5A)<^!=YklVZUhDIJZboDn zjae>BU>Dtw%kMjiPd?hNPV6<`r|lV$tz;d_NcO$5vQ(nSa8Xh$201Getvf8#yEs;x z7>u%h>S5PSc`(Q5SP9`uGc}XMQl@`R7*1CBIchdm>S5ma|7Ssp*HeelgoWnvvAExb zEm&(h(K@Ki?1OJ*(x@%<8DfzrGhQ!F(f;+~T3;`2{k0YjueE4>t%bpBEm&S_@!z^2 z_6Txj+yLPkSyMm9CLv7bk;D3gll)d<{usA3Tb4&o`5W%wiCoT=-f6+5WH!NdLXz%J z-x>vR5x?2r)GA5Gr?o=UsTXxJ77k{k!dxW(iJ%!FBAxv0r???9IfWiWsWJ&U7>;>M(0 zo3j-g>Q*V+F-N6&=qiYvN~QGKUPqem+Vz^=VVG;(J>Y<6+N)Xt1AotgXyLmaDj*;3 zt{ur(QH#rwS2<@2UN+p(ysruO-IQA4*{$W-^}^eUUg%K1a}!x~kwqong7M+uqgGj;NalMN{6p261?5NCO@+^qe2#BSHPi*^ z8+2ev=!Q*Ud9J>+a6^47YLG||8Y6Z{0O?%u;<3e5Hg?31$9(P%exc3SJC+_>%gxqv!U^R?QT9Y8Rv_Z+rE^yJG@F zh7Q$ErTpl5Z5i~n(^U(za5php2~@Q3a0d}CC~$vEJd-U+hpjH`)X$mUDZQMg!<$SN zH*)Gzge{=rgUgnjE>w3ot*jX0n5=?eLf7yQr=0|Zc9`H44bsgU6A0bNI&sem<%$v^9&Y_)UaoIZO1W3W-yV;%*n}3vOM&81n%s;%#?#Sqrbf<7Eh8;xV0?YV zZiXM{?cg>inHr9YRd8f;pp@1iZN?v5NkF^KW*1eLgx%d`SqO-!Nk;(NdLAM9Y$z9v zth0p^tL{a2LM1ipsQr;P%ic_`v5aH0LtNFjZ_~_o%<|-aFJFJSh*tf@7OoK$JPLY5e-H1`eq;nmQKW^Bdpvn>_5c{>u26Pv&s zfcIycR80kmHk3;-^^)}=mI_WJ)LIQ!+B>_e$=^5IoN>>oH?oYHgXg0g@JD<09$E@1 zkPV6Lue1*^)siXA#mS%35Ihu@M_QyERo}H5<3~(G$OMh;CMu^z%sOZ-bzViFg^MB3 z=CG+9WD~b^79>bxsuPRN1vVG`UKG%E{uJk!j*9gzgrM&Ff&Q(F`);STT$NiMoq9TU z=dMIJLK2Z_$QfO*2!6iMm0fR9Q)S=3?%kMqwrorCt*Q$pIf{Ak@0DpK=P`rv7bL90 zcXhB|^zUDYzV(=jKeb*vQ3TXt6;B@MVgd6)ugjL2p5~qGcQ3Z9w+gYgANvb`S93VD z5$+r^BoH8Zi}stjG2?vxdSXKMpW4f>8q$wL(0@4nZN3PVX{?HpuNji^*0;WKVMa3y zcQVHydat(b0G8s#GJXq|{XZa>=-CDxsQ0ygf&V_j4z6ktE{2Y_a+f}`tBO*8d`hQ! z5Zb_dpevP$bo{bfHb_a|#9C|j`@eKKUY8rN-2v45{EjS)#2fr|4Oq^ao{|_a45iAl zQDddGjGB9e0W?RoDA?8;Jf-99>bQ|1*z^~BF@sDnFY-rfBOF1oqfx33BA-zx$3;7{ zc9PmspK$3U`VCP&f-j%=`h5mM9~=(1B%bx@WHTok1SpBs=M0urZ=al_eu}_IBqH*C z4)EoBm{AY>xP%nK_uSW%_jFYv%H%TV)X=YSqA-*hU|Js%9;#>E%VuULDNWd5c|k04 zG=J%zG`b-&9v6pV_?QsRF-YM=ch9U69@0)EY`6EH78ADgcW)!;RQ7ez_xvQM&B>T; z@Rz!K1N@A;E^J0h+3mY^QGCM>tfGxoZUhh-%sYv=QO5gbTvMNJ%XU0l(ZkKPux&VS zZjTGZhgv6mV>e^p>)q6H1c$_Ly#wwZ!#Vn}g1uEc_f_S8r2}}%L9141Wd_h}Bi#FS zw?}V^9F@m?djX_vKOi`_ohkK49oA7)X(P+XzzljYO-d7$l(9=XB=`KCZ#))}4(&V@ z!XHA2Hdyx$G>AEIuU=b2vXkflNry3aL{8Clv z)xPSHknjIS>@5FsEB1Z_?OVemANvtYGN>9vh9DdS0)Z*Q)xWOUSI6Od3l!mpyBsF;gZRGCwm&}F=WOqB`+JB!zS4JNlh3(e?nHt z|M{b-A1}x5$L27?R{q8A)x8M@Lc>q(Pl1Qc=YK$neUA()2rC^k*i{yJ{EnzZqrPfE z`+X|_1NgPis-O;M2l?0QFe_SajdB@{p8XeVHe|&S9;u%ZZ3{(w>UO#*^7P4TOBnI4 zXZ?xLK5{@%Tt)#v0BrgGdevp%7F4fWDI&n=nLUG~^gRL%fs8Q~c@XR?^N4`LNr; z74>oUkE=e%dbC*UXhw?E1HPswl@?iVfqggT|3+U?(18;qR`T;)^SHIuN>ysO$)(Nx zV1*+1wCtGs7`pwv*=>7~6maS0mc10*J=ARS-N5xtTOIXP2*23zDL;#!|I+JpIz|i3 znO+D~_5FYITZL5IhI(l^37`HyKDWa^^|b%-Pd(p@yw-DF;BP&@egW!f{VsAybc+fP zdW4rq5}%bQY|cQ7u};giI=QY?hp_|k1UO7LjCW`3&>GW@snhc85CUs$L2A|U;T}33 zDs>b;C8LE69Xg#4F;;i878SZeexhSxpnkhuenoAtyiy@Y%b*P@fPPc|W^%3bv)bsy=nq~DFHl)gotBkUf34HeM@GLXOe3Y## zPU*eH&p)8a;?$CRawusytd(9t5e04+fn5XUM|zNChQmjx6&ZFZJN@cHUV8Q+StBin z7O+85)F_hJM$T=wR#}x5DmT#2iy(ez>6QRZZX^;aHwSKv>>OOB>N9F zg7A&Zhoy1N7jqDeeis7s=tu}#?GFjk#Tw(3;-QG|JX3sd(4;F<{B614!`RLwKJQmO z(WloaY*s1`v--zshjW()6bOoF2y>(=cG6Xqy~9Tf+pv@b*N}yj?n!tB4j~|Fv+^#c ztyV{R9mrt&5uJ+N7XETi2+uPSx_6~2$v0V&%nT0AI=nzM@6&Pv83RMXnd$X7T z8_Z3DF*IL-8cq%n^)OYfMP~0fip}cq+7A(tiH!_THryWfcRTOG)&A86Qxc;RZX#MN z{3Co=q`SCU%%WLTe9Ug`%ykBN=(!n$#DCv#_1M1XS#-QOH;o5)=~?s%Tc$i+0LMO) zG!G}tnLmn!Smz8Bv=@gya7x1l5ARCZ)8chWz!nupP5lA+Y5VJPoL0ke5K7{;#QxDDbMGX<@-hUWy`BP5}9QF7HGv3lIUoPx!2je+!3gI|4xKdOLSX{7 z*v}ch?8;9UlgG2LUiW=*8E1Jho|;Y4rIrjcxQ)^df^~y!JE2jCcBC%l)H7 zO2;Jc9I&=>O{x$dg6J;Zz4zMRc=_T{cw4w8jP@l5KrYRp3;#~qP@fSjWRwr$vW4n!x)9 z6{!#gmq}Z)*O6I{;-FY!EKiK5v|vainz`LvQ6ZT!Bg?p2nAD z(U!S(ayzHnie~ZaCgDxLC>a%V z8==t88~epo$;_Tkke=eeo*GO1N*2-3$BGdy_`)k(eQe*6x8YV?i{>rPf3{SaKWjBt zDhO%Q8GA&f4~cLT7f@r6O99c}_soLQ2O*&&%B=plKuwN+y(BS=GSaNC5ixbp;2Tu< z`PRnoWNV@*kHFommrz7J*3y;rRd)(|=*-dJUXp^8la!6Le6xl@C5iVNI1ZZX@mn=% zt!&073F9*APiG3gEJOhh`;#zEPl39+g^osICKD+xAAuogrQ(fCP(!b5wu-yhoxiD4CP#VU=_$RYN2vrZ;@%*|BP<8pY3|u(7o1gG=hQ#p(6oi>Gp>9d zNYyWLL(Vmbd~LQ;Ou&KQ@}&dlR(8W(3(FlH@Z`Aa456c(&|4i z2i|FL9^kv<)E`)uvzKTaOlUAIRf=O2tpV>s?i{t|GzO;I*{QcjUZ`Y|OcrE)ltN*6 zovvzSu524Ta9TN|K@)TKV-K$k1+_tLBEHs9U*y-Pje0DrD$gL!O5D%P%D>@)vRmh{ z%FFSnSl+P(u?k~5?CufOY9YLG)dg8)Zuzr{X z?_IACW^iGaW92sO)=}hqd$(`lCiDYYpPEcuFdHHL;C8CC-QbGe@Zdp?kNEoHGb8S{ z{F3?gXBKU2LGkZ{x^X)AfCj1dv^c9}cVZ4w33?vrW2)({f5$Pmfv;;9T2Lu@rPg#y zAS$a^0@;`+DE>7wQJm_hEdq-tu>?8Uc7S}TEp&#rV?H!ENUTx=z8%fJa%;!#v0ahJ zyvKrk|4f;3Ir!(!BFc%Pu}q39=cZ&-^H<}QD#Bn91rEbrKHpq)U6Y*o=@5mX57zPA z7}@9aG7Uk$YBhynSD?4CUq={EJ5}8PMS|W%z}8c&{!mL0{d}c^jdg5zStYH=mP$>O z<9SKhda0VJ`K=549n+j+8g+fzgxHmYZo_TV- zmBIUUdk=JA%*hm-W|UrWzgzq2(?H<<78>!*a){k8dC-DrIJDLwJk2D@V#Xc2e0kw9 z(ROpjeSKB}L?!0L4b)@yw6w9W0dLt)@T%pf z`JbnxK}k=lHpR|fbDltsQJGk$v^DmS2-)0v412!T-z%eaM_ju$RXn#$M0M1ysav#` zPP?lYo|h?XX`~H-xf6738UF6Akx7l=M}>q6K`AuU>U8a4K9_&KVUpStMJW~~*1a8> zDd(-%@_nMWD2_HXuJy&u8+$GVTmp#rF&lzjrOHf7q;l% zfS(Jbr`hJGOXdvrZ10W~f~}!ED?@A8_0Ky?csuko~&Ura(S;QTNAYC-5 zMO`^Q824D{=R-{En}@|@s18ydze-Z)NtgAInZE7B%jfaIrw7ezAElF}j@olypV_CZ zxQFZVP8#RTzd0)pAG6jWzGlCkpI~~fXm5_bEZzxsSpNB+niuvSZb}Gkv}A=k^#_`t zT-+>}Ry2^UO_|MP(;sn;nS6Z8?^9vjf4+fuyYY)$zgc3klQQ8O7v_y1m7w*I&Jb; z;BfK1!@7E+Aw?G!8Bvh;;6;Q*4WZ1GMOgu#3w{kgQdnh6T*5~$_LuCr8j1IpftcoO zxAh#2PY{<$i40F8F>{bP@IcA3Pt?j4%O0iVSFZ8JXI+Z`M046%Pt{X@Eq(a6)kU+a z+r8NX3e4{{H6oj7(>}=*Fl!-xM~yo_WO^};83}aSTYV<^wuv)&h17WX#@e?ouexe? z?u7Fgm5Oy(zIY0~(I{t^Pt@O(+?RrZnE%tE_}ePNX+8d(ej{du_fbF(bsX+GH8D4g zn?_>|Nm|Wdm9>>Hz=f{~6*GlgH>x%9%dod&se$N7ffe_3HP;1BxJo2blRCPdS^A{`;@ajX!E-ujg`uA%mUis;aL2kAso=ex&fJxAxG-WL9C-$4x+!^K4sB* zr6TQ=l{EI1P8GG`&7$BC-s#qFxG>C<8vobP_!QmP@R6M%VGb#rA+7U`Jn?jX5t}^(_T$e~un2p_XD>((?Fe1Oy zWJ+WG`ClW%h4&!@PwgxRxI23m)-_H55-A7uYc$If4yZl;sY5I5uYg!;4cV*MHseZd z#<_oST;AH4^`7Xd2*wL0u2okCGl}BSrdv^9G>e8>&h^_A;1q0Bnnvf<>AqgV0kwwf znj^H+_zooN2Fm)j+*o!gQev4~XrkfZ8&?K}xR;QBinZk2`?V$M@FpyzvM%5d^MORY zVD?vOT6zEhCO%k%FXHt|=?iP}mz%nLoJ636#BS!A3B}ThX~@TnzJ4!8nv?2Zb19_A zbT;0(5{hXj72hA|9OXll-Kdo1yU-1Br|Mk%I?;Hr3wGO)1a^0QLk)`kNhRL%}9U#1-gaU=_W?W5Z*?QlMnxEBI`BH^J!JT(30R^9si@1hSlOqYWgJdSewz9SEPMDTE0C7CD7&-(>~QCzb)brg^o6r36mA zjoDR`GaVWM!2Z9 zf*HWS<$4&F>)&u_O*pxuk_E5cEisT^54;_5bPB-VfmwMwzohJcp#qb#U`hiK{U~|E zk@2RkK^D}ngAhvF_vzKHL?mI1g-wFN3`BfE(b1Z`A4>$|bpy`ZNVHdh#!Y>}kRMYa zr(w8Emtbz{XS|0ZakntZ167LK7LhgdJc2_?anrrBRt!@%Fmq$zN1cc6)>5J=mK6aN zz()jv;3HWTA@vv4?mk}v-UnO~;p7lvnc0OmR^z_T8xq9HpcRV0-`AdU+*7J3_+JM^ z^=;qw0=JhH0H1{NGb2EcWivc6lj;1{YA z$Z+>qD#;dahvf*=~;}x9UULTAgUPhI{ioe0$V7x_-kslLU^mNXv&(=}Ei=5m_aOaCpYLB5bUYOtvE0 zoymXLjr1cI#bG{0deQuzE1B&zquX6wt1326hIr%{#WR^*V+O+{%~Tx_sN*EX2Tj8m z#eO@+JQc1PBby7A;AR`&(ukk~6Dk{J4TXSrc^X9PL~`pa#3@_fsdFCN1W@wO3XV-x zk8ddiZ|^F4W)$E@QV#VCCt&H$baXufTq9TC90#~cZa{0iNc^L{mLzXAn&dhjNMR|j z2Y4U;fg_=gPE@D1N}=i$s4(4YVq;)SA;fuC76CJn_L|yYKCVC5H%Lc0NWo&Fc)~Ib zIYvFqxs0xxe3FmgWl7a=pax)W$mzqR-qPQgdB>U#r7a1oNw;76IBXGH1T^q!NvYh; zeQlSwYsYE<_dx8UCeC^tLY|;or=_`Hy5^YMlekG@)I`8&9HgI_o~ZK$YlK zXLRerxR5pPsb^+|o=Z1W!#((5ky-lcS{uqe zouwxyAbIb2P+}ImuITLyI3wB{uPDf*^5(sosqL{X5^7F#i@4mGR~g4>Plq+X|Ns5M z7-%3bjJ;v9?)^AlfBUIP!S{B3Q<19hivw)1i^+2-&;^#LN?=?UmnvZ;-Riq-+?a)q zRJCKEOy@fVqClzOmn(g?-qjph;s(04P^NJ+)XT5&W{3e(#g^Pp~ABS3&F%6Tup^*+=?W7&X506`> zU()}RAnN}FogoCL-b!rp+_5+NnON%M0SV&3^YX25xy0^oxX>;EOj^+(R8<;_P@-%H z2YTLc`Lr)YW%1{9s@B1ex~V!o2h6Cq_)uOEWm^5gR=qe`2iHq-ONPAB8u+Dtcufa7 z_VVx>u^b>Mw_s;uWEvOkkWa(TnhAtC_C<42A|>dxA4PjJePxblisL95azl|Q{>P%_ zDl12b`=pT7CNhsab72Nqv)3w;Rv#vmQLbW7IghmyLQ2~9j$`?2FS=E%Y*H%5aI&t+ z=+rwNag_y-RoeIF=PQ&kN6QVd9=myXSIyA)pQrbTplc50KLOInL^7$ zVRw;xYru8*b9k5Cn^*Q0Ai_7{w`sdkI@6?-9D>cL_h<(g9>UP7s8=Bq1;$%v02hFU zze?sKPnA(tAf!@jD!12o#ZBQG(0C?O=|P!N>>JV+)t4{+_ZG@|A6^}hVNYla(NE+` za4lkg^Xa~MXG8UM^oQfBUyT3PbOMv^5bj$h9l#+5%r{Kx9*Lh z&(}N1;XN}Pd~kg*42sv34HmvaAiGR>{h%hB1=Tuk@o zlg=f*EFzx&FsdG;5vq6C`(2&q^tQn^<*<5uf9>E-5JzwjdTj2-ljoVlFdy@>0^M>zKQ0z*mna{GAhs%@y@Vf8Z3B^S0+ies^ zA;&&pP|Cx#XCo6Db}rSR(M&QFE;es?iDmwB<#hv{ka^`el<1; zzgya+l^_}@Q%@ts+rA4sMYosT+)K~mpC&shHtXzfhO4(1npW~=pDb(9FT=#D?%e3? ztCK$Y-;nzG-}mT8-9DzF*l|0%72k>s4vC>Be?FfqI|7SaU79!EKzdhSe?8RQFw&VK zcA8i6ZBF9wzJF)M(!=DEuS2-<6RuV{veH@l7&ul9HWaHcDpgI_eOr&=Zk_y8oU9Jf$%X2gt@+q0g__no6e20UAnb7- zXynkgZE+pwEEvjBV+j0IT8blDEcwuor$q8^vRtBwhL-K-8JuVYDN35fV(4oHw)En? zz^HcS{Xj~8JzVj&I(=zc)N$pBf0TZdi9={FzlSP4JS|O=llKP_(@h{+%%t+4kE+*e zo_|Wa==H*`H;_yN{fJ?))=1+lQRX^MYiDKV>jU+Vxlod9LWAA8EIG-sDCf7g4MDVu zT`0vux2q>5=OnaIAD;|aq#k_~K#fGrt!hPxK@$Rk2L`d`HcCc1KepGxs_lsgn$dCt z$t80pALuypQdxc}lUiY5)kYvyK-F(Jq2l`@s{b^WMFVRjNFirc-Xh;S;IQF?wmWkc;)I}E35`~}P%yD#+7N?1U#b0H!+z9Y)xQj!K}~*yJKruB0y;3y98l`JQ5zr< zo>q`NX(e}0Ft*XmQmx1PIMx0VbvN}piLDn((TdA4=~&t7g-{S~EM@(phI8zaTloQX z-I?~BoCUPDJaM|$Oq;1FbgV{K2`9t2K9iAe&|2XC8Z^bx9ZG6y+b#MC^d)>w&!V1e zj8qn3>cXu8bB5ADbA~`2#*aGhN^@Xu>4b(~X&p{>X({-%IYSxD8>0ShLjB!e{skUI z1Hs^4Vlq10-oUg1E(6m&g=Fn!Ra0D>_{INgJW!E&D5IB^TvcVga?|3~)K+o}w{UQ{ z=}MX;6owcWY?M_ULPKEzD@Cm^sdP!0R66|27YJ{VU%!I?Lm>U{Qt2?EbP6^Vbn*}E zAC*<@aVS|iL=65AP47IV68)q$$@$f9>e9{skB7*lQl>0|Ng z9TP&H>#sV(i1pYBOKu;z|3h{Xhu&&Nuj`;GH1(t0>zCyc34AjAz9qDvt) z9yoK_>5>W^kBTlax3F#wLQxfaU+XQ)wdrQ7lL6iOo~^Aj)NdKdPJJ|43F<|NtA3y@WT6sQMFaCl z2-=c!BP*Mr>NB;IzCBWpP-Bo=<{}?wvHWPa1KeHl(wy1=+${Jl2>`8s(&T;Q_i$s6 z3g$5J1mxc4493s%i0#k492Ne^_WQUurE8umC7*TKt&$HN1;+)_7G)Uq2rPW4CQHOJ zFga_47qUH*je&~U?+QJ$sM0ql)C79k$5THvGq{~v&tGVz5zD{m`J$gs8$|?J+O#YN zezWKewm)|j%HPVQ-TFvUYz=i{s@K$gDe&q*>>G37L^Fnj9+{ZywSp}Vf4t^g7M=5H zn4ZM;2~^c_^hCYgZj1P?9v63oR(^`<8?8L_BO$n2bueT~Xgz_?vld>$Z^}h4cf$g; zZ;x-nlr=fGW5biqTCuK{+2>*&40w4|K#T6oQ&$81o_b zJa_zADh}gV=cjtG=L=|m|52_C_MGZ{nI!pJRlT)y4YkT2!^QmkE^KtYr*J22t;}m< zEi10+iWYDt%(AEWJP6)FbTFyheC=Yu^YY(A>VD6yEfNGI)>*T2rON^~piEaoih64P zUG%iroL^@^_s*3(7oyy=Zz~7N?k%_RD;J)#w(0!Sf3Jck`t`~booVAmv4}+XhMJgT z)GFoM(X#~b4?*)uVW7hFrjti{87!9Ev(F!A)Dmr|JXil!b(|>uy@_7t-9M(+jA&!lG0uoQ9c*orFDcTjB~eX+;z7J z(>m@eYD0a#Ol90-T&D1v`7w+9jEQ7k$OHOAPOBT{Txj_*I&d4(Nru=BooUbImbi*r zS080<+}rjW#z^-5UYXDFCv(5HGEt%nd7#9ax7q9wfAl$wd7%E@Z67V}rS`u^eFH7v z)c0P^E2k%1Z)uX7Tfp*<%31{-5Xo_7xcoa6u%dilo84pUoOeHA78a0$^|1rpGCZ?g z5Y_{tsqlj%sp^$aQ3tW-;Qr9;KGqn!+S6;GFq&yW%2bAztOF6F%z^q5Q+nL= z-W_idkKP1U#EKHu`E&d+8~2kHkU`JZHOa)BsgvKP_R>3bd=mDr)MEJ)ndyHyXlYq> zZMtb4giXX4Xjg&L)u{PyHJ!u=p@;j{ZnfY8}F*(}qW226bj)VYsF`S^A62{Nl>k zPm7x`I(#6i%JBsBR&0Lj9XkIe?h=zyYaUO6%_}BGhHJ5Z{F$3tFBPLo60{DMQ;lCb zCqQpl2U{%4v=$4Qo8*qF_J}&2vA(>~dGbjh1JGIvFOrakX~%CAoJoZr7bI5>@on=E z&E8z8UB*mzrz{CtfKhPGpWd9RD@$m>#pMt(43QBg+y3{4I>IwCV)2HzQwzy!H! zQnu?~(yfD)jTcG;jKHg2%%0ldzF}e1+ne%7{>`ecI{lM7M^3eI-~oBpYL6i?AwFh%v0EVbO$Pn0+U3iF+}z4O$9AMnC-1CLH|D6?yUN#-*S?%T zn}x==Tk)h8ALspcAB0m+^~lUF^pWxopMn2XsGeD{R0YRjNs;dSL5kxa={Nn6=*k~N z9ee%1DX}aJP^p_yaVuztX2&aG8IuUGiw=SCe-Xy3$i$kC3|FvMJw0lyA5HGMtCV+= zS_-KqD&=9@>t2ww$`@@rFxZ%6bjqnynN4n*j8@jKi!-Fp2k)2?9nRbTPzdZwEI8|0 zWL=1T-AI`+o~aetw+t;c84=xy=ulhUz^^}JD0Jd&6;AnK#{&wN;OzXt+rp#Go}qHv z4EJ`cGr#V0@o?A+w`xT@Be{$!n&=JtZg=+jhA`{)cJh>a?XWzQX92W?>Sa8c1{5XM zXK0OWs#H%|q4j4;Za};lQ14ZD+|!!0d6U{yXFRGWLj9RTR;qMKQ=;o@X4fv^K;X~4 z6Q!}Y$@^mpSGp!o-gL50+?>96D9;S^nLA@|e;wY?8#~q6wLQq#jtWlAs3mJ1QnYwn zm9lQnDiYfc-qHVt!_DPsZZ*2RE!py9p!mucadNCZzt>c@-|ZV3prrPmcM1SS=q;4B zki6sZZnG@J>oI5zT;o|PtwPk7sk*N;UL;c$J>Ri{sp6&^%+DJ)r0peAHBLy@M;dN+ zCi1xFv@>00a*9B$s#?fN{HVmk++@l~W|%1B$lAD$(h9U*$Vy9$6vGbq?vtg_>nI^< z`YO%+(&jKQ$5Vx7 z=Bq-hAfwG{$fNVT4IN?AYo%G(){dLeNs;IL%?UvErleTsnCZ0wV`VXrR5U+K9}7dG z$Wk7*_|I-{*JYou(oBP=UgLF+sId5*%v6f^f*vy;Lh>-Viym4S$5RT(tz8N}B19?9 zIsHrV1F-gV#5H~`fAO9!*)Fylj|l))?Wu^dmW=mWC(dOT#(-2Vzz1SJQN-MHQ$p{g zsjU`o-^|zHC@8&G83cfC!KZ4z!C#bHon2963*Wx?a_EJ$`9N?4a`U94BNA^*C1(-` zQ>fxyz;NbWaLy5~S64-n zA>lZ<`_n;~t|5n`^vX<)UdGapwGhF`{BF&N@4)za?mAJ93_#NgQ@c&0ttIXSP=sdR z!%X@NpxCu>uB_4&N09jt3uq-{HlE(`(%%xKPvqjPsJ4!3l978KVe!Fjl)J4xltGir z>S5gxXr(UD{SZ3XZb61?9x&O=?NlB{@#Egr+TytnF79iJnEhq~HO9+B0j;l2%ea2r z@m0-KmDA@*n{#4_>uVKUsuwai+#kj^`t5oEc}9j`ja*Q1*{aW_^3^WTju~n0^1V!b z5Hy%zHwM>>5+@fr-7`%5lDxkbtZUJ`&b`yS`VRrqy~zF&n1S0bdmMn4utp$05p};3 zMg+3C`Ol2xO{(cjtqkp~-pv=a5%G^x5(*==jXpuDBf^cPQiAbpZGP8xG(3ViqYG{j z*u`!EYe(mnqM>P|ky-kFwhkRfqcPLNFjUG(umN`slN}cYZsO6WTA~Rmyb_0ks|#Df&!fILbe2LjYc8k^ApLZXD+8| z(p2I1USzZR$Im=zI6j|^zXU0iUtRh_ROTCKVkla`U{y!JFxFfKFw0s> zg`nvD+{<&Q_ptHEHQmxrC`o-UZeq`i1DHLrb5!chnmJq9gm2t9Lzu~3~O}W2C z7X{^2=q zsONFHJAUTau{P%5wJFWMc1V!uh{8-QmQ8t2e~$V2ad}Tm85{HdIV4jB_(Iynxmy}Lnb&2akEb-*gNVX|D?`)}dL;Jeoot-r|t+qmX7OsK; zpFlUN?Qb}2>%P-j{%UF2BJNhR0t_Rgy$Lb$AqQ#6GGAgDUs7c>kvSj#{qlE1J{FXP z+uAF_VkO&TD;f$H(zn!)1mdfR%J*bN4I!Em`GZ6!Rzlj9j=a=nTWi{Pu8BT+TUkl& z{~6Fm5R2ufnUF~cJ){Njm(B1r(nUu{Q?b*r+`rb;4 zU7a-zgpc6+ag|Wj_wC%Wlzl@qsP+#pX=TZ-iBZnJ(t^1@oq@i1#A1$miiVw0u|2pY z>I%|i-PdKT_annZVi+l|vrNaMhJ)x>+Y?3j60jp;Ga_v@cb#jv7+jlHWxof;SJz$H zWp$Mx=G?dCnRN#-y7vnE&bWUr+qS*->dGfOGa|bpqgI8BI74V!8f}ug(=Z{Qb@Vx} z8xVJ&a<fiy%E^w1nM_g@0BvIM2rUL}8ZfWyY@K7&b<=_3ixv z%_Prxs584iJkgP)MjvbM$xl$CnAxF<3Z5`JU;LCIkK~V3Ifbst%P?t9J)!tSgt|i| z1B0$D;)*E4NY=~6K=~5+3VbSpUt0jIwvb$xM6w^QGS!a4*OFbCf8%3{|!{Q|I-n^ zT1nq&6)Kf?+1-Vzv$j(Ona}ZeML;DH8;blH>QX~Jou8xdL^h5 z-}LBk@u~}HH03mOPa8O-yaMs_7Nxhh7iisZnO}HXxgcG+0$!T4+pl9O4DdxRKYC!x z14EdAm9`^FGsTaV8olqO)Z6@dV=$LRW z=&5k1f%kDl$t~%s)Lve(>CjEbuDK*)3LXP4`sy~A3iPjzk<5W%*ixvKRHB{WM@~Q| zO@Z`}JI+hHxhh0N)C4U1R&G%Bw^}^SSMnHDsBe?s8xUj7bj!0amKPrc0yDn3<@sYt z$&U56)>faqPB+-QlwOo)SlQ0%sc(Omf^65+Z9AVMfgkeSKksr#37d~;`gDnh;lN3o z%A_)EXcRq@Tk<#oMQxG8lmTCw=#4&6ko~&P>_($2$<5JU0F7!od_m}(Jq4=}U zAikXwu$v7@zxRVCXk*279r6c2zF3n#>bVT!dh{lw-CiS3cf`cFlT8`5$H$KPKr9uJ z;)WlrXo?HjP=rp+vl~Wlc$(VECc({I9KO@wfZM*jdq*2#wNc{cm6||xDpOemg8Yo0 z$n`-CBDZ1YVKjcx$V2HC^L+(NK(b3Mg2e{is`LYK1y36F!<=Ha7mJfKV8+GS3)f#^ zF)(Ku?Khm(&FkFGM!`<=2^%zrN%{x2W2BD-hy6|FtWMle4bbGEW6&L0>Esr1ic)?5 zZq+Qe(f4bNEA1uFd2WaOvsK+f5UrAzfGx^7X(0iQiSTUg|s(gKj{u1KOv4fMtdw( z2ZtTMj@v3vcA1{$Xujje8Yv~UltswN)j4K22jDR|2$jwLlYbW(&mp6kq zger?gb0U#_)BUENd;h{-WJ`0r5lQr-S6oZBOtJ4p!IgXCVGG)Q*1)S>itl@^9t zP&q|bBWhM&-fhyJVG#u8jyBQpRt+I}m(2UA$-jrruY`ksMQC2q{bEaaZ@4&cZKrxb zBh-%P7X5v`lh{n!`J)D=Y(b<~za zK5Y9Gv_FV?bM2wp7yS717fE~Nmy454)!%R$aV@vv3wp{(s!}V0grDS`0Y(TM#-5V2 z6i}IcA(1p7{hdY(XkmtvO%18%;LNlnG+kd)TPnWfJr^2u(h<(}acF4GUTOV`RDb9n z&l+g{tDT)9Z;80Sf#T!aikUBFo2HdAkzxdmDw4$PF^38v(>;Q+d}iYIh2r{N)*oe< z`Vbtv0pB(ZOG3g^U48U2R>}Ht|C}hhlG0oq3bzSX1W{^&cp#XB%&Nw~dUEzi?p+02 z2J9|^A;LM#Q#N;Gh{mA@zl+nU82eARm>?aHGKY1v)9K6zJ=CcvM~-pZ-E*uRIZR_m zKjV3fnvA0RdF(DBej`@qC&2p9H;SRp$j;KM3*L{sZc!tRVcJB;s!Tq?q6W;RUO4oZ zmwYj1I%Xiiy&>@)!<`95kTgZkYZo3=M@$~2kkymO>kV6m-q-+a3ps6oQczu;e^8Df zL)?!<)5!CgSt)8p_0GDzS7=0!e#Fj&hJv6sV%aNEE4-$-@$&Rn*R<@9X-~T?!&Qt? z+8|0vJ=L&cdwO1GnC_)@PCadEj%v27Pd1_N#e|SaHp-nNcIOP^voAD&X(7>v_G+ovp>KHb_U9iYnxvb%%C)b%;84;(z2TYl0rR!`NDM(A((BFI&yy8bAEuL1i@B=dQAhXS-&dfi(1?!8O0 zo@GC8Y^T-EX{b7v3?s~2%9|UBNr{F%a~kFD`c$t^_vBu=EP)I(P`9O9fMUQdHx)WASWi) z|BMLT|{+|Ip4afU` zLR@{G|0l6Z#8a36_Z1F(r%sPplpn5hlIRhBcz56FlN9U=!zTYF{XE#@|DOR$>L{xA z;r0elOHn7PWY6r(QzSs0;kdP6R? zM?X!Zb5qU14hla`L?f>B1r2_~k$x+A@63NG)6sJk$RA;q%Hb6=BrLu!L$Bx%^&9SB zB0>aQ9LleYGDu7y`N43kHh$ES{#7}Qo8xDpvyl1In`;QZN@Q~5O5X}7A1Z?}5P4p;I6Mqt--K^j*R zLsW+178#YsDmxul%46bz0*DKVGskpsrq*g-yWSbF%$I1LA;&dVkY=VbT3hnJ5x|g~ z_E8tRY9Dn*DxPK59&F?&jkYJ&6T50DjIk9G!DPR$@YTG#zx>fXuV8n{({CY5lUl4+ z&1c@})sVfEHbL8&zvIErp)cXON@Ogyu3+PS&gx_)Gj7|99~QmVYL#tO3-AZ>Ww9BH zrQ=4;?`A}E9`V|vFM%Ze{(yFAM!j##=t$HSAEdHMoKI?-Ol*0IPVS=)Wo>K}x8SQ9 zvs4^cj+K?17k*?uq+3(%LwHBM?_uXO4>RASp{J_lyUUe^eN)!nIhICt>jg&mYaYkn z@x2`0+?EPC^T&II`Icp#6d@?3e&kQ1Xtd`#LA=$Lm3q^ooQ8EyI7{^#4qAO8_(DP% z`EmKXikigRy2uxwh*F}Sq9#8S66k%8dR1!6UksC3L!8*6dZ?UvgX%X}*mY zF|Y2-x}6^`U|4;}pu`PZ&c|iW>0&^pEwjB(TP?4Ow<8VQR-N;^$Wr^*=z+PR23WyY z2Mbxs$ku3{-(E{mQQOcE1?s6HWyeIgwMD1$7y$E~ zFHE411Ya_WPM!r~JOdJ8_pq4*Q`d2bq}Lh%JU`lJZL-Ms?={j0UKSQd7@D|r|Z5!qE~0g5=RkX={^@^Lu5tfH``F!*2Q6|(x>akee18>2Rs_$RIh z?~Bo^ns%=o01^Zg4bEE4FjW^8i2|V;riIz_Xc1(Z(#PsJY*-{#Ql@z;s*Lo%xPV>Y z!Fv-R@v$`F(*tgALMt?@V_y&h+#@EMkkXL&oQ2#}q=%pFQ{O%;6;`JspqD!dJ;IH` zGp}=Nk;v;^Vj^XpG(U3*`&WTHdrRV`JA>T!O~?0kYynfvl1D~^57*=ys+lE3Oa{ku zQE5|<$2i~vgkgdJK%HLx1vNo_uD#UoS^XI)PEI{_P3TDTmWfpJ%~gHZT$S*+Y~e+7 z)Y%!gp_EE1H(h&ckEit`HCCZ!fSv@LwyJX4D(U6G#;F0h&zGp@(m!i6 zQPtcnq{z!#U0j}XVtda|H9Oh4~ZRhEd#(1?w1 z?PuNxZn>)JD_44>0ON6Nlq~E$e|CA~9w;Y+yw2xN=rX6ZHYa$raruIOyOer-YG5}H zpYz~;Q5Ey6f#63IAD66-+jEr$EN5#4w>b4UqVM$#bLYHb`&X@6TWg|H~xFh!s+MG& z)oc}xKjvI=U*OUi?|)+Qhbo5zLPa?PD8IT#qnNk}CCA!XE5If}-70!Y<9e80i##mi zGYY~k3|4_bYZE@djN=~No0Y->kf#|!&pq9H6jWPU!_ZVZmNe^%RuZ|&o~42U&{z1r z(NG5S?Du`kdR&nRG8sVUO2lbQMUu8XF!BrV0g74aCLuZwjb2KXqNpVZs%w}NYeCt%sZ*Jl%he!-$52@K5$(}O z=li&qf>WBh_$=EDTRADMY*|S0HdG{p>lk?w@%j>Cmab<^;g( z2gNh;GY=kz4t}NeM-NUPCF`$^Ge3(4XE}a09nI9Z$PyOCub25%#3aQ0bSN2wVE<_9 z*3`)#n35JNoSB⪚U57nOGQI==M`zVG*{xu>E5kY0p4{4O( z%p+c#x%A7h&6Sww5{<3}zuSkuZ9j4c>&Rv$A>2IPFFIz&T_<@v^Dva>xKsfBT9?Zz zpw@QRs=8%$h}R6nD&npRDXN(117OVOJ`PCb`T2USd$yqkETJ(-mPTH8KLd&Bfv%b^ zTXDG-&h{C7NCQPeaCE6%kkGx2Y~O-=RldT<1_Bj$)42^A9y&@ap!8HWq*_Q_l;&G; zZ((P-<2v3(gbV8dFn%41ulfcCc8%!vxpUDLMlx375x>xrv?7rj{ZVD*LR!KSg<};N zo{ICQF$n_Voa2t!59L^ND8IagN*1*%jR4)7mXi5GIj76I&QJHDHp(xvq*3-3a($e? zZx;D8Uev+%%l&yrOX>Z;cFd^e36}WU4A_ohat`?#sl#Qv3HG%4L~r5sAkA$noL{2U zu}R?jgcS*POxW25a=u@&;NS@LUHP`tXk3K){Dv#O)C3wS?N@~jU~tdxqC8i4wugQu zXfWe)D&86q;_Bq$iUT^MUFN0o0BJD!=DW{7$qrc*7;pEag;Lq96h{c5aVN3pE% z#wAuD`=LGJ9PLByvzkHJR&tSCnDTuTPf)O1rzE18=#}J=1reWLl$2LE@A>9rxi=zG znlHY!WG+pI6u*)v%6!POvNZ%Ep-fyc%C2n*gqxqrw?W9Pj0L%<>z(0zukSWtUxw?a z>72^3%V=D#%I%9W;S{4+r1$Wl5AT`uPUrn|x)G#4>!5An$88*&oIQBQOT;L)V8cUj zi!){>*87wP3OXxDEGl1cMST|vQe`Ka!9Gd3S5RF0KKpzNQBnOM-Lt88<-zD&6-7a; z7{6F1ZoYJtTU)2Y$S?8V^r2?Y!v|IOM}2Z^LySwfx{NAl;PO5#Q#;aNEPJoOnPS(TP8jSt9OKkd0 zFN10yTcySnv1(si8n+q(4@B!OcuCBR2=XYj`rkL?9j$|3Ci3Ie?U%3gJ3~y|2ozHDMAYcYyzuPyaeZ<&#{- zl;!C9`DK%VH`d7IX78Tu4EgQ>#hJ5TF$3WMtNa1({tZJ=eQ7S8mJMRKcW7Zr-DuOC zV&_~*N#>V)we#oCUVrh)cE``Y`I0R!yzz0v0}o z;oK5nxV?9x?a_g+54a^Eqc=qzKG;`?7~@h5M*$tMx_tSff?GWWZ+Bjv9gERwCvVkseS3b(P_rM0_GvhAIt-FY6fMvCqV*A1 zA9}7cWiHJJeq{ur>sYysUuZmC74y!OP{(25#|yE4?yM`U>*>e4EQE~DGJ2Opumpgv zvbP(QErapIPX%TBc&b_~7+`5{@!|2wi<#;4ymyMQl|^?=}8jY1fE|DuK!2QY=cMie8Q!PKH0g9z5K}qZZ)%l_UlGg z&&)0Ml?<{JS46XaT83$T1JZWBl7GaoY5xtUG(5^3u}XOBfF=FCfBx5p^cpFF7a*8< z<0z>eq61>HyE32q+oa|WB>h)5wqN3aRnb4IdlVV5Z;b5iF9{W$=y`3RY>lCw)nt@iD;0+A-&;E@Ea1z7pRlP};>+r;&maLAWAb*`}y34~XrYcDwS|MdI~mw*h_ z(;%RayXV(-mpE|MQc6 zl-_oYD^Bd=XqGo4-Gw!beb&LL$r^<6W6s0n)n+7NhBuYzSfCbx{6f_?IHA!%wBCXw zj*ql@L4zMtV065fgYt(JkNIn7z;FffV@x5m|4yA-MG6B3|`Ahw)6YQ@}!`5b=orezSG#> zUc5QG^KksvC6Yk>Qf3Zpr_-dMC<2__WA9oD9_Da}g+s-G7>(pYij3(86BpG$OUrjge*Qa*?3X zDUwkPJZs~s4@n5`a=FC8_6x=RP;&qD_VRRa_^VKF&c_L+73q60`v1e;dqy?=b?c&_ zfOL^wmC#%0y@=8Ygc6#BCLkpsy@Q}q4NXc25PC^~&>{4qbm_ep=~4s)L{XIg!T9l?e%HB~N_gmqL~Pr|sMNZRXAF8({PvMZhBq zNYCNdpV%h{i{WZ)W+{FO18+}O%F6SxyJMYId6S>8(HS>jfIPOGKYLZQd+go2!lL}x z_`ue2n?0%}=XxbU_4G35spCp!tJxx*RvawAG~tE_fnJ45wv}t#5F!45?J^c~zJb^| zcy_O~?hgM1YG{Sq!*lsB1R{*UWa_g08Z+XPZ;emyeLn;b8<)5Z)EdpH8k59-wq`6^ zyQJHQRXPcGH!b#R6bUb_3P(f_nMLz`?kmw9+zZpbCC8uU6_H_HBa_j_3ZvFx=0?dU?D~ zL0|grfA0#kdmFM|c9@U#XH3L>uGAW!PygqTe-?48q#`=AV2Z8$?-l*M+?$qC8m5ID zBo? zzV-h57lzZ1@AdZM8SS$tFXX-l{fAK|`oB4@tm^y&fBmxnzWvP@OE-M^#GL-!e@-?0 zcds_NcbwMpf0(8yTGmfv8KA8v^=6=%ePUI0@TLk!FP;t6-xW9rD5!{t32xsZ#K*rW z$bm<2kC2)MAoiGwmP1@Z($Ln+J2|D6h@DeZN(JKM`#QCtu4{6G?vX)J_ZQWtcHt3- z!c8utsZ)E;dU|eYwaA`n<7a-a($vrXz66DjN2Gk~_w&LP6}=7P=SSzgEODUyb=2<= z5f}8!-?Zs&T|uN;r8^1Lzd@JXH54ly=}by4c`adxKB$UG%9rY!%Nb2oq*E-xN6VYI zBV9#w8F!HZQP|APA~ZaUw(T#xy&_V#c7bm7dr&i>0H8IJK|*wlpqR$HAwfb)^3d$G zpfLgO=BDVeZ7kteS#L!5%;&8#*ABNLkO~*#AlG(gR^|-D`chYT7x4PTHeL5{7%Uhe zC$5&N?e(~Um)Z5rPRv{sZ7I2$Dj75uVHlqwRB$03iK^UTE`Oy;eI(Q&zR{dkjS+?8 z?oV5wp&2@Hw4N?zJ^8&z)?xADH397c(WrXa-}u;+24ex1B&0Qonzh$B!Q)J1mS1>98m~ov+T^^D?>^3o&5x5K_RM_UtLiqRAdz+;+uru_G0F0Vn51+>Zx+vSk-&(k=wX=Qj%86a znpcuz9BEBP_lcG1QGygu@~}y)vI(pbc_$N^9K27+leE%PPnzy@Pe*G^W)qyh8R4QQ zIe%}MoWG&$4#^tIRCFlEkA}9xiwMp#LA%k8!|p58N?Zf_xfe~|BNIrXJ}4;uexLOL z0w&prlnJcbL+>~!U&nw>HQ%Y&lJsNEQdy;UDa+9e<#Kivm_xLjh+|BgK3s#U90zhK zqSiJ#3^lGa^>&Dn)zlRaFp%d^jXRAwfLW>{xi6F6F$WW9%kOmv)A@ZMkTURz$^Z@@ zir5cW1~bZdRmyeE65ICk&^995&}em@Y13Sd{JD%xG>TF#1~S*rGP701hun1LS13%{ zGNQQ2jRG?^XjhZ2iHo~qWRhLW*?ZYLj<+9O9b0lcUWK65+z)N94QWVfJLo)*ZEx`C zOB?w9=JzRumzv(?2UI#r-I1ytS||Vz>{F;Vl7yM3S&bVW4~SB)^$SOS zQB{0ANzJMsuX8!>d!N{+Kle4w*dU;xDT?1nNA|FdQV)#Sm=lgH;?TTdOdO4-v~!5z zXQ}InAJ$<_EvZY%`C)<<*k03mBeI~T+;qVq785|zlw~)koQZ1>nZu|{6QF@1H{ebJ z0MrX0g&>yS%vDmUo+yj@@b}G~5C9N!2kL}Rag&DHG7Scp74WtopO_zw5wYv9nsx(b zXDhu|zGeM6jMa@;6>GnFN7Ed4I+Nn3YhVrgY<1Sm48Phn{=g#qfLf~-@E%>zPCj#_ zbze~wb}{5v|0ww##B6&0&tLZcW!ol>_qBf?x*ZUAm-S7+pQ2g_y^(h3WxIBTaH=Tk!&t-Sg%24- zUVVzYo)Kzq7U=fLpOb!#mL?2VWx3S>NO*0X)c?L(%l@fZu~?1Z6yyJf50G{Mq7!_3Wmqbt9D3lJuZcx`on1K*@)S%mzhOhZa5PYUo|%#+?=?3Z@WcW zw})u{!egqit|dv&13mG3l5*fzvaw|2#ruo3JmaRaU2j0AA_-%&_6ocuTuK|CtpwY~1YrRpsvN3Wx{ES{WnR@UJ=Yn#;uaqg> zgQTS|z>;tE86XQS%P!`oH~nE&4#cHmUn5bXZgYu^;uNQhGT=|>gl;{rtn-ff23~>k z`cx=F@E7j;@VRd?#c*9#rF3yIan2U+$NbOTU3_jb2`jZVBj-k^3E@ttym(-vg&!eF zyXUJHAt7^K6zf!C%r13)(`REVU3K6k^56sK<||A>qX^?PwL>PqeMPsuzy+g+-l~M` ziPc~_!@})8-jCGx0rx)~(FdF=lS}lQ=_N2M+Q>2M**x5~Wvp9Pbz=7ssd(d}5j_4L zI5cUO|14gyL-ZyCx=O9U*~PJB8p=I+xtMFo|8v`GnPgx4eqYndt|!*W`lO3dzopH@ zt+$pyi8^_c8dQ{fUL*5a*U?{igFK+xgKV?Xy`==3zwqt|`r*@_BnJP8*Zr_~4h2~o zQLo%)GS>T`Un(cXlP?<4Nd&}u`N8l)?aYUk>*vo;_x&DINT;CrcBzhv#8cq#99Ng2 zX{&=kpPciKJN-PofrH=o^aIhV7j&Q3Hkc=w6kg2JUU5F+s$|dZ-P>gUqF&u^)>I^Q z)j0tdp_deeh1Puvrxh#+>O

OQ$FacUMbZVfSE`%9j>tEm;S(O`oMWb!HL0oK@<6P9Czqm%Qu30} z$z~X%Q+!p`Qx0)79gIH@>$Q=pJqM%PvZ^C`;G&@@{$TixuD?MFA%Uw9e$99MeWGl5!C%#1+E zF`DfOC4}L);#~!mX1zB zcC2=(XxX#&l%HN@ty}ldsiI4Hf8hx+X9zz$fNC7XKh4ZiL3-adw6?kxHO5oaMQBCk zi|c0NWWygA{3c=jMCsiRuJKc))FL@K$6*dHcq<)yHp~DXgZZ{Sb<#}aF{%rlo)#*L z(rr@%n7hBhy~qi@r?5GoZ^_j!CkJ88_}#Fh^YYF0aX9x-oA`VnXQ0y_d`&#`C(RmP z&l^Pj$P!ta;EE(;QF95BzO2QHuLetS{AQYP!0~xg0xf-E;R@h6)z!&o5)BJ3eA6TPOI?3=&Aa8Ca;}wtUmzZJ<+_*=Mv@ZD_F|(idnXNo# zd8mjiJ1HadZk@=MsUb!C?hr-am)=HX|50xw*LGQH_*Uvki`+)Fg*9fT(^xu( zg3*KM=VjE8^VSQ>H}6UuX{4_0HNtCtIx<8B@f$$;17UEtCQtK2eYpDiOyz>Vqu2a{c0b_U zYYs-7&?bg9ZI$D~7Jow3hZyum|0dC|8Txo$Sjo-SUFTzat!j`-obqcnUX_^*HXflJ zcYe)6f{TjtI15y9UFxG`>PF+jDh07B(@pP3Zm@3QjMZ*?_%OiT*23AA$?(fdF14SP z>bJg_u(pjsMOkwY=jK zoqJ)gS*;Y7X1B#bf}$8<}$Tk!xrN# z912Am!H;$Q8OUV)v1##TrHMo7JPbLm=4@2uVuFwzwz(+BKEp5YG5jN9H=kkX&N7zC zfhXccZ(Lp66If3r?@PuQ3=EQ7#!?~~3UEo&RlF|c!XobROYt<2$jpcP6}oCUf}SLO z9fOEZf4U7O)K%1w2z({W@i<;!^*QbO2?|l-qkat7gI5=^z=BtJXvXzlJg07dCkr`Z zkCac*Zb|v9tPGQtY^9jX5G-a|mg?03i+(zBPfLjG70dpCGXg z7?|l{TO39-RcM2D2|OJbt2zc2i-KCVQfx|is60v6peWu3yV94C2Aj~lY&jxMU034a zB0q6|7t7wD=i1etXvoO3*u% zD;_TqZt|5E$e+lt_4X0`K0(M=pdRg4EG3QX%ugkfF~X3-(QY7Evey5nk4vYW0$O1_ z61-os=Ea`)!Ii$ytlH$NeccymO>^Wt;#snEk9%{|#TquiOiA_uh@yG(V(aE4#n$*V}$_@okXQEZs4 za1?DBumXRQ@gqu6_--U8b+*n^=_T#{Qk||{q22U&E*@+#h{MqdtPocR$&AXG5gro5qdj3NE&sJ{ysP7X&=Gb^3KrE9}!d^=rUhoa*Jv zI2GICL*!D59AH&^Q35C&X~W5|TiY zc%1%~yZMipT^{5)m|HIKGkAQQtUfii-Ef_}``@>!@g`NdNeC08oLctiCz}&reA3WO z_xQ*BbdbcSM4tu1{84^gk6yVjYQ9qXq|EhzG2@RNS0KF%6NF~5b2&wcj%JwBwp~E= zKS>7gEB3v3iV0TJsi6X|r=}xCr$@@*MAFh`AHQ_(tTHAxSd*+qY3qv@h4-%XI}pFr zF5|G3FL())C1-tltxqj@C|p1U55#SMab5>G`-@c#H9@$0$YP^OnNb4n~opWz|g4 zV>?#K7)Ks_uR6NL11ZgLQ2`lN#)*y|{HQ@5?_;hu%UX;ZApjY!MZFgZHHEcF=6Jh; z`UxRN+Gf;wa+-J*;IglV4W!&=ud9?8GHhkC4Z--58pVq*v&GI2x3YNL3uZ))VqGs8%H( zi}rpUE`2FKMy?UT(?FSedQG7B#S4plD z3I;0iGA^~b3?rBBRSbD*$~!h1Wd~gn^C;I8KWse9d8Lk2b_0?y+mh$tkxq=S(`(_C zWqib@wkbBZ6q)PqB|#`a>REm7+*vV`q(95nNuZN9p0#%n<$r2u9 zvaHQ-xQW}T%Dvz8Vy_d)nV@7%`%f0oZ)LV<94DGnmcnm6D$pLcF@`?W>wJ@QYG0x@ zn(gf!%V+Q9{tWVKY4U0UW3+-v(=_=DZ-y@*+b}PH=j6$MdqyYE8*!7#t zeCbhi@ZA&e`<5G2oJ4FZ z#Th0f2e&XF1Kd~;QCFp<+Vd1E$x_8h-w2}!HjuI0W5V&YY&(I}kD!S@gmPIc(+Y{m?0n!wb{DdXy;sdau`iLz z0^4!obIfqxsxQ-NwVpviGaZXjj#?E@IQ< ztSlO0(>eEx5v%V)98aEnLmaUhb`MJgQMD}2iwcz>a=P9zyIzkjJ2KdDB0w-3Ws~MW@*>l-Oso_ zmt{db^X2?1v_LKzQb7e-(u;MgwYv`%KtH&P z+i9sUO0Vo8X?^08#`bDe3LyJV;m*1!koi7ilbMLEu0F$V>>P4d_g>jdwn3Zuz7+S2 zB&#ttU4?Q7Mt!vCZ}ctuY)L0ka&Sh%FbSJ%KP+I}UjBH5&13pb0;kXYRvt{%$M=#Y z!gFc)6;jSmB^%BPB053P1ExcD(%3ngRcUnQ~jz=PE$?p;qTmEMU{kWY1XY}D-mKMGh1Mt9g5EH z4xACROwkihf*LG?q8wBTdxH+|-j-i72~V3Qk=1QESN#i5d(lS23Ujyqj>t}xR^#@p zFcZDc0((C8V}HTrQ*(N?^eO)`1_Wx7;_i~@RBK@!%3ZzBY8v8F>Je$&J6qg0@Wsv6 z@GYd&?dtx3+YbJsRnTzhyY*tH4fvqkBADBG=>CTV*|!V8Ag8Ur@RH6uxkK;qZW>N5 zWnS~v`HFlPCMl^lA_9x4O%K-J!psb*EOsQh4|v8r1d?e?!o&|UTW>!-{+IHwZ6Ps^ zi4&;osWfWC{zlj-9Pm>a&|3L9UR4WPZZ_##{zFrR#gg|N)zIfhZqVrrt*}+BK+SYD z_ABLgfG)%;>L7$O!5wC()$?r##p9#CAA`|5s4v+jfIcm-j3mpvrQE~D|3cLaEtP|C zD|MtSGPMRLLF48w%b%Z|n-=NFTSZAE^FI*q?s6L|;dRg(_!G{~z1JEuj#H4V^uFCPrPwWBdiE}OL%gbN5I*}1iB$j2@c~+q zIjnF;%veD(t6}gq0w)LR7d!Vfgw*KcId%4P*vU&+I4GF+Uqcm^BS^8p7eL{CKZn)P!(zTRizV}?X5K$;&(b?g zg3utvEjGizysx3*^MBOhU1cFQHz6A%$*!$ptzrJTxj_S$40&dwhQCkx%>6X(%oro@ zZ{lc5;>G#B5qWM_!LF$bh}z#rT^4m|kNyakgwWc-_dp^WNkdQ%xFzA!+ZSPsVV;8g9KP*@E-^Sjyf$5VZi;?bYTSbiekNp*^Y(7N z(5}36qQ_kfa43B|hbn zo-NS&;@t~>^-RXx*oKojbU#|HA^ul; zk$cqB)kP$hHNXe|B)ne*Qx#k=%z;%PSQ%7_ zMPs-$e}Ces*^=?jnLIgHuem4@aLCT|6UKwRwX)90ZQvF{L(l35nR~Dr-@ck3O3S7q z9S#>*MUArU=@GbA(YNy5DUrt+NogpL+oB65+cc@hb`bwj{ojXN>lViB)Pg_69s-#{ zg$~CzZh73!3fx;JPN^d-0Lq^FHIWst+bcgX)&;drgBtEXH)%8}tNh&Cmm?^e^%gf| zfSbdKJsUSqAc4mzZSE4ItmpE2XvD*ZTn2O!k{Vh>Ss|yc5F*g_0QHp5o30Un4maS@ zj^Rq2-(Y_P43*J2OG}|{a++Aevy(+;2fjL?&gT>1y2_V`d=Hka;Jh0s@LH&xN6(>m zbJ}5@y7EkrrZKccdb%-U_owkf;XM^%eIgM{3HIk@Tu&j)y2Dkm2G4p^>6aVfDzlQC z#_3dkm-9v)NTG7Oez&_LAQNtMDY^Aa#fL|AD;2G@L?qD^pa2(pEK_FEtCg?)_sX$V zg2S>ux*6R4LB% zrMY*{JKZ*c!=?rzV`ihUT9^I02=uO~b2G{gK1@Gu3y(zSTXY|&1|zbD4XKtx!!N(q#g=iV4i9 zscdKdl;uG_)1&fw`-Lx677wgUJMIFqs=jWd8dnG;G;b`TelrcCyu;KHWX!m66FtzM zA}`3dYUMKt0%&B{ipV#-gj=Rq3zUR<*W}M!c?Z(bw#Z{tkoD`YlMH-&I`l`^E9K^4 z=sb=+x-l3kZ9BSB3VO68KHqKWibs8CIP77De=|9$bzAm4`$e~#46w)^ePlsh(O^J> zs&##GJ_g8g(Y2!C)8n>gj^U@Ts5hOmz!?dX)uXIGu?Me_EY@-q?U(&}G|rTnWDlYG8(v!pbrr#nSoxVSOoKn_BUi_MR9p@~a#mQ!EcNnT?g zuMz=CGM&go8ejXthNM)uF$s0$z_xS)X7{rd3`Ya2IoM0>qmfI-qbys;--X52e2@B4 z=P<`^Yw9|3m*I@j&cX$@c+2#%rUit=9OL{J&3EHIxz8m&wyc ze3|mCv{K%v)$>15zMq&s=jZ&zNc0yRyHzD1Gb>#*yvY^n3=Ty5s^!sqEO_Rn|4wD% zEJ{6ZYvA#b?q7H`i}D=P=3R=BmI57WHz0>4?u%M)YXc4SaJx}Q>;u2dO2w<_MCIU4 z=YR7=^q(Y4h(^y;x`>xqk__4ZqRX|={4qU1r1)XpdrJ9)tn z2TUwI`q;DsPx><&4sQ~!wM^1Gohly)0LO*IoPtPx1=D{=GB&(O3IXfoo^fePM*2bQboyg3nNXWrfk{80TzlCxIP}|u1+zmy zEgOs3aM}dNnYyFmA*4 z1Y1vEWwxt^a4AX7tZdKpd}uC`+_M(eYWkI|M4@?IaQ_maqE3xW!8=#$gWGXfp>=r9 z&y8gF`|bzU-{?gft&bTqb=*@6dP2>0ePzA+IF7;7Tk@sKIlfPFP3U!}Uuzh`tHyo! zwEx0$R;?V-lU?fDE|buRRQp_Hl^ipxIdzp@n>~PJ4eM>dRLqMo-X)%;NU*|p~yQG*FM2QuFKp@3m_PX z?14#wWj0!(#^g(~=E~{$5EX~zppPn~T8v#{O0p}PlU2bxUC3l^+3n5wN9pdCOt61g z`HtK?gW}%c;%LD~4b1k6wQiDADlBTV)X6AdnPcNE>LVGtJj}o1G7P(UZjtWISX1TW zE2a6EAJa;&-&vRcwEyLnUtIJn6A;dN6ab*A1PVtZD!t#&3P$cI*`C7x#QJ}SR=WVV zfi0^pG-t>SCS ziL@niYSLA~q6Ku`fdaAV0>%;wN^Sa>NrD!KeHAetE>rx_(!3Y?=z*#0%$KCXu@J7x z5P=|Sqq@DW*)Bufmo#^)rW+`V&wJS%dYj&JybyS*3rnwQPSvS(DadT%a%8p@mb?nw zD^J*A>Q_vfY&xox;h8M8jXt)9XGs-uPi%1`lCrgEFg8KK&Attm{40RF;TfCgdu%k`$0A9(q12>?F>ipr`0^Kdt46kmp zNa-s3)?f?8Vl!w3Nqq?l8(o?rc#| zS9u$YOXG=&pB|Q~@8Lw*7=0L;4U4JHo)!& zdb@bi3*o-Sb)5URO8vhCkC+DYH=glvt)~WV#Zh=8W5Z(nVoZ;*Z+cU!Rz_*;)S5TB z5!6H4i`*nxv52#P(=lJH40B=$c6tJCyvQ>zPkGgI2v_V>-xQ?#GQSQR)_6 z&hX zPff|nQOVOy5ex^f1n5&psMe*L%GItQ3Wd>C5RKiwR#iGiMGk6M9A+_$a2lI(zMX{XP0{*xvL4;i%i`y-vCyZ zVN+A7JE?*Z@7F`Pz3khb7!M2S>J$qVejXAQJWu3#QET1J5j+f|gu|7RLAUnT2PgpR z+r87vwB@(X<|(w6=zUk;ihkqqopMe)F=Y(>mhC5^$b{eOAa$wg|14F1T-LI=DO2Cx zBXX}r(9t{s{-xz0s#(zdUp6RxOsZ7xp7ug?B1scDcL$2C^tzzI;V@+f2xM} z`42)A3~gURjNy}IA}N1dhGG3YRoadqbue5RIDl&#KQWEpyp^`N;8H4O7zG{^e)9oQ zVp-n|kdXKd9@AbN`U`JltsNo*g2hUHW0E(z6V*LWx6%+LXMdp}q0EyC7Dcb0c-q0| z`x4|c2UQSNS|!FVA*%}q;SOz-?hXTUS)VyA$1H1(u)k94hTPIx%gm&@kGz5lttWNV zBRfJ2z%ccD$1!LjGXCMGb^N`a*?RvY{Hj)q$Tds*h!5&mdy zhtqVtu>ioZ-YSybI~iJ+8aOSJ{XP#-^JI#27SP1{d06k*#+SqEm0{w+vZ*nj1J$9o zG>8XCAT(psI_=!2qCE!o@=%?7X-sdMh9YU7K+f7yK9#klNo}M7E*_W=-?fGYYF(E{i>BK3Kw(^v~8TIfVs4jckjpO!%+M+AlWeN^SVF>9csVJnc zgU>m38{YCs3>`gG*6Ks~1dMIL+unN9EyFh`Utp!*&mjA(qPO7IDS}Xa$a^PJ50}kU zIl!W{*~E!erdGb=af9KqaucKcjiXG@c_IgF8`rKmL^Q9Vp$bb;kKDt|_wOj*WMr

fltx|q~X?6-R~YbweFj2UNr zA;gJT#Z@tPy}?CT!vmx&K~yJ87cY;%moQgm!%#ix?(V5A50YN$DDqP$8(QBHOi|co zL8ja@0swA{-8IqM7pTcBX-&DKc|?^uP47>_&X7e}X-ysfWQLaKN$R^Ye<36{dO6#; zy_2adAkB~4#8lZ}J7(v8&9;s{dx6bnfUS4x42{rWYm~t}9i6@jhicUL^S&rSp|c0VF;DzI)EiHi@fnD)Mso@67v65QC)nvTwwmR;Yp@Lua>OIYy+Yie z!Ppbl+%)ytZTBl8vI>+e_gVo{gG-zu?#?q(=-5FuV;wKFPkvd|w(XkNq9qpCP-@&H zVtm3Zt#r!gvyZzXgG(Am&xUHZt!AEMFq}B*)Gl(8 zpO+}I7Dw8HurhS6l%V8$nU=o!(w}@}+&al9Sy zuD~x2`QlU6FaF zi~>2|(n-4gi$O%C{&?gm_VIww?9DMw?sw-qWyvgCt~q^#%`S)~ZdY)nTJmXD@ZpaI zUs_pZ->ry$3kQ2!<%s??05MmJI&vQd-p=^)cINQ$w@Y>`XNt{R>`pd(kmN8Kv5%v-H58CB zi$9&6({0Zi=^d|zmhSvrxvRT@G z`}QGXNOElP`wI8-8S8%xxFd1Nu*xY3M- z-W7(dXP}06K-rzY~@$=C({q2|c$u`Juv{&NGBz5We)%dQ^rc z2WE-)~IWv@gu3rEPH&sK4A}IGHX8K z7M+uNdmBNXeL1dBYFYZ*poP)(53M!SdOoLcLFgdF&{Zqvk*Oxl`gMTrQ^T3;_gRBM zu039$X1|IZwYYUu{c(Am}J#`E`2gtW&O52?Skr&Qzg)ZL5U z+g}gKm5AH)7-EmFbzv+HU`)_ZX=Q?|OQn&%G@p8&#M3NeoUn}K@C+$adE5q#$g)ta zK-%hi9Z45MWYb$N=sCnxrBcrPJdn2P_~8RN__q@7EcdA}s~I9D6Ae-K*fX`tOy^p^ z09MsX8cPtQNhV%_pOXtX%vD-e5O$bwqSk<(y{(mkK*(eygQ}?mbj?-N;~x^)U7>K8 zh%wu3US>r29WDXW&}_1u3*n>n3S$-&C1^MqLtS%#FG{Gy?+_}xW}`c znPI*3nP`@rH?32{nC*izIPy-gA)RA6riA)!_nG*_ujyl2qNre z{-t`dt4zdHO697R3HGE@Oo3A*fc)$c`1~uEc2~@jM9Tv)tx5{W1uz!h3+L|0EkTi^tjxpAV$nDelnMvAlft zZeX6?@m{E0lO_R=X^{Fpot=wo*P5}h{e#MBpd{Qip=z%)%?u-zl0EyvUY{Ljg-xKP z5bLiqLRH5z1x41;L0|w~m8Mxi^@bv%Z{_?${wI5(y#YKq zTqiTU{&vsX>K4f!kK$*YsOfM8+U0}vPKDb;H>p;Nuor(!ir;r;`6A^csXSqxlicRz z^p-pCh3(%!SaSSu6G zaT{N#5jplsco67pFh*)O3Q3rW6iNv^h?$-FTv6{I0&2mOxGw5_foBQNrs<}Y3uo={ zNvXd1lA3?j!{kP&($-C*B)aS-Yg)YO!Nb?^In$i}0m}9Q5blEDCNU(`V?OnE?@i0` z7#do4GVyBKaZTrKSNUYksBXIDM)})-*NoF?oeoPK$4z;!qvcZ~)&_`-t0%FhFgG%h zi+vc;JcpRNw3v}O)MG@ZB&Sa!af?MlyDYn+mm%U$ccyS9DBSlRAL=PSt5O{2FY(zj zP=eswU^bkYlv5|tDAKZ!<-}*}wX!33%APfeV7X32TUO+pC$-wft? zv!R-@(#MRwdxD{j%enlWy*=^+y>cAP)4E8S z*GLsV9pCMNyGm~Rj!8A5kk1a9cdz#3&;)HXQbe}hT#tj$ z&V97Of&yBHIv}-XQdSh5eQiq+%3l02*tW~KCU0CT`jZ>e?opiZrOsEE&7~?YEf@Fk zKSu({|%A{-l|Wd-J8>E}RS?Ou_NF*2}q75Gtaap@vR(z-7i+ z``DAtHRcij{Bcvh*+WK2JO_fNeNSxqMfr#X1rpWgT3>U0Zwv-Kksz-Byr^zvxs+%t_RG%ont2(n1{x9#dpet$OAWCR%Y}+Qg%H$DyD6>sQsso3s~` zFG>J%Up4>O*uxTaTCTU^$}9^J8pTfCZ^?Yl(}VjmibNy??vm&f&2A;;ptT+=Q9cC zE2^soG#IZ>-~#u*X~Z|h*|hQqYuCRd!WtGpr?`E4vjSaPD>zlY-r6n~WH+#W&T4q` zwj5KYy4qj8Rz*_d25Zu|W~dhW@|UUidY0}Un7gh86u5c?2F*9*j>5&^_&1 z@CdSvvlzTnw*p-VwtSx=xDn9Vs`4e%H)|!$M9H9nz73?1>U~yz0H<60Z6F2YE1m_y z)Ud~jcI{HpZmt^hYG=#N^=iY>2YJ%hktc8)HL0!iX|^XKoitpbo7eYdO49OR_huSd zTwp~}RbX$W*@=Vtp%~n@;D_H6IbykHzmx*s{e?Fde&4U5^AOvG>j^R~%`{J`7y98n zRSoz~+r$t4YArwM=zlhuo zFCbKDmb2)IJ{ed4$oi)KSI)a<>c!iQk}@Ki#^R=MHl=&+I$3%yBca-q(q(=I%_wp! z)+7v%d|`vN19TfW6hZx*kdEl%4XzRlFzC^uRiUCLy5JZ-lbG1nuCj;3>{QhirQkSA zFr+uWFQK+v5mvjgL42A2q(_`WA87~aV9Hy0O2E` zIgO>$#p8E^zd{SVkDdC{0(t#eQ?mLLfMsS!WeNS@B*Kl_rnEe9KS`;q(oSFxkkKVd9ET&|l=a@KR{CThV5gB#?e z5yOhoa+_C4elz=|i6eN)LLF(X1w)kc>z!c#1ws({!8(MBqd!rLubf!2?8zRsSbg~U z(9aA5$r@Eg+1_WM+`GIO_y;rb;@eK)PY5?=+zw#E%L>Xgn--K4F7oROXZAV$?Zn%FcUUaJ@DpNVB?8#Jk zrG|*VL5j_)Eu|Re^gDq7nTU$sEhj!|$z=<1eV*GFi*J3fuNiI4bY5eVPE2-q$BBHC7%2OO5?Xo_SMGo4T=@#RF4R2&Z zlhCvPc$Uc&+mJ9c0g;TWLhsnpwtZl%?Ln@_J?t%iD@}D^LpSn2*n7*UIGe3)kdOq2 z;7$nc?h@ST#$D4`aA=$aNzmZ#F5S=#O$anDNzeenEx5Y`hY;YI&U4Op=AHL^Gi%nG zdDpD<&iA8#++DTm-d$Z)yY}AKzBb0%BH;z%qr?S#g|mx{akqqe-d$n7mZw~du}r6D zp6}I-0Hxm~l6Bh2Y-F9U(njt+TlAPtTp9!Su|GX;T%iOJ=Wy2aBO73!o~MT>&O+t4 z4Esn3t`Lbfbq9YnJc@NL+p@@M=ujVMzbfz^c?;D||5iR&M+ixzbQl55WZ;WutgwGh zCTbOXh5s$H2|3Nry^8F7#@j}>|1h)30*f1Nook;wI!LzF(7x0tdGW~P2xf=JO8<@u8Z)19%`VJQr;C7RXa#&isE*yi z((R{Yxsfz`miU-8tn>x-zycm!NY@BLG02hqr&9V>g?LLGFKA1n=0S-wM&zO7U~E53 zcQ3MqyRUEVTYhj&Xb;~Ue}95El!t=H6baC1{&@0D9r0DaqeZL5o0yk^T4TsAJtgVP z_Oab?3`g7Qnn9^RV1$N<@HVwQJ_~P)r57yz9=?+CATu%ikynbllTClElqRs=fZGc5 zvhH7cuY?1YfV)3dpp%{r=F(vRBGn{QD3&vFoSqZUzmPNcO4E9=sb|Ve33AQ|E@#1o z=ew5s#t-w51tVp#cbfC>d$eMs}@T${wzC|16u$Vu$I3&Y*aC+^_{=Be(Q<<9f{Qfi2dOr?$U93cNP zzhTjH(14>k-?u&xH^VH|U-?JYY2sU54^4pRAWd@iS7CFEazoqavUAbMdUExwPR{_% z>zY|Ia2L1Dbnbedju01!ivhXQZ}Ew@H@2sR*QyHuD9|(&Qfi;BIc%tKXl|#;eqyQb z*>xzQK~_1H17G5RIxF^lv&_$N@Z;Fz>m94@&RdD6}XRedspWpv*U&4=sXc3w~}U zf3UU=4s6bg?UUHeM-$2_j>x093iO(uaih?0k286U^ZFKCArSic-V=d8CLyDLZ(_j{ zSQwFbS8X--w$a+M#4|hrDNY~cxnZ;MG8bc>)fMM+6GI)-Y}Z8;ve?Z!Ke}2*KtQTb z`?GSH@GT`lwP@USR=RhvWrv91VkhB}3XwrO$eV<>B#Nj8?C?&}L%4QB3wOZGi)d!` z_3H)umB-XcFQT?t776NB+cV-f4$A&L4L#1A)4l;9ilqpVEKk@d-68Gi�PH5^zdT z&T;mWNa*?KUwmE$FQo4!-=?=GzE5n+jW2Y3ix0r>4rxp8C*Qh%Wp?bTD|1s{!4_E8wlqi38ht_V?Y8_20L9RZ1IK z87LlWzU~#S#=jx7;f|Su1P!>r{a?il?TmziEIc|`!J=;iL=F}Y{ayZb&W)2OlIE|6 z(cec4vV;Vw%eG+`r9ul_b&7K_$5-jW!$a7?xnt`bMl8%b>NI8lj78jI_LL^al!k}t z=Z;lR#|dhzq;La=v>5J$#K+EU8lVr_=~Qz$4u`+#)D8iBH1I{EgwGs8FC{1OV9siw zv%(uJuKJ?|JtI38gQlNsBd8!{1KO>Yys;5-)6?$bjz1#PV9YZnT9N&0?8Z@n_OI7n zJhE%392S7dgUW2JWPbs~{RKRE3nIZgwRkvmz$<-*o>W6T$pe9=#jwj2)49QTqe#b- zL($bmI^s7jgJ1o-Ry@c|Ne^QfQ16=a)y|Tls)P`~92a1Aq?3fEznzXRD!3`gMHbjk z+!E(DmTK`V>hvP)hp}la%CW`{)Zd5s3GU_G;77ZGU^o3X%`rA;@x#1wZB?fVOZv^~68*=P9&>k1Wl|gvbr`;^vttVUvqib zJPVYAxu{sfpdpqw5Z~J*c#de!xjw1Hswp>ArI(~{SZJ`j*vKfISvn~5G_dK#%Q|Wi zsU^pm$b72do!~0qW|Eemq0K-k65Rl&jp8@TK=$g%YvMJi9JFfb6Hja|^N6M!cwpsa z*7M{9x=8Hc^qRLLf5EawvFa{=~{+86XLZI=n5#*&~g6bVH&T)GLgfA0*GCyl$X3_!>CQRkoEAQnTijZ^$j5~!@Q}Pt=GkC zjh^w3^Aa{gfdI1zK$hkpMyAGjmo@*o5uj*2y03h=|3afXmLxb$n~uB!jaBLdAAj#1Et3Mt^q?rS0Sq|%E^R~hia_6`3znmtWp6!9Xx{hJfsIG zUiDYoEJm%^gZeW0Sky<0|Cy^0el+wFvP~LBcr?Exg^XhvsGw_L+eTna3h-Y7H#epd zc0C8pc=K4)g;Tr@!RGy{aIx1=67KnQWF(iw=cj4~lUfc=>BNTQMF!;IDyDmvexh0i zr42V;N__+qnRU#mYxqxuY0W0HbGtjV7F?8idcQM~0HH^xS(K~0QIbVxa4#4NvNj5#-d-Z*pzEJ#$fJvBl=c*J^?Av$DzAP)js6JkzEt6Zx73X*&6$L6GKaO#>MYkrRri z^`X>8A@DvX1|wZ=w34#eERB4!QQ`Tt=8z^Ep_UDJBjz{xte$02YH8}NGCv|eTKy*l3otAHzx!I++ zP4p?C!biECs>y*l{rk9PoST#Yyg`)~m_8!Dxqz2+DnSbL>dn3Dj-1V0GH9?D@EDX5 z+(@mi_=MDgi1(O&NYPIMvFzq(Zm-XjBEwYKm}Q;KjV89I-;4dA%<*8T++2)CP*o{7 z57&K6*VNSD+|!{Nh^m$tnB}43XiVR%8g-e_&tN`LuI;)1D8#&2)(d4u;V@DkF}D4B z>DhonzvQRs`zs+QK3W>8NM4=(PZJ(d#yVpAOiXfy5FEYv&@g(lF%!BsJ;4}y6EYH6 zw9(NozbyF1Mmz53ft|fz18afs4h`d5mp?9}I3c+DBwid$N4OWutrwQ1Q#IV0;02J} z^+nwf4hp>fWroj)yyVQAOl@Fbx~Tk_va^Ryoo#PC4;D)WdAr5V)#Od$CyiCzeeGED zI{u;ul<&j(XW;Jk0Y<`)a$|$A2#VQV!}#i1cqi_dg&&2LI%0?i4&mq-S;^xzqfiL& z8)C-qXjD2pBr);7J{#H$hGl?=iYV!egndP0@ zUJQpn7R@{6?9@DmF{FF*k4mK(=KPIe0wDk0P(81h=U%kOzOyi!w8kyNXWHidFqH}L zRT22R6eXLxwL0v4>v^+o;9NUlU)4Elio=ggeK)GTPcd@|HY7j4*@QTZ?Z=Zp2YXX&TCGVcRmFd!7YG zo}{PHuS3$^Sjlzy<2h4Rq)nk*`kn6vi!w;qggTW7;lEzakBKFfd!b7y-Q9H z+#ZoMN{*tUFjsjYxu7vA`z!0X_@N6-F#iY5QGAs1`a6|G;4A*06+Hg3QYVKEK2IDl|I}Po8f)tXUJb@*b=4VXTBFoKjY|Lx?`evN@MdD z-3}qL9a&Lrt=Nv&EYrFTp%;TMe-dsl4z&D}yXf)?Qsa=i|C&#?IMxzgl_eJ*PZ>-u z($2z6Yq>fZzrwK<^VaFWbYNB|mA0!4m5)jJX`S*+0fEcA@Ma`X{SlCK#dY%S{koCg zN+Ui5Vi*i1{Gq_rhL!bf?{%&>)zHK@@RRj-d@!kPxiaBIiu0o8361Y z6FXUl*o6sh52d|f4uj%#X8PmCh;$H$nOUm?Ukxyv^W#Pi?E?iT*Km=CdAbMBL5uED zlYyLrh82Kv6P$X zWiDVhleR-yDDG?j-D93CAa-0HTVt@an%#jfA#YD`WG zHzWUEOBkD8N=G5{5u|xHC&L3H0-cu#J`{I_#QXgWFqa0OeaYI^)ci^twwLqkgq=Ru z>s0%Rln|T5O9%12_73jLNBjqesJ13$6#H$}5L46t#DR*NosuFNH^5X2F%TXl=YcoT zW;vTi>;2|ZZ;#YjuP9`te>3c*fXY>bO~O6hC&fnPRo}3(${rD=f92g0hErHrIr0se z9hy^sbNnIQ_7Vd+9U{ zlGTCAUVequ?Nz5O_&`=S^lt2F&E-B|cSjrGBq!(ElN%Pm@19|M?&;TA7F2)QFe@e` z5sI-okgn_Nd*Y={`WPMutBHr$;kNp$F%q!@HupNQceF=svx^m_U>tVDAN< zL=Vc3mBk_D)P7keKCx^uFJNKv1iHIO5jxLG*_dx*iy>zN>hKP~ie?SFf$}_KzhhDW z_WAP6^IRFkZ%+A?Q`S)ennSf;hB3plz)A5hWhT`F9XTi{(x}QA-FrU~-!&|{cWy20 zyqr3gv(XwdZ=6;$Kpa%c(f!gL?!Yr8zsy}M-NwQfAA}2Ru0@UjrcI7ZP^=71Oxp0? z^ojIN`>>JcKY6Xia`6!y36x~w#MM0x(kbSGL!beq>xg7&rfxHamlKJo$S39)DdqH) zLsh+x!yr+*XzroOl6J0|?*)ps=UOJ9Diu8i8{J-AkHN~Y6>d{Q-Pomtrx)m9sqwl>kKn~1th?;&Wc`W^+Qb8g+WkArjxrn?z4qzG636o_%F%ad(W8sLIHo|Thcm+fuv9lp?RkUuRP(^xSNK$J1I z(+*-Z#fz2N?xKHB$n2_d%9=Dkoht9BiNaaPQ&;L~aBsyK&)#G#Y_{0UQ40`H-|tls zPY)b^o50|_#>msRJr$Otdg5jY6fm^L^6S|TXdHy}-{j0HD~Lr1eumg1jBvjAHkn~D zi1aB5H*Np0%x~#grP2^_64TNjdSnDC;}GKV*{5>e`^e1ttqY$CQk%saJCpZ`%hu!w za$x3YR5gP{MUz)Y>fObhB;N;m#l3n`6%e8%B0bjZTO+p0B_7k_P+<-NN*Iock1;<+ zDCnmN^y8WM_002NIHUwtgrTbqoj~7{LGR2`jmmyF{h|ts(2@f6vaz)1>zN%X$D5@pW`QRk1qn<4ZuwKdw7>)MsudoP2KR<*2 zCI3Mk`ObLtUL^WV7B0Lp-!~HC(#6eF3ppt zm{XWavG~#)@>m6Ykt_Y_y-WPZ0U1ZL!W{bp*6-rh9L4!|vO0Rm-~?yrOI?J7`L*D9oGs_ebegvYZv%2D^%k)w9Izz!al z8pCI;6}(~r_ELT5R*f@f^6if!Rigd5L)S4|4sN<9^}SocT4G+%hM~{oNgw{fN;bMh8Thg9)S)j&8UB(#iau5DK9?kO2u?!~V zAOtCJHA!ADK>L^W8a43^zC}moik-=Tu5}MQunZ>4eWw;vBNZ|h$AV2tZTQ(v)v!k2 zASbUZ>~|5?zTD2Fbwcw#D05MhL7Q`Zkxb1}U*p5^Ma5YukEv<&I6nll;3;wpr-GUJ zeS$}I;Ro&15K3UZjW#xo##ZmjOtn%$!8CKdGOXd%eD3S6^PnQoJ$JeU!29Re> zvppv(cGj<`9{cnlvHGt5v<*LyzPYC;i|L0|1aiVx&}h(8%N;XRcuT-#!DU*2-6y>% z&q8&lb9^7NF5+l77GQW5NourjEpLtadxFw85x!H4E63iNqKWLK(*fMSxGWKqU@u<= z3yWY0N1<~B@;1}1ee6n^3iu1{IQm{m5f#B>KrIwTNvKXA-q@anFOd6+JYb5Oy zgVN&a1zl*@v?mW2k_1MGyoc^W7;Odv#+wZ6Pj@7ALS@LK$3!3L_;79{iq+`Vzw>_7 zM(N|L|08`ZEf-wXI0vpP6N@C8wzLw$w>4>|)GGx!4@8JzF?l1RLY6GhR7rKixV1*ZpAIR?L2 z&oJPu&|_3kZfPQ7t()GKpUUkksE1l+S+suj#_%d$zfsDcPq1nOFN^nt$OIG=)F7}* z#rp{_zZM`J3bGBlem^JNRO_8I@f)hOo=C+zOj8%GxqtJ7Gi11luE1R3cFs_F?Tb1F zb$%=1k-Oyk97UCp)4|<1*@1;CCElL>;3e*E!)L&7Kn~EG0I&8n_d?)aON4VAXJyBr znoIx*uA0)2^Vjy2)VA8q_VnxVMm;?#P9SMn5v$Axud$Tp#1`A#J&p#UFcy9C@O>^T=@UE^&oE{IOsZ(*cp4 z=HPI1H$jB80s)KjUdc-a&hz>-Z(N>ws0F8@Fe|*`iG)mIrB$x)r(hsMFH>GFy*t=M z9cGATagyrgL)l%zFpy;!ZqGSjIf+%2v!Af1Uh(vzHWf}lO^C@1F!2^$uik^{_$I{G zk=(a7yl}M&bMQR_tcu5Gi5sszj}4ZU<8MuS%JL{FFh=^A=i`FD@Q zo2_zt4#wwMS;?sZlI;w<9UXu%`67WBs9+(yYnkOKL1Q)oEWBI%@wHLYz_DEA&jLc>Es_sJRk1ak)f+ol`ytNe zm|-cQSVxyS6K4dxPT`#EaaznPPMcc_#TtwC4zLee+_@)3S1~Q7zote;JF}!%c*_lR z9(7=)=J8~OP1W={3-nL==T2nf(aq;wRy>gsvl|)g;M|h^I?TKec91xs!#nTw<5KzI;)UOz8DXI?YP%q`N|1|*?ltIioB-*^cHlXdB*aXT4d zgPp|+0!@u7n<}iqkS3WcqS`qT^s_{1>jrYJL&-^})VEl*Tj1CVHf8~c74iAR2*}i| z&iH(sKtp@pOAx{Hl+&Qi4iWlHm{kSTx8x}fS@vKRZ-OrQHW4|>M#?EWrv$#Hmn1=P zzkb2sH<9+Q!T?qfPY!rW&RI2<`Dd~$HXs~mLCGrUmNcxK&KdFqlJ$O|ES(k`!8NY9 z72?#__So%vOQwB3Zh6WckREN$GaU@!4gN!9*W>usJh=MEiy@Q74I709MQ^udo)a4Z z!Ml6tQH7X*->FV1iN?CxdH4BCt@+B7te|PHkz8RZS&fj4&JK164`{$6~veXU`3wkFF&$lGzgSg)1(Un{~SW zwXwa~JDt||1}$9-^{$sn8;1qc~LJ_xM8(7f>gW=6m3VfHK> zSqi+x>(DAcnZilylCMrd^r$joW6NY3Huh;o4ra42zGOB?Lf$yY**i5J!)$~8LTeY{ z;QZkf6-k`(YzVtRB+?%@p!7)KQCkPOZ#~qVN@?lKBI2;+;By&w27uGB52tcn-$F^} zgN^RvPr^T3@mD#X0U7PRpQ?Qo%|KNatx_|HOX7&gcxsJG)^$pDrjb()V6;NTv6Gm$WhH6Npamq8+!W93Kks;;&AibG6Bn$00hK;t#TwpW?e)C&n*7x~8Ax z#1Yr9frLnMr5D&q54OlJ5rx%DR_0}7{4_J;Te|#ktKb~sJc;C8?3SsP^T0y7k+yD} z3&G)2TOv%JA>NK+e4-VEG?o{LQuGcRYUkVh=H$GCLBx6KA3J+y>K5}*;#d+dK+8{Y z=a)Vrr?I?xhsQwkqs}~+9EVa-Q0SWK=c+Y)vvX1XM$O!0QkMkvZ=q5tT+;DLqGa;f zJG7siN7~>atNnfAx5Cd7tnW#lYA|q64)E(l=NFg^bLS@N$1+dk$v7AY*C{5&TJ@4! zr#2jZcwqPot*pdhLk?4DfQ- zehDlNJogHnB~Nvu@)?TG*ZiPCjx}l#9salGh-h8zAf`y`4v%T7mXL6Y-~sXStaIrI zxB}mRo;Zz*;W)AZH z$JVjj<#=${Bnu(VL3iW?Ix*w)sQ`bgbA~XyXbGTw()1Uat^h~B_q+h;MC>m#on5mk zCTbe5!eoUT#@Y3zn^Fxobi`1z!u7}3mv7cudWP#FEZ|+47d^jekGyi|cCARQXhPE>0MXENZgk*0#_CdEY@aVPOnzZ{h)=(#UpZ5D@R{ktc_hn7N=5@y|7CE#caHCfJX4|d(X?kH)i>*=J?67fN0D_2n72HTeIPE4D? zYlFdB#Hp6GH2^3k7#U12V8%{xk_)kN&Hn@i!nn;KU~!xl!ESw+yxTMJ_0$Y|D2lL* z|4GD}ZdAqlhl`jXnu)r}rneG-^1N1~S;Fdr>M}X06X^|zb|;$Y(SIyFKi^eNpd_E= zkirtzL4u&0x+@u1%olA|6Q@nl@^6|;vkq>o&DR6b$%3#a%4Iwff{XA>J!Dpf54BN? zjXjJi{X5kgYxaSpP#LxQfbYAeM!#(mnv3b6yFG#W0T8b7%PBI#+5~cRPb1ZjVuMI7F@T$* zB`w7a368gO?}Zz!R4{f5M2c1yjM8QoC?#AnnjKl_K=xLDoMhedY;I!4rtvZ zz|1b1D)7-vCNB)xLd%y=H)z zvlQQ(z*bAI+fVweoRe~z+V#P$;!NQ4l(5`cTw*_iQ8@# z_;r>Cb+(QNZjVTHKc1OA=?RWt4 zw*Q@_@VlP;Hm{=qb|R=lj-G=c1-0Qs4JCmXGo6o}`8LW@XO?#T74*n2>(bfKsmp&Y zs{(Q|B2+ClmFq=^Y44yL0UGypLU3W2$yejD7#}Zl_U}Y9@6dD&q4=!ZRTM48V$@kr z^!8xb?OIlpHeO!Vxct5R&YdM1)Z3EJfCHMZiUJ55K%_2-E#LK?8y4}sb}yOU667Sk%}NEv)>VqAO%ueTS`+GX*EMFykT3@@H0ce*dWxvXhTz9Y&t{EZm%Nm z5-ly4WyW`etol4@)@UIH_m8*KiwMI;zOGRSnQu?ej!&F5CV;lvCwOB;is`xLo+YL)8RP+S3XDwVRLLb8aLOV$C`H3pj5_%7=DL}P9`wOiG zi`6GQPy6#jv;LK{MPS;kU;~HRmiWdKRzs!r)_oT(lex*=9@o7K8&RKI0UF6*%7g-$ zc$JYN6{gqPWkmHA;~)N4hyNS%0OTk6GH;9JQAzu;E%h&nD` zX#3JQDZvLB+^3m-dg$-ZaS*0nOttavV}qNV5rafeUBpnS`2YUeu+02+#GoF|l3~yB ze~>02@Al8zUaeYm06qn7EfS$5{O5Q7LxT3fpC2e+^BOb%AEo*;2u+UDIiL74>$dKg z5{67xMvnH|h-uEeOCX4GrK!u_#o|dTf$ZL6D0N!85q?x8TI>H96E!7Y%*X=S7>L0A zjva#}f6Mz9T6rHw=G}HmrJ3QEkl8%zu!$r%H!3iV_1ya7@*BenvNh6;gz}Xd(X^1} zdEo7ali0z3zsIDLv!nVRpp*jw`7xf0#oI>wJ8Sdd|m*9NwMigB@AVc@r$Dt#nfCDhyfg`z6=;JT)RgpAJ?u87+W zdyiQfu8?N8VizC1+6K{m_cTj`|M2~s`Q?}PeU1(aA-!bx)>x0eLDcFr`qv1|&HL`* zmH_c_;dTxaRNw#9%rHv}Bz95inn%p<{HGWG=}guC2r8HTHPTW?gy`QzGLJLS`v4)q zYHfq79IO5N;Jjx)Io{?l;3{+(JY}!^_jU`;_EHbZZqZdT%C541{>=ZqmuHlR6*Ryc z(VR??82d*y8*|l>5TY1tWAsv}<>(SwEM~s;A)LmncbYUzWs?--)TzSWo=PgFI`0?r z7uvhd56LvCs3^mOn!fMUWp?XUaJZf$;{g*MVsTuVd+a)90LS->xjc|J|KxEcF1i$r zx+>beDjC`YmINMgS2%1@Jl=*2WS&eSn6#qW#fPj#9w#B)o{`!ZxC<91{bFEWi58By z*tHq!pQW#5_jM0IXD^Era(Y)Wf-QqJL>E(AaQ7D)tjR*o%pZ?T1HN}cMt9EZ^F^tb z+Pmwg_B_d*c97-`@*;Eqmnc!{ z#(9g>YmfM*sP`4>-d_bR;4%KL0umRXAoqqJBG5uNBk}^XL0a@Oe9=x7v$2_SY9p@% z)8P^0ToS7k=IteqI4Dr1IA_N_MAKT{h;|3)S}eH|DPV~RwUV@7jzeK zj&Q#oW&xKeS)BDw_mk@%^(Clm@60K?=qRWNLZ}~6$YiBJgF(V}pPf-`ad{A8< z>nHD-u_NpXntLPD$DI3!lJ%5d7EU631NC8AnN0;hqP8+>etCE&FlRmO&a5(0qc%4^ zKqJ}0^ctnO_BRs>EK=&LQbm2e{@XNJGyAcf9w#2!!lr-VOZ zOQ(2iw7k9hE?9LEv%a{#k`}=BWd5vt1L5s%8~xg0N|$$HTu6+&>boQ#bBb2S z4zoc63&hg!Vv_zYqIX96wgFKLC)~eMkEF=hIUihqep`SJQxWzc zc&qyi_?-`Y$CZDo~qX zljXNLtO-)CWvp*jk+p?4tjR6j4QtQ9YPdL;)P-U4tGBl82TUsxRvdP(TRCA- zO^%+w7}!3BDlh~}pKp&p&IeJ6Tbb$UEqPy+AQbVHeP=O-&8H-%fz?A#%VKkykTYoP z;=bYqjKaLL`l*PqTunf?=}T?V#ZNPEHRXnHHdC0}px@Pqo0>WjD99O=s4SaHJD${^?M(_Pb3eKt_Zu4kl4LT9{v- z-OXc9!V~F&?<-s)p(G?Y_)&NteZoK9^sElR=={dbN7H3-U;-qz7m~e~cKNGC^vbk} zcr0pOr2!k8-ez^|tx#b}8<|m8g5Sdhn!ux3n=zLp;ybIEOoLilBc2s%BsoPIbo(+T@G&skXBliO5%}w>ZFwM|;&~HNY2=->YgQ!i^NF zZmJFra7h4S`~!p0B)T|)Uk5O~$)^LQqn_{E`1;R_v@~bvU8YlS7yTo4C-=455Tk`N34Dde=K| zxUZ=Lk&WM53Af?df=}^n;anQVv|;dfpHeh9DeqM?SvM(e?R43|rb*m@IU&iMs+V1P z6+9EceVvts$my!I`nM!f8Q)}yHtz#^XOHV&H{HKJ-~Y48!R}mUDf(oC8_|S|5y<#C zt-T{2@Va#z%3#eaS9EU)IjvFrOqLluW896KmfN>4$mL({i!H2|D#wp(-6l-4P+`8o zR3Fb=lQ|-1A*u<~NoPZO9On(c-bnVg!qY3f2+tm28~wCciKoM$TE~A{XXtkofEQ_9 zg77goi(Jj`qCu^O*vT4mP8J5!9*oq|60@nxgEb<)o&z$r@NZlpZA(7=lmhnIY;UO!i;U?bJ((vW{)%u#OT<*eXV&~K2YEfDju&KY?=J~4J`bo9-y>m05zDVeUqkfI)}bo@aeQz{GIaw>C}8@ zVi0@cP_<0ZasXe8)8J}<*siWLnte=RCGkXFN>)!;_4d)2xKkp}i+rEvkW&Fh*4Oi( zJ|eIi`~E52cPI~bRKj*aMrT^PjdmicBtoEGq`DFGmx_5tKk{Cbvc!1UbF$Yy}~^Kpr#M4hS`d#TUB2rc(+*j zqk!x#p|sZM`%x@xUYyEXRc6^;PD!iWz0400_RuS5~AeB!LW>v3;WE}KC~?(<5O^vbm; zPZL_YCE%;pWBG%4E)Au6Rdb7zOgA(fLxeAX1+}oH5<`pVPi|%qodg#S*ck)F(Fdxq z=xZ=){{&%{uXneH`n!C|9R5<(ta0xsCd-}Wgs@U(d-nx5uL%4;B}0<;LM}5k zOR0^iLo0;thdv2+$(3OS0c$B=j4}0n=91xyhSx(Z5mgpmFbn3dSp`{2-E18uakoH? z=jerok|1tP!6+&%CDB~0ob0qrybgZn{f{JXv#PoBQxpniAu1&nW!zT!9<9<(+v29N zvNG-Q1QVPI*`NFYaXQ)TCs#+P5RscZ8{a8&*e2&HK-D`zc-RL`KoW%c z$>lr@jyfgk-)wO&^!@85&R_#(N^PH2hoV{b-g7AP+Sc$q9qw>IHqAJfB~h&oJj(IY z=~iT<0ggWsnn(b=PezFMKYDkF$nSQKkArMIlZYd?+(J}w%!c;x&@wHVxarLV<##hX zTcFYIhw&exvJ{?cTi zc|I(l{lww!r!wY`II!7NYnbc>en!-1fx^>fxtyKkdInFZ`cx3_T?cNk30v}cU>0uq8|Wh zWZ1i6AXz4>_{&qiFKD+o$dUfOc!%}Mj zJ^h%SRQJbeB)qgP%(e&5q1ON*z$IJeGfJfX-lqjSr-NwAPqV)Ot z@2DuOr_$H&encyjYXXT9G>p_-P!=Wt3|f|LYgl=#b6hoANMiX$gbleDvllVclzXNB zMd&N>#Iw#KY-g;SelLnW6rZG2@@MV)Pm7Vap+||Ee3O5nx%xAN6#ces!M*qkZGrBa zb;ZMvV}*O!tN%{@Yg?S!FaiA4mF6o8I0o$rWvtHvrgke3gsdc>mty?+paEDhJg{bh z+QE4!$5ps(PoG4N`kY~E&}5XSUkkAVl{fng{V$ujzM26~{GIpLa!%24LTne$;^5L$ zNe8Svc)@Q!wmL@jZakaUe|)F3{t>L39UL2lVT`Z+M1nsc=K8AfA3Epd_Fj};(}F@I zm_yO07tr;LbcRI4vm~mHq=VpBD=67gjsr#T&J!+{9Qo+2KP@1j!AGGTbgfi@cHOJP z2_4?|PiDMwBPCFO{nB(k9)9e-h zLKYM9EJ%9&J^Lu&M=^{S(=dl4x`J--Sj{0A9Y&c!ae};075M5|Tzcut7RE9nIDDi- zUr0$Q64f6@lK9*&l0hz)GW+q@Y6b%C$g&Qo355XrJEYNLAz%CLOYuGvw@l)S*aYmy zgn#cdp&mNI(Y#XBt~?osIFiMU*d#rz{gh5*k_o6&Z}!#Fdk&SjB$G2oK`0q&h>goAcGx)rff3djOtq*OFw_BU{rcSZId;<3AR_ zF5hz%jLrjC3a{n+l7cuejNBS5F^PzNn3`l^Qcnq z66|S;(^jI2EY6$}@>Hm)MG9kdn687gg^jYXOC!WlB^}-Ayrvk6l?kNVB}r&a{F9E4 zT%au-Ch=4=tsFZ?BqWoVp9J&zL*J0WSNcEICXI4G;vY&I*)gwTnqq2kyof{X+VMS# z)5NW^3BC=wA=Ps0KB>$kwa@YXJkL`?DUZ!wGCl|_BHt+A+5WHY{|{z{Y4;#+&)0V^ zshoSq=pt7O-o|8pN7$m?U{#b!eSWLte)7;LU?)~vRyR!4lEaTi`dG}``lCd1f?SmO zs7B%QnNj*B;yX+9g_!p6T+!w&0bSlA8S!FLLcNkeHg*9E%u7nwO#L61ZDV>R2Kh?L zeaVFRNW!f2AG{TeNIqD^as&&YIKo9)&=Cl{j+tUQ|20$=+@}X||LQ3CZf-Ew{#`mV z3lo3z%`+vdiKE$Yxk#Mw?B31VcZhiv5{!7ZXn?04GLm^)z@YYrRMV?!(d7CS{pXRJ zLyZJSO9bC0DEf$07Bp?Il!YD3Bk@m5-t2a?sg;ls$+3%-)rm%)M%sK((2=#%SEzKt zmMW2?MzREfNYTUJ*1+{J(GCRO`G6BA1$v_c9*%z$3pTbU`7qcrM!xywr|0-Qo1z(w zZ4D>qaUGz&w5(m=(d5u=*ak23w-$4Wy2RHg`>qDDC6M`g^fKl?D{S+bv@@MT^E20J z$@G)#DIU&MSyL~N$#e?#Nk%U2$Bq1KaaDsiGKfsUtF2})W+Q5iPs^6tuCJ4hV?y{m ze0kPA%im);9-P7RI=rBfhn!8P!DasxELvLlLIJ~vxUh5dEd20q5W+)=*p~YP-xT9U zGzEapu78PG;9a{9EsT!ZwbQ+^h6)tluU6`1`-Mo(lYjZQsKuNkM`&{NeH{G@agoMf~`! z>VB3*ksX&OX9rQ!!e!qI%0(ai*~0DW9{gYPXz3i7!G7IT8B)HI(RTA$Cq4NsO+d7n zeY!7aU*TB!;~xt>lGB>(>w|^U@+BSfDYokHM5%;p%-wPP+Clfm_wutI!A1qu7^%j} zQ}3Vg^pILZ_-?a_N0}CUVTR+yV#U?377mRoj|eoC@9?Hz+^c#sIw_ z&L^JEdZ)Zn4(qeGUBV7AyxOKRQf`DN*4gNp+4Fwa%$0m^sU@jMPX5$>)~svI1x6J@ z{R~BI=g3v@gf+0ypi3jxy>gZ{Jk8tvmHomjZ^R^>kfOd<4ak16FJmnBtm|-rclgbZ zLRZ77M)#6VG8Jksc|VsmWtI&1bzv*<8L^ip5+T;RIaIt<>aQN<#&-DHXCnVAT#b(! zg(QYv*H8tNVOYo+cITF{xn<}2HM7#_$3o2>nM|hopNPf!+C-J9YH5LRjk}~MVbubH zOP(Gt#B;Yw$Zi+^w#rKez&h(f&u4T{J8TOB$5c-4WpL3J=G@=q^=?y_e2xEsC>O3S zHKJtwU{7V(S&3oe)=0Jr5qlt7;%k1`J z-1uGAY7M?0|1a-|1A&_7FtszT36$^r%z5Pw<8{fsJC^(7&J2?&{>R!e@P9(mHZxST zK=)>3HO&9b%(ElETztIXN()N$3Sgi45F{L)+J}OAR|v^#j;fzXe7yE8Q5`Tvc(_YP{h-P%Sef`EYZ zjtWvkhtP{sr3RAFt8_v&R6#+d8hS6G*ARN|i1gljRl0Pk3WCpx&wlsb@4M%mZ_b>V zGvCbj4-C2Q49PF~t+noJUF*6;SEZ;G0z#EFBB@MAa~-kb13Y~@12=fgo+P%I(!N=_t>VyJ#xdM@NQ*aS2pxuV*>bm;g7coQ zmJ#-Xh7rm(7zdfhG8~RpF3GlYfE1c4oTewDd@v2Ey@80IBN|1cb#zKQNcBh0xghc3 z3^Pn=`FK3fc!ZC5;Hdcg{fVy85mDY*nEz~WcyHxJVL2}uB0kL-FcGC9lU|K75L`%X zee>`QMufE+DndB|?9MxPf?RIC@fMs;(9#fdr=?c!!vRge=4+oasC}(4e51J{{DYK| zu)H;^D#^&oLY)Fu_@1ERZJh85g#~2WgK?@so%cEuR#9 zdfh$CHw0Y0`G8i(`l%=qZOq4Ew*PfDQx^!Nw*(DfNOFk>728a#)It(B?2Fm~wW@H3 zdP8jbSE1{d+hI}J!W@chc#~|O`I>Gf;&45V#qH8|2c zk%`{_haS8Ln&cQvxF_jdm`I}GILkSIiSKfcBO34J$`|_Fy5gCty!mvly_=u_{gaJ5 z`KR7Y&OeGWof!Mt{X)Nmpk7Uv+rM&bRmKm|b~NqY&!aq83vwEd9%=uOf5UdU@Vo1! z8=W@YRk=~yj$NOjx8Kc^4fQW3V&hQ@_lvsHAuD{dY68oRZv;a@#3MW7Qa@=&`4F+93pf@ z?Rbny>J=0mHpa)snnGn(4rECp-XS9>?J( z+hV*@v7tRPkJ7CrMNJew`=lVoe%QGPGmv`g_ZQY)PupFU0k8|OQdYmu!hN5MsKXr@H2DyXq5UvYLT_G3fImS-4?#L5% z9ma@VARR^0AjAd@WZ&W3UVf*6XWDh33`-v+k+pxuwu<6CQwQcv(P{%V<=WpQ`z62D zfV1~`EDOy(n06_$MEeH8QFvfqb&R>Wo2i$8(32r{J;9%fA96jJAu1p|I*)!<&S$rd znD?o`$zi*k=DW>>bLM9D9N5Cs0o{$z2Z34AY=_M|IiJiuj`Wh{l)*XNSV~S5Fhq%3 z-0Ka!Eo)V)^W4x0F2vQ(>=M~T@iI}WIySp@kV{TiPfn?BSZ z-x^oM`GtgSLv^T-d7U8f@GPdex(L>q^7F~)xa1e_Fi=_s^#ZOK3h!b?XCgcpfb`Y7K3-BjW%ZJf=PZwyj8h zO_12qEfZ>%PR(HA?3q0fZ&WT6;%+{U-N0!A4I@)4aT+G_`av-b8!BW_!8eCL@DNtz zuH=u3R2m|(&D5v85{e0?9%@U8B_Y(LG0OmvS+Q^fP;BgTrW)m?tnqMp*k0W_zHSaR zH2y(+%>8)bX|8kml*A0aah6VBuAb+pUkZjPNR5iD3tYl@{wQGXYw>oX0L#~6&?FrW zNJ$MinCLE}DysbZ6+0HLYqNLAd#7;vb2PZ}0|s9EkBJVJhe*u_x^BCO*BJ6zDye+Z@YH#1|db z!|%s)HFig+uKEkvj~W$&YY&$!IRLpBE(jGIaOuWuxCGDpuf-8k)W5mi>W4u+$0TMU5R-rTSA+r zdlU6<4X=WY9%)8@+)p2t-Jh(JiS%TBc1iG~H+Sp773;`Xi#uh7vQn*{hP?Nm#=;wN ztor}Ls-+Vs)Ernc9&Wwpy%u_M2}nj!=z##a2sO2tPrn8M>JU;(Q$Ja-hA|WeJStWg z@pZG$%G=#EN$?y7QKLHPF?yu_MM-toq(UnFKAtbPmMizWdR5^Dwzirbw}KSgg*i@t z-4Kko)A(I++@|DF>ON}d?jSlJBWck1;1pw7{}&cR#|@mNYxhoBxo|6J(UrGnnNH8B z#mq;MXf}7~poK*wJUu%N31WLD_aIe{h(Kwi6zw0fsXhUJT>1}uO1tSv4x+*)ueyrx|7HImMWwoNAOz@bya$Tcy zKW{U4QsGXT5&nMbzJpO4e*j(FMz#aylRAFGL3YMxG^mus^>WdSlEF76!KrCYhWHym z&KVx&A|Z|yHE_h3if7O9nOv?>z_;j6{BjnhUTuvY0@;=Bw#BJXHOwc_u=?G@%ODCT8o?G(Nx{$gcbIPh+x)=Dk}O1c8FLwPDXtsk1KUsB;3m|O9k9YYYLJX8ha`|wN_l}IBTyq z1mv#t`Ao))1iNH(@dIPy7{F1`(jFL+0`j$MS1l4-*=&Wp+H_Nq`z>CGsb&F{2-Zk zo*Jiqmzra$V!!r>>h%G<$ub*N$*C}6V8%xm@#jH&()@@}HH=nwr*1uRh(=j}8N4m@ zKEYD>aPG3lrcCph0Dq`?s(7ZcifHC-3e1e3y?BtwU$ zAo<=!VO;o-Yvfqratv0X>Yl+*5 zNN_J3eTvy5ubUCsRx_q2ws1e0s;67kKglU}!;zSolcT5H{cAd5MRc0?c#^mE?QBha zg&|+llK$ce72pXAA;I$=inx!@l~!s+RO!0RpX*%qVN*o1WS>nf5N<`JC=Wr7gdqH4 z!3}Y11v&~5F1dJyvlyCP5?>u!Z7yC|dA5Mpl}n#LmdG8#pUc9%f+Mg>dke5eEJKzO zOhS6DSdEnubj3RK_AOKO$MasTxoAlUCaL!l9EL~0 zq3$BuB2o&=x~)E?*=v?_SxN4rzN=7&)OofbG-ucsg1AQKv~`Jr=3vX#J@2#)+r>*?EFH|e!b7rbso}0OT)h{&MBZ|*#n=4@4SJo-6#rv4J zHzEiyA!xz+^1(TD<}iJrQ`2+^!#mfB7)TW;+j6O6>|^lEY0!@?4Ij!pB1N^?w<(?t zE_w}AdKmh&eG!16*E~PtBpjar_6w+@XRh z+goOB{289_3&KO!)MN04PD1HYNXz^2^oo1dcuzT{MRYZ2V8MkxN>VfLO^3bI;9mnN zBHvt}DLzt?{pi-mQ^a;8NoC1lGi&-dG<#grF=+3dG@Y!TC3<_YmZ5M`^FyhE(0+@g zB__6RjdUZp6>ZBb!CXCkoC<1BcF^T4<8-NM80vAbKSpfU=QJQ!91!I`swp-x0;I{h z9L1LAE3YV4Fnf{VDz6?QKZ+i7vTF7h)_p$?(^oo;3#Z4s6ZRNNjYS<`Syg^j9d`K-u*@sef~~C& z-M;PW9dfFvaoY0h+0esUiVPpw{5dmYg@h(0aM8pUa^O)-EiN5w=6^!H*szuBbA1QGXTGS~>8>Dp+jt8ZSsAx!mHdPx+;4|8F zZ0C?XIKrfYTeckBJVLCv9Kk4-q1S3!R_DU<;PGT~Dj6{qmmY^m ziP0-&$R#JcNH66=VX&!|Vo=BXnSoDZ*7z^^&wQSYX&&&P8AzMI7Cm~s4MFEX-08fv z{AO)#0+Z|S6-o?B=G09>#S}=(M?yB6SV(~HL#MbXfa1Rg|1JV2Tlx6z)F+H^&D|a` zozblU>T3r4vy2y8{+FRhlPxIj|coM?@J48){lmC3#LWnB@nisTihMcdto}=-18C} z#rN2+eRO^S=9-o&*{*+={{vvH3O+>7J~`-LBC;y|<;# zx?NE8;P8@*_UMFI>Fb0X-27vw7PWpgeGC@c>ELx!F&xfM7G(rn zaxs1n?K3SnPVLY8M@IthUdhJCp7H_j-uf?AlNv)0tYAdgor+&L&%23@Lv0Z}%+v(& zHepn4X4j+Fz4;;=9aqp?t84BnvjZ1__5YIJ9j^j9w@Or*>RvzL(T=*Nx zFiObbp~-V+w2x7{xA_+s=5%5}cJ^7!^!ab_O~AXMXWHY}FZQl6m+0=71j3XclF#4 z=)?QE6l5ivk@t}5CS%2)Dh7T8Tb;pWGD3m1%fiA2a_DE++Ii{8%B3pdsM?SZqQ#QCrY_`N_oaKKZAL`mARy>_1zGv^K)S%|T#FjbxW! zkQv_|i!{Se>-|W^dYk0K@Q;U6kBf`+n1s_WfVVg|@~T-5{f}m*3au z(2u@VZO^-mv~6`I1^Fs6%8&~-w_5&E1K^|voW+TXxe!JO_rP9@F(irWA72Y`+iGb` zJ1i#b>0|ewme-?1JvlhCfUrbb>d!3t^=R0r>QGaDVIezH6GdD`cf0o37+}KEc@Lo$ zCRCBsmw=HiaUQK@iMY=VC{65*M#)LkvN2``xDVB7jW7*WQUU6EhgkLX$EoXfS2~4$ zN&l|KUR-va@_xGsE%pm+UC5hCNzXFN*|%NGKcE|OMJj-W97ffQj`&nzXi-xhQ8Xm_ zZL@gZ8}#n3S!QCVB?CV)F80@0j4q8{rWpFw!B(?yzUc!Pr%5-M(v8FRK*7?f!~j?_ z+Z~ZWZAqED+O*k_pLY%6h!-^J;)_P1%I}+-BBv6^hASeV_6%A+aOGxJ^e=htkED8n&jX{SRt>aPLis2*X4a>}8S>39gs}vu*VVkoCRIf%$v0tl6^<>YNLoaSINr z>njc&qB(*enNF{I^1d2gl)P-bN%qkX+xcwVU0d`u(2P=7P}-e-(Ls{B9!@}r}d zL}3xh_ABOheE^}DDkjezy4=CSgJif@q7+I7n@%)?PQN4$Ua-o*>TN-Z=qpil7ETB? zXIGJNEj9IYZ9$&wlBwfU2^Y^W2uP8dSEK6=^x>bm$8*i6$Q5pvuEfX_KWEnucw1jM z`$Ty){qi=Sju%z|En|LY;Y(Cy2^}OtYj|{`>0HY0empe(JBFLLr-fY-kBi~Zxa)7P zCEi>JmN}-^7FI~)#@DfC&+j8Fc(tpVTr8jG6fiELnZd@8CA3DJlnA^SkxD9*stQvRvuv7&$^cW|Z~SWs~I9#3G^bd57R;-KiOKykdt z1_z&{k+i&+Ok_mrQ+O3BqfZ&WHOQ>d&XPjnF2Em1}}^lG}e+QcLtbk$s`W91?o6rH^G%EG4tm={qX_h`Lpox&|CE|4en29 zABt<$73%v%^cl&usStgk_l0`*B_uXnO1T>t{aGYD>q_F1lP>#pM-!Bb**>%;Y|@k< z;r!UV-%7OLt3sTgP_nf&dP$tZ5zn-F45X?}g8Vo+ijLA1T9DR z8b30uv=y4A#Z)v(s@H~BdNh^c7$M%a=ZJA0xh`w;UsxDCeJg^4*%4V5OiPU@2<*0hSO81RNZOI_J5Skmof}0)JG#- zqC5m^>X_-p58pAR2h9jee# zOlz&CAae6rX&$3Olv&E5`q$CxwO`n2KwB_6bcvpVI~U^cRAV6GK%%mb!L3I6Yfo)A znV!^eK4a+s3!;~ke3Z}&Kj&MyJ*%>n?RNMYFj6M2sX7{?QjAYBl9Epogg6OB3kB&n zx0w|W8WB)ZGM}$*zA}fa*m?Q=m`@p!@q+qh;rBCe`ISq6e8X+XzNDNKY1a99b_xIp*n%;RlAi>TI)i=h0=euFU5YS%%q` z@WSnK^tFbpkxkdh6Ozt$!i0F`xjy0i3ZqrQnK|l3?w$HtV%zhsI0s|s^!a@XyGQPY z)xVx}Ij6{b>k6`88mZ!BR2gcu!Vm(#GaFqA(*4vxQ&vd|qYAKQir)FmYYqa-f_~aF3vB=^g0x{%-dyo{x#4*mj8A%1=#Ml zwbJTGEf(@jR4P|dNO*f2!c=F1%!p0(VMOW;H#VLr6)brUrloj12r+4l2=A?^yeG*g zJY{a#smx+TOqJ(4;cB=!wf7D@NxJAh^3|3}AvPx=hAvlhHCT1X;&V>C05Zd_s@1__ zxkzl;C|n zM)3lFOOz9k9tq@1oU7WT;G;>0zgEV{D+>r9NUK}4JKXx*vT?!`;06>M%^72^1KcQ| zWdt~QLtu0&%v8os>fMm|Z%y55Ex~hjWX*46qxgqX-WJ^?#f zJGU9L8mm1*B?(x)tok@>rm)hV_}l-LDbz)9Rm@DR#BY{nrml8Ob77?IUTKBzfwsYn zokZ2iD6lnTHg3>RgBKWT zj3F3RKbOld?3`^*d9C1P1sVPQ$Lm;)=fbfqn)|?@S4g1mztsTI5w#+u_^B}Q#9AcH zc0^DOZ(VvG>|)I%1NXceyN}PvNl$XoD<^wgzyXdqceplID(DhwKQ!SL9|8KP$u;T7 zx8w;i4=it{kq+$*Q4y`RJfYT}aWy?+e?dm~Rbs00$?kz!>pJfKe|?haWr9l#j%FJn z_B^>iy0se||L6%%;JbVeVfsin^LFx8`P@5LkJY7u{!-vXnWdU2KEI~5f&qflmzsg1 zp%y1)fHP9==@(j_h3x0X)L)P3ggwrN0zfSxs1^Uo!b1eDo?ykbRpZ-<^I8xsCx_Oq z&CV%MF6}mST3>yXbUESL>SWGM=(8%iKLAmAa{S{$J-uG? z2#!#>Mol&SG*=YhHS0k+yZR$*Ml+djOMAO2^V(8kExwcW(_>K;S)MMQ=lM>k%>Q#2 z;mPqy%p-Qdo+BUEQb5xiC5&NeMbD@Dw@`{(*MF1OzAcI=s<)yZ&0ZH{M5uES`jqfw z2w7atWloXCdQIFb2K}Pgnet2)AxF*F_-`*AjVt!nLXA5mC1rkd*wW7f=p0w7&1)F^ z4NOuy{Ha$Ys-LG$jf40!NOIFWXVt+UL}6aShxOUU+%uo33SPxr#S|3KnH5kamN%pg zRcV@>^BF=#1Y`=<*{8~(G7>q)l!q8ILP&gh1z;0k4zSt_sNjWA(hLX7dt1s0O z0CUhMnKPXiEAP(b`?dDXdVI>$UIJC%z=PB6lEZ+rH%zLb`9xeWvYX}o)aOpq_6|dU zpQ+%es2r7m8gLPYpRzhmDmKmAu=zTfCZ8lC6>~IWuWqHYN5c9-o4JNqZ%cF<5(9Ph znn}Kz*wn``M9Wmol+{4YU+I?)+++QXy+s*)49@#;eDUz_a)p1*)C=Z<3rKcgaZL0V zxW`!1n*O3h@+II~+xgpn(+*tZdrMWG&Eu~o)u zCT;@gQSCMS^l9Mgc-PP*B`+!S1smfDPoOlzkOT{f%uZg@C<}Iw{tq9&9*hGff-yaA zkr_Nu>Pp^@qL0*A~y+(g>zY4)>%><@<9&Ue{8$nE`Uu(Y&0I zev9a=%TGNq0ipaIy0NdVtzWP?J8VToF~wEq$BO~oRr;b;7bXW886=-#mZ11qsteQq z{OkYCM&CbA`(JGQA4~Fdz_b7P5C6yVzm{Gg#r|x>yq15xvj4k{3_b2D)l3Tu)1y%c znYikwnD^shZKW;t=NB37d+!tsxi^~5cH|;={eVhb3CK^T<7Kk?Z}Q({Rk=?3=y1ET zdkAR|Ny}1YLtO~FqQpezrufmv+GSg8_CCl#SQjKj(|FwIP<=#mSS*o6#PJCWkZ$_= zBzKQ6jggyAS_@*T2aT>L=RGPHWHfZ)(i2s7Dw1;T&07}3%P&mf?>K#!BZgJf!n^+$ z76u?KF5YkK<1FY@7p;&Bt_-5z!=bYbb*8FX!U1dYtJb4j__#7<;RftP6(K~M{r}@u zRMuqX1fb%DEU8>SmJSe@jUYrvDLM^%hc3SHp(tu1b~X_pcBPSx2@}!-i2D>Bfe%jE zqe3MEZz9c~Pc~$O6&3*SCx;fkt!?;8(?_FwI|2T^|@^2R0z z;D9+v4g=}>bF;d=)N{8i-;DkKmPLeO*H={%;sfLZw}f~W;kGQk7<~sf@_ncTs)=+m zQ8~U2D?l8U1Aa21pJkheWQ2Ol2exbWUHY*&x4qxgpn_D7sJ5y-b>bZDj90bF0;s{H9oQ~ z5}LL)W(V6B)dDk~`g^1CRQ1T>ON6l&l~5P+cy`zejVBun$BpV`CwaILwhvxe5kVaN zx}$@4$d9Ew&%SWz#j35)!pQQ2V`{n+JaVUcK5z^~)Co zw`5aGLVrTyBs_>ZKe!oYKBxP`PH> zIMmMMH768VEqqs(O!uJmL$dv&hkTB{l0f#3H>30VtZ!}IJux2@|^*?85@9n+Q zIS{}yHdfWM@~1A9@AL^xz54?8SL{5f*nxPO-({FNJmm_VZHe_{0PVNz#x+?XNsD*E z2Sz&I>(rBT5+e&bm8!Z+y9QDfU0WRLvEdTI?A%le%2<WDob1CwF5(y=F{Dr*s$?pp*9l3yc?Xqck+Iam!8yt!=Vdk|OW8HwXk+-;Y)<;Q4 zo;M$l7Ns#V>}F%r9KY?QDcV=9#xDLxpYD{Vyym$MmfObh&IpI9Ak8rIzgH z{(f^Oj**&sqAO`l>LfhTY=;N(y%NLXt7dvpqeEQ_u6oH#MY%nH`)_9V6j0Q$?R*d0 z4WN~9L?CQt0U&rYaTuRUPegIwn|-&9A6+)fX{0@5$jcl>*orCxf#C`Dyu|uRQm(4keJ;!gqxwBG?5CQkRHJ&0Z~m<-S-?;T zejv$4IC%+uz5#5s%#B|ZN~y8#E>zh2szR*&oxcU$Qhr8dq>Jv&rX#CZKD?fs(MYC9 ztnm_~rpXp?9i`^Wq}P|{4S(DdrCcs7u~a-`+}LzPuxR^e{hmsNJnF4UV)Ur3CdT?I z95d#RNeQl^A}+sAQJRia)Sn(=`T$b%9s8G^F0pYNwB+F^g7%bqT1xj} zUVgJAcMty6t-TaJEEn2#Tc?#4#u`!KT4dVk8`u>Sdm&j}BRLusH|Fdd?O|=>k>M12h)4qjis`Kq9|}UV)Jv6+MZEtQ zu3JIUfuApT;`&*w2K~d`QHXPw$jWilUDdQLsif&gjZ$W{074J@#Z{>G*oBDjTJ=l_ zfl^PgtHH-T(i*yK<=tK(37mWQfvKBUHVd*!{Ee1?bD zN=BT$kZnUUD!jU|Y?1#?Y2(nbFJW%Y@G#n1dWqE%r$nVSsGO9`TkQ1uLVv1;3qrNj zr)^=FJOR*TDcHh6Qb!5Lq4itC%0)EjLfO7SxN^Zw){NMU-)4GkcU0G2z4;P2B_Sz6j(Sd!Bq(#ek8c6v`?|4^x0vqoB%`*nQxSwo&YIY5HtCoQOYIv5@+ z2$dwE?ujo#LfD^h1Aql0L9`AQ!kmOtF@GqX7u{ga#hc!uK-yYZ6BINSu` zC9XANg{Sz;g~86Yh$n%diU+%oS#dP$8b43aCA}oooH%=NoMK-cKh+D{RWkdL3n9Y% zTCGH4G543~3P{+kfO+Kir_^aeYH!`OyOqDNGQT;VmX>+5ziHP=<`BS2`hrJq+3YP( zBF2S8_Kb{X@^i9hV_>=Qy*sZUgR1%m#u9c%-|q-&8-H$8MPRdRo@f50U z+H-}xl#cDBmyufEiwRZ6W!RRG+Ns-6{)iNg#|Bmef=T#Had39(xxYVQXGPG5&HrMn z;c3ZuF-skn?HjXTjuu+nj+2NX48joUsq%Su1HTvO0`fh*+u94WFa@)xh5nnO*e>q1 zc||sx$V@wn()sJNI8*(04lNXN~V@=k2%6l1OMEiiQWiMs16mn#LY6%@W6G18$|RvfX(j zpQxoB(YX?eeY{ts<}&nhXj;Bhr^BLh)a3I{x|?k^ZG zwyk2Fpe$VHhU9d5iVyOaff&txSemKxF1i2xN1n%CGlI9(bCsOtyTRqE!Nms|vkqCc z8tVI*5(xTyqO0=r{a>U(3NgSGi7o+EElJ+DxFwqJVv^co&QtqX&q+g0?4ynwY0cW_ zMS6q3XHo%$TfS;SD+cL-!hIIiBUU$Ka!prhwmV)mQPh86?cCkaX6{~m%hRw_IPFkM z3{X^Nuv4=^z?}ZHn|~^0f;#^{ugP51KJ5y)Wl3Ck;$Gvts zGNwoqh|liLDdB-JLVTzCUYqo_-GsB_qBzZqKzF|z%M?>=TeZ%Xgw`zL*$|=U@09Wt zJ7dVt`*7ehgTt9#MDTNKoX;iEqD}6_cdsyxZPSoe%vDj|+R(>bo2hfyYCoajK(2X< zF<823otqvDWmd>83Sv@L5R7)rmi!^@FPCpV9qd{&ZU-*||K@I&Xi+sJD9B^wNHm2Z78FSmuI^@y8L9u20; zozo6LKDRt*dLDsH7$g}Qkq9k>Ch;94VRKXAk}-(=YXI$%@bA{N=Hi@ELpl=F{(?^P z-P%WW4mV~Op%`%^xvph+EmvT`UXt1HQl(B{ zZv@-cLG}(8`04phsC~Kb3S$&9w*jjJBWHd>Nr7?gLraH`Qi_=@BnSCpopC8n9Xm0r zsfA4o-q56wC_0Q@;gJe4xI&FGrQ^T|Yk=Q8KD>mC%xe_sRUftks+R9uMu}<6H^PtTD6aBfvokw)d4*&2qsq{cao{B~g=22P~z9HZR@nUxw)S5`w^b(v9eDd@%jd`$4b z-K8}tR|gl`9ZvarR=3a7`Ckpf=m<)_;i8vK$E5p7nmxaAQQR7PL->8DZ)n);)7Mb*%wZ11>s;wX15ht3rP)9sE1QuF@RXOWs zmDt=a&t2t_Lu<9szC2N3R7E-k6+yNNZoI8*)zIJLqNV#Vys{WALtR4&+pF zkl#*p0D-G7E?AJiGE&Zr{AtkKOE5@PV}#@U>{47sk5fjRcimCl5+{T%Hp%THn+ziC z*BnyM*N4SzNov=W$C)WtaJwk|rx(LiOI*a`1y9MRT7#G=hB%}QExzt$rACITIjA0E zXbpM7BaN(RR9D;N1O)5uXG!W^0`4@g;+Kmb$d6e4qAe=>eVt$77ARO&a`v#aj=7&E8rY!S@sCa$_yALEnKKP73+?f3E*Wsz6h za+GNU%v1s$gcio}G(`owx-rHF1BPQP5E4;QhY_|(D{$+Kbec8Vo?W-3;c}kAOI&}k zxJQX|2s}}&7|q2}Cq|T*{~V8{cW_ZZ@;1RPBHcDqNnDa`hdk*lsL7KViS!iiwkk2M zRY}uAzyKo{`CY-+=+3M7FYK7NOAv{q(7&hgxOR*92ge^ z7^YGQUZxb-O~YrHfknRvdre=A6RZ1^z|_}3qk_FNL?jE!>A(lihtq-tL4{8rcX1Zk zmJQAh4%L};C+bKlMxSUS>9SLbVUux2Di=m`=6L-(DB*iPL7q3)K;8&0$HFa2l*RhV zix*Nf2c;wpA272x&skg-nYFcyg}H#!v=!zX;TR#e0(z6NMjf=Cqdt}N0(2wqf*7^~6kN^V|>#zhE8+uzZhChIuw9e;QHP+%1ARK>Aq zeJ>bR$gIr2J(xdhpB%v6^rXs{gstxTqOPSKWm=8#0Gbb$59Fr`wte3#({x@bMXgCr;rh0F47;F(vAcb+wk>$@k z8kPzwX4A}Ie)V3ZE+{8de8E`pbguEwP&8V%e~QKfkt*M}0gcVr_|&91v+OdwxDVo} z&FrR8%I}L`N&++sm4xL93d3}NUE|Qto!zoT=v$H)g|t9vila_X5usS6?=9RSSq9m& z2Kl{Ypyw0(<$1#>z@oSlc2AwfD;sFvrCk5C`)hQR8u`~2(ZK`uo#{jd1sY*4gR+!F zZNlIj8&#B1p{=kDAcn9z$;<`FvjkfjH}rgA0h&M$%jj&wBZniOzH3WJ;;dwL`jaj{ z+ElDze5xYQyz3tj9lI$?Q7~H0Mu=1*9%sRqa!YWJqsY zAE@E>6IfBC?tOik{I5KzqB!)9FjTixCW#6M#5^7e0;vstEn;l;daZ}ldx<#!ysmsp zZzjA(#T=+dJNnL>@BTK~Ca%SxB95Wt&9t%N_p;!buCC|CAKQGqbYcK66mmn6z-iKkp(ZZv z7P`cg{pT#nP_L)jjWp8<$v%@c=b*x!RMNtiJj{n#E)Ph@VQs>8MeusRW3q58z++7C6wy0LNhVR?OwQNYOQOyMH0KoO_&@G zGqfta3_VDz+j#sCD8mROZ3*U2PxGJTE`Mm%7T)CJT=R;}NC$hnV(3#B? z{ZNRPJ(v2aUogH~sW$8_etc*8{P9!Z}A-VTllq-m#w%)7lxV_(r!>kNu$)NaMnr@l8>ATUJD(Z+v+}~L{14{Y@QGc|-DCE~XyT@?*W3>uWY8L#Gv08DvPvUi9?lt#lH) z*V#U$k*PnwODWnqCCA)tis;w2pa{g6(8tb94?Co2SP`h69_R=sn5sdJ?5~CsJ7F?v z_uQi|W(=KmY<;B{<#vX_ZIdP1r4^ON<200AFCEOkCu+~YF5j9t*DbA@#FzHrL_Mdc zlKoE5NBRM_7i&fjS!h?~0DnIcMzK6{NT^@vDS2NZIJo}}gAdM0#pYgV9t1qtpZNRy z$w8LUXqjS!b1u)L#@M{Ry!BuZ!y9ZiZ^cKRIsCUoOEp}z)fL>U3d1wo3zhG2CHh1KR3-cx}dofhqKz=b`6+VW}3{ooz3)nf7OKlyeUYUDK?l z%zOc{Af8wcnu2F)%Z8;6v}6rvf+-Rm6B2hK^crGOOs2I6buxfjE5)4WIlrK(%g%~R z&#DfQaV_)agKE?vx_M8RH_aEJa-vIDRerxLW!q*V*VC*`2f1y&drxepptTm`pPL?; zQ#=*E5;DN`b%mSA#lGEekTAZ!jBpQ86!Q7DM4)i>XR6;s?WZVv8tg(qsidRc{)D=r z2d*rC8?ZIIz8M>_R>YZ+1!6^e&emrx^-L#ccTP4YAsfrt%hA|{Mx1GDPPb|5Y1|_pCzCr&eSj0(q_^GtHbtdy? zZTY?7vO#fx4yqPc`Bo)3q>;mQtR|l?0uVU$eCtO~#>aWFC?kv`lmDw`Z=q4`GWCYI ztSbuxjBA^Cxn))O@o{Qe>^ll(P?HZzD0p#NjJ4dehF?UKe_Mc46s~l!b%zXgR7r}p zeOb{_SC*cU#`4Qea|);HOXU!`VcKo>saF-| z3>nqZoV)^>TE7LKQ=-EE4g7@d(#lOBdyr8x%p z;p2%jjB=vJgN6MP9q5Q8YXHbR)jqu?y)TfBZHGJ5wzZ_ik##;`&`RG9;y=Q-JO4ZZ z;^UOj(-S-#*$Qh}R8NjnEe{}n$}Z$XJj7L6l+SB?RoW!W1B@u2G_Lru{&&{qUkTBo z*7KXm6)nSKn*~tcQ@k>%DAmxEi`af)8osgjeu^xy4i5*bS5=JxA>Pc-^VJVX)j2by z$v_IIac2T6M$&bw%QRCTv+2n>JOs&Y-SpXRXky`E2*ismuG)~?$U-46W;_KrY_wsb{|ifcB6P;_u+dtEzC(JV9_ZS(U!|rq+DhC3 z?Bd{HR=!LLd0ydQW!K}&3IpUV+A}=db*EbGT3U6zp{Q0!vL(CA*x{rTpCa9PXQ4gw zN4G&FHD8VYWQ)9tBSYu~w!=+c&Wty*c)~R`58LxZWxlsBoGUr$7Z|+6$M$$BmrOByNwP&uDsjtGrPtFKBD>@U?EUrgKf(xz@WgKgb zX7*YQYa|Gmi<9-EcaqyT>3_~!N8@JHzF9*tVivmiRF0Tf^Q-yOBL}NVXGf$sjmnOj zg{H>Hk1g<9%jA1g_id6z@1~RK0e`nA#E00EQ5r-}fOvrFN^nu3E){wxU~<|WnSw7I zM<=>(L9cQgM5#6;q`*W^#gi+*l|@PTd>~F00M1i+J$1H_Mz3kAV!FQ?loK$}J7;|` zH`S|?tH81>dA>+D_n44D;^yjgRTgEg(8lZ}F+89U^eMCBvEees9#Oid!DByhYZ6hX zG}|%*+#sUuQG6Afh>9y=6~(XFV~6)CV{pEKzYd;+u4U8>Brts7AGu5)B10C zC;%P#p_}32tgl_Bx>B3hX?plS|L7D?fOh=ngZ~fP=@K8#F8_hx+@SZN+=33`{H2^M zy$>}c%e>fyWLeL7J{KJ9p&!u2?D}DF>MdR~9gKe(Mzt{?j(emf%7gZ_eqtX<*^<&< zK{Fjz!()7pJ$c^ish^k!j-$OnC*7;+pw&oRhBSEMXP4?OpMV9cDF?^xu}Xik6bi>A zqM#)s43P#DJRV`7B-DE|hf(ZHPjtBy=kFB~kMI!S#e7wUv}m3=$%XU-E;{=dR4nAS z@H7`m>pD+P-sY14wl8++r@JX$yF{PsR_^*Jm`}M;q)Uj#HK(3kbb};tx<5AU`<&6i z4>)g4^Ums165sT^q^pgLlZYr$(~`8*Fs_<*b&mcOHiLr^o|D=#0Eo)IcK*NEd+VsS zy070CC{UoddvSM{LU9R%LP(Glw-PAsP%02yiff=)0znG|inn-iheB~Lv`8sZsGRh9 ze$VfH&lq={d*5--J$IZj&Oc;jue~!Ed+)W@oNLbc{WyEi)ILU_XA8T6Vhix(wKzR_ z2q)b`TkQ=j+pwyfxEX8u3Ta4E9TL9G03@aGpLbCGKRq54G(+Ab?P|E9s|?h-Y^Lx& z+22W5Ns7CGiE2DCc27skuarx1DwTErztB-~q6WD~r$g>=7RwlG)+&)%vMvAkoHOp)Px)rS!*K6E5N)hKT8XB@@eR@EX{q61Z zNi^po)-Ecy&NG`kV1IzCx^nuEjF85_j8$H2{D>IFM0oViGx;S-{bS?>ssY@a*ns5i z+rEQ^uVOp|b!F@7gRZF^Jl{V9Ri>rhXo~NJvISVO=$pii*mx$tg_pRmBAWK2Agt*eD!bPby?P{#*_amBxl6I!e08favQd3eBRt~OiD zG0*ofU&DBK3?vR-(E6z!nSa-6kEJgA?0jVF{OqZyk^-&3>Zo3ygFW$rbu_{)(}`Cf z`Q1 zEF6q2l3(;&JxKrieqN`HST?Qr@$BFof3&TaTR115UV`F%t)K0&=8At0E4um_`DglQ zPpHYsR5UBlUp8|M2BgiE3wc-E5{mZe;l?{>!K!e)yz|7BVoCScGGJ+xECe6xy=fbV z(e_#-pHj&>0|RA6p?O!1@_T+76Dg0tC$p8gmHu~-O|KiX@mwkLNO+lBt$oL3@7|vy zn}Tpd7Kj4^<%M6}x5;`qSHmxGn6JLztny@!>YrKPgj1NdpGA1Nbc68`EjZ&hPF^u~ zxk_UBkC>LA?dql0+gD|M-Gp$t>Zh(ft@2W2db6zr-$UN85e+;Ho-d*N>SHr3ue5zt zVb}1BKv1dJ^kp?PzlORB_4*-fdiVz?S+QGJSv;NRMa&Oq8|MUVsP5duKU2qnCScTb zLH;c^I(I1ls~(&0uZP|7#l{Cx)1=hg9BRN2;ZcW)9N6wWG>mP|vAvMQnUHRR=uk@< zE{|vdHzY z{(-)Zv)V(qj`Y&-errQ|XLU`*QAb8yatNcK2jn)}%6PrB*rD=N`x5{7S0x#SKN*=Z zHh8CML&O2R8N3=rj6sH?X%pc=VisP8f~Dd?og@EgmOu()v&gc~zF`Yp-S(fyuxro- zZ}QQO*7tU-x8dQ-O*efvv|D8wso=;Rguc8q9V{Tg5J4BE$@Je%cQb3p?0))`O$+7U6AeQ)zLuF057q7y7@R!4Q&V{CfN+vgv zabuUnaxQJv(gzK+)XGA#dRMLz3(>EYHUEkY5TAgx9B+twQq3(sqc5*ZaZ1gdffFlZ zCuh1d^uUL?>K$9I6J4rQj$oGd3KqfQceT6-iBI@|%H`;tL6$C0?-n`mWoB`B9XQId`U&?#n5N0x;5 zqzb>t4+bNN1NsHbC8CN2=BR2VK9`&i$+DG9n}6YP75^iiaR&ySi)ZwRaq%fKg=Ej& zu@|=?8K#Nt!IDUT|5syjBHmof#wTmOYLGdNG*B*OjZS7Vz;YzxTr0YcFyY(V_3h;S zPKsh-F-Q;=WS`xhPRMcf*59&FewAwi+RW2hdG7jl>w#KKu&$x3;gGt-aC%J(mpaix zIn#O^IT2R$S+v>)kgCCvtJq)UqET`d*}!XV^y7Y?xt8MQOTFu=KX%(seQ^T^Gke0uN^--VUh(M*#RQ#RnFsWd?$CXYD>r%WNGl68sT zQ#NSCQK?884i$2x$4n{FH=rcPAm-vObEYM!@^B>7h>B#Hi}!WTQ6?{cXeWBm$ew&m z+^SK%=<73=dOsyn<1I>dFV?nkNn#&Dv;<(Erm!BXnt?QbV|x1m$A&arKgdnK*r-m< z_swXwUXPTxp z7lK*+hg+%&&NHz!xX}NIKOKM$M`^X>2*K18B1#|n%p!8S1=V$$DyB>U7M4PtCIsbD ztT(BVS_j^aoS*<>l(mm3!(njUJz9ft{Q8!%+1MEchh34NhIHy$+x8cb15>dFaKUl{ zP`F8?=j3P0%0zEdxM?NdDQIIwTrmRH3$-3(HYf3On|sN_o9Q-97a@=#ms2^EZVXyTi}l7TQM%C)g8qA71m_Ye!E0+or&)_;OHxP zae?Ao3xsSHnp^?L4B^w6f|rjQHKn=dm3I zyZk+}zE3qj-20#L6+Fsw>eB;-@yJ@~kC)P#4R^TxJuYv;h&I;sh8+&gQ(K}N27f(5 zIPwfC4Cp8OB&M@&G4Ux5NE!MiI_2y!4oQwnraAtXy6^5`ON*zl7da%;omx}E&=w2a zb+yr0=wZ+e&?qvSKq9`YmVNS%*Xs{n*Sli|@x1t9?ZR1{B{g8vDB_&h*>3pB_a`Js z6lWtERYBbduHlxP6oy7%RR?{Ex<6I^npAHm#*&0GNuo$kaud@U`O05dhL9^@<;r%K z^#Re}3>WL3rkhw#OP0^g&?@wanOGnTVDI+M)-BQ`~40z{=_-?>QHd?xyM09^nVyBtSn?KO-p1QLLXU8R0Pg zb77-5TKXf0!in!ffld|9R1weWJ7@zPl8C*-+xd*vnmjgv| zSQ>Xy=DltfAuKQ!e1`uRy-T!cI!e`0p2O|@KtgA@(C&X(H$O>PWv;9Qiw6uq=w9;w z^1m?|rK;24(833D$lLr>)-uFvVgWO?UwacX?bf)R8^*ZX%$b?ZboPBG45W~GWF z1u0r4k+cCowZ>T6xkj`P%L;Vze3Vex_tiIcI?*w}B8i(_9#X5s5M8`Rq>+;Y{5Aus z@EJWYw~S)GIPIZZBLy5%_KeL{Lu4l?JoAWP>?CTF)pi=Rd52)Xednl-0DDC^@57R| z|5%&;l^fA9_sqRiJ4zS#p9qj_8O$ww1lYAhKrH(4#e`E!@I0~)Hdv_HFVic$Ink}Z8o7xJjbrx%CI znnablf9wBm{New{2u2go7*!u78m+EXZnL_36{WA>hSRaud!VSSLpmWHmXwf%m!^T` z=u(?iLAY0?L;7T{$sYP}>aW8WOocY zunfBS0A8Ad`xhG3b1P*M+o!)3hqn~=KuI`C{-|D4LK6!}rkc^^!V zPhB8ynp(?-e1@M|tB9%aFSo`X@7|OPVo_<906{i7*Zwn^D_u{*kR7SzI#T5vuuWL z&Ww*2pl0CE1RoiB6P;@wmg7A@?VN>p)Ya{5D0nfp;d_$&ZX{@+6)@kM)(DJV)vHCx zAWBT0@O_=#RVOys$YxOlCr$;jY=&)O@e`$W=tH0LQGQ9k6$4ck4I;|$5X@|9&C60s!=^%bRLvU2St~(VF)T3WRhO4}$=!NXGOCQLHA7y( z!T!0>j`*7a%I&lXr@t$3XkyvYEVedt?E9J9)*SudnjV6qBq{CP{&l3%BAMtuOn$M! z$j4?hE>Y<(@4#U*BZGkAUkE6a7Ok4ygV=CLWLL+ z!<%+XFBKKz$}PPk`k`Q?m)Rzm@{dHL>}`cfHxK4*gm;YY6B+drtJ|7J3iA2vKMDTI zL>pwC=K;S1XiTT(a0z&&_P?C=vj@1G)5OlY0`(MQl@NdvNY2;Z&!bWRVPE@~L@vh9 zo4@cetK#>(c4Mtyu@v8>I!!(pdM5EQG04o|JN$`bTZ8Q=_ZHh++AsUNj7QbkXILUf zfYaL?-2vq)-XA@xlK`(F(598 zJ*0#g@kbFC$Hv?^uWyk{E}gV=CbK~%YB)cI(JXz z-nk*|X_~Im+x<7CKtShns6te62yyoO;mB<9#O+Am5=u`3BY{9&avNOybTv+2Gipx^!pNwK zV?9pr+nAX1W?RkNl_O&`U;cbs{xVfmM)t$Bx`RxA2Vyf@o4-uj`Ls;3{ceC%mPujm zuaeQcdneH|sa&Zc{obUzl7WJ*rHvlQbpR3|{3~=Y zr6I)apesiwK^eWOO+rf;>@iBMy&dJ01kebU(cp)Ej+?)?%8Me;z~n*iZ*i1eSFr>> z$h0N;*zlzCN@pc~$Gk`G^DCh)sy9yGAty=g{KIu{k406;6K>C;uT>lW=j8v|-EhN( zPYVtQv9th~&Qz|EBpgWjYUS!@*QA+vLqP4w_936OOG z(~%OT+Wf(T4=K6>mTD&I!Bq9shq_`iK-nxSak=S1gzH5s@G(GZL@rA!9iw0rW<=$J zLNLkzBbE@wMLh$j!wY!ra)4m_h?yYqwfsga{LP3oW#+6(n|UBZzP%{ph?`Y{+=;)m zy4t*Xm>X)eOhZgM~0x9TVl`p8&H)SEpJK1J`v z_=~+^A6jtgK;#@+pGneZZ(!EMg_0SkKMZJ3bmWh?1@lJgVhtHAjD?8qB(y+;FCu(2T)@}fLhoxnki`J>hn^U_T8D(b#Z?%! z`)Fy9n^q6#A6#xhmd$_+6_$ImX{~B8h#u-Qp^l{nX6f~4f)2RN$;(DkKSyM)9t}t!2y)C5AEEDPA3-dI|rq+P? zA8UMTJ6;}%jcqy1^n0g=e*OWJZ!!d@t+N#0dWw4dNk0&iA4l*&JerU93x0LSKvUH$ zlg~R?^z@7)m$3ks&6L{62gJr0^_kxL1H6v!n*q7unZOjBiB6J?oNRxE%yhGNbKdVP zby6E2;=c(Knrp`cciGO_plQo*&$>n9h3jt5B0ei)=_wlu#M!y5!L-FkTdzHQp(PpwVIj72N!z`zT(qQg0$64h)Ob>vuQdNDTH)_N&B9W$};704MC}B`8QWOkHMt&$atoK+0M z*9y1B)?!HWXfrtrtf3Ur^Bm&53m_}UmoW$(-a~BdH;q`4!dcF_#Xe^%TrNZL=@*co zi6(U+FB@f3aYj^iCY>p(!Dy3~DxHRD7K<3CFbOW$xcWgckfVraidP|8@;2ep!`g=> zDXobl1|!3b$r2d7kOl$2Xu9bCMHkqRZm)HJD;= z;kpa!=jkaXnpt)~An>Vqpd=f&r&E7#f1ba9x$mshqR`)+W<1FuWUBhV+?71OXfVy5w)B`L0sdn5-pb zW|rWc56h18V$m{qta)B6qCu-}T8W#@B5L1Xj^m|y@oeF129X^&^!LZXQu4bqw^;9g zAqqje&_+>w?UUEB@aPF(P9e4IaJ=%j3FrEnB1vR8`}t$7?fWG%oH)Lnej#-Tw%-0v zq+Vq!yoc$uYl~54%f!w!O@i)o5)FDa1_Y#SNJ(u^<_)IMpSpMV#`~DIZ}9F4(P?Q> z@8iU(EF3?3(6Pi+u4`UYky#2D4f8K^=EWl?$rf|-+(vSqzPfWnd9!XK0E98S1$y-J zYE6|f<5I_E(}F4?5Od$ctI#8B~*Vu*?U;q!AK_RoK#{**w;$vRXe#M5LhF+ z+Q#HM&GZhB71UVSQe4QfA9vzc_AQC{d28m#2qutkj5w+-;REzJNhWoN?bF1!N!O)G z9D)9BA2*g?I2!;myOPwPXa&@^73u%6Cei zAJqV6V$0@Pyn}k$$Q9RPL34?SaJO8OfkXteFx!D7c@xs5hCF^&c98NgJ&ZXWt;EAf zVeiGOFr9~8fz~sF6%B=~-}HZOm3%3~%d^S>YJck>E07*n9Ad?Pb{I_$q-=~d$Nl+O z#O8C6O?&4#*{@lpPyZOool0!HSgy7n2rQIILUrfet*VjWc{aPseG8<$2(PmpwVT~- zZCcqX>?W1GsJz+Wiw6k19N_Gp;wZfhozeaO<404&qwQZ2n$t4+Wv9c&gjDFr|@4Qo7?Teuhd` z`W@J0gMFX2>1%*H2~3PZTnId2ohSP!7>uSDT(5Bp%d5`@Qk&<;)(3fVwJis zpKJPVnD!8BYGC_^(%p9;0 z^-me(7t~AIl&St?`i|`kQ_0U4nB{1WmUM98I|s4jmVi=EJk8q(kShu5NGtkUk2 zpGxE*Soty;8?5rrJ&8gx!k8}s6G=a*ja`JA9biJ}gST7(Rnmm*_HfqNZm3`hoCdp_z|Gg-yV_Icv$D4hB7 zYH2c(qQmE7^Y|PN<2COM^xhY+#85|Lx;5j9r-(e{GaRv7tQq?$zUm-tOGWZM3u0O) zX9;Ra+`q~cWmK_pEHtsb^zpV(#4LK!h>KKVON@R#38E!)9Gy~E{|NdUN64e^S2NV$ zH%>xO`a#R}X=x|*k7xdhTQAkFXE;f4r>t59@zdy@*mCoU?u(RzXnoABZN1@Geq@Eb z*go2$+HTnr?=_q%7*}gX{FPtytvINcyL>4Dv_fEYWAfo6Dw+0Pf;mxH_4ZB@z1hMm z#FPAbNVT1@_LmjvLpPjaI*46!=@YZ^Ah_o{Ty@&Hq$5U4?+a~iN|ro))|4olsz+wG z!}W;!nh^2KYpUD!#+zY$#}DwO;JZwhe&?gX6OCdDDP_m~P)o1it>-1)Nk_31@{4pg zmTjA55TutAyrT9RQb@hIB8+bqmym=y2>n=FB;wTz#2Qnhj%J&4owP7q71$r4Z5$~h z+RI2D0zvh*T}g8@>(%$z8qxzD`}x0u-W0q7JpyG3?e%4ILbNMls}nPL*7H*$Z31-G zqc)6c?#$xvDrzg*FMvIPNfXBFaE$f>RrNB2L%@lv9=(9j-$&lm=E7 z9EP>`k`c-R%WPY1URr^jY`Em?a@>MVKYTKBqO6?u0YjTTPI~=|54fl3As+RiJ}!iQ+P=>eV1h1B4S#eyz9gIw?f%x?3$h-+MH9SEI(6t&uMt~nQG&0JVSMj?$z}B7B zmzsjw#)}Kq`*HEukT1Vt`Hf}};B|I|De1~=N!~^~31sJS#@oWricmCzFJgeL1Bpkg zX7o)0=AUOTm@~}|h!aW7T#2$4(tOhtS(56bt*1SuCZ$8g5rfWF5zZ0>VT@O6U44;y z#k12amTcK{TL`bd9%C>jdI)FN%+(EPBv1#YOJ-36rOhqHKa$u$TFqIn-OcO410!$M zu}ozY%g~~PdI6ztE{&cxy=1xw`7>UaU7*PHOv!L2zjD9YP}@!$qt>}SBtC;tYwv;` zO1CjZ1gj4qrp1$Z*}`J$PE`h{#%BcAH0?#TaM6qul!M-_IM6ASVr-*jpeV?n{g}Z_%FR;qb&| zG<#>zwLmJ!<2a-pgu**7$+;{gD)z4eKlxrHx*Z90Blmh`jdjQ;IpQOk-OL@v3*L7W z1Y%2}8_r;(K0RjGqOKYTh$7gXtXX-==#9aMTcJXwv<}l3)4jQ+*F!=w^JTkHn93E4 z;xhMwJ)PE7f9!|G^b~|HarM4~;Oybfe0EGuk6D3RlV#)ET@biMe;}7t#Ue=(8xV? zeTle|h&VZ4^LtpseOiWRHAmeSl((jO-#?IdP36#4J#JH*ynKyXY_yb2SR3!AX=qF0<@o;F(Bjq7C#=kNmL<`R$0RdO*y-ZXXATuX9Etnwk z@3FBZG5u@f7USv@I|Ira{j?lkPCd}7;)ot6(6SqEZ8U1~+|dm8%*vco@}ZX2p<4O# z2;!FE&94#6V*=9QX@~sWae~o!%{rCS(L48{c}6ZQoqPA~{=I+obq5N11HiZTAofs> zqm786WK}-~3$sNTWyC8XnZo`>YBa-4Lm$BlM0;^ zQ3+^b&X{4vPke41@b#Dc7ZD%S_Ic~Hw2r7jq)_Ofg+fY4fn)6>lz?fua7;C}{;8at ztE~V0;FY4$p8KEbB$m%rI;Ne0=>lhmozIL-=|dRn8NPniq2zNu3EP6|%!SHTO*Pph zRetj@FLmo=f$T5U{cm${J}ir1M|JRXYt6m-2ya5v^Z8x#+nxn^qi@8@pNB^*4F>Nj zgePB#{l>}vsvN!;9T_RNAAezUXG%XXLO5gZzLUbiDK8$1_cCeg0+pZMZ|KZ*k3k2# zUoVqbG}-f~%AjPl(zjxs^24#VjDJ}nT9?W=VR-JIZ{deOu~hj*v+&|oWH5t5<(ErJ z*EeGma>3x`9xu#d38gR1brz~n&8NwkP+%#!D5|+;inI%n%p=@HmJ+j9E=y_&UVI-ICUx0p8s?!6tiz9AQ%zwfX6@sflIUW`DC? zuzV?!=X8O}_$Bd7ruARikzNh2vQK6bVq~N?o52UHcQMNTF~`3cA0^fnka8dKmN|7) ze5^hzIW+YUwIQ;(U&k3!NQRZ*f@Ob3JO0L@SfhUO;MwE!^ic@h7Uzb7@WvjtpWH81 zM%K^w&10?Y`@qMWw+N57?f+^_V2`Za)FJ-r`l~VVPuIWC`R^Lbe=h&?JOBMv|AQm{ z4}Y+SCZ@-RZhrdKlG_eY(6{UAw$XlRtMFIX-;KYHV9$B` zKWrrWwY$HvNl6*WVk9UHHo$|u06=Zl9`YI6emgWj(5M2 zz;s%!6rGN!X^c?2?;ksL&H1f(YVK6D*UL5Xl}EpEXij=rg+5LG5(+;R+Dc0jbu*Hs z?IvqNE{`_`}T+(m>MSS2%DEn!NF0b2Q!69L_NP0S) zdHVD!ee>qVY4$kOh^&jEiTuSNlOs0atUgO>zaxm!!DP^^oWANB$z0&`h1AFXp5uH0 zP0?n)i9ymqH`Xr`~84bFsN8QUwqQX|} z0osJJd*)uYxPOMT{YBtt!|So{4Y6+ao$hgU9*tOuwk!K&I{$qN=2uVjqm-U|ItfJy z`Vm1J<)R=ud^P6jL1Yix&t)Yt?xZ`goGWgTn6witj{Y`Xw#{f_%-@b-gu{q&|8o zTjvI+y&!&SKflXZtBhs{6~=>AtLxR_CX&#UixO0?%ue#ku%z5%mA+RwX(^ul*_EbZ*uFCT<4Ou5_mb(`t+Lgk?WIGHovn#l%l zzdcCV{XiM@i^jgS7e2Q=mO>vF;9*Utc%8v6xkJ{X=x7a%V?z|^pUi8sDzaM&?ajw!VIat*qgzy8(x z&x>HGA!lwx#>L_kRQg4;D9rHS2R(hb3CDehjl(2Y;+ z(hbe0mHFQ_EHFDVHDT9O4B_?y6 z6~Sr>9g`i}g5Fk8KdiQ1dWnOTlaW~DQIx(>B;0Cum*C(C6_!&oLSA#Hx#83edrMEqAw>&qtBpIGaxy z9iyz3h);Wq`d-njm3$tCiY*@Rw66|c`L~=6LK8596c5Hp3`c{?1(@sy9yzaMLBX!D z46P1CBp~GDd6y|b^XsR%IvkMv7`H)JwnVH6athk}V10J(^kKrV4C`i2xKTsXu%Je5 zk=e#t5~mAhbEGAs(j{U5vLa;3QZKm`K6yy)&I+6%$tZC41guuFX)&=#S@**G-q7V} z`mV204FC1{g2%HD{V#8ayR*3_agUN1X#o2Z&vGZe}C zp%joWDVLyU(pL3WfH-i(s_|Jv_$+1i2gvT+1#4ZxW#21?wWV75q{{#z%IN6F<^Det zL$}1>Oz6~FJsIK+$v`*o>&R;@q4Rr_nMG09Dwo7-g^)~RCx8xwZGY+Z?e?g_DMpyH zQ(qYZJ*jjXv)CN`8)wWA)nSs25?taPa7R(~_)}Qfetv(AN#C;b3dPjQ!x*=}t=JXSB-HnIsnk`($aD1ymD-74{2S2ak*YSvjbzoC{_(CmUz14z9p7TzC4Cry*nqhTT?;t4qgM~vGSA| z=H3Ar3VkLVxSgf3rp;Y5jLUtVM_WT=fcuv6c?{r8e!-O(764{5OMn^R+f5M>rO>`{ zTjGY?y#HqNLBjDDUVx0=fT2*hV||hV$G4x!{yGt{18Pw%jUjHbJbOw+cYZR@W}LVd zH&$zE>~BnpCOIbM_p0sxwSN8K82OuZR4>yB(yBy z%-19>FtL%eI+rFD__;nCt8v(m&AC3$<>q%Ue~5z8sazf6frJd*njt&?zC&}^TDz2D z4bnSO_7M9p$XpoPzi-mV_ll%~QeSwBs$GufZtk@k70eYeR!NP?)XWy$y5RZex%)@* zB_(=Q`)e4dYLyvN)Q`tb<Wf`^<%@Q-X*hMTL^2;;*p|wGg)Sh9_FMB>vduAKE7SW_n zkdG-hgAKIr?o4hm-)_1uv)hE3&o7rqlDb&THwm9mHJTCcgP*>1Rq-xv`6^i)mg2li zNDgDtje5I89-C+lyAjJH@IX-iG>JdxIIRL1D>@X}`t_&;e=+z2C!cp@^$J*C-{S6nk=!3l0PH{N!p&7@;r#@)9UaTHI>wdPbr(ENv*0eVYVGQGue9OjF+pH2DSkA$F?}zb_RW zfrCOI>Gzj!b+4(qPR~$(V7M8jml+ifK04`H%i;8U>HhwGL0{4ajp_Zpt|oF^H2Wy+ zulNA@m_#ER@Qu=KwXHv<&0ntub0fnEdANw~fpUt^L?2eR?=t>R{6pOJ@P4!4Qud$9wj0&WgBY5m zH>S1fFF0MNT5C_o&o?^XE%0jSF+*i-L)*2CZx&FlmBE*a-)EHg3( zVs#1|sI}6ZD1M~UpaK`gDyRTR?o#HEF9n$?FKTUh+*75B6qadk$f#}~QS7+YCGDkw zDqpN}W3bHj>oEHCv-BVaL5Iiy06z{k$hHo!cQIxdi1E%(e*k{li-3M4qdc^-KONeO ze&uDcsR3uZwNPSYXz#&rD)(ab*VOY5@+)TOg?qe#s6WJ`Nfns&Z?VmF6iiml&iYhFlvEQkw=5@#prZ8VFj7M16# zc+LDFf74Ywg5T;@y4LXI`D~^+W!bYRXEXoz^Qt}LWhYt-rp?23OVsTwgWu3UCVy(o zo!VSAO#3p0YL4oo=Q!uq6r67z9ZEXJwPrj=N)0dL>=ji5pkmEgS?34EOa!#LZA@-q z!~sYfBQo$Lv9O-Wd*Ugi&(8$m&7Y_I{g>z6(KOL{W?H1n61L$MmdG8mz1wr;)o2uz z7^+!_d9{4x?ksC=uD1+U8q6}G+H$;OA^MG2A%O8(6TQrKiA?ET z0Tr&T$j4Z?eIih_eSB<&#KuMztv^Sen1US-LJV+0UI4IgY8K68<>&83WK}&bI1sSb zEG?o{_U8K;aUEX3qs1u0_aX*eJ~g);u~D4kZu?m#7$zPa^(|?N24uaVWM6ozmfhLE zds0ckG8;54fA$VSRO?t=R%U|wLg#=jpp>7xlbdUkgJEA}Q+yNS?mRGQg=4%z2j}e$ zuozRCDhUJ9#c4w8Gt{yWeYK~{!-o$R>!>m|V{8;nV_gN6AZMBY_&fcqV1eB-5Bcw} zGbd3eHRMss7SA!8J$&WSjF!X3!b=7fFhTFVTVJR%u~ZBy!aL2NIvJlzU~j^uk?Xt4 zh8ay^@=f!J+W&Bgg!K0cow57%^mH0QEnIyL&=JBxtwwb4a;1>^0^4$`3{Tbt@i~E= zl{ILML48~h3yl)^!L899P;KCYi877S-;Yx*7m4xjnCy#O9lnmC9~FH)b(4aZCO`nR>o|Fw@DVyoDW`3ov?qpMY9OgV&+(!p+}7iX7GWH zz~fFB(BM5UtyIH7dpJa4Xgfh^s$thfCck!lm$$%ryXq`i)Ryj;%IWF_jd1uLyG_1v z$7UUy1LtCHLf{8jR(gnmv?k?lZ`MFAg5DIsWl2N6YT8=)OocLrpWj6*?;-KDhu8t2 zGC+53h~DaF%2v&=>Jf>9r9nsLQ=h)kk1t2%TJ;}+GcY@|Ri@TvM|zEWD>(?wN+J1& zp|O39`D65EWNzAzaAxyd;KV9#<#q`QzWl}+|7?RP$SK6o3vCbi&tF`Q^Ihi@nYJ9SD?wjY=gFPzF2pnD->Xhsi*7=rt= zzAeLXoT~U+wHh~Ky=^90y^Fu zN3f!NRk=#X$T}7~{xD`%zkF?Iq_2`PL~0M3)cM$uQQtc?5E>_MyMGoKU1Q_qu3IZf zoS^P_Klr6Lywz$C!s;Ab@QhzjK8*Vu+Y#=Q* z$w>IMv)#0{lUNf*q@n#MEmzED!h@VwAz|F{b_7%7u4|kj5IrzZyZr}|-X>-}WH=>q z{ov}Pw!eGLW;GfG$ghI5CFG@Ay)Hv@qEYAXnaHo=@5#1IugieQ_Z5pf@KR2*KoNZ4 zJO=6@-93)y3F>8bnF%BdVY^s^R*@)@U=U&8kyvR`6tRLxNTlt3_c+L8uvBkwdhm&%p~A~kd&u7{xwYRB)HWP&o^VvnDG7(E!&biBZlYfX zY7y=VeXa{V@;`Y80)qUKXi8Kol$L&KKjFH&MaCqCJMAxU_=4Y0FfzjTl&P3e+q%P* zLo$kBW;K@a5uR|K_$TsxyfupMKLMr9Zb8AKr28UqS^}uc8?@%pX``bB3^3J&+7&%FKAazIL|&Ozot1OELTzV~&eX9#GJoAiz9IB5s2Iewys( z*n{5TXUtob!eWot*K)#0#*ThKbc;0?=%bnJL@&U-uHz-edRfjVpbe;%pkkd^iO7i* zap2EXHsICU&=Txoc1mv^e>n*5unl!^>CSjXZ`e^OF4|YdVBIHW9?eq+otZf^QmM{= z`b)?e@2?T;9HJ2SO2qti>@~B=$1IVOI$&3N16%j^wP+;rNDfPPkDm4=DyXM|3BAwA zG}7@=x-g8Hpm$(-m6xCYaLzdI8IkAFw5Ye#C*AN!YGI#?j|S$`4Nag=T~8$)zuWC6 zs7^aG0y%zB!B}$Yui<owYQZfu%ik1if^7k+{F^9vdhx?$8le8`&wJ3Et^Ohd)4d3FGOZ9HlYGH0%KO2@r+Y45_=EQ%yOZlrJDzpZaD9Y zCml-YB<1#<%l0G#fIztwSd;+mr58x8&)QlVu zlO^UZ+adF6P!9>SpI$@QP9O){<(N23IHQI%e z{1JJ@SFDx&k=bnX=afWJ4~46W%fHnsNWdv&h=Dt4bD`!!xr(OU3`-ClAA~C#)k5Kt zYIPSC#c6aRw&d4V7*}5k9UCU?QA(h6KoH`I^@gIk!p@! z4Y=ReqZhSx=)m%hcp{$=Y=VZgN=r-7hvY0TW_RPR%h*R+dx^Taevt*m%R3St+_8+c z=*P>IFVbYirY71ke)7mk+~d~pVOhDrorTl!AuiS!%1)n z%|Z~uAy_?P7kLj&qI?!ivjVfAr1;YN*72EsjD@q)!?>VW4UY)+6pCRG~()L~GI+G~|@?m}uZv_kPON z49@o^VfJKSK00u%fZme{jQr@|m-m{6=?9$-p1qQdQ&2hQdL(vvlOd1&GCO5U6m!*s zb*3{mlhc_T9(}?MUN8m=2~u0&o4oD}Sh|&{^(t{Pcg8})O-!^jotAPDN?Tp!?F>g3 zWHyGgoGwP>`8#WW0G$9}va@RwG-F+~;)*8fWS!5wi$i=I4e38>6kr8Gf_S8&4kLWt zMJr(AAf3fT<4t|N2;v);e|Dy>bN}VT`~+4hzB#F@Ua8b6V`p#H8O((A0}HIlp?SEt z*}SGwZEh=mf~xyZM@_iQ+^L0!U%~;<7OBrIrCvyddh@TGVwOZO0d7c;jl$w4bfS@( zf6r|F4nPuKuBx&c3oBZ8o?i$Nqq4$oZYWRAxciY{{u_I59TZ2`?F$3JEjWV)cV}=C z+y);Q9D)ZAZVB!d+;xD#-6bKoyITnE1Sb$6x$~U&dCoa+eRaQozWQ#}t$U|>rlxy! z@7}$-_wK#cUhDT`U}*br_Iti0?=OOJko165)Psn~kd?ljS%3JdkUS$iM{lVZo|IRs zXwlJVk}%|9{e||^@r17o)k4-yxLBpY5w;iDepbPQ$&Uf}G&knA&{nM#UXoh;9>67t zdD?5&GN|u&`6#vmWtR?4fNDVVrTtY-v}MOWVlD+}h<9Kb#1SNdEpY1 zNO88r0_WhA8E+oe@k%#L$XzBrWal^J_Q1VbhkGY{cg-!XtPC=2I_#CuXpe(dV-MkL z7At&D^G1R{2>Y=`;Ye#5U^SP5H&y;+7i5xE!XxLGpIZMyuviq18k{GB>EeLGq%+ek z$2+Q0psq$az(Tz!KJ{w45`Tm$lb(0-hO%cVK5~> zp$&MpK!7fFY}j`|0(ix^q6TTJKySZE*y`cR#Sc%;bh7%2(X5l}5E}FI;<}<^VSe7{ z!M`AEN~KUA3#eq`MocC;h%d&M>G8&9S=aBO>;zWDffQ3K5Qe*iw-iv?I-&0%TqMGQ z;lhZFD5KKlM^?$2S~aPRWP~DIkH`1$yXc>%Mfyzcv}q={NsmE5aFY{A>n)n#Jl+YO2IpI|-y_`7@aY zS}4WH&ps^Kqi8&x)e^>_a|(^pDyiAn znZDG89nG^0pCzXkQLKRZ8bm@aenuv8@Mv;Gij=X?R(S+RJJYh4-gYqoLCp)rRnPZV zHsLAH=?tUBs{5n61l%~MnoXzQcV`d@b&OIvhCdUW27?uM15<`qMm75plH?_{Nvs`K zNT|p(cB7>ZLxBtVG&?as5I#e&dIibc;>Q@Zf*D1lBt+`6w1kY};-hK=2ioTO?M_u< z(NWOp9(miSYvZHR#y7ll3Jn7yHX=SiV(xhArqOmSX|tCUL5Q*7=DVM+mA|B-(Uq}_ zaD?~kzf7&_Y=#MMU(BgF#8W=@IeUf_d{Ffr@*=Ioo68&?>RZa&F#W>tnc%v`qav25 z*S+&Gdf3$C;8Z*@+=i0a4*lFbUMS!$nSx-)+{i&WF3P2zOOSD}zYS{TU$(4!92YjF z?ZD+*WahR&ft6Rboc

8kqrTw{$6)tS)7-ZP6rb4U%KqI+(wqJHT|foD_Mv3P_F9E86l- zJ)d5h@zdLlg_H7ViV`%j9oyx4g?yJzW`4~LEZ))h*WNJM(pl}(d-k@s8&hx`w$`uA z4btP_xx)XDsz{1C+qt6%akKmDV+}(dzKjq)Mz_@-;#-LPHpo5%yNl+z-AQ1YtATM`aME5Y0`0;SK%l&^UKw&I^ zbX57lE~YC+T_inhSY1iIV2(THAD($Hh z+1XvPZrdcHrVSw%C+rn>PS-fkRJWDzXxH++ROh8 zU1r(W{5DgnuQ_aklcj&O#J&YIT3Y6LctVQG`F^mq5*wJ`%AQrg2hrDx(lV@W^;CMJ z;+K{N2Cm*;f^1L8KGZKB0%3d#TIQ-!gmQA6EYUM^N(lW6CF3{8)vT6g?tSsvHJRSD zF(%IGMC%sr)?{l zFwVtNNQ^`;N2S3a2qvA|BR;Ips(1ac^-x|iUuwDc%}|ylnq2oA0ZHh5#}criP@7U? zxB*R8AI@`b*4{c$m_YnKacD-ZuF$n@XwW$OhT+dlAq}14QK; z4hnN$Rkt)c@`Vfo_=s{mN_Lsdjiz@E-?3qPsOA6OxvXJuEQv68zXLhNg7)x(z~v^x zb49%0f|xobj#%Bz9c?Pggyo$8=5}iiAjJlFZq|3t0TorzGy^ramu-FA!0-jhQZ3~j zcaJI+;IHCpGQVgmDpZ1Ouoa;jyCsY0a!Dy*eIu0ejQtP7hXd&RPJZ3@XJLy^qUKYV z%q1Lfs5$a_pg57BM6ujQRFayY^BBn6=dqAqg{nGxMI|t+UT4=hOlBsw!#SyuJCP{rQzvH1~I%-hn|Tzt|LF0sshbKVuLQoW108vNy)@(VSSnfTfPIQ{r&$EMny z_xuzdKKJ)3FV&Is*qG&$xUx8{H#)yOu|%BHhTn@RbyErEJg0;F0CHSeX)TPCD*D# z0@?`o^Xw4^FQ=45$E%KQO57V4y6iWX3MpOfN~~NFm20$OIP{e~eZm*<$lG8sOO+t7 zXyVmW|GK{Cc&F8zwM3W!T8M`IGkqE%FdL-bgbz+Ph7+1b#u&`hlSUZTTHqQF7(wU* z*%z=4NG8fSF$S2fJ6s)N$s1#uKb0*oaE)vL@E=K26%(%br@vsz!RDVUHdBi3<@3A< zAlY2^2>qcs*+*GmCZxq3`76R0MlXMSyU^$yVn1_^CxIf4*-X3R*;Yvnar_1BHCwQu zhfuP&#_SzSR^zXtY@;t}qloF&4>HO?@|`gP@FKa@*w$*>ZdavN`-kf!PBoglQ2jSF z3XoShm-AI8%dnLUO@mh6LSj~W!OZle4ZQUc(2pSqwNu8ij?Ze8CliqogwH~uyBEe_ z9XeM29^L5$NY`tMO6%gKyXq<-x@P^EB)52w5n=Z4F5|;WEnkm4dqBDpIK`};9rTE} z9}+Ym2;x6h(H}l`#Hm=(?(S$8{4xFh;po5h2r8p<4MZF7WBFu~xEvayVK0LNm~fgc znHnM-Mq>j%iqfrTShoq4ZPq+I^fio>0PI$od+|A0f!oF&P|qtqEg>7*h!R7`1J!v! zNW~kc7M-5P{Mkwmp^cALiW|G3t9;j=)NL*N993gB&-<&^Yu2 z$ImWSUv$1)cy_G^vV!(Rg3g{Vd$^$Z+eir?lv#TVmUO0KEGs#GALEN|~5kH9{tO z%l%b2xevXjS+x08%jrM!@UOkC{f)%D3_eGzBgVd(T`WoagTRc-I)f`jGx-uj`2-xr zR-%?+bRFfBz7*$;#uY}joaz@PPBA;$74}_#Ilz-THA5ahLd(PU8{YbugYZdp#T8Z! zp?iPMO`n__UwAd!B~b+It?HeLlJuN`whY)cvS+N>u)G{NWOP}=1heIF?U;>L+6pcw zlWjTbH`e^zrG>j%Zl3&u9<}s6r3HyYW_aW;Y1NnIbU{8u;yZBjD2nZKh<5ox>bJcJ zE2h<#%55yg0OhL&)|aL9F`D|6#Gzj#Sg3sH(q{{k)q#wOR*f!fU-E9fT?cS^0q!e~ zsf$epDaF@H#lZF5j<+(Pu0S_H_vDxm-^NZ39dVtFi7XxU=s8iN#f(Oh=3KYP*awlU}*0&v9%vgP|CCE|{mh z0d~IF0o5FKk(p8kwe;0GL`&QH5{g@`@kgmnKY!DpNS%?&OR}?48ithut_KqOi#uH> zY`<726^x*;Au0(+itJ#2OJz%EH`p9x@w6(p&d%^MIo3c*?KQ*9O|rI9TrBU`j#NgP zNa<$Sz}eIS{pii|D;A$^9(EBxo4~M5o1If40^zYB3qQJ^30`OYBnTv`Aq0@Md+r-l zJBeoZ*1Lisnc7Z#ff+9Q%Z|A1wJt1(pe^2@>~uS;ip8BWOng@z{c=lhjsgZ>s1r18 z`uKJX3hjXww$xcYZS#eLUq zNj%)L{uwyb%|{vfcP#~SCDxeoP?W>nfMhjj*HBB|^~&)-gE0Ij z6f@K8OlcN`wNW%4et-SP{Ugg!Ok7G=9_#s8d@WtUH>gGzS@ba7cS^rPGfbT>)IiX& z_&6uW%1W~8A@b=*|Ci#RxUT*KR5-B$NR^&GE~MM7Kdx!|R4s2rP0qK_T)+9#Mu{KUjxAeuU@oL{*g$jBYyc18N&$z5q? zNR|0jFQ#%?H~oV@n`qzKWaP=AocvbI*+#lsug$CgE}bJqOn+JNcxS_qXR3;)5ctcc ze=Ve%pzaxB4*1fk&==mSYe1(j90>K?qcF^t5<+_P6DssV#0C`TE7VJRK_+W zHM9GYOGtHcT8m>iIXbyQ#4c?|xu#q-`Rv=$we6?|<>Wcf&>XQ*GuyyCUNw8bF@iZE zC$ntjbbAAB-4@XO*N;k!i9(W8vF;i{t zMK4E-7C+ZstxJD#|2a_jJ#<)?_Mva$9*gz1`#B6wg86alpBt{Zp00lz*&P1|!Tr3+ zeAVq~K;#t0)BAPI(e763`>jfDN2%<1lNfn(w>1x=iD=%XLeVLuLul^2c|3fK%6wq_78?U*`S;fid?8__7#yQqo1mB7nsXtoG&%nX|9#CC`ulG?;*O?g4?!^Q(AMTWx#ck|nhirz_veoj&giH}-w9kY&c7Ue2F z&(KNEF}M3+gcYQ{8};x}%r?UWH|{NIWH0uNwX%Q^S8FoIau02{HcMg_l?->w@hZ@T zMUN3Jz?G*lnNwLP>AIqmP&qlv+7@x}*gy{=plSWU?>`*JEnqX&aOY0ELn)S!PHtf2 zC|z!VBNbJ9Srrfl3~0dZ53qM*ASpgr&T#gMG8mB;QhjSiy~ zcQi0CI_Ki41Ldu{rG2afFiPL&;>8(OTdI*hbfsi(O=Ya(z_#Lb#hltUQacg0Xg;pK z%wk5<%;tETL|Omlh3+4OHf-J@D&BYZ$#aCvt`L)1^J$&azKCD3qHZ0v&#NB_ZKsg* z9+l37f0xEom!Dv2v*Buz&cuaTK4)m*+C`%>zW0l&w8Gj_aS-w@9x~6$un%%6Ytnr@ zsxr^A`e->TaW%RZ7ZlQMVb?qrSF!X)5W}Sxp6bMrr#B=xD_mA(cSihs+m1qfdte6T4&!u|uw#2U?8iiF^=EEd z0b>)y*24`(%VNdb*R*^Lxml#&v#NO<&s|?>i`eM3oD4T*s1Y`(FiIU;AS_~_BO3i}fWr43o3v)VHbc%ZE%d6=cUiETqGUG!uLH!HG*#>@NYQ}|^ zK2F;Z6g7_Ns7h;pOP;*`Iq+B;u}Y=RQH(k6t^1P@K-TR z8%;Z-LL2Pf%Nx>%t-3llr=0IoSEtcIn-wd?x)HUa)T;ff_fv z?Ry&8-r{VKTKN{@1BA$D)C01}VWW_$Me zE%qrRE~Zs%zd7CTu4i5)Pk-fj6VlJbSY7Gr@{p|top1JyYkUMoEjq@?wJVcP>EN_m zy$Bkudx=yslJ+Rc24QF;){qK^aVyYHaGNyoD!ReJ3Q(RGLbh_n=U5^kK4$=ac&F5tW1(-p$F|Lf2-SnGqp;mdqab^OD z=UNvB^Lnl*4ZG8oB@H6*QlM7pTTOtt%E)9bhS0mtAgzcl`1uTC^3^g@&3+GY6_s_l zmVH8?>;EM{u3z3XGnnM^4(;0rp-wbx_MIUP^yNFG`}>SW=8PoA_GV?8ba^f9=zl3G z267@@qkm^t$Emy+xI=%DJVNfI1|fLQ*k7>eK`Y}NX>JzcX*$+4XSdFP?Z*>d%B40b zoYi%GHLA}hVxt&iJ=JKUwP*l(oA$+6&ZCd|$wRqhE+W<3)~=gd5B42`L3YUPI;5qQ zW{;k1*hs3ACO4mBJMlQkaY6_2W_DDlle6gQXML8&=g1b+pGLvlu;B!KXBa5lOIb`~ zgJ_^yRRq_Cs*Z8Fdd$G#)vT~W%l3#x3YStEWy%|JBr|Mr3rbO|qFc9PPn$ETG54DFzsuFDsWV(d7e-_DCvSw%5U5fGP2|3;83u<)&;i>Ke* zT!V$yZh=vOQC@g@zOr`U$z^hr!;!@$KE1 zCPl;K2PQ_e@YK>@NtIEz#Mm1XI&Jq{q|muSBBWah8OnIb4$x zG_$~1o!`e#R$#?6W=o^;!NfaW9!K5i|8&ntGQr#`>-6MUNGBjjNl+3Ikmoq6G)~w{ z1x_MYTZTf-+x;}dAUUIkNUdYhFcEVPL5~T>5P&#Yg>5Llr4OGdec3ttljA`B?)wj@ z4BnF;tT|sYQp5YISCoFm-B&5@fYeK+!1B?N`!viz}MO+I$rDvz3quT(p0?0HiWf8nW4F7=@9E2|u%vIM(xP((){#sChF3 z?qyHHT~g8Xz=r60n)>v_wE5&+8OKv&5)}0gBdcr#S9j4Dr*A$hj;gXh4kdAawc-67 z%5@c=A#Iywg?z=vNMYv_L6t!j*HlGJ2^k@IpO_axCsTo@0n?wNN4T10cb?6EvCmW9 zXv1@5s|_thO8K%$=Rcn%k1?in$|7~)vh%x$0$Z7)N8~AvUKe>onNVK5ialPDD#KAzskvVAUR+gO zji!~?Ml7t&jaf)3#$xn&W`5{ET-_!6u3z6Dv{Xl5`HJJoX`q(VW+9VZPS5&nNwEd< zCH{U0Bi((jz6-m-%3%25L%0>$%1x)5ppZ7g08X>bIMY-UKn-WVL_idp$7GNI-jG2s z@jb({&rr~LqkF&p2_|0ksM;S_N*BwqknsZ`$d?%1H5}S?dYEP`pwgD?!<kjAGqFus&>}^b5)1jyHk<9*h~4p? z?l+!giG+8MkjU#|fSdMsp%jp;GKnI&DeZ0J`f$ZOVn=?ckockyqv~-y!f#n+Rh=5+ zjst*I4A;ErtcCEm4u!#l)HtKhWBNjoQFlySM{*)M&3co65OToO)~(KwtTlUPYp+j; zkNV9pXHSTsxAG_i-=&q<2^P0bL)~wr8Lh&LEx?~3RC8<8@B5_GD>(Z8AP}6h&@ror zTJjWr;c8vvluPPXws%XV74AUze@Abz~D(maFoGv6Ci@zy8`u}74EwM7w5(SfG zHyJ36M@4@q5WUaH(->wlVZ%#XlQwjui-Lz1g@LB{sBLz>qqzeVJ+1w%LU+?Wo)yaE z6S0dyrWbm&NoTBzpCN*eX1XCMY}veS=1PcHDm{qESn>Y ze+tM!-gqnKr{R{1E+KY64??@f^YCx7VDDLKe#AqruCI{Tpv91R*MXG$0u#ooN5 z)YOE}#+4jo^g&@;1l;&=mL?nng0c+7#U*@o#K3D3U*#fzRt0-02b+OM=r~W!twzH@ zUd))15_n$6#*58GGA+rHr&|CWl_@_pU6~7>t368fFm%Zql$Rzu=2IJ+g9b!{RW0^~ z0Kg;0!rX+opoymRz4awp4XUKZRX8H|#RTEqDXm|)x zteB^vuv9z*f-4gv^^KV;>LXh%*X9zy-1f-&TZV;&iwwCp(3wB!Y`l51Eo+bjJZV1siLnh|H4Lu=<00x| z)2cPBRY+A}&f;v~V2h!$uBpOG24xF0aQ@R! zvCWY@Z73;?bbTz;d3TfUBr-^5rp9hE7sc|XEJ?Qeb2^}>jBWz)n4i=&e&KH@+dDOz0f>ND;ez#oc7I#)`YS{y(-?_L;m|| z`yT`lVCPdtO4V!@HPh#*n0lyI^v zR8BM0V}P+o55UPgRPb(kN4`rZ-jK z_;<0sOF{MurV8t9_DbM?9W~hfNpBf{48YjxqFqMQZM=4@S%Ckq0{<3KAc>gSYI%|5 zcta8Bel2yv*De07a3Uuou$r*4lGCrq@~VG3=U*lCervOrrfCQdj&@D@_v8MmGc{6r zqfDIOFfHnt_P$04CDb?hndqW&zGHU_wO8(cGgZx6jP03)oeL<>oY6ukCA;A&H zh;_rHG}6>qt2xg&4dcXlvVM`uCCs0bUpCOHk&gFa0dqi_WC=e2up}}03jP27{IAgg zrZPY-_6Nmx9qpz$ZygAzlz}#!0tXnYB<=dLQbhdmnzN%05r-An&)QC5=amzGGq{#} zK>y6!QtC7O2SLTLe8W{@?r+KJf0rnawu0AiU`-oilB~CsaNbWi1?cS;12~`ry#Kw% zl3L*lr6y={Y{0hC?lJ<$Ah4j?uyZD;ZT)>&f0qltYSlTY(aWsMSn$#HPMnOQ0iRQ}g|!gIUw!Q5h_rio8^Z({Pt&GKcQUqK4CS6f`#Y3SuTd7=cs ztz8`k<)XNyOx0rK9o$5i@)nKFX5`m0p{ARnP7VUq@rp%r;Zv_nqaQJVvQmae$MXZ( zNB|evjk6wOtr$)LrwL9B^QdiRJwdx{qx@^u8p#da7=U|qT4#BA;t2}2SspO}r@uZ& zenJ|W&@*1C5$rkzB7I?vDwB1tly$Y8{cBNT)G$$ z6ip4`n2W*^_##qLIu9ChAn3u#fCn$q*((%(R z_Nks!;SDT_MjdN}5R|O4!Y2?^=2rDXH2b2~dqx%gjZI7{n$vdzCt=lN;D#XX=wzNKo)eQD)))v6+~{gP24Z>)qBjSc;qk+1sSSsfzlq2@u~syg6MNlm51O=_0B6;r$Rg?MJ9gCf7x2I3ba(?5c}e z1Dk(Q1aHP%KMBVJ@u#2tj3UyqA&Fs}Y#yKTkt_!?u;a>T3HY@Y<%%P-CRTJ!5b>b2 zL}_Xcn9WhaBW3e%s4K_}blRoqYEQ(-ZnllmD?}D;dTXM&9+vF}*XDGLm~r{lkcPdH z`n%ls32m5e3X<)>{W~WbicQ1%6l-=(SE?b*eBoy3s`gHj-|d}Ec)*I==8IKF+?JgJ zIDwe4n8MDPyjiT;v|^fSqtbdij>ptuD%|tsFAQQ8d&5LH98Kz$ppBC7x8JGIRdt$E zv`WAb#uxV1NO3KEcz)>ibHt5+X9@iZ_^Qzm)060wby=&f)J&~6H%#pR=T+)x|6fp) zXO@2PoDj_?xv=R^vJ+ZU0K}`G?8&Mgg8$R1WUn~w94SAU47!*6FR11JLmdUpny**t zFnnGmnuoB@Zhth}0oYjPKam`6qOv1=(-4{8y;&absk7cS$sH0;vb(Ms`1!xtGVv3- z#i#E#IF)NS-k{o}!aH;UMVH14B{}iE7cIZ&fQU(vU0wxUE%`qFpR>#n83r9FA+1Hw zjjJ9>jOEk(Ff8TnjfFr#ZNCwocsc{tFdvOwrLvWhmZPrIu0$89p_4d5hpmTD(Sb7k z=fF3S+c$gnCJ9zNX+jJi)Z!bCU{i+y0T^e9Nkj7{%^!b%pnXD=zZF98zcHqcMSQ&e zgOEhSR?9nBE1P%B_6NbAcR+FLeLvy-Uh8w!rP-skMy}g*RX5L+@Z7tg;UM(Qd3G7x z$M8hKxKYJRlGMvCr5#Wr4HKomcAQM^TO1D#!kFwz<5o4)TY7>JHG0!=??Cj75q}^1 zLM+BXuekx!{k0Y3n;pg|<7n4ybA*@HL`r5Mu{9%20a6>yPjbm?x0m?Bd%vxc>bon^ z@9L^6S&($da`@~5C^KXPs82z`NWX3Cx~zAck^|oj#4&oiveTRYU>N)W-1QWh_6l>0 zH29JHeD*@tx%Gk&R*55H(cAgw?v_VyZ#7lxYVrM zA!vjMCVY46Rs!EWT)eA5cKAK5HVeC34SG9^LGvs-dy2)* zZD!LqU&Kmvl45E8h{bO1!SZ^mCoy)A=yg{Hw2Xzaxz+^8lppLYRla zpK>CT(FR!dQ)fFBeX_6)TCcj2>%iVXTHviY;KWc6>~zYZDh27H(A~Y5413Uy-o+B+ zExhWHTJ;m?^A!sri@5n!nGm=+h)P(EO^c{NW13^LIL{OV^)*5h>BMLYaQ6>(L%EB6 zm#VraTV)X_SAhzn4dy?l2hYmJIx8)v$`R6B)uS0?3Yy4fwD>{rr;~9m(hNE^2I3^S z6w@bT0g*A1fIVV2GO4I~rk_X!lFnug^1h3n@|~rh8R1;Hf|}wX=!;QdRc7Mh6sl&s zDtf|G`6Me2%x{p@DCT9^KOn~#lPspg376?C(ksl1h#PTt`$k#kI|enqb;T*pVHpAH0p{fZeCtu=FDxS>4WLRY72w+P|Zi|26B&UCdVU~n^ z^Ul06nigjs^(C96C9RNLC$|7&s`hh==lVJVvsp4h~P$HgWg#3mCah0Sy>qjuN5 z7B+d)UG$powShknU5%f^>>)TAJXxQv(3yhl$t!5_;S;Jt*z=FpF!M`NRe!9w=9 zRgZ>%Ovu}~?`8@pkXIe(7gN$|SSUa64|NxU1GI_NE4Hq%=j>_ZOd(^CzMgmYpnO)Rg z_<~!9z(Z5i4rjX8-)ph0gFgarfWOmVUMIN(HVx_G*zYMWM5VsOlOT~<3!Jt0{vMoX zc_@{>q;IAg+AfkAHzxpYKpnRA^X9_NRwPi=)t;NthXs<5i)?P32`+Y6t!>Dg;}`qH z<86Gkiz0RGHx!?pPa4;He*l8UGh)UYsh`NC zhVx6~J84Bfrxy!xoUwMJxgo)OBLUj+NVsZpsm437b zuCp~M{HfLyt=XD?WdriWDDn(2or!ykyWjh(TL0>l$bn%Ei#|%@5<7BH?y>nTr7`)U zca#-^DZ6K2kGwTJ!WqQDxb~_Jz_+}6{%8vL zF+Hh(;YI;ONSi^~V+72w+AEMwV&Ycd*9o{mC$VC?=~XE9f`(P_=Rw@qK|HGmC%MQx z8ZIv;--^Z_vC^1T?rG+&11+VXcCQnYd(0m-kvREDQ67LA!$O4{+V<3aAyGPxno7nN z<4PsnueV9eSKJOOTPuC`9TpmNZSJPVt5hxUtp@>e_x)`;&MAdQ)&O&r2;$p;cz$P> z8M^qyD*K6od~m@LeS2u7YzalvRFNEP6+($DE2}}X?r1OyD^DqX$(CA$2G|?TDoXo6 z$dlA((A;P2kkpb&V5aOAgVX-AXom-&e!oSXvci)wD-QQ>$8%nYq!zqlpjh~IQx6~+~7Ce$l0$y2)oY|4@ysPCq#|kTY?L6RZ}G= zp0;-ib^QqLwNQG)LYLyJn!eg@ptvF65YBjXlSfSd!odMkrZ))R)YtXhp7WXElH=Ut zPv50W67}~>@UZp6{FGPG0kLol1la9Ib#W4gw^T>HlEY5WuY>{C^mE1$Zv!17-*@c( zIOb=?e@pn<QNXR~gTr%HJn8Jr!zq%MMwh=;; z_B;cck&sk+UVR0fir&WpBbfYQ6Ct$F#5boHFUz9bnrKLkSV$1O#dL1h+IOWG%JRkXWE=D+JTu5Nsdt^(`*pfJ(76n7)#i2J(^)fgg7cZEf zj(If0b21g#d}@sQ{aX`K?3%TEw#Vkc0OG){bW^RsJYR zy(I0BR!Ap9)N+aa`0wnQuc7&Sfi7r3gX0grh-lvkwom`JKOWL5BzHpRv^woq0vp4zO^oj`ono)ZW)}TBUij7e6xb zz0&tL+}+JKXb~T&iHc3|l6sPn#|<$lyQ6%Hsghu=hzx#qS$L9a#W%JcIL^XHmIq{1F^sj(y$KlfmBVdPD|3f zlkKiYsrCKqzF~ZIwq)=_Gctb8#sQw3DOa^XumXP8euSgJ zd^kih?!t_DxTXZfjD0+sYfI$BoBt4ieafS9L1i$HkMGD5c2#x{9GNM_vXenq91=~` zazA!CeDEDSExT8#8PwtMutUF}J_zKcpw7YOI7kUfu&o@Xs(TKZ=&JEs*W18Eq~=5V z>9s30TSJZ|%*tAY#~ zc(eiUd~wwv{})+~j-ConxRRsBej@067?OT4dp~S zdJ|3eRp*VpAOC98o7YSLebT5ZN3q>F&U4GpSH&zTmbN*OYe|>ywkJWs&zj+^l1&`K z!UFIAAo!dn%K|~hU`5q&alki8RHc+5bUj8L6raL(s5uTleq2gNe2sqx2x~q(nv_O> z^|{V3^xi6aJo#Ma^YGw`y8~glW+c0cLILT2cO>^Epoy$CfIL&ZY~9O%b-i68%X#?A zLpm~TjzuIb^Gm4b2aDHwjXL|64#C)b+uw-dO;YxFiv0=le3F>-xpK;?oRJrm`cQIs z7wsUL9^p^2jtAw#i45R+w+^nFb)q-C7E%Z59m})I0Ef9TooN#q6Upx$tS^Hx52X)e zfYSYonZ|9i1i&)Vgmg1u1jKNmVgz$2ulY?#A6|p&UL7baH3K|1LjeVC&u$Dmb%;^j#*^Sl*Z8#RS=>ioTK3CdNv~PulpY z?7XxKBtn&I-X&mj(KCf#t$DYk(olh?F%*sBA1KLuA!u6@mRt9pk`eOd+q zurNCD$RJ+ zTg>j3>y%;}-{XfWw;iuCgQ7f)Z^3+9^M|ytCfWDs2c+)V*p+oBFmHV2(ZkCjF4$|K z3c76mcgEZFXtbq{bHQs`gy~FPP_FA_2W)DcSk?^^s=LAHKL})((Rgz*zLLJ$v8Fw* zURJ7?Q*vHnep8uh+?`-p9j#;O%_K_dpaZMnO2>X{aHi=$16~r$WdwE>!7#b;HK~xU z)!$M->{L@OFdpf1;Yk&yuoJ|tFJp8kd*}Yp^NVGh+$W;sA;;oQRc|-yk5bbrYTmXJ z*Yy|Ykg_OvXRj=YlB3E%`}r4(m$^j;28WrQO92HH+UIc8{&kAvAKZ4C&P7b-$% z+ls0SnwY(rN>uE2u_c-E1x;5ZeHaHVfD(iH7`MqUnsQQzVNxjjHQUSDUS(7gfH^&u z_Di{>mDF+x-%csH&NL~YHYIleX(AdwC8Z{>0zLv|E2{Tm3~8r1Wu^$7=`2|YoLxmj zy^nQJzf>8Pmft3oqw{MJ2BUYRJA72j4vZD3s_l& zZvm@J%^F>|jB42nHpJ9&!(+`5k^h?i#a>=wec4n$pSv4R3K>s|gw}arpS|_u4?^;d zdT;)b@!S={7qLx?K1?!+uSn5mzIub1J`S0oLlIU>#*60&#wYKGb`~yK-aS*~FR(!+ z83&ZtE9Y@A7%w<1I2z%Dicl+F5|ii~%wCS4b)a356M{;zkF8vaj{0FKdma`foqzQm z+udJH3VMVtj2<6ZngrPaza!=pHAp@bQ2Akg5&Nd=dPAX&_cref%7^q0_7>Z|_#aQsGy+#~niI#su-YtP=j_O9MFvu3ScPd}@Bk)w`77|+wAQOsRJH?ZR< zB3C`6Wl770*pIr-27U&Y+|qd0U=x}=&tq2oloiY7g9EjrsOW#^e|=xar5JOU9vP1q zejVz~VF>f_L+myFy|>}M?%q{+)IQQQvK6{loP4)n;82Bhq|l(SG7{;fp(t*q0L*z| zf)RHneG#dTq|Ezq3EY_E-`MtdhVuDc(NK_>v`|(i?zj?1*AxrG_a)fh#1P+Y6 zdxxKkT0&LdCq+n5bnhB5(PF~MIdtTmwudcG#e7y|U^Ai)hwPv~7ly~&En*nWO?R7! zY0!k^riRkKJda8G>j0#IwTTbZGl#gHIo7P1*UJ9*$s(R`bBvdnWN<*+CQ%{qL+o%9pb?W+A{)RbMSN1% zViU?3I9^6%9Zi^QrOU%H*A*{&A?sqHO8OZE=QrDP?N6I!NEkDRXIT5XBddiBI zin2**hiOj^8t-3T&y7d)g!_$AFpjWwyEXLc1L^ih`1cu#1InjK;z%s(iEc_C-O|pT z-;q{Q?vEmPu}%Icr|lh(A@nz5=9kdT2qJt+nWm>Lk{a4~IC@0K#${$5@?)j^XQ3)# zd`0Q~SmTYukD%EeSpWx+zAgGOfn*S3+J9`+$L+8r9)0pcqc2uAe)9Y-P;|Kup7M?G zYAnqcK?JG4ee_p*G|Ibf2Q*Rpk@loTlX_w@x}3a)+8_>_+9(A4XC>;W@p1VQ<~{5h zA@aWfX``c6>;$d3APoRU82-oMgl`OWKl9Y0PqRZ-_cfP@9#!%$s%xX!2hzk|RKD`w@#WXoZ z8bG|n)4~uJjMBW1ORrj}NR7_L+Y@ZquCm7uzh_n9^VR>uPHVBCA*-tQ&>IMwMlKO7 zY4VlAmZJOp*L!H`z9lB=S!-f+;-8oi9}IHHClr&Olz!`9^QBt{<>Ra)V2RSoA6?Ei z~@E7#jI%Crvt^4 zi-;X14BU`iX^Dw5Ch~7=Ytg~3m~6|o)pLmh$a)H@ZeR~@0`B)n--80R;dpYA#?@Tp zyK$6U#Dc1u15dP>5Px34`vT}ZIfOOl%5A~ZSBhBti<-i#pN_0;!&4c|%*w%DEm;*a zs8puwb0x(iu6|PL=M!OZ?CF=K#@>JU>Y`x{P;fh2$o86c_LG;J2_iW%e^h@A6wdHn zCD1S}MvdnPEeo>#`Ry<{_$XRTMn?i{aKfa`y(%aULo&@mf0V`U9uF za}tYh+TqAWau_|nSP5%^7gOQ8+P{E&e82T_YmvmE&ZrOeHA+BjV3o*_V?jz}fm zM~Dez>3qUh8&@0KU9`*>-LyQ`tZBnD9$n}6M(^6od&38a-{nQ!S?T9m189Gi4Z*0X z18kT1df9%c@iwm|t{+C9>WZ|aYlyS#D)G9s{M}W$Rhnb+eVv@VF3PzSHIV%7oAjr9 z8QryrgCY8XO+NjCk)4?{BW6lNKFQY4WWF*OXtFHUR4t?CC`Wo*)W>QhpJ@=bg)u)D zIM`xK9c;BjF+gk%DxUsZzzj9TR1B^DDRTGzf36g|FzfKCLd-k!324y>n;F?7`(nc3 z7s%U|*UokIQ3UOszxsVYE$wL!R`}kVV#dSRLpn)L+@jfyVTE$`PCr1uXYb3oxL*hy z1$PhTp}~MRmF9bx=x^0i;}Bom0ka^Zz$3j7R8>Rc_u(`qq;i;Xgt#~_tg2+wH0HY3ybk}F1&(G@l{og9{8}mNJ!Y9&?RI}Z z3lW+?T;Y_?1m>-6rQ>@mf=kF2aA!EI8tv7^Q*$~oJ%x>NPIhle-?x)P!5DTPZYr?G z%Fh52SMP#LJr&>4!e{yhSe5tYR6Ww4YtP%x2Ce%=7;M?IU=SVylBI{fey=sb<$xKS z&iFG<#>S>GbLcik4+3^%E=egwHwldzb?gl+^oXE-!A!7thn(N-)-K30oFwqsetW+( z2Ly6wUIs%RseZtrF&eojV(5z7=yF^&X45;zNNR!N>Lcml@XqAv^UIpwkvV>F+f>~*6`S1rQ=hDz++e=Uy#)y|X ze`5`u@o@yC<);of|D3d*DuQ(r8=#G`ES#)9te~S-Epu4ZnhObLKeC?6q*Hi6a3{ry zmnHKdRN%7Eow=XPYQndrHPyTpH&VY&-(`r2MjInf^h*|$!poQa9%*965S%RE*sGY)DdXLS+vzumzHg0t`Pt;j0EQ2`~m)(y73Sbaeo8 zN({|1kLOW~E_`b>2C*!|5LOsm5pJ>6=(i~?6gnyn_Jtk-H*I1FOVGjwT?JNPXpq#h zwap_xeT0&;%{fVApbhp0Wpiw0&jP z!K%Bf*$ZC>RGf0HUQ+`{p zMI%Xr+_$MWLy7i^tk;fjm0+~8k>iD0M$}+DfBHr_XrR|slV$BMNdDjjLz#~A%sZU1 zPgrTd?B_iEnv+YGP^`PFvZCHN=a$z)Fn~X}3HQD}R_Du7aDY-`wU{~6bg2>En2g=C zpjL~hJPLA20#4< zbbNzbg5A(&q052L)6Ul3i>_MVRawZ_mJ=PDj=@F$nqCTu{lFE%MYhbH9t#-ynKM){ zG=-ZakkM!oy^slNLCeTX^+7P>5=I>WbC>}aWqnvJQ_wE55Y%BC&B3E*J_$MIKFw2& z4`C(hGpER+lnU*A$`iRZUrzwIHXdd4um_4pd(1A;aOE1v*vGl>jU)!M;~A=5C=HD7 zeZ|&RyH-~>IHzz@U>GYDEhw?AKB5Mp9cLhiQW=0In(9J<&Nls1Y90*IpGVbLCZnEa zY;i0;6?<`iz#;8C3U+0Jzve$`Liuf*m^9i{o0grYxpLJ-k#3QH+`e$xCy?6`O~_#qsjXz6hL2PmtxbCY2g z2OQ!}nA@^JUd_sl-7Zh3Y}d7|6nP^wI)r6OlI=e^qj}B3ivEsH3R*QAS7JJmg*t{E z%f5D~b}o`8q-{vTW|kzRNPg^(ZQMOT=3!XTM3sMtTfAe-^P$#eEyY?dfP(sp(_%82 zPO*HT8T~yrl%n&mRh|C;aYr(iV|RD9ztq9L<@-Vf(Lqap%dI;h@u@3S1#O3XzJhf@;jDacGjv2;p*9-_;HG#XI9YM8E;M-d?csd!-d*Ewhao^LMfTnohAy zPATAE`hy<6Z_0^_wgm{Pw$xdh`t9DjyMFGfJG61x-Zo0=o!v$=Y=7o+>0Ph>T{+9l z6O=sG(EMFb_iP98J@@1G&tn~ZIwhY-ddQ?QDf7?zy(E+7?M#2jj|w@@Y)RgBass+m zf(xHdF-#AY@odJhs5u&zjlImcam$l@i?t@|eZl>Q&27Cz+`|w+E;gg|gd}0(?|7qF zlI|nuc2MCH{mDHw^vss%xzA);^+>mnISdU5g*{w6Nuu|Alxq{(o5B)93jp5dRTJvd zB-1s}VlCzEI{=4%CMh83qqi#R)Rj8Ty-2LC^j;?A6N8s?4o`A@EnYaR3l%UfX(zBDzY<7x`cx-lprFWw?B)20kA84Am2nKq3--&U+iD^CI67G7w%)M`$F9z z6-?$j)q@0HeR@vP+{u41yMzX?7-(yr7ZQ@}joJ{hI0TPW^aX?WA50$gXaN;46`dh* zGsDPf27sANu{Z(Q3kmxew)Y@C@XlwOvl}!ps%wJPj3oJ5aSL#d^mDqY( zY~wW;D>)3wQOszg?N%EhqvZ2KKHLG{ws9Y*2!nb%!PG>JI+YF*bjV$8vxSFAG9-s= zN=T$>N$^J*o$$!RLez8CRS4{(0MvcO$L)&rcMeAcptdOG{gJR@oXU;^F%>UWQx|vY zqd{3Y_9)=eX4Ii(2I5jlX_xktsSkV3S*zi~%m~860d8Oj1Zyy6y2xf>yx*9Tj?1ZA zo=GY}QIpeor>E;!oKMQVA``pb_ZQC5F=a-|qCF-CP5A+Og`Abq;~Xa*pMMS(wvJw* zwCte-X6ut?U}O$H5QLFBg)_^nLY4GC{D3}GVpbb8!+uZAJ5@=-%;}FLqnxX?eg-uyn@dN@iV zP3)W`QaN#`rYAR#euADvipS#Mq_Bx@ZE_|ukz{zP!*U+1FfN+nGEFqZkudq}rs>!E z;y2vK3)FAc>7JSfe*qSM0aolUaA@YQ;r0tx;j~<}H7Tt2R^pM6XopaDyPvm*swt)- z$*9shMo4NaVcoXjo1LBVzH!;k#=@_OGZoor-u1vEw70aHFtGne`UghpSuVworlhBx?bT_h zLqAYTmWQVT`4pQY*SbOF1)|S=$D(nM;%IbZ)IDm>KCs>u|W+Fn|}S=N4o(A6p-+HrO=g6%3+kxj2a1A%57h6eh;X9YJ(-`3 z7)@7XsLkxhYRWv6v|?-KxqdQXi)!Rv5yys3cNct8_;I2wH{iY=MsYDU=-X17sleAN zlUB4xuJv5?^ANikkH*8ENw#=i>M*<{?;GXk*uMbi*Eo`wh`RBv7KI8h*W&gF;LT6*A<$L(_)0X-Q;ilGM~T7r4iZXho8T)#?VVXf6i!0-Qdl@PR+bi>?rk%Zv^W|TVkA zmfRcIu<9gPciq1$_=R9{BSNBGl0>{W7EF8K0JAgAnGCoto`4w6q?VJR5D_3UEea$Z- zTCugvMzAmj69=s9T?xj&&{RIOXytcl$%N(lXu32SmM0%m5oU$3M!yms&3Vok8AdRF zNY&>vs+JcMXtm79F$3PEKEwg*nsU0ND962oi0ixsU*$N2=di@3wVVlb*p z+@7i?=OJMS@NDNOuHL87s5@}T*AR0WM;li6#zL;~b{Ewb;_ zsfiM9^GK%ptdS?SRnHC0MXgLf^Jd$Ien?8>tSc?gD;|;vIdGdxoXax^j417}7+MjH zWo0@v#}BgxUmUk-kKEt1>CSedR;7I{d!r9^>rWZd4!l2ITAWU;8>Fqx$2(mAZJTPhg1zmpu!f%-4Hw>b67vAKU ze`0%tMPnYS-*!k?fACv)W!%5?)kEvomv<&v8QxabM4OvRBo{u&X9z-LnO89-MB_AG zCMrH^MJzUNxs^3w=E=zO$+tnk5;k4yzu_KqnnzsTBt(2-WVC>G)dC;`t28&w`@-nZ zkug8^Dwu9vuctM$CuMqIhJp*pa>Q$>GUxt0uIW&Yig|<)T?y|Y=uB}pEu1{CMhis+ zB7090VBw%g=suP@lVh3B?`0K$ET>VZ{Xn4c0o5m@kXs3b1YY2smYJr96JZap9nsI8 z7Wgv_+N=P%I2HzKJ7rK$9rudRquy6z!*=~ZjZazXn;1;e_}ny+sGF@uCLf2A9HZ|{ zW(`qWGjUkA>*ZftJlcvgp0{*8+A122=C|^9#Scv=)B|+MT)6JFr4ExsKap9QM?!e; zNmnTfyO~5xIV=0mX}Jm2rTCZ0nmcV&M<>1_{5PWNbpVKV;LQXNXns=;6HZ&wICA^O z2@FE6BPhx%FHwlxmBC+pVl*3%5iQ3>O!EPSGwgzcyn${{ff7hstQ!^NjU--?gQYQokIJeck6UJ5A1+L{ z>fe9#^u-uBPNuA1{$7E71JfoKr^Dq9%bv-sA{nPpo9}y-M>6wCV$TkQEQ>l#p_BYvP5fCXj zl<2D^8bWfVf(ne88GSJ3G_xJ&q#Zy?iS1NX60m-|+QxUKJjA0Gk@5}IPH4X6QoHgD zGa)xR?9*tQI!+?{mYEqxwJPJXlY0OdjfG%5-Ar1J3{xHt`)zOn3wzNxH!&Tbi=>!l z#pYWZW=KI{V^!VI;h213zbxRAH!70Z@AovBRcrtS4Htqe|87ujBS2be!4La`Ms|`> zM$Y8T6eEBh`BvInQU!~s7vH^uFJVP;MDRFV#E#Q6B-*nmn`Nc)khx)w`=TbhBmhgc7=Tgfu*u9T@mI?-44?C1p5zqGDSjx%F%gp;z-FS|6MZVE^GGu1VZbrOMsY{!K9D}G$bb%rKSHsYE zz$=Z#{}?4f19u#&AXG%3YSC`-Q*Qd*q31)@YLmt88%d`=9wBNA9j^|p007Plmy*?K z-0Rmh(Xpf(f@R&V=Lmz|+|QZ(cuc&eAXhtksV{4xSspHtW68PLiJ+#fIpF4Q!&Igs z{0zguSe%BI?zGECH9ZVOq+^-S7z}f{^uT)55>o-7n5~(U+#`zyL~fclna`>`J(Vo4_0s~-57q5IwR=f%nOH&ww>L?`=`fh z%F>E&eip-V4I$JS7!v!6?h>&)(xF}r);uDg1b;kpMbrGfE!yuHaSg zEBQ@^3=bMfVUn&OIWEiZVZ^;Jf22ePAWZzSgzh@jJs8?qT0QH5Wg^wZHu_=XvgN+X z;vWl`JdGmF`N2pSjM`=@Nz*YZSLwZ?j7G<$)M9H?sfTMV*4FIENIziy9qzEXV5K}Z zjN^&OP~6)=hlImsD1?Dji*T9o_)1>&WlE04Mu#$@NeNq$l!yeW;|ae)D}jTss9&;H z$#DartC;a}qb|{HX03QQ`H*IFlSq{)(Z@i$EdE*J1;)@aate^DmY*sH;$jY_4ECxH zj<$nU4&yhCRdU;(^c2L&t`+!P266a=Lwcdpw*`3+KS1D-ZmD43`2N|W&2cMmSxPbo z|0!i}f_){_945{#G_N$-GRW;{?A>taz|k(Z8{=nRv>XB#?freDpf8sNT=HB(QTbI) z8Q&!Hw8BV7`-(sHz=x#W)AuJl-?qZ5sJ$ z1ic~rB*hyfWH?v@*TG^jDfp~Gw{5i*^A^RSD3JPbGJh3a8x(R`=(08q9cC?E6|STb z)}P|fmD*&mK@}YU&}r$X6vW!WvZHPhjr0RKxC6}8_GHU5U_ujfq%48OvNtdrNY#^zz8 zbcTQrf7!R4z5eO7m-3k!=73yW+Wl9N;0z>5&~aKH5&HRUfI^hlh{q%CK+Uwm?&}bGPySQ_7d%rVFn>N|?3zoXg#V$L?aa`klFcdK zHa<&u|NS&SC?@Myx7If(^3?Yp{;VWKau(`yG=0lY@EkyXWlw%yv@7&;ebuy;6P~1( zEG5`pq^jF!k1}-#&QsA+o$E+cv0SZYLi{8}XNe4-I#u~CeX$-MdvD-FZZ)tIOz^Hk z!`|B&i*Yfu6lh{3G!7B`kSW}?y8Y8LU0yPIJ{1y?oFvCTR9$i6z>si7we~wr9fu_8 zN*OK!2Dvet??ELxu7R$hNTie!uoQ`1dZw|BYEt1!I80aAqrZ;*(JJZ-+@>}ly9cBA zjjH+Ggx#l0?8GFNJ*u%_4pY!)_We&45}Y^XSXcQQC3B|tM$))qOYb93`~_5)<^aAs zi14o*E|(Gx&L~-)>2`}&?nMKS1AaUo&&57f%dv_?&CI3^!}OHZlNZ@aV9`CAnC63t zv94qer0GdWG0TF^hOu)Snawl8GuH*os!plS;CvOcR29U`4TCwz=)J(EG^UMGs_eAU z>GbQ;osqmX$Ob)TUW`ZYkExk8YB38~4PFn`h3u^7uL`tTOWjs>ek3{cb^&!}nm>Fzm<6NE zH}mQX&1ObVnLgq|bU8B|D^}!i5pm>t%R+ETFcC>+4^3r?zXk$-*eh+pO?aLyK5@lC zrj|i_gkxn&?@mp(AUcT>6DxP^NWC>}mB|5r1{FNo*wf%xRn@+R1VAQHe|o8vt%WEw zkgtH~yo85~z{RLIQL639NLlGMsJ#?06D<24Z~X*aJIJGjl(8?+I%Zu*^CmBIjR8OO z(Bma~!zA9xI)T7pa9&47o?fe&j1DLLTE2?*l%Wn7F4h*-b^owG{+T-d#oYE|rV^iH zBBPaH5B;hmNC4&#ZY+WA&m|C>n4YcigWD6?Iw;=Wb3@)~3`a_Wjc05Y{6+#^@xp`! zL~jKi-_~4tyceC)M&5BLr0Fdy5o=waen*|kJ~};A`-#ED>99C7aVQ|T z_mNm?;)MCNcBYp%k+O6Z>7u#UG|*s(>lr0ZiT4v@X1W$U6${k?Ix@$hBtBgOnO@o2 zNHUKR@$~r|)u1aEiwsytMq^(YD%x30dRuy;bbRKq!#VfJHzEqypbmq2U%3yH49APu z+UkC11uXqH9knc-B&c9b{f^$WK3Qc0<7f@-&B@rF3I^o^iik3&go2ys^UGREH7^}g zG}uAxIV|p?e)BnlY$dpnLRid?23r>9mzcn3O`l+}`feP@S1L(UQXLoo&m^6a7@ez+ zy^*-$R)#^3yzY-yQ50nc$Iyl`XXnu0C4^Eb%_5;zP77w>F((Sy7bm%o7b>g-MA&cT z0lKt}YQsd7%LV3$S$#`szNYMbi6{fADxE$Fpe1Nx4A1HM#j(IAJbF0ogi0CK6O^JA z%ixoA~j9@MmB><%k66ii$Nhn2mWWb8|1;8ZG3k__yqV7Z* zIJ8_?UTQh-K4)=BBLw38$ZrZ1kwU`wo{P0vV&Lk|gszp{doD&S*Zry7p3N_6K*R#z z6)4j`^cw}QMHLy?#QX}kc)=^W6}&o9l9C+y5+wJ^;cme0=O4`4|f`- zSViaBt~138gf`H`J*J=7>>nSeWX_5gdu@Qd$U-KT4dS~EwcolKrj%1NHXaYjl`%B+ zTqsY(T|Jj45t*vSC$QBpiH_BdT?se%+9GRV_>kXaV(8mK47CNd^LT!IO&0(!D)b+B z0SLKFhw&NjZL&NpGJJPy`&Q^DsXrA$psAR$X_8jg3$~@)2$8u(1C^pbfvj>lT2wi#i;O$ACg>;+OT@ol*TSL_QR#(Bed5#7v>*_WK(+8Y@D zb1>Mlbm^(BevFs(c2N1Ovm&Z8$}dL3?UBKRoRe%q=~F7y-{)=xfvKhBs*=GF@?Kq< zyFq%vUjf>+H=fBv*_!0$bx_y8B>7UmuOuc~c$s!eJn_p@+%J^mNUVpZRN8T zC;f(mgwB?Nl*NJFxRff*!1$Tfufi-(0i(b>){KHGO7CgJRApx$(nxG5rZ~j<-A;B~ z&>@;XW~*yX$y7|CQH_0Sf>o-xWThmRC@xW))|0NT(tSR7;72%zIm1?{7fJ>aTwRW%>F8i>9Ox9{b~1xCJpshR7gx zB$O3lI%fjaeo37c(G%1A0*gy)*1Eb21sNny_J!NX(ygW#UVFL!s(MSGI*j866XB$-d;ndZq&uKkDGDncetW8PK&E!7h}5{)uG*CSuxQpA17_q^w6-RrjQ_QddHu6zFfK=q@=18#Tl z(dW!9bT1mG#z$%7f4FE8F&z@c5Hp82Ens-9)CmmvG+p=l*8hJbce^?BxdP$h9L{n) zQQt~_MwDE5V<~Fjpn!PH8+2j#7ZBIJ{YS^=pmyAEwd-;#)O`JVfd@AjSe(O*BwJ`$ zAJ=a9N5}r4cHM8q7Q8ia8;6cvMTD56eb8@r;p+~4qT85E6eW?<;k!V^^@?B*_}66CD; z=>e_XCtOJuxYQlI^*<%4zl0es=U4d+$5!IK&I}7L3>*I?*B6HW7U(7Pr13%;feU4n z_G7a*qZZvEjP|ubpFj}piKwGlNo9=N<$lbi=$_%_P~XX(vH4Zq_?}crVEbb!rLy7= zMN-zJzVSW7_#x;K+E^}j+hn|I=jR8eU*g|Pc&q;DhX*aq_wQq{6;o-s2|06RY6$ON z0N%z6)4%n9C_=$0%uOu>TO5T%b?g6O3OY0Zf+r$C&q67Xbk6vy7NQ7 z%nudIhyWD-XiDF8n11=;eS4w*h`O%cEayPW9Q$Q{IAD1({2|H+o2%ibG6{85DC;}#Vf5;o zBo6#qB=c^Q=;jUet)(O1U%&`TUhw&-SZ{USPUzb6Bb(EH;Ydk`e!@OiHk zsrY%WLfN3xWUdS;szOUy9JaD1fjZIcHN`sNUjVu(f1o1$$AYl5N$Zsr4fPsfb~`H+ zD14918JqnYGi*M$+Ij-r;2Ys2SH)~DW4BX(E@5|7f#1M(NO>T!XSXe+m^GXB`l<3T zyYObN=XLQ#lNR;5uK(&M1~#HYW}Qv>JK5u!h)q6>CGDudjrgB}WT_k6{`-hoF$8Go&_HkYlT5I=$c4| zyA{q^j8e#%(+yN#ptHzyntF%)T>3@5uhLkvzDj46=|}24FV9ov*j1!zRZ6sn9frFr4WGLAyHYw4|>7+U`4{B3o>7Sw1bo};T6BC+I2%2ZOptb zL@gw4>~PTHX6SHC@|L00t*5M=wffte8MiMez%{1&Tg3)i17yvrm*;Mc2Wy{&mlf{!0?)Z$l>5q!HR>zZ zSI&R@7sda8l{(-dH$;PLGeHnsDeflta*{6?@Qgm8O zb&|~dJqW?fX8vUU#=_tA6LC!Jqt^Ijs}G(v!y1t2209mFE+k8Fm=l;kri9XMBqtxB zhvoLDydmKmEHOCAyO{+H>b|@yd>=WOHMJA?NX@J1lRrBi8?(HrYjT7hmQ{qA z9LJ#1U#rUESoZ;n4Wf5!BmKxw@isr=`ZV(T?1u=8GFdJ#!u678Nq93b^QP7Fann-` zzQi!WZ@cPaPfxQP^`sr$W73EoLtdoSeWc)=FJzKuPzR1UDH=HF9-=%8tZF{!sfOKe zLPf&dAFr(_CUlUZ9ENQ(;2T@tAFZ$45|Qc2|5al~t}Llm*mAui_7#GyEl1mQxh?)1 zgOE;%)7ly{tx&!T#%}i1MlEgTJ7nyzunENXRYjnmLD2#v;cRxa$W^djjKss0K9xGa zqN@1GYQSOcZfu{#JWPV4`?I9REV5r0VQWD}Huh z3m{6jm6hhe7~jsDA0`m@aj&qSIx+JZe-an#9*;npuFcYE(x&WX`FXC5MGUljV(3N5 zAu0QrLq9F`8d9x)<)$C3O2Gh*U6 z8d6!#NhiSRC>?9O^?F8gQx_lP;OtR0hUGh#A!1;;BPD+az$dfexne05&CEOG_ME+C z+%d9je)T`FhKl&P5i#y->Ay12I3mmu^&L3>09jo>E=dKMp$mNEyDi*?5W;;BeXe+= zQPz4Evr}xM1#jaha`BTCbGSo`X)Ju0RbXy;Xro6ZM=59hdj3L$IZzR4{W8Xir`6YR zgc>KVYmTuY8qj%Pp1nnU*C%LTJ2h#cZk9tTlCUOpo-n=Ri#^sd2|-;rBBb^woJ&5} zJM{0bVkv-ceDq<-gLNQMbY}(yd?&mF5hQN{~;N5%6(t6UC~`Lr)Ffixx|Ig`Nf(cdc^Pl=QEF zi~&0u`jE;apPL%kn2!l)``^s!m3EM^U*;w@MKuK$NJ8d$U~8lD4q)EC3A3~!Bw=MC zJ*(^VMrpuBNIKf*2t6S^32~r#pyFW*b8OkXxJ0j}r~b-nU6@9&Z)&J%3(;2B{RHGs z_~5N(ZmL^DvL#{n7VF~3A;1_zFvZ9R9#Sb>E# z$c{8g_>(w7_Ks zTUWS~@ClbnJfHjM%Rt4bnD_fw^XR-MRfNtcE|`_s{kYQCMKi%Wv4lF5T^b~9<1ymr z5ERyL!=8>uIuH-OKKkDz3ON0Yf`Xm=vE`d$W6o60L0+kBzpWrkrw$6V_w_VQf+?+y zQybHtR}?5mr#&ozmB^%s?B{Ab8{=x2Q5Lck3Pl30>Zeq}kRD92=#b%MfxN;bY;&Do zxR`q6t9V=p#*#Alo;3N8&^-w!;j>MLR&#H?0%s0=yo6!2@#G6_d?Vz5O3Z>2=^HwS z*aK!e8FkFaEZyh;YM_tyc`2HoW)xRXgYW659oNxE6$n;>rZFQ}Sl6i$f{W!Sy~ng% z`2|g4u6I;HWy7GFzQ9#BIh6uwh(?knF_rrr(jZ2nvK;VO@JM_BDZXnV5xZEHYd-$7 zk0^-*-FrGZ?l)~s>s~>>47DtWx3pX%&4&F|vv#|0wH(jfVUQ_;MBOua4uD5X*i7Xuv63B;(3PB=7rD#>uwI|1S_CPL!0#%gwxFEdtOOEJC@xTTJ*|!xl?Ys%x9M-) zeCA@GprglyVU73rz4V>=|y}QBia;sU;H4L*_?M#7@(UR>ahOy6?d$kCSU97id`Ok!#@7y!m)YrSBK-*ZTQn`O3E+ zC5B9gHPC%3Ld5LsP~*Ji`9h>yrEdBEl@m_+Ol5DfsTt(I0BZ`K6h#Gs3e@(a5UdSuVom!Gyn-i)zP`np^qMmBDftqs%0{V1<*mZVsr~FN z9&zKoc~5TOpUvu4>xjs@6UBjviHO8u zVnCzUuTNdCO~cx=Td(?BVU9qU0lp`q?(lnO+1oI_53ZF6kXj}e+w6Bc2-Z;;cMoru?GWNhRHu9u`Z^okwe=P1>s8kh zPU;6GC14zMT$Z3u?mgHq`c`o(C9s{S(?FxYce{;)Uy75J=Wx=W7?)v(Rlw2o4eu1x z5xK&ks4};KYB7WmL!X`G-48XwOAeU0Qa-v07lxsX(u@kd_FfSalK^EkxhZ=os{^P* zG0ur|M)<7t`xW-2KYuq=a$?kiS65(V5moo+V{W9lbxyoOmx}M(_av1$~6TNbQy`dM_pU z!6X^}R>`Z1nd&Q6vzh|n$nLY;G%~p)UL!F49m?2)zdv>4A7#FGIof0}sBZSU{Diyo z0rKqRkRJ~!X}Fl;;;IWFv0{Cl$x{i~y1^o?v1;ftooq>vF_6@BElt9n-it1Yvz#Hj zw7pY>kg8=p&Oh!Gpmf@0?zS9OF)0$M3Ck#<{^A1~05|!=XBr`D<_w8N4tzjRVnm*M zsQEB)0tK{a@9$ap{fMBf9JPo%ED>_3SHUh$HH48Te;^;wE-rMm6oAm+ zuh?Dv1`Icm0=`QXK63aZQeFN+JjR_OSYmssA^s9GtdHg`F@G#POP!W0F2TEgB@u9Y zjDEvt?befy9(ikwL}ZJ<8teRWYEgLO=b%P%s^Ps%$JT#%$*WWP<}o7IHA(~1*(!*h z$6^Y6i~rf(y!k}kv?%l!Fr*nWmA`qKCo)<67XY97B8h-t55Y|PFrI5JcWAm&67%>L zO6E3^AH?+w!M>D}xg?$^5i;FNfIKVB0&F?u7)h{xyt3kMR&HAh7<`0D4d`mY1>u?I zE`Fr|W!mD&r)BtQG~+kwM46bCv}fPwDvAlT-(}@Q)41urs{HHqpi6rX>*gETlLnQH zGS|abKMV?s)fagMU>n0D9kq2;U;Cc%hpYP9N1e47rgJ~M`+nr-@1b9uhW_e+rS>XRl_8F(>D*3<1)8PGS2K{b}X<;lG_nT97DTIIfIra=ypfZOaI6zW`{< z)z@eXWRKzz*y7(WUpD}v)2N& z;g3rN3j8P|K&vS-2C&a91^*=nDf{AFM}BGoJb}$MNFYjqyO+WI#HV^Ur{UOe*g8zh z9wH<~s4m$k{;OZ7mc}Ad@%C>(Tb{hkg zfs@9*_mXC z{Wcc4rw@Kz(j3Y!V8CsSv<*KRD$8RE$Qr=_LaUCK7UVpu)(~lBq2JX+84G;CE966w zSolh>8CG?&wk;#&AJ)qj#@9ry$&uJUY{!6qSj8#fpl;g!lWW{Ay(nK>p_&qfs)y&dt#(Uf0Hek=AKNh{nWh@Le+RVY>_ zCB}&%m8WS=hM1WGWHvtD6+D-83#Kuv>3E^w%!Udv%cYkMqzk~rnr;r$u=oBL>b02< zlj2mH(W8`*O*xWbS9~efNyTy+J*D(w0_w6e5~3B++1Qaf(P9pAY6HIx9WCc(t95RE z5QxG((ytp7Y9l<5zj0jx!BJXj91pQ9Ql)@RRbq=ibjrKc>e|n6pW6LY8}7_pK)TbA z49Mal1Pu;9zHa`kZNjNkKr@-06D6qSf>E0e84UVoC>Uf|X!L}`vqzRC|J*68wpB_G zLteU^BTbRKio1R=d!s%`!^)nxN-zZ_2@+zDry83{FCFWk4;6Vuli3U*$z_O!vDFx11;t-)@E=0;Z?(cWBRplNID^BUP3 zbsD3Lm;|$`k?T3Lq|-9`V)1PVb{3+`7uakD2O-7XUoJ(IJAbkm%h3+Ap>TPHZ z8NdG1yg7tqMmgO3XDbAjlm871!+23yP}lPhT|KZgBBVm|f(Z>BD~K)UR}Xf2m|>v# z=WlQh1;7Ns!ChEU>EMIL3KozwW3$3njJ4wq0 zHZ@#}j9jbrkY}4PTP7u`XrWZlvwR9wP8>y-!�A$cjzfy!F!Tc6UYRkj`kR{1sB< z>n2p?qKvB@LkKN%V?f~UT8f)%so>WF_A3d7zotj5{mP7QMNJp!t2Y{HGW@xabC6`{ zbR1dtS_UhOnvIcY$X@v8o~%?DDw)}~oq1I>GRtm?!mCJ%&p{dFLSP-5oM2M)@OQxL zBa9e!r&^`CBK2{Dt2n{A7zQv8P9?`4y=zFrgVFyf)|ecShSHjvSWk3`fmZm~L|HA- zZ%j~msbn3qcI!NQ;}1s=)gRSx5PVhK6rc6FFi-H{4)=`l>U!G{;?hb{! zySux)d*kjjjTF|nyE`=Q?(Xh1?$9_i&_F-Fzn$Hkncbc9_Wkuv#;L4~n;B6jDkI|R z=el@cXJuF^{CFH&VlY;ysr!)|X%>kSEjIzIE5np+0=jGd<(#88CgWB`s;hOHy^Oxc10)=QhB7nL zLJV%|+nLKJ=k|(F#mxPOug`CSi^_fPN!lP)xgS<(mn1K$$zWtAw~Ox2$NNx$u=;Z3 zd+|Jl4pEO9VAhaUT&*H_1$RHEz9~ht-1e_?k6=KE-e$UQ6JH`K!*5_Dhq=vYy5`nj z0AfmaO4I8cGksMxoh^@|KdFhmC!H}+24BA_xng9l2a=yc4gaN7L5fu&DGJ7Ms{4M^_F0Q5MQe^x{J`5>LAW+1ORG4$pAb*gs1WWd$`Hczr%%JDw+KY* zHb9^+AKw{_ku$U$L#LH^cPv{4>AF#eM)ekWqj=fW7@B7;G!V#SIa^w<#Vb2dK0fm-N>ro-ssQgoT_<`{goA5v z=bw%9;7dzf?WXTkq*Scc?p3Bg%w=X5dIKdRi!Y9}Wq}S`qZ@BONNG3O6TaBXIsTAi zE{FvWeM24J6;`VBK(|2*!0Ur}T*dtKe4l=o-`s@8aE)On8?MO@G$|eWK|uw9Z;(aZ z1BahC`rT3@Kb5sX4dT$cF-&Dt)?CzXXs-@2q*HbhKBN>s`Zr}-^;Q?E21V$cJj;5D zDY=zF;^?{c3N_Uv;q85SCv0E_B5Rh9dSh_L4hyf+QiuW;*u>&%q(Nai7G52pm27?h z5H1X&9`r?JO2dJp&37?z4K;#oYwv=ylBB>{k#rTg$wa4&Z7w**wopXfQQl{H8?lcc z+jJxgm9>B=Y+1#tk~C(>jqXd~sFY?-DhA}G3=;=cT8JQ|k*JsbLgyt5@!7qd)u zY`-nMx>XehQAR3;?pGRH9~yzK!?Hh-QfS5m!A9&T5$60HrDu-aO7CIbbuY88O~{%g7+yi=K-_n3*~M}1 z>-OS9g|KxCs7os`({~__bW@L~4H@)LRAC84&Yp-rg3Q zpQjwGFj8V}$)B2w+TC6Q@-cYK_&ME%!_~K{I>AI^6um88HH)}VU*pG9NLgT}>&|zK zqqo@0CvRr@>gcOYf8?{WUH|dQLE1ttwySA8oC-|l?5ymRlk_82iRmEGrN$OLW>XZyLh=^eVO!F!K>w za&HKuFF{_}+WiX%H5Can*Sd%F`F2BkxJNq`I$C3ti|iuL&}=dhi0Y5JyKsY~7oB=T zFn4Vm&GtE#znCR8nj*K|%~1Bi!crV+Lk%z?0)`SLMZEma>GrOk=OZGQaLS(lILY-z zhTcz`5DisX9k2Mm*M`u;`!w9DOXtb`WR~e0ubssd;0PIWxpTuj4vO zjE^c;cURDmfgyAIq-Z>tVi|Bu716dGNT<>93*F3z(=q4kY!=G#|4G*o(Qn-$F`FTi z{|go7dXo-qI2=GCT=NblrkhF%T~?UezEQ#Q5xGHprO|-peIej0Q4 z2-4`yLX)pCiEGyOAb&GJglT658-@Zz-z+lOsnL?>U}4`<&k^T*=Js(&y5vJm^%s@~ zl8WLP=}BmdEhF~L8O~{Cj-%~Vk924U|IwWe7zTo|t1&R_8NJhgc4di;v;z!H5(CD-)01S$G!xE`;-hr-AwWP9vtt!C zSlj{a<1TUt6>l~|70h8+XdElIXT#6&{+*7J7tkMg$71mIDM{W^Vu=JLmGI{sAFSAU zO_pA#Def?3^%!#dbvVAb|GYcvJX-w!RviMyg6o4SXpQ}bi4^>aQM~Kp-fr-yUrj=j z?W&j~T!>dy?7sw#0D}>mQIumg>{87>XutGxB8GGZ0a)QYnB!0(UJ~8+b)laz(%F>= z`R&JODO0tLC)gpKFoa3HkadXpgo`yw9}JMq;s4pAC~~wsQ%BK(;IF- zzaht7Kak*q0n|WxU*&<+JxIcAtgNypGGjIxM=RG*J=LQ*jsh1hC-OtJK&Z%_>W!mS zrWW`$dXs9I9{T6jQtq6-b$+OoNTtWYolk9z0nTF*&X+xGBW54D^YR?G)JLT!GYMF-q=? zepLgH9>IOFx9Gy9v(ohA(@vMUCnL?wBwfU%jH3(93IS0_Ir1b9O^K?z*yxD*oE&wb z5}m+@G$&40v%y5~pzHJSMDZXI$qy-!m{$ey{gN!Kva(U|xT~a1qZc_3vz^z+d=|H- zf&vGdJp+v*fMX|mS6Jnb=Z{NT4>HInIl_s~>11ck&9@J?qBt?q!&aq@^=48udAbw9vTr-z+>=-pzIf||;Gz8GFgiHRClDR>Z+G~n|(}CJG zF^X-mbW-h8C{4Olnyys|j6RSI*@#b7L-thD#+N^5R?IhqrTV3~jSz}@0*1^|B11I^ zCgSx@&R3T$S#w+8pJN054a;KN?E}a2s{ntvB8rQe=Z<*XtTD#^p`5@5o59BOr*QPL zZk$d-C+U9=I=%(}6Uu-K%RW7`loT-Xt z@AL7HTzevVO@ATOb9N>Kg`JvnTEg&*N%Qi}H@N?DaQkuA%)8uhiOPqt5QEAl(WfL4 zwK(vKS}MzqXB1M6kpl0MB~>cIR4zn2v7;xWTrJy&3vwJYpO(MTodPq_#c)Jm862VI zE96^#F=Jw)VWs@2fUIO)Gq>nr)lLaYipSXtjhdsTrErj@z!#l^<6%G4kQFCmKr66r z98Z+8n2Xt{Y#9Znx1fK0DBZ=Qlpn)5lN7ut=h8tB#Yijq^2E{*xN)_GL69^ zCRq<*aqoX9S`H`P^j9u4Q+GO(m}P(3uTs98m7cZ%3)!E=b3rJo()`SPW#?u3s|ybGPr!9WDc${JIChv)@fi_2qUt^|qxc?<{*i}*RE0T|fg!1NwY^6w>PVT!?gi)iON zrP2D?Q2LAFF)X^g$4xDe>#h_J{F20=bQTMUt17LccbcuR;&j(^Rx3uU=!fdqRxENH z>*A5r#d<^(*)oYlG~Io&6M2>kc$#|edE92KV!v>o1xGW9(^*1W+ibd9kUo303%W(r zV`1TwHN*o?He{@r+=xy&sw@GyJe1ps0gJt;7!9bR6y2C>%`ksVG6o%b$_znL#y*~8 zFo|tpujx-)&DHW*aQ6)1Lz%;O!*i%MMe2;pIfeZlJ}JfcXt?A<+I+)pf-ffhL2@fZ z1Ca_vlvT#OE(-hPs&wPkA*~hbST`mLje+@Ywz-Ljt$FfH4ig1MwE1KIU2XF57Uie3LELf;(Go%&Y_{}iH!b}A6=Ko?Fn?g!gS z=xch>e3Jc;dD;A9d`FpN8|P3FdcLP>k!5hIRKAuaOjD&@cNpUTAUkvF=-D}?dVIju z*n-0ARjgTM_9EtL9V5HcIqyGvzr4n)$6`hJ@UZ(P7+>~<>dMfg9JRpI3$~I};ItX> zJ{F+`ds>sZ6Y`T)1XCSQr$|W!o^67WSq&1H#5Q}Q(OvZWQ3m!JT2Bnt&p+iC@Mhd= z{DNX$Vg6X<_tWZ(wI4ns`*T;d@-4y|uW9p3nUjUDTQ%cw$W=VjB%q*IxiE zoKEd;wHoe&BxybfY&(E}PNg0SpE1voyl^-?yOBr!T@}>rYwgvimd=K3nZZ;;Y6m4)v|`QrU2$9 z%#8_N1U|#Z`#HBPc`i}VJ~W(`RKinzo>gslhs`r_gYAu+(_Fu(bft@PeY`T+Qpimb z!NlmrvmhBwF_XtlMT++0y_DH6pE%p(^UTvmW3A`>tqfh3EGMyiRg?B-gEd!ox*J(k z<(&}a2EI1o)sOZzG7jr%5FSRi>x&r04Hp8{Da_cD7tg6At(R9CL8-g73h3F$BN&G$~3ySl4sg$TH(N0gj_BMGOGWQQxM3oBuIOL2a zEzo&0CpHj=#zVBN-XET(1qLJeBEz?xpeK?pG{w##xOGyxtqum<5$REr1u?pGDTp3; z{=y=@P!Nk7%X63*5`oY@n<>GN?yRpiVtA*q|9-ZYG!^4Do5d{1BQ6eM4~};JtBObb z4-eg@9y`n)0zy;GH6YVV(zGmkdczM}I5^s%h|5o?@*im`l*Zk!Q;<%u<){QvOFOGC zQ9RVt-9!7*LELm+^+J1Fi2j-w>vvq_pV8rO2m3v@x5nPjwOxMyMtJ%^5&pvyCeO7^ z|NV*oK=}7f{ZGRDt+DmL>*IeC{#)mtO|1#hUlL>ezw7*e?E@-E6h*G?S^vywZlmsY zJ=NyP`S;z;Z$)8Uq=Kdg5C1Yz;mC*L7H8udVYXXeYr3wQxfS`LlG9wHu^oNfq!5XZusT}zJ zZtEgA;}4B6t!&eoenNZ?bXm7_I>jhLsrj!H!F=$A`*8Dle9ja@vh?&dhMfNjdnzz7Bq$dHRvG{yOJnUOU-DaZP z)rkSjI2KDJuo8-G_~Boy#ziGyq^SvnBq7D*;ZNkV3*<2i#L?*h>|Ny}q;U zp%G5^gpsJ0%F-ng)pu3}%q7Y>1G0pP(kYHDdv{oO*?0~ju)+X1n_mnr3ZC?E!qDi) zDQeFWyNW3NcNb{}9Y-tE(voK612p40^29JdJ}x>g_Ihg13_nAt8EiWug7Ou|Qoi;z{rK3emG~zl?xp^ElSgB*7<-P z%VRXCfJ$4adIol>&?#mQSgww)b9z(KTNI5y>W#D@4pDeVpa2&33s0h5;nfiUJYT&0 zHs&r7)d4cC%?3YD0exv5wgmQ;1VOrpz8W<;vK$7F%s<3D0Qbr&5zt6Ves$goNTrsY z{zSb44oi3>Ox5^>I=<{N@;E)@KO)B%uM{4O6X;YGQ-O>}O1k1qUtU!|`fPw&*8N+! zB}#{1&Fq^mgui_P6PaZCeuh5f>g<$V{PSY^-`_T;W50Gq8YGbo;vs2sd1>bNHeC^x z?o<-so>7mT7C=KI!NH>pcnW-WD(e<6L|_`Jb{s<5G9a-06`))n&kq_Md+HQ3d`RLzsI!!3a!}aQ z^W;oR@T|#a$3J5HTsM1+3DU}+ruN+wJLuVQ(r#A6T6Z1<0?GzSQ?$9xQ}N~b z3UK$p___4?V)q#faG%v$e2J$nF!xhqlw77|INv*K^LbW&3Pe}daNEGUeE9TY9_Rs= znbM_(rBzvwW+|@7zM*sIa16dlYul8uDx#CfD8&yJ>?cO&;{%nsg=2)Y5-axcKEnxX z@zzn2c$s7Rjru70B#Alnk!?-?VywCmuo85OF#}`Ucv(5x2n5uSyQ;C?nyre^GCfsB zE>xWH4~-UfUQo<{;dk7RwUrJRt7QKa$85XzBX~cY6KL z>0v3jOsK3~xF@o{QwdQHA#p}vYL;_Yp46xYa})G*Hw841%$MM`YoEEsJV0bItip%m zuf+Y;A&~;A5bgqMQt*NT3=`;bSLr;YZAuRU3R6gYOMfuy7`f=IYD3vv4*ZLlAgZmc z4ux##BiAgWg>0DWu!e4L`d^^K-DXznm$?#gi7iK_ssr=`M=c*jf$34*U!NsOH#pV2X~r zq_Pw3tOR7$#mEcNxW!XRY)CynmSegof8F)!xdKknTrmaX+x5fBiM5EMYa!}0=N(O# zh$NQ2-$la>CTPv=+Tncn7_8=3PPR{7c&HMz(={kr)GeVM1rHOJmlDivGTVF<7maSX z|BleM;F=L!Q$E~uBvl8vPQW~SgYHUH#$ft#n{gAhI#6h+Q_~+D$jXzQhk*$iB}+~V z0emM|9blBZsp6fC7dW$hW?4TO8w_I&&G4jz*K){x+=!cqpv-xNrq3+vKoZ|^n&`w% zd&Zxd)q6rXtWt63Gur!gcawNR?sRY5OPpfC*n3_T&mAXH6=m-~Z86Fr{h%4NNZqug zd6C?^kOBIsy8CJJLnfZzO}(^R4)FNtabp`haoA%da7~TK1wT3%>IkfKprLFwXSI30 zYL2nu3xL4Fyz7N_5U3QnubS+$=sMMXp%J{uE!OJnE9^!7Tozt?j zzm{ebc8}W;Rb$O2-|#^Gl8A0m^I}~pzr%kSTkusaL6eT(`$u}!kjqMmdKw4}4kHDY z_5|M=i$2%pziH3;)1L}sJrGB}q6~k7+5g`YOy2VDRe9Sw!}V7vCrg=MXJ?6G7t$hd z+w7lo%8Ec1ShUnM0FkJS#k+DW6ZYc1>WSkiW>JEr4?uc7JCaoiW@^Q>6N?;YW3!s+ zB0PI+2nCEreZ3eQlllwEdT7U>peDeOdq1;@ej_GbNB2IW8I3;5SF!83VVyqKLL3Gs z?FL=Oxx8BG4+{<|nD_D(O(d2%XSGW05grLKq+Z>a`+|jy0KP zo8_LENX_xZAnZtL!Yv$a$=7Z>ji;cz)8|5-NyW6TT$Zh3wVO5KFvzBm#gfcWpOood zqI=yLT;ioSEwR4<$LGomp~(*ooh%etmQSNTv-)fFNc$|Z&ZL39PO_dW<2`5ZVPe+pzuJlz`8a2IJn~PujC!yLsCy9)Tl+i8)2t?zSlHbj zVCcI>B<=p7{h8FlVB}?hBFXyMFz8xZwFK2n#4}B`oaqyRd&6qZKtvZZAyGdiJi>pC zIfZ&(Kb6~m^X>!AHGti@=m z7I0CW7$ZVN_U6#{?kYD=;mbIGP;qa}_SgI-mSE!io#rwOp9#w7TMUi6IKt8`f3xNJ z&G3n+Lc9qUfE3Q}sRE_5 z3U!kjL&uf#l%d(c0VH|?mHIEiV8m%T0^GrGKxB`oSx|a<5vLa2ZJ*qdkCyE7f=iLR z7N6Fglg{u?;4CGtaInQ=s&M1yzY|s5`4Dyv5fpR(Gi$janwKagL?#;sdA#xiqq~8f%hAZlmrE1FK6tDrtxk3)AIg25tPvg_lzZW`DJgq*r zS)hjLg=AQtkM8dQs!4H}eAWS;GIOkxCD+U%+Tk`^B+Wkan>RZFE5WQAd{j)VhU$Gj zkN}k#0kMLMO2}WPD(r%^i3Moc$P$izZ_3i*CIrh;$5|}@1@O~9>14Cct`+t%3a8{Y z*s>2L0{HaY&17URDYiErxleIYB# zDeg+9lj!w$+8~~S?i1tudU^6lSY9|77kLZ*ES_HMDI%a9F+BAbkYZYZi~s{Bkj>Qm z>Nu?Xh0{3x#uXy;W%!M4eHGL%LUJ3&m69#ZjI|n-^;~G;m@#u?$^l{Cc-*`reaUuU zzj~u&e^5t6)cmJrx|XnTrWWkvkt;9!yj z%&9cePL^ikFbML0jSwUK8*8~-BgrDCNu>ZcKC2>u`;FW=6?nS%1l`X53)LJOG$$hr zi7YnSJN%&B;Z+gti3ka2zAgeu?`!$$FdmMPsl0+Y0rNTOeZh?94+ZNUg-|@3G&Q!# zR~Qn?go=fq)Emp^y;fn|6tAK3D04@9R9fJAcH_o2XTzi9p$(M>&sUled3!{#sKa|? zI>g~Q0^oWlMEoUmCAY)bUt+VZ9P~ZVgp_#t(5&L~m%cv|m-kLkLian6l`Z8#Tt1>8ZXPvXHRs z-v0>I>i#bh%#0FRn$T*B#%}0SKC%TG6m4I$ z{8@Gb9nP~4H-xLi@zfrwKjBJuYAcw+MbK5@vq{2btKl2ZH2k4_d^*twb{Z#y08U+6 zuQpdD@sO5IL*zz69!-!?qIZ|qkNzvJ(i|miv~Zo>Jprz zV%Pwbxnwa)Fe^+nxNk~uJMw-a zdoNbZxxW42+~Z=MA>a%&BI@k3pj#|y6~T&r06Hi0vdRxhy83jWjKcPg6DfF$r3xpd zV8&xX5~rBsvum0&11ZXCrX*b#c#Q;3DU=uEas>k_)}MT|CVd&cI;-yJRo9k1+a6${ zlVU(%XQsMQQ81q`-?qQv9DE9tmPF4Kup3TV#>>I2{b3YXima5<3pG>}AI78#?0~rf zI#|9+h#vZH;5jr!jc^M17GDtM2|`MjVA(O)CkBJQ=NyS8)+9F%fzToY>UB&#Q5nQ3 z_7Vo3g*+>QOAjVlBB6)bf+89%Vr*%29yT--yvs<;2@-YT=OgZ^Qm>F>ctf{4>uWsHz^f|$v_nITR{umO+=DtBfm}7^DRu`5T zIVR59I4K5#0u#O}LfaTFkB%Bqf-p*D4!SBJSvCt+w;bbgt&+r)@{n$Rh89ke>X3r+ zmiaB9KOX#iLH;eoIyEhO=x3R7b_Ek|kBwzBj)2Q|V?K=*3`a&zcaAVhG>QrN4f$Bj z>+gX+TNk=TEC<|8T{Fl`#|~2lIr}LN)0b<34HR$q}j;{%dl#=fIptZ3a=DPd)&2BrQ6nJ_6PaA z4k5|m&7D{bvp$J@NmoT=rW9tdNdJ!|iDUVJ@!fGs-BhYgfoULe8ekUDKo5Qjz3_8C zQeO;YpmnrRAFy24+Umk2nOIqw`5g}J^{U_?m}-bED^F?Vv&D= z(F5(A7Lv>5fL~+`Wo=FyxlKBc@rjEf$sxuh*98=cA8m%;s(;Tg)qqwU*ts{nhIjWpOjT;Nh5yC_-=n@XhUj@Tq{0 zN%p9SNnsFqEMYcJa@Nc)`&DI0T=g+QiC(?D!AcmHc9R!PP>32+UzB`v0GMHEZj)K5 zUa>hsq=YGJ7pFJbAWqFo&*k!gVpiy$#M^Gp_jtpiOEQj!Pm}Py~FL~NT zn5vBzd6-rT#NnS9{jb6MIZ2lAo_*o~sa1L5OD>;?jH7)W$KDiQo~hRkcpIwBwqDBe z{2scpSIm3iB;F?ros%OjR@?= zO3$FIsO>JGY7_S`!9-fN)0Dvf#E)LsY2CBDwQQ4~Mx>&sZ*UN06KioKkM0Ujj=Il|EjStuom z>V-8%klnVmStCew%wn!3*FLq-5k_}F))uXDxGROh68;43^}V=Yz8AzXbzKeo!RHdW zXY0Lg&0wD2R)q9Rwvs$i;>$#pgM-ZbCLVI1!aFqhb_}X_KH4@A0}7inE9r1i53L~O ze(X#fpXKvsQWS9PbHjZijZ*T5O*8XueaD81(ECyve#4i6-4_(5*`?+}HLJJk4X^$^ z%It&qPO*F-T4_DK2_||920C%XAN#SEnN@{R1FVoHaAJ_$c;Z2es>Ij9AZg1=Nd%HS zSSB^2DJPLShXW8#bu8E8juZn36s{VnsW-Y*k-5Dt-%XS>l#g>6VKsS;k;3$mGC-!h z%ci4(^Bc7VH#_|4Jbbbn$#>K6fkVw>#ph4>SVCS)tf=0p1&=%5=2_w7t4lsDL zMqzbz-n75AMkO8jN4915M&%vZ(l*q~M#csotB?oDR0OUC^L4sMJ6hI)1nFRoOc9aO z8du^UfqD)(!hztg0vtQLB9*KC_{LMwcq;k#2wY#Hj}m5Epgivoey?XqqpN(;yV=RxkqWMx z9Ym($O@E;j*U4+4!FO??Go7yY^#JiVb&B+MLd$=xC3K^I3K9{I6?;pPQ_NA=>hCVJ z;c}w&jpPOKmoRy1V5$Ud)-q-C3SX`k%NHKa-=pxnirTe6#_`Uhf=TCEcx6Eu9sQl5 z5OrWuZHZMs2!5Pg&AFv`4r{(q?#hl2kh}rk_Z;&$LjOV$*Z=$5nMDolx|4|W=el1q;rYprcFYgT24Qb+;;&(mKsdvV2(d%bh~i; zA-G9h6=14Mc4`pM0=&+w@@Muse=3rIyUUAGX3K{fq`>OyaZsM=5lgX=VdHG|LwI)y zFa(2^IyeBCr`X)HXLasMI|UiIv!EXw95>dlhJWfH3eR71me)B_zd2GJ?r9$&8`zJ2 z3MFs4M!n^=#o7z>XIKxBuae)$dJ>FfYKv81d7>%*H}(H^qbIMp@WNj}j4z`ubz6V_ z$Dw;RQOxO<+DtVoj1qBo&G&A3{bAqK zU?UMPsB0@#@P9e)J+u6e@h)oVh<5dJ7Uoc2Nk+l+X5E7}#(*DW*12%tIh>z~9>LCG zlH80W9L^upU;eWjKV)I9DRJ9^>?=#m2Xbv(Z^F}c2{%Tym1zE zxlmI(6Z?u=DFM7JK6HLa%}+YwC;#-b4(uCBV|Y{Jaw1-hQ5gzn#^BGBxn1w-2}cM= z$8rJ(T}JXW1e28j>^bMUjVhw*Gnh<6m8xVj>PH)FKY;oFWc?;r`xaUts;I5Y9B^=< zx6)i~_uoG#Bv(pnW_p#B&+=zz@bVS{1Q+o-k0MEd8xx};Pf+ zd}XRz)1uy)jD{ z20W%*Ig9)gqRo>sjjF-=m_*0Os%kc{7rcO6p(UR;B;zv=F*wu7 zGj(iAjP4sGv;24&)`k|^i22vs0X-@rp1i?@=9xU%i_e!!4K+cWMTOr65nYupZQC3hNxP$kn9A}bjC47OS(CI zIV$bT$#B!(&7#i{rn!PBdQ4?qjAsq5btZyR#7wxfVKrJlkdM2`T23b{xf&8$tV@4* zxJpv2jsAY-`xZwDfK|d>z=3z29jY^`H~EdYIe^|D!y+TCd%IKr8vD=c+6um91-BM5 z74KGmAtFOLH2cq#<~1se^1(`J2xN&arz}dqK9TfgoUuen<&B=jF>O@VrHH~SqwfEj z8HH2QgoZ~xkxg4ZUAxOB_zBdUEBb0py*gCoRy5mh;;l%?V{-q-5iHNJG&2k4SV-$l z1_B*lho@^1F(~8rHUH%NUk(lmgCHR-BxrRB)F$`R!lzilk>YN*1oYSEx__IlgT4?8 z9gx1y;%PYdA6L*Y0JhaS>zdIh-Za|*T8~Pd&~lw^5X%iO8(h{qhLkc4mHE7x|JuV3 zg^${0e0RMnM{5s?3vJ$r_SUAZ6jUlITyNdrhSJJ(nd2ARCrO3|a|{m;Zwu}>RLeAg zc2U{BENRgcFX(w63ydk*vi^Q^=?&k2w>*NgQ8^PGS7JZ@sKz5u(qz-KtPl1n`D%1?|CrihRl7fvPRVrZw73o!5=+OX5c^kLE@eb4XP|7PpaNtd~u0(HlStX``ZjWJ_wDWb?E_ZjSrayT6= zCj!mEC`@NjOb^K<0>;JlQeuE1df6^z6XZw764)kBvQZ}~c3f~Rc@Ho|&~Ym3`*M4__vQ;nPp@rtL!s4(&U1{r-@syX<# z5FC5B&R)>$O^Y04VPPx}JN_uho<@O)t*uc$`exrBGE5k2j`zM+dlZ40?5*@e09o_G zxZ`r;CLV}ixA2-`&J++7q)H9Jb^~CHZK{_Vc)SU3LZ+yv^-QVh3&9)Uq{!Yq}ayjwR>SCvuNr@lz?W=)_-sf&i#xPKiI<NQQEkC%h11X&tUMbQy zA7$+4_@B?}0?zdlVV*pQ-5+>3Pbh)X-xhT_K{hZia;N3wDdrUpN3`!qcvoht-@2jO zMKd_F5;c1SX;qQ6isCekNGtCDxVf7b0fU(BQ)8|G=Oo_*0uw)RN@nqBTZ#8nMVf`wR<7$WpPA zPR#^94~9tr*~3$!^1ocF(h~!Hj#h(gp|>bdKH-*8pUgcRy*(;a#tNI@ZyAwP(if(2(kRNmqf_qpL$JmKGU)dcxH??e?TxC!OPX^+#G(w z_kRJwFm7}_zD25m_(c3y*-TC7e*u8R0+MowFwD_ldX<>R515E&#^m@98dxh}ERnZI zu~<-cR?*|ZfB)l5ZqBZkjCZprS;H9((poh3(DIKnfsp<-{rD+K+$_Eq&w-P&lm_xD zlNCiWmZ1tBl=}NNaRcAd`(dBVvLj{A)IqG(eu=a-=yXtmN0`}=yT?C=PUDZRM6|34H@?rzEe z&U=xzPfmc7iQ;M|xTMEw=Aq24J@vTk6REU7RKcbN6C7naY82yUAD)8c12#Y!hTw@| z4dLwI{l6`RD=4ts={kcbvq{GV-0dogdoH#Z+$s2q{8mUk)h9VloQ!(gb}lFSSG?G! zB2(6Wr}h{(>8N8#Gy_e|@G*6;y~=t!dX&o4U~0J(OtJOia|&ZX($X<@B6Tl(tS%Ka z)I}7?i58mkJW6~(2zd@&^wwVND=U)+h}e{8WycQL;op#ImewIEF;s2QriJPDMtz155B~rKgE5b(U zva9i~0E*q{`>oeIR!7`SZ7;wI3@cI!#yl;L zC*|8j&(VKf!CZ-Pt<68Xof$CUKmVmaUUA2_$Cw{S`?+guk5|mHe9|>xPdYsyxmrf% zSzD_x5{|tjxjC|52P^5gq^%lPs?Je};0MKzw-rg61&dm)PWp}iQG&ymJ_9HP|0!^n zTs0F<|EE%ARkzV8{+~-q(r4|zYwkaNPf4rWh+W2nysP4dkf9;FocK@l|N64t|K2u@ z-}$Mmbp3j&f(a0B;D>6XW^q?xeVrU#en?53LNla{TZ#mc8Pky(>Z$l?w=C+5`PLA^ zCOcIkt3%QvmRJF!8Z|_qNGzZt?mDqZq@*!w`2w4Sx^gXAd{6=}qYq-$B3?#USTH;G z2-mFu{L%pAV4=I{TtdrfPc>1<$cM0>i23ahpen+rN2mqH6OiaqcZB#LSnD@m2^RbR7p`3G1D&STh_E9_G;6I$AY6rp*qklWq@r!Xf0IV z?Wyv+h-+T_Uq-ILiH}=1ugVZ{OMM*lw{qIWDHVc>L4PU<>>?z_M00gxLX0w8Ej~A! zKI=qL_YJljKb5+H!c5dm(tAlb@u=5axR_t+mnv{J78ULsdx;mj1NqxD@jxoaVRIFi zQz7etXoYUQD-_%RpBK;dq+9j6o>bFy+jJW8)SheMu>qQgsK{e6Ipc?$@;UnxuxPgyh6DK?Xm=`$v|C|Qq$IoLp2`XZs}q9TtB*~ zaI$E-X1;C~PCQ{vC+Da+;?(KR!xj!yX`ICcr=Kh~QF&x#h$^@}u!!+x%y+SglGnSn zqBKgB5x{!i`3d<&!j$Zz>8ZnUun?tV>}{+=Ro&n-1&P)m zFTs3FI4on*#o*@d0yXLHsxN-NoPR`pm^jFU@O4!AEX7G2WsLc7*?JM*uIbB+KZ%>AoEJrhvz~cJbL<${>q}v!P`=Sa6;pYKgCI|h zjvIEeigH}ui!9&u*_{d>L&PxC{)C$~WfdA`K~SHL(Q?=%?E9Q(4-fV=YMh*dMN%!% zS&-nf&rhPBS)qUSWd?aEs%4NYIg~Nvc)4NE%u#QTfVWO zIU+a3W!gBA$`Rc}jcQ`hQfz`xi*N)}jR%lx!=xct3w}DnT?yb1PnXO#s02em(#^r+ zQ}L4M&^3U?wpXRLag*SrCVQ#WbbS0kaAJ%0uyzK)w=5<}z!pFJ1=ta8?~RJgp8MUf zL#5d6$07CT35EyKfJVz#;>Y#d1L6n@cgy5~KBDk=WcQ12c5vHNl%}9Ey@uEP*V`46 zyRZ;8z3wC_1rZY<|EITDUlJO_>SD#Oa-N1ke~niS3kQm&qd^H5`S`&O9QH4Ab1)&o zR==NqZMHi8?DYJy|Yd+|HG+y^ItbY_^&LC(ilTNEm7M+Ui7v&%5 zb7SfQRmPrgrN(WeGFt-0oia9`@CnahJ|poLFn=bcC3k51nY2VkJPfbLlDb56Zp|bs zWy@nrAW1Ko4>eD*yu!g_|wsUh9F72UjX)_O`sEiO+6jn#y20GAt#RYq}>LKGkXMrko>wONfRKAoMRaYaKB-9TGaB{qN< z^W%ElS%GDJtJvi$*S05ZBxJh%*Bi%5oJP+)(j{=)^cAHuD$u znL38oP~9wC>(2Kacs&G>N%A5}r4hf(jtIV`?`KOn<8a|}I!^le{hbfSd_gl+v}Q;XED3xR}nK(S88-V{zM(lEp&>rm*U z|A)M<3~sC0wv?Hfnc0qIW@cuVnVB(m%*@QpF*7qWLt@9wj$=+7Q%3pjo%{aHo0(rz zHPxp|=d`*d9rbQFyVqW;U>)_fiYiV<$B2pyZ>L8ag^o9Jj?C_RogrzC=3e$b32#5u z0j4%jQRPJ_GC#yejjN)D`NmDIs%q)yEDK_;G*y7_i}GIxBpE4OUZv~%avVbnZBPg# z;5`FZI{@wX%kaf9T_5z0xVcL}6?B|?L;wGNs<$BG=ucz4QxuRgM?RlZCatNZ0t40J zGpWKis6vtt$cY02hbOSaJ3TU&N(BcnzyipgMUe?b%{OM^PRfhsYnkQzk(8svt))kc z8BC-p9@Q@IPSAJZWz2CT;LT z&Svs;wv{~KI9O+Gn61ejg;2Mv+PgKU9Yq=Z){vh0WVfJ?@WsX%#mc`r8WLDfewi4; zQ8pPg3eP_#LYEXimXe1t_Ox9JVxrl#P;)-kS*!3S>((_>kO8sTgr6881(h#lcAYee1I~N;)w(VRUyGuze7^KrU5u@MMN?(+xMcn&DExrvZ{MK<;k^fpr zUE5JEf{-$6q$a0G0X_bRJ)yQZeU!lGtBM>UmL5duGMu#uA5Dw3v0QQ!VvSpUxI2;8lAchZ>;Qb)I$1TS!*WV3HY?+ELFm?^iz)Ck%`YgztwaNOluUnyN*{{4DkCbU9NC=aY_mo+r*2TXso8n?ns8sq^@5 zAq-kj%8eg5KrVsm@;lr2$a0XjYX;043DY_+@O)+{@(n7V4;@TtKJ4ek`^npuE>Cm@ zYdOfD9g~JO%H0!DJ{w0-4t>HFPCpYmEs?rn&TMiJWgHO0)iTv{lMyk1yu={{CStFKGRYpe68 z!5;v+1ZyD259+((nAdJFglP#)6H_2YN00KWs799@!L_BynBhn^jwUi#L!be)7HHzCQsGL zVL_%G*P+##Ql}0IHzb*1`Ryb?>{3s@9IC9pPKW@4W?L?5{A=`fh09~`se&+h`qZsx zEjn^p3|`1rb~HO`SZ1Zvowy1t!XfOj=qRilaVaba^eowJ_jp@F2*~jCfTxHC3rqJ4 zLlaw7>mLWPsx-)QgDQS7A)a}Z_B{Bf-D3??)eO;nXj)b2q?8aaoD*Z{%VJ!Gv zax(ayJAF2ynsfz=LV_^wwVSkNHxkdpBC3Di*1^eGfc^|U&sZ%vZZ7oK~5ZL{w2683Vi z@-O;oj{0_vZ~zNhZ$$@)!8WoDK$YN{!$U1 z{hu)TIE0{Lh%#FoUR8w!BR>d%{fE{k3rgnz!*)u_ko{D?iX*jlNWkQK95Mu_k2C7x zsi-^zkM|!y%@?DBr*~MKXgN~i4Hc*Yhy*8I)u@fvc3}@gLUTgB?y6on{aEOv3`b%K zCim@1M(F5#7|Kl(yA?(U2@4M&30ot!A`5r~YryyeE8bz$gJE)wEZ?Ln387K^4T5$! zJj@whybm+xzEqfr9(WH4y4FC%swBPamcg(=0^z-CE;&JqSt>l}#_A{0uf3d?ycJ{E zA2dNF%ZPlpa(S)#{u4)c2TMQ*d6q#v$I(%F3I;-9iNzhq=~;THzu@uKPK=C) z6%IPFt@VhTa93m%#wIn3Wx4mv3v;ULKLEYik=1v2Guq@(?1TqlO+3&OE-ED-$a+Xc z^WA_bm+_qKlC*W=Py&`9Z)#|@Ce#KB@-CZbcUi))7!T>5=0L0KOK!h9T{;bO27D_C z331hl75lFB1I{?)KWdgi3)FIitosWFmmBOl+s79?d&^Vyn{0`&#L zSuy&B(~5$z$9Qep$LBe1qo;hZKIBuqmHdbvH2l=*GOjH#9S+H<8)G0BQ4i+XW*FJK z%bB>;qX&Y_i^}C>w|6a`dIBz_ZQt}YYpW`Y!xh<@eKb!<`CqF#g2865uKH3ECSC%I ztv(A08;{&9C{iX~=lr6mU>EHbhSQUC0^X5;5^pJq{@5C`^1_rBu#^&nX5#bL_fI93 zs20u~LP0Ho$RQ3jc4D_dbp#!#;QJNrHNy8696t`*jQ1 za`&kby@h)?KRUyko};%`bQ5_S@1Z&cmbfCKW|z6>X!PVnY`fKIZGV;l)w=d+qFfFO z5`^qp?WEpO$X|sUHeF8 zG>;v=V}R-4i=RxAF{&D8K-O#f_5D{0xnWmbF2D{W+pe5s-y6M<8CBUl6mP`?itXD&(n z%d0^N!TRIsq`dIi&7^1Z!H)pSq`7e*^C|YQLCa9?F=w0!jcLt#<*Pe;`S&-i#pkf7 ziTTEtE=6Tr+cfn6S`(wC7Y*pL2TDKnU%snG2evw$i8PiWztt0(kndPOdF3?Y(Y}qq zxj%}geyf!qbn;zoplxz@DmS?qY(wv_1 z9jhZtRM9R(cI=cW4}j7C;w{zUkxbTh>`AUoM%a+NoU?knt_NMWCnyq%lHYy0yykZO zQj5lR1BbqG?wtdLPd*qGA=;S|f3r;kiC0L1S>Ss#>syGI-f=Pi}gO9 zS>I_zLb#7GY=n1*!M-#5Q0f(D3(2zuR$y`qAIqFi&E_Qd+w1PA4okh<%vSr%!oRiQ zHdgbsp;KVVPoB<9BI^TTVWecUYi(q1e$+Py(S50f#}y)PUS8 zj^#QMG|)5VL+N;p?&Ic8_8o|ulS!e8mbkStL8;kGsPYjW5KJUeUDm7Mrm~3bx!tfFRN>cIV;~Q9V2m(_&szW|koo04kDgf!A zKm68M{Hgp{{rKvQLPqfT2{-o~%Ha=S&FD!j@%p!JFB{()7%T4s;;`R%YC2l5FAKfb z-?BBIl((L2hkZXTNM{ixb)dUa8|k#z;QN`=_Hk4xLgpu@6|ftt8AFa|-Z1N=^J**> zft`=qFnf`~BsjJmU2`6RHnSF0(b{W<#6Qh2X_DUKm*$ziCb^qEs;J(@omv%AO_ndV zcwA7Z!l-31{NB>Y;I4Vj*UMG0b?OKfE=yk}GI-voNtDc~Hta4yBesNr!*G!W#PY)a zbtG;Rg8;z~$3j7EAUb>u&GJ@L-LimCJWA)nP&E`u3bGC-#~L2t1&autAFPJ7NlWSvADOYxwJ97RQQ$ki+lW(v_ zSUKGMx;F>A?@`))pF3XF10lFfzCk`azbOl}T}bIUdnQm#i(IsYZY2K*q`v`3S2~U) z=LrQ4P5=!EU_fW^q z$rrwc-N{@T1t&jOk5R}c6{h&eA@;aE)cDB4qv2QRwvNb($3{VotU|`Bmk~~O_B_#9 zX~bmrg%TYLV~*`mQ1Wp?<+LOe5U;OLsoKDW!cHWDZ;ISR_JGS>)VjA#mLlusAVo@W zF<>I(yJ#ebJdZ)sUMKfiE{^5X%Hk*C^U6nOPkp$0v^_X^;4r~#na;^E1D*Xq*#P}9<=x{!C zqD?>*bJrYMEb0I?*r>JbkIOiVG`0DfzoK%nv)tM+X&CwI_u65*ne-aW#+I z?mt*<8FoJYu=LP$IV8ps(^t{vd{>blGjaSQCs6xGXC_ew*9i^>P;s+z>h^DGIji8bmNNkJ(xv@-qbEI{h*I+dVLEl%R?u&5ND@V70bc z2M%>m-8Ab?)&HNV5Ii5hH~mnk2okXpHqI0H)Q>NeQ5Kua*nJWsXwE+3|p*aywW2T1VT^bl_bMe4@y2mjg!N9zb5nqba&R?=|T0e zIJ}&IR|Ko8vW2)ta-)0ANEopKR;dQRzBxKRmhZ1>dlz63Q4!iV#j)kC<1iT^ zQS0D;(DIThYXnNf{ESE|B8}AtBSpP^6jKE+?)hWxgXV6jh|eycDsdG@Yc1~H6aQ(u z!=5Rt5MFF?2!A z2Sv9JdB~DsG~6y-vsE5##;6W8F!QnHZ{tZf$ecyCO%fNQYXf29xEIu~_=VxP^xZNP zZ&sbfnPzYACRfi-UI7ezSCq*5&{;NQG%=6>cqqnMzIE!{3sMk}o*tMjVtZfp%MDo8 zf;gIo`Ou!;h2&hZKm0A zscOhV@eCHZXmnPyAtv7J5IGkTFQ|O^Bw|9ub?tnf)=&8>5HhplpigGMn}=fgmpZw4 zf@vXP%m=dr^v6<4g{UukG`DKAhv&S*6%2P|vbEB(Juo!(hkRRhCa{LV!b0s|0Iu)- zj<<^O4AP@2^6LCg+EiGma2Dzk;~0CiJcUChR?>_Y*`lGDj}f}He8^qAs>`Hc3==)c zfH7(Q&mRox&*SD6E_u>fWj9IgBG?nO(&QH1Ot{nZgQjkBR%}e>Qz4XmMmlyG1~AbE zr#jLQ?X{P2bNnccl&7dy$AT=Shkgot*ltWE*Jg_Kt zDMsHleJhuzPjPC`YvW%gqPg(Ux~nqCFfhY(jW#Qa{{l%jxsLdvx=nnmcP1|#jqlz6fN~;RY zg$VSy3&V*i>9~B92!tMf)TIhb7RN&PYXYrc039B_m=*BL&kFObxg)Q7#~L)(ox-g> zS1gbk4$ul4YFGtvOmSu_fsRb>si&fFL(t6Bi=g2}S{IWR?FeJ*nx4#q&fRen@BVvx~{|G<5gv4}G7#H+GB(ehju(Cy%98cJIZ4e>0pu?DPQ?6$BVsS?7$UUWAlWK}X*ku9Lw)smU`CWT$TnTg!IhY(VGcxdY7R z*2^Xc3x_5*0Tny?m6YkA<){Oq^mb1|{;byD^@r`ZT|z$k@pmes6)=}Fr1>WxF?mL> zD}N1PR(cekHtCCWxV`;!56IVt?cX=q_e0&6e7jr}?gDKStTBb}^%i(r@Q=kpGmgM2 zX^K(LPCUo1x{mA#tr-F27qR=ltDIGqijNVkPR}iI+cJ|(@)+y8BhirCSo1tHdb<^y zZ8-B_zFS@_?d>Og$pzycTX=Wh+%EM6Y+y)(FK%RZe3lJNKvoU->^gzfdcJs`rE<#;Dfo{|RdrEx)QaEGzjofs3 z>2KPEJdB**vWVos?}gmFx^rY0kbsVxkg-@?X5A)J%L90HY8F(w0&H@O({mYi8oF&^1;)2eP0vI8THJeapQzJ z>48%=^f?67U2iJ?GHX&mVVt314V%Y%XQt_9@qIhRcT^Xg>~wYIM@N#(9v5On{Q_gak=0`Xfd`epet%;6SMKxa zSb+OMf7h)s=&7OJUv=a0o#_%Rr2O~?5EGT8lzvWAoJ2e8Spd$40Fw{kz6E(Ui>Q6%iSk%|&IxiRx1?ObCN!`Pu(16d*V zXLhX)Wb|JcsI8kH&?Xzd4DjbdK;G6w+^#bocBp#sCsK_+5^Gtg|7bAz+z*qK% ze+2Qwb@`528}zsE;%ke(CHv;8!e zBlCfNd?ju^l5L-mm4#Cx;6FvZXx(iajDbaP**bgQN7n_3qx-Nj&78k$G`5n`IY-8U z`wj-AIQD{-c6v}2e8NwLg+N9MVt$WJOAg-?DOU%Ps-Z5&Beszw|nD+ zLOvfNh;u`Jvv9=V*E*{b3nsXvy-(29TzMTeJo#W&KBKra7=*kD3Sm&l;`@-shGdRt zVa+a@gL{KRpz2MmKz_FvHN)tGgcOy{6Pk?J-sLE$ogs^l`B+9SyU9Rn@P>XQ$<5Yo zN80`&%hN*UO2afc+|fNDjL{opgE*>pS1Dli&WHX*y{5G{T7UH;JPEB6+t;DangE11 zWuwpOTp^m$%3gl9Qfz+xi0-q)2FtEqV^$~k6twVd>~bySfn|?~euYE}5dU=^jWRZia;B z8I@BnV6aj;eA05{X8HB`C;#%v0pQ}>t}v@^_GCb*cCiV(oL?qkI;nDqo>>;PIgeQ# z%mHk$Y^cM5do=DYe|9lZi9^=8?KmAj)I;)(BMyLJX{ep3vQ`==@|$VU$&5Zd;6u~1 zh^AUepCy7Sv9l_A(6|KO6bvp18vcjH)XaJiTRA3;UrZmHNz9UOt8nSYR2h2X-5_$T zi7Fs|e;Jv7%wF2R^{Pkdcc##;1FumtudbkjJYSxfp^cwX@sO@72g~32hMslSli^hY zV}1+`fm)7iCbd8rh_uSZNSKZyA5ku$URL_5`*`3x3Syn@Le(*h4%7(&xU|N9|DY0`1+mXgU)V=Rj@?V3*3^wOARhs*V zV6<;l=`uSe;g$Exw=rsX zV8%9u?#zCI>bSf$_+k)Z&RlC&IKmtQ?W?c^qma}-5>#B|p~`(O0a-PJp`b{qn7dTD zL3=S-yWym}rbc8_ZPUhM8L`63I1#dk!%AI31AxIpgIrA5VeGPpifGo-V9J#1$4Rs( z4XI#hfW9^g*UG~q=M{oR8J|BOaSADd46DUO1lT#$(|jPPOVgW7c_(~E0NF;#1B*>> zZBt=f8vWLWb{RcK@8Hf5Jj)SJxDh&t7KvKYES!0g>~$OyFxS1c{{ZAJupdRfMl2K< zqUeu-y~Dd2VlEjUP>gvZiq02$)|#V;;ch8Em4BE1$_x&7qH<;2_&RS0g|k@<4`f+@ z3}7^-^OgSLo4i)ZM&-&XEKO>hMAlmm_*M%XRxrHrtA?tJhFRE*!7ffRI^2|g1djRl zOeJba5t1)E5YuuW^dt@_qddNhEN44-)-)080;4xsu))#qzP*Guc*PRd2gkyxqnU!x zKP}4DRwsloGU@qNEhOBJ*l%QAX-iH}Q&l6IKU7sN_fYnf1ewi2;}&7A-XDfCxXJo& zjHyu3oQM$9k^9BM2L7{6#wx07l$7t^7K=g_N8DIX*Io^h&7g%((!ZuY7q z93E!*{Oq&uw%wO^+^V2k4`CzYqxcl75fUf;m{iZ;)S^}mfBfbAROjd@V;-+vlOLL* z+s04rDSpa2|H3>j&N_mN7#QMW6q`g&i>d`BJX+bW36S2hhpe-d(4_HG0nd4_QjFN8ei7dBb?VVyY~PxN>>xo=50D9I zNrmDlaP1}!CL?JYL^ye(=x5`Zw8zuv`FmAoaTJ5-Qai14mZs|w<`%Xy4#V-`!9o`0 zgPdd7nGqF`N9DmVwjeGPNn;$xgIdKe=W6NGY(3(Gz>{%q_TeJhI0#^;%LZWrx!8AE z4`Tp(O@mB{;^A_<#xcNI*0WF6)r^^`weNyU`vI7x`~cw#ghE-W=MtpclmNvymP!{d z!(pbr$hPtD%?Xo75X>LgyvAWimnx@kg4vIfaWZDK?QT6^zK#U ziq%14Z**0vD3NZOz14XVFcNKOBg-^e4P7JyG4VOXvmsVbgk$#av#P8Qo`fNCi!_-6 zV!lfyQIuCSeQI)KIIYf85p4bq+Pjr>q`Gb*30pO70kB5rGP~J2g<iW0Uk^VJi?dfLpU0H8p!sOt}ZzY-GM*E^ev!nV+%2%3s&ZCG3y zZ8Ff8Hb42RoQ?8#q%PAVGOf5Aw49ef5Af6WtM$Lus{8Y^H$S+&CH0;6wGO7?!`^VM zljZn^wYyl)iYB113A%a`1(bX*X%O(Hr7{^Q3*I1I-OGY~4C`Fulj)RRqMC1!Sr|HelfXb}@$Yk(MC%V1Tu zMf@@zVJQ*fHxn*H(2uAnhR-l;4|_3Mgd8L&L_?-%V$Jlb;g=FMuop}~3l4b+sum3w zDP4R&1Fw9j4e`XaZO?pwgCls_$rBYq&t1MA?lx!NpVazH5RjUOUJ=v}_C{mK zX_vJRXvtp$RmPgLf+fx02l#>p0f;~zVuKVMK$unMJKDe6Y0vQ<_qNt`B`d5l9fe$y zgE%3SP9E6K!beNc0p)GS>k`!Mpg<&L-8&yQ8tV={)s90C#w?JQ6<^^SSGpsDaCH!% z^~F25i`_6+q2zKtoqKm%sPLl(jZoS05n=p>UbwmQ_)c$2Jg-CN>$iF~za+MYP? zyvz~at&jI@=TpTF1!r;WbjI~f|9xr8i3clt&wnMi9=B+=qD6v^aCmW~h7Q9n!)%?^ zC{VXYWy$uJ%b6Q`oJMV8XvfIrM!nPGw?(fvf0EX$7msQo8GaoCb{Gnzm`>sE#1GAd zz{q4Xs^DY|R9F@YpG_yVJZZ~u9H=du9W`yR+5)s2k@FZbDg&*07Ik!*tl-7|Apr%0 zi)-$lWOvqb>W~=JeF#r0p*oI3#y@YB6eXGccq)jqw>n@Wom%4k-?gv z`8u+a9gG-yzR(5A29A=WJAzu}`zCxj>J^`xeL6sn`IuaBnwUW9kk3ctU$b+_Q1aCB zeTs?F+BR=-dC}E6j;{O4-w{-}4D1ZqsgSHEY-SDOUO4NnLSx5NJEFzLeXQ>h?BO3jD?(%AVf(= zg$kQ@BsiDkI9gmwC*o6Y#CEZAchoNYjKS}8CI0K!f-s4pOD9FOixc@qQ2{aV}8dAc0nwiXQ&dUka-j8s}RgEcrr_&|M zZ_@N&xKmZneI!j?s(yJ!B$RM}Kfc5L`yaqS_h{eZi%SfBN9eNa;Okd7_PpPOsKstq zZ9i5urOy&oupu_$J{Be!y2kK7?c$8YNS+UYyH8&FJs5b0O#x4>OaUL9B`8S&FS-#W zPSv|({qvO(K^4DyLOv(21^XDxB+~^E4{(m?Qy zk@P#-iOWoTZOt*;2z|R!MJAu@$exZCUs2uvAgzR)@@N!1s5LPgIQNs#z_^)X4&Wdub7wdMgwkaZodjQ!dm=ViY{cVXQN!ZtiDi4cO{kM$R_JVl(&u@JpKIrewpI=Z1lr%NiVet?=EtYoZnaur zF{ysW76D9iR?!aR?9pVEU9_>$GgpPCDXFs};z4gMvUP%Dm!FQscsIy^J$t8e29 zTg=6w*!;Q&!ek*}B_!vHWgV%Ov*h%1uZSorpr>I)K<@wk$FmR!MJfm<`ADy|54AuP zH`~M=MCEmCY|Squ^<_Q{p8f=$zf!HruvzeH`#zce+)UaeQeH^Qo;^2pHJfLDH~?Mu z)hBdDXpzH^k^=YA z`QFJpGA=7FwpB$s)wo0gws{~Rw@J=2GQsf>W5iSh^GI7bg)KqVc0X4T-8m8~vFw$V z(>bZ|`tLK8)6P7603my4aQo(&tP?FKyF6yZqFFrZIwCnTqP>zodffimabA0*0tKp# zqbz!cK9<}qK2rwUvMOyO7 zP>!;4Gp-ekX@twHwT2fahf#n&ib2}9GyFZyj5Lzbfowv2v$Z)Z)r5T-_7W#qlnZ#1 zc7&_=A)V{BDzg;w5As6+>1AZYs+hjRY}3&vaYqEsc((E@_qBmDYhwn28X<=00n%FJ zsi9oR@p5VGKn{**$|E;btGvUc;VmeF>7X2Or8c5{e6U&4XM<$z%oodVK(&C3&2^SP z08R*$be4F{0?47Fg`!0dc#n)_>ywf}ZZq&CBC(!5ZCOptx9dRZh(It`BnIO#z0cWR z*qm{4jiTflp=kYvv|QJ>p@z|H$)Ypxkr6nJ8TSWJcCKrFRVz(NJEQehcP=7eMLTyq zJ(Vo>%{Dn}7`2U73uPLgCf63P%vK-v_yBS{Pu8XM&M#a@<>xy@7y)c%@1|#&d63zT z3N5K-x!wsRNya8$yZ^B4$n>;NUzmxom~vaf^&#fWnL*{=2OJotFteLlrGrZn%yxEA zpOly~qBHFF1oM^X`ehI@N|?X40Q<#;{`<=Ie${}^N1cyLL!XvIUUYaEd-^D|gm%8H z*HREmpt#5IZ6c+Q%=+H&Ee=(O#Nxy?l|_*4S^EmoQOO?f<;8GdI#oa-LkjAY9R|NL z(|O0SEci`|HIlx!RX`tttmQ)Yc9qTu(ibYlz{*Y_ha0vVH(_4YkE^W=hF$9O!hG>p z<5*rCEDCY-#%)$n%eT9XDcIe=dfkb36XQ;asPpz7(f{!yS$+m#^rDE~0mWr=9 z`Ba6UwCNCq?mH(2Pn;kF5pm)PO9Mk5mU0aAndqbO=&YmtNl9XzqXg@c%2pxPgwv8; z5Y;nfIclWHz#WSI)#h~j=!A z!JSbCh-1trQRzWdoy7o-m7vIF?FEPlm1OALvQ76%e6OY0x)SNZkTO=`Bn4F)kE%Kb z%QTDcZBT$3D#c3Xrhc7=B8%ytx=F3tCwsZ`0oVgY(yGJ$7m6+;M2kJA)z&H*~_atE8M9 zXD~4jeY;T#>$MZf6rX=DPawcnUC^4 z$E8U`(tRb5&Oj4GJTNyd@?*6E`wT?doLT@$^%S8DQnZ&zUQvaMCU2|a2Sbh=^aNJu zN&W_$ig6>eU)pUFxnK6zjY}2!&h~mwFGT!~TP8K}$vI8~r9{q=`$Q=!qYqZFuZBKC zMJt|TD&0-y&`QfRguEvhDwsAv&UVZ>1U0Wg|Xiv`Cwj%tv-YZ&gzkzki01}z3JVU;tXFU^LL=&jmF zCisjUv#IamXJ}R@;J0w55+s5EM>x<_%}vyKc@p&gYiNeg;~_0>9pc*7XZtai$@Aw@w1bi#G`@X zEeESWB=sLO!d)~<=4FK52-a-1zAEK#g7#sq4skN*fs+ip8hu z9p;_Af!zHz&HJSCrOH*wm=kPpGr+uz!0mKM*C>k>c(8a+J!P$sK{C)^IlM!6hAJEz zaJ7uU1bqT*pUWTLe3T(pz252KnlWJM*?I=Rr|@we+@C*GLi0gSo)nk1v zgiieZ@m%knDvrsj{c`bF1v~K{K%FV})Z4|vz3RhL{dSLl_T@}l`vuw=dM^{FT@_<+ z{tgSVn&{;F1Nf;Iat#xg+dSu(6DKb$N@2TTrqHWZ!?gT50ww|XUgR@{nQ&2)%)pM3 z=!vhdMge|>F6@|1+&1F_D>mULOkMc2;x@caABZGm?U3XfIWkCc zJak@lHj4RTlT_%=MJzcGjN#W&jT0AHWS4!f^exY_s-DWVVAbw@DasRcqPFN^LcyHr zK4c3HQ_d=w5y-l#&!bp+)-=qRn>C|Py~J@hmXU~99$!-whp}=}PcNP96QM%AzOT=enS4q~Y!`+~rJ0IZG}{Im3>l$ zGjE#NV8XUFi>|d_`%dSmmY*`nJpHJebaJ~(cbeu_=|bO#;>$zJ%WS3w5t6%=@cglw5SveN! zOCvlCyy%Uc7DJ_ne76{>BP&oI{s4F7dMwv>y$>Uy{N9kOrJPyE@=f2v(-Qd@iPy+O zQ8X~lvo1>1w-VX)KVCY>7C}3Va<_A(qA<8x%|B52osxcRTs|AO6f{AG;{@x59jAu7 zI;xv>u{-~YZZheH}Bj1(T57LbHQ#wq*L&_oEp^o6Pxw=>1 znBzg-t5?RQEOW=EJfL+bW0YZgz{cE)g1uBhi? z9E&SF1z^wfwGFGX>-r1T$KQ~Y+RcqgRH1b9k6geKAcr)8=BpL607}wO&CkozjmW6j zvwIQ-g&*mY)&7a)A2hROd=zrcycqnncu$moL z>OX*!{0Y9AR7WJCcgCQmR4TnmDYgWLOc!k@Gib+8xK#pFwaMF!#68?COe*!4iQwV~i8ea%63+r3eqAX#2I6odr4T-Wp zIh4it)2Za{S4>O-;2ELoKl@>?CWCvh?vS16X3>O5S>hTP866D#Xi}CJ+{|#vnbL`v zk`KQ^{8rGKLUOR$O6yE~B)Zn@{jYK$ppp}Lgkv1;=S)LLm{lg*n#WJg!-NRrSSN+n zs{a7Q>wlF7-<=im3>p58X{`gc;@i*mQue?mOYGt3ua#ZuGX0jo!YR9y{jm@P78Pmo ziL)7`wkhKo!Oc=~;2TeKL7|lW1F$-LL--CP4y011E?F9KY5_A8fz3H)XL=h8I0DQ^ zF^mC)QQ{*%EBF#{YKb|+%M%i*J`kf5)`7k8xd)WsSZC+Kl8DArdkAW3^HYy9Oa}Ce zT(bnAyNLqdj?FZ>%2~~Lek*(o=~Z`bVptQp&~5EAOl;UP$|gmRbORifCc&)PZ`ZNh zeN2gu`|}M4CJB({FaT@LX_QhJrj4+O=#f$7&3p;@bSOD1cp5RtEH~-YX=8@v?Cu^? zNS>ad%N0i$kC`clMtscE{-(-3P;VlL0b8yeUT8YS_h8%}801>he%Ue!6YZ^Ri3d3o zOU@y9GoP2oE`dOzNFXWqi#pSC9HQ+7>yxY}o|26My8hCcTql%xT%LWbGvz z_P7N~#)(rKCpq?XD%qM1Pgc^~n+8YM_$>)SlIq4t`$%ewIfF9U=~3?VMm0&A)ac&P ze4FqPy**0LI5q}nCKk-jZ?Y@=UB7g~i)>wr*;kR5beHX}BfRriPOz7MMdWxNxR)AD zWb83p!SvQi(BrnVariDI{@bZN^_-O)M#W_^OAiM%o==9^C$B<`lSi@(1g$J%w^te-}r;k$qorI>R~nvpY}MGfLSm zo+_N>Iyc60Rrrx7i*2a(v&w7;unh1R^=|5sxr&Dp2EtenFXq2!a&i0tL`pY`+XOZj zo@PkobMEbI<52R)eQYTS&EWV3b})kgCuq9OVz z#1U*_`4A8Z^;3c!b^sZVl79oc*t1(ax)8(1Q{841=AB4o;u(LjC z^*Nw1QWxXX-nsRAJ|kjCX_TuZ(HS`Fl5Z!w%Ms*NZc3zce4j zH)V}_(!u!&lenHEYLw$>MW9K#mdsYc6cjhDgu(cA=!IPQMQgy*m6R0J60$4B*&IF7 zYXg8xq=EXdx`y5)`kHI7xMlnDc%dAg6i9u5UuXazWH{z)PI~bENCsk_XJ>~c{=s3L zGP9U=^9a;T2$!~ULUTF+I@np!%&<7GTFa{5TlDWCOrYpXRwJpjk*5GMV0t5$wACkP zJ3aZRU+a{Gd@aF@0Z3_*LJj(iF<;>z5a;_X2mUYxq)~Vi0Wwd<3n$H=8F>0TSiQCD zxh(6PI>u`M1z^FVtaH|%?!|I?GZ}9Bma<|RnV~Xi0m*E+jiV%1`pi+Hk zvN1i0d)Q=OF?GnUdb={)J$CAEr3I@)L#b}%&Umow9r~1Hl2s5;J*1e>W1?9pI{1vH z=;cT6p*KZ`emqiUbf<&z>mKxwqo3>nLkNWTn2|r!)-Qg1gE4B$%~E%0RoIAI@VyU&n~<3MnL7#M@}cR%X?{IJ97$K0@EU7EY=F@M!nlhWl)WLK6_-kYBi8Am33 zP$h>VP8>{U)T~x7PwUrHgpfZhJeusSk(iS^<}Ae&H0i{WIF`#j*5kZ?XHt?Tpkt^( z&{O44tIKZmuwRz{8N*rzwbKH9KwBAN5Kn0*O{!c=5>LUb}461&x82#pO5?*15M6wkLE~t#IhmR_OIFvt!nQjg}%Ja(qwCF&=rPk z|M|L@%p4MhMh)KWvO{F*jQ{y#nSiK_wzrJiwpa>_Wg7=QP@ej4N+SAM!P6RZRU7U? zP$$hgmsH>&Ih5F|eZm0YsnVa|Ci#7Tx=<+W|9}0GzRFJme0XP5)|VQbay!B1{|pKB zG>WJWj57#+BCYd#yxjOK024(au=Rvw;iEmtTsXrISR@=QbVr>3R^UNsmW?190@k}z z@SdXV{{t`%Y#7+sJ%7xbFzOmlq{^-MTfEk9j}Xp{ip8B8*@tYj@GFfq6^Sm{y)_l29~gAG`AYV zmrwtzdJghI&~&nGf}8!{Z~a>~fjj#{aAOex_A;*8u&>U)e$LCAGGqk_12}Fso!KWl zh^*W0wBcuiEEr6Etn%k802vAb011Es008Ls3xfY!#|-v;;@^^oDuVZO;1kaOw$|a^ z!+%@mzb$b0KZtt|xF(vQVLU+Sq4y%v#Lxo-2@s?sy(@|e3WyYy5*0*{F1c6WAW=k_iaErgT`k~jwD+ALT} z4KwM0%>3sKW*YOK{qObYcR4+K+e^=X2d|muh=ugY%iQI?S5zRNCJ+S%0{a_C6S@#p z{0+3BzV*Ar{}MOyd5c9ZPJ9FXcRx##fXu4yzqaCgVT&~$tMG&%un7kA=zlN{ibY#^ zERzXCJ)t|nPuB0GtkZ5UO>9mO`k}OqMUwjix0j3%gNNmEhreR2s<>1?|%tOUy{6WokJR<$4gWqDuN8xp0&34K8;Xg-RHdq!w^kbZOD8L zMLuyf^Em4Kn*8M7fp~7bktIfdi2K><^(Ex;;lEe?pkY2xy$8ZT2$p*4H5owF_@A?UuU!jE!ed3v z{@%iUQ#T<=I+24fn45zU_!z`YB>N2AZ<5=T397e3ZPLGifbsTyzBL&j`5$!TM{AMz zvU*@1HGI8-?;Og;A_v=$xP2D%uV&oAV~OxsK~+t$iZ-NFCw!TLUg|El%r64uiV;Ec z3Mt^Ifa-kU&sJt8K}9du+26>1DFq5&Aj_73#@`;ZM&e}IQQ`D_THF|FWIqikwilUA zh-66o+elcG_a9vUk_K3ME$tiVSB%)4W&t>GIS;%3MGKU#V~=#g_fVavcC3O0HNnmm z?U1TAWEF-R-vx2)LdxuOtrx?Zc%aL@$Q45bzK40nkRD?M8#F);_QEk(=zfcszxV-4 zzkv?^6ZDs@A@SXK5TfN9=r5{S@ST3Al)Xpr_1kZgs;~lPF-QzR*1g}+1Nt6h)qzW6 z#ro*zvoK-|k?b%!dI1e~e=KsU1rBT4CvH+x+xM)c#@!zsuUp?me4tn7cAXsjm+F;b z_(0|{^c*mdN_g;feD9Dzrx>pa1tZ?vx>=C@Y;CUBWP zKyly1Pk`O~5`=Xjm)pgv zFvyh?V&xca)V@cf{NQ*xL~|r;AQJZPPJnrze>Hl(<@>|V;};K4gDCg{A+X_pDHfqY z%;B-e3FYH#7EN%#$luqd7u-1VA2=`<2+qvCa+eo)3JTrH-U2k>x4!PTY$j5|4H2$2&O{ffd~tT?s>KBCt%!;2q5FGQU$NX!2319gPjKo z_Feg(nh0D3046>FO={LlkG_FE{^;&f0hr#Df;S*i3Imbb`G?|_-MezwgNdtDTjPy? z2}D^jHw+Ti4spjqDzK1BU~&_rdto4yQ-<6qTEbMGF<_qQL11X%(=mWb?rbbrrh|Ez zKII!|A*pfh(6-U@?To$cNezbB01Dn|8dnUhe7w`AKBY19A6kVAiXR<)Jne*Vj9Id z8KXl)8jIojq3)+l!ur*|uWS4w!Pjnl1N~hyW(kQ0;HQAD-|&ON-X=<)=M%~$BYyO= z_aUYk=j;xdnSkPIOx@{W7$6V$#rp%V`AwPF3`gN^n7QY?x3(SmUFi>l{?yfV=%=Xv0n!1;fe!uhWF3Z6hQq0lb0GP zBDE6Gi$q;H0xEIJlz%fD~_@YE0 z>LFCsz?c?D92Gq-2!^8M{l3Iug}=ri@h_41He|Lc!ZiT79fmc;Qo2ZgFfMtuukRpz z`HSu+@$RDewVb}lcyd63((;csNn@D-Zw7d$c)o$~Sng$v*a}t*h5>>wcSSp)jRffQ z$(c4Xlc(`#3KF2?|NrW-f`G#@$Gg;X~}#z%sU+hBn;KpKPR&hCIP0tO9*_!}Xo z1KC?z1kLvO^C!gLVwnJ2ABg)}=CeNGv!1^5Lpt)Osc}Zrzb`5)W7eh^CBQ|i6R-%5 z1~Gskpa&^~;YD>pY=O1_qi#f)xbEr@X+TpK(!C3b0#^0bgT9NL#qt8zLV&3UA`yd8 zqJvAYF2|I;#)CfXvlH3}!*TV019=fVaACi__;Av}f?GGo8{ejOj`rO{5kw=Zu!w|S z7)Bqt5(V&p{i)f3aP3Zr(?~+xZU>IsxMm2f2?PsI#dpf>o0@jK~;-8zYzLf@0J z1L45&AO^SBLvBT_W0eN4HUO#t=w@aur@#FkWyg1j_n9sn=EeK zyK0{jXwnKDfd#j}4%irp0Yko@8`Y2eurzL2X?=ocef$DIb^r(V#YLsx21;H&Kh#N> zu0S0xn0y1wssyIrETE6lu}Je77xY<(`36ec*{|ltMJBTjv)D<^)n&B z9QdxN90JTmvC#)&D@ODr(QsQV91{Y=(ZWe|5f~#RCK@#OW10O1k~#Hujk9zLJYZ1SdlQcm$+Hc?W34FX&|r2}7D;5KS?N ziZ1#Ky}Vfw?C#C5Y&vL58~1^cpFVb_yX^gog}_d|kW@OzwYt&)Oh70g8Pf_}D+7TL zF=51u4?b=gLr}qlMNR?}ycf9qsH~PeP^tu!w~GTfk_kiYLkVJ@& zqyfZ#0yojy*JeG46=2S`Bk_8$87wodU9bvEHw2VujO~vuVh{^A!@{e2j9wdzDeg=7 zOfQG;vLOfJP|Ih5?-o9u0R)aEg5JwZ2U-p+O!ue8gDNEo4)3W5z);ZaUWn}pWWskL zlmIcS><)pYcYkJKkE7zwUwnNKG^oMm^>`Ik86aB`01+ZAK-RUEQuf%n(004wJ>Xu1 zk{{+J;QxT$Fw#coG!TuOm{)p%Sxkudlj;yyOE?e#>&o`&d!X6j+%lcOew8;+J2C~4?djp#tzX|;v|FL@RH{usfr=yXihfBlqh|Epjm&mDJ=*j}k zUoJFo`=YuiFg=M=Gv|E)Aw3W7O>o)vddDrX-Q$4`>_^XEV*jD#U#GYT>+ROJCu3=x ziPxrR#HJ&V7QMhX#>F0fSwburr7rC%lN;>lJ=ysoktoxS)`KAB`1oU<+ci8KimAX4 zecd6~=KeuJXkf__;rO4!r>5`8D18e86$8uCs$ zW&J}%@3?!3-2F!AsWK-Z%l(I_%*h@2MdX%7!llULJ-%mTw z|0su6h0A81uQlG7RgOVciawS3_-f;;p`r&82%$e{z9V&4>jQh4@rUd2SXmShU4J8x z*~ce-ZV4UiNd0~U;{1M`_|DZO=B5GUaag+Qt{Xsx(1=yWE$!4yh;RRElmUjip2fGo zO__&)5+K9+O)cQsgM?ENm!B5HBDen_KrTNausr!$2&BrjAW#bkL;|E!5D*9qY68PR zAR?O8KP3ZV;g!*Q)G~xijDKx3?~i(?M1wC5cz18@wwTm6(6!=qY|Za7$_P`n>w7dk zsKCWA6P6|eh)nZW;Nmd54O*uBj~)Sf36MbH$F?6;5C3fW0n&K~sKvz$9c!7mX>1-; zN988p=`Z=6YoAvX7*`}t`S&8^Oq;@G>@hPy#L0SmuLE8esfmZk{6jP$^g@H^Z&vki z1NS-hW&I|k%Nr=k8o6@4t9}W%3Ar)?taf6b{&2^GBoFn6SSVgC?tLRO9t%+g%Da%2 zF~=2skt=$F?pUY{V2oZkuoDAFTE=oo9o``uUC$MSguBAu;oOqj%ZF#beBeAh)WA zkTW=21Kr=h;=ur!En(9$JLbGiD*j`yh6fh;D^>rgy=!(c2AQQ!;r^0hMGcl6$eb1P zyI#@IS?_~(#=jz@K+9*nXS$^W`g7610}gH{0Mg=>PGmU-;xP;0+y|ulNA4o4&VumH z*;mSV?EX!;pDKSKz-~Xr2v}PIM&uxHG$J71;YSXSEW6UUY73M?Z6guo*z%bh-n?x< z7oqN7KYz~}XuJWF0{$fv16UZK6y8Fphkdj@7gLXKum2_Sr)Vs6B^n6}ga!0+6H!6q zRAHvk>LK2LiQSkU=a2xXRC60wnwkfZVD>a8e8D9u4CQL@?EHZw^=X4 zU$pK|2B3XF)xi-!P=r&&U<`2FcVM8X3Me9tHDE&Gl0X7j<|?2r23`(a`(tTlx?qDq zE#cx*dGpJ&III|6$NhobDuIiF9Rp#;uh#?kzacwXj6nxTi^Zlj5qLlLW?kkLT`2II z9+gBOf8gTxJ40{h;q&PL&eFt-!orvLgEyWIm=Lt>F24zFKL^lMj{jDImHkKf{m}v_ zzEd!O77U{lb2Eah1c88kd($q#N;0H7xuE-Z`-;707+2g6DDvi1Qov@!AeMWOTnq*5 zpkR?6lE16@C*Eee^F!00Rb9U}9{y+8#~+c+81PYo>`qrTus49%8UXq0frlJlX@oCe zpUy1RZ!hg`e?Gq5aAzC){Ov^T%+gaFhYRX}@n1*W$NhUo|FiZ0KM(x9hh6mt2oCG> z9ce$9J%ncBx_obUHR7@7fGs%O#Qq&04=fF;3vl?wAEr_M)(R41AOaDQ5EE085D^jW z=X*p9AU-M8izaA#UKKb0xQhPe6_|1J)CE`ISqKUufJ~YG8>lXvDDWGID=ZtGcU@R{ zijTuPNB4w<1BL!@+1E1{s#BKHut}e{uN!VM#c*mmFu6< z1n8FdHAU5ydbhS3*D`AjXU0ogQHs5}jrTM7q(8z+0o}EU#w59{V!B-jRxvnV86W4demK$Vp<13uYuqWJWIEFtbv^s5!ln zS^)#GXN8$Qt+*QT6|D6Kj}eJgh19ZvVtpd3=@@E1eK+lo{tmfqd>6n1tUVuA=e<~# zGZ}SsK)Jk?eUzrg{WJ-_BH{OpIKg&)$rl3JsT>9ZOOgUuPh`a(2?4DJpcxR6gTPzn zj&`Op8VpzpDB+zjmsT1*IEP*C4U3Z2Mc$v9sG;J!U~DxV4z_}N^#_uhBvjsS8}>i5 zAUv-z^b|_|d2<4-I`IrGL&v3|-|PxJnLkvT%unX!s!%cqw1Nin9FDBXFcSbGX1_5n9FB&q*}g3iTL!p)s)GS?sr4-B2+R64axI7zPb z4_$vO!+&X5*Sk}X8ZwRbwUsdxC@$C26KnJnXYn7xta&f>Wxi+Tb*Qwu?lxm2kI*2V z?GoP+jrVOZ^!;@1;e0`~I7Dei;|2W6?a89Im&DSZ1vJ^@+VOIfhn}0t)-ruv&e#oN zTq&TYF9=P6Wd$}l-V`<-J2gA*&IRP$4phmUtbP;5jk^DkXj3rjOcNGhy;%h-%Dc= zj!_Mb6bO{cpbEy>pMePMY3h+)G>exade1fV@VzH*3{@PlTCG`NT=b`0bWZcT59jS= zmkTYM6O$tK9;f{@o?!&+nzrx^Mc|lLZP?ijOWMi&``lWjS-o3x%ehU((hskIc+Av^ z?G?Z;h`n7J+$Kbsj-xCBY;TI4$UW^5P|t1a82*?f*yJdAObPAf>ypni>Mo%ys2s>WIWu%XSIDU4vSIZDbtj*$W#1p>QW^HMEejzyn1$Nu#D!*YRaYUH z&E^S__ZdrO8Cw=YThDyEmrBpi-H^A=|9q4*Iy2RzO+{*ZL~&J^o>(#eN(wrizJOR% z;AFq$-&httlpBauok7cIra3G$0*`r`e@Q6n-KDz7McT@$Q}=MbjqG(~&@81Nw}HFS zdLedKG{NV5m$=4JIj3j*l*BqgK%I(Tc2Ll)%#<;fS@0sm zUH5gHx})1eANP_vo|jJG6&F&x4jEc}+IrJwThVYSJ<;&>^@uAEeXH%xc%~yAZDNPI z1($QD!CuT&3-VNfpwTgiVArf^UP#P#=RIwjgrOYTl`ZDxX-ztGf0Rum6`!ftYo`10 z*ZKddr1&1E{n56PBW}u>xI&eoj-VxWBPR4`2?d_4HnCO29J3ud{`Q%{9$%`j=~3?C zUe;<27MuNSIHrdD8)HRB)>v_D-WE?QADtwldGg+)&&IDWi!a!V*?pdYUqwKnjrvWj ze9vEpCnd;*U61yCo$8<>XD?X>Y3<$HOwmQn1B21`(+b!5#7DO zyv_qXGOmudbIfU77)W@z*R*C9mTzte%ErJj<&pRYz9FT{H0s}R*EMTJXZq{4NO#C{3Dx0`cY-z4QrS3D>e{2AkfR*tZXe%U z0AJYg0`|-X>f~O%-Mj4LzNB8*xvOmLlVy~6+!cR4BSiuou_tKcL~9f6T=2=2c5QG{ zZ%X9_C=gQuOfD7e3nnezcJB&8ZEMLOG9E2|wR%^>F;>0iXxrD^Ur7!}=6v1~;J=bV zc^z&ExnR#P8&zwT+)?*s=*^{-WAZoNE%l+5>RyH3t771h@v7%I!#Z_oT&GlL?<;Xw zJ0TMOKElP%22L*+8B(a&^+7ls{&eWqAGRy$cenev%^0NvZZ+>jp<-Gf$G%)Sgd| zs~;f+kIlhs(uDa}Q}Ws_mN?7LnX$+iN@@(5I=L|0mpffDu+bD((Bk+qLMx3oI@f|Z zvCdQT6bJlR0E6wb-Tc`u(Ubyj-R4>S^B;~hY-=y@#M%hcu58k>rsV zE1RaN=$TlO$Iz{fAEMK%isM@><(ovR3e3p8n{%kq{D78HbtteA9zdgX3bm^fR9Ps) zJsn)&vQ8$2^{WOS@;K1`&6KGGR|ZIvs?NfPS*p&;3b@~IdF*T-uWJPybMu@mlgn7d_0$HOzIw%) zre{*K*y|?L$e(S1VqUL)1BEu+RPr5f5+T)J4NiJgp~hb^nz{#3U`TR+o>f8z=$2^( zF}Z9+uO__o!rCGVZGG5o=h~oO{1>10WuvL{Dp~xed_VffxACPuTxlG-2`@12Kf$IJ z)F1!wWX;+CA?Qo0#(NoG=_}>-#lq~3s8jqbL6Ae?{8at}!e$CE(<;(Sbh;HvBqFBG zN+&}XXl63!$Gs9l)Y|p#R@?yd=ZEZhIFv7`)l(pEv+F&rD`*B@_S1k zygAc%g)L!8^XA$SgU`(1(&lw@Vdi4+V&)cfiLQo_)51C6l_cnA%nVu(+3WBotE9<% z8k_lQL3F(DhkmuI+^3MTH#zw@BI{2rOn1FbW#*m#kWDHeZ7$)+5$|6=Or+hNNKs^A z6=wqFNY3>v=11E&am0r*sIUDl^U4Fkc% zd`s#sL(aBWI{Zo5dz_$VkvY?sfhUE6R*o%%EpFILg9z-7Y6UY;&MmV#bellBdq(G$SlydrWmNm`eFZ16p zZ}Vr^nSBoTi8HDc{;kM_Sr8%j)kiotF0NxM*`BN{p6-~p!kNNy4>>xk_)EFG(z$C$ z>g&+QDvM&uh&}l?WF}et%Yw_<+Q!DE*P@<4)=W2DO*XlZgP`D0KD%}4$Q_pYm1 zKZ_!!60^h7?mn8gYkl$V_M;Z@d?CcTi&mC8EO{y4n4$Rrx%l(-9c5E{-8XWVJLxDt zsd=j0rr?p-Gv{SFs-kwaG^OI1q3r6Fr7{)7b0Lh>OiXL-QI1UKa?};`>ZGT2h8B%+ zS8l2wFJ>-`R8I16h2df=pr(elztst&HN-llYnz$S@f<{aKQ$RVwbd zdy@L~v2~G?3mP=7#z%D6o{zfLXw)XKU$B+JsE)tMDQC{qu!dM}K7m+ZmUgwBFA}w# z{%AT%gGqX3r)Sfzsp#FyL~(nND5Mk~vhJ1gP~joBRo=w)Yy}EqqpPXO{6 zh3}8Ep)W=Ccx<*`!-la@_c5ONd`xraqtW`a>}&qZGbS4My-K^YF`ICf_MmgaXO8*w zw~{CD_BpwvZ;@WweJQ5@SY3U$WSZ%@cyDfG^X<3KtFK>Dd`9Z3rRSEB zqg`wi1ZAl{L1k-+fW%apJ>ObCLUsjZcdGJy$mmx0xH0P@QKdSK9D7GP7SZ>)M$2M@ zlQg z!Pv54r`W7}c+2wm&E?M!EvAa6h0b&(e!}q4m&Q3d>qq~~tS&R64YBGprCQgfQv`mgo zI;&@qcX*Z--RZkLFc$qVNtmv`fU{h|b@juYbFch$82W+?bd0eom6jaEiqd>x*GICJ zPxzUtPWg(+ztGwsvDiE%QcH@SeE6#7Ay+iAvD%;h$fO#Uq0!$o%Yj$5=^Ka@ZPVVd zG@pHR+3&*BTHy%QN0^qbPx^~}F3IcD&usjZQn;1#C3V5{%wEMJ{FDL;8_coyFpds} zxsS|lY4noVCn0%*vQ*C18<}{^=AI1iCFY+(UeChQ#V$KkNe~_N#oa~7aj>URw5=@Y zb%4~YWJC8}VY3D7mXvFY9_y~eE#f3w@~O?smtEN4RR*MdQ=|G@dJ~2AoY5lx_aCMd z6Tz%qj*Jx93&wT&LsBzm7|!r!X45B$V5-=o(eN9-hYA_yqnH9+bXyjYTMcOzFkd$<&3rb4yoHj*Pmi(rI_!2E`;Mg`X&a zUSU7(+}A#;?v8Er;mwA>r_75@noG#NQd2o(c~ebsjG7e8u+>W&%7n{_Q>!U37hz5g zFc)E=diGpVO1bdbuKcT(rZm;|%Vu8T{t!0F!(R+EkGSM~8XK;?9YJ-g6_sgvT0QY{ z<3Ol9qy+>Ag3&NZY+oq+RG$ME{Cri(mE#3!qHMBasEP`NmflvmVF;pSwo@_**I z^cJsNZd@ zV`26B;SY|d88#tX+#w4$=DK%LouqothHuc`e@K=xe4fgzmIxCgviLG4)pJRYN{{7> z@{7BDq2$aMbzbe@VW{1FHld(#=&P2~k`}@sBkBC(wX=R!p*LXY( zFdjG)ww$|8dvc_G<7_!8X)0H`f$N>wI@^Kl(zkTIAL3L(mhZ1RZ;hPDmif9AY%6Bx zi1ZS;VyNV}UIAV<-f*Ops=M@HtbF;Y-!;0UjYVd7BoS<5rZ?WCR|A>x`l5F|p=A=Ny9iM7-8>b^4%R7Gx% z>@MSWKpk_kyVnkZOnt3w%_$kxx>KFCQrP*AXgrnlGleIK^W7EC&X3}SAL{s==t0)D zu~%mzal5U1<+IvhsB~?{09p)jo!ej|jjZj#da)7|qdk#*HE_H!;SdQt_L8e|$>pMV z+%y{vsV4Ce)Sk8L+nWP$K82Q-bC_zy0yB8tWYzf1b5z~Cw#P`T$Bs4jvT((RrAoPw z59Ac_sxY=(@$O+}2(R!yemzr64W_hUu%{BLp~8_ia;V>kiqx8|a&!o@|RB14PZ{zNMs z6`McPo-ruFz<0#ShKy1X>x-dcXg{ScOI{Rgf9Ddh<>;;9!QRK|H>&X^`Ds^?3yvSp z8Yz4)yv`f<`Lqdx)%0A~ht<;#8ZxaT-4CE+K}+0oHP1AT)VWDlYegyx9hdO7Os(KW z*DPwJpo1vT;M|rz88V5Zf!CaF6f6eDO_Vt^+hx%^r|){>U&-V%U-)dFgB+C=%gn(c zY=`fEq5_${cm3WSP@DTtgA6bFBm`IvVM6F4S7xWZVD zzCo?aPOS71sf<|U)z^XEGE)&Z+Y}h(DE$u=0^vsJ%q=Gp*C`pLC+RVCIIkR!!ff)3 z$pd)L^r1t7jx3p2lqqs>Ou;u%OIKhNcn$ms+SMZ^Q=wBEQCqH`FgX*CRF&a|^c$SV z4A1b0sz_JmSeph=*7_xZLuFGAZBQ}Lfk-8!B`d@$jACOU&*J?YJ&vL4<-!ERnkAX4 z2P?}>SrJK;*M)B88ZgHi1Q2To&Q8hR;OBpHR40t#p{->stsi|UEbF?zEIi7So?gaH z(7~!Zj*$=gn7&xLvM=|N&Y6xYQi2Zg^(Mw)jr*FWDv7Ld!XG={z-|}47ZhFY1^i|qFp&5j~o36P63#HfBrx=%j*PFsw zSfm4FZod&3a1)HSvWFcv&pRp+E2e2fEZ{sE;NVZDLGd&&Vxsq9EAx9B9r4>bRkEGg#m|h?8I5)sIGPHa9&13DvbxWjp zjnAPM!UTrJ6YI zd1@2i^6yIB^G0%`zZJ8^qt^aQUIAidVFz@7&bv;w@#0Qy_j-zw-E<*c@#SUrQ`c4H zxCdX@^CwJ7<&xTw*x9OE6pRLpF+t1?9cmu9IEB`k3HPm*JPDk-#&W|@wzGi_Q%J&G{pd-N*MA4HhSN8$9qyn!0y}YdmL&7E zPz+hwIN!Ma8Ut+GDRtbU+(CSy6d zh?K@@HLQBvkl}%6?hu_C8twE#C3*4ki{My)d+j)3{~)-Bm zY(f>^;PPgl@VOzi1)s{@uQ^6s&JMoxP|X&;#TFjNXpqN&7pC3Wo<%0(kD6bA+V%E`1>MG?$uRazuf z+h&t23rb#s^iX*wgXZwqAsx^0x|HGEi&Dk|Wi@hSBDY`8Xsdi;?crBTd=Gr5Eo@3V zP_9tZh^j%I-U`tdTQEm>Ei7YYe9ILbOcclsY_oR9QR~%(1mC9*I$YWJXxCDR=%TBu zgTH~mi3&9yN>m~a{0XLLT`ztqUHhCu1&*`??qF*w7*4YC%*U;r0Mpvz*zH{ITbBBZ zO-hPIOQXBn){V~VU+(OdP1J0!HQpKdUIe8Ss%gxsa3&pDL97GYkDxZ94`P+}2@8KKt6}VTkkict%u3#106=E-nb>}}$!pY_EC8y?rue!*M#oiBtWgz>&uiL1kO+Rl@5wZ3_xorGNPIM0(xl|Zt7ehra{=oQg8wI~;9gI;d;>R*o`o|WulalLo z;%#&6#{>{_S2ed-Wrl!G8cGd@t;uBUP>=d6)9QMa@?{o3vH*Uve`^g-Pc$VJ!U7J+V_$Nj3~?JFBf zcQg)B(P@cO4p+UdwTT%! zeZ?!vu1C9}x)*t#C2pDBKbby~$SK>6FSNJYTwy_ob2ua&2A5(`DTe9p^wB18mgaXJ`A(Mg}r$ZW)U4xw|p@17Y`g?zmk z0G>FQ=65S0Kce|P$(12Odl|3_p(l*3S`(=SzYq};?@a|cg>ZDs@3 z!W^6fz-apmC`yGMEr9|C&w)?emjb9z#$w`f9L!pNXRd&fsrg7-npMU}q3xl0yCDx& zMJH<&YM(PrG|8Ri8oq`fXXqHE8TG{2 zXCalN{ziaI79fsf_+nK0do@lA{#WCbrgLp{zCn0k3@Sj(4;MIEu6mH4B3I7%1@1 zhl)jc(8koY6^Y}4bJ=Wmb{0+Z=c#DpONE%q@m9g!Y~-if2PZZ%JlO0T@CHFl{qCry zuQ?|uhlLRw%Mx3r(jx^I7R1d-4rkeJa2~zPWvF-U)A&q?+wko1L{+}bh(n~RBu@Q0 zGbJkgwwbSEnP2%as8G;`3V9Jr4LAF1$+$z`YjjkxckpnTsq#gv@i_OD1x(I!gA1nF z6>H8DIewfKZo9PD2M2u~#+<(V(T9ou^6Y8?lRDyE@7NBMj8*aR!?~w}R*Xomkr$6W zKc!D_MAE}4~%)wZ9!CI z9L~Y*JXwul``nXYt-6Sf)1aIbDz+T!lL_<|u(i{;G&NECx?WwyAP$sD-RhXzm{Kwm zg5kMT&r<;0D3kbN$VIkeeOP<`{vLP;o&M^`j^qmD#+fwqMp7I`)#12r&HJ1qGz~no zJzKYo@;{}Hpsh)pbzbs99<}>9Z9N9=p+&5d(~1%mnGiwb#;tN?e z(lECriO}+o3sf2Ktk0Km_ZO>DY~l2zH;=}2BdxxHA~b;?dv85I9yo3y#Ih^!|k_1fQ zWI=9y?DJ5*(Xxc6hYyi8i)vl3AtrreV@=NLXKHzGe7>DV%q1>%u2Dv zOIKSWoat2ZI&8M3=o;vm;8A1Q@RtE)lt-i_-*y#H`6b73A3bY`mb%NMw7q;|B#Niy zOkmK*#H36mE9H+QZEsh?_`~Vg@51Jk-8RPBx-K zZe>VkI-Oyvj@o{7{{2_5*4?32#v!|5)%*rE^_Jesb(?2>w$F-k9f!*CG*)l=Lhwfn zGaiN(>i5NbJl$hN72E^1#ciGk0`i^vJ;0+$=ME?BF^J8bGBk9f%<(Ou(C(NMQ9q9L z%Hg%;8_^upt|6^F?aw;?dhye6bZ5M}uIbL4e8J>Uy$EMtjFlSNbh+SM4bkRm60xGk z!E6D-q%uze_-#l7J!SaO+o@j+P{~IymJmi6aN+cfzWI?Lj&xj7f|5m56FNxJ#=pSa z+{L63cz@XIY$+9vFFW~ztG)ZdU{5rH9wVSD--CA+g=M#-V-N3QM+%pfT zxxYS}Vb9kQ9d@=#yUj0MCgeA>I?qv9X?(}ee5NXoZt`KAlPu->?8L-9CR^0G7mD|VgGtD_mp+(SK996`_w1zD}WI5 z8j(*`D3xHWLjhQAwWB$p*Ez|bz~EIs*t|xd5-<57vXQvu!>%{&zka zQm^US3(khT504(9WhM#_p(1=#d(?x~gH_BCz=uYdT$YxFmW51&TZ9`*!a{~_LAPYK z5ePsw1RUKGi!z=vo>E~iaNWuH@u(#+2^rbH9<_9fL-#M!^ZHjbufSBMjK4o>N&Mqc z%a|Q^BuR7o9uwEZHSwa_uNH3er8BS2w?BFobJBz7(W27HcV7Q@#p#c%4nKZ}tQ`(^ zIQ4%Dre|(VXSX6vec#naTy-t0E-Vi5QQyEhBSdbr!P;HCM&8M;)A(SlUH>l-n+JlZ zCF!_iE92o3-J$Z}$U>QTMg>SB$PMdFrlPZ7?$L7`Gb z9F|&?jxS*gAJ;O%KUn(;1XSdSWidFp;Ujn`Oqj=>1i=L}*qt~@TgrEl&lxt&x>yd+ zyT{y&ZK}jeF$m`5geD*gTS!nbe8pTOc;;$rdbXRocXNZ1&A2s@qWr0NKDi=5#!IK`pH(J)(i$ylFN@}U+5 z!`;!#0el5PON9)G-uYQ5hQBS+L<*JEnWMwrQ2j||SYK>6qj68|Yd!0k0 zHcR1abad{B<_X}Xc1DS}t7k#lwI3Wy?<8Jm7Q){>V{y#;mi&#|);8G;9Q8Y<%Vzgw zmQJNzTGf&*9@E)`QBXB9U-|uk^ zKPh_!P8LL($M4oxdSwIWKEL+7eQDdM4JFwa*SvSiNNP6Y1@Obpol2wPQ2!($XLlBR zO88$Xn`M=gfLGvmh->M%NJ7^gdYWZZOQKmAA5UYRSX@P2Zr`8X%Dz8Zi>th)oAH`T zANUbkSMU1lv697DE&P0(>;=sgE}7bib8lAO&%XNN6EpFBMhET`XFO&06uh%d3a2;q ztkNDBN}UC%obY32X5j%%8DY4gI$S(ogoA?)#k{aj%z=PN5SrsGKjt&=y*?0&DVEB2~o( zcoOc-!?8uH5P3Y{@R=wgdfQJeK`r#0tnI;n<#tmx-qBI6==B8de2NIe*nAYG^k2t3 z1pGt%D&s2|A+XoWp1(5QR5n)TJh9OI(Q&tQW3Ml`Aok%cIW$l3M2s?Y-HB3My1uo0 zCPt)-UMIJUQxgyo*SgK}r-jo^hf;QVzk!73IHp(W6l{4H@W%A1w295F3gm6F3H^0iDf`?vNe+!1>4?gBA2eEV;=v$3SWZObE2W9Ua)h~M0*193r(DY` ztov$xXVyladwcj*!P0Qep{?QUM#~|O1z8E#+jJSqJj|VQK+xO?D*!tRbzJ=_$~gTb=T=<2B_+>H5rAApNubXQTyIity0yK z4!U?Wg&}u6x4)|X?qdUwW*^t1A%K}FU)-|odW{j$CW#&rSEH2 zWk%n6s*ZnnX7$Bv(m~(i4(3%}8#pMF%a6AA!`IB2SC+I&C;wOfpaGdYN+@65I?KbB{|%hb@cXEJ0je)F2b zbIqnr(I;1+6cta1ZKTZiXz3Z=zyvDwHl%~lcSmn&lX@mgM{JOg(OnJ4wjqQ`qaHQ4 zL1WE`JwCp#41RG_MyjNqrI<)!Iw4FWIws=O5t_>GA$|WqN}JYMaUFjNeN3ez&hKR3 zi{Sb&p^t5mhO#5iDBf^vvJdyq$3-tuJPW-mcaHXK{q18f>ReJf=3abu_xhy!E|ogq z3ipgSlhNiox&hZdHYce%f6uqiZru`iS2|W+WC#l#!&2GjkaaI#jZ<%W(i&qGwF{Khw+rvggOxmuup{=<@Ji}JcKIJW(%_{`cPBW+ueoNJ>A!a3H z4~@EbUWe+F5ZrUnL6%DOg1|*d5or%$dx*Y&C7+00xR%^m4(f+YPWP)jSuasa49lSX zN&-O8(`_}wj{Uks5)!2)GIL9@V?>C%SI^Hy>YCZO`CKX*nbUww2M8JWV{|4}v~#3y zynr_AMb1X$vrez&6`X-iP z704FHQ%!g8$oSY)NPbGZQsZOx~KW9?Xzo1A-ph+YiHjV>Fr%YJf#00us`kK$1R*eQDtYS-!`3tamE{#qZ z({jrkSrItYp`Fdhv6O!8q+^4h3umc%~ACHY9NC+*xNKgJA^8cKX|stON5qJO+Y^R zoHVtGc8<*e|72Vejnrh>?WpH1kWGj2~t98sU0~bbI$tpfjp`3CP}9EjZjfiQf+zmo$XUq zfde(EJW0iqt0kAobDa@NqGy^vS@zQD4!e5&yAoQXGqlI7QtNv28z`P)7C)^njBPy< z-bK>H`Q~{3jJU^FdP#fI;so}|&ANS8`Uaw_y5iK!Z+1-oMV`>>$WN%)xWO(3MGeX$ z5VtUwP3;(0aiJ+MePeph)~F3lE@}yu_^K*}B%(0`)f1K>jA~Y3e8D-LSmn{wBV%f+ zw4db$!i3u)k6#supZ2qYVJ{&e60GQ8npshyz(#AF`Ts}Udq6eub??HVClDa?jsXIM zD!r=#0)!r_ROua*-lW&idoR+9fKsGOhtQEG(o{N#fPg57!0+O#&b zFG#~f4*uLa8O#CX^{shNOM35~o(%oFdrZCto(b-87>mlaaqqHTuDr4O@iZ_at31dn zpTb1gB|C)l-QLH<uGDdOa!WI1h2A%rAojwRvR2Bk~({22bq`UN;yx*TSDNP^NXRkeFL^XLF#a!oDuGWKu|EzEc%D21 z4LA-)OSxX}UxSU32!X=iUqup68h=hS#J@B@MjaogC|5_?23^#fGy8ni)1h+B`k;n_ z%+s`(^( zsBUC}An0?QPIgeS5(V12vM|)9d1169K<$H(Y!1dxVEy44YA5xP0@XcH7inUI%4>v2 z8J6Cw??ZAPCw43CAS8y2qeN?NiiT1DI8guy_`~VeW4VcfFkvwEY$A!@WXs$$%W2td zq-~1Ka0*LaK$$TKkO&)AQ)-_ulRmYeRvJmXh6FAKcbOLsGFf+cxmt^ko)y3dtDUoGy zuF9tz`Su?~7|6o4Ku>OLCCbu(Z>-47(u%Kq5jVN0K zDwA|e|BY3=qU}RvPBu7}AJhU#u#xv1u8h$*l29$Maq5xqQ1&I)*Fb({96Y$S`jkPJ zROO)QILCQDR}MG|R@WooX~`^d9^rxz+K2GkFU~t;oe^Q?-v}$e7DnRXVwWe%TAVfP zCB9x6aB-$kdEo_ST6T|yWXvZW(S`x#Bt9Gkfgg9@F!vVRu!*jp2s>pV-qf!q@F`zWi)uA5i#;FIZxvzi&N!m z`4PVZs(@QNyoaubv(sa%k3?$6p+v;d?LQ^l1GT}93}f5xCkmyIX>BiztOfn{bSI_4 znW{|+(DYJipH3MxPJ*tw-DyiB35uoNY`&X7W+_ouIP|?=2+6(@JgmA#XYU~kk-MmU zb1&@A{)$LUbjPq=v94W%+d{zO3oF@UIJ4|D-{EH-vzKl>>Z2X3AIX$AeyjaKz0ALQ zbjIp)lEjSke48MdB~5!yrtO1(ud;jId)H!?ALJQ&vk5e%vZ0mijN2S7ZoGOny!zA()xUWI)UPPwK#KjI=$vkQ*;kY}u`(f2 z5S3u|tiDKb_Ki7fd8cY}zD;cXW=|EZ#(C}8x10mjlGRH{z!E#xfLz*$2GzmK$`6Xc z43w6MuVR}QhRO&IvvxxBKO-+*y$>x9A5)!p(JKeD5k`^-sWaLkjmUH&?Gmsnm0y{Q z{}{}7yG_$Bg6A(bcSA+%m$+8Bc+SLvd|t+S#cESXOinkc^Y=0^Pml2})=uAr)E~Z* zBWL@X{7y>uK3HGy)h)j1=Z*|GXML1=C?s{B8b`$ua% z)98c5U*>qkMg{hdjFoPBb^ZwHVoRH-7k!oVvCZwy0BYwUw?Y<9(B3qC+U@3%^WjJ| zhBosP<;>4tIElmC8YglG+&7EVMcx+k;XtKZ(Hc-xPF`YdYMR41as|w+Ic_}YH8jh! zX$QPD>$SJdI=pbW(Pb4@%rFaFpeUdw54ZYev_&YWB#ZCg)z`|Iy!LK)Zt~;a(YyMk z5rk}NsPSZCAF&uFp8#Ll;kMG@>8DeGE^fsi8DbA$O=B!oKuR|7(YkQWvm)ePxX~Yh zWs3gZCy(hD+w}jlqVi1w3UAxF@Tix!%J9k=Q9suCem}W;p7^1+eJC$gbNZjmTSYva zwt_bLHx{e#8s=Z2ZUd;5DoP`Dmn(&KvXEz!Y8{%-7NIXoZbb$>tHIQTvI_!@C4`UM z(5mm(?=#RSgGFWYO^p)Q9`^U2;w*MBMch${Fkg7PfEpi5!=!g-sx`VdH(qmr)~CLG z-)XTddls=pZ*E~vzH8%E01LX*>3JJM)uPS{3r+=Tf0A9z%csXU#jf zgn57qBv>H$uzyXWl%515EUm~eq%6k3<>y+N3K-dcAYTd+4IEw% zEu`lldkiJyd(!2!u?_-au~bpIwbp<}>puDq8O`%8FAaoF0JqeyXhtiEaa-cfU33bw z_NddYOv{zhYol6N9p>z0M~N*p9vd40)^fIlFKR=mYraV8KR_x3n^Xx?=*+>^!XA4R zIo7}BCRLZY5cZ-QK|qAvXQka7DJc~TtVf?`QnFF1``NWFMHpG=^MAk=tN+H{V2L?D zt<>!;d`WBO$`$yO?ku9;#AR2MM8?AjAA;FBHJ5=PT8){!6WOA`meenAJEjzpwv&`3 z4K7;E2I<)fjwcq!W>mMu*isG0%G?I7DxI6xl6YRG{2Zkw6X&7xsczu=+xSq{e{!+y zsJq@dV8ySI#Y2%7vt-x@p{?g6wqZ(`U+HSXp;PI#o zKaeU;8?$$-b4(L9?B{q{Hy>zkekOih*ZDW9#qS;|r4JeAU$H;1yHrm`T=#G_s^JX2 z?Mt@&H98(LRm}P4W|z?Zib|5+(PM<2i1dWN3PC7#_=ICKXH6~n@!oBwvmMZb|sv| zhmxlHC%dVCOAdeko&^ajShN&ry;q& z8`#MJdh0~u-EMgrHva%XYmRdwRG^;oY6k#+m|`}CXmfN+54Tq(x0sbM47-fY9gAky zHQmuk$G4yAj=7FU3J6tT58Fs)?AKrO#}F>D3x_yAN@GHeiv^W=Pz9gqKqxf?wi^A! zXP(4pnWVMG?b~>dfZ-AZj{Y>I8-}srCanKQtGYYf;Yj>kdTY`idx=-67^$)~M=^?U zi6iHI&b^!O^6ifu4Fx;ec` zzB@XS*w%b5`tC&Z>6>3=H2uM=$^c)ihQv1`XMBq1%2`B(NEhE^xMVC9sw9<5*CBDf zE(~^r8jTXaPpe-8h_9*5G{&@7E3+N6Myv(HO*`u@Q#Ow;{CSx&>&*!V^uB51nsTPv z&gj`Jl}>xU$*oi4s1lk}dP5n_F!hRtHP@DX9{p*cKw>Pjf`G)E-}ZXjTs2Id;)`Bb~1nJ3(sG*VLCt>@DNi@O{rBJm+y;(q2j-H{2ndVuBu4fb*y-c z!iZ%nj~*A{%{9n zmhLXSlj~ZF%`y4!D$%I@mNVaLIpUYvbDiILaZh_zI1ih?X?T#xyR&e+)o>D$*tiTb z!}lS$o=N4F`^fzM!AUj-b45#YsIUuhqp9&5&q7%OKm2#p(fVwj5Qyr4ytRC=8xqR^ zm>q3El=GNsz7bp?@Vkx2vsGIEtUEs4CIpf#$gC8qCG=BLE|_Q(h0vtz{ZK4!uvl;}Tk)-I9G(iSj0R`Z8J82{Q(x zWg8`JL6yow)+32ksPJDmHR=%_K9O2gqF)|V?TwD-6vIkeUXv@+EBvL*Bz*88prerI#TpZ*kvfW0Q zLe0TEm4AH?1Q&&_GeJ50Jj|4 zEh<+}MYX?o%PA7Ze9W^pGs?~J5j39A;GE$i@^G!G3B&V<_WByS6t!=={I^*o&erEy2TLQ1deO^YuNX z3nsCztl4I|7T5dF|07^g5llY+C!L}nw%wQMq<_*WR!0g zd4}YM9bSrDtE4=-ymj&(llAugU5f$_YrU_D!-JU3S)_p3Do1U;aH^&5xDBxyysHn- z+O!T&B#4Nt1*$np1B&WNU$)}t6a@sB7J-BnK%@$(Eyi>l&ySoz>|XM>_mjNm2QfZ) z8~qpYQl;~*22CBq@|N%Zz{=95Ad#VQ0bxP}&r0Y5|a{~$! zy(<>e3pBOGXQchl=n>Hhg!@mUDFW z85=d&#&^p5Omd#MW5f12x`5Xt`^e}8e_IsclqfWFoXUxxs#P?x{38K?qJY=y1w!e- zg7TdWPRxqA=%^iEM2xa8?Fx8XT>M~fq<%6ljQujNydX`&EtShsH8I!C0EzjM7w0*6 zja}33N|%X}!DNkh7RhbWEsyuc^?Lc_cgT0AMv{5S=&F`f2}f+$LTubu<0KkY7Hy9b zkk%BGJ8(b=}N52W2Y3!L>bk9p@Y z{Jd2*xcF;R%z^4mNa#n96!^5~o*|%g?Vw?{7t>2OSh8Ge<(qCss0og?HY%iD3(gF& z-^ImOSe>I#knGL?Dkl`dOvWsP<7_z8%=Q>8)m=TDcafd^ju}q)9BEuM-ud-3ja9JR zk<%J_31prql(U-tEGt6v%DWy{>~Fc9_CJ+i_O%~$8S-83D_W|wTtXk@c0W7yL`mXHeeZkGfakLO7SGmeX2zZL zEWPt*JM@D8VZ5zjcOGt|0mxXH)yHlT|$6zu+OHF7g zcbN#kfIC|4OaBgmh$dHdp$q4}B-i?1$J)4k-v%~2!bGPh2oWgSN zkyT&3AR^+h&r{-WH@&{n0u0w|Te4E@^9U9%0wPD~mb5NQZ!-(T*>LxB>}JNPcxhZ8 z7(~-MG0ofBGObY?U-jNR^TA{_TDMUv{3vQTbjgNZF| zmxoLqrbp6Ln~do8j~$7x-ks2L?FYnCt^E9HA+}E(C#_C@Tr+GH99e`!OfyR^4lrP8 z@8)JI`b8X~8+|Hbcw}&?_-z!U)K%NV8Xt2&s5!~-vc(c_m+^n=vbvOeF_%((r%Wi|f&CgAeC z5qyx=Z?tZ7`+LOlhGeG|ymB-%s=!k7U0K}K4c{5~+c+^6MZH3=dn#+x*X8(z_|^!%O1>}iZ}y;R`h zV;srTGLxuryu>+#_}1EsEIQRwt6yv14W9|7yIZx2u&^%fOgXIUysgL}6HfewN{#xj zNA+3GRnq>( zHJza8vxK?7TjoOD~`gJK= zuI{25$r{$EqBF6pf&bdHX*TZn#=0wnkp#=nTU%V6MqONIrFfTO7wxAx%?r~J_N7_b zW3pllJxpRW$5UQZu&LYQZN;WF4Mk~5&*8HC!|ekb#dkz6-u~E#O3kz**$c{d6ZyQm z*BgrU{Lt$4p%25+gX*?F4rK7S4f5lYgg7$fvytAXz?M)h&{_X^De!~;%#=pyraHq2W=xa*A>LAJ#6q7&B7+NR`IAkcN^Ig@c&L$!j$E4?Jd5~MW zYT3L1y`*?&iopxZLAjHc0I8IYC)5mPIB|1Us?MzQ<^AUCbCt{X)d}V$xS488+(_H7|-hh3cAKG1K>^^Od_`P}Y!q`X~z-2tTDQi6-O z4PsidJEZI!Yn$Bzb>Vb#-%Os7tMQ+gq2Fs{uVCWT3UeW#Lcuq6U1gI4-*jv@E^0*F z1JToiiC~<*6hK}*6ySdDWS3&fl6Ut*{O6r$c5q#B@O?hps_TgpaN=%}_$Q4$of9M< z%&GyG=ib8c)qyzD+=Ovr+Rpc7iW(Jux;FljRM{!6b(AfP!DL0xiA%@5k2z$-pR&D` zPcl+aZpd{Wuuy(Fmvvzm%<$h<&}tHZmjQJ7SN063w$`M9tkv6)|FM07>Z%4%wBi^F zPA6stqSE7FHGp$J7!g;xPbmJ#JEIwFyE6_Q$c+mcLJ7MlsSccBF;m*<_JBz8^1l3=BgC)&(6AIZWECSu_4nBe|T9u*_L)#?cD!*mV?Kuf#qbAqsHQ&*DTj0KJfxo&)!MTau&s0 z{z}AjGtS0wAN+)nkH{x2^Yen~8ZPr!&fsfj@QnKe$Fd6S5uL#};K73mX8vfQ!^Gdr zaJ8of#&KPYUY{=i{I`4k3ayEs7MROnpKr{FLrkws41nGqc*KA(-_(`rTG})U;8Y(4 z23}RX+Q#k;6m*Pk<#Xd$JRB~nH>k>?N#3@qX7~qyYc3?{hfTBj@-$H9)BmmSJmC~A zPoBe7rT%?ro&rDS?S~hUvB)}4L_rPxaC+~%iv9QDKa9EqW?T8+pTj)X=fUbNYD*s8 zrMv2`=Db|ID2026SZK~?JTl+(>wFMN6(eO>o^T&%@6sjg=g-+R+#CUPocv3DZv_vL zB)f;z-^W3{8yx7=2-qyyt=>0u;SyZ^Yx7=O|DZzX%ZnKC`mOsIy zD;dTKGrGU&bkf71W-i7i6HA88ty%RAb}oGXTN{~~es@w_p`(QOS-~&LE1APO`l$x9 z>^AET77>es6aMBxsV)Z}FPxuSZQ+2oxqL31uf=hs2Jo}$`sM$vh*@)20Smt^e>zuZ zPy(1&PSoe>O|nof3Dd%bC`0b02aEHqSU}}T!xGJfGhS8cFd-%Q`M>OIb_lS5d}V^8 ziTEheClBu7?E;X0F$v@065#&}-~KNy;s0O~*0F+kK*Q5YPlfK`?FjUo<)(Yszo~@p z&HqIuT;FOk;I`>u!GE)q{=3oI%7(o9XHx`%;94QoJ*wB@=gN|Kd@MWBKY;wHwlBCh zrGoT~6u}%}(K`bm^3`#|QkC)Zlr!Q(^OHkjPW1Kk4jsXC& z2*oRAHSQ{u?bUx!@V8A}g#eQdO8Gmxl#G0-<3ytLWCE6C!nIJ8**gL z@X%U)RxlMfQ-GvA93B+AV0kykFoieKc6#h2R9Gf=jSI!${iaL!=ab-X%)0tOv?Ue_ znrbCYBH4&hHDBTXVvXOiV5n`nOU#6$tbm+zBAoOD?Fq}Li7m1R&`#M^21g{PBGg5- z_j2b76@is7hDYXZkPsyV9bX8E9N~ zBxv37!5O5I96Js*LhH{I)aM+bQv)KyY(W|tvgspP`_|}nXEfl#GR}5IGOv(&D7G=p zmx!oTXJp;hiGs69{qS9CBV1y=$hfL1Cbgeriqa>Zmqz7Szzy zvh4(5tI7L+0OS00t;Ulb^?YPJ_9p?~(%zcZao?ZPD}k;D{yiG0k~*39yhQBC!IkIk znTsHvzfqH_-n!^SoELmHdg*6kdHO881M~vI(*%h_z47i`w(p{p*!b|rUoM5>O!|gN zy7t8M6#TqLf3tjl*n-mblC&WlXb`h1#NFcDHKG|BCNQ8U@g^W!OQl**MrOpPO2(TU z5frFG+4J(X!Z(^J6)w|ye^CPv0FefpS?4hY1!E55a%QC3JiNH=8;g2B?4F;|!AuAM z5m9aGRO!)A6vXjyeTJESd3VdZREOCMtx?rIw*(~?P+Atewd$w8QZ59&c-=98cBru^QXs(a z>S|G$d@UYCbx{P@UZWVG@{cp;XZ-|cp}j2{9%uA~1X0>v$g)Iln>@fXq{t|>;LFTi zFIsW_X9x5*!`g$JE zZHU5@_flSA0*w^u=%ci50i9I6D zyy5);P``HX^47BxGlI1Bd$I}Oxgdkmrn@Q+s*d309-n56^ItxkJ?{M5-oExvzpwAU zTC9El8yx!L?eVYwS!8yBO8A^ASC)oC9h?)ja4O#YFO2__|Nq|*;V0Fy$z?o#H)8}a zMd%+83&t+u@DhpfF+AUD^LoKrz7g2OrN&Mldc|juh8w6_Qdo6RuJS9-(VZaM(ZH!p zRaws&(mGa@@`H*?BTvM3JmY_7(U^9QMuY+M*obU14s<&)ehPSmlfb+Jjt945DR=oQ z{{S>O{sC0or`;@|{sDZe2)&qpa_vOzhD(d+P7{pk(%*$^og1RL0A24}!HOOk)3G7@ zHxav&If7=Lp8sbH;Q7pNSaBtQl@Qq2$bqkL#AeqcOie{Boqbf7ev8YgYwG;LR8Z5jRC*q$$Q#O`+$F8tjn9e!3b+lM-)xvgY(xR{RfTsRfAe<^*_N z*m+o|U}CM0LH8nhLnDk6H{wz&+nB#vOxuTk462dmd%aUZ^G)z**FWOX*Ez8``{|aY z$D*)C4Z^yxk7_^zDB3{BRyF+HUSXZ}paB6@E@3t)D-Rs=FbWO(n5$wn7#aB*iUnZJc}_$ z3_{-hX?T7e6Z8gP0zQqPF%*aXLB_sT$P7~d7B)8~9@4Iw)rojj0;Gd_Y~hx~+pfL| z?jmo*yd%h|xm17-ti*SAMWvz2buvqUN+0@^qclXd<{Z|_*|CondBZKMV2H5;jB>@8f#Xwn*?%9K9 zGo-vcA4lccdbaZyWx+cdmLBYNe1jnpI!tBQ9mho$Q_WP(2y+6hvGQABE9!hgU9~dU zxrK?&=ijfb%P!?6(-jJ85IQS*3%@+Y38Uo{v^%n$@3tRwm{Mb?J8gs|v2OWFFYZZh zgcsO=e*}?#^x6(IIG2QHljsx@1Z2PdP`*5-*eL z^jC-<#*4nK=Fg5<=w(9}O5~OJoC@f}jcE4RshhzO)bE6OmXr}VrRve2Zwb8* z-X?tz=%_wH1@kjg8xJ{7NJEzHu|m&ie<#Wdk>cpU<$m7iZ;ljjyy6Ye(ORW$2Gvq_ zQ@)|*^=P7Ii=N|buI+qS*m2e=lr@W7ZsL{xM1^R^k6`Nc4j;#58q*52l{F()6f4P= zv`YURwVugNitNq7S5NiRb?%~deF_yUhvei7;>bdN=cB;HBefG2s#@uapck#*Tz}gt zTa$ECSVgK8y=#Hgnp03E2m#DLzp+u26cTiux6RJ*F{t3#mP1PxRxKH!GXlysk>ROP zs~XtX-wZdJcXMFS!@RgiYQCuMK+VHpsayqP-pLq$fUUXM=MSh1x9;``zw0c%H4BhY z7|?d(aJSW8@RwRZAbJlvRX2kX)X13GUovj*;9?J4ki%spYMIO7Fi~vz5YYQumcIjr?9MD;Wt6Wc!dwCV z3~C?NCSNC;kOk{2$%WT#o9HWKG?ti)Kq|O1D&alD74eENNv+qj%i+NBCb`;hEi?%{ ztoP@{l5rg#;(!2#G3Qr^w>)9~Ef?p&g)PH@GPnUvjd?o0Tb22*J0j2E?@1r@Ilq)} z*4;ov>CI_$S+-l_Pw;0d3Foh(1j=!JNgWH=EO&=1Bs2HGq$*psUhYksG97Bz0rT+Z zbz+lJa2|hB9_<8Vgwb0!{Xepd>NriQnPb^N>uu8V`t&e0Wh;sT83IqWb`P=<)4ysl zXWXee#7iX==ICO0^3x z3U+QXRq-lmnlcL9sDwDrk4xVyF!GXSebo9Kl&lr()QapGoQImzH2QB0JUjK{4jCV5 zBoEULYwlGa`4E5$Bk`_73A&`}G*beNfBdrSoBf1MXZ7}E@!D>s;+QnTGV5m6zyI~* zJHwW~Oav9TYdXh6foMO7MgT`=1V~5g2R?hIJ?Qf{RtfJeNzN?bbMZvQ_ax;J_|2l0 zLMRB1PT}uVYCap1H&`~k{sRVuUGUU|*Q{PjnlqiumX*-YDB1aO3Mv??N&h{BHmo9n z9U^$paXi0B*aDr>mej}NfY0tA65Lfq(i6TL*-;Sc64EV}GPhzkA}l<#a#|H5a1M6a zSc_(IifYI<96odkjTP}-tP-l2V&$JlptPI@y<}WxAgZT{@jX4VxTYoEQP6E=Qv1rZ3P6b|;6I z>w|)Z;Eo~p)P(XgRsi4G+I5tC=S3O)i_Z2n-dfO4z_`wzPjj%@CSKO8OQac8kW?nZ zX^EfE6)M4%(wPfs01C>gKsxHJEVgpqvY*w$lM1Fu=jywv=BXQLv=Egn!luZMa?%N7 z$#h5C3~EbaZ)TA@tv9rd>j!>TZ}>m>mY!rIG-1Z;2QDsa8u1ExxRpH^u{tvuI8q}l zPJnc?RVpo-zvOkqMq$Hu?N^g!5h!t~+grgvMCw6%6FKbMQHfcLc*y=i^MxZiAF#|d zwx{HdMh7IQ(6PS-h&=CC;<-fsa1S4(?i)02tT0W-qfELd-B-(E;Nel5{IHb&BWLd8 zHJyCXgo9#sOT5Hs)S^o4!&~`{OMf+207E2)6HTXaage2DExxuW%DTy8vK!2&4lq<1 zp2@j`Z4esFmcM>15<`byU|Ehxtpg|uUV&@BWErba;o+c}p{6XlH0;`d=&+A9y=kga z3K-S~k>zAZ!bBivqz8wkX3cK;dWI!@iVPF;yc7ydx0$yRr(UYyS0jbyi9#wm?I|yF zoJt$7H3eXuX}zzt42_BGh+ZzJpoDfmG8o+R%{5Ts`6Y!l;|hraKj`zTsj)Hb40+^( z6-@Y5VG|dnr(2whFpTA=a$ptdYb;7s*|AIs^1i@oLIW7xr1K`d0USN;#cfxkl$70q zCvqq@U;pP3OxO$E?CVKQP@h7IAzq%@6uUBNqAz1}Y_6-Iob{uoW>?#82JC20ew<;K zV-}`*gOp3OOoCu4B)`_|oR*n|(J~Ub!Xm~wU-1IcRv;oGlnU0SqL$N!>yl9)1ZVN0s7k0vyD6lY8(UH`1S6b;blpYgD-G^4Ik z1$A)b#akF5p+0~y>PF-w=unmR2)|5Kge&RUMx3NNdbCq7u!!HIN&AmC@>8OMvbfM# zt{^n*1gbdP0$Ev~C6_|&x8aWC)Hf(`m|?eBJMFW4ZR=~A)6hcQVn&oB~aCV+p}6rqskCJ z+-MA1w(xjE%n-CfhIMo#L8`u~Wif<&;Yo!&nB%4=Y4~CS9+so8TNG~MT-=?X zXktF|t*R4CrHfR6D@u4POMD}y>X`khk&wkEj2{K!*B2y0!7}WSBFwc;RlkWLc6{6p z3bF81gC=M$J(D7BHT_0oKNO}Mseo68ithk{woHPF<#IKnkl#N_iSo1m#@*1rM^O_s z&v-uJL=>`T!(vpMUE2$0WjiU!K&h((3A#F;7M2evRn?~>;8anii_T_W_@}0T@mi)w z!Y4GeO<9&xHc}+$FB^mP!whhIYgaX0OeSn10i-yl9+kWtcs@pd%{CefSXR$8O z+&u*_L2fftL2X=2Zs$Sv+vS5A9?TqlUR&LgUq@-#cizd0N6ZxLRU~D@I01LlY!Bx@ zo>nY5t*CVd$&0Z{hQ7FtjY6a%^>Isaq{Iw)3~^reN_ZD2Xu*(JmfxCLuLn{`ulflh z!^)m1++?Wj#3G_@6%?`_)g^?6Af#h&DG{^szbJLzRhnH@wJ5Paj=i5HwZ?x>CGtR* zI*MUmj22C<8BR43Bcny$%Y_-0@n@Lt9+(3c&y#|(aU6AAmeD%HF~i?qzjJBjfICNY zQa;&+MgOZ}VnkNlHA=kNj(7K(O{vD=K!Q`=+od#@rFvw1CLkT{wCTAr1;i8>;%aoCQLR6C~3;o@V{%tMWp7Wl$7Qx=T!D- z$V{O>q!4?2z2D;i()jGG2%(D=IH*Gl8&B@v-9MA318KB!W7QraeV`&D)wVU)!j7B; zdY%I~!vHB)i~p)mKP=G@+n6e@DkhRSXdwX*>c=DTl(?uk&tjMjQ!7%*H5z0)^KWY} zRdDcBz}3}=GC76QAOstE*e)GdF`e0u6Y^(m2>yk%gH<^ip3`xZ?TioqG z1DBF6UmclD{iPnzMsC3Iko0rk0CC;Y}Csn{mfrL zIR4oV?GQyvka6AL)j9!61ghRDkA(@4Mq1+^`M3&`s?SL@IQZw|Y?p;pCTue`-8_y7 zs~?9rS}Mq{d9({wu+)5gr=-2N-ZaQV;iX2VD1v$jpPaI>yh(J0sU7r4^lZVZE@|wsFuFRJ-j%N?kVkkhk6Kx%sE1`%JfF` z6AFTy#21V^#8dU2*U&=a=o8?R78-&ISBTKiaf zVw!}lNFYKG9i7br91K?-gP*R47vSX4TJfqNxvO#zRa3ZE&l|O{G7PNjG!>%>&8ZQX zn1}w%I-%B`BcR6J>A+W^)-SZ(whm^;(f#en6-n(wCmKgb01q`x`_x~WT~vHg{qQ5` zn&UC^L5r9O**#ous_#rAZ#gwN&;yfYXHolpFiu15c`ZhKsj9bBd6Z7WSY|;6Phmwh zY=>Q_5nz`7HYD+zoMNv1Tsf?Z^PO4$)rOUIKZyk3T9p`w$|XZMTZ5$WR=jt8nIJZH zIEcQ)wk8bsaPF2yDw+W`$XS@O#4l4M2s4YpueYwkQ{yYZwNedmyM^K9`csnP++e67 zke^c&vs4TU_phZj2GWCeU}IJxkzAX=Z9+#I{rQ%f!Rc6`XC`l%g5wLlNS~QsUa(1; zG=-ZPYC*8xr26j$$C~F9g-wkrO z&G8Bw#3B@daW!VU+@{e2%s~YQci>VZCU$scUdhBR)U1|8jo2^N-1o3@)VYGO5jsfI z@XGS1`7~T8c*%8_)@jbWw;(eU)gppC}z~im&%O zka~mHJ{*^3W=pyi?D9h4J?5#!S)F+>fE+ z`iPKeMn4!vOr>4Ggtr^tem%o$73PWKgr-9T1P(o4ejSxpPy`Ie$(b{3UI#7tvtL1V z?JM68UoUpMQ`CojI;h}@CY!FmNzbi}e@hPB1i~Ih_lHut>%h;tW8Z{a)Z+z|2)-$#T_bmTPPZQgs!-Pb_6BMy&zsGq_bB#94CCQuPw6&PnWmO@C{vD!1)N(8O6!v5JKXqz z!VBC6`bcR>So6l1rihs5b?6BZ=V1Dda5#n2AjAmo6TfwD$( zenC~b%*Ecl1Me4X+1qOIAM%@3jmZEJnwwN~SJjDRwDeKU+FRQ}3J?2#v|2=%YGzj8 zf@s!>V}`~MPPCM&fKsWr3r`$;g<(6Nr7Au)-M9Vegqg398D>)bJKlE++8uRMT31$!CRCnu2NK6ZPpmLkF68WXCsUH1wvu;yOt{*w2&(LSdFd8y~cz*1mhn zwH@$}7C8#qGEqS?wWK0E6Nwr%w( zXc}*+^kEX^NzwQYV}~0tlT|A51p6ZCR(Bk^>!2zKR_498{+U+vVP4HEh`QJI%P4H- zJ4e|N1=FVJAJamXB4m`t!j*`W2bKbYcGkb7N7U^S`m{ zV|)??Pu&T+_6qlp!BxFtybOc`z(JV4tfr!l0w!V}K^9Jg0nDJzL=KZ&KzQzMsxtiu zbXL%ro+?5Bh$*!uZWe|;H040W&xn>!wWh{z%fdUrebE*WE$y|V+y1QC(8jIRdy>@?_VQWQeAZjM)LW=&-aWHia5ZM zSyHrntazA8v?4AKaI6a?FIYlo8r>gZ+n!`$;Xks6B13gKv8z6neP7cjbuJ1LC!Tm>9hLxLoc*^=*qp5U(hQ2g2559MIHz)d+%uuSRMqF--4R_F^-JJD!?y9KkG2M%Xv9nW_ zO&Uvyh!x*uy_pTrBoR~?nr)@jGG4CV40=x)zdbxt;Q!d?nA6`LL(9xg8hyKz)aG(B zgE~o=7J0-93ZryxZrMZ`zDg<>?4J|#m&JWR_R(kgLfG6wM<+jpvbCXfr~E!Xacszv zo;p9QZ!Yu`T}~705&VlS3nNny_Dr0E@;yFS>}KJG!<>g6hXtU=v}DR&DScOy zuH*kjnE5iPf`d9>~VOxqSQ%7VK&_oh_!Sz}+W{#ni)XXj&>=1^G070c;51eK?cBtWwfJj44NE%1bSg+e^!XL$q9M@r- zX#&bzc^7qEWf(g$eO<)ZS_BBl=83THe3r>*ZKf;e>AW+0BtQZHtk6D$d-CB-Stf@w zv91u{qTV5=bYItzJL%!qGffKGzkstF19LB^8svx8w1!gj=YQdIaU1(7ug?QLjl>*a z!}htHz_)We;5^Q)3nKyxuBW6PVe0YbRYvffCf+GVdFP}sJ-~B5{+O3KUX!c$Ou;f& z6^1fn2=&!eWc?KlAvK%rQ?+RBTIO;%1Fvu(u{M%30=v_Dp#ek~0UCeTps|mzlwUs8 zj&HQES*}H^>h3?$9JZpQz(+F51sM_uxz5<0G<>Ur6b?WM2 zcCgX(=6+fL;7JN#k3|EJig7$+o#PKQyaJoLMvrp%VNY0}a5r0H&G_ zMi{X~4Hdt=QSArUm1;fLrO&nGc||!AZYM>Mr{VffAcz3`L`|CtL8HQN9=A1y z%EhJC5Sso2Aog26oC}Y%*sPdyxUxHV5*?{XEvg9q#PtAKW!~+Kta}(Ex}+f$+VeY# zAj%LyXz;W8VMLF5V#Ff81o=%rRA+aObzG4)JE;&l&BjDQ8-LHlG*Ht90ad5$Qu-Jk zcKxFFO0~kHT+R|)w0Ro^R<)<-+{E(@wkaw%3o78zMB86iVCe7NzdieeVl(DXEz@&? zE(lwF|HqOwT3xO(#c1h+^Hn0Je3S%Xcy9!6JuYRLSN_6(A@03{nhN@`(Ih00K#(N# zY6t-XgkA;25CViyq<86EM0!!tBm|@?gkHqZK{|pIMd?x%5EKv*5EM}n5wY`L-tYVC zo4NPSEpzA2oH;wY=bW9LIdjhL^ZV`dJWDS#@=;OU8E({qJ3Xe7RTp^tD=*&rJ$`1h z`?W2EUQ}~Zvff!q&XcH`4)vJG@SfOg41IQAywg4JGo$944ins4Lvqar-6VkgRDx=h_yPjM-wt+uSCKSa~)v!p(}QF zshdz-5Gq$*YgMutkxs|{rf`F&#sEu38^ISHw!Z}buwHoNxou`%HSAOJfH(ivGR;vP zn)tBSrACpM(QhH{UM@=r1j|hNCvPRk87Reaxibk_vYTW5@XaNnXl%2|3@0E6NZh~W zXAYj`>v9I6ZecfP4ZtE)XGM~zkFl0h6(oJ7(@aX@#fCY%wU0QvlT8({4}WQxYK^u@ zF)+$a8u}z)JHzXT_E3MYsG3q~4>c9 z5V;%I&*{s_q7`OR5>d9h#VW-(qn9?~Ior4rDp-Q+Z<@#qKnN6BGfOe(+9zRev z4c^&7hvi=2vr&)Mx^F{845=7?`i2Ch{RD8O!){&1H&DM5g6S+S&^~5MAcU!^#ma$v zmR%A~&?$Xu`zzPXnOB(z4Y^k*IbQ`XxC9F}Y3s>#W?!@Lr5d$*{0O}_K@p{(3&D|i zQ7u2PcV?C%0u|&-tRHiO(=rzeCkq`Q5p+KpufhJQ3GcV^lw^^61)L9qQg+6K{ho8m zb->Z?nSmf}1#NW_)wKCLbPs?S{*Avn+hs%pV~E?s9I%Q-0|DwF2fW+K`ui;v2L{74 zJIhx|_hi?eEHG4^31gJhXmQ5bid|_q3_@w;TD?8lcj{N7Ekky?^kw_C5Two>w?3br zaR(#t`R4Z-@J7CT=iRH$vv)B6+oiq~VeD_@804qAaG zO5?KDMPj^%Eiai#S@Bs+II#gS?_aM4m~tmmw`1{U*0sJN)$I#`n2yca_XcJ{kker~ z(oJ=`gh4^1s=j)Xkvk!mvUcRXZietbn|q<--Ub5-p&wp*_+RV_c!%1Sv!P?z%jr5E ze5pxU@Vfq&e3$qvURZ5s&U4ySb$D@gEd04EDCPgd>SnMK{So(RQLRqs#TR8W97CNJ z-%Yo#^nOeFH&`zH(aYj>e9#9gU9RCk8<+8(V5O{O1a6Z5{98a&NLDJ0@S%B*cCMZ@ zwZjMisY!(ww=0OEDw}U9Dq{-`*(W&F z&clNfT`m?5c07;k3_nE{17wB7SRdLB!0{7X(e^J`MK2yeuZh=>uUBnbF5VFuL`0=l zp29Q5T`JRfDRihcb$XUFO;X!Cv^o)&1Ee!Fv^OOw{N_(uXZy26c36d zhCq3N9OE1lIn>)f_2_#YF*jK6PU>w<%^E$$Q7dJsnvZoNe7QW!F!s&irKGX) z;5+xK61k)S6ly1yTRRrqx^aGo9s03YW0~#i)?bd=IteFuJW-uGv=aD!k(*838p*pr zqoB6yHH+H54_;5N9T3uaHR7wwkG4H9tQ(Tai+Sns|v7qO8XkuG~>wg_j3 zWhbFdHY&XK9JGHOQiTZDTm;}MhypfVpG}@Zb}6Y4dkG%VNmvMd+*>5+{kB@QZvUVF zDZ*A`&Ox>LLY?`>YJqV|xmM?>o$EF~EFds;z<;=6a*b@a^XOeeMiwl#YPk26zZ(k< zCWY;LlTjS=ss@nI`yvcYOYuQ1r;yokq%0dFa4m8wiOsrTxYx$YL%~*65~Ll_Um+M? z2G^D?YdUbHtNiHW$Sl!%bUSm z!*BFOyZ9#cdu}xL^vA~~N{&O|Oa=JsrQiL|exoc7*<9DSy*J&z*8%S&)TwAWaLOh{ zJy`>2=beJ6U*Q|sS|YcXXRvx!YH}NsjO`{6WK}iebCHaNI#ZgLwupY7HzPDrJx6!w6<72(V(%W4+!i>bUHd!|&iMfe8FEGq2<+ za7g(Ih=Ni-;j$13f(<3%)DI~>1m5w5L&4LoY>Wt5@Waw(qaYavLonKdJH8#N4yO_b zSLaFD@oX?$GL|j$ic~h}_NQ0w!RNh=L&f?-L$$^QI2ea0PHSzfTaRU%umDIED~)3_ z8NOIthcEf`vC1Qd6s4SuT@tq7v7I|By+cKy5qY-l3;}hdw_)m54<7e)QdXWtPD$c0 zNBKGRa{HkBl~XvJ)DdB?@=n#?Yj&1?J3rp|>aBr18WGUt-M{aTAEsCj=mv2?R0%GQD8q$OpG?GeQN15Bp`5&nxD}X#3qAM`rPTDY(Q4-?I0VWA~i&=0%vqnTGANZ$TunademFbsJa!gT9fnmR8Z~Zl??NSui_q?*s zq@45-u(bpXu}rafFpM-t+hpfm`gOsx>`K;&3O{zhbq(h>6U54r^?3d#Qo%hx3ud>< zNnop?9xNTNY{9T6{+B&l(=L1QW%F1(qM}k#WK=Kl5{b%4XIJ$`JwCscmh5Z6_EFSI z&&#={ogU6#s&IQs#vzKM0iGLxtQp$(GBqE$r!u9Bz!i%;J=+8hZUK}71BG9x?3rJ4 zfTrt=T2_?2If4*~IE$z6ESe%NsZzY?!sTg24+yZ-a0gJdqD{kywT|FJi!-n31hD8t z&7%=n*+pNl5bcKja!LvqJ+No(1XT5AI4#=pt+amN?C)Pvan$Pb`Gj`n%ILt@)nlW* zolconTjd$5BPFa>TSnA;@FZt$1W(AFB?*2tVUzlL4tC}9M$f*+KWtRC5@IP9!pGg? zh%Psiip7y3K>X}iWImgevUHwruGJ4yOus7CNL{du9+xl=5PE1zx!RotxFlV&nvYTc zIV+Un21zTJ%j(7Kp#i@6KKsnT`{rOEgXU?xvrNdABafwaH~!ljySiDLX#q&R*T~y5&Rvv_9npqLkPjLUNDfdH!2%E!E;*Ffn1D&W3&ucB|U)<0#nimimn%CdW=-nKau|UoWk-F)& zM+MR2seIhIGbinlf7>?eLSGr*lenKvnLOx>(o*UcUImV9wu`p zX1%m~^V~qF5P#BQ>w`WH4VD`leg#Pb-@)1-bCY{JkZ+P5aFtD8UGp6R#W<={Kj9aRcbWW5=Tro>{9dl3r{N06n zw&;<&q}U1F#AdG0H}=0wHfEGx`mn*{8OMR{x5bCJd;o_4f_1EX8jzndSNA6i39S_e;AC>c;Ua=VG z+wEu%MFz$|th$`QMTOU%9xBGCk4CJc<2F*J8@D4wdC$~Rc$=m&ECv~YsRVPnV6pFEqIVt@R*3Qq+N~=*O=Hd!UTP!AE}qLM#3Zf|euZ12B@hNZ%u4YY zV?BZVcUqO8XZ=>90RBC}>B{&Ut^JdnTxuN^?zIq#D1fsG1otI6uxG1H(4v#Zw5Q^e zc*&BPG+c;&c^^r^jMI=&NW@5w7?&Gt0;t$^^D1dy>^ra7Xc${P&pb7f2Xrj7kPVcT zhRLX$+OvZ=ai%DYoZY)6&SmCk&{xB2ZE-=S-{uuC8m8EOKSB9Ow_s3t+WgIq=YiVB zwC~Z2r=d5#{z*Wf?(lXI*G{P(zWT`_=*rC==`5P9O%%W2zwy}mt=I1X3^&o(6*XWz z&vItY6W4j;D&toNBP{YiV5wET_o`9Z_-Djf0C_;L1$hH%CIuSgGkJlPGw~Ht7hUR; zM=)&O=gLVgzBapxt`gAEP<_y!fclL}%O$1_Gazh|7o@?v1>>N>;Tb0+NmEwq3gQ&4 zsP+RHk;}yh>l5B7X~gDGA>92ffUJIe#ernrQc#ojn=p zV9;6k5JsdXypji@q4|QY^UjYX&GArpW{|}S?^6j)&?OfGDVTGwZ3rPGA2M2$_9@c_ zl4?q;eK;wG@T*jhG`xaaPvIUkYm-BKT3(n^OPYKqd6kJPATg-93@>zFYaLqPwq^1@ zua)!4$82eortIJy93bR3TKDdyq4s{N6qIB{Agg3W+s`d!xDRjIH zgLvPlh^ubW>Y-N07s>Qkzt6GwfQLwr8-KvM1*hGoq((+I?6$h!L8Z67cnR|Vx2IhcBTwN?QJPWVAlUhzt;2N8$oqR$Iuj#uw; zCFcTWPcd*Y`!!wX&$G^5j5k-{B6p{@4N|K1!=i;Di=u8=VI_NMNe}=`43_-YU|rAQ zJk}4Z=Gk*l=kXO#n6Pyo*6tXOb8T0C9wy>E)*4(0B zR@&S1;p%xiFIP!)l~p&r*rypOf_z3WfW*a|qHm*?B0%p?uOKa2VTCmaLM|Md0>~-I z(w%o2!*~X&?Mnf5?UH3y5GdK1(z7Sq1Xb;?7@bgaIJI11@xJ|5QI(DnP2ovU({*3P z5p5f{)aZJ>^wEb>P=oLYT*HY=&Jo%v{@KVQ^VL@`SQI6Tk?H!lcea7PbdxIy#t62E4g?QRA49jD8sm;{C{jL0y zPC^g-gN=I!h^p$QqU6=`N~xLd`0qTl+t@?Zdzg{8<&BEavZqfkjA}|W(u{qb0sPe& zjJAw{w`A+8a=QJYo&KFC`{#pxjS@!>BJBCB3fs`9iOD4w^o9MPxAc)OZ^SpJ)rW&W z7>P1-l(8pme+P0X?_mYv{=v?p=V7YfZ8WbZKoeFg@_ImHi^Hc~S$so^uIgM4LydyL z)fXkAe-uE6%xaj9ZS!xtzjdg~Zy@a!fMG7yXH}UP9_G9v$W4TS#Clp>Eyxq5sTt&x z9$9}dbr!KoHY*bn1-`5f-Qpp)3~cJ)saqWTstMdYHJ&?h z$x2qw4r3y%|D3GRuF#B~`WgF0KuS62`>Nipp)!xtym1EDsu{mPd_W8!W%)ug;7_`29)Y*i)z0>CAe^IkGv$!aYbk+q%uB;oIQ6M zeLQbDxm2L>fSuv{ zQfOlT*gcnvLry?|zy|3xb_J5s3iV|tm(Co=#K~xR8DHFBnoOpQ+_aP!7zLW{l;z!! zCv3MLuSOa08?wOk7!V_G`r8D$p|wdVaPzdSy(;Q0G}UNwFU`(7v?C=Z|s<}crw#q2Ertl)tzcf{L zUUpg7%ajkQg1}|_$xFxNrodB9cp)twWDOOjqBo%>;U+5&j#!;&?Yt>zRUV7>dGj#r z3o*NpD~B~hbfTGh+N76wCKaipxji_n&T0+L5@A&;y+Z(#g;I?;m9up%l*{QbwJS3F z1u|e*GF6G~3@7p}Q4h~>v<8C!{SAK>dP%0-JLg*CR{+f0=FFYjq^-C_7)l@?LD0-R zApq^q7N5(O7h;s7{Xvg81aeJrpst<=s7&q#WyWn{9_(r$#31;4+1&;4x@&WX4)`|dSWueg;j6nhs~YSd?3 zN|5=B_FM6x8eq3~I4T%izyx>thT z8^tiNrh;FH6JGxTsAlQKF*NSCXNhk<(34c!Z_`f;xK$bUJ8QT1KLEb~t)#KH$B9Pc zXZRG0ymr^YBerqz=gUEALbTGmJl8^Nc7?F_*-3I+QxOMT{j9syTtAEE8b_4fCNpxp zET}2gII<0SMY$m*bHev4ib$IG4X0_8+({;9NoOnR^Cb}n*iB``|03^grn7L$Rqq7K zE>Fbw+my|{pu!v{%5OaT9f}n|_bo9qI)u$U3aoQ`*yjZwYxcA3bAebnf_$G^3fKwR zrYtOFq!8fnpx2XP_OakU zt;&4Sr;L%uBB0C8 z@Q$&Iy5!q6W9VfG6FG>`bQzBL*LOA=*wxMtPu^imv{5(n#P|){%!>WqF>5X!4y76Z znsBLLM1P&t&5xvf4gv6p?r(0+bk5ai^(=Jui_hd09bCI_ZW%5nEtPc(t8bT_%1`an zakCnDaxA%5fIviO2ns7#%ONiI!lLTX+tF&o(iDZHHrS1620avH21fe)_dTwW`BBow z+x|RiFMp*n#)NM*L`HjA(Kxf{VQ+p8uYXgfHc%GpC6b3oLNa+JVb4-`z|rQzU~q=s^MWM#arG+H0{fv(fWy@_DaX z68S3}8-}u$BX2dw-LX+NPZ4kfr{RHSMuoj;(C#S|af)M9-MqggJ8I3GvSDM(dA9{1 ze7+$5>%Cf=hTtc@%H8SNm3XK%#%8$R5Nw$SH*URSoK^M0K~k)SFsF6~phF{zN}ap6 zv82vG9?g8bL@gP`A<%$%fes~M2MZUQ8LlZD|{ zX|MsZ)!^=-z`^!3@N{a(YN|3C=10*;t)9igumVu`=z1|$k*QE8pqo8cm0q8MXJk~| z2Y-VXn!wm?;-9PN3Xn$pMwQG@&-QIy6%~X-4XjRNeL%TVS2Ca~GnA>p>Z%$vkRVPMtITS$!RD(F9vIWSkYWXr{epl9oHme;~o0idTM+B=!~Mg<}rUMC6_)SE3}&wAw+}$WGYasiTBnf*9)Q0OY%fm-f8%-~HisiBd7A z!v-KU=R$9sVq~RzQj4gNDktd)6Wq=~jIKj4a}Q4FqUSsIXS;C8Dankus?L(M`*B;H zC*fFM?6zFefT@sGTRewLK~ae19UG3tF}7f^KzH9!=0~sUjoZrD%VAy2vL>0*I)ezz zn{(ZImd4#h4-f-f-?~WW5PpB|orX;xmC+uFB*T*>%x$Gk7*efVi1!NA`DYD47;@S- z|M|CfJ_QcE8em4fj520B<($TA5vNKfFBUZg7(>uW)WOzLNI{BpYKNvY*bs#qmlPK& z`P*kX96xCkv?N^lf-)@gG|fx$kYGAHgk!sdJZ87lARoKI1i-KPjVtYezbu^v48i~ePZ zllZ9i>&XC_C9!6A15;lk6iri;lM=cry(IVMC!&rEUyuxT*1z%P&EsqTe(kVuF3u_o z;HpG#1l$R8|02v%(Gq_sDT&GvVprQ3CjyL;lAIUiY>r`+=&Kv#AMWq|_}HJB+nFz7 z!-lL|9tv$;=DQ{B;_wl*UwIxmkZ7WMVIE}!iC{}i5m@31UElWgPVTLnuiG{^_Mo9mx0->YXVR z#B)WYyHX#i^wL|jvBO5n>RN~Zk8*Hq%g?I_Lbc_A6o&1SEVHZ>;*4vDCvn5IKIPW@TWD_iy@ zK`F527bouoU-dV(1-!^?nci`KBRZ&=v99^vr0F!UnjOJNGnWa>*r(?UG@pNYF*sPS zW$#X2?k5!^`nP9hJ7?soGqdnTeaqz%q% zR}}?a^wa|3ZNJ{~f0srk!ffD4Ge>W6S5FhL9d#3{yj=xdScV$ef|gdDMat1arAsmc zHXkNUNI%Zb;`@#U3iP^!7xQNEONvMpnQO*%Cr1J56FuM2{yL4nIOBAYPi;kIGZnjS z_^75@+Xn@y5`bGM915dpP&!bU&!!E`3*&Qx!Q35Lp@28E6>bcvuz^R8f1OQm-2ART z5}jFS<;-Bo9^YWV0er>b);CdsiI+$E&4fNKM*O>*wJGrywEOv}?PFU7MX^;Pm|y9G zf;DV=K0C2A>)MvZjDn<1YCWLGM&_&2&cMZUr;ZkkTm2_+r{^zeoCgh1a@vszM3Xon z8v|5A{pWXLWyX%;=ui+%5((?7G={IiGGoium|(_>`19u8h+%Cw~gn@!U~G8S%$oW5j)Z`{b7RySOzFy884U zI?n!Yg#Q5YRrndc%tuE_(*v_t{nnG*#M}a8E%%z==4TJfgQS5Id0W9!k~rp&*Tp>+ z1Z-B0)KjXjtWCP5XU%8mJ5!dXScsmqO$iZbPrFCxAw4pipb}CYF7$jh=^)zD%|cM- zyUY{5L@mxz9jw2?`II|~^PH(C`pb3ouO*tB75Tsf`i6syqv^4H5|_#y54oDupc@Y4 zi`17SsvPI+)zZd(!8K^eN20#p$IFX%e{yX+mkN=Zh1F(spzx@O_w;H+p?^}$6`s($ z@$8h}y65h>NWwR#*FSC)-@a-^pkwXUO^uAr0ODzOgV3=z;U6&N$IRfQf~A96v4z1C z1Ga6Eo5slH1Gcsfd*N<1?Y@e9P8cA!`9#5Zv~^)yJ0P>8GPEtfnp;A^l>oV7=QR>3 zFG?D62(pgL3pPl6{n3Ft=f#Y^_XoYC-5=E)v7^WOC2MkFk)HWu`bHzTGf&H*ZcVCC z^z5C_e5_adkW#SaPbbMc@)0iq=WS77owd{^t1! z0Rao;71oKmOOL?P=OQHr@9+jbOb93yQum@`qs(unI6znThP|ok(sDgivk5PWjhg4F z-M!~8eyr?&Sk|Z|p>FcP&iZUvr16h3d#r6Jg0S+0;KvRI%8jJQ`EST-bwrNq<=hyL zt|?t?nAJGe^pP~1XM{E~u-PE&G%wzl!+J7xtNxDk2>SLg*8j8e3r;5vXr?j*WRxWQ zePkb%5$l(e)oV_9BbOz-#+~iRFIw)3!e8}uIek}T+(_n6pC{_i{FSAR(dqody87*b zb869;lZKW~!T}<-Tnbv<$J)$`c0lbN!e~`3Z8_@*qej&>eUN^x5^yTpwIXGNTzq$* zXvUcgp<%b-HLh@*9z44ojo)a`Rd4a`&2>r`Vb{ssFCf^e$);5oab>{RDS$-~*hH^) zzb1n^pV_p87^R~sjlb%v7gK;C)FRs1OIds|3^zQBnC*BkhRm)mb-_+b6q(AO#FUMw zVy#(N7Ozz2X!R+%Hv)VNc~mvMMLOn!PGN!wIagG$Pa~XX?PbjXtEk#aF7@!noc45x z&Ol7yT{r==NOM5`TAt2@bb5n4jfR~&#GuWDEQ^Ip8)zd|iYc*a2^jK{$c597ouw2vxf4(QwIQ;^hDNiv z=-#ce5Oc38g+UPdQB;_vZC;|6kG11qy2Ka?ZV&|3;g9nl$3?i_xdH1cs4t7WZ(en@E3q%l51w)I$Ifc%BPkcTg)k*d4vB}ns5RrXddePBNASm>m z@Zi=r=(HsQA!z}{8I=c>33HT}IZ+IJh>UT0GJT)iSNBowsRPbQ3%h4${bS&m*##@V zG%%J4N*u$D6i2%HAtW&_nDzJu2PFOlYFab@D=G5yMC$ou9y_&|R)lkg^!-Mm<1d)O zd|Ow5AW>|ArQWRua?5^Ba_yqjFXXJLHue|pE;coujp(FMHc29&AaQY776Z4Z!e8tb zVR6|>@L9c3Br%Z>?KH7k#gs9wGp9O*^_`AGqreqTs$dY1$8gY%?gX+cmbz^EHX?|r zkP%WDmmEK(YAu6I1i-6NT0K9r->lSpd>@)9a$XQhPoK}-1Atvixw{feX6-lUV0jMP zc5-*`cHm;6IYDN3eV@t>{l%afa5jKhK(H+&}yb2!Z-jPW8}KiZZAZK&)>e; zE?XemN^~xZ(WwsfaHy@Frz>2blZ@1 zJ8zV6sLO3er%(?-sG1<{sh{)wm@Jd|`dswq4^QGF<^S<1%sT_59R^I(b}e+{I4@|K zR5wyJ3)o-dLtOYKi!`MkuSQ8ktOK9g)%t1dxJB}{rCPTF<>m{7*ZLuwxmL4Q$eYi- zUk{iM7&|Il*Lci|u`YDf@|#Xxg`9uHdpYvR&Z#Gf8_&GRBzz34MQ>+mz9O2h32LNZ zW~BE?x1EDF`>L>gF0Re)^R0#3qja?DoSU%^OHs$NYDpbv@Qz=ZLj=4VT1v86pB!lH(Yvc-mP=fOOK@yr^mTm84y&U`GyP5i*> z^k@`y93-w+)tyc*1LY4_h7{Tn#F%I{t0>df<9#WJT&#A+_fY&sL5uEf! zWM1}t?890O)pAHty}x!VekIh7A(+1qdh+#}YNwHj4ZoyKn4USa`w!ov7ZDb7gH&mS z?&Zyc#;hN~3T0QG?_qr>@Ltum3{J71G^U|oOU4@(w_*_>_l0N=L9ga@GR?n|^Tihy zapBpNJ*0}us0%`73?puk8|p@X$*c(!jset?wh`1edRL&7_!+5P{*2mi3cH+fR*G9X zCwvoHE2=pVw04fKx^Xp9;wduseQuojMiB<6!dl z=5GGVXf28(9#CR_fiHQ*b9!OvD>pjLJ9OM1oVBPp3GmdX=O1;)=ht)^$gaHHaf^ND zcQ!X6OjBJFW9v{_(Pv#5B($|U^Z`E*DfyP0XnW15`h-34lOVzJ&7U=0TOzZ%pP$px z(lgL9q(HNy#Ss$rFyK>g{gsWKcwv^#rw%C3}D-2 zUb~xoM^Q)5Dm>;g)&rj74n;7jf}}RoxFLgu&<5!BcBX#MeYOh{K#&Z`{hZ3Z?PP3` zffPSFWhsgo<6G~F9LeF&9CHfLsz}F=5yVm*UUG6Sj@*7f&3msO2@XtR1Gm~eGug;k zcLLd;cQIG2artlDqd3g)-g}jx4Jks4egoUrIf`c+q*tT6iZ9o}$QtsAPZSq2fc0%v zmZlcpSm$V}tBAI3Jc!6>W622NS6qs&6tOJv6`tu)M5lJz#9SO!$TPbA~)iPgE;)RML8Rw%cntY^sO)2C?HGAsqTQs&DX z*K^^VG74#C+X1IM2gQAdHLDFZWt5b2y8@rGdr~XXvVFKY2Zey)&0QSAyOB zo}_HMSeMMVA;!ExLyJY&Cy23;2~1v~j)X-_m;MJZI_^1!n#^tgxpMe; zrON7P#?1NM<0~z!>*|j#Y#%I}jHEvCmrEu>jd&r-{(dYjGp#R@$eEkoJFw8KC@TpmjGw;i`R8`vmiIQRE z9>dMe{;4IEnoqslQOq}CIOb50^mkIuX#Dfs*8A_(E_V*Ks zt33(w?jDqda3SGX97jx5YRM%tr&O6W$=x%Y`@bC?_I61NM?vAWDwQQRQHHG``zYCZ zthd$WgU-!`t%GW_Uk#gz!=TYT`Kx>&*30!KmppmAy|WhJ)I^qaqI@hs(47ZqQ%%Xc z8Y&!=47?dU?QA3pXsJ!9KW9vdUvB>Afofx(+1svX)?KrL}X+r z(~JjsKhm@eeGyNPV=dTpK6;{Z`ew5^^{eq(^A3CX zFG?@){Nn*5)A&30AHQ0GKE8PBs{=~^)6vzW3Q}x}lE)YAu=SxEC&d2m6l>J{E_(qg z+I3C(+qFv5kL+sm*X#cM`SCv@KL5uB{|nK)n{0u5NB{opF=-E6NSCfedAq;jlb`B6 zwE-KD`lRd$?S9QAypph?^D2x#v}3$5j(s6K;%>F5dX{ z+2&mU#wvY7c$(m5@O~I`+OxuDKS{9&{{!L~vVyyMp6L>>VRcYZU`ZhXrHedC1 z8dI*MTTbQ5Q$3#C>q=k4j}23I8@*1*QZ{0LqrYzUYn_)(*_~znh*y|4YbVqXE4|Cl zrWD;EDv}HfFRiva4DmN&r2!J^BX3SkAI$`O;6c$AiaYWfBEk)aX`xtk3Fh6DnZDPX zpDC$D7MpjUFs%|l&|e=xh-F=Xmnk+V{>Jihx%lyS9lR^W)Nl zW%|k^Ys=g@#U%#Ig5FGuCGehn$#rf{ziLc?{e+TWLNoe4y;SNm-zHr(XEekwh!+IS$hB2H^N09 zB)ybYLL62LMMxN-s+jmH?#V@M-rsvaI~WWKXej@f1Um)I-~10?{RRJ;O;%f*c_IV$ zw_&QX@lnHBB6y~tFRbn4v`NZM0sFZOZ*7p17HDyoY;_YZI_HNm>v>bGhN5#*zg7hj zwRg?SM6t}F-?LzZ!T-+j+(O;fLj8Z7QU6G2I@EJWNSV%)pYT`H*$@0=?AtcbZLQ5GhlFg>;NF`5`3Hi4OrG&b+zBD`1T~?DCw%;HxJ0)mjl-|_B^Kl zzpp&GuJQHF^wrpJ9^!vr{$H&8KOEaW&$O|vf_ALXL10*&h1#u<)x{;Q4GLk!WYq2= z9)9hNcw4*x8Q8m|9<^yrlkMH*BK-%ja(42f*9y)+T!ibcCS)u}T+&;kJBR!G_ut{+ zE7)rGm!x|1Sb1W4*F(|OGPx*e7F#L^jisZrgV8m3%97KI7NtK9OlpBxnWk3mIFOC0 zag3ki>c_M?wFj&W*&4}v-%{%Tc^J)eO|M=js=s^RX^vli^{b_9$Y?D7rsoRMf&x$6IN{@&9J|8XZ+1V<2gasoh!DF{po@H8i$R^*&&@15W?`k!g-$ECLT0Igc3d4N z{xr(+!HmZtsGvHAa2vC6%%HUpnVGVPQC$oLVxS7G&46r2LVfgHjqxDct`;)e5hTAAf6Rg%hbsF{VC&c+egH*gt{lWS)D>*kNv8 z%}PEUBwh^8TfiqS!r}wcjQ_uYTHy-L}!zE~M;enMs z?nyu{Ajda!HEVKXH!xB$_pheu$RDF4c8%&!-Ie_pP>+z22!p z#|M&_?=#LaG5kbyMvGPb9vK5?0 z?ZEQ<*QbYb`3?GpRJrELg$A5pVO}&jAt%qi!B!!X;`W40HhN0OT33oc&~wKR#H27N zEbxk&r2@6bmq}3_q@uZdpKRLy_6ET!+~%^xb57(j|MJVOvz0ykH>j-`riFW7rSDrF zueO}NzU3y<3(lYg`T{wz6r;s_O~ONRwfglzNak$mVy5ia(wRE9U&O*|dpA>!5!ti2 zC!1a}vVc*_M#tR)oru=cSl8l%w1=9?g(&|7mSrya9VG|kQqM?D)IisuO5bM=bB;)! zLX3<42M|ZoV2zLczW3U&HM9O#IJJ45zWw0w=dUL6Efsm$jt=E#W#%Q>`YT-}Cv|@@ zKmCIotZX$1U)8;?Qm@0+%FCg?tvYc8%-ZYlhSzz8^t~Z+>r-FeKMq1$itqgTy!u$y zYVidK8pVl*D9L_MKJjgfoGTq_Y^vrfl6nVlF=9*^ULYA&7adRv!fVJKJznw9-Wl$$ z$pC+q{Pq$l4`5sTtrd0>DDhyeoj92MU`Ew9%u5!cs`E21M(VBW+;@lZKxggvg?s-Bi9CH%QVd5KQk zBqB9gb{Z_u?U9{*MT4_{#HA)ZS}Pza8^0{P@bGxp^bSMK*8x|+2%iC0DfI7=Yu0G z;2ipq{aLwI8#Ou*mQ0d?WLLacNi;rP*7m@R>1;G;m3jF>7h}^c-$^^A82wnjup3k| zEG@+3r}>sIlj}2AL--b-#mB;uB^Hz+^f8>IvP+uFJ%#&bMsCS+sQioLRd8ToNy!P^ zotE5yxDZJ%-;`pF_eITelJ;b&cOE~K@k0c#CjAIo!p zf+90BT`ue82|;xD+1N*KB#SO>W%6#Z^_hNf?uqaq9s5v^{xwop6m)^2k9@BT2zjYT z1@Y?2+ye-PpAEWJ@L=}at3JVcXBR$v_PDN;`6uVCnZm@E<27ksA!M>I5cnWE-I8V! z7$uZi>QJ-#%9tMBshq6p{6yF?Kw6SHR@lw>phEI8zC<)1_-((iz3V~XdB6&utuC`y z3QPH{bX&|DWj$!@emgd`v}o*&N6}0Gd7{w$noN^@CAWdzs@>yXssn|Gb!+8$n{0xP z)d3+-3V5c{3=%jk={ijkf9!@(x;e5lraN}#zRw)t!S!>6?}djJ!|wbt;eXH$@N~7D zhashnIv4)b<{?L@WL~J2WV?3f7qc!XQnXXkM#Fr@-*+_!LRJsm_BYhwx)(CHoeatB zTe^ecVvZ=rhIyW!^RyQ7LYp1LG%TOfHOfIsN-GH~LRtjb0Gv`-iErvesse_Ke#z=o zq$G%Tp1l+3^uD*G4Gut^uPA0_ngQOA<0QB3Gm;waPHMGU`0XUeaGeZRQ09iB(|dqV`#9N=)Vu>(kjPu@Y(5Vlc9(HBQ-6RBJt%S6 zI&JBA4&U&x{j%gDejtid7?oDcP_rTUFlU`2%q#-yj%_iEd?b@`AMj>Hd6~G?+om&I zs=^mT;){9QF+k5wdN~H0o7$2tD=X8=J6Zea-?RD)-EU1nW;&fk>7#jfHftE|lt?Ui zaNLT#kg4%t7_KWdEwj;M*hPg#Jish(;$77kEcPz{#E?E$iGloZC#HVXTOn*v z$53`(eG+*0ZqtH;LuTl=nE!52pbCxF7p1lhnf|MGr86$+bh(T96{7HfzCrUhtBYwUk%Tm*K(j73!nAed?~|GeJ;tn+827qnb&;U zFX(ms7QajgKPQ(#-jr&Zz?fX=fiN?FXWGPLO~_9<2|~@ENJ$`;sJ@Atb*@r21W+?g zr9P=2BgNK}hwWK^q%lT87hG-XSdSVChQmyK_BY;#rTgXfH@2huJgeP|Y z15hzeF~RY2D4sYFRS{Ri78Y2ic0K=bS^r9GZFh602%l$&F2o$--_l;&sb#m>ztD>- z&kcJ*dhy)x##8n3I!HpaLkB96)c@sIPx#LxIN&$S!hDfisP!nUSin_9SEjC#9bm zY^qP&qUc92tL!zI+dAV_bslC~2<2x_g0mBJC=Y7haW8L>=(B4DX@HE3H0YGrNbmh0 zZ+gBf;B;E^rGKDO?l-+^{ZXFqWP9w`LPq(U&_^`hgP-ptS0wOBT-@1)zRTY~zHD*- zKP}>|z=9uWS-u9Mspmqq)-?IAF7rpLmSn>RWl-aewTfjMZKCQ*j;DiK2vmFO1N`Iu z|D`*X{H5$cj`>PzSK_?)X+J{l_1Q!@HV39nXz&5~)prp95LmwP=E5e;tGAW?u0ut{ zCf>HfZRO%uj9J2mCs`LACjM!jEpBThM-NMEe|HeQ#OuI2$NRrg_TEuVb#J$55)w$L zA@m*~K&aBBg8@SCp@S%0Ktw>A6b-$%(3^w~(wkD1-a)zuf`}+ZkRn#l_vZbb^NoA% zx!<{C+`m@V&Q3Bi_F8*C&wS=H=k&#&t@J0GpI-|(PfPJlx`rP@J$X0qCH9HkQ3gZV zt&YxqGJnvil&{$r%q5?7s09>_a?KRR#Qfzew$?`9-S5VBOyXmX1b_sgN|&%af?iNaQcjxtY5R90;73?V_d?aBAPma$^xgu5Vv!_LQ=I_E1e*ke%8{ zF*wr!hNby*B|EtZq^UWoR8}$X^t6LX?2B&{XY)v7(gAL}=7oxEBU%jdj=vT@F(j}$ zd;+K}OIp#e4zZ|AZzVXY;y2!os0GBQ4H!VfXsGV^IS)aHSq!qno`Hp`gj;fp2dI#6 zF-lBd#_oMGJHtA4*KC%72P+!~TGyJ6>qiIZv)Rdlvh1Sb1{n^~&gMF|l`@! zKPuq963&SM}$mej!+Z>N6 zzv&eNfc6@(G7c)7Sb2Szz-kzZx{YPCC4Atpzaxq29wp~0i(}1X!q0qK4maDO? z%*)T+t@1GA%dK5F6pPlN^q*496As?;5>J0{TurNCoWdacd04;VGfiE znn$JTlS6QNbm4?c@QSuN`2Z9!Pe+xFuiJ1O?9`*U$5Xne2I_kGeR&g4{xZhq{P&|1 zomV~=Y}3Z%^b-6T2emB%1M%hhQl;eN1I}47g0Y#xdKnq7XDK_foSeW=nNSwi3|{CD zc3k=L8Iom?^HRHsqq;VFu)>Ha}^I23mBw;n>S$L zt^}8I$l%HLf9`VNAyeJ7?8K^iN!Ta-Tt&BSp?Bj9ysJw~yQ!k#PTgPR)E|({*#K{o ztVCl4&~AKh82(ZNm;i2`dg3{}nv5;c2iHq0TMM`K$qr<1UV+=c#in!#PfeRG^_E$6 zu0F!|BebzY?t({N=JmAp?$(`p105kZD1Pm9XSoy?J+_g-%2Yswq}tr4d4#&#V~bMgw=ce*mN$IlU2n|I$-O!S95iW37X zBk$v!PoqrCD)qgeTD4%c`4@8^+M1{wn#JPUC=OzpPu4R<~wA;v@FQ;+z;)K#mQlXm^;b%_ih={%A6@m`$tm z|30t;nob-;ZjGP5;JE*jZ#P?R;wtz2v~ydgbvsyX>w44^jX^n`-K^xE*asHa4`M7Pr8s^3- z6@Az!kp2=Rt=`DC<0H9z&uy=+o83(W#CLJ51-qpVkBx%~ygdK&<4XWZ%u_z)6-e)) z%*FPydPc2sW?>Vqx%oVE#4bNJ%6u0m>#r-5;na$5{cbV3Yc@t1rZ;`N)BO9>M=K`l zb(fYmylRdu{yZ)H|8oeN4NjA>_|CaJEeW3Jjd#e4_5$3D(ajLVx<8IQ-FQ5Z2fgNN z=j+yR*znWl!c}{kuKj~W2G6AI!Bo}CZOZken_gg(#}%?wEorGFg2x41-DSUf<4$`d zRU1Q3Y;|)hMMkc%PEQk3DUz#DRL{=B{<4-7Db%c(ibh==em;IJ6vRkfdw5!*=)&scVNwyg?Nx6E@1bVeGh-yG(lqWO`mFA`Ho@X;uMP8O{J$ReB)CF3sU+*aI zd2^_)9rbV_&E(lZM=D$$DK9V-6Q)hKXrsk6u^wQB%N zFP4VEboJzy0(hkS2M9K=%YxgwK+`diQSP(YQaE+4OWyZe7u(v$B48#P z`Z#^|{$^hTUh=V7n}-Z>f=u}xRHH@~e$Z?Jz$Pb_*hx_K<7-VVcaJk4u)rJOy9xPc zCU*b}xfueu-GPp-QdKVs74T!pglsW@NLP}{W^JT`o*$0Y?SOIr zg%t|KAfimD0%siZn58K=exO_N2$1De~obX+2c62|if` zuc$r?{BaHWR>yolCZKhjff^XhzcoQmeWxE(wY2Y$3jq#aiH zOYxqxz#>75F}|5Ngw4^$FgmhM`=l#m@H@nsAiP3_l$IQi3ymLwT#nCo8Ef9kjU0G) z08Q+=vwfILCkMPMazjf79gt@z(5C__`SY7G(aPGYfl}6o7)8Zn-+gaY$|(pRMW#Ul z@4V+X>a*VbR;bXZKSTcm5UyJhOUoV_?a%qE(0KH*`HtpDyoyTDyPojs#t#wCHjAsG z`XiJ6lU|i)sJBJYUq@6W&evH}LsA-mU@NvWOj|39hP5!ADcttal7Mf1$8t3s9*7WP zz#?kQjEn(N*j1L3mm-SF0)mRdBmt&~`rYPaD;~@UaDOmLI$;dz0yD&ED-jr}k>fHK zwMQ0%R^^B1M)*B_0|d*Ewj#gEv)V_J-mI0}48*qb(hASVC>g8(C%%G$HF=8~S!AG<=DT-aDFL zk&x@lI1Y)D%h6C|u9XqZi_~Mu7$I!uZe?2nK>!KtEPA){{Lm(=s;LY(1RxYdMR_8p zudKjMjzg3w@R;G~rimnX$vDZvMZ77L9ObmD9^zp=62|ZGW2kv zj=}{fFDAYe6c|KRNfx7!1B&aE28dQ|XoD&`5V54?D44Q%=g$z<`DEmp?=Sz0Z$n|FU_w4*a!MrU_R%M6>9-E31y1GZPm1n@RA zLb{l*ms@iS^K}yP6L*Leogc-JK^9anX;J2DE}^0UpAX)-WN>t5RNQiAlaU7K*D}Zw zx6*XxZV0gP5u)CMdCY%WX8apFrEwxvCq^Pl$#St|qPpwS8NA4@=LW&i1MG`YK+tcp=~B+~)%nYSHd^w30D96eU1sa6 z)zWGr)=(i)P--oZ8Rl*IVX*mU!?I-X!lTD{p-8nq7z$yZs=k8HC+X$e9z~r3A(C^N z7qrJCWz>zt|-}F2FO4)kwF>~h0;D~d~GldCw)h(lR z?qf0}fUQ51bN0*rPPx=qv8Id9q=F&4X;`Y&Y*KBys>~x(i)crIBaK~AB!y9mruY!@ z{WvA4VmWu9mlE4keQ|NK#}^Lfc*|kmU|bF`5YrIXb-oYlA{Vi?X5+9hArM=->nLqT zT77roMaMDD??)yY7n-yF{9*l6b|T_4|2WUfZ1RWby|vHu)9UB(gLUVaa1Ssqf-eJU zG4zHW^=u^@ffzC{>;9dHeuu?s)6;|h%D?zC_iXy^b?wPE$+V9G@3&K(*3w12e&1m0 z>9059R-lB^+o$}8K+VB)duQMSmcjF^>eM-Y0NObx2X-M?v)K^D?Wq* zGQLb`nd;DFqUfL3w6atFM z9aHhV*fvlx0Kj_i?>_PEHD^bqO_%~D?9G(H(56k=c^0VEv|*oaqlui%NCCX? z$~3RdQXHuUjp5BvBd!=CLpb%9>nYc5s!^%qnqHhgqt7OR&Ad%!v-TeY)k&TjI$aZS zdsme0@Y9G0B%zvJ06o?1a&9CWvzw@bN8N ziOht{y*tICE7<41dYZ~U;2HyAI$7Ya$=h33-;7=F$1d^i-kT|?W#unQ!Co$zN6j(B zpBeY(3dAd3T3{o?VH@U8GUal+73~-m$)x(o2dC2YZn4-mHR>3sM0v=EF-cl2u{fHt zNMg<4yCM z&tj24M-YEG3SXrX9f}2tmZ?)+2K}Y|DA3bYDZoM!2afFygGgUI`|R&uvp?sMB1ku(6Iu?QB334}xH)(k zX|zj}RZSWM8J0aA92up5PJV_!NJe=_o6Y!k?37A>%VcMO##Cbfvr(ySFd$qHt}b4N zJ#vI0A$VEQRL56DvRV2|X?{v@>P}vcY>)&~Zi<$!cR_|59%LDQsI4dNzb_E<<@n~r zq@-JA9r*@R0Jny&Q%q7wETBUp1#2f_Sr1Xn1--F~bR7WL%?OHxTI}C0!ZD}}Q^llO zu87C+xre3XsCa(YzrN_*WUJ15bvxqn36()2dhTO?RoblQ{Y>5IL`6207x800%^sIG zj5sxOqkdAk*pr#ThsY)=p8;+m4Up7~rX)()&=*$wyH=xrzj=t}Rf*TZ;Url#{aXk7 z2MRM1>T$jpYIUTqij0L;f#DKO`AGRQJ$XLxaAob&2XMi`r-1dmS(!3uOeVsRANR(> z&c=fQXKF6lTdz;Wi3DKrVvL|=9I6SBQz145`T9v>b^_B1ol&C8->965S$@!7Z zl^-qcvbH!xX4Hj>XFX+TU@|X5lZ#-gjSTF-WS*Xe9cAMXFP%03qaR3?c z(|c`ia3>5=l48M>7L*ZWFu{qa@l_O~JWvn3=VdOf5@j`^S1wpO)G$7LkL0&*y=k!+ z`RUzM^p2r>?b_=0!H;Z&syIt98K%B-|-lM6V z(4TM5yLNi?fYdHSpX!hV0D;CAGL&p`Zp%*PaA6#dbEUTS`jx9*52txXdo zd%AhL-s~xI(!zL>kiLbLTkgawx(+g%L@N4OX-X_ zZqr1i#gSN-5<**j#cx6(b8*OB9RUYHqoE_FKo;ks5Dz;QYBKz(U!3@xjy)oF3L&sJ zkB+s~7wF)9$wX<-&YS*Jj3lN;_MePIw0vTp@qjIux%EeAzBjzn{ZjV*FnC$t+(l_( zl{)?5PvvZ9xj(DCx|^UT&Ajzd)~A^gWb~tc-*)7$rhun$w{{fSbyKc(y(x}}08-Ri zBZAy=-ntk`&|OAcUMkU5JJe?C4ca;;H{SrbPYvo$islwV;v8xxFcFM2i{*q0M;b2` zI_GF}uAkE(6vGRV;W<7}`vo5l)WAgRO{Zx{ngL%9{{he^=&l?@a5HM<9hBV@Pl|McEsw*CNlA!Ap}^BkZuOv!QbZ2_>h z(HiT-3h}h67GOxwU{QODuLol8)?;;vdLf~ulb~J5wg`)VwzC5eno25AV*d?TUeVMG zI%eMl1Qb6^%PF|c&0!#0B-Q>QKvUDwOkc>W3Dn-9L&95(ja%6h<^cg3PqYZAD%cEU z0mx3R4XnPdH@JS*TlxuYTBZ?8$@u>oa)~bXjZajjSUMefD3ZlMWY-kXcbq94Z~P0T zC}b5b!l|v%qMqw*H->$wyV)50{w7_eZG;%~t6eO9e{qXbRqcbwlC^Yu67T7W7>Bu#rlcuIqLq1M6+x+Ht6mLLFqmKu=EQRON5cY5)=h_C#`0BEpF|Nbb$MwQ0{^#&A{jnMCck$)lbo7Ze(f;#DRuXaY1xzGA-=^l2}tZfUo z(=m=R^alUu$=%P**7O`+I3qdiE_|D;$D`a4ifJuZ)l`NMbD|8Wk`7Mk1`nJzPPm!K zdc34o6nianm@KBfTn>df1e_J$?Te&ClgxBuP7k)nG6rh=6K7DeAG_Wa!Qk+Vf;x0E zD@fM1Ry8O8iB~02WWcma48|iaay9*858vt%onrD!NQtWgLMw9Hr%s!p{=Dv!h z2fuhct%IP>wmDCz3q8;&g}fNynfX8XP`k|9()Rtr2>70?006Y_21$uau|->V<6IQ# z=;=C>iblNQ$^8q1(^?-phBFK+k_@;4L+ZFb`4*^d6_e6{$okgOk+DxJCC) z7T}oa@Ag+@Pqwi|b0{$bWVu-Ib!5)Mn^&V3_gMK~l-h*3VeEvBa?b_2qs=;~ zmzRIGU;)g4#bOkBr_A4Uoy4)!WfEBb%}?in{qt7MsS~dSvR>I|z_qHN@08}2kA_}8 zYTl^-aVP?Q6Pl4%&|Hk(P_vxeZ@Q-SDCZx*-R;ivCpqVZkMA1KpI2R{6H%+dfZdMt zaKDa;Sg}YaXYy@;@U2qh)va8bdR+nTj3Wf&FOe^fBtp$BR6QST5#QAT_A#P~-prwq zCXi>wB<;JSsK5RJ-L`M5O!ObLkSlW6z*Q2Q@$YTmu{u*}v(dX(P^C{1A zsT-&a$_!`!0JO%;OpQi4#0c684hTb2jmZ}z63cx(cuk6{rQzm~4GV@6wSz3LC~vRHO1DPm zT4737tkwqo>&$F}?6;plf*&8~K{sgqcM5jwyRUg)P&+c{yQVIG>6^=kPw}C1`k$ zW6Nt>0m+Gqf$JLZc?f>*@iwUla%Pu%CE(P}yLkUoc_K;Lj7qjh(`hGyw>JY1Nk z_was%=3ik>lempv+E2RtN21I!8o%?`_7)fxMR_zF#VIGGinnBThpq74*ZcLawDe=* z177&%TIfPivj}yw{P8wepzmcwEX70h>^_FE4w0_>>7JECUIAMhPx*IAv}p`Y{2v*` zfRM*R6!yGad_J-T1OpA$T&dy{Ou50DvAm)R?J0$xF4d+<4!QYc*OT1CSLPQkeJrRp z{MM>KAGK8fP;4a(x#6p42{T`6t;c;D=0G!&aurQb35QzYpmSbwh1Jh=Hv|5K>7%wjip&hNvbD|Vi%{Rj(*tphF7 z7jlVS&jj%k$6ncao5F*<*K2hn@+$C8U71IA1->vEQ7g|qdZs`}_n)Lj0nPM>@jipACG9+Q*5b7bT^6p-YIA9Do8?~a zDbhXa1XospIKHT@1S26fST^%frgkAzc-lFt_3pf{uPKuDFkSjd7t;*)V5M6&kFq2g zADa-WMtb=u(gPOl3n163twqjmjh6aWJlyhH(}lFBqA21FexFeV(%87WPRYo{XO zGF9yB)?y|fNUq0s&W{b!*u0I##kT?J`D_vAxYO3{ltd@N)S`J5pZ#q5cvMmkv>!7O zrYcO(0}qpHsc~W7$U$iTQwCPh1Sn}aBCwsuZuW7>SP2OGE>)pWv>mZP8r2zDt@ z%Sv1X*<$hsa(kAJHXBzx=GoKJqG$>V&zPH+L*u|Hoglntyv7{uS6I4WA7ShtXd@a}7G|hSuJ>xXhFmthjnb zTG$3*W0@ijLdJVutS#beS023NLmBLjQ3U{RB8ABILK?=#K9AN%O5r+Y3+0~?h2~oqAAQzsG(p*KH7~(uv+97s>g@liw3vwbon87b2jg}V16G|+Tai6 zt&(hwuS@M-4@V;NsRAD)QO0GWjALncxBq`c@!1WgpKjeOEPSW1x#eCdLRbhKSWA*Tp zRc#o(E%3`-;HO(D&PS#?TGXw!_i>pEyVqqxy=iou(cf>^$ z)`&ea>j5sZ@=+7yBiYc8Gpa)4*c(bGsq`(EPNE1_6cZT&zd#n(I?z~Zv)I@2(MrSF z1|Se+>CnGOY_l1@2K4FrbI=-N4#1EK9VA8*B};-sf%@GfqL7pq{$-uO^%sJ=ErcoL z75)qZg@b{7=x*1-l;`_*a65VV4(H}9C2NDB>|QPAEqPs;OuRjE#OKHn08?V7m}G~8 zkVXhB&Jmjku7%uw{}Dl;ue&z2u+nieQZG-#;bwuEWj$TPz8H{IqLINw5&D5%1Ssir zAMw;=uS~on3a<0En(mJ_*N8fp51JtR#)=~>L^xQe&{4}mtP0zfAB_<qN z6=$3rzfUC<8hi*08Ud}h-3uZby29uuvVY8HQVEl-8G#7JMyr!vRLoRBaeM0Nx7Z?S zVW4b#CPfhw@V?w4TI5FCgspu?mdzQCeji|(LamD$C?3k#XaUIT>s`03j9m%L%{i4R z*@eU~cN$d8dHM<^OHFHIeQCu?+F}iMicl&RXl)5r{h<4m!HAVjt9FQh@`#nBl(+s8 z4b}2*9Y383jZYH$iqwV#FLmA=TXwTC*54QUMLs2bb3{HQ2nYKI0Ksnjos#b;O$3yI zyLeTJ7oxPbg06XA*i`!d2PjCYd2p%|`Q`Bu_}*BBlu3p!hj&Hg9RV`@4?tam{5G25M3s7Hw=9Gb9L=wv zANdbpF^913l0~TGPMqiDHWXk(2q1k+sCSu}eO|UKy^WgsVZWw?2HKYYNs1&#nTDB& zxc&pM`aZ1B`g(TSFfrVh;%6n|M2)}cHW)9o2X_7}vi5LI^bE9gCa@2`{wmlrl$N)| zrK#wpPps(Qp-h6csOi_kx_^vO6^e(Q_GGMr;uoKP;{{)`ggR z5GA5Q&CbjD^@9$O_bJmh-|6y)s!8j(IJ;ic5Croqk3D8ksJUOb{Z5we8o!Bu8ofdj zeLQ6hMdlR(irjch=S#tuYf*|Md#iG#vt5he&#cOCR5~Omqx*h7{x1xeN4ki6w-GN< z3#t-~nNL!nK~m~f(y1G%wA2Pr5sRp>gXo@`R6D4w4@7z_dPw$udxp>^M%Qw)?}Cy0 z=AssseVB&frm{t`IHiW8#En^adQM_@s!4^QRJF_;Q96 zg~vSIQzGsFDSaw9iE7-!+on| zLhapKr9#A+(i~H!$v|xKbKd>h`jTQgM33k$ym~kfg{WaN!rS>9rc~7Y4}ir^1JUmG>W+Q_ zZ%>kyNFuRTXr+qFI8ltkY}FLTuKkS$w=R3RY@qx^G4-cs_QtluLzRYKX467}eQaNO z>q>!|c0u>-D140yM4q;F#M}x550#4r_K0EtU~;mm#di)lg~`lpM9Jur3wNjQQ-#tO zSU`Q5nOEBS>WyZk%ivTbgfdrOJL}xlxp;W#lIb;IU*MN|Neo0U;JmT z2D?i7uXjiXG-ChfaX88T9sHN+f6o4E@QT?Emr@-|*|^233VFQ}qa?~!~ zof>ht!Sl-RbR)udX?R;LKZVW^p@ zO$bgyQKtw*OUe`=1N0*w>WTP)rL>obu8Qpbzo%SOqYsV0JTQPZ#3Wb09HLp(jKG_# zmq^Fo*vVIIYq(kj0N{hgva2ey$t=Fd=4d@*^Y}eg^$gJAB8QEzgL=KqK+vmoO7FL9 z$R8pB@sZ@GGq-Yb69w#=)xg#?4hVw%9QypN_;4$(VW#6QQ)CF&Nf>~yeOi)}F-xuM%D zb2y$NW#^&jD~-?|j=9X+fvGLE+83-YbWKjG3&D1UR_ke5I=xv=Ez5)+qaW`vh4it zlXBqb+CKmhmoB);FYAGqOsCgI!>P`0gRcI&2^*bY=D}^imBkVGQz6OxIDDrXQaZ;o!#rrC|*{y#7>#-AWO8?tmq8R?(P#Yf9Xm|XV zk@Lp5*Fy8-vl_#&e*ih+QOAFo&e9KTeLi2^^^oKUuf_25Uw}N9-A>~J(v@t|vu;*w z3cVT(DCiB)NNQ`EAP>2>saB|v`W@v23JJAn+*Rb1chz}Sp5{{PZTdjLj+z1Qh~~>V zSk9mND#ch62X0LO@M}-qwd&N<2kOR30)r^sJ>$EQ>8E{;s;O!jfNTy4fpvWc!%1VM zcl0@i2HsCQ-nqHBc48`;9p{?l=Tpk@Jo?*(DMmVmL^q6x5UvhXO+C{um@$I}OOXr! z343qw;kCv2x}B6qlT!JO4}W+vIzL)_V`muF5gQg;8MWYyi|DHQZt)gsqEFtc{f4P3 z0X|ovv>Bf?>Bb@tYojl+3y1^61i91G_ zkEak~O*ihqU<;m*(og}0NN*?ZvFLkK4sQ~pDFWz}Y)K7VG3-OHG!&qHnXmVwvInY- zFBM4EA2+M|vGH#sg^8#y-L>-jknORK?Lkp+1N}Fs>EfDA%Y9zSdPk?&_?S;!hU(6r zoH>#c6nCtUdRU+P3Xh`_@i_tLz7nN*wY>p1xT|5R!ZXVIU+E+o`yr+uqWi^OfLVu_ zIMYP zS;cKq9r(y{bKJYLuV@nkW>!mr!32EBzgy0Zd&?%$Gw2ModM>oZ8K>x{fU#JPlsE5^ ztEP)?u?n8rlmef|ypOf{MXXVo)e(<5SO2&=q8(|bevg4hFF$B2DeYw0xTt|+aeAWh z;PuNzYF*)W!KP~$kNRkb)Ee5(h5gRiF6tWzjZea^OQ%y@Pq{S|e=hE%64l17RgZ2C z%EX9{x|nS(zjExS5bdmg@WZ`D$m#WUzU|%*_baNY1`}yvs;`IlKD}h4E;ZuRioc#9 zgVv^|V2;sZZEmv{mDX^V%mxVCYD*7`LET{+QMtc5?p}-_Zi!dP#raw*GCB%8tAKa! z48Y2T*#IdXvXE=!6YPq-$5cb29}VQ7-~=llO87jeRk6@!H6eQ2w_)Nm@67|6UE)HL zF}QwK9Y`-?mB2nWpMr=qzcw9n+Vk_CzBHvjySob9d`*BZTs73&0fNEym3-yL*laL&^)#!MnKNc@~ zUrC)MDpEpOvF|=hMV$eMYK6K}51;<*AF(y3R92}NZhHQhCo~h-Un0!Bwa(LslNUMS zF`S02PGK!yw(tQ`drV0-?L@h2I*>ZBl!)Dw;o-*9L_gGhaOgh(nfK3rZ{z=y5G|&- zOPG2m@q3^R>2B^(h4tjS_fUq%LhWEj2?Tz1yG0KlUJpUA$~arU{fQA87U)`3`^H`r z7^w38CIcV={y5KdEdRWT-YeU7=ib;vFhFc$zjn>E2!ZxyVXKw8ZPAjED=aQ4wKygE ze)~%Uz`^vInR88o9Q_ZnQM!Sez%NSte)Qdj?{mp#p1|5lM3882cIKOcI#@1Y;@k=# zZU5pg)moZ%^8Hz{_V)!C<&c=qDvUxcbTdEAx5YeCmg;j=h8L95jtFXh1*P;mkgwLu z$tA)T6!C@kYPJ$iYvyx~S}tX!>C|OfWWLlhU5gYJ>D`+m>vx0CIEzmiyMEQ2cb-S8 zQj)fhhlP`m7~?MQz}4&asjH=ySF{P)bWUC%WUzDP3N!hl7Dw=}=6iZoci0b@9WVc| zC2v?=g1PAkTzAox-5w)&5< ztb?7jPp$PWHCM@ox^xs#j0PRAFV(T%sjN_{Cm1-u@50`$W(v?hMtO>h~DT;q-7(!oyNdI-T5#6KEL zvcg0sCS<+%-DRxye@`h)-!R?(>{e%_?p{y zG?Fr}g-7#!2qJf)Aolo)cW*yZ?)U}wSym?t0inZ{NN@LiMKFRUcn5W?b{I~AzFW|3QL_LshR!{dkkC*eSu}%pt7UdP|x8*vut5!*AbQ+hdlcy+E zBX4`xu!eeMb-GFT#oWZn-BIE0EooZYS+Grv)KNP7r!%drrj9#;v-H)1rwM0LHD{S4zzi3M zr9ow)Y(PcOkx~#s+5RWj|I%*KY{8Q30sLMYuDIhX;Rp6NCBv~K=Q++V^)#RV<+#7i zFEJ&7&LX%5XtcKCZ%z+b1@2WmUC7}@ByOlq-4cT`xV~DE9>|d-zo=+`c$PTuMb-j# zwyP7D?_XjG!<>(Y3JET(R(SBaby;i=roqw%XHU~<3>7M$HA?|6ZB>ck%VTHzTpmB(W|H6lTq;2bm{ewj`thXv?& zMnhAsC|=diVm^0so4krO-U5wYs?;J6RD^suguI-S?4Ko^DxbwLuij*&Q?GhD)&J(( z0jRT8?>5A_rT^9Q*bt(NV0dibZ>-f(uNiO+sGREc55TMJ82p2Mn@{s-{Av6aktP}m zdMyF@6#hW)GxTM%PtdQfZKx;dg`mJe_3+$8sBpEYrF`kOcz5f|Oax(2#|_wj%2A^iT8 z-uhuDV6#_m>&xdn^9W&upDYdgO?J!x-w;QvVneEi>GJby45rxA2-MKmGPt4I4UfC? zkG`cKZl;h;P~_cC*}S#z-HXpt&z_PcOd@5bW=*t{^Ia8?o{b>MP zziZCa+lT!!o?o_Eqs32>DB3?-efY4M*amr%j%s`eq=v}9YC@D(RTi9wHG zDnik;Eov8O4NgiY^QS!HwtZ}eG|3%T`iSbCQ61kC*3xztl%8R$P zxfwLnnd~or_M`z9Ovn>k3bi=&oas-P!0_l-?DG|qkw^t}nanunV!il!f*4y6(Fu6L z=&X8>mvL{=FlhVT!N&dH-#SVQD#7X;{B{}HZ#FZ2Emk@V8FXyZeq_00;IxTtCfh2b zY^j&CeByDYP22_e2o%oI?huDnFUkmpA^9c(Nh4+3^#xXbbOO@ccex%|8?$gZuQ&4K zfn|elr7EXaDV;J7>K%%3L4AGxyfzMg(sA_bp?RgM=lqYldrvHyvRZJZQN`Kz4L@b5 zTJ7NHtGMskDlB(aI?^&1x8;-{gWpxOcl59!<-KMEdY#)xylp<`FGc)Y0Us{de7Qsh zFe>b=J0?71c);Y&O=+t#h9*%Z(k$OzpEih~`aRa)_Qznl`9H{pDl_4uiz(+9@Jemh z3q?L}qgroo>PwMX#HIRN+ko?D(n3lT8oqZmpD!&$B`LO_&v##at5(Wz&%+Lk0pP}N z1#3)B;VuwNu@rkaoRn|8qk}M?ElpRu=a>MyEVLG~vY1^Yqi@kN%vPM;D%_ef8b^hm z1ipZl(+udE#+3_el)JPdDb9IN0s{ecj&B?^p3?*=Sv-O(KMUC#xjR=TID5Ul?k;6b zfy$#`SC5G5rjU*2IxVGY?M=?zdP>e2GBj`0DD>6%vOE-F)4iC%KK=@EartU-QOuB4 z%)>Mi(u%1EHeRpi`+dHT!u9U#oZ{9X;LhR5`U)qGC9?Of6pY!}oNf zztHjHM1pgU{)`pr&#?r(day>KFxh^x;dYae5SJ?hq@=w78)!d&j3~1VGo2Keiz;N{ z_T-IM6V1hdS!WFYOmVaf$Wn(GV!xxuF3>9~i<$Yscdckg7MPa5S31#cl@V#QBP-Ki zgRMXphiCY;+F7a&_+xN5hb&jd2u_?V_`0OlD~tuWmnooqj&!374t4p7*nfcgL|-9K z%ZWoqzN6$eXdgZj4ljzJqCEYPeL24LKQ5-MO$I&ci{MbI`em%9BTXgYSFs(XXP6#X zquzOSj*T5Y=0jic?l0SaYNZ;lpZgcsF}-fhuF$)qqnrpZ>8p#*hEQ=3Z6r~k?-Bqb zh)TJSxW`XIJlm)7i5f}fH@?G+2o<>R+TtIBPM*q<^u5^P%GO8kq4aM z7~427(S}8&`TR{XD3gml>-#rfqv6}nQWXYEf76B=6V+!IRP|J9KcHy{;fBjZ!6fcM zk5^ITgIp-C#HyU&+C1k)QbE)rL4wk|`d6MloOibT(WjG}Rx-(d4xDeWEaZ2lyyOB? zdmUqj$i^bRv@&lsbiKX5XVpwrqLZ?rvNCNP0-Co@ih(UpGgCuaLS< zhO*#I;9hFf`hG9tMFZlX)g2w>f&l~9ZoL0b$6uUyvTiyEGyyh{2|?T@<;4hujW?U+ zsq@I@@Muq)_^2bL0y)Y$W}aR@d?>{kj?#k_Cz~q1FrgmzTileqNuZLF6VQYT8SK-` zBwnvMpTTi6;7k$X)q2BaSbujwQ~MvlJmRYc>B6-?2WNN*Hm}E{J%YY` z+Ee4?QM!W)HM*IpyXyUC;p2O2t;~6}gRS6u5w)~-iMdQ;JewDUzrnvW2*1UE2xoge z_)uOs;28(!OYIrun-3PCFnb7C(fT%9lQCIPF{>S|)xx5Rm+UQr5ha;>X+?2>xU za&&8NYTV3W!K$B`hx;qO*Dqq;GLpR|F}xWB)RtS9hslz_I;|S3pcCuhjRtw#o~~P= z?K-OqjQgUCQ!gMVmF{cf!lMzY?*Zt9p}V$oJ`aEU_EF33!wW=y-9Gwb8ocYXB(1+$ zuiiTSZB=A^xAn7?=k_5&x&tiZTZ^Z?krRE#fB98FzsE*@En(MBHY5M}_(E&q@#l26 z+5*CZmcOCzOwrD@*chKNGOy@1PLSo1h~B5zGp0xhqYwH$en$WXMgoG)OF&o3Q9#P= ztRlf@dBPOxmZ2l-lM|CLuVUnO*#Rkvq4~ntO*#4pE>)?Y<{LD?#G#H9KPFTiu`qQuO)FN*KLV{?Tz=y>k_mA}g z>>iP5n5n=NV+gx`Wr8g|z0%sf%E@fqG$dwYbkA)1y~-onYZn5K@)fD({LC9hDo4Mv zetlRw;%&9d9?eA$*7^Zay(E!thLi10`ik_}~mWF}MgfZjo! z4DV&|%gJh1k!P5u6LyzBg0%$cwH+lvh)W}jikPmh_|1YR?p2KMqxo5&MUq2~+_q2!`kUwm?8R+A`E51j?eg?_R)lb-`Dw1Mh4O@xv=+yhhO4kc1FZFLC4be) zl${+cxQ5uHhm#3)qvB6Bufqc*k8Aw zXD*HV^5KIHz1>^cK4QXxxS7PgeBA*{{B05Xt!3f?bbm3DBO|~$X*5a5Pv1moN@W&Z#H3!W}8|&h7fL>9ZD-t8q1K z32PEbEKI1fCzg&79Hf&jQ92)n+%Zh%FrxslhT)*h_g&jf7RDvD9XDKU-WM)vU$)^7 zqh|flWh_>1xd5h+GDPak@DWl8!iqRZPm#!Q5jABp;*ZabDr<=7)Rqpvo4ep6#WP@w|&6b#Bs5g@Wo< zG6ZOjH0Yw+dEjVc z7MQ=cu*{v2go5pwnuToFPAB3$t+jO4%nJq2dzPVb=~OOWLsWl{SwB_a z=j`k^RH$a-7;AHiPI7|eGcF(-fqtHWO2moM(i-!gTFLK0B~7gQvDotIwxpy!Zu#r^ z{fp0!3LN&ZbS1}k9u<@UfqKd}ir;QtoQiY+4|z}X9aH?2^laRxjQm||(#&PuYh0DN ztuoUp9Q{d-zRak{`h1yrs^8Z7g?sNI*EZd2yIeCO78MaF%Qo2mMm=iXJ~8OfPJEI* zKu3Ru_yZ5q9Bi~i|{9)}lCTJ&mth>jq6D1$vIRs*Mml{Wfcam_|G0$jmYkWb1M>~#5X zi|F`uq4VX~k<%=D7q@O!6!Fu=s9LAG~6q{28HV2+qlDmqm#XL4QlH@ zuo;XFbb?a0pbXR7ARGsPO3_zpzwBi8W{)7CS(?!~Kz&+|S7yNl<+LZ=<53cKgsgj2%sA>Rmd;o8M6;+Ah9nI! zi<13B#gj4n^qvnbpH!vj80@UoDQM}@S%sz$2wXzhyK&Nl=0%w_-Fp5mk=X35g3h~@ z6Ld@@6kpI}A|S$<)JS^lHMd_Kj+uIY=*o;=q`I(?q$8Btgv>=M70nbkkW8EMbGX8c zT-Z^{THI0#iNQxANp&{a3wT2qR0K&ZE5D|}9L&OD-`v#wL~sXZb}Ky#B4HX}F4h8c zN{>qY);1C>%G<-o7j#8{fOWKOlkRnmCsgh-Ej#}MAYZMYV~G^(9-Va496!0_cCcU8 zbD$&COI3Up8)MV57j+;m@FZ2j(v6K1S1^`&A#bMoD zT66$<%VqRM4Y#$|-d@{wK47}|_^5iy`P3$%Aun1M>?>E?LdDE=HGsWPt?j)}mB~}3 z@>vFh*!V(Dh583jkyW{JPYpu4_&z$)!O!+2p_LgY`r+B6VD8lWJAP3p6FhH@gUHoC zfVus`7abg}`pv(8QP<6uw2(!!l`ofEqIi|?vo=Br z-ZCY%ZXa^|47bO5faL>9asa1;j!<-&pGE#O8&YIf=j`_U&sJJi+9&_+ba=`IR<|&8 z7YAGXz|t2wy65pzJ-iO2e<7VNsHZ?;yTUO3Yi7V|K20qKm(5S8RaB2g*MYVUHOwiT z8pYY}k9NW(fC$Vh&`MbnI)9U6&6!^_uRuL8JLS&D?VkW>W-3($o%L*~qA3z&%PMo% zZ<7~j!0T2+(Ov3LO8O#G1}SBj4EekQMJZQ?QUaLu!7MDd2ogs+FzPW)d8?Grm%#$2 zsF*;*WA&MDOO4Eh%ed(kulKO8wESTEK0gTlF{ec~VjX`g)kI;>1f0X9Oz$xWf1;Mf zLRFfWh53+v!i>*4GkGWb<^~8Af)b-{3qC*-&Mpus&~GROt?}^#&jV){%{o8|^+W~R z#BVJJZ=?c^h(6I22jW|B?%5QO$%ohr$}z#*BBKthVa6HAij0G4>4l=BHxX8R#lV+S z010dL-Zlwa`V4<_F-jJoC`z?skTq-ct~DHJvL9rUPZ1s~%d}H~!~-GKdU}i@42`^c zIyb|M=#@Y}Q~oG@v4nqQ_9D2nd8MC>niA9nK4y0I^DhX9mPHe-mor`MH6anFI%<>mma5u%D zl&NII%(_dcrL3uFddB57G+K43{R2?P)1KUCj*$x~<#(+G!nE~#ZiRU^V@X4?GXWzS zXS^74$S--eF(fj&cXM!Marf=-6I-gIiE6U4#Yb<~u9{-yRjdD{cTt*mvTY*L<74nL zm6US)OB?Gz6bFZRg17mZO+|5x7*o)9oAYWnMI}i{hM3ORtFO$y8W>RIqmCMvLi#}p zS|#BMx;pPhw;hVOx)vx@MTD;z#$~dDVsRJlks>7pZ-!dDL<*&I7i>I^;XHr?H|o4~ zn^knhbwxg8EsAXPaeoeCWG}Yj^}nX)Y5|Z(%#_o0>=n2e8=t=5kG|QR`=hMS3x>Yp zP~r<|cyIjgf4!IQFCE!r#R~9usvANTJ0-0Y1xWH zvf+3#P(kLI5+7cUfPXglW1I3tc6}P52O`jyt+IFzJ2&hhzB~xxkRp)$}u( z*fO>yi9t3nkUYb@7*sFXa|$l$)}#g#_e3HbmuXTFnP2g;=)J+*vc-&st1QInG!~qZ z&YIX_a;U@y*^j@Cm^ZsQ9s*sM8EX)UlpH~#fF)ZObe+Tva&lstN-31aJhq@3)p?jD zhQQR|Q>H?~_6%Jm2YG;^*p%yyJ0%|6A~sd2HE-Mp9pSzX_4Dgf7UmUZ!?s9j`xd}> zHW)Bim_ERIpGWz2kXHefP$%@32*d7ioNvI6N9$C+vDHa5mGkrQB;8U_i11*p@Ps$XZjGJ>lq}d&5;d=BZ2%HA!$GY#gQCC zgKw+sEiM3o%V%XKAoTpV)m$D`ha3L}{7u_zKQ)lBZ&)`8@g)kQJJ9jW^n(D|g84u} zi>WjE0U@M+T({5m$HHGf_ui^rHavOsf4hOVk^x*CT<+NZzd!QdzVS?({yVk(NB*OA zG#WoVXX5rkhZrq1LMo+6D_o|eXw@r&R$MNmn&hYN+l2SIXhSNHb=F(QqWKPV`c*E)OB0XM2tF;;p8vk@143)fi@$!pi zzT*ZzxQ?UE>D`d1w!HMxz0l@nA)T*^aFZrt4+Kw1`i>oH7Bl^Tn?{S?=|HA~0u_LP z__1FKVFB_VZ~ApVW2_z(ebs00@%_L{H-SxPNFyqRq>n49dK#m^M9_Scj{x^H$A5CZ z2?QmKQ7=S$MZj5*NE_4cC5w46EY^!rSg|8~Q@bi2Ps6j-fg&6G3Pz)J$eYn2&ZZc5kbz z*J}69qx_@1UKu<*uO|2KQKOi95F4TOQr1av_TE+zV4a}L7s@#JIbDfbw;z)Nt)_|b zUBG_bVQ(F(<)QYFW+i$XMRTgVfY(Uyx|EgoKAuaw@21kpA8k7_du@1yOzY#&(a?!- zr5-{zV*=)h7Fe3z$s83zMMHn39sSwmbJT+Pfy~qM<(e+~i%|J|s2Ii(7yR^zE8FjGX9O1~rX4#*t(x#uUA${H{P{LtQb1!&9OCMlR4>2Tjg@me zwp{rdK_Z|o%yIanft-ARe-GZ$>|^+swIfNj(qZXeEB6aaEs$Ul`~8z`7D*l9;t9AD zC6}`65}}~AAKZOHR|2++nx_QhG-=s3cl{DbAxezfUQ8>kZTzmMAiH`QF(uoyq?ul1 zp2jDy^!xGFfctcsxZgj3$LDI~;{3B2fM{Z$=3rN%k{+?YSJ$j4oZ}|#jUcB(KT`57 z{P=oFZArhA+|m)%ZM^yds800{;C^(fHv4S-zh?py(vUYVzR<3Ce-DbRma_fi_7##C zFISUcGqB2tDp}^aE@|Q*<@{+cXhgD>Xgz+=q>k>T*bte0xRM(lHepG)LxcThCpSV@ALGlY(kvSa#`AG2|4^;T`y!07|4014431`5NJiAHrP z2d{M{dX`vEtQ`sKT}h(oLf`vpn-3~h{D0AAh2{bc5v0cGWrrJlbLXy zo9eC;FMcvLEszY=gE2q(X8bRjbCOCb_Du5UfYd-_!qjW*y-qsa37^%{wr_q|<)tBm zO#yt)LT^jdFh`nU#g`tB!!DN0vTF@?ooxkC6~nRwyl0-G7dS{ znC%bnjS2;<1xwj+ahBhRXe70#5lRk~|KbtGVv+ftjW93x)M}qQ$f{i}0Od|8w z+gOu64gqNav>*an=q!qY&*l1|L3H(%q^ww*Q}wo%at-U}^d7bTFbENfbsiIqj(bq` zsq}h`@jVx54J%O8vc8n(@wT^^^Bey{6WzTpyr$8sxUK96(;{|5|F+3~~;Cm_r0IB($_ zPHB!t7_?|<_RUI({Se_|xZ(F(<%Afw)k6Jn%w~hsVyDv7rb1HGG6ou;9>ee6NV2|i z^Zq9C9KqIsPmC~&^-4|**_D$w${EXWPf;NUVaNIiES!b!B$4M9?|#elP6G+ccbN)sMH~be2I1&I{3??2ly?@^#7n zNIH$SO-qtz?%?IgEuZIucSH}=k`!{$#@3EFx3hng!VE2e>=1xe6u!Bg0lQh`Y4Ws0 z+U!M&x;jjVMMOHh!W`BfXR<-MZ*p^U2VugW^wjpUgk@o<6&05J`r8hO$R7AaP2u7p z*>q8X(9ze}G$>DTHT7ro)P@5YHb1((euVE&RpCnQ+>T~nucFUMf%*Pz?bHSWGk&Wvw9N<+M?n?HAe~Tnv zKecai>5~QXo6Jd0knm2+-i|){Zohc0^(_71tFd}Xp4c$~M(|sRoBB|k>|7>Id;V7L zD>fyXGwfJ;Ig^U^b*5Z}uN%LN0}j|z=9F@Nre6L5TW63RQ`Ssd~_@0cvKiuR* zI$P0pF3xH9*^c@3*Dap`1XU8=*M#`xVD<#)aha!MCR`KB!GVzOcA~bw@HAI1`Ev^MOELGVhhY;^3Hk zTo>lFmeLv`IQMJ#)5GI+IHx7D=>h@Qv96r$V>X@+rj50%>xkJN|mDt@`49 z^BlSN+m?(M=X+mm%HbmWrb9Q7Q3x1JZ!{ciF`m`Aw0cdY&vJRK{14!l*z=CqBri)w zrmDZumtohj$3NO#W_LDhd+3zsty+~``qv+mx02R9ajB?Zyk-7u<@TsCqq14u9pWTQ z#C%+(cp<{Q$t4kcbfxug^$h7`FIQFM(SEeexVSW7czk0g3T9=2&5Rvk;g7LKX?w8i z$EZWAA26$Qs2C>*5GXLLq^?YKYJS_`5sfz_<|Q!wY}&!lTki8QEUD$wKVA;ov9M*l z)eELPHo=Cx@uyMCgfs9jp}uHS<-rVyT;A!^@-PHZ_@U&bCXxQqh1AK&VWFmlb?!Vh zaZoN^KX{4PAt@k)Pw?Po`U?}^uIaleomhqusJ_9(k<5Aa$%lNX!GDO*0UoHwz1$m7kb$Ea(+eNQAL0s*k|=%GWQgr` zM@oGyVCdcW-0yg2HIAU8%2ufe@sd2V&$@z)OL5TdB=L}2<>gH`9Z-VJv@T`Zei$Bi zU=uUE=D88%12z6|LS8`#XCR;aH@#40@F~)%d3kHs>S88?ya_b>vvkn^Xs}5<(*?3z z0?gew!|FPi)zgp#yNKxTN$dZB0^2kHk}cWpluLP2I5pL-Dw$S34GUrN;AgG3`+Q4Kn-b`+78^cUdUje!||-8OhWA%KYN5I1v~*sIaa1(~2(q z)`vLOO{Z@?oY&oO4~}Pt&dsV^=>^~PD8#%|vQOjyE5SRk`(y{{L5Q^Xay&3O^owti4LKD~iQr@%tstI+|N{qO9 zFv57%6J*BkK<@WnpY;-PBag^WZT2K_I!<0@_E^L#;A32F$ky$QNGz`C#%F9C9B{_ZAaFtYOw|M`}!%qunldS6GA z*eK6jux;GL1Oos*5O3W>cA;l9hH&@PtxoY{98}Pqnn7;Z=hD?q`;+9mFMN-n8Ls%O zQ$K3-Zs(jmDKpQOR!{QbnT+mRc3-7{g2XXzsE3VS;9n8Cbg$o&LZ^MlS|JYtQj)Q4 z0e|y9C9WDq)211#g5Eg^0i5_sQn6ug-xu&JFK)bi*Y`Pz4Gw#mJy^D3R`Q={X5{)w zZ+1iLhRE}%JxVk^ghQdw`4ybl*k%dN9IV_jwwVS3!ifnNN|S$L$;J1y_1%#&RO)v> zUtBo8<7_ZweC5pc^48v4!^?)OM?O0r3;)k=%JfN}`2l1Vsk|+m!6(j&OXz6&`B9nE zpLaeHE9sB)&2B|FRtdpUZ|9kZD~LkTG0kM6CXU%viL zq9)nE$1rHWFD#-tcOd;OotO`H{IdP8jv28e-Y1eP3BWLtFG8^8>Y}a}Uk#Zk$bN8K z2t}JMBY!Rspei-&;f(KoQJP(gJpN&I^y25jZxM3tTI@dnJ#wpjlZYpQi8(7Njx0Wl z6UUYo;nyr~DIQ=@W_(%+V!qA9$BQDDo@&#gF)Ss50i?NL2k>4PaNB&=WTdND3RIua zqvE~EZ%k6O3T>cy>NgV_>^~hLuJ_n(pgYZ2mj63B+&8%fb6aMYVXd}M6ngWCh*C7K zcU(cGo6IRl29Yg85vX3*SC_n~L(B4jil;=THV5gS&j;t0mi;XS6eoMbYwLeWddhhD z)PXf>zXs&7I`a>}l?vVJKFIe+QfS{-LBuY*5+QTBS{h?XB5FYNuT8L6Ii8LE8uBfB zZL4(r-e>fZgb0fP{-$;qemAl8hpq^kg@FwCS4c+p3C&5|dwhisSe4K0hTM=dZ{CL- z-8dXR`*i@(nn;`zje;q+av9MJdB!*Q?q%&i=I;3N^1mHD#}s!0^zi0V)H|$P$?K2P zi8140+B^9V8WZlyTJtsn6zyIYQjeMcvcJEa4cKY_@Y3p(#ja3bSCG|4yfN~MAqA$z z0m)P&#oB7>*A4=dm}DHBHlATDGv!;KPO~E;&GZjz+O0y0a#Z%YRaJf&tiq9A)t#b* z*`1%7?dFcXTSPm9s%nW2M#TKs{fhso|E@7*_%Pz%sykie5B_HTm*K(uL4Hba+AE43 zU3goHU|8IY*lS%2yH=H$KQ#Z%6G4gFmkUHCi$ABiy+=|VzRQ`qTPS_$MiRh)2xA_8 zSTlNuU$K95J}sO^lxMAw%qr6I>(NqY$@VC@M<%p=JyO%fP;LXH}c zNg;1q>gMxa$l=d29r8-z^4p{v{l#P7r)cS z*@Y+_-i*rnHXTod%oVz(NF_1uXudp{RmdF{VcHdx66~n^AlHngCbGTy`I7&&kH|Rn zPM&|Pz)RDkqKl|>jm8_3`>@410V9uqGC`nAd6JC2LAraSKl*2 z52;9Q-;9H=t40Xl0Hs)!Y1n;gv$~s*&6&=%|46719ouDcqUSS({cLV1W*u7Y1=M4f zm4AWU>MRpmRHFk_YcbfaGroJ&K}I#6*xGz5%bXmYrNmy@J2c3q6Q>=_`YcJ|@5ldc z`}x1TlVPKQT+i=3FzsTm!R~-D&2Fzr4mLQn>1vAX2c7d=_4|$u=|tO^61dXcRIpw) z6dwn~)2gFyc~!eLk~AYOTP5~2kKoOL7>^qguDAiVJX?U`15$x;g%UVC)#x{V^|_8w za$WJ|8{Z?-cn0HhXUB9QD;PuQfJK`#Ci^M9D~S)ETWG&K@eMEi3P4x8kh|2bKW0HBz%(}V2KB-m#^Snx5(J3~0;;~uEX6NGt6c9`3GsywV2L`-bP5jbDNz}5otXZ5ErUx5& zhuMM9hOvk4HT6yhuwha3T&rlonuFYxi6{=Yp1TUOJ+#%9BgGf4`yUk*RQo*Upr$Co z6-*#}C{Av<$Jk&E4;_nhWo?0uOQ=6N~Sz62qm4!2k;R}Ip^znCmK^x zI>GtBJ8@laz=CDSuR@(A7k9Qinkm+*6vA!u=ELW@9`5tixF=fmN(+ifwT&|B&%oN7 z#0^U^X@C;1Qb9O<(tS)KrqX*mwk$YV#NaJgyTViTOM{!_~Gs{K}l+hFqE+Gt$5t2%XW!3t)%#(>T-5wN4K7l zS>Hr@_IsO1_Er1$UaJr?HOk!axhM2J4(12MX45)(UK&z^Kt0adIrgc-2#3`Ys(g%W zY8}>Fx1PrX>ts2MY|NDQSLOWj){v)j`Oe9){r`#Yb3Y+-@(uJQ(2FJC{}m&uF&P1! zfz>$c^!DL~>+k*{|ItG)PRc);o9|gt|80}g==el@9}&-i{Vi_vA)f$mWMT^3&p=B< z-iwnrBzp0>nv~H(l_{P`AAT&4Q923=GI8nQEahyVS>)yH!cGjiO|2;@2(j~dd~H>;Xd0LFPl@VNWk zn~x!~yXwE+2#OSsNi3s386c#9*1SnHP_||^jwqT z3&Ys&Ixh6?(EFVqVn)ORf39XU|5-;noLMD`BF)1EeF^ABE@hm)Mr6J(%v zTE}OaOu)N4@rb$kcpFEpEvY1;V+B-L)PafE~B78DOm8(cc(RH(Ms z^`*>gw{f&}F0i2FpN^_&@@gH(Zx-)16omD!_o9L&?Ku3z3Fv$W zw4cpp$uY>~7cjZt1eo13n+A$pTBPqs)F_X(S55j=szM7w;$B4Cs?@%3S69b1W%Yji zAX_aR6uVnvI}N!Ue70|rr+7a5FI1}&W!h2ZeQqbbKGUMaz>?Zz)~1RZkcgp+yaA2u zqutWGpcIWI$sWimd@wJSt^Vqj2*5{6-C16WeIp^N*CM*dK|rva4QGEda4bY@!=<2fH*{gi%@aEaHFRx7uD*M5LnN#pk5yL zt9u2M#bjZX(g{w92&xk^;C2gM$Q^Uqd64A^+e#r3?~{xu);E4N8Tp2v+&w*lD0Qqf zh&4o+*=j4k`C861uC2;eVQswGCybES&6Gp~sHoG1%};L*8i$n2I7N5JvKQ-kKv2Fz z+5O0N^H#H5;k$8|B3<`7Eoh$!#TJYeKst+g$ld?*F-_dt7;t4CKRspU51vZaXsN7a zU{QMb^HY5E*o)juWhAH0+np&cX)JhLJJ>9S#|O93o!*+wn`MW| zFVYnhgq9!R$o-nXs+IrAtFG|e88$tfi{yC1;cuJIe4>?3s$ zKDinI`6gv1eV?&uvU^bL>?>bV#?w8Im=c+xEFUvDUh`Lu zvNpf6Y@q{`u4}ibVu9YN^fj|cK&_Twa2Oq+S??CRG!ul|qJ6#UF9yKJI@sA|6yZmP zMeM^HONv%D@RQN}^a-YI6Nfk7J3HI>mlB-eBs9D39svnn8@|=d$Yu^&hF%B}obQ(J zs;~BdHZ0Fj#|HY|W1)zmkkH9|XKax6kQLHQUFOq#WKqRe59a;{@PJ*mn75GX_8hb#gOP%i*r}clZe)J}r-$@KZEs6lG)S7(% zkdnhMbQ35pF7A*|Accs%SufKG(0rX7BQ064Fyx$3E5(|;_>58^!x=M_R723)ogenLfT zmn1cdRNt?Ea#IRRyg{j&Fc^^N!;ykq-If%-rr`CK?gZC}9Gbzrl3NVC>X>}gBX%mY zLMUW-OX<(QQMRZ@JqAAFPs3&VEtlmp3NPTp{o(h)W!io!8(ko0t?t=77tsCw0j6r41F9aM{xwp`qU z2@>~Ls@ydFQdNWPLy5uM!RPRTcuGTET87MG`xz&vts7uT*WW&Sp-V$Y2k{0JKd_-i z1||;=7!U?&DIaJ?@77BC*7gY5o0{v?F()OZKuPB$DCjz8b3-4j`9i0I2Bt}SHPOhp zA%F{$NAA5Q044L%Onyj(N4{ujlyJD&c(KC>g?tS6Z`s>{k2VR*`jg4>b`ySFpW+>2 z;{EL*K;DvWHO@Q^vHIN2rE(%fx}*LH&%V?)U=y2GWE?nK#RR5T;b3OR003Ignov>jdDt#`#r^6$&DB=1mZ6Kk#; z7tstfS9a?+Oh?WO12bh_z6F}3@ynPOrJ0T(*YIHyGi$%@LHKl?1)@&{L zFJ?_>V(vlwn;kMkx1c^{zkR-QMuXWUQe^KuWVSa}9g z>I%_4$RWqKhMPmAsqD5f6;bPD>5Gg!9*x+|p@%#*FUj6M;!Vg?nXl_J6-xDQtmyky zhM`PLVY`>uU}C**Yf7!8tR%<3@^n(&dSw^wybt0Lg~Ham7xS~C_1H}Z<_2YHkU+HcH@l*5CS?$jgAxD};GD7qY`A zfswOGItAzi0WxtCkV6GIbWG!`{$|9IKEZo2`Fp%2^X2z(wmlnYks3|3t&$7z?BpmfBq7a-B+d+vuR3#*LTJ`{{Y#=ox^InzF6BlbwI8d zd9v7jCNP1%Xx6!(EXuBytHn3jmMSF{S`a!k<6bk9RL& zaGz?E;^$SoK}8W=CLI>}z`2B8yy?r&(*}I04d&_-7BIcPOaaL0U!KZKvEl-Gy{P&I;^crH7`buuON0=`(tKyQmWrLqE#NgBprCD|V)*i6?wHj&Vf*3bi9w}i7unwj~=)EPasE)nN zeze|7!c?1UEtC>|!tzaBKyJ)pH=?FY{EYGDdgvgh>nE|{{#yxlAGf2q)YZcHPJBHvu%1iSY6JSNVPGs5ttjr*K_$&onJOx8%LwwWx(_$0$9sy?a_2HYv0pX5vuPjw4kO zOsk_3lTVZYfavK{Hl(_O-c(lo!R{mBElbZkvYZX4%1IN39vuR3ikbSR+edT1Kkpy_+j&&pU%RgdC^r(j z+WZoqR3ie_x$cGjQF^#>!ld2Sqvv5F?r}RsIg~?QeTH)wF41fCwYg$T%0(ZYD=x|N z*&)?K3n9L6i(aev7EG)QUJ-^>_ln9CrbcATN1HKFkg(@TFzGF4%tGuzEI|)4OW(lV zHYFhdG^+&bFS!O1dBzy`;Z{OChd9VSun1@HjMjdZR42(^F#bb14vlefbJ1|62wbZ< zuNPseyXooFz+V~_YRxMF0CbQU9BW2-Wk*ldb?<&i*w#UkS?2h);GCr88Cc9s?i;=@ zJ$8$LuBJJfGdcerWebc6mH-Q_%@ua@+U%q4!t({p+m|zu(ZXhb8z7=PDESdFH_Gi=fvpd^k0|d+72=oj-{Q8cMEG8>4 z+o$)o|1-~Tvc4IeQ$fmF@l>~1Aii*SEXXxmXNBPFj81O9mgHw>{`(K$swo_6*YQ`F zTBF#Fd*)Am=rhjXcj|6EusP+IG#x@myb&GqY%q^vk{pdm{4NRc_TuXSm)pXnZ#B zUT=wxb(apHqtQD(8L<-1%ok}GS`+hZ50w)~%_6+EVatp@1Q|G($f06zE!09_aT^bI3xUcM{kicc@CpRH4=h)}>J>o$%29@dJ z-JElNM`+G+{N{|I#Uoe6e)pfQw)b)?c?784#FI2Psvi6(0mscsEh>~Cci*~G+CU_e zQ}Uj!ApMaH2-JY3SCmr)XdgyvNx+Yrj>}+G2WFFOU)gNx#?hz6x?ms2J4sQH6*zjN zlmX(M1z_9>t@fF26*3w4VCj1Y3iTp_0Gkf6gz%bv;XvxI?ygN{zaVp_j#(?Dp@DFCS>D&2YPBJQOgG>Cjh6DgG0&fQ zgNlo)8ahky=$|&}yG7T}|6?mt*jEwl1UeAr4nBo;?Ep6X+{Y)|(q1%j8qG<}b)$TL zzA-BnTHqWtho0(9^*E~E`E8K~BK#{>Kl=L9?j1^5G&}SsqkoU1XW>Pj*dOV6YzLcZQ zgKJM>u_7M0LiJtCX{SeZ-O8d8UMI=i`=Cc$=S<_~2eo-h*asuP)BI9+d7v91#PO~^ zpHvW%6oYY~D3grud;Ug9#%xjsqpmr|N3fC<*1}or=2@I>^Nns8%4xCj$47cE-=`5{ zfX>bZk0#)yU&Y0F9%ExtD8MOBQe~{9t0>Lzq~zuueoKeLPmIu1ZMcnKD?FM}ii8#+ zF_=HgiiX6bQP@MJA*BNv)7V=)_Uf*8_WPv#wr`;{DsTW2&jE~XlYBjLX<;N>$e!Iz z;K&!-y1U|s@lOGuvyrPJR5Hf!j#xHr6WR%r9%GT(Iimx?j?tZNAV)4h3)Cay1oOtQ z7Em{9-@_vz6Pjc<5VFDEfyR^{4XN48m;>oADz-lfjfgR<^0)J9_mI=LO9rC#5r}lE z?NXdXD8;T0L()m3O)#>+iIQU#%5+Sr(kN#F0Fh|)i8?bJ5{+}no-e&4Bjr^TahY(3 zdkegCC(Y7$r|}#TNxPa@t*a~B>R-&nrsrp^_7&!Q-~UeEb*0Ac^sf|ZLo14?pSRJg zl=OY3l~AfugMlXb=OKC7Jke;Wvd3Tr|E+SIB7~P{P&kW>FzU}6QL`8q*#}46xOyRT zmhcbY@+9x9=z=!k?t{^JFhSUQm2~KyP8Xro#0Q$C@oFI`KR^Er6lPI6*HB>^YG-pDjx0hp%BA5GG?&B z!1*F=Ik*)pZ;ep^nyT(uPA9edUKCmx`^L*KvGWg`Aaw&ZICpBs^&f8}E;x)n96}3^ z@bi6(t(CEMVAO&HV-5}ULYTR?53I&Qcbciv(mb(9(`|HfFZ9n7(BRFVWpk2& zf2qw2oaEhp7ageUFc$HbOrFewb9`lD^%`z{fZd1s}$ThlS7?Tm?@tV)`?b?3j^r}6&icIbr$bee*}8;@`BT^tJY zLg&jw=I(vQxar{1W%TKL!-CdMA(&Uid?!{gH_dO{-xrWMu@PfV<}b`u*M)?rc%S}I zJDL0UYya+5)$=2?L^p8v$2_q%@w(GEF3RJU;i^QMr;@Ai9UWjidvN7&Tfgcc_*&w5 z#oT$vqrCsWeuP}1ks~F|v4I&3tbh75r`WgZJ2}yYDPYXMs?|pPP+*9AoQ*)Not=_U z*h;fJ2#rulr6{Rd{$2&7whNa%T3>m$R6x5vJYF_@UmvUG^02V`;14e-a;Gdg<94<5 z!EQ3RcPuuya)yo>R@@l=e5C-%!^y8hKaLws=u83F+<(`2NQ&4iC;G&d&uWD~vx+nO z^7u`wvRFd3!;gJP9qB^#%jaum_dXmiKfir`^`H6GKLGRZhBMlnf~xPAoTc6;YbwXs zGziN;$xWP`?e>w;AVBXw04D$65dRGHjD(j7V#-C-RbfWDq%7f10ZMm7diMwT@^tBR z_+jm=6r$ZDQtIrlU9B!LJ!xps-3;&60C&`NYw&BZh0`L9^k6B}70=zX_?oR*SKP8X zRG$?5T@TzH%ai!uAKB7YH;J}Ai9oQER{CjIiz_v1u)&1Y!eETE)FamfDF`m3P1<(l zCyn7~L9hMmUrMx|*J0)OV(pvlCxV(-dy_@7h{s2g+7O_~rtq&Lgg5Qe!c3>122rtx z|7n31DzqVTQ^nZaFjv48zeTBq0dG}w?=DxLNn*$&Bn*{BU;fDqW zwq!1EOh#^hFpQ@v2{k^q&&A<75s3!`ECfzGV$NCk-`-OBKBba1okGE& z5oev|!7E9u-SjcZ`B5Fn8UyKt&-oW;#8x=;8qVsR@UNqH9WGh8J)3hom3l3Hsy@kIEHNWBc9_ziBx@xZ5ak;J+0k0t-ol$%0KbS4OU7X1R!_68 z9UbgcgjRC%-G2Zco~PM005gU>F>RXj-FgP(Jqm~Plur>$42y4y+Xj9Re-_wx`@gLD z4mZj)oi^K9E^hR)`YG1(BEltQ=D&DL&9;O0`kO7Izs#}P6SYi&=@^$S0b;aS^l0;a zHraCO_une|`~q|+4A&q3kILRNs;Q`p7EMAZ2`vPqD-c3&0jW|95Fqs4K{^5=(u;tG z-Vs7?8hS@SK#E9_-g{A+2nZ;Mf`VYV@%!!__ulvZyqBFZcFst~IeVP!v)0~otu<%O zMfhKUk=8#1;ku*K8}5-VV*1}Mxou)z7+XCo=#t~!%m~1Y^>-&xt7oJ3yykaJR9!>t zLQCCE43%$M{A#{V2021Y2q3SKit@V~^%}Vj{@%0q{ag=sNU?3VEFRy}s+X4??*MmCZ%5 z8dN(6B*aCJ>Yf!t^ilD^NJEr1A9uS1mgz-s3CoCi!uETnN^e5N{quX0ST?ibeVwxq zhrBgSHP;WOIPL(3$>fA6&z1i6)F3kY+C-PR_B1+w13kxp@>xBOUSLN)STLTcVT6>^ zM|+B;Ld14lqQZ!of}RVbSEa$_SB2xW&g?#0cmbNB_-4q%2i2rNqLGo~Am2fyk;(_{ zQ2RfGWXe8+8P~o&AX;kzTgg8WXWL^n02wZ4Ss1X(4Bo`Xg{7=35u*l;_EkDW z^&(_+drJN@k94-cqasF;n=tgsT(IRU6O<0GN?7uO#r+zLfFfOaO{XPGN zKWqF>+Qc_>2m?E}e{*+rw@aKR`ZN~TA`?dpCaun{@-8tewJGdV(PJ5lr;!aC?!hNS z(`6%Om19{%F-nI62!!70-P~DV(YY}#a-*>fM&hOlkMAxUHT(<6iO$-%=dWqSqKSwQ zObm_3zlF4Sgr9BKQlUX0ryj{u_W~DaS|t*ejLZcq2#q@qa{Q+0JA5+H(q?piQY4B#74OX$L^!o2 z2S*VUbF!T5)b6P#F89fZFQAEsM?M1^{L|;E2mHzMnkLfT&V7jrTX2W~XN(uvCISBa zhD|BmNY{1_qiR?33WY2S=bh4xxgnkp;BEROqg#?M@nh1CH@y=OFWSk(P~9B2ib9hR z{05mQi%4Aou9Apc3eN6>nOf&~Kxdb+v6+7(icCbih|$T2Oiz=RN%~&VR(<&Hx zM`2wFTN>3?zN)utL0Utod2qzEKL@6rZWR}Q|5MI6wJROv8k0y%ucu!=>7Lg z3EREsE%All$?OTu#_Bz%X=luI-jpL!iPdE1z&Vi`#h1!A7gc_J#DM9*0i~L|V$%k8 zeeTTM+`AZb1MJ~-pR7iv9t;SXs-6r$fyuF?j>kRdKbg@y>0@wKg0Tw4Z43A{{wSfs zxft9t7(m4s{#g=j)$UMS9p;AF(BJ>V@pukj)}IW}dTq+`G7y?}ymtKgc4g12tM3=` zXFH;=NukcKcL=JTwR%<-_+ z!$)=;Z(rYTRjK+?7WB}B_^y`yr#{TeK-viO7akT5_NdyugxWoB{M7gKs<$lg;JC;^ z9RgvS#-#XjF#B{o4_5g_`k6|!?rzvIhca|!>0Z$%n`x^=EZ&LWYr7c#ji#_L2HfBz zeFDrRJF?a~@3VNqS^nLCwU)nVP-4CBVFcIkOd+S&Vrjdl87lDF2q7ROh90oJctht$ z*^0P3>)ylVo6EzAAEnnj(-gJO%alikZRfSmD?Xvf1xXXm%TjAFT)S4+`1vIFLj%sg z*g0R?{U=e3C}1VJ6v9eZ9{RR%GmAZ1g!`uxm8R@xgo;5d{{$Uq?BJK*E(UZLP|m^m zvK2&J1eN{sU|QN#^kkk9Yga=H1Gdy4lFP?oSKohdJSDhD(*V;x0toabNLLzx2tzdG zH)9XTh$tPWEsW`=@SB0Px7nj%cvY2TdX?U2I`64jUkChCC{+|rAhL!t)cu>VSvsA4 zPX3ah7Hx-4ZnX6a%aF&nHM_YJ;$XhT8&Bi=2Y=G$Pr@&6)4h0pWAG%@>*>MrW|6TM zVmR37o4)YO4={#wr`eWfx9P)$d;kWJ*toK3WvW%%we@EF=b^AmpV2{^q=MfNYl^fS zQB`oO+V83KC9gh*LgOc@nXb9D-e*c3USntkcITb|AM$GGsmWH)YTo{!R_iSN2w{+m z$|KNy>;JleHj#C!v{{63O*Rqz3jv4sYq+)M1KOVsaYWaU{PdBFc2U(U-MV%mBQMam z+TfYL0Q1YJ>&xa!k2g#7Y#rdTzMP@!~+E-yL% zFCe;ti;2&iA$pLa@X+%DV!B3lB&TTyjD9-EhyQkxr{EJuvPrm#N6VNqKT$PGzl zh`U-^oiz2*CNS!dN=8tbUbQSONJmrgKjhq5WG>j+2$122;0Iexo*h{b(s$)77ws-G z+FN^dIZze}lz4EM$5w@&7JuZ40h|Jx`?&hwWLD9>w_qephr1l`ig2aT;m{FD?jt-l za_W!@JzyrdQDt%!&QTEJVM%x$e1?Y0ET8BrqUmFL{+4I~7(od*Yf~5yyb#fiAyaeJ`qvf;U67+*LQsT2G!!X zN%AUrhI+waNpf}_wS77Z$~pOO(zoz8>@Eo~z{+72$aA@8Z{5pmGbmCS)wDF%I7!x0 z^sd`S!^eE7$EZM&>M4TAHfAdE0!=Xn17=l8duh2*n$CINxG&33n(U)_M*hsqa|uor zP^^+~<*F3tI$0p?xtd^aTn_EiFh}U?HtJIZm1*NF+CUe&8R|b?{RTRyk_ZV(LzBpQ zpZ{R;n-3+!K9t3@GLPP}0_3R5?d`j4^L*xSmjBHaYk%pf*{DI9ngchM-W2~B;fJum zrm%hHJt3-zs;G|(W_IVSV>2k>sK~i`_T(hS{j{{)CKQo3Ezds2m!+x0S!~&GeXc7; zjS4+n4_07C4l-~`D4IchY<@&)h40;h5nc&1*ZJ#_Hfwv=9k$lj-Y7A`LxnBK9lHw$ zcrC*^*RC9Iystd#q3s#w3xmL2!vU|XRwKVdPi3|))iVZQ))&LG&n5O6Es-GOU*u_7 zfoRfMG^hqpuf9ci^kxSel<3Z27N4g9v_`#s9>4uqkHcl&)b0(?wvsnIqEBBPeiubJ zse!9_R6FSjh=cN^j-Kd99?-!iii~f~Jd28}6h~~ZBF5x8n|?aLih(#WOqqP|=zdms zpX?JsJfoDe01Ja?Tki*qcm*k)%hIWIpu2f)n%FlTDe9`ZUiitAoOiuu^{P0Z+mw+r z0YqCdPh^MFig9{Ds=#$Gt6@lon%=m4d+#8SN>zrf7+NswM&&HrYr#tiAY~>Sl>PSS zkJ7C>xz`1);`>iFei^xF#-v@(edVqwQV>-~lK1KO8D_>fW<7N!q-nqDu5#1YlkfjTQhA-|j{ z5*H`@ZP8?d`Y7AIbRPaCfrrzxMTx7T6ZdZo=>HJFJKY&!<70Rwdwf?&;+?QpmKc;Bv*);#!6ajezbK)Y>Q)&mcjf6|!Z67k!XRH%bv9x8$x|nLa zLWRT|lStQC_jD*6VRQCpwOmQ$3}~alp*Q{B$|rcu*qkF za`|3BEw%^4O#&flPDx z7c!bq3K-_=k05Jt3BoBJ;Fc3m4hW&U7*LtxPIe!m$A1?=aZhWEQ#vi48)}3|_s&&~ zm*!a!KJH-1+Vea$7(V!YVKI<2IMH8{`Ph^)wCB zhDn_%Qyr1wUiLC(gFG!d{R;w*gGd=<(AVDW>pgAWyl`U#8JVl_oGXU0a##XTwx!m9C3TY= zXG9m5tE#<5Jg?Wtm5MksQGz!A<`=VINolW|d5rPH11 zBf-a*Ht17dIx)8dPvFBga`5ReU+!Mi8z;q5s`&2qLGAq%eXf@>(bNrR25O(WzRpqm zj(=#~Wq{;%nc4$PU)axOzAG`JN1&cS=ECz7?-KcCD(4i;-=x6ZW;OE|dw<+E#{C7j zZLuTnK5zdO@Z*L&Z{)3hdjXM8KNhb%ez!il?745(R;7dBd{;MK65gTcY?iGMB7Y_k zTu#F1f1%q7Km0vx9osN5nJD{KM~4+F@KFuUNwope(@Ksa%{y>N{$ykUk>=u$U|=m5 z_tQDWS?nQV#o1Ul`@|UxT?b)UJcME|zmuW-%kk_v>QC* z18=GpcsU_2wthQk-T_SoX%DNby~SN5jIeV*pJVni8~;)o6l_qm0ajdV0Ce_3=ZEDn z7Bu=OLFVWf<&jyc0gKLKZU+`R!8){5r9}drk70$Tgkd}&f2})^E==BVf}3%mAOdk` zo3N|wXQpx%;iGDr5zQ>avcG2HCUYYf5Q*zU75tK{yj~CRPO`JQ0&>tSe3*xT9e`9Y ztt}g%!#mshkZd*ns)nxHFVvZW%Ddls0tDVb&7aUn)VWB#{CzhK4%=Z{_A#Zf>~N_bD!WDBJ7c5HTlO#wi@E<0Gn#e> z9>N@!&1EP2%gS)<&H}9M_QCDA%{kr1-y&zqnJbl`%WrQuw%luC%JW^-WssG_tW9{M zoTe`$#f{!+p5$L?Ku|2G4y;1U115u|vvcdr#0b7(Hl+fs&Ugq#qIRi2KNW;X+Qvyz zh=OwBkNf~BJ0E6f7ks@DfhLWZl%Wk}Y3dMyTz3)EFH`pO7++(sY8 zcjE$uaCE}TfWTXRLY;kYSS~F}@p-R{vZZdDw_;d&euw3zx|bz-8ou7+&Q6b`)36az zp(xr7*UK1nm5mHr%*LVfL@QNI?Y~MP%FP$jX&-%*CcRNT&7eaznhZRi2qcxuxD6`r*Tz>%G}$nm_!XKK<}@c{O;) z=1J}O=A*=5&L5uuLFU65#TorH`$jblcVZv$w4S`{+%VDPL6gl=PyqX96JJ1tVQXcI ziNbt7sC2HUXFj+h#5bFQm2%Yn2hYq`GHfqB^f_Z0<7YE|7^JfAoRqnkO*%?dwx*$+ zZY9;s86>=ag&1PVEvRR}k8<`}EwF<7(m7cY`buW10lYre0=D@6=h&EzrfY;n>oY*PJ#0G|6bTheI}iN&q@XghbRJlOIt zEJ>jmA>G$*QPfzFfD9lrm18mRR2|gP$SI3{rsN6|9|yTc~=$yo518(j#X}WJFd81U5hLY_a2lhW8w2$uaa0dJ!28lx$~TST!jel05#hzl z2N`Hq=5#I$T2qFBP6S5p=D=orZ*IvsK0o%Hd!(3ztmZStXObrrAlds`vFf)Rp}q<_ zRw{G!_u@HI=!KE(nwgwT;Mr2ODOS}`mG(3MB=aAL^0Yt`JJ!5sZ=fvd@t~j`J&P=< z;UVSY0RXl@Q8!aKZ730%D1;j4LkU*EJZGIbc~!x^L@-LLoLNz) zFpCrsGxXFO^0;gqP6<2ydG-T~x$PSRLd&2k!%;yeUKVyl?z;2JIXRhEqvDjWFgpc! zqDZXoC%yUAd!Loo!*f!m!!LWn&kr8^Jzo$16>^J30hM9heg7hRaYBby@b~(o*vhngw4Q}-45~f+U?AR{wWzRSM$0u%Y{_Q6R@KsN;8>u5qSbLq`2owQJ zR6bZNJ7QpqO!Q!NZ*__+mj^uLK;$jxaWz)u0U=9mC>!#+UipAA!Nz{Ii-?Vwso)Hl zId6jcfl@DYuRChu2Xfj?^v|CiszQ1A#OBPNtn@mw^H|fJRT<<`E)U+8`l>$}W_)VL z@aumcVAE{v&YwT~f%>vC-w1=0IJ_Hh9pdim&1~x7mD-G zdz5k$a*Vkj_r6Q?;e$ub-So>@9CQ|r)SH^7z6=#(j@Vr`1^C%))a-pXA6fp_WNJFP zO}|M^9+pA~o5iUg{~;&C88f5rGxseHx53;NIgPVUg`PG)6l3=E);?R(Y%$ z+r;0drz!56a6^0ummfUAupC6699Av|@Q00xjOFceWHL9-`Fd*c6TyLLI=yKOLvKV5 z&G#>FZPGJE&wgm70J~mdM=|dUMS4t3&4VkT=mnvt-rO%XH34RB={QDC6z*{-h#cktJl)0j%N+-@YB%K^ct--whEvkU2t*y z4ih2~I{D~nva`3*_xkM|cvuncdJYOZh1uB+lWOzaaRNwX)K1C)Z-;QI@>{&L(4?82 zleBOw8Q-C94dYWSp(bP0;ohG*19lU_>bFFyQ80-M zJ>)JZkJ23_Y0mI!Rj$58#YoG8V8m39+wWlZgN-AzZ=Nz_#~*0!yy`a`kp%W})+74DPj4Yidp)?X>=6&aj1uym8{Ll4E;X>_?!f1@#K8pRgcb8Q^$|vk{^wsaD zwEmyIOjkM7rKNK-_!Ff)Si9R6HCR>Jx<0(l*lGQ&;H`J}xtPdzgl^v_`PC23UuH(P z*4?6dObno(jUpVHs=YTo9lGbW#8Xf+L$jN!w^WabI|Lm)i#OKSDRnCDLi_t@SnA`G1442Aq@=H1_piY$28o|}$hFbc9r*SVPq3(j-X#JrYk1wzWjyiWYWU`t4JTe)>^sWrJ4i@A+s{x*3!c3gVa zjWss1j;yv1V_EF8Ds7|Y$RFzpg{<;3 zXANj`cY8g468PkE>Bx)_oXc5P>!;Ax_xF~f_JCKW-J4a}LJ&@~sYUVv>IxnJJnvra zJKh0=ufYnxAQp?#a=xjZl;!1?CkpZSv#d%7jIMX~Gv{$BgapH5*2*bqTiP6jRO}1>zK2+w4h6TUxx|>%O@L_T+)K_bf{?qj_x`X2`P>KLgT4 z%1gPGKi<;;Z4?UFH3@k_ywsu_Vnn1RR5^iFT-{=gjVn`~T=o`hR*4;NHSCQ`?XUS? z60xESxWP`G-y3IM4POFKvs*IV-n9u1Y(oBc=G;km4=Y&`i6#Kn_z)RVD?7YJ0ASw1 z6*oAMa{b9y^v7a25;Ky;-~2iL3B{B9+L97ELzav*k?!3bkBf{VNFFg;)Vm~VFc3yb zV*xg4k5S)N0O1aIW|{cAND?yV;ZE(*WH>S~Zsy*`ElU5%vG@7L?`0Nq(|)pf2;RW= zpn+GJn2PbaTxxcQteWUM)cx#+PeteRqyQb3bp@F}k%yy9VFm?{je|-FaCy7dOzZ92C-t$4TKL839cqZZf7}oHkdZ<@(5Iz4hz4S@?>C zl|)xd%SX|;qrRTlKjXX&DnEdG&#`w$9_@db9bD}30D$@)kpBaq7p`bus^=eAzNF05 zZ*xs4o$=Hskvx`s@$_z{mwu81CLmk4@ymCG7uQ{~X?va!p`t$imaZkkS#k{khz_Y0 z$i>ez@;mG5@lU&#M2g@4)zgK&5&QLg{r1@J`CsMW57e8xk2e4Odw~DX-oOxc{1+f} zuK%sa|1UrQfQD*#V-@0N2k8pfXv{UlHRfVeC?I3n{nY;~zt@a~F;+k@Q_ne??ma&) zU8g!*l*hsC{m^%WqmBIr3W#lIP9qe*7lCoHAxRW2-O3iZVtV7*1+>VTKjo7RcNVOS z0A#f#Co3G?t|?!YokI^97)MBA$5$&sI$})*JEO|)XBv6zA%yv=Ykii0jiqizEB{oK z028v6R0^a;V;0;p_A&XBk3?F_8VIX-dqk_6u(I(M>x#k({wOQg-)V0=KapmQIzKJ!N-^kU$o~g?fKU%o1P*RLUpJXp$4=1Vtyd}2oC>yqtSG6%waU+O zvBNhEja=LP*wls=?AwGogpzHa=OSM5*ubLUOh2HSUYx>x%u<$?_;qpV6D+UMxM1`1 z2pX_$qvuK>H0}>63f+RzhUA&|K4vN<+QFGpVFMIgC7Vvp6S2ttIN#e<3g1W7UOAlq z>9bpyyzvT8hJ6#&gaM4naLZN|uk5QrmsPCwp>D7FU|jswN(XZZgYQy_kiS&c693Y< zhb56i_qzwsS7B8K*c}K9>oN#rpK&Gc-0WsOAdyEfDR&1(gvlH2a9*HGcy@XZ0FkNO z#ojR%$7q2uA5I`Va6t1+4f;S&Bxjpp#|Zh`E^hM|iy#$+6&X=tGjo4X9@FrZ z)H92yrTT@&n>vr|(K1OShX<_(xubMRv7CBld+}SnBrR}3K1>a2jgTDuva~#s5qhM0q{i8@YGF0ZmqHD5!NAex~(nyi_@ z%o`4)fer6$z5;(d^jb^JP58?8$&h=Sy5>k9uM4hoJ!qM;9cABKb+Fi$D~3@5%n{fm zUH}NdNt1IpZ8(FON5b0Gc`1;+Zksq(V`fhyLkH8b*>BGQ6n$&-6b#ov0xs|*$UU;Y zyg%@g&GmnJi3&_k_fp+vEJ~4wzB2O02RD8d)APAE$xU|JkPSbe{m(Oe&XM|htbV(u zjM?PsRY{Yvs_*8d+%*Q~XO_0VM&(X)A`l-9V{s{L=U@4G?ANPbbk3FO)Go<7cPyCj z;M*XhTlv@ckQJOcmxQF*<&tO|d? zmNkR5%`URAo`BBu8am$6&FjNx#;C;73+qUhMV8+}{m!X=H=3qO0|e4_S`q4%qcb^i zV*+H?Ttv^Qfh(Xi`$M?6GhW0w_D6-y-hktdTK0lgbRvY^*@Vl}!tssHN!wVmq8kRc zyX2YG%Ti!WBcY;|x+@teMx`8VOYfV4f3A~EmrY7p%#c7wm>~+k(3aBPw}?yrB)OG+Tt+CPO5M)7uPHK>OD86x3CL0iupf++ zw(b73b319qE}ME-@<643w^$&}h{0Z6qPZ(<9-ItUtGZPlUHZ)7hAES9x~+K*CmhbJ zg)ddA>j*{}=@=@KI#1xenhD7Z3+}P(INseP?Qc>ba+!@Tl})9E^|lsP6Ucq#Kp$+0 zx>*jV1)^+2s0Da(q~;M&QXVCYwkGUH&e|YY3YpN1=^rS z)KuQDSQ}A!en`mh< zYEYjzJm-XvFa>js#<*~)k`g^c6}beY(Hd_Kes{>z%eBF~Q1}=o>2Q^&eZ`yI8 z`;B~p?%=cc`!@TKq|3JwqaCF9bDdrttSo#s63{xnpehR4O)X(a+{aP9F!Ff!x=UpP z|MiYChlTrbwbOXj8g=Sdg^P=T)DEwy!kzfs=T8poIOfXOwH5ue=x&#XJ=ThJFCLRgz)z-A?K*RbG$?@@#5{BZBMtF*07-e!ES9-uUM%M*!`NkY@C5@o<8yaMyInUn1$t93*7&eDZ8r&fGn`q@v7NeKZHGE_c)c`48X0)53V<9s3*;8_J9$Q ztJ&4hgHM&ONk*_5N;Jx0(G$GHAfb2`3&!`(ei6S4=o}T{(;{Wty`O}Zp@=X@^K$0b zqp0#mv6Fg9J07&yc28PbMX1mf^%ZV8C8X;EQ(BlPhcpW7LS59P2lzVrD@Lgwkt8-s@$~)t zwD!wOMR*wRWz*?SW})&a?}DcCF`X7eUD(`h&5GcBwcO=LC9EI2*>svoDLk_z^uS3t zdw1klz7wN(^&D>3HS;-o@ceJ!`vzsMbng-xa!EhD6i&X0}0Qo?JbTTfcQRe;E2?*v{qsBlOk{gU!*G*WO+wdoq7ksc>m5vLSyr zsc)=ZPV?U9i(%8Yb2NYSI_~AkvveaIi&;@w2D9#mPFhY$V(r{z z)eqWQVm2~!;B*i^er~X==aiW7bW--T*S=SkQ-%a}|Lk3aK>F1;YkVUzvl?Gbk zfGpWvmG!|eqve6H?Lj+rVcP%&zK8YFyQO>n#rU*9`Fo=DneN7i#S)oZQVByz8Wa}_ zIx9_$Dlv{e)#th7l&<)nqFa-Uy>a9(BfmnfmvtX@bnE_9eN-?k^Wdx#c`^Cj{q~w+ zaD@`-pwEunFD9u*l8`T&R!wiPkt)ugxW(GOi!?3YP4bfwo)u+hGqa>V{V*jaZo%(q zTie8Y9$fy#>0#yyv-F9j(AnFxvZQOL3iyM)dpff#ekCNUADWbb(MaN$h!9e@Qsuv> z)*aNcH$YlLiWc!{@W%~9Lt7<08$4!L-f$@P+Dk{u9Z4VlnTP97?QdoFlcmGVWC%TNg<1BM1B?u}?-piKpXu*4@ zG@@fpTEB=I&6du6;!%fTO2jSAC@9U6?m^6eT70=H#H+y~@(>}UH2|2d=B2YQ#$pf@ zXvrMejSn!Yp<3I;Rnp<&NB-W zt8OS;0_RT9(HJf+upv*g{X}F+IzV@P1@5*4w&+|pa)Z8PG=GC%o-7OdSuOQa;o@}w zD(%hvKWfx(U;hORtLR5>nnf0i4WvBu>X7aiANvSn|`KaQ74aL!USuC5PGLpBI z<6P}^aq?2V`4iejxML*H<4Q)Fe}PwuTlTh!mwx-))U zYb*Rs+*_-~NdgnTg^IV#9_vQJh2hFUe7cEG+dG9x1o~!l#AV$dpIL+0Gaga>W@KAI zSF=iLx}V%YW%aVvvlOKP`|tK5;jX49x z6xpMVfRr0s>F=}wh1ZlTswz&<1pc|s1d(BlL(du*l2x(Tc`_s`A zj1p=k-^0G!YjTeGkOi`WQYK_rR(MJ;sn*y(*T=HL+6Yz#oZPtaZ0VAF6_Eas>qE(g zpFF5>X8g7cuR0he5_%UPlO&wr+QL)-0I_d@ zubXOJYGreep&k9B3plHd2S6A?Z=?}>8uiV1CJFJ$>oU;DsTfZYiHOj|mg813w1}^! zubGp~dfDqg$^1u`gkod2+PI&iFBh<0k(CkXZkqiilAmqAM+xlf z=mJLE*AFr@S$cQ(&QH8?mqcln2u+2L*qPs_=}Sjl{`DVXSx*-;npa=p!as4PKk=r*L_LN`=#BzEMY(NgZ7wh^u*o z>Z`E2AC&{@Ifn^#X8NjpXF}n0F9Y-pn!pwv70Tl`cVC9rx&Nym7`d5h1fl3GC4wgX z+jX#3E6+Jo7a-e7B!!DBdDmhg@QPZ0j87#w8|I)h8`JKB|m}Dv6w8kSD8j{1!zY5GlYrrS*kgY$#HP z3ka=ne85nFSvrI=LG{{3cgm1~i1!`xSbADT?)Mr@Wxk_~;3Rddz- zM)tzuP%s-*UH%<4ISnDJ!%Z>vGtZ?KlZf;_ztPKtt#Lvu#wdYtY_Ns0QnYJFi;lj$ z*z8l{cYjw?x4<0UDhI^`64||Oh7FG7P^u9V^)dUhUNZjo@rUho=@Wd%^uLvp*k@(M|RYK zCK^o5{DTJp=}s?ih~~PFz&i-&wl@4IO?lcn+$*8lAK_+W3$&F5Gbl5a2^73&`U_w- z5JOMT5E^5WNPTCbyB{h?bAkNiGv~#Aa$mIrK-z8y@0gd|jg66d{LJoKkaTaw?h~!p z`-tX_YD;@uE|3)Knr=Sr^bmwiGBE~+5K(bD=Z1X67ZTk0Lt=5&pH8m+0-o^P-5?>a zgW5ZOifjsTV$|@_>a$rP15U1%Djh;FN*a(Zv{Be|e7+71saX3WY1FJHsyElMWX8!k zKGIqJ013ekX_cS5>snr!#)$8H1^WQcwl0ugA9$Poj_`7fd}n-nhwd#$Mu~p>rWvt% z@mQ$p(dWe-yUkyg)wkDGZfsJd9e=?_()Y<~t(A$$KqcmfdQ3t@4?*+3)lOm{*foFt z520Rh3EPd0(&`z75NUR;T5zMgWU+yj0CPHWS7;E#!A=sb^K5m87@qibRq*ru!ykct z4GRs85rS==7V&e$?N{u7vtS-YF%cwATpB_35!DdwB`#D%FNFGZ{7i?(bz?n-$ewr(RA9NO(R&d*P6CKKm?nJyMV85XT9w9<_pu-q2)S(Z;d(5W z^13L@Mdb_>`&r1qo_4WWQvdN54LvHeQHvnLyi>VhilIy7MGvV0fW?kw19#bJm7OSb z=bm|_?L001sBGmz$E(&(P*J~cM;gYO|9v~#0FpqLOWL0-=5u2VnM?&Tht0A^LHK`c zeQriUd%`b_8!5jXEfVy+mJ1s*ea*;{DBoiM!pRx=xq-1!obJG_=kz?~v8-Rm1_-wD zCY>3Uc$NL;#d@tTTG#9Jr(0$BpgM}RUYiC26AhFAem1;vBH3J>D`3`w%un1s&UhcP zv*NhaZ#HZby|GE~UOEa()jYoq#}KMq?E46!#rZfjmUdxHWF;}L?w6(za_Eq4RRtYkL3mzN zpaXTh6^n9NcUY0@PkonQKu*e+Rh*DZJ#(YQXe&zQp*ttp<(w!Ak8)TNB+W{d0fKuLb)X=I2sf^I^yJps86Wl9nW|A6Z~Rl^C4k9!@+Fn(+&=s*7YrWSIx`XQfv=Mm7hC zl_atYl?H4)Y%65(ePO`T`Qpe+3g)1rx3#d7aV7jTR)25Tp+3ic?Bq^V(XI|tPMPhd z7W|_YV;fCq{pPjg9cQZw)AfH3#l3!heh026R&8^v_E-UXNZDy4@*ABG*MeIjz>I(b zG%_%+)fglI$9>vM5eamS{yKp7Nlf{hV`zA)gJ$ce8O*lo#|lWM)Y>vdCy*9Z-R;*iENmoT$~$>_F~E>-8NYO5&_*oS_2>qC2Sm%C>KO_mx>nAP-*Z4 zM46o@@zUe`$l?R1xd77PrLN;N92tEoNy#h=v?n2D7mPi25DYOuQpmQ!ZfN`i^fl$a zys}Z9VCQ=MY#L10nruPugt2N)h6mad+VbZHbffs%#U^#h=0#|Pxf?_=&HVc@F!MA` zS1#}88%}_7|GJv7mycr||FM!MZhzu8wT>7YW-m0fDQ9yJdH1=x zp84KUSRhhy8LD&K;O`Zk1pZmPg8I^T&P}Icau~qEUHkwbu=s2e98-7WS^C`~!SB`& zb?Mqq-VEM2cn^3GHA?$1&P1#9b$`gS73VXF$HL7!85<&OU`=Ss^vU&o%Ppp~+Z2_B z{2edXKV07*hXQcYe*rjU&ztt5$f7}olldSjnb_-6fC;wFx*HGkJ1797ts?Atjv| zQ$B%UuEW*C=!_ru3Xq(k8-Xv()kF%-yi+E^!-fhD(04o#XxYnNQ?A}sHp#Nm(Bt=i z0mhLHeuSH6v{VWPm6Rs_y7NS84HGx?G(Uv5%M@=jK^uJk_k%;z%f(0KzoKzJAt5by ziwZs$|6v|@&k6K*ds+2jF>_=JS)ebe)PFwAtj~39U~5RabPzB`j2=jt#J3UhImx{n z#!ac=JA>VQP(Ciu^WcxfjZEVIL%!ty=~49XxT9V01aSK^z~m*{1iT zv)+f{EP3eC-B6>ZLgNsb&dA#dVsVE%Rgp3Qi9K}Bq(ld`H9aHx_f2l+98}d3NnUan zOyemqi@a(d?g&t5shr?x1{HT0OV|m)rlzJYv7jvU8@!eJ6H7Ca1f#R>P%|d|+t8Yt zx>&*YjfB+XyIMk#l=%>UiJ!(I#V3a~HMO_JJSeoSPH!3R{YEjH8b*rOU1Sl@|`B3T(5RU(Ku?sV>Ve&U`Nx(fIOVAnP#U@Fae8S_?uONQh51je z?k6ssv~8nU!zCsm;^q{n0{N!DQQ24iN){~7_!p4y|BLwH|4k7=E$U;Eh`KxUftZd& zWUsg!IgGp#D)!NWt3;$N&UE;7z9s z7P4cbe1+OQE84#RU0~@{BXH%99oyFHUDS)s>n%Ge7h$Vw~|EK{X+ZzBXZ7tl1dI1?cUGg9esW>tHz+gR$4Y<-ze@rc`YIqxj`hc+0chE z7^+dWLErH6ktQ4WsR)opi^_zDnHZ`uXn)K2+NfQAV&XGWe(WG=-zI#2yCjjHa5(9) zmKFHSKajmb6CX5q{+0M!L*xsUng&y^h9mN1Diy~e=1nD5{t_IW$!eFzVtdo6l~w?s zm_PH=FiGOoW(3;0-xJ)T>_GtOMD2mVY+?@OzhTp80C&Zan=d>ha4~S+g$B*EyY`LS W34&~|zm~^jK>rLAqxrZ0uKo`T;`>hk literal 0 HcmV?d00001 diff --git a/en/2023.4/assets/freqai_algorithm-diagram.jpg b/en/2023.4/assets/freqai_algorithm-diagram.jpg new file mode 100644 index 0000000000000000000000000000000000000000..4126aee65a6bd699f0543967a8e65fa5ef169e8d GIT binary patch literal 501262 zcmdqJbzGF&+AutnAV`ORlz<=zC?MS^-QC?S?a+vXbV#RwAl;qPFmyKv(j^QbAo$*c zy5sD<&pFTgKL32*#riGQy*jU4Yrao^Uj|`Gib{xr;NUtga8i@a{wR$ z4haPb2@w$q6&d*!${o}@chFGL(9kh(u+TBEG0@Pk2(Ykm@$m8S?_k~~B)}uY!NbRc zRRV_qltDzgjf8X?4;>91?|%&6+d$YTh`qP(BEV6C;IZKlu;IRUfJlK(0$qdk_ZQqk zMnr-~K!LjrBohM(K;j=#L2&TE5N>^+1>Hdaiee#P0V6#T|5f;}|35BUYkn%T3g|4p zM^O7e4eFcuz-qLA?qE-nB1Vhd+wuREz)%GR2$@ujO!FWF>UpYp>znz#!wkDF-xlKk zm*Bg>!Lj~>*?}}pL_rYT2GM%;XZnfk|Dk)|TF9@Ff;v>*dyq7-BbCFbQ8>6dpc|XB zF9^)F2eVU(YvX(p-g_&N=qLio=Nfd_#@YXW3aU|x6a;Uawwr2N5?dphijYR_wJ?Xn z_8Exn)#^+>bEbchc-4yG6$2uy5|oX8v|+im{0~+vXi2lbu|+1 zz850COEX)ugzF(G-2H$^S=*x`r`f9(mwHv~fy3$f^-z+Gn}s!J$RMttZRc{GQquFj zI=Y5PP*)FsSwpyOxEO}=Njv_qgps0l8G8Lq=l|71PLUI1bH;wz7@y>py5ar&CY`s# z&FKr>K~8}s<{h$)XE_ugZudO+?Y(>(eGI}`RQco5f34=`n}q92P<~Nx+Vm&rvz9S< zR;qCE)1`!NSlXkaSiT3eWVR=yoZ4kOxP!ved7lxN2AV&-?=-sp4+>*)iv-q9}9s4no% zSX%s)CI~{}a8PpTT@S&$wfAsfuP79p`SD|eq`<9fJiN&sJOQ1rz!9MHTuY!JaxY$~ z=&SNanDYqMli1`(Qt297PHKovNz8TTX!Y4)B!917g&Rw1)EAWNuxIb5r*V#Lxv0`2 zk_W6~9fc~@ED?47;rZYGN9=z`(Cm+DB&`YgdE+#rw}RTF|%VE%^~N;0G{%84#P}rvg-&rKO0d| zu(tcSt|x=ZK0EgDGh=t2Xh1}~SZG8|YjOUNG-@#GCbRNZ@!>kAoDWdfU)RHx0*K~J zNhMtdkj4&=+l80((F!#P88>}ieRqQa)dwLSJ9=lgXbU#8ILL&Ux0eJMXxAY)X72>B zFylZ)4r}$|VU}KrKC~;KRF_D^Z-TgoAh`~>Icmz@uRgS2R*?ic?@N63;bt%dr3WF; zd?{W(uX=A~5}x+^9wmEn>KziH4Q?6LwPqlyrv5s1J~W=D$rPT)&-{6gN@@F9{}($g zFy3iN%E1s@{PTRq_2_NrRN1#^cf0PzkB5i5jvTDJme!{s4Jdy;~Z z2OqDh3}vvhq`?_Zo#?dz#wgA3c&p1=a*W#am!ea?y9~M+=Pyue$id|8NpyTu{WV*Q z&mKfV4cJ~u2X`y9Om*mx0XcUi^d57ze zz5|&^t<`84vnM{UqEO;r7}1b4jMG5e;GB$Y>L!idh_f9NX-Y)f{5fb4DC;qEpqf@D z-+d$nt5A~V4^}M-s8cpi$uvUUIR=r6P*Mlv*z8JP z*>E)Rj~L2gBA(*oATe2glLSH;M^wAXNFA}Z2F1rIyVB6T=W@< zF>}}dm1k^qDDxe3$2!SIMGK#=?BfQqcMhVd7zseCUo|a!BKI)zX=7D!&IItEy7hQv zHVh+R>&&EOS&vF0h&A52AD3C^pnbC=;B==ID*C;u4|B`cXN#`=yxjx9AQ0aC1wsw= z%gXujwKLYz{r#@-==DHl38+<=q4o7eIGDxxiV^+RHM?a^RO0=XcaQ7uZ0I{%$uujA z>DoSUrdhz<700-CeB-+2BJ6m|*kve1c%C8s9h5`P&fw#N+?H0G2l2ICYI@knDpuCI zu@<3$4MFot80t?G@hqR}LC5db)MUnYp7gMJPT;nJ zK!`ZT#4w6Rc59gwq^G0%qd~9CpMs*m> zv4Owu0+!vP#9iW$l})xA3vNMz7)n>DLmW1h)#s}VM*#vt*0osabG!LmIMS`^%#%ty0>d1kzxo@&pjv<%xFycnVY!6#ScCr5U7%!H?>u96=xWgoTnORqks zMYz`z`RXdg2cU*SEixQ@^$x(u5I&uB)GVcA>uaXfTI{?=fG5@Wgb;cg7x-HcN^}^| zZoRgur~zxg(~>EMNQv_I9z8ODrAvAHWzd5aiujhTdbwnz(VY^pwycP zZ)E}*;z{sZyuApHNS>=FR4kmIBW%9mA4DKru9Bb(e(7_X z5{Vn%^8r2OSq6cKm7IX0qHi84RdD&N_Fy~6!#}3viK?T3!~ZJT8rDl?7Sj~JDrYiK zrFHdF_`poqU@Eu1X%K{PFerYz$AiE2n_m1-P$@MKi*td0H|9I&nSprGL6av*x=`(r z+)|@gx|_ROW(|CGjrH|H#GmGcJP=HkMv{d%j!JOS{24 z1AoRltF6~xyHQam-lDvv94NTGAAzawIVoC-W>!x-Yhaf#|AsM7ZyHHwhMAP23i=MB?8;*x4octv%k@P9jh zx(0nC;DxOZiy~LN4Z2lh{E+8}6;Y%OXM1lec#9*6aUa;Mq%Q4NEcOirH7^8VIk`z_ z8XM}tGCb-0Q-VK)Oc35eugtv9S5M383=pwRkIGnub>VJX@{J4#EAu-TeKs0!ZeX%`rqCJws@i<-VMqUHRZymMuk!$%ki`}~4*dk8P z?%yZqYkG+f)=j@L^ALSNAo^r=kq1ur7lCrF(xM^mnTs0(Z%U_j*()ka_FNz+Ku&UG zu|Za~r8gx4E&O?qnyeJ9Qqc?`9snhWpyqD{HZ9y28PJ+c-c@64H1|Q(EtJX~a%nR8 z6S=r1CgmDuY`3?uOCpQ@fcSVLuKXzb!|(3tPTFzNT!<~3+^=l-P^zX6QgVI3uKu*C zY@Kyf+2-mwS!gcxQ{OF3KQIOpz176E9v@XYadeU$|KhbVAFl3RmPws_f1bluaqVf$ zOWHfHW{Mp$O|G8L;(yei`VM*&Gd!z>+VB`vaxELJIq$!nbvQ|#7k(01(T;4q?(U(o zX}#D}xo^ACDABkAv84yzs;`=f@+XBPz7a_lMKw%gpmnbSc1}|F#s?T3FC-!5fd?fmdmkLEm znZz27{~ zfzxa$?Cm-K{m%c<4B+1(S^gRR3j(P3Z*JUK`F2(EaMJqfvAXK)fw2z=gzszi4%=Z# zO_frJ0z|bL$!s8=UNDD3A}fC9E53n6siP0_`0JhX|gC)2_$2RgS(b;q)_jI!96ubr{1v;$g`p;9M67Y zX0$*RU8x85;I`r5B#WF1e4HjR-RzAPznxDO`#Xg*$+B|tTFV!ai69WHXZS9*1a@zr zWpip_dV8KozJFYgig@gpnUGV9xv@hEL6GNH3QsO1ejj8rz47*%CkqNnt(J)zEh zz2T-d9ZV{3nKho_|2t3bj;S2(yPF&nX@=#2T+~cNxQ>;{OwYch3PwS@wyhId5sdm4kG=EK#|W zQ(_T^M2m6*POlZ~U7;-C=TDLYS{3DvP_#A;jas^CyX!?tbR2 z;keX~oou5l#Q(gob)z9;C7qV9S^=5|h6^XH|f#8&~6 zsJqBZjbP1mFVt6pmtO=iTlljg)iyzvE)%hCEDZg`WHb^B*+DX%(*82Sw+k4K37qh} zLXXW*sSuBaLpLA4m$N3bs<4E-P4s{U2yMz$@-uIexG7m{D`dis7NtJut#zCRZn|9F-_Ufo@i%GP9_O{64U-=YiFX&C&? zO43=68xz~PeoD-M5gqN|oHq7A+^9(x5?Mj!%c0`5Wdov|+UZgU*~w-WnS_f~M*t7b z0H^!EIAyQhofj!iYz`uov~I_7Dfeb8K%do>t%|tQmHE?h4#d7)kA}qKlLjW1QDP_C z6&97|HdL6&&0{%yNQE4VxAfZCYjoR1_#4s3TLiZ`(oNw-WFE4;4u0E6NfV4_YZu%t zV~@%%WvZGT?kYdm3ZillpZR=X@^x%_%*J#{p=L}i-4khJ33J2tDy<@S;?$M#*7ZfC z^^qbDT=s$wAYxid67#LxU_kyD3V<9?0HZ?EN%l`(@z{H#s_Eu!wfR)>!es@LjA>9Z zZKq${%KwcjvfA4VdC3Qap$(?mR5{xiMsyI>#lAHMao7!7o7IV5FDi_BBBPT=F&0^* zDqnP2cwcg3&PbyIwR#j4THK9=5B_wUu*RMqVxm)%4r41dhV!cwu0#D@E*?1yckeO z%$>IgMbB9h7vcD-72WzVYBly)RVYokNLGWoKDg0*Of8UWl34c=o?pnt@eVZA%<37W zKDmx}+l>w-*#`vr&vF9@TT$6h0zIE^&@wl*ceo@>CcXmLgB$#xw${F~c=f@yKVcjtL<~e~hPZi5a(OAhz z{9kS~k}WJ+6-`~`k-7?9rBA+UX}*#LoN|%D?=JfsQOFAQ{U2O$`*iM;W}q8C;IASJ z_}Vc=RuDm}JnzneOs`Bl1-kP?y&$K`U`Xjn&bePP$A0)NCA8As7gP7dqGrJ6CY?%Jw0b6#2V00fjT*k}+OMt+n1`D0ThO&ZCzL zE4*8Dah6g~fx#_PmUDid>+|Y9>@LfB!F~}|bpk#c*7f{y?Q$@)!^5X%#^;;vDel#K zZy7j&mIGO0l`O@k zU$5Ahb!f17yhm^hy2E{$Gt4yx!=paR1m9i+pF+6sJPw%T6~DTA)drkSzwA!EdbVEB z8e+vrT)HTRGq%)nPaYg^pYJG+>IUY$1l1d+cd>Z1HBo{pj{4d20wW>H)Q+ zgRn+Y6Cz4u<@8}&$7pw!NNeE}H1OSzG{alYOgfe|TjmA4!c&4~xyp+719XxIs|%}# zXCgscTu$F)qI&V1`IS#O!ac)SeXGa!O0tHUUqKH7p*5Zw7Q3Eurm1zDM`M!r98jlNqRTYoRgiV?l)snr@GcIMQhdy~-7;~2o-)*9I`lNI9L)Ch{M`p|1 zvjz#>r{087XOk|zx8j9QJF?Phb_bukkx~Vf9xs$0_pv8UOXyFlq$a3dJTyJWNfxas z$GHtG!=DfCy8}1MFOc|>9|w1+qylzoL2zgBL;0^j7r?;hJmghHEW zmK{7D;~y6EM-VO|DuQitx2oZhQPF~htAaJRUf7IIxeo|mr9fr&m0t@IXKm+ybX;LB zAo>E;f~x<9K>A>EUp!Reu@VG-2iZHJTC#1vQ|Pwa{ZF+=y05asmojV1K0-_0^M3V^ z@qM@(%neqWUgSD`TFO3~T6O$~lgI8=Qhs;peYdnt-O)ZX#pL}v+C^@T+FxY%_Z3*= zW{$ZpK_p-U(+n)Y5?wqU&|v6^2AsO1b}S2+Nc0n(J24BTJ`7ihopFgz#g6saUyGb*z-MhTb$@Pu1+c>JC zFRklv2!@~jZk8sCdpj4hXp0|(${t@V*c%TgCsH!h*Tmz^R-sR*MDNnKkW{cUhdu+h zJ7=<%T5iogxO!1kRvTfFWioMlW+l4#!?AYh=|s1KZ3ckiZUDY`pl#rg-=zbPPEVU( zmCWyyJFX^KJI|55LA}uc-tdjeAkYN}2{^G||`4;%6hSjvwSyxkb zf1gZ%Rm;E!gpX!mFJc@tg4OiJiFqYBiwotklqw_!2VuSZ^2minB37i>S>^OYu%XOr zVclPq|M-7{^sRc=1gr}+*5jRV07)O1Mfm$;(!KW5yBps@k}E|m?1*>wVqQPza)F=^ zAb4Z)4-|gfn8WY_5C}Z=*9O_KMp#|WljY&j>VT>G@=L@)CC?6Y0%oQKH~Ri&rd=-a zcuBrQ_(@rPA~nIvtLsVs9991r-NR3+$9MS|O)?+LXiHM)=uC*QjDt6Z+Zf-yihH`B zUEO1xmfFJ@J}a-YFg`W*w73q4$me~48)-Dda=|5_~3cW{wC?|hADrxJ_=oM~b zj-R`5!=h+L?>EIbj8-+qCkq1A{RB#C-O1oSV3Gc_5(PensXN_OWCSjF1NSvUZ!drOj=sMgy)%}joac>*bEmYyM2cZp?>bwA@Anib!EtFG{JyfF53J7b-|)SJb- z*mqEADOGtJtwBwGFGMbt51jZB)`P!Lwuva^`SZi8>78mm8~}9!)QpKY>X;q~ARG*3 zfE;(X_rlikT6HE!4L*ROv>{q#{A1DCMh{KcR{POoLwfn3d^ zA5;2Ub|fjbi5Osa{QLb3@4i981*bmK*0dyn_z=uq%&ZxJqewxG5#~&mUmsl^f2pu149@L zH&cz-*zBo{1vl*VO8JIJA`Kilo;+w6lGO0?CDX3zp^Z(&k^IDM8&(xme{C5tzc(IPx_q!tlA2V{< zNBx5-Zie9G>QOk(;t>BAK2+%N@uQ(H8L6!EO>;QibmeIlXGjxh8pV^wPDA2p;QL&@iG%r)9FonCHwdT2k;B z#fdFO+Z85EnDNdtEa7ex%BJn7TWvmU7!7TNu6F!>3<13-Y!`qqOVS1T&z#0ba*bQc z74F`|;rpkn-z!S99S_eI=8i(yeomE#_1B0=thA*SqkaKJU+kx&f5-mv&+{^5-T<6f>K zq~%szDLP}8gigka$eJ=zmmQg9l@w^%5ga$o@BsM%qie!Gl>t3;;M%qEjGbX4ZlBd; zK}V-k%0{I^v9@*#68aBJ!mqnr{N(l*7v22*F%cB7vEQO)s)*AA5EBeK`#O^?Yc9~Q z_Ms)2fQ`UoZx1JKk59l9Er!`5C&F%)Ew%0=Q4C6ic8_5By71knD9zPQkJ8x1B$N&c z;SeDSD2*|tDs|Pg&u2+<%bpvtx}H3=?nyXv*|l)1S*b63CU<80w0=gvuyik)VpJxN z`HJLAP`Y;eC!#rPi9id6!H2=72RyiHrk_j!9B$}z=5(niB&zFVvXM?k0~mZaBa4BK zV(4+vqulftkQR{G(;#S}X-wEJU@>s&8YjNov)k175P9&$t0gZJ$=2Dff zr^h?nBz#3?74~}H#f$Ptr$>gH#6SldR;F6Rv|rz?wvmpPzo{2QIYL;I0azm){<@F@ z1vLr#M6yEb4Fz^hysoyI-y%3Q9T_TyaOU)|7L^IviDZ^9q!NX;Js&KBp<-KQtB=Y* z?ef0gv>;B;cnPjYkv^S-JP*}eBmd98mB6?lN2z-x@(zK#V4lK))2BJ(o~4XCkR24 z5=oYecBdVxZy)t&gl-xZPrZmvCb>9})=O1(N-0}qYk4|PL}ovj*mU%eOVWm4r87LK zfJZD_B$YBhfUQVH4D;Es{20#1nTfbWlo*wT_a!sRvmInSP?8#FryLD$?n|8CG^k?> zV}SL72s{#>=8=$0>e8J7L+a#mEh0Y_1?gcLgf;E6tJ9fC^H1?ymB3S9wz|G3PpsFv zOy{u7pCiN-m25H!1%oe?J`}>BOf(r(pc?t+Ov2ydsGoQfJ?#NlyMgpWy2gc-(Om zIeDgN*9fF_+45y?iETvM4_Ij#!fw8|e+Atm5I1STFUUs!5ce~9SAuub_su93W3_gz z!&aq{-RcqA!iJs*cnVNq&51P4x_J#qSe-r{MH|r%7H*3RdTTp^KO~cYts(#By|0!A zuXuI7`-b^KYX~hn_bJc{6;O|ok3mM{zgeWa$ zfRtR;#B61L-Uo1Ym(kWDv$?%rJca}|@RwqxUca3Jx^w3XP{0a3@wI6^&&O;IQ@FGa z2nc8rGzM_4@X7xM`w<@TrjMlARf(Kk*q4weET)hdxxD~~ZGBu9NK5VD9fFRgZX|&( zenfHPBL6YRkIxlwOxEZ(Ikq!5tHkxc&2dwuU%{9sa?;}4b+9Sho~0ZUM5;cjp}sMN zaX-Yank(&h2EV>!?9?{QhAlQHj-+5n3+GrR1tRp!I6fe=p!G37s&&+mrMiW?P2Tw- z^_uVnR@aC`X15u)WIGK{5~i2>MVOAb8(7r%u9{prgK)&D zReaj)7aN0_U@~ySRapWD5P9riL=d)6aGwNJz1&3hy)CXiFOI7bN6t1z<|d|Vn00QF zkqpcx5$A@;g2tnh_Pahngn8;RF-1KxUAj&^<=Y?|-V#tI-lXU``twwwOuoDC0^1{m zS0)ogT0RuGe!is4jl`Y36su*}_BJ4A9DMP$j%`->K62JHS1SmVY2)^_3Q(;54L*(u zV=o>i`U2jc@$2o$c^cn1Chm0-rK|OgRmi@bhMKu*7;E5lH_E7u>s&icc=Mxjgwu}R z2h2ozl<037=kT20lb51yU<{8%?#_6DWi_iv(;SLzYl%w3q*;QVD~^s3*g`O$fd1{# z4{j=3@wPg@u35}~Wij!|+V$G`SmH!t=C!Yc)SFzD$2>IYEM-A)s{<}?UVONRl13+y z$f>9ZKcG~>YLoiR4OC+GmR-nwWjq?*s{AN0>oyac1|`qe^UIL{DY~0xFb2YH3;;}8^;rJzqS2~CR zl6Vr7>BX32378X~X0fB2o{M^l!Oxh46F#h5u1^i<-%ggNj;N{nTGr*q|m{=c}Y29J|Xgs{g{@`oqo2RMe}2Y!Paf=`iS_4{~krHl<1u!%#nE)NvKcX zyDmYEug;2&4qQ}E-AF6k+wFj&cQzytsh<;?B{tplYvw(4-(1^>qH%Rue$S@b^wct& zLs#M%zzOs zS2;QfJ{^CsfK4K7ktnr6fQ;%_+Kd-F3kbY=60u`p{VY5f&A{R1TwOCHvgP#w0RRWx z+}wd|Cp$HNOTLli$N`(U&6+5{<(VD8Y~p3~m%=5BOxVOIZr zov0`_d1h<5t-BbA2<~NN57@V|Ge};sYvJI8AWal<@P0W|s<*sfHQVug*<|+PAkZ4@ zg;U)PTH3M7pY>5yoHk9UuEf{5-XWLUnmil% z?S}00r|~q*-8v6GE7cOXaQWwy2UO?Qy`#NjQ6t22)SY$woE)QjhyI))@QL3iSz{!d zD1i`Cdo<+QzL4ppqr$~Kcgenw3(K*0fSZmU67*Y6;+5rq(e~f4gV*$@-EXeSHZDF+ ze8Qk55y;xNBh!b?J*Ryq|KVkvJwPdM0J;63wR3x@i(9|q10y%Rofvp&)M+1f752J> z3%1@%9p^iv&FOnh0=iYbrJ7dq8axJQ?WIUurz@LujC&D_&)1uHr2}XYSWE71%}vbB z8lD-4le7YrBg?tAXHlU&{n3#9)+Uczvm^cdJY(P+=-37SQC3q)Mn%mm`o6~kzGu4X zvF_4gZt25|9hyQW@ll2Yw)sP%uv8P?N#XWfPXWb#^~>$=pge-dah~Whv?gd(RQ5a% zE7Qpg+n~r9nTTz9W&AM$Om0kMMnVUY7qmt=RnAacdewo~cjc1yzWVmvWeMo$J)KKE zw%`andG_Y{WeJQg&e@l={S1@)aEpoyEa5b)Y586o<*QSj-}5Cpr^VuHd~4afhv=>M z1p9IM$BY@%F#dW)3Wxba^3GUv)pB`Cch3)DhSEsRhMpMV7rb4~FIa(*quS1Y*z5gu zm$by9AjD|k9joOo9Q4h=SpWL=vN5M~>h`RnIN^7Y?5Tjo0RDV3?uR|6<$J2+xLhUE zh~Ct5@5)(u0kJHrMITn*zhOCV=#Dk`y1K<+qky)cEp^-LeEe9nt7w*ho@!V|9Gg#s z?ZK;(b6o8NSt0{^It`X?e=)RYsVbI4DOql#&*CKpB+fYHD9wwXA`O%by@{SM%4bZI zC(N@mQ=Q9{$PC>I@e8F=)bPaG95J?`m@^a&u#}?3VSOc)Iehzyo8DU#JT_9f-#WW} znyTwiP{nEgMTKC24*&4mV=3P0DLaX8Q1j#XAr4G&52(+v`yj%|x5l~S<6N^OF0#ap zb>^+^gzj82BeE25sGI$-Pt}3MwT+&G9h3_DYcsrtS%oae=8Y@$B@}zgxy1TQ zc7+^Up6};`JN445{(mcyw!o+KFkmE z;b-6zHF_?O5jct6mtTxw$+sTcQuMAy1Nzj-QEVF#C-+p{9bUEx&v$Ca7B5!Tz^Ams zAMoDkeoG=)l3gX^$_km$-qzrysSn={H$3Vdf0j8Zs$*bG+p?xWO%puCo0dgndgxC? zWBDCqF2&jDw3+@TJ;W;em!l_9aAId zP8K$um`2d)84`RiI`D4Yc2+KD^A_2M$V#2{8LXLmlYS)8js=WQL~oOuNIarJcd3lZ z(pmoss|;W|?8oAX$d}Arr#5aLu0p{Gzdc;-)SW&QsOJ3+(kj8MC@E;$?cd3j60^E9 zMqro{MSzOVh1y~$H?jNWGDP>7Kvu=Bnr7Q5yFo{u<)fN{;(U`0dCrHJA zoaJ%DW?G*l^Otb4_73^U&9XkfMeMj`HFgJ@J!w^WpCBMzU&X`PW5s_)yD?bPg#E(&sS6eqDN)~Avg`fv ziMZp8gfWeG+QBaC5Ds0(NA12r$mgVh2b<*<^Y>alO(Dr(Bmr`nEJjlmZ3ANK|FQ0BXjDF@=t&eoq}ZLhtfZ8WV7 zF#~IsV)x6Jx2TAv{1APeW$nddYu-;q=SPEsR5|0>-v*x?m>SbSB6@D>{!gSKZ>7Z) z5J}X`bj{TKXMpPn75ioWDmcej`FmtWkhrk-!)jxJR71E@-ch2Tmf8YhW}jtZ3gC~k zd|grcdU^n#Fh4Ux%|q6T@5NY&&sI)v7dI?Sb0uPtQ{9ZE{U-9Y`Ce#rOGkx>QnGPy zJHZSJ-YJr7aNsDv?Qs}SmDT{+cTmCL=61hXTvKwWv){fh`54~>z#4+!#=_EDJ0a~~ z&FUKj>Ghft*Yj`C04YhCT-}0B29SmHrN`IQ83BHuf7vB^ni|f)E;;aGmvMm`Q9FwTi9rZ7HYV$y#iCs)M=!8%fvB&{=Qn zVCe#`mwXDmQF-ulG+HyQjhMH01-0l`I{7w&-2B^TVk$h2!2;Dgg!2v8 z#O&kK&Z0yqfRm=e-DT7VafWhWpY0pK z^nGw&Rs+w1_(O}eTLE(`sylqH$g?eS4mZ)%8}0qX#w{@1SV-y4lO->5$0v0}L+OG* zm-r1$(Y?e5{q#Hf=m$DAIJx)gqi6Z+-9l5m(!@y8Jbi#X7-{pk>uRFrKoE)CS+Pc0 zO{j6?fMsK9GET>0+sAq==^q;@s9o+` zO`9Ma4%Xp6mZ=L@c6Uc^?1hr&)ZsO&H1h6P$RgR~*Suy)I36x~v1hQpL(tLlh+A)B zVkTsSJJRRoCA}*}A;U#$s*SUjJZghk{z#0BHin`!>-J1ubEr(gSbX16FYC({L9`{k z(7I~cBpaDFLPw+EP!tRfSkR)@cCcwC$>+MYO8r?aoLMof2IHW|8MYc^-`y9 z?_qjHBY=llqb5E*;(4N^+Wcat*t(dSEb~|n2dzBuG*)-gi+*7IGYGj~pvfq{e*|`L z2^YRnkUMf;wqg`IGW)RKAzP*Rm!tFb6tKMy=(9rfKlbqWPP{4Y;)Yv_Dd}MWSjIbX z9PM;LEYn8J{$B?5Jy;Jf5H=R>!}o2@5htq_)!fnTFBX$~0UH2kDk;WT3$PzRgy-1N zO|=COj7J%Z=4XkHJ3CTLm7Ajro~t_L7_`(<+OphM(kl+HPjH)^7Y%Bc#>!vysHLM} z;~r4w*A!MIxWOxokcASETR&FTBw6MWR({nDu@Kb`cnRu0`1EBDXaV*U2Y*-?l;`uX z9)J*4mANS^ux`SD_jjU_wj#1l@qSRCSHOrwG(BPO7ygb!DIA z;)&7UO1WimD|SzaRAtXOlXO0?4^z#q!jfj`7OGs!43%iTY2qJPWi#)LMVpoozTA}| zEAVyA%k3{daDg-7EZk6_)GZEEic9FjHGnp8rqr&SB!)6CjAD4OgfDaRybuH-)I8RU zc5TfqrFT;14}p4T<}LVFOj=(wE)P-glvkab*$rhOtc!Aj5TZKcr}aID2n8HPY3JXC zK^mQTL}=+o_Xim3WafQt`q#1FL9{?b0fbTaJ96Jt#*_+#

VY#&qkShKkOK!m0N(UQ&_)rD?zMA&%o8Jb=bnQPl| ziH6*)2A`-SZs~#Z6{!c$pgz#jFBwM{XRqjfg0|(0|99t6^RQsjs%XZ_MYUV2VAgVF z8j9jE+hG40FEu%}73v&suq!#+xwpsrDaykIPq_{9*(!f0&{zFSm0KUJ+UB7e<3lbc z-`%xr#oI2yEBDufw|LzP8Wt z)0{4)?>La^Vqx%>Vbt_zRYbz<-S!wQ;bNw0fDH={iq=2^y5L9H_o7k1#V_D0ehIDJ zyPSSfYqEmj!9_HDqbwWB@>tuPy-}7emhdDe^On|{7>4hswLc+$!cHg#A{T4*9QX+1 zVn8T)ce;b7X^Q9bZ4mT-T#xX$40YyQOTEFHNcUYwk(S2h{PD;3jkMbX&vg2F@Z}O@`&=#Ztnao{aA`&BKdqmyK{&Z&+S(`#?3T`4qgLf3k{oNgwNwDO&KYkCJhu>>T$*1nNAUbZnX_=hiysS%SkN5xD?K+ zM6t~($gbr2a=@amG>y+qmo9ofHsZA3wqW7nsq6AF5XxJ3>{nAjeSmQH5lJ_)iz<`C zNLcb~F8d4E&pN@^^|nqEqw$wkOEb(fwKD*#^z}vi+Lv7j?D{AOwA0OE4oQk-Y7XDW zR=@rZQo0YYxSATV0P}LqF3aXJ2*StMZzpq|T{LV^_X$NI)1G~Iy0ft73pr*Dc;*17FpG5d054Obm{tl!{O%MZ{&NaJcQu8401UIeAp$BBl~Ui zN_tCX7W0*vp-fhY9(dzGF!;1B1MOQ5TB`20lo7!_G=@arI{F>va`U6k*z*;0&6lN; z>L(8ONf}8VQd=kJs~?ujoz|2C&HNK1gCrK{g!%a33ZD0DQy{aY_zLCWRZ;cBroG}v z^sA$%nLvjKfdeKX zJ#`9z=jxW#Hva32*})}++Bt7Rce{?XS?HCmQ9+wnHs^ZFGn(F~dd5ciEoIM@l=OqU z&3LdXhZGGwXV$Q7_IoU43)N zj(f{7_dmukah7&#$Ua&~w+7grfX~j5R+!bShMvda&lufYdbRA$l{(tssf8!3Y69X12$|7O|^kjT`Wj9crn=2WIhl z`XGd?mW(cBTU{EBb~FwY;4Q}xFaQKh)_xN|vunrM+=@e%3u)^sFZ&)4r$ZaRiwA_? z2d-lcN}k=bxqAF{f1laC9GOfoo9RNZ;ljP%upS+FupZCfTv@&&8^+k&toNu|A2ug{ zgdeJo*krL zvB64Sxy()|INr!Ocuwv`Fe~b6SZWfFu^P5jtHM2Z9|hMeRly=s7nnW#pR`I#o77r% zD|Qwy8)llli_*9onTYu&U-x#+uMz9=v`&~>hGjjpXsEm#)V@)!x6Iie%^DQ|1HB{g2%xwTAok?%^d(KI1m6AgUF?zZX*gi6i{l zkB~Gm=`boBL)*O{p!Fci(WrB+{R`YVT00DnI}DA|S+xjwE0=HieYSaC8abyf@HT{c z!MI?%VL<<7+TG#M%>VyGX=lF}`yD3ILKKRdP!cCfp%6bFxei&1AD zwVn?#Y^!(W-Lf`4hql|D4``ku$MK^HgVe0Sk~G?^!7eM05>M#Tv?I855Z}JWbw;9K zoPaYXRqDItqHg}wx<>y=BjtCHV!ZlKITBD7mVYkX?K7_AdNFpds8T>%oqknOtL)*i zM?)}H^90jzlCLYX4xR`qz#0Z_NgDr*dIeQ5zRCsR?yD zx_4z%EXO}LgbEPG4(-hNs_5^Ohv-VFXm%2{T#lQpw?ECkv?|(Rn62)DzTuqS&JszD zxV;#CCH?576$2r;J?+#Mbr-UM7L;P~=0ygy@88FmRDZ=2$b|25ZT!e~4(mP? z4l{gK_`}>Zd_#LBB9l3`_(=l&oH%h&cXNsS0kIV|=v!MiJcHy_ok*F3tRMV=H|6dSZPj7_Rnj7J@ zfCm9E99+Vue@Z>)tP=#D+io(KmqC3eJvzt4-VE8Pww|8_6wXDC8h7=A;ILz@WYU;q znOli;7Wi6$NAG9`!A?Gi1kQi77v9iK6DQLR;J9ucY(CT?sGA@t)MWEYCFhOW^z$r( zk{N7$tzx#t_uLR#CkrhyB!H{Pv_qvnCO31&azm|nrIT~{Ou7p#x}%^zMCoyfFT!>f zyGS?Xgp$dn=Ko>uEx_tZwsp}31PKs=ySuwvaF^gtaF^f`ELebryE_DT*We!9-CcqN z3*;@*o$lVfckjLLKKI>o?)%>S##gguF`0ACQB|XA)EM=T9)=73?EIR54>?tqz5`>g z6_Mvwgz|Sp+NGO7^^6*@DuLk*0%dqwnm9S%ol%KT(;Ckkjk-8?hR_&S8ib>gvXB>O zxoV=X3S%JSz~w{8ZA?%az4@5!^LuCcWmWJ;_XS|PpZ%Hl$Qw)^=ph=}K)eciF}KEF?W7|SL449x}APa_yCWb|FdUA<1c}ab7@C#PmWu(tBMe7#~`C;TIjbRv?AXn*s z+UcpkmV>NFyLStZI<0uR=pG1m;wjc!Ec+Z`7Va|5=`~dn&Il_hcHeUx-S zT=ptT-kQY8_-R(#8YNw^9LdV56c5SQQ(HT;AFSGyDnNiNb}-h#xeGCmC+TaTT?==N zcV97gcuqSY(*(`OmE1W#>xO-FMxWJcKe%YKe*)ccblHr>3tMeqPjXE%O|?Ez+<0GA z(^+vzPm229L>n(a+}_rFqbH|qgo+`v4BG~K;DNrv-n8Nx`9nX1ip}B<)bKyom zhLfr()Q#Cz?e>)%!%pT%M}jOhTror_>kMdR3uZ`e0ww6Kz^OLhnYjD~@23&y=kdo@ zf_UudOTBatuRPN)w23|!B4&+5Fe@V>US>qClNUUYU(NW05uqxj_DmDCulg%8pX`##vre zxUyP=r%s)lW{6VRS+`rb?mkO*^qr_iTXK+NH}N-PAv&sm&tSaDSls|p!&>L&$=SUS z6c)_G2B8VrlRb0I;x3)_u-qa(!t;CTUsoLT_uZK$8`^{5D(=C6ap8tp4#Mr^d~Hu& z))2dI%=X^Ty`^?kdW^dGkZZpD1Ud7d`xb;b1L!{O0-O#U9K&8Gy%I3#6)LL{YO4?c zAPK2_z@CZ;P-&utuh*D!!YfNP|#;dEm1WiYC`(Lo7e%95G^Y0VuNC3XuC|$pZ zx0Fuw3u`mm9pxP)qv68`Dr-vZ>@&qC=AH^OZG)}~-YL zY@AdC1zzjCV$SCJHgJjH_MC?zQlmr7@q{~!^@9ct4QO+riDXvrTqh1lTUeLnvN};qFkk2qsGb+O}x}bT8-K8xcMaM7jv9!L)dX_Tih14u8+5A+oU=li)?_Dmh z9eW&cM({7_Jw45hXVS2>e+9n!QTK0maJfFHp|t!0VEM~}C1`QXsu2MBZrU(NUgpev zvpNfWQnM2+jRGm0IZDNLr?X~WP6XL`8b=!s|ER>&o(SS31%}d56UWI(-!&eOCHLKHP@auQ+L7 zdI1k_4WVqM8fDi8De`cSd9GE=t(+Yu@Y@`#hS*5inl)k7TpY+rMC^75G(GE_KLkzB zp87O`I6>ddK);mBzU!2owrTF==&fETotz0L5{#^W45gCe(%q2#S=}` z$G~1`WUhWn>)Uz5nLF@rhgs}B&Ertv81c#PFOmA(b{6-B4vN@CM)ie@T`l0As-$F6qv(V1 z5Ksraldkki&GwHn9+zw4%wh6o{PCR3)NPfwmB-xE&D|lZzNZy&P6;xKF~9k_Xn?>V z(4?-)W5FGw?&BP~r>B-FhK&iXM~t3Ia@GEDA#1>5euOi+hoz@z0=|w4F1<{$)N3jI zdA{xs+!W!O8lWlIU&)Xk&G?VWUue`nKy+`4pET;ej6-NO|G;6t(R4qbxK)qSiZYv< zuV*EfU;A^fl}3p2XXKkhoCk+vkZIwUkkb~1SzzBC*XIzwv3ihHk|eG2O^>*P))qW5 z6IseRvt56$zSyK#JG1z<6D-N-2zCL_Q#@GK)7VTZT!-aG=&GKbNJoeda(8m9k~X{u zB>p2V5%O!41Y!?(L<@EH^!~K<-)i0Tf;!0$K0woxz!Yf~B-zjX!p`?L6D;r9 z5H-T(B@1Vydz?t3V8Tb8v0F1|JQiR`6Re2&qO%Ahq1xZHZJE_Q%!yeU$gzE6cZG^q zL<%k48#j=GB3J+8kRs4@gxEWPM7PqfihK(gjCm9FR=na7KZHos)kJrm2$H-LOO$!Rg`T8R2S^; zs5jxA6_X($kk$!;7XW94_1$u7x5sdO<$+n@(Vkll4bewycG2E!xeQSeQ3S;23oh`D zZ0<`a-qWv#hoGSQwe>>P8eO+?Jf)9@1=4DWD^;@x?)p5f9=yK`*&Ld+$V+>)U2I1&F7S6TBTz(HBfC8 z-0U(xlBE?dxD{N!vQ44r$(h#FOPTLWp`Fw>QTd8hHs`isq0uyl-jE=d*+9B@tWrg> zpn0jSgXuJ2F63TgoJJ5JJ>X?zPpUQDlcA1GMW?Z(ajBi5DOXX8#QcUM&x)Dh-*E|1 z=WZbcfK4wjxnBgn^mYZn%av)%9hHTj8K!yaxj3d?INg5G!}VGx;0GM@8$SR5Z}*s+ zzVKu5xL zzmL(=%h);+(aq8dL}87yp-Xban!zt-iVHZn^&-W4M@n#Spe9MgNSU;wR2Cb$GWes) zQhLQU*;z>&xZk&L<744zKZ^-&=(M0sYL$L`OY5mMAl+V!aSQCozn@(hJPP{Rx- ziZ+vG)0L(L_14+5;UIj`lx5#kz)BCZeNiyYo=p-(j`)begftWtM7|+LtWgCB8T6Nu zUEwxh5%j}O3Z>`T_-9)de)Hu5#_ix-6n92`E9H?UQa=vvEQ>>S{@{U#H$W!pWHd@& zk&+V5xbiIJ2ti2l67#HcsdLI)@`H!8uKGgJjDYm{VOSQYz+G^?YK$|2DQyM73HM9W zBS==lJ^5?pU)rX&g8a4U(yKxKT2OAzM(ds|%Xn+>R#D~7R{-2(Xx zk2?Ug96hLuZ(pnr1vzhDeFY8KuimQOihG#SlpIHxSdBJtaRxR!RDb5|G>M#SM_2je z?~ahz_zCc_q|C`t^%H14u~9$kWP5TNTRA|-Z~!KPl$bX%=v;{8Kgfn+;TKptdTunT zcm8i744%H;WFxCK7d2TZmp7tcO5NrM57kT7|Ldfom55pWrn$u`dE^H+OGfwa*epLC zcIT2nt~%4QZhkTfGpc0Y8o?1Lu8LiL!-J?xE*^-zp~XSs1cGQlINVEq#KMeL7j5{eztq3 ztaAA4+YVEB^rr;RC@`6W>OXi6=$$A{;qyF6T`wt;h|mp|s*nU3t$Fu*J658+y zZLYJDAAk#g#Kr&FHA_IIXmD`(NW3iSDb>aRU`2qrG@teyv~xTU8of(DH zn}0Px0iaYFCi(gzp$K$MFz?SyANEe4YT(B~IgsugI=N5A15w;ep(mEZ;wcE6PatKD z*XGv)-9VzKn^?-{aCzEWxm{h2T=Z`Vko5k^M zq&0#Su+?06-7cVK@tj4gjMQ4LBDU$UvrmPOtU+YJfJY4@MMD+6{WIzSbDSud3CPy@ zWVWcv^2D5`Kxh!0nX3Ry-sKzmT$%pUM!4eXKxk>C_!oJQe*B z`%{fdJd!mMxNi7#&+f$=CG=f_nk?#cf!)BdU%u;4=|2oj@P{fu#$tih+{bqtYU`1W zF@rVW*sIi<2YY>J3yL#ST1ENIa!U={(nlp3SfCJnhzbgf^6AMzfL#KDp#9km0ARdf zg@B-kfXMaj$t0BXZT-n4%=!ax^X*p(v-z^3dqH;0 z!NZ5l=NGkIS{NtdI=Yrb&f%d!Qp`?9^mfdCiHM8d%C@W**k|qI1@GB&Ji&GhG|=ja zx@V%URMP0qaVG8xIP%5i_1)PV5#*v{Jv zEHB#fE=5?=JIQNQONT(9*4ygl1~YP*z9YAm61)eN(2PX0=&>>4CMxsZIb)p) zip%&Oq~ko!*iuCVic0>>o|OJL>X-TK+Gxg6I_N4~mfdhr#Z@_ZVNx1b^GSODag`!& z*!zAih#uH}c}S+^d8H~}H}wNr6~sQL^{q{Si8BC?v$a(K005zmLc`N85&p6aEe{H2 z@-%~aS4s{t-_9vHKM1C)QNZBOMO~{X`f3J?0VN?jX&sd9M@pKdxpaGg5%JNRDmlFeI* z?Wn#3nA<^Ks@*vGedCCGBRMmI`6;^D=Bwi?PyTWs!;>!@ceHW%2X7GGi@>jSy>NVR z{L~EexAlBjD(vBXz;MRq1fUR{pLb)Wr9HXYLE2G=42JFf;#p-3S|I)909x#3{>Xus z|L1)Lqd&BT28c_2FCcU|V^(x5@XND%_Ky#6Z=93#aAg>L000m!+WXlLNr3?0&zUuQ zg{?Td3}TbbaHb{mS8nrYcwO?Q@ghyyRPcxX3Qe!XVS&P)f#m81Urq#ffl$3j*b}#? zkDl4o8Fd90ty!Fi)-TE^Lm7l%IlKuG3rmI-s&GV9?Qr%~iY1ieoz1dNBiP+#Krx|& z$N%i0cY=;zILPVk**D3^R zyWin>Xz~=b7P7sU%RZJyD1<3fEom$4L*@y!xJEd>Q^r1QX5gZC$NeN*enpxpq=v{(I-#` zH@c|fN*9!Y5t!L!yn9J(+d=FvGekM;6-5^cjGpz8myKaSeSV2calb%fq8-(+HP%ti zfo;D9@v%9dCNsLnCCM;udTcS<_mV#(fe*c7fXB6XG305&@x;^>#*(C`W5 zS_5VGHKv)Zx9eEk2Egc0AxAh23J~&xb3YSB#pmXPKQI2`{Fh@y&H~78-S6Zo05a&Q zUM+M|^lB5oUi47a+IKA=L$Q;P-35o6)B|m33}5S zgW4Uc`x$CIPofgd@1%Hh?UQq>(;2z$W;QIV+GOLxw&+K4+EJQorCx}&xRM=l&?>^t zV*%^nU<~?Y)YfMuPHImWf%?Xl#7$HnM89|nTd!qIyIU8O(b=-(6h5%7@(QC(qqH@^ z%S0~BPshw{ffTB`go8atEMU4e#>$0GoDP^=a};^B&(oeqdh&zBczj%oFfho_*Eo^_#cFereod#+Oj>g+l+B_Z7|=U;LNMuWE_w z&SB2+8q3RkUgT(|CaUfe+dbvvT-z8GYh2ID!vw48_*l7d-5E|m;4<-^4$Q2`t_WZZ zA9goTZcOQrC}$md*c$Fmy0{fES2MV`$C10I_c9@yqcnF@g4s}95)o5Ro_*BfEtKZb z(7i%oh+O0i)j)gEB0~1%xog+(`wa6_(Ff|DuKtf0m`}v7ohB+!3#nGpZ2$hOr~4@i zL>^Lbp!zvzdNyg}g>TbXA3JWpH97ly1pp0!;lupJ;S(m-O`iZuqv003Or_k?Ynf<} zkF2j<2`#9_642JXgN2jaw-IH7yBzHjM5qy!owN-x%s#=lsWSx+ZuQOWYV1p?yB;fE z;|~Yz+eNGM-NW}-`KePe66Oso*}lxC>Q%-D3=|n-<8ssG>1{HGOJyC84pc)4xRwk;aJ4`R&Z(4pBBY zRqYPWL*Zf|bKFr9svca3NUL;i>S8`jHC>afX@GLV)FkL}h|}!RR*b&bi@VaHH=*1! zhbH)&jkd9#<3@A?{UYP=#}nAxAD!<>x_@izA#Cm;TLtyMurVrx<`Jab|Fv%ymLG4F?eFbw<1jxf{|4- z46g_`nfS+*7IbW|S1?dS#)NYLL7=g-lw^CR%-Lu}ryFiW$1z^BFeds^P94M%rgQ2| z@$0Vi4}6hb!@u$IFEanIpX@Mw`u%oHw={A}uARdRgilQN4T%ix0D$S5C(E2L+r|1o zz>>?xw&l(o9$pH~O75mj^Hh#?~1#&y3I0wAs3HOpQ}v9=vH{=H0EO3p1kF=Sa|> z6-_Rx1OoB%`{?$<6ACJLERv||=7sNqZL+&dNBEvXKmZIkw+us5tzudbn@(;}Rf362 z;CPM+83MTUqoL@zs#KPF4$ZD+ghq~wz2Y}Y?=+cU_Rd)kZnNceXI;C>)d-Ksm_r)W zCR!_eBRB~!j`PjqV!Yi$qv z1MU7B)zpE6GR=<}Mh3m0Y}r#KfoADTQ>PIlj;1mR5W$ro#byA)Usf@wh1m=ai>Mo;HbH&S2jF{$N`Q=CRqoW4=96wWO5nt$j~t`!xYw7C>f zT%0e7;dNB}+&1C@HhqnF10}~UxdMftipihEx61D$Vqa-rKZBt^8?VxMfVIuXvD4Cw z-@Z@hX)NxMY~5pBLa`9ug5ud$oEzB3AWuN6%NXWnp{%*j#Mw7-7}{Ewki>la-Z7<+ zOt6)9gdD0%^Jq(90MCeu@F#{vJ^BUhSl(06g{iRd^_Q!EOMUf=F2_Rtsigzc=9Tb)Ae?iyZ-#Cr9iz>Ny-5 zOB~n{ey?_5=3FeqCJErJH*jDNv+Lq-wkt$@e)2-3is_tVMvCz_P3z|${&4=+Li%ax zo$aXPHGF88w|V^SC7a`|J5`76)s*ktx(b|%3Y=Dc`9V_Ubxpiexm{roOT*OK6Wvkc zXD=v&s09dYO2xqbNd%%|bhQ%n_{u--`sH3w>TV%ueNi9^T5;$@jF6CY1$5<}hET4Y z7FmyzOgLSkgjoS$7P{)2VhleZt;up5zC&>`RzpT+$&MA>r~~N)f@DLua$KcNjE$gZ zlxuw;lu|^uz7(JVS{ycoT`)Xux`@%Zy>Qu*ZJ74EUGtW-&iv*>($GuGh6ck`TGhC+ z1`Y5!4dVJ;gOJ0r0ko zH2MR(|D63Nu|j{l6=VyuoKO7l$#b9wdjz3RTb;~a{SyEThWz0309Ts_7pDgg9(Rmu zkd)jGuVWmO)6V{fIU~kEOZ*1fOL|5vD}3^f!YylT+ra%`U3B}pJvyk2q*_(`O^-1T z(id{62{aFls@hX;vVA~Sc#8B_+xF}Mo(=^;W=EXmKd;cw^eYz7OjoW zIk}WvYXN1P+&+6AU2C*qH#^pP3)O@4vffywm;usT6+f}8g!(!o!4Aaw)z$wx>j4P7 zAbkU57qH(Dx3DxJma0wB(QnJO>Hp#vMzRolRiB%h!u@$TH0??!A(#5rYVz6S;Za(L zgr?ms73z!xi$r+%a3^VGp>uAKnbo)G%%o#>BLu<9ULL(Vb~{yTS~*B9!sVa|Ga@b7 znTiwNc^=>R*b$bHa#QLh)RU9GS;Pm~ydvV(iS&t-7vP4jh8@lRT zws!)*J|B;7$$%tUy(R|VYrI_oV0SzGcKd063s`4rV};WsJ#?uk3%>!&5FKboBFNtM zpC*s%VJ$0_=4JUi-NPw?8a}_i{U4pyM`3GA>4I=CV*GhL0o;!EWT%+-ByS`(#7KfM z!$}P|qk{gEEZ4aNi8*0dt}k0$%Qn*V{8c>RKYBZ;40~B7rx=$ttf+RHJx6&VOv!>x z7G&xB|4vq3?3!2yOKchJpUo})f;s<%VuJ7803JjD7z6|W8U`F3>=^_Y02~4m;u#dk z3jrDl85JFa1rY^}kdc{H4zCZs~#~KIfbzJTc6193ji1}FaQ`N_;0*oj zzu>>Z59f>RxXP;V^H!Jn*>`C!jsM+=hSjko?`5}Wuf+DZiqKhF61E|j5&KRS?f+93 zm7!|1n`nQ2sZ=(xVJ;urY!&;6c5G?N%*dh1RkizktJO-Ktro}sU+`b#hlOmf`Gkdr zwnzBX!br2%d2=dFQ)hF@e|J(Ah>I)0qpS{H@sPsUQX=D= zk7iY+dhoym_>Ro>JD}?%WzpDx6DU8B(aK}dby@eNPB)o;%)xuk$@Wt7>H*Vsu@y`3 z*2t+@v}`?JWAlT)20O${!?>!FiymIP!FjEU`vmishE6A=7?it`^eLfgZ)2=TOomtV zJP!&gu$<2Es|M<23(+IW_JCrjA^VMT(rf}aGuz30Ba>5Dq|=8k`!{0`yFE#^wh}cs zeh_{^R76ZZfQo`gXSp{Rl`~@lkXxCo!QTNS;|iUUCiRl|E9iT;wQPc{n ztL8;-my9Z7LzcEtYig=uv?o9dTx-9{Y>ngpE4#0#RJXtfZ7V8T(JbOt;DCa)_gnc_O2 ziLov7-sFVd%+r`9PvBwk4S`o-aynNNGBMeeEGHt{D?Ia9F6pY4KtiI26fUhr5<=$k zBoR#{hL*?_s!ziS2^2txrBG0i>;CkZx*d*!N5mKajVvcqVrPzv!&8|$N@9q}%nHzp zu;RnAiNp;TBIBJfgX*X55?{yEG(g)?3`%}~(j@o7uu+Oh4M4dv#dkK7lH9@xr z#?_hMYNi_r#EY9)5RA&dRCqS@;wWYxrpxd3wKRXITTR#H1Ns9-oz|5OjK`a6w|nC^ z&|9OcAF8Jym_F^#Tg6}&Bvhe|M|5wECJq2%-l}`BzcIHbZ%;BlmqSJPc2OfBvSl8$<;6qd3mhtm$ zFMfrokdBw4aH2U86O5C+C>ie~P+f@a@ddEe2;e9%sNyT*`jh%pqUKH3j1+v|Z%({D zHxG5Yg&t45MIGS-Cc<06R2N~q;{))G6($GWOG5qjF4ZtQkGI`OdJp?}NWQPkWiY$>oY3Hr79GGxBoPTjRng zon@>z4=rLZdi6exuvT`7M|5bXg+k2RQl15YHSvO~Yo|)pb~BkWgiSKlq9*C9I}K)J z-%H;Kp5lZnyq~>pqt4o}lBDDFSFRrIMeq-4sP86~DX(GcPzR#ZwYH1O6Ur?Qk0&U; zh0iizZj1h)s+oxdjcz@BGaFwGe}~1AAaGEjvbE=Z)hqOVM6(ub3$Lw-Y@v` zY6j{cT--&FtZ4rT;;2xw+o@A%dV{KeG^ z5;-(5{X0VSWMe?mxJBhfkc9ufm`%YmL9_ScBBq~!y%2ltdrcCF%~r?Z0h&8gaz?C{ z)CmttLa$IBR=nobI&N96*N?8iQ>|ZQ?2~Kwhr-()W^V-v?lN`V^v~@X;%u;%VZD(_ z*6AW6CZwx^aSY=m-Qjegg1{rJv4;`>$N~JEX~zP?az$B#HsmJIuxehTGTmYhXuy|V zJ}?R?JtGl&WfliXiY`9F{~dr9z!fh7F$fe~@QuTajGySDh#!U@IPw7&HbzG1><8xzQv(LjJQxpZ*LI3KM%4V0%c0pcpjh4+ z0BXzsP^61s(Skee%2jmcpz@y=Bimy4XX#CiY%nZ7o+Ll^W2|59-UNsWw}hPOVdLWk zrI-)pK_W-GmPO*34B|krT49!PCd@=6p?4bla`v>K^mZ?u03I z_yIgw4ioKb2xi03RobtOzLjQWZXR;dJ|WBFP`6#2wlt8kQM40g2!hjL&(Le+NuIOC zn@Dlkq$msZGt$ms)%TWfd0a6K=P;AzmhdJP2THm1VS1=nl&ny8)lnKNby~WMyB$Tu zQ4wA=N(aJufw(nHu)(_fn&G3f`B47LbyDA!w+O+c>MxH3CNV^exv1bL455z)Y9NDO!p;4J=a?8r`gz^064A2?R*}FHDSQG%K%y;Sj zUO0c?*4=09>|@r~vQ1vTLHYHEO?zyIcZ{p`=KPC-U*3nLu^#tw@1-jD|UCReM; zA)AS8V*=G^!l^s)9i;`My`oV(F0)h=^d>n%LKJcg`M@+;ahxzNHWx8nb&=k_=rW#$DTs$>)8wZ55eX>f5+wl%cA0+AQ$`=tP zu*@OS$#|V^Qh50{_;3xWRSFSHfZI&j(a;z%_>`|Ty@<8pgZ#VPS^UQ**M@tNta2f9 zj`B6`aHOgkzWk2syblk)xct2OO|Pb%=0n>Qpc(kM{OV7%Q8uCG{U4nlYSP4%^h%*X2`0NR-XsYnV1s0>q4Q9~P zgsz@lg<-rJOQ@wRlEy6bMW|4of|z|T{H2gWNR!xy$`F4)Pa*RhRO6SLd%+V8HRCKy zeFrQ#mz$;*WI5g00p>@jOb}>{A95J22%U)`qp|0jIhv3ylgC@p;J{n4V8`weBDQB{7w(Ut z9LEa^@@W1X{(ctge$R0C$;eL~();@i7)6%FlITTclqTS_s_mmE`y6n;F^5m5^bjyw z=aig{`YuMyIM=w}R%6~?m(Ii(++uLcD|AL7mE#eQO1SY0#(rsz9o&ba3uGVf5c$$4 z9OP(XvG(3rg5zDTOL%kyixR?yfg8G{kY<*jNzWrWseS<`>V7}7FBv^;$5F?S676=M z3ivp1h-U)A5WY)RMYLgF;T?hy4Lp5x!66M z!WV}4uCFkx8w52iv2=mL)`)Ki_LKIv`PYK&}mKAQs+k?Om4m=e~JJhw{zUN%@ zuQiPFAO)foMNyxfi5VE$sCS^i*Jeuz4>v(wQhzq4>nw?_EK$=8!x~s1A-Z9F1NWes z9HpTDw#3$7`mr#hPUV*wHRVx@jD7W_0I%)m%((wy6dRZ%P@EWJC0GCP*KupND+ReT z#=uu@&==npeO;YR&jfe5%&6lfn=baud<=Z|po~xjR=QoK6ki9>XyIi<7;@feQ3g1+ zWLJkwf99MPyPF{amfy$hB&6o}7wH@pBTz8;q3jua{0s>;&#t~LyOr`r?T3ardiEH9m_vurvkqN z?ij=Ey4kP8Y+YymuS(QaxfWlMORtC7s{c3raCKGc%?y*@CS4QCtCU@o=@CIP<$(d6 zTKF>HW&gwIqb4d!vgk9c@yf*5II7obzc%rcnbAn+d}hI<*w*8)!2aGuyRC1kU(+pV zWl`Vc%0f!GX~miymt%wH%cn$JZ;Jp;(J?iW^0$UqVdu%1H26EVwPHR;(fawOTN8Qr z*Iqeyx5i4|8A89e*FSE3p-eVZALgW5G_6^X#ilt>?_s_X6k!?2WgMPot);Qhyl`4k zY#er|Vo|D6RQq&W@sDgS`**;2fOWj!wM?<&VEz9^Nzt)Gt>(2nY3wOe-ps5A@_^Jb zbLk}Ye@9|pQrYApXx%irNPY8)-0K@(j7&9OE_-Q=(c(SVP41A$CGj<_)Xy0cMkEJ9 zwxU=)HoNK+u;G5B!xOW;r`uA(wrr4mL1xUu4a@Z^5h0HFc4N{+GSZmZLIed%%BH_c z-OHZS`kQF?CP8{jL-r2pr7~A`0E(NGY?bpt*u#SrJWS73 z+2+yW>46IkO7&IuCf)QI)>c_fm&IuF6)a_?W_{cn@M_F@Q}eTiq%N`ukL#IwWD8U^ z826nWzmJDo0M{T(q}>ZdfiRTJLuK#Qiun@X!oCwgmd+Bv9@)O-y!G+DBqF*Ady^f< zb1|s;fz184ap8TW8)__gEk-%hLp|}{ z`bFgk3H?oa5{E42)b9YX)zqIW9dq9Sx3b0+Hb4mGV0*hIppO6O$?P{ew<@~qb2*}L zn5ceZNm0Bwb;Km1TUErXA-r1Q)MHijSrjW$hBrQRoGCgu)*-BKAigR_f9Z2jrWjDjM7&TMzjLF!FIYg~t-U~%sFUpA zysAisZMBOG*(5^i6`~3CSHlV_V?ok3A+C<>oA&NCpb72%t*BvG)Gje(lL_rVC|23= zrL+3nMknaIM#lD9wwGRi9dDwEF~T%~Ay%ltOLe08wpnUpbEG&)W<38+BAly~qQR`C zj3qBx|KJq^2Dw~1i-@06GIx0O^K94t0d#ocjm^?aDeEBY=R9ykj$gGY?d*^{KBVzY zL_*ns6BrVkT}ysi;0zxMlwyjMdnOh)5t`&5C${>fSh6xJEJu^%?vKr@^LdltNmz~< z^SF$oN5Gs}Wa9Y(1N|_S&&8RkE=oKcf{b>JMz)A!>NAH-X2EmSZpTPehyJ5s?Hnrj zVP%86*YOB3wUB4A2(V+4<~+*X`|_98x4((?i0^>dL6$E_Q5wY(i4|2I84%HOU5K$1 zGDcnRfLXgMV-Jn9H*Y;>4B%Bl{Ha?I$4JCsrSu(7+icrOq{#RcAq z(AB)Ht~ecCo^p3Vem!&qMzV$x&_aamgv@^fne7Lo2MR9|F_i-A8O5}Y-5A>7cay0! zp`JuM(TauOKIN1{-QC}f)hv5hWxO;j+^!t+@^71^OZ@G3fNUiu$mn+`gV+}JWLkxkG`^6GgR`QZcU@{vOvbfER+^-;@6uI?iTMb~%4h}x% z7I>jky>kyt@0P zwKXvXOq0+wbbNetz^4JTfwa~3V4YrWO`L#`{TMm*1z!(3>FvVaQ1s82II^DIH zvb?w=hax%RzWd9+8w=#`C4c9+;;pv|9grgqNXqpoVHz;L0k-HCzfUr)l^>y43To0- zoWK60$IpK|qWsNMJgRh0;+mr31=9mfy6)tne>UFOt0 z1O8Q_QdRfnH!l47m2$xn*x}VK63taZddu{BFXkLoiZS?Kiy=qCUSiNMH@*@fD#or< zRENxW~O#b|FqB&c3M>K z={@^4N5-gIoP%~GrB9QcD9~AnfcV#+@2vIZ4!!*dHW)re>Q)K6Ou4r>j;Ux{#-j1x zxEL2w&xF4Fh)4cld?$O?=6|H*;DE0vZW$Z3uubaRD2Z40|K3F&t2AB@AF3zGSpDuR ztKzk9N%v=8Up*=nZfOj-*mVp&$D%8wKRgm~uVtV@H)(DPeQsiLAp$aZfK-XO^KPPqbeeLV znCsba*pPVP5Uv54xn#ay?jv1q*t_GUF9~!j)${^x1*>6H;Yw&1{t{$d)s73cCRTGF z$rMwFO{qo=c_kTG%~{^2D*H`V#y6|(C?QiAkzl51vm$e64`sKzE%8!I@k*@@M83Lt z_(KH-{HVaWDBPuLLir^@;4RuO4OprAWnX!`=E1b(V++GEe;cq3=MNPTa|?F|*`!H% zC!S^d`8WO8``gC_m4zeqv?x0m7yTu|Z^C2~Nd>-s#G%i#nL4yedQCA!cNA4ahvCQf z#zRaQKUC-hqLNl^jFKR^N-=Vg+B7ho;7A(N+R-HoNf_9R(FwB}gTv>FJ!@z*)}R;S zb)@NwoaZrZNHQrZW{(Vg7wkM7lpY!U>Bzkog8HRcAOu$>1~{&ya5V{WE3bvRN2w4{ zpD1%9l_i=4!G42RG-+R28d@tL5e(HEYhHZ7r_3=8P25_AJM*?RA^jJb!FDP-dOK6XbTK66xnFc?&c^-1^a}G#_q=RK+?N61bWrk&^c>l%!S|uwXr7X z$h~51GK_CixS8*X48*acaP&BVS>u8Upt+4h+0H7Mqja(P!zjC+yy(PlHj35#2oBWL-M5pl zoZ#wm3krQ|mTqvJ{;>bMVs>laR(^u3<&&=ibc{MEos?dkw^h8JR%>hLE5EB~uNjqW zMg{b}L@{B4=7)NA3v+4VJlu4)sSYH=H|MsRd6grH4lX)I96Gn(PMvczd>h#Z1;*vC@UkN!WFpNhPd(VUav>XeXvJ*YHR(>~U`vTYuiIx_OgY;e& zGI%hc-<4{~mKgR~AF!Uht9ad=%?@}0oGoEK3HHeTxj6ST z(fWsR@jB^n`7;udUyZdmoM$DXozg<3w441Dv?{%Vvhu4;X|i%w>~2Pp5^}$)sxca1 zaT_hQVHI=}wJ5~He-moUbQ9u5^|B1!Hm@SYZY?fLy{ACpWRjSz(4-jNV$DG;sKQp0{>Jto)iLEN zVV~pR4q1MSW>Lp#(W2HngGu3$nty?@^LQ@MwmraCB&j*Hjm85W*V72qt7hy$a8&@0##a)5XIgbcW4lD=Vi`v4`hP7 z^(Tv~*ajipK8MQW)jM{Z^UyVdqzLo?c-Wz4Pvj?KO9If38A? zJ@kFEoPGR&cCMe!F_qr|=NTu>MtN8vn9u{gUR;;5fo_K-x#at%pYD@f4scrqv)NlC{B=jDmZ#RkRG%hOYFzHN~xjHY%rw^f7D ze`K^K$HPhQn=)q}_HUEMWyPVA52hmGAf~~L;6a7IFFV2q9 zjubgkO{4v#ZygHx2jY)tP3#szj$j+s5gTSl1-U|}r@3lkq5Tm6c7@a!2?AN5r|FgB z`!A95eiIJb6WB2&_z}Mdt>|g1W`;D&**n$m0JaW*%g2$vO^>_|yS8BhcXCBabN2EL zcx7^3T2rVA&Q10`8aS)nP$4x;|1zfM+6KrC-Tx19Zvhq8vTY3$+}+*X8h3Yx;1b+j zf(L?oL*t&H3GOZl+B6zEKycRtf&@q)f&881~ZM)mpRW znrm$)?|8Kx+^kVLj{gJA!s!VCIUj#})lZt2kqfLKK45a+n;p$U;<1yF^JS|SNEsk`hhIJ1tFgmNx@#s-SNEaKl>zxpbP22d8qzW=pF4Zp-Q zGCl`Q*lGB=mM>_>B+5b0foSehH8zgSjw{-2mFU5Tu`=a1PxbX1Qk=OQ8Qj81EuNbA zr!mX(5n$wC(+`p!`=#t;V6Gei;7OiBq6dx8qU#CJ%$e~!HJJX# z&1d^jer5l^!<_Y%<&^;cjzy!}cQ`^_m{bK(*v$IGWKH(O@X-lw_j_!YdbSsd#ttLR zoiH4~vy>sg2N57pYH|UR)Ge{M3ym(>x}CVXv6W5VBqnAgQ-n2=g&3-9J9DJBqC;2O#R5KSdyZ_!wW2t1d6 z+q-=z8B@`o_^hvgsh`B#fO!Ac_m0=FZ_W{%{pU^1W1-A$N<*<5J-G_tyfry7+1CL) z1a(UMwQcP8yZI)AKwTw}i7?t~hPs%pUpk~^;naOc{`d${zbRhS|4^y2dcSCcz=0oe zZp4jQ@f)!yg%xE8VBeUiHVH7)w3}@F8KXzG+NDPk$y1ap0_tEKtJ%cdKz6PIykkQh zi6;bMVSvamnCT)QCaOq@2G2JUnCLYcR5c>&kCSUvUWjHo6^VvMNsKxS07WJTcQ;G2 zWizZ-VSogY9esW(6UAH zXSPIdJ&@OD^63N|sg@1Y%r^tzv=CFZ5d+Ez+bMdWqe4d>ZVj!NDkX!{pqF?x)^#c` z2x5gB7|w6*r`g^bccc>*-B z!#VJ&Mgb&_(MVM>P6)4D*`=FBDiwz|`h{|dNuQ;%%JyV5to9Ize7)k&fT&0cdFfi^c|oWnjZOe{V6vD- zGwI+W*isk}#aW8g%T*#t-`1~tpsF0Thv@A#^H50z3#!It6e~dV`A`N=d+UV6WLUa0 zRLPTwaFd^lt~5XG2E&e^N)I&^fPpzXU>Yh9pR&r9-Yt;1=fUo=gl&!4j&X#56`k5^ zwZ?@tK}rkhiDGB>VQXNu3$Ts!TH50LxIWPd##|{a5qW4nRj>NmA3bAMCa!aWF*yuw z#cy_&FlFDi5pW8|;| zftQ@jfY&2AwFj}pXix9#qP-av!JyK8j)OeTE=4qUgc(n;I*#Q_A9(lDwn=Z6L%cmL zDn4y=ZQ!*OVj&|z?iZ2g&Jy|t+r#BN4zH;woSZ1rHKP{^%F0NHtEiZCmS|HjBiyMl8VxOBAK2eC?&A5`>QdlD@QIRX6EUC zOs)`sMLQzYFk! zRW|IV3U6l5l;Oz(1vYK826GcDlbvk8Y+Iy;XZ5N7dCMvVm4~tql$n#Yx)W0`+L9Ld zYl)D>e0<46P=pLWiSg(GLHn+}qbDEk3Vo;S=w>PX%5&wMfv-7qCfypf?x2{M1UIU# z2Zuy|=ak%>tn4sLf9{Hd4%x|l9@mfuknuG{AV*imOYu_GO(C(R zL#&B9@4QXXzA1lY{-$Bse8be~amY{SVwvqBjTc*xp;RIWm-r%=H!8>_pwZSVt32_>tKxC*U~rxwFbnuI=sQ4un4~tlq^Y z8_B^8+LU__xc=wrhOkIvx^bt z2a}G%k){q~B!Sd%!+P^-X3s*`P~<)lk4?x(3sNgVl@TUT($xy|$I`bBbQ7K)Jq*|S z3uc}ju$vpN?GEBtj4ChEo7VFT|EY$RkbN8Mbxxdf;&27e)ng&;F(J-bG=KHu{|89Bz6Gf5C z1q8?#6|3I+F}ERQVU|srSHH%B3;Hy?o%Y8-d?Zdl7tG;a+%fNQH1Ga@`7%CS)x59~ zeK{QN{ySy`+V~C^AbbI$jjzO=o#4`u&ikagv#1&u2J-(R4t0wEu+NmwT+P+T@#n=Q z&9J1>G{Gi}oS`w^^4bex-271uG8*HJ;h)qz#VkkNLsisRERcs4+SMEwi%TU?+T;sW zVxb~PZzbL=uc`^03)4u5NO53T6EE|n!AS21wNOqfBvi#XU%v6zrylYwgW~&|prv_xxo*aovG30nNQ7@8}b!IcK$$rlCMJk2`hVoiS7c3k-`$Pp{ zvHLa5CnaI&RF5F1Q1 zw1zyr5wKGwqq3iyy73FSWgj7ZhjT5r?@I{z4i`eB8WCGwC>{2+-H9tIGc-O=)z_i~ zv(mevRHX@#>Zy3d!gsj0oqt4&PKw4ZQln$RfaJ)FvG7|}CYgK3#wdHL77(pDN7J@> ziuby0d(y&3p8*sh!Ge9rwtM%--2HCn-8Crp8RSDKiR`#mCg6M1rf&8k(NyxWa8F@Who$UC}Zp|MzKe0IV`jFX~ zPU9`p1Bk2eQePyZtLOKI5F4t(B3xYM>Mj&qtW#Uc>5k|W$$J}4hvu+HdbuSJMQXaJ z91%anXEYPc^GqNwu5sNhwA$QCOMaE2%{)Y4w;>Gz+q7?$VnT%I^81*L_mi3BEXs@C zRy}e}w*{&m9^YIeYc}4?B`q&JmG(|z)zGM+cj)76p4ZcEaI!tYxJ-M0tu5u`IC@{A zP+Yw(ge3zB|HJxmLA~!+RMo!XfBp`qi*f@&<3L3lrkb_iB`x)V4u+&OxYgRo9no)6 zaTV^6Su0~EK;9|53?i7ER?;X(X6*Vz$BXMJJY1}4sKl$LXU7p62kA&>W1?03aMN5B z?dDacN`X1O?##sFC2I@Uj0(OD_#;5om($nOi1Ems_oCOo06C30G0+}1bnO8b`D*CY=YRv>`yvzCs7v*J}w>0LDW29Kkb>_lVfT@cr+F@JC zwP90cXCT|6Z&G1ov%*5rZLt!oxi$|Syb1-3?~GPTvx8w}i!`Bf^^pOFi^~fGlmw$~ zx@WGhg1i{j2`+smjO;Q#bd!FSyJHmV%b3Y0J8yVYKGMQcCFQPdKZw#p8W{9URnA>6 zY-|uGUc{1kY6PUxc|Qmx92IrxTW!BoOB>diFLzr`y^!!LK#tmA)Xd3N@R&WHNm&BD z(pI4%L;NCUR>ArmPEMl%W>Qag4x^X~Lsy^1>XC+~xqKKMM%I7sfDqGWmo1Kl+kJlI zRauV?iH)b`$_pT&;3Z2cyO?9}T$7*beZos2=(LrT=R|oPIgu+8+GU7MoNq-0$CHR* z@fLq)l?PzSGO&-&*t2gr_nS*NhO#}bU8RkdNxYb8=^0(LP7Kt?*$+{R^spXiLM;^Ewpad|9Xy3V69+OD~E$hL%>R-NZoIctmm7Vl+Gq zCXG=g<=yoOCvp0g_VNR6CkqmB_LGM2td~)Y34jPm1D*bqKSSeZMJ=t=EQ?FqI9`|3 z;+j!P9vPufSR5)no}5Mry`1ogFlSVHm=V5RI>Ww+I(@oNI|-U1CL^1+kcmG;s;?xq zl?Mt#rV>ApX*r(yD|O}V9O^7Aq1O^eEA1C1s%XMqvZ_hu3Q6qeE730k0!PSXu57>A zw78OIGLyy!362rT8DHIFq`OBad$VBf)s-7lBV&%!K2NsOs+UWrSEi!vN$=8$sU!_# zCBp}Qf=Nyb##OUlCxZ;%QP6pV1b00jN%v0J5uoV-SIETF_k!k+C6Lh8M!PE0MZ;Z?kXP z9XBJMpAt6?h+a)|{LkBUKzb;?)WhH5uE^#8D%$y$YA^Zrp3`xS*_mmQACzKEBeU`? zBjQ<@xtS{5HD0}7J2%=G_v{xkY{#R=A<^#T9J>sn4L-YtNkY4bDqakV_CSnLQ_9A` zA%rCHN4!2;&$o+leSO2fiqZwq*P$z}vcGq@)>lx|8;l{Ru{f7NiIHj<6y4ykM)i~D zYq|-kB7&x8vnG%=s~{R0>Lz&kkZ5DvbC3b#HRN(THEH^!?rbj%DDCC2_q|WkZ-N+~ zJ^BY`3Wc#&6BPEYri;sE30Adx6<_4^G7Kl}mO7GOQVYo1*)=-LG~xM0OZ2DiUNlwO z)>GImSU$T5VxdFl{1>_?e#~RSRppKw9@p2;jLKELt1X=^sSN$FmzZ44MVFl{Q#w`9OvT>Nv-<=3IE8t9s2 zqIJ~_V1M)x7s$^QfCqI%s(B3u$g>y#TuPT%cr)1}$#^N79!U_X892`YDqS?9QF>YI=o< zazCd$8%ydpAZ=0P0=z6xED(#qG9Z1fm%0Pbm)uK5!wq>{>7K`ve5JuzZ9_s&E}a(8 zi^lE&`P8=K2~kGGdL-o73=2@gIBeCclivvF;0FR))z!wFR}?(nE5(65tNf!`t868J z^kfvFmb^5*hq&L2c)7=~zJ980KO5b2p{>sNMdXu?(WS8b=kdI(yco6Wx3@nD59k^f zoplE;@rz(Qp72Z3`{YH9>1JVx0GE6H3m+Iwv-nOL7tO4&Wr6n>xEdQ%t z>8t5ftu|L;o=b#us*hD6kBL=!7Nn_VW&$;G^2!!Yp%=ugPjx?l+~(E16kC*OI)20P z0_zUKOPJsxCFBkuc^{Es1Hb8j!h{dyt~`tf&r&g3cA&9;+mSF+(bZH04Y zt1SzRi<0|neNM|A9VLxp-lvv_WE+mI81}RGyRBj0aqpE7!(@mrev*9^oFK2ocmdzI z2RD~Lw21+ig<^{N?nTefrHzzcZL=KyjS_!Y-w?9i_UJElBwYGbrF#D4kXFi1%bJpq zdqgSp3pFbtRmqXt&F8EDz*GsX+Af76Ln-1S%ItDivu9~T^+IY`8zu96Wum&R+oi4O zcewPZ%&3_nA_pBm!WzM{Dq({sl?kSDkWL75BICn3J_rENaP5CuK3wh~10TI!-cF zS4Q)d%CNaVPr^8mvVq*k6b#GVBBIJ9^(IOel+DeP+l=!VRKdl_-{C5M(WCUMu0odr z31`ijAFmymTr&krck=~}B$h#pT1vd^d6`ZHd<|d%N*~i{p)&rmF~@>Ag2x9+GfOix zD0Z4o`OYGl!QHTw3Bcb(4!5p*u7+!;W<;I=6~E#R2c;|D>Y7|WzJKDbG-?};brKAL z=V}^Vv?*gG_SviE!KI)fr(wYxs1}Wl>T#;GfpmN1w|I=eXfWc0EgI7F8(#N<#|8A; zr=&u(8HP>()#vSQe2R!GR`HdZg?5oaHLEW+0$tU+)^tDesbgMParKJj5kp`0)z=Qp_{Sta&02gw0O&vUyW;+NP6 zu5?s1B8Q^bBL*XgvCPY|oIDh679*wZ0d%_A+w0%qG+;)$uG}_}kBk&E?ZJI)`m6JV z3DdRXlU~_+9IPY{GZhD~5z$@-!c`FlS4Z-EEEi@3jv?v`hnS@@>W-x1nkJOCt<7qs zIiLBh=hduBNukaM7rLQ7llnnG1liz3N4t#KdXhn+zfW7$&=Z%HegnI&?sJ#F2pJ1l zTJBeYh)?<0~Pf zaMH$}|SiS7xX&5X)-I90TqNC?U3 zsauesXPODa3y=>O?FE^tkL=#SNdLG!-RL5BX2pjhN|iCZfiD#sSh<75fgt#)e_C;yZ90T^tQ*f=1?53(kh24Sz z5X;@=H>pUFOjBrONf;f=5;{}qttL7?=6Gu-_9&_la(&W}DpXkExVe#+GHwq*&TP z^sJ6Kg@M7DB)xVIe5I^PPpJe{s;zg0r8hhw4fu+`(I5UD&X)hLnB}W|DX}Vc3&3TB17lT&VRGwqN^4(N26s~32lvDV@e3eD4@mOk?JfNz(y&=L@!nM4j&Y4lr`#;^ ztpbM^at@l~An;zCBQE3ZI@))*IXW_JlxjJZdZM|oX>)O)s7BgC4ZkZRQo@4!c6~O- z;0y(xGp=f@yh!P(>?6kWm1qA5wmf>}kS672IT{==*x+ut@H1QwGc6TMZlm<1$3s8d znS5#qTUMZwu2VzgeR3&(bOdT^0iOx1HY=wQTXa#q(2(E{{b&NuoyOPC+Z$g^Sjzq^ zTl|{SUk%ZGd{g`I(xA`!9`Lud&JS7NS~oN?qcB`+@68#qu%vp__fDH)`Feb#?H{;e zuZ1whwuwM_;|^OYNTG(VdZe^75`0+!NTnF$VmU9642-{yuZxNfugFpHN=ArQZb3%+ zB^WQ0uqqonYevhW`r6z=jR?=l@cL=+U-aK$(JD!i!Xo|d@XF>dFQj%zKC`G3DZ&?( z@`e`RaEqFWlRr9)C-70QdCwMUoFfn)0rW9gyiRzTAwZ<+#^;kwtgE3nfWjsQp87cP z_IHHB_1!TPWg6oPc+?nke8z0820wY+8Yi&e!5iercwpbPdpUtTWcbl5H$lwx4u*0X zPIzu9nFoyn-pr~5PCe~DAfEM12?;Jf?y>Y^JRMy-|=90UAC749S*Ter{%FckJwb0r9vs8+LiOsz7Ho;9=9t3n@y zqau!QtR`{CMH^~IuzBQ>oyNBqk22JgT6TILM>CVH%dVRd-)karYgxsXY^V3Bh~7?3 zx|HFbjru-^=KiD4@38P}U)wTuMMSULmT<}|uN)Xb&VO0=XSn$-ol~B>OD`4G>1gxV z7CJC=o?SS*{A40Xcws$p*daBUqnJOyU6=T*6Yul@XPMd1z|Jd0RlpC2BF5$h%o8^> zoDiUUxPpcqw0ku7NQ)nL$`}|-^a*u^M90x+9(PSQeR3aAQd=?t55Pe)!&L3{n(Fnj zrvx?v;?EnjU)y-(pWYkrR71?>{yHdwrK(tISz&o|%F!&*K-Net`NOq*GzGu(`aq@D zcvN512nPrel)bi~75Z4k17W0Xgq>q+CClkO=Wwe}uqd6Fcn-Rj^aEc?&M(G{dI6#- z+pD4da=LnFLrc5QA3htaNbEqDlBg(slZOzALG|QYwB7o)47XN8xl_Q>W$bA%85?m_ z-h(0I_!-Rg4V4oBg!8;@;WdLE`d9`0M|Sv;bcXz}?0@XCM30Bx40ThdzPhCd3r!DF z8xul+OX1xe(0)zT?x#|$ZBElqf*Hivyl-B?j&JT_XE zR;B>Qk1C31PvRH5-7+zG;7VuX0a`**Vl74nRPZGdorz`Mj4`j0Tj*bFSU0|~mFpmh1u!R|<50_p^_w@eg}PU-7yga{|29n0 zNoy-TLtt~-#p&&1C+6*yWXyOFsdD}oVpDOO9{Hxq7e~>uxfjJc5#->-T-?>eJ-q?H z^oSI7`dMbA4MhbTF3$2iaL|8&Dl>a^oB$N%-mCt-DtG@i{ zNIYlUvbAPC)Q~bSYOOw>`qI1!l z0InW9?wbDKTeNT>UN;?JR`{_)bj`!Jn~ge`+>TnDA6fWsUH`=u|E*^29RhsqfS;f6 z@6IP>;{no?5qgAIUhC6td|z+!zGlP1bom3NU~sj)>dLp<4B=crqLXI73C-g- zzb)O&Tt78%zAju*(MuV58Wi}f`ush@3a4sM$ii>$`^4ijQn%CB?{Fm@2hFO?J}rlH z_N~s(YHz-Jrvyb%)mtgM!qN!$5&gUV{Xy%ksLOxd`+uUR>hdH3Z@Z$&YbQ=!KHKj% zr(t=eqCFSh08Jg9|0jjmqMo)#vxNw)q+fm^naYv+zovX@E`sm#ZX_RdLcVhD4V(YZ zA}XW8r)`~!M5nN`7yg6l>R}zdaz2Z2QD?r$?6Katjt4B$@D`L++r7qBaQ^XJzQ~Q(s;?Jjpue) zQx>&kDN0+uBsM{#i z6;`natb74*|XP-i0XTZwgaA{F6k7 zCJZF|ura)7*zs;}s+8)_kOJ~a@Bn&6m^-~Ry*Bf=1_Xy9UE>hnVWZVk5}5zrX%DXk^FDJdfR*q_RQpXJU#v!7nKW8_kBv1QB-Wu`saMKj#qc^He|yp znSOff$x>a`_F(Sdx9UoaSv+i3>SZFl#wfyxvk z?K|Xkq^d@u*lF2gDdVLpkXT3(o!wx4S{v~UAm{AU zU2=tzcOCJV-{Hs=>tI}F?)o3rQDx)yZPrigSSn5LuM_Pr1{D~(6v(^umL|Dfr+*hV zmZ3p&Up`{K=fkKs(5NnH=7ceVaX98pZ_( zW>Gp#Co5~Wcfco3yP+XO&bd2n;R@kVgS^h;%ecxN)tcBFvT7!>YkEJ!Rfq`DQgbH_ z)nz?aoX#znVE0XJV9Ks3j zhdy=|tjD&tGT*&aavSpM0j@4Miz`; zi-4mDX-GZZh{!G_ke3>B#l-TrS&cLYJ$xM~AftrzBfmVS$djdJIEE7jSk8#D>N916b3d8I zb0Zt;2=#0zhLSQk&&qI@(Po5m_Cvs-a7Mh}QERdp_=R8mY(&WmFzHOQ6KChOHyB+a zRjnKo&x3@kNkk*E`a4|mpyys{Dq?N|+zwaxnBpLLe0>=v5UbpYDTS12C}#`R{R8$Mzha%Q&@0P;!%MH@$*P%)HelK|z2-05b| zYQT%FQ4Wiort)H#_r}O7gK9DC^{h->MG4%xqMI@DhttLq3bqQ!6sr~@(?`NGuMsde zON+V^fcUu7y#2ely3p&C7o2O`SB1;ws~5IdLb*89gmylyf#3kurpny{r%B&zE|$7% zQrW0WwAN>mT_Y4FQThU9=Ys#TiM^xPFt9Iv=fL*i4)as3ds~Q^(czD(64%ZM2T0!> zk63{iW>YscJ2lsWE0!L{D-t2Kg1-TeYTWIzCM;?7)h&WD)z}yUB2+^9vi0YRRzi@l zXPdo^4E0c@CQUx{1;DT41~VPqQ3urSz5Vi5(QdNzK^z+|x+Gf^qXHJ)QaMkhqEwv# z`K<9V&Jj3I7(+O**~3(~x6`|vPV?D%ct!a|uLYC01ry;2;iYD*x$v!a_}8vi&VL&# zbxen?q3oBvTCbICgB5%Y$2SL>4bQBvYZ~4CyeZ#eQYE#!xg&|oMph4p9me+JYDzTA zN@UibVmKxSZkezE1L{vH9=3m-(6SJJd8YV4;KT=bwvqjqD)sDBxE`zm6Dd`$-6FCb zm)c*z(Iwx6L&K9Nxmz`t{L^*rk5k&+Z+~RE{((P$jG?X*452;F7)BEZ62}5pehyiy zug>e=#NtVak)JrK8#K3_Dnh#D0{4vgoUN=p5IU{w`49zH9=S5-hkrSE>g!C}T#ql` z5+dIX7PmJu>^CSkQg0)C&t2w!&Y(sO6bxIs3b5V1n5W?*Q>)J>3OH$Mp`TQ>gq8hc zu%C>_4b{z^q=02;v2#SSMddW}Hg-&bC-Zw`FGOg_%TAepLs-H4-V={}SRiYI+j+I> zi%uOJ@0}#`6H4uo1)U7cDX>kwF;Uh?mRPL^`^y*a2NN3E2N5+?n{k64;LGIW+D#h~ z_a@FR7Td%VLizC+A84D>(1ualebhbddO9&yTA{^7mXMKJ!}1_GP2YVP06rK8joJ94 z(J;k|?%28X`?)CZG*7EvyN`atq5IpixeNQVdn#@Zy0!(*n=ZeS zm-xyHb5anszN;JJ8B?y`Gdf-nkO@j1!)VeJXUrSBok0y`)6*y((jo+Aq4+SRq4?R2 zDUHXLHU;qL$#LI3ys`?efQty^J8v|X7%yvzd(DrCF~Dz)?XfPg_#pFC+jb#{@d*u^ z%>xFL#_&voaYuZ#L_PebtTi$OvNJh^?IbMb_w4mNdsBqu^GhfZnfV@Z8NnPup2|-v zXk{nt)`hhgSK=1IB(iU5s{ri#OzyD6)5nX%BsMQGlDA*Su;HrhK}K|FzpxTN*DJTd zg0UTB$cm(hZsOvyCR=XqEgoyX8KYzv=}7$pQGX0$O%Q$w=D=Zc6^KyCQX|HjOo=mO^QReTfgMe?0SL9mr8@bx>YhskfU!4EDx%S}csks++W8s6*f86VL-(NoBcXRE5RSodLJ4%Ggi_Phl zwv((XkAjf*X1`#Ew7f-)$Nai!&xBEk%c*N${WYuz*B`$)BxAfh00k0_Hi@pse=h63 z_(z@htJa?>Kc(ZJ;afd&0?2=OhU-dbt&I(?S3&Gdyy!CtO3}{jF!eC$U#>lTPTVhp zxKXJ38QUJqxAH&jKJAphA)!X+upy9pG70G9W~sldV82163Zjyh%?``B&>VT8{fJSr z6vJwd8Z3%?E*9SVQRC~~TZl02rj^G4l8%t{Oa)Cs%|AtlcIh(FZuS>~(yZBWYnIsB zT;|25T@r0RK`t<(3E8X3m@s0=u7@?38FneP*_K>2{8uOZ+w1CF_Q>bQnf@mi)wasC z3;noCuqM;>;S`~iDzBcN&cO*)`AzehKybo_Ma@q%D2~RJVLj{kfe?5Tm*t|{5n{3T zUclh(1^KCVLI>58f`yR}J$(TNv@=3hwoALm2$|2mR3a8Qm^sW@ia%$j-!p0|o*8ks zWN#O%OZg5r8MA$|JD13FY}u+Bt+DlpHx*dy)BI3sgnvN7-#x^V%A)8Z1gp?-5|e*K z!^aPZVaZ`@G`;vX>Rhd)!~TTIC-t?S$(1 zWYoIQD&O&|a3A{S)GcoND=pIcnKSN>;X56gv(^f%u@aK5X}LX^Zp<*GYD z(zQPYd&Fa2^USJBG-HVghYug{tTCfc|y z?h2?y%mmS}gU^w@Vm%8}7y)y)2~l%~B`| z6A>>NQ@W>#NC*K%7iwX;myjfPk{~j^w$*rdoOOiZ zEkF9K$YsWWn}}_45%A(jdE4cmvgV}<8&T0TMcGO?p0eT)i(Ko5OP(hN*=7c=uC?Av zcSxT1Apo$;_2@z!Ta@OaJ0Tv#`$tU>>yLFmogw{Z&r#P93{jSL)lYy8a1im;9}7~6 zl>r8I123342RR*tHj*_N#NY1vc~Re!tgXkk`toofcxGcjXn0cvij!J$J|dM*GT^%{ zs`oI38^O7<36gkKKeQ|3$|(j7i6S-`fAx~hK&erJ%5z&VBT#~ilfdWcsk1A4@k0)K z3uxp#m~OOeS*P@Vmp0yLV%U6IVQhUfKuh7#BEjf5`zJJDl=l_egR4j6=He*$*$fn5 zzgQuO%t+(czCm{JLPBWFFe{CR_pobr@y!6HS{D9lKWOM+2>4V1i4Z74-z9>$avx-c zP^ewdp;`S)$FPt={L#8Z?;{!4TjhZUyG68C4dq0_MrR`v61pyp+iKVW>hKyxdw-C#y7ynOArF_SZiQ)6j@L7hQMw& z|E(4%bAt48mGyiw5!c<<%dWQCR7|!d&@FEt(K^|;J2~_qxfD(d7{!^P4fvbOUYNGGoSx3X!$W5-XvjjUs*I`^)ozOQkF#e!or#>-)IB5=-TN-~5j6k0o`?^U@pqH{> zQ(Mk6PATea`%s}aUfH`qOmYd&!h>2N=3R&e3@jU+kugv3YI{#0vJRX z@-iJW6@>6R@BIZSt8Dll8<^v2_C7Uh>+L#dTZ$z=dRuPXv1275a<+YQi-_4okkR4P z=nd7nf-VeiRw+i6-UUu=88 zPQlWVr5#{nnOOWpS;4}-+vy7M+3;r&P#C-07k*&lxijM=$ac_!FPYa+uBVGpa=?~= z)4|T{LAyD&b{$hPU|OT zzbv$=(S#E`&6Xl;SA8)<$q}g55nc1Y_Ezmp$-*awpbRHBpRG*ciDr0~_0Kn(J@$3{ zx~>24He$~I$^phnvTT1v$8fWc^(IJBF!;1GBH$L1^6PY^>~(R_ zX0eu=DrG*`&%ANKU@d0fDx>4HM2)i^`h+wwLF7vXI$%OCsAA!jsWot@SJfa4z)$xU zxxUEUIbi@T%r?DcNXLyoNYC-uQWLIN82oV|&l7wsMzGLsWld zJnI_v%7LrFTpuhfGT$`6Z#dWEXW33pm*&@}2+|yMM`)xevZQ((;&SmTux_gO9nQE~ z(f~H$E;HwaKlro@o7bhhkX7@ zvUIEW^#9|*OIBjtHRfw){=@Aso(|6Qp@{~|LkkVAXCCS-k;R(lRHqO@&@vx77GhdM@KVw;l@m&+x)V7!IYKObH`CDwHPnluECN7@ia**X|L|soo=?0K@W_ zjU{9iE3h-|DHFlD!uZ3%fSZGYKeOFa3{NJZ=-oJF^0tjh$e?D41#{KX7Ct2}53;g7 z4TM?OYsjEHAL(5o<)=t1cmTd?J#eark2%LHL*GhvU`$!SV?l~h*INi0@=ZcVKw9cJ z)Z9mVaA2RFOwrFrKEznoFizR&X@NK|)I8bbf%|T0QOE4@OXCRK<7@#xRUzgAr_2__ z!#w0ozcS(sE~b(glpJ*><&l(e{MQ1hSTy75VeeoG=n4!aj6*p3pj%qMEW!*PWat7A4&6^O^SrtRMf}FSsBvX5=%w#>Uinf*ZrAPCku?UE|AOz3w;6%Jnhp-)7E-#+E8?AIQY57+BKM@y@%zA(H)iP z`!DY=e%V01@Xun3KhkSyLci{Ymw-EncxN~bh9vqpzQb|;G5W0zVbW|Nw4|Z|3d?*h z_7^zY73wxa%HHRtq#d-n&&y-PRX?VojUij>;8afXkMJ|3V)Nu4n+bSLo0CQkve+i1 zb1)GTNKEp*4&swXYclxCTRH+!BCE`Rn<*x|mpP_aeVLt#-6RTN;v!@k2I{Ty!d*Bp zJ_(SLj#t2y+%om=Mwmi@mr)dwFw{=4j`{dYM1=sV>ORfaR4QT_R`ubHj|C|XN9 z*xVwPsbd3iI=eRHS2h6W9;5o$=Wb4E5bWT!ZZwsLeY1S!XLROX=a+mhAR{kKJ1G)zCK+z5&KObImt zL8nD^MtW@@YOF)%8{7IMPHH$TEnca8x&C1@U!kal_dtJ|%qx}+hUGzpf=~B`J`xs? z9`w_mXmX|F!;LAsEl&OlBh7y9PbNdCK^+Zxno#xyencL2Rlil6tq=ZLK*RBARRPCz ztv8mGm38M?!v0Z{jCMKw>}-i_UxqNAuZ?R#eN2*&9`X)2`j&l&r65c0Na4%J)Pb*k z#H^T87QRd4+p~Gc!UbUGOh(*1@nn z7_GlXPd}}6z7dopRgCdi@iaTuKxk3lRPQpXK$9Pgni-Cn@nET(Ke3650y)!5`1xNQ z^z|5g;0L_{3EiVJ(gMvw`)Ruo7(k+DXeZ=EtwF#xNy@2ePJHyybQHC90JpO=l+|3` z5boB9+(_x?iKX6#$X5) z-x|Ou`00iLq?K9N>ZHZq*}KL%ZH)CWDlbdw$p08q(VRDtbMx_8dS?m`Wyq#&JVoHA z4wr?RDwpn@y5J9n-OUDW?w$tf;n5AZ0sYkzl8@_Oe21ez{-TC(|2_=1{hxMmTEGcZ1p@Wd(i1rquQ5``r%!zbdabVk!x zxRV8Rv)*!zN~lN}4g=k(lw-&(mN59%u89S(B_R|R)|EMZM$1*^fN5(p+Tn_WI)2g0 zFB)QIQ1!UcEt!jQkL3>kZx;L3FRD>q$aWbxH1%it$pO@Vu-kWK`uSZjn5rPS%dw_; z;hphT%kA<9iyGJu+mgs_Wn5j|L@})7vO}8t%f-UFv<7(z!E2j}fOZ%YH?h#O0V_e? z98kmJ8Wsa95(_K;L>2~A^IxG@Z;kp67k?-ueJg1I-+p#QU0pof_h|zZWY6!wfR+}K z7WjU4Hg%{7j}3{?q|i#_#!L5H7T>r9S-#?U7mK@HQ!PqJ>LEU2xsxOBPAlz~Pm8fB z~SJTOI{8A;~nAxtP*M5I(LkHIS$k$V9G#uc|Puv^px= zPm%kwyi!tE;-FBLfz>3@=FbS=PLP-^!LFnuSJ7b6_+`NH{57ldMXu zCX$r39qr-Zq_Ta+Ft6MEC&$B0#p;|I6{Ovh8-M!u>dc;0_>)n3q+zjt_ieC@+|jm> z6$<(*l*R2W&TMTOWnnhPh72Zg80}IAoUz4sAKTr7a;6PQ=np|MuD?hg!%C04uhbWm z)12_O?@twGZ7e1Qb>ob{sy{=Q;8toG0Q6 zYte(1Y!wwP7*j=|*s}RF4%sQ4^bE|0L3HPkgN5&Ki`O6O8U#<=>V_U1J-;qHzY||x zGJV6*JVvS7=@?e`9WFcMS9t!8%qKpEqO8f4#}oF1uD^ic%zgHwQNxkUK{^^5M#P&) z#)^w<_iaEPrM|JswC#Dulm}f&Lalowf@5Gfp~{06e0HT4fqfkrXMkNdTDP+DR<2Dn z>q9a)O&Xr`NWX&1pbdpi8_v|;0A~b{EKL(v8LlQCkRvWI!){HU9O(89mNJ9`FqQI{ zMb5qdHzVUrSJ3;Oes)O7o^NgtD+J6RZz&Awi7$u zp$ivjwaKqMIY=bq2;H`TFMX#d4>Bg%t~{~B-?8O*(v)dxwFf<@Q0t3ae4A_wwM_+) z>yK(85CqC15s2>So)2e`;c}bDuNF-5k4p@%OsYK6`queH@YM}Mr36_E{LDF{&e7A8 zNh;ma=`0SfA-i$}1xFhFS)#U1({~?d-@C!mVVbo)33TkaibZG~&4wd;Zmkv8_`}XPnXbw7md-2N9sa$SE^A9m&UwjszRW3F_u2f?UJXEpUQYa z2$`tFq)2kmw?S53#-EVb7@_F*Yt?ZXR2p4L(KjAOz=1{EG+9M<7g&hxbk}u3H%RzI zop!I}|F$2S&QeD@>$@_}f+ccp!+*I3(nS-ck)^1OK22V9*(pQ*|A>3bsH)d4dK3`} z3F+?c?(PPaZi!8IgOqe{x=T{JH?;}r+>|te(jg%&<+(r4dC&Wv_}@G3r+dF{hhyk| z_OqU~=A3IzfJ~K1;6mq=wTMP@ZF+~GiI`U>n#4+0LH_uuzibVSykXNPE^*{OO?|`6 zKAQiF$bLR(?@1Y9mnyJ*$oxXM_=s-ymBOk*-idwd%t1rjX1BCFIU&E*su!gUW=f~q z%HXW;*@GP2YcEmvs{O)I5z@$fIZ#ofHxU_l#008(Y2%kGSIaiP*_fe>DMDjE&?n1- zdGI}>PIjoUX5JP*7h$5<#RmNho~eh8lzMr(53F44VfI>mVa9*!Fq@+L1BM%9#&O~5 z)X_7Y7|NwN&+(A3r;~h@I{wCr(DyQ``xrI8b*;$O;%6nmTBMvNePB-lMVfd^n6Ih2{Qz7x3S*%1Dz9`^j?Fhg9vxDShQ+cWS9ic%mh^BqYf|fLcl44hFd_r5 z8}pvd>L~G>VC$XDuP)v|9=0Rv?w+HuEU0QE7iTo>na356X3L5B@b0tcRHvSU6BAkD zd*9*@Z9s6?=GFpEYB2Z-C~H|Cv!I6^l2b^Wz0}kYh#$;u7PZ!x6WhDti~4Fgn)CDQ z3*)c7nalR@ho!u>DItwkX8GaC5W%MC>FXO4^|kji8o0`B%Y}P}*rv_xF>sn$kam$k z9HSsSleb@ZcibitqXd!|HkQ=5`Yz=aQq0w9m$=CvW@SdH)#Pz9AHs3=uiyN61U@}7 z40F(3m|J@Ueu)w7j!0F6(A`Lf+)$vFvVc)SYP-$`(tOW-`&9YAUgSijK9~V>ts3G) zCMJD@yyL;ySuyK9Z+?{2xa*N#4!N58FbAe(tPpY^MhqD_4rwgUfG@+=2+!y@$<1(b za(8TOGwD2=8y2hn`rKSSYs^T+^&`tZ6cHOK=uxg>B%h*DoNp;nv&YQd{=Y%G1@C#g zzE#~1&xo7-_IL89oCxhM)w2hP3C67&CvT4DYpc_^xU3aKT$sR?<;$SnIP7-1V^2Qq z^R9=TxNY8Wb@7TK@ebfOoUdZ(r=7ZZxC5SlLk1kXj?lMDjPY5tqHq`4(V*=h;hxkU z^1Rc^#ct>pc>WQcA*gNCDVfZgj2@8f8B*!Es_^qc$x%#;FOc zLC8V0eUA2sEkxFZ`ZU{guD+rr{>kFYB1P+(rum$bDIXI1tg;ESsBPaXeHOo=cLsz` z<6_yWcf$O`jGHv$KG))lUn6mKILPrgUu}Q4DO&X4+9<-lmib)FX!HJ?+UiY-B{HW+ zH6%nbv_Nua98Z$v$p@{#y?9MShCQd*@R6J}vk4t~8qu5^WlSW!Lbdv|uc{VmiIkT+p7A6?nb6aB72vQ0 z&s`G%pR#|dvbKR04R)aqSP-14x*q!jd>WpXJ)p~tP6%oA|QE>`^(;Fc!B>GhtN-}OU^u<$R%d6-ai=j{f3A$Gr zn+f70KYb62w!samsZ}iWlFN)gtei*r6^N2Zi7>Sm$xB<;MqYFh+i>kI^4M10u#T#@ zYu}J;&&d{AUyY&KNolfrRT&x9bt2$+knjw5LcH0In#+|VGHcWj*-(6ut|3`Seo7>RD9xok@)$>r=A}J zv7U%{{Y!&IdlO%1FE9kdWK}R+HjOt>D|m>RUa2V|XlJ>>lae4>41sJ$rbiFYsK61! zyE1vAgvP-%)t@0O%)Yt~Mp~s|`n~w7uH0+HPHaOsn8A4%@$KL)9aL(}o&FY~6mLm* z&^m3e5L}s0gu8Dtmcv3|KMLcnJqA7DsOe4|hMsK`q8-3mVEhI*;yLZOxC}97?j~qW zU!2$5gB)(>cTc!2S6M=KKBl}%5k=wW5div>T}|epmD7X0^ePRXyO+PgeLisfqfoCI zQ~GCYyrhmTOMAG3%+FS>YTM82Xw(D*B#={!l^54c7y>^z$ht|(slh!0w(Qc+ldCmd zZ@YPWFQZudo}s%V*%UMM0KokBPawG!0Y^oANBsK0EID3M9NkaK*ylClw}a=XEpqE* zeF1AGD@U;)Lmjnbm^P-G-7=)gY6aqp8s?kT{6xav`)c%`{Vl~yTAZ=_4(%U5WB=81 z!c@L>nK|lQx=;NA5w@E4*YB!WTcEBlXQ#vp*qOdnS=hhDnJVA+{web-!O8Iq@{N_l znVsnPZMy|Tw*?l+LV9)kBK!3+;kOW_>NmF(WXbzFgpEfe!8tJCvkDH~oCX$pw6 zT#0yUOgN`DM2;R9IUaSF4OI zBI=a?`^&UnJl`(JAZ@8RG+Ft^%J0Xe*<8OhLW(nxa~8B6xcWsKRQ3x4JY>o+_qlT^ zxZ%O;Q1k@c+C4pj&y{~V{y0h#k@c^YJ;O#I+_FEQr;6_kKxFQk`DWBB;rV*|`<5Gb z=ZVK#r~zv_0Q$sLw}y=OmCWVZ#=JWryb3N+ra##FagxYZO`E_HowO@QuqLJ z6&!c+5Mp*2jQz$xU1mp3vd@e<9R+`uo@fs*Emd9$S@dA3ylMW7@diO_mshAUnrkoz z{5sRAC?)lk#)>TH5ic{xAML9TKa1%e?0=YX?`hAc3hCAWUqAcdI)Rf5HmcE-?`00g zLQh-KWc!pQXo-mIpt@}eX0l+#>`}$E1U`C8 zP>i$X=vdZx-isKXlxN&n>jJ9ge?QsTo*%#p*gc!0Cs;&0_u^2P z>^Hc?Ul{?rnFvpt#ij0lW5CHUJzce}Kwwa4}t1 zUq~F^e425-+Xs^3tG7y4t4dxZ>!e&?vo90qt+8?#60VB643n%ZeQ?l}V?E6~HK?gQ zFiU@1V<>zujg1aQ0e1}Sc}C2iLGC$#P`keCxn+w@+a*WivQhbDWR zNI`T@2^}9T%#(g3Dudr8_U5#0BGm? zu})1tChVmw)_oS49J$g2lFomiTyrXED#=E zHjUf=QR1CdSPon5d~UflwcoIOX#pO+ifusTBGYqD=7HH=>YR4E>QHAls1_*4w*7}@ zb9K@0?J0-x&bQ7k$Mu#vL*1G{4Owr}sA^Fz7pp|C?=2KRW7iC7>f;fBRzZ^LaLXs9Et6B!@+3-kZe43BKzDk2OdHD;H|LzN15oimVHag%9!%&O7WD)r(EoM^^7bW#k}A}7 z*m`CWa9XMAgjlC9I`JqY7F+pXYmdevDZf_g4?zF-4VWLv*0-=%ZFn%tG3A+G2zg{{EH(ED@$e$ zSd%62{kfKsqqe`Aik`eB%uSl5w2*E@Ms&+=sj6-F?ToGxuuFXws#zCV)XA4mf;J8~ z)Cek8b0(o-Ogu-b*GE7Sg=4JW=w zcJ3DL&}->E!CJ8iL3@EKm;Cw@MZiRk1M-i4%9PquJl_03vmVurBNpNqSP9%PUeZ{zqsDl*q*yH3<$QGQW(oxV?VfeKmA(&o zlfa$jWd_Xer04T1`NCbOJ59QiHy0nA&z94!45!r}Q-BUC;w~MKF>6lbU9C2#_#Ovn zgJIuvKhO)xj|nnDE#x?=6uY~9?H~*ZRi$ara!mwpuRrD^0boAjc#J>qHc?^F8CtkZ z1}9;^=QP2}sPJcd+o4@~UB%?5>u2ay1J&_8*1@0uPwA1gYDW$gN|J{OZs)m<3U4KQ z4%2=ED=m_y)Cw^exv97PfBdn`BDI^(-*nl2CSW@ye(L4eek?SXvHcI}Ix1*d)kqK} zZ*)v=`_k~UZo)6@r(~UpFZuW#n~P9t4EZ4rtJ)NVUF)Mc!mNV`oDPvrb@P9_GZ|Zo zZ(Lrk?u^~@0}lZ(#_gNj)%XD|F1SbfS6&kIw&tgQ%$L>+9rcBpSRK8i>y4k>6~OgT zSp0Q3q+WFmKtuhF)ysLQV6X9d(OImnvL?XmvaI-7O{((~yA-yE$Or>~88Nv#wYTrt znx9>~dw1ZJET(<2kZp2%I&!Vg41s7&VLH_%HbiWY+&xpFr4k=@5|9r@`7r{XsB4O| zS;P!Go?<^mL?B6Z1di~essoT*gYjD{fL>^{Bv@}m0zX?fR*hw_EN%0X@|kU^PxEZ2 zfvpJ}Z{7c&e^63ty+Zmo3{^bI4c5n9H%P}}1?8P13Ur!CH8|C%a5b}wbgM}6H}VBtXLxPtHU=R__;l8~xb;~EA-#s` zGW=8ty2DT1wc&{oljbCa$u2L=`rEzYiFk1(6uh06qk=R7+E^Sgd3p4yBOwQ1S`OPHC??pqL*;_ z+xI6{f=(-#F8^sra#n92i~!0ik{cL=`s{1!6BmUgVu+jiZZI}PBdU@MC9xl|dP%oM zQMJnfmhj3=#rn!??XSWHd2?EPcGicskHa#VN}Eo`6Q8gDWBOU48vLCl@$S*aKbH1^V!DMH8 z=~9e^KgMd4CV6ISrk^DFzVpzLLEO$3)!xvNskKv|<`)!E8dJa?^?DJ@O{9jeG=?gu z1Zi?hRuSSA%ahP1=&e3@5^9iTde3W2y?Y?x0F`G@$eaS!aM=^t_@JVm~tbzi5yY*hj{@}cX6>fhkDwXU~L&ZP-FggUum zjttx7zKxd?9e=G|5}foB3Joh;OCFqeuxpSmKLl^{6y`YD%~xp}fA8pDpCqqS8_0A3jY#;Si8nF4$% zOe%fs5OK;m^QX#9oRJoQ+}VNk6OIzpD;BX2%qdu4*&Iwh`jX`X6So7tGQZr)PM^bS zXliWonN{h!8z|B)zW3#$6=h9$9-cuz)het_vm zr{p>H>fzN?e(D`hjZ-N_2f2<4?jBpx;NZUBtC`;+n!}wq+^-0rK_jPycjD|5RVp6O zn0)Zzmdj^pdqQ_|Lmu$oL;@}xP~Jcx5TKJbd39B{Y7mq}~eX52m5b7*K37WGel z`O+yq`+%Pyt-&LFZd~36k!mX_JGLY(zE1>__wl@?J&=ElKVld`!X`$ToGrA}S~9$L zprKC@8H>*A(Ni&oa=KB+Kq)$TG7d%6Fo7hpy_Z15yL@kj*Jp#BprmG5^bQ! zuFU(N2KGxF9>}K--6(Zad=uPn`31SAUawvpSSRImI$7H;Y2_1LK$~KV*K*1Pds>W( zKGJgyAN*p{R3u)N?1jawQ!sRvJMPsfwZ)CTN^K|zBj=KH8^n4I;DN3Nx-5dP%k@*HR1Fc(-A6BP|Z3+a8mc%w~*fZA=IM^syM9k?x zJRF+2GG1IG=NUiEMzo9`s60Kf_h^9%+f-_6WMmH>30>CEje44MDNZ|@fhJM_d!jz# z0}&)T-tMYx2u8j|P7w3f zA*wUrZ>m?)gaLbimdVv@#U6$Hxtc`{cUj}hA3t}$J&#Jtb5Z_L$ zRK&KOLB7!t?=KxRi5Oe+Er24Td{o#(G_w%7>=VX{HF789m zxfqkvgmdkBgs=TkKbcA19HjEr7 ze`VDhZ-K`6@WaP)ali|#=@<&5CLz6`AJP*W7~QC}rK)PCpkm49j(W}OMoIpZ6_=Ui z33V{&J$J7(^&JQQ%#4r+tXhm#d(guphYnYfaT&S)d9Qj?v>B%1@b1xb2ya<@bo|FS zj92396bi}wPIP?vB0f?j)znYlyII*!XwnV@i;-sG3afE@W$Mq_X*8z}$s$4d;5aE& zo0L>_NAwyM3C%==qHG=%#Xc&C5hZ;7(^UC)?~ergn~^w>&{zVI@0%Vcw$eSpWk47S zeM+RdAZ|XJ8jdS%uzvDLt>2-R7!s9DcJMbmy9*ew2=D#0d(3q%oda~BVfS}0$#jmO zC9bYJa(5WOxAzz;A_ehjB0iJ1X{~--i}v%=qFf^g2g!t5R~dJv58dhr9$im~7Z-lH z#ppo363yIz=(N-^cU9b1NjO{fh!+NJzXjN(l%Cs*iuBrnv&-62VCWtNoxW|shj0S+ za4M<)!$y@1hA$pH|I;r#pM3sAm8ii#^Hg)#^C-?6dd<;$&hs$akoWqbNf7A7)ZP}i-T2X z%=8e)^|7#=MgSW+i+$Fgmz?|pw-!-n%@HTDa_HCpl#1}%0*_4 zFSiDF0F20%6KId-)rJ1H@^JpNu4QTc3RJl;zw-3am{H>|1H3(j_toMr1>Dq}ME$?c zZk+A2qBWH}yC^o*IfGjZ3=VvzXrh1vgQ!S%c$M)OT81SlEb<_dHc=;r=fz9nES`08 zMM+H3+Kv*v+hijPN#>CZmT8Y(pQ!QP*u{uw*9f2@JZN__JEx#YvUb8Ym7dwNL3fS& zTpymPnDTlj5()F5VIq`Wati@f)=h_2rq_2FQ4w0Aj4MngcGrY| zwB5#3V_CJ{xwjOnp^R|!s!oDEB0v6FIsxajjiEW=5{tbn}3Qvf^x_stQE!!#JE56U>3zkL(NK57 z@F;U!m_Gj?cH(lnRA^5J61+L5OW-lBrDWjg{{N2XXYzk9c77IO<@&A4#O|rwah?*L zN8M(_^(_EFrd`*8MNItG0Yvvr{wf1-&0DhpGn$jXfQU>d{cHcd=KqJW;Dbz=hw0;< zdl4EI<1pcY113Hn91!X#`O#{3)e>v{5#BM~XTat18Q10CtP23QZ=mrB2|Ss2tg3f9 zaDL%XwFl(YtocwrPS|j5+>6qsZ$h<<)?u%k5}uj$faVbnkiWMy=V;Y#an;}|QPN<( z{+dqfW}M@#Hf=gz11dGfJ%J>?#B)d|+)cdXGzP-rT0zj<$9V^?7MIZ2h zjzNkXug$B=p<+F{5KoV@aSrvr5Ez$f$1m5EEmsrvKBXcxd?m8*=bm}>1|?=8aOLId z_7q#?Fj?h;uk$Z$`^|9+L6^SpeARSF&<=Cy|ELW8pxhEoQRz19m)OUO)jNn`bh|^H z*q6{qvS?AytT)=&lRQ~7+_faiHYtnSv})gBykf`|0<@{;DvE7nzoXmffraw)$4Dah zYERLh_tY1LDC(6@(KP9|Ci^1vXvN^hhUG4~U*)sxm!IJp`A)mta0icI@qc$>|8B7t zm*%22$jP|j>o9zRA*^;c4w#(+;+avm?S_dmZu45^z+amu#YQJL&~Dg9 zLC}ffH3;E-E@M8b6L8oJG7W#IiPcD-pZ4N z(LtKT_AQMBR1D)j@Q&6ylMBU=RTiFCS+j~mebIVOcs%U&f?KM_T%J@Pg~PB?Fuk8r zU^{Tl`Jl&l%-mm%%39fU8_`OAjEfpxXD^FRaD>LBQP^bFvHkfDruNPAUmEgd@?-{R zkX^~pnN-^VN}rdC#w3U%GdHmnpf{}rpDNGKm0R4+0l)KL}3-FokiPSw~TIM;3o^VJL zs0_|5q?x>F;b-D%kYw+!{(SyYWzMLo6x6>fm_J)@@(7)IgU`YO5s5>+dONW9f|+~VUM|{ z@5U1fSLxo8d9d0wa)*2hp!HanrkDPa{8sF(_H@y3JhEicDxcS;yp%q+yTfcREwrz>R3lBUBFAVz$%KjR`7R3{qZyXYcEq<`v3J zsVGl1O~FQ|LN3KJWnI9KdivdHi}HsHlRVgjpM=~a%j>&!d>>ZS=|*B~yH+benf@_@ zJH%&~NFsSs;nGHJ+NPPaMr$rcr)p7Qf=E8O?tD?PrI#yCFlpj6EC2p2gppMLm6>dq zPf510=tpaZ;>ko)Y2M?>1WVWg2W5h|5z%vHTC*o3P27x}@9Bn(n>`O`s*y(e`9w*R zxAkxC8|(=hBcHV4y>LTVzD9h~F#?wQl0yp1&12O=d|eWUi6@6KM3jMwA^TJPP!xxA z87-Ev4||g#-#mMGM2scZC*>)8VT%*n`}XI&6KpFZR{y}GBo_bH9QQ8+Vvkwx-{5%b zHtOz#P8ex9ANp5f)uFkzgrJGLeBk&N*Sr|}$7br9a zMr<%c; zQi&udb#4%JxMXr=?{LfOOCLb3*SYtEyN)qyDn_NV($Aj64q~0(F(^L111zuBU8L!Y zd^)|<+b@Ckzg~w?i5r2$DGziT5?;zn>;f>*e*_L3)%gJJ_U-hK;VLjo@g9dLKE46B zrw9NYvn}H5hXL)=-*LeI)AxRD_%Pf3shHsPti7`ssX<*!3^2Yr(G_7)nvHg9=1^5S-37lq3 zzmPEwog3qQa!XF;s=yyta_uy_X{%1hUwqr=f!ELZ=_@^1<}4;9ST@%tpn{uChF3=U zY>!m-FCz5|>7)ju96y3_oJ z7@c#FQa1u=`EPL8Z{~j~umAw!oc4IZ(spYOsr3RWwXy{Fxux6iI34p>`|Rvg7h)le z5HMjGTUEN_Z*T_xsc0VNFoB^5ut?3z32k_t zH&FtwZ;ghw5Z_b1dv8Db0VC_hJNXT#EtKX4P77-sQmxFFO~G)Z5#e0R%4!eLQu9ep z2WnhW8GKhRn)7MiIsp3o<+%X{4$~hq+I=;r5xUnnkc%4eX>{vH+moDzZh=# z8ys4|e38otCPNvP4VR*-qPt`^qe_2C?1#8TrO5&Az;Su77QQRJkJ@DcE!yCmw=~cf zJz=IuEpleXp~n;UOgScy7BBkDf}haEMtIEz=ot1&GP-?#Dj7cbqC9QNqfFMf0)`J# zvJ5!F@85GvrshM!8;>Sm(PK;!Czj#qmY|CT6j?lN1ZVK>00}@>^XQ0DDJ^Y*SASya zG}P>PS+NtNZ4ag1NW4y#<8AetjF?d`hFL7*QcG79<4Iw~il27sdDkGHZQS^W>8Z;m z^WSguYtWPyt_vahOJ+1mF)J6~#*(3ChR8;j8^}BM4>kvBHl(|mNT*zGzrfU`L9xg+ zc!g$gPCG4iU`!tAp3^H){7;lLTMjpo2ZIE&0_qsLP&1P%ldU%5ijfc_9)ORmIaG8+ zio>aFV4#!%mPM_6T|I-B9yeqFwIB>*1ofj((QUD*;<$t*kKu4|STn!nW(-bSW?ZDk zE;bMwRCSP3O{Q4S{K|cZ2ZR&Xg0Rg>NS}&wRnR0|@9GjEARP2tbV&R-c&heY_O-;5 zux5nIcZ$7!`AHYvLVEndS5PG_tdniU|rKp=ZKBO1v2LsurIuCHa56 z=RSu|x+Lr3ej?#qcw61Dd_MW7#5%oY-e8sg60e3UK!@r|>v6ugpksW{%8swtQziqs z8}3;re>aygizqAX-Q#o1mavT5AQjdO>CLHn?!oA@?e*Fn{l7tv zB_XQ~Gv~@9mfv=MgFANTleiVb3t)$A84yodit+6=)8KJ56?fJR4d{%am*_UROz7>y zVlzPO*VB%)qh->oA=B%+LoB3p-~eafL2^jFsndCK4{~uVM+Si zG@ZLMl`mq$poGL}g4(!h6y0YL2yGXVAB{w;zIWopz96?o^1x*Xt<2R574XM#LgU34 z()P(NrF&<>?&#If81MS{?RTA;^&z=@VY>7y4AHi>i>RFN_v*bpfNdJW#^rvpLTf4& zb0anzkFMJaZ3m_Tn~5pmwA}L!FB}1E(7QK8)L6ylE70&lpTma69da@FTlj(Ex;TaX z5yLc(OSZKUjNuE~ZICA{CgkbY@}?N4`d*m8li%R}bpwukW0spo!kRi<(P4t9y|kbW z3e-f!m4@e4`Z6=J3OP8NLe%;h(WnRmV^V#Hdj5uj`lZDP(yT{cr>l69A(6-NHgN?c zKZsoEsZDc3lnx2prWhM0*uE;gMaj*=Q?0r%pl{M@NZkCloBtD@p(0_)ctK zA2SeT`)|Z?WD{T}Q_6Lues`oK;9HQ;;KELb;MPBa@xp!-=~@Gf|^iQsvMJp-^W zNdUhJLE^B%Xl6YqbAjo;T-G>I8rM(3tOO%? zf1iJvjBsDb5mAJTSq1uD>?%wdH%uS*Lzw&|G? z-;Qeq3+zIo94BEyTs3suX%RPR{Y-pn9UwEKrpe#A@^H2B%kH|Iv+O{LgSpZ&ZsWM!55N|;rScbI<{!Ot$aHd~{u|uh zr-wyvyOTI0?@|A}cotw>E4;bjRXxt6Nrjc1WcS>)U2mmaKvZ2!t}ldyHllTK zAKb9Myfr^8n@j(wAkfuNOgDolTI94~n!bR%mDU6~hd^<3pEtF_VhlFA{c#OnR z+<|fb!W$+qyuGM-jq}a>CqoTKc2at;?U*Vv^y>=$2AID1wNSjAE}~QI42fq5daxP)91-bD z3oEhFj@AcQ_MOD0=bed6X`mh4MF_;pI=~x{R#-;BD!*)AdA|^SGjSB^MQ~)eica62 zIZr9~Y<(+zu?C>6B7?St0sGfGp`E5}^_7O=y*yL3w$Pf-^^_a>Ir%r-6zEhlGg*qyqR;`fr8d~KX4V~<9Sl~cAdB0@pE z`d}}p)Izx9aF!%G$eEOOnraM9HU3aMIOmRcZU=ENfL7^x8qWh}b`B;j9EJq7Cs$bg z(7YS1TK>lWDrDa*{n=mQ=^{qNV?q0lPYx&>5eWrneq0Np+iBC$x<64nLl?ln-=zUO zJ0?YTklA10^SmZ9;q{=^?|er#)g^|@M{a{QK6x!g%bBuGnU|trj})s?cm&_8zT>An z80*NvAsmlkiI)*=gN>#psy2EwLpU%gL~QPusau^fGme&MArv#bn#LYXBAxCuS8c`i zOsUef{TkVbQ2>FA>tCaSSmSUduc+4U5xwY6zkBt{sR2(U=jV6h#=?4`NxFDKgv$fX z`C!*^L}7U^C}V1z`*Z@QQ$5PZbR*kzH}4_~9L3?#e!7?Q-gGi=Wob1J=MDmv%M$}x z2s*|DW|}VB(oL$?Ab*&=Z|#BK&p2O6e2<;V7n1(yki7WVO7FYpP;x5AUz73oSE1UbY%kZ%U?F*AJ&9pO6xQ3!kGqwa?Peu~F6UX9r z+ewaW$WKwnOYrfMkg+Udr_>253TARk^PKFvG1$rC?Pzt zgW%peq=kq&4&os9t17@KEu8jUh9*Jq6Vyk+iPf6V#L7qbHK{b3L@S&nmfL`pz0ToY|MYb@;Qz;{un2Hkt6tbzscN(ojsQtxk++7 z5mu_o@W}$Re$R~Ri$g0!^@4R^U1iBX4{cD04Qo8dmHe8tDxKRZ_n?+j?@zrmK8sa~ zOMAlMz2Qf}EgQ@Cm01+~JF~M=ckCN(4(ieFgc?HVboiDHetp#B)BG5Os4(tuzK!b! zQfyXQZmJU**9;{V1@_kis%v8`?U(<=55!86gt};uMzNk6=B#JQrP|}w->AJKOQ_Yx zEe_oc6?4BTg1BL^4N2paFO34#=ftiqtfj{1l zsC`^J(wEoln5sa!RYsE~9ymvRpD1|IpxP$X@_J}SZZ2?w9u4;0&p)C9-R(LjytjQ6 zFkC1JA>^0lNvq+S70&#CY7HKL`R_g3snHY!&T)zhd4n8p2lG0FuT415WD>!GUTFFE z{$CJiNA{OK-E-t}j-B07dQX{T!MzIU?Gtx;+Yk~u#SZ_ZPU69Z6TNRc75;K*XUj^p z_-jUU2`1|R1L2EQf86(U+fr&_?cVIt#=dmL1L18vQS}%zrND z@`_4t1w4ir|3iuLNdvq9GBjOikD}KcA0@m*t6EG}*MN445%zsa^}EMZG#j}rROO9v zc7#$zh-n%Hpy{bI5_|LpV^|2Yk zuYKcnq=fqk#s7Y%0QGiu&uydkqSKUH))(M`2T}t$D%UarFO4@avy&_L{x|#o+rCT# zB>DZ;lXX`7i~cFckZZfe#qFNks^8$i1E=>xLDRF4zx<~va+6jy&RW}(aF|?*_2yi! z=PVhHb6NGhtjdH$$kFjgY2)kPuBmV_PUNKv3blxx;HxEVl%yMd-K}HVTG*NfDHYNm zgF*^K0JEo5+QF>!@F;NsT&89DI38sUDzD-gUe@*VuS-2|Z`&58&>Hx-ndtHHY&+W6 zXUpg#OjH}*HMBS;yzt>GdFQ_OYjvnu-EKiT)k>KcTL{QB+t6?=ODhBANKBrarQ@>8 zk93Tx$+8}$W|~H#8+|>4{5cWR>u-fa$=INLQQWXCbgxCrPopZpVU07S6I#b)&gJbS zCp4;UeN?C5xAkR-`_AR*(wO$#SguhuMJ~6)e=P#JIw!15j%1ye`EIbyBbGO-$3c#R zg2?(M2AH%d1l@q|EJDD7IF&D{zM8hr-D|raQETR#8?qS{Vez>~mzQu=}>ZsQ^nrNYPj4P&avm*%@&$2nx`W#VrIJp4uQLM0I1apTR>Rvw6=FGjf zm;SxgJi5MIZjn-PB>emo(`@<&dOe;uJ}>dO+<)T3B9|CC>b!2TD61VS$mr^G&^HRT+?){#p+H zwR4sxj<)twk0C@SD?oKK;5t73xfQkM2CMe$t1{rg3SXOGDzI;ay&Db)df6C zMj7vZXwve;p-`H0lXbwJDQkM1SS=@i{5)pNS{@%qc0dhU>%h_{+&%gVVQ|Q4@1|M( zqO50Lg_o?T*6bd5Hml&Oyds^@O5g4l>O@ILem|1iu!~}JHd4*|_GvL+F@$rYT*TvK z3QcQD!3V|NT^qv?!36N$H9^>puG)k}&=$5gDGjAE{|THYDh@SnM$C`qI*|IpwT)y^ zO0r0)5$=~>kmH9IM901$q)scrmUk_@&X+R$DwrC)QvHMppD-UB{XiA`f-s+7jdi*A ztZ`hUE`tImbyzk!s+|%x+ixO7j630=B1ZE3M?1|B^7ivXKtlJ*{ki#rK4n;jYtjMc z{zb$Iug44;wf%?nqLryaNBKsFQ$W~ELqfgSO zdcCDjh-LbF?Bmi+g+!&42$B6C=mn3MdSI$h0bWDbz$2UDyp zjYfMhh9isF%wbeK->FP;;}d#8nTk&(+WTV~SdjC$orW-7P)+NjQE}HTWATr86H@%d zOb5Ye*^_6jC^msBpN0Asu16#4!lw{ftUR0SR@naMP4 z83Kh-^DE79)jlPvD3Kqfl~H7E=#Bd!7~wsG7G`L*BE)s`2{x0I&bX84|IXOq!*u%o za^F~+e5wuoijX!R@{UKng3@|zy2Lcje>u*l+fYR=)f~KHhtJR7Qne#_VTZ>i#qb8+ zMCdtzug|__J#U z)^$d%CIU14ZVQz-hn=g4mInHM5?=3*A6pyE4T!9U=xN{Iz`r z@YCBtxQzom;ap=9>aUv8ANapSG?Iseo&^hZvxG_YhV@e>HEg^hz7SU?a1G zTF)4n2&!`rlJcN^p~0Pg${$4oPo8f{eH@*LLu5`uZh{Kc(ilWg8_v!vJd=O=#^M>7 z!+Jkqm;NQl?^?R}SQbP~D~etDA!f7?89(I5&>0gqCQ&B6cL-ufX|V(PUYx29bzZ!~ zt;0~ut6;`=MYUOtlI|wU|1T!;Ky$2f^+s1rLe+Kdek}s8Jp3_irKe_avH}p#|EGD0l>v5D^z{e0sqQDJx`jD5xNg$Sg<58?VZI;S}NCJ9_T+$ig>oKh$w|AFFx9)AekkG zU*GaA@he=*>(8>lx<3s~XvYRsAoG%il@l8Rk~O+pdDQ1pjxkA0Tzxp(dlGsEx*H@>cw=`R~aWvjU$V zNq&RNAGh|?POH9HIUZW6SvnNv>dWP1_#>1g&`5Lf*A9v`phrA?hy*e*g6NKG8()sw zA6C1qT;jGsz03lmSR~V6hko^t~ZHy%$Dc@AQFx|OGluhGbz{DP|VjO zwcjHFUg-}3KlS!lZ5 z4?eBMKUiUXrTTq7zd=widjLcu%R-;dm%tpmgjBBSL z_OD4_LV?xGe{^eGw~4~rd9$|0alN|qaB&e3t7QXWAHSAiiMxdUv!(V+!~#nv(f0T9 zZ`*T#kKVv{*!Z-t=ec|N53!fL_+p1NClP2)TMYuvxs{*^w*b#JO;NHKMPA!^VN-^jy|1mQ&gr#m)5=fEosV^0IalaPG6=vYIL{ zuR5Q40vO+Jo_BcfvSxmJEzd9AcaDMPm-H?f)r1n{FcA|24ZYX!C_<`$osp6!Ns;g% ziydPu4f%+1HNDQnkGsNzvl8$-5|7<2g6NzsK^@r-X<&$pU~c(t7(Bi2;o=s>*dUU= zX^6jJ*O2*Y9Xl)^mc@`kANdA3aSB`&!nj2&-MF=%KD;{ZewYCa@i62C^;5L?RZ{a# zBkjC65u~;t)`O88QIma~g6@|xnuLY1zi*3UDiZghBT(L|ut!8$!F!5Ef0XJ9Dp4O| zwJcl}sfFxb5Yc6q0>$Qq6-~x`LI{SWii?6lG{fLyMVv@CezX1O<=OC@GR`<2vqdIP zJB8hEH6kd%8{0_Y)w2jW4om_W!%$-av)3%a3ml5aNAFE0?&F%Fi>~-h6OOR+1-mew z%V;s7y>%FTo|Z?KJBDe~P%r|$N>vBJ;uoy=W~C3tKta<`MmNk#E(v39@Yr53KZ6Db zr^p5ceI-15^)WYK5~8x}wl!U%+3w5M*X-oJ+Y$;x9Q_)s(^NGKwo=HR2~vyD*IH2> z?Jem%7IMCJC^p}_*DkKFh+4_J8|i=0Zd};|mZ#4k@(jad^99;1i)SwS`+F^Cs%rCX zHdvM?%RCxo`PUP_Iol=3SC^hI%_!d-MbMGfRne9|^4Q(K^S)&5j(UE<@*A&|B>>iB zn><7hG3bVImOSj#AefIy1w~=c`-drEk`gsPgA`|elMfhI*qns<%wU3*(L-rzAvQGN z+uL!kzH{z=z5A~RZ>K=!)fC=o@7tQGo(KFdT%=Xo)qgY;x^4SO2Vx}ADa`YVz0_OQ z)U0l_nr3g!a$h6vye0<0^Y-JEP1n(0ztLVP#Xg`L;U&X3^B(9<^`%??Z@MfCfW=Fl zrjFI&)V4aH65XnGIJ000MA+3e@H*hR|M7`zfeEILW_Hs;WJ>3Rx^eT3 z%9MMw;S}wJmc9>Xz4Y3Jtg`jMdl`cX7uYX3G#m75OTSY^}@e7`o+Yfi2G zg5~3v31vq%#)~k@J9=LFQbLA-jt;AS(2XmD|EFaMnT+ED+LFWUMM8m{E!ujr}ocDlra2=AY=*IyT}m4 zQ4xqNtij2P8VXsU=r(p}JE?F(9NakV=?I3v%%8xaJJNmzamtq4*BP>(Z%d**Q`;O% zf2#b|;+41I#rskGEOPc4z*n$wL{;&ETI}-sQft){)4(=*ti6R1Pm&?HdUoykK~{2h z9T4*vb2jMx^>`1VRU_z12CS#?vgh{w&=ewugO{7i0f*;u4qc!gKufThw8VWg|eX1OeU z>=+Kc5y{>fduULJ4kimmh?OFi?ewfFxtKpZ(-<9V<7K~~ah5*+J57#~VNGzF6V{ts4Iz&1&*b z#+n6ejfR>9tb!)KsYme+4FgErcDum`VX6jUH_x+5NdhzHy_ajQR2cEu{T+5j3RaMuIRNHH1 zZ@EJ49n5^vf~P#Au3uaQyedO7tZ6)-^5v7~$G&*ftKFpBwjWP@V;E6Vns&aK5!YUw zmsyLru~2zi7OT`SV_j7zM8lNk~(~5mxi(jsNw%;D9CKTFILa#0S_CUa@_}z8N)fc@PuejLKc_hIrS9;!v zFv7hAD%)!hhJ=IJ6!wAWPW&{NCBWwx;OY&~W+r#HS{SQeW3ZNO&A&5rY$-b9gJ+)E z7@03$u-vWRo3(5{jYNt|Qc5f&D8};I>f7oMvKf%X_jQ@%j|!q*;_r|V=9GN2By6Wi zu|fW(KJsL6JQlvIkM04=E09cUnj13uq|=h+I>Iz&1c{YA-TZr>7{q}?+@sq>zy!kN z;uW|$l7%E-MuDZ<2`BV%+0v>jb!Erf@!TKg$??T zU>zJrHR4kA=7KNamLKdZtCmgfzg%p)^TqY&$A|L{tvy?)=drA}P`kt$DcKcCCUlh?D=>~B`Ox7w zP&!UaZ*uP|T^{e%)-Lzu-@s8=%+^|I)*DQ5*OsE($w5_tkn1hArA{x;r?!B|LC1hd zYp>(7>h_gQ{Gmgnun&F8yIXCV8`3K^Tf>xU=pprv#3-`~j3Z0`XU#5`lqVNX$#n%d z;`%F#Lut>`9??kCD%vAELt-qMwZFHYwWGuLcgC?L@eB;j*xDm8i#DKvelu(2`w1g_ z?}oF+?}DRZ4YEtjY=D^PS_85lw;@vGaeR8R-ZTwCxOxGh7_|aqczcJC@%=IMwom;< z1;QiIn&d;QW(a<>YlNK|*#s!vvP3Wrd`_yZyZMddjk7BVY5B1@{9uPVkqNgi=uJ-j ztveMc;gc{M5b(?9>E2b;M`m?T+-g%2W!F_wy15>+#&&}jOM@x8KTt)YfjKdN^%)z) z#|2a6iBCA!Amht@9Zt*{rzIBjxE{_9hkLmcjpn*?j!JLt6CJ*=Uu*EzO)5QZ7qN_# zG^sUGNH+-iAl5NWJXDTRHW6yzRynRQWJ0(Jx0dxlxS!d3elD|4#!gL^;1HvlKIsC) z)Ow)R1LM=GBTRY4RbH_jPKBH zyNUC=YVJez3^&@5W*k!;RMv8sR5kx$nhui$yAz)8P*U>UWX7Ax&3`~yVUarg#|m;! z?ngkM#5OfC*)F*v_H)Tu?$C`(P4`Q9sf7;kHmY69q-**qh6@jk;ml(#KzrC#3l1xw zh_IS+4s{FPyp4OEEV{8aKvBEc6o#y0%?&Cg!N9o}VUFfSC<;CvB00+}hQWuGzcNk% zmNO=vE6f($51Pl!*?_f{Y?LvlN7r9xW`a?vZvG(3M|_wD3m^FrGzOzP)=ZMEaVilc zEyhX?Z8w-0h$zMpv=4M6!#fHoT5$J!`|e&_mPJF>loKpS35&Rq;%Pl%G9>C5HZW_w zow0~Qvc%&I%tJgN1BrD(lq_NlT`_{zNttulPjoM4E3M}>ZdYDMv@wAesqtg0 z3`fF+7^^;iI@4<5&ma&mk+`a~1Edv{hg-bi%0URqZe!3)F<23SRr+|H;$0s(OC;jW z{bpgt>3#1%=?~-f`Pazlp zOPAcN`u=Yk5=yQ!=9tGMAtVV|RjteqkJ!x^W0J{5i=T^^smqXwWIb=ra*8l;8z7}D zUPi#g4`XryvD>VeZMZ;f!sdOrqtx?>(~}!dZ&1Q~-A)fZ&qDUANQjvI4*u$haER>Kq22^Uau~lQnSP5R|_RZ(*IuehRde?2#u_Ifk zw@r?#HZ-Go*Y$WX=&`T;FDe2>+?ME_v=2|4txuQ+e!X2T~Z#iX|Gf!!t|X7X-8}EElx>bbZJPWWVfI3IV@v zeF---w;rhHfN9*N!UNw^r}8uT9oDo{eQ)fWrSA`ISGLwJ=1B1E!}n3Pv4@e-uDJk* zbJ1=XOnw3QJyrM=H|XAc$yd!u*Oy#Zo^jtqm}E)}>tb?vBAIoJ?>oS76klvB8>?_3 zOReTnlKG`;ubM+^im@NzxGX@GilLF=>pe10APK2{8`-bmkfJvj>zVG}|K=#c5V7;V zEX6oWzn&1iV)A&&JubwJ6LqsfvRq-EBtj*|v6~YDknG@u+?TZJFF3%B?dxHJhEs!( zW2Hgtv0UCIsT-Olzv`Vt>sB(~{Ck0_aY5~B zoYawIzFXdSvy!S`P@z?nkRZ_0fDQ z5S}6RFRlf+)}}vs6L~eKpM$Y;2H3>{FGm&v&eoI}!y0Ao)sPZ5*I%Dg?B7PGpn)?6 zA!_+$T^4-LxHtOwZ<~0jTx}k^aC`aD_LrRDKH<7*Lb*QQ+&``U48pnhMZ3B*Y;qHB zo+1a=(ykS=NK`~6vMUv%jg z5>-1fltJ>em2FYDNG)l+{5c2YSN0fE3{lQpk69S42U41?;^ORl0+r0^V4f;9+$8eA z2IL)`%u%s|)~4KFmXfH_qS0#F7Ky<85TrbaWDIst z>T7itygY0#_T(3~HunKmN(AB*9C)>7?G3u?Nzc8M-5}Sh_jKRs){5)8Hk&US8#B_A zd%rB1(p=-biU`S*P^Qnt`r>Nku+lFEg*T%frr`@MBsX7ZF6){Tl@3;d%XSfFl)tPT z)ux$GdY=CoWYhrI@d8mYU2Z*~qaGGx*+X z4D9n0N};;k-QMu@E<3Ba2UeJ@r0+W3lg)tBnT6V~Mn+ab#R-+^M^%U1$J2#vCakmX z2SM+z_q;d8o<~x4NIO8|gxwmlxW_P@)l|taYU8ym1-wHzpIXxRVpWo1YS~pIkT@8G zJ-=1onjb}AjgLF5sFD~js`KhO{V!9+cyfs~fb%tF#&>UMy}<;M>UeOA=M=ium!vix z-1z?4UFwzcEw!--E&B_pq^5}a{C7;DeUEJsDs?J0L0IOHiEEA3Kb8W68`x@2hF7W@d);${myj=^UFw1N||> z^%*2DO0CZcR~`DDX@SMx`}AK5JKp}lAEU!j9e8{jt{}dU!#=fQQG(5=$+K5p&9EjIZZ2HGBYQe-fQg!gb zTjj;stfqQ~FOw3_z--s4sO?!I5aM^deBv`mAYZ zYi1~>pLnc;PT0vq6$QA~V9P_EO(MeM3oELTZA>?llUNC@Hrt7YY zk;p5}m87DrwvQkM(1ae!Uip|zP@Z7>s`o4mx)xx|JIF>ni21}_N>FwpkxwcTG&y@! zEQSJ@UUjBPYUk->wi0K?WXYgE-2J$5s8~{sshGcMGao*ndX_D|xn%MP+Z)qzz4tRn zGcdSMz76g~pUq{nOM?)x2My($)%RUXv+CN&O2uR~;Y8Zq~Vg?SxappuMnb^y~0GBvSm zQK0X7CjzaK6f1L%TqSqx#5)P)H9E+u(5C7!hwWIpqzF?h@L$`3g$ZH z=>D?ob5lXH&Fx1#H!-clddPz_94W% zx&|5+@W~qE-|tUN4D#KFH;>2bix2db0i{a|O3h#9&17tR^{fF(j93hpkrb;_!GYYo zI5P~MtQtDHUYQQ%QHZ8rLe*1)OJO&C2U&?aoq1{&3iU2m9ftL(R9HteO=G6?RMOS) zhgMo^P%%t)W#J~Wst)VPE`E8HPO0RBbtWnuJJ4uv2@ft?zb$u8Qxv7^+zf8RR%(e= z^&!&5UHMb1N;^kX&TjFI%~IMrWc1}INoxO%f>+qqZsAqUV(L01v_Dk!3fuUP0^~%s z;6a)$9lJsMB;1)wJ>ljue#d zW+2NSk7Teng*bVT52umyX?FHB$j~ZeFPA?W45KR%$PlPkV;lp%Yu`2tNEpy^blmXyfVi1^tRwn^PjoQ+eOPUMXS?c)yD~VX zhZBIc7x1$iBki5pRlc6oTgA;!MbslgUw*gTX|m7nO_hHng+hu94FuE9@jeT`sVet7 zHa+N^yBhHDumg4|=f$?t0=6l4!4Ii#c#Fiu#<1=^b1Kh0WS3x~nn6+OvSJ~mZS6KDT!17<<_RRVQhIPeb>B}T3`0+wlf}e17RG$MT`d&h> zLK5XJ{6D<Z}g zGrtfFSmz_5;+o-j;_f4kQ>1Y5qrdi`-GnxbC7Ak~3 zq0t-5gZih4o>*sB3{PG6I)E8t0}!OiT^{WzeeC8EOUNQpFHTL!A?A6jlP|YS#^6Zo z99D4Ld=Mn+WjE~CpWV8OFJ_FEZ(-DEFi8em98h>Rq{d1}q%hf6$#BlzW5+gK# zV4Id&b(X4Q#hS~Bw7b8jolCYmFpwT^_2=(@`^L|i#PC#&x4&VIO<OIno zi?KL1Xi{3Y>xAT&P0y=|^^qTt3N%q5qiNss-<04oX^80$#@;q>PP#=-4uMfl&Li}c zeQ-JQZC<}Qqr2s?Yi;%b4&f)eQ<|=&O;c^yUhD^wOCR@~y|0mXo^UPzPJ(q+nw0LG z`eb(9v_^U@0nTF7kN1x6~by>`JV7IW&&*{k5$ z{VziM%b#9J;BfD{tL#e;v~m`UoZ3MrYB7)_Xz3E8%hY7TOv{8-lFT-4*Gg&=Ldx^& z!>`=pCxBB~s)W1JoD;RB=R6Egi9B|DlAF;Kv>s^ft&6Mt8i-p&{R3xODlWKdyj?)z>b74)rlrRK9JG z&b(Aje8$_&fw7q~RfICt$iJA>5X-(#-!Q)i0{OoL{}b9b%-`Xq7`hCgIS^<|m>8*u zYQUiH;~_*ibn$yk7?^?)QNdcKaO8@cjwR)H6c$;LLELEZ7Rl{3ZBab0V$iGmHo3@S z_lVPDLt*JRDD_Nspu;1;vcUbTb6zLowX+B3Og^UdtA$5LDm`_cKE88q4!D4;kJ+j- zvic~-^f74w(k4G7FO+sn!zp&~NU(9$&$%0NHxywSB{!j4+%hzB1|7eQnnx`j zcjk9^YD3Y!z2f|8ZtZsnT8~_Q1W&yuQ+V#c&Wz@5-T6p|biekUk-+xcx5L>b1l@Wk zoy)OlV7WM6?E-IGw|?UAsn0Q5PIA7V0fNr*Fc6)Xf*kQ>Mp#*z(Ory^y`zdfY*`rw#~}o_;kZ{^#R$^tzoE7 z{AP4v{@yl5i8P{XJKA~zr*FS4^fwUuWy_IoZC?V53B8-e8HG3L!>hD-w3RWj+!PM| zJTP16+6qaOQCMSQ(uc7qY!hKx7guDusUpvjBg9*kXrPsZ^kk)uy`LS)p*iE4Qbo4` zUR%i@h+7fR(}pDNCZU<{&g+vz$B-WGj)A0_lFDAdkkmu6B()vh$BR1{HNjq`G>Vv~ zz)6g4WhZ=rI=00Pd2C2~VKJb(UR~(`B~F5{v0FjMOsfScv(R_7vYwZh){<-!18-~@ z?n*;$Ms!bBmhkvq*i2_Ci>RMNR0*S<2?q{+ie{&*as&>v-7Y>R)Xc#1slS*iX3A6` zzoz1?2GO#DaFI3NCtR3`Ze}(7D)`~m;T*?QiN!AYfEnGvY(`tJRnsn0dD)9wr*s8h z;^eK9_N8!hdU(;bPAjOQb1_y&ldgGQK;<4QE~y%1Ed!akPg;isH4dw*rNk#pJvgWg zHu|-~8w@!%bU8D0wb2dG?w3LeoF>p2dJV5wpPNrsOE_=X7LZ`^@Nir4361zAW9W$V z(F;XZGY6ddW4clinUM!ziZm+gz7!W2@;KjUQxCs!yiNOKW09@=fE?T6#3-wf=K=^f zi5DG3dEpX*;p+cVW4Ipzy!Me({yZEP@z{$*q)ku^mkiUV=_DmOWqcq3q)gz3Hy@OO z7+fhQVi5$xBsvxDX~kn=`xeOP)ra>5MF$%QViNM2$AoYOgw|M`UTp>t&((3i!^X>% zA~j@j7g^#3#gop$Kuf7BB$DZiFbPfpIj>3#&_->S1Uo(t=Y~C+G(OpiI}(>5BermA zVS5~C*p0y3=iye}I|dGkqwS7$;t$K%KW%*{Z8p7GyX_}d*PG%!QiE%@jrjQfEQI{L z^rm*#%UbiqVmVFqdcOXA%0c$!xN0~(JwomBp!BXJxQNfxPlF9QL08hWP~EKCeGIxO zbx0WpKf-n&S?z%O`Hu0fe+IdNY81ZYJyW%4pJkGLO7ao0Tp;FtcW}r=PqC_QepA_ha%-LNgY~xJG1tyF~(NFxh5JPTV}RW|scCN~8@CoX>K2~e`=Z^HOgVeKa_gw-mV%Zf$9 zt%hdsP`Qz?%|`Pqd2~i#`wx-VC2eV~ZEqa>*~NzpotgSf2imZ!B|*lX>6pQ)wkR2k zUK5UToD{VPnc|)Tl{IDRqZbKolLO5b#%erFwQi?9%})$<_?XHGl4pKK(P3l0`)-ts zWp4r7+5}GJtAyP@)$$w4*Y7$y()@;OSdIHzeYs=z{h(>DIw6A$S8{n|yeqtOqg)LJ z3zl(p=F~AI8V37vl8H0QtxAFD_AnGAf_8hE&qYhv=Z7dF^F3!*YNfoxT*&4=9a8{G zBgS%Q<~YO7_+}$)XLELB{W;Wbd{Yh4qV|NTtAglVLdC%WYRIr9q`uByAZ+ePrJf?x z4Z$kTEU2HOe(P@+(?mZLKXe;WXd$%hLn#dp%wPkyc8qUoJ)Lmn{MTks@(9_8D#u zlVa8$Ewhf5x$5!2_|V1>5!)NsX4S`u4mG)qijFb9U>27Ufv!MG>$pISk_qms^fa26 zlg939rmYEh{kYX3*4jef`ew1z%KAZXr!mrKig>Zp{*w;}njEERK6GsyOp9_^DiWz}H}4kBA-7=i5ho+4-r0j;t(q zYmD!~t~GSh*axbv!53EqE;aMx>O4RwW(=IK$1=8I960=B?dHi2KvfOlOM9?2!UWv4 z%DgAOiNQ@zQ&!3cBcHqk?0FOyd&BXm6B3NSXY(Hz6ev(=?w3I&Vv-m5^`Ky3X>ltV zXjV1WpD}Z zD0;1}eHLjqwEnbM`vS4$^~lI$?IH-VrEV*fpJZx{(#`9>= z*+o$KokV-Seoi2=58E4~Ku@Gwt8p&3@1K!RF)=nk?}az+(KAs5*B#5`7mID9SJ*{a z47^(;kBs%48Kz>0#<~=iE|{n@k?qUQ@a?Z>?8*0ki)cjoBscbyY(0s9SD`6P{*AZq zPPKXH3l>=|^m>mVcg4>jr|U{TH+aH0a7t&sxr$~d%Iq`9cFp-i8GQyW$1)URW+8bh z_9gZ~ZA}`TE4SUMm1fFwjs4AVj>CZ&80ad zTQQz2lP&Vu^n&A&ZgYr4z;Q4To+@y}qF4rL&&z1$3BKI7QIAM1Q;d7$$jV2%oJznz ztJ=L_A`izLFB^=8F*jyfE{V+8CB{KJhylj>jO=Wg0EfvHs;Aad?jb9?92|>Tx4PXU z&FClFB*SFHy3H1W@~p-0@jBE9Tvr{R9C5X{r7tzh+3HYml%uB2gbeR?y+8Y9!43{# z@^$pC`_#VlMAB0ewzHWm_e7ilfd?tUIFp>Q-}KNg6sn zewbH+^3RJf2KDlC9+9f}lW>r~drW2Ih?J9ej_QhZeQh2Y?w^TLitE?-b|_;tn1UY_t|*rp1@pMKKjj36n!Y`#*$rRzmJZCP;bNIG+h;T}~Xsal7(HMW{E& zb0h4-TTEf%a0hma*}Z^2A)y*Ohh;DH?qeRXyzSA;B*jb@PF`Q^ayB8J8`S17P6i#2 zTt|*vV`0=eU{B>o>{;8GE%9rxjx$b%^S+e-@*00rt(HMf`z(28M(64KhuB;o2CqpA zJFwwjpiHWA^3JD&!v=sFprV*_YPkuzfg^X8;r3_vuv;?c@ieu|61MF_Qi8bz_z>?I&b^}xWp zA&;(v*DeP`+&0cJ)CyDOZTG-{>{=CiSP5gw_`Do%34>ryL`)oXC{Is6VO~VILgP~4 zo@Yikgn%%SEQ_ZqKj$JM)}*v4lm7h&NH~+#Mq;&sE1d5-rMQs#7a*A!*YE|xL#fw- z#N7P7FGzqld_>bH#q&vp)Ezrws{`s1fXRSHN&gP1rk5}68=Lxros+x3qcQB3dP@fR zv0=S-jSqj=^)?kM$=I5m(i@@sYepGP^;Yhz&7RrsJ!v@S)AArF)YDac`1 zh|9dK0$yV+c`RsKY1GjHv0LN;7Yp~Y4h;u&(HV6J6KM{Qo~*2%b$0Ith0WZ@EF)mh z;?-$zYq9$`0L!dOkZ~d_W8g6Nb^Dob&Y2Jp#x@~H{RB6or&*k5Q z29fNYpFut)eC`5yhztS(eh&f;1`6Wdy?Y>cKlebvn81-yP!%DN(3x2tDk$mO$Dk20 zvI;8uWOopfvKdsikC6!LIjj)d9e$n%fxZWJ@16kQn7ON;RFo^!Sg$}8_P=+@pm`6e zj5pVHbDyVx+9fn2r_k)h{Qor>_F>+2q(;w~w?3vJ!46!nPEIL6h6IB9k9Xf@seB|< zUp2ssH0G!<^=zqrX7ZrR=)(Df{L8Co+9M`JamKVcmaOL)az*_VOh`qtgO{^_l||z*Rb+QI_sC^uRuppk=7lB`2*$SXor_v zCXq*CZ^0-A%N?Go{nql7EK!R7M3eVK{y1CF8SOm2uHYYF-l8^8*vY;7$3H@_VK4@> zoeahNL-ojtB8hX!k$z7j=B(^wWJsAwuE>X3NOu2whLNxKpBU%DuEYnH%Pi4;XJ8w|= z%mKK_sK!4Eo$v0vfw9btjdXnT@|hS>V0H%E{0~iHVmT*g;t;We?qVFFk9!2l1kVP) zT{iXu-F?z$wn7?VTKw*60Mv$_BM0jjnaa9^@N0X-?X<4L@yZCt3=vbFZi0jhBN&`2 z59g@d<%GE{Yb3wtvS*y&23VGuxg=w-ziNyKow+r6?L!}Fve+_rYJp9&%$?a7slW2a zH&sdOQEVXcQMD^n?Qj2ud?*b;EBo~yg(h9KfzZ3 z$TP`O&?J&Y-_n1k>j05SPSh7xVWOmb2u~V4uB2*OngI8*ihleW>o zpeD~H;l);4q7Ta)(f?(HhiS$mCR0rN*8!yCjt`^&9>uc2p3u(ue*B5f?X(g#3ymrt zDc@$_J@qu*$yTx;=Z?bD0zA7#^`es?Kn_G{w`zR4{mVWMZsJqx;s{pX@uK#g#Jgjs zu<>lM0FqnnKioP%}#zj_* z*PI#KL!eO#BdPl+o{b+Vrk?g;xJ5UF)lPXzeD~}!Y+-Jr3j)^M2tU=HF1trcPrUD} zi-NF<@gyrW>kMEvWrVEjkV}V1zh4YPxZoyjwJyJJG@(H2Aj+G%W6=2SKaFV-?abm; zT!BwS{BB<>PFIEwdiaD+0bw=S<}Y;kImO6bmp^}p8mf=9bw8WZjN}=u@=qi9x8Bv6 zT=kq2Ko{6SyvJF94ByT8M*LUf{WolF@Kw*T@36UHej>-0;q*hF-#y=cmPiKp-(=IT zZ5~p7p72-N@H;=+!xIyXotnRq`sMudg>YT;sa_dhn~N0)Jlrpu>k-#dtCAW2m9xvfkx*UqO~(2e(mEf+Ow|>Z5D#jmaY^g0VmYhoIx-iT@ZF8_bmuh_sJO%aiF_Hg<#*zqQ zda-6-AZmi8J5qQiJYFSHuL-Rg?;6|j8XSHilX1KQYRDL&)kP8OhmDGPP-$YD0Yk~ zUPjZo_3aseDbK^_HAk?1158cqPaUBR@1Ny5Tm6Dx9k)-LprQQ!a|dFjb}0?^s2Vn! zc!@ZrD-%bT=#GuWB7KJslZ>xiEwHqjYT@a%=L0e6_l4n-T(ScOn-u;6!uH=n_;n3X z&JkL}HI5a8$vuV^k-CR>cFDXUV%p|I}C(Xa}$Q3X(sYbSVC%- zI)K*8cl{?I;Ko%FanB@i#Y6i=rj4Y9?thu*ziZN+@PZpWRx;=a`}=d}Sgmx03Vr*d zCJ3cH^nXIOTI6@Zh;E1SXAr79;q)OHmypOYrg?H6kzTXpeN_A$!0n|eS~m~&KlyG)f@Dv4Ek_2&ZOpvlFaGebdamG(DYyj|j*$56In(7RP(!*_ zzh?aSN3zbZo&Gpwq6v*SzW*H^RwLpOl2F(J1ZEv9gB6<2EM-fDqK972dAsPo?~h=a z^Y+UtsNKB>*P5k>AaH8;Q)8M;3Zo4yiB9}gcR7Ncvjr~3{ zF)}zI^~Y|PgUz-|&G;@>N8*YkT^uj&|AFGUoyAK!4v5R`+tZIRJTRzNDyEW7ET!wl8d8wR2uNHH6c}0|1k}0`(stujtW(D z--;6S0KF0#h<_Le<}eIiu(?SjI4BB5hlW?9;4t?8R#v6xC>&XK_%leTHz3sILk5Jn z_n@Gmz6f={NOb|BE+EonVt%NIM8qPfpk&_;4xy|clRbupPRz*4<^u?HD_01Gev{~e zewFC1ngxeY$Gyq)Kc&d#gV~Z7dV>1bKag1lNU|fkg?`z~{*PXeN>n!mqcRfCbWtkL z_8^J^)Q+&Y)ehy{bzgo;h@ojNzbO@=ZVPT7Z}W8=xKZmW3DG`R39ceHqD)&}&4=^2 zew1zN9U21BcB?|E`5h#l?%+r%pyifYp}wtK3ghFa7ec_GME}b;qKO#p#`5nJoXX#& zAGfI-(GaUW_0y-fr==XEx`sqdxJJ!X+i4{;k~*lcA{#z`WE7_`WWs_S-ebs={MR!m zZKP(db@WIz~tnZ@=ag?p!+%O`Hd9!u{$?RJAxgmxW% zl*9Qb+)W9ANBMOh-omoK zQ>+J#Xh5@moofT}_W+?MrCJ@}9`8={{XS^t|J){*i62-9_wnUO?H!?Dz7PsvrUv@X zzm5T?^Eax1t9hf*{T+(|xA?9vEQs#Pf1q)Q`gE531-3E6i;w8<0M2zu#rf;yIe!|5 zf2!YJ+y5t~`Mw2&VgC$_zws2vHZ&E{-#w=~MLmj#{eSs_jE<2_L#*mt2r!uX?+m6H zx<~rmxK*g?_FR-usK_>ffMcXo_YoR`#n4lz1`S9?qkaAN13js-P@!!yh+}BDC$>L> zkkICz@vk-ma|gq)cKG*%CDL%-K7X`zuJCgQe#eAG0p>G>bIQ4MLS&{IPl2%%|Jzs! z_+j44K`+$FGM0-nhEB*ZKm806lZbM;d^dK5_P?K-6k#XI<*N$CtNf+<4wZP>0`vIe zaC6AeOr*(c_}*uvW}e6m%exnv6gE#(D=(SPy*OaqoBWvSV&O5&!}iZcNN{Av8F1o6 z#@Q5)Fc&3)m5kNBjE|2`GW>py4OY!~@-%^)FP>Pf0wV{y?Tn#6$$;>?N%%(uq*W#D zoL^H}=WdRuOZ_!%e_jc`qv+2w|DSI9<23t6$NuM*zgS}b_X$hyU=jByHE)Xi1)Jnc zS*3r*2N86~u=KyTO3vT*4?l7dx;mMX%0Dd)Qpdjdn~7fH>q__CBy%?qRIbH-+ot`< z$|~N-zizp1m)_Uy(~o`o$Nu*pzCjR4^7cN}RjAi$Rzu;3e=rl^m~%}h7{V^{WzInV z*1E5oyBy*_p6_lCd~OY{J?=}tEr{v#JgiYcbdi$yLp-vp6m(Pmf4g9g@#}6_?J$wH zhi1G=tg>ixQ-w->t}yF`)W|UJjg+d3`R0yNb+x>JQbI&e+)LypJ|uO^N0@|gBO-qq>*t#9gfqNg8ScvoML06qzVM*Q+@QOIg9ma8x3kiIrX zbx|W&#j|bhI?|%sP-|E{KGzEZ_mFbgN9-Dws4-l+0y%$vItKG+0L?l@mSHtuDI*cGHdh9f2xM#j!evnC`c>i0Wb_Q81u@tprh5pL%?5U*2PP`KqL+O%u z^}{7y0p+>pwhUP>_TF$LjZU7M;SO%@c(5XDoSgBcLLOgSW)6e7d)~;K_%yeA75;C0 z!HSah#ggz+7%_X^>~0D)orXA15O2{o%=+JZJf_GNK=6MrodHwcm3l4tfBPl5SR**r z`$!%TI2a;<+yesz1%rY|gn|Nv1i6O*$Q!^xkReb2%PKTtW;Q`Zdq^ZgeV>@P`0UE6 zF?13}7FKp4B?m{R*!Go&3I@);$?cv+C*;+9 zQ8c>;0x1Cc8RXKO>QP{9cAqNEip(cLU0Bl&*iXv*NNOD^hrX?7qyFfJ1)hZE0W2#J zzy=Ze-ycZPZ7wi`KoRGLQo`5ZB^=s{&;0my~>?4m)kT^CvWC5>}-|oh9U#!;sa5QlcFUWs5ci z-cM>j_#dZkstJ4s8GzrO8r3j`6;_W;cwsSs0#kc1zo#-Gxpi}&UOz2u5d+7)UW$#9 zNUeUGJnnfu}A!??sUr6<{nGfI!eWe8ph+as?UmDY1F|AO9NK&)32HY>Ne}7 zRdeo(nrQeck665WOQ1O8q(9Ovvajx{RFbOKKbC?G87KOBc+h5whNehWyH0grXd=X( zcZ;KBq%kmvd-$asa!6kU8FW}9O2bw?we@CUS^zz%l~AX6V}j9s(R-zN7cs#MD?6Bd zMa^PqGu%hj`r?jKCZcs)>9E<2#w-L^tYzRATbvUi&QJ!?HZ%zOwY_eAvDi;OT#dU+ z<^+!?KYgeb_!gEeFZRlc<781mqLkEhA$oblTwl0fGt0GS5xdH)rhnMWQgvf7=A+st zZpEw?L5KBfY$$`w$MTte_NG12RrDwm*<$&lb|&ySE4-W0u3Lu5DT`QLU7Iv&A0>ED zGOZtxijnivVVOh_L3TYz8ZrA2(E4KeJf&r6XjsZHV9;l~D2@7w066NdOX% ziHi&kO?mE!BD&iAt_SgP_On!$3PLFKk6d8MdG=c&$LBjrFORg*Nj^O#n$1kbfACx) zia-e8i>tawa@ubp*SeMk(BnqjhUw3qj5y|4_bjDH%Uec4di8oA&UN8795?}P_&;--J?v6@%LIY8Lk1Xv4L&$A+}#~!7zTHj z;10oIu;2~}8X)jq-Xrh5=X`amPSy9Fy5EocYtQW6d-bkeyVvS|R`=5k(9PmeZa4Fr zxw&4K4@~An&$kI}$!hWN_3OYG2T!>^MyYob1yuz%yyPtl6svIfvCj@IcUmen#t)1i zVJ7C?ts!~b&|YIO#uyQ#;wYJ<5JOWMn8%!7MHv^~Ao1Q>M%O8RU!k~{pFI!Z7WdMQ z(Bueb#z1o#Hv^A|(~Yp(NLkKd78m$=*7x2{?zfAl;I6X5FGb$=8F56n)kbQO1EZSa zJA=1Jr?@?gRXBfmnXB9vqqfXB%dd973}y8O#DG5{gHp0BLCLOv>qFq zH}Nbj328xQscViNGl>S(tkg?xt0h&}-Zo#TvO%}i4}&->#H%Is40CbXWL zUl?~fs`e|>u{w^9^#V821jh{P8M`y4YWElRl~np#%a~VQxPGX_KD4{OmFKo9K0mq#O2Nn|6V+nF+v<#j_UVoz zrIkOpW&U|$uuuEp)Pfg)&z;#MUzU+ypk}|a>cH*Me3se0BQf$syud-U2H9d^kMDnl zS0Oe4OfG|8emI!;JxhK6;Wqq`Kc=c}m#XK!N!n>H82*V6{`$p~Y{9aNnlOLj#Mk`z zHN7)KP}x%9^Dmjq40`p`YAU=`=NIjXTy76dqB+ddn{(sGp0Qb0)3!v4z-rn-1tm9i zI3(}ILtdvwVIs}dRh{Me-3hGf{pL_o}-DJ`6vYw+ptH#ZK#uR@q3< z{O3IX`tl+c#c!tRXUEI4Bu5d?1nOu{F6?5$RKHD+eTdKyK-M*kcuCey9dW(ZVBz81 zd7AsfB|nU}Rj|4Ba)P~WL~_}E&A>$1BDJ`_HXF}hDCO4@PHiXKkO$MT_EhOT>2%T9 zGXt1-eBQ9B7XJ%EUy=;Q+xjV=cnq#4V3SC=jpaO(rL6O!(ol_Yiz}hL$e9YBZBG!Z zY$Avs$7OQpeN!w4=XJvneGoq`x7!=CxD)NE9U;~YN=ew-Ul=Qt;j5y5VGMll5gH^K z_JINxlIM{6)0gA}pMxMItl1PPb4p*Lz#O(*PYO-kT4%=VD`}U5>d3Y?UyuRZxm4Vk z9otN@Zm@fuQ-7{Qq~{m&9=HR25NoCyebHt8sA~ZaOzpIu-0nEyUPGsTiHSRR-(r;X z*3755x@~>qVp~dC?%{!b;Nb`T1mP14Lc9JB)&8o{7eGuO%3p?}_(){ODr)?FWK}5m z%w#h4xPmFe+^*0vb!hQYM#kM1l8vXdeTusutw0oP&XdSGf5OG9HfxQSiPe;P6Uw8-#^$21r#GD$aA}X%^AjQ1 z@*V(!-s5(nQdeGW7}s5P11zLYMPhb!OEpnD8(Yb4N+F!rDowq7Bz=%Py14?!=kl~f zM;WJ6RXIIcA?f0v%wiF#ed0U~*^@ZGG+oJcg6_;$%U#_P zLdh!U>A}TjulBlX+7qky-eBHHKDk}7!`g>sUaq*RZcudsHuMFmJH z5A2(&FS>i|T2f9=k7fpLLh#u}YhUY5We;>e?~6l5XT5BO?YF%?tSw~a&f5O+(gCpa zfuLssXPL@aMR&zv>BqF6K2;Af|L!0FTxdL$xN3_c+I&9J`ZY+u(jd^24wCp6hIRIr zBRfSVy%Ae64KlJ(A^3uuF8372*4()E5B_hqBBN(9fPZlr<}Jo9G=rr03ER>oN9p2f z>jtuCieE@+pe(qr854`8rYp((7Mz9$?Vw(KI$o{Z8sfOt3SZ~mG_=p+!?*{53j~Oz zYxQW8md&ejE)?=-pXbZi{WKU)mWoD&DIb2iZmB3?H0{!ld9j-D)pCRH0bbGQ-60jhLfzAz@ z^>Subm}i3$9HgLZ-L{@fbNvyHj;!38>ncv68^i__Zy;;&XRyePxa@#KZhHpY0-Y4C6l50(1`ibk1JZHgIL{0Y6Uj$S+1 zNKIJWb8HplIHG8}Zw6e_VN|cy!=ik7~=4zR}Fkx zN;hRQ7pK3n%}E~wjuOCl8y!%RyuNwS9DA$(XYoL-w>gU=cDnb2kHArb6Q5Ihde^T% zLTlA#BG7Nx%980ZIm>#=SG*z#g#dR&Cbt_uZT~s>I=jOvv~(9$OBUK!l^;H?y;lpOY|4I?!rrO1*E)obCmu zh8yyu4YlR_-vHhO@#V)&-CR;EC>9ofx%9XrHsHHn{aHk`DfZJl`LuY{tnF0s3v%X3 zqpS$-SjI!Y_5I4SL7v}D)7wzIAc`MnTHQBUDvQyRz!cU)mPK3bg@<6G*E-CpZ1Olk z31ML1T9TjH?aZ!6NZ=IwTij9F<+TsiOD{9A1_aVfmHf=p(P?NCsx@ zB-J=U>z|EV#y*mtlLkUW)CzdV$9 zhLPSSJ(g~d-`y~u`K1jXkAs(5X2-qAAvIB+zXr5Q6O){>p7y}o@31#R z_el?&4 z<3V>x6>|Ezwmlgzo+0Z$w4Tf&CxMK>DvHCc6+#V1?soA9wi@=HH*3HS%Y{O!S zbQKQ==KSJ)bm~1S_j#pHi$qg=OX|M0=@%34N1RvNF8EOU2gdt!Gur9PmpnjvDq+d} zxA*}K;L|DRGpzOkxVI%(yYzu|CQaZy&nlFM`D2z{6=ou|0$iykA?whrkaabc-2Ib` z98+Nr{L;-WTW9s(riDofM5n>B9bk}`16_qaA-J$v|3Mis^`?uXvO{xAB_g6WHGy84 zm1C|zydfuu9k=gQnt)c{MH+wbd-y5t2e#*%&sofs&a+ACsEu13eL7vl;%1ngTXl24 zeAJEGo#H8YoJ9WenJSw?uCQH$l#)I1zWd zdH=cm^MTNESJ@r$Nx`3uIRro{`+L~DZjJV ze=Auhj^llW{sTqrk*1^?xYW7EOy%9h`q&k{(TY*1G(YQDGq4|Di!<|vnyMPVO)$|W z{1$EyM4X!E$3ybi$>5D@$FEYU!YPQ@q?z*>N@lxCpBQDto#PdFFSJvYij%oGW-TC4 zV$NFU_EGqksksX*Ofw_+^`kT>K#nvv^ZLuh`6=6M{9jI>~n%kTt`;0U~G=1=u7732uhO1H*S z*xz>%n9ISc+$#c_O!VE^)~gf<+wyw2-D910!(}z?WvoO5vaR5BHT3qH`NiE5G=E{Z z=!Lxws$ZN!`!9VtFzWUlMpp#%o-`a3KAE4m%QKACs(mb37qRNN={6x8dgpci>3>~` zsX(z?^&iv^^bG*&2o-7^>epP0rXqi?+@CkLQJMi^^Vu5VPDIuRoj>6Vh`9F~{V7eq z30+c1j7c)QgAAqTP-F9@mK6T+UHs2#M3UT09~|~StvLmp|6ovQEb|0Ls-=x7#-QHZ zCHn6}i4Jk43GcSEWFxb{J)C@6&V2ye3}-2kWuLntX_9qlVG$7 ztbWb5&t#2NwhH)o5)@7HZdR(x0Nl9OZP8jZ3>=(9y?1pr+SIY@p|(Z&oaH6BNBOj7 z1@C)f%bb@p4J@EQ2aZrqQ3Z9j2$5Uf;l+&dS*8H8Qr2EBW>!w##IN!vzGfR4HIA~z zOVYqA1<7mFLRyErhj`}58HlD6J!$D)5r`)(d(sb zqqxr+ue0B}qjI;)d`Y9TOg}Kxuo4W}iq8g8Z_zeWCd(Kae|_69q!N_!(4^Q=O&@qh zfQPUxGH|8Ai%$_MQM!ISJGWZOez=EZv?N3kQ^V9nR_yWc!;Z{4YSw#vq~4tx{dNL~pP{NA75280 zZe1u;<>>5TY(+ixrRIv8%hsVH?53fnBUX?|Eei*C>WZP}H4F(OVuLVk@sI1xKIng|7kA#Z90#97co= zR2hV@<=II%&pTWWP%Th>qin}tucjo|tr?pVr zikyxcPU@o6a(|laHT=iPS`Top`%`u1%a)c7Gt1dolY`6_ly;B_L=qd&PuoIPLIt;+ zTc{jhhp@byF3(i*vCt}x%_Pk?9j$p{N1kM^s7;{KHq9Mbu!6NDg3*pTia6}=G{={d;w{!GSo75kYS@#(aQO4R9dfoR z1)A<(Sd~%O(A*)%(at7NyU~*Nx>64WA-?4Jif;C`RoqHJ7|9PCE(K$b%aN;kwGN4k zCaKV#CnL@3k5u>)cASgt)e(Kmp~qo8V#1d(?@+9P70iTyuXXsU%#LO=us5`8X-F6H zMT4DZ+7*Xxza-yexD7E?yeMvxtz(Tn$!9`Uu%-B!!vvaoTqSY zdM5*~qLbC=?M)+yqU!n#d#&hhkE~pUF#m|EMU$deZ8!!V;Lo=+26FnpuqQ+{{CQdP zE~^}|F*5Kj_pZmsq9RP+5obXeYvW-~rBE$uPulKZfvU~|^#z}kgbFGP`wiO5U@?!v z1yYEZZS|U@enLd*Z}pAL=(YuBA1?wFLR2O-vNsHy>!33R)X$Zo4^d9Lp#6*$F3_S` zu0juML+HC~Ra+y&2^3OF0jZ$znjX`F>Rk3U*dFGGH0zmuBKMt-G>ugov%pxE52q9T z(41<|$izVvs>Hm=gkZ(pWLmK_Xo=_%tX^SdvzwnC}u|H7mq3j8!$Kpj59X=&#U1M8F)La8WHv8j|rZr6u zgi4+T*) zRw6v5c1AIPt>bBL%wkO|NJxCh8g~N(oJ5q0#jUM5)zYi}Wt; zPX3ehicZe@#J(uHB(OfdVC7CY4Laf|Cd*h_EPt*1$2M5GJ?as~F4;3h0Dpt@M-oAE z5pzA+uZR#;{LL|}in+iS-11A;mj%Knj;nSBReor;$%wOgSObt8tg!{dX*5?V!(7&` za6KH(>izVqji(Q^ahj>2*V^v*p37~hnLG+AnN5NZ;I`Zr#6=t8KDpx4@|534% zCnwh9tD?%-{MyQS2aY2VL~rVlIoKCi%o!5a^77fH*vcoKUN+ew5F;biq$#dUB=N@NABPG=h3@^{DxVm4g=#=|FUpn`@o+Xx%Uhh%IPml_*A#GGi*B8Q zgs6L{0vMsSh6MeSWOXT<2Q&?BzJA*~a%QilYR;9064@4i$d4apAHleZ3L~-ytBr>p zpYgPqP<5RZvHi?`e80{voAhDCv%d)nn~qFgmMK6vZ8_8KUAE~R zX&`*Kxe?y#hwh22?|(0O{|15tz{du~y(Xj__U+BlO$2#?T+g*|T|g9*m@>LZt<=dM z2!0lh6M~y9hHD!sEL&LU_Xr@JlaU*lDWD^{3e%Aa@?vofh@n)!Wp(AomhljDrn~e$ zT~r&j2jz;^J>p-HMRPb}ky3wls?N#!0+Smm2uix8Jk|>$fg#>-fEYHHl>bnB)DF!9 zRUYd;c7QJ{OLU&i>8gf=R=nWl!P?OUdS1P~X24G_OouY!=P?45HpxWyDYi@5Kp!R8MCOSa1Liua4}?5u55RJcuT8*o>VLYmKR9Sqd>5S85AjB+FJG zn<}!f7P_~p$|n0GQGjqrq*P?S#;Z!r({Y1CNygWxq?lMlhM6LPvMy(9fyHj8)Odkg z14PaW;Cr=CJ5j&yik{mcZTO+)GJ6*fU0!6`$8p_ju6GK;sjf007P@zNRzoaO>%}3kF5sMFRm> zi*Y1OdRg2&G==E8OH`uNm;TU4REAIAYTaW9Umd(iA;EKF9+#D#M|KH8Frxfg+dib4 zJ$gYN?!Tw%_gL69h)@_tj}2%jtl`^w5r4_fbKS&A+NvPv<;U@)tR8qAry(X^uK5Zz zwl-nZ16Fs_Fw{Y_*6;AP*gu`YHC&pg@^Z3Gmr>U7xDLGR=V6Df{mNBPTJ`;yWG7W7 zGZSHZ`WJ?6_IV&gbftLq0n+8IA7g(s`h7T)1uKYz4?U-jqm|bko#O{I=LLL{GcazM zO(4K?VF8axAPr$VxVmFdD`Mz{?+b*@2tf+=h`VFtb#-=nCxY`SZLmUM9YC#5Z`v`# zZ$g%|TbpdN>4h7Q0J%=$cH0H66N#e`jXax}y*xnfsh&SXBFds%jelsX1MyhuG|E_b zq7zi#pdEc+_(hzzbawM-@>Q8g_}f5cTjNcTs)-JzV*pNIaN+`RVbs$#r=YlUNtyGl zwV(#0vLutDR5xrK7=zgn^ND$v*QYXb`DJv&AlLUMP+~TVhr=xnev-062Bvdu0oOjD zw0;ysKW*-Pe)BT^nD}$?(?qtj;QHW4V@QDF*<7a!;nSF%im&T*`1IrESUr+dZ?fC* zLcN4ON&Vm3{&%mWIwN)9df3Kzpuot6BqKiB1Zl8GlLh$)WGaS|q^S-6o#V}Jb)O~X9NJWyh#VaMqDa9N%Kr)`PYJf#Um+t#mx1D^$= zXB-2F1DG9C6a+mF6d99O6-QkdWooI&&*|5;UivQE(^}MmaqtfZp}M9qxUUIvk{92x zRZE~vHAf8^;gMn0UQCM|c|Ae7M=C+RER3~C-eLSi8V+S=2fe!Rp>>jV??gVW%*euk z@5_^U$SIGM;#uLBFD&1F9p*($Zo(*NMMPIl6pX-bnAl&4yAG70a;RPhWl$OD?4mjS zZg&nGc31Z)q{*Qi->z}DOoQ&o`O988wiH>})4b;2Ag}TVd#)<$$RE2w=n`9=aGZg< zjqx3XD`)?Y9EOdeZt)v-|2tCL%8jOd`u=@?KR=F~@ge<&1uy7Ra88y=7UO!4JeNs` z=&B?v1YY+KJx`o85_av0vM-7XFs| z3xmh^d=)!qtKwtE+Kc~t&;NV}(O<3VO*XjzS|HoYrg|w?$?Af5curEzoR5A1@U*n-auUiqWODb@Ju?H_RrLNiAmsKc* z$W#*tTOLa$FPdR*P_E|Cg&+cZj3TN zE}Ks$J6Za?Dz|zYy35!eyV-6axXj8xps{gO#Y&V1WzM4hZy0>o<3@y9C%nN1$TCK> z7=C8>eM-DGw5`k)Mr*dIFF?2E%TL9_U{)^cGXEo%h&VTup4Y^F1x{?YprPb{gvS#( zWVlI6n!=kU<4jorlo5Ex z3sxKQ5lW3)*ERSdgA0a6R|!#+pq;l9v4)a&grX+)T2tF0~$G>)wZe7Iaa?AvD8j%vn8Od z>zio55O0Juuq!$&5RBFHOtFs|_yQ3*GIc4bpEI10hIp~iK4Lx$s1euFErU!(954x?Sao+D? zhPp1w_w&5QB`^sJ|FByw{lhnqvt~H_4R0xky63W{HW!O3s{pxHj ztkK&g4XgHX$~R3X$iB#4DrCnNi^74OVYmi_&f(=s2jIg=%fBu9iSO-^AQ>ArMUT`N zWN;VJQgdhUb(~21@RO#i!OZ!d;ZL1BM3w3bFA?8&KuS{dXl$;3UBg#A`(5&t z=_U<})MPr6{W!8BZSqtf2u`XvV8<p0{xB61O*C7usG1on@z#|A@uE{D$14)LdUz#-+-TJf< z8@NS?YX#5ddj~^6%uKJp%$iv28EnmIQ)@Nr9=gXZlf8uQk&GCsOJ06<3b3@LrS?q< z;eaW#zwi@>gX2#r%KL3_8fSb`LM%v#N&LJCY_-;^vzi*UuICEb1klQSC{fpIW2{L* z&ZV!-nRFvn5gcM`OoHzj+jhj*$4#~Fbe;H-FC|4-CDAu-I{f_va`i@v8cxHdD?g%$ z&;OK7uBkV3BUe1)abGG74)yOJ>5AXQ0m2|Qw5 z0J^vm-idfMnIuX$%!>b|T`&4tYI>zAO|v#nlxUn08rQD=8*t1vGrhu}=koh|29RFX zwmFvkv!HREfsFI_jX1DWvw5O1KxqcWbgOPj?DC{}K80l{`qT*6{Y5sQ2ue0%i4Eza zOOWcPhOQ-M<9wMRiI@-5b0-QdkJqoN z*gxYok^H!Ct>bA0q}IG2fN3)FH5%)1zV;R_tnw=MCbaO1j_Ws>kKCtzW}^$Da5$<> zr1l6>51`4F{P~Er8{G92@h*B7az%upEm0q(SnB~ID!{|*HV9%R1T^+6gYM~B3iBm3Vt8|DUO?L+P*~j|F(+#gf zfOaacLKt6T?cn91(!M|xq18vEXoy4#+^;TAT+vw<5KAZLr$924u9Ls3#WbY6S1L-@ z!y0nN(!UO{eKNR+nP5uu$O);MZzP1GD%Vv)rFt;) z3XngnIbkA2uM-G!Gu>*-x)D9<2HYS$2HcgBp} z_Rb-V|5k9igEerb5_iK|vaYYx+jhB^zln+wwzY}E2|%=vkHXdK|H8O9=1<^nSmIkj zlLpwkXD?&nhnO4os%8$`Da@WNfAWMfxofQ(X#;BV`}Ozhv=G2g?~J247;B#o0qY-` zIP@ns&cyKWGzuEWuRpa&JgQTUdfo)LG#ra~%m+2wpmn469u;R~j3pS02OuZDxiH#x z@E?E2mb??aNM73-hNF!2;s89QjzRR62)Ee za->Z~WKH5GZ#CIOLY@~InJ|!x03?H(V{hg@Z5Y*0ZnX$Lhsf+N1xdu%i|(jf_;5#5 z*VJbDu8k_FNm}B4&DT)Er7+lQZ=SN`PE;T1u*a!tVC9rEaeCgJsO$O``0np(<5A3t zqDo~)pI1&D4mQ+_g+f-Bqnjf`FWUm_nnZwA0|2$><|L7k`oa&S6#|q!tPo|w zvfMx+ZX8f~diu6#Xq5;PaTWn%Yq)x&(PxB(Hb40*15RyXG|w~AN^w1(Ub65&HZxKL z+dF%U&s($)d0PCTRY3}_+mfc(X5d?eD!4_Ut)(Opm5KxPqRS>SJ7LC(TWCju9M5Jj zC4TeX-7KEC*$}2CX8fZl!k+fV;zP}CUJ3+}+qWg-fIqgfV@!6pd6J*L|CRamEFFXd zS8;6SPanL6uY}{b=@M(OYc30O{8ht-4$0ZB+%~B378Q%lhE9GuWTN>O29bD+D7pV) zF2(Ri{0db^N0wIgu96V)dHE1w6^{=f=$fZO*IiCkYr0^r(aY$cHg z1k~I^X(lRQWg^AVYyv`(*4~rx@7y3XFhgqBputq$JdGV_i3754t3APNluE`gG8p0$ z7I!GfF!8#7r3soO4_2j?4k|2r96j<;OR2gx2#|jw*8PX1+@(Ul(V_$5`CXXn(C&Of z<)88sy?8YbIP-S%=e(`lDIUh;;2uVi$B7%nl;GY63@0!_Whf9S-drk@KEYaxXiLMG zG2)>TG8TZ$tdy3>=w@3;4C$zeOeWRlVVDIV;1mkIi3hZ3>nCkgv+=D9F!-|Ey$~jp z@zLwYI`-Nk6tC4)bkvi+^$;W(erobT--e$}d(t^HU=pqth+d5oH{>x7AG_r<1eVL< z*>D_EkUw}$GKt;THgS~;B6rXJCAz}{pKr^wCgsBr7Fr=#eRsc@>$yJeqxDCPtuKC7 zoX_l61br&3(DUQE6ZR=tGv6zjF~@!Ah-X&(TmQ!HLcgnHO20Tg9q9ysCo+yiOcS90 zm-Fpp_vvkV|J70cTct@fW$i9NF?hkXEPr8a!|_<$fnyfuDZ3d6mh}3}#~@oVeH~RW`G~%|q&Hbo`}gBBT&{|v38dZ!{uH>`Z!VMJ^gKr9 zz<3Eu=@iMCyBhpn2DiQS{jy^=^v5br%}ni-kU{8q^CY4ua4u@Nd zPZ23C`auj7-UepKXKtm3w%C)e_Fc=KC4v}k#hg{k`dw^6N!C3qmj|SmG8_r{6PwMU zb!X-u=sr?b)IJ2u7B zZC&?}hAyZjXV6n?B?(u)e5iQRHHd1Hg0zeZM4prjM?!?p6^%)Hj04fuWj5Nr$y*lZ z&@!!cJ)$PQGhKE7$>Fv4|nH!D~>-P}dIaR?jGm;zT6W(VLZ z#hwAi^1rk#LR)K#ZeW8hl00LP-#l$K-+oegfl{tZ%#PELpOgYz4ysJh=9j{O6F3`% zZ=2Ix#P#e{*p+IADDWAq^-7*EX>gzrd=fP8lqrVoy6?wU^CjjJub})38ap!C>cf_a zp8J$tlhn@5LlMyZM{)i$*QeIZS6BfMsjH%FD1f4-Ynxd0so)EgCA=qSI3NO zObuV8UKM{NW(*;1ay9kkigLomRam(Ij1tS)>|WJt`x1OFH6}Ol4i-w*TJ*5=FJim5 z)2qotIS{w2Cq51l=B3ch(j0bGA$1*&HsNr0b7FJe(3pB7Ee53011dGQ>so6gE z&w8B0w%wiV--1gA-CYgl^df*y49DL1g_uS8o(S_qnxoR@qouM8r|aE2r~HjZRl8W% zs`HgSR92tK9x1<6D7oSkGMG1Gqp>hD)piyB^7y_v10d$XG-(l$yc`3m*o`>?&{Yv+@_lzabLiqLh;tO$sftW!!Sl z>B_T|MWNOj4*^+Zbs5`pTX5+=q702C%OQ=jOZ?{5ExS}nm%=_ah=2A%9ujW<1{c3iM2ck5f&Cg@`z2z&XRw_r@>rxO`aYeX*pX{f z zg?ND|p{2ED{%EJ&=-LzbY9Sz|%U2!bte4INOGr)rys4_{51SKDevPNd>9@B zcFV892G`0@lm8#gci z0X++DC*B+rR5m_cx`PIsKg)`thqWgzb7MiqyekaCx8B=ZOS3YxB31V;v)b-DivHuL zIApS2<)H8)UN6ly^7x%O<*$mD6_H+Zk#l{L_I>snbiy0Lzapr9rT^}Y{5|St(dTE~ z&TlC#Xiiy`^B`z2{vgQx;X#mG@NdDF|L+C=R?YplYV!Z7I?C&B)vbT4Hv3!k-_Q83 zs{cjszoy#uv4&u{vWmw06thvw6-2jKVs2bnoh^YpGJB@OJI&nW`yU>2bYo!Ys6jZKz@o)t*Pj_Ns2L`>q3N&0); z$8VKr0xGk}sxf0K>nHW9w?t64wRLHyk$DJn$esJTh_y1AlpAQMwv$nU^+okvG39n# zRBtBlLbSh9{je@Kq``ogN=H>oTf70850{DM9Ud4c(W$CL=4u4iSt}(t>)g4|GPwMC zN62Q36*Wp}h#xRK1+)a8zNd`gywNX*E@Ecq^lYhFtJ>kMvRzN>4awJe?ip-9Vhvt4 z3Qr!s<@GXUnD!P#b`roL6AT=TRp^=Wuiin;lNO>PxUU*;Rg8fe{zlhLd!glH!L9sQ z^6Gi);vJ4le_<5Q$RsJ>;ampi*OEq@>t3+~avX6~WL4USFObT~8@SLOD}^w*tcLBP z#wo{FOd|Ekxd@uYl^wgvKguBnY96v+u+{cTX}06C)XVd2$R;U=N0ju+^FQL1rXVOy z$ydTR=XP?D9&W0(v#Ly_?_Jq0=cW!&HeK*{t`BTYyz18?)C8um!=9!Q75P~NhY61g zP?NP(*IuIdHN|K?l%wPAfpYK5 z-g)Ru(eBbnE}S{mVWzmC`7aDz!v)zdsv1ew>o>h+|JkMpXB#UPOwaLEZmHUYfT%ZR zGql5un^XO42=WT`lUN^p(hON*z4uR-1Jnk50_M{5|DmEm6aaZ#Ya% zynV#^HC`qj_Er4fzdyq=*$xw-2Tqm;HkPT?a;gVv7R`^1wkLD=4-75H|Lvt|r+@V% zXAA$?#Edz4*nVu3SYo*Svb|ueUS$2w5W~dN7LpSCQ>L1WsRH5#T!w+`zY79XBtDn% zTRT4_LhlqZolA)pl}aBgLO2kK^CvM@J(onaLKlZABKsnLr22%gwwFmr5KMzJC5|1g znX>X_!+?S2(KE;v64JO>!PsK@oLyk7#f~(|p|QQa-8kY5_REHfCL^#`D6Jw#!p&jx z;FdUlrtu1tB6_q{hq7Nxe4)*re9n5lJ?OhwdC4v(a4a9B%=XCyN$q%~G@!MvQ{~_6%3U8YBE}80C1P(xiD3@adMnyL+fK(-##Nrh-a6EO zC@L?Iz4)NJL-fAYFg7#UP`%=#WB6~9KN5GapRB)OHwQmde;!=SeM?!JJE-`=*IhAZ z{@6UkyuAHM2SJCaAI?YGkInyOadjwFeApQL)2};F@xyn}cp;3d#R;D7NM%{vTxnZ7 zd2YW#_p|T+CqEM13GL`Ns6x&&&&QkoDZ&hY)Ar*2#{rXihHaZ552H3U>&4?d*P0P= z0Kb`>?(VdVnjAzWVN6j<NL4YH@Lp(>N}Z=UR?2| z*%`E{SeoR3W315^;mBz8wZ^Es&Y+#f2#3gB#Pqk}luvgOexOnu1R@(TkmDyYSopJ5 zFLcjUR(_T?JRb&dhUCEC?u=GF^g|8 zP>qrE+pm9u3au2M-p`9aQ}2}pF@;iR>x9{}=!dx^aRp@Ln3d^;fyMmTXiR`HPTuky zGAi#@-Lk`bn8EwqX-m<&?!ckias%f&2ilvi%B|Yf%R&>EF2W~p{H&YABD`*DRX(0P z5(R&ebM$`I%u11MG}zot!f5hCy>a1n3ycB}tcA|5cOREx1wpE6^Z`w!4oyv_S;2Mi zaazMPXVUnD@c|f!HsTE(WjCC>Ye56bDRZxP>B;q&c;-1Ka_L5{L#E|{lMYDlRDu_ zb1VRPyY^AzXCDz3yI}mWD0i`;3%D4WrB>x7i(CkktPx}|wM1l6f|pmh)LM^n*U6qc0l z7ZnaTRG?${QB9~(93@}CU=HM(&)|JRcug3SZepK@m@N2buU~j5~&a(Petm5wbstD z3!F9*Vqg$XgC?tdT6k4VvIRRAuX=bEHN}jpPHDDkdPq1MOPKQvj+%9N{PFDxs zf1p}DlD+?*DAJU3XE(=qq?;|8CQW~l&o8o?s$Oy>^?YiuZk#Q77qoYbF zr)rJ(EHqmq{=gH8E}D_RR||r%5PX+XC6H<0i>*>HNf+}vWx{QV91Ky74NyL@-z$2c zzW>cn3}LmMe4P*B5v?$aFE`o`RxKkHWOPVQIV)X`Qju*_Z9hM!d~R z!*-7Ew}b09-+W4zvHc!hNE;JL=Bs2q#jnI}oOz zx@4sc_Ua6Kr3*wbn;yr*sO`z3gy}IA$dMb#3-aH|_o^15vK?9s)nnc~$9@AN!jQUy zCp{Ljx|=;ZC>J;fE651y?~Y#m;MS~gUAj z1KMd^q|Vwcj}PRMZKhf zKWabLP5Yk9=0;-4*pkxiO&afUbq$D=e1*rWs59g#QXBiWuBFZ4qb2H57$l3heW|Fb zRO7{vnm56?N#ilT$EqpIr1RkS)b4eZ^0ENYy+6a#%eR|tO}09xNS918g~bYy_*)6U zAz5hEro4uGe|kfiRmRT*=m%49k!i+TlkoLa+Q(=9Kbt5)XZOo!7yUwqKMAP@q6@r#K`?i(8?^L+}8>p=b$Oyg-rSULU4 zLyHvm;8xtBMT_;>dEWOs-+sS$_L(#1kA2SGv;SE$Yh`6-MdrS*>t|O1r$`p+c)>@* zt@bk`t5F$wS_Qq0WPsRZs7Q<=#Cmw4@mXNPI6zE9WZjBGZCHFjUG*=l>fEt=*U1~; z^GUZU>CJyD50|JY72k$FlSPO?Kr>JBX+1?*RM|12XBc5}oJ9_R3!}S_elTxAN@s{J ziQ9ux^gBA5<-piFd1^u2QH&B4i&;`}AM>RuG8T?*aQaZEtK^5lvRJw51U4<#7>#{i5bbn?GiCH9btDguU*w3cqW%>(iOCMIP zr0^FO?YsBuT!H(VD%a@*V`T$Fs^Z}rj^Co7WkSUR2DBFkLi;aw3>BH-kHvOZR#Sgf z%s0G;z%93`vKdF+}M4tPe9<=0)0yEhDz^)QtxAXKQ|SLYtpRX zo9zP+UO>Bgo3XiH{r(@znV#1lq<;IC>EdHtdp_{pgk7D1wRV1UnjrCmd1|-!CB7pZ z8*ze|j32v)`FyX8wLr38mbpZJ98m7GTPP8xW5b=M!-e|UlIM(LjkvbP+atq#=(JPL zOyih2oEVryGLPkkES-#K!(ZTd7duOn9&K{I#ADSj`QYF;)WZ5ot}avY7v3Be{Y@-y z5l6Q(De;LJ;WE%ZKf5ODTcqM=SQ#;lb|n2{fi?iXZkk9c6vw2`b?#B9BIjfMax(MD zPIbjRS*h`BclZvqosq(yxOVr*rsD9#YUye^g zwU6+C0lb%|L{3`@Jr-)z?!%=<(S+Ec*27O~=6$t#6froPQ1WIS9`IxI(f0PSjzP4h zopvz<+~vtdidby+`DzL7lZv-+D+MPsPsYlww3zSKSbt+}U!Rd2n0CH(?qt%*k{~vFj-_@z|9`N+ZEvOtd&#SjQD*}`1 zCeX8ch6*yPic>cscC#rBeQZnQ*;x%Z6BFyX0Qgk+bi^hRL9AttJrp#ofpHUutmfu; z6OuF<$)`(nm@uY&8ig)5u`+_Jtl%^os^)kUtxA;?7eLOmCy;kUdD2MP@zNyq&?k2| z0KYIEpr)hV}Ff4*m<4zU%0J59V#2%LF9^TMqtiAMcI+Q8(yPF)OJSrK}rH@V2eJxhzrs6sQp%$N0!Ws9@79V|g%U zxYs7Nc`{Y^*1%)(C-ralZ*j<8_D)!fa5lxh0NogiEP;)krJZ&O>d(y4)WmxtNcEkH<~Vu8!ZE3&FHnU1&+}Ixm$RcF>7T}@+4bcQGJ&w(^U|jh zVU;VL$v#o(ak$oDQ%L{uku z;C{qEX_Ve;PHx}KqRr1k_<00uBlqoAoP%2lw9&tjuDURH7i3h{Dv{d3F|Z>XsP+!8 z_mL*OVJ1Or`6}OtU#X|-`y~}1T>t5DfI~ZK;f`@xlJJh3Y)T0nIYbAk-8Vd9L%iV5iWp|G#oO zxRz3B8Y}=E%E2#&f?KN$K{Y)pMoz=s&<*oVj&Gi7-!P00^MG<-$X+wU6l4TE>hC8pw+X=7XKXV?Wy4`P-mXBHpi#I#PB`OvVZH-G^G#1_ox2KT1p? zk(@ipV(In~_uMZquc--+>GlhKAuy9%g8aNdAEF2CH153ohesdXwI{~kXah;YXAKo8Mt$&(>Be&hzU^yk)?ho81b$o>5%T=xon#Mp z2Gf9XmwzQjdg&9#Mj*<5)(e95|eB<5~A~^9mTXKNnExRC@zp5DI`^m|ViWlaHff z`9nbFMw&H7pS}m>SxxMAzRQU3@k|`zq69zV!8Lkz|Kn=ypHQC5zIL8cuHk`j==FW? zZi|;C47=*spM^%eMzLnHHQ7_%}=mjy6Eqs?ocM)e}1Z3`~Cd}Lu(_&MY z%O%xH+w+S@Nj|bEC{bsK9;Zto+!G#7N*PK(UTjQDk|&TtO~s5iUs?Y$zmcYFmD^%g zs3Yos2nHS6taI{hdkWC1w|n8VAn^z8+j~p0K4Qwgu62G&wq=3`%}-m_)GmndYUpq6 z8MVv%x<@%;BBB`#;%CmI4;P#VlyG{}J~N6%v4+;}84l0}q3q^OyQP{C=AiNB2zILy zwL^6D`-;Wl(}h8%ny?6EwX0cz$KKeoJ)aamr5y|9`dtiAuxIi%CH{D|pXDe(&!x?F zMEj|!@X8QjjBp;39q6>nSx>XbAj{iUjJ@6Tz+ZTwW^tBG@kAB!{Zl9x{G_Tr)+0(@ zPTvwyqD~SamDVQPeBpSE=D|o9$?}ki_*MBpg?1gyJG*^rb&79oBlqO^>;yXW$tY88k6H+&Z`5=sU8{e_h@ZT2K!VE_N-EnYfOoFDlw9f zeG@73mdinNE*U5@t#cS9k*ICld{;T?Efay%- zs<1eo&oUOW)|)&FR!Go^_xXs}%&~9F&P;)|@blj1QFwWIJ)dL!Wr@Jz*9|A@rvq`# zlpj6|n=9=0xE9vV6;XZ5;YUTXGiuRUdby2guIX1`1r z^_EP2J{J+Is=aVtC7%3*{%$)V*itLaPGuyO+6Nfv=(!+#&~8q{tjNdw;`ptph>OTV zg<*qcdSo9cMK@p7c7G(x>I6b4Do7nnC}szLgh=kCVqHzvL;Gg0P17WVkNF8$vE^!| z4GWvi8f(RTsJfr-rS6OR3v2m*;u*iX(1wI^B@a7#OGM=Vg89WGD#E{5X8}RhqHM+b z`0G#K^>#}9?|am|#HE>eeYbd49NRHD{~j=u+cWp`^S^Th_xArq*FLL+{w!YnMdypC z32%%tJ6MDf-Kp0sMYG3EEKnO7ONxs6DC@0mOP1y3DBIORdh$tEl!$vs``~c_p!S7% zfQ!z6_c>F@37K0Z!jEdA;(;Z!)m^)W6&8OB8-|<;?B@mt-qxOlkS=^jr{C68D^1q04}m49J5S1}h=533@BtzU)Vd@j5yKF#Qey_{TkbF87dr+q#12d8sH zZw<2k-oj#JZ*hB)u=DgJ>YxAlXl)Q*7=zp>lWUG+!3bFz2|RNi{e!YH-&idZ=3LT~ z?WfE%IRt}r6XkB@(t&YSO`MV~&nWPE>?bWaX0};>meXN(Th-x(Z(wt{Cm?)qa&RjA zgabv|Ea$x-;dOLOwpUTp?Uh9gelS=&+1_ZK`NswvUm~Ys&24tqr@9AE=JK9wk((vN zv~3DU&+7*zBbo!)u|$aPAOC|#f?1VG4qg@St-UxsnxWk6HD*8D#h6(ipEP)NLjR4r z`M>*Fgu#*MJId!yz*TIa(q~L*c5~KEY zF*H!|MZOcv|DshwYkQkwrKj?u&Uao{S{i?paZ!tbmyBOYqW=iLha|dNUZL+jGp$g3 zvRi^*gEZgM7^#%C7DLIJzi%J{_t1T3VFBisGh^33OFm2RYxk0FCeM!AkNptb8-E?Q-Q7EfQ%o%W9P)d`FY7|%7w-N9V`OedH@_f$F^Xf^tgi+VnSj2fRrcJo;=GA_G zNAlfVDacg{fZk^R_!aV#$6kIjXeB+8CmNS*;@+WTmr^l4&RpzZB!4 z>YHRhHhvSO)P8#S+t0!{LS)I;POReG+mCMM)!*t5#VWhk6#0Lgjkro-e$dojm!uYL34!vD<#wHxovKQ znb^oz9hn6SzQN&iMIf059O`tu6s0T2c(Rp>@!1+%pDTPk_0kQ{s(;g@=xU`uQW56e@pu2vP7rIn>r5Zr53*K zi%a867PDB9St^@aPP_4;gS2VoMM$M+^cr{U`fu?6v-HHf4RiF_Bd zw$&hK$F{xdLiViT@hKa&J+kqkrK_>58+lO%vUwW%f{#y6`!jVs{KU-`G zoyJE#rSN@=Gy5vFTsLBJQ{x-$Q1fb1qbL`HH?=)!aFWqSY|#Y6a*u1qc$r<8K%Fw(DJkCIWm`v5&=-fO;yRrni715Hy389ww<8Uct&MbBEt2@i zT5Rx`xB-klU!F4xZuEIfZhC9YRM39fY2;N&4*d?_lo!mulvRmZMa1sn6#n=%ptlq7Sa$_~~tP#2}YgzjE%1iK_bhOd{oTDhGCxFnxGnRzL zC4cE;1`vWvzl<74Rdn=?GYk)}ZG4*ci!76Jq%r4_iI$v4-Yujc{Q(ivT)y6hW*Qg@ zZ=iEE;6-4Q6}va{T{uTO}3d z*^FOcWl(i9MIW~Q51+@0C1-72C2}|=8ZfNzZyyBsvw!cVep4E}Of5I6(LGf$EVbc^ zd-(jIRc$D$e55S?NO}7;vNRN!G|%3K&$t_Z^v%73Rbfd|O`P>vFq84NQJ=qLDT4!4 z>prXx{-wKmAF0jsJ*8kwQ`xVea?$%B5)-sdp;X$0ApWJ}Q})07S&k^FAit5A4;(UN zj+NPmSpIkWrGLYJZ8h@ul`_#9r5YK5Nem@e4n#TJV=}1tltntyPj7*L`g%Ljg^?D5 zs$})3K*YGWONLRmWz-ws$+)lVN?L$7+~#CP_q#Q$!ybTju#*$Q#?1N_R9#*cH#zlj zhKtU7mmLoeG>2Q$Q*1+~)r0sI zV+$3ltN{`MjV*PkzPAS+;1|V0eL38voeiABeuI0=at(7AAapZnu**zlBi!oN ztD5%s$pEu!n=U#Hyk4UqF3YC%0MwUY`D=ddyJ>=FuHY&0ltH7nVm!H3rY3C-bc9oi ztcU~BfZ%&y2$fL6YwULk0+-{9lXIxPXu;# zTc-=tnTtmphWBe{GT6d48|oy%O~|%wW*S0fYa~|-52Z*^VdD8aMEoF%x50*-*4g(F zrLv|Hi^=oZ)bqb>Z@yw#YG3s(hvje0uVY>R!ZOL1bs`N)WfeWD2#AOmGZDhX5i?v= z{dxWCP@?SqN5J#R>XYTc^W@|I9@+69YPaD&T0n%IwsS}?Bsk*Hrg)?CIiQ3HGlfXq zcsuedLxo2q$lOO1VGfqWQ^EE{(My&;&W5TFGf?g#C%d!H|Wdap;L_+mV!8IUJisqO?FOA<4GGE%2!AoHB zjwf0Tdy1GTKV-!zGZc19i!CC1?e!8?J106tPw%;{@#-as?T)VO#1X=%C8IIxsg?(N zDDLd)r_q3ssOlv65#2tx5Nrd~EvCxE>-5KPS-co7OSwr{{QOmlm2JOm@xjQ#(FB5I zlV`K;)JxGftBAj2!E_^|CohgN^lKTkIfmGA{PxGy7-DMS0EcazB+d6uxaC34#+b+; z9v;5Sn}_Oyutr8XCkclIoid7Z?Vt3i6sQc_7wRQ=DBpOkNdTWZ9IaDklK^dZLa~XQ z+na$yO@xM|2NWmU9>}@}n_OxIE&V=75w?{J7G@W;X<)FUdZM*zl#MP|a-QS&4-c6Bu_?K*_uCg-Ga;Hh}@i|>^J8?#}`Qq%||8E zFBB1vlEK^f^I!`dw8*JPi3o6L$}wvp+WW~Lw8KD0k+7NW%kZox+6NOT;z2!Oi&MIc zja|6|<5*xFjt=&qlCAdBr!p0+GXtpax7e3v4ItybZt;;s9&qwFP94aGC+B8{Pk?Q) z(`oEqwLU~G(gm2P@SB3`P7|e@)vvSHrFRm`>g(N_lWCQVlbyG^e?uZkyn|W+ZGifs zAX2Q74;}0rGI7tfwJe49WHl3p%2HYi-gX>j#$xorbfxni#~vRsfvPokm`=9Iwp|%644I(?Aer@AES9jY5w8v19qhNDZA}K zYR$q>#&R-(o*&;oq-@ondy!61qAPr#2v5JFin}6off!ncsYL|O>ETmSNvA1^rd22spfZ1H`rL+)iT`8acAC8xoP{vVrKkUUbHHE zy#;BbGWt}!kxsi@q`XDfdfIISmGjmp$~cpeUh_1!OWmT*ms6zUNwNBx7zb-;uREYQ z!aK8q#Bm>$W%we_IrPqoTCWDDCSlwsYqAoyu z?{7;~L;5jR6;6uL{EmH-rfVHCmC2W>A1^Teae_N;3hi)6Iu`?{X1Q?Y`&<&uAv=9< z10^xsX?>oy2U%mx%@Hyl5reRXoFl4g+`q6|zHI^p?fedv!;CR}UF_v$L4A)+o{|fv zN|Fz+UD|gV5^j$IlV%c@;!Vx1Nuys#>C?KFZN_AP5`0bvR$B-V)M8-lW5yY zm?~0;_$~oZV<<6u?OAPB)A}2Y6Z53E$Ugu z<)b3q+uNU&;_y(i)&6nZjayA~3Nq(Vw~mf+`kJOHUq6g?Q&Q=pm<-~4R-NWFD0*TCTtSPfQ zorx`l5&Z)+>3xc;ES@)^9W#P)Lmx%weS!%ZP`WftavF!%hgAs$Jp4_BT6>R ze89h{xjxZNoi$o``K2ie#+$$MNQJ?%OAftBevjG`XBW%}+iKcMgMqrg8<&L7!q7clR z-Q*a>aNBH}*gp4foTEV~NuD-D8+e2zSn*miJIc_Jw*WE!aN&UJR-%-AZZ3tDi8(CG zlX15eI-`{iXs9Tyo~eEt_Ve1gek8FxSg|4}Tjh`wWlx+qLld@UYqRIx^I`dW|Br-t zGn3D*wHws*#RZgT&x{^E{;dGJc8@>fO-CYt+gGn&X59{nIcD|Jr}n&+m1t7+Nubrl`7tWqf=P zzaG_IQdpzP!pIzk+QsBf8KUA%4B-XYMhUewt95;Mu<7BbuJbr4qz8-&i|h1MPYe7p z`Nh0qB7QT3;FGipR0x8`o#`@^gJq>{IuzbcEQ+>+%ujox_gzpMW~|C0E76xHtL3T{ zcv^CL%B<#SpWigX+jxavyksQ)mfyibGP_+DaPS9BG&Eb<3G zn?P;xy+|4P=WLG|;g+x3M?#~UWBtpG^8ryUWp>mLsl&*3l^#pMuN_f|a>^q)a?KT0 z{F<{SU|T~K>vFK5vL_R)+^nVuUb=xvZGG(4W z{zSq4+<3S7gj3J6rR7$q^RidAKb~h;4;Q8YYCg^AE~xK^*N}`GYL_@etqzo-Wf;Ts zRXxnr3OO(#)V9{TsC%tm`TxA-Op=Sa%!DT`N$1(ofBvPJeAWAWibI~BlWygnH-AVA z?RTG$>%nQOS@J%uYc_K2vxI5f2H#`2*(PE}G-91U@oP%TsqJi<9i3$IH0xAu5!-Be zD6mh`a~3njC*HK%P&{`KNHugu%vV{LXr@!aa=DpUA9ryOD)@c&pa5*?o~q>wW8;-; z_eWtJ+MMO_pxgF_adc*AmlV6BY}}mr_8$&B&^trFdWY9v@+;9un-9UhJvPwEF2MM0 zzMCcZsMHhW0Y2hu_x{^~DW)R*$A;sTM zMN<0Sd$iQf?gPZ}$xANK{x2^u@9>7PB)s>x(mi;)STOqp83MJ%SAPo(1W zG2n?#a;hL*d>RZFLe(zW!#LO6=wH}oc_|tgEG|_Ri@GivXAv0Gv-$`V%GTgm$c6Hz z6PIibc`udDo;9mk__nL@?g^m(*r4bpBDpzI!3`&by|GS|`Q~Gs;-(%Qm8M%SuXF@h z)5^IeKYUc#9c-RaUkEE!$mgIB=}7r<)yUg#hXsH6J60(%d`PWA9^rYkS-g&qP;Hvj zVsC%(7`{5I4}YcQ!dfXSP|vb0E}bAX2qoI7aWNX-b&mkD%PuW_nJJ~_GBPT2WsjK^ z=fhQa6SfD=G>4<|VpAWceShCaYY!4~Gk71*pdnVhxL-_NZDbyC2tYFZj1&%E|E-eX zekf9yveY&Yk|JOFs0b#DK(!0HT5^1ASk-MqQ~Nm4Ftmqh6PBoB2og<~_v?@N=d=}Y zLzFu!tBsY3n0aL;-zhJYcP$iK{1!^dSLd2&TM9{>8rW=aSc(2!-qi9cU7M1Y$ms5H za8^Ic#DmBwJvv^%^^v1)Ouq}di4!!%#<>~mYxh|3YIrTM|HGg8AD>0{{+2&qYDuKU z=W`1KnUmGQ^c-l-h6}3zLoS48BB+efgq4D@}Q;sGu2k>}1?|R-ENP9HM=Y^eL#BuUSZMviU0YqkgWm+?5>}zuHs4_@7 ze$=SkfAxiD(dhU0dM+_i=S)f-sPo)FoSM`m6SY+I*QZY5K}0e7sJgcvu5{4tSac-^ zdXwWwtv~QMDM7)Y3RMlLGtU)Enq(EaX7wwlA1GI6#A9f3WfOI_U)etLs+4RX!r?`) zb+JE$BWLsH*#it3E5iMc^$Myy|8p!4d4Lb+{kOT-gtxn<>I&8(B88TXCyq=ah_eTF zN#VB9@n(i8`UWl9{_gjGsTfaS0wi0C}j@*L+cIQz~7IJdtW_Iv9=!kBl*!F&0 z0!hwN+JUys&G7>yKLySZ)QfU$xQJL>*dDZ$UU$qO(J1vNR5r2R*Fhj}BsQ?ArpgMe zzCZ3{zY0}Z9aA$NbP0n~MpP;rrYqBl7N8^hl}Xvy0WNwAWVQ7G`H+|D?`Y*iLz>S+ zFVG@0BZ#nd6KlXq2tX3hxS|p};9#wCC}w2RT**~Of&BzHnunyI;Lc1^?@5ehC5kCi zJMwwCS)H0s;t-Y39Hvu3IfV4?70VG19W700UYs}%*SXCYXc?WF@t0jsEE;_WVbGTy z6IVM@?nG0u7se8K14bDlB_4IEc{iO-8W!9&cu0$asVS<7BqBF9( z#jCn}zSPdu@zJ!xijxmbbT7C*6&uz43&0cm-&>@TuZ_>lRKOqj)81*)PtyU%BZ-w% zW5+TMz#zts`UnYt^MJ*Id|zCVz>={8QW4hM2w>$mzeFz_jr&r(Cb9lLSv$|pX3w}8 z!zu3ugPY)K&6j5FF&Ypp?uE+)ZA#gjHXa0 zU2DXtx5`K3OQ%>~$>E}93}X{J>jN&<^O3ST!PVB3cdr$JhC!qv$+TkrSYNC-%b>~9 z=d%tO5b*iwiiFatB(&qILt zJtRcm9S7ELZ4G8M5IuHlv)+zRX2h&w-1ti*xgtL&I$wc?6SIv=7Ay1R-mT2NdDCfE z@HIf&nPZkdv#(RRI{}{_DUwC^;gkPBBQr!~rD)BexrJj;7QInAoz)9@If=9XOjN3> zOhiS88p6Y9O`{86azt`$iBdKp&12pK7KC#1T(mmOO6nw(Pv+Ja5A*~hVSL7ojVa5e z_QYBE^wcf#6|>`)9(LFHiP=jvIiq3uN9O93GjZ8+`rZ1a2-O&^WnhmJfzswe5}87{ zbis-IR#Sy#W>I!t?NCJ(Nz8`xj{abBg`^Kq%l&I4&(_i!O-e$QAEKy1xx0q4`gi=` zWKp)Y9V7yGGfuruFuy)E_4!ji%IU$(9nr*#Y}RzsM*$2Uh5o`a{iQHoFAyTk z^{B4AFgTL0zy#D$Giw^!)0|-4U@b{566Flur<>$v5iFfg9+v7)bin0hzBHAK!3037 z=g-)_+Bf2>QjF;fsdV~A$|{kXW`+%$^Ti=O%WaTur+a&cbknlsm3YIa<>!C$FN8NO zgS#K-EFeDNckEi{qo`rUEO%@%-LnpB+h0%k7wmbvCss$QGJ5a>$d*zdR@A#zFG9-a z6F`HSA{* z&{CO{u!V%v#K}4~FI&m{7j5g^Jk?p1Jcr~%zlTAQf$Kr!hkE2HRP!0PN(xF#4$N26 zVdSwar%37f`l%*$ZkpoLb03?Ou1H6;!&dqG0g!c!uqQWtKJUT27gH~_b$w`HUCW~Y zNrX+Sw!RI~)QdrFLobGpS#+Sy`TC3$#&<+h1rNiYHEe3m`WeY!$^xXSndP|>0Gy0m z94cKcX1#a*VShL6&tQEH6~s+QPY;*D%oqH;*XwG**7^r1T6zk-bYl55oc=u7C606B z|Lo(#o2}tS){`$?^;*>Op4aibHx|W^5~(PQ*gwRGa6@#goUpe z@zonDhK1$*#JTWP>5SV&=uRxtUjHy$YyknJ-*`z?n3l@PdTg(P8JAP$HqX+6j32@{ zIhBfsN$&IS-9@$GF4X63+ z8g`9Rad{ic zZJjd&z9~1&h@`!vAq8^-mrYH>QQbRU=B>SF>(nY$6ff&JZCXi%uK;1Z?6KnxLsfm+ zw(e(a2{WItgl;l4-851J(CX*po2ccMBdtv-WZP%&w?gh^ufEa2W{ikx*emQmv)3*L zy3L(>6e&4?ym^oyPdQb98Z&fGha3A+w|3LF5zcYLK^RC|H$ z-j$0*en9xnY9+`)E~C7NZA)es)Zc*>FFM|8ytXf|jOs>$O%)RpMU0a&=N=5gnyTc} z!(UysYP~;ec=({+enyiAuGLAraFeWSik-N20?TZD)i>Cb}9=K%~eC% zmng}2%TsC8ROB2IBEy~;d&TG7gePVt4%3`cTZ2`GVDWO$ML)k~T}F?YGgsHJXOm8i zU=$X&OA$eKIsh=4Yx`#6QH_#cUEZI)D8J-Y`G4kS{>yUf$)HVWvw$4f?lh67@C28!*xDiTKn2K?q#d+4o6EV)I(QUYP`<0{2@Q;h|Y z83`EAk0kctE+OEHNSY{Gy@5duUa⪙XH{R@AwAHws=u`k)&-?2-LI!HwUp)_s50O})@WqYS zto&6>#v)`T^$=`ji!)0$i0jlI-kmo%#;iQma?4&{9iO9iGE|yB%x~3BBIX(EBrsvB zcF#&7@0HfQR&C(uoX_lGbx^%sS;35LrG2uAzb!7T?XvB`cU2pqu3H-sYA`)YQHw6u zc4@nO&L=?oUP;m8Wx*j@oEIXex%F~fA@MNf+WyDQSA*kslA1nHPKbY^~Aj{Z`uL3E6cf4Oo4&UVTrnJnyra_!;xCc{P4mg zdD5sPjo4e+^I?%xo&0Y#m;}WFKz;6XgY{X#;!^apYA=p0HS~VA3TT&VQN2vIX{>tB zdYgVB^z-Ze(7y?yM>(q+iSGRIBHORp$8-hB58|sF`~t+^IhWOje~LW^U8ahr-Sv2R zUi@gH+o z4SUMgm&=e3G6Xp2jl%ajCWIk8!9Ha_Pf&Gya+tI$*P6zRTE!lzShts#cb`l~Jd)Mf zfenWm?C;FcuVT{?y1lkKw=xH2icmqLyt>+b28E+^{H8q2;u33>4R%)I4;P9WjH!H- zc80Hw;R87h6fUCl@AW`+F&3oVtWGNQ%(HjFDi>yJ1np?-CNv0JIff-y{UXMD$1 z%fE@j4t?@G7tiY$m@cCNm`!wU4ISFlbIDpi6+`T)YE}O)Lpg`d5%`|dgVY#{WgbpPyIF(9Lop=NIf3ne}I;A;?a&5 zi5g)h$vCzLo|nDwsopW6$rX$c*sOGX?%Yt@lI9r6D_ti=P-lm~vgoMZz0rZ1hv)SA|@DMr=*BmmiCS(m8=+HTxp^0icC89hRt{z6*8XAxIwWU%#7#8p8K#LvY zY1BJqU13x0&4Tenv}j|+aa4}JXVLK0pId)%7;_sK^4SkK*tqdvk#{Oi`n3KDkC@$B z9Ln3?qTA@`@$#1ozfF6@UWYTW> zH9G=hNE1*$U!FjQ5m03`Nb`$5bX<7u@EBdq4poygh

eF}RdW_p&OK(2A)`-18rF@C_F$Dj)sJT*F z^Ay&!%hSNB#JIDnmB}fZoeMOj?b=W?4K18^-_uEEVi8gzV0VMccv#x1p@^2v$<}*_#6=*MfKYCW#;<`s+NgZQ zFc7aoq0gDa#km>%!BIBUn@In~lMI#jj)o#Mjvqx#>BT*%XNarg zNO-{MHr&V!l=Kbzg!AL{HvQi+9D3)U!0`MQVf2p-D{9Y=L*c?m1DD~vdAMhiQ9{!Z z3vWg46B^zL6=NQE6_!x@cit`=uXnr#KH1j)ArJg`l|m2GZK563@khOAI=?+#bX;; z-ApTMmcgC9f=(84uN4jKK3aUu;|V!6=6bFrBwjTKJZ*wHF8OD$TF7d<>C{6(!!u_Z z(CiEzMxcYI+(eWE{9exuBe0p7C=u20sdt%o4EtjV8*>T1v)CX$P9g(uKfgfTx9+#nHMCglNvq5`9<^W!A$L&M`>CUsFaiPEy@ax zSP7ZaqsnsZaA7PsJCo~!+3p=Vms}v_$BKTnMW}gI&gK5oOcfU;c~PN~XJ-KWSH1EC zM6)dxU#z_*{+SM|7;f?$@@hNJ5AqRxJ4zR3q+JUB0447pLEs`(O2AD98V3Ef9LUDS zHCBMgDWsA7dja_oShw6uYM>bW0Z+Z4g3vWVI~PTT41>@&O?OaH7BT9cVul$yP4n}| z=S&TZ#<#b7uIEXf6^do9MGQ3KU1Kx@B6~208*6*GA8Q1=35dPB#3InbQ{+Eu0timh-kaRMgx1bqXHAP5fmx2xK5%t1B$5tx`k_AQ|C$>LGtJX+V z7HzO91@ok|rs)vv8@zt3A7|QW*&7K0s(N1#0A%R$>E7m&Bi;qyTP$)FZ+^(Zr3pMQJAyC><`8t_QsK}8**cqa#c=wPTQ(KLrF5w*uD zk27XkwwNj7U^Z`lBUOm>ss6QmmM~ENH$mJVhKiyoFy0y*(7w#mg)V4 zR{3BYc{m(5saYs62M|T)sh7ey9@ac}l9^y1PO0f|+SyOWRFpoyeu~llqF3F$_?U97 zaj{rDu78Enw!hmL^yQ)!v>EEzWN$FKV>?NK{Ue$IOI+eV7Rm=Z@EL?|GLwiOF8Ry{i_La!WZ1sB-{l^g`NHYY zX@@)gx6vNnpMj_;Y7LYV6b$xSJOZ~X_UIzqm1N^%eab5u1j%CXi<+OyKa9J=I7FGr z_*K1#dr-|ZSu8wtjV!W`_UHhsv*QZ?-#l%N7(QSjnbbnCUftZ%B59P+Lu3P-$}h7O z+R5B20m;m0;r0qgDkN&<8|J^#%93hK_)oPKKtltYmpna4T&iDoavtKlG^Kd%Ljb$nqti} zTSm6E-dtHUHC(JX&hzxAhR@3Ozaw-)2`i50rt=7>xqq{<*P=AzU3ddK0>(0XpyBRhP7Ai#1MK5a(Bzb=znJHKDL-N!D8DQ^85E?XG+NrdS zi<0c!7HrLxB5%C45ZJtCnMvD+Gi!4RKG9Jv$i8T{z?i+}=K0!tTYw9JRw|Q{x~Z(p z;>s0|_F?4>N5OrpX$IH$!G8)DQ_eq>II1cdPDiA=!V4aGex-9>oURN`WI^^yF%YP; zFbqSppfb3@PBWE_I@=C~>pSDtlXPa@`f_uks7mfohz$zCB>Fd!rxJ`$Fqc_4Nq%a; z%v|khoNYSIVqxIt#Ct!;VqKTvoCZ@J>N5M(K1^Bkte$cD+oDP`0ZtYGH9$lj5^aRQ z)Nj6m$RPqg3($->@06d;;@MAc&6ihC9zxJbDYt?J_G+?WmaXgq=7ulF0QA|00ffoe zOA_{ud%;dzTmY(uZ{*F-UMktV!H39wJ7-s^sbS{G?Kal5mgve<*(Asz0`F&$yu5P3 zg>P{5Lo^fGwzSIi7N|&d+<6-}B)en=?eYB3CB1-pktyDvnFezSs@*-+z2>7JT-l13 z9za$DweQ_d^^=(Z71J$QfxW2`n%@D*uZi)x+Y^laG@6R!)N<|?G4f_)DR3b*{Ihe| zdPBka8KV!8aJbkEqn7)i%bRz4FVz20S|V4k2Vb{z>W zVBn5;&ZV!fg?PZ~+A#@jByTHo+7hWQnJnrt&u2(nCmrhbK1}r3pVs7$9m(vUFv;<> z*)D#%{*u~`XKNQ>^nwD=;EHW`rdN|@>}e7oRzKRzZp{Va)I^I+MZ5CoPt3bGNCfV~ z*NruPE>`dvz5j6xx=zm1j^bWyzfK*b$lKt@$} zc}G;26=o?C#=SYwx_?JTJ{jM1#GA9+K))PhGni4rWmTdz$Aw2oHqi*JngRY@EoLR& z)Cx?3(fR$rce;XF)*$xc2)%yW08@YeQKMGFu5~2*-ZEq0#O`2&Xzk~kbk6|r1J%;3 zu|eo>?So35t~Ii-o{eSj` z;@>D9~+6@Y1T~}9&9yx55%)#{)(k7PO zBtcjdT)-YQg9sfULJ*oVm)3>?N(!5<-?ykfiO=fcust8dgj_Q}nRmq`VnlSWe5U#+ zRK-g*N*LuD>@uV--pB$jDK`+4Pd5Lu0vgWRZl&Xd6|*mi$mzS!K2lYMB$a0qPSrnJ ztg`$eYsxCd3aZBrVXr^6(K^#9+!VmTO1`&5@vdjBFcOe z9T1NC1h(doZRNO0>X=}Qx_C1)Ci)8&@;SxFXkp}&NnFx;hbGM2#wT{t`^j9p>GslZ z)X*N;U=P9S^DY1i65d^677ry0J93O33;69J`WIIJUs%5itB0i@^t=Pm>J};ItJv-U zD~nUkJrul@r~ybW;KFCFg5fVAPv)H7CpWjqW;&3)%|rML2Q`q2*uZ}^6l@Zo5dAmq z-a4qQe_#Klw9rD4LUAkZR@@2{CwKxu0P5;RD0hZbvr zLT|dyIeUNi-sjH!&6#s&?%e+}$*eU=CRv})dOy$edd?M4UFdGo);L^7Nx$8KcDgqVM6Jh5rZD%t5^p(x%K!VIUYN(W{>BWUQ_2fbWtBc{V z4pmjW1$sHsUiMd~Y4Fy?O+Oh1NEeBa`vasCk@|bwr|3b8OxRoOPwqs`20r7dbNvf) zFDJshTfp2^2Ob!8jN1cq6>x9djcGqRTxLcG6phv| z(d_PI`0eHjxr=D2xR@OI^s}|YP|-wm`_?>z+=pSGN->?lzps6~j7*$t#zts}Nw31fxjx^ptTq!TkF@#{Q5#96UFX=bN#N1N4R;^Y+VEnL5^~r= z;q$G4> zVev|wd50XsC|NP6HdA9|QYv*6D!?5*l=om7VC1!Ks@y`1y#mL(GTQukxZ6`=QB2to zc_GDPuVIIFelR2bY8*<)+tI3b=rxOw&oziTQrVzYGDp_K7gOjJt{o6t!91WyebWKF z8BvFtGR&A6N)<#_Rdhjjc9oIJEWE~I%N<(6ZqXju@42BGks=GPwvDvmOW3or zyMGT=2o&WQa?=42JeQxI%tWR0d1?0)xTg7ps;0*)IE#_<&T~6o7W*%;;jo6X?%#92 zt^Wh>swPI*Mzg3XPJ#0i<~D9WT4zC1a2K4I@YFW$HsnXie038JMXk7#!Y4%c%@8|9X}R(yTd z%q*p(q%QP8>>oo5kE>bjk0*;iWw+BQM6CdmbnB6{NkA|CXa^}#pGtWe^K5stYI;D| zRu$d@CjeJARYGDdm_y7`hZ~nUEb5iWypfX{#ZaX~l?RE2wo)`V!&-c)wt;~2?c%~N z2L8ZKpqBxND&LJOu9Nhzo_99cGxzxZB$a{gB`S@2gia0kxRYYGJk}!5n;=wJ0;;~G zbKm{M1^12lHXD@0Ek+D93JEB-Qr67QwMMUkrsxO9z7}$hbH4^-)=N$fu@A^7%lk$t z7m>*rxAyXx_n0N{>3vJtU(_Q^G)9N%A{shOcN1pyrxI~~UV$?g*8oD>Ot^Up=zgoU zJ-o?UV<7eO5rGcu9`AkZ9F`6y%;>X(fU#PhjKmMWYso4{g`eSMwylvM&{lP1y5|m+a-CarqsfA__!eo$*dA%%3>0E^bjN8Jeo}D=1hdU5~{Z;zX!3U zrLb`0(-NWez)mS}`Y4JrSMpWTm6R@uDV-+V(4OqMtqpX2tweU_=;YK0C2dpV+iFj^ z1--UrZi1g?dE~vk9mZJybJPK^a{p^+ zJbMIZlRvy?&??o>t>Dts{l4l&*_ot=Yo^{*+W=@p!?D~Nd-fjfWSuEM!)0W`_EE&) zh7(@LmCnvzR2}X>2Mlw6i<|V;gD{`*#`@Kepw&7szgz69=qozsjUckADNJ-%4mk~qz8nHKZn14At{7Ycar1Rs`c*e5tnOEO= zwM+~ipXd$9)y&#N49uCtA&GL3p`s1fQu@AXt*5sIz1iqPoIvk@pZe6&8Y4_&KYy_N za}JhapZ^UdS4f4qnXu3TH>FxFW6%QxRnG2%M9Hmgu+@jyLKd@D^zDB{Dm9}TN7AS*r@Qg?q3x!UCyX>40Sh88R` z^p=mQIg!%;XiBN%GP_${5_9cN+i5ylv3BG1xfHlzLzqho00N6^NpZq$;D281Xb`FD zx`}SEB2A-dvcS1?(nI@fh;0O?!prxK4`Ic%r-M$r_remqY8F>Y{~kFR-l&tSUmiL_7q`6u!j<#Zp?CIj0q1@kC>Cq?5WtVM7y-9OE;W#Owpx0p5Rqy++ zH$x=@mwsa%zQ6kQ(TE$9Q`B6=TY(xuCx>OCJ+^M2_$@w_b^-$9u0EA6&^}7YM-eyN zGUNNL)s}dO=`HI$3IA`%4u4Z#|LuFCO1}&|VlUpd!5=NwI%Vs$kETCoXVGR-{(z{* z(q4JS@yJ6HtzoV*-W1{1J(w!3wd>XmWCkdBWSnFTeE4ufcB-~SEcSWmpfr=)ssU{v zEG1}Uh29+Fh0(qtkA|uAAy+4LnFDZJ0=m?XL8MvYP8&F7?SN~Il=TI1h5>;@_?F%d z**-gOL5b{XMSzwI(RF@aN1B7|<^b|I>}s&%*Ypz#SIO2VgH8(T;<9bABm>Eo&dJtK z0&=QbCx*3@G>FHt);uDA*Kt`!zUg@g%JhR_zG`p=npYmeUY0dpO3 z7oSx=>B`Hcue(d^6mSx=NiP2*_|0Doc!l(pgR zV|Rct7%^%((U_6F6(8895V?)V&1RMf^be4fKis^xWFTY;+lHdV;Q8xXP(t;nCjr~| zM#l0|!VW9uW19~Ky0;{{ECm)wlF{O~KSd9!$Xf;9bAz4mOn@5q`Nlf6H~H^9AX;M1=|K;JTKfIg05}hkNiwp1J&ctKfZf^YVZ@5 zK4DZgKfP7O`5Ws}N>0D3`eK{}VVs6*{$I5857@ol-cG~@m(><1B;r8sP_ zHPlZs-gr^V{G@*%Di4Z2FpK0X*e7~OmgR5{Q*+Sep-;WH`5O!U!ljNi`_7Pq=($tJ z{oI5p$NXc8+Q%!^Vu?~b)?Wg^yTz>~!pvyTKMBUGTuk@TMu0te%9rb znlU=(#QWK2zTa4pTq(b?PM;L+K7kR4pXPw|Ow^9Wt^*~J)(i(jvX);sZWNDH{$(d1 zXR`LU88XM8hOxiQke9;PZeLtSmg052Ow!Fxrl(sXM^!q=10T1MJsw0bPosW?Y6!Bh zX%~BL0)ib_3{LH5wc561tiNawm4rn>)Mw;g5lxGQ)7iz2rIrHd#JCJbkM&kuR`CJo zSA-?@Zt|elzHS8(mqKYjn6;PMu1m|lx>O4PwCv@QlWeHw`10a5VLQKSdx7CB_R&Gu zctQ8hkiq+>zyrReDiT)H$5aeAF!Af@_kDE+I&QppRgD&HyY*^n^~rCtdKSXjS~cua zZ6n5rx(8X{2R->E$v$220!Yo ze}Xy{IwbElZKZX(JX-^2u!8pGfc635oD(yGGaQkFUDHfU*tu;ig{I{*qBQC?h*I39OH+)6<5vcPQS7K=|}hnG^GU%H28Uo*ee9F zG;?u6m^ghTllYe#oPB$Zp}j)Guank7;n!2jpZ&alH6FC+Wrea+81x6g=b?slADTI(HF=Y&+|pQo^7X|6a=PoD-|_oe4$LLLN^q*M0#El=zg>J=COpo@x-^e zs%GlQxJ|%w8Wb+bqhb~@p{U2gIHXj{bRU0yy2}ycDpWpHVf}ztMf;eSoXv3)Xr|nI z9=G)#wO0rzmW2G`c_>M>1v1s4`|B^X*O4~Nw`fUXe&K67N;c?jf75a zx-ck^ll(yKG2C0|7CoPd-Q^Rh4LVG#bn&_3!B3gth6mB{soE`Vun5%)Ll0cXCvSwCKvI{NVqjWwAdShA+XD~4)+OKEwM|mC zl3lD!87cT}=nFyIo*%m{LojB=w+O-&_8b0FWOV1#=5BqUT)&qycR0l2Eq_5gMnV^+X>$>n}vU4LGHyWAoq)pUjGwP?C*9- zuZI}sL1))bHKP_0Qk=Ia9APFWHl2n5+7Jh=i-nZoL#D0NF2$g)ld1TRpQd;r9?Y(k zvHLxt2N8Af$5?-q(}E_tQ6Z-K^*mVK@-R1)2T#4PYUP494jK)WDR>XjH<)mah|{;P zu&CoPXg8_QC~|h3G-W@WT*kP9_-iiHsI^8fCaLtjq8~JYhj8`LR-s~Mc0?4@xztns zWjWwDeJ^-(Gfz1_gUN&nwB^vhFNu^35K9uVXZWCTPV0I)v2=1$Y*W6r=vr{jZqvz) z@NMxDyal4}Mn|4l1x{!7Y#Pp6I2?r(yEnAS^-jbrU5)SP75>H|=Ks;54ot$+xtCa< zTx!)^|IGtV3)BbVX?HuNrfwyi_~dBqt0qc4%jBzwGPWbUI}n0C%lIexDC#D0MSh0* zJn%%W!ba0YEX^XRj30*a=^$Lse2q?WAT@i*TGU)aLm*Q$IQ z%(=LbQF6^OtVoSn=OJ}VS9zaJFDEKDyM08Ym=-fyX{5UlJjNw~E76Xb+6-w)vzBG= z2AVyyYH4`IXw8At=(8;SUR+D1IK!g0)t`dBJoNb3-2L=IFDDjwFoVj&UL$oH<12&4xUeXk-ubUCys zyLS5VYuj-C9{!%)+v~apaTYX{L7I3Y3L(U{#$WFt+RiYR!r&@qUOTf;))4sPZVV00ZBzw=D7`E?Oga}TQM1I>H(0&PC0%Ss#q`?C z;hCe{55K%rO#q#S25%KxZ#Ek{TD_b^SGQmak2+f%5Q|KlI;z0T*L}O+BaLVLD)s(# zr~3qf1b4RnkTJvv7Put=7{P)|Kkq-o$|*j72`giEj(JZvr@)Bp{d*~Z8uW&Vkx|~ecI^OjY*4-nvO+K+=$uB#%N9afu zgrU`eDLyzA-Ug&eiss*O@Gm)Obv&RcpB?ck{%>wTEgf{t=5_n<2?VDDFLmh<^ zdnxU?VXU=l+e4E>%~+Ik2qs;({HQ;BKvvf3)0w__p{O9+8GV)@mu{}_o`rI&QTcSuoDzH<7v z4T99fI_Hst8fOA))APYpFi`R{b9Mp^ETIERT*lo#>r&?hLOpUJSC|v|#_zMWMuu z$2L2Kla|EDiz=hN5SSLiH*)D|9Gg7&-lBcUCfBLxI^sFdM0+~hvD6HfOx+9T(kx#r zIECdxi3)W@Et)>hWFLGqo-OVZW>8zMO|=?bf^*qUlt-s{c(qec8R7BQFTV=Km@MYh zRE&poz1A}+r)PS;o!_%%5t*ET*6elcZr0A^o)x5sX7j_`A5QTkjcq5ugPMXGtYl`( zUicW(*s(e2ZD0ycXQnfea4wD<7~1Y>(o2e?veW&CTg@#Nq;p`yOj`u*?5vp<(?cvR zv%KZlAIO#3PsJizU@M6=Gp(?n_e^1Vm{hQlon)GPX9ZxYPGS9@d+q)UH>9_91N+O^ zr|xfv*KXPOx2oq3ZgVbP7WJa45wT@We)?M+k*)1d5p?1{v`<83_Y1uaKLqDyK3Em0 zaRYpQS}7VWA|B_T$$2?=V^k52La`a|+K7b4C(NGTIFzpSYpWQ`E2!Eh9|^JK#x=ni zcB-Vs3E^q_lJ!aX-RIavETD#bdu}N}hy0{&;w!k3-@mqPG^+DUoqasShlI{n7pZ_= zE;SjUo3&~YL_?g@CH_k`rFA+5&zL}koSU(th0lt-a2#EIVxdMun9f0xeq$4Xfgy#Y zmG|vxCUqYznD5kl#zMSMH;9vZRiqgs(}{LRV~=fWO?KMvVV3ew*^)b>El5UV<8c?a zsZrgRA0sLsMN5{E+du#s7&9oCJg#g$y(^m}YKF!Ijgwk_>?3UePn@Riq8dJQf;Ywb zZ&bP{YxV>WE{))@XustxB6pmrDP8YQF3|7gZ5lah`t<2z@QIIlot#>yAFvbiprOxy_fw`ce(dR`>d?hgPK_~DRe zsa-Y2oyqx!u(~El#+jJ6G@E_=%ZAaagUuSmqq7n@7D9JLJ~hJ!Aw|0!&1TxN`^n9B z>$^d1>ikEfltBelz;5o)}rNc2Tlsm&$JOgGZTg(EL?F_8ljP z=q1;p0o6$VcE3tyPg{E}rszph^1kv@8rJQX3V{U98krIbdom1`4KM5qp=*jYD%v#q zicy_bIPWqpc3(H<06bsGto3*?%bIA|;x1-u*O=Iqw3VSk-kO*{4UyMQOdUoCvP5_C3u0m6Oe9 zrV?Ot!`&;c>2LgVf?;u_IlPh?C_hu&Nhy-}On_~JPYvfAHY)m9y_v7t2uw|X-gEK% zhBm&*5nk{xn?b<=5hC;}oEEseEW7L8ui9vyuaxPrk_okDs-CtLb;5J8k!^ciDzyg} z8~QP5tnI*A7VxDZEuY;q;-F<Sow};g>>tgtlZa0(NUo6UR=^Th- zW;oN_$wy{-P#bWR`HuWeI?it_-ItGW8~Z1&Q=pUjHBjOm$ZoO!}m76pEzaEF4v?E)YyXd_mtXw32Wkd+>Ra(EKU zG`E)+Q9TVux`u2iS0Y*p`+~mBuD9*B)(6|wA)!W5;I|`6QmOE7anOosu{R7YzIf%1 zl5N7)&z1WM#10WHAMfB ze0C)mqnzpdm?V=yfAj?JQZjwuh zu|?kZ*6z(@OcAE!cg(royRg*4uznY_=sq@Qi|*6HwnYvt zG8Gnv9Sd0amrXn#c+l&;OJ@n3?@QoF;4+-yx$mu$KDTK;|JrY{W9W@DdBhRUU{fx# zp+sTe65LW(uo$V4b$(O^#&%hx#SzODDjKl*v`};wc`Gp4PRcKz}P1FNfY#DJ0a7eV=E<;`+V(uj0ggt2<)7_Mb zyluVK4}lxA=whd;|KJrf?{lLdCkfP&-njaxYdP06(;f8q{)JxGzc~3?0e1uB0z;61rj zn=O(v5D#0eJ{qxC?Wh+MlXY*`-^e6#9x+jAB~2Y(pnpHUkW4;07_G9Rn)K}PLYal8 zi@U{lq*_b>>jh$7h_V<8PpvlXO4zr5==73{o38j=_>?gE1Fr5H;MfqTzNSjWZfXG7 zw0II|)`qFWYn3{E@-U>3FZo(+cB?}rW45Dn&gkeiSI8vbCkKyWt_{#1iBCDK@T~T< zYTqz@aa8U0iJBKjX}I!}<;dSyu6}W%ZFG~iPdPGN3~{h$$RlJX>!lXGDN;QiB^y~D z^GJ{NIyo3ha#5CtrV27*RgY5!Fim$1&&1v_Zagm3P^kuMugJZ87tu0Rf4&969z<8x zx2fqK>Yp?Dm?^=^AHiaGXN_p9GzmN5jq68JIh#G3D@4g(jMEB2PRkeabJ|;`)W_VN zo{?>Bz81Wi_~51}2Aa@vPFGk2;3lGtIMK#cqkQ}vxwz*rsgUT+Rzm?%^6KXgX!VPs z#+fjpvRwR+JOAxHG3O}|B;@$I0{rQ;Q2cmim$=CiEtW1Z?3@KG9I^(2%F|5on|<4g z0%jj!D%%EEBr1sjkr4vVd1wumeb%bRg4B-<#tK1@NHLkz5+`d7?NE}#$fC1mbj-}i zQ6*;;)6V;5reeIMYru%u1*F2=+~?$ED%(>+^hKlH4_$9ozaV7WUW#P+gQ*Ef5;}>R z(yg&#lU~F^wTH?}bp+6K<}oSm*;j1yH8j^c*O@m(wkZg3LOaK43f{UYf%wb+zL zmwL5;cXhYzu-3I|jOnn|<13^&%oG@TKV|ldXJ`WhoV08(N6H!l&&a<^6fVpNo>hyN+Il^&0*U zB=2TH81UxR?6$#WX*pNySwPg|F?pKGd988s`-V*gj~mfnbo!!x2J}7qRzO$W-f$-I zPxtB(o~(f<d(pH%3DLbW1aPyHhEPg5N;kl3%Gew`9~7U{51JR<#~2q{PG z;cod9R{tAo>?YK!916*Gdy2!uIy2T|wL#Aw!$I!7O84>8Yq^M{H>IQ9pTAjFX3lZ^yJ zUv@4sdN6N)v)o!x(_L4ZpLmc> zsA@t##RKJ{c0WMI!erjoKPuf=5=VI#m9|b(Jm!~&?3uWM>9`*}mtG3nK|(VV%4id* zXUT2SJHMjSNA}LuG$N5Y1z>2l=lGWLc6Qsv>GER z-&dg|f0{ETviP>KrGP&;J>M`bzjuRmpl3tbU12jd=dwOGXK@;(v#{z@N$x-@pU}q; z7EF@!)ungnaCp=xE{Nt4EzSmUMzz9vR-IEnb<35SJ=^kgxIju>Pc<8E^K*{A%r>)l9e5 z1FYC@m<|_qxHNeNcfyp$OsaGqS`wyDf0wV`@Q#)7V~73yXYDyX%_>!&V<(br*v-uL zR@Z)G{l;q6_M>@~)^OYmPWo{xTbpp12p8b>GfgdU+QO})U`ZFk4nB+{&kxYyLR77e zhG{F6^plDYQNPvWNH^FM@tO`1rcVl|atq<#scYd9uPHZ6Nob9s7$0~s;3B{VGlk&U zg*d}<+Ht=tzlhZ`HKiwn0&EdPA9kKMk3GP;B}dt(cyNJMlXNCTQM1znXN1p$fL%0> zFUlIZn!i)+LF&yXAAEAaNO#3A+}pC;P1%_O9fxA#cj;zCgDYx3)=6)ZlhVERYku^a zORG#!9-uO50P==73Z2xRr7}y0T88oYy{l=fx(3Mfh2hbRtqIShuoWV?I)>sM3}7sH z8@on;+~vx2?mNG@3A~Dq5bj_-`l`=(DADW_`zo1HCF?xF#pOO@@WroY>{QgpL2a%K zj^9{=9oOHGihD-CvAiS8c^Qm$E+%$a4fUVrsF%Ap8=%5j?8qQZ-&O>t?fbL!do&?Q zz`bbOfk#D6M|>VrOQprC@=j?}sa#JA`~M4(H@t&j+2&#y)-2vCWI0!G^{2+x7W#He zwR<^q`%yHhf!*(Vmvpwv_3ER_>(xpCM)V|b>6)2Q%BC$#i8Sm`ZdRiFTcSm+uvFbinpf7}s@@_x&F2biI!d5wd<{(SS7WiQ z1{gJjdke~SZ2wD6)7LWisdg!P(`2k?defy24kaHFvx!5gZyWv=0XPc3UC3LKRHXQr>m7AmyZqdzQMJwVD;xop=CW#wh3P zL23bmG(PwV`p5WN%N5Dw04|L~uA;(0H+z?RpkWP8r_!_F&=1WO1#4`*_n9%x^>?A8 z7*muq^@zJ&_tI)DFi7pP_0UfrkHlI!T#k_23IEmI4>D=b$Vqxel6rn31{P6?aF%qj zN}6P*O9Aa%CK|Ru-&wp%#9du-nuh=&8Y-!f+Bhj_u*(n=y=g3FB-vm}GVG!CjD;#! zMD&wWFDr*t^9=3LF{FZxg(Ne4tM`RMvsBonRBK-`nmnxAL7MkdLL}d&6bil)x>Egx zm3ux2)AYoJPfM0OU@LpeA?BMTR3bzn6!#T-yA-bM?6_6C-4JM@*yJR>HSOnBNe~7; ziYIn!rV*!PqZOz~d@3;T?kU$`S=5ILpilfWjqW91%@jaq>g>`;L**c_Kw8YP2`*X) zgiB|Xh&5G%Y3+FpMIslQ#$^{yTIR85anpG_m_2(r*mu&lVE0am>@(6l7C`$6Tb3uO zP1B@z1psjo%@Y#Cy*!ksQ^zH(Q|%QRG;ua2+@ z+f!yJ6QOBwo)VOYk>-n-9&}Rkp0Y1)?IC#l$Dl=s8Akcf5gY23AIeul`}3D z^tOcl+>i7kNgtS$Tn4tH_eQJ7vOBZz!5=}0)jH!A7Z9ORWf=*d+kq?sED<{*F-p>G zmxcB7upb_&L@TeDmGCtrqni7!Mx}9e*}|PNSm*Y{e{pHGBs9$kA-HKG;4!D<)c1uf zxPfO}x3p6phNba*clon~OORxzw1x(ROrgw7i>s3Fm>}H&HctGKq=sY%aR{D?vI6sh z`t?`Mcs)`BBQP!*8VjYA*3wsxLL2uIlrIiMB5^_od}t;S0(^@d)UlGJmcC%hM)PU< zo5$~EyBeDIzf`~S9u?eZLYZ1zQlO)rQ=+T%RZL&Pl~6<3Zt6hkMueW_&IT8asDvpn zD)hK!D=(#Gc<1K@X{Bm4iTRir7NTUJDoMOWt z%l)pden_I8Ov9)Y5|qnjl2WtFVMeDaFK(w2uW>E%K}j%C)ONspv-ymRNHdBKhW+_d zQ79y9Ej28{DFPr4Qrx6(eztnW`j;VwV&IyZg8x|{D2DQ$T4HpI`0yo3@@XmnPcM;g zJxdAWXf)itHym^KR}#%~mA8*yW7*5!SZjLW%s4qux`!9b_kFJs@wWav&aZC4H!3An zAtlTkd4SYu+D}KEEAwAV%nbrfZ))l>tOL;qfb{~{vbtv_2R~L6(E@}7-02A|G^{%^ z?5>S(d%yWU9Qp+IV)zA7m+aohv`+evfayR!f9VI;6}Kioy-0o-{6c;^U^%M1bXc8J zT{1kGN@Q36oCoy-0zYn+zU<2lwWaAaT~#{#IC?nyKRe_92M#M7Occj=(<9G?8~b>8 zV0cCU-6#J)f1f0m^|jkZUJ6u9DYW~$^fj%2&)tk;n8kf@Nl_@QgJ0YQE+U0&SWcOEbL>v3+ODlOxK?sL2F!pUV3o+Bht%v={vw3$` z2^uRJREhq}{j6u>?#($Q9-|A-Z2xKB9`f{-btBN+L=Mpr{OPS9*^AB&jO_+fGut#F zpys;u2Jq9lGUvK&W5`peOF{$u+0?#2tz|l%&#jWdWF0Jn?5px|h&4KyrK#Jj5w#rt z>&s%jEc^fZ-2cNN|F689)%9)uFn8`l{Jb2f?71Cj@p=X{fig{DVxllYb z0Fw`xd3vE5@YqOoK5^>?8usS9&8?Zo-~CwXeN4sDu^c=*+LL+8-drE#D91`^myt}& zn=F|M4Fo37KH7o**DDi{78Cw#ZjCSSOuqkVWmx#1wK9nPDRL!)XQ ze6zwyQkx$Lyq=kEt68ZCVz}$o=@}16&V2U8j63;1?S8;`Q3tWzDxVSwq)3?2P*An{ z^Z>U>uA4TA=r{b)bL)fJ*Dd4t(4T_K@Zfi)&z*3Sbasz9WZ!lxPOu-g87 zH|@5jcfc8EEl#0p7qV~loV4IqBy9oYryCbauOZ;H1eS;9-jmm zXv!Xw<&G$Xh>YEP@lo62fm@ zs;Zt1x;^T$+$Fopr^)%^m@69k@gQo!=pm-!wSZl|spADV+HRHEfgsHZ{?Lfpr9Szo zQ3#Y;#9y=)KsI;8YAkHV}Nt{3t_Ib@I~+Nag0&0(G% zH5cr#i*n4f2MKO5%>spA)u}hjwgqeq==Bh|QuH}tmP_YNQmwa}aKXi|CDq1{+wcBCa zpnQ}-coGM3cNFE9uBUyn;Hl@{RDA_AxmgGfr*|t9Za39oPWNN!XoB~Er;A^|vF}tr zMsYd&og}AzffL4O75{^4WC$Yk7{Na#o45hE^5PEn5PuxJBmwPL-=%~^ugF~exm?}8 zi}GZd(ozD}&b>-YQs?E!E^4WH5?wZJYDnBWwO7LvzP5XANZ$S3^Qtmx^=DgvL9gpC zGueu@mN)$Q25zQulsXmve?9V?FZz4 zd>dUL8fm>rXp5TN9dY#Nzg!?Ojjod)%DC&Z=T5z}pc+Z{emu_Qo!h6EmV7o08wDnf zy^LK0Ag4T_<<+5!*vd_?yrjc3E!JvdteF2|7#C~ z{+y8gf9ZsO&hxhu{*T__#eqR%*#!;j<-S_C)#Zy@*9)Y#WZMw@AF~gmwx_JSYpd@{ z=VG)k_Oyl~YRZf2GK9j4H`Vndz1u~nS_1OILj_3P9=nlYa#Z!I;*KZgS+UfOjolW5 z3O-y`2TH4IEw__oDd=NxXU+1Fn$4xXtaep?%7?WPYQjgWmme6Kfxod7bPA0=-0gT5 zReO-}Yw$J4o4liiJuhKOX=Gn9FPjg4a)5EGfCHH7muy+_D$RW?*bFOt%faFumc#au zp?(IfNuGX6DaS-&{Sxh7?^T`+=(<)RM($#Hg_!AuZ)VJRgJXeM;sSPU6}QosyYj74 zk3YjKb?@h)9qt4YS&^8ge2+QREaNs+Zc31>tOy;BLRH0ykTPh$Uhb2QcR@b{O9?fJSFoDK*=ujcp!|2Wok+1=jDipRuH9zfo|~mc z4n$q3OoH%xNZD$BcY(Z13Bz(dxcmi#R?~mpSYA?e$oeq;i<~;4o{c_}V<2MJRq%{M zvK*dgQMgND3wyeEUf!#Iu^H7*4{K~G&3MHj*ipk%5qNO8sj-TuZKi(D6kYbMrXrDO zhfoIwzxOM}4C_)rZ>?T1%*cf$Mry!p!U-yW-j?1*>cMxthdzX%etC~m(uwRvn?@n4 zh#jn6k`zabHj?hmD~04zBzT4MHRrsD^Jqcg}n7>0lUZk;RDpKisG0UlIFzkQmB}{+^bSNAiOeQmyQ}jS0j*| z^lzg=dNz9ULzryi(oNdT~$2?qE|^(ww?}v4>-X%m!ipp6{s{T%QaYMlYB0+ge@iF^O4df|dFpg^pQ+5xW?s5Yq??diiKh8fX z!36UisIFN$BQecp1E&VxN~se$1eZIj5W`0Ob>9j*vh0T!Y#t1C>PqW(t@TAzX>Lx6 ztc}HZEWlSMzNOLkQp40uf+DOMV-267QO!^3o|S&bjXKj3>e!#5Y2&vu3aJtx*4O7y zTPLhG(JxEKG^C9yczhW1N;(>une#(p>eIfiz5@nzna7cV6YgxXr?Na$GH9Eg`I_qY z3Tt_0vp7<*yP&C?OV3R^)z7JEgR=Z}m=;5(IMDrrq$d;w@vUF&U{ z8oO`^IRU(c$Ja_q$aWW9AW>ULQGNwr<{g03NIM?Ag_9p)P}()UE_4of`Y7U|0-#^( zd}^|-fJTxK;Q%f~+%1Y@!*}#&CLJ6XDh-Y@qN`x9d$0#_t;>kxmRiIcNhmz2UO~LZ zepM$6)=#)vc%Z28Db;=#ahIZps)vfvHYMzYlff*gK;}pyQg1_3QHq2d59Q~%UPH@- z$E|k={7yC3Ai>I9#L`d%Kov+D_~ueR8axQO-^h=s!M}`8KouPu5Wr;D=z%0Sm^l98 zCb(=&Y|q@k)6+p+E82^KMZl}}LCwBZ7$qCmO?BXq)YQ-gGPQ@nf{ceKVY^eB9$5*8 zgQ>53^Dkk4ml;)evOX#tz6q(&`A_@^KxXsmMga??<1>*V_X<&mU02PYJ^& z1m8{TfReGFJ2aApXT>|G)rp`?B$+0V+^_o z08CxnTRuS4Q;)*aI#|+WkbI>RnXTH@+rX{%RHq|dl{DuK+%<+haEQ0mcu5E#yBk95TS6YwrzLoGR-JxycJ_1%a7$jq)x-j&N* z$P64&u_9~*9e2YecWfSVPG#Max;9_{8|I~aZe#E{uQms1>tJeZ)}bZ*+day(LH zqzJ4ekLAqQB70{X!dh)tGa(GFc%>UdcFG$Z?5`x?W-@(!#auJ`seLN3$i}sVf;@*{ zof=jr()>I8bbFHRTUn9h9b{hTY?SOvS765S=A;swv|Efftd7%sq{>52ho=|)h`%J)-dkWKtg`QgLcni7n zIn<3R9dFLXW=X*A`2@teeeYKp*i`|>%l^ylo&EL0)hm^zZM@H?WRDzW@5b-8YKX&+`M?l^e5|3KJ=ie>`j+k)*o z+J46I_Uj8loa*43y6>P19nkm>RY&dQ$ZFFKxH%kpJ0>3Th0%D}8Ytg*IiN~ZlWq)Het<|d}X@>6W>22&3mEKg6S_aZ6z%ZtV^zzBYpQ;@lXZ`1Ij(6NS zq~9}Qpwg)WT!W|HbN>OA9+$j-{$=*_ZO`^E1Sa;9)h9b}T;%ZJT1aNe8L^Oa6lG#9 zu}7^tvn^V|z!tmnZMp-=enIQBF>AE9F>04Om{r19xyh0I1+)R}IvI~H?Yf|x@e4Xz zA6q;9u4z0&1kQnZ6G664?`kp*%j>gysNvqu@H-$bF}I~_{Tf+@Lb->1HGotuVO^sD zBEf|!@8KaBuvn6hGoaK|VZsst;2G9?+4xPrd{+b5_C|!tlJnZ?y9+xh*&)%el)0F}>7P@Nz7vW#syv2~wSO3wI4m^}lHLEqO^G_Zur>Cx`7Gi9;p0^ZY@}dP@y; zM6FPD)&9_%*b39lZa=~>Gsx|7JrY4dyS`jgw$;D}h<(&r0*+OfMI!oZe`u6D8S4Y^ zVLyC_-TcW2`r7i}2-k46xOdkiO;Bj{SLF6gu9vcC+opH|I(9Eu(}C+aI`4Uuj@8wE z*}b;@rSTB5K}KseJv`%PM+pR~k(cZx-*t<5KKjj<2VdUPoy4-KPQqQ}pew12HX>8@ zXIKQ1`-y%FS)J&;cT7q3=;EXq(a1H~?$$5`o-wqg)UwHau6r|l(KW0~qiiW+OBs^L zdn==wgYqW2d)S{$9Ba;@<+1ny>za!DU+7|YI<)LctpsCAN}R*$UBlVcO!|WwX;y%0Du} zz510uQ9AEv+%!p=xIq4|!@Vp1kK=q-G>B*ancCnyeG_i;C^gRN;cY?l%jcB$ZAFq( z%%qYlZA|`h5mA%-+H-d(-|PuW`(+0HiPg05BRUGzQJup;vj>rc)ViXn45VKqZ;1bK z8L8t{oU^?FslG1rEeUE8b#=M~bF>x?cGcY{P1?BMn-mhCy;4&~ z?jGeOLSdnjqD#?~i3d12O_1b%^L5X5eeNw&D={bGBNl_2YWW#Cg073FAF5khg5Q0b z-=!Bt9jL{K--lIy+FLJROzh|!o8A*n{I+G{CD+lC*m;y^ehA|Yk&g+HJ#Eu3NqM_{ zKlBSAnxx##n(M=!T?*SnuI=gB-Zz!s$%AiV^X0*OOm4=e@5qOxNra2;#pDgA|vx1_-xw)Izexj_r+GW0x|5!w3&ifv~YZCuC!D5`JQ-6Bz3 z-8dKI>BbP3V$FC*6l~7xWXkznCIP^y_sro*1Y9CG^C9&+x}p?hPPtkurX)1{FqM>D zVoyqx_|t7zhGk&z>V4_OG3n784~FjlZWj$px~?;F+cuE)FVn8q$%Gaw%avJLjmY(+z@^RM>(dXUwnCPTCNy0_54{m;nqUQkM%><5 zHEWtMhJ8eBK+40i<`;DAuLZ6vJ-%dAnfVMg)<2R)B^NsHem(Zm1JVN27vu}-7XW>| zB=4*9Ilr7%*EA#WcF6FP@M!s^7D<-NGBu)Jwgly}_@?B|y^odn7$(~W;kO{IBQ3H1 z!etzBNc4H80r6p%Yizzg8BP8@obub*KBUd`U~e=FcnT?AQe*yl0*dX5%NT9xU`~on zFI$YJCTO{8Xsbi2R{z_tIjQ?h8TbzudkJajv#9Z7(0dK|$E>AvPz**pLmx}fsY#^> zxubSUrl|%(LQ5+5?u#vd1i6i$jCXfwC)UFF2d3XNrF)7WIEO6|v>&ovtkf#a?V8XS z1La*EP-?$;o=nNC7k<#1naDL9$~6ghs8YUw)lzAFfcFM8N5`IU4~ijv8rBo9PB4lE zfyj;NeE^i5&QR?YvQ#PVRn5DY?2L?Uh(YXITT&TweNQ5P`zBFUQ4Tvn9!&aZC%uGs?;XRT2Bi4(`cyOb<*Xc4oiP@3Ycr84FgIq)ovG+Nxwar?dB_XuJL#+ zQQ|$lt$hoLJnEPto}|SQ6NqFI2z@y9vx!4)o)N9f{izbk zuuer_os5D_Jrz7LQG0VJr7VN!DDu{0jH8bRWZoN``L^g- zPFW@ESYIo>2%#M6taL1J<36FnCSxHayeLIOIOk!cFK(3faxAQaMIdCW*`tM9g+tc8 zj#uk!LxAiRo{k{dnJS|NKVNYQqKQ9wq9&On12k`pW4vNLc)VQF=v&11+5Z=x|g@7)0@wS&j`3F-$;rF*Vfvy=&N4CTq*P3gEpEnhhpOQ+s3k|um zz6q(uz4A{kjCt4mbo0m;kSwvXZH(SK5Dcu7#L#N~Faql0F;20pF4>WI@qGf} zAJK9r`+w&&#iP8Q6gpLRSpDVCiy*KwETpp6rycDWh5uj7y=PQYYrFT0B1#cNn$l~4 z&_Q}p=@5EC4NW=$0#br>mELRUpp=9T(h@oN;8Lyn8+SJbS%+ zy!(uC#yB75mpSg4`{8HR7;xXQi#x|WiD}^i86EL%vT?TZeG955Q5$ObaaNIFvtbz{{M;54 z9@-n@IF8S~<@LjJnR&@uMwm7O(IzDHOkMH3^tBhOqHy)6JG9n~r%kw@@1nB3UFPAo z7X)}a8rN){;hICpq3_;)ZKX5e7kWFaZ9ZHE0g}9W@O4H7rmkoP04i6Sd_)rgJE8s~ z2xPVEo}gm2FRENg80#FvaJ=DO6%_Aw>FTDHrPv!1JC1ij7|CRjq0Cs02Qs_!#m6@4?z`4{3fs0mCtlnC47n-`R-y{B6~~9=Cye{josf&B+&m z5*NC%S3>o8G~z)b=nSm%b$mj{xcRUVZI(t#R>q^_qk74T$5nl9Z@Q_fRt%#YM{juC zz!;Z-;<`9oYPF;Bpnz!q?Vkn+gP0>C9N`K~H-^iaKcJ{@_1OZifG?&pE#ZD_5w?{C zamBK3RG!W^n^{aETloUa0zzDS78roj`TopWpFhO-0IGjr>b1(6(pkS=&(D5cIa&0! zPSD5jTHSE-_D$$ZN(HX1`a93J^+s5aE)ckGhfTL8xqz}ub#;hS&)$JU)r3hQt-Gt? zOkY(qYD+$qH!4obz+tafb*cU%%PUJ|asz+~Ku2A(Ra`5(Y>|9&gxcXa?=AP{m28I2 zs^cqD%dC`Yn%|brg=-|ZhOSB`TAh_1x&h9+R|sK?k1Z|BoB8CkJA^cm>D0~E+?ZOj z=YS9(>1}vxJ!V(%j=61ksrNS-R`S$7hPjXF6~)WE;-gkOM+-ao&v|bQQt_4QjFoJO ziM!D3)&Pwub#baY4OLH@w9oQ&XD*G`eS#HgIWrVcIvMO=*-@S1<=!iGL;%<3SGwC& zYIp)nnQ2vdb&&xj%I$941d4fE=Np<`bhtH*rkqJaQ9j$+3(lMv_jN2)j#yS%c}ePGaCxy#}pCXVdV&y-)wy{erlMnaa=YHfl&C z+>l6S?87mzwMa5tkpH}SxN+xbmGmN_E$Ck-t3%(xfkdyhO zoNUPIn~p>)-W@uiz_fTC`kCLBP^9Im_W-uh>|^u>14__mbW^E#5!jGKvTp3-Z&lSI zqfQErmwttx_Uly6eCgK3DSxGk+pKh2MXvbPy_QgPjOOL;(i_@;um}GprAUIIRGQ8=aAK<8BctHIm55 ztVJeym@FG@b%IR~D4ew6YD4EJ$H%^a&Dj9x4V5Z+ros(}IxWYFE}8v~d)4o*bgJ&| zSKpFMy1y#0oezp4c>hqUo$d}`(L z(BqKm({5F5+`N=;t_?zPm5kivF@3{wps5WYH*W06BlnB`ZJsV91BoRW8ejzJJhc>9f9S(voswnd$*s zmY-Oz_Ix%w27bBKxM9w#LS+>gn5Jm>Hba)Fa{KK7z#%c+Upbb1tmULmtK7J({bzQ3 zGd-U?W*3U1P}&NCk*`Vd04b+(dBugEmn;0p^rZ-vhWV+o4xACWq4fA-S=o<3NL}-; zsYvS1!IOLZhU^lRJnQk@9m9wIa|uq0s|c&L@Eup*{wX7?{d&EECB)&o_K}cD#*>xG zX|GoXa{QTl=BJj>N?;?}dr_|;=HQMr)Z3d2#}jcaflp>;+WU}1mZtAN?}ncKge6)O zr0pL$SgOYqiMHgpa=pyHB>(+dc1y!uy5Gr|M1x)j!^ZTjXEfmKfrMV z&TZc&bmt@B0UEsoczl&FWb~6zMuxu^t!XU~#8;DkWKV#}%*I3@BwPYZ= zEwN%}Pw^%r@BBZ}R4Nosn4#U-~Qg)5XcjndLwI_605VaY-Q1>()}{^qK9+agS6zagq3aNb|9ad&gMDH1QSJ z_q}Rt#imNBQ1}ktOMv*Y;McpFZC%B)EfF8tU3o>{d11dm@BD76LFuU(tq>U z*7VQL-M{%QxxmBYcA{2fUTgU!ubuTWHinv?XN++INpIttM*3rsx%|y<#E|=!2gSYx z%D49ZcPGh+BYEFOAcl0l*cba|D&P6vj{HCW%zqf^oc8xb?tf17--YxKm;ZNn^#2V0 z&xvMCgS|#9CU^lKhca`eDP;Zb6&1J6f$y3J78yF%{O!zu4DLKG)Lqv7pAa^_@h0I;1JoEXJdtk@*DHWkpPMH+BG+~0xaTe zf$vgpc!&-yawyO^zkor=j(gWrD<6~&@DXe?#uw5*WU`ub7?bKp zf6i=xMD&63L9=(medmk9P=`l|SjOy-@X5}Wu%eFU^MmR@W5T<+u3E*b7_mvh>j(By zBcR|(8^p<;p;2+TSvGxLK_@SrG`((SKd;l)0{jiy?_$A8m&sQXq=5)pqgGB6OPrw%Jl(kT=ljg&8 z!H+*ksKEEq&EPId!W5?30h*sQD=ey5OV+ng?O84W3~gw3oNi{1@R3Zu-X4)H(@`sh zqf;2Ivx+>%_3=ufux35iHY^#-X3O&GkXwA%qz)0n5;BqW^ z{c0;sjCJ_92PLtQtmM<2Jbeh=M1=E(Lf?9~z;rwLybnH)|Q=QYN)C>N`v=-jU zhsx+GZlSTMP-(OuBaJnL37mfRoVs_;Q;|{R3IhMKZGrG|X}%96=R4h*+3d5@R4wXA z$J@F!SP)2QN~siju=oc_+&Aq(=y!b_UM$Z17w?{MVV3P9;N7C3)2fI;!2Nk=fvBFH zogTyk`dc|=m(xt9A1+RO0&DPaY?NQq`7(?CXMHO*2JOk~3p&+zoxMzgT1oE&pcgHQ z4QgeVZBHyClU~V{xpQQ>X^7wGirZ%~GNm6wn*6 zL@>V)Q#|laUf|Y#Zfn?_Yo1f2QS^}#SIY<8>O?u}cc`%nxe1FZqo|O9-D{nM=R%nW zscJSkEWx0SWhI$Ku#N|$PJu_A+`8KQ5-PL6Hej)R_r&rEXl@$#HsD~gm$)fQ1tusk z1!%g&uJGj?3vdAib(FS@473pgiV2x4YABd3bSgArMz)fXF$8!^^d?{lZyHH!AwgX$DFmu|*h^N=*8NYp5g$b+h1zNg&W_r4__%HL` z__nk5$rSrm#LK?&VUX9Rp_|!B?)Z^y17@0#=1laEP1c;BiiCUQ-qzZTi4iEZ{}t^1 zA3*zR?Shwmn}(|yZ9Ht&`Th)^KX{Ag?(()5jWjuFgEo-;9@d|Q(`|vWdp%B#1Nrn@ zGI#3}q^!Vq>iPHOuAc23E<);tA2*&3l-4U+3yW4Pf+Mx&jPqmS9rKc{u3<0Sq;^lt z(J>53p7Y*oj-R&D?(jK?iR)~&>-j{-JkgXr<4))zTNVhACzxJ=->`fr5LGizbM2+qMXdBj?(t9$pSf~t|{TjOj>0YRhj<;Ye7Z?R-Gk15g%Ds(kUyQfG z@7)WWkVVXth=rPL+-|8)7=3a(myS(AD1Q7l+staENA|QWxU~&;kWhr$N0Lp|x4Js4 z8SCUH4nsld9Z=IwamQ>L@;ZvOd0>z9S-}W?0HB!NOPXQZ>8F&3s`?a_4_?bSd}hB_ zoE@Y6_N!dQKt4(%wg_QZ@!m#|etdE0{?K`e=Z&{3;Y+mn@iC$UYg9-S2_a7iO zv^T-1A=!)`IOPMrs+Dg%VcOemiGwgB?|J1ibQRTS%*s$_La>L8zp&zBsX16|sr`DR zTjP!*o_!?Gpi zS4w=~)}5;Nz7u%ZfO0@pS3GSM2r8p7jwZa;s_~HJNSS6aWfUD!RRBd|7fmBQ7Pk3c zYM&XGA?vOFT!B4?%o5r1BQ6_~nWv9$ek zsN>rS*p3%!vnvmz4S3zx6kAb-Ycnid-=#-QZ=tg*cS!7eMZf#-RWwjUhM7PtyEKy#+g=;@inHa?yN5qYNTA> zSuM=snsJ#dbOBIY0qQi8W;UArL^{?1wex&x{E%MheWMThjc2_xjfXEI%eI#MLknrs zxVuCIcSCjuh4w19wkYN_qL`^Xq}qkR*s4@Jk71|-cizX;QHoyF;b+QRVT!jeC!#8R z81)=vjV-7YQ=M!LbTgPM?IXMbi>!<&$bG~+Z!E8=LfS2-iRy)1zn&v;31dZK04;r+XkPn* z@_H=-1jcu+(x-p^otpnggzq-P?|TpIcoS+twLg)5w+=AN@Bbii^#5&I!%$=3#OX}5 zONtK9=kV@lj38}y%rE5Xfhp?u_kiV8`&!(8RjRQu*&o_x7MQ}AeRFc`*Tz{^AAf0y z7+v;GeEXH$Gch6kBnJD{JjH7E%JQqsM%`Q8Q#tfCzN-mbLi^aW#(mmFP6O)MMqHzw zUOTqSFBNp=7#2VFd$=P?^4&wji=SPvrSHmlZnvQ0@!n9z{G>w)?2M&@hqc~-!~NXW zS3Dz+m!hh?@alm?-hzUjAkavw)6}*2W;au=F@`D$wPEW35ztF49ykiW7C7%zTU3xa zPG3^Pd(N@Jwg(r;GHD{Vopy`aMFubhTp2dWzr~v^#I8v24@zEOsJ@w#z8ctE+7)IId=$1mx zU1X1N#&$~)z|AAeQT7C4!Wi#bG*=xigtz)Gg@6|B@J4!~L$W%12>7nA7|#Siu5D_sG~w-l{*7_yo$JxWp~LNow2m&zf7uTkDX@DPoITcj-T^K98jZ$Exd0WK5Wi>r2o) znk91B9F`qkc3Tm7#5TCJ&C@u?5cdGn&&R*6eZzf7qQ+(*fLd=T^+X&^I1tvf_N%yARmdAAwtrEh>ZVG$%DV%%k}SKyjf{?@qRrk z*+G%97y9ls^GUW?#rZW3HK7mzIVRcc%vdX!i!Eu%om2h;`i#dHO>S84G-_elIstPh=6YD?CM)8%jt zJM!2V?xmvX?(whACv6EsLMRg4Wv^`9tZwRYy46{M%g>aq-%)(iCh9FCMAtabz124a z2kP?S^E_2x4VyITHoyj0k*nP!>&J&YcpuB3FbHj%uK-iyiQ2-1Xmd{t^h;G88fdUY zbvA*5&)2%Bn%-gYwowqwOi?)&TExNU^a1CkM`r+j5G`%}m9|j}U>z^*i=TWv6!xuJ zpk6S%_1m7hS|lTPOe;gMSfM#PW2Hv$_YjkgZPa7)>%E(GoOISHsDAiBd-O(_sc20q zFcy@&kRzW;aK@T>vaUkQxAMZUO9OaMwM|4<>qV$jmaU-{Kq00^9e|QyudDL@0H12H zd%Hh}TD_na#RRN0;#vg!jrwP$>A;0Q3u(0_QC$g!$rYKR9b3&0Exa;4^eBUrZPP!T zUtO3SHF>gCYO7=S`vu!2GNsl+|OB}UX#K`H)8mb z&8&QE-TUE$lP|7$rfz4ftGR)Fd=aopJ4vfRLrjvlT_dSr4_mLBGwkGri>dBt!VO4aMH|$ zPks&=*tro9>;HVB1D@H3+vSaQr0U!0zG`>YxW$g`0|$p9fQ$Av`^`PS^}5U^NVeH) z{K?N?8%`ioBsK@kbY)v;hvlHU4j^^TD%yUd*^}ecIOVfvxIeZu@0STCIoA#(cRhx= z=-L6VsLr5zqHu8fRO(7hierm^u1fl7F@{=Q2(A}L&ystG%JvVEW)vSXsukd7$*{mM zYATx#gn)cJ$akvfDji>qhp?97eCVQ|#Hbex>^%Y{Go*ijQFa>6I2p9s`W|P%dAk7t zf0N4UZ$HQUNGYGz*)Lz%5EQ8B`N|pc|K(BOLUZrV&E#Su`S%-whg>x0ewiOSd?`qG zMtH@2m6T6Od`W2Z7JGR=Tl|-!oOtg4{b82bT6LCY|E~_HwK&lpUK*;~Y9gy1Hr>ik zAD1ABZB`;w1>vZQk*P5?$Yik$hepSVK-=A2N9CBvXpZoLX#+`_4k<1Jm;&inBzp#6*_}LS#LjEs2>j*SR#0N&Nwwr@xlmX(2hUn?@l>n)T)oh^T2|xCbKo*P zeW7ke@HCuqo9h_ttQ0?8)Jnyr8?tZo_)5fQDAY=rNADoLdv8Y=UVRs^whF_v;g2}r zJu;p2iU9&3zpBBfg2o>gfS-7#?oQjZmvz;HdQRx4Ln3&ObOu<$tR|QCA7^}j0*qse zaK;VGIpiB8o-^5b+N0Y+gS-nN*QtD%uTh&tzC(Q{JYG zt(fOw&A43#A9=+fEHL?66ci$xt1BRuZP;#Z=6SFTf zKQmuAMn>Divzl|++PFA<;@|cqFIvjU-tsw4NY_<7v`e{mhViU?V#l`q z(6qufg%7^B&xotoj9<0oZE%}OvRe)rLnEKODqbj>e4;Mmh@ zOM8fqe*ckQZ)G3l;Y6WDGg-<=@)$Fy#e=Q2THYSgp!bY|BG@RJ5uay%BzZ0|3uM0M zy1W1pZ;dOkLvs(TvJ9>c7CdV!$lF&p0+97RBa2i$!cA2*YLOOpN>SYn0$J(c6apdN zy(TE!2Zn4JTy3Y!4STu^YV?8{Xdh3~sb%gWMJbQw#@|R7tdXHz*2I-rXE~Rzzd#y0 zG$z6P$rqqEaCq58%t5|4u`bo205})Ag=Cw+e0G&^vY{x}_D)luE!ioZ}5IrQ+S6$e#G_V!EDG2R((% zlryL+aOO%Bb#%nk=*h1Jl~Ryu{Df48U8Eq4Sg$Adts%CvRru2J4_$ZnUS~E524xw@ zDUz0Aa^!`24tHfFQ!5MNSLKst#B9bt&tJ;cK`8UaZ+~rf@6_o{c)~jGk9Uixd_UO| z*I?s!Pj};bA!i!EF)uK}pOwakdA(Ab98%=EFeXsN2=R=Nz5Qr3u`lL9x@UabxXkQ}ERB$zR z&hxSwjspzMV}$QqPb}+8W7dYblw@t6lmAOj^LHiOWUvkS%V ze`N|D5CNc#%t(5Fho>`Yl+Ub|3J+8wY0+q5dosQQ3vH*4fT#fHS@$2H;LAJC0ygMN zn*jg1S-P`}fT)V}mNv=lqkmrHKHVE?Oh2h|zzEt9vq`7#Gvd4A8NI~Nyy8Kj$xAM? zPCTsMfL<3yc0_z{otC~9x4vXS8r!pM7t9#Ov*%(dI&Uh+UJa5BTaoL_n$d|vo9~=S zca?2VqA?fmoOKaK%02upwXm~6%9M6Hc%gFUS%%!ih9G-ytF469n$&G(<{M+?<7b^K z@ha@FR26O-bvs*9RVGowhRyMUTI=2e{)+=C9IXNlZW)zQg@&bWWul4;=PV%&^@fV= z_0@VN9Rh*}2wC0{W7sl?YUUMq$j&Z=#lm?0%+0JuqfMqX2e%U^8nd4uN}$2!N^MxS z_W+UiSBedBWK^j;jy}zGscp<2pg3-fc4p2!-CRb+C+5v)g}kdiA+BsBRbM%yE7=|i zA{oixO`pC@uZ#RaGT!hvYE9_LkBCSQf04_RWKr^c;JnFZ^;tP6C0|^c&|JkfY}Xl~ zZ&o}t9@9)1%!H3Anc>Q2fZ3%qO8Kg9oxcR)_M0gAg_J;+dLBI^>@{Z)Izqo+3|G}q zLPqz?mMH@U;D9I`W{y>h_3B}$4iY!@Q+&BWSLLHDRo>xSgYJ^!z096(tAxwm;-gy6 zy+=Wa`CZAFw;O-(K`1nrrAF%j>U{T@Va$cgQu&6sm0|Z5GfWv!w?sMUG@oppws{Q7$USD%3HXUNX(n)l}f>0rN=7#a5fFu{*Hc*|so-`X#7(9PqO5vY|cq0mTWw-RUd zn)e%WJC<1=Q|ZYnr8J*d&iTKT(G9Ih4s;H#@>Mlf!>!zYmw|1{vB~K0AA!_!U+&I4 zGwj_#=#L@eoF_7mUT2r~%$BY#uD^10*44{HN>wyG=urz4F^!k3C|%YKmV?V}EAAV- z$;qdJb5)UOc72Q8qR%dkZr7}rNnmbxW%+r5k6;CQT(NMQ@h9p|MBvKcm!D43uBcJd z9DoZz_(u5!-ZCjgx-_~^%PC&yhiULbriQA8Y8XYqvFTf&(wBQ1nVluA8N}uK{`&5E zEnc(`Q`-f3g*yd@SueUI(gK&UW%&Zc>I^s_?ytWnF$3iD(C z_|-k0wlm7t{0)ariF61@8d+W+#@CojXje}T@9B*qbE1bl@ws7fz)R3B!Wj|k5N-Ff zS;9I_t=g7@4<{`-IHF51N(Lw}HaWh%%~A1WM-I+=dtuApt6f(nZak2fvQnq_2M|AUK2ba3x5wK7@+Vink{j5d*g|IArn;O7O}-*I`Mmuz}7RqlFGGKbBLFROogR-;8#fpj=srR!hj5@KW$K-%QeT`Nu6J()Jo zQ4<8)!2Oh3w@-=hh6gNt$LVur-^t=%pa-SO1%yXE6lg-fy;)wH*VWjsABJM0*|B>V zie0_XEB*6eS%{Br&TVE2kovT6OzjwKa#*rW&+)+M z8p?lZmRC}xjl-&0cG>oZL*1=U)4$mfqj|?pjm5RL2%#!4-FMxrJj+Ox+PnjF6Ti8I zu~=&9Xdm$+vT)A39LkhlI##b!7Nn>IqTvJ9^IcPqo~ohXL)! zru6m+g? zwN1W+c${Y7(99>&e!R?_6AViko6BTT-#5+s^onQA6G^&W=b^>OVngvsGT*qub^Syb z$X`Ah6XV!Bel@1Sso7Y4<@~L>ZsUT;%6H0sylzt)u+CKtyqj~M|59{qC!OA5G(0r0 zV1Qp6cYNFVg`>`$W85&_Pi@yl74hwUc?AI0>kw)x!la&L$Pw@Alu~LHFf40K?asS* z^O8P~+f!)#use7AKq3AC%BCVeeBlntc*aEtYxnVZ(Il<%V8sNQe(yRP4dmdL2>vq>6j!wxT(}S-% znW6`Q@`-p-L- z| zwZ;==BuA#(6P^a`GXm-r8%)PBE=P&3cF*&w{pc9EaY>9f3qD(VbqkO_p`ix3x5;C3 zjwtpgR1Xt@BihJTQVzcaEN$&Qo$Fzhin7f}c=kBgtIxNT@ashQ=m zRq^|{C*Z?lo-#!?Ib$PHZ{()QY1``?7NIVRr|vhK52^*N9L==3sj28oCiK0Q~v zr(at2-dZbqSr$baJ+w|m#k28i){1Dg>!QpcD{{lice^z=pSU%1eE%+jpQHa4wW_;T zr!^g`-b9@h(JDt~@U}87N_v(Kufq(EP9hQ3mAV#86L!ve?R;NUVwU*GW{@qsV&s_B z`2C1U3!!L9Y35{*0{F`8GV*%?Q5l9fF((jnb6}rvpo_w*IV~u+wnhu)wDs!csc9Y0 z6@grT-g(330rH9F&%LsE;FWOYZ@w{VH$Hv*W=!Iuvqa{d{WiISbE_Ja{p0{aUU&5J zYpc_Ph)5T050F+P`$LNtk+kX20>&WCXQkXD$)aC>5ems{FT{m}g#wZoa|SO@^@AoF z4QE%b`hXBslhRw*dUBEiAyYwu$%3xCXL*T3g-@H-v%&@KLiMU2wFZONP}fjiSDQOm zY>lJ_54330*C$Fm>P8$;wMAU!&Jrt(Q^@C5HfrWCT&Na|)9EKI^mKIhC~DEIsIxei z_+MA{7!!UmL`cK_rCHnx6xEmoE-X_hB;fnI~s7_=z(R#DwFgqtwO%5$;8#ng`xq^>j-W3bb_nBFqsWW&T zQ)j?vrk%d~zWD@UImBhy_XAMwY}bPM3SYz|CNb~SOJg(rZm_;I3tR{iqhzYPeP}M7 ztS8(ek;;N+R)3RJa`p=Hl{u){`=?C1!*%*{5T(2vKGXstxf0%ksm^FGb1OS*p_Syj6!rV zEllzhU#VAB_^EDEK9_zfD}?0<-t|&CiK6E!_1hf=dLD=f{m6&s(rDQeZ-6pSHmQf< z9aLI}ez@5F91Y`{3FMaSu~sd%4mW3MCmyGsN$j=&SuRh4tP{!AG{=x$WWd8UR<&Y~ zKd7j@OLc)=eX=+SZgD*`;U4cW6bo2X)BU`#V{GafU4kWkmwGHnt1+dR38ifHRT5*wKviQtnQs79z2PiGF@Z>o;$(#f}T zmcj65l@TY-df<_|m!$J4h0kE2}uOkhmP_F8SY1#qacppY!Qag^E>zxH&XqYXTxlF-fpcty>g9@_5HMY)eVNBw z`pG@Kbk+;TPj9bO2y88-HH!`PK5uPhfnWPs>sK%cfVCEOyxFzEjyibclHa#kqmkbY z!(u~{NE9k=@`{B*PuI6!w9}MYOd!c(0^?=G428v)wYMcZj4;<{S$<#{R-K)WZIOxYE9UmcreGf`J+rE8|v4eyv*`R;VYO{Y9wjeVn-`c8EPXxTEio)0uRS8RQDrLVU{)@vEO zYE$V9*xA+dTv~JoH%fP7q$-L$K|$7+e~`#L`bt$h9@{=w^*~8WMBK3J4qGlY&6ffS zwb}bALuX|UUdo>wZVz;hk1jMe`V@|N`cCEYIGz@mj+#*TVvZS>JTQ`C5xfEneh<2D ze4<<4Yg`FMsO&Cagq&7fcDH4*H3Jp9;Kc{pD7AK8A-@@-m@Qtjf3$ZS<~Vvt27PZr z%gPXR1(8 zlt$VMi<1@P2#Jx^|Iy;2!G~=tVp?uqsJ?f1KtN5(%g#gT8Nk`2chYt&l*&h^y zWj1x`-5E%qxoeR8L6Mp**PAYN{#k`(M*NmkfrBnY2l3Sz?|9X%61LwDN_)rQYx#0U`-8^p zzoe!WzB+j!O-vT#gJ9z=W3c~jZzsVqaBpnYxz~C5(Z{hu5}Ng1kGS~j#A=EhwE=LV zKc>(Hi);V-<+aNrMZyaggLY(nHy-RCyf2i`-u|j^tzt2c&(E`BjveOV?3Tr^nEQ6B zIJd&upHxG7m&L6Jna#i2@o*8s5*1g;`0R2ly!L+1!i*BwXVh|ZASXkP>3of;wN=#l z`LxGVGJk3(y^aardjU5asI0+(#Ih)f4V*>N3)t$S9v4)4*!b!K@k+zI(EiBrPijs;fiH#AuELBdu` zgO8Aeb>fW)Pt^pVPb69j%}9&B(qik_E}|gb-}f#W_Oa$xS0w zX=rP+ib-~;b+rulU8TV$E3Tt3Q3(d zs^!}cb`Pa8lPv3csg{7guhVjJOP#~-|a|^HbssZ<#JFTxc(J3I&$rcmPvVwkhBj8!8@Y&tW3-@CcQxeEBy-x$=&;Nnwf>b z63ylUL(X=S_H&wCO@FUNypeW)0j&7js=Ax)C4whZ0`Co7*M(HCcq3->!mlh88Jclh zlQ&dp#BYww@HxAxoJT=>%;u%}pEOzKAHsjG^_JQib5xr#TRuZJ*5eS{ztY5}=H(KU zsCm@v9}k7Ajm@mAcM>#*Rrd6^p2qLAS>J&av5&n!LN1xpv2D7m90_LWY%b39#6TBnys6|}hzn$M;&BS?D=<+jxZv}u z_KNO2I?zfv=RoK|hjsU5IIgaAn@hRRSR?q``xEyf3DPNf^mtff-_kAll_`-s5 z>$#=^4_`R_lmO3Su|{ph-nfynW9Zk34-SLZalyyyIEn1xUJeq&IK>X?p|16^&J(NE za>%4H`DgaIPOV-&I|iFsRxR)1b@& znOYQ^^LfekRH;aD`M#Ihr$cjV_|s)I#Ikl6a}myv1eTPar`s^vtA*#ffu0W#|vkxu6+2iJu?|m15$1jvUJLrW(6W{0BeI_gv^`l zX*^oNEE-i)&-{qB5EtJPsqy3uo;Mp*N$lYat~&jBA*w5wiu3()imJmh?rZK3*Ov}Y z<|eeB^&2QlwWp`v-Ib=2cXk#O6cOfc)S5(aUV{nE^rAXM$34^^t}pJHFnyknooZe~ zs8W{#KqlLa;x!3C?uIkz9zewa6a+SI@ z>T-p5bb0e1J4Svt;g=jAJPaOme5%QrLd2%}5N;IVq-5PqcF$SAVZLD_s^^!qZor;= zX~<<@VcWW* zM+vup-%4Lx6(11_6o_i#aPm)2Z&mK@U?8h;<&-rE)Ia2T;j2VvE?ScnJ)9N?s5n9z z2yO3;PlF#-hybZ8@1@Aw@=9S$e_F|h(VY&Ug;YaBL`e1yA^d9K1Ts<`B z>-tu#^(FXv9X=PHm7kF)A+P8z;JF#UdQ>DugS2P&D>L~=AdU2mTjBr%8^$-`bavpR zTZ|sN#&sZbVJ_VPUxsajfd`1*)9S$@AoLE!CmG{R!&JRet67TPsgMxi2K9{CcESrt zxh%3O^&FsU0r@=OcP(aLj&GcpdNZ%nQa^#_CV@7%F^t}U^^KH+8rUzD%mdP?igLRL zKv_<$KS)w|*v=ig;;=OOF`5G>G>KX=kv+^(REjsC*`V;XK;N)rA)}C&M@#(p*gb;yn4Y^r z_M<>Z9x~{=)^5&aCX2>dKm!k94Q}U90=bs zwvAI%Ab(t~H2eT|BPW(7?`YudvZZOY;muCZFr@$KA0(QB4=L0%pRq-)#vQ%7DE^l; zYe_&#t+_BMCGOkJ`{?&I(~}A)k*9%+-*yk@B%LaGMNr(ltBwDtWq0te`is)L6=~Xv z)_x=aXB3P^`4vpWkX07*pIXy!wab>UJJ$P#+RdXN33^ZEi_MZ+ZJ~Mkt{{`b5b^NH z?5@fRkCUv55n7VndZug^YkPh;B7c!?C3pc2RpF72wikm3mZ52pDX6(+a<#&6uN~n# z<`8UV+>TV={9C=$Ygr4zzA!kc2ht0uuz++`&~j>>g@`!^jr}j5p^uQm3YAxAA!hg- z)P0sEeQ_G*pxz5iPY|i<`TS7C=R)BEk93sOFWi}dWflBfKzWZQYfirwES&BI@7y8<5xIr}PUp`C{(*nhmw~ur|xAD#~g0 z-3>2RSGZLSuw;v>`VRQYf>QQVUpTrmx^wRe8FYSTiSWs}Lh*!HIbxgl6!8PEhvS;Z zWU4)7uV%K=qN<3#2TNreZSx56!0WSbf(dF6gNvlUW~C-&WVg`1YZY^uUn+^%sWDd-$d9izDM% z%1P{!X)sY+)P|?w6Njh#YnftU#A4?lv|`ZpS085|>DXyjuH%BWW#IJ~EOuc9Z~fFF zu!~Hd(C=IVq%%hA^l zI+hL?2PML6IcmU;L#`6$@5ZkBtpB1Lwv=v)`EajsgrLx-$e2&@v3JbRV_RoHpuH$o zAHB=vPvaBF+4-7)`1&s;sV==UtMr&Z$A$+Qp{e}Tdmz<0EpL(QcL&M2@;@eEHQngp z;FgX+$5J6H#KW+ewyv5c!5OZexB(k(Uf=fg;@_S&BnGc^1|_vv2lyw2tW%XcB<-1xZ#z(ElTb0RX?o}?Q662WpC6=CSPmLCRuSy`?YhinnCBm5>!sILsE67U= zn^Kq`e>zm|y*la=q~xRAX9{NWFWx`&JcgtLnzP^ZJG6C_afU7)*KgM$;wNMCZ{BmA zYf=g{7yXv@w9E45(kFDsXfDUGcCXl_KZo_){jp}3H~OsT9*SVmc8fkB$>G(-L&z*K zw_59`UXxZ_!mh{nS#<~9N_40}CjqJMTr=(9VesYd_>88B9db@yXq4XNjvw@}+4t_~ zq}rc=xX{um)QN(Zr)<)gJexk$Yn{8ZA9W5~+}^L~Eav=umhv69YIvXZ4yz;Bm`Yp_ zFm#+`B{%P6tKSwkkXocNph(~zTbH&mtV59mDSk8UNNu^!OEWZrEd^TdL{*yuiNG-V z=k7`h1228-Fnl&i0(sD!qCth278+0zCqtiuem@i;NWX*)$gC$Iv&JJl2+y~N`%%gX zEL4V*`PBGO2Y4*{o~Y!)2O|`wY=K4Xh3*)|qNa!xp6h_S6@SY1kF6)Iaa-IU52LZ1 z>22ZY$?s!VWe#rSn%&P&?p>V}z;>Rj7=LY}5GcerfVF?Gbk9du5u%c%=z=r2ye|8i zUbi%TpQ~|Y+xKM#i0P+JxyE3RSMeC*U3rWG3nT&F#=?w^R6Vw6I!F>@$V}m(2X=RA zoMFCIyme(*&Mb48ia)e>@5RsMMKW2r6uo`_Wvn88h0)|RonJEvpn4M&Ov3H_X{Y@) zfMCeRJC>zTgmsBMxg#cCTb}ccJ3h#!<1#P07u!sd4gG;Ac5PIdp`Kjh788=nc^>$r zq>4QxrAF&RpQ~XCd*57It4SZLUu11TfBC(uj!S|8)veb;15ry!FAR6o4Y$Mhk7MxB zg>Hwtiw+(GY=QS?>?0%F$`!m8FoW?pRg7j{og(z{-f_x;njh-L{AbVO1$qu;1sZ8g zt%O(E^k-SB7eqEwWBa)A!t7zdd3CzApFU9}8$a=7^}(Vj|F!Qln^|nS*1FKl*3RpA zLBCWFLW<<$j=PaVe1!ROFgRUAL3aV+#mGx<|bRV#MrkrROk*oo``q7AC`m=gT-1Gyji>ogvGr`)k=Av@_Nm>AWPtOKYdp#rv&veCol7u*(qJ zRE)I;;<-~p(My4w4)Xq>jsJ_i_l#<)?YDlhOBbX_l@=g$kPaff1VRa&Pz+5VgkA&$ zmELRUMOuK+YeH9P(os4Ipdh^}AW9KW?&sO>-urp>{f=|a*LS>M__WAK)>!LW>ze;L z=MN>k^d%QBIpQy@zpC_9CtrkCGJ>Cv`gMzo{7h^KbsS5Ty18~kJ|75MZXDeOOG{*3 z=)pOcEgJaa5ono4-0Z>Ze;aE46%9vA-7&b0Jb2ixbhWNwwpGJuC`05$RAqYg@}Tl9 zVZh~U{D509p?>jvkX6_vN3zbQoOr273 zZRp_n-$mhb>too(3L=DnfbxZwAN!2ki&};NZRD;jE$|0<=3BNIy$8~HixN@k`GenQ z#Xcj;z6LE;1X!mq>WHRigJt#${f6*MXu;8R9-1ZK;Kr_1f7>g@2=S@oO`XtBx9+|1 z+uvCTHf8hMHbc#|lX)@;{%rHM&dSQBW*c*|r{<{Qiw$w!q~){I9rRO0mP(f|jstpdkL`KjNl^}*^VvfS+onB(QHAlk2}G+FN=sG%6W3z--?_un9G(9KJ63R~VHl0WGI;Ufj{gpcQ`(6}L6nFtE8Z5}_ z=Nt8NLTE}iU@Hzp2m^W!)uBW+Pod^E%5(?LA>D4kcr>=_n|f*~Ah4z}tGAhf^TX^A zLRGM7oECHPwcQ!=#zNd-K@Z)VnPn*Q^?C*)ohJ85*{o0mDV0t|Ep$w*5cBmh0Pf7? zx|{~B&io1*Nq$SaE{@{K!XxA&=(Hja$!Jrxx(A{|+ba9>(%4@rF;%2TQgFL4BOVJ70SrN3EkNk@B{-XDYfe zgD$M|#yDo|CA?y5@>9xZhg@^l%iQ#2`IhOQ(}TwpSYxJkWjlcOQ+;C)oG}E} zT!!x}m{e0Hl;udyU7A$X)Y13l2WV@Y-fBk)^521h+sJ)&Hmtt6@hM&1*g+rJ*?DB@u7Q90UiR1hn5ac2H`qfEdq8vQHj z)ZU3zJGP~V4w7$tgE4pr0dwpBnE4fm$Vha3SBcw$vj&j|TFzvFluH$`bZK42sU&YH z0^^gza=rVPU)QzW`r=9v-zL3* z^+mndg8C?tFRLr>AWMOeTp)O`Ua=du6Xy>&RsK#-hifL$_jW8)#9J8*B~O-A1|^M< z)wxjDy!u&nqi?N+ed}&aVUQOZhOq7v4yaBH?9YA=%_v_=C|QEg9H_y&6)h!L#a6+( z9M|<*o#U%YPjIvQ%jw)IdKosolXw@0beU|}v!JSs(3ZH}IsIkwie@(H)DCy|JeMwC zMH5`^q%A+ho79nxaeKb~&dt=R2r zm|7u-+h6G}8AS}l@vI=+O@=_xB#ir*>BxsBYy))W?etO2Mg)f#^)=2mD6c3;Lve(&%haKq<8zd|@V8RDeB+w#qF)ul*1-dbl8W@y_Uq?>_6pWQ0Y}v-I51o)7y=o*EE)DX503@8+&8Zx)m?suX30w-4E<*r>+U6ucrXFa4O#HHHuOLG_j&u0*5JQD*pDNs(wY^I%WDr|@Sj+r=v z{Dsnr4^H{4I!kqGAMjRvT4`-&p1G1_*wxi9B;jxOhx0ph%sny0MET8h`l$^Ks&E$g z(T@hSTgIY`7anqdEM0PLb_^C0(uvI_5@xm20;%=vMfSeiYdmv|pCUp1nGIp@^s z%hNl}pA7D7JHqVZ$8_#2e^hhl2SN^lluxq2pw5W6LLiBzjS|N|LtT^Tujy=Qc${ZP zM=`(hoWI%+7O0RcurY$Sn+cT9NLzSa12E7IJ#hR~Xi30N9Flmtm1e(o2L-Q-)D`${ zVDJ`Q8A5-Eo(`2xI>^aN&Ou`JZt1mZgjK)X;XF%_-U*G>Pc5bv3Yrm(%1!YU1&gT~ zzoLS<4&V}(S`uXLQSl0L!{)GpAJTGi(*kbu)#NB97UP3ad|6M?U!vwpTy!HWXsfb> zVP!cMRw;nPpd^OQ175FOZ%AHJ9t}TdLVI2n!m`is$f`RA3G9bvk$%##;zw&+7`HRZ ziiZrDj_(`!$kDl~m-2H$xI|N0s*?SKeEKv_P(LS)a2D#eRbD>R8BB;@M|OD4zKkOQ z8(2AQamJd-T(KCQ!1UZ}IrsI}&jiGX9?KNF$t7ICW7mkmG7JW5JsIJfAGOtuX)agB zWhX|QgjI(#A{-eH@|y>Q>bIi5Ycef;Yul*!_Blq9D%o|U(R?t9QtA&;@H2BAFHzBZ zf_nIiPQ2VIs#4*hg7umwxU0z3UF>MT&fcT1yGAYmpNzXk4rk`S^5w7}ob;GAqlchM zNl(mjoV~sc)w9bNB0<5NLS5M{OOKZuUQ<#AQ6V#WYXL0)c49dpBNyguk+lzET-=fo zPaQ!qn64lR6Bc48i@ZD(>XzJXgJ9p8dye;y3!t8Rcu6K|zf@PrseVBjMitUeT`VgO zn82_9DvG0eDPB4&->B5s%U3*;U9e7`0^uF1Nq+gbljMw##P~jxxxE@n^2!3keLm6A z8d-qoZ}zI;H6GS;Z%nyG+L^Tt$^&C=x$foV@*?Y(nrPzoB`*W#R-XP6J&^rLP_8a& z|1QO_N5`SSIfzR~%S8EjL+F7GEO_@xS`WF-8Ax8Q5qn4Ysk{c>pk z;=#3&?vY~%8T$LzO?uswcS68*4!2DQL@J=Uq4S|MV|Y{79PT@-{>1qcbHHmS1Gc#2X_P>_$BYQZetoTQb78Z>?L&Tr|6FT=8yl#g}c z5iQZGr#cz}CHN&6Ua;U0Aifs^6HN|JkQ-TX!G}Pc9mz0jYb!VN?b--5>L9Amrh^up zmou?I9x+`2Tc~W{maYBld_Z6i1>V#s)u&?dYOu9S2ytMEo(t4*I)-2~-!W)ToMJLLtMVqRw;k9K=$jM~*Yq}0+S%(~yxJsh z8VfPH`g_G|K`nZ>h#KJdvG_wUqI(^`d-{7ouF|!mEICquye4aZuVoexb=%8Lyt#(k zjm|@FWnSin;x7o}Fh%Emo|pPRnfXLq z1$2BcTJkKWyZ=p_dRVTBXiWlJ+TVHj$qZ1MpUIQiqO-EAoLZh^n9w#@U2YNR8~aXS z{0~v?qTbeTJ(WL152-Y&u8g{4HNcFt({B$YFMm6Tve)nOwS26JvN|8v42EH#x(?Mx zB!MrG19wMq{lyhIu~~>}hD*&0Sn{~9%y8a$%oUT#?LOb4Q8mLHCeIU;6 z-UeNSL*A3bXNCo*047_6Av1~ZLnq( z91gPMjKj0c=`^9x@%)`r7OUS2)h_yurEZstsHY0yJJ)%2id+oU}SAq5~#> zUGA*LsQR+W)X-n7VG=))H0 zk9a{Blk-_3ujTrvg7_U#9!5`tW;76TAMxXZ-y0_^;P(V!+Ofc|@J)%O8LYKd!Q_v=-dY%_q5Q;kDpo z;)-G0YY_h)i$A{!6^Ov5V?$oAmjedJO4M4|tUx^a*PDe_42!7q-n$pc1bxZ(SuI*= z`hK^jqI4*Jo|g+TQoD0tcr9MRev?JrBSZhT!G)TtMX8oDv+D;wl~HK{f(s!GY)kz5 zu~}fNmx>C1uR!rMR=4EmlG?6`W9hvdQ8h7|qRP>-Q7~@nW4~#hX z+(UjFub7q#Q|Uj5r^X%z@sF(=rp zDOruMhIe>6mu0N>USIINFbfSp+fy8QOF7m>wSNU-@qSiOXPstGHZ)c;{B-LPC9z+? zP+nW9wi8QLfkhCFx42`;hW3724?+uVP)B6fDwrbaIp%umsRYo`sJ%j9`832lOf=Vop)+HP=N z+kDIwW;*MmhtKMLGi8GAF&tjV;;d^|*U4Y$IIQCAJmjS9Y}|9Jm7|@|7?IsFlt@$H znda~m6`Q)gU>RKRZJAMZKGpHrzL-eqPvK>_g{BhHSm9~`7 z0QZ`Ky>O^?{Sn^bDN_a>tr_IAjxQdmlJa-3JLshLBwyaK=3BY>BIN{O*lg|t?b>Ve zv-ll?Hu~L_vgG&Kqgj$cE`7#whD(OLQ|)Vdkw;P2H2<|go#)9o=`a!p!A_IIEE50w zoEOZ$b|@%+Z%q08obuFX2z(k5$H3v2t;jSaCE0)XJyn^)~_WqrSgIT2cO9Ujh9|7n43uQ6lRSmo4{5TdUB`4L%2eY++E)Gsplz}sDp zlm{bws%ozFRuUItDZm0SAV~vIK0zG$p%`1v_$hN)W#6PkJLo*0Yncw15ZNq+{VH{2 z2$&s-?_lU@7MgAMeWf%v>whmLs{iyc${nxW*OhuY*0|JRLcQqJV|{3@tnS*GgI?sk zAKNRuC8e#a?t$h5?+;7PR%QyeCXZ_VG*OeS;g4a`9U*z&cW>A< z$(SDIJP>}}3MIAJGsr{yu2fIOX}dS>b-|=xNbf7O;cQ-TISk+KPH2%L>xU*rWjSeT zf2vE-d@R1f`@YxkniPaXe!o=u^$6x&0o>@h8N7kPFrJt{Mf~ty*n1)ybBSmfc8pCv zRbTUcDu2Xk_ibzU;rcrJsXKh%Z!3^y(J<}JVM<`c@UgsZ}$#LX@| zPZpe?(PUDOqY19RJ~J_!7MYwdzGx&*j$Me2u2%N(x7E0}Ah_@*2Vb}Am>pswuu zBuh>u64NHRW}lO%9|(^}$220VMm!l9oo{3cp8j+(qP3;+RF0HV@v@8anvWnoF?`;) zE;7K1=eI~kT77UOpSj@ZN;46S`W&^VdHqbLY58n_y{vvvSbT~?Y3bGsHkbQq4l4|> z8(WQ2jaV@;$u4!6egSMNT>@5G9778ztF)_;;f+cbt8eTbp1sxg1u$*2eqFZvL&RvK z##xp?vSUg}n*=`giYNv07~Fx?JZUgEUcQie>+hiM$S-TmZKw@wW%}s5j~a}*F7Kb- zg1}rW=K_=heiD&-kRQET2ow88w-b zURxe5wDn}Z&t%3h4W?@kxvQZeA-EG;(wp}ph~$q z4#Rn(rY4|SBJ)Xbr$#c&Rv`|jj0aaoaDVYOiB%CCbl;FoE>f!4D~cTGlgulQHM*W! zWJ@V{6TrqsdNTo8xX8O70hh73e!|SF<>5ktY0yI{Onzs@#TIko-Z$St z=AlO1qc`5#RDQ05f0AI()--^VLZ)!C+9BNalocr(v@2;kf<}6Q7d!)Z zkCJe*)6)Jvug$9(UM1!$$SH~isAN96-*eKSkC{)RP`7Onnv>lhYitPq95>V>Wt~7U z4#z07Xo72eZ&aIzw(`hDWA-p3b<%}yo%_pu*KnmpL;xTik(Xi3YAqeC`jsVPWA2XB z+MEV)MB;@UovNwZN@My1d?085zg(i6&)#~UyV@yjTG2xI%Z6e;IiG#X z_CTm4aZy=1Qx4qUtriuK=wLM3FsTj`5j-l0U5fu8SGEAbGDD6iVBVFvmY7GfpRY{| z6xp={7o}C15@w!xG9=8pkq>p9-I8Ne7LS(|t6^RZmob<;@xk-a4!POsJ}?N=_6~0> zMzLnC zQq>`6BVGO0T7uR+r5aEN4-!61^Z;=|Z;SrEUguO=osE4clSQC&@+ZGqPbvaoJl)yX zA3)A7M%tEkXr4<)Qj50nL=}~13?$gQSmn;S3_yJ&Fkc%>Yl!t{3;j$DhrMp^1W|sL zyY|VpIUmG$7T`=8Ha*oE0Md097GW5{F_$Gm9oM=*j@B&k!voRmYH(m8`IROKF~?Hn zj)9f~NY#Nw6ng`=Uz?M`B$oy2FFxc(0QP75=JO+0y&I@4$qe0UJM)7C-%1*6o=Ks%BBDU=MbIL%y)}dK zQWQP+y&H2PI4fpLSIOgZ?3-sJMqX5JJ_{^d)qx%A+IiK=racs#^u8{c0!9yJrC;?IXtP*%TWQUwhiq(3d{u58{TiflmRm%dThgbGXz$l> z=l0*eH_p-9Fgf!5J~AWLkH&FY15gRFgMptO0{R}a+!0)afPXIdnJ^~7GViO1mKRLQ zht=5k;YNKoArP$x@xB)aUwubB(KnUGm=H#;hb_BQ`^>V|)6i)`thCc$exAo}2!M?M{d z0i`;jWJq@f4gaGe?U8oYWKADuejE60%d^%aPp5{!$GW z1wZHBXyrQiYPZ+4d9sBUmhRT88-ii0T=dOa;O&}Z_D8l+Pi*$A(s__+~13y=`8f*e{`Bc~TE~q*1icUx_uD}$>lrCR|YOdR$qKPxjbc6*NOiS!nK~1G>2Lz*PWm56n zGLB|4Bl8jL?tI(7PRnkvbinS^c%$fSqZb-`C$@N{6>&1&4JDa$R&jNl zau*Ww#%I)9HAIS(Nr+$W&B<^IA*Wp^nHULsme_AqjH++BnojHp%G+3f3c!ev}1dyv;O< zT9T=iQ@+F$k37&>7zTL*!2ov7cHO2=+5@@cYet3DrQ}yN#fv}4YH0?Fv?2P_{e@0L zax(5zXbXR4`0G>wZu0g+|<*5flG<;dxjoFDC7Nl206>Ws9;52~-m^*B zbkq|;)paIg9qq*Gz0SOCiRZ3uHBOz8OTfaQvMjXs%h#4$$Yb`!U_j;?vqI9 zDuI$yX3ufPqG;hh8aT-`XqLthD3xt8easpv5q7TocPje-grWSy`*^@UFq?VL56J!H z$ErwlZTCY165oa{f=o1ESS$G#)cUEzIS`}9MK2}g z{qG^GL1S;Rz{4Z`2RYwL=!F{SwMzF9|Jb&4uvo$0_+TxrHfMYao_!oVaD4a*bpnxf z`dEv48Gzl~rFd&d&DEa4SvmyD_gG-d=&yyNO~1BgW?N3BJJ0)}}Qgd$>0?uqZQHu1q_=-e%m}Zwj7x2(82lJX#iW$9I~b%L;PPh~AD!J?e>NW;k{C zvuJo;@5${rp2gz1j_qoSqg!YuM=ompO62GZ$2aV9@s!5!YqbGKvTO{nu4Yq`e~6AT!LmIKQBY_yqe^^dDUW2FWIr^g-gm{m zi~4%T?op_wdhbB#@UbWbin)1&6D zRg7+$mRHVBvA**x&Szi7#p~C{eD; zWs79}2$>S7E={53yuf|4#zNGlA(D%FI4##~!Lzm?h~k+RuuNUm;jP3EzX)H2)?-f{ zW0iZ(ZZ#VujP7knBIh|TeziRd!vb6#O16q+9G7ewqiNyIJ;&HGZFYV~d_}V(fWrM{ z_wY`@M;vkrF*sVEueBVDJVPkGv4i9Tu-J*llS_?tU0v$6D7sy$qGzrN?2_6gYts^M zsrIzV{njbfoIxK`CJ;w2`T9~P%c}0QO9AsJ;Fs0dxy)yd|5l&SOV^)SGvqameDk~6 z9U^~nY59oEpT;s+uSN*a%yx;-FGH=OohSCNlp5AQ`gjpq6+&GRL(eF(nlaP;G`O7w zpZ)8(uPGS6Gwi4%o72N@a^{2NfM3e!sf~QMmJ>IonQYRW_NSxlG>kHpbxW&xglg&9 zqdH)%$OkLXnh%f~;1*tUU_PucZis;nT)Sbd`&<2yM}U9|(yRU(#0bTo0=NMaj~5DI z?0meLDmfkiA<2jwppD2=qcTp})PCRaCLP#4r8D|Wgy6()AS9#1mdd)&UiRHx z_o?tJ(sl3rcI4%u6sM0Nz9KuUlSQDja{Y#CZSvO<4DLO^cH}SRyBrq?v(9{y66p+U z=UF57l&5g|+VUdli58vs>kr2&1lX7{`%`96*KLYd-MiUcd)7nSU9dRa(p6DQ2fI9M z_nnddU+cu|TWf2k$RX9+MF7SCK6lMx6lr*=>w?*y7C+*72J1j}+g-KP9o^o{f)u%X zo*4@Ph0ti6L^%fKeTx3X@SOo4wBr6Gwf?7NP9_I1VJ(fD5@LNS7G-po2>+{+SbEhHaxvt*)g!4GkE;!HQ<*Saj*E_$ zql`T-k;Iv6N6hfJB%I0eIi~iHEnc~H2)~@pBJ^u|nw5WY|BiwLDSX}vY+tIyNayLt zuRf4=@;Q`bJ_~qc)6rN@2pngVB(b~~97vd?Mx=8cZ>;@?Ok>yhhv-9}<&P`z|Cn0U z5vEorazvRqD(Ks%7VcWhJVEbg8e<#dKVrVehj#)lAzr>lDSb~)8%3<_{3ZMDoQZ}7M3yBHC@s?`n@L)CIkZ@lj( zMC8o12$;t5fttx81k~s~jjeAUf-nQeFGE)*p;>i^N~jaPMUMWBHvOj`sP=@W`2{%%(-H|nqLr{CN| z_I(w0Kw|ItCKs>zTTDbBJ1{ah*-P2g1(F|{oiOME+v|M}=o)c%hzTb2 z7sLkB5p+mno*Hc9{wdwh;4TNjXMEH`T5ISOp2tyLx0*3!Mn8}=;A6zb%;ySeM>1ID zqd`wyZa(Ro*X*iHA3uHa?B&=wZP&~PiaxH&Tyxy+xbTR3&wt+#EqUHK&cMAQI2^T^ zNN&x5HDO8vmb&)Bc_E3E|6_BLdVF_M@y%PG>yJ?%rL9Pgufp2qbFvyNwN$R=d{ZvC z&(Y-Kc{P}1WU1FRLs4LxziSm2fP1Cv!>t?RK`%@Hw!7NJPi6O0<0NRd_9;+^ZHa{b zP7Nv(P%(D7X+q6=AQKP5ERnu8*%cY7MNxEXB6!xR^wd<)Mv;pA?&(qs^Xy($E?6 zG7jw`M|}ra5tB^wF|~%#ME!(6g{Qn)OY(-Zcv&%UP9U{##X+fsJO`L=*Eq$#_D)JvUdWw1u)ER6=6DBh9aXFmQcv}VXpJC5>-AfnHCR5m1jC#+ zl}v+GC(nWZaLm3X7-oMYS&=LE(o3j1@(bPmSg^g0hS-3LDjAV6g01c)K>QSV2!1U< zpLsS@e4B#ia_>#POw>v}1t$O*@p!+eSt|HdK_E5vyFmGLxk5KKeT#ha1!QdefN&(U zj97Y!!&*0s?#rBj1f5>kitb1{!(wb>kN$l(DuabjIP^V_PC>I*36(1SUvQAR9Kx+` z>}{0@GKdEG8Z5T_;G#}5lRxRQVL8J5si#&NN+6O~QONh&!H!SqP^{dbJEaJ3}DyGRy4itM}TIoB8t?dzn6N``aTw zbi*Cq3c`LlMu}uZ(U*?BAHkvZVD&|9v{A@$(^8uD#y2bV2XhTs_9kz#GJ5L-e;_e~ za<_}Tz_zK@v5fgq?p8?!+j4i=j_<*JWbSw=SzOPC%O)s%zSghS+|N}W*6P|Aka+IUY7ZpT&0!)FZW3O zg61RYNv_M%#^TbS#J_Kb?EE1@&-Kd^QYJ;eLGqugMgC4*l~7odTq|;p?s{~tEm}P1 zgFLD_?|R&ko7O&T;J@hqNQh@`>#SXHHf4tFgqY8BMv4l1YjQ*pufK^JbmZ6RY+i?a zd__!?F2@QNrRfZXrwp6F)su3Yna0j_T`l>l{1ELNh()>*ZWo;lk3}_u0(}s`>2dYq zK0e5AUpsQz{CRlfD?@bTaxlHbk2C%|=?no2-c(6;P3tCPX(V<%vaD>9GsmgOUA@O> zz6Nmro?q-627S>xoJE);v5edHEdD&!z0lWd3T=SooqLmy=!G;1w6e+)r*CbOtN~rz zF?Vlh;0(CmO_w*jaY^L2Q7wTpLM2OMqoT!o36bvMdpL;DtCA^$FK=(#1GKtd58b^c z1nsEmFtFo?K6qis{rRF&M&r44xf&-lfw>FyBJFbv?3m8&YXaG3zKX$7Y@d&-?5Uqg zC=~OJCja}sp`pEE3QN{lW1O(iB{RBpuE|Qf;43vJZ10r;`x+GdGE4H*H1CRRUR zSHa#!flTpklD-qPRJx)WdD#ZH)5Td}F3joXnjCQ{40+rZ-=#P4utRmFiJa4#>0!Dd zQo+`#k{1!=B_{&^uA}4<3Sr-Pziopq`CTMLJ;Uv;{^#B=x@ZAc!c<~=Do|8*sn+x-xHOhf{q_>*eA0i*fe_t%l+J_>z z;q#W)2Q63=$9-k1cIg|if$dfCvs+{Ahc1>y>F}A#-x92kQr63GrB7v%JQSxM+FafojrQ;2?_oxUR6p$<@JoK75|1B=^!^!)-FsA}%Q97JQo8 zP68jo`C6KuEo2nFw`i)#=rPiGBKWlNV{aXn&WpuIEVSk3{BC_U2UOnyOPs;V=%39? z&zZG%FrB&brL|e$#T&SvoQo#pY&i0CsojJ<4yHk$1dN_FX z2>Vw><2UT@v^+FIHoO%#U`(i$iX|ewR!`BKSl)2wJ_AA))ekKj3Mh83IG(C`#NH;x z8cU&Dop8NoqH%vxA7v(lItEDT26hq_i++(ZE$yh(K>r9 z0BPF2#{8Fmf9ROR?RtK5ONnO2evAB%c)l0mY83rX3tJI6a|;t{_SX%<cmSUwDb%#Wk1iUQZ8)@fo!> zk=##XXQaxaMjcyB=PteugNmN->@=9*+gx9cRQs*Y^_M=B+WI$M6BzuDy0BL0(&4#7 z`}}`ya7t7b>kgP6PtmzkKt-n_)u-COMvtQ^LX{N{5Y_B)C>=23T`ArSVwiIk_%7rx zQ{D;xMK!@{9F3X!Lo`gVDQmztB#~#p|4>NX2@2^2OUTBro(8&{_~S}{6+VSDjzs*u zqII;p9NS2?Ak1Cf5XnSJN}+Vl`q}rN(VDy8a|{m!c=(;@c|w30dXUx_!9woeH;TLY zY;ESe{_o%ac?Jq?DW6V$+h9n<#J*u{b2b7RUEzA9QT zj`lQ)N1|GY+bMQ8u|Y74+~XfVe&itL^l{5FP!h?}fGFF{we*`n zSVOwhY;b>idz1-bZ57uVtG;6Ke}h3VlI%~w40xPQEc|+otWS5-#tb~kDK@7-ETJ)!r72v;bbiC2mPd5KM%>@fmej~{Uy-qgZE zsUtI{)0HjsQ4iBSLq`6c@mfXz#%ys%wc4COLtxM$D}HaI07hQeMg{taTt1IpZ(SGzH%0Z4gV@$hz6u_BE5TD<_fHEmY|fAxO3_@avbhfl7cctoZu!ovYoKluI5UMspD@B5`%c9phQeB&e!BW~sWQfH>!1`?0hZzZX(rEpdKS&CX@1InPI0#6@`un_F6(!Ok2#tg+HKi@7e0ca0f!S+ajY{W>JE0h6Q>2syAOGU>cTV-}9)(ls{VfL?_(cm3Z>7 zR4_1fZD7?c`%!LEN?t!u`R%&yikTveG%>tZC!4K;Q|cBQd083kSEGfSy~)YBx^T)s zw?{W&M#G+^vobEW;v$#bT&b4WiNAJWLUro%W4UkjuIoz2{taP(g$6{z@=cnVN2ocq zS03tOo#{9@or$#Xp{=c>cuV6k*lCz1zc4P}`0=GCJKD@^Q5A`;aQ6+BRjaz`Vwu&d z-Gn!+1^IRqz0G;;jUh;E>0wz94UOlDE! zBf_oRoUH$KkGXFhlZ=H#gaZowtTd3vhQBJaJZ10P&=cv>Mwo|cK0f!H?g^0xThJHO zPxfEe63OW$129BFkQm=EgwY1Q z9Hq%aUWsydyh@CPFrv7$j;zO6fyNgp_*+EDoLL~|%s7f!LF_w&S*Z617~fD=%aFlZ zMNlnO_>sqtL!Z~qax3+T7{Kk@_Z;Mil>ja;*l(|1;$0khD_ju2+#dat*s2{syn^-*Zr8Dg*&03J#Z*Cbvol@v^`@Lt#rCq8WbGT#H{eQkA_pdP2|yL5 zly#s?bphD1ss7lV)6sDz%P9Gjeebn!MTEl*bp^*-7pM-%VUwyy#qAc&$)svqyc*w^ z?L6YQAw}teSSjtX=Eu#X0F9_m=JURkY?oy-o@%^r_MAtbG`nmr)D@(4aEsDQjy-xU zbvS9|&?k?>2)0--q-OB~mW-6>tItOSyK$@cB_*E{_8r<{^Rqo!9ulT03bUkmcc<)gG+2J0hi z_}UWO%GSi!5YKedLldCBQ=v2i4m2KN@U9B9`Vixg1fS4Od|qPYJo^`)Ip+K&Bjk9_ zsBe`mU5!taxrLRp?Y0H^`*M_gGZ%_SJ-u>a&Q|$Fx}dAVRFa!A7K$+al$wwp*5l_* zK5XPS5M5_IJT;aay}x(sMR|+GbN6BuQ;a=JR1Wg%U;UxeC?+Ca^#mELwMlci+1RIv z#hRPhN|X+x`qV~>`*;V9fCcAIgYir43AxH^&W3qv(g_(;Ve;;PIxob!5Zx?H37=+>nXu(ZRdpp1pg20^WpU%X_)ij8@ z;w9LBf9>L1J`8F7DY4RO@_5OD;99a^$;x*)`mMV;@+&!Eu2el`PFy zZ%}U#kfY0&lM7Xd>4|O&QL8SL`zi5Sl`6B{JRm}PsipK5oDfj}iKdYE0ml`q_NuKR zIr@paWGD(R-h{4@GD}BU6cCuUugS(6Uwqo7(f=^~;33HkxC)SfZ;CT1c9kSGsyLF( zJGDx0FXtR~D4VU~{mgglihb-WS@vzMPbVPI)XcSPwCqFfw`u_m*LI6O24BB5P1!M> zevTe%PU^sWtcY}?uW-eBHkRU(a^v0cA$LUcG3t z8p-K^i}Wb<83ZjGMJZ&lk+_Pi>(a-2Co0 zfdj`(g#xUxJ>?Crd5EKjtte7=NMixAO=eS6Ih@|B(Ih#aozriG>>o`s>XIyrlBp;}9ti1ryAjM@_T4Pb^BEt6g>eE3 z7E!8b@q*EHEU{WCtn0T}cC_1zpYLFU}eL(=Dzv zE$@*rK&8f58QGPF*hg8pUgsS=_VU&7&8aO_ggn6fwz6*Q!d|LiaBiUNLHALk_&=m_ z$*JWJpe08BXUjmzx0<_{o_u9%uXXZt_jCk?C9Kh&U^>Q1=a4Hr%F8ZL` zx5QWn?qU-AT22O0Y10ICHKi3Ge;Q_mZ?)#+-I7)}wzr6xCAD{{2JgX!JnhB?zlxa! z^htT-5}-|jwgm%HK-Vrnqsf(*l??tf8E){0abyuBCq99HX!wi!=DNP*5yT>NG0Iw4 z+*19uidr`_pDrL83owI!>IH0#csQ9*4iD81G?a@(Ixv4fLjPWosSMfYI*S!h1{S-yQg}I_VAk_fF=VX1 zv`j{OLPiXUX#jT(KRJMN0G||oFc41f;IZsxSoJ)Tgd*5wW4)S^!HvH$H|e1_q1bw< zFZA{=3M-$|77{xhl+@H9Ts<`);z!Q2yCF9RO2xhwxuH`TSI{cKOVwK8;$>w#)Lw?X zJ`NKOi~ifsRV}%F%&A?Y@{qIY3-q(zUO)Lb(aycJogyt|nA5XUn}YVjWK2c|i>e1c zoPaBKcf)XuR10D=)ZC`kqQ+L|Z#7wVabiwY8Gn(E!mO zS)^tD*|5(T&$uJrmzo}mHDzgLFPOIx_G&K*m_z#r;GR=&8%TjwVX*f)Fg^AxN)~&wW~W3_L-`^$a+!PYwPB8 z%%J)RKuDtLr?7($i)&mdxN`CF}+3<<}XA$0{-oSCgEi$ElyBtzdSw8qMW zPqx{`MCgd71J4JVq)e^Mc%i(2uezNDIFQBP za3o{O-~#I!-Xx`bjN_oNUK{Z#I7~Xc?H%-0dW`2LdXy-9fAGH6Ok%0>T|VDlkB<9` zBa3+Zi@QzQq1SXr;LqV_p*B21uACDD9Bx>Ub)R;KiE|eUV&MxSk1(A zaeD;a?#w*Fa$q=Nw!=JPgzwObG$U&}lnu#A(r!9E)!@(*vslk3q3gPa^9XqbE5TYM z0m47VK#xrL9u}1|mYIEbXRV^{uN%_0`<|KraZq@S))`w_l$v6zQoZYIU*QbHB=X2b zp%W!L*9rgGqcsm!}mhV37LMP5;o(;;~;_ zp<2G0waN3~4S;T-@my$rA;`zTj^=5t%qnEwNmb?i3=tPwN4niTShX<>0YpAP0tbi` zQ~uX%j5;!LLZECYEBB&Ut+_8s=+^riT#@gPz#-b}yR8Q_2#vja#3AdB4|y!3C6FRl;|a7XCs+* zmf?K6hxE&m;k$V~T7VKizldj_V1l)IS%q)UT_0(UF|Us;;eTdu-Sd<2f38o^+9)oz znOUs(T7PRNF;>+GT-4ZMdQNY_cgHugYz)MK6IA8w0swB23H%-EU4LtwDzc#}AZGS< zSlyhpw3QL8>l5$o2NxnJrPeRv|5sOXrFgC3bWxwUW<8R*@KAW^x9IGg?2!H+?sL}^E$-n;X3z2TQ2Uh%X{qU&UM-icK{k~NDY zCTW~}Shl1L##*GI`ys5od`jP)PUm-~gEYlaH4uW6s$SheMn&o1g5z?A$d|s9`e%2U z&LmOyvGAq@Q4M3J#sV%X*|e(V3qtTjBZ~VnuD~0kahgj!HZG|CV1~!O@zGK{J?JdT z04Mz*jLHW{5G@jKSa^k}YLR#`iPi#^n!NON;erb9r;_ySFYgC!VuJgc(;=Hp%a#lc zSy_Ia!w&_a4+hR%L6BuFgjZbi@su$@ds2L+DR@wzd4vPYv6E_+ofoMK>&>tUx`YPg znM`Kg%i@$q$Z1}e330_IhOovA5~X<5Fkfdq116s&vW}NP7;%gpZ?(l{X3tf97%Cjm znn6hoQfwX=i9uHc0VV_BO>%%YMe#IcTBk*ktq+d%<8BAjOpaD7l!k>?Q`GYZ~X{XQfQihvlCd1Q0uy+wfbJXF!QIQ8X zrZy|!;(;IQQ14F5OeJYON9R)|Bn7_yq#6-BP;uVyJl!3m?52Rc5F z>54PwLZ~?WOIHn?!LclHFw?Or`VT(OB7pk3Ca(?lXn6@1io^XrgS+PSKY07$-NGn2 z>|a;D9Zzqb=_if629zNS-V-8Gyn#})iZl#hSp5Wd^w4LEe{&WRU~Pr^Jl~TSWo?cO9FnrwW=d)~#tCk>EN{a5gK<=rFF z()X*?{hHF>h+nl#M~1Imx-)rIULMc;l*Kqd5CY37k>;Yt$YT%B0*!Ny7nLP4i}%&YV3* zzs~qWS1_Z!0gEJ`X-XbnOQ|%mES2>dHF{&y8^A6$>K@1aT@FldsS#jOPZLk+h0t;{ zNZ2ySAhR^I(0zG3;WYUh^8{Z)|7R_Y2Hh#ZySsik8w%0YZ-pt!c$i96mzMp5fBr!i z6>fXBfgA_i)F6IW_O4b3ot@PlXi)vmA@OjF)smS&%z(^riZ75rYbk)!LaB?$(ho4a zNVrIzA@!+%(y72vr*g;N3#zFm45diPm#{(CbjGDARSf! z3pK?w0X5KU*}Gt6sZT#S@(mQ8X5cx>%nB(pU+G|!w%rZ?@Tul?>H5qc z%B%$dU9QK0D_tplxeYo#L^AQT=x}(J)eT^K&xgRO^H^T5 z;4vN=XYpEJ(%ZeDSVe=4)gpc)^nGq&gHVn~n$I$Y*v^)Y48_^qwQW0%6Iqenz&4-n zv){9JiKpn@V4uwA5$k+^voJ}ZWe>b-`>OY5Xeqt{!ngKqHG~t&aI^_&toL)z``V67 z5TbO=;&VAYYJX6`rh|_fLro7_G}3tXJFvhjg^!h9H_y$7m@?bjbmzw$BAcx~qC7ak zE0M%%Yo7Vv0+d)#<2Sfqm{J2g>cp(v7@;p=%(Ct)%JsIDWc0@7Ai zl#P5J%xL=P6>dXcF7h0}y|z_06jWM}NArojRPKlnXWEbwgzGN0tEr`#98sGk47oZ| zH@Cs_aLq|HS0w`T#UNjO2obpIaMQ(w%DnE}tgF14Ip8$(&}V0507X||F}nIlb2Jgq zvX`r*@fF`a~SG@E^Ly0kUm29 zmN#o2T4!rq^eFp3&vG4jGc#y;1AVQBu?p)ia&67!h)*6DMrh z3#m7=wgOqS8#<=rCZSrru+rS#k%&J32k+3*f@6*50rjplQhVPRB^J_C1x=^mhtYm% z`L5@VyAnreD&jEMj>qYvqQh$chEh?S5;o=_W5WRREYGeKt!XJ5 zqpr#vNvO&Db|3uPe%#4ww}%@FxA^g@+G&c6xwbdu!8cp4HAq`reJ#_%$87RqNn6d|>g}uDqp# z-mnOH)2Ia1{2-#-Bk&^6Ef=+@_HEW-(P}6KJxhQe!z?ngE^kBk612tPa<7n`p&ESK z62fL&hv1Kh8#z!CYWwTTuNB{dF0fNOf=>B9gDpe(&n*%y)3vtVN$${Xpa6RJ^FR=s z*@VjysiM2tdNNlpx|&8X@xliI9!T03zw-)3#G{GK{)rM7g^qP8e%8nPGgBbl!30Y` z#k5Z=!R%u%vUY0i=_5crr+xCJ8hO{HQapY4S3GK+)_Q$kn3u$SY53U5Jpg$n?@B}r zQYFNEgO2JX3*6+8_pHquGrHd+_cr5=fkp)8l3$SR)32fm*#4;*7obB#}tYd59PjQ98Z-!1NsY(4EEw&(-Cc)~(+(`nDHdVez(}qTgO$O(K>8RN&;w zJk2zysq681omPB9VoK~zmd>B=_tWd$@IFf?L*WeeJdXhL%Dqj9=(1rq}C^Etbu$n82kns%iSIJ%GJ1OJ)96h#0#v`8gn! zNryP%17u+D#OSsdlHxKQ&J+)L4hCrlvrwAIWMskaiY>Y~F^U~>CmFc258rrKB0XgK z>(mc$+g`k$S3d9L6U6oGWL&FpB`>`Rl1H#z8%ZF%`U zZn`&uK0}DI4IXAN6G9)W$Q6uMb*L=; zbL2H8nJm73*t}Q&%p+2;cSX3WilxelzX&1x(8B_kn9WR9oIx}duJ zX9M(->+G!>bfI`}0)VD-3Nq(W)E&~3*J_M=Jor*Ox>}&wlWeRJ!(Z{`$kCqWBMDdL z*5~%um96+dx4R|-%XJGi!$yByA+T+A)V%ywQyx|+ontD(5s^OUy zG+_BXIbpOTJd$ioO+5IW?dG)ye?}8Uo|JuG%#Oat*f8u%(z3imXPuOB_awGVHej$c zaW~Qu)EV<)`gL#mwHYA=SL6jXe}9;D=xVjL7(tKtrjB?nJ!TLWe>Cii3Mu`+c}@SN z%zEk?ZR2JlXe1GvKLKj}E~np1surl9v4D840|IBaG=p5Vk^ zIF0Y4$!`6w*<|$_-q(7M`0}$$zdmkZswPiVoHTzOCn@kv{gReg$WS+jeee2yiP^%l zw8ts!E3fSIhy9k%$eb#b4j5wlQy6?+Eyojl;$^8b2P@G?d}N@{XLHtCYamfxU4grc zvZAsdHkUn|wDzKnA%%N`(uP$TGeFUtp!5%-5=@@IT{!J<7NjIWS8~NYt*FQPjw~jr zS1kO@DkqWrV2{?rnrRyvlw6U0ySoHB;tCdSip@3IwQG{Ax%2v(*wJLe1AejAxFg;BBsa`(b-dEPRfIm#KGGEMOqC9 zhilX{;jzLpt2So|`1tq$wz@cLfm?xYuP$wznDf<(1c3lbe6m1=YXDoSx=7{(`mZa& zO5PWI7O)EN$0u6hh)M$dh#|-KfhPk)0b{`fIrS0zM#UuEPhnNN3m&dSj^Q}OGr=xp zKt+)(yj^s&qR6o@*}FgOM|gzmh+R9CGmWW8tcQ9dMexw z+f6hD_ZI%G})YDsDA_>c_&sVTF)CeEZt6CJ!XON}1F_Wwa7+d0@Mz)N#)MG-fdRL62Rurs;Op<%INT_$%RaGv zO;g@&GOjIaZ8H{5ia|^X05?pM2%hA4`;OM;Cz(mSW#4-3X_q{>c@f!zmlfFRqH*_% zn}se7C#g@Y2uAg~0{R11-b+)OC0vL5vz@=bgeAL`>8C0=2C$_x^VHL?P3Ra7OD(!H zB;i&|_H-90F327Ksn1Ulh6NdzLK}qgQ0dqE*;U}>^ z3*JXN@;@`jJ99svKZI#n`c5&Z({`3!%c<>L;V5bNKia+@pPnGLOA}|`x>M4W%1cS( zWbxQZ%P_003V9CiD^GyeqxzkROtn_lnpygk8QEobSrE<&~mzPuO>!k3xbbKzNJpX);0w0Nr_z)xPi9OP+DPn_MC&#Ot}ozx18IElq)QBOyn+MaL9?cehB zFWsVNTFWw;O}NnfF*T7NEOM(?8!Vtk0L|`2&9EIm$t_x^Pm8_fQSq2ZLNx!Hed$Nv zb2K=?|GD4|vmvb?pjOjbdrgD zn07O%>}`CM^D(miox!ga_@(TAWi;c~AF3~0Z;)PU#y(TVZN+B12}5w)-PN3T+aLq3|ZQ&3BvDj;LJfH-{N7D|3U-BD&UD*uY3+?~L zLXT576t1I7c3I~8Z0&o#+e}0lVz~Qeu28lISKX!^O20;78~RC*TjX7BniLyl@3!6i zRCcCu07?{S5+dnwGWC2-EYJ~?A0H+-kuo_6_m*@c7<*5H(kTT?j>-yShV%WyJ&^I& zvP$dA9uK|cqP!uo`0Cz%Wyi0gwX#Wto0zn=QVlBanpdNqnE5`By)AFo*D8O!?aPjRH7VP+@SB zp^L9(8+ISu(lIz%uQ;1c+^;tb%yX{E^I09u1)@#*#WieTVxzC>dP09- zn)vvbZ56;prorFrWnpEk1RXRfAnv=ta@8_m!KmCmc3U11b^*~Nv+S6yWbs!kKSAuG9WVLq&UWack3}$saIli;l$x|%$NRP>MeQ{^x2{B+x#?1dc7l= zH0We{;CKqlCmxUGTJrb|LeP2?BT6og^0Wd`} zim0)JRCzpAo}1pkt7i7gJRcm~qT5S>diitnoIYlkJ1eyewn zXk_){YAdd94a>JO*%;gHxI=X)O_HB%0x28dpK9uYXzjq~!VepUH(jIb%K4=X+|PIO zm8kWXVq_9U)H-6|Y>x#_X9Z8QHFH{gAIr1pdLyjqY8(vh4L?}x0jYQIaP@u3-y!)V zOXQWQh$$$HFxu{5#G~p>fz-@PfL(SRLWH+}?5`_!u$kMUmI-pycA|6$^3h6Ls8n)( z31riOo}OLIqg1|41?$KLNzdE5;t|^gSYGJFriU3on3)@1=io=er=O9i~n%NabNg;{Z6_m zjFw`H*vtHEIjhQ!+>qH*V}}q-iE>zh``v}H9KA)8kaz5#MfxvI7bV{e2IWgpKNn%j z@N%cLR0N26eHAUjv)1s>Fs4|CcE&q8!Zy6_!mH||t|^6lP5GJ2F!iF&%G z7Tg{Po+DGw_k&7081!i9)ZBbScfujq-(6f|Fk?R-ml;0YB!T3GvuL%?Dl*tlHv0g< z&t1+@Y_K+AeeD28a`A97B%3k9Vdvl`MuG-6`HJM0<}X#~vjuM8o^ z0^thEl=m*pgdjD~~S> z0J72@;3e5p!d;0{8_r0YDc13}o7{@vQ}=i#8CRV& z;X^9Js9ISgqc2mSyae<)?9XBLhlcG*X(do^Ql5T0pvL^hf*;J>%SkMH7pE0dzAL_o zpEejjyxyBzE=ul9d!78^mO{osCXa$aN>se?jfGEuD+Mn~w+cU`Cm>fVsXz?b>pE9! zJWN9!o=Do!UGJ+u2wwCN<&s}dl{zxHQ?IAD#+GLtpw30&%&d=xaQz)VV)N2=0`WYtFN$*%3Dg!TvXzM6|7rqm3YB zcel$eP%RjZq-g}M1l&4JvAwTYq4T4hi&y%bS1pnM;qX2tU6qDTA6>A1F z7L*x`0}16+nd?7GKz>eU?M}^$d#M?YA7hkG681MB9-_J*j9^RQ@^F%>_;dXJk=gxd ztpwtLL%+$m)X&JKD>`JVx^tnsU#6{@%@}*hc!fP$nGPR7qOi`LonK(|ud-X*t5Qid zT5qB8WVPT=Et|H>eX2wfgFruQ@#i=rF^NT_DJ$Fuzdj zK@>kHJM`B6Qyi0~CXwLBMgt{rGvDbX%zn!e%^*j!zt+8G$sze>jgEzC z{g5owe5>@rEMc&j6s-CJcW9eu#?~?$ZwUvDicS>m)l@i+Jmk|L$dnepDbcK!gjy>( z#<+AJgdfk3p@VqR2zOS>zZWv@>y5eMucmyLlP0mW;SK$zz5R1+np`Xl(~kn*VzUqB zWo3cc<9a*yLUA%&e9muU4Pe27GqLS!q zW~B6(6EoTfp8Bn7igJ2qy`%BV<~=AE=B zEo|;TjkdjJ@BPp54+oPx2|l5iFkSU?=ayAS99A+OOnJ|*?gIFNZ#83ktTvnVT#MFe z_3Nt4r*vdQ8}>!&N4M^&EKw{mX}7{>!WT1(d=(VS#B&SgB^@V5;gdxtQ9U=T`9s#r zM!R2)n6w(q?H|Lo^}Bq(Bs>MZdMt7(J>GsFxtNSZw4sf4$sla?Jf2^nKi@T*E1tPC zU&36Teoq&t7Z=QAqp3q5y@Kip@RSyz&l&H(WYoxbM1E=q##dED0@Mm}|HT+t6Af9kMe2y%vai_@}uKH}f#xegi9yRP;L=(k<@GHrnwK==kK z7+Hd7L&D?tO25f^+rA|m{6T*7qR*q8Ez(kMaL(d;w|95tdUFe2vF61qSU}knAmk$vVSX1;}Nyg}ZaLn}I;_ zYhYqUcat2})tkQ5o)$5I0hq+p&+`*SHxmU(|W0@{NF9UedW zjt{%svHG1KAvMBf z49po6JQhEjqI{~k4d~xOY|MUh{Gvz4V$~kun+c8X#5+J$Yj4cZL#{LuO8^9+`X@PpJ@z1*t$T z8_MS9dU^X0dP)g7W{D+XsTS^ET5N*))sR>?|B_shxf1vG|L!CwFYEtpTl-fhxmao= zDk-P)+StP)*FRh}2~l}g8>z_7pK(3`TW!g$vaDdcS$1X!3ACN4;&%Vte>A-Pht}2K z`N;F#y+!;AgAs`whiH$l>EGiPJm*?gY6%zqhDamJxU_5xfBx6ZniYV@@7wf$V>j>o z5cuUXaUrJv8_nrN^7v`t=`GTn$V$MJ7W}CyZi+uhczCJfi3Md+5>oqgX?EIFlBQJ9 zgP$vg&!+|>g)21ko_!KL@A~TslwQftLncg{!e-UVMLaLE`DR(kS3)+}OY?lH~xcn>C3paKp1;*fEDXb%|AG z<5MBc_QRq4rs_^Nb&Rk`bG<+2#^+Bx<^tSLwNjc@r}Jzt7Od_&rstlI(pBSsTBS4z z2vg27rh9AKT&npRb42dUqETg`qq0Ad+4ZF_UChZv9+XLI?hK?pe!RQNh0566#d^lv zj$Xvq6udj%g_J49j3y&%_ibE8YNeEWVX#H@YK(})KaY*O{qco(dIgVL#)Kf41|+Od zie$)|%95%H&EPqVJEhl!ZdYT!vZ}Pr5Q~_w?W}4o!b4itOH8pOpX;ABl2csH&496c z+FAR0OIkdc{f*@>t?MMewDms%U~LS^rpov0EbK!I;Nv0KqYY`Xue0c7G})ssNI%xb zT+(HLIeIeV@P9_I_NNFnffyt?w{42o$FY%FQV19LVdBR;ZH* zEe-*+tcA0wjG~2}GDgFUVD2nwr~{r_ic&ikYB|%=mpXTfwq{1c2JV6vgqhveamW`b zHN8&2vDnQ00MT~2O*c^(K6GFAk5%4x9 z6M2$G_Yrs1r1o8$uj;v!FB=PT#03W8Y4T%Dl#Snct7R8jjAKh|&KKKk>owP=Fp&0G z8pBHYkpZ@hbGRJL7HJ1T#$xD-tck?e>c{m*)3=OLCezS7{uLiy6uC919DZ$^h;Xd4 z%O{{(U#xF!9@PEk2+Z>fp@7FcMIz+=Mxf!Tr(RVBSRxK zxjwTaB0w^bOWuaUl6?)_f4Z4~K*@#xHa3ms;G_uT?_$>%M!fVOWfHDxqVDOnnRzB- zJ2pUac%<`iQ0>TqvsL(=@e+e!;up{-!lFZ&_)KjI#MC}7e*Td0rCM(o30L)B006~N zlP7M5#g(tJvlX-*ZqTt=72Jiu%fD8LeS7AxnjL!7T(U2Ca=DxBV_af~+|;Qle{?7} z-!mgT;N)^%4d`<+%&e!d@P{@SEuE(q!0v?YqCXBe)r&Hx6_-haD361LDAK1-BV0CX z{f3h!fLeG`L;yfz?|Dn<=K+J%&xIx)14DL8C3olh>GEJY$e&LSw+I#!_=7$2KW7N# zvuopiC2RDrt35S~)1+jP?7ZS#=Kps_vY@H_g&78F5MUQF)43J>pv zL`iKB7n)eP(k_E4LXz0%AAfBUj#$Cl7>8P(rW)7Xgce0>7-Vcs!t*z~)^3L3xa)QZ zg|mZfOz~&p)+VX2_D?4-3Ub?rk4+vCJ0pAfb;wqEc_ceB6iV;Vh{yx_X9;!yscL1} zRxSUegl$Z%QMCay>dnsrGgqp+I)eDI)-%eU`df#55(PkmfG9|i*09^1RMzLr{YZ?o zkU3CthLbarVFWxZ`cufLtZ3wXRGHc%<->eQ<;Brkz`f*t*TG(xmPxt4pPA(JUZ%CZ z;B@n2U+P&UACq)>f}P!a^uvOv{@E|Y>3;UO*+goAch_I1ysjip|8-??W~aEG-JRK4 zqb%;Yp9Y&RrThbzS=}pJ-fP9>vm$ngTq(>?`P{JVEFZ_wfMV@d8cmjzm(0dU3lp}3 zfGa}8AMbvz{nyt|(UVj&kl?Wg7FtL7%V}&$ORd)#zx1O{XlD?Q=@~dCfURV`89%q6 zRLKAa43d`Y;lq8~FYVI~T=wB>8D3$Km{pD!9W$>5WWMDn zzVBXkv>_H>0Do8U6_v7E+qHOZ>7~Yar8capq{|D^vG8;fGvEi$tkpjI16b>R7mkg`~G2iHEi#J<4hG6L%G@tf{Rp+ zJvfF_@30iw^|%SLCnQc-Qu5{xxA0hC!2`8iM0KsYXp{nWODlxSQnz`8HIbW{E#ap? zMv=i#$}#Bv-4${gzL6(Sw|*=9_f+k`M*!seEjM9uVyXE~$Sjw{x~P_Qpe=ROL>_yW zyjiDPVoVKa0g~s^HucoWJL8@E%u-33d*haFDD|UORrx+4^Py{{_^f*!)#_>x$9jGd zZ3dm2BADGdmBgrgFQx)z>h4n);&>0UTE=msz_nVZr*+)+;=J6R2q8Kryil9O$@Rpx z_xK89wm^ogXiOTO??HE3>FlkUY?L+Ki;QfgeK`q*cB$uK^LTBnUL+`0NRFXu&`1{H z!OSLf&6zPn(>nGw8mKU)V&{*gB!+;K%`xLnPDRIBh8auMJZkl9biKSslAWpV>$yN2UxX17im5qbH-bHL zJY6N{_wn}h1u~U(YQu~4Q2^NH;wH%Gn6j5%Ox$z2Nop$XQ|bWQKCW&uK26P6Gs?YP)@s%U=+&jf?J-C+Kzc51v9HpMEqX-bd0q~#Y0pCd_cB*j z_wdA|1@aBCTHOt2H77iiPj1n?jJb;D+71Mj;7jxhIFmtnS7}-=Fv({qRMXaXNCSul4EH$I zaI*uO)2-BWag*iA*m!#xWJX)WP{Jtyjpof`&-;)`2sk$5McEtLO*~L;O(8zbAF)_v zVxU2&WLZzPiD|3lV;kA{Z7Ly?9;l|`+rGI2v%1KzviO7UUaBn)RrB0CH{NCs zJx}?i4yQFe=CwlG>5m3qml5Xjx;G${Bn-RZsp^&y0YYwU-MJIcu-`@mdD*80xgL0G zi2|}v5;(SmnI_>fn<$ujO&O1-h}0=T%GwFfJ(1|73EY%qiP`IS{17BW`;)KxH%7%; zKv+G?0&B0_;TUWRU=1u;Ah*mmGih{^j%N`^Ifh#kk7NV-tM=IQfE~uqr>E@=8hCk& z2;%Ya5gQ9`k&3#5xKsq$$GT~~2=b#@Ola1e(Hg1S8t`aggm^R*)rSlZp!p`%r6}J@ zpQSPzCwHR(ir_D^u^a12R50kCj-~MNE&zd|u*B3xOGzmh-=QLhkfDDZOOj4DN@a8W z{l4oQ33~g)w`84`N;Y<#S|;yUf+mmSFmxGb@~~*8s7$Ybo1uz)0dGz^k4uzAKz0(@ zeLD}R-^%afvH&zFoGQs7KkeF{T2c!~M5Er1e3i7h{p>R=&$iNO!FIqUQ!6!sJn}(6#;`kUT(^aK3K7I&@@jrb zdc_?*{%5%5Kr6K0B*XHATqv^EdGSMb)D*Z+OsbnCMUd(@f|7K++^4$Ko6DF8AZ3?_ z^(TcHHg*~brxufdN!lkbMyhP=5)Fw-nuQIsW(@C2Gu#s$_si5F+`LCuw49-#RUm|S zi%l0m|6KU{SG!*SdQ=dGyTD@AH=Z)(((W2Vw3wKqM53!1zG9b2`A$Zw1~PNRLn2N6 zY4ueK(qNMomX`$LZ{vv>QZyu2M&Pmwh23~jD1oy`>IrFz}4Rr2|} z8zGrX&oc|vlCUUY`}YVo>Q1aetkYKfr;VbH2!b0_(3Y1t`0YZlU^u)VE}(y&x3^U; zUW93xysz$ACFggAwn16te_kY&|Gr3eD-Exd%&j6i)dp{w>Z0>_~oWtRQ*qcAv#uzFaAolv}EBWOgU9F7d&uGa|=?>X&3;ZfjTJ3N%&0dDD5aq#`qK>4w7K zD4wN-V^KHAuBr9DASf~N*rfU~7tY45&qb}qBsA*_DTdH2ly~YLVOMED&mhs0?1QZD zpDVKmwp{o9tS)Ra36wlGYz-n#E z-q2F&GKo(pp*O#S%iFeiIK~|x)nPe5R~YirWkH7fe!A=aBWQ!>3r90L4P0-b^pK>J zx(NMPso`fYM5;H_r%3fC?r&;Df*$yG=Jc~gSG{tDWp7AGah>b<5S8@}BIqJ>GJ}`u zyxB}t8Juq*oS*ZY#l)?bb;>mmn&tRJxM6#vJX>v}3e04s<h_V4ZMzxpmGva&qnnmmgtyAf6h>htzc(|BRe&2|kj zpk?F-S8c93HSr;7e^qwFm){j!^=}PodYiTT-(qD4Wd=d5$a6MVAiGfMnfSFNsWMxW zVM_hOE_{UyXHL^FSVzatR&U72eAQ5^FF&E@GOgL1O6-$>hx}m^@u9s2@9L_m!I4V! zm-*%XU!0BYk$!R zLLSA7<8B2p7+09M13HnZwS09xK%nQjC!TfLvEm#OM21feD7LFxf|_$ZROHmGfqcHy zF5RkYl}X}a`*~+|i7`M2)Q&nVt7qpnc_{yYj4M_+Lu!?cshr3fOg*&Yt6qa1=}49}#Oe}jH50XFWnIk8ZjeNL_#W!|3E;`DCE3k6k|U_Qm_ci- z&e-Qn(MT6g0Kz*G8)yl%ejhEuB^GUFfJ4$F9<8kcpXJLW7hqN|-pc(r3|rEQ&r^y- z;c{p3kzH#4v2QRJB$k=E!ALq4M6R~dFe}DHx6xweb8M&OgM@?q$dYzomx|wl!?W_l zj;FrWONf!}qdf%WMk1+Lj}QIbaBlL|pl8_=yc!e7P|@2NPbbC}*VJ^0!Ze1NGY&M{+p6dz4Qt-G*2 z#DD~)u;UnL|Go7h(V|3pL`mIAoGmp=w}TsnxN)9*vvlk7;o&T~7?p|}ZI%x0R1S+~ z6-%!firM*w8iR~Fw}+R-DKK8wi#~fJBD47&o6}J^u=M^*THyx?ldmj&{+=MVoABC+ z^gwm1Cfi0k2m`XvJUHH-8+JGt0PPzZpKm7k1_I9^9@cQ(>*swTH_={(?>@OoZsJvA zqUj>FgUTq(dNT2u-S?d;z#bAm#UWtN*L8cdw<=^Z1x|eHDQI>)COru7U(-|Hp&n|Y z^^QPcnRUmGMZ=-3qgm3?bl%IU5A{G{p63v@<8buJY>`01t` ziL6iDM_yb`Ws#=9ElGJwN1sA$VUB|&dqlN1p@~rAkoH|_B$u$lZx$kUJZRgV~-N&xAWkq<}8P$IQ;ut+=3Im%K?eP;4i@BO^w) z$JfLYBtYXO)c?kvhGq5%+GA%CPx7m zpKPk=tlos(1|a0cmF}6g5Zmk42+t{n@$-@pj$pZBnc5>WRUW*4(-EalOkAgTK3Yd1 z6V#5K0TRbm!Vfc8*{Viitc{#A9|77QOd@NzIl=ZCeJvyz(8Iv|((_Uuv%0NsW)f@# zJC!5Z7DQp$$`G1W{m*f>nvViok*K^c;Q?X(BO*`T?99owGk^HIvZJMD-8qA{jwJ!7 zK(w`26xbvimatjq(D7>44y$Q4*O*zy@{JXo?!D!2papq9$%`VVI%vTP`P;RJ%NP`| zS3ss~UJo@RW#LBN_*+XD47vU`d(smx&8J@C31fZfDF5YI7T4!Vh6mBNsh@R?Ox|fI zE2iblhftTz0GMA$l_D|kR?-STmIN~7D7Gd?umP?Jcxhfnf}@$Z3|ow!>NgAXf30TH z;{}S9ab1YO9bdUq_cGkrvWqrq3=NbQVg!Fzu-3nrU42H6{3;81&I7z2A5f71q|o}} z+I)ewic5P358Je-w~3h6HJF#0c&RKPH&&_iI0}H<;7S}leP)W3iC;0^kgj$5rbUP= z&7ByCTRhq9oFOTA9^dxk+X*zx95DNFelp9!D(}%MoJoM&^?7;7*zQ)<{p(Wm){il} zQ6iVOTK3Sn|Hp6tUbEU_C$gq3g5BI(ajA5DNz0|ts(507&7b|klOBa zK&d%5NPqGm(wXNgW=ys~D6mDI(ojYstI1#(qCd%7E3WPPmd*%uGs$VobWiHxzT0Dm zGH#~q#l=hcQY0akN-Z)DDsrp%v(uJnEsueh3+}3*#aOr5j1X$dOCKKWNZm>>gJ?@W zR7qfkdl@usBV!rFLvGexRlD4pHKgn0$rwSrB?jj8at|S7AyMBu<~jo)jYMEsE$F*X=I(z%K^yZ ze5jj*zCN_B)QM__K@Ex60LsziXV?f03~$(tL(F z=6KLTxu^KjFNQ|aZBI`PR)=4a|L)>kIQ=8~=p^RSC}lKMuQ4dm#G*QLV@%+9XjZ&!%a zHvD!5+3H$-kFyTi;9XQwwb4W;du*M`eu=Din@RH$WTtAShLkPy9S_PB26=kS7CquZ z-?Y~Xc8&ZRKYYzE18X8dyCOlf)TiTF#iuIIX-qbpi=1?E;n6F-8{0xg2A?!Ii0sZ( zo>YeWRyu;$^tsGyOv~Q@d||#4olms)5A5{9{g<65E1@{~e_VF)*G=30=H-(K@A+RWx^#|{G~hgTa$^X!AM;j4q47O>Tw;s9k51L z55_^~7B9Clh0?+HD;~9i*%h^M5C8o&S~JNhH^&8VkaSsN&h5FFZ zV?M%0Qbt`$fqo~i!^c9A)b7N43o9(&r_=iPddYd-dko^ybh{^yruysTE? z*ww}?Ys5OJ5ZNk1;V|C6qm?r?uLd=r8l)u;7Gefxx}cXHe*MS8W}aMM2&4k|&f648 zIPfb}njHgdaYzIZA1bQ1{w{AjVnAFxJN#fB97*)rtW9><(YRs~pu`3$*lO*7=OVnY zj1PQB`m@#a`Jw67%$J^*3?q;<*p-*rKlbLCT1%(eZhKqOx;{BXe;YBOAkWMWqFwlG z)|qEB*oo2tAUko-pjJX?=YJRi-}vpHntl%hh}jGL3P|u>zQXYP>zILJO(+_lgM|AD zu=)-;f?yI1My$6h_W=w?tf5nFH`>A~L?RfyAhTQ*_9)hDVAisE#HBVo>B7@RU`ae> zND`yapRWGi^`QqJC){i6{lmdYr8tAUuW`4}G_wtVexNA+SCvXq=Z!4pxxFLTYk>3V zax6mq2&_`71oU}lDg0M?;9pcut~0Z37#5|0U54+oC}LLK>g4m$Qukn)*Eb%P-6WT} z(5lAFu*oY(o+uAVq(?pXL$kW49Fedh`wx{0jIib)!A{c{{)_%*h7@T4w0%ym=9^+n z+EJFL7(fUh>cE~5vqGC1S0_FI`>8TjS6}hkZ>W!A5z>0zX=wy?OGVc91tGasS3v%&*e#Jik?CC1o$!l4#{nzJGqr zwDOjX`+Eof zl-*aVupW@F?|SS53;_!GExrhKYMtvre40MwZZwu3dAp8#)r2*FS1F|@RdtYFwgT!v zPv@!gvFlM<2I394Th*aB8HBhX^K4KpRuI)XA_V$)8k@>BjOTf7B=QyIxZmZhtC5it zl|fjoU*xhIy^~{8YiR#*efh(kWPEgQw`oyw9O3k}>{F^DTS-PE_nqehBVMn5YiQWe ziIUEph*s_mSio$-rFZSC6?&$Fh49hAuBj_Go?Uk|lfup}dV;e35?^AtJ|PT-f8FZk z9dERFr*Xn_y1B4i6T$VVnV%+1?kL(_?ELh6hPUP3lyXSjUOIMu6NcvwtGm#ZxoeK8 z!Fl*12f+g??rU>~in~LIQ)j{&lytxH9pf=#z382hIRC)=OZbhMG1AuAKU-^Ggr|iv zC%$x}NnTW~WDq{t1k72f8YtY%mT5hZKKF0}|Cy{jO;h`>u*ys2&hUfSHd0f8A@&^J z*Tq&m^wz?$b8%OR9XRH0(PRTEB~`jF$qFH@P=VpWjoyKxIWXpz+UYGWAz8Mma|~i? ztyc&hxR{;+Xg(mr0WiESD$)_^!J74_psbu*V>WZ#54(xvgPA)aeJn)!>RY8}KKrIK zb=Z#EGmU?k1;e7PzE4@HG{^+EmpGU8K5g4@=IM@Nt+z?9-x?ml{I(u$FLQ_V4gr)8 zq|GV}GXcO7jrmu1RC=2Un4Qkrs+w@>uOjwbe{Q&gPbRl>(%{3~{{ zi^H@W{tDPi?U;;jyTXrsmo=QZKzCh5{-?NHx#siDBEA~jdZmwaC9A_?9kwW*QGPGs z*jo=w{3)M)X;VICwCgwtDNN+u2uOh1gWGkpiFPJUKrbl)?pOPldWD5B`!h@5W*>1F zU`yDP-qy#$HhvQJ9$LDeyo7xCYLJ2*L171Myl7TXHtOPEV4}R6OImDwNF&+F!GKu1 z=Ykn1;r6xe1`)|rqw*!s2*43cFa^(ShBXe-8%FWzHYe}Mzn<-+7+{usLivTi3*BQe zD;kt!ev2QndMUfS_+xrv^xN?Mq=GY$UQa_4>Wxy|9Kioy2bld!K+h^?LL*K0_OP*h z)d@&w4>f$9FL|(##ZNKcCjrB7GR-5p033Sy@h2+o`L#o`CG(lM@9iQ zxz(Vv*xf7kAKiDbOz+$7(A=9`AA05iqQp}1njZX&NAquifeT2Hz5%W&JL9b%ob!yh zz2~&Lc!I9!tf=$j(liPGHtpDu237f6uNo(;!8|VdeAV{T>7bM3j`*QJ z5zo#AFW_^8ypK);eyo63ZI6TBgqd`_P+!A&D=7Wkwbr=BdL5?Cih5W2I6X75^j5a~ z)~VTEKLxhXSx-&u7<8AlgMGE0*|Qw>1X7+ zsB>Yv#{GBzx~7FT=*(2e)Cc|fLq1CC;eCMDNUEUfoyPoGMSVj&y!7L#hC#o-Ah0VQ zGHOy*Sb{Y`52N0q;DVw|D7eUEhQqUh;BR&i{k*-F^yRE6qsB~UT8r%WSJKS0%2p2@ z?`1AQVHX9KdmATaEn!NHE?-8qrjn;&CDG9~Ar^JB6*8ZG=+^)b8@nY{G70IlS3>Le z)+<#?-0M&M3dV|qub*}gnL0Wg3F}7#qBBeMeboSu0K(6rohp}_E=v7VgTQsQ+bRbk zP?uG+D?fFtChY8wP8*4D-y0|ax^^p}q{(Xcn(E7;5xf5hq7hvL&DM=^LfSCDb+Sf& zLgnvD+i)V!c$uSMP0i%!+hfB0b7OX9U6e=+lh^1Qoa|^}C25a2GhufL>om4ZvBg5P zI2+M!@#bxfr~t_#MZ8}=?BS4yuO7qzF)3O3+^K3KE#4sfr!r^Z;@N2d~+Ps*6u}?&U1dPs}=bOr_0vn757@DbP*?&OKM_Pufugxr(wr>u-{2N zyF{;shfyKO8d_Rwl#Ztlhmt~t9f&CN)3{jtDxn;HNi+YY)f6F%vXzW3E-i)csU;czty{K}!@l=}el>(cVv-RE%GCHhaN#VfK%sp~+ zgQDDB&`8hZmiB5Q@1Jg!=x}Yw#DT`C%Sn@NQv;#j%r@()Zt?dhu$$dN5}z6t2||sv zcX8h;n7-F|zWqFzv$R)HIMD$QZ71!>XX6KLv+so49sm?9oa_hWa`m5=x~m37k4Kg* zRxNI{RYb=!?KKo15OPUb1%qhDLPA9aF6zC~JRPRheRdiK*vI3q zlMUw52kSR%6Wxg>-Hr~vumRL;Nk*r zM{B*VJZO%@3_(IpR}c>A^u08!Z6?6(8TZ75Z9TIaIEo5?C98Z@d-xnz)(i(sQ(?Pe zll$YNrT$K+MnY2=$20dY=XAqIn3)qSc>7um#XKz>4mZCt4u`PjPRi+PF!=;;KaJsi zFf40^qYV|m5*J98$`~SJ)idKB@q3`90cTVlA89^#Ua_>>YGf|VhMJU};;fttygbIE zU!qbA_fx9#?JQypq0^Zk$I=8Hn?$AB0t4^HJ&NuRqdJWf*7sW^n2^a8KEw_EP(Ao`{vVb}5+`2p`OIn;kzLc^#CL2>g5Fk_G+Ju2N zbzrRYBY{j*VXZ|B@wSG>6M#{_uL`T9hpCoTC@@LHN6r4X+Czvr1-3^medRes1!ckSh3jZ6rN<0~dtm3sq!0;BXE8xOP;3pwmL9LAu9 z9~k;b+Ln6zi9v zdomnQ6W)PB3EdHsb@n)!N32T{Sp&|_&io9v%$0uoo0X|Y+X^aT9uI0m=^xiR)tJXI zH&j@6*^XpAjAAD+@jaGT>k85zVn$U%GNLgn2>H0KoEjBo4LP`T5s>m!s1Ye=ls~RL zP8=%du5_6t%~1iz4Vqnl#%|lkGWxvC4I24;Q=XO9X7m}NF*iMVXS|_7qja#>S64>! z)`mmlq>YBfQ5?6Sqhg#7&R_FM2`M8L-jHLYj3>rFBu)pivD>(6S8;snTZ1s)(HyDJ zilVQ0O?I~z;xV<28DG%l8#4Gj@-Wj8k{ZcEBJL$uS@Yx%ooy3rjC9C)V5kyRT~(S7 zd)brqzo?`dR4PSfsW?@Fx^k~sq^d6eA6M^92*b^%QM#-GUzmu*4400~1l>H!fKU=pZsS}T#d5tZbBQPOr3RSHs1c|nALEG81jb1WBHVLm`j)0eoC*E zf6v%6hvkd4Cka~@4hEwsD%`G3hoL9>QTT#J?@V?5h6OA0@fK-=_%|xjSU9!II=5l3 zVb4YfuZyOhC#XyBI?))uaXI2w2l|;)+S$x*5zU|zTlX8S;C;Sk&4G5WQZKV5oUd;# zQ1T&_hJXS_R#!~aT190#PLUwUyQn0@si2}QrAHLN&j*P)7JuNrNt8-hA?BSV2ph#J zIDc|ZWMqs0$H(G69Ohniy*xK-_<&j}29lsYs(niFVAT^fi#t9E=?Tn@R3kKdVJb%}=BSh&KtOAzQ4JiWUQotm|ow>BBUqiBr!*RUwmwFw6=fVv8 zvb-`EHK&a$LzddKWqk}FNKqa+|M=GpLNym$+WgQ zs|7rqelc58q(7T~x+4NmVKImmEOw)ym$Y^drg7hY@UyeLbl4cWdr{HL2V@49h1mbGvlJG?u zFA|)3`zwI``XWn%`3XpU^Q`gZltmtVQshs|Mk9c0=!Z6xl^1Utdln1H%HEy>?9KdA z>1n`tGRyQ$j%?;(b=*|8%z`)M`l#KNwPUD5an}|YQ`2MP9{(FTO-ZYm4^v=+ImM*c zrw6WyzO+*9=AU`XpfO(XYbZA(Ye)50Itz${15dLgNKzQzMaUMWw1=v5VQrP>VD;eq2u-+V?F)tv&QO*p)xZwxSodkJLL zv;)P!>47)eff@lw?=MY$Yi2bzWY%-F9l1a6${OSruW!>Iyp3S*Hydee%Qu^kdZ+#| zUM}*nzyQ}zs>8gXo3G*^k-OXA$IOGC>jcy_J4Z!#-c*7spjVN|j{er0oc!*Di3z@R zqZf!_OUqeX2)AIT)PA={k=%1bJC!Vj(C*YM*u4Ik;Lml zP3t9_Pxr?j*6arSIZ?41pV4#8`y%PlzT10HRdTE%XGs!<#+I%r4!}Wo?KM$j)3D&c z7AO09m_d(8=1Eu{@ty+wcNA- z?Qf1HZy&XbI7EHSDtz@fB}GX%$3R}V3;Y`LL~_NevtzVaqlf$S9~+$SwI^CCpwgj@ zJNXZe(=M}JWrGW|_#)@r32AyNcw;R2|G=y8%}tf?LTb)$VYb@sB+C$_?~2%`VE zb)XfIP^+L!2+jKEKD28@cS+47^VR>i>3lbt>uzxh1%N}`o@Xsc>GkbA_5W{Hgyr9? zh$dz-(2)|$!BloHlUi2o;cWY5rZ=XP>y)1pE2h0r$a&!kesO<9vEiyH+<7>)bQ zvi+{OXqex(`()8&%0_^vPNkGaMb)s2mYHySTwQL(*D`m(m@zM(O{EAlBg%GzEl*lWqG zvinD5i(W6QKayO_p2fsok}uB{ZW-AdUabx3CsNqM>-6!Ls z{cb&;EoAc@^I`$1ju#+uk&M71EHV47C-O?)Ym)6M+aGwC!ehkC|bJNqi3>;MPI>=nyz<`a|IqY>SM9@u z`JAs+Y?hDyubM4-~37e1}VWcUynu^%sbN5kL>Z`(VJA~UJ_;5$baDTo+_Juz; zePMGgcbvaeX__ToxI(Bo*5ac^*5y}zG3cwd>$j8ZkVwsuTD>n%RN}v5hd1WcArs!U#N^<%*XqnNIt+G^XIm!Mha`5Iwr{7 zJaDr_a9f9XiPMAVo=}!>$nRMfI#~FZ05Z z=fTr*jKe#4>W$mLu(=s)lCM<7Cf_2IF%4u|KS!)FuOO+>4T7_1I^ETLH^>s7qs8Y+ zus1Qu&UxZV=f-93!q{l&l}Rq=w=a0CIO{6?Yaz5fi(r55c`N9=yP4F(6jnVScE)w5 zI0;h?E60J(V!~VX>`fhquMI?Au*GtjpkC_g307gYE@9#tbNLk+FA9K|L=@|BJjh?4 zKxX9qo~^jI-k%c1p9}~$E?;P}5j8|&2i))LIxbo1$=yvl_ilXlj`CN&9|q=|3SA*q zJe$zqD*t5A)bL_!Y`Id3N811Zuq%$9e^iZXMS)Sna5|0*gfP_P`h#BvpG0*Hw0-O+ zZ&Yo`wVIa89tS*DiMKC1J~A&?FmPGaO78|RUAE`oFsU(g>*2}51C+kvhPxitM*7(k zIU7ySEDpDh4vzw+kQw|0`s^Xf+HvrQEZ)D3G3{cgIz96$Lp=Mb@BD^BVEbQH09?id zm?cE3saq}kLGW=zO;i;9N{Ga*bBWiexQ0OF(oKuCil8AH6KE`cBO`uA$`Z>J;DYhp3)fL zCJr;9-MBVjWyBlX=cuaBY7z^r-Qk(3luQokuQao`%MeA@xAF3lt!_z|XS_b|=~1QN z+=n`;3yZSrSsvKnJ+cGqhP(RqXT><@@9Q9}Me&tpUU)|3JMO;(xTz{Eg(bJ{wbD5F z2Plau1=MYvZ3q-M@P9bkZ^i(7ewWaYx&(`Kpe)@fjq%wNxS`#J=I2(gB@%8KYTJ5y zxpP@`4jsllD0T0?>p`Q#eVZlMhA|t;W7J{Olj60c7#}I2eqCIEvk6WdP2yc90nUzJjDONroqjwkK^EO&q=2lK5b7m--zz{8+4|TeJP4eOPN7%I>kq z1wrC zGQCQM!XK~|#GgR?Ox7A<{ZumkwxV*V#E^S;dbh*vdCf#V4`^JgXbs8#2D$i$$Kf`c zN0m0)W!s)I>c<#=OX!YNjFSi)G73~mf1V2y=b6Jhq?dp7)U!c^a7^*=wFYoU`Uiy; zdR4KlPDYh7!!hP42Zy>MVgIOWTD>Z_k5ER@mC3r^QdPb60LD*+0IQH?7cJ9F+gXz} z(ezx6wo*jFy+#a9rSOWb7i7gvjw!>{XafFr)2$J_>xT)d{wr) zfMz`3<4VENoW&my53@8r#lB4mg+Vr`-(x()hrL7ov)rMbDYQUd2FD=0ag4CG7^SI7 z6OOq_KIF5TNv_5zHlu;VcJm{#!nIn}VP;J^^gCNPmQ~QLeCCpKGiZ2ojYq2`MV-mD zV}`B>`K`c%u8sVrJ{W3=`>sjlKu%-IdA7tU?Xg~0S=7*x&{J>You**a4igl96XBvf zo06Ir;`LnG(2q71=RW(zT6^^TI|U>&$9mCZ_JjIK+OOhxAyH^TKusj@2|KaXPJifQ z38#&;g!(OS^qRm@HCNBwLNrD0FXz|v`}v{9BgZALnmZ{a2-l^GM^6HxWXC zCGE@|lEJ`4Nl8zoR&NuQKUsVYp7cVnWrU29YX#LNb9UBodpSDFgXAY-ycE`thx`6i zT@!oy#yX_L$nF;g^;TZ<7bduhm8F;m45?$~k~LuiN6w;mrWp*qeM332>p!o~YfO3- zwNrbe?ekq(yBt<;nHe$d@7w6*S3@xw&5MUHQKbT95qU;vbG2rL#6bP$-D#bwj|%k6 z2K=$eue!Mb1olu~O$2&IF_5ayV$TR56w_w!&RWuv4hE=#C(66J_oin%k~*;Q~amsnIS8qlbC#ZBydl& z?S4K`+fk{DJE1&%^)8sR3{2R3Kic7>(k$$E3z5bqk#^NP;aNgMWH*Xnkp}6fY8@?f zY0GP4@dovMJvDTo)uX-4N{ZF1vEGb?3E=9=jl6EoRR+LcI7JB;htiw|{{O2O08bqL zNC#(UD1WIDEdRaass1~NsYkd6+N7J!+`EQo_v9BsY*$+O8xHr>muVOjq!l*4Zfyi( z%%$XmcB@=$lSL#i-OB0D`V7x_r<4&|Xrpp6U9vb>3t7a&*)CU2;H(8}b- zT_5u}t6ZJ;%FJ!Ryufmr*D8urMX#W~L7vM%7MTTwCwRol6!E;}i!(DaDICoW4ESPt zQ_BGn(l&0yLDsJCj!xmN@;x=Bl!j`BYZ*WtsT*~nH~kr%u)JZ^-uDZ=h9Ke!$`c4~ z7!&XFQiq$b8!zoA?2yb+#wFH+btrw#!)ehcpK*yY#G81drDAZv<7fGsw)Sns20h4m zP16gkZitaxSs?ywx9a*4FYrs6s9uUx={pZtQQ+a6*|#rGRg)xCo9ipG<~^088J;9A zo@wiQ!R3ZUnKTl`KUZ#oS}1FhU(eN;Mzx?u*C#!uPReiEYG1MRb>OPF#Gqz~X0?K- z{5d-IJtuWy6n_cyFef8lZ;8N+d2E+_Zil1|stPlWy%KxhuyR44)RJ zN{)STLQ=dqqaFmw2vDJ`txHSLOy*?aVz^u~2qLsE zi6Rt_o_hEcI6~nz!nCK@ZL@+aRwYjN76WdkYwFQ*4(Ih=VGvTR`~?2-r6qVe*rRfT z_jbNbscdlHE5yB=5&o$5Sg{(AB>Q{zS08UdrZxt14y|*z+7w}&j|Yalg+cS+`&W~C z1s0MCv;QTJ5k0o~jWmDo>9MOIsOAr=A@b=*D^@4?OhSXh3U<}LyJ6<4c4OAD6hJA5 zuviRF-o?!;nsYlG@;+_7)Vrk<;4x9TyDmh=+h7%)Jw*f-#L$)BHt1nB6;h^ey(4mZ z-Wa{s^BtkJ?u>F(Wge*8Y@gl9^mkH__=}1)xSp4DpvhHcj@T@&H6R(ge*GJONvYOt zAYbfSs&}BJX{rlLOZwMsd%F3Z$#K3r`%k34442%Ww3k#S0|t4#NC(&CohE;9aA%#% zCb^9N^1lkAslGC1XU+47lxX!bPUmrI;paoTNCAg=ElplG-&TXplJ+Za?G1zOgzUWB z5&W5~TqY{prLVUsH;lV+UskJTD0R1=4w~s7I!7Twr)Ev?9%BVl5$th-Ae%2Z!@h~m z%7z~XB5cDl70|Q?iLQn-9nkb(?-6KfOA_GWC*fd-Q*@c~F0i^>{rbfpM}0p zUo)zn735&i^tGf?Jx6_qo_a8D8$LYzN+4$PZbA}_&hLp6IzQKdG;ZsqCy#E0kbke0 zZ(LBOyWs|J?qQ$!SR-k&wNji0{W z!&>HRomJln<-Tg44=fYZk1qHbVMz!j+`gYBw=uUk_Y?geEZ8^o5rT6QsxL^5`wf*( zL%PmXp$p#OZ-X0)^@Bj^0)wQ6b%E&`@Fz5&kosFS@EbTM+qnzVkon2sosAR|X|pn| zQCge*hcKs4Mj6eSXf{fT!$y6B6EBh`Xeeb?d%nDrT5%;TFIo>3g(=sp3Jy|OkwD!L zR43l(F+jcksh{A5sg z&TOi%b!rjV{N?EN1t@E&&4z}YSDj{+FW6XqpnSaCJ{Wkjyt|Rd#A}gsVisCkGvu~J zG%1NLElS8_mKkt2XkT}R$XA>6)=rT+}WR5n=|#4rt5-243$skyfz zh}8zqN}@CBUQVu3zoW4eFgL08x8+51$vT3v1QqWfX{T8l>Q71seTyo-ZcIpTY>@R> zOG=N=oL7t?Wr^n7d~fMha4x9K&`HfyI!bhZXcC&}U2sJA1t|L%)$x_o>iNT2$>iA2xV zO6EH}-7#X}IMo*)J$_N$wpzz9aDXmESkn3mi!8OBvg#ddjBoiby5hY=O|h>^C)nO% z;L~KLRaUISeySPmrq_xaucaW~%v^JYnFcOqB+y`U9bkprG44KBT7M*X1>;|X-Okn zR;{z5Rt9=aN2OP}zOFpgFq2!6<1a8V=451`Q#c?cU+nwoHFPlgJYd;XM1>bmEg;yK zRj$shQ2yFJ%Lx#+w2_NxE*=d{oOT;>)rnjFUT$tu-WkufLzY%ODj5oaRQbB|lE&Cn zfj&{db-?v&sEkl3*=jiYMwx#CV)$)KmPKK?T`HdNvR^~LCV0W1AJs^{l*@oKTnX)L zj3F@>%ieZBQ!RV#7mU>PuaU#LmLccXg>Q(x}k$nRI=L@!7=LhT07vKn0 zxv{Di0;<`t!zOpW$`wm(nY|bZ36^ADbwk}yF;4FLR^IK+!+#tub#KZQ-pQ@z>Hk^A zi%oy+&`Fa{H~X4dtC}m(IS}O@JHd zHunz=qEP6&EaLWsS(tI-frc7Ao-u*Y=ufE|4K-FN3CQNvmiVX%YGz2O*uMQjEvdtZ z65W+uewF$~P^nlw-qFY0&W3`{O5LBS0l+?=MQen>gyRhT1fS0~k0xGLf&JcIG+;l7 zy?ui@{J_N9tV|AODKjFGn^H|8;TFY;jk^PIU`U01QwVZqRWYi5tb@pZj(zZ1dTGz0 z<g|ZvU%s`FJ-hN%O7X=kJ8Cya&rM?`+Kk;_$MY-Ypq1G2$rkERvQU?T zYgkZQtzM7pGY?H|Ox@4dB1Znm6OOp6Tsbr(NtoDC2z(f{OX#_LP0kdh>Op@l`Bv#ml@5e$i$d)1dU- zy#8_q?*!)TIt9OHg(_fZt`)dLe>&YteCQQ2)%wtLQ+C~GJMfEO5f$L(*G!I7a~*&G znw1~B29hp+&hSHRha;c_atHWh{-l>Yh%hqK+4wCyoRl2flId0|%%2%;?KUWRUG=1G zdM~%}a1AoWJ6gu*Faj>f#%iWEWKOZ~-@!!8fRlL}od)1pU%hxfELtUaifDahZ_2$n zBMcWh6E{D0Sz>`=THnIXT&?G9{A|rtYepNmH1bid1Zs`5`N1gJ`n4cBwkXnU$0Ppc zmWnstT04tAl{JCi3@YnvMr4Z!Umpm2AcZ`aGxTy!bD6Lm%w|l0cl{lQ`DWD9Kw=ID z^3+z$S8$uox1|;_ncuRCF~`4Nkn_RFX|Y5rlnw{-ww*U8DQz@qsi|EbjxZVtAABY6 z_AX}|YGW~4RNK7!JiBYj6h4ShmJ0sEf<2boZ$e+PMzupgniw8Bxz_unL^H0R$VPdr zN|0QQu38W{i|65dO>3-_4e11Mnm@RQrB<%a$a(HQ<;6_NXB@5N@CwTwGIeg74Jt{e zT>Vd1`=Ty5PwQu+l>9$*>>o`1B=>(l5R2ffWC3kkX`F9kkDFyZ-J#SN*fcUsSs_6|8(J zv=g0(rP}IKCIz?I>-b9kE9@!Y%zFAf8G%vM0!OEMxD$SC5!)bQZ)qacjHRxh$`5b_l2j2N(KBvXLg<6s6U*m7`~3$T~vXP>d(lV4lGriIBPnPCSaF z!x|T=p9~L#bFZk4bt*PvuAXhlaH&gGx~SQ>>!kXZf3;I4z4nE(JNx#{!&6@ln^RW_ zvAlN!Sch%#i0ju6dC+~l8$1&`ZQkM(R&Y+vPe}=hka$}i(8yvw$qmhXYgK>6&a3}z ztYG4jYiJhPI~+q5yzX+=uH&;E5px#4SiY=g0`7alQBd(NWK{LbAQ>rPx=)7~8UK>9 z{=V6bvy^kwqEat?J+*`-GA$P(`b3W#IZI}2;;&|J;BO>H`f+n54eP>=UeT{ZT%XE+ z@AMkX`YqV{xphyZ#Ri;a-2MjyXE*IKXpN4GT>$E)WR2I+d%fr z1-)1MNWya3glTa!ORMDe!sC8;sP2%CU=@GQCv$8Sc^af-D3eX!W8?4_RcI=!s10+5O=|NjNGI(?C+@_*s7B>pih@W8hYjwWcRjt3 zVu)dKrRFIZv@I(2N>4h%d-@;SF1UWA4b;Y`xqgU~sR>cGyU$yYkA%LCkZ7I5walFs zA6Z1$f5tVjBn#vqjIJKdn}=zSR{HYJ@Yf%>eM2Cx9$xjFF+7OvnvS~hO(Hbs@bwIP z{gF3cSo!qzqB1ec;0;fyAsnc{IPSZtiF|j|HKC_ zar>1YX7`{tD+BK(A;ykP>o^h8rW!}|9k$7rI2Rg|mqUsO@C>Xsqtv1M|D({wLvt*$%q7q;UJocZu6`F1Z3Fff( zg`DB(t%8v8vLbS1>BQ#h^G)3c$S)7$86E?t`h$}dkE$4&N#NPiermlGoTTvnB%!oe z)TWY~eY5szYbA2mw&-$68ksPtIoy&Bw`zmE^w9>+!LrB{q*>+sB@=SA43l(Fz2END zZ>`dVQXtgU;Col2F5NPYa~1_bn#mJt5}R^zFTx|& zt4RKv+5%brIcLzg*#nig`I=aZ^L#*J+Z^F8a)se+q9%M}GX9pWt*NofwRL@hZFh?# zY6vQeUWNubmp^jv1#@!A&^#(P010B2vcb>=QYHpz;V{RsG=5QarG?Td@ z${v584m|xM?gDT7?FCH914qo>=kFg&<*ZH%?tGW%cd5 zndpFw96>brGwL(!ayL2QzEw#2L0T@-=beVKRqB%TSm3#`j;T2#vei{TpTBpfIJOhxQA}TciQM&v`bT?5 z@3PpsLJwJ~uaCa;4O_ej19`5~6t^<_lF?yoDtW)ggnL~Q4u$yNNtV?%`G!kBaIE%c zeHjBvJyMA(RoNj)dIg7S)JuXDN2G5caAL$kN{>7gSL4t9=DYqvsx&27s-If$edhyG zmsM9t@vUJcCgJ1{joBRJXDat*WlshFJ*b!wSZj=1}H0R(AZB_QewF{pOh%c3&HLJe80EEy~>qXY^`f%Ba4i_T&$~n z-`KPdP0Jwl9X$p4!}z9eFe@hFU`d#A7@D7h&I`pyqgdq0 z7HZD~F4wHjp~B?NHyxhr-;V|eBg5w|)6R8Wr#d_$71z4TVg_j+j7|OeC#|Vjt)yFG zHOj-c=yGP7TH@<^hy*Cd$qTs9J~X=dOWg+)EA{a~b6%Iw>V$$> zV;r!D5`mIJzh?}ZdTYT!>b8iwFK`L^wIIuye!5yHja2jOhF*L9;ybn@>;V-ook&~u zhkOLi;Hu)+q|Psz-fh$;I|{SO}iiPIvte{q^)Z^|L2nvXndKx&E6 z>ddyYmC@FTfepGeZRO`dzqKF!ewi6F4r@B+^C%bU{?NzEPK5iZrkC*}SLko1*%^X< zT}fU`tx1^oq^@Q)&`DnoQ<7rivjPO;{N5mNO#PbG=|?zfZA*pL8i^fB57g?-&u0fB zZ3Pa@<5A~A!{J*E{pu|Tb7l}kS#t$Rb4HNQ>d!pF5-S%pJ7X>R9k;pO z7%m$#J^R7T2@19Qyt}cW*1@I+NhvPV1?yaN1(n)Y?I8V%o2bE4MKmju0R4dmn+ZE; z2S31`PXyQT7uEIq4jHX><^yP2bD9RFJF_HpYL#&+c1r#G;nk&v_m)9gGFfiu>F37*v5HB%*TKy71*G$?HY^LvkXITnL<`L>@Y7h4*E3?xxBV}w zXve&;(wnTjOTjHR=*M{XUI>r9@3kOci_#-keCX`2^XvfmHq@VYP|u>D;5_k850bAN z5+X*_E|xdT0Si7N-97a}lC4ES&<{|2R1Jze1a`WSnga>FetYRLEnrf@WPf=X?xiBr zBkCBc`9t^2gFE!x+v=YP(GS19J-IX5DR_LX0492D`u8DGveLBwXW1u#J*`=0;RlVt zT?^@wv?-lvw8Z(y`$vvnzp3l{cB9wf=^UX?xVf~fWo-OZ0{atfgj;JH3l>e&2mE8I zGXZu&yVyr`EVlwy!lY0UUS4h7SzSW>^o+t1oVTN;v=j~R9pc%O2^&1qE{|dx_o96; z#a-Ln3|=Bmgt)xE81ho6M}T&B4&p#NYL}O1`hbEA!CIq(BcQKDpTMb)>WJd@e_QYOBvKAG0~L5g2xDg+}pCU@PZO&WZV8az4z(Wr3(Ys86gxs4`J)bVEiC2b%mCf#r^ zifjki`d(UpNN6Fei$5Su45mzHeG5w|5)1aYZva{tO5f7sYQ^a-PGO*^fNq~x@?jj$ z;#GD^$ou>oWBtBL(L#hF6sCf2FjZ0sXY2(F7$j{ZxjY~(-tuPv4fRh{>HAAX`7MCv zs&Z2X{hB8)v#4k!ChH=Jj5ESitn>c^GQLO}7sSE-Z0eOI&6{-GaL&{iW>&dx5ZJdJl*v7J z=DS)J_}=&*FQ{`Nl2MtmJARc{+!qhNJ}n(LD&_TNaLTEg!)ra^bU4ts5hPG8kUxVm zm*JfdHyGiFYUN*?GyHY|FL9KnK8Rc7=G|^+d9PlE7m|2dv)@9Z#QOP{O_tsE7wI(1 zP$f1ksDAKV<3)Tyxh_`I)9lQ3cKklRhloDlBiYo$ehuL9XMof{;I__76(noTj2?LrCbCwt@|{ymIEMD)NUVdKlN`HCJ7lhTIWIs3_*m-FOfE_0CHyB0syPd@1wA* z{C7#Sz3USjEWVRIcafXYt+O<}KqGqsu)t%kIIjvHy7SR_0x$Df-QEy8We ztz|6^U2^j=HzFbQUUniJP&hB3_Ek^(h^(vb6d-?1DM7DuDU+S}rL0`6Nr}yrlEi9O z9nH=>y>$t3SA5AS?m*mWG26s3qEvgi)<#pAEMK* zfjte~5^mS#Alvq4ym9pfBzXuD{r`~n)^TlhYuYdsXn_L7Dcay61a~d&79cnj*Wgm1 zRB?BQ;t*Ve)8bA`aCf&BmqOq4dCqs{obQ~Od1t=w{rzU%`DbVC{YrBMHCVi8u+YjDX0D6}K$@j>O@!3AIRUJ*>S1Y{~Z z6UDI?r$gRr9+*tSdq=T*zRgzbxikVq3_${$jXG1n{S6kIalv}P*tlm`U@=oqt+xqdBY z5i}OF10lJp{OoN0Pw-9a;@Tp zruX42(q7+=)37^F^B>o8tGR@*@BQC zCJ$qlKFgTSeI}Cr-Z*5fy7hR#?RBg;bw9-5Vp|PxntAY<+zKMVUQJ0g>LaJZtj^=F z96p{$zX0f3oan|Ed0NgKvK-(dt@aQUO8W|6s6p6{wHloHSO=A#WJ3pn{-# z!6S2Qo}=cDjUl|PpoamLLzEC_XCbe?ZkRmjYOy}92sAV1zS8OBWPkRvWF6>-D07BJYy!8ZYr!?Kv!3pql*BmZ22n~5*qnVtyA#}~clVmO**(ES+4M{<`zL?DP^LW~tiZqe-JdTPl zOW5~GjXtfOOaWIfnq|B4dciL!aOF71)lXarV^lI17#vI+;=YAcCI+(292~SZ$FGZa z3QiY=s)iu$K@u`&dfIS`X`vmhv05JB`p3&T*S(3K>{v0AL(S7Ggo)Aend0q*zRK{2 zZE02}W~C2Dm~GBp1B2N*Q*a;>tb%|RxaaKE&ohSLR}PIk;jI)SHHqzpS%zV$9~L)a zy|gddTUG4ZexaqEipc$5_Eg)>{gEqGJ$0?-w4mGbJv}simWQp^d4qsr^E03zP4)i1 zX|k+DpkQWxEqg&+AGfrm-4A3kLIP6`5C)>2~gZPT2TTIir9ciQ*gR+cV* z5OdutYf1p_g;_4%@rGGw(MM5P9}mT)XoaC!nxFPCAVS!{{tfm$7yb6F8vsY=Qo z)R(vSETUxs9nNXm`Aq?~#nqx^IN;4|x1TRbkv>j9NRn@7GdQqC7$@URQ5E_zX3y+3L?g zt@3 z6XDUo`*ZKuT}LxgY*o)mj^0|YvNcZKwB7aj0t`Mr82j0-a-C#_`oKduU(w2}jE1`^ zF5%&ZXw6n3)qqY6gPvyX*fC`n$EN;k`T}f&iv}qTC@gn-3UW{Xj=b-waLQUVu6~+EYh1#JhEKdVmza~^zE@fia!;29XOTe z_|jOja8k=@0W~OlB03~fYA5_Obc{Fey&0rz1(9o{;#jJ@!mUBf67ifJTpu#TySrH` zD~z1$nhjlr%x@wIs2Uo{>+DL^_RFss)6fhLG4OJYRJlnsU3K=;Ju(PtEXmJLG=t(z zrEX_;cX{5v^`cc59Eq}&vUPnxKmlV-_Et?`xj*9y%M%w|@3DMkEM4Y{Ra;+YTB6XT z9n>Ik2T)yUsF{nkur_285Szye!BP_yfa8#O88g!Z5)LGZmMhEi@5Vn%Ry|X9X(Ur+ zBvM?%UiWCfd1(Jxpp>EUZdA<83&&FAd0LTYYe(n&kAIn*;g0k!ctF3p&l$W`>-m0% zE%4ExtRjC96-RzXS6ETKKwLXHW}w0%UEleH$!@B;VIiv3(6M*1oZtLNhob7DBx3qz zqG#W4cg<7m%%z#e=f4U--lc9+vd5yxHFVbAwyXe=`qvM4oO7qdc+U`@@Ff-+ZSO(6}qq#oE zp0n2LNXCyfT$w?I4cTHdE9eWb(CPbtlwyPoEWg4$Hw!7}!)+;7DrDKpQ)T89`r9Yk zw~JPlz}!L7Udpja+`tW7j1oJ<;u(>Ef5XQ*7EFSJ{rKGDhhgF^{@$ET~YEh8dkKlt!=RmCzb= zr1gfvx8Sc!zoXfQBo`EK64RM*Sb4V1-J5>;C!nQGqoLQ0>ANO~$6JHL=t~k8KE&r* z-`;@0BVCA2HUIvg6y`&F;vm0_&(gk0ztHmdxBjoI>i^E|DwUb!xGZgo z4g(_~vHUw5MV}xjx5v)ECuBZ!UJ;OgW9;`7F`u*fDRQjOV8}r6okz|M`}FXK@Yz@1 zJW1sB9cwk7SVk=cu_TCtN*d%>1H$VENa9dG5>!MJk2q8jv)e@OoTu5@-`PEA_T1rk zN=w7YjptQ<1QsiQV~iVWgGY0{l|VX=qZwyeesPi}s)X3ZaV9~Kl-6BQv8=@Mr~=0S zK{~zxOe?k%eFH_}`#vnVVex%4Jmm3=`+37tXxamp)Y!Jf!P^r) z?NI9BkW~mjmxTa(+?%qq<`-J8Uzk-F_KpD>SqA|+SbRen`oZu^SC>bt64KClA#AJ9 zHZdi~rTS0cEXgkiz<@0Q8NPL=)eG2DB9L-5XsxL&?KlfLrMvAKSA)x?ovpxWI(KlpPQPu#mo~03_m@HfR?5xn zBj9I2klCjdX>PUSEm~Y#=LY%PgX6WVFohy^cs{G59|G->)f@c~s$krvk$?_zyllWh z+TldA$9QbIlk|#oPQlUZLfH)MNSyn0WO`3RNGN{9@qT?#P|)fx(DQWX9#>edz{6>v z{5sXFF~JD+{!A);!>ob&vkJfZ<@aCb@_=17^FVa@*5gcjB*&20y)G*YWCKqs4^QUo zXRO_7<$`e5kZ?mOf67G3MmYv10yj*}S6roD60dCU;|($+Mh=C9bnuQ3xu-*a97glx88zA3b!84 z_1Qm8!uiPCBJ*vQ04K6)ufnW~8QkO2qcqeoI4fQBpwxl<7n;1NgW5KpJoL9yrp>oo ztfhR>P+~f_a&mM3CBI~rvrv@}fR-LWZd-<(fWRC3K-kJF**8m9kIaJ_b%qcNABt!<;Q_TGYH%X zm2T`1uviEA;|-EhqqByDkXW96%ykmPXEzfWf%e7jZj2bpS+(8UA2UPOuxqoJQGISe z*4FR{yMzKjL8zUwE}I=e{R<6GV0f`mGe~HR5T_cvzCdT(Wd5kzq9-=Uob<5M6S1UC z+)#^nU#G}N%#YV${iSm{IDS$318#z}tE(1lz95O62?5N!Ct6MJEt=y%e2f*mCrp#1#TsLl}IaKyX_tSfe8%5lf6$*f% zs|0)FK@-h2jUP^O38MyO{a{+Q1+GLcubrvcF0-;kzaL+=t9@)$CF4;m zfYXTyzR<0>5LwGZEnvIL*sy{=5lH}c0R22Ul!~Gl!+GW^@QOt-YBWzF{mOfRw%x^W zBJdCwo@`lLJ{TNLTzFtomYovJHTan$t;h@b>H}kSk^G7XFw%$K=_Q7KWc!-GPx;kn17y&?dKO+s+qrxs)c! z@$Ft_l_YNj*=d03lC{$)HHuJa2j}X#n<77b&CV_C+w*y94SgzMAE)mp;S0ME=iXE^ zqwd6G$23>ucL%7nH*mhmw|b#b#@fRcMk8p|QAL&v1q7{(uILHr(Gwg;@!M<5Em+nm z_VEOiW9opL`tU^NWPD;tvbfzfOBNg~?KMM@*mS6Sx2B{NQKGAl(_S`YEla=-!}kzu zUa-Cm?#by%A;$LTsgDFp02wFOiwHJA!iPC~+GTzojp|w-8$^qSgx{Gf1eGT zk4%G9;q%6!3mapNgGlTwrVk_>a$OtEKU3C*a8fJnvzv`oXPt3%hT0}+HZhh_xWDCZ z-Yr;@v1R3$u2xSCijX2FBcytwYAQnIOX~Ow?bM!dTka<0pLV0uh>Pm!v>BXJ$2KfA z)p7)wf&g~PPpiLC1$Y9tbq1~PQpk8{a&&tXauV<8hsyp!dyW_#@h(BL!sasR7ZO@9 zes2nXKO$dB>vqs&woPqQRuJC|(njb&s-rvG%P12-LXZ318D|Ciif%3yE0q z>NGX1UQWk!y?gBXe5}YMonm8f0e+0FIGPO3X0x2=5UQ9=H347Nfu!s6{-}J&T{2okC$$y@P=*8C9KD$ zL9*sSzp^ zIvDp6*+5`pzsAI3si{)uro5w>E^8xIdjH}FoRN0j@A@S3-W=)!#)Bmr#tYZ>SjO1A z`?A757~^AJrh|5hHTw?=nqvH)77i(RFGA2eUP)iD+2h2LYeWV|=Bpa0(?=NZIJuzs zp>IG0za@ByuEBRaxe8^*Yfh7U+hsEpT76gxI5G9mB@@@2ul65~+2lAksO|==$DGE9 z!NJ(0g9dyso%+OjM)vVTK)~Q&{MToR+^-L^G#yP{-M?_nffdM`^_mT9^*LOaI2ri; zruB_Z({H?Yv9=yPjK(4|BGIU-6y5!JQ+utHFO%Ho7XXteFUk={$Bv+o%W;TnZJ1K2 zMoN;y0FUL@AC#W#IRMYKA_zvt7haIJSFUs4CTM8Yk0sx@sEK-!fI72Bb` z-}Zp>PZv~mL&j$`TFZqE4sXr5z895s`6(x4JVjLUQh36%wjcymGHvN2N}9IN=u9(v za$KHM?dsJyO2JM=;-pCwV{2e%z;=5tJ&;axz{$B>(P=NqNRh0HEL=O7aU-$AsZ3;o zVX}L2&JH5yfF(=E5LhshxH^7PK@Zj$lBpXH{*hCeXr!lEMsLJmRAmI=DsmYgw6UkN z%uWy6WYqcf_g&_$_?LD##kKXq~FK`|F5HAAqrCS`U7JL#>l zU_}QCYQy!|EkK8w2ZRENN;!I#-eIUmV-L^AMmJU?!AHW9f`H9;eCu%O3Tu8$m*82O zL-AHQJohs?E;sQhLwu~)N59=XKK~6#46Qs@+!|H*19yf~K)ttS@?MdyE8XnVkA*Jj zJtNG_?^MihVq{t$Zkbn0EYp|K{ool>iX=KI4Pem|{@LXW2aSPL3+t&fua8ep?ck{b z?d(0_g%8t?20l1d6r{#Hlfko*yx-eD)NC@S5MG}5BL%Nv^7>_v2QhvJ#u;&oM4_Hv zXZxGdy-ZU>5W@Le$#rXG1^Fz~nCBdb;4V!gb$fn37 zk(PxiRzxgRjai4m?6Dtik9Q#tcxwe?JnwoOBwM7+d}suSg>Y{UzMd&2qg_s7!yDrr zk6aJ#ayTYUlblw$6qyi{61yJ5))$ggDHfT^t>Mxt9rW0WCCPE4_VHnTRqAM23h#N> zq}XiMR1E8LnVfUvjlr^cnn^N%8gRr9nQ;t= zIW@sjX;`(v>6nICKDEqpfscTxe6~iFB=LBcF)n%TY z^k+%r%B#;ChF;j+DDw2R24nUs9jcBWZwDF|;?R#Ac!F6+PHQGnlY z??2&v{|Pv?%|HwW1-_QL&4`m|OB#u}VvSm9t`#NKUH~Z?oxniHI<1ZL2 z7~o!?d1IjU80b1`7`hhf*s1@1;;>{p?*khipi~N8uaev3rA6!AsCoN(Y)1L_cr{-# zG%x!iIyyGjv{t9e3`%keqYiU#Ak7=Bj(aCky(h?OxH0^CifZ;;OpGC@m5hnb)mX8% zNbPJ)J}78FI$h$y(iZX4T;EhDGA@pY$v}5BUn|3=8>6Pd{La>-uN~E=g0C;&(;r>G zXN>Z*C0_MS3>U{`&K$H2R-g8;r7P!Pp|#j&!V+i96#~$=_YwI<2cl=wAHa5F zVlHuXyai#!MAbUaH{Oq(Lcm2rc+T-d6{5ItH7#rTkPb5aZEXjRBJ9*o&adAyq?J|8 z=C^<3I#HQw`t+>4rgk=GRn?sHNyC@?IvYL{$MOTPKq;a?JaflDW&NJWM9>;$7s;SY z^+K8(#qQ1qTwJSym+Ou=!jU8ot}mNz!PZ5bi20yCvamf>EWN)e7LsG5?6nhbMx_CLm z@v`{tpeF%y&BbxJf{?QEGMhR828p7RL6D}Qbb-Nmw&FOe~` z1yxV9BYK0{wO2b?(3Nz)YJaZi{fAljco8Q_;*u9be2ppmMJkl2QP&_wYtg@77aCn? zEwe+09R^8NG6gIe&bfJQuWWZ^3DU4h59&fwfM;VI-y6Z7c}sA3j&q1Bj0d2cO{j-V z@0-CK*T#{iEo@%E?CN3{;6vGT?z)5$-5x4sQW7+E+#HeVKDRrVY^~ext+sZyL~J%* zpTRO6#_O}m-NfNt+|L{k z0o#JS$6oeqKwx968)2U9GxWw{HUUobY{ZN6FgMo+L-zPL+?gRo<)+D-t5G0M10@R( zU`Ph(8~<|oq@JgaUB1x~K?3#m(5lb%kQ{VWa#Q><|87&~)VA-KfC|=Sp)1bpip(_? z)o3KYH$67=2;1+`FSKH>M^Nu4t**zDUV|D}TF$*fx!$^>)qxM_#7xrCQa=UlfANkW z|A>kIWt916`I3xu99=PtZWL3KW#0A~=HWS9c~$c^-o5C?)P~L%#RYWpC`vR?+I^EV zOyXN+SnB*R*a>LQCUla{MB;6y3J5gET~SJmPQjG&u5q`YP-zc)9Wtp^RxFY?t~c%b zK3pHGrn08WaEE8E29V6<369P3IES~GkeR#m1o`zoEMi{TDqfme`?nTuP|Y2mMWA{@ zg0s*N|Hbnuq71Y-3uttWp>q?y4_`U@hGXErD}5IX(7-8Gwkm`C|~RV zh2Z^B`SyzP@=|8I8jN4CqIko$KfTax8=bwC_oe)o%~i>|v!;Y=E>T>BB{PG_X8G)I z6`;H2MB~IxHPDCIKca(X^$k_}npAqqT3vsK67@_pl-&Ex$k#Jb3dUHzd8g>|ia{>U zooVXuGU(pxPFg<17AJJ%UP zACNy7w4Xu8WCPdfDOd`Pv+)^DLzyi(!^Fu!o7^}ny95652O8A@6B^vW%agcZK*9W5 zpSnW6)AEH9<0?}9Pb|vaz1;#nG-o1C?`8D&=(Uw3{yX6gJ#U-|?)5xZ7BpdP;_s zQ^+DPp7oixm!!wRgG#Zz&|BTVpT++|lfw`1S0D3@BlcDdC^Mfeq_A>t1T*XMTpz}} zj*ju2ddmX8xQqjW6}XK?O76E|n~&I&eL=25=XGjFkJBOu&gL2~ z@BGCjMI--a<6&x7XS?r|MOqnzbN-@7(AdKXBTcalw(n@;%MfV8oyJBVlV(~DPYy?t znc`L@AiCe37n_MZ;r2`-&?Ud8+v-_3|s-ohMs93}o0Qy?^3-_~IX)s~ioOArm z=N_YOt}%9d<*9RZou|T#hW#wTj zJr{}jS&8b7$r+_{g{QF#_00+PjJM69yD4P;=nAs?dIQ0YV-E0%w3kIl>k@@*d*0~M z>ySlyY9lz&{Jx^dJv2CAa@sv9<>BmB4D>+jH9L-N!jMPxb$YC->^Gc7zCHL8p)QQH z;LU*30M_CwN20jl*~0IzUi2AT7+X+ueqfr~forBp{C=svFr%)Q^q7lwtIuZtTy>6M zh%foR+=7zg)A7>%1P7r@$<)qdPa@W#9WUAEs@nDlt^LMBl$th*VFsa@^#*(_^mx)E z)#K7>1uT+2<2+Skd-giPDX?whg-2ietGreUV3aQoS77sbV3YqMdn0uO=a&wn-~=Yw zmFdS!y1bZ)GaIHLFu zYckGh?B&CMp^4a!+wOm5F@$3y`=YDNR+|JL+2x5h@$3LcPxR!q~~dU zOjCrGOgG&l;3fk0B)lXdrP3`YkQfFvl6S-LOUw@}MIWeSA&Jtj-1G;qX6E6}JgHPd zT-J35^QEHOHBVS2AJ;(cbKeJ>VR_5&hADT2*CX4l8G`{rUm07q0k2d+4P?>PtJ4Pe z#sf0)g}UJpgONgWMY7sSDN6507U~QdHIbae>9Te;jr_9ck?j^^Hu|R}iZ4Lop!!1) zN^kBgy@Ry1!-YFPShb6N^Nz;ruPqCOZ_I_E&TelWO7UFyP4ERO7EB(cOIP@BEHCLI z>E$|gE!XW8ks0P?p}sAG9x`K*RxYu4C25d=gECtjDMWT9g>@k0X( z8b%!srTX5*`B-Y^jC}gA?eHwX?V$e3mVBhR%l#F7!&7UUTIxyG^lY846~2gcFOpp| zS55}7w0vh8^9Xy<0t<;KFMfU(S>khOd`st~v4z!ixk}=cx+?a^RRQ7m>nZ39Yvkhb z;MCs6Ngv_VIS+KKZQ7VXg$~ORAAwPhpIRU0P-CpLBVVslUV^Z=d8Nj!nrGf5t2h@~ zyE86hcAcD0_~Eu1w6eGi^KckjgH{@jUZXg9yNfKXFlU41w>{Q51V8bANj7;OpqSks=e!W|+1zwh}ie+WNzgSm*S1xuj3 z-;6r5z6{kCNRZ1J($3xEyI(~+NqNYCVLz9TArzNU|Csb4VGVS)r{H0}z!0(?iDb>1 zsf*3qLpegh8{D_ZGK+D3xS=oooc60H zy;pwAqZKZgZt&xVxNEzkXu(tO^ba3<-1hY@hgrM#H|n_s9Df$sZeY5Ho*O zlFtpA{UL8H!%zqNJhD$O75J`xGVdu?S(W6wa{}CoTU0oCL+qoVpL-$CU%ma0CwF_% zKd$^j6I#_3qd6&O_*c^O@BhHaQ`4Lje*bC)z${}DgsIN%Y>}T$5xn^R3oXNi9uM#8 zUr)FHfnlZjjrk*O*Z%VIKT}-kN6xQc*(+vw;-%hp6l~Z6bIg)K4tTDdSqRFviUAue zh=ElBW6rB7{Ei<{BMp(H;cq;;oLkIK%IN~uXfj+V*~xzH1+Tp7-7=KJ?hzOI4}nAQ z=c!R6oDv?S@sOH0TlU=lADRA#O7IV0Iy|pO!AR=5@sRtgUDR(Te3`e zghFfmzXzOq`J$ufNqi2@Gy~~4UknPqzg>I%3vJ3SB?UE;r&6`}ge2>BUeK+@xiPs3 zxsvA@+Wtu#af~|-il^?oD|%y*ux=W^sZrRJ+oh3cSY>lSbNFnXYU!5c@BbRVC7M7< zt$8r6KaTXK%MQ29{V_ z`_8Qw_@Dzb2B5zYj@UaB6jFWJso={g-#D4}hPN|jS;cw%6``RZ^}{J#x%dC&zR+#f ztHvZ)>?_xrXY9%CgSa--9*v!7S4n+-GgvIXSmALRYr#v^JjY#vNqYEAUyJAi8xd{} z&?XEV0&)sWOveK5)EVv)H_mj?Hq-s27l|#uFxJ#iU+ar%)m-+~PdI^is|e_0+fF9L zOdmZd=kpFOhu;;8X9@ptZ#P&}Ei$C3_50jF@W#RMw?AjT>YVjUC}_7&;%)2f*ZX+9 zOpZP7SEIvoWzYJBW|$FktI;sl_T_u`!b#uF5`WfP!%x?`<54T8L1sQS?pjC3w-3+J z<(Tdpse6()Ta)TPcfwq5AJEvabfQ_Nd*mZT)LpF0h4*CCj| zj0!g>9cclPX~!xr3Eqmjn=nN(n7QUOMmMYnLJ3rBYGWXrR74GS#}<(pm|8;VWeQy5pXIUZLn5=}~l zAMlK&xVMdG$&&y##y!uiYs9n6On!?MfBI9b80}54uZS#7zXrkVwhs0$v}mTwX0~E# zbq4UNTB-lRgnB`JpsWv3A%#kEw4qC0(E;>){G7^P#!hM!NpypxWQH~Gt7a+iM(}gd z>Zf4h0w%RTV?vbz=*QiN_>-p3JBuNG06~ml{cVwR8Z656_myV$&99Tjf%1`w;a(*E zBG78$GTTZL3Ow-30#6+O?2Oqt6*5OwU>#M^2ham)2~ycY(i~iYs^snAlW7NdRCA7< z=ZhIu=J-6g1pmX+v8QXl#pmgvPi_px1nbS)#5}b&6%KW@exYg8e53ks4HOmqSjE6c zY1(cE6d++;F?@t~{;YV$DfMlc5tYl7*w&lj?O2UG)AbGY!lu+Nom6FwR8a5DrnhkJ z_006A@qJS=+cvPc&&N9gF~~3{hQJ~pZ%Kk&{HI&izi)JWMXbyw6KL(`AP~Mn7Fc=y zN;S=rFoUascUf5pFVeQCVW=Ur^IKM30mu$J2KHWOS_nkig&hK643t#oE2NeGX=lZ3 zr~o^E`c00x&a52%rFDkB4ULk@02t1!=;Y3Ix2-rHhR97X9-y4v!|Ey=46~8=Lf2!H z=M)Fao8)C}B4XO!Pjhz71xvJ(WzN~!FO3_B&9c94_JjsnG+0rG=gJX(m?5vz-OOLM zCJdf`%9Sr4?jB0ocH?K)Ii&YEiiNP>8_VTctOM8#^h6mRZ197TRzESZI>55+6#IN< zUI{Zf*p)_J>7ZrNa{R_ZtQk_~{`!^n93_a?9ZFcJI1gy!Efx%}fQe{1ruUj_fq zg8ie(pI3?cqhNnj;*yQ-&w~BMI9jARb)5 zN4!3I!lZePWnlW^PTK<=tr9fK7s{`u9gvo1<1o_Sgu!Q+X)W8pg(8q=6j*UYk@<4)2>oIp6bUXJl=L?%xP@)G$U zH5dT^tVXue3&?HmCC~$nFB7yZ<>UlHIv~cQD+Xn^cYqH|!>oKg( zEQecHV5N9OX4p|Ipa91%EO}>i+WGSd2w=4vJQ$l`r_x$t^o8xB+Tv3%3!rTImyi$bN$98_{Af)t}Y_CQq;;H4(>&!7a>_@qV&`? z8EQD`xRSA!?u|}rr6`S7KSqZ#*;dXR_I0YlkO-wUXlZ@D#h^^#hJw!(U5~6~G+$;Y z?0_)bnsgLJpmKVXc9K^DQo%m&U%&sU*M;X7+N#@`|J!#{{Vz~nb(??PjVs7>p818g z%#2@3GafYm)jZf_^ZhThJ}9F%>(bAnZ-s4jpT&2>rS=bf7%M+#`O!gWOVmBkNjDG1 zz0#i+xe}fRzkLfjd}x0aAq%aWN~kQ?)eQduSDZPRCKxA z+lrnKv68VK9|V22%HB&s%eWThMSA6m#@se2!%XWi`L9^yu!l9Q`bgq}B|P_Y_%a0w zt@tsn?_I1lbkbD}xa-X53YGOvmLfOPkhJKJE#X12)oq9E8ZXU~oj<@$^&e5ZRg9Ib zjw{kVi_7kW9Q&g)U~=7eqmxH@9J*_z)3TAR&qQQnzfek%d_SwD_7KPDTFUu5BICTW zuHiI1MN_7sn<0We ze3@fb5bZvqG`m9QeMSn`Po@9}rQc9QyKZ4BeD4{vfeg~~VCkU3XsDdm-8!^i0Zu1x$FA?x){XJ{90(}v$VO-luk z+9^|nm6@jcI~Wtq!uE9QMb`B{O;Q8lK8ZJlCu*X)GcI z7PY&#@Q7!s25&~U@hgrvF=2L=(=F`Vv}M3|Sc5meEW#e54ob~N=sCv0uKF|bwA|L`=%189H-7fvGz z;v%!LG>!8j@Kz8RH%Yv4nV|u9>uR!chsqj!-F0C2;%HAC7O`5IF>c_(Srv_ST*gRUHReWMSV;&+SOR^^rT24d-A{TbxF=D1A! z`UiC-I;rO?6drqT_@k2`q45SL2dKrhh$rMv%7(~n*#^_1iT9TC=`4pU0Htt8;_C9Z;k+K`8i4UVb0P!5GdNw2`=J_py<&yx&5X{sh{xRlzXdjP}U;M8?vr3EV=gVV4ho5>xKHB zkYodMPTcBR6Qepq`>A1F#JYaJWkP;e!)$ck5o$vov`r&_R`vKTU{1-t8E)Rj{?pG@ z%Qc^N9VI*Gtk++ReCJIu{eE!@sbU^J7$j)t_0*#)QEv$BbBO_5UCMBI^Y0&DKOp+y zG}3l2SqKm{0P&-P_%P%4xea1dNF8r*4M)Cj~T5$p*HiNS7E9N#K4EzW0 zsW!z_F1%R0%~n^Bi?LLB2dkOeVHMb2U%5;TIxl>s5aSf@ouKxjhm%zbH*v6pPq|Qi zuAij=f>u-%7^FeQG>iNWfF3LvdGKMm;aNn#4y=R^C>_cK?$)ld1UoY*AWUIC@Fb2a zZtFf6OZDVlIiS*Wi7#8Xhx|(4sxefgwk}trNxRB`Pe;pP2r}zE1L#S^fy=ZftRQ|_oEkiMYniMlYyD9@T)i?+;<({_mFx@N137&Ro)cBU2X4in5*WHZp>M!b zAY7U2Zk^07Iw{>Ii*-kkUAIJY!%W2t6Pew@N9l?W^Y)tYh+No3HB`W!)W|wM!nX0e zxZ*tJ_PJ<9lK3RzXF8Z|R|<*^(DK5}&^ur-A}HSQje(+)?c%Z1i#^m9ygrRNQ3gYO zJu*V4)a8sPYc}~Mpd&+0qcuW4OmJS22hE?Xt(QkN`DY3IdD>EJ6WpOYK8V^!16Gf8 zFGjz=z?ds5?L~5F#sOw{SfU|wrlYx0khtwc0JHB)+fa$DK^6CvdCiK-0Ak4!%S}3q z6_S)uX_rc z=q9Y31b9Je$f|kB)}UKBZ}fi@re7uAP`qOH)u$9qAmp}N=bx70pTG!jLd!{* zE{F|T)ec!2bU)2&JO7#6?;4?U+xs)O{|BZ|iXYta@z`*PaAQI~Ar#foPA2zHvfmh) zOJ-hA{ygn^Edg{1T-h=r-#Dtg42|bFH}tDV=6#V+K@$3%o$r-OuT*q+`sh1|QarUt!?x?oM!AO3 zHWQg1)>iMA4Q;wR#ne&G1`eDfz7Odh~9B zN3QxN0r=qBItL^`&UcZQ3438k6em$>KvAzswfUj(&m@eqC}GhK_7Hg*EC^xZ|@ zky?15i&aYG*$1gIhLr@9DPGmM!V&tRs9ec_T7`vsmUK<|2 zj%^qej2mn%(%`KJ6!iC1#HaVZ7VI^^>k=gMlHQtVY(1@}-ZBd36}Dm-@k}M3x@5X6 z?zGu2M5ap&Vl=$89msD+D=_95UHcuxwOB` zBnadJ6cB!r-Z(4DP13F#=Y83rT*~WTbXaH%Q%0) zsynhVcJ7%}6hqk5> z48sVwIMT2&b?%g&&wY_b!$2kS_^a_0k}Iyp(uT~mKo=FHSc{f<2KZ`jW{dwa!G_Rf zJ!>B2NBe?*VE2B&O)ZrvOS(VL&bcAjqPHHM;`Psi_V(%`u)VASkjG4Fjn^&<0LF2ogeahnDOJV z_!pYHM?Npb{VuEEBVd9cRZGWOP8-Xb7@PfSIrr-3zn=t||HdMIWB*M4joDwnwTb$L zhHWnN3+-n*QX3Vs8E&rg3oS17vLyVn5zfZQT}9N{M2xt2OyfMN`cm-3kUTN+qnyN> zA`bUoXr6wzRkL@^6SoOJD~@}-e3lPgFN7&Pjt=3wp)3l42dQ0pYW#E=v>#T*;<)tR z*2M2koa_>f=%gxsRbx*o4QQftaRkvhRHqNR7=BTA?|B1kBp1Z2{;cM)g%driIPNR? z$;|$vS8Yz4!|VWz?-=UG+D%fJE=%Yz zka^4;Uv1@*8rXc}hdcw)Yyp^5*^|3r7rCFa&y;Q^7W+JgZ4+n0xNQ<>e45X%cE9oo z<#xs$kHSx*dIu#C>xbUL{2~#n6)3NlocJwu^0==Mj*aDz;2m^qf{9NW2f7fqq7vo- zHcu?h{v{dr!Um5!rm}Z7FM~RW&O?v;byQ!IT;6v}ZsBXZ3v=@>}sy5nUhW+1J%K+wSvYIdZykLEw0(^+=U-dIMXO_;c7T zLii>@<6Cop5X>HwJWp=@#8r$kg&k3t<|W^fNTF$Oe7#4|nTJTs422XN|3Z6(M35)m zMKFgZjgMi`EHVg%HeXx_c2R(r$<%(pj*V*oGFUV2XTEKzPo!#rz4 z$s=%nO&Og5?}}e&1uF7)H1eKI1wnGVVR!F}RalY~iOC-+%ap5GV^>gGrqNpZ-Zd$Oqa85AwF&hpPC&}{^z)$>(iGz#N@{i+ z^pVQ7^?X(Lgy_!~)SH_7EAQy0e!pV1`U?)W5(t&&t$QLtX{Hq10$DMtgPn#^DS4!h z$BM|jcd8}$D&_?Dc}hZ)YIP8nM2j8q>%HHv@U`ASOc%FJSX67|rM}i8r`i=K@9`8+ z#1BRl+S~qCMYSOY#(%co9{k8JHZaC<1}-n_Tod(B;UQ|H{X_mL~d_Wtb~z3N+k z*8?ibY72$;m)_j87*T1Wij1!tbcA2~#;S*D-6)ub1K%c{h`@}$%-&WJw&Z7M_NOuw{LidYF-b62Ay_9^645$Hqg zpFmpJCjq-#jt&aIS1N|B1EBBDyH!N2ydl7}rKq*JSq@!mIaEk%AHuR($_r>t@_crR zqThwAvtuFG?=lY<-TospZLF>-$vjSnLTXpV*slZXJ!k}{3v3827r~_zmzW`CLg{}@ zP`)mAgG>Ih&6McA_U30bdpe#QB)&9lhw)I`Md$t1Hlp;7yjYVpNbsg9b^`{oF)u(tR(l*rA;aZjl z%$9J#J{bcPa^OM|j$^8D3DYxXV;euz;%G!dSlY_u^}es{WJB?y=KSX(-!xK8`Mywb zv7{eSfCT%v`dH-^ZkfF6GtNTqc5`#}AE`f|d!_NPf`vufz0tne_C~5AP+=(l-YAaA z$MN~yk2|bi?Uiuik~HOCz`?sK&!>+731Ys+~x~hEl`&-r-w2cf{!9i_=OBq%@V8Z=_q;3|o>?lkq3PO`Xi{ z698g;@IEcaJoeQ*Q->vi0sbU93HhDQI5zY@AyyV^*uO%jEDM)eW^ zBou;|?f4}RLYc6wpqU;Jlw&!V&dT(p@2wba!~-_6?=Q`G?&I>1k+}EUXO-G^$A;P6 zdcGBs1h&F-%;g7bs@<{8sNW-&u(7Mi9c^92_a&1zkezmd@brEpkvfJ1h z0q2BfGzwIY8DbMqe$aD_0bluOePsI*LJCXFUhP}7Y-F1XMO$O04Y8SE!wr#K&zp zWBfrPshl-5J5Z8B5=Qo8Zp!$%89duh!HI~PO#P${fmydm@*>P!0mm}Ym(E0Q{sGuE z>647D*=@_bDq=4VeOHRti^MCPnx}rP<}Ebtzmam7#j-t|kU^F1)deTcS<=+m$8o2Oim%eVsS=E~ z(vuHXkE*9xSzlJCRO9O5gu4RB&t@x~2<$pyqVcNL&_FB5rKA|oL)v?91~?wSBU}*P zeBYY{=VZ2Do?Pk?N=R7YWtZ#(OBRLL359&Y9->C|PtSr5M)UX3T45@D5ZOQ2r!5A` zleFf!o3010fxXT5I$)1S=WY4!O=U5|&mO8njqC;t41XS+5$-H2Za(U3$D|yOTukKm zCUR1ylqW2)aFr~h4*q0Bw})U^Lob;TTObpp7+PFVx3r-~+q=rGzI8rn!wc(k3c8BB zy4#qih5;u2gP0Z7h#A;x#Ec9YZx~?YKd4gC%_+UjG;c+kcaeBh8#Vc6=-PN7yz_WA3Jk+rw5Pky5N$E4MASXvD*YGg%~ zq}d(FbQQ;1Ci4=$z8uzoHTO#!eFn6AUJIAbrluM6x#nyn0H=KLRH0Wn>({vM&9yO? z@hm+0@%3A~MiCS1`6oRr%&V^f`1EVIltHKdc}x#I=e_Bec=X&7`RQO--qgMoI5|D5 ziR%M+PcOvE|i#id(rxBHnzD!-g%p2&s}0xLru2MU3A1J&i$|E zc^mH1p&zBx)K&_tvg$r3g1^Xr+`10;em*klnEKkTTGC4)`J-w)ZNV%%8MfxDrpTF} zWAMlR>%}5>o>y}^RVDX^<4e$vBPtvndC!VPsSm5!M{zh_9>rxYA*By{!}VPvM|e|B zp{VdjK_Z+5R3(d3?{IF=T-j;zt;(;szgT5Ft9YAdO`#SP0$%2G8u_8yCE z4t#$k$_f>YvL23sR<-D6tA^`ONE&jtjfv8JvQ6-4dY>4`*j^#wcA%EFT{!}i*i#e| zAD0Q!vMyn5T6#lqc1R84b=m6%W>e$FJ~>kna$57iF=}*CeDCdZ_!AU>KIomWX<*%| z%k)F2KjrfW9>E$0W`#hlYNXswYPwGb8e4p?ResSwCS@#=?NFXtZ@a~x$>EzB@)IA*sr-7D$x?(0)rUt?mV_%dVfCC zvwJyK{rXq4-nTJ^5|Z!i_^NLWFGb(tI149%34$NE>ZAKnQ__qxTn)q_`rih(de->K zgpwjHLr{cqA*{t8ML59&R&IO|W1N1a*z>hn@FS69Wx6XBUfAk=Qc1HD-21jd`B_*y zj#xBAD6-VZww6N=>wJOhNS}C_hTn1pT%Aq~4$ci|vVP^ZBV}_ow!<@CYBJ<~EZxzQ zPTQh*qKm85bfhm<1I*mZ1Q_9RQjhMpdH-}e!abT~~A(t*l zTCRg1PBJd)k+!ldy+~WVDhA;L+?r!y84vZVhx#}ilHaqw2zv|PoZoR)^Lp51xR6pw zhC*FF0$d53R`?nm?{)QQ#Xk7L8GtuAhr6Fdxn*N+_`|mWDF@b&L1n|H;<{*JcfVL3 zj^#w!FDth0X{|?}PB*?)%{a(Sx6QQ1Hq<&=4YzcCJ2&E}7ok?`+we~4n)J9=0gGH2 zNSP@wy#<*837YL?LG6=#i|b{;*K~dEVNj_w3mXxglEwh&vhr7-BcjdDu@YD}g`%)v z+817q^GP_2BjDh28_Lspw_BRXYNi2zhSrKV2Q6#JFoLKd7e4L4FRvigOJiZH(avz4 z>QhER;uW)zxkMX5FXEO|+WR4eG7*a)hG(j=3Ln*yIpjh1ifENCw(u5sG7Hh55f^DL zjvCrY8{^NEnUrJ4%%Yi2xEwO=tY*Vl5vRv&zd`K2ejgcU5MP1dm8ie%3?itU{utBaKWa9tx?*z@2%167T0AXg^R}fCGXn zoRV;IBr?E63jFYF-sypHA2=-|9tDYigc_0GI6M*}`k~l=>4!KU(rP`A%x<$S{qltg zEc@vvhhQ#Xaw?FrkL4v1E%4EC%kzMlU2ZLX%ZZyWSZILh76k`@~*!F#n zpZL1%jQ0-Ovf^@%G^nnxS_Qg*XV{dE9Ph;Z6I+E7sBy+~XpXcXsWQuXC5Dm0rHot| z2g=E8^b$R1b30A&{uvlW;+l!KX3>uXwUtp-f*)C)Sc5e(Edosi#GQ(GzwG~*;5v>&8y0qUg{ihLC($c-Ft*sCYc`@9fVstWvmmNpS%f*_3xxZ~Yy5u?~l^;|VC z`i4Mcs|U>5ZE93glEtN%XcjC_QniW;sp1~eF-w>e71~u#Oy;kr^zj8h)t}1lD{O)R zPLi&&K`vZHAMCGvDlYF9gMMsV{@!Rze*>=t}eGu?r50*-$2<^iJ^FTTE^%cuoBe-0~il*wln!UZ+`I6Be!=yfhx#o_3Hf*C{9q}<8 zE3R3+30)7a77ia#8GYvSI0fL88%xCv7IzxB>3P<~xrrBe0ooD19SC5k?fz593JZK5 zb*EqzVBAafu+>elxIub;`grj-qQ8Z8-&LL=2<%`X_2{1UDlgyj`F`Z1d?n&_eH&4N z*gW{-b9%ej`P?Z&2TM*Kb(h-9_jYbCS{Vsz)sHvdJ)Lb=-2<%VzUOw6nkxU5=ON%a z;sB-+^DS;+f9**+&&J4T_cKn2*6gU`jaANQG9hhe5Y~LQ;lv7#&2GXn69Nf(XVj$q z0}j(0e?OyKKm8U?de_zh-*IoDZQ|soL@_ZJqaXN;DhXYp2EL7HwmmcmBKy5cA8JZ~ zv|5+NJU`MIO_)oq6t2pRk{;>ZR!t-)?xXW21nKD~?ra46MpW@)QOz=}32y&Ja_o9* zXpaCQL_Wl^Zy6zb=-}>8yGdO&0UnI%t0s2Fkd&0f92#?~sR4EiZN2dDCdmA`>{7cj`KNo_YbWIxWlTPOKWl&c{2 zL4bBGXuxED?Ddy+i1*xB+M0YZjS`s`?K!%6b7*-f(NSQ^3k!@N-;FpwrlrMXoSZ4O zIpe6{(O+E}adHrSNLYDK>D%vSa08AFXwek6i*={dJ?yANcJF*)P0^@JTx2@p$qcpf z^I2e^GK(287gg}DkwIQFK93rB4qYI>G3p}x;lOr{Q^k8S{(}i&Z2T86)%dxaN7QoH zb%^h8BqJM{PP(5GODX3`B#gpSapb>|1PXsjhvXp8Dt6{i16vHMX-^%0QGLnpXLJeI zjg$zS14qj&zP_*g@1PYC)6>P)<)#GL>wOk%rK^+eTkmu#dVh^<6mc*ZeQ~=C zxe6s#;#f%HawN#K2Slap;L+Nw!Vd*Z-iJsfdnKu19iT0}1`J2O!se8f&)HYqfnwJu z-Du&o0Ho2tu)%oJmL<=yXe$|QX;zIlc!fL1z~-DT@?|ZcC%(O_geO;Sl+?|#MxPI^ zdfa3eeH^CT4y+YJfC=VsG}>&zn#cd`hMFfh)z#I_CoYP}*7~IBZmB>_nr1#~i-#V@ z-oz@1>#AR^(owuqeqqxJOW2YHBCm9lM8(Gg0uT*Qe;AJaq2tlh${ zhEIr@{S?}hw7q7AOoHw18DSQD5iD&b3|oO!pRj(RoT7ThVsLAEH^ju#X%0FFVLOYT z<(DZlw0`H{I{oa2*uO05l+~YH`-c(zqNP#up{cu&K>UDM4ocme%!ymjJA{pUc_zazk_eDO!JKMOzv{t)n=mxd_8-Sbrc8$hVawjzGTFWkB8ipJwU^2Qa2C zVg+GSf&kzWx_D$ue>xNMQKQ*z={~ckzK5+?dWM??r#82@)y?D|FBGQvLmK;cPVSC~ zQ9YyF0{DkXuC}9Z-(cKkYNIAr%Vt+fD(UEQ6ge!16l}Ymr5{0*M>jkk!eafor}1hv zLZRE(*Nas07a>265jZQof!>2|Kdqm4Y>U06{R>rn%a?EA4&|lLu4$Vgk6~L>K{yQz z-<& zT|)8Y;E$qzXW*%|VOzWEbKUZX6&d?X;1^@tckU}@VAl>vbyplU=Fa|_)+32C!vCiJ z`iW)Sx3SqWcAZnIqfifHsJ>t>PnE|KSQy8+?%k4o8=PRvE|eXe76?GSDCPSpfZ0G~ zs;a-Mbucx5gL}LPuQFE;PTN&b2pT$n-$fjR(f^FUY$R$Oym<=yB9y6aOTQp~?3Ke1 z!=ok(hSN;uQoDHLc_m|a9eQ_8viywF`svi%?zl)ERu>s$HYjR1Q#hH;QxVCKLr{Vn zJ>7N@;HnKkfe)y!-YEd1X_f$o;v_e<$(579ev{zUgE#?m??`;RbtJVV{4PUs;z;CD zMp#HS1OXB$l)!~i?eD}~{zi{!bH+kJ_VDad(3wNCdbR6C8?bB3IoR@aeGrz+>q=F)q-c!y*1Pf zV$jWTN9pm|EeGG&^JTlp!-0vE9Hy-@ER{q(7xu8>ZzR&qjBrKT@E`l7Rz^Y$FkWtK z?k{Sfj&~dbe5=-u2CKqm`|n%4Xv1=H&JCY@)l4O%Vx)07RNb|>xiUiFJglFG81*F5 z4RGi%1xUz|%$RjRV-ePo};rXvp-L?szyMh0i=dYtP@jicYwtP>Vf#iW|_4MVrao_U8{I{F8 z*8$S`UH`oiS(4+Oc|P-RTmJTVV7_4&ZI7#=6-Apn^swB$4r|nUIumD&-fGl}-1ZGT z(EGwiG`4i{KNRrNM=X#Ytc&DZpuf9&D;mr8A7l5v@cVO?#ed5GZw3CdCHv3R=6^Cz z#O6{A*`{hl3K6xX3#pBMw^rcD;q?B4Ii36&5tbKD2qK*W_y0T?zfF^OhX?OZs2XA#7dYDQ| z#}mVbHFTOHrjs^qg1M^_-}=Rds|Il})z5eWc1$@sBIh^Cj|RR0-mIod6;t-uHx}Om z7M!ErVR~ZZcJ%S6th2B%*Gfl-x9IWheT#f3AJiK(k6cgLU3k6n6ro!}A;OTawKBjV zFItGZ^m@hH=vkCf)Rsf9mnvaSi#MCi^sJLu-XYMsGz11wyN7z-R1yrbh?LOeK@XO# zu<@8pMNe}bI%*$Y)+ElJq1Hz7&r5!9)qBPep?T<~^eH)o$p?*p~=C10^4REv^H>T+lj`mhyl-Oc)FxqbV+R(t3au1lBa8Ynv!w0ky41>&1 zQN;bg)Ptoe<_23nkP_o^xjJsoom!iV6L!ZXUA$33p~!^nBT-a{%^n%q2^tqcn7TO0 z-uR8~qZ6!&pqjVd)Wgi;%8V0p^_26q+_aiH2N`o|@~D2XLLXa&ZA?lvj87Jvx0etK zEp+Tl!ZF}dzWnpr6@+jWn#}+b(O=oGM%Xm{l(%WF{?Z)xwT3tPtG>Av`R9_hZr&hG z_W&!DtgX9lbm4)E032}i)*ZXR_+3&RB1_BrWCBunP0xWAK4p`?CK9r1rm z|6`nWYa_tgzqdTP@K2pt8-Dxo^ZvSot)1%WlnvchlYulM+F5S@zCe}6{$GWLp^vGb z6pbiS!C2N{T8SSsFO{+M$5Twic!aB-4i*fAQLiOiPK%VjVW0OXo-}>=d^NrHJXSQp ztZ2X9n!mV2KT~~@`Da)VQa(4s!hB{k6u|5e*W=RWQHWHolA*6hVU(DzM;7Y&VR<|c zE}vshJyo5)By}}{&kqYgcuW>6JKPY3K;BC|e!w5_RNR6IEsJ3K8P>X5v?m%E^`(!f zw)D9)87S#EDGqD)t3$0+vUfpg;V)&L?`>-Dww)uN@h6h{k;6uu*nJkK5xC>81X(o3 zTz%EYRFrGoeF-hpiGW(M1Gpi1{`!D~uijS$m^3Ioz?Md1`QR@Sn#U@+;y1h9-7U|T zzV6FEK&co0PS4U-x@>W!X)U9FcU>gRhW|n_m0ek(EYTa>729yr)l((-BrZ~LA(msb zVq}GV-?*YHyx+QRrd1+;majpt>jkdWOK!AP^GojtG$N7t8s2e5bQBf-YC5k&k=^bN zmHLnA(zh#>!qxC*YypB5S4T4V2Xoq^2$-p(_!}chPBo}<{Qica>d= z)URkhC!T|PU0x*r<$Sii8c12sXYW*hmcBe;+lW~zxIpD? zUAZn|-G8z>qk-DHp{j<|USx4du9~YAZjTzWHKBO_YS=}cM}^uF*rW2cRV|Z7xvL4T zoAfBOQmbyAmh+YS3lh$O&Z?uflyoBT4Zj`VX@Siu%H2fMy_ z9CUBx?JkSHeIp6y2bc+YHQc^|;dG5kF8Pz*hgqA!X;pppvAdi8*yytvlXMBGmnl31 z$M1;%eCs)V9&@r!%56ouQbNYF#_V~6(@AZ>K9T^a;+)rj*_yxbtadGN4AtxR)oeI* z_Vq%%PqT`GKg+8=Uw}E_tPHU$#z*j-B)d@qn~$n2Wo~8b877t#Z2Hbg^7dawVTpOE zETwV}BWr!EH9f3h>+2MKobSJTjkrS{ZTa}$Hg1iHH`l3u9c=|j_Ptv(*A9Q<`cN79 z?HunnlDGSF+RaaA|3M~O4|ioV>*bn5c{5AvQPsm+WQ2A3UzWbgp4l63>+OpqwCW(q zYfrHDO=`CbQusJ6lr)V4nszozKU8d6ct6*bz7dKC0I)_+{rjrS^&#a0Qcx($)=g!L zBi-N`SK5Y4)$J!I1g^}mmim!Wl-w-3H{VcC zfj=A=LE%LGQFoO)-KgF_Nq9*Qt{6!~V3;Wh8{4HV@GZ-TnjQ|EQ&Gt0D}!GntL6${ zQ&xqGD(qU1->lH)j@a3F<}2U&6NVcJ-dq+3fGPUS6S>`OM}|hDKr3&UbWwhC1WBsV zcR;J`^+&kBO%5;Xvuxgl+*34SO4y6EngccOm}7U_Im^qn()C*R1|f3z<%}ck((CK+ zBU}8wJsA>iSLXJda#N4vl_vO2mrAsRbBfB0<)Hj%d}YQ;KuwT--JMx^r_i8Ah#-Iy z9j7bK{KpBH#>Y96b9y2n4aG`N!Dx3b;EKOp(4LnBl634Eo+4l`5!&JbS+UUWnO030 zL)gEn>lzW1I%C}I#DcD8hrsx$;hX9XKLAfzta(0Mz8+kpbIRogOq5WV6-wQZm$(C- zc6S=ypv7*K=;)CsU3(yjSC%^iZLm|W22D(H0Bs0dUhz}JZaB-6;ujnomsy`Ft*@2M z8a*6!o0$;)rU~uS6<^YgC#!BPJfxk1jg@4=`Xbq$_BmJ|@hJPEQy;Gg$Ss#GP|y9y z0i>6Da2|Dkp1j(4+mRaOnr+^0+p=(Lb#dRZX29r#dn(#qHElXv-PKZ4&F8ZK>kTQ1 zri{HrUrM>8mK0{7m&yt|gtxX852=ov3%SNT_71 z1}-hTWaTgcjJq-?r5ol31X7e|35CA2gk(z^2s#;a+dgvpR{ou7{pz3GpLHbk3XR57 zInn5kLQD+`f&$V{H-_a6q?x_RM(q<0*{z&Ej9O^@x^kVi&=92c%{Wq)FZ*?0YnZ%J zMW(?;dP0e>;Uuw&0Q{TdG)xQ9nJxwsL%7O;y^BnWhE@JAU9&B;AZ;FeMQL$@ZOr(UDU*^nhFMSSV;i!^kLX!bW!go7h? z^5wuPD&OIXg;tDo?qum~%LUrnXX@^Ud<(h$*OIDh5$-T?7u-&+wdouqmN><*6%QzE ze;Jvo?;V58Pr0+H{@e{sbI4q$V&#fLrfa}j@vwO3a>gP- zwU`#$Jv(}PQQYo|0UL4JQtJoiyC>!JMLnQqBYqxdilDZokRYggf(J9F4%DS=Y|Pm; zxx|^>*NK8E?DeNL3)>`qu~Ut-$msagtG#IvF?!U-ypCe)R~lt*=>WczgUM8Y@=gDD z>ZL{fOWQ9HVo^oIOtPC^1EuR*rzgsFE%jWQEoqmc>X{%%Oh2{#b0b00!EF`vk~}Vj z?g&oH`nRB?m#0w()oeqVgVGt2Iqq15yGTuNsfaOV(YvTjTaST^uq{WXdj7F8j_!K} z-Hu8E#jqDjPKtKrLxee#p8hXebf-;zm_xdVJy;S_nN3Al=ZQ6DXothl*(6+Ti>jo$ z2+Q%G&$@Dd(f1}Uiw#8JAk&8Tnh%W?J{`^V5U4Z(>WN$}h*R2U3yYoGMvLSgbn6Zw zaQaZYlfamdJCoZcZ%Mqd&bzh%!v~SSY?EN<{UjyR_ttgEv)<&Nje$ z*WUPT4vpn9#c-;IC^QOp;xDGUO@=X0mCfRYt(fTo9Z@qR~DN8&5$b~a-J3d? zI>VLYM9lBs`gGjrnP}YpS)_{>+F2O^TZ2SU*DSMd? zP&88~DsK+QsFz2r7Xb45BrluKZCp*%B?7X86|^_Fzwhjv8j(w?C-m^pq;2kLR-g?v zzJ4-fZ%Q+Lh>SxQD^A1VcED zcCgiSw8!xw&DPUql9?I2h;gj!eKw}qQ(r9r1~uF+=&~79)V5m6pdPiCn7)FvVi;Gf zHehGNJCRz*<^Muevo=r`&ybf?e!)CJ`sACQmU+L|E~$j43pY6V(y+chqb%eqdEuz+ zOWrTKqt|>B;}P`DBUt5cEK_$~8guVC4A(%l{u}Lj^ldPf)@*K~+!n?nCA~F>uD}7a zNTt{^$|6R}Rh6ZJhkm25K^#eh53UEnXQmEgW6Qs>XF{g!?ssax8b{7EsT zQS>1Gk>+3Y$c?t%*(3bEe`scsFbF|!IR=ij%-4c6v1`XuBSz$LsLfNKf-xcQQ4)@q zCRk2g4zqzq(IPbq-&?gdXPH&=X2W)eaVWnCN1)n^u-Gg?lV*b9oB{rzumM~)S52j) z_Qqr$K@hVlWQu&o8&b+bN#rmJ^72DW?+Ws)htS2_`DuVu7P@AU4obn1XTIfSiHV76 zH@F*BqtW{HJB{ZawzW(V`-@tgG=2u9-aHzq`=&H&B-c49{V(Tpmtwm0MkcvuHB{V} z36~s}K{0!ZN6VGXPB9P3^|V~>4)X|9=}O`mu*b+`GM|nfsXdCM*9vzjCj`m>*h7hE z+yYo8^ggZhOuhm%0q*&OMO=(i7PAuCE?=6WRVtPP*_pqWwCbfZO74o2wM~|H9_|^o z%zlj4fYbas)CQ>1BF)x{t+XR_e-=pm_RD=7zn)l@#5#{u_??Z<` ziM2PbP6p{Z+pbmi?#>b=S06HsBn~MY9W~{Z2Q%F@?^{kwY`F#0xv~o-%f~H2G2z*6 z6?`MJg}dQI&r{DeM;s)k88CimN~RmnEOwtQq%P!xzB{0ezScc;Ef@2T?Px1E3Mf%X z|D<`YBNXE(iqojV0g9A6L?HLGwAc6bU>u4vs7V(IM|rx}7whK;mn%6Lt4SO#u7{Ac zYT3seVe9S#x@y(J+YQ^3s(QKvvmq{idIfc+99mWZdd;sjS?=8x;+gGd&`By*7crXZ z3tDy@1K1Th3q18iy_9!I+zht$+ekx!bNz6B(LYtF+-@qg7>Za3h~Ti3EQ(sUxZ@;0tJN=3Wvq&n{a%5K3y#+dqHi;bC- zy@Cej=KNMo*iAkupsBR(Ufa5cGlWvIZ?!gt=MD1lKj?p@1)#$j4{b3aC;Iz3XS#ZG zM2Ss#Kv&!k-kY&pJm!f#hTXCVYixvW1zpsiJ=MBe*c)xkSIEyG9LSNBnG^ouICb`@ zhBI{Q@Gp*2H{Wyw8Vd!*{A92^J7h@8m%mFubW_>)L9@?*2s*InGCNfnxeU?umWf%~avn74Z8rptm zR%@?$&pxK#T8uu_>^?4f{5o!b=e_!=$5IKsn*MwL zh$drF|F~W-`s}57G(10vKN+1Ra2FQn{g2~6+XCWPmn9I{>odrviS~7wrQ&*!jP@hs z-@bQX(6n2+JRpwe+e|DvLW>WIo69uPn1`=|bjP8Hp#x(W3iB`K?`t~0>mf%!B=#z* zaCvtZL{G%W`6LbnP(_r_EyZJ&RiHe10WC1e-545(B3d7_eOmg%VfV{8`=0ifduBOP zehO?=Vvkp8!ZAct@4L5jrC3Z&P{jK!{Zh|cmwSeW3x<)H37STJLESAyY=$t+y+1VV*Tjtxu7`la9 zCA^nYHK9vsymc+5_oZaL#IK%mAqScxJ@Yg@pR|&8L z@SG-EVy^OkKHdhX&{bPQQM3g=?;xEZr+NYKSJ8Ofee;2%O4w6^N z2;?v?n)e2;SV<};4DNGfh<;s!p$QqoA7cpV3z$CL<)_n7DTCMI)K2l8KAnRnHxhH1fr|I?3}&d=ml@iu zrd(!q#l&6po@VMjH}7|?;HnA0sJ>=O66|p4vcRmfJCa%qevtHkCF#JVY7%EtTc{5? zMoKB{iC3F1@p#fml5Xd#yYDW`LdnFfoZcC>8nIz)lSGR~#Ad|k*XJVO)`pj)fJCJm z6vWiT{t?f5ELTpWJN`4V_q%5B849SEmvEok!UEIU-2%g=z=UYpGbmcF6eVqYBmm^w zB@PYwfTpa^*~F9*DCRc(N|8sL|+?CHSuXe(Xw3q4|ZkCxp1c!rDo7>6L{d$HQXF)OoGRPSOy&o-)OOYgt`MM1%C*kx!pp z>gX1J^Yoz(l(rkz8Jx;MPrzW6ML^IEiZ?X<5F5$|M1fBRH5^IHND8bTHCXz(fz>1B zf+Tt9yegMmY-u%o-4mT&G!gVBPfFvkT|Rltzcs;JZa(W|TCH6A#BTVO(H?3>tUi98 z&N(Kjdn?0ns9GOm&~}TxGul}m920F`MDclCVW?zwY?IWRM!$DMSm3TIa{5wpIL4+y z{$oLszBivf3W#-M=tE2PwCs~Qz%%8NAy3*t8@-`{E_1m}AU&h&;hQE(V4S=;xtWZW zFv0s*XIDj629YsjGQukMzPt>mg;lasH_X}`YJvfc*rm6o5S*S5$=M$6dM$%Y-f!dV zF|W{`d>P41D!>(KoNDyTSpCkI(WX9ksDX*f#Ym^@uEtmhk3_jEAB;@9A;5GO&Um#X z-6jdWS-kxDH0%D^vn07wyRSa%-Xu?IzAYGgUyFMH0($*S^+MhjJA%;8)V8^QBiS#P z|2pWMI$*7ag9^v!>|a~?m`9~zy_pdQi)AI5j-ttNdaIUpFA=@># zu9y1}6iWT9HJ(C$?}D!snHda?fEDZ{dF|%Rm5D`Q^|#ykpF1?gOD%{R4yTKW_w8|& zrdQUaR6T&Xk{FW!SU}1iT|Fuy++sj3$9ENLdM(Jpyu(2{r7e|k({WP{KbT2vp?#7 zjiwzqHkv@1t_8^+R|ktlS1(pnBc5*JFKRC3<|3a?U1Q6J(8YR`y-Hx}8AM7UiH%vI zK5Q)l4*YpK6@{a0JL`+uz}gqTkv^WzoAyL;o$lV3 zT-1ha4_^LA(Bl&hm1IkA-rUz&z?RT?xkG02Vdf2Vr&x{`^d%%oC5}C$zq7Jwx4+GR zcYjn_HY73(VfX*BuzS4Ldy{V0a^T+8l1R7-Hu~AcDcxI&pt`o!kAjkJ{Zk6&NY#7V zV@AIY2F*0;v7*2@dNNc#Y__ZiPGpuXMWK#{Q$-M74+STY5QV`QFe)HXD-X>2fUZU#l3~$YG7P;pbXC6H2nU%<2@aly3)A% z48|;`NW#TaVb0O-HU-*rv!8;$YnmJ*tzl&pvGweNe9(NG+9Ps5vHMX;zNZqX6YfGY ziD+n{+N_p6>Pi&C^`q%GHg1C}6|{V?>_K37X*Pv-lt}fPXmm_@cW`KfC$3Z~ zGAWZ-H;XJiE z&2d(a5f1hj6upI!{a%IIZ$yhi;&U#`>yd%XHYv)k3(9$pK;T}6IqrOQ@{pOIiLigf zr*Y$Ck9T{$+`FGJLeMO+v4io+b!)u^i5@>lixH(Rk0sCMi-gWO^kexMgW+)*&`hJA zO2k7k5+|4^PDymdl&y_dk1WO(;YK|HAKM)^UG|fK@4Y=D@J0<}*j3PNJvKB;y5`Mx zm}(HwBt!U-jO8x~{^fAc(1G3c9%&y}F!88qLE1D_2Q?#tEDT{W*9GQvSnT)~?@?>> zv!Mg2gzyEII9_>T%`RFU#}RzWRT;m+Ikw!NX98F!HZvtRoYJsWJD9^*qKISb)O_qK zBpZQFRREh}uE9}_rb>nMQGHHWP7iQsHlt9VXV_O1V8j2ElRg>aVRt#xtrz4RqdNQM zF|DZWiIXZUNh?^N=KWK^LbOncr>nB#AanOOMZ*liCu;oMyAe^bgC?=A{JgoAO;#me zyqL)$!CMrr`fO&PTx^>(@If(f^nLDatyt#86Mk4fJ+S?5pUUV<;$AtvTv>|FjIPv( zwSFOC zIVvJ6UYp8|wgN9T2=_j0QVa_3^q0_KFX!LH*isDg|227*29)--!*Tg^zff1p8;QX0 ze`hB7AVzbBw>jz!Jo}9_FK9sgYRgs}+=Vnq@h7V$jsc)yy@o#PAmG2 z^!kGc=Im}QUy{7azP$SW$_LXQ_!1@4s&Vn_Y^WMadlA~q03z!=HCc<;Iyw9({|hUK zEjtPt@x<3@VB4$Gh6(=2_8~t#2SOji)s}awHxU0?tXqy7-jQu9J_YC$0X-My6EL@N z;RSs`48KQ;2|s%epE*bVYVEM>N&KS>Zty0VWQAF^5}IIy>}tkS(fe-9@qqaF%OQc! zseI^9pC(NXbNxbpvM|8O+e?7WfaEe2-j9M|0)l4MwY5-&kA6G+cJ61iLX4V0;SAPr4xPV@!Ifdgj!S9g88k3(sN@-)j__iSObV-cf~LiwK$H!O{ybtpf7 z=&8S^y&O3B=F}E%`=n--*+-3jeB8ou8_hr1-6u*lwNs4+y?bl22eG%gZgt-73Vw-u zO7oM_OjZk7G$`d~sLS-0zgz;C7th<&COY~_+qmex6cWw(Nx2at^SgKs=T;XN6~=>I zqZtNhf4Pk>llV&*CXTDEr(HJ$kXn=W1z^sjh3;vjK(NmU2+eIkOFWHMSg?>$dmfGs zsp6Gy>c@78KzN-e(K}K4xR+}7m08@OXnMkLL<-!GrR6TfCAWwp>`hlPClWL1IW6=y zSO;tPqc)GdnU-=c2FG`_me)ITP%o)Ufnyb2hEkI@>zzu;kCm+mWIEqAEM-m7sVssD%U5GX|cZ z5;?9SKRLuOHVQvnML7`fjLJHEA`lTN^#vQzZ9OnQ<=}e2=jZ++Z~+t*Tk)`asWCIP z-K5V+zdi%XsYC!Yg-&N6!Ezn6P8|`TfveGVvPl z9LBnR3x(V}2nJTLq%N1a8dpiEMljJkd`*4H#S=`QkhBOE37HCFr3W}QyEotv3SBmv zu2|`ObRYoWdcwhYO&^O0xMJsS%oP($Jt$0KbjN@W^<+~n9Xg%;942Awo`qp03~YM+ zU#-WF&_W^2hGZ-&_y8=%aULR+!(`oI6V>aMkhitwTY4*g>rlD!t z7$%Y6v1*Y%W6&LJN-bJAZSy^5+03yZajo!=s)Q`QtMD$6Ec(zENGBUPm`4t+e`6Af z335S>VTyiLUIuTVNH7noRzp>vm_eNk-8M(TJ9oR+k0K8iGeUP~8gmlG;ybi%;uS6r z&(~AlL_-m1R1Nan#julY*L47dKSiob`fRVqB}5V|UA-OC(6*r2aY2G7pw>B&Ln z9?)Ld^gfxw6lU^8!IuNMWGd8i5Fu2YG4M`XoRCGu3m`1`Jry;;p*`r2@{aFB#^;jjvtu38*tgDJ{9xMG=enUWAwKPq&>>Tb@%T=7EL-l3 z=8d{ut2i7Jh8L{xIu{=;|D!^AZLua9x78T&dc@Ag1x>EBL)95X7lRCc$t)6A^P7(B zf;{$oaC&@HL9%3!do6}-j`Dq`hnu#2u#0IP;YpTKWRi*7@n>r#7DnL$l*7+hY2gYz zU0;~Sp#COYBIp2*!fYh){2TJx0jn|X041E^Yxmxy zKx(|h=N*}5hMdG?7nEm)BKohe6o^F2algCun{R|G5AEinR1f5driI$DRh-Tg>x1hRLIgS;srtb0_8cjbIbx52wSi7TW;s-3c zp};b9pHpz5p_{F5s8m5s+)Q4yIGMCS;}`}qvwa8n=@3xYZp6Yu186900*WvRmS8dx zEm3@LoS>mHla@e};u5h<4&R9sX@W0E$&7joZgb)Ka#=uvS&sD_O@TKg2El`xjyC-T_mQ?|Wi#bvDB5KfkA<`T=+B zW#FSo#5xO@@LgR{pnHv+V~|g@hz^$3y0Y-Jh}s*kR{_xH+4M60+G_b8#+Jjxy2|{i zhy`7TGD2;27Q?1L%PoPj6gj~IEok>#_G{u>hfDRX1O@&QkocjnYvvU{7n%|?bfaOo zgFD+mZ&)9V)im;8ZX?EpA&0qtd>5X;{bq@gkPvc6Ie1J`0cV;>b2$8Xqwd7`>i=Wz zEr8x zP7&o_uZE9gD)L#JXb!R7-MXZv8wOvMOGR%A4!02lOKOk+ z%55nm=BxcEv#R61Y?BLa00{Q2lpz`vS;)<5)TSjKDyAq9j~+6^l(JO1d~p*Yi${lP z4o|v0RkM>dKuW}gfI+Nc4G$nk*sZaP1tu)%%8q=?R5F~NCG6%Feu_h|qJ)gYLkU$D zvI8ewOL*DJ(Uc6r6ivT@3AXAn6s@M%+qh9qF@c|*e1jsrHYpG^gPlKxCT!nv1Q3+x zs^t>`?&*Dns)f4GB7bNSbW=mR$HXF+?|lX+Hpae}h?QU#+yDS*FV<8{^ zucyP2=+OW^wl9^Bp$j+<) zn{KYZ=bd>mEQEk;5iO#I=2ujh{N!Mo34?mQN!0R@)rK7UR+68}w=p4tDaUvjwuX5` zhRLT6XTiTPCwEZYe}FedKl!GG-7|cf1PBt#vx|0gyehIa|NDn3^8vS+tm)o0AYMxlJ zv4A_-w&vAay956Kek?4kY>YuHL))$$wH=pkO!qk`-jEPoJJ{U_u0sCgqsgY8tWz$L z0&#ip{P(J%eolqT#n#)B$0NrUpFynO7J;b4xKT^|V|HTiO*Qfv+g zykS-NgoGZ>DMx)(%Q6Bh0sQ z&VTgB+g#2bOYxf>lnI`S((zm59m7FNt&U7qdrN4R8Jmhc(^PL=Rdzu+NfMjX6`mhb z?Q5(G1Fhebp(5R#)T?K##tHkfmEx#<>#J6_lcx=%RR!i`6^ic4JjLsx)%5UrpgeR+ zt%4G$T(GWE1*;{8m8vRU#%eTtXvKumON9#SB=y50bijdd`L_e;jLOSxIMHh&scWeA zm8$kF(9Cj9+F?;_sSYkgev7kyzX9gDtA^qPh~Y)+t1yVHsjbs^ZCL6Gg)I~^l+2*O&||ozw|09r;ScftC7|igcyL88#Ef)C&Ecj8u3+hCCzoZs2H^+%1N{- zA-D{7ig%mtHVltSOv^FKD(=1Gi>H~M6?Hf%N_Jji2v}%KEm7F^MZDd*sFF=!yLAQC zT6*Qpa%E#J*L#5>_>x^P$baAbD6TS5Xy93PnIEL#kXY~B3IxLcqNM3!tX}=Ybc~R( zPjX-DXK^YC-%Js|kUFBr+%oum@*4C~7n_-F3h#QYZjB6WuMq#O{as#_%z@r#R6Xq| zK0MLEWL=vt>Eeh3n0V;PRn45G*YoS2LzLZOlDcmzQVCtlCn}uHzt5z+h!Ma11A38z z!K0IlR*D)*lM+zXL1h>BP7PY325gBgx*q$N>Wh{Yy#}`7=lP#1m5Kn}^miy!lfy|u zLI%qO^6EIn70H+hCZf^yj;eNOo)DG;9Nwv&u@xS7XCC+a?C_3zMq8#;%U_+OQiT1j z;UFF|cjWU_H$DgUg$P5jNvp%IR^Ae>#$vy4-4S*ke5f!k2r54+4)O3b=Rp-?npgZL zsv`1g3LO>CY>fhWn@I*ONQX74ce~WYKaL9$$Zvpb&dap6qAY9$#5@{|XH}=eD^X9T zxc08W%VrUeD&3-E&J)L?v7lg>B2P)@$!6Dz1RYx*nyoZPLBw@Du9aAh_3|n5Rr!JU zmynTN$ou9#H&pB03P+*{hD2a>Yt%&9TlYbB=g=ZK6iGBpvn!;74+;t`(-`nIbybio zVtzG@EtIktd_zr`CQOX3ITA3Gk{(#z@hq}LyO+Gs5}Z|%swFGF^&zh}Ab9UY=-EJ{LFatrjghArY-x8981u7XFSNrf zH*c3;o!0N!>Z2+I+y2@3z zdl@vfHsq)AUjM>lO;_<%s?vOF^x<4#&%1q*2zTNqQ^r|buJ&CNEdMKM6YQ`(d9)O6 z`YP+5*ovgrTR5Uo`;S+1ZWPxyfMQgy)WK4t!=MWM&J{!6<9$+hxm^FvL05i4as^Sb z9hErcb8@hB&Sc&SDtwG_+`E+a)yymTkG58#t+E$;=O03dO z@_EDQq(LklOg|zb8#UM3A#PwhmB562jD1RwE+Fdct}J=VL=Lu|y4|6lnEtue;8%jX zaAv7y2v{yI&!mVGO}2IKR3JUk)}U+oi(`4vK8j86jy3}?)NDC|^K6meEK=M{&C~MH z0~JcO`EP)dM+7RBIVXX6p4f_9Mw<9r{-(e8AbtFIKA++(36p3`?5y9+9!kNh$%hgvJEaoX) zPN({DC~hc9YpoP#cId22pmuQBy2yNAgOiMLG_sBIaYl5{>Q`X^s7sa*(KnqxhB^H+ zjTcOz4eSlqzaw`af_z^3$t_wqzg8D=AsD^z^5Idc!4T3)d}xUL^tzZ?TPM06tsM!VKb=i3$B z@;N9Nb7~#W-8{Oww>~>McJS)w)Xl9O{RbV2u$QttM17h>-yyPXYYR&To$O?H$*nUT znMB9$fA|*XVk|?QFe>JvpHbJkc9-2pwh{5n9#00GlkQW&7eA|6)Ra0ENeo|L%zLwI zz)F8WbiHs?!w7lju)2~`0L?LiK%<~JA!rfYf0Ij7G3(Cb3l3RS01UbHK?x?y&joud z)=VJaWi_EIF3EEd+hmRqwW>`*hw|x;rwI>W1m{7!QB_i_?ta8NBwv@^=k!FEH#SEu z@rb15fV@Ryg1gY4T{%M<(V4tkuDhrkY;zv}U{G%0^c%q5zpnQ*K*DQR@hJ5|Hy(Cp zE5AgBNn^)dvXsTaQ&~^MojNurv7Ff==Q=*veV}=3hu?vauUm6`uAIu=q{F^zjb0t?nWbVk6^TU;5ETwgtdLwh2(z;A_&N09(NI zQin?o9iEc&hU+LF;Lvq(j+X$yfwr+?5?IvTUO6Im(6Vtv#R@eck}Qpzbmi(ZU2=)} zRLfh4X7TfDtWPNaBN%Mx@`g z=%?TGcea|{Ron(~j56}riV}H^r9c24fF{N!{&ZMafA0= zz1@^CV+P8_a65+My(pHjtcnzqVK_&p+XrYI$BEImBOLiG_O!%dT7II`VL-)CFrm<` zB{UX(BR5uWL=&BI+7b~JgVLr{DwVtEON|h9K#PUgGk^Fh2}k%CvZyPLvrY`R#4*O6 zJQdgUxrnw9^1Wa2adHX+SdA=CQvs|Ul;{@9`98Z(*U}i ztaYYBN7POYAx?2WmrtfMnP_QR#7g}^D3dfcj&1MPdeq!8n0ZDCwY;*dSlNrjJG{&? z(MnfqCX)4Zr_s}J1);(h8^1%g5qM~1WH-h9`kY;p&(y3niFECBm;%E`bQu4n%R4u= z^vKyqPAj3HV8uA`zuYo=hV`48XVD;LzO_Nj^*z;xE9FD7A-dNKoYTCG2304ciDJO+iRbk2J1VMi4b6 zL`tE>rNP0VGmXJvLFUz`8<~1aAw^EZoUKJ8dJgq=iz;Fv zq~TBU@He2zMEW(I_P5mT&4f8u%ZtR7lFC6Z;o!g>Bw#p=YVL`bBp%o#F@d$ai0)U8 z`USsRyWIt|=C$OS!sfX#?|#JP>jGfqe=)Ac4@?MO{=pi#WAevEK5~oX`@omg@Q#$M z#d=>oMT}TMMjW{l@^0MNyhfEpci;Tul)nM#)u>!j?=mv2+pQlh`$U>m7pj27qSjo3_`0sJ76N1Sr3{EG=wv@vZO%uJ0zB5z z5cQP;MCcgK`6zp*CQV}*{8(N}qZW~)xYrgJWOBqQEgBLOwkLooXXzHZsbT{hA@q)M z9t>#mnO#jE@(KwTiaP(mrYwa2wnd4xcg6Eb35QfA>d)7B4d@g3(W3q$B+QRpik|Dc zIS@F>-|VD&pc4S|08ur@4nhvj`*Z;PZvA#5hNN64Mf2BM+3B= z&kA}aT0nJIO1v zN$c~unV$g@OeDD{2kGGeH{jOh`@d_+81T40-yCLIE_sjFqVlTy8VN05T^ zN84yOT{tYb+S!Eq>zwE1YR16G^`EhUUCsC#kh#w-a{TLL9d0VAY_N{>$$vlj z;?vN#w98ycZ~ET=c)*ycy2op&)ONT3G1k&5dY%KY(TIkO{>Vgjo492FaXQ_}<~lL% z74XcNa3#D*Ijkf@XrpDK4JLe4YeZcK8L+l%M%C&}xR^pVgX_Ni+?P!ZL^N#~QdL${ zv{*{zbVQ8+RFvz$bjXI7%8~R>8Xs*6x>z43=ctr+mqrXv<<_kSESvsrO+uOSpgrA!pH{>UMU?&kP2CT8IHU?nR`YmvAEBbZA3y?g*7QV;JE zy{n?eM==)C&T7ds7d>ZVM{Au{ftA2ue~=g} zaCQwA{78kjhFUfo(NmG4K@$F+Ayv_)SPy6Rz!faPCj-p`XjgVo_h98TU8fpGYmK+a zUPlCT5k1sfOTgiucfeQj=G_x=kXiU?2Dy<`r+;w$H+{{C##lzR2tnz;(dmZPym=@2EaC{MTuQTbDM>K-c z+{?-aRn;PS7osrrvkFDY@Pf2l1q~xPU5l{j#OK35jltC67yE|gKEHp;edu1Fd))0Y z6k!R<35pvd+PbDIe=M4Qo?3-by%g@pa@r%}phT;$^r8e~ISmQ-F^)v*k{@gVoztG5jr8Iw7z z#Xa=S#SuB)7*Lbkq}X_h!7o&tK2T~g)hDpeG4rz7E*nQHJ%p>{Y%zVpl67yM+Y8&! ziWWPhO~0081sYK%{j7Q{T6e8y(kQJ42L>*Xv>4k z8C*QyYF6fYU3Ds`oa44uTqT4k7u9UWUZ--(Lekm&fRjpJXe28(ZBm)I_`WYF^T4BQ z3M&&Q&O-N46_X#5{0X&DfPsdqwuwoWa5>{NT35lIOd@lY&Ez8w)`2=q)K|2R3(9o3 z(9RmO&N;?u;cD4zgEX^&{^fjMS+ zxe;thv-|Uz@^K?)sF5{gkcWILa^;SSZcou3-$mq7b9u!&F{!$86gWnBvU+?ZObnaW z@28DnwlNKxhFfPrd&dG%UdR?XR%16XAj=(%nU(r(dEIwm)A%TW>;nP|vIeJf?Oq5mD5g16xFVClnsMqy=lu4H8ATDiN2jET zjM8py58Mow_sXW}E>(7)IJE1Fq(*cq)5Win4(kg_Ra!}`Q71#lor@6!{06LK@>@-g zNnbzDvzUe4Ubt|4lxCIUph(+MESRsTo z@5U%7#F+YP=TF)*^7i9e2RfL#p2u3aYSu+xC!ob1geL(1(l>$oq)CZLREW zEY-dUr7n**fKQ+7huTEfDaDh-pAU7E`wbw7e>HhW{W24ww){xGPD(Og&9~{b0?;qL zA|HB2ej4uns=s+bZNAyOM8EWWJ)Oi)!`B0g zFYEu&0oPyXm#)89|F<^6q&?H~kWeYSPF@ia%eqsYIO0j=n#fnU?o9BEq_TIL+WckM zn`>ale$^SE{HkzBrN&DKi7}mM=4^vrttoH*pK#6IA_^3bt&3$_#x9Uet^wga%EFul z+&8*Y?n=jkrxG%|78k()yZ6sKOk1T&Fw`IO7wOS7!nXuuIqTMKmw)h)MAy-0!<&^o zRVg9flYegbf8BuQxYo^MDR!&#YISTrtn#Tz4^R<$s+)~DV!%ykxyHr91cz{t=-_Ba zX{T6~f@G}~K19WJo(Gz}id(1ymKK46;?wd4m}tK9tDTIG#1BgEbFf{Lv>ZMbmCdcT zcnMVh!(zNGeH|ejj368rF;{n_t)N_Vgr3GbhGQ?V_)PJH-sH;0sw`T12&L{WMe(#< zip|>l0fq`nl65$`T$f2|5L>fY(~zio&s_4g&rQh-&YG8q=#VTR2DNBQ%TSOUm{zM@ zyEEP-p08hlH(sfvB0zQ`3&1pWPzs~vdV0sK&|c#Xde0#J_0?)hr+| z?hi?bO6hUk`38wSZ~Q>g{bm!UNA2BF|4~C%%~#0J?Eh}(0`ilnf4rfo=6_@7zd5ss zSrfX%lj7`G@Wr__23(Sg&6~@SzuWQv_>?QNil#?&Ao_3+Q>fQ(K={W6yE^|b3B+w5 zUbOQ6XOQvpLc!`AKB>en3f{k~+DWY1lV(Yu7DeIffj=>P0yj}(G&ba)9%{fV76Z}C zw&+$0Weq&aEk>_MY0&&iRLFc);ZY{YwD>NH+dC7UTVg)QFF9r#R35L`hdfvqmr7y0 zD`pR9%#;z_E3}F{Se6rWH!hRBsGk^BHOR&B&s`gS5NPK-x(EXSNyMb;!usr39Vm@|6K1?S}dI<4Wc%7 z**nJq{ff2I`|kh49bzj9B007~A+11|za$a@HyZyLquC+eez=A^WaMbd3B+s|=HsFo zzngYSTO5#=QywXMgpz zV8VvhDvkp8gnG_1<#1vr<+v5^Bt?Z_&%e3-oYK5POix53lZWo!V86ruXy$gsWYs<} zgliqc-LcDgsD*!(AgwRdqNqr6AtGr@d)oL_7=B5DCG2>~lYhyOQK^AwBC%0um2sM) zIR*|fI^wR`0E%JQsG?3*F*_)4lxreQ5tdxSerAcw`wIRy0QL`;dB%fd%J~hb`$Dsd zySgqwx&4}HO8qY*3C`gQAx7WZOUdfL&LLxb`3;~}5l+X<#Fg!6U_gIYs=^s0bV=38 z6Mc-SCuas*d08M`4rg_Z#2xE!oUH?^<&l^vOW_1wmwB6{2qqOeTTEVx3mYD$d6w*? zWz8edX)fp3l*JD1l7?pbc{-dkL#ooaYQ1%zi12IKq1e@)%1rMoQbIe!hzm7WbT+xB?fz_Ej+bV8Ig;O8wjR{E<`)SU7M|$`TcvOGGPxC~|h#}2Rf!LKDV$wC= zBwkPmccWMPam#G0EkEs!=m+XMxc*RMv4+C+MY|L3|D%p#Rz9q5>{t3={bvkK)S8fF ze)=xdqKsilz=H~7R>3jjQ-B_1RfmO4X{^^!r@dBlu6bYmWAz=RzwJNiC~lBVhhaT% zUPJmLcw+sD2mA*>r8lX<9%A+Tq_RZpQuy}rdt92y9$yrbRG=qTntS*Yt9T=h88F#7qV_J@-Ye+h6HogWc%>!@kVzOD#r#=4BT`a<4 zgHMQ@oG|xfDCC}e2 zF4{QYrPpX@xVzcuA_6r0tfAhIag{DSA!`f?d5wR0q;_&k!aTj4EQW@0f7Z5psd;6L zny=yE>dU}Y&w;ZZ<9ow~ANA|M{y#sYij)ZX4G_SPCRI6BinNC+fc*6n%gQGY4-v-x z^&gaulD_-F^`Q>a?+1NUm2Xh1%KaZcAS9hWHDVVd$7CaQ=D9*T!4$f>OPFQ6wLatx zgTrYGC$bWaMJ1a=@mdV@uT9JZBon2O)%@nKWcpVu^gkRNt{@5cdtaMpU_>6dmcdCM zNb~-a@pI)H0~$1EEx;z_5?7hMJWcMZ+jTg#7fR1+oEL(RIu1M&buPy@txdKaq0Vxa zR0V@68%xO%7Ld@}B*`!xAWmBiIW-i30EvT{2y20Xy=4)b+W}WR2Fd(|`lBH8kgxvs z{tw8*Jj|P@=W_3S>`&E&`m;m2dp{BnF=pL%3Ka?Y%BWkMt__aFQ-<5@=)3rQRGVR} z;3kJCwa5`Nu%C2>3+K4uvfxOO2Ker#Od>)(0iTdN7H6^nJ|fOFVlyq zFt-TO`CMTDWykJaSg{y)AHJybM!Lb$I-p3ekfOQk1Dcvv5j4UNRoWv21-qlV1}V&x zOcM`1OiV|S!?OossV8quM9HrLIOU;XIoQ2q?(Bq;I&J}ZdNY*S#Uk2U>p8Qwq^)E*}@g7%1=*IMBtFVmm zL*iRqul}xd1S>CuUj;YKnSXI12KFpt37E|=E_Ic`)n8LLsx?N~7t0Frq3~=M*`vFA zI|S$|n0V#|)1~`}J(3H47uXAj98QyQ$HcP5o-DaOEx2D2WvHC0(CKYA9=)nj4`dBsGSHYbh#DB^r+3*u4XarS( zTQLbBw`Jo@ z=}ZbJSHGoc| zjS?+BFTkGM#f`T}!sy_j095vEiZram#E0nx^KTLxy4x2aM8EAXB(HR@ht{%_>@z=0 z=w`tyR`?BwTi98Q`oXZon^(NrwG+3H@j|BN%Pun*FAK`uY_YsB#~6!F=-=<$Sz;dC zp;;;rtxUyX?h?$%2+!j_F>Py?K+izgGIqsq&#GY^BB@DTn3JPESTH!q`|K41X}S}9 zgnF0XdbAeG{;kxbs9-sS+#ao2Vn(JWNUD5d$3Pdo4!5WBrb@w~a#euN2CEnP^Kz;a zKqob1&@R2j#~l|0LO5Bjp;(*2;~$twFNN0R#^kxCRerJf$2oX8JmvT~Vj1wMG6+tC`XaJUh9gL#Ae45v%5fC@7$t zJH~ql#=y{D|B|fW%UcB(%?*JqlnCufO$6YA5R&r%|k>xR1noI zBdfd3kMj3!RJOT~G47n?63Y#C%@Q`8Pvw6#f5ccr(FhnL5^EpJf+H|$>e}qOE)_Q= z%cEACyfDE{l1iC`R9nc2iRo9{;#+wzb%$RtR@ zWrld%UThW^#NK{JXsT4mjw#6wj{<6CRwd@Se7fj@HOzs8@R%YTc?J8Kb5|x?1l@hm z;q8SMml~Olz;z9(LB=JDP*R6)d3B5*^6|JMR%YizQ|-g!zbule_X#UpR-k zB#NN+xwBhkXfE1_J|MN68nP0d`jQ0m2HvaL>Go-L{YcK2G zFRPvnX9V$>a%nu{nNzJVVkntVre_R?QpbY%FWMF0-kk$I!e0ZUIcDKFP_CG(<7Vw) zQQ;q#lL74kE9};u^p>BCX=#dq5&Y^FXJt6>s!!F&;e<`M`v-POlE{pSQ!qJSHNP4j4=z^N0#n(AN+`RksExtZPxQe#bZxl-L!M_B>u4gzhkJaN46PmQOqm4?p&# zTrme4R02j#gVvw{&)!0N)V3`3MeooV4#VEqri4zx=1DZ$CMQ4R#5$r-y1Z!g-0>%3 zK%4ems%GKmiWA=JuvgY*WGWtHh=v%9exUB^=mZV@>Z3|5GYQ`vJ6f6LZ5UP<(%Ciq z7+!c-NlD>Zlgl&F`}HMM#@g`g1vC281yAQ_uCh9oDoi}DI3DIRtd|Q*JWZIyr<5Q- zH00JSXqL(z)Z(rWaXlVxM3L2>ZvB85jg9w``zj(v-NRbK-;PVtA-b=rxm345T~I)# zL)UT>_0q-eA~defd-tvqSyG*zpjQFmzVd)w%w4SOJLJj0how_CYV1^(w?|&Rt{OJO z${lO%POCW|L1_3JuhU-;h3pBN#7ft3hjb!n5D4|if$L=IU~N4Zm??ZZBM8?`fYH}h zPTQiCLEEgK(r4!1zGNX$FXz`MdO*RFI7*C`q5&j^Hs*7^*EBRLZ97srkJ8WAA;vz zg^55eZ`uCs=&xSWh#0@OxA0>XM}L+*_JQi=J|4#zVi$pv(x-I9uT4DauOV&NF$(*_ z@N@=M1b{q@b>@5;f3o;j&Gd9NoDoh=cdP3Jjh1e5vbB{ys;%!16VnwD;k$Iv4o$KL znak1y93KL;GS=)R9t)k{FCy9&GB+u$kpG-*QdXjMA*WJ6Y8PhYD&rG~S|XDeFvmDh zx|PkI7$8_YHIknxpcox8+v2^?4M%B6L!n|s3{yT`vvnoXgh>bE%{7Q8^C!}PfG2aJ zkz(%}5BQa!L$MmgC#$OG2=Lc{oCr+6yP@*rQ;C1V8#5b63#*BFYMa@%m zu5|Vii$_Fy+Uo7(5=k=8st3f}VG09Xi6#!gL)< zbtu9&dSOxWitGuh38|7U(jH8xKde?*F)T>bnj&=vb}$@p>43(r0!YLsWFC!@fvzHa&{$9M6W4DH3n0d`8GA=l0%)l zN3H(ihC3sn&w=ZINcYw7bBtTS#SKw=Bi2ZOktJ_LKD7$VqliL@mv_uSgt7)jS64C5 z;uNf7D6Vt45Cu)RT`m10?hrbo1K0z~N8f5N!LLj|rfCiJ5mn)g$dEP<*>ZV9whA1%@QFaqWo8x#Z z(kv|N{PUC*0(k!-et7o8t(I85$aWT4MpPQkqK&rK;xMD-2*Cs(5~Q%QjzWbq45~f* zP`_0HB$b5NO7vKNh#6v`OJJyXl*xBx`JKRG7`eHQvL2vvrinJO zQ4DT>Nku4WZ60*X{pkhuZxrNfT1%?1>VpuT$Y22Hpm%cKs&NhCZ8mf3mS|zc$@(ex z52b(<V_xSuB8ACp8~4f}D6V9K^o zr6QcRP7;M*;b)@IdBdCKoE$bEE!I?6p4#xa=>YFOu$e6K+tiIjzSvwl1_&? zYfHr39PMNTNJQF)0O%maGHMteRzf~sCJ285^rWmd6&M}tXxCa-XLU9`oFyTg^a5P?nE=m7b&)r5 zkkUlbwJ0hHT?1cF*hPE8;baO^NkQ=o&L_{q?ell33r8DBw9^pLSlp=O>JI}n*J`Gv zgI_u4@>r(`vUcSHSm*q|yoWaW-Z4n+R5dR}05GB7EyjK}Zha$1U&GO#=}v`{Zx|x2C|g0EmvDmn2yZWhy%9tGt57(@6$c>M8rSTx`loX3$rp z4i2!J#B-FPMBCu=ESC1X^}Jt>(ZKGxrT-XndJ=@-?!V%-$hhwxLDT z+=RZaQDw&RepgG0mX`mPKn7L7kja~JiSjri--;fao?`uN!KAKu8_$fs&EpeYvc)Pi>bHDHNy6Klrju4%+6CI4E_d*~! za~{inqEtQh763 zh@f%fm1RSovkqkkVds7o(Kv+^=PWgjZEEqw!jFY(H0;R2c*@+~T^%mBL3LD<4P zxuiUHZEJ+VSs8rQ6I`NXgU%n~yqL-xh}htkL-A=OSk1-4q`F9Ih-oJm?lD8OnT{IA z(`!MU0&-j?!p7dpK)c7N5TMNWcn|c-r`7b5lIhJLaXJx(sGnhSGcV@N8gBzBk&QK@ zr-Po=q0(UIGRJQ5(b7_hi)O}4ZTlJ@hHez$ma}SRn_I-%bM&}l(kIbk1^U`3pF;by z{gJ#0sNC4m|&lavjcD^AjLZzpxMdsL$aeZtnj2QDwG(T5Em)jC1dN}N>&cA40Vg{?g2xe4k zbQDng_ZCgrAdHe^iw`Jmj&SYq6^@#j(E|z*8k@zX(B|6z6G`Q_Zw*zbOi&ZU( zEe!0f!;GLtoKv`}mPN_iV@O0NMI?82L$pn!n(Yi!#}h|!dR~O+>9Cozke?j1zpzuE zl-W4O1qlFl*czZaap_h=r)vvw0Tw9>VL}W)vZ%(#3o`^4Iw-Fx zzXC_H-etqe++e#aT}oCnP6L~yeRiNohrH_^Q*onJ)NX@^(}>40^$MK7@yCd$x&#!O(E*J_R#>W2b&)Iu90~X)kx#ANPMT zs*I$qP;rP#$xh;<<{UiHN0^<6Z##FeON*zaeBodhO*W;MI=VT@CkU zVUKYz@u}7|Yw~-k{8)8*jLF#7X%dpo`${)mlBLcDAYq$0RCqZQT^bTf(L#+Oj(4W! zh8o0;bg3WCmPvxc$YGI>g2q{|hZVO(#|Liw{)r-TwSFQQRBSl9AC@GHukhn%&m8Y5 zbXSoL!mLFT1dF1OF$G}<0o)U(mg!s)0VW7)E4K1TGTZcSah5;-eJoyZP(_f1I> z811Y`jhv^2jI^;JR}Lgm%E5f1$Y57bovOC@!tGY8u<5|0w2x!kZZ12;aPJ2tDMj4u zyF^7)FPFULJ`rQ_sws5E)F$fLx+B>fc{KQo7W8VPGgxMDil$e8Z89`9J0}2-Nr;-% zyhwgJ9-|!ZWT`4BJ#j64kp4XjhW4Tv0FI^7dFeD5(C9G{9aAISiO@@l)e)zw-8c(e zX&>s+K%Cw+<#>CN#n&@|8uzG%`zZ2i7=Z#TNFL!7!uvSrD-huT+)qX%kYJfgxg4JE{x3c?_sPo9s#C zLS-U!Pa~B6SGyQDvBWpj^P^ldI+hBmI^J{u#bWiUGDC!$Rx^575#!?|UInFvDV~C) z1V*1FeNJ$c;<9$gasiXrqsAcdtg*z{|ey1FgY^oDb(=qK1WYC;1es{#4h@@^=JSr z`eQl_lsWQ3Mi8&nrm5ADEDQB-K*=8pxC7^&3~XKCH!aOGqfNPv`z{L`Juf7?xfhhq zMD?1gQ)M$53WY-<`1gigc2QTpEN#^Vv;BI0=4-)UC)5}*Vu-!U2C1g!cJ+A8Qb2vo zXqZ$=uy1hWkCc6a#@=pFS72SGadehqU!j?V0u)E^h`Ny+Vub;8HdJEHc7{a|MB;YS z6S3uj(5|QuQ%3XTe1y@D9M~ z@rL+9fgtVNZUdwF@smpt2rHnkmfQ7DM2!i!7PB5ruOX} z9bBI)v5a2JlG7Wc$TWZ&i^wK4YBDJzFN3ayYaQeFPnHOaL&%(?P~L#qrN;{Ok%Lc# ze%>ggF_>ixL6*8!_$U|+#|yKzABgh6rxJ{59C~X^yFY~AcUfp}BQ9BDZvze3jSP|; zBSd(u8<)b#KWL{aggwKJQI9RI3bxhhz~JY~jzA%8q|TQPbQy_~EF#+OhYqN% zBzt^Q$*7+@tgf_8xt=uT9Sy9Dsi=)XkxXyWPwrh8HMmY&OL4}&QQGbLx8J3(q7ETR zuaRwdI+v*;`h$8scupO-zFkMuCf6huBA{GSbE^Y}EV^*MN3=xZ3AKsG_cnpHkWz4u z6cO&q;$q%O9Z;j9L~!X5W@0E;Ncr2WRfpwmaB$@}6x4AS3RdDY7TZG%16!PY4h@XY z1a#PRba?VjzMtxh(3M9)mT_iV6#lV98yIGnImf!YM|8upR53ZXv*TRAQ<3zGy!4&# zuk$DQH?MzIiu`DMKltVnrC={qv-)H~t5A7YxqtV~2RRETZ&fc= zBEPTx=N$GaH(yN5s7*?#TXR|ZdZQ&*jmEXOswa-T)(>TJqI3F11JMl%LED zhw;@vfV{yDB;XrS(-ry^!SmUdUALI7c=qo&v?X5sX-Wa4W(h^b)76GHPsYZtS?~)e zP)+pfS{;9wpgiykg3Eg?Bs^w9&#uc}2UH=KSn@6$zr~?RPNy{1M;K(#u^`#K$4A5? z#@ok#*JsjR$@)Qr2uGylJ2mu@0hJlFdtTUAy1?LZJN}Sw2nDZF>~tpZ!p^^aJDdnL zC)5tvV)Uu;!oy0PSyu}+swT&3JE>A5l2f#=oV7hMGh?;Ap9_eX4kyQbm5>cK6YbEb zaR)0!R_+9w5Xj0c(f)Lt`sj%6UN6NeE>fL!6Dlq6{_agE>y^^2B^T7s<;cU73^Awl zJeLvaylau(Gu6vzQp_y{ZO3%PdpqN5smAJkfT-->(hfr9IDsr@C zC(KFilbhG`uXaS@u1>O4*GqqKw_fqca{DT7M;Vw}Y_c6HEFD8AKif|D!$(7fCy5G~ zwm=U(b*aok2d<+8Wjiezmd(gi1z{vgPZN=uQQhH~t(K1@>87-ymNV;2aW^8LFNDJs zF-rlmSpcdl=Z{^RC z3`*{up?K2jwyA+DM2X_zd&?0Sd_dsxWihC*0;>C+w@j|ge(H2_Tb2;BR z=l(ta%zy8lJM+%$WWR4RnQU9@S?hULC&p`iySyWbbOg`uQAlGCbA{q2w5eDllw14olAtm=5F9R% zq~jR{Ldi*^ir^Ec_#YZ`S@$pg8W~PrHB5>n!F{TE`)7zwy5yFVTX7S?;G)KhcF7j& zD<>*6qM!0tKC7TD&+Xq6Y3`m7`yQDRla0HknZMj$hISc@Hz@k#;^PB?MHDUaELo)>}{3=%GWF?9@dL!Vv}j_MnHEzpS?VCgHKZ{l}E6#2kk% zy%Ey3tjS>FcG2zlLH~yMu)rm3W3vKGo5AX2B6`k#B4Ov|cqTGbd~9{irJC3J6EG9P zxowNNw(R8)XpDolecn#q&_f(V!@lCC5D#|e@0K_x(QbxH$ zwVu1|aD_5za>Wziia?USmE`0!xO2;NZBLSZC`NNd{MxQ1{e6({gNmrnr`?Nk25oUtQQr$Xp6H8fF@dtqZ zuMGoOu6VG;MUb^)R;hPX6--x3#5x8m7*-DDs;#S^ zEGSb1j}S7dE>Q8yhm|N0;6?_dtFmWFVtd%hfVOS&T#^kyfG9?htO+4}c!rW>q=4x5 z(@G&$+&J6DW+U)^GS16u`-wWLb`OJ+*Zd9L3u#^_sf0=yn0#Rv#5SJ{hVe~2QKQ4n zubJy|7k<66Y+R(E%x}{F&iVeFMD|fC>H1*#!oX_6lqg&r_R&A(P|_So{{M?GtKRuzwA9i>kq)xsekb~*By%$G_Tr!ht|Bn>D2F@zhB;Pl#+-B zL`kbf%+l{X<|6=SErXS^13GL{1@%DrhZ2M*>Rp0vq3e&jPGp)cGf2dF5z%ce`8Gh0 z6FnT`GuQ-NFSWMDm+6x0O-NWg!Z2hlfsXi48No>?L|cp>vrqZzS)Xzb?4^8FcH`?` z7{7&lKH9`+%x6Z$qH4+U=+b>fLuAeJOA#lHMg$~*1lw6FHQsS|5a(a8fb%nQ7&jX* zc>_Y3c}_ef@_!DX9_{z7FsS)7mk-j4;k|{hu~a18*I43LI+ADc;#^-qR|T;2$_^2P zSMA@=F-8=!%^i`Taxs8FojP!45*S9i-AX;PI8r^=i_J33Q^GVX zmEGl0I9j?cl_#@Q-syJtxXVH$;)j=&rjX2@9zU~|_CH*#bBix|&(xf?mI^TLQojCW zrs%oEN?TBb^o1l4486SJd$an}UtS9<_=g+6bmIiudHG}Nz2(``C{`TjuBEQ!?2w65 z|8lJ-X%>Y|Z*=8@TPS`myH@Rdl^rA4pfbIAC|gQ@4z2~P)??lR+~~vp#ja%-O7x_Q znn;Hjlm^y89O6*;Oar|hgO+ju5?=}`7x4E;uc_RTdEenem55P0@6X(r6R;hRQ7Ry8D2;ayz55I1-5r*c7h(jJ}eSc*f+*K}^CIUCzKI$ZqV>*60-0lP{@ zTs?9wKInI=pGx9$c(L%=`~3*DX5&QyfB8f3qS}T0%{0Tqjv54$5k_4}eBSozMeHPo z8HU!Qxf<)-<7FanDIDTwn@A){Wo{Thcu@;WM$b&xc|6T;3HpX#l|JSnYWyE+;h)1D zf*)okRQGlyq~kBI<+4D5y5NiPpW|(`BlyhtSHu@uJ!ni2y_;RWJI2BAP(p>WEAfkg zdOH?(y8Gjh!;tagWPZ&U;fU!5KAAgFRsyD&C$C>{Q}_r4_z>MZ*k~s^jK(kmNG{r4l}G7d%9?XFUbhS4KwW(3^H)J%zAGKJ}Fx zjo4g~)=8pOB)lEO#9SuYpLvp=%Hy2GRY~^V+7Rn^f!)X#dr9z6!5#?4m=;OhjhCg& z{sAfg3s?KNMPh^Lx6Z!oxifYm?>~eUyss58I1s8pR|cwlA?5xG@kkxLsVv=B=BO~v zLG}_nbqYhZUvVPVAI9>!;NnwZxhifJtKV^SvV~$yc~l?x#Kxy}eJ>IOXI>7Y_Jl1I z8&`BKS;ea^Es%|x)JIM5XMsyPN;Hc<@v3mki|sMRCDAmR0;AgI>=MG)=Oh z#*?NAQe^4XqxIasgE!Dw>fm18vWRH-elx|xVe_XPY}9$a0*4o!$+ zhD&MvRnI;}0lkCy{3jSl2`lrWLLlEAK2eqb#r3a*x(Cr^^v%1@i}IOQ?L1aj4gayZ zYovLX$hL?;4>m~XyW92`80#+`-_c#P71!UAJ|pPyBRJ>PR+BtyXTPGOTjO3lmA*{d z9elF>?BmOt87?3S#h)Zm@Ge0yqUS@(EzU}r*|d5Q@9!Xu)mankDQmw>twpuxQ{KUd zDLI1QxA|M3*Ay#-fC|^q0B&c7mqMvRdiM@4x9Z&yx&U(-ss2#o)tqWMF;Y!gdZL5v3>0KDPTg&xEnjie?rw9yR&2 znn3SgQ09ekX7DAZShsp^G6Ru7*bPTxbC+MQX*>l~X<&8oyr0CK_+V^!81QRC9~m*u zu4I#aeszKEGU9ZEDP=ItZ5nAS7yn+6FmQ&VhcQJDPeesf9&BO3ZY>h3g&#g)O0ZJ4 zP`O6}kgmY*$_9K|ic03ck@Rf%W%IY(*u&6uq$E$Qtdlu0c*5pMFMJ3aeA*ueL$vu= zHfn96nc_SdDH3%V0cx*}_)LdGg=}0QV{NY&yr`icYIgOiCTqD^9t)#a`Y>DIc%z&jW0FL%gXPEl|GnlBMxU{}5ef0?EO5q`R z$n4}t5uaM*o3*Msy^Yh5A-5qDG)@ajGn!ekPL5J7)UfxDEn?4Icxr5%p>S-VjExl|VlH*4jbn4!ODY-<$y2D}#jkA3v#*h6fX-IB0J zpJ3g%^Yt7KV&`&Lru;A@ArMQpeI63`+X>WTDD)Y^-DEAQDN1QG`ASgoy{W93Exd z&P*E>n#Y+x83WjQd~();b!U5lHIKM6yY#guSipu6lJCCblHNeYL;<9HpWX}KQdO^T z?q%~#yXlsEavSAX;ybJ;3NW`RR43_CYLA*?mqA61TPV3b?wccAapHEJi7?sCiELzU z7}heNLDVRw8p#;9YaA81HGFT+vFo3H6PV<}e|N0%1A7?`oj!GCcJy1%YHr8-GNQ|r zMkLbCgUM`a7;cktf4=JPa&i~`SGqYJPVbdFFz2izqqG%d^#o+(#6|2eDr<~9y~JhX zXqnGR%O8rSawv-n44tcB^*Z?j;3>hZ6^BVXnf_>ov&(_Sz@GB5@)gR_DXDORHV@eP zeQAe;*o~`|qt+8MW&V5x+jYDwP**b<{zwkz%X)CmlW&}e@+&810$c&0xnOl$KzoiM z%v?RINFqi#&07uMr+Vj9#eQ)>)oJwdrP+G|f+$hZAP*kYQu~%yY}Tt*v@X!GzHA0m zMjPj$_1tKU-pNslL+piX?lM3oS|(#_jp`+I6n`4`Z4y==yNVhKYz#VgMQE?Y2M*sH z2G?|o5$HZ~CZQY6ev7IlRBk19@4d`W9QG-DmyBpEj~1 z3Ng4Yjk38v_|8)@;wdFgf2)fs@|5Lh?3cBOIng(@lh4&R!JBLaR)bq^;&df5A-kWW zT{;L<(F$m5#8{d~Isd_MZd0QXn9K)3v7TJInV0@%u!f!o!4yAwA{#EHWr@8inn%V?k*r za=$nbBoxCRPn3_?x%vw;^q}tzGE5GSTvAIFhrk@A(7^W+WYNRptghp)uXI=}@m#h@ zKhx~y-$B%iWl@J!XMLwsNK3X~c9mC-kV~6 z;ib@zEv9_@t_c+Jc{f`zBGghG#TgEW!S73ctN z+T>i%kn1l2&aVz;IGkLao-jR{lBJ0y7inxpF17$#7a_^xGcw-x?&uwq15#3J3kFIe z9zk`0AIL6kg3R85p-GKz#`+L;j9wsI`OT5p1k0W`&nVw^I2GMRk14Hga5{=}L1~u4 zely}Q-UA9O3%CIIG8`92A59giWM9SA{v69Oa~ereORf|S%zg(I_Q(aOl5mQWYhV6l znHz$h$J~|xed?son*;DdOyAUq_9o}Y@D5luTXaZ2a4F?N6`l~ay)@c0ejGA8QyrwC zfZ1?mfBg1*J%K)y(lq~-QY+PC&mx?EW~Ao1zksHyM23AKBy#g%;$eM!mqI5QaIs_o zJWLw!(ETVUDlm}~(sc26|9@-7>Tttv_NcU^iT=K^7sG$|RY?Cg#)JV1U8OW$v4%6T z!XE&>tG2W2Gi>V1U&gV43qLUSY4G@K31&+95xzbt{}B!9fBMG#K^bqQomuoA8^DKC zG^FiAh__hVtCGT|cSZ7cCgo(j{8VgqjF-oWujg-cQJ&&P3`9Q}l8!+E;LKW(elU6Z zC&TA!G#mPVWL*vUZg_>AA-{$UKQFqla=bTMeW-@_KfHsnt7Wo&srbJ?_<#D6|KEGC zr=i#?Jb{-p>$*M#Uo93}T6)&dp_QXpehvRnsef`RJgKk#?eEnuvimL{y7VdVK~xpf z&2>^<;Z2&1vibWmy#H6WTP-1e{C9MlxLG;bqyL#GR~Hk+$>EDg!YKWp2zBHXoeB_L zR%9N1v8=H>cy$B1 zc%t}6*P!y`W!kg0@0GwW`L?e&SArIE?FTVTB^lV$*fX!$scY4+%9(HUr7mKY&2%Ox zsC^nX?=$%s5LgXQCx2r@fN&1{gjY9 zQO^57r*OU$q)?aZsy;Feum|KLl91J($#l{v8EuPN-3;QmJelWOquNJEj%%#^OOu9Y z0ZZ!IW}+67?_& zS?PY;!q14vZl0}Vbn<|S1m+98wTo3p$yG9B(a?GZY$h^qeyaWu+jybAfw9GCzOtA9 zm@yPzHRdd-c5=P#hWh?feifU&+Yi^b7%pqczvYni8wiL$s{P8IVpkia{)^+FFz|n0 z-{d_gY>V$dNnXAG`#P%F*`H>-@A;=gY-;{0Ooq|e7Vl#U%iZ~Z%}BTgz9!8?$wUQh zha4hXRPnR3Ud<=({A&>r%`=Kk_n4r?S)j=aW|j-t86BX6c(DmEXNdeQ{I+?)ZdAQv|C0<@uD6$5>nNsE*SU}q?C9UO@Ii%2if2TQt z&O=$bKri8KYs?G##nuujSw51cml>Dt$ZXx0JO4 zFir6=XL%jW)cTE;X@(h>8flUyC~R3HX>6)l^|~T#uv*J^Jf_&e+eA23q+&#=k+&;a z8@gH7q#%VVQXWi;6@{SmKJ+ROz{On|37jniq4p~lq6px2F3oLA{(dl z+S`-ZS}liru=1Sn;)TRe;LaToP|XOZmLVnc;uTynEmU+#3*mZUi^}1&3Ao*ZN2E(^ zv3-2J9Nwgpc0Pr`uvuS=iRX`35McBIuGx=6`{C53D{m$2C01|rfI<4p+1&j$!-N_< zIIOz-dAfA3rXb|jsO$s9yVu97_f5}lL)2)0N&e*$*3a8r781o=7VQfvyahe?lem|d zm2Kp0c=7Sxv>U#4^`7LT#&dgph97fZr;MLVkGsZ303Yuhw92_z_bH$ZHhXIQvtd#z z&?GtRfXrOuWp1p)dDR_qaKsZj<4rXb6OhPeR9zhs-Dv0`?^5AjA9|obs=E}&LDhFM zdkHYVhS3gE*7>=_QcF3=h&8;C>-EqECMwS-Pnj_$xA0;gWFJbDLK~6d_C&l{Y%4L@ zn~Na&J7tPkH;8EO&t_d1PbwOYRpY7JZg*28ofA7>Rv=gTpkA|zk~`0+S%DQ|F1X@`0K8Ea_xUVHTxZ4!$QM~GEU0R;rrQn>nYxeabKf(<7# z`&cLjB?GkdT)6`}0FA*`H!D-95S38@@rYe-DU-}A^a6bf^-)^+G-bKta5p7jA-HDu zI+!MJMGUs*r+*32%-+ZXIV-?4BbJv~<+;<24s+ZdjRHy%_ za&A9By^Ebr{#=ULWyx(Ega{k*W-y?-qxuaconG#*-kNO&=Y8$eOZ!V#Oj_vz4oF(u zX)akd5OTKP0Ib~!Ba8r}u_qxcuA3meCVRZs$! zy@5pS9Q(5r7srYEa`S~S+k2nv9TYP;&anEL0FWDAW*aME6EEoXk-ELL@2J$NL^7cbMAu zWsM;4Ygp349kGT+R`;`ZYQ|NCY1qEYn*9%fIAW3V*sH!2#Nv#ztGmhP>(N6}rm`Z+ z`@NpO7u&8MONe)$&3)CKH?QsYg=L-shbjrXpqRx!0CWwjouOaEAB&>r3bcul@IBPj z3wq5Lp+*j)p@?*hi(V;7YZ_rqNHa#}Hs+`lCsvP=jHgp+-X#u+#w$Gwya5bP>#1B2 z9(oQ-6&vp{@h);Li}GaEYA+NP)0Mn8seNAXYOs80!FLO6GCWHnQUY+yn7$wqbQxYP z@W6SaXs^I3ve$K_drM8&6C}8=q}~gq{QNFs_6z7Kk2ucHx#Buv6e>nSvi_JRdmC>|2vN#M*#@`+?AADvX{hzx)G`G8K{ci&X1n=h$GSBM16rd zi&)T58Ht8ZQAI)MNg8457mX=Ey!K0<2A#q*)8a@zg>4j9Q$W0K2F@$_^7&)J7^mpz zd;;{>FEy^g*FR~q7aVo8Oc(H+`a%U#!E}DU`+PEtJRn75g9?D;(@crE#|U^GLnQbZp-OR^|P%ANfNKm9$e z?Ud%6t;-=BXRuJC+}a>P$(vM1^@Yy}s9ajbarWpA7T=+h>QhEiE77Xm`P5j=<$C49 zk`;_B4p3uy{7@LRy^{H=W~PldC+6t}!Hy_Sj^<`ZW zyR7QkWLXxA53U@LY^|g0H@7t;7PoCZ26qCl8q#@?I|=KTf&9KRJTQH`%pgN|w=?8b zpQL95Fe^2C02DJ=#KkSE3?3uve_Qs(SHjXBXasxGBVLi5CN3lf&q7Q+S&nL+8>UzWmG+tBDq814ab+JI0hfFPYp%f5U2Hnpl(S7t_o`U$oOL0hmc(23_%J4)Z5A_?9vQG;Vn%mrkby}-n9Iw; zMbGYO)~7!J4=z2FD$7kFwRg=yOKY3Q){DQXD<#z_lt%C1b!GtUqS^ma@jN}=W zB04Vp?td9lGrvGlPiFJ1NC|_^%c&bRUl94r;oY`b=P_ot$Iv_)xLjhK-KHm^$a|&7 zY08OCo~Y!8mka=Op(TS#c!^(H5QlGQeTJAqHuIxpd_;*dlAzptG|fi=a_Yk_$u@;A z)6vSaliOa8%480R3*Fkt!Q8Z5tTX$~hZOU^FKO1s#kk;7^LcnlsBG^DkXC<>oXKN; zs*fHEPLHtZo(|Vr>806eHz{uwCd6sdhH_W{W}1mJqa8?Sda%Y%O`7+V8_&v#EhSG` zL-{KR(KSX4M+Mv;$|X3I0c$flOYn&UyI6AVA4`%zUvXmMs(oxjc#BN|3z*a6P5acl zOVu=7#vX_JP7ME*xa?{E1+qgC!thJMFhyFscKufy?wP@x<{NWd`j3krM9xEI7h^3- z^6C8Xn3=fvC|_A_-UM9?q^4eqE-hCHV8(g7uYf1hf#4V^*s0c>Kml%89t}~_F-4)P z&u~6fji-WkWTMjUdkN3?b-9g_HE|!r zCf5p&$M~7BH_0tb4-fR794K*=fAlY(j6s&hc~9(RDme%b@0LyXVAf`UF5u7O zx!`{5+7iYIl%!1v`o}?#J~Cx$cUS%`T>1w9B%W*HQ01kQ9Q+_W`#poWU$^T)E?A0; z8|WSMU-K!`mrP0C*3PtmJAy6F4iw(1i9tGT&!a5zz9J}doSm-xNO$%RK#STwX1C82 zPmPO8ti8aX%W1V%OISF_1MWPb*UhDk;(>Pu`z6fA>j-lfP4IG0p?jrttqBVTL^h*( zgD1?GS@yH0^BrzMfLG#p&cE{Go)HdfQ);anD=w)g3fC4KFDXsv==gOk+HS{0#ZTJt z6Af*Qk{K1R2Vd3EPg1N3J|sytDde|GJh2L%+pPRb*m75=6~*^T9lFGvpv)#-`=KL$ zaS%eW1|2ev;vXfw*iHPNzet%YzIQaCF(T_aOfcB$ndfDnF)zeGLoioZ!m<6@>Q258 zlOOh3#oV{Jvrtt)H?969aM@yka3x!)2A6PIX||Zt1>Dpk;@tA}NVpao!RZraSbC?z&{7MKeur1|M~BF+Vm^jvHfl%WCl zkr<4w7W}##>`gw!U)gtIgAIaHo9ToWv3lH1Rv8A}ByU5{JOvf;0eBI)EadXK;SWF^BK4z@{-WnQjN6+hQ*STBQgLc4_>t3h(!6*pomKy+^!!4W zJ~CYRJvHQg+3|L5<~5w)T>VlU8qYe2m1!lBRGg1ZAT zvXqFDD61_VlonYY<2v-GZ~X-W4dE-vSkWMaf^P6N*tvWOJ-5#BlFQVh93sw4_A75`)8kEu)UdoLR30$s5>hmTFwOgqARv^{U3lV}$R zQ4-f{E3>BwXBd+eMnU;|@Ef0vU8ejzrwF0^#RBZW{-%HarSAAw4Su-;PtD@fvSi*r zkK~YjHW5mnBAa(wraY(agK2?vZ+^4HimvO4{%t3C^?fscT@UdWWm}M@JN?k|(!`oWDZ<}G= z4~yWi{m-ed>4A{t_E8oy)p~Vd@n74Z*cv3ew@xWHDlYEG?V;zLP}Jnp&8kCoEC z!)L0ws;>54xt2xB{D*9jY>^||O_%iVaua~YZZ79xyNuz;@KY0xG2%F}Ffzf9b{@|2 z5m57sEXEJ-)YK?iwW+1fvWsw*+?a>azln|V3-fmz|99-f%vmu;=IG_@S-7;;V@to) zI;BFFXGU8!-;SSuFM3h+&x=y}Eu~=T!Xn3FU&>|ew|ZX|e%w6%`9HoORzIov7K9zK zIRC}(|M=kJpQ_5;RjWr|gPs>g^pSA*iAvcbYfZO^jvP1R;`$Z>I^8}bJw&e)wm9dk zg-V&V!pu`hhq3dHhk+=gb3$d~9D$v2Xkgw0b(cBOga#DFX5^u!{^Javm-_0V7yk0c zu(4+jb2rZ0#pB5j;o8MfsTPzM6+Q;Fc^Y_Bqu%IBmQXBbH&ahpbeTd8y$J(|3Yu|% z{Oa7ADo)aE;%?b#J8@I2WlA!Z<&AxtjiBfwSZtg(aeu5RYoyl{J8m=CBnF>8vB|59 ze0mgme?m_#=1PjII!M}{E8W<9IXnR&h9q6_H&mJSWyU|0Zhnno8mx=JxQalFnm>ILH{OLDb((h`LH3MuOyeFPV zDLLjJb3}4gAwS&{-a!L+eOM-G`lq`fy=g4$IvK!<9&ib5Q|8lTFhlh{zx^OCf)_(YDuwm%=fKsDKf+2!a6a*`>>x*R&l+C90qWAJ}(88m=d z3B*YF&G@BOP&;vFG`&gCY*?)TeI_N86dPdkz*+w+l)$OGI>6jhqniD#G+j+AwE07r z!vXx>R(tTxh^44?tBlh-@+9l|v#RAj2*!Q>H|2k-) zg`SA2s*d=kw1#L4J|Ai9ei&qXFXfyvGtFl$O;$<#%P$5%f2wt=(JVoa zf4XZ5uUX?o$BH9W=AiCe0uiCL!W}FRKbA;@!I+V;@eTl1v1&FoKg`fm_&B}b0XVa0 z%$St$ZmBWC(QszP9Ce814$31hwNsk*!^L1mW#G}kZd%?pr?&6SUt)Nb!ZTeu^`%5> zOzwgkFO?Evcicc!`2>Mafu#2!(*D~gV9T9Xyh?(3jpH$Q}L-AU-P3K+2Sr(YsJf_gXc zSr~4&it>Q;u-(dm?^oVnUBasC8rpfAdyA%iP-Pcau((L}oX+=o4c-UE-c>;ql0v@Z zAoFIJtk2O^q!S@(T9sd*6o+-J zgdNa>CL5~dh^jk{?GowltU{Sq&W^;$vxKg%WDm9B;yalO$Cw7dP4W{~)>;Z}O6X)r z?K5`u5N4lf-C?2EdqfGEl>&N)vu(i)mF9==59Sx@silI#;<%fJ6$vsT#U$}MM1<~U z&n;?E`$(58mQWYQNkqzyjid&A(a6r&M~QK!HgM_u+94W>0zBK#Dl@I(o8+-gnoRvD zOG^`U(VDew^v5|Ixut6C1%o8T`^bvmtd-lN!T}mNqnNgd?aipDiG0l(7Xj}elv{>5cA76M z))LfAD-1IyIc7Fjhii;H4K>Dx;v*<)-jA^YK0Xq;Dj|nvcv`b`+_=0saRMzkD8y=` zXE&yjPbt3S=uolIEO0^`U6L}}5=JkS@zXE_vAkCUt0|&!C?g?0#%NxIq|AO3 zqHu^wuC(xM1wd#!jHGBM1PenRRGbKqWJIb?-g{#tJV$g4x`c$KYZ7@C!1A|>BbVdBG z5d%vYmA{)ldKTmflG27{51&`m0|d@UsTdgZ5V}blHRkPvpY2?xs?lFz85x0)hk9jX znG^%-CDUiv#VtyT!G1wak6)D_H1r!CfEUra?-QTZ=#$owyXlQE?Ioc-AX9^TyG*Y< zhqih?K184^MuaJ$2!5l;C$iy6hI@)~9xbW*XnN@U-?;htj+x73c^lYQ;z4!-^R^#Q zBh^d%x#jTQdjZ>g+g)lU0oY0xt_rT&LP>25MEPPQ)t_Zg5rzq~d?*6t zo!a8NMGn%xwp1#zkl-!uSZWDU3&#bq$Ywd$Lx>n~(JOLZp|!Efo$&9g@#1yrHdYkJ zE+_K=kFDRd9s942b1jLeL9C-v_TuX4AL$C<5pUr<;M{sV%y7H{SvvhVajMiFU&^(r zQjPHvQf7dhnE40W1r;DXZn45LK0u0|mqq#zUCssH7Hh6vk(>A|Y?AnH zsW3LPTyBVxDi4l;0(oFa=lQWvG(? z?NRLbq%*)pFLKxGvShUQW!@1RACdUslnHZN-nWUSs*By4b3B0{;0Uai=C;7mIh`@P z;^g=SBK~>A=S{e;kkl)$T;=+PT31E0$>~~1B~tKPcHk)LvsV@mBv#y7^vjUaP{uwn zTp_2>v-s~7&tY$Btn?dS8ZB}jXN2r<9wTqm{@wfEI@pcdY+PD0x&0i17n*uiFxo?e zDV;(ouz&l|67*o(Xo>TL>?Z#E@N?TY%~r;Zb^q!x?*TlRZ*z&6-Tc&* zk9%4AtCG)}FNnMsWJ|83T~j`3Xg+cbrhrMPvr8Iw?Fo0GmfiPVI%k1^Dm>d zYM@Z9+Ym+;4Et}rtQUW-yV?Gze))-9a|HTsqlj&ks6CS299+o?JCZfBmK&O6s~{d* zQsX*OBiWg-hTP*B;POnYk^+F}p%-7dkN#l^{{e_|N-$Tz`ovD(gAdqyVY%4;J4-EG z4c!>ex$68pC%5`QD6MH=D4m2#h)hTbN-}3~R>pUfJzMYZs{xP&`(N3KS~w~`z}hn^D=X6(dpmz4W{0C17R3NVF-o&d(|WNp23 zHshGX#*D*(m7a!GtSC7Q&Wzu#%i|+uD4Tv>#rK{q;@hvy)=@mE-(@9a{sXXO2D@y0 zg%YFSjF;6lg5%g#@il#tbizEPfGNeejFOBvd#3mXEVIf<>D~Hegg3r*ZcPbt-P2A% z=@nInGwU@iBP-9rOMNOU*rQHA$xh-_n#wcXDohQT#b-nfoM~2CoQ(1OS!Q!0-v#8P zIdFNrZ9m9)Q&HV8)G|pZBkux_TzYlNUp})H$>i_xsEo9iP|_AD9{+Z3 zZco_qi4F6I0Lf$CKLCEh6G$`adKgzT<14n;{F<5^e3y81T1k~q*&eb=>cl&zVN>~o zI_``~?lMQ!xyLN!>G$^SIF{zV#sTq99gOsn$TM7+D?23>pHg!f{W!uQcL7y~jhT%zMpwjN@7m@Sw>Vk0ggny!~D~7Fl#T_Kb0pw#`^;xii~t?p`OBTDo`u zF5!oODcU%Ol_a0SB!2&vY<(wHg;jtKoHm`HLVILO0>ZOXCbnG;`}xh{B{ZGG41k>b z4r4W|?zyveEYsOHZka1w7OhYI*e{?h{S+#h9;npk@og)J$EK18)gS$)-FQ|V=-?fa zor&$h$;re(wT5>pG2^!l6HNnG<;?aCt-u_36{4fDsgUpS(>c?rC^wTop;zn^@IET7 zdnU)ACeSQBfpoi+YU_|WA*I;|FR>j{aVWp*--%ugVP(v3#SP8Gi+=jVyz~(lLl_D8}NUtWk@UOZ4KT<7TENhkF}O}`jA;3Hn?)3J@nJ4HW0^7GL_Ke35MyJ=yE2&vOtJscVNeLO^6x6jmGueb1 zr8millBDyb>m&<2o$PmDsTO~bwbgQN2?wpI>vnA7f5dvvwb;uxUSPKc(5Sc7avJ$M zUqDiCG=d1k9n`i)K*IFfP(Qdt*AjPyEk=<{AhSl*X^zA`si&HYheyZ&Y|U)rkKcYQ zcCymTbRpm@KMSQeP(c7trfAa3wi}@);62|!Qsf|(RJ;s@?Q*<@09Qo7Vt_dFc2yh{Xs*{msp(iVoYk5En;c$=8bXD(QPpJ`kw^ufOIgL?vmX9nNs~SNtlUz zO;%7jjjf3?vkI#5T3(@H=8g-;Gt=wWTMxzgxIbGGev_8dl+g!nCsE^xr#v&iEdDA( zmP-x?;Zv{;D`&E=Lq%@jXcWCs>hr49H0^jp!pOZr*6683 zZi6(TAVs0F<;2qHZRp$zCxDA?&54?m@4&hCiv$Z=#gUKh>v(`$P-r@MzL9g z_A$q%P;m$J#UWubq+IYn0HmLa2cLH-h)r|suG7+P>uyaz+{ls>K;Kz2(n%;%Gqt$o z&Ug5H!^<6vGAB#agZ36Znyo@s2o2Ws2d;BbZO}qFjen4S7lv2ZGqmZ>)KT2)*k}mV zKO;r?qM)II1@tT369X(uV=@hvKBj$Gz~tbQ3t3FDjW2mFdk}x_+-sZL8PA{JY@f=$ znq!T&bePd-7Qe1jy6fK{_9Ajgmn_e$4$!~qqof=hL##!@PXYW(JX#W|WtudWH z=!+N%yp9CjvYsYmP@`3Ra#o&S)Jg|nYl@yWLDdaVA>2vnuhmvF0>WD0T{hEQC%0eB zHcN{xPH%apRz2_?9n0rW^JPx+Btf0!4t35XecyHk<)fTOqZBh^Pvph6DJ8R9)?Jdl zVWcx)lA6U}opALeJeNN7{l1S6L>yZNm*T6zH^_h~ zqMko;0()-~H0nL-mlRg?^Mi`}WRwe=ib1m`UaeZE>Uqj6{vMYTrgS_2ZTa;m>Z! z_7wUN%A*o6rcj-$$w-FjgqL}FsN$KrNX@e(DetE02uhh|2Tky4k-!#GySY%}Hq(BF zFMHPSt}XPCk!rZAXp-6HGc=cH7O*5@&qCN9la@i#z) z{nU|Y9FJjh?0Wd|{rl*^Taqa5UrL50dOI*)X%40XwM-_PWDRaqJEX-Ul_T8%b0;_v0Q21Q^J#zar0w!SP=D6y4CU0rmS$;b*u7YuQE;GvtgY3@23^jW^; z=Q2nZ7glh2mtzEM85@uGX%&<=APc8JT$FE2$HsmIOujE*Z9Z$m-;qi95{-jVQuZc| z11QUvr1u<~v)Ha%YhPG`w|PYaz{M~%A-iC_CM9985ZWxfHg{cYBo*1tqvS~%Ei}aC zdOkaK)OI)Ehmd<2IKMKZPigONW>}e%ReB)tGAf*XNJVD2`k~OLvK)k{ded317=#0T z&389z!9PyhhpS24>Kgd2Ca_niIng+Lq@z|>GRs~rbqHpG!<_7}VOYe7&R@MqsjF`8 zr;(nB7StX>ghwdD-{7u}T!_C0J(G4O_-$3qzQ!zM|B_zI;Fro5YUVQymTC=54 ziErKfkdDF&8=9|GK1;5gyjR5x6(a``M!}nn?yk?4M{ob+GP;ajR*GwMAPK!?+qQj) zrJypmP2%K2u0%?-o&GAbkV$OE@rGGDRO7e=!!m&*lhJX5^hpOP1CF?4Kil3rEW9CmwNKbBbN<9QH1CGFdedR8mkm)2lBilBBVtFf*x7TqH4 z28IF(^$#sm@*VvcJDbLl(a=5UG%b8#H5b3rkc+-h{&zHUAw+G zJwIWA+Z9pz5k^64#KYPIC|DCNr>0^ujo9S-Wbn&oeqc$H6x5ah zMM7@Q3<%G$T}~!wt~xuh5oMTrfk<`$B801Uj9vMCsigj)MOD&bUKgxYuRPz6dV*ilj68c(}j3qePw_OY4yD|iq(YvyM9bU3Hl}j zIs5FhpMB1I@3(u$$cK#R7|B}eKj-`z%86QA5@3kA zCs8YUcfyP%)nqZw30>hf_If^lXdTheXEmPI+oF|sXYV? z{r7tJH!Yr#xfQ9zTiCKWm$VG9>0wrczyJChXe?(EPoyB)!KMvnH&bm+Q)Lm(hI50z zlH5b@d?P%eigI$w{unCS19_+?q_g}qt9y&H^uF_Ehqh%XyJk4eXZrf#;6$!^C1nMS zz(OaLaQ2WPy`uw;(pBPAunprgLT(SgbQ8B5&ajh0pHn*?VZ_%Lje0f^9TQJ4FsduY zqMBCcag4r8M79>3A%@$Ji8J@KnkFrx`zGjOE6p9Mr(TZydw2R!BSc9y8Zh>yi0Kw0 zBsFyaXS#FKcJq4EBG54NUt;=Z<#$`AnfhgInn)T|;9G-=e9<(gyFDtF@PqvhH?ymV z;b2?1rTv8D&Z$w1xyA5rb@7B{kL<0I2+b2E(nx=R1xalvfsR@uUjyn#onGSuf1nG% z?KO!K+st`{!hO3OnT@m~VIVTk@x;wM((!9Lb2w@-qTbn)rSiqP56? z{g}`Ym!WBRcful;@~G(=>y^T1_XaPvE+oU0MLMR#5^69y7PRZHA8L9(8bsnCSwXwr zoMjdb;Qg#$YHG+KS2y(?INeq6&#Wh2wXAlv)Y_R`sdhwC$$C|8KZKz_((X0$vVYF) zVlm6HJTYnUm|2OIKe~Rk`P5xe1PN@GM8h9)N_XM&Eb>6aflapoJ9qwlz(0nFvi2>@ zh6f1cS6gy<&}6VPkf|yaeKr=>DP0cB3#qb|zAvlW31=oEz=HHYh`7wVusPVEW{V{i zd&g4pf@R5K^F-qrj#Ec8@3x6(mQz=`Op7^BMlW{)M4E+K5c(i7EvP~hL>-brrG&qm zCW--4`jx$-SDR#DGOJ&b^cHoV&0lGxe#m^Tv+5UJS^2?(9-=2>+^a;E1nijOITL)~;B@bf zEEnVN96Ki%13m0qAZ8?WU7I`!9p+?Cphwj&BydR@SNUrAAz%*s7()%K^`o|K?b>#gJR8Z~Wk$Q3kYvJJjIr^66 zw8OO$cbZ}=eWGCt;l>aD_2gfCpWCCm6?*|71FDfC8;l`H{0c%WuT{LAqG{+*!7XwzyuBQ3>cHvzzNlfWe0BjjFjWj-I@&J0TIY3n!|S5# z5`uyX`3H6Jey2ZroXkfm5X!H^{`S1TP|@RDm}A^i<$qF4xg zN{?Y4Wzfj2^bB`vmCOMzKQ3PdoJ0;;@Osff_41>$i(Rhy(Lh7~91t0GuyhB=>ihD( zP#97o1F9+QCbM9JjaGfLT7@H%$mr51TjGJ>A_zBW=KuGm=5_}ruw}fKeQAt4L;3f` z^JpyonO2HOPQRDvR(^jU_A0mBB1Ont^kkl9N2Y^He1`1S3X+c_9;Q-YGplfy4X52o z^w7P)3~b-cUGiEB15}SuQYiv1YWnB-Tj~`Up3t@wTU}XPn$kLtd5hJ?X%!_NQRHYQ z9aE}(mFCZv)p|9l_xSbLPn7I$qAT2hW{@$0gA-q*i$_#*iv+J16q^+_Q{75;8(2Vc z*DH&Ih~ble>9%9JBvG=K%<}uN;aa2xQU(>oA4npFA1ok)wPp|-`cX2b>k|&Jvw}7H z0*ENFnj#mzVUcdH+-V6ZGm9-A5jPQGb$ry_mXWm6qf34X4W9N^-}$tt11-$=IkU8h z#r(W*5M`PAvNI`Fo1@E(@LfwRB}_t)$F(t*fSzIQh(VXTRO5C)FYrj!rlqr%)~b-8 zpg45PIR^k!`<1i&sjvL2I#n9RslQTEGyrSUyU#r-I#v=`SG%K)XU2B_s&tkgMdSp| zucde4srrmfWV{H8f=;pOk)}R*IiM2fQ}9qqJ`pmf@<56K>>J~<(7k7sUoC5c$Jss`D(zak`I|5D{{KqZ*2NDEFEXD=qKEm$TG_sW`g zCz>b)macB%D|j^L{EmvcYC4zhMNc<{CRk!BgyU^7e8x0jJ`8Vu38=!LaQ-Dj#uLw`nt z!e_X)3S4e9*74jPbSrkxdzHnv&b@y$6f|2z3g2otf{HdD=;f7```&wVV4{W5nPzo1 z8`{==`=X+}Lh-{$_uw11m=S)>&;ck%yWPPT!8iWM``gTi(9l zh!huY6S5x$67hTcLvQ-fmQ{>d-Ci01-JwEoGorjLn}ND=rs1JxRB2Hms0(Fpu&BQU zB%ih((6N%-OX4rcIWS_WQWB?QY<=L@Rq6bVF+1sFh%2`&)xzuBfOs$y7sc{D#BVav z$u>er-R}Yt!j-!|iFS7cDWY8RC!{>g*Lr9@v~wpI!_5osW1XFoNHskq-8Yqbr8!3* z@V1_huP5Y=`9*@K=rKXWAZR27EzDmQ!x8|U^yp5`5F;i0$foHi*wE7FEE2E`8Wj16 z2zqCj{DUlKLPelSV(Yq}H37&GI|gxt-6dngvPai?<1J24dtnO?Xh0h44AW%=m()HZ zlM5d|_y8CD^B?Ywc*x&8gOnBTqowu@XMq>4ff=?3lYA);7lNo zPCnnh;O!RW<9w%n2l?LrE#=ae-~V>-;`BLS1_pXAv+W;s-lFYFH4W@#qD=-T`0%od*44BfI}H0%jwX(^ zyjTh^SiuZ^N^R0C>!~r57+%OM2g$Ui^8*W^gnrkt*fIwT_7}?2I?<-Gw4s-WoCo*h zasQ*5)~@fD6#{QhR4~1VD4g81lJd9BYn%IQ6H4DLXP;CNvWnh}&rWDOz!>g|Ja$v9 zC>$b38sIyKy+r!57?B}a^?B6!t*V`dwH?&lTKdd$Us224L`-~O3Vby1E9gRi#^f3I zwU17-MrB23`wbJZ%?Vu}JcoCEsbk1TLGO9&n7h)!E>D3bLyNv5Yt+N7tBMaE>wF)7 z&4*t$ssASBb^Wo3_c#hC&IdEKaHpTmvNYSBSt`z2totZRby?9tq#)+u9dOlj(Wwl~ zE_97FBE_iR?4JF2dn5E1Ane)nNJC~efk8pkO}1Qli`*sq^PY+ST=BjA;ISf z*{Lz{)J+g7UBIABIJSeijC+<|l=Lf+n$wt~pGU-&CREqzFlN!m3X17U9*%9sr+FPV zz(Cn84qe?2n~Sem!r8ka{2Tt{(u1P&k%x`l-)_r&HBK(LGpTjt-U$-@b9`r269Q5- zw--pRV-XR|8m;DV3)=hQOzJ$9w>BZZ*>pv8=1#IvIZHst!C)0Z$oYy7CyW}px?c@@ zfgND`8hTQsA&}*hc>jt&2T80l%TD|BN?^y7QgbZ8kBbxOg4mP)s=bbfuA`R8D`cZ- zw~j4z!z~L!q?kDyr+A9O1d&rb4Ggu}=9-TI(RwuLqAv3jS$NjEB-Kfv_Odx6l)=k^rD=7 zopn{^`S^e>VkXDafQ(oicvBXO?Wu0|q`(8v{65~e{l~&_I=?&I70_`$ zorNwg_XF(+;Wljt8^46EvHqb9TToRiQ<)=YO;Bk`Mih%@nn9=HuuGi~*R54-7*dh7 zcJXDlyowrT{7FT+W2n3_eo5jpQ%%7$C^Jt$EQ zZyLdIg5c336}U=4hQ>TkTPq>_aM39_H$Gb&S#BwcLuNLra=+YCkuh&oVdc|?>6GQD zIhfaiF@ah<)qR))-RF2fOBu%;53?U5$DxbBa0bi-xu|O~>DQ^cZ~B1NsrdfyPsflv zluldLvljiTTF$*2!n2fC(_Lv>DXiFafBSJqN!6Su;%EXlsUL`<>+8Sfmg zR(J1^AxQkzrE0XttxcwzfHNCCVxdqA)z4>entRMi38N7bnM?GduX+bmgDRL`3QRa@ zw$TVljUgY}I8=k#=L%1(tRb=EMi#^|@^HXVX<$L14ovCHE$>oXDH)l~G5?*bQQIXc zv3}t5jaQ9GC;f|Eh1ilW2V~lb^vDY$fIpX!AHIR#3JLhg;7$}C!RK>x$z}E37CPy` zv=J8@yF=osfv*FRgYvUsb*GQHrF`speP zk||gNJTXm--N2m)L%1I~A~PZ(8oDQ@!3Kgl+|Wl(OLS3yWu;Vet4Iy;Y!SUz`0--} z`}?tnRG0G#@0&Ftdi9Ow8JrywtQb~en5jr-h`7g}&v*NSF8OBMp zws*>b&=Y0)78`D?7tFluUIfV~Ub>a;{*{wP0N#+{Aw@R5uq5E-{W8ZI@>Bvu;)BLz zQ?d(M@3tFJ-ed$*0j#=o z#|>W{M3{zNOM%1&WpyG756~e^aOD!Tp8%!gPW({No8IH*)xsLpbERO;0u_gb>mPy1 zOfATjE=qJAVzGnv7S&@URJHdLd-nQojL&lm_WS3LvLhfa1x{b}qhAo_Qu$NWgew(V zRFO{(tI%hRXJsZC*d`yoZ`fEYS8XR7?_AEbTceGx)^Z(4Zta{CGL)l~pLGkUq!R~Q zMJ4-bgs^jqE4f6K4yUJq%N%#iE$MfjnT%Z6`fwD}W(w}9_q36wi^SzS zrXPanXoE2&N;}JGDhtb>DmgSZPei#?6T$c3K>jP*;Hbw=^Gaxe=)$I8zH}0ZhB>UA zr&5=CEbb@LNyMcw5T@wiJlQULPBqb!jnB-$n}LDuT~-s}>)gv@P$_6lc9aI9f%B2y z6fu2?%zPLO!5y{omi8Y@7<@|dZxv;IwSj;+&*|+h@ ztuU2GG~zG1jZ)yfAKG6+_Zck@TM*j4lAruXH9km;*)c2HSS!8~A5x+M=ia)TK47MuZk&w%8|BtIHtnF;1>1E=**_*qO;g;w5w|a)nSTx@cc`6d_ z_ATvA>I8X`rm=yi#Iv_zLZX1d>)8i*V+wI!qW#|&D=T5S&pdCShsN>PZti+69xg@7 zyEmt=gmgL!roRjIq9FR=tVs?@lYikY*!;A?*|P2r8VB9W+6`t-KAN9;>l z3*Q-G84T5VB*4LQ@=ps_F6(0EUul$6Q@qpRvkMZsi!xpV#XSv-I(XRC*;icnz6n19 zVJ|iOxl=1AM{R0CR#8k7K#2k2A&?3}z@?*9qQ!{ghG?UF~|CpeK z0XNy<8>%f`&er@lxD3FF(Wf^{1b4ds5cI_6NANSw`+<^$HShPc8eylLm_s7|l>E!j z;eV?LA8L_Z@WXD6HvJwc6gkz5gEPA5#qJ`5^fBN!IrSWh{&Y;U)Bye=D`Vy z>W?2?B%WHX{iiAJh3;c!EPMP2M_}3%fWh1Gc3P5h>Ud4V#-9PANyy{niDZLaIgZ%v zFPhtr6v!jhrAShJUWM8i9B_Zz_)2>77oNnI`qtA1s=c0vr1`)y#AwIDlvR9CSRETw zJ>xID7)ZFn)n9m<2195Mt|IZxL8K+^PD127@T}g7duFl67oZwy- z)HXka!6q{1#p9AMDs%1Bj^(0yugZXDCImOG@HTgv_|hALRxu3)hqVqv9e^F#lUs>X zEa}meQ7j|ra`}zYk&1&wwFa`~-wHb_P% zZ&_Trh_Gv?SpD0p|2kwqzkz?@byckB{_uj&u-4HP3~4PF5mxvW*daMJM5P=yjL-K) zDxjRWMh#g~5v8y2crjDQJFWaVg)X_`iY)zfx|>-9N*udI?0hlW_|S(^vFSFwBi6Co z_P*po7s3YLf;pi{Rs@aI^@TwF7l3;_Kp_pv;G7|AvofNrywO>tbcX~W(w;7>tefe~mRc@*O&R$M5mn$-( zik=?w6f_NGHcAF%>j0z2S8^hIw;A3skZ?Aoohy*#((`1N!rp~+x;OJuy0QSH>>*}3 z{sjq?)9R@H-R}k-46-$MwO73e^FxUY8#?G~sdVMQbVwaH<4M^T{<2sJO=&bcDKMdX zvh@!CqUXlM@H9c?O{zZk3+ZK_#GxAdj_jBYI&m z622Vh!CFOJIh3 z-aj6Fnel63OGqI2yxC=0RB1>qjeTBy=)`^Cze<`kR)Q)cm>uOZu1It&tIkHt?atc8w2V zoEwbvmxM!r+On#WW6#Z zPZ!l58Hjnd!Wr_`8LV%U&h2_~Ms=p88^H>SJe~tr zp8F_AQ|q>3R+G^H+=Gn}jj`xcK>sM*6~4dH&bLme1%8M!#_8iaHLkK54dVPP@FFVzNx(v_1_ICeOEq z?!J?wQ));uSAx8aO-xLPMLwgrz1+t4m<74dyl-vZq2gx!=f(#MP8yNmyrCle)-{JV*s9SwoiXDTgzy&s?YB(XlY*wGL<+@JM(9PSQ_(zE+A@j>b^<#vaXFHCOJb0&%D}vXL zbeDRm$;&_3`yaS7#2NBSwb&wJW*9Nw;r*OfkB><>&sCduNr9~0O!@L1)}=X=HRH#y zsim=Z#fwjYf7^m)p^GnxR4Xe7HItGisCq6h}$YSF*m+#^*GCKA&7TBU)!;|E*H zs8Zw(=*0z(wkuMmruX0#3+C1o8s|AG7MgRLaZenI5K!#8>&BX}!6S$@2eZB|k1@95 zS*n#l1&gkp-v1}EQ1-VDEv1}I^}Y?Id%*=>hP2UiE{r1_&an?HY;=yKDb&il$>s3Y^Q(Upl7rSY!uFUG)5t)3*S zQNgv3nKd=gH6Tz7EtuRF;+l$Xg&&B<0Fjevca0ot>lN68%sZ^-lU5R)W39j5TI~Ou za88?-T#uRE94x6mXaUd=OUl22zN{l*-CnQF&Q^Rz5b#>H7V@o?WALJ z`h13bDsz3_)mD$hJXW8vBk(B4yj~av+b1uhIiJZYss`}GwhZ|H!XrCUb!m8>v8|H0 zvoNwhb^rQr%h6T>Ah;60l|wuvuKfZq+OyMiycE=_ftc+qg})3ZU}b|EtQa1lS16=- zU75%F0`79sFc|j@z3I*<#5XeOb0VgwM$0V@B+%27#1;-4G3k^0c<55*RKPLw9@5w+ zwKEPnupKw8-%7D?zwy~d4jIuDPhrL91^Y@rba(-3A_|ke=WgQ5${FQhzr_@YDY&Hl zva7aA^k0dP{)hLj8<&bgv4{rY+E??0uH$|~ws&UyHq5$OeLB;TRt4-u#`8X65N-!@ zlpCJi8#8~XCSu}b&~sxQ)p??tx?nD8&`E@`x0_5dv;qqo0U}3pQ=f8VUi`;|_mW1= zm61+#;=4nKshGxQrn55Ho{fUhm~NzE2071jKHX3;NEVw>~v8Letk(gOw4!L=6CSKph9)zRlAQ*(~qY|=GBCdgm9 zF4IVtkc35RG#mE_oEU%Q;Cleck%oPN5!PzDk-(7qNiaVFwd>mL&;IsdenUY^0E4Xz zE)P_*W^{tRUjA|XB=COkbL$69{~tV{({x7@6MSI(F9Fe=q&C94L-k)gY3snNa#^(V zFN<|j5T@#?>fJ*?>0BH7NZ;uN@n#c6Q=Y0JrD3-CDqeHLJziJ_9nq56#p(dD*4`5N zNAwfUs^p^EAh|BGmCj6{A*o1W`~h8Kgcn%TU1sIYovn4X8uwYVM1C_h0Kw7_zCPjW z?~ngZ0kkQMQo56Pq%EKk*{HsK3)7TPR^e{o+7HA&!bnNjQwid=NN238t`b$J6L$z7 z4cu&YxGVgHhbp|PC#HpdNXFejHBCRia(;`i*=NsBqB#rPDW)($oAnj?4u?tqng*eN zn;erUe)mdh?R&0%=Sd}{X9dA(i>bq#A&$rwC8`uiLmuuxs!O`15sCVvBeP#{Ynq4r zf~&5{W7I!s$sQX($zIEAMh2f0aSo6Z1Vo0F5U7wPq(&4V(ws?X41E2h_>`_H-PVN5 zyp7j{c0c!Yr|zcavTOaYewu4rwRRNv|HGY;&p!-skHr4_&z*LQX3Nj!aIEvy=N&@N zXCoDM!>%6Sr7Qc};$O?p>gV_irT$#dsYGl`IJ598V6Hxvezxle_}Q{1q}?+#C))7! zCXE2Ij{P0SGhr4afBb+}TUtR|Xs44-+MJt>cxz+}v{0#o8FI?lvp)~BQvEHRaQN|a zcjD*A7cX*~p+2@Ou!n>=33hcH)1zNHt<>#CB90QG+oT#DMx5q88cGK7poe4CO>rPn z^Iz@4m)oI!f4nVqc4?vtsZ-H?x?IjK4OFh(ZOtd4--_%L!6G1Ej`_FRS){aNI}5nR zlML^;WI9D=H zs`9D5;_Paai`y!GavPxuKbDMr9TKmEtk)_s7pP8#=NE~SNjpOcdXk>S4c@xz)#(oY z)8{pD0g^K4oO@NhnBB!ySHP4O=RAW@zfB+aUA4*)^{$$wpkhXgid)lxuA}A%D!Pd>BnRJ!>NGHghVxw$4?N?nTKx7yW=v z=Mo7{40O_?1<*G}o>Za)@A7DHD~*IuqNUcv?aj^mh;^3Ta5?23aRX}Ci%E1(5;m?L zu@82UEh5aLJK64M66CeA!evp^pEW)}dUL*2yz17)D+OfUMFH|w)Xp8)ea)lHYl&Z~ z?Lh*zLz{}bhD_TI`xp!y?Dc~{_S`x8c|6apg)exuY4HSCjpo*UEp$$^u|}06Ze$u- zewkpA($Mm-BMv+HYK|Bk@anTLIqSa@F=PYoOG~6xsZu);#b+v7g2h=}yodxrtM91x zTakQ4Z<(RSvzRPK@QdXRGD`85Q=OQaye&?<$YGs9C)UgZ= zMcz8=aey58h>~9C51eeixKQkLSm5~Kq^tH#(TUmhStx&ad`YYJBZiS$@{+(S-&tOg z1I~a+mt?vkt}Q)Nbf*ZDTAmo-)OgJ9ITGesg_{+1{PK{C#ovsNPKh$h`?BiRAl4po zxKVO_tDiLs5x;S20jP$V(;t~(BLv+xi&-Kk6sU!gwp5%>f+!RF13wc4&-mAhMzX)v zS$Wx$6@Rybdn<9iZ|>G6&_2e`F37q76FcKH?;Jv<79s<2rf1QWd#*})+)MQ|Ap$rv ze-BN6M_e+`);Oe?kBXWF70=HBpR{MpIWR2+EN_5kEGoTi7n3}APhWw4V%`j?R%Tac zT_Y4?5P~#;Tdvqq)5jlDdNS6M5MfR7Y-;S-K9>ea~jZ5WF z&;$_08XEt+xTc`3A>Nr^7Gi5y0{#nrmnOuDVc43F)S6l{n_~~@+i5J_| z3DY89;1MZNX1i>W1SjIe3GQD9Q+^GCa~ctn}C$90TZOGW{TXKS7jT?P#4I>9N4OP7#_k7Wag&NODDw|$J z_#`IPhNy%r7=;c(A4dU}zSg)9gjr;cWPbCyI``MH?qE?ULt*`s6K;*iqZZFoH9{B!i~2I>FIXWYB8Rt_^X-ucYD0@FS}PmEj_r4?oSx&=+aw}Rm#aqe zIK>Ia#F0k~+zXWBi&hHDucWGn2U!u$EPvb#!+8|BAePDP>8Q9|?4b8^tB5kAtu{b$ zbXe}&;2G!EuDWck%xK@1whO&$`&1r@xX86XWj)Hp{A{lL9{$Gp?pO#tCY!RKp5{wB zdOzPG+WQdCOX$7R6H(=U*aKN*s#Kz%-Wu9damr~9fRd-T)U1CA`8#Bp7;vg!7KMzR zy*MrS6h`z*#jQI%>dV=4!G`sQB}^7!6vWV20zMTP21(MamX0+vM ziGMAV1n#L>&T6i(e|;-0e0Zj#6bt~_JM$Pu@;=xdo4qPvbiIl|3cc-#EiFu#P;80$ zz96~=aLCa*9k89?773+3c3171e<{tar>r*nkpLp>=QOhx`avT(O$JiW0CFOadfBSK zOgpqLW4InlIAI^T^?5Fr_(A0md{scMN8iscp5UJ1qFPn#sOQ|C^IHur@$^Wj3ILg! z?V>~7no_$V9~_@Amm;Q{v>6@AAX&bKW_Ow7WJrHpHY2gTj~6rI`K?cx^j$wdjl{4~cLkF;uwaxBlXfS6V$k9o0juIe)qxz6Li#u(UpzI3L4CWgNr1kAp&_ zxSRLdK>)Q)Ef{0%oS)U{dV`C!xp!qOFRtPyT-)?mBau68@->U`G=-H^dcv~26}zJ` zh~us?g&5F@)BRuc=DQZrm*+ga z-=lbd!1})WKYBk4hc#^31f6*vFgsZ6n=qB2@~i^$;eT4u{Uh;`CI^33=U60S=&r@>{FhS6?4= za)XxOuJ$P`R{G&_BZr{Ef)P8tf zg`8(|f<2zoOg7c>ni9&E9d{9$#VmsVq>-O1qsA&jehDT3XKQsYwS$~+Qq@2y81DpLQJh& z!wHU1M;@cQ=(F^j^f6goUSxGlM}BQl=2nZ=)8#n{C?Gb3y2MK-yOUw6?#NJMer`0w zpESy?N*TyjaidW+&D*$QoKqq(6`KX%@D3ua<;|cW8f`TiX`?#l12U?r2o-0#?1XSI z0&689F;NGmbo?c7hoKdmy2|JW>*nq-Hs(oV9B_-?ulue^}S zxs!~9+d;An)sIxNcqQ?zJ5OXrTf+>L6?g`rdxVd($1^4_xzn^@mYPEQ?x&wn^$4Q@ z(^g?|VIBL(V#gqH_ihb!w}G($&@w|xbX?uM;Xp`YMlKM_1* z(|Mxr6SIk|cxvE+-`V;XoxbakznO{dFmNGg(G*!aJkx79JlX3hmVEz0;XNekQAwp# z8MKyuRdC8o(_mR$!1PNmIeA4)^~W}Qp8Hd!o=nwG$kBJbKMp>}-kg2yU7NkR%?rF` zzRO|nDtr$y2@4ayVuQJ-QFToO$v5-jpaAFJq3(FYdTErxnq!9uao1Na$D<8mQ^qBX zFIOgnK6X#jDd*=Ij#L=*KLUu*`$tzjcB-?> zBu>xHuLTrO&ejsz+z^3+ds(>z0C0WFQi=iCbtr}{dC0%qyU_c+@n z6l*dS>vkXkd@!56`R)b>Kbt589J`a-2Q#tNPqtHPe$G137yRG*{%Ak>*BIm;#L+iI zsI6)uJH`58T`z&Tq$0j65X_N>_J7_5us}@vn!#f{G&u`J#=)&{F9w%jI*cEL;ykT> zXn2&BNkK-Z7eiIm5`%(AW|wI0^qzQ}R!fpk>RhQPfZqJ1eruBsUi*%*)lXDcTDKYr02H~^SX`#wW>%5$pHPnCl1YO_|WbUQ@rZ0 zohflV@?3uxIVmssx@(>@h&y#CqOL2LMe?-ex!qQn7D*)78}!}Z5a#ehY4bZ=ydJ~| zh2ewCjTp~Y`jLjWX$C_uQ3*E_&f{R zb8MCCj$gMktY5~8!?KEi_3JaX%Odqv0NcA&qgdx-hHGY_@kGrLRg_1qY2GU54gi;b)p<1p!vw^VejZi)4{Nl_9vfI-0L47JZW`#2#(b&VZqjD9Uc~$RvaEb$_$e=vCUEN zdm*kZA)jdPRJm4?geOV5E*`NcMh&&TdAlCC8I2!3zvD$UvKA`Hk5FcQ5MuO29Eu?T zljutGBx@K?d=AiYS3s$lr|W!KN51h2r&s4OtqHG>`5186dPO2W~U(zVK|MG5h$N;(_v>zIrcOb}b&TWr=g z6Xy-^K01y-A<|w(r%IFA2xb&2%h`DN7?L=Ve%~KAcjyEU5o-VPclr!U=*!k(%T9}~ zj7aB;BBLv(Rr_u(M8oe)3?wFfxy*!Al%`RC%F&q2ov{*%q?CM6NBO#z6BEbf*E(dS zmUN>{kRK~+Vq8ok7)9v|7dEp`38_ylO*)xaCkXUpI(%z}709}jun`DN2s+VapdMlF+cqz{I&I1lNbuvb4QR@uzW;sU_4a}XD)Z#s8sGQ3f1d+u9!iUU ztWQiVY$C6n>cA-&u2AX3CJ%N9>x1ov=E4raFJ&JWLM5IB0ejz*c8cD{N`Cxm6Q}m} zql2eE6^kc6HO85;iXg=BpjC2K*{aih^nun|V^pZ{#7E9-p7r>)xn13zTvn)cM$)A? zQS@6K1dRvVw+B^ZhOu^ZhhwW%R0V7A4G0q!JD#9G6_odQW>6;ZWJ^IbVjj|(5ZHid zWVR|0-C+lb_MdF)Hid{OK~w3TYT07TP3s_e9&j94Km&i<`=}}+d{R4?6ntzP*Hlpj z)Y&+tl0p$jw3(-J-LQ!4KKicuPd3BT+cV;vD&{8P&%sgTYHbi{SCR&QoVzUX;mAJDGo7=CUo&6 ztEwECiqzJ1u%&_Tsg@2#aC>e(zq!@C0q$E$|@z53W;&m(v?XL3rfkJ**a8 zxH1HHjG_~X&T191IHzl|r@vJ?8I}j^b z3Mj?j5)%-ogT?9{c?+H?7jNLuZ`-Iss3Tf}V_^Zo_B)duHjCD)A#p6Mg*d^;Lv=Rd zqTtMuGI&vsfKF-|REuC~_`gX97uz?KU*vg1SJ+oVG9(kZ_qW2sx6h=*O2~|;mVxkd z2f;M~8Q-Do?yCmS&Cf{I4{`r2&h;4En3xT{!AEsCLI!Rx6`77(5=I73CG;4ENvJE%^3FLp+qYV}11wFKcyzi@Xm3 zBlSA5J&tTiMZ*$|79_rA5xpCpRs#aveBK>JpDI=^FQ8?x*pT+2FVrGWZ5TTpJWEf3 zjqPUS<<8QikHh=)Xs1_z3(u6d*~z~%VL~$z!w(y|2UKf1V50cFXNQ9ZG1DGpngyQg zt>P7a^wJ(&d)=8`3wQ(EfLU)+D7Pg;^Z+cLm21(Qq1gdLchz!+=;#$aurA^u~>XX^O1!es2+s4e)V z8R+3|mYMb3CqS zV|(bUv;PJw{&$L}V`RCEWh_hQ?)+wOvB7H1Wy|Dl9YIg+>SvjF(MEQcWVWmcf1c9~ z1MxjP4lAlTdvi;0js@6*`pu&S)I;6e?pTTHkwzqhbPLGb_@?$}EKsy@p9(f+ACD(l z?wF)HjS7vRn_x_T%|Xc1LO=^Kb0%tJ(r9UlaVCYnM*EL{4P6=A$jB+!Q}}>-a-y>a zKulYYF7)&+wg^jE;hR+vT@X=cKNqOZ!MMGkWc;b6GzU=Sm$a5$`Y2571Ur(F?9|H3 z3i0J952wZEqk_HO_%e>BZ>&&g-)>!*>}{1>KJk6fDFXKp%wM=<{gerNj5`{hbCxyh zFe}8fiP5%ERp6$9EQC+9B^_6XtPExf`x1-{1+DV}YQT>_ zP$$LXiPR#6xHB6se~&^{J{0EeMtEc6rADUI%`H90jr(QP9s0La2CUiD$>Ma(jy)e; zV|$d^S{i)SbPV!zoh++OuWvJIcs+sY&MR)#8DDl0$xm(U4okl86~8Yhry{yt>Nyv%8|avPkHo zFR}i0rk6kujeLv$L)=?Owej}r-naz~E~OA4xVw}>2$EnSc#z`mR;(@VP=dP$hho8@ zxI^&*r4%Sq3Z)dO(*6$5KIgH0_Ph6b|2p@~TJuM;R~&&xJQ*swuVFJg6KV5wN>~a~Zk$P~@cbM@&mSs*a9~J%^VJP^N)uV=?Dw3H>}S6B za2kv+!Us(JRf!WFxm7S6o8kWDj?k) z0%)ZZM>5<9@#$2Zr8=12gyB03FbnbI4}8Ns^pBGHPGT}PT;tX0)&=Jpguvw56%49; zd>0OCGCh{s7Y)>*i9{VZtnx7Y?rOZ8jMQBDLwkJ3kaZcitF%OJ72zgN_B+zg1PraN zf?{-c-GE_$DMmJ{4v{1hk3F)CXde<3=MZmyZ(OhbnP`(9X0ZOg8*hx1>v_`07hS;- z&N*DNdsJ~RiQ%zc>eDXd#vR+!s%XMuduir7ZeHf=3$5EA@2T=~-J#xNV1uruw`#q3_ zGBj~va7v}gbL>PVdbJv#gKdn(9TDS62@1orIabpxzg&c%Mr^A0o#moP;tUod^V4ZQ z)2wHf_c*#tr=`*#W>C=Z8P5z0%P>IgfF9{^(KUYUX+}T6D1#6*dFqXXHcpK+6DOWs zO51gv%E$Hh51gv?0V+r;i2|ZWU(h*5BgSuX+NZct;)|~UR#O=Sj0(C9>SXB?o)?vW z;qaxUu;s5~$n3{r3-O7_L@Oi7m}7dA|H4TZ3my&Om{~8^Tc*TEr)rxktbAYqjewo5 zl!X6YnEAhxnD(;}@8Z}y#5GGNpMCsCm|Y0{YLusUVkb*vmsL9JsUq4t1<`aA}s2BRJuL;fgu-6rij zrAKe;P7%@hf+cAUXMtxbC7~V5^Zqn3JbvWh97(-d&2*|V7Sep($UqwVga5*TrQAtQ zwcd5T(mLhWIBz#s6sdIFYHm7_YT_Z@L=BOHR}(iu@+tDoF4LdG+N!u*0%Ev1*z65j zo`%n#Mn$>sBgt;+L5J}^1IoRmR*l=jj6Ty8PiXZDD^hHYPqVB~dFLyS+bnrL9Lpqh zw#1*0q@k2LNaZ|zlAxf=b}G*|qUdvY*W4O&EJ#SN?(J)-REbc2LbH+LD z>$jkq(^_s~>B;xw=Ur6)TPCY_C&bRq9ZJqY3<+`ZH&N>s7UG-x-^JJu9^Dem2JZq2 zff`pMSEFWV;$C)%Ce&6#)?!%O1}q9nko!e*<99}aXVfi~nJ%ZZQZ7k*qg`PH1Lbvl zMl{Y4eu#7F&wpIscM(Gp4@e?K2wK6}Tx$>gn(r`uXGomvYs~yhT@Jilm|;<*?Et0O z7li-}9*?OT%QKXbR@%~-Jrgwt96Ha_%1%vZUjM-{=yMW_NavnZ_B?A$x10y)`9!7z z!3G~!2CjVWOWNOLQc`)E190Hf)`#~G;O!*$i`eaXvxS9YMjknFS9QkTk(w;vd8IEqJ{T3+b(3eCe(f-ST z&?bz56nfBRy}qjBwASgwT!)?(PkCK(mT0ZV-p+d`+}EL-uusg6R?SXE3V96LXhmBE^SCQd8{IMKr1eFy)r7Kfv>KU&ayeo_4SMoj?`kBtfoR)N88#QvM zL529j6cuOIVh7Ut1|H8qG;kZzWOdz{s6t2>u&nXVq z8tkQJp2UvlPz$fn^9***JQAXWv+;D2hIcduqC`A?fZgdx>|SXS;TbT2QWk~8;OkzI zFFop)Kjzjg~TT~y3eT2yeGn^1acvNmydIH|Qr#8VGL#=L)sJAXQ`}l-s@CmAK zedpdoaD!N=04#C2Y+kS)d3{OJ+GPFN@*#HyE2q-QQvx9!Mw|hLlp-yyD8c0JBxn`4 z+xKD+ZFt#`IDI%ka;N`Y3#Y~6I7&XHCB>RO6o|`3bjsYPg+b26Y2s>EOt>Evai)86 znnxv^3aBwKND%V$hbR2O796F@dXM_F-sUHw(_9-b<9Rd(J0>Q$z>PC+Igp_{9~-b&jt zr)N&hbEL##OL8_tfkWHHssu(8;z2~N zw>qD4matc+ZjdgX9A#xH!k;dvdwHS3V`^SLLCRC78snPX0O;A5Tq>T(CQ@w9Y8BZW z{8p7owy%Qk%&-@U~YWwN)1gdI5@bSdx#eqEv8_;uB}k z|EW)8=+Co4$+2Ak?IS7-Lc}%-H64*eq&iZXVA)qQ$J8@?M##*SH_D zYmJQ+nI5>m71tC#09L9|*HP9S4QsaT{fv>k+7%=i5 z?fBHkNK{5i#c(p3&W(gx;Lhq8BJ>}X0*(xotM#VTx)vPJNmM{VDS_ncBco~ZF?Z8B zL*Dr@%7~TbX1TbLv+g9%Q^I?JDJh)es$*=#Ld*~w@UIn@OZrzNq%8B0O@Jyyp1}cM z$ba5dbE-ruEt42h^XHk($)B3w*GcM1~a#OR5i0V&~%WCR#E!WjWNvNLaJ;6YhN(fQR zX?7_OGxrpwc5OKeI;oKX<(Wjjkx4cZQelw3Tewc{Xu8}g*JD7#Ze=d6i%XNvRw)X) zaWJ5hqtaA5!FfR$4+GrRi<4OxyMvUml#W+*bUL?zwqwuKUBV4W-d5%TcV>`B_mL}f z5Ls;|^U_qiE-V(j6QNY_Duy~i6vD1r1tkUHZHWezk*!s%vVW@rRL)C#7Cyi={)WZB zVvBICBy!imVkeC$M>n?LY?<+J`oAHW9=c9*j21SjOU1rpat+me_xW?rOI#uAT`?Fx z4dYG;t~PlyOkpSbwHY;0CV(6nh&%jPNMacfy7&Xzx)(*qz)MH1Y8N#cwLZ!%C_E~z zeJmNA-NB!18Ceyn*$~(2;vqc`QCakoE#|0JmgUO3k9rR7=zy|!BtEG(vkb3?EPVss zb;4nToVZ}IC^MJHVPw>BW0m|0f?H!n;J%IZD(rQ%7y<{|xPJQ8Kq%u^;{_|w^%Rw} z;G%*&Z^@K&Niu{)eIm@<>S+#OsOXDlsM4H0{_MmhgGp+ z30<~}(xV7>vMMjV$CY|;o$vVMNg+KDN`wT1l*Ovqd~$}}UI#q7e5 zylORtE*+9t7Zm!NdG{eFdMj0hF!e+=NBxT#iFVcaBx!e)i1G;V4SAqVEYt@v0a_8SWvX8PF=bJX(bY4Y&Y}3RQiZ~1_LYoyya?1vWGc5Fn&r;=@JK9xeU_z zQQl(p!U%H)MuWhLn(?tRCW#N6uY9K~;139em$<7)lQd|YqEu$K*Ud|zPOOpyY#(nC zZeoQpgEx2d>@YFExQ>uFbUNbPsr-KZxu?LxVj*%1L!Km3JU3Emu&&pe0;8k<OG6pouWcKEwQ z?g6)-{A>vqZlA+0x@#UK;S#?uw4o&@vc<0Dh1Eq83>ts&A;b7VhB2=RDX#=&p^gjS z7kf14djvgWE72(DOme9w0~?BC#0xHydU~$nB~6;?Lh%)DfyyOB(A6>leyM9ob|a#G z=8w!M@!GI2`QCvy6=RVstTnCrtpFN>>1cAgkY(QIzmE+$&Jx0Fb}G+EqHV;r$nU^Y zoJIBSFljFC+`2-n-K_BPp?gjXP~#x0Dm2`CKgq$k(XdL&GaHCLfyaT*ymXUd&nEJiAB@Cz zK_hMWid|CC4Ax~A_{R86ERiZT);)s+oy)7|hkCD!^%m;Cew=@#(8vDjf?9R_;Dhw@ z5cD+E}qV&re~d^Z7v^4GaBCGtzy$K z(NUi}+<18#zd0e@cgcyCdLaRrZ!sJ*O`t;M={tixa*BwmkfB*BnkLh?d9pi2 zaszgi8_Uy27X0_K&zb8yXaRZ*qiW=3ww>^nodaGd!|Y^w9~y|Luv`vHy!6Qj5or^= zq_g9wmU*48$)6&YuTjl)p>i`IvzJkP@^)qkC(%hQ%*fZ=bNShS0<`XGH<}389%Zm( zR>nQ~aS?HeB|M1WP_Za8Tx=k}{+6B{>Nl=Ajq%P$7eS30#dUs`IYzJyhvE`bJ|AYI zYty+vD3PZ-1BB1iDkmUWv0@C9CPDk6<>yktHVpE_B;uMBw%%F}YzCCiivF$unB0%u z4T&ePFwFAccIc$IQlDcwZff#^5e?&35rGeGg)X8X(EY}xD^kluHDI_5Q`<30g|@B%^t_;# zCxJGQFAzBp0qE}+birKL?FQn_|9}Gh#~h0&49zhsYYlSJ zcNhF zqUE~u=kHNI1d;2FC(0;^egCq07FW%MySz{9ccm}+ZodC+tEWiRg{UJ#N`!Ydt!B{BW797Ekr)2 zu|`QtH0M@F#bS^sF*Da^ki#TVP$ch)p($(8VeV@|l8kw%F{ay|A~x*I^txr+@8Ff> zDo)I8mX|WMtG`AOxs9HU)+;8-R0$Fy3T5w_tVq3Zxf#4<$9XHuj7C00AO+#nW}R@e z1e-d``&M3CXf|vkHEt?fG6n?i((v>Ebm<(tf0*>I5=!npwqi+Y`<)63R~}C8j%hI2 zc3I?oy17j};JSeAns@KFI>Sz4`eF(l!%!;@MFm(fKVO&2T%4o3aK&b|T-3xeJN#5L z5*VjWh?%dIZ$`r$-qG9INTLiw$?(VmBuQf^qUIG&eb>*XS!bo~hHjRD{*PnHycde* zWjMAIJacaF#mw6rqjU@F5R;TFcsX0sp28Nx1GUc^!2%@DnCu94Bk`N~Zw2Aqo!Qka za9kIJ_<$1m+*ohR{BqiwRI>|>kJW}D#e8h4su1luHe9n)qNzurE+c%vgFm&yUYJa! z*NBYRhWk{p12^?*)LoR=+n zw4}q1_>c8n^|)8Od=MuJ(|*iM+AS&tfrk`dolv9z3FX*2q7h(qbl8)E)Q3qy7F9bT zO>SM=5`m+chIP`khTA{wMH;HxO7ROxzQ|!eyLA_4lT(CnJCv+OMLGv@73$zeWm5L< z7+8K52)#!{5pj@Mjup`T_+M-%`SnFXrVS6M3VxK*gKz$2yN*lx3+Ju2@b^u*ml}1P ziuO@|7&B%Q@u8ov%o#jbp!K}o$|}XeT{YW0Y9SUU;V+!cZol=oDa0}g^^1!}iQ+b+ zgT?TB_XWnzEJu6CZO<6MDMF_$XIAyzU<@JdMJ=ZR+Ttm5mRQCffAJ7%eB_4JeaAr9 zv$4zffE2HvA6rY5KP(Fag_tU8gbCkA>eiwsUFZ+vqSDJ~XJ?2xc#V)>Xwj}jT{QE7 zaxAZ!o2F8_&=%SJIe89b442IA_?e#!LEm5mY+JV{Oqq>+vW4tMEMM}t3;FA9szz{f zrSu`ht>$;Zp-tnTSWcV3c=ptk<8K%-k-cHu_NT3%Y(-LICQb%ieH5SL|l7WDdKQNAiWsCcH95QQM6WxiP4h}Rbas+`(N zlSnB%vJqh+zgB${OY+PqljMneBQ_d~wR8qRf@Bqv7O0W1ihZ%`!*=eN9@5%u zKjW{Kx)m{R%=Q5OF>hU@CBm!QhSgf zhh)K$yh`6<9s{8Z1cgI(vGGR4`tKc`ORnuWR^(SkGdngC*kVxGdNcxN6DfnB){m!@ zMi(t_^}q4ZT~FAoa^1HuFAs7kD;AsU>wt1BlP{lWCGzuhxH86~(hynEg3E`qn{}O4 zz^rVOx<$K6uZZV&+9RAHPLfmTlIjd|n7pUb+{(t#!Kjt%sTKrA;;f=Oi$=l;nuf0M z8i!3MHzRD!vT#He=h=>8r{Zo-1b$~JCnq_6`Ql4`RglKoVneYk{J~yg6}gUal^o!n ztX1RFYgbVVYQ?JLgrqVEapB^gXm`R3t82@N35SIUIZr5L+AqHTd&&!U2Va!!P7!0X zfKk#mMGtGDg)bZl*5*FUs|tUeH^oy7;6lY%TDxL$gvAe4+Qtdpw8ERT1~o$w#t~!c zibZrLfv|wh(V$T~!nWQ;mv_f`lA+`CF`eT{cLYB0rlLRhq>+8hD zAiPMMns^gt9|O=HKfx9eP--gs(~+AX{c}!wNEw!HnkHc4Xc3B(&KxFzyai?DO-8KMpY-OKRS+1sR;Jo(C%|fkKz=I}vl+MT1GgnkHq#(Q8iN!T+ zm6oI4sU_33VHkP?YoKIFW=w$qwQ&YP!=_5#CaOMxoig$5gG6bOKM2s+Bhs}7226CZ zY@v@q@me+5LVdy<*rMx~ac#FvNF2cAJEF^FAu^3%77vYZqsP5=6w|{pg_TA|c(BOV zMyk5*t3YjWH2j4vP?_U4+V)%~hbP8y=t$Rc5FEyyCn#J}S)$YYBexPoee0t%v3lpY>Qj-~vAlEu2MAx|`7;yYlxq$y!<@Dyc;8 zjAYNMigm57iVTyEAD|k|u;yzbrRn%O7j|p>%n2(}pE$PbbR0`KrU=){7jsFD|9Q>+ z34>a6-8ty|3XY}eJ(P%KzP_4b7r{+-+@su^;9{_jLV@U4u`mBq17lxXT0qCtXuYw| zH!U7#n3&b+1^bI(Gard(GUf$Zws#3^cPH|&Gl@*Zc`}zE#N;V`8f8_bL@)s`CNswS zS0SnU(k1Wz=@tDAsj@)xB?0Zuk&VSm?{GWd1GNuieW) zBxc5b@A{tK-wR2+eeJ^%Sje(W^!ts@Fq#7BLQ7i6pYSAd;KN)F6th%=qNht?ZQ0uO zN2orNz*9vWZ8eUp9PJg%y*~7eo|D6`!b$Mj_Fv9&fE@VbizSdvscD)GW&AV|jf9Ek z4F=R%T4#8Os5yR0^4iLn19VwAf{NLiVt?`D}#PRYS~B2s?35Uhdyk{Cd0Q=$hdxK}C(|3Y6Sr9FNZgTqR!% zD4WJiw>HYbODvER9x@JhX~-n5hsP?xzz6x&uF%9I@~k;J^}xiuX*d5IM&2VCq zTCRv(T~N-$O2+siQ(Zncg(d1h^=|pVdwMtn7O{5;wRscjyGy=0AKM{M*)Zh|cC1;T zGV<9wVrq|B4x6a4?Np8>{hSJMZ=7o@xNW+f9XMWQlQRyc4}yNRr#~BeZM4&B(T0z^ z(Dv+}0t<2|f%IOJS%zaTGB?WEh0_vamNz6J(0>q$FfSm|EGnBmHJnwu@>%g#@X!(G z=O~=hZBQVvx|`5U-{X$z#v4@;5?Vr+g(c_2yI5csmaj*7Q33u3~STC+u!QG6@*Wx5R z(+*Afq8{t9@gk^MJq{ig-;3rYd=}_1N2D~j+-z$QF(X~=sY=A;rKyn3&bXdfVq_>q z@3Ys&tFWJ1&55&Nxeg?ez7ixmSeUW%8u;Ya)iG4#epLQN?_;uJyN6+-uc|pMQQL*- z>ja++{-ygb@C;H(alsRKP@1KWOnqp{139$s|V5DR|xYG*D8FFWl`VZUjs7Y<1Q2n)09xuZiu!gGQYA|Bu z<1DTqF0PuErZA^93;{FTCgNQy6T?g6Y9|0dGI|MO;-H-xtExWdZI2Fs?le|X7gxbz zx?N5wPCiFx@_fvN6k=m;o*YNkSZHgxq?s`px8GUo6x;aRYvfNes|@WbY2N7Lo;Q5p zv$+fO_V>5j2iIl!9E-^o(}Zv4E#5Ct3BV4#-TGcK%05Wgv`ff|0Ma&v1g?vld}Ubf z!^^lFvJ`Z!o^vO1+Knm=;F|~{+KlMEOYAe`q9K8p8DpluQc}=H48aXend|*H48+?? zZy~ScXW_&uGoxHQ_8dck9{LlCmS#20E~9 zldmsx`5_!_Lpu$Q!%&ezZ>=$m*1=jz2EdQyA0463RoM3`)xOioOT9H-XtLq?vPG!I z$SQkKOQXR|K?)KQ0QNY&;Co!i>!*tC#4u3zh`@Z06Y3UCb|qtRXBXNNjbJZP}?NIRxJAkWju{XxuTqG{BZR-bUYw_rFjv9atV8S2% z@SV*WZ8 zZE?M8mH{0u(@E4#8-G3N1y>%-!o3$iUfL$$VSs+uNAu>dq`goh$z{b!<9#~L&Q z>{1b-l)x&RgimNN&-=?5=kN;g^d;S07$G!xzpFk@J2-7G^9^^Xl+uJTp(q+&(J`GA zv90^!7oe@f)Q7_^eI#}yexvp->KvJ>QZS@023IQxT(0>un%Co802E!U;Gc)up2i4a z-+uHtJFh+ANp^)?PX&7tO zYT9N{1?MG=POstM&qN$y`Wn7fxHQ9BdK`GHI|f;D>6Pr2PJ(=a7CcQQvFwGH+(-Z; z+m%L)jV!Nmf`XCR%NyExYy9rW{?{VVj~*e5-RubJFKmYelbpU;1tazgDR^8lg}&H3 zO=m6X98cL|3%kbCbC*v<3?>BFmW~*QVIaB4=c5)0O_!&~|9Vf2e%Nn-$K+Z;P6B7_ z-A~BAD@$Zr7z6>YtsWfC6O%_ykf2zPik(V~;DS6tOVZ!Je#pHGmQpZC;jXr6r&ByJ zF;}XyMC5e&OA0My3m)Jl!57Vw6t6-(n5;wB?-BW?)6<%O*@Of!I?ZoIQDuW@7_>Sn zL&#OKmXlaw>;!RRb3ecP)UEJvz(JC^gXJQwM|_d#Zt}Zma4F75*~^YCI;AfM$3{`q zxk*#Y;4;|JlH{^GnrVsk#K@qoSd7<7nY+3DT$GfT)!Q2zo_q&T0`j6BsMIw~moAN! zVDCT4!;@u}cKK|ZM?hQyaM2LuILc*R8jI&b`nxPyv^n)^Yj@Nc<>i!hU+$)E5TRq2 zLRQHq82F7?XWkXNM?_jU&gv9mR704L3rOENJ{Oe^e1E4k`%a42X5lsD3S6uNk@6Sp zpHj&2l9;G~){1`TNYW@9lcspRInJim@3B`Ut_a|d_TP&TOI%*e+U5JglmK#GNURe;;C_#n?AeeqE? zmDKB62x4(V*bmo}|Frk~1DhJqa%{DH3Zu{0durvlrc$|-aA=^PK9oC>TGMo5O-rs?CJUdVGo~G}=2~x>!7~pHJeg!W?B9y zuOwNG1Mmq4Mk~!pkoM@Ov!fL^@a2Tm-K2lX7!M*`20u|S5^QaI!&xKVb4n0GG(PIk$%GPK>DBHbj#B z|4_lmZl1RthXK)*Te=c;cx=Uta)u9ms8e@{bHquPdPUR`51wfXOsGvQnZxy_Cj9st z`J5T?pqM+7ztIT$9{YD!SFol; z*4sTz9;V~2-QJJB@$b@LIzH+u4!arZx@^J)56?Q4F;rPb6|aOV9>o4j{V`Z?j}gb$ z^~OM|YK5av79D1CS9c{sJK8N=Ybh!t)T{@afjJuB5yJIu2?gXxF$YOQIRvIgvjbzl zlTnv_;dZnS=T9d}A6KNI_i;<+&~3bwrk|-bY|tR^A{8MK0cN@sXiYaF;H|=Rx$0DN z<-{2FA4K2yAnsvCG?kkI{zT@HmleMxxk1D3U@jyQKiC`}?@Ov(sd6N<7?aqszEUxk zZF6HQt!9zrvrmYU@Mpg~QnL?6^I8NYgc0YijYcQe!gDc~6}pK(7hBGe>0eBQil7+U zLT{7kr7X1pKi5Mwi4&S=%M)>_tp4a~+r@^P{D-TYtR?-1A~Mn>U+b=UR!t9hrul_k zB8$n4)QotzR&dbJbFuWTSfV8O$WLqgO3lv%az%VD+jIzldR>q7@ zcOOiAB%R&Q1}LG3+<#0{Z=YqX6avf2|`lL$V_B&ki}L_ zQBSp^h@zA;gbM=`E z5Vg&qbZ3yeEE1ux65#0f>vVL3hjKYa2F!zyg{d^AOq)}{{>7`hNNARZ-wLWX7)r70 zYud?~;LXBzf{>6d?7_{i<59TA=BvinLHt^V`#l$uPGITYQ%B1~5QA9{sI{BxVmEVf z39RO@%rN-&6^02Zflvink4@YVmh%p$dnpx++>S~$Eb@uS6w#c93od-;BH0K`qR|uY zBh^!F0nPLxw1aQ~izNO69$!fFY{^8NGQWBolDUR|_28jyGyVVhCue+xVYh&+$x z-dCTAfQ|TPrYQdD$Y9MXa-%4I9{Mi~10@+f8?bJddMxL$g4ev$s}JSy_p%mn0a-s! zJo*9JGK*w`%{<>Ci)+|cYI)5kljkmr+!GQM<}4;8+RJ%=DMsBSC6;B z8zfx`JP{31R-4H@|B}_38W3KaCLlWj7!zovqx?DHFEBFqV_eQU104(ySMasCQPS`b zb1dKBe4j^Hlvi#I-pnz%mt3@!*MXZJUBg>$I#R%}9`TZ^tr`hhv^2+!qoVrTDkt9i zKz9;IOxJmvsAk|ivUZpPpol?F+)U% z%ZP4l6dFn5$wkgqM2;8&(^H9bxJh+d+Fvw5)Dw^!`)W!HcrSBN+Loko4Fh};ZAF#k zc0R0he%JDS>j2WKdD=n5qJVHlvcG^rWf^4qr{NzTBCdPmc4t^8-5d-ao`meGoHdPM zo1JD`vy@lV?y{!Q+>(PUsruA9b$G3aeJwei^m6+}pSB6;VW78#VM4pLwsb{YUk7QG zahbS7Yg-3Nb`7-!_^wH3p4DMFa=_zoV)8xj)`OAAQr%6rJ2L_|A+~q5_#BrF>63U$ z=Z;Fh0&;O;KaR%{@QC3$0tkh=d}iDzO^`}0f~~gWnhyHR8Hovn&M`(hK7FkXm&5V_ zhoXWy?y5hNAE-k1wI9<5mTVU^&%N0sllg zBbmo(=*U;@GjCmoX`%=7e7Ho{+hYAEmxzhS;#w8O6_@ux>lL_>L%XPP)HJuFBm*Y< z1adCvLT2X8?5_3l62kzo*y&7bsZZmp<7fvRenF@uu;;`T&NBjHOC9`8?X&OEhvn)Yx?F8FNZNV$@w4z;d{>R7oA^VP~fE-jQs4C=zkDZb=s zkS=#5jaMC$y0(dONT;8J4)=SEGC+1c-&RlpzzE*PS~yvh2{Awg;~L-~yV z*!bfGLj@x=7o7q}TXxBK*vwFZBq9SFNRnAygvYQ4>YI0E3x^pypl+*T8Uq%v1RlqJ zPNd+nMPhFnS*fG6U4>=^1MX0rzMRSFRDn-Tq!B95hWPdt1RzomVOPc%~FO{R+0x+P@l;*%Sf8 zCXEqNNpT&j7>VgP3_ak#r?rif?DHOFPY^X?Bjt>;Ys4<7aUIgF@JL2E3*R(bEz~lq zq4k^XBzL`mn^^TmruitbDK}}J*QMr=doHvzF@q;ThLVP3>AV%-9=j_H7egS>)Rog? z>snXOS)PKu9r@Cx3^Fgo00ONda{W@KTJ-WNde>f0&5EI3WLH_JGj$4h1;BkAz!#- zjNc7o@7Xf1opj;6q4_tiWEkTQBy#&ZKtw@bndhj_*Dp9CCr6)1TP1pY!K%&-D{xrK zaMwFwwP(UQh+1Ng-yfv*lN^+fN~Kwge1gb-f^j1uMfdwgGn|u}EfA!9R!@cJ6Xm!) zq)qrg@hngjOtcJ=JQjn04Z7GleNtoLKa}l0OOWXqZgSW`QSsEZ=I`}X_3}0Hq+Hy; za5QW5iWLG*R{rP1qD5T9wX^Z7M?ceUURwqJPrFlxkGSdIF=7ALE1w8IB+Q6u$2KyU z57>f8Ca(PcKI63ZIPVJ`6P|H70WIl}YWe!i#*y8c((8t9*`?!$i?_K zE0g1yl30o1YWYM{v<@?Ij}0WjYmMJ#GK|}hR`OttJ)*|7l#SQ-d5ah z28Rc&WuhIg7WUpr#RUIM5|Dx1C>_A{%fWNp^oZ)Jj z16#jm56*d%oAl1c11|&l0S!XXv>)OP=3AHHSLo2rQ`nc1zf0q#x&cV~oUYb@`Bpo5(=+Gg=y!AD4a+U{0 z2({q!V(-)vLMr2^oq>-~d(ue-UQR<`KGUaqkkaDjdeU4gfkTjYvQ+GpWj}5*d8TE) zA&RGh$4G9YsYim+-t_F{YUBiwDLnl?hl*mL&Q+5xKlAwp7K-nFzE?pY=?VM&xP{z$_!%P(-V2jP+0fK9X!J$kKXAM&kz0gc!)sXypWQmfhFh<7VVy zjF8IN5t?8&tIFf^*lDh0bPjia`T!I~MY?5$G3vTS_3ivow5OMDdp9nyabe!7;b~MpXv@ESa(>e5(U95?P}=mV;)bsW|e$vbs#n;(xClWsO==4>7C5v$&E+jXI*VnQeZeE zG*cX4h#4grwb$dS=t#5`qa{{g&uE()Ay5@^aT)GZEhc?j;_2b?5sFQ z%!*n8WA~x-8(7>fJU*>Yk1J}!GPhaUNHg!^I)O5rtkE*^H)YL_Mgu6S3fHh1gB_`f zFEA+SBu+re3)_L>x+D@c?AqW5H1GU#1z)*Zk}_$Uq_FC)B((*p1}$`jhC+gy{JNu) zxcgRWKv*KR)YzoLx31nae3{1R&S-o!KR!_6GOsl*pTOI__=DSmzMa2t;mr1RqO2 zWuJc$xW|5IJ-!ELcueoh-5|@aEELTSQ}b|vKpO>!{m4^OSyHEJKU3Sj@N%XHu70dy zW@37y#~h);di`htbN`D_j$_rL-icI5_1UBVby38lcD?~Gzq!i*lTIo_dHKHOL*x=H}WX&%H;Fh#(R_{|- zHsADnKp)|Z>a9bR@{-ztS0iGFb$ku{m* z7fzGdw^+ko7;94!nY@gOB{L)(H$TROo-d8Q7#PjEi%NHuP&@unKtGz9dtC-bM3a6O z6JE1Fc{ar+=d(8&Px;>*n2@hzGJsFzp{m>-FE2I!yRZK5&+-_)V8lp;t**R(zxqe1 zE7436;=peuwEW#)D9UD8_2SK~qIRv7hZ6Y1dvq{`>zW>|IQ{B<947a?!tb`Ps1Ubi zf?K}mI|3C}`5A30r9PEraEc^?>SrNW@VD_G;on9Lup#TWq>eikxt6SQ*!RVv7LSiY zRq~zegb}ySTKZ8|fk`mIV&RLZS=VF0!j0Ib){C4HqGwVb zUPr*MVsNs3jOYEp3@`y{fyBrRJ$l8XcfU&hKDE9B1No$E#a8^7ik60(*)F7LeNR;9 zO##CJ4foe20(YQRp=zmRQ&r|EFr7YQSTnA?q*TpjngXjJxQaO`mab8Y*&L^5$?c7y zYVLOGDfL4WGX~aPNYTA}ApginuFf@S5g-k*LR#$r?8rZnU<$tF1mf?2k-U(-UaVK7d2=f&A>17crv)1U}YYzKxp%1 zywoya*+5%$CZ-6oHD*m2P^HL8B~uZ;h!gm~*n1DCD4K6ybjUd)5=JC5uCCQxRb9Pz z{c7*pRb5?OCFjKcv{|JrYSy5?ovx|up@I;E?)kuQ`dN{B9>c?s)N#Co`{l0*HkcCp z57;aqKYIKaLVRbnIs_!YyNm_$pb9aa@yKmR{6eOt?Qm`lFE-PP#~w~)y=CFwmDH6| zLg+9F3z6FV99%tumd3#!E}Ibuo$Lc4z3`G3p=YGC)9T~5tLJ6|f7kK&ik{_xhE?ly z_+f2#6jf;KL|<(~A)d}Lk9LE_&PCIXM%DQP`lsSPmlM^_9~#WoYB*L_^b3+x>_*VL zXCoEMlFG813QZ05nmcJ;KVAqlfAD{i@Zxp|Ge$mE&lAim5dMafr3q~*1!@!RIIR`_ z?LdE;+bwClI-Azv6#@37l1jIpwwKFVBgaZNvVT=dE}kD2ZTsrFA%$7Os1!r4%+Y}! z(ihGz$}4h4d|SZi!n;5&kn&!uu)6R}0$q%HrLediHY$xW3%WXCSBHjiY|SJSx*XaG z%N1z2WC#mwd95I`o`Xx$;PJGq#cZmtlDfC<45||mNsPqP(fnY8=wWm(PB8QH_EoS6 zHpk-SPMi3VCl*6YyL6Yk;fsB7MCEUpE03OIUSfiib$$W$FVc6=nA zkl62xay+S7w*cciXF<=I+dn9>2j7PmB@y^a6P1N4V+1lh? zcfD~{pn3Xk@ew8q62v($&1{K|3TcC~zP)uLMarYyeV zP2Q}4%jmCGiNc_pt3$#mza2KqLOY4Ow6(?dOBAARcxDxf(bG<+(eJ}N2W4VB%%)Sx zv8P^6?UyO0GY>vYg-nD!O<_nQo502@AhLomR_5P(HxHUet3AqL>`zJ#Dp_kA^@k6< z`xzyy^9k#DdCSu)P{X&`BSWGeTHhY;$Qh;TJ&EY7^D(N)YyJDDA`o=q+p1LqhFE#7 z#^%&1|APpXYftTiKU>T^Y|SknFaBx$k}^}Fd}90O`wPz2|GuY(Z~=2msC-fO>&rW` zBP_7xc6JY-)YB=DSwhZR&csv3{6iGS&7F4L?>>n_sb&2|6rX`!*_MFnL`Nn_U2~V7 zoA`k|I6%*M%!mv#y6dFKhLu#VR9??$OZ?2Uij4nr8Jp>W>Ey@Iv@v}_RPLg+`uuhJ z1OltK0=`a+(_g}0<|r?A-sZ9%5eW{*d!i7}X$X%*)%ES(xkdIYH|bO)xY+wRy5Z1> zSv0$NFP4*!TTr)l>q4}h$6A<0Ra$~_T!MSmi6y`!JV$ROX&9+_UKIH5r&+IG0SdNp zKP-`v%s2B*BMD<2nfY3$2~(^Nz$mbK87yuYdm{JN;=2`CqR6;1kuJsj6NT}Zy>&fO zuI+_hT6(K~f6DH#Z_Pw61H(kqDGK-n%6M^WpJcF&f@9)y z?lWgRSs_kudatSf8VVwo%(@TRBQf++^cx5wi`P{Vy)Qi3jjqR3UuShT$ljR1^(d9l zY5PQ!(Ouj>56cAKPu2(LDGxmc*L_Q_B3Z@?Pk~xpdZM-vnlQoFwN#&rL30_lhB%t; zPe`I+AD(Xi{FKg|6+T^*_0g9o{Ece0i|U+q{uOBK&d(X&pQ1xOqI0~8Px~Ifqqf0K z42?qD)iy^bLi?dE`omA=E}Q2YfgmcPqx9xVu@pEryr0C?B+5v80ZV@%BBxr70HO(- zxS%VwupKC9c#EDm{4mPQufB?3L$*qcBiLgh16UNRpqFmKaBjzj#gZ)`P$FKI!)Cf` zV0FFZ5ok#IsLa&cRxgM`H&io{5|u70JbA#yewvEf#%W)G-jWRr!yYTVPaYeF_Of2~eF9LT6jf?0y)sy^`^97?_1t@ltg z$3PtsbjD#d8)E%&_{e!xlaax$DUW(&3Xju8P-R3=)fV=O5ra^>b@nA;DsNC21f$2(dYa+8WRPC}Ft7Wz!6kV%th7Dc5 zVvYh3vFyz|$z0s(+p%rZd5lROz)v(|A;trieC%sgR6-9UMDEY-BM)}9ebUE2&Loyr z2><>pQ+mLxNvK=t!MG&}-nNgVXI2sP2&bn&exsyB#GXluz%#4*EsgvoPX z512#?g`#mlz9~|Qqhi^;*;giSw7B;KEEpSE-8ma%k^HWc7Q!B!ljv9-Yx|L*+tgeP z-w~D9bISR9zfzBAz4c*Uiefjp)IbKk_*>C-@|lZr^F)r+@L=%39Xy<)SJH-{+n<^- zK*B{8=3@t!)eKe<~3Af8kDL%g%WLbw=B*J-_%kFztx%+tZ%tlr|Y# znQ{+aWe34#c-jv3k0oWB6wS+7E%Hs5n!fW|P9Y?uP@Y>_u`0feHpx0u!h=Sl)jh>~ zzW`lxfHp5yw9O*J0Pn+;C9 z?*aUG=s;3=e!_gTo-DC+l;#In)^`oVDXN91`d=Nr>^hy&tdtaA2WDe>wzM-OhE1~x zP1KFXeZ807-CwMzeg#T+Ew#MqU7!|AM|&m3CLpRbKWITW(s_S2upFnM%%~I zdBgb551R&*ehCwn9JyB$g8UrKcZgGOXVt8J_B)sV0xTRR1@zz0fu%Th=Nf0es@|25 zdQZ62p?*mvswiNl`$Uxm}X zY|1CrY9qgXoyjNCruBo-lkIGfe%YX@L>C;H;7fkV9j3~2Od+CE8hE$1kE1RtXrNUH zq6q={KFYA_?l#U#AvQC4s`kk*n8j7Si3_~>+9QaCBg9Z^j4xV7vUGTh5k>?0x=GW` zj**I#1v&Y?XpMnu$w_$5gn1X|P-c#={ZTJ#88i>Z5zAU)~2}btEUJ zP4MA34F@zS$~>%vCxX>baDE(8ywjlh5@s@jX3gM9nv}Mr7|P7PZ+)w`Om48N&!}9g zG}OMmXsxHu016FPV?hsaFwdbk-wKdSTR^T@x@TqZtM8`e z4?WKpkx9qTfX^+ypOf@=9=3icU{Gl*W@w8ZM@ey_1vBXDneHvIr%bfcxHlP-)_qJ0 z40nGeGn)Cp^W)lxIa-bbL}cM8Fcp0 zC)9&UJ`#NKPyRWR*%#n_U-b&)X?PZS_c#|TYFbdcy;0>M*9$>-=7rbE<39r|z;|$Y zWl)|3tG+PvZR9y%w5R8xO~=PaLU=N#c-`D~iFIqAJYKnw;vPTyl+U4AiFJ7>r|@t+ zr%)vHUe&IY#_~3f>)pK{@3-InSFYY9i3KyVzH?3cm|EkT?0%8$_$d5NXAvW>U?o#`#RW&wY{74`LeECwwZ_UYtMLK6RHlk zmq%ccJ?-eV-rfi0ufO_YvU}~f{26gsl0yn$9Lsyi?M3%=)xjU*U2t*?8SCvQd}_PO zL+`pzx^MTT^GMQ}8FXnKL>-@W$uZ%WGdB}At1l|Bk(i65x{PChh`T37V)(_8eYA?$ zNhpUcK@-6C>lM-Z89y16j2@AG6!fKEPO8fy`j#k_=srOLth8hM+;)!7Kb6ro=qaxb zYjH38gw+^C+BSh^uLo%+GqYj1f=Pi%(af2J&=LfcaYnBlOjRP zkMBg4`w1Gtyf`~WxgO=s1+17x7Ae2hTQ*wOhGL0QQB6UuL&?HXUVIv+)76x=M>d?{ zPqR=#gH?Bw)?^xQ)w&HI+S6l4Q()>yROie*+|<+lnP!ZEtVe^ zV@3i@Hf2HeS0c1kp~zBI;Eyo0d0uv>Mz26=80+2x(g$a1Jl|LhKCM834xia49{#F& zpT@~|lAIiO$}-V{7M1SQx7BI3cNwRW6M)G5e$9X4xfkybrEj^9QOLV=H1BW)>U!;{ zcv*|u`+B+rlj2pe*NOcf6+OpN`ta1(0JJI0Nz1J9SUAo08*ZOuTR!TT_dM3NSTq_1 zWy!YqW0PQ-qjV#`Mv4Vz8v00iJp!nl#vh1 z!Y3uJ9L9Q{KjX|;gxSfD-tu}zjgsk+mUsTTaXX>Dbau8Cv2+?qNXsGUd|<|X{`8U~ zsc=Jw>3zb7x6Q7bU8z}d^HXCP4=kgXfwjv#@_9pXm@@j|*JYVR~Ho+lCqUvR6sT*3^c2bG{|T@T~;pkFwU5iEO>V*@7#A84Psf+ScC* zkEcjFk1ReyJz3G;!&vu1936FlF(Lh1ZIc3&lO_Osp$hC5`6)+VSS%FbUPNEt=K zpLer^QxaZ058_0d&>m;A>v@S&AO4v9DFJ3~TqHG5On#D&hd0bUh$6d=iaBt5YIW`k z#8iddQ2BBRPJQN^l`Rp^EX0N<$-CHPen8#CZ+fX-ahv5NeN(I;_G~oNu5_-x2w9;x zHRIMMK5^&q4y%Kgx-KgSM1PDUeqtkX-umbBjRuV_{uj-Vju?KCxujkLpaP+5sXlPW z*1i+)z*DL9M2MzmroZ&%paQ3a0NRphRkn=(OWi zv4c{BqD9|5!~D8XQigmSA`51|`?4KB=$WYTep#D?ddO*}Fzbgom8?$D0`iB%3f>>H z-d%zA=y$XZ-gO#PKM_qR)(_i0Z+5@mq}2PH7Be3_ zQP4^sm}HVH#T7>Ce+9Z6%YsVYiR(Q7_1*mlka_Z-p|o2_i>2JGpUvSBn-CgODf2$| z^uoe3U~k2u-G1Cxwz%H6wew%E$1T<3j_lRSk#j`1qaAsopL9As@+d3GN;c5ovW{*E zkZS5z1UJ6VuQmIT$vrUb_OknxkG-0kJ1IJ|CjEwnE>#Bw)kXq2jekfvri5tZY6pC{ zctbB3ZQ6xZ+)CK4?^l%IY;@?;@-hQseD(Gj7pvNDp|?H2Z)WC{{9|WqxIDf!7jp?~ zBBSJcKY4=0?{)WdkvkzB^rI~&rKROI zy@e&50f*3S^*@!jU1#6D z2fCkBKQQkHH+2h4wr6oyM@uHAs#VeX2hXpx?AP3z6sB6f|3(O8FJ}Sgx-A2&@5QgN z9jG~5xcAXGkleVp=kqgBaT5AsfqqEX5f#|I zIV>8c4WX)mtcseC>X>QcET%m6h3JNO1aXF@t0L$Us@Jmup2yG<@ST{n*eW)XP;v=} zJp8^YoA+$ENcm}UcCrtic6@Y&P%y^0tk1YZx=Dh7ok22^4#~GqadRL1U<8bKi$TO* zkT6<232)PNO0h~zTn2FWl;Rv|zO0$je#L`$Q914$a@}W9@?c>%9B(=gqDX=&owph$ zrt+@BZIS30a#_*i?Bw$7$oiSQEuR%X*ELm3r>2r2f2!e~h?{>;oX8HIP5EYL=75fB zngF>7+EG$^Y@y! zv*lr^pcu$X=|Bs-BoHe|@9y`JLP z4oNm9?zeWqEt+GOHF{6{Bu-xZV%42iRUA`KJ}Cz|wSCMOSi`tjf^41pw7}p}F&ZmA z!m6G+?vu@QGli1c5QU4B2}ToAEE5+gJ!C^ofN);0h(0=u_2Wc;nLt$;yYCP=?&NEU z5ZtAVr-}`9+;HmTZ_&FIwzoF9==)l5(1errob6NADUm^BDE>k#^-p^s_T93(NAitX zMY-J!+LFp4dWPv3@1la<7w%}9x-5w8TLp#i?Rm3NWBRpHMfq4+u`w2}N~~_FvrwUV zOV4H=?;g$5xRk<$%)+?9YFOypmpuYQc)dlB=kk%%CcW4(GnGV6GXHdcu0U={gKb&6 zx*|QpX{Tyw&n6qw-SP0pJsvtbls{C@Q^Z$~37$vpVEWbTK`vfp`cQGa``FT9Yldy_ zG^~n4fiQoY$hf*727`TW=+^b6C71xNUoY8mh0iDZ%PPRzcs9^lUm z3-*svmO~X=-VVVJ5+(QRh$q{=F(jHuk1<`TmoVhB&A_b$Y#swj)8eUK8j9<06? zq8YX&n%qZ&9h$#&#^PV4d69L-4Q|zXA~k@wE-IP+b?z-*jH1J~-d&#TvtA6kcg#yI zR3{%yE=wxCv_7`J`Sy7F8Q0gw+lK~Wr}_ThJ#^eIOPahi<-U37T&K}&y7yY^i^v;B zfBKpNdTdjm$p9YVcW?7Si<3=`A`BW0h%^0)QUG~Tn{iR$6qkS#_-YqeqSehzDc$TL zoCj=@iF-@CA)Vr@*Vl0mvkCkiNz+v45?}6-@alm_cnXId?zHx3^sByLC(w^V!SI5{T^uEp#+D2!aid{Xx?0J+-~sh3GU{hayhs#_h^*afOHlW3~Y|T zwRd#qT>M;DE%@QR8OwvmDKgrW>u6s-LArBkH$$p6o5--E-P+yqMjnXu==Jpzk7Ym# z^sU6H#7Oj{E14OSq2M=*?W3Hhr*JfF9N!STFKNTls44AO6MEjj0vKE*MiaEI7VP7# za{>XeV*bmfVE93lqVbd1$ui|v0>rfPHK1lr6IL1|RZYUf9JGx!(%>Pf4v*=?Y|?j% z1O9P1?b`;$QXg(hHe@p;?+hhH7KdxRAjnyi7wxhLph;@)3PiSG49<7$T)~tUKhVk4 zc=5u`q9>)GT@&12E`Hz}L(K3Qd6KlNS011Lwl$W&A1{$|Z7~$@rKC)Tio^LEco@>; z-pLeUJoa~t6e%pMbrz{OU&9t+{@~nGx~O=)E71DOG{fzV`vc8km#Dq>_WV4#zlr#W zUK02K>)|cOXjn1I#T*x3F6wM+old{1eO&h2V?ger^Y!Ox^Mi7gl5tG2@#`B8Ua(#m zX6h79ZykD|y}E@+uKE5H`{kdwe|SPB$21TmP3*JEr#ESY39+Pwd2mQSgFe{ZSr(}u z*H++T%>n*z)#K&Qmig3Abg#(m=u#_B^!t`~NJ>0JUjoatL}`CmMgr3AsrleJcZjQJ z)5nrbdfX;&(XdZz+HQsNYW9Y!mYcW}J}io;1}maQ>JN=a-^ycOt8{usQO5T5QCNYyrs+QLrv*ih2IYh<8C^Q5nfRK}{3bCL;RZ5S z*%Jk6_dafXwYMv+OBKr|M*1JtMy2KNhX#DA)x&UNbIh7RO_$MmlF~t}%ohED_N~OD zzy>@gN>pYl{7W3!I=)|ewLESroIzJ0j>J+v;?JeezPDePH8NmXasD)q&@&~YP5BZR zsmmRw$~0F{k$;RqBK^eIgiSfgErrA<2{c7lY7#%Nfx!5jv$3Tbrb3N@~y0vg2b z5pijC``a@6RL70;S8R4ISkK$zI)*<)-W{wsE@&(Fi}f5%Ta;)0DJ$3=+>cI1mh+-p zMYP$9k!SU7mwb#1W0pRtSSS&L+~b6ZzGPu(;RvEz2_hl2wqmw{Wt3<4(PWpllw70p z1IDRq3Cs!8T5+E0FgnrF2u{pcShXGZQ8eJiP-z;Mr2C3;`vypct#=?{OAY5l+T%*1 zcHXZZY4A0Xr8Ib;WvOIAZb=y^ta+P_amIAPS+9=Oz2prAc*Ljx6?R^uO z{~fF7uy=2UJQqK@%}M1_Xy=}aGq6qE8`(S_z?YwrMpFQ9qe@-7phrEMPKS{Z$+8C<*)!6{Ok3gu2(#tKNA? z_YOVf7)5rN$T*0JCHtt$&=G~VEJvy5NG$U3gONiNHYZkWul zGAHg=A`ts+Y}56KEtoZ}Ow^(DDSbREjs>B8ahVj2t3TavPJmG*Z?45WsUi{Sfd5pa z#$ua}<3iU(y2CD0CPCa(lzQ^&AVyBimUME(-cZbU?IuQ}0-jm-$9g`P$;gE~F?9Ey zQd54PtK%lXeFj`OdcM#3cU(TP``m~KA}$>fb~Ja8GY`Z4zctwZvcN>QXUd*ngK|Ee z8KaW}HfMkG~?cMFC-oR>DL!@)S0u(6XnFuG@S&w)VA3~$D}tIzq(U5a;n#UORA zdQBm)Tq3ISz*}srU&*(+Tw(r6X=o$!^Q{mZ`urjSDJeS3XV#F(_mWX7J9Ao<;X8iK zO!`EQv5C*bx0)a6IPL4`81Rk6&8cn45ik~(zN(yB=eZAie))Li$Gzh`Lzo*mtn|H( z;rqi6`R=HJj*ud3!ie@4WkND-)xZ2UweRX0U4g#14o=P16?MxfqdPULv!@9vx*pSnD?pf$PN z;;>m}%=|k|fkKfHOk^W&B@u{mc)QCItL2$ZvQL^e@lQIXN`q3(m=0RmaBTJ%QDWKE zfXv%^7KPot`sQPOclKrGc%VNa-8w;w)2`&^Fz-RD>hJGX{lzki-75iCk)rY8zm~`f) z8Wpcf#l&bEEv+o56q&`WJNkI7qx>`~M%ghrhat+J)KMx1fQ=x%I4fp~H~gIoT-g|H zQJdsmsH(@lx58Cgv2zXRl)Ud$X%(Y8SSFw#VJ?oY_vF#GexZbxXk?FZgZO-&R(ekf zp|fW57l!Pqs+mVZpSFdm*DHNj>z);#qLz6aHyIR(@3NQ?qf2()dcHX;zaQ`aPQP#i z%fouz-k^o0SE|Z%?OS6zoj%b(1lQrTYl&vC1rB(IK)6}%vMu#siDm5@aobN%LgpR0 z;-}!PibJlomtiN5O)hmlwkG`gj{#|tlYhi(RDbq@*wC_Hf#SbMn2M~K<$t}@j4FtW z!98)arEI~4V+n?@OcY`7_$*)iL3I%-k78s?%xTJO56N&8_i13oQy0my*~wjWtWft~bbN(fY-< zqBFsmf!Oj8(#Z8%qHUEaG&2str}Do%iu8ygFpN2@Sv$o^rmhPn4VRC9e8D)T<&?2#7BiR}w1S?Q?bWr-OjCD3^l zqKkSGm+&kX7O%Y&7v37Fo+Jwgwn25x#mgF`XQSY~b^Ddi;#Q9gX9DT;gfWvZp6KZQ zbDke=GnII=yV%~9jB)Zh4p@~SLuX;tLB&=hXZrxE*^IY5z8k45HPD4krMERurcPoh ziu%tM`9ozy4+;{LAk-75#GqP!M`gVsy5?*qdik9>hmUo^F85W$_nd*%FNF4^Rf*|L z{Q}c-XemO!6MWW>er6*r=swcucVy6h_IB~&gTt#W_|g2P+pFPi8{ixKMB(#$Tz}u{ zx`*$w3@ISjsZ-aC2S<4eWfYUP;KB-o=7JT(Q2>2kNnR#L7kragmQP@l=&pp?>Ds_Y z-HKK8oC_yAkfF0oUpO}gW<01Jzdk6psjTL7GTs))r0X#rTI?;jao@Xz9#O~ z7-!OG4c~Sw*Kp@UpRZ}QeU8suiX?73nxPK~#K z7!`;o3xPe94aUM9+LDNnZugzRn4y+2U5GOoh-5#B*^V5hV5Q{eI2>*jAFftmI^4Dys z-D(5+cK$Dhv`y9WKVDI*AL+Ymu>YGvXvmx@wtdjdjiY3Kr`Kex1~|yzE(8zRxlh|r?;`E z6Gw5cKaVnK4r7MrsJIFQ42|ro-nUlUYr?b`wD?)`OKO2tgChVW^uZR_uh0GB1fPbcdGCCWXo_zs_-XX0iq&r*OEcxEPd0s5u#d}_d%UgF8MLX` zBOcf|C)W3XUVNQ$RQUeu)m+{AM1kV&ID+kj51|p#B&tT54KsOYqU7A@F;WK@lwGpu zKV7+z$yS$318EqWLmTbd@`jVc#wTH3M!pXRD$e-t(q)qeOkp(2loGm#UxA+ddm(;{ zj}2-cre+{MoOCKJUP1DlRv64xPe#{-__aT3FI;>LciUU(^*7!%SsRRfMT`hUQNw9_ zlQ^*$M1p?@-H>(N!jee*!r*THVHPN5!Vv^kYxhw`61j6-TT_!V{SKp#y~$H-kdGLJ+$j^skMRssn>n{~ip>j6 zcRf3-C0}lE+SrM_IL?tc;C$5Jv^JQ8xU4ows2F08cb~5W8#FDJdR12f!!o@?fA^H4 z^LA_QKlYQ3jQ@71F03T;reRcm<38K#0#e@ve$?$feZIqUqK6#kLgANW@{{k`-m{XA zE~l~~JX^erb#~D%CMTI9L22>Hw1qw3waIK%&GJw(o`|b%Q8u!@|iNmU6M_rFq_$NQ7D#CWYEssxy6< zm`3xQ8Y?iaoob30lC%&tEz}oNUHi!Y%_?-tjuy5NLnk>Ero_s(z~BKHL`L$?LBER) zRSFzTGu+sFY+GtxOzk?Vb6^$_R4bwozey&aJXsH1}Usk({xdB$FVditn|^8=4d=NT=NpT{@AlvZ9lJ zmq?d8A-OSLi7Bk1tNa2==>Ms^(qQ8Wc*w}+ZvD^l3dB)Z)~i_d9N{ZaX73AYei`sz z9bkre08?D6(Zldu)MrE4ts`;bFH_H8Ys3(}^o}8!2ed~)Rtm(|` z{o@+syQEn&y0x;ZyO?72HB_v^-6Phd(TwAvGq3cL!o!{rSvTSEJipfrG*aP;75;@9* zeuCH~h4V0U$N7*uE;+-6m=V6Qwg&a8p6F>nt3m zct%bhW{7{GB&3iPlcOL_l++~(HY+8Lln$;*KC&W>GxN@?p}9kwFScZCzLfX6S2Dk6 zU|n6-DcD`@7Lvu$5M?)MVhopxg@sG*k>X;HX%m|hLf|~M+epu5ik)}8u}EYBjmwyi ztXJi}YL#qse&V?E@(uOiY@@`XLH6m}u?ro`S0yzfTC;cGd?%QG7WDQ1H#dq+s5=v& zSs%E!WM8sw=WcxX)kwW??m7DRTS7#lMdhLtJt^l1b#vSrFhskN`XhhcZzSq7yfBk_ zw0lv(agl-x=G^1*l=ZYP<^5C1bTu57761M0ZCv2-mh zxgnB?LMsHzdQKY z(ZhDxiKk{=a$25svT4b%){1(M>Dh>#fK)>};o!C{Mqz%()9-qe**^Qg92%s?JWvgJG`k$%1UIHbPwY0(eStX1B`81V>( zp?R_bv=V%+h}SOQfR0Fe^G7XE@(?@nU#g-5mBa0Ec)SA{DZN?FsWlu7mAbs)b7lNN z;u+{1%#(afl6#@JoD44^>L#(kHZ%2)ebD#ngdyLMEG&(NS-`U7(DpPwih~f0`S+w# z#4?sTOGOFeD&%(EriC5TK3O`mEWbiu=44tH{#TwrzWl(6=`YQGdmmp30v!yy$z}(+ z7bK$t5^Ji5*wrpMPv zBW(#;(mdq*xlwqEqNPw@>s;VH;}BTa2sMZ_euPr8o8*1Yg|p9#vN#6$2`vpxc@-Mx z+br2QxyrU=bFhMFT=>M}&5|(?UZRhBGNz+CM`%7>OprLv7aQJ(TQX`)KUoM_Ku@t` zInd}`SHC8^pxj~{QWiBF8`mNAd4rvPN|AFu>-V#QPO$m$nIi-0;F4ha*q}%MF<&OE zSZYIYwIc7+hPus{_oq5k+F1;a({15~yFHL^AIS8ojU(#vpT*vfGwa2W7YHDq1+`a` z{Y)lEsU5Uqi07_V=o|z_apL+Q{`DOr;6bZ`cG-9)5Dw=ynuPeT0+7&cI{9iPG7g{b z&e0UIB}k33F7?tbsZ{1+Snc2C>)m`V$~xGCAwF;C9(bF@uFmGW(fhF0-D+$mxcqAT z)-riRTKk8yI1#UC$4ocH5>3H{GI|1mbsoK5`sd>8@c0J5zEN+PpZe)p9Whp2wQK^< z%(TYu^3~UuYt6JRnxr)D*d-_p71Uam*_b`yquUeuQs4_wX13GZ4mOE167|V7Sb6#H zb&O_+uW7?SdQY;194GcY7j|`{kGDOxU~13X7#d=R_92M-WFY(7eugBTd3jVpiKi&C zU&ss#lX!{!K@yDF?o)o>u$_HR8Sb;rDfisUie_+HGMl21=9Ij!tdg{kwyNlvEMzq* zgx*%IB19{i)XE9b;uBSsSqQ=GxAPwqooMGFA|6lwf*nxtzTU%%z(Wdt!8;k_CzTUg zUr7Wh0UvDW-Pz7L_{rOYJ1C8guV%s~hN=oq@o}_4_mTLugvKPRKShTOGGe`rJ8Z)( zt{7e-USuPCiO$A*;i zjUAYQy$XSa6JFgT==o0NFl^5^s$XH*CeO0!Jz?V?duM=LsX#+KPR+8z5ASF(&YTWu zP1SZJo5@Jfl;Yl0K)VnF+0)cWG|V3q4_v^T?Q~4!bfdx%N#CVtNuk_JluSqoJmb&+PYr@xADeNV-JSM!S zV2%V>VZ?Gat;)fwV)*!-lCyY>mY8T;yPeAwC>TxcF{~a)v*`F**I>AA7hm)1)>{ z<2EF9*vTW6|u&K4e#CrnX+*ZtTPDypY zZE131sR}C0cv3cspSsf2-x-m~axu7{Ph>Jyf`5o}EKxq}I7JI0bF%mFeorD5C+We; zg8?^^6CS7byvKl9|0vk{4#S;FfeBOgiKGy0!9)Kn@7Z$kRjq%oZV*oPNef3tvw)@{ zzB2wrWzpc|9C!doN=qPd_6*N8*j%rdiu=QA3^~2JwVV z@WP}qXEWqQI8;AYZ7ZIyCJ?BtkBLK}m;~~58|iU81H5j%tY&pQy>UG9L>;k^L|Jp4 zvippm*0>o0X3bex{qzz<4~7|Jn63NCgeQ>5Rt{4hxS;Zp%0H}S>?V6u-D?u0|IP)O zj))@KoSru|ZrSxBcM$dI!?(y5L`lQeeVn{6OW!~|ngtB*vPQaV;M*)8>t=;?>BivZ z`;?%$-Z70|>9hXfZ)#6Gea9+!{nZI>)3~gAehqEH9+ItQie8)(3YwVoTYGVhI9Yv3 zMmbq%?hqxtwK>wKwI63vhILy<2dB#l?s|{_)BgfEQgzrGhf98#i1A`}tjDiy z`sbOI;G0s-!ojZrZfHM!L=={+AYKcM!~-`V`|8aT_`Z`n96 zZ@2=n{cjRtwOAMOE_Xys&;EYb+nG7CGvsb~B`lmiHlJBTI!Q5VFh?yL9^A*WZu=`} zL}dHv&FKH{j{yIp%cp7rnzse)EtyArHq<(x^{LG%8FkL)dQKA%A$sQd^G2QbyCFS^ zEbUpwu2-PcjBFTgNU=WAL}JuX>G4k?^?dmR2rIU3guFs=KzXM+i;Xs4Eq(d9(f{bT zbkr8V2mJw8pwWI*9myJkiMjJnTDYC%%o)6;)z&Yij>^kFY#*LpylQ>?zrv%_C3nH_ z;ILpJ22iH(@AcR=C&G;72W~Gx_tZ}WxW%%#gwI0%-#p^ZbojR)eJFD_tR43Fz`rfu zuP>|ddaMN58!7|L761Cc+2g%@qj-ael+&xw)r(vGr)+UQb(lBbs*FF%%=tgj|NE4Q z`o`uhY(~o;2IhOkiTmY6|6d)sngvR1>vF~ zfsjEcAP|V~U=;p8<*q>gD~_KBqu4V4$@);{k4NFVz&G}rkLtj=?RI=S9{_0o;2)8H z-Tm)*c(gnZ9$%MA6zK|3JSxDPBt&q(@cdwuuu~C69|ql%0y*M1@u7ViSK;gI(frSg zFr07lx^xjl2En@ko$KJgp1I~mNf9xFxLG);&O>nINH8xVB`eV4!1wSlqkGx+I^fWu z@Lg{SP@>$an~oCiX&w*Zvfk34MWx0A4i^$I;~lkoGsd{UPuY26@E@WFH6A zt_*ef&$A}-_=Yy(DKPLX@9)R2G47(kkw;K{ANlL3VjA1wfjWOeo7i)9X`I{P5_0S$ z@`OZJpx@jMM%kSQ;SRm1PJ{5OPIlf-+`lsNFCGNQ|3v;yT;~U)s8xO7-6@re%g4ZY zLntMrI{3Tr1Qy$qWMj!_Ibcg>RHEjyotFQ?z5kN=A8u{d0O&}ffGi>uuQLk1LU5hk zL2QHscQF~Vqr&McEPaCDuV2G3X}_O8Mm5^ktwM%5_2NRipkZZ*)GLVR|AUL;zm9|< zC=S6UuM;$wpYMr+iREE?%A^i>q)UDv6I@c~KN56~Y`#3Z7T;MgjW4E&$Q5W|RA;Wl z)c#MtbHkydn9C#N;uuhkh=1bOn)Q1PM4#F zmk1YrFgiMawBV^Y;frUeb)E0usdszfj-$BTo$Lq_|8VW4lwZF@6B;`Wce{?dO7o&LDczgHbd8j0q2zhU+Au#FtpT}86@87|~mebW&7jxebEJuht z7k<;#02nCpY21MLbG}9l`JK5@*ue-q7pUV1#K5!QgJC{|t?7YRU#CY@N#?)#Abhhs z{|vtdl(g~j#VkTju+ziA-*mgkQNrO8-T%OcAcp|GAH)znz?ahpljwnw$ODlG&>Dld z(0-UI2D@4x#5?NdDL}M&*LgS?#iRKhSHBVo+b9F(v{X8e!at9GQ91X#VW ztxwdL^}#L&YR$%bM261pB@;^gWl8V>w)kg%s39N@!uD@;3`ldWX4k?5>;T`SB5Xw& z-iZfY!G(7MwX?+L=fP?7&fE7J&40^q{>IFl*xWB7bn&_Zy*|AHeYosi1oYi`6dEmy zn-xfCDhiz(_WV<}2r({2K_;VsHVL*80R{Bux9;}9vU?#$=+LZ4n)Xn3?kJ#9VLfdT zF1fK~0P&mS!*AP`Si|arCX)GLrvwrIt#}955g^qG27X232FiDo{R4tB!Xa?B!}q08 z)dDf;qEJD7h{zkIUBM$aQRF+oV>sx9#tb3icQeZy#@OOMn;msg9H<~jBltBTBx!$C zL@1^L1k10hQ#oKXHUPWv8L%GifdALdZ+_R3pf!hL($tAXAEfcqN! zHG>h<+^D}e{r2ofbpTf4FTj9D<1d7NN&#_qMC8LG*K~r`qiwhZkq}`&G)tN$7l3r_ zPVC2zu05?cC5kwgtH9e;fVK%dIrwi~4hYcpz!63*63D|}R_{O5$PpS|L)^G1L0i&8XECDfI5WGV8e_S(NxZ!-z3*H5u8sy{oS2M65bsIpT>dinj3f_%E zU^o2@6B9)oGD*PhtU~TL1Q&|}1Go0eFV05qZZ80=u4#Yx!xB)|^BbU#uS*qi0=hu( zgAfG8{RqVQACBv#J$(P42Dcx$dyTUf=wYz~ZJXl=*mRVgM*@fkZ0kbPN)-FqD8Tl* z0RKh`%nZ4PT7Cl(NC~W7anof2*wL%}lPP?^4$*vS{7X*$)XP!a@qT!YEX{;G-$6SV zNCrUN3LrNff%T=qH+)@deEI)K^y4+qw|_VSYILyxi1UUO=^xK^aD)cYKNvxDb8g%| z!gC`~Fn2HR?jWQ?0n9xLsRrUl=&nF>>AQ$71MUNxi+2kxjo+`4eowqyK%hMF`oj}( zoxY|8U%`cTp#4=V5FVbW34j@b;%^5AJmWx{sf_Rd4q*#N;Y&e?GjaJvNfdNX4m{q@ zbEpdYrOQv)3uh3+&9|SrhMD;{OQ2|J5sw@U{K4}F*nSk~oQv-O0}hzCIAz$%Fla zBwG#>IS4+63Tjiip#$F~=-9`?r3+_pL}!m4gj5aU0%ik1mt~b2BT4A>vA4F+Gh}r9r?uL8mbz|shhf+NP-W;i z5{=0K7)1$KcLbs~d{=zO7|q5I0rYQ{mq1>A+;qPN2%`hRB=RuAQA{WZ5HIuq0TisR zz}^u7U+TaV^M@_tfp36542S!mz$ca9JhG4_AHvFT=)pi|>s%@CF#G->EV`G^VH6JS z1$w6(n{$cTv%EkT|2k)T;e317iVDtF;dAH&S9G#F4uW|RpsoQTnnTzD0c;WvHWq~_ zNCZHs^d)4>MN~dyMaW_lY)pm$g#iAb`JzM_BB}(m2Vj!xk_HkVjNbZWfB$T!#Owb4 zP7`3>n_AEB&`bihOrCTP12=bwri}n}(8rF_1tdiXF-H!bEe~15z@_bkIC_I}facB%j+7WNJ3JgpFr#$Cq(L=S)$kU)jfDfzR|qBG^@JPnTzQFkQ}P)GrA zK-&boYytQVdSM#B!2yjv-vKVrPX~|zWpe|rNg^E9sSHK#f;va??RBH>4MUY+YeUaR zy-v3Y0Ade#M+E^^2czz17&AMLJHa0g5aPQ$J5vIzPCg*fh4*i)!n#N?U*qgxZI=Az0Nr=1SbIhZwl8PB)-W}xVe{?Tr%-` zGyBR2e1qD_Mj5_D0HC{3@#}NcHAg@Ahd@G}X`%UCpbTy9nd!tCrL zb;gJP@H($XJjK2Tyr8K3ZXR*~OEe%}Uet;%PwwagZTF_xKMN-r#N3nT+a1NsQD$&L zVQ-g&RwIZbpjGrkCsd#)IKa!sHM<+hMZxzv$yIyVc}Li*dfDT8aH|G7f$qdL9L(#M z;F>}J7yz5c+JtCheNc_0$rpX0w0Hm<02RE*n1DKD3+~JbOr_(#vy}b$x(y>hCE2lIT`=~5kUmWC_&PY za}U?t{sVqK~xNYT8<%{!y0t)Al~xL7*s;2I7WdK zp!_BYP>4nCxKAvc-x6R^Dx$az6cN)fWa6<30_9lLbn3S-SmtiVV%;s54!}j>e$aRb zgk!e)QHdB-B|tZzlOAkWoBr~5`w<%*44W!&;T||z8yg(k(4hII@4w{$YJARZq|JdK z1Oc?-8w^;~K&1oRR@kus@YuJa*f%s7wjw^PwTW*WhX#P&gmr-7#M!V-xI^9*3HNhj zdQmP!+*SbqC4oZ9>EDL=J`ElVB!*ZVLkuh`5SUnmr4AOgbyyUlotdHyI@F3_RS_?j zhvzl%J(Sws91!2q!lJXGbF50o)EEx2TOt8CixBF7%V8kEU2Fw)@#%1sKcFd^sEueu zJhl+aHG>`o!5)U$4wTRUA7K2Bbq#dT!|l`qA3`$-1<(ki`G7zVb2qRWAAoGdV*3IJ zxPR%w|`1l z-15Ruf&=h%ARO^+$$Bi2Zxtaz9SjvfG^qo%4ur#WP;=ctWC6=bzGL+GS_V*oI>`7g z1^|M5-Upkp;#UD(^v6r~t7rIN*A?VE$6u2*cr9-Q3vO4(JUF56#v9u>OigZAQ^p_K_y4(+m

7V;Z{%4Ni)<*GaJpY6^L>LDz_q!>JNdQ0|uB~f=PvYDOqGeP7wHTulS2jUkzY6 zHo*70UyJ2i4}1d%OLm~JvWHDrbpv8ZPJr*yff5F&(D7q>3q)>dA}Y1{<_1tN`T)dG zTP?#_z5yJ>F2{S>UuXdk)93@UT`zkFZ~bw1-#Y!41GtHA1t7O#DQ95Zx|-NH8rC`K z$9jB#iy|L^!1QBRF5j%=928~MBEF91-cqKCKb~N|R3B~)Fa!*6m&YP4aiEET5NBZA zpM3j{z;~))aqHkKA3zC4@d5JY8Vk7DSiS*z1uSP~N#KB|j5QG7CV(EeL8$mF3350_ ze5;#gj#zxS58xXxQHP@Dx_}GjVPeFVHUe`zMm}$iEIwq52oA(xg$KAPx=}Aq0Jt%u zdngC{#mm)cSR;^CfN=k%Oag4@)(?RUhKOTSuzv6K0Q;KQHU^NEP9!=O;iSaAA&&$_ z98)b&zO;vLvjd5>M#J)O9j?b;1m2OGl2Ma+<0HPPtyL2=c&S-MES@eKe=5NpX91_2NO!-L0%YB=2LM^?l_(K;yC zZfuxtt{>lNUjRVIhZt83D7qNpp(_cLqO5lFvqsL0}v|IFcvpJuS_M@ zp8j0o9-q_#c^%+I|CaXGqP7jHKP0_cjEi^4|0w))vCV*y43AsZxMJm-3mbHMtghY}0fKPDi8uT&3& z4*}|8%>W5IfYK%!enxl%a8g=y#D7o&+jbbJ(5(sB7})>>a~lngKeMbH(g_L$ZslOO z$zbdjQ0{dNc9&2nU>E+$SOdqn_4BQN2LpLLb8;ID^!P@I10jQHB199QYcb-@s(gaY zqy_`1`$|~Uz6JaNjDV8O*kjbb^#Ir>?8)Du1_XG%L!yB_0h%8*PCE}UY+wMff>(wz zRP_KJ$}yw64fAvufI74CZ~*>ktlhyfb^VwG%s&W$_%`&P^JCQdP#8^wfhOPuQ)1~q zfGTeXf~jaA%7-eC<-?ks0rBZbhP)2A0u+^xhR6Yxx>0$3+{4GX0S&$+R1}O%PoI8P224SATfb{23Uy z5C(!!fD0+W5j{}B9&v!Y3bEXlfLo*vEI^qk}T03$+#!fDFY;gxOB7BG^Ti1M-mU!^RF6b7=KKve+(8Hy?sL(C08 zta_l}Fe&e!Lv$bh8>sal0ImVUZ7|dTIQ^DLfvIOs2Q?h-1L$pDpZHcpt0#AZCtJ*| zxre(%zlM19bKHM&@@Lclz~;1>13-8=*s%cN{5B#pFhS%-fh^@9_Z0yDM|?D^f7;}_Uud{L|t3+V_5bRDhSafIyc1zq1Z; zY*Vlmx-$UWg@Kb6e}zAA_qaNgfF8K@mr2HC;LO55{KNo4YW-JD*0;aExes`G-@l#y z6^rwip#BXOe|76?D~OK__{|3%m;izc27`g0egJ=9;N#HF>w2Utvbq&R8_Wo~oPhY| z84D-iCn8WVP)&yPE66{BH7u9=|7iA5!O`LA+i%IfIv34Ei)o=B&GqdyRcj^uE()DXf-B1)_C{f8C{Lh>DgkY1u8Y+HvI%d>vcQ7palD~Nva zM(wT{Oo=sKHD8~HGUD$~=xaC=xP52H|E99TO~N3dHybmY^L5^8_PcKeqGQ1K1V6El z2d&7{v1z^Wt0~;Ol)R=8;pkH89gq9s8Ee4?V~1*3u8%tVwdk53O$kYNPR+6n3ku(j zCw~(xSoqG(PU|xBP2BDi7adpdQ|tn7Xl$Okg&d*nP`)8ZLw4jI-JXXkbdv`q{CR5yqF zAX;pk-GGI}#pJ2B&}gRf0Z{~VhNt3NMjLsD-C}%`ODOj76o+D!x1nG`im0cOlNjey zohJ9V{|M&_f`;`>?3>~%N9Na;5hLAZy&U5$f6V&7f;8WaV{^xC;N5DP0^2SBkpH^- ztg_JLM5-M!-fl;T?z@@I`t=1tZP*m?%CK<&$iAC&-3^ z!xmdw|J56Gtv-Z?Ow*sG@=F9C65Y5y@dx0-YZrB@_HVln50PBdovtVf4>6&5^u%J8 zYsojr?BzGuCa#8EYm`OWEc%B`8LJsnja2_%wiCMSQqU_RL-L4iDCI;-xjtSj=2exdr8PULHXUPv|Se=hjN(Pk+5Qv;{ePp`G&X8~@s(*wXm9iAP8M8vZF zqEtJM9yzo2qssK4)@T4Ode+YPF)}h*CiO8P_8`qm5A0Z-V*oY|ed_z*7^5QEabYms zmsjr2>qtD(B+-6syMc$bUMwGcPKuC8S?OPozE}Rmky!cH0Ih*~;dD6YblI64bPZKc?y>?iAgsB4ALz!HFE@#8a>ukHV+s0Mrkv(pfG{m6HCV+qAL zMpJFJ_dgUC3z46t2PodBh@#E=-Bhe-9t+H8lZuk{Bt#^-ZDkC6Fit_f#K^}8@6hYGx+|h@uH?mC2F5@npyFegQ^U&#Qx4+{AVBa;^n@xL;OkURS_~okE86kH0cCmISXU4 zC$= zHa-3sVwDKTTWOOj-fYj;PA&-E(7BaEv-z7aI9<JdW%ZBUk-GBSYrD6U{G0gn+XFgiTg@#B@hMgnLr}f_@`eP>- z1z{jJB@7QIKsH^ZeV)Z!v`nK=v!B9apUP#>o+f{M10kDAE;+ zP%DTO4a;aVaXZulp-}OMLMJPlN1me9IY;g^l1p9VQBND!!8I5I+XGKulFTnlt+Cz$ z9oSwsrju^6Je@NQbQXTwv(A~(phy>ICs&N0|EEKQ9bpAGxA&N@gjD>U*RceFqWoga z+MM2=z68PP1~=4&>I%n#&%UQTxNER#sq^fwPG>b69z7#6FI`nQw|o%ya=&3-#P?q< zP(^l?Eq#DJO0wRaUsX0S-vwo8?ra`XeA)9)UZyM(q4_4b7u!4lxk0R~7 z7FS2~I1U_d&i8C!$xx>{l3|qw&6L1VLb7me?uEJ!A+XPL83%Ay+(j|~ktS2Pd( z2o2zT3k`q`ji{9L{*Pvt&fMZ8479lWciEZtU+$uushXMU%Sg~`@j6Vu4PEsB8gGcUDxR!e@p{Df%H8Z-famVa>ZLUoo^#@ybMmW14R?~C#fhZv%-Ji-&FeSo zH{cmoP-C7`%2=$)4whyBN9Lsr#!jiYo6M_nm2Xp%+JcBx;%M^ z&ISF}1b?Kg66mEo;LHEJZ0M1?fRYRw zIX9s*%c(2%YKH1rCu>mggr^OOu70U`OGZkp=JjNsN0L3IEp_+hjpd>*pUDBRT9S_x z45wfG2A9r?(1*F!6_SqP9p=t0jtWZh!i1N~eUFVZi8cHG<^PU`GGnH|YrOSizi2e3 zu_9DD+g@L$QvrbZwD6sTO73ODJJCZyM?U{ht+oF7_i8#k(6=Lbs#&GP&J{xQ$=_rY!yj>a9F&`m~t zW*}A#N-n{#@*Q^w69rzTP_p}U`F6Odb?bbXO6d=QA(JWhP1`+fKXO1$cIAOUfExXx z`b*qz0u4FB@He2S9s{kQmB!R4n5y^CPe}vIl=nP}rBWYza<(<-!s!!Pko!MzA^Ucd z^}c=Dz02}^$swk&U!a5$ehD%2F>wc)F+JnwOkm-#3;3p=IDVEX<|iir^a)lq!t;NU zG0gwosP*f%YTz45at8_J4S4MGe}Zv&Zyy+lWITo92SH=naf|(zEhO&}7yO$i%*Efp zk0hRRt{J~ErgIXYX!vQkIOPvv4Yn)6KV|>w-~5+1&Qm(GbWE8@Y4u$!{+M5%=-mH#_ve@hEHPZ>-~7x zUwX4Y{)Hy(%?$xPFFQvSYbl;F?QRc)%FLvd!tWRQp|gK1_e-1`+YMgWR)81IU;rl6znrrJ~+DcXhq(@L>I zChB`#F}MEeuBrA9F;n*2c^P-rZw2kIs*=jj^+RCTWDe{)g=Kc+ny?m<;vXnDjWszx z?P=J+(VI&6z?V2DLO^=~wfUyktu|jl4j1rkjylC501Ib-=;&{>W559nt+T`r?flNx zub@g|Fko3~GX4U)|3AS`7hb_rGHY?XJ@YLCE4@tSl{YZD_U%&3Y|apv)P7?EA><93 zrd=?~>OaotB}UVA|Gv4X6&kR9zX{3@*X4IWWIDxxS@Mqz*V!b-&SZa(>cm6z-}csr zl`X>Wy`361d`Ic^OldRn9d)-rk~w#So=@)H-(K$4Icb|vD=Y%?`~ytD?IHl#)dfEW z+%Tu(Sttx0!u~dwLfQ;&qd+^@?k1S|pPbN%y&SYlzej-Wv*S=LF){b%jnV?ayD1Wc z#u3T)%^!X~;lAL;zi2jCmvd8oHF)AgE%nb!Me7%u$QXH0gx(4(I1YXq2stxBClp(CG<3^h_%@`<)ai=c4ibp(d;Pjr`Y^BCTU zNfG(r8NS%7qkP#*m}2N|R#fZuRjxciY1%&&YB=P-_0GFZ^|3|i+VfxCVf5&!E&y~# zH7dt>iD{IPFEtdDexZA`ZPz$d@)e}~*XQ&T0+Wv_Q}&sbG`^5*-#-f72fo9p@P)@O z=P0Rh5O|BQ`J&%_fo+)=|D=l8Qq8g#56MA5Udnfb^Vv7Q(8=XCn+VGe^`L98epiP+ zxu=%|JX6E{d_l^a`xQxJM)g#5eH9p0Q%1mxefs#t=;z$iJWX^Zw(gBRJo1&7?&Y0N zYB2tE4tYd78Tp*ri+|f}v2N&Q?wZp7`qGzUdr8Fn+-4h({)r7o>gEZ-3VfFQ#pOK{ z^I8MvAWIr$vcJiQmK3@CUmE6#(Em68kL7bv9~uB_KYmhRHy8&D?37{mg~7mE0l;J+ z7C|{ZQpOzKVVZix^leh_I>m1FAoH_ z)71R&c(?b4zwqrCd!Lq0&gvepwyCTdyRh=$e+vF{dR!*8DL=eAl5FZhpho3e0s9Jq z2~V2We@AUq`e(`{I%e-F==Lyf*!VjQ0`EbnU9g0>PAFWzqQ^DZj(BNu95sa&vixvuGUXN<4($P8m16`&*(2>&%BXawk!mU^c9zsUWFRHrn2tLHLb{a`l zCHqE)v|`mUf%^fEs^RHz3`r1WJ@tvhg*U9_4+N6?;4Vxv_paS2K+O|)z%qT71XGJ* zpOK!JzYJP6i}auuDY1Mn_g2tRD?R&>CP!6nq)F}>`qNF<`)prdF@4TPBt{5+OQa?DUJ$pL_lz9($yY;|_ZPC8zY}0IA_S`7N%UmPAxP7cWhDcD! zBAlDf4JXjvxA7o|qJm(*wdjS9#B%1=ODVtR7X}UGl7fDGB=s^$uQb9!idMQD*RMac zJ0R%_MGmFXQ75Xo2O*l1@%+o|U#Za~mQbEJl@gm+eS7vf)V?K+Bh)5VAbh0rfsVR} zG0~pPx|buTm!nuuOfZe^$;kw?6$699y19gi?CEMcWzFdRgAi(W2i*4-^F(*ll?=1- z?ERa_6v@@7!Aj(BMy|XmGe{oQ)b}bi!+|-R=zd01EPY=hFOZBsBcff5ir46!G;Iul zOD|^B+5PZ>A^a`A&@&;9)RQ07Uxq@N?Q68Pm&T|(sUNZ;EQ;;4o`S>nV-nQKsRQ!| ziBdxHWkp@i1RL(gyHq0MP0!ScxyEN_>kfRNjqKofo`sr3<-q3GCwKMsx?|xQrA%vx-=CIUyHvnCuL? zzwo?ouVbvb@;A8TxnW zS?`?qe2Ut@#`tcR$X+t3k!i0^k=TW^n0*0W|4!qgC?B>SbF+hM`_)8Us=j!H<^vbZ zQ!~A+=l(B0wNyP8oU_FTyK?)=UK`N-m=z2>f9rN;@0NQkZg5Bj7wft>?<-;k16h+y z_r|Dqm0Vu>q47(@68Y~jTQmxns|KqJ_Rrqp6h)0z>$y*>g}uaiUzj|UTx?)0;%j~B zdMlp~J{zy&`Q@E9`z^zZnMWe1wYEf5+aY;jXYb1P)X_qr$8LmIq0cAnDUqIpPwwPz zb}qT_H^1@E?Q3d@kvYg6rv8XOgsCk~SQVxRg{ ztP@_NGKaS>01uvyN=aTt&OH4PWidQ~e#Lcv)W|b4>>l#4_1Sse{7F^ubH*v| zEU>W>0hIt8DI`dz3>HX)7l2m-0}0zIQsJbg$Sf$WN|EK}vNFdIfr5eyl6W1WT?2yR z-n|Jj^03giD#P$t*Kzx0@VvU!EeoSQcjt@=UO+Nyt}6M8Y>ju!wyij&fd1VZh{!t; zr>9DJdkLfwSAD? znqE-|;^()lDGwHzbA^n}-=$Eq6%fMwl&Y zL{AKL#-04slG7fRm^s@+hI#)kYV?#c91VLaJqs=N6f#sRGIPkt6J5SKGqBEkzltiF zlq~y;iII~Lm-0S1ua&UoRFm6iHL`(w@4}#+*LkzJ)~I}c2X1af(_<^HkVr=_mfVNP$@NHd z-`7<{`-gEeDpuw@doiBtx?E1Ar_vy5Z7QD*JD)bnSWFSH$-gm{!Yk(%g$n z?TnjnVijbejkHUdRfBcF)7oVhX8A{TR7$P@k)y50VCEYBsUXOkC3w$UzwSAI-+Xbl z=V~8Gh|M!;z*^&x1N6;N%=>h*76WRJKxFlen4+3n5n05u*EXhOCie*+@_c%qx_yXB zoOi94m$Ap|zl3h#;^aMyQeK_;3VOD^Iz^^qxRzD?_Wc)jn&JiOw>W1JF*=V$%Nr&&^+U$w#T_!T!jJr5z&rU6|BkrvRvx$W9s8&#c?ODokPa^WCnxFRqM2 zrSLUWZXQ^`AAhQo8>_i^0v5v+XHv&Y}#y zaE8#%Wi5Rm+(q@)MaAfKm(-Y%8Izo8+@h4T=%yP=SMEeK;fH~fM)(W@{gWIh=7sQT z!$wG@M9@SnYxNY7%`e3%DsCu^nbkZnyUaP_NPfE|74tHpzk8(iO-4tQk>@n4gzA~1 zt|GzFuKV)I5SAPrqQ=xn~%60GHh7v-rxNCyXSf04tXyIaRHKd*-RLxiM z((?$r%hal@XhdY>;t&IAMvfrMU};Nvy2=#;`8l+k9lgQigpSxFFM&qi8up$*HT}MKAR?_K($5KwPa=zRujqlws`*KKbiZU$Ys&n8A z$&Kf=)dESM=^6s`LKtOV_ij&WZcjQMLYHCXTYaEMPhD2)n3zwmHxiugxfffVJ{v$~ zG)Q22i}v;h@bDr1tLSNO<)a7)Q@__#m9*nnR`tq&$&NK`c1FkNz-D?CuHN?*0fem+H_WDdN1UqlLVx%Pb!wKHW4n zLb3KYo$d%OT^_knay3nWs^d|AYq&o&mGKG_$1t4!=@K}@{k#ov?_dtM#MHAa;O(+b z2!T(C>T|YhICP-Wx`zg?L6_D71|pOZV=QZSaRBD6kOUird`^` zAkv;Yv2~J2@1EQ?(3!7jjMBkdfLz~g8glJxey(Rw_Icf)uT7>?C~C`P{RFhR*2OuT zbRwFDc1(n&B2R7Hstu`(E^ltE0j&3_%(t^V$4<7-+oaPn8?=WOiU9ch( zKPIY$2?cGdVZ2+aZlHj-i6&j;+y|t2++#{LdxDc-J%~Ri8V84a2BXn+zRA#HGIsB| zSfaF8`qBB-fDnJ)(x$6@+lw47lpZq7Z_tuld(IIl^EH_vG$oRti4LbvZn8=^B8)18 z=grgl^)@v&ZOyq(LCGu}cTZOq#sr;;7MP!bmehmprIgeenUa?~Jjg4eddJ1D9U^WW zk;{za*p)b&3bsYPQgMee8n%rU*@sBKJS@}+x*4FBd}S1N$?B}^g#_w$k%VLdWvkfc z>DW8A4$Fm2f~&f(52#>kS(D)0I6+aW75;F=>egXo`gPfYIG2Y<%yJO7SX0ogZKtWA z2_ldb$${}b85RdH$hAXN0V8Ec>QiFQyPfhMqa zsM8EFRD4RLwa?)0OLJSwA1lP{ZOnBckqHxLq&V`12*mLzl-xa{GByc_$|XGpA6-Ph z#e2kkq_Q(@-fl&rNcc$QOp`8YYkKaffstH3Bl*RGeK1#s9XOa{dydr0p5ST`TTswO zUb*;luf=zvE+FOf+Vrxec|A(bgaBD2hdZwBs7!(t%z#j36907Wk~{r2AsKHUakk6@ z^vMy63$T|L;SVc@-d*^*=X*d5VtNCU) zIkgMTG?<8K>-$P8_L69Xu!WqIIQL06W9w-h|DKlOY(pVGDz1RID;}V+{tB;)@5uGS zSmTH|UP;E|jcThYmp4-uxbaP}v{B(lOD4fh4lsr{Vk<5EUiJNR}F|nt&p>&QeZUnMF+T-b>&vEGZrCtZEFe z%a19IaYqN&=o%D_zpd50X3lUC*Pr*kdwNr(<7L70&qSGW8CM@{P)2cprt=B$LZ{DS1cKk9wRzk1DOGIemp+?{OFp5ymN7<@4p%C(AAu7y`}vbf7#>pC zoY?&$Z(Nw&9=zDG2~SfL^}Lg2pMIjqANm3|;yi@`C44B|e8FYsZNoK8TKr%{9|GGV z+CZN??K2$|`HT^1<+^^O-l>}9RhUl%l9ok;b;K4W?cd~pYUq6z)4KH-=M~U!H}Dv?$WCA^>T)0McB~OYlIry66}*IV zs&k!m3ifLDO1&$(+4235VR;LLy*;?>ZMY}6FB0@3sOXZZ-bq7zaA?vQ_o;xVpLk(c z)|iQiF9G|V%ppWtunS)A8=KfwL@xJ}DcwqMGE|f1k2(iSt-LbQe-$+M(gV^mOp(3T z&3D4?n$tZfI#;g+|4kx+WA;Q-+6fPE;*%%CkzGW@hL7)hOA-b!mn&#;dh&OD$t+2e zrH2wvwJ(uA235GigA0cuCz`YaKgbI{`2^(}+cwpNzdp_60&I;_DPPo)g%ZAR+8mS1 zB~G!Y_hiiB44M4A*rBf-f00A`R)g4VyxV(iy~!O-`4J3WnX%b@AIKCeo+=x}&fyQf zOa$@BM3)IijrNa@ayX*%&huK3CkQRG)}n)mhXPA@F;ewYJuOyOpSZM7=9?lsN0am3 z_(VaW^0sv4$Oc7BpwwnOk1YM=m)r`n0nhQs#r249I1gw*d2U$O-FTa%^F`aW`$9RIteB#?%s1sEZj0jz z5F+v3k?j+kOPkkRo3QA8^cFQtwMYD1SF}KcHb7oIB2A0NUub2pI|;-s=Vo)})hSsh z=DwJ{MD%^15SlrWVBUZ<(+^@rimG`ia=e!4{ASCRG^yIF+gj%j=x{q^En;+^KS7!V zAuqEam@H4?v0oyWi$G5ZP)_FG@z4~tT zdWCNkb6Yf9p_2EsHCTSWtM!R|MruDw{iNI=(K-7XTi-z<@l)CQl1+h4qsugrCZcZ> zDj9d{r5f|P@A7OAA zt_2#QBlmyv!9HK{-6y4@%DM9|rjhDsX0ttnN>vK*?bj$$75e7_Y7d*)c)%)Tsthx> zGR(OWVwC2#;JT9fa7L$(dtn3<&or~lDSdF1n1ms@roSeCTeCQ3f-(z zDp$j2=4dJ{uHT?^Sq=Ts%7HfCn8jAfn`1{$>Pc zyjCUAm7*ZqA!>0Y7B;7>_h@8BlG3;H zu+@c>SsE8-PY}5FE%gh_*J%w!D>Cw;IfsAda{>*LcW{T{LIFQ5oHU+d-h%kP z&YM<(D^Lkym9jyLd1}|kH?vZJnAqK9-Q8^M(W~59rr|H7n2yfRxKKM&((PY3EKbp#b)neDOPjV-%fhGcyhrkV8T%@doyt%3Yy%ze zwP|H{oO`D7&^I1t@|JS=)KL`0D0$oVfQ#oH@l^e!?7~{=7Z=3Xd5|WYOi6Hw0AKe~aB|JHEfBf0gno}A$!1p5dW)YSp(Z!GX zRy+`cHMn{#@(kC{zkFs4JX?Qh3uX`y%q`1XP;JsrDkCUQbTx>Dg?JPSVus*C80m6a z()8=@h9rbUlBw7NQ5c#pd8rGoN@gYeH21ffqXGlz=$3ABiGviBSw54xBR?bqJ$YZHgJ~@j#q-RG1-PW)nr`j2*hMR18K#^KHS+gT<2^^ zZJybYVwm>}mZ7!TA(@xBJCU^)oq+pE313UG5ZFC_b=Yvmjo&sKQm1IM%f? zzvJR9A)S%cuVeEU1=ICYM~7CyVH5|#W9Icc=hmc?4+gA#xBjoa=wN$~IdoBEwBg!V zN9wy#ueEzyea`CsvQXLRx$BY+7o6^|43^P<;W|ec@dcuz&pw>yZFTJWvTJMaCoSJW;!b`)7$Lrsccu}Hy5Tg_Sjf|lrzTAye|XPn_coRk?)UPyqKyz3Ko-Ug;J*H_id;{ zWt+vh9wJ>vbeZw%mL=bP7CK11-D?}A>ByX}HNAWz&7*Jp;i6UHY<@;I^?s9e4j3jg zoC(Aj*7cXa0(_vRbtyWyf=3hc$PBWvugdV!-n@fz$*Il z3}YBo<51T8QPudpSKf*H11^3SR&_o|cYU#WCiCDk*Qft0Z#m_EtXj2ZHMe@S1|^dG zc$PsU&fT~C`dQJC|9%p{bL-B13Wn^+EY9&76arS@rz2^_^yKkD(-(L2yI^#{HttuD z9`!^Y_?!}TLig?i$-K;@xAh)BLb_dd(}jc!I>FD)nqGFGxH|&wgU(&h-6d>^;N~n@ zx_ZFUuka$83+!YSfAO8o(;e87q~#jiDTOytOV_=tlSPq>OtRAUDhnjwSusyGL-@&Q zD<@Z30WsToXya66xEqmV#O1EV3D%kGHwil^pdRUyY zN1JiY=%Lz2;xun7RW}^RChwqE{sflt3-^!8L$4Bjp^{wQuV;G2?`F8U$oYOFLWu0< zJVQC2N`q7hW0#t@BF{NnVS=uR2SKoiLUHt<2<1u~d|nP;KrqDcuDy`I)wy1X8Inkm zOu;)MdT|GjQ5Ah)n}c$TTMiBcYdlGau5Nay_tuN>0qw4D9}bQ6ECBG$QEiVZdUn1P&xD z=DP@a4!QMMF-2`^O%7CQA#0Kf#bLk8`$Ar>W2rdn!pTYX+UpcNGibKoO~2xLmu63hnF{px1$!^SU&Wym^e(x7SlB4DaXmo(a#Ng3J-fWn7UqS$iQd+P##uq!4lZ4 zekX=bQqD=Ilh__S2IbRhLxq2_?znJ97UgeRqeC|sL84MX&#{G8+2j`w%(uSCri8x) zmkh>{VhcC6lX}I(oj?1?B1fB0Rt-M!UXd4;qnQz=;9hd!W?aqH52M1@s)`L4()^r5 z)9&1YopRkr@w0uhlV#U1lDrNo`U)c0?=O2rDx*B^@id>`v`sIOu-ZX+vH-tngxRl! z1kq6_#~#Eee=5Cx^28f*gS?Pp1qyZJbI-2uSq!T_c_9v+N6b$fx#)A92O zaEO%P-XYfnYfLSR19O@UJ$ky9uLZS#}o%qUXcm4~jDT1<}}+(Sa3x&;i5G*1d3d|;m_hw?5y z52z{f5S<>hndN z=yPNooeu9#=U?yzaha zL)qnwsH;c#`Y^_jMw!5*7)?TG-NFt!-SRJL)A_yo62EJcyJa)Vlz4XOuyIBpsC0ofcgv~dXm?L^m=C<^DT)7hGzNTan_<)Y3Rsu zKmd}Jr71elPTHE6@QG@vUEB6!ONVTlJo{0tsZ5sFF%*D}O7TwNz|)1Qbp;4zwz1*6KQp8s4~KT|(9-R{I)RdFHWkj61Srx!^zon(BjbS*>~^U+1*<;)VeiF#gw2FVWvi`*9@?C9}Umi#YTFboCC zF5fTBR-yr3PV^qceBX4c_Hg$eNG)4FVvS5`0q^Z+)$Cv%2dg4aZK@iCJjAto*;s2z zewe&1gLjmj#IZ-pw#0tqiDnT!uRWXt9bsPp5g2h^lPlzQPSAPMHa+?o&QT+j$U1$g z^ChGN;>)b$u+zjVHJ#2zZak=*=PJKFq?07pQLt%Ku;o*;RBdZjkW6Uz~ z#8%fK{}5Dk(OL54tXat>L+j@qQk5TH>Wgl>c9SXjbsN5PP3h~|qzg>0itn~bNahWYV2Z-I z?9g>a2(0Deb8jK@g&ANKFZA%0#1TbtyrC=w*~uaY>goU>sJ1lK=={&S^Jy%2ruvr6ybR z<;%47_fOhZ+-J2nxOJj;GKeT!TiWh57r8_uG#O)XC7ir zG@TBu_|ic5f%I-fkw4<%T8QUY(4O~7EJm6-{cO`=>SrC`d+S7##z_L{5p-*J(h87d zS}ci}lQAr=as&2*RCl~2@5Ap1&^mUz--PL(z#A^Ry!V!hL)P%b6I>#R^L!<#_B(PM zmi0G5tbC*z*Yeyd`Ugd?6U{D{Uq&xAQ$kJ(Thmmu(KbD%)RRXBb>xpmgq4s7=IeJG z&y`D(ZwoQkMpfLgp)vJWkkg^R%er!jn&vC0PeM+?-{5m|)oBbeht!!q^oZ%{K%?-l zAP=-37UOqBSiKREoEdDSaWxHGE+UOcAh5x2ZzD5X2Sun|e<5^P?nR&lEL?(bCg5Y3 zvnpDyMO*2Gl*3GY5Iz-W?;I1SS$lS;hkr;qp2#Xv&8=1TrYC}`zU?I?}b`feZ(EXNL$gSrYvDg-28K;mi}I4 zs<(BC1q>)pUony8fDp&2>T5a^y>hl!tA&#q>{luhlgS40=0rw;+{&6l8$`b(vQs)h z0?9`@CR`10Z_z_Ig&#LDpOQOJY*H@M42zpr2*-0)tJ3EA_|STd%3 zE)PDQm9vt*2R^4zT8{2@?>KeW$u`MaJIpzHe^qu$Mi_>QdOXaZp6Cf;UK7Q9k$wWV z0Hb-D{z|K=GR~WPe!iCOgHg0C*J~d9QJSFaPF^FJLZF?iYf1paq!aplc*k{drY7+Y zB69)m>?f?!Y>#Jl11Zl68^6$akm+4Xr~c;ZE3xRg!LE4_1nGFA-wke^tPWB&qQ~J6 z_$Xi@&U#KWq}*mPgOs9pXf0Y;RyM~2M9FxV(;$7kPrA@?{v7DlnvarkgWm?;UIg*R z$2MtS!yG}fYdLtI2UPu1mq@mxS&nS-WqRp@zT_=MQw==7jj*G5C>44~oQiG0YAHN6 z`$%{hGbB97^xg~ia6r~CbrEO-KC;P{>D?yS%kl?SvH#&c#E=dy*Vb86f%~vb2a(Vo zmd~U?6anaa+0>V=>nuR0q48~WH$#bN**MyYO20IhS(QG8PK`Smd6+_Uui&UI#xbCu zMV(^y+*bbsLFD@IN}Z;G14+WM0##mH6&P` zW>D@#9gMZ=@tInAMY(uJ!DM%!GP{V;)X(rDiEt34Ix{cxRl&FF>b+*~nQTC5-jx_6B z!Q7-xcQ2agy)K=(L@eq$$Q^hQBsq7|gE3EA-2k5yJu1@vq8j?TI9TwM{Uc z_de(S@}7I@yMJy~S3RrhdH(fUYmPbQm}BA`pi)Lj=67Aa-U$8}sKbt3FD8rj;eHSV zJuf|o%`6*g?L7DU;CJ`$bkb`;QLIwv>?xHjV}n#y7=Q|hjgi2>|8XC$MGzO*bQVt~ zCWEhB@YaE?qv1CI@-qH!npKOXj0ppd#h$t@^)v6y{(6h_FVuYvtNN9npQO(wKQ)}4 z{NENW4Ki!UtK#d}&nxNs|6YvwzvPweO!FJh^LxH-cr)+!?Vp7zeYZB;5mUi4i$Pqg z;POZ|$RS#E0H$5P`eQjzTs+A;?IH17?{u5Q5BlE#6G(A=QH)9^_=rq!^Bd37f$7!7 zZvcy5^6h!bkIvi7jhjL`7xq$Dw!_iHwI32OlZEC?Y#SaZvgd-e1yCJKuz~ntB)QZO z=otcZzL`g?qeA+#*PeIpERfoyZAoNlQ+D6=xS{eJ@E^Evt3*Z(VmQ;hL#lLA z9LJ;YtFi5d&rlD;KdSbv<+j4=t&qg_R87U7y1fCQQc~hMTJSC!wce5I!ey9ML zC8iOgKt(Wfx~V@q3Q+_Rtt}hBTPiw?&#V7eM6d7PHLr2b<<+S^wI-w$iTfRd z-%mL0zN5v#u=SfIwIxf}kZAu6Q6bEZ;87nPM=fqM^G;DfRq>%|4dgahFppO2FLA95 z?d9_&nV2oh_hU#9@n+~=aN4`!1aCOc?ZNDazhmB;=l>SO{v-&y{afdg271T;UfI+C zPOg(MOsctT?+hhF)~5ir|kJ(ysRX}9Mqr`l_aLVX|9 z!=Sa0KrYR=dqh`1zL)mftr*mxmHE6=s^fU#yK+>s^cO$0Yo0i|j)$3T(@rkq;`{Xm z=LMLed48|qO4#|RO+C_Q!i}L379a6nH-OG#=vrL(eK%epI2Izv@Lx{-EZ}c6-+)xs`jZ?*C*JlHb)3??!uN_k(<-WU0O$!zF7AL9!II?pg$lm`-IsE=zNDVNtO_v%X$9{jn&*7Mj0L3!xLOmif2}&7?^^JiWsx`#G z%1LeylpWEA|BfURPeN$9eZPgeJ(J!iVT{MS_sK0;{te*$k6=H7GV`49fgZ=aCQ+h)8O#F zgcuH_WDF|+)NtTqyz1EaN5Gkxj3|aupwCB<1RhK}MR`IBM-j3HUBU8&ihhNqku0IY z>GJsQ4nyZ=Hs~rgPDf&`OUBQ=v@>pvwh%Dv1^ecZ5FQJq7Rdr0z{#U-#6ZW9cfNT` z&wjV)NUXJ>p`+~cDtgtsuWd=KFVnN5gu~v;o)BUea}u&xRKeJlS#fJhy5nGCpk@TF zh!f;h#EOV}nVxmA!=1dNCIjw{u=IzxDVRbBndR^sfSVi~(|{lIj95UevGR~lWBO5v zGSh4a+AQ_NcXt%mP4Atqt+D5uQ|ojjB|%AG zG-z2NTP{?OE?W#i>-tH*P7H0XS~8K`b1IE6XiKU$ z!j48dbeXpBI4}NA@Pu|innEUPpN)7rs#%5?mn=fl`NV$~&RU!TY;uW@IZ%esY2^C) zDrp6ZYtp@NWnh*>ofJ+)PGIpPF|5Xq>(h28b>I>k*N9ty#iJ|=wPhnGb8g=yFlOXur34EBUfwI?4F_?VbY>E(L3=qw9t zSR05EPWT#ecP{ZBpWO+*`6CW>s9Wcmbi01pD70r_4~XhB|MsB8W%?C@kB%0&1MpSV zF|$@chn;+Ue@~;A=hHh95rk!WPuC#pTsu%z-+*)(>-#Sj!;-$KJ0jH{+3(h3b`d_h z6ToMIGODUH=|B$$VVaFvN*3u4e*W~lz_b555~IDiULTA}ZjrVXs2$Lbo1~WScATw& zq-;OfM@>+-6*J5#eyKJIboi3Nm12q?tT@kpqF}6CkJjr_?@;oc2{X~W{%FiSC1UD3 zXL&&RGBj#iKka*aA@Z_?-`yg;#j|28{l}&vn3Y`yx@IaRQ*u zq;xVN30(#rdxbS=di$=0avu&N0P|?LB<)a$wP%?jTPim5h>;I_S#t4vSY8Pv`aU_} z)9tW%hzn2Tw-8n}Aq%{Peu*G>65DO&3*4FnK&nXuEeUJV*ID&I5mwf8(z3|2`6R}_>h{m!&R>6|77L(|OT<#A;`<{(>3pKqa zIRau_eaI73YZp`(IJ!RR$*)QZxp-k-;|5lmsI{<~1jdMQ1vQXb(piZKxMYNzv`Ja{ zm6Bsgfh{Ic34yZTXdh6-Jt&n@GCFSVm|dmUP~K6eS&!mUyB3mZ4$22l?DmWRxrYiD zQ@ahmVsSTL2Y_S5#zI3zaW+4sk09kC~6BDkjV|(I!)GaZr1n(o{O_3vr*?|GxC7b zJ;{3gomnm?#uOdbYprBi3ajG{-p=@&tlNuA{)soH3+Dpe+Xd&&k}r@UV2;E|**1?Z z{Y?EcVLNb9c9HIQ+vZ5El77@DyXiROA@$W|bwov9V<8b4IbC!lrtG^N|3}DH>%m*I9LU-m1(a zDJbJ+09mCez-bdJ`nx!5A=%kQjZgZ^`_3iWFdtqKzrZ`ev%Jw=0FQmSdZR}aFFB<$ zPWBf9VC`;di9E)N#;-9^*A+LKIh;C2c86o?Z+Kvsj=7Ge15Jz1XB%+`R4_M{fh4Rr zMB^izI@$tTb+f**z30}Y)7>k$qTZCQpetQsWSSk08;|1oNu96&UEq4+J-P%aqnS$| z(nKL}M95W8%G(4|^DRJeB97chW4|0F1*xlncO1@xFitL5*L*UqaY?xRk5jYA;=A`KB(rL9Qga>WgC0AvEfY=<_Ney`$ zAIP=CZvc>|!QG67t^$Ms;TkvKqkI0EsV1T2*{9$>iY2Mhz<~j325T^NjOCyRo1MuN zM{J&(GsZJrF|=gdOZ!R{yFx621=5lm-?yEACKFJ{lsdmbL63ssibrpI#GzS2FJxC$ z;9TNQvFcGlHn<Vb=@LHG?6+~+r|jDo$6uNxi5SNWp%UN zSUXk<cV&1Dlyun8e_vhfxA>0Jq84 zP+&RzO9mnZvXprP{MZ6jV`3Kv`~?OAy?~r|CQMb;zfgzv_}M}$y|sUgDeiB?Dnvk> z5T6TkG+%l2%rr8P2yhmbOqdcyf$(L0)KwlCcWnt05+?wfDf&gIa+0lrVeCBiU(3@H zHV7A~GUN?64#U;x zw?!V#^3L9!cgNG0=~=t~pBBMa#_2pJEv~zx$VXM9K-o9ZJ0c#MeEf1(g6DL{>D`rD zPh54MRvmfyq&r;m+}+9~OZVTKPg0(hb|8*kHhhpueM5sr<+WJjd*J?3@5;D3un7nnwlY7o@!$E?>-?Pp;#y znyiNS_U=Io7qCw-sqyg{A`bhx>c9lVyPEvsDj1`$x^xZDfw}3cmo~rt;BkBr;Swxu zGVUNzFyWO8FW#F5pF!QW@m~R!%a%$p%#GFaMyRX;^_jQ_OWBn4)HpU*EaE2ttN68< zBQshlRq}i%?cy67iO9$it|MtT+IK8dEbkmvRhjC1=BTvq&n61 zy5^5^lsF$CjO6W&wrz`N2nh`GKSZEwl6$~c{E0-M>Gk>3nJzqg#YE^fsgUWdJ`##o z*RAAlpO*qr=~qVsFd%;zz zrPI90VEA=#f<%;i|CUEcon8gm;al$V_=xHjxu|~c?j}?-n+5v8=OWVz5b9s=J##MN zidVSJ>FZ}b*3uslx%z(t@>%aui50R#_kitzvnUf-&glh{d(s=f1f`4?r7hcBJ3}gvx4yAYgR7ff{^bcb%*KZR30ErwKPjPBbzbe-B*v$MZ%1 zJ&;6Z%s$2QKWWuF$-vDW+>%68C^C|x{!cU2f7^L%biG4mF=7d@KjdeNtf5W(6N>x> z)RFx|aa>r@1aJGniKna)1n6`v#JkV)M(oRTU?H{4LYM4q9PT`_OBA!EKgoR&>HMrs zcU=~B9qTL)s3dj;`B(4!@oftxHTe4oza0K(?oBDlNAbU<2Ax)RTYO$=nk_KPnHZTGJ&?XH5}hUI^l zlYi|YX#)PKg@p|i9e!$i&;Cxf{mnv2F(7Yc>2o;g;!AZ_F>qgbkR-1{(r$hyg%9C( z9_F~~qQ*@PH_BP2c=f^SclpMMSyKIUV2+Z8+fhraM2#gNr2ycjM(Cfzrv!xa!rjNK z{iQ1pitJ}*YEhIE*YkFxBH80HOSsMD{rqov>t(R7pE~?RxqQAVlRl)285sW2Zil{G z_UU>cj#9f#8l^e>-&J2C&BS%X>GBxY_V-sma=@xuNb7Vp5b4_7Gq3c=zIywKpp3zcM6G* zaGnLi;49F*NS7k2Y{Pvr%~h+y3yrQIDzf`xIfu2lFmJ{*a#Kus^5*FRWD^@aOalFV z;h|(>M>Bj~vKtfQ^D#m$zBBz9GhhKmPX#?TMZpOB2#cs9o6zt)-TGuh&R2N8(yMn(+QuNE`i=FXd z>Jt_n@xI%_loHni9Gw>^5rm9&g6@ymKAryf*{u{{0Y0~-IIvV07or_e79q~ha=g;_ ze9>n2j>`jo_skN4a>Kq(%%hoAoeNVgX;CdM3asF5S|T<#6Ky1ujm(B(yIBE@k1 z%noHh;drv$ZVE_lsM32s_dk9f>E@aT)Oc_I*Hhc|Y|2Xf0grOsdNIpzLtZZbO9;RE zhtyE>&E-~3*av$WBc?z+qR*qUJ45U^aIFz5oMWVB=2Yk$(QG%g&ju8)Tx>tLHP_ea z`;gRY&lmS*Sq^#x2H}9GN!H<1S~ZKyd_P5>c10;20(L1jo#YR5qg2W>pX!qt$oA&0 zELRMy5^d<2T`vF0HpM`DY{s0~B}m;G&xX>yei3k3xr zU*S|)cR9DlE>1VA_LP?5%o1h+PEyo%nHT*PrC8*&z0GZ<#wV@4h=I@q0x&ncwfj}n z4>UVo1&O-D_;l&)_lSb^k{;%>?d3fX~OmcBtdZxMTXbAY8 zd^9guSEAQ9#L}e)b1h-2Y(>5jO`T6{Nw&+gK~>JEs|*N>70r1MNeSn|A@W3*yTJsB zXF350u$S~A!J+V(+wgmTyK`(+3Bad-yJ8xPbb7iMFT7D`pZ>+<{&OG24yjXdP)l13 zZ@HY8U<_x^XVu_j8$^7H5{1M>O?g^}M>Gld#}489r*fLz(@Oi^UF}ci{l7gDzs>;! zsDDIYq>Bj=otUoPBR{&d;~#i0qwr$)-1}8T!J^(RRQBqh;@a9)vpHJ`3}}q6-6$eI{OOIy#}FQ z#E@hzFY#-%rDxMp;=bvc-ZoWYyV0o%Ln$khG$2hU z`x_7HKS?hxr1M0BcQ7_yKR0QgrL96mj))05;o4v^2~bPP*<^ zInM{Hl)fT++A73(Z!t{3fga+r{Su_bI2P`ntb;GlN0J-5hC9%OQ=LI^NCiSr9EQ{c z;okPH@((_fT{fP05nVA)qMRT)9L)op^&%GJ7~!Wb=p3=E8osx@?w2|>>I|qtA#7r7 z>jF+Ol+aAn-ngUq5XPE|I8S@K1O3k3+{WAu} z*zZ3fOOn^2Ald2S?KY|eNPvp8QK0qga00d)zMsZ~YU@< zwA1Euq#B+gaGszW-dmt(Q@!IzXg9U!envMS%viTM)OS##ra3q&QOZ~*Xs)xdoNEOc zwo)D*B>(`|cJCfXtvR`havZdx3v5MQM!`JjfL_>f9k*4HOPJy|o0rv#2zz!^e3TahdIj&wnGEZ7L?e`v>04o-SgLTb5E}=o9F{b4&!qC^+p5Pn6cQ|# z?SvB37ClzLEkX{lLVJ@@rm7=rEXkVJJM3$A>HNm@?NaK*Jd2Y~s(=zI z$B{XA3jFR9G|oj7tMs50E@!6qMHtamN-hVfGWsOty` zK-Lkfi&7nokWuJ&(` z&Jodl<-TwEP%jX&TmaQ2&M{$yM{*waA4o)b#{l0acod8uml4J zZN9n;`jba6WVsN$Ox#FQR8VFZdEg%cNtq@-^|5G2>c<%#>GW4rHxBsm^}Pjm`>0OA zUb4u&4)pBT7IT?exB*aH=1KhL1!8k-VTWpK{vAL>)Zu4S3yONru>c*;c;uxf67n6F zSYYBb$a01gaKfqmXiz&arvmeAZ&PV+q57|2#1x887bCn5Db5$i)cQv#NS5-ZKY zu=AHg*Ohmv3|>7#FYWz2H8rz6Uy=k%#)r|rOLl-39-5bkwy9l`4goNXC`rgECf>BL znt)xHyXBY3jl+UMs24)FV*yk60GwG_mm&9+yl#qQGG77 zyk`yD1({jkhB~tA~{)Zn`Sjna}8yzvO+&q?9NntIdYm>fq zl;m=oQZckZHj|;_hx&{0)Dg7qP5LY*+qyp=tXYYd zU9f8)>NDI_<07gmf0d$Yt)rO(l8_gb`Pat;*l1QdD^i#QyFV4&HWc+l- z6;j)l%!JRCbtT(r?r#DwD;s)W)wk^`HkCLQy$6-340c!Ws>7Om0K1di1|iBYF!-`( zD}~fu+Sf;si3HSwG&@Tedt_nCD8(aEOo^33~BU>$gd2zD5!{;u2e2-C=+bSb!CsZp$!gHPp2?tEEF(hR@ zf_5?72vh|VD^StXZ};sdgud$H#s1=4I@T-4t*q%0_2^jDHnnI}d^&c}4;3_qV%$U7 zT{)&D4-jd7pz21Wqd*R>5Wpm4-()O{eF4pV^9KiUU|N7uto3bHcqJL$G$Lre5xJuyWvf95iG zNwQ^hH?$U4q@`hDbWkvQ~<7^;{h+mqsx}MQd!sos$^sL-$gmA_H^I3Rsd?%e4 z-f2O!vl}Avy33iG?OG)gAHQ(y*sK?CVh9oq5%~eer%|hQY*ko%#tgv(;q!yG2ymOH zPLrLAqp_?>P+>cvE*e^vjbyW9O_=v5W`EM<*cOW6N00-n{eX9g`E%|TCa?qoMh0vo ztdkg>N7-s-lf{vpzX>GJ%zvZ`%!Im7XN+JzQJ=9>g9!d|FZuQIxK>{{*j9hKbK`oR zyc+#!*MT~^lIoiQotx3o$``b31abg5u)Zb>O;0|6(d`x5j|~buj0I6EG;GuW>{jk% z%OK33!tY+IzZt0_HVvp$+i=sM3ERieof5;Ae-(5jU}aIqhw)T}(QOI>XQs@32HSqP z6ZBbmHD9y<0Z)1H+Y9TXR#mUAbB~nZ6?rO&px%wWNz;asSzt%OjFUX}+innPj4yqe z)C3Vkm&XOC;p&n{08iFzB>&8TTes)PQ420f_)-z}n0EJJgCZ>{S* z$Sq^aFO3YG)rA=sw@BG`(XN?FtJ|wap#Y)E=e1bIO;bqcEbsTm_QJ>4k(BM9k+4qa zCI<02U?jl?Yz{mS2Ov=&)H@WN0iib({(pHNh21RMRgk9IJHW#)!rhkU(H7kR%^QUL zMjT+s)qIYCNKNm$zjywzG^t)Nh?&DFp^_&0$3*6c54-PhNA=Dy>7!1m_bMpJqNl=*Tj-2k-jt(29ewMV~f z6l>c2$GCm6#*(eE^wS^g-V@-h8ezkxi@wZ$TZ}fTN*d6x32ItRpgBb9#mxT5;x@OC zHs4$E2B%bvr7lkK(jSMURC4elLzK=1g)9s-DA+1L415Y%E&OoWBUdQo8-g`C3*yEW zx^}J~+1xc2p9UJ?I0&(cvOX?NH79CB%$84#G=sm1u1yc)&K+o0Re0*=XQkGoM1H5YKL z1BeC>@42@yipjqf$x{_W>bJBDjVwAoJpyATV3puka@z4a1N&0ZOO0aIB?F;!p@cD17~S`rCTA73`~{aCf7bF%ECt=3u(1@c*tBsg zlc#fs@M{cs<%}xmmu@!ZP^quVGenZ2Vb;xeXlrWMqUxTI$+C@)VAL&Uo^I&Q8dLMK zb$NOPGohNK&`I~D&ePxf5=(8&fpHO{%YZn1(taVDLzxYVkhij))9eUd;u=nY^{05X10olX| z*n`G^=5j#Kdsmbea&`avJAD5kzVQQ;ox5Olsi`Q|rd-lFWUf z;q6&jHEPV@394(WIOgr01JwQr+pJ=d zO}DGZFsGDWK0o?zqlR0w4@PEI1JWV3A@o+oHQ6|yd z$!8|>Q~Hmh0G@Ip&SLZw1LBdhuaaQJikNsYZY8zL`%^rY)w!uT3&m?>F@py4g&vby zh)?JCn03WQ-o0AIK<(7??SRg@EOU-3<|NbmOr5i~nog)JpeE2FQ|{VRv;_rj z#~z)AmO&Ca@ARensf|fR>^zr3Z7D1`4%RD9gdn!k7HtIp3bH}vUYIgbS5_#nExtoszDUv1SY*yNxdaaGoa#&6bo^5^?;8X&Dq+M{HclxRv4X zKybB;wrZ}!llDU_^SVZyTaKS^Sea~8dBF$(@eiz6$&|=xi#n@W(s3!kNI&pd+;B{! zhyX5pL6Kl4aqCfE>Z{mAzQZ;HMv(}#BLr*die-He6@WfJBNUB6n{MicdkluEC3ZP5 zgfQT9qD*kV5^vQW?M?lc>THd~W*w{7U>K{HCCTkD(~z!9Ixv$@LrvA82=1wQIg>jE z=9=Pb9){p_MKxXq9LLg;aFO>Dh$`xtGw|rj;%`xc90br6_Nu!RQ8GFN;f_gn;1Vxmj*gZH4gJ=Tmqhg7kA%O-l7&ASrS7lw zpOZa*pN7xtBU8TUG7Oxbc2y4@R4su0t7MuiVa-OdWhhx3F7$;U1X=L;x(BS>u0Sui z{+^N#5gC<~l=EDMjVIGi$Nwh?#N|-*^I}&rw#zn*>^yFpZM1m=YZ3`f3LUL4%fami0ILJDyJtao-s zrY5V>-u0Gi+uaMaF>vQy4z0|=UB?Mj$1aj#Y&9{(cp+wQfC(im9(X9_U?Q7p>fY4|WhPNC-yGU}e z(RDgB9lshDLT@({Ic2~d6mukjHr7(lEstK;Ze{kBzLh%y_25|r*Q67>9o&^J*=rAO z0)a-^sULZFe;UAtaVTk`#bhw`i z$uu}5PV^>!!vD+7ABAdaM_9OY6FirLPej-t!0U5sNsK1joG|+O`tCKoGRShkrUxdg z#^VMIE4;Og>V|8iT=na}9GQ4e*``FGhKV`!$m5H;aU)==SS^j^Ljc3k%vsIIa;Y!M z&v(>S-N^yu^$e7XM1}NOA1(SbBIVP1?@@01IlMnJAEC^JQoVayyoV7dA)AlB8yL2* zOwsVMclR$D`&SW}YZ)9B4W+4k-4YFJlLUjNsrtT9&Tj|l-sy3~uM71xJo@8MDLgxj`q#Q7gKW&T1@H<;0Z-hWkZ=Y z-Lmeu*2iYkq7F>7cl{C*;z-|4(Z|ZMefrpF)0uA6l=7VzL@kUu``JlxZCG*&(xApJ z;*j2;I)QySlwsFrV=;I9Qqb9T7{BW_9}Q*L0x3(4|3mV6A;0$&*z`1v2udGrDLGn? zTch8HvFz0G;9xYQw%z1RV;HU`i6lk4mnZ$FEUEf}gJIfVlpbjS!2vx79dIKuK#X!| z`)uvr`606P@RN%!N1u{>P~^+R&X^1`$!DzEO`$>>)?e{;V?xRTzW6lzh1^duOcjij z_Ild31Pl;|-N(pes7?7%GeD-Y_?u73!kkOM@MmjIIl^+dcAdPIELP$c_S}O-`Ijssr}rWXYGagrt7M7 z=L5X}eK0*KquAafk((ZLP37Wd00VLMmQbzAt?bPFgwIx%^4>`MBcHezb~CTKkrvX$ zZktgTvKT0U00(HZ4!}$U+@;GnH=cXP!vZ|wS!>o0#1Vaq&YFAsTo}93O1^H#7nxoz zfy_hM+c!B8WakOpIGP!^R3Kh93;%e`-sv+B7nYJ_sFf3AsvP7b+X^ekZ+9ebfRGGh zitf0UUHx@Jpq%0p=xO-u7}QD} zk(8ijKaX4zqZ{Ko5MIw+3?;!{%En_dk_G?!MXWmqHDW9P*j=R1{?tH!@#3M51+N9_`h3U!}Vw%2>!|$tr zkx#4_KQB%Fta6&B!qqJyw4JEdpBo8d?Be+>vy>XMpQ~ROBK=l2mm`_Oxe}eAr{1mh zieCPV9$s0*Xs(u30FksjqjygbSx+mb9=WOkfuiJQn4a9{{ZZ10W=_xlq_Ze8!I?eO z9>REE^h9iXt5ofsGjuKXOfr&T`*bn2P2AKFs9Yi)c70kM*Thn61w2u_{uqHBByBRaR^iHOZq5 zmCtB=;z@LACqo3i?6w1}a_AauFxcTU_QvA8D4u8406YQ(8YFJ}FxpllmscBO=wDKlU61tDeuVI7m zjBWH(xQ#2_k*sm2>(@HoCdWD`rfatdU^LOJysUD^eILxE%k?M`!_Fx4w&(keM}M`) zZbqAg`$5~=gSK_N%~Hhy2+RIJoohvmRvM+qpB=caK=J@QGvO9-$ zOmS5g(A^xfihzZH^K}_sUJUPc3YFu&xI)|jy7~cpV39x#3wpF%TeN0k3c=txd0@)! zaG$GExR&}XNC7_p!GTehIB+_HHdi2)k6YhoK-|RW*6cOF{^&xQ1~mc;%4n-0XWy5r zob)l7BR?X7A+ZhF<2s<&meZ3F5ZQVmusPnMg_`JQCuX%@yp~|tb0IWOEEG~{h!?sucCxv^9=4gT>_qlZs1^#d>R#SFh-QHPXK`x>ie)O0M> zI{gM5{f`;Z|B^M)kw`p&*Rl;!TxL|<%mQpBo8zZ2<3dmhZ z`t3jOGofyHthv)M!3`U_mBhFyI2vm?UNkevo}o?o6T{9h=@a}7uxqS~1xU}QSn0NJ=R$XG+Nb7i0Kp88wHET0u%wTf!+-P=P)gMPM37z(&D=6)0$-cD>u!X{w zp95}n;D#Q({JM5qcxw`{7$+@^Pn}&`S?t|x=ff1Vr5LwAl%KD{$S|)$5l3XrS6i2p zovoJWx)B}PipuUf{n=$67f%k)Eu zUI)3M4zdKgD$U7gcEoqOexNAMf&zl-u|0LfHzePS@qs?TW%ro7l_>GbDZzDHJ5E2q z0brFVb2?cX4pal1@5vp8@m(Ir)85D)#tzzE?Zy};1!a{Sz9Co_7$AmH7DN>|sgtsR z{w?s!g^~7wn53;xEiTz3Fas#TItVM!z$jl6Tj2DIc)!-vuogeh*ki~BD%o+Zo$O`qPd~CGt=h&So zPqr~^v(q7cU|SNh0VAUT6BJwgwHQPU4?_>W)=4ZUI=f$>+tfiF9(gCuY^^oe7c#8! zR7+xN56Ro31IA|xFZeR@Z1G{@s@ak3O-?zP4+0mGXexYg;if}U<#dxmSq<4`A6-9U z(hvd;LdgqLRLrfU*9BJfrihzSVEbLc3Qa_D*<`L_Vc3sH)-o}-jCDZ8;s8Ac+-uya z*$OxJE`qM-_N9-o67y=F`9D%`r7%ugp;0)#+lh*2qJEC^IR;WcX6B4)HU1T?i ztS-={_hQqFSGgGM4Sb$a!h!g!!|@^g59YG_n0#fsL|^MN91o9v6bz)OEk`sMH}stmexq((&2bVx-y}_{G&Fj1s!d1Haiu& ziirEm#YI>72j?9xNA`YOzIEXI9Xitqccr9*87)C?S84+@5B9d2UeLMG)X?nBD~|7M{;by<;kEU?Tj zyF#JR&C?6KBvI7@U2a(wqoke~9fnz6x=m1EtM6~X-?WNpDEhtjNlP8iRlC_45~ZG( z%>u-bMx$bPQ014m-Y=s1u88NnC|%spGd-$HW-4?{(OAGX!Iu7QF&b|*{j_(*nLizf zb_o#bAcq~rKg9pBY0(Kk0Ur>DGgSrsfrW%B@nCo7gxk}Wh0@Qtws`x%*8QPCWTzsu z-gLzJelyqZ74Pm*hw`67mws}pz~j!u>D)#g2D-D#$Nb}iYGSsC$4E)0)(bAYslZ0r zMGF-Ah{X4q#Cx5WT%1K{v1_-L)vr70m)|SyAUT~o+OVsA6!Rifn0|`*HXyD%gSJ42 zheUQy2BNqjQNma%Ni9i?&+t&l%rlXTUam~a`WA$qcoD~O=o6Lb;p)r%>9)H=E2LA-Z#;=JxI@n5c;bT!Is|u&>O1_ z-}RI_kZb<8u}5KD*XO!TO}>hr3>=>%1HlCN3bNx1uA!14Xt@BlB1Sc27s}KhM3hzi zvve<~m=(KqgI+E+Fv`b4^Tc%aq95LZ!ZHN$gAc$M)_9DEI7ZA<>C{WwKeNlbx<=0- z{_5Hj-!i9m(qyCw%+4E%Qrm7F@lR3QA66`d_t1f<0vrJ~#zyrzY+5{_B3&jqt4&k8 zyO>JYCA_GNo;bky=V_?DkO-GTz=q@Jku}(oWEoF;&of~LiIn)amTtiUNXl9BS7M2v0*@pQgyuNXxIwJ}3%JqH^hT@8iO6e+h) zf5vEj;5tv5K$eS$n=Prc6}5#=DRVJnt!-SNX2!%fbLTG+BspaLd}8UA_R7Ffd7-Zd zx$fZwNE}d9$0_-WX@UNtK+_I5gNTF`< z(v&baOvht{kY}5|y?)nGtJ6eVeIX^j`qp!)4dm0Onu~v+SPNr}6GjHvU0O*^&21Y3#oN9t0LL=iI8)@XnRo%p|H>?-*l5YIetg z^2B@?Su>dbr5U7bs(%`3;Z}8(VM)+#mXr)$`v~+Qu&Z^MBiF2O}Auqtg6HPC4GOL3EZwr3evN(-$$Z+?x1??ra&8BDoVvC#myS@@yYCvI0-D zyrWOfgEE|VUCkGREm0knIPu6x(m3FF7XdTMk9P&usWx=X&74|oMs<4;{H-R;scp4h5ZbmSwN(?Wnr7m{^l!jEKj7?#g^{6Hk8yh5No$*)R9u z`wifH$c2oakZ3u)^(;!6pzGp}yH;?f>I*2OvNpn2VUhmiRJR&1Kc{vqmP$~=Gjld$ zp!6Fs{XGwVu)2KnFR141Emb`{bRMK}uJ6k8`3l^)@-bB#J!V$x0dV~Uj(aCKPd3)^ ztoaWi#32Zz13l_#N7dg5Tf?j(-6?)}Wv2vJIYVt3I@`{eVbq2cqmbMpA}#zYRq5!9 zylNC+7H#r%utucsZS3D9S*D-hrG=|3W0(&^@!{GV&E4^wm>SxV9#d$t-vlW$vDg1Y z*?ULBxwmb@V;IJ0ql_|o@6kJn(R=T`g&}$;#ONi67DS2OyXZtli6GHS^yn>w2uXz4 z-|YL|_xC+}-_Q51_4?;p>$Llp~iO`04xiNcXcV`-ECY3 zbH0YBzAqY+$WU{y9q&z-`(>e6+lNXUx(!4fRehlYhVauTM`XoKR2@}RnP`}#_jNRp ztwyd;w$?Yu7#Z)_M~oidg?X_hQ86 z64WG)1veQ!c^@xF@Jhjqf=JpkrYWl05Wx|Bj6J@p-!&T3RA$0BP8W>>pMDOmODnRa z56)N(l|qHY;bEvx7<;5h--#H!3?>)meAr8tX0;Q>n*8Hccst|LOLN~Q@wkD72jV1- zmN=vUQttU+t;5t>Zrn(F;w3dNm!v^RUu!=J9n98$y0mV*cQ6%Y02Vj5tzPxnt!4KW zn8uQJh*m3Hzzs;#oJ3;uf1}$aVPE?pFbv4`{ar;;3BNmiXCHNV*MJ>{fmeGVILYrv zYSWIEO$Eh{V|AT1@{q7A4(oD!&k2z>RN5j~`CMT0S@^Oq{>3|)k)e@ImbMzPddCar z^7<9~R~>q$I!KkhGOe+Lw-MglVdp?C-4jY(?w6ss5nwD#H9q<`0&M@ZAM!>&e)Y|> zpE_^0crW{zU>uGOSoFq^qPOY{@tHL6=XX87WuAokE9yispZi^2x#saeTrWV1ocW3r zG`?UiIdtdlmlqZ+>z;E}G5Vycs4$UAb~J*rK`c1lXo5!#*h~+6P3Y6Zw=j)lKpyEf z)5lih`QVl}Bz|RraABcx>Ou_PlPjYV2)K1Z-u(RZZ;?3Lgk*E)C~fx@kcyfZAGiyua-v%PSR|7YEkO9(giP{IS?rdenEx@;zl23imaGF zZ56yIsI)O*k!YJv6Xo&CduBT<23fV>sjYizqh?Jbwri8i63Rj5_)21X*_#u^ zE+m$f^Zm}$nB(da0w)OF_J=bw+Qs$K|D;)S7+}_L4G5L};=Hwdk@^Qf!c)=P%#QD^ zzK#--oiyt7<9-RL*dgS1GR}{=1);n6(SMS!M^Z}rO_*^QiaHqi9}9C~w``SO?(C=h z<>>%EM#`Q^8Qwmk9}N#GQ2HfONYeD1^TGE&nb619&8llY5d*$AK^R*gu>2o@Sl5g1 z%Hib$`4TUy<4)F1KNwA)V}nUr+BO?u4Yl`>^_M9?hDqIui*Vc2}_hzCrh+;!YN< zGLA|K=@dmPY^+Hf=2XH#8LMcZsoR1#eU`sGBwDe2hb@?T3pfoS;{J$D-L9`w7SIE! zANAxHBCA<%%DX0dmfYH4B?LB3hxY z)Q5z;T?uDz5*8b)TzF8SLHBt%XXqSdgwO`|OnG{$hX;m5X?olNDXD8=SVwei3JK}e zHt=9y(P_^{? z-5vaSK0&(K(mydsjBVDMNI@vTBdU%ch>P#Dok$l(Ye&H>6^6!p2PrQPA&Q}gN>Xe< zK4N9%O3UlszQ?@B^L2>7_s}Uq^ezR_tT#~8+Px5C8$XQDdbi4M_niC2&MtUT)^_wx z_ZCF5kVqjj6CtUV@B|C{*ZqIA$H}fx+E(QZYzHR4(ALa(%-*xuqP9CvhMC;!+kApQ zc|SSdq9l36x7W>z_l)#QwAf6VR1n2)W3=7$2vQ$qo}CLd!wia*1f@n5d25@O`>nSs z8v$C(fEai!xV#QROZ6U(6S!tK{L(t@tUR9&h@*&RDJb*1w3{A#30({!-<8L9I|C8} z1EO#SV4BZ)C?6x|WT7T`n+6`~AN(5?O^OKzZu5WHkVNM7INq5qGG<^iF(2{9&;s12)QN1 zZYeCdxSgd_!||?1n3e#Kh?ULM2rFalXvZ#wmom9cxuDwX#Hcoy5UGmrP>`_&4Rz&S zw=f={%BB+5aO&_AMBXdLX5q@}LGY9$HP(t94v30lUizKPc`4wO%5693*&6s>s016Q zLD@6JD$oI&>1ZQA%anZP3%Ldehp2P@TVD+IDooegIa<_cdQS?+V0owa#^J6$tTiICLoe(tb$Aep~1;GUZZPPe4bK|?qt?G)FH`KfO<*Mxv5EJD>5c?bWdDxkosYQtuViJB zo_7kn{<6-V@R>PG{(`KgX9q&(RM@l$uZ0w-aSGyufVs2Gnq>(qpSdO9G10G(l4tEe zM*Qp0YE;OJZZoIRVlbt4RZ9|AJOwPByh%X1TCGN2rc8dCy4l9hg4;Q+MwC~jxygeJ zTUtro#N6%9rDmK z6iK}{Ou~O-BNfw2IkdyUdSo z7=q=JkEbswDA+1C?ol^Xt4b23?S+gxy*Y#ppsVpleKWMG)c8A+(BMTOJPV389PH_b z1ZipUeFn#8t%qKP9`yQd3}D|%bmGWUgHTCD?Pp{Ev6T)!wqwCR$d9*m-&y$*F^Hix zx%_)Hk;PCA?_Q~De&!7OSJ?p+3A+F6zD50d0Jdz_ez@YkBq%#2_8}+^B`=C3`KdIq zU?(osj&&PPN-n-Y2Yb66J2S(0Th4eArR~}Lc26_TFD+6KtC&dZFZJbkgvRq<204dC zw*b?z^Ab5-{D%I2NH2fSG=PO$uB?5L+>x}la{w^(n$C0~Gu+^I<}XR+E^(C!bVOE4 z2~~cR^lECre7795;}zvW23FZ1^ELoa_}oGY(z`_{gMWVuhuZ3A><(Sn>PSk0TnSqy)rbgkZ@MY z%AK%D?m@lGq#rb_JZI4;19u*}9_{rck^k8}lTF#<6zM(NFl0Fjx{3W_ zc{29@IVxUjFX8eF<1XY%8DiVzJ0(4FC@X#NLprLJw`ahJ_-O~2)z_Gpp)cT&@gD%& zo(hB!Wr5`CG<@ACyJ@D>m4J5OXwIuGE|SO6pQatxDXXy@j`<=sax3}u_3H)KP1c|t zW2wAu>**gE{AAk4lJykH!HpNo;)Rl>B{&Cxgl~nV6%pksG8!*YyzT0HM+dkh;&F zmBzsIbs*>-FY{G;j~koOBV0~8pP0%smHTE=!5LraMor{EN6|BCNNwh7x8Zwh#U51{ zB&qbJi0M0CH+_zF+MWZ#AnEJ1a9$V5A=E+Ylt3dcOASM5Z4N+NuDT@`YP8k{j6J6F z3I!E$5{(}dy$saQSmuMCbx`gNsY^k`7oKv{`|-Db zzRzX9+|JojZT>8O@vv3=7BG0AS?&0k41TJ=DtR1r&Ug{2E)A=tvk;dNO)!3L2;q%j z<0Un-CYkO0j%M;%!%Y6}MFy7d+n*UWzxx}Swuj@$q|cx<`0C?xi+c^#w)Y5|CxGvR zpb;8Rn$90VtpXUco>qCZH<7Ci^Y0wG43OxQ6JP^=Ye)GTj! zR7s#3Wrn+;aYx~_+~ZgCb*ZUii6w|jw`v;N9xDX-o9r|pSWeP`jKcabrJfgp?DM)O zr}Wy6z5ZA|W@-Y7hjw@p49UKLEDZC>ALH6C9JPJWBZ#_(-G-7`ZoxXdCx)WlJW?m! z00H2^K-ws4Qp3K(KuSLMS%n59Y`WdB}Z5x8#TH`d@!pvZB zr2)%>J}DxViQ!Z**iKu70I!%INw5VoH#W);EPLj2e<#~6c_y4~{3kbSq{N?Ofzz3bOq2us;h{w-o=@wPYz;OAp>L~7!qTN=kyYAbA9_Tf19WoIUw*b z!GFt#sEiWA@%Grw3cugwY$n9rJR^}rf^$4P+z#wY_I z=_D%o^W(VLw^C4=)~Fp|*MVjoI0Vu05AFoAl|z~kR5IlP z)9-%ApL_h!s$3K|BB=Cvi8Nl40f^~c>;8SxNs{p1?6emLMQ zikn48N~-#XL=@$#R^FRM4mP;kMm?$bXF4anqK~0=?|42d7E5=qVG!7h=dYwMbs3dV z60mPQ@+jx79yiDPgPE|VG7zO4Q z6K=vNUM7F~a#C1biKzZ->%IO{A{$eV;Q{AtJP47_m-}_`T#mToV{_tW9R#RSQsecQ zpA?Wop+a$QLFuunj|cy8Kg3MR5-O3sWRNcNNP|M=xueyI#38Q0vPL6Gj?U^`OO6M7 zIE2g;$6Y_vndpU}UD1{75+wBsxC7ur_X(T#y--=oT{ffrg}YjSgFIKGw4+V>a7qjm zpys3T23*>=UecIT>Df8%VdQ?JVgDHihayj}R5sQBan_fp7#72EM-l8U%B72t3G?~} zZ>LAE$E`*iOu0%=*7@vMW{S@WHy20_pgb z1&RO-NVxW{^eK2A6f6HdD=JwS=__S#0C4+}(H=IVE>GYpw}3P@W9#h^`6 z-+^fEupzl4poswa5uudc+(-a|h-OxjTmuJ(mg~}S@q@&3!E=03fCn)LDWiA~116Vt zA#bRNAbm_lRFn^t+co`?>fucKooG?6vYe#{QB34 z0X+$s8lzI+BF%FG!x#)C9`M06YNJX_AMw;a{{tY`AgYbjW0J;_rcNCUYRd|ku?T#B zZ=?=F>S*vecJ2=V;dU;31JYV&D^aD9fK*}IlDQ)A^2&ScM>{HC z{T&*VtYFzK4O)ooQ6b|>94(C4?_ktAdOrDGnuhSx{Wwiq)6R(%%O@Y|Q{R2uUsZ=k zght-^e5=cme{d0@@^3%j->fx-GQIx8c1qW1?E!o^|x~bry1_OeF;_ikzv2AzL9=%>=ss3Uo?*Y9nvj~}PXq`WPK5P_!t#+i!pvOL9r(e{5I)uPGV45|x? zKd%oBv*;&J!;v6>VOp{W-kl%G?jL~EznNRi33>%#KiY(tUX$&L6p8)6&KJl_fWtjvW@_@e&UgTB3a#D{C3;29#+@xB6-+sd%Hy!rSlB^bMug&>yoKK`M=X@xUKD3nI18;IUigyz&Vv-|GFrqFBZ@ z*rJFfpth+(`wp#W4S0A{K;ozK6w8kMIfWX_w>al%DOJE4;k!~z%ut_OA_rPjG-T*= zz^V&tx^+*oKf@#3!$5;Tz%1b(vgr`Lw(-F1yho1|A$-EU#3<6%f;kM zkT-3{EFpX==#Ef`Xtot7>uy}SyEGA1UJ{!QdY@k;f35s(189{oaGr+H1{ z8&s7&9wAAb#P)HxUTQ;AO>eduJzdR5ug6LR3xYMyL&2(44SE)8sE!MRFs=K4lQs2` zLdPDPqP!dyls(CF19y(WcWalXH!m^fqkc0OD`?)|Q_2zDhZrkp*xf-X#+v*8{45W2 z!`O>)jownCD!FO)v-|!B;QRm82Rs81+x`J?zsk8mJzr;B|G)Si*t{5LWbm)EqZ~T- zq%;8SD-G(hq5mA=$!vvsFmaNo!%{($SwI`YWRyuBOJV&*-=l#zRAn5CH6PWNI%CX_ zDT_0&(V)ps&DWBok9K8#=UP_W$HBqlqKpzcUweE{prP_n8`E?Fyrau(LiBuVC7Vf_ zm9qMiVI~y?6Zj)Fg14yzL_TA7kq3%geGj=j&AZ@p68Zz62!^;nw#H&UY9Kcc`vFBF zlqR^1;%Hp7pDJka^D#wDTf4Ob>JfO3{SsW1Ko$m&(%c}~{mJ?m_Ck>f^yByE>df`D z1PhH1%xAR8h&yQ9(kQ=I&MCL{ZW1Vt@<^8weimY7OGFtE5tIcx@D=u2V>6{I)=_D6 z*CGvGtdUSpM>w8{Z3-|vipQ0NmT#X@=8+02J0)CQa-j`NH6|GNm~rS%SqDxuaY&qx+))6q5!RYenbH~9h@L1erhYbxB0Yc zh#8>ww1a|~%VCTsK@S4qpef)KGMbQ>)$@sgYq=UWD&=xxiLu&|Zof+pi7JLOv42v_eusP>steY>npvlhk<#Ajqj~(W&^+Tr zCDmpKT&TPxFMP?4?GpLUg2}*Dc@B%T_;pK^ytgYpm1E3dv;}di9`|FKgh0fJpftv- zA-Q)!yCmWVhP90S{}RhB74cmH0c{6P@S!f}Pr@>G@F=fj^2!5&cM-DMGX4mf{oHiN=ANm4H zaSFm;MH3TGLSZvc;=X0K2ETqYe*TYM0^>h8*yrR`v6oa~@!E&~QH@jaC*!^AHRq+d zvpkXCML)y;71^>yuiMLpCGn4~cMbG>+ zw%o#p8E(hi;~`)@^EwU)d$dpsM_u`$xeib6DhKmcpI8W_^h>)DyLW=)2ahbb0svNG zDrHr_Jrg|un@l#8nZih`-K|lPJj%nPvdio1mg1`mlP#Pi;{zoj%bB+!;U<)|jaKC0 zu?wg4r(G?M!4j|3=kWWrMcFa{yPCJ9JD7dwr;1kWT1G8&s;arrO(21gA^Gi6nxI4>vG2c+<-c zP^TCV*_j)1n7Mm|jQ0L4Vk1kad$`~2i>?A)UWlFB;E^QMn=l9R9P%5QS(2`4q#Ytc z4*p`oq-5@^uHK0J>Ezh^^SYGrD_w9wxUzwsrx|#jDVl&1Po-qP>Jkcgu*cBk*r`WG zL$S*t(|#k1C9W9Wnh^h-(Bm{&T){7@L~*;i75?;tkKanT6fTJ}*0APi3|(}EbfKVi z>3J4Wn#F}*E|D{gnF|st??tTsWEjh^tb;e+iZRgbJ6PggQ_^v#|&#lqYJN1Evs2&%h*`Yd}fn%k^Tl~MtXKnCbdv&5s%&Oy+NrdKNHfTLT;w&lp ztsiNGS`MDrgQQ&esC8I4MiWxVvz#oYOI_F`YY-_ghA@;DoYUfT5^qPJlLDqUbHE85 zG>R<0E_UDe(Df-`POGjXy4GIMm=|9UEF+v9%ula|9){*X-7MnZ4@v8zRaMWMy{LRI z5PZs43%bgkJUH)?&nPeU0RgQ!*f8{ z;jk2Pd5T~ZcSnSJ@FM3&%XN3(^x(lbmfZp02F%}^I8eQim2pD7{OWb@=Kxedn%Ps$%b zvO}hbHm^Zb`lWr|fgc)pvIKFwMxNMR_bqNPG)=sF3<^T*v&w4<-} z+{6m5&LR<*dycdD;i0~NXtc7D5<*bevnXnU31=Q>Z-wsr*n>?H+<#EyzDk)nr2Ue+ z>&IuzmBazllN9yzyYz$U0v_85Vwi~}{|eEIn3R5TF#wR1m-&W{k^ET+67db!`Gj?0 z;&-uGzWJrAN;i`8XQ`5sf8rcKQ4NXW6_saRI{#~>G1lWbNb*@=03%vG+u6q~G{zf^ zr}9v;h;~vOEo*+UbRu1XLyW(qH|7vEp4_~<(HR>qeaqZgYyr+qiE?mgo8xOd;>UX^ z7Ht`S2W2N!4_k7|v=ONyssEsN!sQA#Y16_j!-c zg(IKQbl-XnnnWN)2$QJWKEB8W%Mg>0`)95c7pKoJlI+?RSlfM|Dm7q;mz^Qd!7!p0!dUYX?UFP)@ilV;-3*c;Yu5OW@H9X-_EceqfEpBAhF!gJsqvXMk(Y7Vn zInDAT^~%q5zxP;_NXD)E@fKRAM5-%t?oDNMr(fN(ftmO;=I1$26T{-xTbAmBc=HhM$s+XZKdiqveC` z7|~9mnbOVE5g;Iu^mB&FcWeUAR7+B~=}SeHU(I|SWbx0ujiuNF`@$d!*JZD1+l?HE z3F;o-2v%c<3KXYZ{BW7@CloDCc*dHUAomoZH;CehCg0gK)6CK{;sJgAuR2S7+CIX& zx^(COH9oYTnhRw-_pO?~wG*t1!jrSX%!~hqF$Sey*pd>nUCZWLY$^OUvmwgO+Tu0%QKyc0XB-S5y)Xr`Xt@W(f%kohouPc&t# z(w-HIc$0|Z!7d4u`l|gS&yCydrfGVjKbEl_H4E7%<>kJ9Q>pvjj(a-f9ds<@dclSn zKS82b<8Dj?zvA5=(W21)h~>tvbUgaeioJw$<%a_lhP$WM7XIG{q?9=_Mn0=yV~eps;oyebM+Z4n2pA~T9a(i2ETTK;%HIpfj35V#e7qy?u8FQ4~!kPdJ+ zM6Cn~^L3FPNvW!&Mr#J2&-5w?^C0wqt(&4rP&+zdJP-g2o2}T37)NZ@tzjaWjwP$E z*g*~&>R?s6nZDGLz_f(g9!9^FZx^3twd0Ke-WqY68ZCs@Vo214K4ZS7;8DihrbS<1C!q=>~RZ@gPcO0C%kX7F_`PqfKI7~{XmEejp%pEeb5{yKSku@O6ZFgq8)5_~Hc zcw*4nDruFx$9?ySx8u8_R!XZ!E^@du!E=puX|Ap!FN!vLOhyP1`}({B?kwsYhS+a?u3g57mh=YK}wvex8W zS=!dUw|fx*T3EC!#gQVir;dlo17G{wO>1-oFnPi5WlKi*XgeJj8L|98V|y~8l%}O& z+4Y-u>P(*NDq9ZaE&8G9tqMW!Y#+Am)ie9<)vmde6>mBy9wUFeV0`oS6J;qIEiI?s+?@)a_1uT~WeF#Hb-&tFeyQ?NbdZlOu;*9BC zkxJe5DpH$EEZq4f)IMh(4Pt-%RV4*pH!m{S7?ZVS2@h)5FHzR_@1j$W7j7V1^IZSR zpp6Ejdm1tvvK?NN8xe{2!T0-rOhw8lpa8KstO+Ety(j@__JUA10k$g+qzhuH-ebe+ z+}Ed+@GaB(0t^fMT&P+aB+L^u7=!GREId(=i7A{Sup&(2U~M<2r$rl4E{y+r8qcpy zElc5M5$iv0*S``SjGJ?B#wWK_ro0%&PybbTkjN_yA5Zn6rcp=RMsF=OQ~_7U-fg$G zjBKm#;rBsk12S6SRfQ+Wzv+=%j3}*5@`$uJ-B`b`0xe1+EBXY1kZ%T<06RGz9FSqF zLGI(I(Y=k>QX%4aZ^NNAsB&RXt?|Yxh#GfUsk(t}HNBkD3`gqceT?qH?kV1q! zZ7tt#?v*_~?qUM_KiK^DYT&O0K<&e;zJ6jwL+jut!T)nbKsBi=@HB^a{M#dcoY)KW zeYQge(&OMKc)Nq!ai3Oki@v5-mb@3*$NQGxA3^=h@{&%o^iJq7ygpR(<{f68vm@~8 zW{UgwLBC7Z-TCQ*KJta%n4sU-Z?}>k2+!-?KK*7yS;gy5(SN-g{%!nbn_98y*Sp+j`G1ldSA=bA$BlUI2!7N3g$?U9-7=Y|#r1+p zs`Q85*al>Rvm~hTB@Qn@@GX^I>Y;B3s>bxCwZ_Jlo55crzL1y>nd<1vUP^z?cM&oc zsBrW(vI=wh?}MDjPY;>nDK8P{$DJlao`LxWCyKZG1@0T+ZEk!WPM;cuD%J#T5^nKg zZ%+(8zU74@$*BpAJ>ds?-9i_(u^Exp^o9exCFtZKfEi>5W^TaEg0D39hW25;mPI<> zqsOe9TA%pJuA5|AhL3%2dVA z)Ip zOms*Wq4bVxxk#FNs3Vm7b$`VFe0WUK#fTSA3@4MAKFSYQoC1}aIrNsX%@}GT#RB3L z4N`}<=o6mN{AA30zJ8B4sj;{V4I(;xE-;y(KfbxA-l%t+E0qFlEq%);`vBT zVFYYa=R)5MQ_)nH2Y+U=tlx%}k^J;0lPE=oYfNx`PKuOYfbd*NS7%Y5KT25pzHX3F zV{CGZ$3PI#bn7^+DQN0XN+CS_(dw9NvDDOzM$^6)Cq>yv`?p0gFTBX6)IrvJyCa)B?qf@c%}72dlw%Ws*30%aJRR zT%bro1xQ;{%%PEFN_6@}!&rIEF3PT24W+V~Fh2fvKeJj1Wp_@meH!&#c&S%s?{Vx7jE9qbJ|i;EEd z=^nK{6CfDgm2)Sb^XiuH@pYnafVmu*{5z*uyMCO=NurVj2!pS_n8rZTSVk9Vy7@yc zS|t|x9=7M`M~Ox;o5QAv4*#~blmx2e07dC5TrK|Xm`23)b`7}Ss11uSi|cVU@-&s! z(LtivFN=HGER^{u%O0m7D+5sTDwayr;xH~hAuT2DYvYAU5%u|rO7hUi{L^DGDTIfI z@$#s*kCl~WRcQ{hn?N36VBO9?-e6HL>@5;+mm_`+sD7B$uvWf|-i_kLem=a3r*j97 z+?6?Cy#Kl|_1fytXBjov7-V>G2a*=-=jTW@C5(LBw@EMORI!%Q#}Eiq{^8=46&~WU z`9$mX+0e5C|6A;urhA+^qB(9*2u+^=dC#UYGZ_ioRwFW_b*bS3JXnCJ9_QAle|q8GJAq55>)C!K}ZT2OO)@c zE>S2!MzS|XP1P+-d<%Q^@=le&kU*WuTmahzZi*I#;`lFOc}AN*@uW?W12AOzQ7%M% zONd1Wb4T^a{Zy3qDu)sJ*>G_LO{28f6n@q|win!6`?^l&R*N04L1t7H{9H-MHHt47h1{r=gmHQiT_9hs7iHn->)fOU@YSmQoYg=#{r-GnLYSX9z0jhyILN zgC^s1MQ+`@wYH2T;p>qg#g1hmVZyaQ;br(RhTopJfqC!yHK+fc? zaj+8IoQ)Ev&FFMpz9^sXa^PfV7`<7d%&*;Z>4uC0}PEPB%i8ZDQwMyR)H1=3NVmV6`?_8;?%xIp}G3Ym7i%~ z28ook2M%OQ)u8fv->buTYq)~grFD$xek`L5`(TgyczOH04!(EaWV7D_IIunJx_M(QUI;MHNv*Haesa~r4jp}_Ph7zBZmXS(a=e}3;iOZlTNK?U zng=-n4e0WvMgnAU`8H$?a*IJ!WgQ^jLbIh8iELdSbcry5J=-*^=cfeMs=F44KD4fV zF{&taAN>pN|3;WMMP}Y76=>?*^z44_W08k@Nt*95Uy zrRK}K3X*OREZR}{4v(=&0%m)M55aYFJxMLZlt<8_U6(OgdG3ylih~ix3dg8;| zV*@*dy!V9e>(!$`p2UW%Al7z~4b5N$9O5!#n{Yx z$|i|@IRF9xGN~URqL5_AHseYYB&1AcR6WA;bQWZ_R+`puZDhnN(>(y?`d*4vH;5n$ zZ@Hfu-P$Bj3*sf9ib?VPpb=c9`vaR z#Y^m5P18_ba4sq5Xq$a&`5_}lG+DCLI;Ij2wEU`qlNneW$@Z+D5k^{3{3xF_Z@7g6 zQVn2&j}h;-e#4JTtYa2*pS2Q>321Jqg%2QkQ0#aR(~*D7pb`>JUkBhUEv5iq?GWj7 z-3J&M>Ru||mL4YYA`wmUCN3?6__|fu3{Z`Y4A+~&WI=Xad#5Y`O9Gx~#ME$jzf=r8 zT*`V&oioT&i5X#orNk6CFPfb%^s6fjWXQ+JWomXP`TZKnTm@=+-NpjaWP4uULJ%y~ z7Qo`v*hLHeWN0;#x%Q0MG=oV4NzL0YHF!~{O{olk$xByaF_k8MWMC#7UTCvM2t57u z2q6^*Mw?dZ_2A$BkPJ2P&HLGXKR%In@h3sT)7^esOwQxQ1LE*koAlZ?OAyH|X!@HG z5rD0NBArH}wsz|8RY(tnxt_94a1&6{cBlZ56-sw9L+;a}VCcx+=SPh3@lYsuzbn~N z7ni>kPOd#DxW5&pauOwGS&?PN_dEd>h$oVa{h6qY>*-r2*=z6k{?d zC1VJmzh;*_+;@veWp>loe9zlNFwIa~zW8t>9jB?`$n?8<#mW$L_Z{0H%{YLf9`I+TPpUIrmF*={Zn7~|Wv9c2U(t=#K25p^R-gd@sQc^+T`GGNn3eAlzZlu7WO^__ zgR8zE4u4g^WS~rrXaa##)Y&^!2mGdZi_v+-^Ix_)8y4UDpWng1y|u44w?-ndT|(q9 zgLBC$09j<@BoPK=`X!^W6?)plJaIrQdd}hk=(839uEiDU^}VpTZw8MR!r>!9g`~J? zM7$n>waYS9O*u-v{MM)A@j?%i2dsHbi*eoI^>WWW)<(4QF3&Ppm?BAWl<1!5A)1$R z^@~A@_)`ShxO%+HM|PSb=G57vh({d_R?kG!EqXG7UL8y_f6HHj5*bq%mL-Uhm@t`o zM(<+W`GUvK$43gU=x{~=D1Yo-w!aZagE_AHKix7Fp!g+ms=mhQaOzCsqE5p0|GsBi z{w|f=%#({3w)y(NS*KKvIys>y^}q=tsHjsEbN)A8+AzHs{(SK zM`a`S2>Ulm@fK5s~%8I@S zBED0RB_$k)ibky4yJMFMl)&j6h}3Y@%G3(!l4vv$82uq`(xx>wEB2Gy@Ksy(y0M_+ehe?j@=3iq4nY) zmzdlq!T&QPY|C@N;C>|B{o^P@Q`8J`KWHx$_l_(JRa>v48!Ip013K8t&zIb#s5Yo# z1g1u{(3jbFn0frvm!6ao3}!NT%UAJ}g?w<>upK_xGWu{;;n9z3`{QI=w*TInJOMo@ z`j)S~xmhgT)z_XEErLw05xvc6_ld#JF7^|P5hQcWyPyRuu5NC=h8i0rHb+Qq&U4lY zs&q?2-{t*gfRmnwE?kI=3Zu2!!oVjg88VZIhD4hYoCwdpX+SNHKO@#M5eu=3_ogt` zc#mMWlW<8{@S}nHoQ`x3R5nuj#wK-f@-QuI!bp0x zHB@V)aB&*ljb7uV%%*eaSoS6&4p_X+1hHfOMn?=TYkrF>u{P<}fAu8TA>_rn8C(cx zYTzG-`?V2_4ok0Fe?$9sp}6}E-MBQ5p8Cst+VOLvA}$Tkp|Xl_P7@JcEA!4iHRXpn zBXHMyzF~H@{A?{zE>9oSbnM6aXCRj6EG)T;%}E-fL=bHHZw@@XM{QXu`%jgia0O~; z83>CkI%b3M^gJamhU2|m(OSJ7l&YBmxwL<&RKy+(gS_?*QK3(*Wdq081+%k*m^5V9 zVMj@8AyA`yWp0wN&JS??VPn)-A607+O`JKmW0Z2SX&P>?O54o-+GNbv zV;9@Z81itN8(Gpq4`vSGGL5Is;5qdzJ(adUm#IMJqWat?MyY=f_)puonI{g6M=LOu zBU(m9<;bnOC?`k1{1l=#l58TKj4M$#55db-J~VD$&J@oFz|wL0pA0b|9}hVBut3=p zO{6pg4mCrFgqW+!CH+~V!=g2Ds}$5TixOY{r+2bC$kVymxexXqTI z%0a|V6WZT=d4s;z8-AYo*D)XcO@ay##pam&nbVo!`aFA+x#Hzk6b~I9>8GJ5?{uJz z*+qD4_0aN!WY=xcAg3@-o`h4MDjs?3Z}nlFzz!qoucB(uZ4n7sO}MGg-8kde{ZVR$HI$t zts=9csAk6pcB}>uY~53(kE+}X@QD%rtpD z-F+5pQmE2*5zXBuU`h|9*`do*;%A1{k`A?ijj+p&Mxs>5xXcL>U@skdhp_k#3Nd4naWy z=~R>u1x5bP-1l0?`>uPrj^o*SW@lz&Hm~a!=l486C+EI6oIPD5-geNuW^fgjO;}4) zwE!kF`d(G!7f1er@?x%E)bIC5}>Q z8)VGek1{_p1q;!o=c%f5H+|S?o#3f%zm=rd&75WxGpO`KE>|gY2h5jneJ_%8)&~@s ztCzAk>(Ymn|E<8Ei3JNm40yZ)jp#i}<0kl9SZImYiLn=v0&etR375**@UHbTDpoq& zv$70UI@Qko+-935Mccm^ZeJZBakZgOc~Y0%fCVXdvZ$cGyP^m68zbtP6e%aU$=V_s zQ!X9M)P~f7qls-?tWrT+d$f*g%GxEIBME5sfPk+(&956dKxH2!G{j71UP-bB_xSpK zS6`8w?nU~TBXWjde1j2&AqEDEP{aPbU|-fsvo<|`+uo$Z{E`O)Nv*+H)O9?{zN;p& z9lobxOCm@)o9B8;ZmA@Ag?IRNs*IgC&}q~=AG0WvRw=f3BAgxu1=o2;#Y-E2`5;l= zq8Zc~?w%2#wYHJlI$?a}l1u{tvFdi(@(j^qOpPibUqXyIsL-p8PFPUkbf736)y< zX0UzC_3D3)%bi!R-5uJWhIJNp8ISl8jcQ}Nz&$?^M}GQtnQpthH>9&z^XSv{gxpWO zw+(@cOd{?Pe;kM)+)!al6^|k!rM}b9Zoe5t2+XUq-^(~%SaGbxI&jZStIh*PZ#HWK zlrk7WP_z&n(2)Q1O_PQjTo^TK)^R-=U21w(5(!cd_y-_-DQdUi2>{?R7vgLT8$+i> zYgS=R3tNfwFDnRC)9UaOHV4)aNfmWE#|^R&0^j^p_yYQ2r}Fxl*PRx3F8 z#MdYWFe>E9E+T<|nG0a{2E72p$U3wP!mCK`Ifs2oi1{;E{q0ffM4pj3I&6l>gcu(H zfQ~Cii;ToC6iO!5PCd29R8X-N*x4C$iIEw7j*E?XV6NKu@pN>jUCu4h zuH@(+ur^`Idap#yB!n4n@=b|qy*r5!glsC4v`@=@C~WYjjd{Ls3rcFbKJrv`Z~rCd zq&Mg^z+Zt>RH-yo8thFwTIwX3UUWzel6>0!0Ym5`$)OqOK98^1^+g^Vj-X-OvLlHv z$TepYei9)l@alP+HiccMO7NSkhe{8E`tl!($7t7zRU>H+qsSC}v})a|siUeh{cU|Q zXh}2ktynZ)OT4gU73q5iy2VXr0_8=n1j26}Nb5*+ZK-mYPah(2>pg}41htX@&g@9L zxru6ANQr>r2Hxn>Jx@lDJ_e;cO`b3QwpD7kul~IWJooI(@x7U$ju`ey!})y55%k*7 z3XfuWf*P|QG(6+wi1el0@U}qW53!;pysySD>vZV`L0w|L((l9yQ&3Ei;2afCwe0AP zw$S~gl1ff`IA68iHQxpm?q`Q}B&u629VDYA1S1_+U_9YM*prHNKM`L-B&~SUh z#TvDTaeOc$633UOT+?7Wxx$#fTp}Z(+%$HY;M3l&9xo(DZMkEr>`}UHlVyUXOD}E` z8*L07^3~z}(Q6)ajS3^ojnN0dAeC3Z`t@GP%`u03m*N8`(cxio_uGfVfda8L_xysB zny*=$lCX*B!`K~+J+(LKXPND=jhd==ZbUv)ZSr`akdf#*427Hpt zfRA&oDG}P)YnIxn$~+b*R%6Z>%+sjKtGE~e*lSU!SMPLjgqsl%Ng5)ee0Z0L^0gWy z&N;aynEJbNfj}HQ!5JQsM<+d#HF=}mj4ENM*4i0thrMDZyHk0Wu+1@p#& zNV%Zr;)@>SVpM}KfIxJ2yNj)rmLN&ud?n29tI*{d4PLWsZMXR~0|a-#cgDlFnV zbmI%_8(bEzxIC!wGe#-F6ySm{wQ;y*iST*}-mJ>zmrNxU01Q9qJJMm(eLwmu;*FLL zw=9%+A=eFzg92$9?2~^4f+teWBBdsqhsh#WdV1ebER{BGqqGn(FkU0MSlT2>ByPRM z^HZ>k^dkBgit*zemQhjX<{I3sfXC-ca!vfN=}oc2x$X1$s6S(d4VE0Y^B%$$@`0xV z5n10Bj%>L>SaVE}f8pyxZ+NJ~wAEeJh^B1teXgT<-Ggg&Bk;M1I8KW}=oQea5J^cG z=@wkgK<+1gP@3cN>m{+!Yuw=)XyIv$mx-IVauG@Wd)j72rQ9G6F1Q0HeJCro!oI*x z?~A=20W2ihFmkbpExcLHk=wAXXVo}x{ih=Y=4t>2+1B!PtfB}F7R$IF9;l*=?@L-&TwN=i`FeWO@+)UetP|U9msG2EZ~SM#n9M%_^Y`EG z#}7r@j?AQSfYyIU*?Z+ztlc!{Dg1Y%S&*j z5mE_4sVig={LY!cuVb5aEmRyWXK3$v-Wrkn>c#uWMpa?am~?K$IQIyiZ{5MKG=z_4 z|9+iD*7E&4wb}JU94f*Je3825(TeE%C{*RN%w?ap#$8kGEN|UOtH6_)ZFMuNt>gY} zN0$UfU5DO1*LuO}o_F0~d!-_C+(Z3zXgf)bjC6T#BbpqtG8+BqjN>N;kuW8K)Dutg zB>|1lHFHOKJv%aFk2GGZUm|s(lufSv*z4)T!m-fis=Tm0^xMh*5q#7|k*|NtJ-2F+ z9%{2Nip}_PxGeLqjp*ROjjHxFi5hc9`XJFj`S^bTQh7yn$LQEWAxhI)Fx}$3LRN;WEa%lNR%2=;;z$e{PISD^{?1v0rk7HC(aAbJfuGk zZm18@2;l0?P89XWgyb#Gjb;Zcn!M*1y2{&_;9}9&Bc$u@`qCHE_VS z3}k;jxLvDo_u=|=!EgNY6sP*(I|SqevZy{fO8b6-U_S>-Wh&pQo($lslcma)x zjZmhf;mYfW6nSfbzyPr!^S3(rbyCU}xgWEg@)7quiOK=UBd=jOP87olh&55*8A{LJ z%oe9CYDzh5P2&NodAM2brx-iY9Hv!Ekg*L%^CQ5z3j0zVSoI}b!kSRT%^KT=VB5sb zr`~bv_yf-ApsNe>B|8h;1Q*8WH`--rM&mmE3-3{o>JW*9(km;OVSM)}?*u}G3$TqU zTUF9i7hS`vPe{BG9L)=E0I3niAN&wQL>0vwVHg$TvxFu9M*#e#{ioA=;CEd`R0Q?S zCFc2S{{SMJ zaOPdj^?oRn%(*X}9ItlX#Z5UCeqSbUBT$m_ys?^K&cvVQ`Q2qJg(2B~;iPhexZcP2 zkVr9aV%R%lLbNBz&(U;NQp>{hh)&#L^?=MeuhlSe)$bDj0K{K*Nu5qksxk&el)XI3 zaD}wv^ab<7S7n`F22@;(!Ffxi(IxN10;YAAlB9$uaU}8b@SnV-1K1wx6CjB_sn6M%7a!_&!N>DPhC)b94_%5dI z7xvr+C-U{Fr_TL*^-^%|KXf*Si-Y~y$60g>3DoJ}N4GyhULDPg|Dip2T_!g!@Ed6z z-k#uFe7LyytyvWl_bv?g$ul<}M&MM73w4eEl}=rjjq_QmnBa8KR9w|hhb;Hb{M31+VI#7-}_ z8zwMvl0PE@@jk1)_FzWVw4GDTZB!dqOQ-aMhgKYU zo!?-4v&2^co|x>|z=RrcZeXp$?`Mo@#69IPw)TY!zc0LMeABNY8~Wqj>E{US%1EAj zh|fJFUkdyNDT52q^ruu?7DRl>1UZePKq2#UZ6IT+i(B_POIKw6F6ljIXU_IE%$RgOMe(d2!cNYRAZM zR4ozLXyFjaWq_LvUsbH0)@`b+t3h2FOq@aR{?d}XT1-)HWlGBZhWKBa1!a;Sv*gLWI$`VxMcXVC3E=q%7dJrH@?*9bKbm zf6o~V#NO-Ra6u0*$wW`q*e#5ue85B|&jtprbeVU@xQ-K1*s&lL?Z|^vU6}MAp{#u~ z;i$2r?>MSAHy*JvHp|wav7AI}olnr}CP$yGc8Bew5A;ho9W=uQz9Xqhvu7bF&q-rN zD7Vu2y+919RuJ&7=IccjX}Rll2O#GMD(wK~{czy11RuFpu{}!Bxca-?G88?Uxc35o zh-Xp_LhZPU1f>a>|9EqFX|FhH0`RM6W=3AGlPe?iGdjbrnF68;-%Zwrsz=6_5QX~h zsTQFqQTu(5aycI#a|z@dzU$}~6g4A>qGDTPWk{xd^241KoRnSAieW+{BiXYM{w7@9bGsgF-zX7NL?&<i1j27!kLz*irWq8bTZ@8PxCdmoB6q`Kx-FGOeW zdO%pxNAjDZGD48VlnYNEe69bP4a_;#h)-qpAKI(<@Wed?qIXSQsrMM;a9s3wY8o3E z!+vDqgZPnu@Kg|~4u@L*e?&vM1t*S4OT?7V1&`ssCL_AF-s0)<-v9pnU;kn^5$#1+ zOF9p0$ho*`fV3`kxJkEtb1_N;j2-#>VuO-?*MrM;e_}aQM^8GxDFj#@alvcX%Ca?=Rl<+}0L;k$<2DKzFmTx1ERhK`U?6@HuKh{!*l=?z_dLFKsU#07?h!f z?Wroipf~7;ORmO^-qG`P!78&?ID$;1xO{1NDc`*G|0>bYYi?ojP7qhZAb3<))&CVZ zj6kjcT?+q7s9eqsd&JDH=tki|5XdP0yt*HHpuZ5Y|9@_Hr6?>b)9scDVV(Bu zozc}o^(2>JjZ*GtB7s=^&%a98DLvqaf|G74LyfG-$-(T-eDA(H7W%8axvwlLW zkvWty9M$K8I07R4RmV$(r3|w`JLuCVndv4ZqhwKftZ&8a)SbcE3!a$-V5|ClVy(-h zd(c0fwv>Yhe%a+Ah0yz$546*ps@NV)t1)(^Gq)G5`ljGHx114Uf#y9gg7@;LHW8jq*)ZT+bHy3I$HT=q$_&u&x>_xg7JdH*f&hMO9 z6b$-fHDIP*ti#0@U!8sp^|2|*NrgHZ?kv1D4a-~&VLP&qY2`c|`YCkl9r~Jg_SQsM0nRpFQaD}^!NeEyMQ;a68Y z4l4%*sU-^&iorhYm(+mk9OGiV;Dpij-0-UTPq}k%TSa>O+-tHIjUo`EW@@V@03c!s ze<%XLKs3EI{Jz9w{S1wk3YhA6opdOc(l0Cjnl@a?&GqH@>!gD-N=ujm=Da1GD9Nnk z!<%?MP+HoQ2^OL~9HHvY$p9SB}+$pWT?fzR*ex|-X^l~hm^Gwd#;X%WP zHJ!tQWY_Sr_;gl(pPJkH#-Q%yn`bXJFQ`&iZs%;?{JH7PzfC1!tHKZ3Gg*$C#)SG$ zM%Dg?-D2ln!^|3YYER3q&+AOgpV+p@kZIQsKCHYi(ioo%ednVCZR2_|3kn#L8_Pig z^pprTyhSZ?Q9E_iX1uu%M_(uO?C`h|K1g9x5W-zvdS+cWJj$Tr9cOc(Q9$maSe;lN z%&YI2eNB`nH^5$50(eV62&#itf3hd>+|L%^=FmZYsFHX#mMDmFNi$Kzn4rzbTnCIL z@jyUKg?_uVG<|?Iu>Tv*?}bF1qG0_5qUfzLXjGtB-S^Je6R8(VpYfPHym`?~(cDN` z>?pTCrnoLrEY^3eH7c!%_b85gN{z|#tDaO9(;UaBP!pDMi^34jth|3mPLud{%6uQe zVK45=h@NCyPUTMjCgJ^o;U$pOzA8UUA66NVbGA#!r%SHFxrmh=pC)*ySgM$2d&}7; zHqVvMEs|FwFgWA15_=7zpU+xdYLHP-P-#PU`e|TR!uU)=}6J*fz*!+n*qm_)AYFDvc6SnMhj~J z3`k;guPEMpwSn9-4_`kYpj5_tVaqYQ+mXQzs35vMb=Vvo!wAkj}MsjJ+7d_U4qoc137$L}!% zK;;_A(IbySsvhJo)M#=)BxsdY_mg~YWuQ^>oAH@Ec7Z6T_*?jYkGd5|on+-4cW*bv zPCsKDF~9CH;Kc9UKetIMxjczeO~=KJNUXIerR-WQH|o*(Ui1$j-Fk2`i1;Q!Ds@2Q zz$Y$Z>c$p^aFJK{D*Bo4cU;(t9p7v|BbEmg$8v5M_)uPq3JkfK7ui@6=Los1=BQ9B zE-~8gvPNhk$gNqP;%L@E;p>}xXPmJk@FOV7SK zlO9w{A#4QsKqAoXm={V97|44~M05cpa`K^eN@?&+lE&4;dN8=SNj3GI=~k;JkS9c} z>SD``^>~)j=denkrZe$kf6cRFigl=^v`{NywfSez@c>&IrWS zp&5OahhPCnnq}8<;WQinHziI1HH^?t8o?LV!J)IXSe8w@QO!SF}hA1d^wG5Q&FiIcvRuGJ}r|XX?+8+HPqM71|tx-2KQ71^RC(w=wg9ACawHGUEMi`4;fv|oiI zv%!TUW;A?sK}HpgPxvx2?S1|UJEb?6{IQ1}FEE57nh}D07;c5DsGAASI@WuGk#uyN z)gbNyKf}ayUbg(54!X9vj>3dc)I`<@M~o=MOAdH1dzdHtet~}I*cu1Y4ll9dt(5>8 z{H}8oF&)D*^5q+v6b+CQLHD)}b?#`x^V{I!rwlp*?@W4fSOMaWEbz-VHj@wm2=P#y z#yPgn>{B6Qfbu6XAhNbw2c>3WQd+zFwkfQHE4QIgInF{4vt+u5HN3_@*n>-aiTac( zs-i>p@($lyvqnzK^e2U1FQZST8J@9o-JPwiT=FihEChf-7y9Xx`_h*fxoe)>Sxa5gvVD5M@-ds1G(rqXL%*epQ>KPgEi*O>Mo>am^GQt8v;TCoeAUhwFCf(()B)pmDF}Uv zarW|(UBoThT658%utCT|*_7^&8n!QyKE(RCC14t2 zwPLSWRLWdpe1N_T;4vqMXR*7Ui7TO%=!k_sUQNV%!NG79?n@f`@zp+Uk zx>$MLk(pC4wz}u_yP~;_=e%kK5IFv$k|`Gj}$v)mIc`-SBc5a?A`I#r6J-%6GS(j5Wb`U~LhU z)@6Lni9NDie=MGwI=qr8MxqUlvlQORk@Kvd*pfr8VOZe{Rh;-!emZ7N3pd z4-P^Jo;&P*U`=LqWk_1p+6YU{?h>b3_eT`seTs^GcdSxPvY|oSSGjh;PM%=fQ$P0n zdA)~H1NX*bskz*uH$WUl+_)%BT=*OfobF@Y9OsPPllGk&{IY3^(<)td0gtzNyP)I2 zMF1-mLDtLha0)6Z)^3Qk3veFbKa^yD7CGEV)n5N;V(vj>qw{m*n+2&n13Ur}UE9PK zV<{t5`P|EUra8H`cTH~=jj)ZA+^LeJk0N4GB!Fdc;bV{6Exc&5&~MBx^ZpT-EBs&+ zwcHRzGtjZHQz1E%7SMn|hjWn)`Z33wp5=kv0jWuOv``VpVmM9d7h5}WFE+`r^}QDY~OiJ)a-lc^T4H}Z&&XRInyi#DD zK_o3ic@$=omB58@^GO4OIx&DsUhX5I0P>6y_U6w%G4rb^jWiFQw~pom=Kysrw$^BD zIpGJ-6>f^*HvbBu`Jep=DH<{IRA>lNT@9qKa= z23VXJIrS4?D*$qidZaR{Ly%(?4Io16qNF%=>In%ROd1W6;71R3K;Rnq-hw?u#m{p5 zF^`G7I-~nzw7;V#uv4*kXj8CimBE=0JDmM%W>$kF?oHqxNtFNbM>O?Dro;?5=wP1^G;T8uG zLX&s2b8PhQg-^|wLE?%ycYa4A77yDV16+c>!h6Heh1@q;iHe8e+!7?U!r%_n`EqlX zM$Zt=6j~p6Y5us0ePOcNYFS$aOK=*AXBCOWEWYp~m>6_4($B7uoO`gv_Q4HES(~vW zv@f?j0i>~{wvXA=K(v(bdsNX>vkQx~X_^od=FoTNDv7_3R>9Y6d66%u{tuwiD_MtX zLF#Id*lQ`)lmPlcj=wTPti_c7=~XDgH7h=iB6!cCOKYM7>-P3^uBpJUVY|hbs-iZfxAF$5803|f4Jsv>ONd% zAt(fJTuxuw7))re2hBpT?yDD-LtNRA*M;@!kM8o){GgUlXaK$Nj31hJyRWOX@A{*& zc%-b3#mxQ*x5xvOQrF3xk|aSkpvfdu9^~>z@&yE5rF`WqdrLQZ**JIUS2nN-5qlo3 zexEBmkMdeMe(?F?Y>z~={H#8eu zua-ctn=p#Ea$-~|@~`0-chBpb7R7AJ|2}hGy(5jL5sZC0087gPZLM6B+(z1w z%Hg#PRj)Rw?x&zOIn7`fXBt~6&t91{fojpb_G>sVMDFST@Tub|;5wtlR%S~zDxl{c zgwxGEcxs7@dX?=#*7xz*h!OSl?aOzh3on<8jSB4VL%sUIr+lh~!;CFqr+M%%bLJ|^l`w71{Kpf={DbYzXG(?OV zB?Iv?h_e)hm<-SHaPgrLkqlDyIrBgSRT3r6AS`eBOkpPub-`tXu%z!z<2%^m&ovII zP>2)Yq9~j|ULs>iP(sN+=XsaZ3M<3&q<+~Cla=I;!9neD@+9l5-=ix_@PPy!2Kurx zr3k{>-H{5hJg@M%|6`2k$OuJ?nTRgGKA^Ql!JwPh9{_hFAJ;k;*?;`onVDT0_s6JKdx%XW^@U}ZkWuga6ldyx z;c1}>fznevS98T67McQ1e1JT*$jyHRHMw;3-FsN&b4K{kGhn^xL&CRasH_8NI~18+ z^GclKr_lT1qpE*6d8cX*)@DOO!rHOBGw)OIaq>g~lQ6};hwS|Yg{nV?m4Wk0--@4J zi+%~mA0}N`{C!In8!?st=~E9`NyMTmYg!^j(X4IZ)PyYT1QdgC1>$6}@-7f4!Ovoj zXt9I{$zp3~wj? zdjcyN19KYd&LbGa*fYj{!@vrdtxX*|$(nHOrmcz~`{1a&pHYLu>_=fz+%z`5mAeM= zjo;b@JaSrysw&L)5%mcI8y`%>UIj!)~>HEe<1vqJ7aGo`5^fjJ;oUOMaAbSK@qp^Y^ z)ax`!l6M8(0Iz!i>(5z7{Rs4{KZy_!_}(c>te~}}`INyU!dyD=iY=rfVY(NY1)HM2 z35$fc|82N0R`PdU;iLqKV?0=4*QrCx-uQyTyONi)7vKX0gwh}C5D{Wx_^&OkJdg+| z3YRZA?+_j^EvVwU1tSHq`~%SX`;PH99=+J%rHPq)VN>rfXJ&{9z$92ieqCaxgyQHr z<@pxu`P;NhW`Lc|xG&ks&odK?CZodG53iEv&loTEelhj|V=O^`%=Q_(EoqXnLc%m~ ztzVKeJKjVu{(8(A`dD(58oLx>RlG3$3@#{Rmt3(p%wUY?>PfJce9yC0u``Lg_2xRm zOoW(>e;ZqqWVnLIzoY>43AVGm=1Q=;{p$fH1iC3g6EuCitxLcC{U-cN_1K-r_qh*8 zR+ryi<2_B?x>l48Yn_YOai|BDMxiN1mX|+KEc5cLk?X>~Xe#9VBBsTmaDjxA{*Y#v zQQCs54XTGx(eiEun&hBk%9U~k(|601q81G9?RL2L>`@iWI<0x^(nwL&>I2p>Ktr1I zM?C<9LcI>GVj@~J0VbamCewm2p%hg1w0)Hjoay;Cx`xDq#_u@RD>J9Xl6LbQeJgo_AAMVq-jl|mPv*3k?>wne zs74|r_SIRFxM2=Kg73mV?w)0=&n0PqDzMmmL(p8=c>8XNY7eD9)KS(Xq@xUkswm+T z!qkDMjVkfeMxQm&Gaq*Jt4AF3@M!C4=4cVW&;tH03@fi?i z8AORV1rp}?(K-;&B9iGjT7!!+7==)j(e)9tE(DxunSzZdI>y1-%3@7n&0fj-V?Tw` zNKG4R)K>2AYfBr&E*&?rD#jY1a3XI95ID8e|k zhOL7&loLbRJyvJw>z9GTC&*4^sNN1S;gcNI(ZHPtn*xt`hgvP7JbX^J}rA7xh1_ol9ldQXtWdJL2{S*SQSv9rXQ>NaTyi*6OaUCP48xcIUa0>9 zvtga)o&w*4o2K!az7TY{01}qwnQs>)$Sb7m8SJ?kyV#VMd7}FoZisXvAJzpS?=_t+ z45Mih|4d#bXJ3z#rNfAcxjo4R0o8$e;OB3}q*y%3j1b`TMHE061mgjxuq*V_f+RGy z?^AP1O)K5Yrl=TjrmA6QKS@x%l<-=uoZyXKJ7?WG-DC&{t_{*Wwf+E}uG`!vEGKrl zRd*PjFqv$1FZ!0J%lvP;W|)(*a@+tLj^2H8o0%YC67Jh&at=}Nb{%rG8cOR`wUtv;p4i=gzK^W{8NQfzByYJ>Tid^`=JxgM(ao<{_j82?sp-?D zO0JkI0RR_y>)*FP%J9ecL_0#M!+w7oDt5S*hp82>+|Fa*=Xm^*YXC&nrtF-$NJ1pE zpYAUhKl&5QHrDJ0f(~Xc&o&wE5HN1+0|yE9R$j(LrQwWj%iT&09on}r`4l{r z=WMa%>YfA7?ynpOCXU<%OZslGN9K$5ifs3!Nxc}4@M)gN0qSKSY5)L|i~n9o8$>k9 zn5By61Z#q-8w$C?v~vtfEEyR&9|SP-MQoh;B%Y9(0`iL=AG(|FI;-jjZE}>_cIdu&x9+G*~cPLpudyB!T zH=#fvrRu>3Y)uCG5uut2Zq0@5Qbs<~ERjz!Ftv)r@vHRGv`VA{db$82CAv&vo5;{tH|jGuUCtcD|4!ATY1bHA#aU^3`5OEZ3@i-=CC`Rwj z7B*=+swrd4>+iz2pGDuMMM9=a$$%fLLd-x$#u^&z)+~Ni37{-ZS!l-BhS#3T;eT`Y(=m zCs@`hSxSB$N8Kkw#HOx{8gsrN(bl8>@aHi5?))cz_^sK`)XY2NY3eTXBl~`v#l&Yb zX2qxW1ywiK&ML@c#y2Yz9ct}xoMkE3`Nb&O@nSMnNBkTCMZVY8e)t~>1VgTF_P!Ox z6|tJswa*ToD2T9WhrLc0%~rYnA^o3c2Ks)ngR(Ap95=b1ZSYZ;a3H?_XE*1!8miGD zE^)Z;u|4QvZChjz35fz@uTS#zbPIEJGYNKkS*BO+N(UXX+@PsOGBHJb^SD|#8E@m; z?A8EWkFC)&SX-4eyu5pa_DlBkWb_^`rzWpnTOQJ*yzM1bb*TjCGD%7|4c;@@u2Q@H zVz`f^7G<|VUH5o3#PnhYMflymh_~XsB5$2$ug@Wq64sO}R!1%?JX=i8s_wv;GM?gX zBWymB`0C}H@aSm?u8cN$z;GMxogQy2Wxnn1wOu`%KL%!6U;i-keOd*CjKtWynq(No zqQuH=x#@pZNH?M)T<P3$pECNj>JA5@v}Dl%?sTU&*a`TLnn*M!dOfEt80^ z)~DWT`jX=MV4~b40B=3^(%~|^zSC?2>AMCmzYEZfG6wzxz0c}>#^rgJ|T+wSp!#0YGAMlY6@UQZ!gJF)4|; z&xo&RPnDS*6pDRKZb#?kbW>c{ZRiTgsP;dbHdZ%Tt3pLqHRXY{)&7nSlCGcMa4x98 zc-n~lY;jAjgG+IhINO}iy0nDW3--`ibrhMF$#*y%^ZEaLi4Kbweh1s_m0{tX3-(95 zL)*jQ>c1~)ZPbM>hX90qG!K8=JiAX*v_f`tSMdD}jV}FN*c0UR_dx&&exlLi>f`ff z4jT_Lp4KhS)RIs0*v5JV($VPNvwk`42&q1d^E7=lHbf-E{BcO;bpbWzS0&MuLV?WS zFWPsV3%OAB6JN-59Iek$Re9tinv}8>&ZT$yz$nQ+Rqi{RTbcA@reH!q9cGxR(QN)8 z$T4-x$D}f4GZY%??1A+u{oIU_e{&XwIR8Tf;2ss#ESoMcaa$oEB4HpuA*33kDIrl@ zOu5EYk%Z(1o)1-)YLv?51r!9+UxO#Lw8VBHoS&R^I9=%=O0^v~z_oIN?fPniT>&b7 z7pnx5!HXp)3{^Sa4Fp5ASg`A<;aEnEoBs{xrBMawz!`k*@H|hY#E^FNsW$$-smK6g z){Tg6;QS-7Ow-k~HE(*ZHYs%Kx z5dbw!T4pTmG#|_?dp3sj5Ur^mcX)FDh6(OeW+Oh_;2gsAY=Y+f)7B-~f90YFZfS;0 zDb9QXKlLUBBm~MC3}P8ICr=m-17!PdtHkk${D!sQb`d}B)~*?FwQ7$qB~N9K^Gg{E zrBxC)rE#Rf6&CxxY*z0$lkSR<@ZI>3_^x4}f*okx14o|@wj*hAnq`-$n4Ejvsa_6i zxjzp6KmTE)imNiKXow&7H&-NTqV5ER4x8N-RrFv$n&s!rJ(PA={YomZTaa18VQ3*X zS&yjJQv&f52~iKnv_36Z)PC4^xozZvoBA>dXE-NKgG+N9uMt&wLCWzN$xsMNQAb>i zoS(IJ+J(pJdU-OrcZz&VFZMPN*gdEl!$ z$2zu%yaXT-+8St}Ou7fSlgAwv$FX^JJM%v^G1QmM$Hej_uQqqYPqyUF$t-#;l@GkN zD^^}l8jL$QG7fxm;yIb$3Ur^w31ks{dbz1gIO)r~{$zuzS&o^f1jAK5k8 zDBc{@Rrs$5=s9QX3IRIo$7D2<;Hq8o6qr7lF?#*4qkp+#FV+ z6Y%4)Idg+gWoZ^%)ti!7EyDSbuAHQ6?$=o$GU@8`8nF`Iz+-*EK1=Zx2<98)D#cZqRO1}X3Dk^lFv;nj{|ZXm(cyVJi9 zq!;gIyjJvoHGaY~f2V;T#PR6693Hv$N7b^4{wi6W+}@s_fN*cFxq@ldMui1{OUi_+~=EY=fSpI zZ^JAij+p3Lo4?;Y+-*O&MilRVv5QCD(Bq%nGLSzdHf6o!L^zQ)4L(~J$H%~?Z#6>_+k zcOgXCmwxJqBMjzuqW@?QN`O`^S^XLt=>I^4Wjy!KZpkmMCO=hReHXf(E*1un5U%zs zxuK?Uhsp-1a8*WnS}riox!-)cr}V(OH!~&HFZG@=%mohbA`2OS={P}vQ`=8Eq}Y}! zOyj#6;kLBD?Ng{U#_i0&Cp(nzfaHL@H`yy0f%reWW54ZHail zg-FLCCqY$tLyOP;eG1gX{|oed%dpj+O{m5h9~6BSKbp767W{(bINa9z$^>RX=FpT<6gC6>*N)=d>fi`Ets(8Y$ zEtoK;`#tV(y8=@X#cwY4;+;7RomI1;)O#O@U8SqX=t1c>*AnH`qWx8K~_Dz3j(2*_f)vK zZn6bOEcVBUHNv0@@S^ak&rH_ipVMBS`^f$bh?oGoJSPP%9wtQV`kytGygx)nUy7*R zxr&37#>RYh*fZrSw2U4oLqiR=p=@8xS4XNb_|G930WN(%J2QhwFZYLHxZ|@*U-Kkia?Sb7G3L1M z-<`}?PyLMAoPW*$EB@7l+qzkvKkSPP-{{XAK{zmYNvLOfV$_~KJPWs_?fx4RCF=Ba z$|BL1WF0KlO;$Od0s)u|MsDboCpheatP%~X!wNsc+isr#G&l+2-Lnm^ND>pD8ZtvB z1GW1z?|}bQHHG+S+%bGDPhgzpcm2kX+HZf@3$2QIp@3ZrsuqS$U}{Ts-R^i!k`_;Q zDh+6xVmM3ROLX3g{+rxu=4c=V1M@>l&MK~%w?j>Da46rzx_sLVXH|H2x>I}cv?cJH z2r@QZoy%k-5~L>rwAMP=$bD0?DHWZ$zn{cI=|Eo8TUeYucsUnuvBdo8=F3ii%qSxP zlXdaVIhA2(_lafe+^e0KQse7>y`h)8kbWqbdBSn!%|qh%!1%BF;w!E|pTR)ozkndT zqE;gLD+5ZLifCQ7M>%x(9$}r5acc2gT{Pl&Y72I+X-;>#Pg>L^O_H#<3>d$2G88*+ zy>d-dVfrsgH%tm#J*Kp3n7l0$`}r{Iq^e7%lX2rM6Z$jQAlM56;o*ujmV!1R)?sPu0^+tHbwI&VRYmf$Jm_4LK=7M1hRA= zKB4q^I&pgUp^m%>mF0rW=`_e4~ zTw8=El~uaMT`MfPGwNjkn)Ff;Q~Q4Iq;MotcM^RW9V_<+&XBQsm&bWOSBI2$n_?oX z0)Wv^I{&<*H0w$aE`RNIc~zL-qYy;c-$Y{)sDk`s65{h{a`;mj+yhmw0y@>*sC-_<%K>dcpxh-M3fh3&h!jP2m$lVgRoIqWxH> z!do-ae~bRR;uC)C>LpBIHRINJW8QNQP9F(q&$NwbDYL27WO~Q2cPBA!FMQ1zkqGZ7 zd-VZMByKfK;U(J?*w+G(2abgOzotm;bT5+4UP{Nc=nO(>Ru-d#URIiRNjngji z!U#T3-$)02pkVTflp}vAdXj{S)!LD0J23S%L^6^qG{aN{Kg zsg(dG0wtRGK58$Y>Q`rAKhHwh&aS9X@MZGI_XEL&x}_3gW#GLCAM#CiK=Hr8wI1X* zZ%41?z94tJZ#3!XfYD?DJbSP}6=`nshD6nVRA-xbZNMt%aokxw4v&w&7rheS80(Y= zM;Hlyf z+N#H~B1-OyEV%?)rr^FU_qd2essf;2`pwNne1;LLM;_m)oBG!@7)l~TT7#Q^m{$K1 z6)PGa=9cRb+&0+n=Q5KaGZ<~yW%@91Rs=s_Jes}Lh&}O&UFw6~?-^cYo#AGX% zkGVgKnxi9E&`ygQff7a>4dIJ}{Dp8B58iIlwa_XTMNP)Y_tVuAM)$80MVXK*Q+8N# z`8bC-a*i+KU=0#DCiOSIOrvvP{JkV0)&3LbUlk9@YD^)Ek zrj}f1>>1q#J+)Y7-ms;MKK4;U>N0^Cd0nHI%TkvTks(kfVaOqweWpJA*I>O4q0^d^ zS^?W=e7p6!ZUR z?4Ey5Y5_Dr^tv$WHBw9+$sFIm(*AY!!#kIp_F($ki{gw#xSiN7^ex$emn!+s>pbs^ z<|<~sf%ZQXoNrcdd+&++%6lj0{{_Ssz(I`obN@Z#uzI8Nto8C@+T2OD4+!SS>+0h* z&fgernatBw*dL(>q)JU)(~S=@`cxJK(Q=isB>iY*Kf+($x~p}X*&ChJ%DL_a4_8Ww z!W^lTE_Y$38YE)5t?3e}u1fPi*Zl2DsPG{%M>P}?%rM|Aj0d#^kIhOH@MG&wFVwMqIO@8+EqgF#G&)R=(V;lGl7EGU)V>7RSCu^kd2!*qhVau**(EIU zLg3B|_}Y=GfDuN@T+q%#iyulcWc~nIQtGIMltE<}>{Ss_!Mg7)`9TScuwR++vF>T( z7B_HoosEJ=8gg7mr9ytNsx`!`;-^urxOh~)h-TFy=N@)I4%MrHSmF93j^*c1B9yeq zDXgO)Zg&GfrFnKM!y~G9sFOtYjI}mH6^O@GDXa%12#YNH5&oE!sCinG{WFK$oUF8&WiCH(JC4*T_gg zy@HT!YV41cr4c(v4B;suDSHN=wZ7K2k}^yeQf?oufURCLh5j0ZcO)0Wr@MBrIK zB{@?zT09GX`U;(JW#LudJH)zXTAeh0T<$`M>#%Yc1ruJ7V)&yzlWmxw4j96krS1#I zrlP_?12F7};90Ui+^A@xC9So9i@E}s3x>>+JAssN|UYVS&1p{>+z;~8Wn9x83((vfZ^P88XP3OPE_vO@aOiPJn z3U9=fzg3u;;D@a+2qGD7bGv_|~HDnKk33_T}fvK=v$>SJn=tp%c9TWT|f=7?*?F5F|3s ziPzBF;WxWP*g477S>K^@-aK<}QuaH}CpK?{`Uvq13#zQFOZ1f;f57yyUa9WgXU4le ztZe$?E%#e6uJ)hCmWnjK;g0xRs=GknayI3&o*5NFrzIUa3gX_=uL${ljUqOa*Ni)2 z3bP|g2xFbV<(ZyPKn3}66%fU)-2Y5S8OS31<(ImVmZn+K(nJya@=_hnK?3yAdF>Zpo1e&-wevx6uQQKQSdRu6rF(~y zYZ!?!8yRw^gGi7%BR2|k3an^6_|?)fTuCHzXWP`WSn$4M)w)i=z*TLjzkosNE=S&P zkend`M^RO`%?cU7IrNf@fD2B3f2S3KlyL$8Oi|eD9Jron=(v|K%uQ|^4wLG zpAn5WhXBc?H$JjFjTY15}m_u`Surw@Nr5xFHp%Uz9YrsKCn!cI;JzGVJx3HReY zGsNFVq4p)5DKh48$uztB4DKyNc$`*E{gW3oBtJ8{9bPx_l!wh_xZh{X^-1;fe518gz%a@|b%b4V!&n zZ!%=G3jakKsG9L(Bhl9dsslNOAWRcqzB|*?gOy9#4RO+xH=SYsI`aG;+$A*s{Ke(% z{uu#0bv3+nWT_VBZl!JhdBxZjP}ja6JjgS^Ktcw%^PkG4xVz{jP4zB)WEGm8cwt(m z)(MDMp=evi<(NLPyA|&$1yE_cHpY9X$+lVd?o4eE`js_~g?hXm$#X@hQ&PJ0QV=Zn&(;6l@%!TaW8%oGlYGu8#)#Nu>}-#COwVHJ;8$aaM3SsVgM;P|kL~20v?^ zupRFLFe#)bm2d8dYv+&e|1l&tIW-Gf((XN)t2p$SR#8pd>7A7Dq#Kbrwj}`xM@Ht( zEo(IrEA%sN5Pv;fnCuGp92N`kv>`Q>n4!IwA^9>OX}utubzz2zXL2ar#%yl~BLiS( z3B?OPF8p-b=|^*A8~o@yoX~dx1We&hzVmY=z|GkGyT$+jk;l`RO!rq@?nuLHr+**4 zZz+Ex-TxFaufxek^e`Z@`k0Z0k*Ia~Di{gnR5k{5=bmS3x_uR68A1djU#KKf0C?UPaj7|8I${@0KHhhyL;M1e_ zS*HngT(UM2EN=OSwuKs{VcL}gfkFL8 zIJ4z%4;U!}!M_N{FTMJ!h%g>$Aj4xv2N&-oWvf$z6KwM4?7R$ z^LGxacG_W2%4kVKQJ6)?csvEadl|16Kb2{3G!^{X?7gFdE1`-VvZTBcKgc9dkv-(w zyv)Vq?c7ITTeVS!xV>aDM47vm29HVECF8i`&0;MuLQ{d6C)s_G<=|%NpA>6MpipNN z1!KmQ04W?r$PNxt)mD0GQ z^iAm!h>LJ0E_p#O!G|hC2)V{={0wyt#f)x&@v&j-=Q3T{@!@@%>=T9*1f$}V_iY#Q z5wfyBnBrh$RU{QTlYl%cHyY4#Rana+06XI?3KvpHN!p*uk(6tCDDp0X!WhXIE2)Ll z@&FidP>-Ozt$z2efkm6LK~}D`W?GR3!ZyD}S_9WC9$~W2`7;x1t?-Iz%Rt#dGF6}{ zu_2PkTA2}9n^+4ki%=kEEFw-!b|U*g%>T)HYq^oU!5qamt;#)zuS+yGox9As=BSjP z3C$#49R!wbx>L;w7h2ArJ~%|97cTEUS9XJBQkZStzfa+N@! zI+?ctQl&A`D!OVVn7c6Rl)Ez{0{xY}%%5JI3g3#v$m*ckB6_jwzWz5c=T;!(snY)_ zEhIex0vBZ2pFJwndmb@$to{4xi>FtA%C>(nJ@k6|c68>Xf>te@GeqaPH~Oxr5f?gm zK<5a$OFr9k^(~G`bw3ImRjQkQ2zYP>03eRf)765vU>zhz=e~4-$B|R=O)w$qx+Wc0 z#oty;|Ki|~Z5-TwbA9>c@vl;~|FDl*8Hr(1Cs71ou3##lgzVeJQ3D-i7y(jox?a@f z*`cBSn%`-CQ9-EnU_4TC*glJ)lD{O{uO{{>X`a7;-JG|$ZbR6nf%?!rx|7^D{HWwg z$GV&N%yGMmO-@9B68F*B+lQPr9(+7qOW*<$gV0!#uyUR@(Q_VZOgKA*%_R-Jda%(c zBbGpOLqcxht1xQ@KGDXuU;!OWbSW|W!?)5WT(h|C$Q|mtn6jLX&8%|9fwq@r{T-#U zTt&K}d5t5nC)ss?HA=1)6KS+Yk5ro@Z3E&*rgjuKnf>PEHG3@j&PY&imt04-SQHFb zv*}+6@>aCY86FudMNtvVH4=w>kuhYQiqN(IaUgm#(jXxFaMFI z=%D_3KlKqWO}4rMkfS$FRaI+4F74DN^`#Au)KjuQFMlMkR{>Ij$tcIB@|*4Pgkpc4 z9mChPMMcVE>l_i^btCOWB=O}ugB8gLp>g_ziN%s!ahUt*)e70Nc5mt3l}Xy19(h^n zYpRBbBXT0WskYF-0XpGs)xZ|wTm*Fu^8jlp1NtT=cbUn_VfV17$Fw*bc2DM$J zAaZLVMmsq)E5uU=>vmfaUj1T||4!x7Z)XIJ4$2tF7JNKsV>GL}#hs-zm=ijc;?Bf=9 zr=X?0zAlS6C%=DHiydauq`w15Hd`2ETx~^Fq`$3_B7)p4S)+fREJ%(6C5@T5ld8@u ztRD(eyy}B&88_!=F@K)E{W;1E-R?+s*I%fL{uco?G`dObnC{P;yA%0taLgY5Zz-|Q zAryVi8&2aI`jG?V7e_KmIa*?wM3Z3}PB*e3t`fDhy z(*bO|ojO!&mH2w%Ee+5kxJhD5XW8FQF0SeUxS6tIDaM5}OmdIxR-x8uBv%E67$l#2*cazv(DUD27L z=M*)_bt3ZgM{O`ZR-2Ieas@C#X7$@sY66NZ9+9?&WNC(9=7UOH8k-rE%0Va}1F^6N z2;T*KLsEHN)|^e6gki*oGy=x^>9oU=}wy4QfH>j*f#<JlVfr z3f#v3^9lgW0RL@AvX>yA;z5G5eWn*$5)mq%AlZ**wge?RQT6i2n3$F)M6Q9{E_GIMqnSsU>0~K-vG>+k;18Z^O4vO(;RQ@bf5!JTZ}LD{>(cV=p^u(r$hc$ zAeZNWj;&0>JbuLc$kJVK;pOoEeG5HTj(i)vd{8hRj($+A5FJ1)tpUPM0dD`)f$G%V zoSQtMSxcWDn;7#H5AOb+qPmc0xF}q|w{*+ko%8eb;-`o9)5_mL|J3Tr z?D<}?l)ui$#27SUN$eV&Y+bCYCcrTD{Oh4#)37J@ja<@sD;vBGd?&0u6BFfx;*}K< zge%->R;F~Tdr5fuSOd;VPEDpXvOFZZIvdwjfXw7bJo=Gvwn6oRS=>DGoGxtG?6CxW zm$bt86prYuNJyI|NwTR0`v@;}KGMAR&7_bj&jBoMp+BVCFNtVf1x;XKhcvb>%|LJ| zkq>&p&}tn4?%n2ip3!cI0HYErG&E^}fhS(kpE4rm2$0{Slv0BGMAb1vrN4=UM1&(i z#jDPSIJ_l@o}~k~F<|E?z87Zu4tK$_9RiY-=Npi4XQu$H8|!8OdAQ8ccc_cZ%7{qCHi8OkxRd3Bx;#Wmxow5_^1d}^(3{=kx#1=t{OytN zL1_#S*rZV5FHqbLBzigvrPI{GY0&=!2iHpKa=Kwx(oG`FSW3HVYbm|%n4OSB8!Fl3 zx>w6bw78GKw|Wu@q>G?-g0Cq(6k<*5!fUerme z>20;6w&|zS^IdC!SfzeWb-$Mcf?3K?kva3n_Qj>|VPIBc_A%!P-OA{>W|1X_h5*e60E=?WS53rJ$1YxeBXD}m1#Y+AccwvD5?x^r=pZTYnj$(-Sv8Q)&$#6M7=kwKZ4h4&&m&o!GtfhXE!2M zl#mqwKV>`w$^e}ZHp>)?F?Yt{e~)&HIoKVqrd2+{dj;|z$z*1>^rJ;nflP>lxU2dG z37I$-b0~^-Qo~@99bOGjNTE_-V{&2crz}%`R3hM4bQ`q_rXQB)w!YkaHl)ouL!H+) z|LB&eRcltX9$uNHm=|PUGZ{}zt=(+Qa5YIFq?)zA7wJdaseE*gqHYV9AvR=K8vHBe z#1!4$-m2!!r+1e*mqMWJ%>m9CFDY54J`!&%4Kh|LA%>zX8DcKxIyw|aY3W#mXon6I z6eb1wPtETk{&-t(&51r7zn7eTn{e@Bu1M#XC>AeIv~s_(wR2>3?`;U_w;8TK_wLHS zGQlm&qig1;H`YTh63B<2-8%m!Tw(KZ(u(uY=z}KnY;i61F*wWzZ(Oz3m48J8AdI9S zZc_Pq-Sypj9}9bIL;DgOhQ{2dC5wJP{rctA_0xYgISkO94Ps`~r<#d#Wq?eh9?5O8 z(wy&)4WN5X8_Z6}f*Cd3_l3h-Q(W}lssRk$L~!D6KS2oXb;QM<%kv93153BFX*<4^ zWrh`_rmdM-pc02ijAL#%R?Y5v8WOP?}kXeTxi1 z1IW=bE?0(ew1Abe2@blTLDjOnek@y9fIxqIu7@MyGq}S2?ILnBb@6-Z++YOM0HOo- z7cS|pzis|A`jzl3XIq)%Y2(fO=Q`mCkFS8!K!`YmlVh&idlF~_N{?_Nf%rTMwXQVpo-H0iYkJoc-cZ@JjXU>yC1BmgaL zFMnb9q@+Q}zQ08IZf{n=hb+PwQ{ED%kZ9p<*o5eeosY?fY+*B50AtzZ( zm6P1^&+%RZqf0up2AEvwzw?95Ni+pyZX3$~UAOHKUSj6c6l?j!);@c%qW(SFBQ} z?u{ser$jvo7VrX|%@ww!&+21{op zw+K0plq$1{U+anWv@q0-;v|;_6VnG)3P;{Qjv`*QbyoS zd9ri&jcl+7^k-5xYDN@3{r@e(LaN+;Gn4qANqj#qUnRT`OHR>$H}MA`O;DREDg}nQ zYosW%YGCO6so#H|@8TNgo@H&P%Ju9+kx0&^Xf=sd0 z=lsi^j=TM1Xrg*XyRj(YcA2)rdD2chrxr=~+c2tR8WXwsoJzZKIgheph0etH9=#39B5!_O&By_0*Mr7%AOQ5XZX&NS&8Xw4I+n^ zQ@*}yH@}j(B&c-?E>=B`OZ#6^lu_4}H*8-ExO5`llzOE3ADYtr;b-{}lYE&!{kir# zD&eF@10|X2s4m^jbPt zeR;?Coz6_O_s|Oq6z>p~+fkLp;>Nc>z0zZ$3Yh1-Kzi)#{+-EqXM|T+5f0wQO1swfTseD))Fw^`l zMn1v&osSgG2E|s*aTTDAozR$$TS1b5DuE?#S%v2h&)xCBfIkfVCSquu&ByliK@nVr zW{sSxZsO`rO`|*>#Vp`O+vd~Hz$8iS)p8+Hn5dO&M}raC@m7Y#mJb#Dc?7 z893ngsKZUkeg){G6^v@CrnhD^ApZ2E3x`_1X6kqSa*nb4*puwvv?K}D=s<4@^nCrd%1CM@{pWP%tw$)8^hEP5!J=6iaVkh04N|C9Xm6C2E_`I`FQDY#m)JhOr8i(} zL;RzauwuAIJhY`0l&2p&vpek5i+=MKii4=J+sg1#8TL`#ekP%AZHRjnzklsuiUrm6 zTgPj&4~dtv`aEDC@rW9rem}woK&!C6}p;qQPitu zUfjlZi>bNo=ZD&R%jfXQoNky1yfFgq<0pd(xuxEIO5b4!Qd`olXB5WJAy0sxqOn%w7& z_#$1p&a}i1;2{VaL6$M~i%#eP3Bw7GvW?tTQ742U2k}#bfat8mRif01Q1L|*#%UJO z-5Ye1jqLz})*ZXpOB$g*PY7&@ewIJ>NMB{S*6u)JD^DxFry-xxW~*J|S_ zjKitkiu8%f(g>lX9(J#ngX7Vhr{m*84Hsp&K1SYY)HcFM z=sS2!IPb>?e+#$I8J`CUT~Zy@fG`?6DocE98_}j>of2^i(}!FwqjmNOqih+o;wi1K zjVt2y?Hyn6Q`htKX=NayNI+TnU83nkJa-CBik*{A9kqt1KFa@snAa6|vlTvuB2VIi z4ax71KSf<;3oG6>8`8|lvN)10B4|kMM0)PC4(vFFmwCbLwdmJ`;Zja+F4UlfF(+eUO2iO z)}I0)?4svtp+zN7s@Gy5Fn1uiVFi^A#5(HxYde)s zN1;%Un~-Nu!~K9DR(rZ{qzNjxxSjmPeqf{D5wgT1PnF4oUaBN0R=oIfm)${O_8iDW zu0g`n*XnZfeEX61rKN7E*kVpyhRw8nbO#ANEmkmWN(~a}hvYF0iJkY-T_cr&I zxmI5f3)XLy$Z$NQTa)|fHM^|^nd&7Ts9ejMd%8HnAfMGhDvBoet7n7;Nx)ZVgf-4^ z0SU62(xizF)g{80f)5FT5OzJU{==EJVf=Z?D1+78>gANpO9I)`L6Tn&-Q6pT#PF;@S?U=)uCjhG- zIFi)hvz4<;oL!_Exb2k>M$X!3FpA>FHnL3?=vly0P>2_9w`GU!_B(Wb-I><+XDRcO+h-r>uIotwHr2kKj}?SnxuH zHe5`c4#9F@t$he$WshkpO!W#{9N+10wo{Jsr9`ceir{ne(@-~!HI)PK2*`~!9?^T< zeCU4Q1T(~GbLj6?63ZNOCM=(7NYKrz-ZZ008yUx{uAV!DN zbqAL6i3-7C&gn`(P~bg0moQDaRpuqv!3kgT@bT!{t@7qqlwmzuT_XrZ|cY;d*ngrwCf3%z~@ze$_jg~CZS_sB$>gbM?+=eq#I^$jWn=N=CB;k;+F)=mO*ftD#JbzV5W(x5XMn_b!(SGTXFC4GITM@TH z!fb?lI>N@sc%&2*X*j`EesCrtvSB>TCP4BmZ{P`x16f&G#e*>>Q6cXUR2g zYU#1mv`Q3ZwP^%{tnTCkmp?9++9+hlxYq_zL z06#2+Lvn+0+^6D_xn%lYrtNVb`afV!`rqX*mrmX65cAm2>xV(RMb>*SQXGFIqqWJ{ z_HM(uQxx*qj@yE_zK_1t^ZmdOGiAtN5FyPJ9aA}Fd+`P*oB3_$GU@mqE;UWfbsryP1F5tmtKbn(IgtBWhRE=yYZTL4zm#lzdX*z%`gTpqtYNC1^emx|OK z+jJm0E`?J;+~bM(-BF$u`gTnO!PdU|uSUO#NA(JC+=+L|P^XEdpET|CZ=b%fJSujF(&5tenQ; zZ0dqGM?1Hi!D**zs;HI(UU#bfQpUITl^$_jlC@=z6~J5H;yzVw+-gJX3E)US;i7>7 ziuwtrwZ~FO4_FWoBJSN7w2=o<%5*J^@E5)-=CM@gJ>2HHo83H$G+s~(i~^!c>PTsC z(dhm@PqMupBl}K5M8O>(INtQo@2esAH7#CQuK+oZ3{HZLY<}XFeA%OdUj?b^^1De$ z5H4h}ptqexfzN5tc`%P5CJapdBNCkHuVq}@$ag=A42~00$aRk#UUIOokeoA>Y-0zT z|It$q$q&IQtit*|em%4JPV7X-Nz2?jSGPF$%~v#GZ`hwvHhESlS!UCc&znY-a`m3+X`r3Zl@u5`}>6NSw7NTxyLRk=9FdIbjwl ztCt5KjKhN}Pc_RxG$8$mg_d!_&+Nyt=FuZYHzZ4O3Qmp5ELF#dp7jr?9$ledaU+0) z%XCJ!cz{Ac&mF5xk52l4uI4S(?@k|X9Dpu6o_w!vh};}_L?h}Rmw5N%xu2 ztva#;TT?^d{!y_M2Ou)}L;;}N>yXs0LCgH zU2f;Kur4a(L4BOfMb^yn&6u+5>`^0L&06qkVfZr}lRMKzMG!60a<~Vwy{tdzimQE; zf301I0jKK>^jqdFsH2(Ym08*_h3RT}V`?G-vqI10@e8-zHR`akt4@000Q zCjpz!rZzBTo)+0hGuA_#Rb9Fw8aux65<`m>JWHKnfRv?d<^wATIjB`(8jw|(L3>tk zH)zPkkG6K-k4{TK>Qms{aY-44m0wdAVAX4*taoDQ245;E=?)>lWgn?N+;uD_qXkvg z?Wp~@2|4RRMl#Aup45=_l?&BH(fbHATo#Bij*Qq8Gq@z)l@`ABEn%*~>_3wz)O(OP z(2akN`SC{;|7eCK=GzN!99?A6c+T_GgO-$OHr0Da>zC&*q&??un6IptcPB!-NOvW; zNGn{yR}aq~{C}t*{u7eP4pmk~_4w?ksMPH9T8Z;9a@P)4G>QI{=h_aL;4$XUCHy5%i%ci&5jzTLM2Z%3C8?g4^x-zLxQ&Z{mCy>y5z z5K~ls_p~>ByeNO_;(Ijc#RJJtv3~)8M)U+DG}@E);ZYY3ne+oQ5WFb(j!_B}OL`t& zNcKl_3w-38aSJ-+=32ZCr7pY4c`kpSufq`9_QwbHl4FMW4mtA|FdzO?D1&28WW+rZ zVap~d1G8*$4C2{Jf!i|Wter_&DJlhAY4K&T{%FIoo22t@BLRGEu^wLN^H-vSf_H?b zWXKH|Q77J8m^Zt5fFsrMD}KW4J;V8un|)JCl!m~3MyYkAJr8b>NXdC3B@kb3>xqA~ zz59>`boiM*;87_9ryq5n=}^$4BT6XN9dPmNX%na#@BMK?i85kO*KR+y-d)9j=V)=c zh!*5e=O{QIPnmi-+0E~%M5Olkp=Lb@`5`BQ0&*_#K<__Tmqwm+V2KIzgr04eiVI}Y zWV=i04>ZR1hE-&l36X~HerKLnP6U4))1AU1^Bv^+9J#>R!7|uMVAf)RiUJ&{en1^gjG5UHdCHFBCs^} zov@x}PbfDte4>UQ`l9pEzA~gn}&X$9UBY;}?T=tY9RMQ0>R(a0LZFtM0~r zEV>XeYASN5cFzSehnjU}@=et+@E|W;x!h!!tvrcoNCXbgZttChv7J0YG=>xO!UA@rR5W#kxCLb(e{ zwEE?fFNPt!b_0m6P?)j@xp+~~<6@^+!XZlaD$0yGbyRI_&6jS3-qIn5-q(b?`T!?J z7X=6aTm>84inTtEkKJaTCe`_5H}BcVGWU!@=gqa))C&jWJ1AlJH)dn-jhsdgGN;F~}@-R;iT7AE59e#w9R{-y}vDdr#d?3il6uN?> zBJ8%BOC~pI;a0OGD#hb5NHO~4F~{V=K?5e~g=?kE_Y)Yv`HOP_S;tYNaw2I$5G(_a z4_uP==zdABthxXpcv_!XD?fg*v0Ed*4eOyIa+dC7Xb)2>99oLtoKeh4Mkr=}&0RdVb0M60*6)rW4av^-Dk ze@}5nwq(&P3NiQK=A%n^BjJF)Xt=BpXBqo9)-ha09{9oPZKo0_py)N69{xM z6W9I;PFovU8_tReP}e}rBGExQzZqIJ7!ffY({O6pwv~_KqUCzyZp!LIPc=AVD!PQnAK0uniN)^f}Y^% z2TY!F@~uBUO>R*TpcCeZcR*G<&3ombJ^M_OWeFKJR$iDKg!E~2VUC`>i(z-#$pOHf za#(&rmL_4r(fQ!aCrCcu@v_>od|Uf%-gzy31ysZ{jsl~M`8e&IRGF64$8gmYKbcp= zSJr1-P9*C#q6dp)HAJf}*FNKW@UYL@-v!Av-NziPAWJ-}=ATvw4$roqy+{ zCFdg?efB{A-)g;ZrJ`xZGwqsk2j#r2P7_b?pNrgpa}F(*2Z=o*_ZNK16jR9fc;crU zCcH=^o^(NKrb3>p8gB@Po^;md0L8_rYj>#Vvym}|6;3fZQd~6I0OdjmH!=0u)ft^8 zCAc$)AdyR}MwByGw>)1m&Zc7O!zw&~O!CxBlQxI8%wHrwr~QXuML!Q6?>z0mud*Z4 z5W7GyC!?0SUJ7MEY(x1^^8tsl4gpdb*H!OYd+4%+2NgjR$IAXtT1at3R1O53M`oKu zX@>=TLh$wvBlHc~*~kLPGeQrt5#kUHIyO)v-gh#!G`!!!i6JoF#82EG(UZ~OVHzjc zG03`*;)q}jOQ%krO5YZif@=_hvziehL)2`vgbQn*eD+#X>FwbQVcNZh6p;iH#1>7E z6U%@cpo*THZ2q9Z=0%>mzAY`%ELZlCA0j4>H}11XVAzOXHEzvdoMB zqA-Qk)%vg2P1i>IE)4cZ7c1Ymr10k769+%uZYO@b-t|+JmBXMmw0ogTWO@wF{=5uy z|EgEHmLh`1+D1sy5^Dp6$B5Ke;69lr7F>pIk3IVG5l*z#MUo_lkR2Iphyc#o#JB$}r^_2w zB{!mVj|rhO`PMd>C0 zh!6X`r?*G#wBrXTQe9#FZ7q{thxPgELwm+A+_#(m9LsqQtF;CXEGEzHgj ziNn+W#D|o1zvfwE1M(f2+>Cfwo`x4?L?}y!ztWdkuPt^BZGk~&w$Q!QmWcS)5?xNy zRxO;bwj~-1?GC9YjSc=Tfriv6xmlvK*P>?t(0twl*Sr~N{HV!{m5*eBDu8y4m+`%! z93`>M>y)+xtrF+T^gwbTi5z0J*x#r0J0$;5Trt&|2_NC8QD-@`;-Yj%*OM76rk+PV za+u@o40_V7=w%56D##h*_jWdfm#oHhhHCm+q>T*&3nIUC5y53fX^zB`pKamHW?^-d zRj2~cV&gY)bi zYsn2~ECXPPX)t^f_1#A-&``30SEd+`Om2z~r>{O9M;TFwPGi{C`k1FC9#%?_lcemG zF%(2d4Rkt_!7WT`^HdZQr~+hD@XriLEz~T?%I{Hnv3L0)hy^xjaT5ufNx7!eqX9O7 zj-@)Ll2$o-%q{4N34!Qlzzdt5K^w(!$!_jURQfZl@nL}8nq+`-hWsU8 zV84Odks+h?RVFmLeEzMa(Qe8hdS){L z&=I7tj(uZUZB)<`m)WJIW$w$+;Ql$2&L;?`8*H49`fVlv?pbaY6o?=qhf9?RQ zMPz5}htlR-&2!@v^ZZjje_#JD3v?(6aO~nHy7p=l*Q#4i=7ks-dq`=>QdW@VokdXD zfX3uWQg22y{c59Ss3KyP2H{B@0ls_VBp#5e3#t+r-Sfh=r;7>nt0%zY(Gq9G@>+Lx za!6psg)vU4OwXMc5Ky}Oz5IlO*8rSo099u*zlz&OlO}i8x>6=WC&O!FN$w2g#|25_ zz*!RJghjjRukvF(;;OnDJGJhjtqK4QS7yrWK!9ONRxE36WW*vqFp1Ck`$=e=bJthL z$xj6ORYf_zIr0K20Ty^y9rFvTHIr|K%)!0w9=2R@RK~#$?$S#7dAi# z9p-y~ZjRLF4nX;21-69)qeTtYdS45S(mJ@G7FpuaCZKwh!GwpWVWDs_SoU5GSr$>( z50vxWCFshjq<{DSu=mzMaeaHfXbTOErg0~YHx5C91#L9A1SfcK3mQn$xVr{-cXxL} zfDl}Q26sq+gyix&b7tN-bEfXS^{VF8{PDWFdad58x_a+jd)fESWMVorjWjzWu(Ic0wT%`cG?UW4s!r7EqreY z?C+oGeQBvN8z==0w<~UKopTqdZMefWc)04CpMm0*Ita`?IasiMI^^V;<&=gzDD?C@fc4>=mY(F2echBt(Wmepm<3pd~73tC(=|QN>7W_%G8BYj5Aj( zkdompFQPO$ojK$I6vYz}vdsX}Ms0QlIS2q$!-iyT(5|41(l$nCw$9r@{1o%1)gjj*Z&%;~6HAlKUAK+tk^m;c|}z zMtbtu8?ZS(^)FU`uP8el;mT~blb2UvPud0E3LL4}?~i9(aY=UVLj7 z3kPb4aD{Lo&WXPa3T+JIk_3ZFS?6uP68f{k9xO-G&@}LN-$(sx0{ahO31RiQ?c?7mTOBO;_G!8FDBwy_7>5Msq63mBVnraqmue@r67ZZ}Si$YT zXx+b~LLu{KNXxu@%pWATZ(h6ZzT6t9-2EkfKm6qpQ^*F^YtoO2z`xXR25o^vG62RVUGkEk>D?urF zTEGwenzHwRj@4PO!E#sFKm3S@Ak;O*ieEHWR=Bn3I009;!rGN8cZEsOHRBx_YlHGt z7yjnl`FhR+RN`vn!CHeAO>$&lxTuZ)080%k5w#-NJ$G}ei;cD4A1Cagrv27JP)bBB z^x~O$54HP|bO|fLo)R7{FjrtYWbesGE)Uu=_9hTdq+FG9qXn(&jh#-t#vyihm`)Ab zSP8IEqQ5y@H*Bm>o=a)LL*631O!SS&Oyt;RTosazDI0@lK{*$y?MI>sXl2J-P^A9o zxa^V@46o`k%L%JYjp->5RVUzw;`FS}pintFCe!n1^Nj%i&ucr=L-=3^ujD0$z%aLO zs9eDjxvD9!E$W5&8;&x21TBBAVoD#pI|&Vrz@6@L!O00)YE6DXpgv6)7x*A041SfN z9Q2S9NN$r-+|$u*JebRs$4qhC@6eXB!UVA^gWP!^Gb569rURNPD`eCH8+K21jmA9tp+#(usav?@07<)7sb&Ub2kGUrqh<~j=em!_WAu}8 zyp=d73{%)Y+j?KtZD>fJxlqT*MT!-e%hp8HiR`1)5dr@0378HRgTBmgA^a1s&heyE zWqpQL<9ts|!S5EV10-s^vU(}90T(+)S&xJMtB(Dz;HR4! zZuRZa;EDIt8F`)9J7Jm8x&o}g(9Dm#W?#1&=mzW&f#$!4>iI9|VKO=Nz zg*|W6{YQ9WwhQ`KqlYWKg>RojW`i#_&h9*QTQ?Y<#49{ZGar$7{E%R;{GARJCpZ~R zBJU*qtqK2TCMRBH1;Eqi!7xf)(B%~w6E$k)_R|XZ5zmqL#uY#!_I%6)h)*3hPX#7C zgqd&u;{B+=vLIUH<}u1u{>_BHSk?X|3|l&fwkozqJz9cJ>(~vq)#`mIr%5&Ghfs6N zr(^n$l0mqH*PHD$a{fEHO0ICySJdY&DlGXeloBQ_V?VzMFRch;Pk+VcZJtM`jkNF# z7_y}N9Ll7{HpEIxxmi=!uz#GT*^p5nn;D{Hi$TBu2hcTRyr)q3yXf5GgpqE^@D=K4 zS9Od&RJ@yAYa^c3%|JP0Xmf~!;9Bcw`k>=fUbm8qCk55hTBZ|-vQ&+$I;Y{uGq8-~ zLm=75OYjQD#olAn^s9C+3hg`MK(J$z*E%6QU&8bCC*uNclh>m;lew&mgS(g*O88H8 zU!XT9PTUcjyFPLrG93A&7^|!G?G?Fbjlf4FX(<02JB-9k7#{*nt8V=lw#I@s=3M+1 z+?Kw{J4OE``?J2J<`dx8R4Jj}R#^)~D8mX{V)I?ucAI9Yd)S3-c!bwR86sorn;$I} zq)~vNn2ec$fJ|Wniz0$Y_azNx+jg8z*6g?KyRunQ2ODkC+?M~hs`OZ_8IUX!N{N3xH zNKGcq-cd!A>?uLpa}+iIjStY4W!EHj@ zr}1|cz9eU{Matx3Q%Z$gXs{zQ>*6e2|1P`O!)I84 zB!49B4vCs{`1v_O;0l??I{X0gbCGav713j-S*9l1nx8x(t8`+lZ1iR;N<1~hXtDS7 z4Yy@7osvYEJccT6Or7N@3np9mSNg{R5e*Adjw0C9TInXay%>zO0_hLJ7?@Rn2OWg( zh7;!~rC)oAu7Y1X;{?qu*B;3WbUk^`c`8L136?87tNxyBTuxg%m$f)39WUP!s`Z7m z=%t>O@3BTBYvNE(> z5~W2-Z{PwRIvNuhozCMbHRccgl^+8OPG(L8@#^YiW+{S2!`Dwx$FZtf#1a%SgB_R> zb|Y>qGCV}n$%W59TL&j$ny!{8&J)2dCfT{yMM~hG>PaJ)po`D7s8ReyZ~QWB2~6hb z2#J<`TjHai;9a5W(1|!*qOtX(P9qNqFw>h(D$(rmi{|dk$l2duTW;@x4R} zbcKxq*amA6NXn3TV(d~RdE>!n>kWUg!HnRSO?V&GAF}ljCwgs6h{J3ZhVdlKNQ%TS zw60Lm6JW6Z`k2`jEuz=9h=El$gz#FwgEFFI>C+EC-A~wRSQ~|V9-~j_aIxK<%Z)>{ zbhvb7%BB}e^*_3)X^@*0$IYW}clY#8UF0~R%U5Tj&il9$8HyfX53iVE0iQRm5ORB4 zz$=<9sB6PoMg$P{?a_5&QR^;#`Hx(+!;KamZ%{kS^maeb^A4zN@5cJq==y)<({i@8 zA(>wg>)M2VbB3=@zNLa&P&C@8{Ehx|=ojpQ!^a;5KF9=I%Wv@sWZ6uGn2N{}iz5b^ zty37bP$T^PL(plZV7~^oAKW{L0D3KA?$gMTu1R;^HU{!S*Y$Xd_+G%?;5~(%HVD)> z3u_6fKEI~Glneymy6;5rgieLEM59%3K?1|!pKrWlt9aQ{Ndf_l1-|Em@f*3L#zxIp zx6<1AAM7G{y?u~ZC|aE&UdkCH{y8j4gjm6tR{(==X&J*TQ>r^1&x3UC23$MK{nq%fSZeq zxIzsMWZu;5bksdH00A|O!KP1?S-=Tc)HP?yP1;VH1ullBJ%lfOSC796dh&meD)1%Y zNA{k$&4}U?N;d4c__X)x^`c_P$K~Fhx{}4_N=Eks#)5J7b=CNU+8pr*Tr6?dG0YRl z&SYiqaX{^U{w2>HD~x(*)o@L$yh{OZ=%QFC1T8e*n4U?XCj;Gs_xH;f>eyjNHN(^k zU|^%J7jp5#?}DePS7EE$6Wv@~I~mA3L!Vb)d|fC%1pt!dO3zt+t4wIjgheyH(MLFI zFR)$>wAh8_2I{aWles@QW+BmUDq(5Spqbr zQ#wqs5Ifcvra$PG91@8&!nwSOr-R7DQYdV|Fg%;G5wl&o_tw#C;YG_iz|7pD7yeA& zGc5BJ?>Oj1I402@Gh?0^7THt&8X=3qal~g}a>pFh=ZwePxdS{ZdaC@|o^}sSbqgG> zK}F)thhbJI+O+7G$R#~-uo*5YF_$z6LTSy>>}k$p(4pYOqi21BQTVLXG_TMTS9sKHrwl$GUmEfRaD@~;9DqySe$_@ke#1Vhb`axfmt}0G}q!kNMZ4Z&I z&5J57nl_=#Aa{t}IUAkn{|fRh=kjR%5=Gh|oSr znje>jhrfbYOcpZ;z}@kRH}N~)j(lFR(JNRhX8%KIgogNFMDqC;38F;l@q`R7b?+PE zp196}gFP0cPYdw?C@*nTh5(~S1}h9@gGNUbd>s(4paX&@QI>n_YrP_l=B^+i&iMA* zpuj=b$-|)?rHSCcvE6QGbPD+^H50csv}quTZmjKWcsN&Uiaf&xXQN3)1&~Z73m4(2 zEd%r3sa7Y;gz%x`^IC7GFLDYtGemZg)IDZjCj1j^yFR?8 z<}u(vvvsL_+5zt^RQvK~?+H%%)s{7sutLNvtHYttVIdd*AXCQ5fl$z+$tg!)CI~j( z&qMvYI}lacW~TfB4gTR|zmvvTcJKg{4kM1bY95f_nJVuN2WqXj?QpXz&VVWR6Q~={ zn2Di4o1Z`+LV7=ZEUC;yYYzk<8N=23QuPVrrENm|&#Ljd@eW~{VMlGlbXjSK;CzmAMmSj{``SSmEm4EdIU{8{_tqmW> zT~pS@m%x0(7*@*TQa%OYPG_OSi?k?Q$D!=4R1r2Y`3%eVyqrV(kzdT#_7H6WLKxTs z9S9u`(>QJWRn>L?4>z6o+0t#Px;I5+$Hoq{po~I$t)M&e4uDR;002C>uaMAV2=8DW zXlY?xU?JGxjC1$QR(%Yh%cWz*#{J^v1Tf?3w)Nuye#(#sQ4-MfhjFy`D=;YFm;=#( zn!24l->nA_x9AEn2!?7R7!OoAU$0-D!ylcql5d(wtI-V0K+@7WlY89J4+S^jE4%I` zBG-%MOU^^kKTi@Gn?vi9V0C)?ig2X)6PA`XmTkZMWG)=0(``#@#MZ4Nue@ooN z9g2PI`YOIoTPZhI;SQk=#dK?7fm0o&4GMyxqZ8rrMwcyeRvb1Xz*qbFwF?<+U{Gaq zc$0wN4v*ug2L{@(C`Myv4}N?I)fT>~X&&@Z$x8gNWA2o1m(lG&fxo;0ae-SW=Yw5f zI@Lff`fbf;rYRNz48a;{4Vr(q!STnW{_xfK&T25^&V_)ES#D$c>6f~zyG+$!>h>v~ zXyyXO&4%-BKOrCtHkJDBhQb9uRDec*gXO-4%!qO!gBi#7P7rV$V9xj1{P#j|8n|bk z5WdzQ?+zKo2j}4`AAcfw6Dk3z_psD6Ih~kbNGn7O)E~OXW#S>`93>vRGt2&~DNI86bxg=x8P^s92*r|321N!xokR zDA&Jeci^nacA0ymc9^uX+G2pa4ZvIwfG`yqtx0y$hF1%+DaS#!2klivV3t!~Cgc3N z0qkJ3*En-&1h~dO##)$G5I}i9DDc3r1{XYSE_pFOP9SNT1f#->+~<9t%P6I1-0CDC{?6>PHkeubBPRm2@1&I{ zXq~?ycxqG7%*eYg)s^mRWbA*TIDlo~Efhk(cyX&37!(B5DDNHLI-sKU;b5HqCoO`% z;k*)J+3K8#bqvX7R6P2d7P2f?*0|u}?*Ur4Ri61Nhnp9I?(e=Fw3Gm42#+&>S5JTo z*}}!Ia=&s-<}PnMFvQqWAP1h_w-$ewo#=C8o~+W;`hFNj_6z-m(4t&+>~wqg@Jq^9 z!Px|&;M2cA52BnCG0eVC5`A%#2clU;V}j@Z03KYSEa+a0cfK4HkDO&4pMHGwp!4&O z_vR+2S`;Oug^kbtSyYJIXU4jr6sYx;-0;njRXmFkaQeXXP5U-YQ^@Q5-XnZ8UTe3q zWu*$k%;YdWK-nIS&ttgkjP%nNL0evNw_~mtSS;2txDzr#QZ=h92H|$Gm!$ZEtCGql zX8~g}d;r!~K7MPmL5PHXaHhT?m+riMhxdBLFNqRuVuxw3HMa2SUq4rX7A}qI<#5{8 zaH&qD0rxD98z=ac@OV1ybYO6J-B><@aV7;yVhJpN7L5(G?$0d9s3d&5pcwv=2nV|a zbG}yx$kERIf*!Q>O2X21`@2RhhN)=FQ+e{3y6z zkM-k2f}T(c;izvv1u*+Ri%|7xLo^`j5x5K5AyO?M7tvTE%mImT`YjI&wVbG6f-(rU zno*|c6pBcV&WO)Uaq@@kcFZG05k>a{XaRkHPAKY=&-uW>5&+aCt60QnI(Yeu=wI=o;n}WV%w0((LiBTN z3UVE^r-YHcvK0&0ln7S07S&Q1BczzSsS>B>9k;&%4h1_8t4Y5Qv80HpNtxcyuwont zEx&TAT-qT?5QXH3u;$_VS6IdXnTeWi0;Wq8`{p0O6!+mER!cd?#X6Gc72b2~dL1Yi zCSA=3Iel(5{WburGsbz8V$s~Ujb($wC}Ff{98ntC2QL8Omb`T78?=%rMy9h547g~R zGO(%V+|6M_H^?H5OVoNXf&4RnT2@E1w=9ZXj&c^B#RU-1ML?B3&l}f5Vu}1%O71yU zF&vFVgr+uj$cyogapjUET#-4~q3Qbda+BEYr(qHS!3=K7Z1+Q8@^B~Ia|b9v(GUN) z$$XQ+g|M8NPC?jh39;~9p7LTqEFMSVR4AJ>c5kfe!MlK-;58PY+n`dWJ~wFsK_P=) zj$dex8LySN^O&Gd9qE3|Ae|ag#VL>nOyUsf#b`zS1+~4qUa5dq} z-{4E1zVpJd`Q(`&a`*?J%MC?(xqxlvi^cOI=Kje!b~Atp-=Fy;ioPME;z;{VU;vUN z7)zNg-YwBuln8fS2L&oI-`hXlk|32cIHsCAFbIRyBiR=_-v)aN6>3Mt?)f`Q$^fs1 zfy4TOYeO`u!1k`9ACwpS!Ld8=2me5IHOZifWC6f6ElnN`PiDe2QrRU@nAjElJ2^H7 zlGdg-bq(F-x`s?FN5&e5n$@?>e`cILg%=1zUh-5b;U=nD>g)}!36*GD6CdXuZ0

w370P4E1vc?0R;7(4bjH>2xDkQARR7K0bNZz@m6A^Y6F!XE63hi^35DgZx7)p8S zM|8B7O6jgLXb-~0k0y(E6x^StVKmMXu(sX zIOFM1x@Bs*{bYE9!)Dv^{4XL0xJTmJ*2^Z0PwZjzHDZHc&k4n z?kSE;Z-c5D-~ul-NHT#}4l1)^UQEN5_dsRBI<9NvgfH;GsY&@4HD%SyTflxKZV6E) z64zB$j8e?Ww#lSEU5uL6)q>4<6ir#S)xSai8nGBY$fbh0iVi5HUU?h=Vk!K~!A(Lo zp))YIsJL&NbR(lm77giY-YB!K$6&2^uoXmi=R(jZn0JA}c~t_5qrI55AmWzVvm07< z1eC9mO!Ca~O}uf@DyG32Tudy`>gns!#))jUVy=;Y@CC+G7qKtf3CHW@rlO05pXd3> zuyrtPeazy?=;af}Cw=L*%hq#!P=TaOe|KXqX(Q$IR8=U!+g&m+$cUQqwZXeudfyj;24${NL%feCe ze8%FD;E%LjiiyJRdIeogmhtj+1hIF2O9>|IXY4~0~YxV`&SMpRPSBkqi{P$ZVbc;*gH3%Fw5|AgM$Jk zKzcdTrIx>r;Q~l#^5p*bAex}Ic}&s5Am{L2#i^wKiR5})6ytG$PvkGar_nnOvpOGVjUk zm;L}$gG(1^NNFZvSTzDItM$ES+|h=aXGlWxGsO0T5Fm@rx>TFDvs(WaP>p)!v^fDX za<3Po)YbA`^*s2CD7yTa4yQ-DXmtYB2vv4DBu6A>XdQ9oQhl$v?kb-lXjDJCVFI3n zxmD+8DqNGH=Za*(hDE!wz!b02tbSuA%$$Zsy4`4}8>ZqS*5OBPTkSJ3#er|s-)H|6 z`SD!tBEoa(^}q9c$-j#SoJ&@?$V3^>E;Sy$HoAs7vP(#<&y!U=V;b@+)kUxg4T>;1 zW64T|PFH%3$+^P!Q_-yFKCGe(`cd{dH`vKCtlQRr8Ue(D+|ERe#&bm0`xPqoRUcp5 ztQAx968)G!QXgRYrgF#7y1Av~KCSnz`HpjtKAx8@Y^DGBS#gTqy--aczMYA{Q>oQU zR2&@uA{&V1+;?U|>6*vfTi&j0hb6fO!DPV#9UMMFm2k(Ntd1)x zm1F!=7b@2JB9x=IR+;>2+$({`6+@e%{lnb$SA0S>W&SM=S+3-X+I@PDNEqs+viP2f zP=h#iPQ^$2d z)o&1pn3ZirsTrW)Q!#RG+Kqf5vA&8jBWH``!FA+&t04xZr)m&hp?)b83Nf+JYF?6- z;$tB${-b9PB zkQ0Wl%jEm6EEf})5hzdU&%x8;_5j~j)7NyV64!~sfwGY-Qm%6w{+Ri=z`O-z7U}At zQc@&qh-vb6jepfd7(H5S&rD4?-Ohj@9uM=LM0`X6pBqAvBK{Jkq=U(5-VD$i@UDxo+@s7HH%QDW&Djo7qiB1@OPZccOT0*}} z+8KS5YNy)T&DbeYrcx-`CL95}^l($c0l^RmMadWH@s_2=-VEws5PNKoDSBgs5)s`0 z3$3I)-kBp&qZdIB)D0YEqAXMrs!}qNi`IJ^+NU3gt{+5Z(my$b!p%kI<64Y84U!J( ztxMZN14HrK?PS&zlU_omDcjIMCj_ib&EVgbIi>y*m{O+rizRSwa4l+O02*OU7rL?Y z3PHXuh-r{ERP^<;rO!k1p6um2ljjP~BAo_3RJar2Vv_mo0lfOSe$=?wHS+=gf`q(% znjX>jc1zSNeX{oxZ2?^tM{`k!dI64FgPVvi)(ys20KjkRrZi%kS3~-~rBdeGFSoxZ z{{^E9J$a2VCNrjA(r>;__xpBqgVUfyuKeneQit7^;ckkYi1j5(rj|bI1#jQT@mZ0I zd*(SkwHS%jydZj_b963YuWIRU-w<(L$z&`v8I-3kDz zjNt!Vz|Kb0J}<%*95a6z4!>!?pVWEZdT{#VnbGgv^6r&u&!1UU4iklw2{)iMZ4rYDZNq_fK{%{? znZx}D;hQ}82faCkaR0nh%(V>a6)WWzT}0TjG1aJ=17Lem?uw{5+g5A$;tl!Pqb>m8 zK`YTN+vsH`kTILPYz0nnKut$?RBdXF(MC4=qJ(5g~q;OGItMr z5SsHGc$iwA8DK}$$6%lMy_1L=J&LQw^=palg5XW;_(o6AVxmH&*aJ)F3N;P4?ESt= z;rrAD$!}U>_7qgr6VeK~L`gqZ2$Z1)l;-S$lyEb{(`DQwDA2B+yDP%<k;aa@}csw!pC$OOr~AYrMftsK*ZP<3fq ziqKzAiK4E1F0g*L3T)?_fNy4=VRs2(*N8nBQgAP9f1RF*#KT*X$K^?R=Jfkj+ajE=05=5xd$VgH(!5E8vFr#IDCKkKe(A_DY=4&T`Nm2g4FNg+6ANV;mPs6z1*G5 z7W=;_LRz%L4dznfdKpVt7R2udbr+n{tJj1SGeqw!r{=cCt5mQLC63^OWE)*t_0B-r zti9njo|8zQe47963SQHXCah~xqN>ejE3VPXckYua$LB=XDz^uj4+IpRL7sW90}OfF za$PgMHkcQEo_L_y@5jMg3yjiQ7#^1+JM$_(hTp&skaN{!om|#I{^8C^(&QzBdJ^On zxM&x&G*u1k>ynb0|06UxsZTu2gLZCNK1LW!H6D3JZ3pC^UlH|xIp4?S zKyQ6*;b3^K-{jmd<${o`J(7aQsjFqe~`4n z!Gt|>kitO7X6`5K$=(Jg6uEU^hsGz4I$~m?O2@s6yH?0V!q=U!=}$660BDfyJ;7)4z+D^Y9vaDF@wPzBK2(LXJ_%0h(QEG?xCb8dzUlqRh-ha9FM(0OX z(*5rL|Kz_4 zAHBPrY4??RrrGtkNHPxS=pyu8xYKU$WbV#@OM+C88-R6>lmo+&>OCj8$CaSu{ILlP zD+p7MsJaIQ6ikJrRb6~?-9-v;03ns1iBVOppBx)?h5-V|AK8z!;hxU<7O5V4uQByH z$eEuzodVE*Jn;tfHvZ66DrzCgd1h{pH)2?;I8b(BQCP*}GK(>FY0jij5NzK7SHs0A z0`fT@BNFQ_tXNknO6udWaqJ|wbQGKWk7r*TM0b>%QWv3K&xF8tH}|nOx-NqlLBeu-+-g6TVqkt}3-(a?v1?M5H|r`5d>GsG+7U zMedoF^719!@$1zA=$ZDmm2s{$af4E5fTy8`F9oGMv-PK0RSZ~p1)cc#>W!!7TOZ!+ z4+?+xe?31a+&j84bP>!?J$Ipy-2nG_u~3-mgX@b44}SAd2dp~Z+wb^3kX=7HJ4t`+ zy?#7r5Bec~VpOCHLR}iS;VteN{UV54j44^w`stFj{jPo`X3i(>u6N-`L!!Ry(eAUI z7;p16-5<&S4V94gU(g)8K)KsVS8=@)xx;58_pe$XSNx#A{(ny7p-I={hoC!Z=ff$x z&p6G+a7XQbLxpdfDw4JepX%TKyEbBb@7}^o_|wbX;SB3|*V6A>kN*JV|I4%c-<RP5pbg*Zu71?hjyIZN)P^ z5d;v_LM4y`H0kA81Iy1`X6d7}FMz%1QN2lz~OeD)LN zCO;e`{0DF{eRc9MvS^)Vl`e}7Ke%#p95Yg&QZhtr9jtv&BEuWbdWM>B+LOaJ>V&IN z!#Ef!Yrs!AY~V6|Yn^c+n#(gGEWz?>#R( zR0(k-o|B3rdfSNJLXcun*s@r&CxBg>Tyd-NOAT- zqoWavZU=6v%L|bR;_zS;eKhLs-3jUX|CATx!5VcFm8geUBObTU*@T10^3A z7tBOUc9bI#Yj21_kcRlQ#~Ha5vB?5m$wnWZ+&ZHu0R7F)TTEy9!w;?iGjZLczbwZX zW#}G+AG@9WKb|~>=#AF@F5(Fv&-F(|bqeobF#kH@!zSK0Yd}*I7E8i`n)-cBw1CTj z%}eU}Sd7VxSJ4v*@OAq4R_T=mqsk+19I{w5z>=~I_EY4-(mLcEK4j;t=WIn;6IG-k zpc1ZXK0Ytm^-w`0c?k#o;_<6b3kFY%6)IaAK5bl3+&Nk4FAj~TxBRXtX?Pw4EKm+h zxHXx~{f)C>fn~A4A6INsLE&N9dzG#mhGER?@cp1KhgkZy|3+q7tCTg2r4lqap$%yx zlApr1D)vLzc5CYy@Ochu;Nyshr!nI&(1aMU#v#()m;y{XI5BOGMR7!mX+9^*)bbr{ zvlzJ=aeMf6peH^}>V+ByuF=|a45)~((EPyJgQB8edg0N_4YarF17-XJgW7yd3}^Am zIT@uT%ctt`>X;Lu}*gccH!dy&wNA-%Y&!OM1K<&=cSagn)GFZ)qMP$o6Luv#9T z3$}?6Oxy<>6ybx&MFbjKbMee$!o;WuXr^VNFZ9BnkCEUrpMPq15)DXXRmnK`qck@v zw0%#7K=fefNBE@)-7~5++df9l^r{0yyqqi1^_4urzdYxI?+WthWGRa>N`|Igi-)wC{yA`FA+c6y^IlL+!U5Qw1Zfc2;vJkQV1W z)XE!Z=meAj%k*_W)>J>j&+s1{r9*cP8^d&Ku3rsf&r{$~4+7GXm zw*pq*Wb=^e(SHEHkJH35mp54sGB0OEUM>70M`V~x$?cm-Z)^61WU((R7Ghebg&z(b zi3*oYWKvaGqe%#B*qP;oGJbwBX7E0dV)+f0BQ#Nkr?CweR*#EAMw%q=nk4W#gw`2- zk${k1)~z?MH|+v}3_f4a&`2pL2}N&<)NT6eIwc&5Y%`bOPC_N{luY?27_{5KOz+a@ zF4p`LHB25e&-j!tWrWzLI$;v1?_~gcqttkCwV={(WTO)rp=xan#trK~Q2d+mdsz@;>sSe^?8*%$$P|3D_tyn)84S2AM$>W^Nri$I))#9i>06Ep)?~6P}k9NOE ziUm?@SnWO!Y%=}>po%iCGms>02oA?(92Fw>2B+|3q`JAhH>_}Nh5D@wd%gG-lt;ZB z-cg0Yzdlk*@VP1IFe~g$1{UoNffgNnzgL`Qc<;HbQHI!lBJEBO#!MgiHZIp&!N1Mfd9g8F z>A<=lm&3`Yp~j+e5qeEOI?YI_-9W!l zhExE?8%0o_IzvyMx?qlppU&IMm#fRAc#U9ik$?m!OdHd{hPzp2akIfL7y#pWTu@ZV zboc_&@+p#^t`9vZ_;`&jQ-;Z>mayh@17Q?V9$lq8SXt@T{cjx_U1F=JLK4bK>bA|C zE-yu;o`x1GIkeU?n?&=ZqtcA*eMUjel108u1=D;n04^f4`>ypD<-u>4at+2NY<#x< z*&gQNpL|$rrM*(yCOOooM_aw9uYtAFL!;XxZC|2o;@8%V@g3LR!}?RzgrMF-;S*1c zQ;CEn=-avo3Z8rI47Dm=cF@u2otwqt^z=yat9~X{TTl~_PKlW!6QS1=fUuO6orLJ3 zaQHSRLh9dkygG1M662|$Tu$b3>{6S;wIn%*ki5-7W2v1Fq}o^=+~Y9^1+-ZYhJJ@6 z;0>Wkg$Q{^4&FvJx&khB(h5D2iq0~e({N~lWS+%GPvKArnk;8wQqUKuq}o^CT%3`H zpJpnjpb`)BbxZMX;+{Y;=b~4kFUztk~0P&p=d-Gu=*KBLB-199s=4(`Osli zD{UoFH4EeVFxRZVx;@9fjYuf$di0Xcy=MZ5RusS2R0Rry;Y&&`zA*YIVkremyhOta zz~Hw9WP|lK7Ua{*+T1=<3pRBMPe(7(pkabAOrY-dk}c8YARBm}I}E*F0UVPEqXawi zU}LUZff=_ojm;xW~QYYon`ls_hM9?rNQ`DDw3sr6QwFbQ~ivzH0sB3i{Y$V!yjX2(L zD;Wv?6t?sxuNZZG1=Aqup~Y%qeC0?p+xf* z+?J3^e!!Eql6kTNZBLnwOT>lSrza}7Qu;j$0Bb-UdO~TTd824qS<5p&FdeO`Ck0(& z?$!VVy$m1-Y^r|VkY=ZSpl^nVEgU!!dxlhNSS;w{2o4PG0VRE*()ozldf0Z(; zb_V|e6cGoeS@zb;Jq<9XLS^26Bog*)y39r(q^`5{gjOS-T|R_Aag+iYo+w}|?_Sr* z#FIo3UpN-1n0c5A@FN&*kj;-jSGN!_1XtCoVCja8gb__>58Bp9L#AAT)ye- z=g?a7Xs@N-6^Q)~n5Os)i;^(mkZ4szQ4fvgYBHg3pRr$mJ{^Z`z{|8G)Ja0Xg@&57 z*5pi6?Z&#o?M#?%lfQF!pk{HEv@bF{?9FR1G&Gp?Sz~RT`kIHyP2A4810+dsT+1u7mvU-XDYw zL++uuHyeY6ZW8Nb{7@61)+DWo9O^FAQ?PB>S{Xg~N7|BNwu(nn6Di^YMm}IHa zMVov~Y^k8^SQgAcz&KTI=dJOvB#0j)>(TS9W3~J;YvWX_P>BGf8aIS>`N2Fb+sK#3 zb$LD2`=wL7OWHk7)rn4f(hbWiRc5Er=ncZSYqQe%t?xHHF2+e~=|*IEOi-gF-fbDx z)40I!XOXA8vf5~6crtjCWTE%p;4dI*%VQ?ke*kWXSnrr292BuJ4%UHjZY-jxfmaRIh9Z!3R{Kv!$wX3qVcKom8iqj-T9oG4fyyw2UT9gLut5x~ z)80s)HAHq3086S!gn^!Df8c5OFe`#f9a7M8b$I1MCdlVkrbCBhC+<+I()e}fkEk1hjgX?q~8@R3kR71z$Ak12sSX?ISq zCT0YuH+hP@H=Q6%NAWDF#I8`-t<(|+8D-QJ%69DeYkMxyVU*`~k$BSR{N$zPT| z9L&sHgy0H>M)|K=%4KRQ!$&jaL+}U%-L@?tQm*Bw?dT++;f$J{d!#jg6-*Ei>>nt6 zQCWGcPKL%5Zc1n-(HK&IA~7;kVDM!<5JgO{rjT(5t^d5Q0+ROox8kJX%bx-tw8#hP zHHhtl2L-j7%yK|5-LM)TVWt@=Qve^e`ioU=Os|L$qQ>CX(43w;8;tFb3ZZ)|tC30S z$OO0Kx%+5$19bulBx;hW6eD6F^q`+rnFWR|g;o{Kwn4!rS-T(Vfr0)2-cYyhd*16t zuYTxm0K7U7xEh&-4Vf)J*o_exWe62b&w@aEvZT<%q(J)6Vhsd~&-HNK%$}6%VMza) zKatbWu$hqkviYlQd5e;qx>>%FLB=xwyvqfm&f@I+}JawJqb%<>*Sh?0b!*xg`BaWzgw&$zzZhejIGW zPEZC&8#X)^H{4+z6haoBIEAJ~+MZ#!OZQ5oZ_kw?Izro&pBq+7R3u2}2qRX3XeMVM zLthnCy^L3GW}<+qT~6PDLx<_54uS!}uJ(tjTiMoH0NJoG6d1AAO|Nd0wTZIBBgdZ8 zG3$flCly_HcR;w%G8^51tnhd!A41pVo$B8*(|%+lL~?OFL?tC(D9(*2-&Kd&?RRGL zw$WBs_ZNy4kNw=;9X@?L`76)yO}6+Sfbu^-j_BeT>V2XghRs~i%J#$Gi(%549jlen z*U&_)SAH+f=?n#x+&0>9Z|T&D3xeFvvZdg z#Lm{cM{S96mIL6$3G{+0MAZ(@GX8+whVyq~x4zwcr2e*D1J?C~1=zW42Y;FdFg~sI zg7=ntwL?WCPyvkK=QYx+r3)Z2nJI@+`N>m$o#b>c<#w02P?{>04S|hZHA`V|!v?p= zYlM3j>I1wqQA(KhDatn7fd<9y&BJ+U8d&mZbu|f*%;nthJhsvZ%AYk3>uTv(hjecp zKKld}I2gM!%f_qdTRdpnO<+yVd4se@(=W>hk7(0-js_{q6r%LeTv-Pa z1Thde?xfRIF^;8w|#$)!=N7f{tbJ@J$J|R zt_uNfQEOmt-`TyvKDvM(Q&mj8JuA19xYI{QfRBgA3!~^qMVTIjSZaK zND8KSi#3t&%-?>|=;LKUcK77WbNL*nqXm(egC;l8{UL&cUANY3FoJ|5^h!@64LIsC zn~m|eez%o>o}H)b2DX?@A(lc9k-U=2QJ>kZV9WAqJ3n~xD0=bJB+%&FK^&zb^_S-^ zUv#mnj>|+x__9gAii?SR{eFM0miIhSkh8=c;NGK!S8{AfWmDUx`UjBokI~HC<2`$* zi@6DzbQkcEsxC`kTgyywsJtD|7;ThxaO(gpRU(}<>Y4lTMbmGt;m~W5Y*4oJPFRxi zd!cXL^`i6NU#0#2c=>WM7KE6gT~$04WTE6c8jP;inN0_O`+oLZI}%9SeE<07xM|s; zY1uMa&+@NN1BWKPzv_Q~@8>MAWyIab1*gjQJKR-cwHp2b&;~r;I>`KG+3GZckVhyu zW-Z`-+Q4E;q+?2;-45UE3usBtFdZdEfRkLD{o|RYY$a^?H?tZ(7ckB)FiLhfVa%x{ zrT z1UhJ`h}OY$-YF{D0+0X#*8c!@-u`1QfM?k$7*zhtJbDIA0p$63s~4B!w9LE&5TX`g zTV*~2whf5ur~>3>T;ck`895J4aTokhi+*sn;0<=Bj)A-aG*Z&uaL6+Wm68ws1VHDL zZwu>c2r`!fo0LtdCO5qSWQ24ae}BJ5N0bq0FLu?O`l3c{?0`|Dw0sZ(WqUh~j@SvD z9wwIay>?7Yozykwm3b4^v+u8yWng161hhR4H)>}|-;KLC8bcj6?0D<;eRk5#*X+j) zuC?0hLN1HGAt5h11b@#yx)*)7Jk%MZ6!t};zdWR*BustTS|A!6pLf`&M_K|bs?qe$ z3-r$>Q~z9Euq6l~G$CtuiD;>sax0Ut*IqY-OBP5-1)=vPFmhmZgGwsGz5Om|6rM=< zf7Meo^C`sS#^4L2(ZRA_g*{!%71HY1XtO3P%1`69u%GW^n-MX&ZyUOpPKnrgkttlB z1Q394I@9J28=4@T;@!z?Yi~xn*IWPXDtvvTL;}RK|5tnO9n{v7b&E;@fe;`BCg+^N zV3VWB!Q`B?O*YYFn;cCJCWFX1W3oYDqA?~XgE7%0V=~4V@Y8Q*zL}}|>fSf?>fQU- zS9Mfu1vLAHV|z#wW?zjbJZ%1A~N-jo#zxpcqOtg6qw{kF>4f2!$yyQT^D9 zxSfij6f~us5M&~rHdj2(mpT{b1oqKc2`mrOSm|f%(;OO*U=_b_q$(NjR(Io>@peBn z3-91h1J-623(S0`2Z0T#t|x^Kn{!m~tQvUqqFGZzh)Meqvu)|RC~qRGdB_O^7D_kj z%FNFy+Zsitc;GZ0ka-cShKNd?HiUQgh_QRi<3L=s^r=BoloPoRaon}`IRn1Lj#TAd z$D_x2Qt7!A@KyACt*nPRb<)TExVS(Yr#BW+T;qI?kfpL;bX@UYm_e&Cw<>DbM zyRV4HKA%amIhN<34_l2dK{NuhQ}-)dtur>V(qMIgG9vxc9H)$@Ta#z4v+4rqW1;SN z_q284nEsG?k5_Ts2ti#l*q?C*dIP00H@hdRXj~k#*`xM9{q2S!5#xkU+lZ}4<1R37 zqlK-!2g}27gycAOPyu_!bJ~aRX?O?#CZuFL4U6Fit6lTqw}Sxe-I@WmEv9qLMxNtbI*5LB)=^ zG`X3Mq*?uG9$hBa3MP`eE5`d@?kR-CP$j!6X9|SG0bkI99k8I{42L-F4Z2-2zw+PL zcc-VhXjuw{9k&oI>aGE{A_-15D!(s5buT64Y8o+xXOPqLnNFrxrK}h}N>Afrj8D*y z(1tQezWW*%()aN(TE;ZHNdlK$sX^vz|9(a@nU=0cjgjH5+7lY3XP0u8@F%k!Tr&NX zA@!ol#`-HE{v5k!KOlDcNQn;Ej}5czyeHH0M@l3}W%Ml1v;U=+v7qWJT9r(q2)l#k ze(^G*RZ0}fhBlZS3#|Mqg9fiLE!Sqc9{53E}Lc zSF2V|SSf@L3h8>KM54u|HEFgIZoPl#-%GfOiM@gjDlbQPShKDnLamRWl>#v_1%J)F zRPt=qWxJ0j39ZK4S4vcU$dlnWdAN; zM1pMvkZAL-Pp_l$u<~)$z)Ump%HM69mZk>`78N)OUIVHD!zn2b$5{`7>}PmJn)fq- zw%Z9z%6=~pKDWu6WrN}ak=`bRxud&1?OIx-<4IRy;`lD~yiuZ49s6Rbm<~h|pjM=@ zk}>m=?>3|M@D)0JNiI;X-(+-F_K6oj#9_I>v3`n%2{Upan}WNu8H67<^<{5haAvxa zrPUWDjB3Ptsb9pdXTJ7;^vC0xRn92wC{#WZ;4)&UX$flv-$9Kq*^kW0CLP!tx=20) zSs;kGu2aN*#_mV00G-ld&ZVcXv6?~1!r@S@>ga*upP}1FS0sx}0m$I9`+u4;Ps!ef zk3u-sHB)ld4+P={X=&k4hI*WsGS1%X#*|!4`U+lBrxHMU;Z8?!`?<>YJ%J2iM zjX0xLeK<+C`kzf_=FN8~k`HlUr{sn8q?t05SYSD9S#)T55dJtr^)m0##9T_yk4E8% zF~T%$L0vD{OQ%3J$VVQ9zs8632!tcdFxV`yfeN}Z?Ihy~@m8b^5UwhBleBZ@b%0NL z{YRZ1sz(YspwvB!>)@x=JgfCZK?H#`E8u4O%sy;7?AV)BzM( zL9Zsp7Rg5~9T1ViE!>cat8}bEB(ZNT%f8k37A@Y^>7}q4hMgiD;z;M>jw#$1EsF9y zR)@T13o$u$5wax0+X$-9b=>kGauQf6hUWwj+P?4(MBYgH1`<;H`ce*@8sd&L7^W<# z6RAh!XbBI%Wn|)PzW3RdHmi)nWz=8FSAG-f&-1Ii@7YuO1L&x@IFZtVGgkS5+Bs4^ z@rcC}(|tvDabtk~EYBlidqnUr$gfBxBhmQ=mW=*7iO#arPIBE^@%yWXnGA}dQ^EA4 z^t7C&a4jIy<>L9T8Z(!tDGtdc4aSNl>&UZH1yyH60g$N8HNBUW32GWX`@69J1w=gl7xvKmXmSQtkEF1_VC4C3!Ce3FoyVGr&!bU} z)Y2R#Ov&*%?Bz*#?{?;RMJ7?Ve*qCBpR|+gB1&ClFgXF9+?!n|Uy^7*XJ14cebk)h zHTRC9tGPn!oeREQ+J*(JsUz%Gm~-Bz+KB@`CUchbeQq=MqL2>cif~bS;^MT<#J+*K z0H8VZLra6;R4bk*B{$7vg$<}NR$^Vajnl`!xp?sV%X`Q+L2BYn;(u`!Ifht%g|C~ZwO=tB#UA2c2C*jSs35P7+6h!Ws{_l)2( zzT$pmLM+!V|D~_<&g0%8yy!W|8AP0UL8>J%?mhc4OFJek$B>I9)Q@SM%2eAi9$>Fy z;z*4DI5cKvQvn;njw!p&*3!>T1%*+`LEs8Ej$%xvkacR;xz``&$DVBuV2_HV$BPWo z=z{ZJT9gP|2@L2@Fq_;Dt0=d=aoVyE(^aR$Q)(S(n^2eBtRCT{G((%(1g4j|)3nb( z4}l(%Xbj&VPKzb~muRxBq9j$L;r_c=USr!Grb6F2`S%3BRuYp>Ia5H~c0z z(LYOo2;op=WahPN#)v1t-4mZlz{#w#O*4)g)gbGo8Ut+{C>uwFlmP*b%QHbOLCR?D z1t44|gUoMf+Fo05f%uV|kuol@HEMic&(PJ7U^Em(4V*e5=b5U4=jz|PcO?oidFyZl zG`cf8rmG(Oh7gWZ?v*6F;N!W|gK&GwsYOIgBAo}44zVN>1Q*TyEDa=`cp_U{R94Rp zL$L>!*Vmx^MBV-G^i0m3Tpw6R=qYB<&B8q1MfGBb147_@^Jb5cDMISr3*fd?Aq#l{ zelZ)QXD1>2i=bo95D5yCNXesYT5oWtXI(mlx|k9S2CsUh8`Ec*oU>WWKFctO8GPWn zXgbNeJ%PUj0PNvJ42O-Jv8D$=-G@nv1bCn84}PNl09-);3a_^OVei)fzquX}0 zBTan}$==R}6v9LjK6Wp3QFJkLrd02Lx$r_+x0GO=YM=mWUELsy&ll|;6jCG`2_=5` z805G1&c#9st=x}m(GYo%1IDe+ruc1BdlHFvlS~~oQyi0U-ILuOmP@`wpu0A$uT>I{ z)}-x1AR~xGeukXo>r-<~9fTYt_L5g*-(1g$Fx!{~jAH1JD>yt3f0DJK47oM^(6j^XkX{d`dA^3f zvRL9^!bj`A@5)+>=eO<|V8wk~dtj1!V;3EUT{MKGp^RK+3s4@KS#Or_W5yd-n z%Rw>KsSA6M{B1%VQAqNIrUW3NU2uxS>0>gsakfct$n_ zyMEM?Xk#vit$uHj`pdZHj`RL;RnW)72pf#bk_vI#jVQ-u{U&`L{col{`mQjJhJT+x zZV7n&^4pqjg2X&lZ*IeZ#X(8x+mq_w2HyG>gcMk~lRJLhmiRl_g9G3yMGJw`Kfx2U zq>~C3F$^ZaL|2O$3#5WholQrAcYvV38W5powZ~siH4BtO#r)hg*^V&f6$!;`^HK3s z7?v2i?^UEcppclDVqO{QwXcl=l=H&|VD7^SrVEmvS_QXTBQt@!W6X^Ba|ohp>JGw3 zAwE#DalwZElz|+9p@qRT#KYjbq;l};$eP|P4g;5+t{kEA6h1_*nX+503qI2)hU#Vx(*(ukNyu zCh3(q$03^`^vhlyqG*LG0~2EX|q?X zYvS~@EfO_i6@oi$k@w}Pk%rWUx52FS82&zJp>q>u#e<=V)%@fbcN`%j!_2RkplEFF z-_^R@^W?Xe-FB-8T$#q|9SiGc!W|GXv4%2IU@~1}pdTfkYi#HEdxwXI3=h>rSZsHU zslF<^G$#{O)Q;qk6Cs*vwM*6Ji0{yk@I|Dd?zkw3#yu~ZSQ^Ue{0*C{34`&-+X z2*>69FOdjeNyI z%(10U91<`5aBHgZ@w}V`U$@t+VZuv392&!*yPqKL^u~{}@+S1Lq=34;akvn}O`y}l z8R@Pqs?7tJxLU`p2&05baK3RQ zk#*qtc$g=E;C*%L_K%>3_Qz;Ywu|+S4+eX_jVk?w_19%G40cU-+ZDg7HOcnBZTJ3u zvQ8RtC-TqzphD|Se4rcc%19epVbmWl5SMmfq_OEcx8cR{utZ!_i}HjWAI}kA zTLDOlsWZ_=B(rv{oRCgQEqghCBgCgFA1-TGWW}D=Xs3YNt=}(PFc4FUS}LvaO)h$j z3d<9$kMpQ?amMbA4q;%x83xd0?HhSFFihrVoY~fZ0rM>?Ab#Z7?ALZ9KM5CTWM!2A49?-i9!K#_5AVosYI_=U%o%-Xp}r_CMX)3R*EkD z`3I1+FIh-H@y)cMBIzp~CgQMtP{y6BK9>8#UGpw2Uv+f0Cpn=58vU3NELu9! zq#gB$Zq9@wELOIEKA`Gzb#?V7w&PCRmy7&otU`&0vm=_GLFKMXpmTj zc4#W$TvuDRRHoX*WLA6o=ov?qocnA}gV-&3HFAkMtE-0pGn|c>x(2tl=L|!kGn=%F zt>Ltb8T(TbD`LaeC+F=m6c;ZhY9Y+(5XxpQLh;g~05vZFi*vzN6C_GVE8@&7eiN%? zrCL^mQD+m-yS)|`BZCk&!k+fn&qNFv<>xsrb;0MqSmCPjMc8q$aQc!?_iX34H#Gt{ zA>R$D{KN13ytnlvjZNl=X^{gB2IhB0v3?liU>grS<$59)z^eRK5ivV>kJ*%}OzR6! zqtajtNo5`M&_thJg}9xKEmLAQKipXU2Wgy*o5}8HXp^x1HmdK0!Jc zO$7^YOn*I*LaQIaR_c2Tge|#_PFZJvH3gU4bsFBe4iV(gK1Uc}xU#sSXvmA$r^wi` z&qf*4U6H(52z_<(9^Fv8HY7)>vBx~#o>_r{KZwT>Bk_yrDPh|lp0ez=RE|8FnyA*}3wnj4{H<1V3HJtcp^^5gN@YbsQ2!(tD z8U$*;PxA+G8_@7i4s+j0<&%UU{M|kR`wbbwr7h#N`pGPwFcpXxb#g|iknv;-g~fZ1 z2(s!_vjEmGD?$Npued=+tsF*H9!JlkG^=-np!nzRwzZ#~1^W@I-riO}d~9o-L;ywo zLb`gYW&NM!#n(tl8jN4lX)_l|-~9oU+8Siq7_lVX z+N17%pOnsOx?vV8%EmW63lI(>O5t07^?Gstgt%~E(9v-H_tuEou2LMpv#}o|q>#u} zYA3R_fOr8Dks|-#W%2CCIx~kDp%YWiYggP7XULQ3DCMZ`AE=Z?dcO{qt8)f@qdykGX-oP=<; zmJL0ygQ|r)aFbqs%8C_M?hPjkyd^vNSc&#LB&U2=@P_R0@g@Q|$&e@)jG8x1?W}B1 zqqLOdU0l1<`0e@ibbU!bYB-GUr8_IS&=g`;-ncD`irR*;b^ zd&x9@KX;$<6}{G1|0|^Y2Y>^JRVru9H>C@c`8j_dWO!dOJ+|+E5qvomdsxOsVxC&` zb$l^28B<&S12}79j)37({V)$!rRIz4RD4``E&227%YPh82`?XhaHrAvwOitQ3c?o` zpG5RCzHe%v{OtUC>*q-P7V7u)F*WXK`SkQ=GSS`4*H6Z-oXPTwY3X)n_(C|GvF%hR zvE%yC6XZUhjp=?Qsf@OhrQK8Et3RWd4~S2yLxISmPE8Z0z8Ero~okwR2qpxUjd_?D^>gfsg$8vDUe5o z6ghuPBq>y~UB8lzR?Ks~*?YC@hqyb@Y z^+2)TWSfwMy#iHfrp!|cI82u)iD+6&;pmhp0BUBp=gP&a}EOwS)R#>c%}^!bBO3iq2HpvJ?>d5@$~OJzovPlB=RZ;a;Eh%^^Fdyjap3T5;Z9na6Yiqo$s_=+1kYV8>U{DP9Q`&%>)znhD`bP7;?H6)lw z9yTq;i6c8N{=F-BRbJkPkv=wJh>rs8Op~T0IX8pTTcX_!Xkh;qS|Np|CF2#A`17we zF7kvUlBHhG)}68VYJU1RYG;Qsjlq54fLxOCMD3U3+CXR{006=Uq8FJ`Uo;Q#;N3ql zH-7+wN8!gYablX`q4$ODBcFV6d$Vsg^58er{&Xn&9=_nu{e4qoX5u|rE{*=d+&`j& z0C0w#(~Nfq--qgaZiz--!M-xTjaz#peoAw_-1Xv@>6Jw%&2{1} zwx@+y^PP{X2o$*kM6wcdT1Gh$=fnJnKY%Tn4q_V;mEnvaeSYmx=(*sQKE^6e2z#>| z5Q5P+plG8ML~*{rkWl2;C=fGRmeobJ`)HPCijEC_N6xnQ$sId47dzMmz)&I*EPXzhc0!5q zkx6efcMVF{{1dYuPLCBsN@B0rI`|Z|EJklz3t51m;N7^!I3}E4P;1Jyw-$i?Giva6&F@K{|GEU-{lBf8DlD7A;OrvANGlMUfGde{SrB?NiMX>zzG|70 zHr}JFy&zZ!z&`x{)DaG-sY|1DYHnICm(q;>@6X{-zO?99k!$)p>^2L56IBZYqe`Pu zufIYTmtG5gV)8HLe+=XINn(yM_ae2XD=#0M|APFjRTOwL(Fa5Gn}^w`3GV9tHYxly z2mE6y_^;w;phX*z{_~4>u#X49ljEu18#FXV>F|%Ah-P(%D6TwydIS-F-jByY+=&II z+U!)Ne%&iliejwdKrrj~>p+Wv>WwKhU9l0mQUDUHx}HYx8@BJY&s`pbw?>w`^%L;O zgMk?Q^CsFy0RzcAG#-Q|9~5A;`fWAjuqhjanJDqJngwbyS})BL>=vp<%xuG6Dqjr! z4tX%bQ@f*$@Z3sRt#U-|IhO)vVswuw?p~W|3_iv#^8H0EUHEG=Ipq=um`G@jF#aA2Y;g?W{32xe;u90EEl? z_z7^)gK{GBC3OvQI;ch+!gu5R7+zFPkpl_t4BiD?9xnbxK>AOLgA^K8-%6B!yG3%2 z#DKLE{oGX6cS6AFQ|(D*bZlU2{@=If@gxnke-l!s3#;OuIy#<(cg*Jf0W3`YESx4R zun9VC423;a%9YN!e)qWVQ)Lt^PdPpp=_VlVw|>F<2f!4K<^rI@3@4m9Wy00&M-32w zi%kKO`_^{(6rSe|!0%WZd9>cZIn9MSEq?AbY{?gSpOumTYILe712gjiml+n4B2&UT zgteQ}>b;Cd4i))&Qs{QqAHk5tS|nDyjeOZ5zz|g?_$yI=d(Hqxb{ia7-8>Q`M+_cT zH|B@&6A!DIei0}O)LWCbyZlT9b7pxKB|9n@)BR+KG)F@)919#VRYu%=n{B7Z@i4g? zvvSyKOsze}iYxowDZXE5rN%A_5&||X2Lg#rJixmU~m1Fpg^Oxg~?sBYOzU z;E?G9KQ!{ac}<9!Iu$ON5-UcWf43DvBi?zXdg( zauNB^lMOrRmiWd$#Uj>4vXWbs)M4P};KJa0wS}#y)r$KfkTy_6B=;~|*el;Fp;=#+ zp0m!ZWBGz;M`7lii?{J}s!q6vxwA{;Gd_!Zm~%#pB9Ydk7$m&`dYvU=28Ix`eoyn? z`IHWaSaLV9zN+D%+jtLumt5r@KHtOcCL;H)r7|P~{+2{wrecgSH!UCpAijbm?YA_< z^qQS1!-SdE7MLtlvDM)4EKprYTH(kRam;#J#5AjtvR6~8eYsjA4%11^8K}BekjDZ1 z0F9}IT(c2%jV*m5)y(0hn!#+GP07ph=W%)LZuEVF*P2F&L5{fb>6*JaegE^F@SyO# zw|Cj2nHf0hZ760^cXmp^siBKywebn-ik;YOFSKk)g{|Q8jQ`HV+FlSA4zsDpZhhUs^J=j5bs`)JR{~?}+JgYOJcmbwo8*lc((gWvt3TRXKM6+q%iz zcGpk?+%L0#&s1YcBr-)GfN5HsQrg@o`%&{v#^;=_jtD>RR|k6VF_TW3eyq2i&Z?ks zhaP6w9mq%Nh zbfgu)ft-%yM=k|SS$QMn90*8Heh2rKeOA7EsWpJJinsYJR+WS7IXNJN$6po(zEiR$ zd<3*x_y8b9q1Z%JE$}ZGT84`PHd>fTxHaM)z9ET(4C|@^=U`)>3KfbNqgNWSE6=->#Z@j=>CUhZzoy4^;LJ`13GCw9{d zBc+30@P{`KrdQvZLE7;cy&zDd2n)R)ej4~YX!ylYGTHtV$N zG$}Wr`^d^UoyS*wGgn0~s{g`OU>gvOc9vJ)g{GsT zLVS47g9Xu|(b`iIjRKDQYzFnt(D!(k_4YZ+^E;f3&wA6S5{zP{N- zf3k|1b7RU6VZ!jQ_NGPJ6~Kv&p-|u_jPyWt8514>a|ujvWJk?n-Ea4**>a)dC{tYy z#R{?1+)Zlsy!XpCVa8*M#HPB%X0N@RH|3P%`6O(&n+{cuO z0lvW^gAmN@kNLrD`(w|OA>@T^+pzRxE4gvX-~n?BNF0OOQqR{zUOM;p*M#hf7Ht?c z)@Agx?W+)Trb=0}a3b*&AzWwN%?}xvOyP)}tOM_+64qBmJ*FPedwM)c%x5cjhlVUc zoT;uFaXatPf9JAr7H#}eH*a8c@aTDiQh>y+R2VMIs5Y;9n{<~NR%Lu00yU3xc@ckF zOF!^-%B{8p}(V%>Zov~c~F6;CvuRi%F&ipK4-wmcR zKk1mFh!5S=*_%VAcRIhbPC{tae!vxBEJBlFiS>~M1X12^=0bhv$?&l?_FkS~TM`kQ zIhtvBnBoYu4tyUJ#j_rCl5hDj`K@UnxDl4>#ckSqZkn(tZH`ijqnuHB8tSKe)(?Lj zg+KopGl{W&PVL2MvvDG{5(iLEDrQ<$$@@BSyzZQ?8g9X<>XW~*QKDJ;WkzsYUo(R952qV%7vd z@hXj!x;fH6csa;^9CD)F@T9tIiEx!XhaF7#slk?zbKBqti8CR37H(N|UY8%$6aR|oDqQOUQCru! zVOrXk_nK=Z@}{NIC)q3i=7h9B zXbI|WkhFDDz;QEaTAua=brLJcJ!HaBFZRE5fu%||ooKy`0;#I|IdkLMM`E$3t5CXY zkyf4K+r;_e)_Yu5jRwcM$%f8O#Su5ywye~Hj%_xdTg$(=5mNF45CHnnQ0kcZD{tftuV_#pNo7x@ zhw=?sddesuK<+-OH+$vZ4GI(QWU#hoGs@r~VVZ7$ZW!YmeqGf?Fs{ka`sr0yG{Zof z^vNf6Z@c0WnSkJRrG3=br`MSaLOBIiFFUPx`Q4`y$IbWT8et=N6ge#+(#g_!0iQhJ zO}Em2j}wtwypGXsx~2PN#fY;!(bfCqQL={Iv!~=ZaSj?b!3P)nvlSZoV&4B@fshFJ z_0m0f7oAu5!2DkR2OGt9U%N!+)rLyv*+?9nVk|oGlew3O62PWrAYM_;?&HR2j{L;O z6Tu&?$NjL-6_Ajg%SS|P5j>#6qLV+3pDZEC4@Sro9X27>BR@f!oPWzhfko2 zqXFJz8v7qGGJz8P2uc8YGt+vRa~;R*dU139b0D{R0`}GNcIweNh%rY--aGLH{;D6D z@HlRkW{I^B^Q{p&s*=R2Cs&gkVoKy3Dx$@0hF`EHRqbbZiLq7t;{ZKs|6hc=cHWaJ z>gn-?HKXH%dAS>gXKI8TZV_TPkl* z!-57=VMqx|9^wuIo-ke+VZi>&$hr7$)g;%YGF8;+0`AM z6kqO|d&N}n31@9Wu>Wps>SsC20V_7_=@m(jd;5CnkC!9A3Odj_zKe9<0=8}^&4RNJ zlg|IA?|<&g-=7H~fp_xvkIe0z+YdUvCGzxd2RY1(S4g%OtW3J@wwwoUS|IKG$Gt;dmxO1m4zV8np*7DEt{{!WA BufPBR literal 0 HcmV?d00001 diff --git a/en/2023.4/assets/freqai_dbscan.jpg b/en/2023.4/assets/freqai_dbscan.jpg new file mode 100644 index 0000000000000000000000000000000000000000..0974550d27edeffc4be117755f4c4c6ce126b4b5 GIT binary patch literal 67965 zcmb4r1C%ArlkaWYoVIP-wr$(CZQHhO+dXZ&r!nnm&Ags(ch6h=_q_9>>eQ{u$jbah zW>#cG+^qVu{%0S6C?zH-1^@yA0D!(Wz@IIEFaQJ?`0xHjpkF%}Bp4VdC>S&Z1UMut zG%PF(Gz<(JJTf92JQ6$%3?e!r5(+9B8X7DD1|~WxCNe4->fb_uK)&t)1%m&YY1P6kI`eI{!F+lzc z7XSbEroCO{9yTDhGo~7}>Ojn(K-R0#Qd?Y=kmO~W24m*$6MLK z84S}&tpS)P;p6N@xIfag$`}4T;9ph> zk+HhsBWJqQi-OU&amBTl?&ZM z0rYASCQp!6?%pG)Cy&#KdEfHbCGq$3yzxoNqT2^RM(Moz zbnam#{2PMf&i&M9N1e+dXc`)+s}o;SS*zzYEIpy%nHa?D=8`rIr@zf4}R>s2iv zklW{7C{W^?SuCKm&;@UPGn zMbrEKd=_~8;wi;OL{;h+6Mdg4*5Ub*>|qbjz-M;t%5*xH_Wq-HqbV8!JKxw*4c)%A z_h)=L{qftLp{*@e@+R0F<*hQjE94!&@K}dqn;OsgRD|g-KZ9-t%4MLlBF5yRS7A28 z(y#8LTAa$zKz|bUD*yluzT`l_cJ-EZ8?Pe%i>HT@>VC&4s+y)AZ(SdAU5e4B zG?QVue`_wtF}Y}Fatbv|0p9=7eh9Q4)Ey9^7_{48EIcoo&FQ-vqU(B7%$LTJPCxBE z-eQHXE{8$;_|#Y35!o3~EN6+E&szG~R1W3lI&;m=Pen4pKVQ)NyKByIF)r4m-q;3% z$4Su_EBRMH14P~(av4iXmM zN}^ggXK>?5V&1*M6}NNwiah1k_|PD~-X!c@Xjee^dSw!75c%bpU|pQ}MO|c-|rU^~lKvUrM?e z>;8S$VbwjSDemV(&ll{OpzHtuo`DS`Y>u{H z`aMOBHJw-vQ!Clt)YJKO)Lz@Js8b6+lWw-?P%gKRx;W(OUi(cqtIJrDKL4cSa^$L* z^SECa({=X4XTVv0BhM~*Pww5;X48_7^ z=-L#Y^xF$g48EU9CX*AcYv&&Tj#RR*kH0jSJOj@il>;FRbA0kf^nHLp@-~!Q$}`=p zc84eQEKeTTtM7iX%72&TpKEY7G0H>@Sm!<7KEQZU)2`?M<5XiN8LMY7Tm}O%{P*uV zZ!$S-CCqwWwA+{)c*dEE+rc^&#~s)oHTHRIs3xAb#d}~u5Z6CFb!_De80kN5t7o#< zu;t0^a58S>B3?8nUio-O#zEy>E$9wC%}LN6A?BM`vGhcA&8FK=izM0G13CQ4yqET! z+Z-<$IXKH=tv@H1!;ou0!5GI}EMo)IaU>aCQZspO zxZ7zC&EC+rKBGOVJH=?X&!D~OJyn6boaV*RJ{J=yYmAHK2kfm|*FOLTU*iWeySy>I zct&%V!&ijMzCM~c{Iv78-`?8U1p4LDNAjgy@+lpC=|A53{ARlOZFh!sb+~xa_UY1_ zPfw-@bo_L8^<`=9aJKRVAIbeEAsT5Mk&F90BjF>w)I!hRT(WTZywmyhiEMkD#~GLR z-5>6EMhwZP0_t19<2`DTCbBJtnVh0?&s+A<|EsWHSFkcYGL4ab8~zL0ogvB53GsAPV$L5BZ9ndm4Had>Y$wZgrY;MdndrM)lF8Yj3jrZt5QXfXgsyYiae9 z!=YTb{qQS)T-g=H`N4&xpS25c`4}Ht^M-g+jV0;k2&x4DfY4vEIr52|w&q>;|A)Lr z#~T&x-uN`YtEk6$uVRe47`Ip-J^s+n4bD+LkJ_E~LzfFNrI4EfGkF@zVQBAoczdI> z)BuG1zQE|q+vLxKcglN3UyVbNo4dnXg!q;cBPy$Hmwj$8!)=*2iyX-O=~^mZu>3luJ0HJ-9sdb5kvr z7Nz9r4&uG$AXKLVtOOtlKMt)qyz++eiFkpq1BmMOcLV2LsJ7XiQKz<*;4lOJn{Z$p zuNvi<9-Uw+>IN@+wy4KjU0w(@dp`%}2Ir{2No%f@HWW+0KR24`oCH?FQl0myYXEq7 zN2-d?ZVL{3XR&g3X1wgDr_0gl^V!il>RYDTgtT9zrj9g`S8QLOm9CGfx{u=xpcnJ{ zu(Ca@Tuo!_MpDaGt|yE zdvx>8yD+ZdwE3ptnf8JG29n`ju&0~88089JWO_88`U8;hndd?d0PK5Zab8Wn5dWLO z0d!XIf5qf^G}5bDOf`HUuWddx&{_;ls7n@V3`6_xl++Y1NMWc}L&5-z>9lmbD2hGp#NF zkm(@;(Gm6ANTL(@)zlIv>NhXx|TD(IG-afx$t!gj$ffTKF6HrLKZ}f zCtmbU12SvVyH^I3_fqz=4sL|>=8f-ePX~wJ81g)aUq~}rd0aMkR zV64-9D7-JN)(pPGC!w)({w2kK?jDE%lCdYP<_|D9{MIgs~tqp=t{3SIBBJQ{6qFT8v zEuZ&AmnO@HI-3Vg7_ICPLYX6*GcPNve;+0w5bW?5W+$o$Hd{PX#U$G0sGYWo(tDy; zjt%7)G(*SlK2*KYZ*%S0`)Rj)bT2E#(Kx*lePX-65%HXYYIiM;P1C6#wr5?)%_ZCq{KW~(du3dyO=nmM0>OwCF-pma z_;jWxN>T#=5P!p*V1#YW{>o--A(sQ*Z}2z&7lD&!2Eb~ou9z%3!?r=!ze`jl8qdra zMSm9?#`*&YqqFPhsI*&58v90LH9a7rbS3U)(@|FC65zNKJvqR<#|^Vs{%F^&xul;8 zHv28wL#S_2P#f#QF&1jLydnODyPBhvjCq<#003~%1-~Jk%IPTTm4GPtTeH(As(L?xicQ39yUIK-Y>mJgB>HqxOZOiyb)!*=gsFwbdO8@YH)mVY(u#W zzASFJ4Ny(WhYvm7q@tZ!tKo&OzR7m9CrJ`xJ8f|P3e@-x#If&49JFUI-lv%_vu5$= z)v#(S4!CS3;<(Zycfp&h3#0=-i{+~?D3C0e-8#G8#NOH>xQTx z)YxQ`whk(F8y4T#&mV`TMrLV`j*Zjv5S$pBq%mL7=dC6YXR51@>8FzE6pGcNtQ%|? znHpfyFKZfO^8?IPWkqwmE}0yh8|Q78zsaX%>cg;}k`3!Ogll|N1E;qLD!KLOqh|AN zqTJd@(2{Mpu1hoPS$OJcO61g&FFSinO?uYF13wV}1m~W4f9cp{kPCU$-ULb`2M<%k zZqc93t@9hBo9odv!=1uUj!oJTXmJ8jUPcO@44W^@DJPKz5awkhAdZ$TH=MA9tVDT? zP^i=XA09q9Z_W%c-zxj8*HhA|7_OOkM=RbY*pxGLD5?ch^B1_QT3w!j*}=EnSBo+A6vNl-*>}BW2w{5N+(fJEDpp4} ztj~bItqEZ5Cee|rl6I7$?h9kn5j5C*Hfz1%PwDdhDLlDyVo;h&H^ZSa^uO(3u z@8sWtPWp6PROqSC=eL% z2SWx3XG8HH!2Y2R(%#kXM)CS#yDK{T2u<9>u;zRo`kc&Nn_7o-)Ah{MGR_OY9pSkV zw?Tea@xipVws-_{R+BN7H&{m!b%EX}63M8yV!V4;WJW!VYCpK~55NLx3-q_5c(cQt z4~QLChf}tQ?lfOBJ%WZUI-W|eV@(6Wn^0LddqltRRa~Jm7}+{$utgj!jEKXb^9e79ev?s**B|4uN*+KMCZYL!Sn51Am0pHq~sD z4EE<7blD{m)XJsFoT%iPN}X{V)*be z%f1cbKCnO!CYd3@Si*xBp2+l(I7|I_o*4XwrcpSdDNthm`3%=DO88USk0`c5+~-(; z*TN6gt%($cnb=fBS4?&G(UU!Y8IqBq? z_t~1y25BeFBRd;Lennkp9@#Z2X@GRN?+6d~18sAc;eGJ&m5i4xU)9v`*w|XW#Ke4D zg!?zT5^bb$bTLVTpLY#EkW4n8S{fPB0}I;P$Lk|HB^Mcnr?e<6s!9Yot=*VvFPD<5 z#8--r?Sl^%z{_bUg+gXyTOLg=6zv@pU9sE+_BN;C{W$P98Q8w^uC2;+ha^%v|4ayb zh4?V}12}`_H>5PO#K_~a8la`51ZNL1S_81b!Dp&Kuui;a9OhITlN%*QDAp8Pimm9s zm7Zn`woy5?uOa?~JwU$c6zR~NRl=P-ZfQ{3m<4qW6DzoAPRS$O@Wwwf`kmN2`TK3} zIr$HuDgR?5+@ALfv3(7E8Aenk!7_jifzaejwXq7f^MHjUkcI`j+Ic032=e{BPKmN* zSVhZscvGw{$!q`qR{ppkUlQ)B{yvPu5+kIWg8g)nthLVXZ+U}Xpwv?xyeided0BhhlO6vDz()V$ z+iM_yo-vT%QE07`P3G5-&9wHa4YP7%NPQwLO3X9XN>^;UsnT zI>h-y;{qEtik@<57Y{sR$*})7i*c00DAIhgIdStT3=KMdRLo!}QsdS%{&2Q?}R^@xD=2<8NT{K7{8s{+15xibm-P{mIAxW=A$4 zx51~|Elv>7;O(bHjn=9b=CeF|#2U&8p186}S#-K~d_;sPtJQ#TsuILkqw(Tb@H^X_ zX{ntY!eNUc7BzIc0z@vCIAcv2*r5H1lpn6e*BHOWwq6Nwn=@ns zWLuy3(?|y4T06TL(s(m9nC{oU9v8`ZzIN^+kp!DJ>9Hi$R3n9J^mc|jN7wXSJ~Ay; zNHd%DzCM1lNzCUrjSJb6d+7DkZyPcY^cG?Jdm|%~Tm?jzWUc0UgnPc7`I;fj^SHRr z@;Is4TG7rF{E18%pnDDC2IS>gB#Q_qc6?e4!Cu?+unQY`Z5sFy$iQD# zw^|cTg9vVdiH#4-CRda|1IN2bV@c8M-01ZRa0&dDcM*#zmp`Y3jrF7 zO8KVN9<;c@LXGARkGPGFd}uwr#ZR+`H3qU+LPCyLaZY90-39S#p?URv%()$+oMDBs z;!R7{7}>bi{j+B=HJyTVdPMr=|ZI2v3T z2@S>&2&;d24@G576WoglAGYzlCfUN09x#VGNpetAKw@%>utQi>YbIG;ut#`KQK%?$ zrnX9!Br3|^jq#yoyJP+mx_^$i&Q%!TK=9sRTRhSF-ktoxW+zT!RWQfa4EQd@d?JQf z%30X3Dh@Y3GCO)it!(-;oGLUU#}YDdcUCtgym)-fG#3mhBQE~e_M{Xjom}Z!=t0x= z!ke_3@u{GPWtLMiIfLh><6` zy_9H+a=49L6t8L4axV8+YQ$-V$W*4Ses|A#vq?_qVaHC74H0r_Ol)VuI8mrdZk#sR zHU9Q09*bM@iqB{ovCB@6qnXQKkmBrDEYjH7P;z?P_yDTiF?NO<@?6*stELgV?0)Rv zt;QNaCWy>N{$ny_rbL!t$f7o!ZML&Hh4Bho{FP$=i9EsKLSC9=hXv($a#Yb4W7rLw zurL4?6$Iv6+*pH^78!Di+q1Z8cF8_ktypbX`B)r!E*5605hG1nwmCo>m|jVXv)OFc zHTJi5#9N<*`b>w#`Q&j9wWbp8H>#jUWw?{YD~BC+NzOP=T`jtST?w8_%Cf~!D*9R& zCX;AT{x9rM($m{yr$fo<9li{G;V+~$Awh~2WXaRYX^Io#rBxwdP!bA?j`KNPURZ0U zq`tXQHSSb1>d5|y%2UnS90|9@5Zb2u1g+`6fnRF2;-L_R)pj^Ce;uvcsQjSH)%$@K zqh$hSfED|iA6=)E=P|XLG#|vo9u^Zm{4C~=JrctxaFP+Ig=;}Ma;p72xw_y%NJtG# zb9I$#h)NZDg#skF6B{kVvl25|Tvb_O0HE;WINgK%dH`{oFAGw7`UIyN&gxJ(h4Fdi zc$>+hH}ZfDwgK)dazHtU9E3l!Gr%{e`lVa+1fPVAnIe$%Mo2-9Ma=(WXw(W_g+NT+$bgJ4V5IPnr{*IQP&ZMhgZ_| z>_S7^JU*Yyub8{$q@gq5n!AOT>PFj83dH_+&bE>1O=G8hK#l<0^%?ACn5%Pjt}e8n z&gFg$0H(^Dr-~27^Y20l&%bmEz3hEoy_(+ZKO&ts+xR z78$RU`}A=hH=uCFZaCkDnzIvR0E@ z?RA248RsGn`w>)8OwXpuHh}vXGiMzf+3fwoE*wKboh(Q&PjyRy~ zzTnZ*ajPYx-Q`*6<2N!;35=h9?O0v8TBW6wCrn5|jUHAV49oh+fC7Ys|F6#%AKgjp zVofVe_h!~|)pFKoCUo#umj2L)gCv|pwjYw))he~fee+$tAEocj%h1=D*g|V8()p}B z>U@T&^I1Ek5!>0>*m8{SuZFk*ft{_KKSQnMskoDDx9V+JAA(GEH|!)jU{96BxQszI*0c~S$z5sZSR zjv?E_J65YK*2nI${16j$ILs&;7Db^F@$W-{K}2fCG_5N~-$W9-Vq7(Y%`q=aW^FJp zJai0gBo!$GPCk-~A7%LE!ui&P+;Fzng>2WIZ7`MY12RCk3In)**s0f@&?%v;!Rjf> zY?(8db9`+I%oX4I*Qz)QxsHvxV$2Utx2B6L;syjxS(2G18X{Gs=H~fYItd7o1LGIv zC4UGxZh!#{wMLb#d6R2S3UCEBiq90+LFC8?=ixw|X<^(H)EzPfLWToH;z^OpH1+!u zyJv(2S%U2MMl4Y%kdRKIV~Ti-YB{oIK zk$$533Xj9sDUf=GtHMU6D`O{f1^aQoG7Q5Fo@N^$Y=E{=*;n^&XfrY!4mvEpEx}%d zxeRj(!|FDaozCG0P_oi0Gg*yogAo^X*`W-_9Rf9&F7X)IpmG>J+!1LG+f+<<#%FmG z?Q*iWktat9siDLVNCVg=B_2RWi7%Ukx& z_N$KSkat6@l$-*+85wuFG)Yk64c1JzRR?@h_%An^#0YB&QEMbdDsl&p5zVb)^^p>- z`T&8LBkIFkc-aPav)D2@EYlVi<)lt?dBv#+M#(6iQNkQ14rRcVK0}^Pjl7Z{rP(AW z-2o$W=Fq^Akp;nY_2%$2au1QK(a{l_WW{Rj%U0R!4RZkd+xT?g7+-7Kp zsy0lHjFBP?1vxaz8VJ$c>Y3vYV6D)9@px8@67iHtB+1qv&)yHmjY@Fc@Dw?m!h#I2 z(EoJc3(5`z*Fx#pqR^qg=rA(f_BA!0`#b4nNsO11Y}9g8))96H{{i5S*n+wZm#A|b zR%0EZ6|v$dg(;yUb03EKgExq*_Ms5t-ER)2`PezB!|7$YffhbMOh_D580q}NdBDL& zZL#wa-d9Fqnmx_D;~ksDBs|bq?=Q4kIxwR_;bOH}Jl8e;O{iPYf7g05AQ9l5TBBvk zRfj06qEJb;%B)#Bzy?3b%B9#M@*H!Q`(IV1q>oFs(dqtsi9nnG*Jl zey50VfHzL4B!~MTqrSk)invW;6+;3vVvH%Fqh$>L(r$FG&B_4x?y!a6Z9f57kv)lZifjO`6>$%YVS$%L zfeqJm(_{TDdafA`;f9d20TO7p$yA@ZDy!D=?TqXfP+J%SYP&YXC{x2koC#~AQs(p_ z(?crAs4me71B5Tc;*Zok?t|{q^zcKHj70F`0OGRMPir7eCGortJ*Fk-!cYNLIHD|-nu2G zYsw`Bv>?INV(aj$4z^3iDCfhYejJB&`KN)$;ZKpV6X0a7z^|zfp$wud2s8e|A9xmqFHD{ytm~)tO7<6F1 zDyTALB)MRWe?Gp~X|uSPaSNm}yib)2vR7 z`OSua(1r%>u*CAg18*O4SS)a{@89rlrl9%wyH49?Gm8&4-L_4AsJX86hl;(;B~B!? zY^(ZpV#{?>+h{i!;`QhX+Ih^^?fds2?eH=(e&+xUrTZvsOui{ST_LVZ_!IdtAIexG z_fa;+w}rq(2XUJ#u&OwOe0SpYGFj2dLcVHplOZa;WaMSBi9^kpqUxh+Uz$q;Hr)9U zP9=#hd-&2>Mw1k0S=+^+#SxkC3*|5aUY8QTzvV^CznXIM^Imb*`f4R7WkCcQpLzT# zz#R@pti{bK1sY3EUspwc^bQ?Zn_uceX|^V{j)`(uJfhz|EgM=jLSxE0o7#iQO7()C zIOGrDgO-SVYnJ|)JtjeJx}@aYBYoF1dN%Y8b2k9{Alv|OipE>%^^DWAPS~EzF)Hz* zxCAxHPRPpNA=!ZVZc$)lTwPYlYu7~I;Y*G|wP{!~8Rj?1PdZ1lzif{(Y)%xFCsdUP zdRnflqc<=}2$*i%GUBmCoxo6nHk453H1z$sE{D~0F8re5-{Fn%<+f9}-YD5_l`X=N zEh$=5rpfO_ONDL?`xLv+r`WyT$L{edcArhL`@D+Xrbk?yM|`i@=2Pny`hw=Mv*oxX z%bJ*YkYtU{&C&eUYJGTnZ%K8NY|NZF$-R>&=~ZKWR`UZ2almP-+)uk+XvH+udu8P<+0eP(4HTP>$leObTamY^8|g7S)QR`h~=3*^h{52B-USi}(bm za%6QSIq68)spneX?m-+|g+}mj;dw^)!9`&^%CLbt(u+;!bD%z8&1Kl*zteAn1YsBj zPMtQrwkIx0OcoXezRI`*C)*IdEyQC4{6TTrB#)3(-|xIjER72jQ2gP@r2f0;HJ{ng z6l5_at>1@ONHy4cKNP{4%#6N4eQ$J*BfvOBE#Q`;{W1@`YyFoJS05gU6S(5-{oAF9GlkOdsB!Av0 zng@=+s9JwS`^75YWF06lkv{Jcs^(P--%&Ao>|#)KC>xScE6H&1es%+y8>;2@NWx=k|!4--~dSFhibn);o!GXIx`Sx2a`VW9+-G6dh%t>QILK+(z$#38o z3BSTnvM&Cdy1F{_S7G@C70Q`iUe{K9|T9MW!mH6evux%+%P!AEg@0 zk17+D0)WQKWvWsDQ1}*n%ASL~g`Av#_QKn?5V9?dpA2%u8^EutvLNn(a*9dy1z$lh zC1{MzERHz`W#DBv*#`Iqk}S$RS{NnGBu$=zSZ91(3pZ(o;Efy>Q%@cBO&y)5zH_z*8^0>Or?8Rf zDDZ{w(?Q-)rVHoW3a~Fo5mw4h7e>!u%CD%mmdP_%&0}F#8(8KuOqJ&|OThamnfidE z)HC*SWVgiGj4(~UUF#+zQOq*b$Zt)Vn(() zNOm`96=JpFMJhO(usC#46O2$bMx40aztT6upVMsBeQ+yM1`6f~H*(hql!d7Z36yeV zHle9*$Jo6gp6rUKVrR(g<8p~czn~Ibx^<#LYjyP$XUt+xu(=rj>-k9#S55&1KH~#OMMmauzUP^jya7L1lIqjW0qx4tc1BzAh(F8J{meII8 zIdYiDQpTo4 zHdd{?OY~*pS@wG4T9=LZ@RBhwj(3i!8yJy0O*9gA#E)*E4=T55PhrR$PL&Grtr-~OYWV{HNd;qg5s{Y}zk2US)4=HZYE3A zXNOTgE?)|(8bu!`Im(i05ev?@Ta&`g=}Ph7+7(pQD(TvYGNC~Zf04jsPh3D;gg!?F zZX+pFU-p@H`0^p#V27z-jI_iWa(!$O4pXEhnkAJA+3<}bu}j|}zr~kTdO3l)?AtQz zmJNQzd0Oygdh3@f)Npz-lp`W4Ncnh52(~vD#|y? z;noBR0c))Dur4qkx`8>HSgLPKhdWB3<~Us(0orEzUN9-+1Z?Sn%ibVICp|n|iWZ*L?2?1d5)sKWNP+u>^d!IaPxEK?RY9zY^s>lr)&Z_~buQ1; zK`x=sL6PRtq)}IYAmX3_m*)hAlhYx(;c&Jir)P5+2IYl59|k`HwjBl=N;obizqM7e z^Mpj@?Xd~5LXn7aLJuW2Uz_+;Qln^m$Y#G&#bgxY#V4Ii$!rxnX5>szM@UGxks4{d z7->v+)4Ljtv>e^iN`FIX`W*b6gb#;{Ez-*g{=0>ctry&Ip|Vki8y)zCssq^Z63!%? zOE~Za)S+&c2C|9 z>c)umsMy9*x*Y;58~2apaQ{r%`)FJHXxEERb~TTy<`?pdS!L`vd8uHzG#5YN!@=Yu znp&ncp@;PeS{MoRUu&sA6*_L<+^pfxXTiYW8Kl)|w#_!7% zS$!_T4kNRLmu;o&boQ@|pREA@1=Ij}10`H4?c(?3b`wirgBYDZkkhGO!JXd#W-I?T`}B0oKTfm4>!asH7i@Hf%Ra2 zoLaMzt=c3NV{~p-X(8RqtEQhDIl2pHDWm9!AyNr;W5JE~j@{Pb0Kz;`F=(C=DKw&55I@;B}Z zOhhyjQvqH_q01T0(?Z{J+^Gahc5BHv5w$h1uFw^L8*1DrGQ{N!e|A*3ImGs5oNOJW z4fXxUGVBGo%P^n-3-UUQoef|g3a_vrJ)0ANTcl?L3~-ACIG22*R}g$AW+m6<@r{g$ z3(A0Z31(Yo=-U+v<8N+=1L;j~fn_yqQ47L8lVP?mm-(xt5T2UJ3vh&i1ZIa7Il!Si ze}vMr0eSk(icD4FFMs2l50)1-pJ|3|NMG9DLtI~r{Yid?7e2kr)G<^vm1yB!9L&5E!tKmFs6=W zyX{7e#T=toQ{91K6iRP%Fu3g<1a*ahZ!$9cnW(at93nk^*0%07hpwlj!&{P_!XU`= zXM$2U3gbGE6_}#MXp9!4f@9EL1$*d5StU!qQolWArh2$LGh5P9$P~NrEHbiR5h9e2`@Ys6o{;z!3^pK|!IZQ2A-H zYsDYJKEDc)m!=0D7DPp4Zf2PB-fci7H_^^eO+4KPJy!dIU%QXx6)p;ihtQ`4q1djF ztR+=rv-aqzDWQf@;($ZwA zXdrA?TdE~kqA7Aw#o!!{L6VLWtIu{GR@7N?{N0r^`fyqBV!ZYtq=ZOQt};&55O$)_ z`dF?1F>I~UG%SZzvhTNODDHwJ<3tMWut?O92^B;o%6%l@e2V8_a;minbsrgW0e3dA ziAz5xN!xAjGrp|oak+_ZS(W2oSEELawvw%sCep;HzJwUkmXNZ~T538oT*%!%vY)D{ z=dnrdFyJvPpI#c1v;-u(uD@=io=WyT8%A%*F62IBDmE`Yghr|WjuKTJ^z-0}*Jz~@ zo!@@XM~kZGSZXDD$r5?_l4`c!gI%1O3G-W}vnti4eONXe0fF6$I-VuR6RwlgySk45 zfQL21n8(U?Vr0-SzAl<+*su$5GZ%@MB%=}46!TzjycBhX0D1@GaD-v&vtGA#ES zswJ2dBAe2&&=f{|N7QwtDO)<^67=LLxSdiMZU_(Xh0GN6E%cc2ePzU11Fl4|Daptb z6zLLNv2E?Lh}3Wmi|!2V0{XBrJe?xs;*|vl(lbgbQ#k@P7O|Z-x)6QmuHl$04oYOD zuwDESEfz|J?kJEs(o@Ch)NP!+Z8~o8B9~aUhiMLBI?7UQIYyRH8Ypo4zOcM(TiQfV z4gn9Mo|CC6H^6VgA0?Yv7NAv^H<5m7mf^tFa(4*mQr8SsB`4qsVs|q@8f3j9U`7Y< zwQ6W)q&iCLQjyVOw2+o1F1sxF8cq(0hP5dp^<-vwDo;IdJ7>h+PIcv>@M2}#7Mkqk zfpPklTol6EmF_+qQ+TKvH5PU3H#!e}{2F}9XCSit6b~U6D9}uinws39!z05y86odm zzZ1e1_5_s3*@SI>B8oZ&Pu|~7Q(B^{6Sct4XLF@PqLeUBZxiHfFyUg zBw)`V>oj50ro|6QOwvqA&~q>1^(H*~et(TQb_ahY)t1M~N%9>e4K1iH98p_Ss9mJn zbWJINeVmf*J^5Tn*;CaA1;xo)N%&lpIwA+Sx_TeH2gca zk!H<9ArjM#=LtPY>g$N3FG?vz+BkhhId=tNq@q2nhct{g3D}I+#++_N?Oodm@D zQ&O>Vs?kmr`atm|(v5kns*@N6VP@-F+l}~p4AXsvZp+9yiuBCHAHa|2k!cD;0Twn3 z$|Pr7+TIagLci~}l&|)8A+*fFY`5}`@0hl!=EG7L-|-<$SNWc~!?%ayC|A*{ahl%0 zRtX&>K#y!9BIBwL`!1j+Q4tk~*o@}k$~a(c5|L>-nw^lPFhrKJI6JfBamN@ZDQpRC zaNpt{Qsz;lSi^nZancv&7Q0YB&0I+vKKKzSa$$6`4=C#t%(rYPc?c#t>@Gszgp=+r zB16jxMsk|x`ksK^mZw-qiQZC>|A6!?S}H@ND%sIcikdAoP8`p+2xwIT)DDq-CZW)< z8}9+^8DtGtwrnLMNwv}iS1!bGC2BV-iPA@EueX`b(w)-bci4w3^L`AoNO^p{^=W%XE5o6y{@x9V&@aU{yYm?HxbSF5;b7iDO-N4jxfV@xvSbb*mc!h-7r*GVzwp4 zVRZm~v|@B8>zP=>MkdCA6W{YSMIv8;vfw$rAe0+SY*m}8C0KsDaR0ot$s@kQ!c^~a zz?Voc-7uxvWV{KPs^+Uq7m&E!gKwp(I$O2t4nHL$%P}Z~hYrCixoz$x8G?-}Q=!7Cnu)MG8(pQ-v1 z93vxr`dRE=Ff1&ULH$e7;dT!S$f+9SfZ%evp*$YI$jJ0d7u&vofAcugoNq(!bCK+~ z321$hF9M{L)Cw~D$XlRKVAoaUIL-8Hp(#ei0rHg@l5phLdJk4H2$Gp#o5mce`bz>! zPk?0BVDR=*WsbSb>Z5yG=1!1PoVq3E+ z?#;l*2^EP#T{2jrL0~oyj3%QSn}}0nHJV3$!_M*T4yp^8fohwo0dBk`mXSDb$#74> zYqF|>pGeQOvq;Esm9}znO({lEM4=<4M84kVm7$Yt#kTkw3TBQysAzjG=3nm35uO7Y zhr9@+n=)5dN0B(B5b}Ccpq$AzWm3V;Do^1JtphO?dfBQa$Dk0oTXv@p$BS@mq6!CL zq*!Sr#fq8k_}mgNO9(x zQmx1w5=~#W@cJB$6z8I*yk84T4z7$vEnIf-Fo!F}8oZG?RSxC&a@yo>d&a)7)|b`O)?5F23=DqiGHhm z7Cvt>(!DRvayyCF+i|U>EnA9Rm{MgrO*jQ9;``VsMxRm1H7c^ZY(H80)py9iK!%gP zk%{3md<&_igCOhRKwsZ%fCL5qXXWtMCli1{kU)`9zARk=A|^p%1w}(t z24lVGj#=a;hv3me#C_qfpHw&S5F^$S0cec;z0z^%I&M^$8)m#%jG;LK z;{-JiFu3lIwH0|7bJKBZn!3c&_)|OaE{FX6hMX9CTsG#^aWhgUGpiqSrG}mzkV7;wJufXC2tt-w8iBi6qwyNGuwuXR``AepGr2(I0hBBrIp)ae&#&q@;d1XQ>qD3$h7r zQ1sgk0mL4SNzpkvAVeo_P;LXnKTQDCO_;TmXuGrEQTR$l4-u$0N^zFAF$cAQsm2$h z%1$4PbGjD2xH{vmnx2jnZENbf_^{la2&v$E9@vwtA7PCFxCiQI?v6XmUu|(*o-BXU zM19gOWiGPpDt_lOx4;cnkC#q(71G+)FiU7s zFqF#$Puz`LY6_;xx!zg}>8i)1@^Ipeio!xAtigR*p__(d-ZrX6eQ^;Rq;AmJzjtJ< zZ`U*<#^7QZHv(457<-v{%y$=UTNj5+*1V*_51xBwLm?xiW}cDinmMRv>MmXaltf&_ zXk6GToY3OBFl>O4$ zoO<7X5R{MlN@g7VrQBBqfGflqUrZuBjzv>1j5&5AVcK94yJ=j;J%DdiHU56sNZE>cO!|FznC}YJGBUvf~lWGm*lFAjk}TS^ZRS&*Z^al*X(}B zFYPUKGZ*t!7Z&lId_QAMfgb5MCG4s3jl2~Xdc2%8V5A6+EZ?6=YnjF>dmD`Q!m`H2 zU{nogZkVK!K_`rRu7ZIjRu?}kZcRyt?Dd_j1U1v!#xvWX(@=nEDl_e^MR@pLQ9@nQ zE9Zp*PGZ2)#VX@N1~tmB{X3QUG6F{VTj=E!D_c)f15HK|8582I%$ezK=1{i|z3Tc6 z<*=uGS_2}Md6=NvL?0=F2xM&fV@GEGlC0h~e5M}B+h6vfR>^~3om_~q(70jr3BROe zUo?ATr43Ou-UthgSZw8Le~1VjL+k!5t6nuC=V==|Df&X2IcN=2CM#ETzt-tp!c`4@ zrtH{T69mydze>JeNS#?R@YecmYFf0iqki2=B>t^wkbW!M+NO=-YYd|Y(AcZl*uQG5 z`J6a&_1S~&W;oPZzB!hXMc8!j=+ot^fW>?GZ1Y~UeD9DYFjai^=;ju6 z&Ozu;5~*Hn`LuxT@3|9TCXM9jjf;T(BL9hrRmP$TvFp6->WX!&JhOgRzfg?e-~#u9 zh2*0F-9_unR;6az3M4ayyVJX*bf<_?baKnw1%p(!BKt_6cwr;j8QqP?kw_RtT@F8YB)bXt1@xn zYGprX0y_v~9@Cx1K|x57Kh7hV)OL%Jm4uQc1M?)X7adGz54Gse=3VI$hc^b9#KS|R zK_Ar)AF$&URGGYnp60&FeT4EYj;D!zx8l_E{Ol7RHnMB-KWL&0$fn8%ffUy&rYcre zCzY7)JaqkQ36!TNw(FHMDp#Z^r%_o}*u&-N_tQ{wG=g3FLxMa&VnuuAjkFh?+AhR8 zkpOj@EfTuxIAf+$w(W0L_Y%;$o3HQ`_r2LhVb>rEGlfkj6A z%8l;%PS^*!rlV7LWz7FoUF! z>Z?y#B~RGHG>(Wx^PhDf7Jh+{qtW=o(EQOK94|#m;ZayCfs`JD|M*=`w;J|_9B$E3 zTGjU7sP+xfVNz;Cn?&uZF?v%NLd@jPZ|DYU1*?5zSeqiqk zBpt)GptR48ENvcOU&eOP@1M0FPi1ow^9Rb39A`Je4VeU z`_J;zD?gYdSlIq*R@ZQAs3!(IfUj2}BZ|{NsJQEAA?>}rQ~u#VT1q==j*_f5>e}1> z4=7SWnJu=E84(#D;Pr-oy;stlaJ7M)OxR9wQccrL zQ=MSvyp34u%JU!nAV8$FF2z9G{X4!`|<5}nHE)T!D zdz?b|L5(p63Ap}tm&wCj&QGcDl^%f-)JtrP0SY zz61;5Pkj%gLz{GkZtY0L&@$GSK5)S@p?o9t2Hs!t5sAC^F|x9*z*oz z4!J^lS)ui60xnjakNN&)dxzhVg?VX|w|<1(wd|LobO6{WJ2}j+NV-l-$E)^Frnu9A zAg-L)AN+r(Jl;BO$_8BTyA8tCgN9Tt=Gr@xx#T}g?(~28emDeVNM>0KH#t$X_kI^` z4T*%jBf$Fyajk#v)8$@Tf>JSemka5%h%p0`!Z;_%M$uWCVkXDnmx9V$qUi}>LJP1x#$b|s5#eTMN7EM0Kp8{Nb(j5>toT}$ zIj&G~+p=C7+Vw^I&-$msIM*>LmfWdKNY70G_v%{vw!N;b*GxH~<6X9wn=~ zBx%B!Ok8y^sP;FLPIayl)iQ9txJTD(1=T(nk&(}HeSp7T>ArN3eHP)hjL1xCpZL^* z@`Lm(T0hM}B>ALPt=Tfm3c(H6zKjNgtR=LmWZ8yr6s!6qTXbXAutBy*-}q~2WvBqt zh(RVwaHTkjnbBY>JytFv>B507^7{UehyZ+-}k99sut ze;Lj9%)fO|NrrSjMmuq~s+PA@DBqZ1)j}Mi*vt^|_D)aO1k&uJdZFi#tqF>D85DEQ zN2H_Hp$aipt>DC^zNqJW%`1!-zMuqG(O5<&HTqAuo-EFA<4K%V^3DvX6?7*#SOM-O zEDwrNh;IKuEYKX)jV%?5R1aJE?ghUlmg)ohKVQVDwVpEg-P?q+*ODv8Vv>*brn+ef z9E#x#?`N%g#2Zhsb1K0F^|kWGZ^O@d7QQ{yW&vdyv7RR7h6ly^9y#*iiV_q?hHJB` zzTU^i_vGt4>pTKb&|)F53Com^uHC|g1hM!(0 zW5%PJ$3GE6>w;E!`_T9xm3dZf8CLH-%yk)3rpo7jYcEz$5iK< z`v>t}D3~=-b5TYDkE&Gov`4}FC`GkT>;p2TSS6*Zvlgk>eFubQsoWr0ENg{7x;^0) z?`L?FC42WnTxe>%H_Um6>19=Qu`ak+i&s4T7!meYX9}!NjC0>thLBu)-YIFg+Y)H+ z-x0oy%T!ha>$)e7#W$}IVo33HO8F&*_$ zyPg6DPLX~xaLWP?77*xHJ31s5U0l@Tqw1*UhgJ-=PdZMlyNw%41?A*h=q*0z1enNiw*Z5M2#M2h1|ZvSUv#R zIq--e-;`=L`MvyFgW3ve@2?e9yQrGd9<2uPZ@Q48jiR%4^;>3`RpKkK)mMBkW zFuK#~g`)o&fqI#k_g~Km`C~e$g*TlgM*nQ=wx^Zmwsz|o^7(tG-| z)2W0~et)K&jOr48+b7DfnPF@qt_SjyIVw|5uHTNUIaQ>vT0gl>ccj|O8c-(e_by1q z999X)RcqTW?KPmCEtVxOykLH$RzJ9 zV}{Hf^eEEflq+AovAz>8p~jYsfzplbAw026WxOMTxTjf#H`f0>D>nsmlABuGu#p){ z|MFgYZP$R6)l9W-rD8BW3lXg9ir&{yImqC)zgNZ?kAfCr&?3w>dO~A#6c@=9S$NzU zt~GKbL@aE?0+~H>Au<)Fx36~cS$a-Kr+#XuXB*J#w4!>wjKNL5kPt>`CbI@GlXUTd zs>Enz!895Ubo_t5ErYdQ;sV3_>+Hj_)J9`d&^Xg`0Tmgre)Tr^K&ETn4`_gA~*e%EVka-TN`&1;Xay4f6ASoFmI3u z7Z14JHq9hcx1Qq5VKXl`FBlpIbd>Ughg`USW!ZZ12qF=Z`-3xVxavB1ojq;rjKQh} z;gyfE2z(u%^{SxJqCRp?OqIBoz(_D*#lQAwTU~1kXL~Qm*x5)|4!*|g)+1S5diKj$ zvL7G+vle7`PJD)Y$z_H70)eK}S;SywI@Lgv8jKPk(cP{jn>htwo<)AETrk~;Py^*V z`pj1li)ZlIV>bVlzcGZd2Fcq9Y|a}&Wbd@vYG$^K+MdhzX$K)b^?i7pn;xJW#2vFn zo6&S_<&RcD?_6k>jqW1{rBxJYR_xm_(tFiUvtr`^_gO9}8?Dl6#@l$r!@$A$QQc+~ zh>0}S;$aT*CPO{);;Sw-Znny?fYseqH(x@0w|3P2Y+W`Cs-UdGndqYBs1FS`E3G{> zNX4?sXRvftzjUDez8w~y+vXG!XU&u}A8!c2iwp)#_tGM6vY<#9P}4Cdk2JXIV+)mF z3vpSSt#`OF8H2QD;j?l_u>~U0uoaiyQ9A4yRxx86Y0-AjRmJDe18-Y21pLIyGz=3P z5O6pRu`bbbk#3>h{=Vgt+rcPJw5mheRWW}%Qd1HOKk)hifU<6sY`ZKA=4=fm>0oo3 zyNpo!E(EIq5zZ{|%C+S|l4%q=&U)RUIk1y7i!wREUBW`XgI0-Z=mD8!ufk@fI;w78 zw2d_JC-gefo!A%!xYF{I)p`V)2?hVvV(ira) z>wjCZPtfx=xf6K>?AxA+)Tsu2diOzWRg`oL;)Kolv!wI5JPWSeTYfG`Z^85zqXZo- zK4G*pUEm(WU$702TDBaU8<%8}e~m3w17ewMKB?A}v>-h)+w5WaZ2?VIHM=uANX9x; z#T58>L$?G}-o`{N|Iyq)-2h(h7SqzvXh#>19;sm04rY3+zR5e=G+)F;uS6aK>B3l*kEmxx96EiK{#OKpY?6#AW_qG;MZwcS02>}xF z9DG^|pQeuq@|h424ZGD99!Yj7Aosyaw^S#*(N=LCwj}I#Ly%x+*J<87MCe7gL4(b|ZxZz*e7*0(Nj|mMM>yg2 zDIlDOmym}P&wRo}wUWSX3;k;d`RwMTFD9un%mPiO)qNq1zMu7XZpX8irOBf|f%ok0 zEb1(i)+vi$y+_@?mq>>O8m=l1X@UH2MpvDax&qeJgA3xq(6l)ObGv8%hFmf1bxh+7XnP;a7pZRhESW zg$`-*I4j|a@_gaJn+)tutxj#1;VFs-8bl;Dc*!);OCxzQty%RP^6zT!#NLJWUY_82 zy{~sx_m7U|I#VPa8!Q)CWd?FHNWMnqiOHqd-<Z_Zp| zg|5D&#cK`;)p{WyC{KBEQ3sa@d(c2mt>WtpID~Vq&tFDDirxEtb$$Mb zMaXqMh<)Nv=c^m(hCD%w4++UP66@)fc1Ip3ow(c*;Jhv*#6CR=xAPGNhE%Y@iV5(C zouk>(}o#WX1e5BIjON5ft)OEc>Zx%|6aGT z9bb7ntCpgH(ZL4ccc3ROp$MHz)PuK){{-T!LO-;#-*vnT?QLIlYN&nxb_#H-BsnMv zM4iC7$hNv(wBE*K#+H3rj-X09Dqw5?6sb^xUvaxK#r-?-58_W?mb?U4>mjA3N-M^z zpS%j^)oJ1ggN>)uvD{c`3HTB{v%Tz8?tYN{vFoc5lgiC5W)f+Ll=-2}N?Cclq1L;K z>DS_~Kg;|~4223grDG=03+Wxv2$j@nH!EzZnx@_0vA%Fh6Q<{UYJmq$rxNP_K{$@R z3{Rdewy_wGX7@7)zE}AL5vH2z+ZBj)cP>4PbNW^Ps;yhvTUZ6Y`C_CX6mN-zFL@?W z9#!LII;(`>*ha7mudj|iz;#*uU4O;ylnD`h(KxN4EXB^|>`$C!T+Z@t+0&S!|B@J{ z2}?DYm~U)gv9oNt5733_F6J&=3C(`4Hk#s!lZ35Y6_evv4|wkt)>zI?B?4KX?%+Li zON4t1dNErHr#n5O7(*M4$y);9HLcyW)?2tqVTA^m=mql?l@*z&*e?P@7E^FC13o`V zbGM#_u&RmP*mxxqzfAMTtHE|}UBKwuVPG+%fp{n!TM|HEc_*deN=kW<`D z_j#yc)|Be27tpwVmVn2af_M?%5r~MZC}gfkG|#|!_xeapo$}JK#hnA-uNa%J4SdECsC%;E?ZYyQx(UI0L&AP_S-2}Zl|`z<$LzkW2)mTh#pd5 zS9!z7MQ(_io8{7iSWqq3a#TcxB4IF@Pe!%Pz8Vme%Ml;Yf=fI)Ja_(p=aCf}XjiRT z$@U<%vTkDTe{Gy{Icu!F@po0n3^2iMB#|~)8XcpxahsbkF)x3GuokKQWhq--W~*9! z{quOOs0!5c5Nh&6HcHNL`5#1uQw6!`dipUTZZmx= z{jJy0gS3=+b#2y@o1c#|@>h*3)~)TfBG;rqw6m2DyLOonliq@L%xF~R$xIuSR*tw) z!XA@B%02YR9N-PCM@=i$e)vf|yCgLQ6WDA-hCx_18tFqYitY#n^ z7K?8?%CU{EchbuG_=mA4&7*5iD--4B-zS}DXZ)v*qqCCj-Jtl0E)qzw$1GY85KA2+ z7C7lppxxs~S<4?`Cx57LAubS;VZjGrz1K-r zTcaY=+4J~uC^1d<)@5Cw(8U8{2;8b^CBo^Qv-wppWAT@%wN z4*SV}F(v6HuosUbGSoBY{_iqaMw3r8z$APm8R8+pPL@#&^->LRq_Te^^-NI= zBO#g1AvgTq`GV1l(kMI=tUCXed-Jz8>U zBs9!U(2=^LO(D*C_K7(t1=LW1+NPbH=B7ocR*#ktB(}JWPp5h0rA)%!-d~kG=~7+B zBu5;9&O)CzUl?9|y7qU{@%lS~q8W*%w3g;y(Ma|U_r@F=XG>e9i*X~U2WvH7Ur&Dk zmdiuLEH;UO>SwuSX>bY0hn0OfNie|n)CdEV>{7Fp@#VeyVF|RIV3U(tl;$S>1gi&R88(qNtPNQ65FV`<>jFUFrZqXWQ7?UT z`mlw<*-4m%jUw$X)|2nV{EGNCTp%)n#nDGmw@oIFKWg4Z#6$bUJN%h$u6+;RFEZpB zl9D*f

I6k_5Zw1oQtvxa#HB&xOYDAPYp}=#Cbw9s`LKqWi(-O!-<|M#{RizzczoAGM}Y@ljPUql z!fo}viurS*lz#Bf6)EwTid+9ws---w>u+xAjTmKE#kjGu*c)xrz{VSoH*V92eAzh> zm0jY&tEH^3e*^Jt9M#j9F(g4iyOvpSU_BmK$=Yuh!I3X3w{7j*aw4(grf)24%oFwwh*dTUf~x*ONJcx^Y+~M0nZczgNfR0W&}Fn(1EA3SSS8$M z8!$Yg$3Un#X4ST+akfLV(8mwXfUmdYF+8R-pyUejngE`+nj{06X3*`&F!s(nHbO% z3V)7K>x#S@SW;%^>z!mZ=~N4%l*MGQcHgMO7kf5vWKuH#t@_fOM=FXABYzo$E8X|r zUt>nNBG(5*ZRg04V!2a@OG7^9r@!E>wAM^iee=>dq^L$}{cy?~yPp(9?@LWrPR?o! z%l&+)he`%BCD8eh?&z}%mu#V~=Gsd;dQ_w3VIAbQ*C(y?qPx$7i?7d+!dSvy0R>1y zMHS5o-WCX{s{Y1GK|pgG8q_3DaWyVck5Fk4*GVs~Jcufd?5r_W55l)wQ;artw z;rUaejF5la*TgpQ^n@pvVd$wN8-8%;l;!cz{;gm;d2WUc5Y(S#t|=oPPc&OaeaKZ(LTO~3cLLQ)F}7f`t3QMyW8qJ8ei-66(HI_kbVPZelFAq zNPd$3k`PgSOgV*2*=^S~t!n_A?cU$mf4~`1rSJN1g>`F4m3dhFqq-InvkJdOED*p& z66*5V25RnVQ+*D(<$* z!@xhJE7*}SHRZs|<`dSGS*bP)vEq)pZfs*HL4=paO>jl}$!!bH;3nx#sp0;fRaDkV zy~~8oPm9#tSwW?YNni)V`$nV#fiZqiMM0}8t6EMUAqs7kr6gSyrw|uEThtF=(>n@v zE&&h3vj?n^-8rVdh2JrisDzJj({4)Y^$5K^;p_HYl8a>MP9n^u5HlP>6*-wd1U)ak zL#$#1KnRd>{9-{RGp1`R3$dRJ;;_H_=;MfOYmSkM%g;_#I+&-*=rU)cr02{}V(I$H zyY{s6I((7P0gN!kmC3B5oL=HJaKN(h7WK4SEe|f5!iQmmMkWFjyPf*8z{=QXSLe;4 zHc21^eCnXl>NNc5M;b$>4M_G;oRr63$sl z%t3)icSDngr>F!nEy=181-Rzpsdny?4E*fuo6V|?A+{Z)|E2~gaZ#g%qSCkD-(Qs` zP+D&zh8} zqRMZi88S2}FhD=Y-9KB2P|Fw@TUE;v>tZG=)cG_^;M=1cgz@Q>G7p!KRBlSZ#qbY8 ztBxSeh)35}73B*PHO$xPo_D(tzKDbj;@oRUIWD<=^fqhdV`VoHjE*%voUTe0c!yTY z!V|2wDi&Uo67cr{N^P<;HSVS?8tnfzAmxi8QdZf z6gF0BSOOvWg6DWcEtppW*BweGAsf-)H+#B1tf)A>ZY3`dUWE78?>?2ED!vdAJnB@> z40}GABg6a$0e47EiVb&i*`y{)9+9X%c6Z_DkU1?w#<@J2p%C z?O4yHAG|Sw#b<2l-Ib~rnslvAB%)5BNIGivD&Q;EY>nZkg9``7SZ8ptjogwE7%Pu_ z0ipEPhm6}a9?Tcn_H48)5t%Ed5P6C&V$l*bRdIKr6`t|GSuHlb=1)=_9V>yw&v{3y zat~yfOS04V%sEeH)&pAYf`~Q(3obW$E4_JC%V6^r{4$tdI(x5wXh^FNK7F+x(nMw;&v@mC*1iQ&cXbU$P7FGpM93)`!2KF zDNL=sjvB{lO4^OJ`Y2BwT`Q~ToE)3kQss&@CWnnG5X9dKZqO_T`tgd}5wdK{pW|k! zInN|v8m+b3#K0^e;06DRXt|0nMWYSLb>c*ec| zH=HQ)o(=UPDyfHon7Qx!zui$vR4FO8_RlJ3m?@5BRrH2+TwnbpUv2=^VUF3~vZ%H^ zq=cm#m7LQ>z^LE0Elu0|GXacC4E(L_rKdc6TKiWs6iV`;Q8vy;8jIJ}bOB$jd#MgA zgwa$Qy`O9qs*iU`_-4u)&JY2&TpZ-$S$?{j!GGGCs7rGisEaF8*O?T?Myj{<`Q+Ic zOvWb#EOl;=sdbivwFtHMk}9RZw54n8Wy#`+Bbt_c%`;4dOj**#nkqg8HX{{S2s9CT zkd(&}ggm=^72@K4pse~>Lp3Z&em~-Cqk#UyPA&FUq|$NA=Z^TMc@Z8FFRhx1>T68*%8?7qT7e9sa&z3o zg(PZ_oAvkw2CZ&`I%mfz0nEGV**aF)``(UP{=&w0V=Zr~Ns%eIYCKh$NlyVcYze6{ z8#=qRcUSvNIQPe=Bu8>PZSqZL4v2Cz)Vvi3uaa^)H{~d4vFfbHbLnX*Dkt%n3vO3? zIUm0Tznq>n=icI?A!5i1|2$_+#aF@*P=RHLccA;*%5Z=yB>Xwz0Xdi8X66Pagi_Ej zkJ<8!Tal4tRRz^@cYr7sTX8w)XUB*@rFd}S;%Kb@RkqmCFK92yfD~c{()n+4-}jbN zi@u!xycRuNfx)Euc2bHf4&t3b?srRXsC?C6%7D)ZW@F-U1;ex_DS5)xrCAhM*HsR& zUnqg8$ZUEeuutMbMcnenaDWYKB_E1G^J%GV-nZ5>R@3n8H&x1tEK&61DS9;Kzp$_f z$(?~eyK^6yv*{1CVbGPd?#YIjb& z+=oY~4ko4cx_;>Z2;bW)CvYXQvsoDA)^a`u{L4~|r<|NCq;9e%%44GrqfVj>w<*TU z^&`fr6biiMVLG=rF*TX${q>{kvQx5}<%#6x!qH;J@R zytCPU>hWb>D7OASD1QKDdMdt-PFs z=~tYCxsc)mdpX|*U+PENstAP&zYrI(objV-A%rnttuyog4na0YYWs5fb*BiVI#YLI zZ6(+B;0?y5UfoPINp=Y@(CuJ_ieito^^mc}1ADttzctGuTCht z`UF=UCaC6b@h4f$&b*|!`Cfpl`O?pg8yw8p3O`;Zn-%Pe$>6)G$OFAXE56GkUR_5R zyB6yS|8kr};9jC1wNdmnpKI{~n;olmiz=N3uMIqr5Te{_DQ~_@pxnkIVH83TjN|OX zJo`9M`utGRjZ#AgclaMED1GBzafQ5tPw|U%--Zro&=DqLrN*K?lm&ZsUrbJRc))sfjp)U~#HG)Uv2P%57r(p)o#4{C7zEgx-JK zE0KMNxORQKJ{@~=ReW_-h$YNaXy3iNz2lK|Z3t#>~U*|xBYK82SA zsCs}sv;DyiiY8i+DX{jht$|#b058^B#U!Y+9-ogos)8S@U>kB(Y@R#TzHhN9!d}kb)fDF2P(ShhriIhU($Qj*+Z0fZ3Wq&J)TtG|zvbN3L?ok6E zVDsz>U7^e1z7hmi+Ed#rA852Edx-($NlmG~-wgL_9W0{0Io@RzA*J5S5xdnCId3L9 z?b8E8IKMlnStl3}^^a!_gwHfUG`t_#X3NR0eZNJaDp92eRfvZ0uLP}Xx2GVRjPvWr ze(#~)QLZn?XfxJn(ODg=YOtoO&{Z(QJ> z|G2xl$qZop5a*pP^y&SaYHEgug z#mjZYL;n)@2YAVVRX_Wd;|kS-XnQ!f=Lq~L9%|2e@v!7sl!;|=HThBTQC*E0IrWB< zxDe}XDgol_y`Q@zlur=YF*_odc&4D2q$ebtI6N0d1S^HY&Va27|B?%?MbQnYXh;oC%SC#;!qE4Y9S6 ztKZRDbuGaiP%QYTBymLiK9ZQ2)sB5Vd#_$+a+J#xOn3UzKJ(IeNn5)0sBJX_eougm z&eu>fXiAw7tE!fbEpg*EaWTs;)R_n^OX+IG96^s$bYg-!w@h(7_a;RRA+gjz?xwtk z(i*E+#lS9%$x92X2ccT|YRY7r`rds4lTZ4w$Bm&`37NT<{g3k>sn<+~>O7Tc0t%U_ zVrrQ3M|6l7*wCbOB^=(^(&+S&_MCeQ%y6FR-*-}GIe!;seyiTfyB%W8TFrn6;5N7k z6?Esrbkkp8s+-dr?4UYn#1XoALr5`EtFtZWmV`cRgXmGp7R-q6ee&t3!Ms7jfu%$| zstIa8I=7w~blZ9Pv2YA)t1>)yp9UVhgDdPB29z#`tK^US5aG=aPp|P zfb?`oBiJA?n7u)8GzKes$7Gx1<^ooiP83l&Sdn5Is>$TRg-nHtS$+j&jdFcAb`~>9 zMX6pui2nlkNHzImDPf9yEwNG0Dzcv`7gV}|^LSgifNqS$1 zrA>ryDG&DA7mU+z4QPelw8qv(L#w(ZW^$I$TC}pg3G|anQV0HOCHL#?Dqvv93SE{J z`=bLb_=nyR+g0uYJZ(yzpSAH7NVFp^z?J8b^ldbPrTy@b?JDn0%D>GyKD>#GRAa@NH zCoR%9(GsRvF}ZXcDzfSBUR=jPbg|(%=2Eab-TeNve3kSaTs`k&WG^dMsAGexp{7iZ zv^Jl;ngokcp5Rjn6)nKVK);I8`b7f}Dmopa5*G$X*QD{xb5IjPM^KF(Nr|Pd#7?a znpxeRq--OWv0Egj;$=I1&8+^iha_!{`k2t)#d@^-2=3Ab@?ywB;9mJ(C*WReJ`M}A zkx*lhfoe@Yfkcg*JP+d^QG}654P-of?fj#sfi6JILl#r~y@O4gVOx=lZ}pfBo$i*b z%GE+bf;>iQ^?GxHm{L3?C<9qOGCF-aHeInYODqgi_nTxdNE<05bf37C?z@Tqq%7HS zHhmok;XZPzm%s|%t^pA&E(JjzewtRqQoUU!slmDE)j5FE_dEd8DUT=VGtCcG!?$U!W=B1@x&385 zL$lYuw3T^`HB4rSurAfbLOQqpA~1O?A@H{0lf)I-yrI6eH-!CNXbLM7n(~NZ5|>C- z9467kTdo)sW8t8uCrvMCc7D64Z-+u-JYalk38&m2>9r>1Ur`Ei1#|Fgnt%u9BHh~6 zcdWTQZ{j>K-x}rHAY0LREH>FtHm5kWqT-5!-RBW{{qV+t1}q^P1lZqaIC>O?)Zqy* z@8VQL(b+H-RD`8Qk=D{`7PU)WCqA6ueLY2HUuyrXjc5k7jEW7vI2FfNTLep0Lv>O^ z3eVbrqOp|8I+LcRdL{u(~^o=EMJOr0Il> zYt_4&8coiT^~ZlqxtL?%2;B4XQ#fus_<1`L#CgUju2F3P52@2DZ0K2>n>##yPMHlA zWfvAzbf;`N?7@#59RN9Zjh&_nxXEV@8Z9iJA9MkIVA}CeT>q(^>G2v6giIAUs^jN)pjlpy!G$zfe_~yUD)YRvcz(J>-ULl>)*7z^vy{ z37UWwzw4B0tTL?2xG2VBSSA}u>X#MqUW}zzd495IIxr98s@zX6=gF*j6+Rm6h-0RM z*V$;1cx(OW`;OTuvb5YZjiL%AjnNo+GG!()IN9=KH#jGEw_rX2i~LzX%AYYYD!+_9d7Rnb!zr=CpDO!owl11mLB&_AP;)D`7NS2YDvWix?;2A|~O6kE>e*2d)p z#=YG#0HU}oyvoj1MA!|Aobp(v0OLh+Pd(!@ER&BI`E5*7en^w>Ozf|xrl|zyk%sC` z31(V4X4JZZ5`tQ1`cXcXPWHy$6!M&@C_3lZ&eLeK@54KqL;lPA=a zv{dw9RV{0SrYG1+I0w@Of&+OqlsOQ0kaJiqO4!mpqI_ft{X`^h#>p(~*rL~p4BZTt zCNe3iu;XS$&R=)=tz^Jh(Pu?nUIxQAj0K(XY`4}Lwep#ODtOR0kbL{egogBfiYWuc7f$ z1c_n80%Bfia_Fns-Cgle>ilm|lAjZ>);@QI^1=;4>4%{~j;G?C5E89cx_j|eC&l*1{XXzhccmM*TbaDbhO2aw$ z_~2Nwc>0^`cJX~5w>>lKDj(80rCR*3QoZ0_5n|_B)oIz?c{o#L;mxUu&)zd$sT%NJnEBCZ$Tc=qJLf8PZQ@Vv)l1oSWp1nPwxo**Iwa#`~*a2BE#bywm`; zgSqU)34fuB5C3V9na8BYLi~FvgZ0Gnq#`BN-dHXdxmW@ZYkz;-xKb#vXK{(FCYmYY zc}Q1VO@59!uCMj2zp9kWTo{~EAQfs}W>n#z$d9kKneMsaZL|xl>~1hC$*Vgj(C5Gy ziWQz(L}(z*@SkbWV7|;qQcI|%(BQJe#Z9w|xggw|DSqnHTmX2vvc-?wSj9Clj{MQn z#r<{xgyhF^oL&9`Sk&xM_&LYFkz5vVLF{y@Y(MSQjiK248*528#sG%IRdg#BRq{XT zU3W0y!vuioR65f~@+U8zkeH} z96O$6S!FSX)7`sB(ec2tOULGiIwbf(Vf+Mer2(fohs zA$phNh~wKZB&6HgjP<_JH*R!ioKnn0BdZ!zA`eP=a8fb#O*LWsIf8J-CYM3jXV*(1 zEE`)YH2uPmQ%Jj(K?CwXDsD*WnHVPS!1?Z}<{i&7gU}Z=-^wTSkl}P#_bX$57_l#)8VMpBb%n zy_c+C1xI4uM*UKzI=XrL)QbZMF*dG?#1j5%^XG1H3QI481Ip9Dd28CU1#O6Y$^J}a zz(v{w7l^mkE&jm{#pxR4mR%1(oA+|BmX{(cc!6{fx2imMRio`F7#D|WWT8S z`=ZuOuxZySr`ng_q$W-uqw6oNRWfmhs`TA9vQ$h8ULQ%)UV%<=Jo0(#HjA^NDEc=Y zL2-31hM6euWY2|fw>M?Vc#`kD>K^4vIb-~@onP#;y9t%;p7<}o znG$XqN$ZZ_*L+$}WCIq`No0>CjS>%m$IhgMkYDMPZf{L4n|3BKm^D5TuU$lC44<<6 zo+m}g&85pzmEnvaqpbr~(bd+&EWE{5f4{qdN958$_!QkB-(Vj-GwY>nC|HIci zKu6ZJ?ZUC`i6=HYwryu(%*3{BOl*5%Clgx}bZpy6$C^LS`@P>f|9Q{(&RVs0@2b94 zyEm$9*L_{QAK=YY=u@A>l4`a4*x^b^dGQFz5faRg_a$FgNTK6LXZL%O2E0Pn6S&;% zynC*5%o$qWUJi`cAjhGiZ{j%zZGZ;sjQ#6*gln6AUY#o zMTuip9qM)$feOGh9f_s~6w6N%!lZ)#f;H z(OJkB&NL1xgOKn898|aPo#tn$6$hc;s78=tPo#b1W|oi`Zo_Y#&IO5bd5_6d#Y(qC z82}eNA7n`?E4^>uxz)cel=gR$U=GfgchAmMK$}7_R`gXW=l-v2D4HMDdh|;Cg-14oTvHa5;b%7ZIZBSosbTBP|6`MXtPZu5<{ZXu1oe2qa*^- z&n@>yx60vBX20Du<733wZkbY8f&eFx$^2JdZ~xCZN2fK{+TGSa_P*ht?^>NmNT$q0 zwZLW4&T$w)Ws|z8`f_O#&EQ~&;7`t)W5IgtzCE`19MNAfFupbJW)RV3Bc8(PdIH}kA}KOtg&oSWBCwf7RG zSABDa4^u&&;&7HSSX_UYT#HnYr+e+5)$7Fv3ZsY!Hbp)%V&9(J*+^MkkffAlo@&w@ z!ie#hBSIRWd48a8_(*nYp*R5ag!(O|t;qbTIiR3Cv*cA{8RlWwAHC~$_2}!!NB@we zyPty%ttLbQW(lI?C-CgZqjC(9miNdlPesS-(PQ z&4?Kaa$XyPH?pNDp+pWBbqip!zm?^GC`k^~-xnvBfUf<~_AM5^3TK%l7INPI}V z(uW)keAJmOtV;1ZbD|cjw=U7;&!$>@(27MP?L7J&=8uoOlfE^A#o%h#ua&T9NcC%( zzaYf4#gyN4Sh7}sS#DuS;6d+>{`f@8jcjuE zs?#IXCtO!=ow11E=RLEZol>xVTV9Evx_Q&4(VdL`M0~69F#8v5sFA7P)BF=IrxZey z9;>fA>L~r}vyqtdNS_NkO8%(8HKsooxg*NS)sX7W+((3e|4O#-kY01JdwpbQmIBnhLr`4X>us!Zyf1;9*xrkju z(o5nnlmf@9d6dIs+=HD)x#$hS1O1qES6>G%3Gs^H1fJPoSm4 zmQn;YwBI_E9*3kcp>N)7HlE0uZ;E+u=bO?_bV-hRv?Kdbk9dtWjU%ODDiXKr@;F#g z;~o*zN{w7Ur)mpMTXhHGKN9~1yF=Pc>9a!peE%)G{pABjK>t`lkA8{-c2Af+C7{yH z^(qGci~p~!PMxhc1lPyy402Yb)HC;>E4m$Vk(*=cC1VlLTvRobQE`x`i7sceZZobx zy#@h&>j0zE{&%SC_sKBJV_}tA64)V5r?+K6!6}0XW%JJ1k-62_sSLSpr60|$;iJ#1 z0YHgg{JB#a7K@l-28zDskcne+z(=eW(UCdtq&Ffw;9AstWvr_co>1)ztfgCeF<9^% zd9|j}HfALlLS3`82BSytZpmBJbjw*XoWid>)lcps*alAz)f!z2r51=#EX&4@^B~Hm z)F}SzUVPkdN69L`U)SWJCx(tUrCD#H$#pSB-Mx2fR-kP!br$Q6lBTDr)vA&9eg^i@ zP1$jl^w=fef%}*&IDZe^s4FqO6B9$)TD&@GxQbQGhj?DB@*zl7Q7Qd$r=&enc5pD~ zs|}p!aHk;+CI*)2M#4Iv#-4*DwBZ`BWg&A;^#D}*!G}$0Odq_pDu(Oti{fh^FD4-p z6l@%%0>@8REg};c&g~YXiMyP>!o75(bKNeyLOUjS8XlCB>*$+^LfD*AHk4x+-A@io2y3-u0~do)L2Qdp8GRer1$9z6Wmn zA)Y@!=^Xj1gnov+Lma+yQk?MY@A^jJYC9zW8Ic{#nQaJ_G_(H7~f+W+a% z&2BLbhS?)ytVY^)c>4?HI$M3-#O`>lfFjH@72By|;yYUHLw;FH;GqTfs8*LFkP;dD zHO8?bH0K!x-ObD3o{Vag;5AoWl~{|+`7U@swA`pG;RO}IEw4!J(1S1tc=acgkK(MMD7*~L2+YzL!4 z7g{bbIZmd?%%RBtL2Fq(W>znmm8)=3?ec6f;M!p3!l_taY%M9@PC3zi?n>-5afyS92WD9j0Ba1E1&;EGOccqroWW(nBKk#Lx(^)U#j zZf`tnARI5hH|$_!vSxi6ebuaVKAEep_8N{nSINIprpK-npXsbtJiD91(2Zv|Uc?#w~MCgW5*bnTp$ znAaN8BN)8+`wo2D7w$fMiH&~U#lCVymb;mW+Kd>Tj z`2yE)+0;pAp7=SjS_}{_(?_Yo;<>&Zb2@zy!*-vrU$@~W4UkUhv4Y1Re&0J`BK73w z_DV#zH=m-)p&3BvA1J#LA^nZ!L({-XNt|lJ@CzbBnq+Zr(lkKth|9F(SUhFEC$5PN zj1-9kbfkh{C@tSqhop??huG3fhtOi(0G}>KFI+;Yp6a*c7YawJ3=NiKSJD7P?WP4c z;k*}5u;8w)4An~QdtugZY18xAzAkRKQJej?3h3npYLn%oVgEEKP_7{$NLkyRDPSL- z@%s6=k_kxxEd@U{DesU;wvQR9r>{U39>a=F(8sN|HOq~}mTLb|$t!J|5z*7Rgutm3 zFOrei=aPpvD_+v8M!raaCRm%~+S&Q8ba zj8CVgSql~UFBtX@nOzPgWhfcKSnl=v_qH1uE{NM0Y~)WaWySGVj_{qX)qbtLi_KUk z5kpq#0Q?F*#?9&-Qc|H~$zVpS-(KR58>-Y;FsR}~)YW`p-E*qNIar7JQm($GDcVex zQ68?C9T{`)`SKCG*a!l5mkvdUCfsSfLl`I3GCLNpl`X0n>39?Hg35Ux~w?Y83Bj^Ne$A~XB56Mjkx^==7g z1iJ?!l5^F2o%ISELWX9^!MUm@aYEYRWyy%z4nPyRS+oEbm_$2n;xSh;`wgSgZL5vD ztzxmkM#pFeH(q1rNJ!jKdxTtmeco{`OAEz?6g^?-<9tqrd}%K1dxC|s$_pSW*K|QC}WDE^B_zGq7Xfxxnco=0uM}cxJhBKpThehmmn#d7QrXh8Owp zD2n;FMBWP3$e&j7_Y9}^Y#LD6jCnu2yb3sbgD8y%7R#0|WW1$t9~C~eHX z$=J*jKn`b$7s8hDom^`DwQT{X0<^%q>oDqS@wT!*1W!PzAF<`G6c-9zG~-twPDBVh z`15XJz{XEQMvJVB;;&R2zrt}mG_f4M{A6)2eL(t}uf~2F8;8sAjTz{rQ7!ALTSM=! zS%WnUOt)5BNaEBZa}2z!yh?358?l3)Bd~{t{FH032_*DUMUB0AN z$>^L{U=p#nEZVI;!Xz~S+3CL0c3V;qXI|{dCA6Rq#(_Ar(Vsv>RTx+#D>P$$>7O9& zdr>0AiFf2rZ>2*Bj1YWU3_AGEaL(J06Tpupl3Y&R%nh~ZizUWvvSGQsv2W<@S%MDR z{THmP(fHS(7LQb>`e^em>V+&Wg^W_1Nl*bSoE#Nnnrw5K67s3T$F%h@4dRXWuc> zK7Q7$^M8Da=Hu~+1z#e@>AW5J|M=RKqn%*tHh9G8`0xUk*XI&;={Xo2AI;0T?g0l8 zmy=8HOewb5&+?sO)snf|b}7ic!F{C-K}68ujU$G-93h>iZABb$m%qD~Yzq=f9I(0Q zdI{Rh<_bS&YiC=%m>=fnoA6C1c~tPChu&V`f9T74&scpNlWgcQ%eR*s?1yaXx#kZ7 zB0uL{6d~-(KZi-=991XBDKPZp%u;+7lu=BS-*vYER8#Zzp01?cfa-W$h@WV?)%fb= zj(VI6ZB3Vp>9Xby2qO6+I<#crA%V7I#q+aYJwL)p#1rR(Dk)|3u^NG0goyB?+Afz? z-|;4rF_W|uIk1tVeSf50wBW{~Egbi-d-u1lq8D9W0?H+4uDd)BO9%F=D+&3sW<)ZN z?_p0QNo{BdzBs&-mEe^-#>@=wl?R?L662CCOH;zKR?QT3-Vj#rv~6t``impqgz2jS zpTW#PhB1Ed&|}a&4$~(2I`DBOrHov%t8FiJS`mxj!)`(C`Ff;6+yYdgk8|L9&`C@c1!W$+waz{vbaVQN^>BX z`@Z7c?=6oLvGzF#NN@$iI@}qSGXt|RQSjyZ5`)jx{ix7mjV0DK;rcz9lw{IvlS5UK zkVB7iS%{LA;`4FG1N(loQ(1XtVCqhnY0KZPuz*dI10vRSu9Kc$SwN_WCj`3e(1_41 zu-zbGxY2dlBuEFTNvla1hWdx}72|_`fHc50JQhq?SR1*~^{uc5B4=yv8VORELUyY? zBG&U-J*t0tf9Efl#CS^qw7vZvXwObbYnyvT=$_ zdI5!XzOB0L#veLRvtFbe`1<5`G-1FYJMqAg8A2cqVQK@IXSeCyY(PR6&-Ka)wvS3V zf1+nWbjv$PR2l`D!2`ZiDc-9@wTX1)kj(44xDPr*z`o!>t$l`;Yr<5A7HV$tO<8pV z?k|{?nVD5oJMJ@$Czn@FFbufiGVzp#;12S#gWqnOHk8jyutLY2X$pe%lYCN?wVGH) zM&~3OH6A4IcmUG25hs%xR7%cBV9G~%1k=wkP}eDOKcfoyF-zz?+_=wOdB`#)Ut#uT zg)BEGl#cyrEwMVPb@LhFs@LxG>DFj*(iPd>lTfWf!`^O9+O|KA2+N{RYh~9Am*E|w zHx>7yp@_im2HQU!0^vVG)g<`BP3YjrN*b%bwRL}E%f8U@JlFBONV?%oeJPv#F=a0$ zb4KBohn!DY;kG`gW{v!OGZLlM*p2a1`IiGW4@>Bu2yX2XeRSwy(&xN~qM^Q*;@<#K z&F#9=Cwz|Swlf2{1;oO6qno@HITF-InLohB{h}y&*cS^k!*bA$7F|ahPeRTJwcH+e zRxa{A#zx$+meQM%kH&%ZQc@m)h7^ne`(8X%f-VM|R*&OfFpB`;+h4e)gNRR%Wo1kd z5c*acs{!WwPe;JWVK8BBRHfAlt=W7b{|F{rVcGHSyxK*J zJ}3$W|D*K6pR;W$s|>%Jt4`{Ws#II*YBy~j9z5326ZlvpITOtcy0o+*XBt4u4WN6d z+im<+P%>Cd?BMbSs}y=;TqiAEW{7G zlOCZp7*4Ia$0@Mwes0{awC>nX&hE~pke&s00@iXt<}Uh9MSv>)9~ zmtCZ255Ojkb5mhv#J9BSg#&M+_GP!`C@C7YNu15jm|;u*=uB0w=&O6eL%Dn)qLz?{ zSAG)F^;Yk~x@ei^Cge?ar0e!v4Xvt0Qks;F+h4FeBP&JO4&WdSX%rJxduh;FPIh+? zscqx%YJ&jDG#R6_-h=gl8Fc~tcX|%{rG>a*`?)d5p3cQ-Q$d?I*cayg-GvWmDmrStS2OFe51OPcqb+#JvLP3NykMdl$| zPDcIo>_ET0Rf$;TBehtDtAf!?rbyo$s233MWK%*f*=r|_SMRKDDd!fT3Y|>U?MOV! zD?Z3|C)=FpIT}kb-qLZKY$m~|lci_k0%qBQBOi-KHx9440`I7>w4kRKWF~1B`KJrO zG^!MRO%S@l36l##mpIN))%wHGF}=T?1_u&=vzDMi>-dZ^<2zYna~6F|M(rk!rt@wk zSv|_sKcIUe)=#F*ZNDDd{HL3o0ArKk7hw$fBX@;`-<@crMS&oBSxLWBsKbtBO7Aw) z1Wml3j{E($3fj*>QkLArEhQ)9s0IWbS4CrXMVk_Y3fI2T0P)y?2W2M`M34?;!);}m z1pe_n3H@%aWX$ke6s(2W7muEz6lk#Ez+TQj(ho(eR_)>ap~E_nD`TIRnO zZ_Mkm3Y*=Tt?b!yXYQ2_^=3{uE7f*{^vq$iH<=%uYEzcvU`H}*0yX*8mYD7#G+n)O z{6NyyhdHNh_m-Rr_9I&FHFQUpS8|IPdwkU-d~P_*4QOZqqoA^-o^P!d6?3SeN*9aK z`Cl+6#JN*(gHDHj^Klh-N{II@2UjOHJHLy<2kMw<3}>(EMWZCYoy8&!+-X4;J8bQU zh=dLIEk$2nTjP2H7*?bRt!0%EVx>5L;5?Tw_Ly>5_X9;LLUDVIQN2EDZ_gzd#XT?= zppAql)Vlc=!uci?L%ZY(@(X%Qi#X^_s?9>yq0plhn=;aWN``ux;kV^UWTt!_Y~Qb_ zXpfT?1ihgdO@o4{3KC|E!!CX*)X?I_l^U0r|LVM(#75t*$P*_pNpeMgH~ck39;}&X zgG$@9A}u&)Dg|NMbH+XVFVpA4KyB-soI$ zBGMG>S8Ea_{+g@=X@SUDLT`84A7lnf=QZ8hh+xzzgTW^Yk@-lpM zt_xIK`i|^kOPt@PrLFjK>Mifll#h z%fuI`^0aqiNY&KYg-+zr=j>TJljom=fz14diT%X=;6I0Cri9?S*S#dh1aa}d??l|F^dFIxzDg5UzrT1RCA(vw*1MDhGZ9?m z8g!SDw9cYPNz!@!{^9IBRC;+QLWVOVV#7v}=$0weG(>rWM+`ue)em>C-W!Q0=GSr2 z0MD0~!p1irF_B%oo7R_qq#^3JW%zxzW;x6iH6h?EmHyiy2sBVSB_lR!Sk%JD(J%X& zy6YN5IbI$d0{3q4pgriB+N&o?6{4x#2C+m8!H&#T5t~g=tdomXGl1zS5A|ju!;Q#H zhfM{oE-eaEfIL4Vy1&D97NRSBF?zHbvSO>rYVrsCOkq)7MJvktgl#t;facDcQI+ta z8nmbN&`i%^i$KU>^Bph$5N9Wn6v0vwYqxk~qU&b1DD}gK+K@ed;XY|m@Kls9y|aFy zUhpPl<4VV|LkvLiHDqvc&#I8VAY#I0R!`h$D69GsFqb7|k0#uD&4ZKN=TScNtFJ&y zL04P3Ywr{S$QT_7$ZR4Z$_kbSR#SPZYy*ra4YQ0txgso}pVO$0Y1Q6qkS7i8lbCOUg0pw@0e(D_3T!?|8yo@-Gc-NmoD3GsubL8oP$>$D{F#b}k z<;#B0CZr*h>H4_q7We@r-)gv8R6?LkhD*?cT60kaSHSGX-+e@oEwLt=5ocO(v5U$O|u?ysNxSLeJE zq|xBK4GPolA@7$5Vsbyx18@wJPekDORPd&CBLk}8q7LX3^u;l=l*jifzoJE#RyRkX zjJB5=ugH5M52MGMtU45e8yTatkEfN%r-&o(j7Qq@@s2ZyL3;n5rA(21Xa}}TH*8g` zUOC}tlOuqzu#6Z`Fi)0>6)AqclJ>?k!&9dr0s1=8*EhYm?nwPP1|eP@6qCWo!sZ<0 z{b7s@%Dg4rmz~E%Y0yVpWgR5IlgtbbCl$xxh6lfhw$WiqZXy6HkRZe>O0q(MziF^1 znl>JXWV%bnpK~&(=tJoIpd~NNcjB=ocI--Ca4KEz3hk+MTr*jcd(QS}@8HDyl0WYh7>6Y8pyodwK zKJ(U$umZ>eR$qJ~PC||8cT_oozsp3qYSE09Teli`ezC;M7Uea2s%i$iyY3G%5&fb} zUxUxCjE04T!6})zshlLm_`I1PjUTzV5hB zI-oo9iX$$i+(igGGhNXs>bGHys@ABzAdEp)X5jblNO_7X@-TMf(|1t}S(xbCCt@Ku zO6};v+3C3zSf2^xAQ?=n{NKLQUXglV`jL*Xc|IZQrF8}#Wg#r{R!6)S$_>w4c>NI! z4#iHAFEkTjymNniR7Uk==qi4Wd)07V_25iu=%%v%5`Q{g*S!{Ntaqf^X2#rESdCqn_s(0OOjJSFup}r_4aMqZ8$M8))LFi>!o9x1`IiUL znPe}`Xh`G}35#~!^7>--56m#FZ%T11$^lTxTvQ~xMfH^A-dJdo!_o!dB5*avtq2Nq zV_1+ezjvwqj8LGPHNnw*_WR!NK#tFrwy7D;Q`L0S*n{%Ay?R+sxCCqveQh&Nt zOOiXPElATt_(8-fpVMzq5*}W&Bn24=_szNJXL%r0{oK7Oa-r8m{sDUQfR0lVRA76A z;tSW^4s>ilh%R4gI6EB}oUW{y9^N6P=cOaAhok`$MBNVszNQya!)_U^9TJa;tvWx&+@lcB+0t9N^A1zEK!1LoYdPzt z%5w#h;$UHCe2)+0Z`6Y^F%y|Bg3J^AT9mUIO>Xue1t$ZF_U|g@jh=rk$ay< z+2tN<9iUF_YdAGCWYPOcVQD$E2hdYZN$I?D{j7Y#LWyQWw88E`^@m!iI=l<{SIy`# z1?2T&vx$i&JcaN}=6L##j})X$%UBbtJQBU?L$XKcIv1FkdD3Xxx|KOeR`2o}@*TMf zqAnUC5XlhCW$=-vrMA%1hnadOOuZNlN%s$;sm8tjV!?xz0BLi6L07P|vk6GnCi#u#< zmCXH5?i7w8_UM70G+DfxK$%g$DFB$k#A=6wcwVq z*80#ym!mEg3dQw&Wybw(r25I%SA_8x5#V7H8Sv$bNireBs^oBK#FXSdPR%-vu#&0b z@yTd4$=n}yf?`zti^Z&**jK{2$@yNpMgSsCbP@4bGB^|9X3tvv?WIiPD8EI!9xL7; zB$P6GzuRED!R@SP(>=WTG)`)$PHRJHdI?0YG${_FQ+)7_gb_08fge{9Gw;q4Od9~w zoC8-di;H(`s_q_Al#XpsC0#-1P4?2od(lPy$PpNK2XPmp+jfd$FKM<6q3iAxQ6zle zQdaV={G?V5kNqe}0~z8MR6oi!8@01_)ASqYECesi=DIU`#nGK7%FM;A8{uYl%{X_r zv6yg&#szy&KVUin8!AFPrT#WE%wHq^7HDJcfbCV}H?!5|Q2h$R^{*kx3dp3MM5FEl zpLwfC6MYogUaojnmClXnJfoW5R+rdn(pzq5zS6`Buu@N#V|VIh9>!C@Yba`!3A`e=nm{U z|2ET^2o`YmmhWT<*Igl_1g$S~2jdh_W7Dwv{$xMNLaF^Hp2@KF6cP!XK&4ui-@Q`8 z`&vjV!i@_;NfCCj2k9eJ0#-~{ZQx5R{5{GNC&KKyP1jOqs5r5>!unwTthaa&@Ivdk zv8W7fjpDYuf%Y zkrXj#cZE_erNbERYV5mE*)20Yz!EP&A0ebK^dQE$_k=SYxX9R-V?Y*4l1F|s)CmsK zp9*i~8t>tZI5@t)V7Snv8)=YK%L$lj1QM+ED$xqMEVc)de*nRp#%G8q>-?vZ0Ys=n z10OSZd$`9N$B_Lidp6S(KLXWIRw_#APz07jq0Ylq%pD(`Kgh``mQc=T5uQ|uk;$(v z%`HrYH`jbJ^^`;flJ)1#c;1|VfwTiE=*84v%X@$#b|6d;*vt03JQ=g%F^-|zZqk^z zJK}CO8d>)}xv3tInB9C*l$CXX94(bcMKlMk3$ZvU3MY!y73}bVljz>gQtd7K@@@Tl zm9x4o9fpc;Cborh0gn8a0p-ne1>-Yek=dUPiOU%3L6kOobZD`Job}N%7B_xjZ@?Z{E%0dI98ygzBoO{ zmbx=)F3!vhql}&;q+4D;PhutH*1ePWq(tvD`5rKWK*kj#KexL21&waJ6C22?onH6@ zj|Yj=G}<&lNPPqnaMa0-*v*SwsJu~3K5hbk*LiF0g`a;g~&@wf?s zSDS0zo=eeaLT6i#mg4%5XeRpf$Y!d1J3%KsW(Gc>uzOA;kG$VYa z&{UetJw_hrQSGRPk`hS!Q+9x*TMO^5%Dh)3LG;?0&?7tgGS&`r*7T(zAz zdngJMngL_J2HPk<_&ae^-3!H-@Fq7cMpDpv4R7QJv4!a=NBin+l{eEsHIk{x@2$q( z3c4|1WAr>zwtFW+4$~Q5Hud5u929hZc}t>bXlU#E2I*q}cj}Sj+cTRARx)Rc8|mR) zEy6CqakZBKX)`aN*g?6ksi7dzOO8%bNimkNGGbQvpx!;QjIB-{p=WHYhb{^H;rec+ zw_OXpyc69v7n(wI__?rx`L2ytwgaNhSJ|zKuBm!_F4qk#9J{Z$BWnBep&<0!R=d&| zmt3_P)^DS{*lc+qzk~Iw>%@}_i0F?p?3GNM{7dQB5ScKpIB@MkSJFCfzS)?o`jpbF zk|oA3F`(1~UexT*kMk*)W7yMyPGACb0E7Pt&(uq=L81UJ<(Dq)52=92m3lW%aM91j zb`>&{V113eqDyhWZg==GPr_n90v z_e6}8-<3TaXqxj=c<=Muf9wA5&ZemUkOsjq?>k?0zf|&Y7mQ3OTfqfC(k&UyedL2-YL~8w)t55OUt54Pc(4$ET|EHxhMd&_iN}rPpKkq282kTF;jW)%`Bx)jt9|!oM|fv;kfaRH$$ZVkG?Ui9>=zmS zhZp{zO8hs|fTU?ww*ghJ_-K!x1?Jpol?i!|J58YBluv87uRj0b=ysZa2bKROi2o2q zz0@Xz~o{|Cr6?mtvOQ>36?LSQ(9YS8x2>29uxuTJi*uVjNH z`b3~9{^o!8=s#minugjI;dWAj0ZkCrjKAgUDzKXi=cc!ET=C~knWhd9uKnjh;|KNg ztMw6X_tGTN@vB9CKIRsLKyftRB)+7_`+S1>H#BT@-oD-cXLQs2s^Ng7F#dr^z;zhE z+?j4FRoG>83(_h3(sB#6l&0p8`pW%UqSgxg6_NFPW?P-dpxhvs0VSM>tsw%>l zxD6+KZo7OyN&^1>M}(wGU}D4kpefkdaQ<68`3F1Ji7uCW8;egsZ7a|;6KI?QG~W52 z0Zy>YBJ4{ut3O6E_nK@e}8ucz#eMU@wk);9p*E z$Iyz9DT1RY#ALZer;$jOI@fADL4CJ#({yJkouDiJ`teTQX&NU*ZYNpp;Az7Da@rU{ z!#;@z9N}-?gE@^Se&|p9Y}a2C*}KX4FM*Efr-4sLsGo-e08;w*L2(%2zYz!A#?JHx;wDE%o0)+6FFT*xn`yM(c&HmhhW>r-+a;~p zq2}mme$d45>*w|lE@+zJdH|_@gwT0}-|4gEzse;4jXXCf80$8kPy(7lMS{Y#y7M0w zuXoe&v`N$a!vox()qzi$sefI?|I*r?W^kScBq;G0O)IcZ5<*Ie{M*{rN6)9g#4)KX z%U3D1pYdpW@c->-3aN3Dp9~fXG|Jyl;HxqyAUN@l<^OlNnS=rj68@J=>x-HD|8Qpi zhmmWzf2;gQ`~Oh=%qXHq|0kme;vbIF|1G0P1?@AVh%}+#pNt}9;LVP((Z9&WKBa`v zKG#VjlSTPM)Ge7@70){$Uq4nC+lO~?fVtVcj9j8)LzLQ!VX^n)Byx)tiL?C$Lve9e z&%azcS7v!Z&A!1PQBl-8jhj+xm_;BQN_ugA6K<=~ATxXC*5qculw-KYFUaRMSNR&m z1E$fElYy95$Cz_TRCMMFTB+m)aMdq+bbrW1l`2YZ|@Uc~ApsxV1W z_%T>zChZN(>TtumyO!^!Gy8y2RYpqp7bYYt3Lf?W@32y%A2l?g_W;pKgC9|Ms!+)2 z1=Ei<%hpYP4XoqR7pkSJZ`gRlskO*^l-v1|XQm!8tw90O;>8NcS|O6pMU+E&XWn+O zf2vXFc?^}13Ke0ZCG4FTD7`APN@k{9{+a_Y_I11R z8ls^H`S{`aregCQTdxhGyk(SB6WVWEjz=i)W3Q4&r2j8~Ie$am}d5rl+Dy z417pZ+gU|>duZV}$4U%gsKs3OaWi|l_;^X@8J>%=1+>-}`=CmFU>$ z{b3kkq*o_^_Bt#JUs)Ltv4Jd<$(5iHj*q zcOO%Hpxk_Q&zC}0zLoBA3)v0G7tqQ_^u z8uf&t)>ePjEW!O6PH6?UI#NTgw5-YCq|CzX0v8%WQRw6YoYUv2P)Oe8OOm<0J1*`X>;F%S_niy zsx_PdQxrpJ4 zt1)=hmx~|iU@Qw2s0<4epcPDE14WzCzQ(i7BbjUS&7p1_HoQ6YK&h&6M+jD32jU@B z2Um)wvZlLPS+=G~4x@o%DB0a}(sIRFm1*H0rr`oYZMeO%?h6EI>e7Lj7_Jpm7|=uV z7)u`Z&`gN643~ewlD+zS8=H(S4CN9NjqBV$dbIcKHhj5Bf6gZ?e5Pw4gQ}30d?gs`+Lr(g8;fp6_Y-8uAKP!Zo5s@ zsd6UQ{BARzF!o8VUA55eougn^9}eA2lP|6Wj3_>#XlXVCgUvwz>K;{XHF22jTfael zK>78&{NA)ZuTFi1(X)`}Jh+(JFVNO%GXK5|iuUp%>t<*O2y@+Pr^wDYDo|b{5~@qwKmF$wGBmyHHGpe_V<)p5Dh1 z%dCRIW|od**4qHRKlP&#q`(8g3uus{?%l2XjZyneV@VRHcbOIA%p_nXMz3IlMTze- z7GDUl{AafksO$lt&X^Z|NPGq*ew05Q8e^k-eNysVd9rz*JssPnOot~MaDe{2DM&hb z!{m<8pEhYeo2Bk+(^|@LWk05Yozw=MYMEA+bjh}~1t7r9cMIN(M@YuYqd7xMPGGm; zS5j1U5q@#5SnVX6Q&t0faZLAS3Z7y<^Jt#XSsxyzR@m?5E_opi6W~lrSKjrS7l27i z_dzhtnt<}2vEvTUthOm6K8mXNSa}gJg;E>u<(Al;9j9TjmzE*T+wxqa4mo-qRh%!+ z@lJH)AEA=ij_fKg@FYa*fL|Gu(r$AOxTg-?$#d=!LeZ+d@tS0|OiNNq1PjsR+j%OV zVF==-qT=$LUzkrOP0QLnkZYvSs}ffzB#KyuCZ^iXO)IZu7U0zb>m$#-1c>8kyq1Y2tKL~r#5I#haz?jTt*r6o{kD*V7QL< z=Bus&*Q^rSOTCAO2OkRw0Lcg({gswYxV}cZzJ=15sEwd09ng)d*mckZG+A2EEtfsR z)z#8Hi=7ye%YF4?gjcFZ6q|Ab&I1u!)xH~sXCB9GBMEjUd*zD67Pt2oPz4wY1F_+1 z{WatgbCNw;5hW=Hzco<&+^^XvCs?doabyQ;QroVLmKodi_tEnnmRU_n7f(&G{51>0 zJXPu>KZM-9(kfwGp>bJw`TokDE1tC|1UvDiUEbbPD67bs-l{VR8pgY}qPUsgiQKvc zqMOA$8;bmsfN^+MP&5^b!&%44XI5(O=N(GKo#g4R+MF|osa(hl_L{(eC-x^uTY*YLN9S5JO>1+LZk z3LCt|WITckr!$NauxpQSOYYc1uKH(n7Ke`bWQ{_9K9qV=lc=Ci zFdFje!gs>)#fxsDhdwiG5%yg1gE_fOC6bDT z>S3J^sfg7LbQvb69!lsj%+>NjLLgV*B$|JLny=EYmHpD*Fx7pq5bqKn56>P(gb@(t z|1c6DJ|(!dsv*~v*v||6Dr(|0|7B*106YaN8j`;<=tafJ*K(`H;bmnO`;r(dMHphJ1 zZEPGNSEH<{yGy=a+b2Lj4OmdV0vMo*B-|N_?%ICnLm4)q?`hrHLv{Zd=ei6P^?vlu zCs9-uL8)Hw44GK|f=yk;JVzVb7~VAttM?CsW7Vot7b=s5>AvI(R-$2_+Amy3d5}7sSx;P=D(El)LSBBwtM;I^gOl$S0ag@KS=K-8)#Mcl|KU zr39)eaqz(gZQox>R;z?{iTtt73O?0FuuT<84zD$z?Fl4-^96ySs;PXSr8KMwk(^1L zOXITZOzPY}%Yjg#C1XO!PaRNH9q_y%iaki`ln;U2!h}IM;bw(}tA~&g{~U@f=D(s! zx%=bCSHb(CRF`7jf&Ib^TdbkbtZS82n9aXnD|S4+ll!V-tFmzuwpkQ-ehmyp`| zJ~D32Sj{iQ+5+X@j}kXl1(1*N_t=!PP36Z2e!#(=U#jZ=qJvH=(zbk5qMJXQ(`-Tx zo>a>hRmNb+(~XTh&nJhTQHDdb5fGmpTJ2;XND+Ddv+r>;Q!yY&Pn?NW-~kM5wg^tr z4$O<()yWoGg-Gm7LVRucWGy$K%>HnxKsvjUF1Es%Tq*BW>tcNxp(e}Ot!vSMCO)bG z>=_(zFyBb_`<}t~UJX`oYtz}nVC{jG#T4|@;cAYRE70wHe^c2=>;2w|Ff12!*k02| zqnb0PbH*kQ3o8T~W44)bG3*A>nHbJSgn5gkqn#EVh9W}fhG{T-tf@22i0e64pmhPC z<_lpn1uR5_jeI&7bg}X=3bXiJeQka^?cV9!yFE6`oNf?Ppv#E9wPgq>*^_%s%sP1S zow|lm4M)l(Z32w>*PE8jFm;+Tg4RQthapJeHNFp9{5Cv4y7goD&dx^1>Cxo~BkW`mLR;le+pro&{9>a#*$b=#*B5_cj|X8-z1r=;I@?r%$Lhx*(Z@r0yMGrP-m+Pd5fX79@PnI>h1=FKo}ibi0L z9s00}q8@0ze;sZlbLw*JPChEfTuerS>#_~EC^lCLx{Ew9Iqvw%e8A%sH^T&t!*x2B zLx3m+qtJVADLni~WGv?hessIthvWh-(F4+jJTOn>Wg*<*la!?6|0C)x{F?s4zyC36 zG#j0c9^KtAI%IT9mqX2)MYw*ysku#RdTSk&@=R*EK(6kawJ&$4+01R3o%)tJULC>4)PDdVE5kc)!+#{I z_ogXEkc!8;@_#Kb*w{E&xc~oo|GzCT2z&DXeP<)2Wbk<0U)XK^KMM@`eZFzFL*u5m zZfW)7U;lxa4T_h5cK!_5hQ^ewPI4GIid|;VXI6K8{pvGv7ZcenmP@^^5PB#k>5TTC6+!zu|w` zQrxm(9PaVbAMH3HSCrxF=C{ElUF@+BFQEiR0ON~6Bo}{AM$e|R-iqP`BD)6po#mz; z9`90Ak`c033wVDV*Z#z(TRvlc9MCrN!1%u8V~H7cqY@OPW#6LfCax+{~R^z5`nsEug?Wo zZbbM8A{>K|`d_@9S4u-ic}y%s=?ST&q$3a3DJ2?d3L5AahHRrLZwGLSiEaATzJKHE zeqbo;1?tUfpeuigLQi*p{nj-m$<-AaFS^mw{QN+hYqf8f{b9;nX92IZ0>`lanUaR` zVae*jlO1a~@pA_W4tnyxyWKYV%r?)a)Au>Eh0opc;*I1Ez0x=0FFcb+^ynBC!PWV% zB~b0(V=FXniOtp(lTvrf7V3wXl$4^Z+JAt7i1+{Kas*pE)QvwQ zm4lVzXTKY0V1hM2JdZOQ;br_r#a=Xq>*xpF4GzZ2L{{B35AQKKSUH^D;vQht37Epm z2D`&M6v=CykydJ-l&keYwg=y50@kYX|25r}eNJx_`umsm1sbG=_6Q;KS#$jlfKC16 zpJ9!_Gs_`o9ckBj3bh_pi<(->+z-spe8lv0KL*LJ*!La@nU-1+QzdHZ$b^}@(^3YV zQ*D>mHwZN^Yjpk-+Rz*1jp-^t`4{cxImd&U8YSgEEAdCg~Zi*?H=CTu+xrOnkn)wDQJkdaVz5JfD3*;~3C@wkZ9yT1L$4i4RX zN>zN*I*5D<#J*wyj6m2@xhK7TrM&)SfD;bkeDXcYlu*MWOrBRqG4k`)Kq)DgQRP#2 zUUfD3EXk!J8R^qFq;Vu@s7n$bYOvy@nwIb^4VC@Cybk$oIL9FIGzl5!TXi)SC_7=6 z)4kDjaY4Orcw^v@fc>(GLylrQ92=yW1pY_8K`yE&XWSo7KL5st3?#EIjXpP#K5G1_ zRd}%RbY??pSs;C{QpNnpD;SOPx~=z)~#1z|fR_@lGvee>m6sE-L*u!Xj{ zvuCxI(hYI(@|CdoAs^j z_%gq&s{D=sch~8LE0<)Dk`Pl;&2%&8`JDa@`x)5wkpca|)|cK)>R}M4MRM`v%|7(B z=7|4an&(WX-{1cLYWe|EfT=C*CP^=yObe~%&6Nlq0guSrL4qv_Sw@*GLh&Q$06&c7 zH65+GgSHP^s)hvbfZL9Odmea6wrrr|>17_r4P2-1Jr2yqfFbm+9#Tdly;y|`)A*do zo~-D+C%LZRv#r7Y^?o*KO^t7jBB6yz=>Uv z*}PWzl$Ol`pf*J(v$4jv0)2uYzLfOn1y(FN2K16hRv7zjVGAlF4CcCSbc^%@@#WEB zGcCobk_TIta>ho;YTYI-#&5(}@%{SS3p6G52qx@!D?+Z?2sH5rmjCjv6)LR$;a%D* z{f+oA-=yYWm`)`=K#-1_)YqCuTYuh_WZ=m6_yA#~x&~6tMT?Ebk5lKhKM#PM8Gdn( zgHXzj6rHN-Tmv&v_-og~v))QDfl-bvt+oULf!&c)%I7H7sZKGQ-pBiZ(#1j=Q}D)c zo|R{R!U^}Pw&}6n&_LxbWp08*u|o@ieJo>@3|`!5YW^r${Amy5H3UgxnjIoa^L&fu z+agFxvvlBP($HT;3Ny(U6d4kmH#&1o{z?SI=7zI%O^~TArdMy`;tDO!pLN*}#O1K= z!Hy@lKhOIop>9eWtAq3>o#KbLb9rTQvBsme#f*J52;TMHl7-gM{tK_u|Mt?rIbVj( zy^%|n2ERS}ISG~f5R*qQ>z5)mCno}GT2UmA91!F(Xz@xIo55&-g=&dIgrZ!zbk2!% zH@S_b6G)h2`LF1UVW7P(?@kTM<>V8i>!9=hq|Zv^Lm9A0#RchATQ<&XCIz33Whr8l!qB^yxTyd}bl7mhJ0+ zHI1fA`rKd?A5ectJ$aL+-B;=E?7)i7%kh_A*6nex@?VTpL0bvg$oSLU8sCcNnYaIF zg-=acL+D2c*^o(7Wjs9`M7@{D-J+Y9`BI`zV)YM(C{t`pI8K(8oaJhgUB8@C~jZe5Ou_-^n;Uw%>=N{ zbGq#m?n7Q?^-{zjVQtfyyNJT!TrC3t0QS1yvUk-=f&Lm?-t$b|jNqkD-cJ&$<{6*H zJRP0Kv%y?Gh}scMG6XP+MdAd1Z6sOy*W8uIx>=C-;NLxyeo^v6HJfJu!A0UtrqC%996rN*vNW_wyIfW~H?4C(wJwX1VdI$kXQIY8)UP z{_)G6dA1AS|3km6re^8%eYJjuoDOW3AXR^!vaKe-Wsu zb1LD?W-y9R6zWyL$=>MALz&`t!T}l2YMhs+BI1->y6^hKekJ4>yJhVY&R=6ba*w#B zI@VU--iuMeV}0|(*=!Tp7yynflmZZ-sW&xkQW743ftz?ZpsbNelIp&Vx~ZJ9%>MwV zF}`>Ye_zafIH{Iyu)2xx}}hz;Hmu7#ZY8m}%2 zU_qx1f)bMJp8%FqQxa|d^0Tc3KBr}%k;@BG0l?cF#Df9Chb5{YL*8c~<=#Pb_gpYmKMmV;Vv7_H_1}7()rJiX#-dmQH6G*&9tY=rmL#sje3v1avo;Jdi$|PAES_= zsxQUh(6<4kAXdJ|85w1EDjY3|2eeUK!spgd9*5slu{uC*64tea!>mJGIxuIHV_(Xe z?w=%WjAjzN%s!R>i%GHbMXhFo263IL(u$&NQFRJmou2eKZBI~H4hiYRwn_yG>u(@8 z)A%a=O$kar!rOSQde+?%(|mG(VwuE6ksL@xIKH>OqmzvDX8e%{|1u^T`erP%TG)Z2 zjTt(74qzP_E4hG|JaaAPqo}_UBVBu_4%;ufDSs=59@`}PE?<3+mPlPgcc!LRTY6@% z0B4v!zqR-cnUn=FfW_W3RMLOs8KZ%DfR3dGqsaciy=66Ca}hj6zX%8SQ@c|Er!&61 zl;MwW#}Ls%V*%5o;f%EO6L+|5s@SVsKdVDbdnr0WC#)5@U%`|2BwAReNDqBF-Ep2x z%?3!avZ~4x%6FSkQn;V-l+1Bg`Wm(eg;n(iipB%i{^KfsD_5SJ)ZxXkn3VlvLI&M++N17nZ8^RE>+H$ z@@qF@)0DE-;oLGS7T)=E>CZ~^;L~OC*F9G{dhkxT$PoCKh=#KSyA!fcwKlTF!S=?R zQup@FiL%qxy=hGBe&R5^Vrv<7l0w3h<203QN0lRE6ce}va^I% zM)z6)@-Rd~wPyYOlYkSZRn{1|tvUL&_q4QBG@b#mm+_HPxE-c0=Buevl5A`yafd0qwhrq=Wjv_i zUobvi4rr0BQDX?ExAJx@lbnR)nhL!4OLA)}4qZYZW~8U}}$(;3L`G?p9e#swuOu&o5z6uCCRSmM@yOrLq(EkQG@9#4J^FXOe9X`ia28>i?(=N8GUV%U8$aIEGW)$dB|u4eJ~XAL?@4sA9(HSJuGXm3 zxlwtim<(}u#HUux0u#=Ro)VKp19Uhsk*~}TT~%6`0X~gE>oCc0CYHHEIq^jKF>9bQ z0XUUs4fvDo!AvY}nYyfOM|HT^;#*xt0s6N*0fUY(>X!i=!;)v`fFnWAQ_dxjCZonV zXt}sxgVp=NaUQlRyk};IGcF;PZVt2O2@A-j`~13h+iC{%9rncawp3L#7p+m*;YN@o z*N_(dM!UD5lCAS<@b;KmNfGONF>R_-EZj3`fz4}bCuM0qXQ!lJ$X%=};BR69&2*Zo zUoJvtB1Ryjs2o=WB$i}^Y{0b|p8VCD6QsX@ifakf3D$NVLGHSLj3t?R7iOFqBF(Yl zH7LPP=Z=LKweky!@WohfZ4v|Q?t3oPDHm3gp)JGZUVXm}`bqPc4jEq7eBN}yT@Kui z;-t+Jf^;q^lYPO{_L;m2jj zm#d0YK$c-^#0OdVZLDzvsEDc`TRM7;)3t8DIO*pv`PR{iMhN_rzEpx;DCJWEXO;K( z2C{0X7YCQtTtY0mZBn1&W*$kf`XowS3TXi+I{Q|lkBlm{^QwordW%@|3Pq$DYMpg3 z>*@cQ^TfAKNea#S=S8ih5~;FVKF@hcEM-n>pQ=d>H+HhZtj`Y@iO5YB$MpK05*X$a zEeF8>1&RSli?Q`B=;w{`7;q-z;f9_)>+5M(j40Q z>wrEw!r~qtpv0=V0`jEG~g&`43QVl7jPTTPF4ostE((oD^6Xe|OXmDeHCveBD_tr1(axlHt=-S%vohg^35 z;f>_=2DieGpmT%Nh-nlLzIf@z62_i;<*b5YBZW1hj9VApUJ_|$p)xIA*$s>YI9uBJ zbr~k@!5iH@2-qx?9i<8)I1}0$VR{s)hp{byXPXv)MmOK>WOM{O{H+PVa=~l zVd8%=-w?@m3LI^lOV~!8?%8iypnjQQxYqHHnI_AV3S)FmFL?qPuo76^jX((U7HPbG zM2V(aDucx5VWsivRz7|NR&{3eii(f^1Wmxs8`C4zU*5wQ+!2Tvt7jrx0qX%wQYOE2 zD8m^Ab$a?;jghkY;QfL56;8w1Y6TJRcYXi>IejSb;fu#t)lZhTj+sM>Pn`Fqwe_G3 zGl>@&2lzcKno8|ODxo#EELJ|HXAXESOLotd>4{g=N&APhEwTUXe-Rjhrc$q#5^^SQ zqhH@MckM`Eq9k;a(3-3~{US@ux3X6SKj@9ND)LaJB7QV&g3YxGT(d!=o>mzSCa+C+ zdPu!fcbxQws{sU}qX;_^OWi&lZ&N-G365|=r5fpUl0jB5keUZz6ua6>fa37in?dYq z1cN1BRB)VElU?y`kj|x!RQ5Ud&$pWAS&eILzS_mnUT_Ywq@0q{=6K?@ZfOW3@>4Q7 z!S5{vLp0PAu_{T%mUy(_=40WX?4$h}H-u)y0VPXDE2->d_U`K&6#AM!ERAL(HRf+~V_ioQ~37e8$TFdb_phDkA~+qk<^1 z`?U73-M4v$x$lMnnU*Fj1@?bbPNgdyJLfT4Z2R>3r}u}ff_HJyG?@Z60&3N4a1~pO zfoWWdt0sfWUl9jR^L=CD7izL`QP<7kTw`}UZuZLifEuiTn%hVT<((UkP3vg;Mk)4R z`@bu%qx&+=0i76|k{gT1O(R*X7B1!C?+8>N&mu{>Y*Gczd8$Wwv4E|vkwOeFFHTBe@jWs;X!xOiU9)2R3~4>XnkhbD{mur;X9AJTsR3p88shq z5NJJt+oT_sl@@MzSU5PlYqCr7GAjm|TdfCQ466q*gPOfvp^RPgdyjIRA zgkz(Ea`asnm7iRc1Cb#<=e@Exjo51uRxUP{OZ{07{v-X{bcnmNWrODYuKcY}NOUUb zl0Nm(`b{1O>~pc^W_v0iOhAyPk*r+q=KTfo$~{I`Pf_#}md;0sPTBhSwO@KTEZTOc z`3+U$Vx7KTAGFi`REo(c?98~pg)^Qt0wN&-*g3JdIK4q@61JCCK5c}SYTIB(Y%;~MkY-(B#y^vd|{Ke57!At5+v zs`5AVY$+fw%U`XY1w4|GPbT80Bf{d$9n5DpU9lj2>#k8Qc)<f$~l^T5fdX~0e{>Oe;^MlZ2KlTD=TK^w)|KENzNE++6m z(5J;WK5G1Cp|rj9^Ipx>!~wxt*su_PHP#SJWAeh~CXPNH#B`G$PAbEw6u6t*EI0QR zz0@D_#L{O7Phxg|+gi;%f@sV}r>28-n6S~DBjVD-lka!>Z*GZb{xDi)RxyreX;dsQ@^!CIl8#p|aVQ~}RiOi%f!Ryk&USIk8qc1$}uuEUxp z4gdbteOay9h-Vg_$&FVuFEun@;&-C#(LF7*ab!-$BDS2;NsoB(E(2?Z9p#p!&g*Dh zV|4Fv%;M4GCyU?2o070;iz3mo^}U`++816gmelR)VH$}ZL53Z#1j1@_91KyrP(Qpl zG5@#Z?6x#URypNT9&Y!a0JbwiWt_g$eoUl49nxj{lq0TGB!sG)=AmZg^mf84(EHON&05lCA9|SN?Ct`lKp(70VoF->>Tf8-)Qi7; zcifwsom6Vdd)$3eo-*F~b|?ovcl6jshng%hFb6B`ZwIs|&D zJm2Q=w*DE&#{PjQm!xcp+o5WJSDxfk*jTTtbTjE#31s@=7X}br47++tqUv)l{P016 zf2Fp3)PzjVv@_%Ff)R$bE?;u@8K`HGz6UH)RKKTP=QXwrG`$R>ca?L@y;a;kG{ePr zJrg?D<^w1`p3)>~9F^tf6OIJR)Izq)ICfDgY%*OvYV*sJk{*v1^d(IAmP!)6I{8_T z4p?Hfhy|~H9!H4WsPd=%Z#|EJi0^lG7D4WWKJg4p3TkiD!#_E{O;2i)P4LeVU8g{} zd8WBwmOUbTzV7lM!Cu@KcO=mi->c}A5K<3O5|MG{O~pw=RTOMUw$=EC(nbQ4WfC$r z9y_`=(VVpUFL8~i6hvbTE;UYNO?VDkHJwcvcXMP_`DF?sdP~&p3i7VYY#PI@GKZC9NkUu?bJ<~ zv&EChxNL&k(AbR%qS>6Ivba__=0bf1jHcI1B{sGyck5k#6AWZ|?4L#v!6c1>9eV?l zuxA8Zsrlzt=D1N$bot`bMsfJ;4mgLu#+D7AM6D!u8ZK$s7cuo+#)`vdKEQ7I)#I^I z)32B0AzaO~E=n&{ab5;B?c*sCm>_wnEIIfIE!xBgier(;klQP*2sCs;wPim(HoJ+IClN8xQYi z;xSCa=5EexC54ExzSgRdsu2O_y44KM^Ck0X3N?vlU=iwmD=)(=0|FmMWa*joBO@rd zFP|EJw3FqvTs}>D&fbf=Jf0};B`n5LJYkc6I#N@;3ZtMy0KE0TsUf=cDY~-j!V(#B z%--&p#DxgO^c_BTq>knTNoe4yZ<|`v3jLymcLR0Op3Tv-A{*V9f$I@B_j5WyRZ9+2 z;*NpEKlSGOHp;SmPE`B^<1#?aZI!->>$Z*MlyGuB|J&sH+U{gkh-UavZ%}do;|G*T zaUsji(JyXV@}VkeijJw;A8Mvf`rx=`Y)!^b)(DK@^k1?e>kNrL<<(;u&LeVArWxQl z!jiq4)QW{os3u3GyLYtuveq+Ep5l+X^JBte>xK>0^V>=x96@YPa|0%rT_|`Inck$d z{FqGPdq4vg87v;s7hZ}y!+-onBH{5A6E7h0C7=6J-yNWtfV%_QqMzsAYwN6@#>i5E z8klUeUY3uN0xc*#8Eu*>9~p`$TS1BiDuZaXV$uZkwZ$*nH5nTT8dHaFrMOFed7pu*h&zZF6z#k+|_^vrv zL3ZwkLT53Ncbm|+oc(u?DFag#4rxk7rV5o@uNbBQv5A9A#f8^I-*2Q^Q3YY0(SG_7-lpnEwFnMh?YNmOEFt-$`*P-R2tY zTwvJB>#kT-k!$!_VS1G|%#&;-D@FWfP0D@pakhLsW>C+>$9FO1?g>)fg4~6Oj_-bz zM3T2oWFAzEt$A-pgH=9AC>Q__*POkBVM+sPevzsH&epT}DG^Zn7m?MEgYGNTN8#hU zg$ukmc+x51w@&%<`$b4pz(@C%{>eT5Z5Dj~Xs@xo3Zr-r+~|0Fp8gEkfb3 z-Tli1y_Y2474PfaN{YNS88QpTeNDc6u-P}HUEau;X8!}x4)Avyg$gSXfTbQG>N181 z)2b5UjL!C3{0?0@=|u^fV_VDCKNN6oZmYUg18J|TY^mjqV#6Dm4;!3lEBERD>6FAK zX9}1m7}Yp2tfOu9P+Zz#RoLV)j@~4aX+`Axg z=PB;Jj4v&={bxX46P4VHQ@w8z?zbeh%Q*OwG7PS-w8f-fyt`??rKP-q2k6BkO`I4O zr+SjZuT7Gk$|g?jn^8a7UMArlUWkT_k}!B>HhtU=&?7jd@Y^>?Xds0lj3;2tGO;VP z8eE##U70RkQ5ho`2iU-LXmC`TE;@2dM!(1cmn8!CB`D_Nul?fCpekQu)bN(&`f?*n zPUoUbpJWVfwORs;%XbV@dZsK+E#s`x*w+&XQ$$xVzTabxP*QERlMW%+$n!C}EF@Un zVVFca?dN&;Im9yyF~wNuc`{7wBEGm+Qnf>g4SxAIz+J|?=m!g*yXolAY72r3zfHYv z@@w6{qV>?&{zqJGmjyiOkqX3pOsXs*x8gz2C zeel0wXp9M^e(pJcn4>3QN@JTTi~d`~^U)-DRHvN)w4jve!i|$2Z5|B!ZL-;EDxM2d z;g=Y$FOcy+m}5Q8YX12pNQD#E-0vOOCNT`=h9$bX{daSJh%#1W`CER`ORwhoC~iIP zQJ75Z5u*I3Fu^DBM2nIc^)m!4-gU9CNy;D}JT%D$MpI|0`Euy97QxVVRHmLP-eP%k z9G7WH%N@EuL0?>^^RYgc-%1ar5GZS~{X@*NzsZ}S>^x}0DNThS*jz+8i?9s*!_xv)Zb)x|Ol2ASA*el>#5=fYm`&R78}9 z>5fwJ*F2Ku#_5MnIz9Ww?co8((zARHj&&=QAS@@~;{{}hLWu0?EkK=39sFuJs8~V^ zAzT8wead!B150)8w6YzLt<$_E&_k}a89^@I%?^ZNB3XpssqcyQd9ZQ8MGmGD>SkAX z>@Cqr>>q`da8pZ7ivI&RKXbiW`YKNPYo3J7P>ayQ9p=Hh=VRN9y|zTOUKV&`8Thj2 z@VKvMK03^}GExtYQ4_gVJqeNjC7FYBQ!y7)t2C4xoMRT285TNOy&~0RAkvVsioSbv>DjHgrF_YSKVj;x_Er*scnkW zOgTULt*Ksndb`D?k}J=8>@v}EqdB6p$S_Ly(Et|z1W&w3chX)-M&@}BMfT-&a&xE) zo}ejq9=>k=>j9A>4)924)Tj-f4Qb5sifvTl7y;M?U5z#+UWj6Uz#8c>ZHE%^1w2w>G}f6<0w%h-JP6ptoAFTki6_>cCwE&l z^6FMnnU0;fs1Dn?!y$BAzj|r zxe+@d+=!2xvh?4D&z(XIH;m}@F5X;iDc~{Yx6-yRtH0wlRf*)c7#WOVfA7ZX+vM2d z#rz?UlrznBEo5db{EqgOQ>kWOuSktu`5eiSrgi7`mxB4tJcnE7Cb4E+JjTlo)1#m( zymv7-jdx-*c-tU_H~za#dBnXl%BtoJvi(#sF7HDn+_J7<3|Nj0y&rCY&*YT`!ed+@ z1$@#M1=Bh|OS`pbk6DBR=QE*(JPxCjka7SCa#)w1AWJn3`nBERzB1}WG?n#3gia0FH(?L9<4xTE}o;fOh5#$YL~nh{^uY)hG-FL7oc@#1&yPF9fwv0U1kGBy=W{wT$1a}Q5TcCXZb8&nmq$|i|+ zxqPd}E<0dmRyp}0 z6kTE`_Fw#@(AvlLsp7002*gz6<5+1X-uZn>%w_ZWmX^s8( zAZdFDOG3kbS<&P!5y1qZ_OyZ*$KTHZF#REr{#a(-(0_n37XetWM-P@V^^2#e;yk=e zG^3-dS+?25#^sW5V|^EyeunN0r!*-y8kHZKrJ6coac&l)zs!%Vgq=MC;FJ z$Y3R2GpyImhw{B6=t~BpsNwkUia3|u&r(G9~ z#%{^maCyZ;iG&86L8Q{Vj_9!eZ-EbS0usS2&p3 zcELH{&U>9O#-n2Nu1*X!+~vW!f5l2_Yj3-I&q>+ul(xq9xBZIb5ghVS>%i<{0esNu zML2tcctq)bL70aCK0x}eMe($IMY;@w4yMj(2Mg#$WBq5+h9^;lz*pChU*|H)Hnr0K zr;0j(cD&`k*Dv7k*Z+uk+9TS`X-Ot_CUN<2tZ*P}-6^liy$)Jw@^3s{p5bzmf%oQCU&VWgPJ$n11NV z_=oJB4wnSRBWtZLm7~ne-G9Pd1RrsWev7_9edU7s(J?gSzlk(ukViF9)sQtK7-fNEd)YjW)C#Zyv) zbXk@16G;c@GuOX~=cs_@+|pF!@UnL{)%x_i@PuhRb99kbdyoaiYc4fo@ENyplzV)y zo`at5c74HgE8ZAAv64W^JPAZo1Zq<^;q^2LkOdI6Y}E_S_bu4_iiu~h%i zt3t8HqXz-*jHsN+r>;GZOUfX?ggZ4IGr-}M_pAqgE#SYl(<6gba4&9RqzRU|5?5<< zOS=m}Ttf5EY=;@?#lUOJtZdgv@Rtmy=7}#IS^blFbRYA*PSt+`_84HzyNtu*_1v$l{m822~{hdTB!_f=0)v8wiExNCj%0@46NknW`mg|uZfoN zo|#$H(<*H=c>d`$kFBEJ>>&(rx$1+WfUo`7fr?As}2|P1`GnAXb)@7=D zbV2C zi(PYDmAe$idb!r7Ni?3U>>0NUWFG>4S-kZC#rV)d`UuSto-E8~go8dAx@8uvJb*^6 zs4Y93$;8Y2#PD)_v=u&O1JQPb$jujU z=J4Hblx<&n+n7sTw})~G(;!w<{Jrwoi-F1=-D)mISFz67nl+>DX}MxT8g7CNmQYlGk@hdn45Bb{^@^qx4;rTgI{Eez(SebZpV#&UxNL1{kgatMn{Zx zI@5%9zspVeODs*veG6R+oIK@4AX0@BnHOSwS6kpXZ&Ky)9gt&naSPn1_qq5fH9NDA zy>5tf=~U|qy|YV$U)AAA&zGv8SB?3&;FIB;2c_5L@Fg*HuKcyYv}Mm{a-cm#C+HO5RJCO7`{P1QJ-V_L6I--i=FL2~}p zR8esfZ#xik$1)~|t>7hbBOQ(T7j4`}<)-bUp#WjC44%x4gjmiCEIVAj5$QA!79W3d zmRB-sg&1lAfZaRFj4%!|;k0l#tqe1_M2nCUWn;Kb$P8CmbSWLco`A@*rQ|;VmnHnWS`u)bI z*?Ayr&`g(!%#jq>RtNR${}f6!=-N%Du}pZOfxUlS`BCzS1k%MF0219}bKa;o-C0|3<6*QAzx>kaSzT)jdrxi*7(q&R6pBklmXbd&$s$S{c+ z6`f!+RlnL$yU`P8lWZY4S1G-_U<~|l*oaw_f(G^OsE-WD*$fgu_FA@ zBQsaP_@clA^>`P@P-e4{=OwF4Dz)gfOQ*;y{aSwJO@ymqs+bX1=a2pt&+UlaVw^V( z4Ci5N3{t>5zGaXC+-WlZ>bS=rQ)thygD`Qr}8N#`0uU$Cm#$sjE+t3Z~W+@;|!T$lO{{xsj zu7{e3ntQ{#6e6?B$J}3a!iIPy=dN3b2$N6v-y68>;Co#Z**|BXnDuLEs2#Xqh=zH-8_PD?VGbIwtj* zL&G1nY4~TWe`E8@Wng&$6RE)Or`+4HM=Iza4$w0f^-Ve)=x|r!0_Y=jiN}&Eu&aD; zj>i5mRZva$A0WrR6@c9ajtFrpTCk8!w9E%bz7c|g_hKSr=$_tuMUJD{%3%^hEaqCt zq}Gb%AhwufeNJ1t3^3U)+i%Gwa#?obpe>p#20h=$sY%GtDc^}G>&AM)v+2u%{)*wq z(S@>vpECBRru+y_4~;-&l_DM0-4@f#V?mIj*PWU`8G6w&p#}ITs0;E`lG8#(CB+a0 z($*(2^09}2!I2pTE+NQyBoPh@71hWqC(J!@>bbqk&21NbC6HoZ%9>qN63*UQgbxI*SKgUKF zhfDs=wQf?pFo*QSFF$M{?k+L3;pyY9pgMojpa#FwNg5O-m$AR!nnGu2_S3j3NtH|!uU7789;;7R6+CmOes5xy7QJr{c1LdPWg?OlFWdbm z*Gjndoh%F#zA6~U#sCCCcctbp0DnyQhbOO%W@O&m8*S;7R-kT5Quj&>(xoxlw9Xo`4`{dfkWon})l`Bf_m*oPq#)#@R?^2uheX53*gBa^ zO+8qCO<0|&siH}qm7+Zd<~i|0(&98is>FzBidnHB`o=Y?G4e37 zo{1tiFU~bLP$rpt57$4n$Z7I$_|>Vqv2{evZ$>-nW1*44&_=EYhJhy&v~t1RX2p;c z9P!-Dw|e~bQ+w~yC`dHVMNHkwoNtVkX%up@$0pU5753?%BCV-eQrwY_W*z__uy1Zp z{8x|S_&>AWupYt~-WS|wo`!?0(mV)i(abw8X?G1FB=L)f?VrM8)C|r3_LNBshSHj+ zhCA9tbQT8}Cfw6a7G7VoU8w{zee3?Aqb1kMPE=K{M|I2kG1%}!?5i~{24d;mNjm|>M30SLz`G`mlPmuKDSisU z7SKKjp(1e_R4)NPv?tz_$pSs{{VEbwn_H3}imszO33xbhL9B3a1^_UB=`Sf86@IhE zX%r$mKAd|E+Z7(E?WF)9AlUs!`Gt2Wd9b>ZeN+G~zObPZ;6?uc^&nsg$xg5xj(~3_ Lh9NtL{@4H6IJ;bq literal 0 HcmV?d00001 diff --git a/en/2023.4/assets/freqai_doc_logo.svg b/en/2023.4/assets/freqai_doc_logo.svg new file mode 100644 index 000000000..23b116883 --- /dev/null +++ b/en/2023.4/assets/freqai_doc_logo.svg @@ -0,0 +1,304 @@ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + image/svg+xml + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + FreqAI + + + + + + diff --git a/en/2023.4/assets/freqai_inlier-metric.jpg b/en/2023.4/assets/freqai_inlier-metric.jpg new file mode 100644 index 0000000000000000000000000000000000000000..d9cbbe8a1569ab56130a7e610dd49058d0613dcc GIT binary patch literal 468744 zcmdqJcUV(R*EbqODN0dM=^!E?(mM%NP>`;o^di02(7Oc?=|y^zj`S8Fp(r4|haN&y zAR$!gC4gVBJ@@mrbA9i1&Oc}Tv1iT9n!VRrvomGQY=2Ds_yV}Bte~U-ARqt$2=M;^ zKM(+U0O7fFXTXQ^cw8X4aN+#<3m1uriAcyUl964ycc;ZbCqauXFwa>4giV*T^rCU;DofKiU9SNQkmYLkS6P0nS|^ zAiP5GqXWQrCguf#GkO06V!Vh%BzqQ1m4ricH_dtCOvTN^fpe?Ad$3Pq!<$k*}g ze+WaNvLgGU@E^@Vp|SFh;-4g%_pRXbpA;q2#olL+uHjoSr2qB3BH4xc(T4QOvm*Gw zr|zAoW;yUL#-H5y%Q-H#2k(~tr10sxY|oNFjQsge@p)Hhd4XctY90S7{zcu}`7p}y z2mrYF%oLpBtCDe)tvg3>iAp6d|D455S1)h@*|MABl?>m>eM|!1AxIrPhLGn1Nc6Gf@@8kCV?UeOs&qWGiL=V z?$?~o?KZgAoZ(wWEopx_W)eh1p#>j z#k86S3!gPw6<*5!{hV{VV%!^6_kdi5svq$tRW05M4KC9dLKv$QgB|pw!jkdk7*m6? zC%Wfm`c#C%UgOh*hVwQzRBkr~nFu|X^?5PwU}Yhly{3cBlZoP>7IpJ}oV$F#cd>X1 zG2zm9_wphCefmPGtZZZNavvcz3NlbLQ@1QpaSoMGA{y*3{L0hyb(Gqbh>UAjpRL__ zlN>V_<%R~;%h0c8eL7)My-A}fym&I4cBNp#p5#=PB;Yv-u_mavWR2kB3h~42=^!{H z0Ekepix5hkX4~91->jM%UwTfS((4+=Q+C*ZP2%3Umq{4q!xc$6?8XAQN+;4r#dTq@ zOB0BCAN_FOkX?~SEo2K#Dx466*X(y7*^+1Q->!V^UTx2s;d2My+o(gmueupu_b4KY zKmX9<^-52M6(BB@-Yt*BLq>SlNrV zo`*cQhuoesPo7Z{09s3E31>Eb&71}%I=u$@MGDY+#OQ4dYa7xsXgzh|({fqZu}v#0 zg@COzWh)zJJ+HXs5PWJzb|9f((M|6>x`@RA_?4JKt6MRC+k_e%bN{UH90Yh}^pAs3yN3-$oTW1Cj)nd;S36t0SmU{9kxR7eEo@V#(5O+lW`2-r^ z?-9!7vDpK@kyTB%ST@n?F#8IbF76@2?gYJ&5kGqB*#Doy9>4BCbfO)gef3fF{7@{e=5Lhe}mRz zv#q(go6!|qn#-tMfbD9I@9Z+(#F%V;otm|nZaVK7Eo?zF@-p?XfBCMT%%sB7r2y`= z7A}b!1WR+S!n}2(x7))(`;j=Wu-eS9Z#j?a8<2f->T6NPpqdaIDs=G6pO^7Z%utws zeX~?SkLnKbjD!U;JmBY0d>9%fvWC=|@h-`C4#`iOx|?1+yzdc8=a#N7!~jD=fI`7O zz9t_TGNvcJbC{!`nS#7IQPJHol*qZA?lr)e=w|!&;HI@p=#qL_ZMyaZ1P2`QM4U2& zLmE3{`@t6?-ntmeZRox4LDp-S09gyMn1_82GuPCvUSC;L9d;7Ddan=Wx5+>!ipWGh zWTv{Km-_T%dx;n``FKMIoCF%>-qZ{Fr)OJZ)C!f%x_ zWu4-rc9nVs`AdQxMu~Zlf)6*&)GFZ7RvqWD9)NJ^$nbdwKuQ}@Cb-8g``dySvMzfX z#xk#6QQ)3kcJOuSaow|e#KJoTfr>}h3rl^+|5AtR;aMJP%S$05k?B@n4x`L?8ipE7 zsm$k@YWwLi%pMfwHi%L%Z@R;%M1rhlMy=ZFIfgfEC;TzqZEi^_>$ldqO>)6IjVw0| zFw=DR>NYnlu)kL{SC}h4y7)n0=VW9ssap0=(l75>p~_-Q(^~+5G8gOq>N0>}|4ZZJvZ&|3%6~Emv-(@l^zh}9 zN}jtkWx-F-%V@;#e*6o7LJidYvS(y|n##1`3jlCAqTc{0U-A+EBvWHWy46uI=3dax z8~hI!;dvMEjq@wDjN-4gDHt6WPcuYjXT^9@gfCgUaYaER`sKtA$WG4RqiP_axlgPO z3@kcFGpQctf>KI*i-?bce)HcGl1M*C$8UJy5FGkM& zF+q8YJQRtF<>Q`Oibh*{FImvNi6RO=ad`Csyt+LDqZtj;e0#e+XV!gO!QuVALhVM4 z;@(mjK1SDtfFNMa8CM%1Ibw-eG~+f}&0tlejPTaNidlJ?h~KnTwcMylhU=c>$~CH% zE3Nkey!#6^Dm4Mt}rKp@1FH->eagWolnwd@M|KwBv|z+jBAYU+&&*wX#YKBRnKmp(q;ZSY$^8(xw+GZIb2hiZTZtJ1XVj<+QE!a=wM z^#lyCTg55s{?EoVN=&OKaa}IzQxOK1h zTX^3k93eC4e*ELM4A|$40+)TO%}}fK2Z*Kzh{Ef8c)b7B>;z2r1Vws;zZ?5}Kb&RI zo=Mm8;G@}FBs$H429kmsO`sw{giatrHSlTjtMnJz_z9$8H+6SPxoMP#N znbTQ0_KwGQ^c^XVz*VQ)gcSo40aQzrHlnNwT$_FX(wl)Ar2>tt_+Fko7D(uxQQ23f z=+9AsIA{OFa^}o3gd0N-sUHCV#6zvpK>)z@yNG8gPpNJKm?&+TZeTt`XKGKc?{ADB z?ue9MJ-Irj>AY}SKZ(A49v2ud4)Kf;79w4Ko~LkwiuhGRKB?G~n^A4MI$s8}d&b3U zb|kcB)VbEwjIHrwhKPSQq0xLMW}KWg9;V3a;HN#4+L-hL%B3n?G+`9~FqLO{stFqA zqugDEC@cS6Ey^a&FMWe$UH|Z+HByF}ZcKB0M=duPm#ZwynWTvv6vx6MlfR`v)J6u~ z*FI=-T=7^hsblaTOT-Zj`X;UwU5+->LH#%F|0{dM^F3Mr$N!N9uy~q{XCUrn)03vY z`HPi!WC!}<7YR65reX{R{Q&SBtF9ar)Ng!}l|GU7KLQYJZ*)MZelz{Q9T9xlSI$Zc zIL0qee+KCj-bzzEMX@D{CEeAMpu73@RA6jp;NFpt^bXeoanjobeELRS3~{4VB8`>HlT0MGv>3Um14L z_r80)ZQPgi;8w{asHo;e%hbVsGJzDkG#Y^d#r5WQd}J)i`^?=0Z~B66Ou#>2V3j3SmYPXOo+;TU@k) zG6fg}{gIRyz66!UE7QI*)M?b>r}_z>05EG~hA-guhu~*xe`)Vsw3+ArP{B=n;(r%z z9KC?80Aaa=S(A5Zr8Jz&b&YLP?wMugJZpe^fc1N$@4ttzE7K?K|1jt&=wHrmX87Epe;h#y*T80-$-T3J~Es*ZP`eE3h! zkH7c$e;^2H6*+UL_on+dOCIBP+8mSW`kA-}G%3?&?87A)$+JJ1d+@mBhPF(T8T!pZ zpR^mThTO%7l_h*{B#)V_PRStYG28aNQrrUP$NBG$yzV&JGXbZGTul;s|)ZcDfI;u`n5{RHPx zj;UxLu7)5w*V$yBESSxY=dTt7ySr};&A#qp=4FhOV(<%6d>~xP&c;W-Y(nrw_3k93 zQE{lsgwsDtAg}s$0e&#{!!gA1O;s3hc=Efkq!A9ykU3skrp$7))Nd{FWm4yLOpA;F z01o#xMJpaf%rdMDv*BCeIoU=L6;=wk_e}KXc#eMtm_sc{a%TH*Q99=>M_N2CF@A}mNLLL|6@v`W$P6YRh*JDUWxTzUsES z`|-demZE#x(y_cX1$uhYCV))eebc^JXoq`k29w%;*0)&5vQJ7g5btoO_4DK%|6zWH zxv@UEHol=mqkT#%{^htk%TU8fzVVR#aM7$Q6ZDLucmbbI_jqox6&AJ_SCIA6d+`-gW$wmMl5s_*YRmE87246ApKU_0``yRyNZS9vn~JjD z!z*t@n7%o0=xjmmy73qf06^fjf6k+ri)o^Dcj-Fhm#u{YZ!8^duj}|0tM^2`$~9%P z_cYD7E_lGo-Q;xlDnH;F$;!hC%iKqZI&zY636Au_ToVA{R|c$}{ei;F>8Gc&dE%{V zpr6xeynL=2Pqro>&{;4VUvxvs>=4VuK^6zR+nQcn{i1|02*<>`5QP^lN!_;5^Rp4Z zsulg}1u4umxZqU^wCytX7GDc`a7g%gfzaM+;RHI{{72!xF+uh+EU_N#-y9K|@1(nJ z4FQ%&NxzH8WqOD>PA>C9?|pE9t`EV#b=hIW>ayy3^xY1isfPn=PA;pFxo!vbT1nbK zo02l!*`0%iyC^VNl3fTG+h9_<=gZcq-EHSOI^%MvTDLO3{mcc`s<37Lv`c5&L8!lO zAez3$@bUQGaF=7WbjRqNUJehh(ij@q^(;=9{==hTu`-Qt>;g@TW>ocvS$b8!&5l!T z^Tv~*%r38S&+`+N2qrFy651!gII+W62g&FtC)-Zt=mm-_CIO?yW*Bls9R)kp++KNd za{C2TQlCSa2VJjBVds|pKSs?~<&CpTp=~e!n1^fWT+lBX_uc2u88`MGDmge*!pQA}s; z@bvPEx8X`i?eI*CDb~qC4=(TUUW1^y%4gEqh1Cl#8j$9_Z`}6)*#Z(N4#|jku4*g* z5vcU-LTJg=-ed1n(HWThP%6^Noir=T#Xi-J`KpXUMuo}y>_$PD_(gt+

`DkcP2d&o$z@6m+>z_c0)+aWT8*;REV$YWo~T#d#f4~a0a=d*P>?c zj5#QEd9$}M-&>3_8E;#|J@1J}&`JMLUg9u}w?R1Gk{oJNPqi0ro zXA&Eapj}OV$jzdGtj9BU@@+68d5BG|_*NcIX@Gk$txi$0CH zQ0x}ML)v{FTtxNZVt4$l8Hw8yX;h3f?PwlS*+2GS^7oFRX`f=)BTicPKiOP&0uT+A z+8=^dSqr~@S@SfV1WMyJLVKD3^jal#Ry_j#hbt9dSZ0~P#C`iFQTwP@h@s+m>+!va z#{5t4O@0WK*y}}HOxj0+I$M~QLG>{9AUU2Yv18l+jaoKf+Py4^*PGwX@6Dgb11@{8 z^1WHaeN2-jXt2SB7{qiQ>Xb`7tX}~#yRR1#mC1}#fc+30`DFv=RRC`uzqb2r1 zez1M(^?bErIO_HSt3Sl~8J^;VPdacAwMA367`Bf^TM-@VKiH}EjFjM%oZV=s@vy2i z*;UhhJzx9-W5vd5--v|bh&G;&D|F&b-r)fB-Y8R|}J#V9c-+OA?3vlXlGJw?Sq^CGgjSfE{IsljZez$?zu)-o_AHWQV_9> z8>P$*MwWq-n>^(%=2p3O+gFvpy6K#vorO<2snW=e_2XIM)v;34&Jd22&BmK-e~p8m z0GpDQ={?Y^mzf&cm)ux(T(Xj4eQ&S>x`9g*Q*At=8hUFe!x&NrubgQ<LCwF`f0`)EQS;1N!Hk_xvl=_kfd}FgJST`r@B9IX|0^D$cKw_0em?*vk~iI^ zC2zW%!Q30++n4GeZ-eAedQw^e+V2XIrXWwBL4~8G*=RN0Z&ybR(n)P&iWVl99Gc2M{m~*ws*lg z(S2Icc|4I3Rx!xUk`Lyh41%8-uZMMrEZZ>UIvR;2-n#$U&>o0zQ#qdd#Q(PBVM)!i zxLT#FcCU_m=+c0o^GS6R&Gq+UzNh&53Y&0ax3$XBo4ersKE2cLpIyRht)E|Lj8~R0C^(?$2)R9?b<41@ z1Q|BkQ_Ze&>kQMLDj`kDLtSAmSEb6c@`lrv?T6Er&H#~|ki$+BZDe2^Jac#sPhnX| zKw9I;Fd;q${s4Hz@l~S-BYyyX&INx;NeujQ1WJr9#aP`h81!&^h>xQ@<`!090}1q} z&jEmMoGmI};&&`;JpL&4H>Mk;>r!3D#(2EZdlEN?(CD&Cf7S4*hN;fniY91!+ZTQ#Ofzqs6nK2Snfhf@nE5-N zX2lG847rR+s7vR)?$0nL?~R!sfbJgv+MDQajHl(R?z2p~PKej6{9i+wF$($hmM8Uc zu6v~-doK~`8Fd|_iqh#(5=rwsday%UKV2YAqMCx-Pv{@|T_0p*9wZ(4DvK!(#zU52 zax{G`<1T>VWLWZ8$1PNBC#VkEmC684tykJA94fi|e7|T1v8W?)tGZQs$ZuO)igXaB zBqGurD@*dnbkct80L6o*0zPf{`5D~y3P8b(G8yVQ(u#hcPT@<~v;QJm79i?xX4RKy zxQjs zyAx^O{tGbxp~Dztw$Wk}{{73JM*|cf)O&OR>Fmz|LtljU6?Vk@@MVjGxz)gmqClXG zEZNjw*5AZAK{#8#jJB#Lr_So={VA*Ns1gFJuR*vy+BM_NSsl8|*=KHa^#i8%jNy`x@JBcK3MPci&TLi}8Fg`LTx) zmrQFMJ61>CS+6HMfS_8~SDlSP6`R=mvfCkE$7qLz>y(yP)qR;4p;OU`#MKP1E*AAL zuKH5QOK!UCd@k!@ekdcf!t{Ir!Ce~2zk=|fksHh7lMq$h+9S4~ET?)%4ih=#Bd*=X zE?UD{O5N>7;(%taqga0eh*s=eh&<*>2g#&U<+i;TGQ5O$Ne2Nd@_}WbM*tu$+xwsr zOW}mY`&3rU`IILaIiC5vGRHZ5rL{e-M%8j>z=XY7x?TZaA>O9wU)udMZvz0f*a0SM zJF}t{0V)b1J$z&4N)&ri8YG{x<4o_&9gn5w2ZPwLT_5r!l{kYa6_i zu=eWqk|(yI>^SIywAq~+y4JgjHHACR2ot?zzj;}w&AgyCuWG!FYHMuIJ^Tccyu+|N z+_XDfM1mfSN5k`i2(zHXS>Y5h_!%Cv^s+)|eD@b6$GS`TFZVux!pu0$ZqK~GFc_m- z=+=K}nC;7>bAPNydF%Ok!;raB;S(%nA>7 z70y4s`;=H}{j>cFX~=|TETz5G?6WaLOqBueDkycoiPSrgxv zN$e!93geLx%@GF`eF4L`!FaaOj7O}#yl>_@@d$f2-7HWf^!1NOfOAsSQ)P9&4VD%{ z?Lw?%{Sl(zPIc^5CkpRcB^Wv(xWACubi8;%AawI@m0!d;Q8=0`Ai&L-9XFu*&A_ij z85_Qc;CbovYV_syl5g3bvYTAbJ@4-xzT=!DSJ*CKN@!IZltyw591eqD7Z2?v?|c-o zN)02_*)eP+ew?uq!(KR+Fld3w@2kUvr%d)Sa9au4*j>dGo&)cV@IJF3Vz28YO0Aw_ zd$p^c#D0|=8?L377sx%c_}Tl)z!6To6$n8ZdP&_d{@g#Wwuif`sH3Dc@s&=+>-ZQ+ z29K9E@*35y85~SDWUJuK=QI(DZ@CnV1NSer*F0#ekxzpvrXAfQFkMX)yPbBRE$$YG zAN+#}GojR((Ccp~9I;GJAlZNR0O|$&!BsPP)$xGFHi;WmMXDClH`o)_FcsS3-lDp& zG$_Sv@*}|SUjED%e!zYcd?N$4%)nfxUrw{9Uw#4lMSsJ`ayyGyFRy@P7Mag9@70d$WJ zJ-T1quXKAbJs3{JcfW+DG=BV;a&`h-Ag10vZ=PB7Wqr0L?Mk)o=wl49ygQbjZADGJ zJlx`gsLEVZ$>{ODgxJ^`dK5zj+(0TJwm$isN{X7ZU)n(8LwMn>oyHe%&aUcir2eOj z9xFkwXq$O+$KtOPbG={?(9CMR@S8efkfer_v$4tTRa$$Vc$BJo4A|TAIDynP!xX)> z6DsZ(35PJu8O4CTU`s~aG4MQwIgot1-;&Mr3t8~K?DnR~C%9Y})5O zrKoC6Wm|eb9^u&9u%``auu1^n&*blL|C=AxWf5O-oF3)~n?hD_m_inoLWZVMnx=`i z7>Y9Raq}x6LK%Oy-S^Z=NE6daID>w}4}h5aL*@X;fjGUtCw0jjGF90gIr9)7)!TLN zi;uUqKn`^1{lT;)DCCEwEKB5Cu0Wr+b-Q;{RL$hmm!b#B7wJ01v|&xPcAO9eg|LLg z$n#0w*X-+6lf}Gl{%u-YkC7vq^UIE5b8TD6bgmyaf?% zPwxC{uK24mVN7<*55PwgSm+UcNihq5U6;9Bm*V^b;5v#bXHMt+qI$@eBNiF=^>A+e z{2;?8_v_7FFAP7Xm^JM0@ToL7BO0h=B;}8!au+K?%pJ44`*Vy;Psk^zLVQfsr`6>m zd@(f6LtwEq8EO8LX$6aRsYdrEgFtb^{uI}Q$vc+oGMN-;)f@-vvQkG$l6CXmtjhtB zOAu!Sq-nmo{_^QtYj3lTEn|g6_QztPo_keR*W8+3)TX~;+E*7fy0fmrb2K!#>|cXn zXvnQEa!N5foLch$f|DGUU=vV;inCH& zg}$~8Pf44>gRGW#88O_Z)0;ui-V?K36SE{SD9uS0QyKvFp~?m16%zdL`vYLq8WaGqW&q(;o|BE1z1I3zEq@QGn ze}YHNktAJ7>To?F%mJcuBm#^xy0Z3Yi{EBJ0jX#p2a$)H5Oe?_WT#G$?Uo#;8n zCxh@3>q)JxS00i+7HM~~tfsr5EGkIZ0?$jKJ>sZFC`O^9P`S!byLW!in)Fdh@@FI35Dex{Y0UX%I2**3c3L zb2|%UoGK$9xq?lU%~pgVT}ti00A&A~62JRcMG6RaM%-XkoS5OGa>!Zj9DO*P$K4=> z>=3jZ;>iJcu30hl-@=9#+CpzMt{T}cAv-tH{|b9fpt5gj>B!fj*B)XJZqW^y z5};iff|36Kh@bIXYX!C{HqtC3y6?LGt@Ddu@xm2499LALi%aa9hZKRPP_xdcQ`;r7 zvL67&ls?CIu1DV)lAs2=-jYpNZN4%Zvw|!sBMkxQv%VF(%0jhicW?e1o42&=LOInF zV3LkLQd_wNQHM6z06c zlfcqCx#ET|B)X*SxfSnA@e$DCOvyDOlfZq%8;)>06Q4tb4bSyRFj_h*i{% z!|&~#1u|K1m;83|h%CX!jQt@i$(qpFmDqZrObw?T^ZtCsMO_z)9>(!?Og5H zj=s6`RKa6s4K9#4B`n0LVuQbp{CjV$z^(vPn2pZ9!<)6x+*_>Yel5NeH|L(`R3i>HhRuxkIz>3#R)qZan*PyK8|dT;%mU znIEy4%}5AIjow5mYbrFe1I765G;WX=2>W#R5oX%g%rN%-QEEKTZiH@pVDoFR`0W+( zd4W3ESg7N!7~|HBR8Q`&h*I9XP2z>RjGkN_l(0-{SKzSN31Kd~K2_rOW-s0&S~8+< zR9rW~OdH<`zW<8(p>)*aAAq4&Nk9727p+H!ON3qRT&+v=Nn`dC!1kxn%~n|7Ol{r} zaQBjxWYm8hxr!NNQ-GHST|nXh0ia36s*H5ee;6sZSNc0}dtazY=Be%;Os+mCRuT$- z5)L@~ap0?P(&`^~65G(6WEqEZhZ`f%iC#>p!vj0C|6gRc%_yTKin5U%N%F z1=B39FOWfhEBqvck8ga9*alVVhX_e!g51ly6xF?WRxVUKO)?9WJ2W8QB6#2>53CFj z{s^><+w$1-r08r*M)MbtyQ$MMa(0I-J$$|-lL_Ie0lt%Ff^t?Wol?>wiVaD|mO98i zf;DuWLi`FuXV{WgKkyGZC7>S{bk&Uv+N_NPN0=|4x^~?kLvyRb$O2nTKfnzKHNRp9 zY+!TsHDn43b?Ee_6@()P@Xb`2Ao^UpN!_FieZeI&R8-`)z;>FX0_N2e)OjSuQ?yqi zv9;23=|OdT;cw-DEAPEm-s=S6mv-p+JXrv<{LOz?8vuO~L<1EQu39Z9(zAV;snDIY zEATGaVvH8N3=ubwjxL-{wRMY9WhDsiF))#6KsKY0iadyd4q zx^z+ULjRf){49Aur@ofAid(3mQV~iO?_8*A^s)JHJ+*4{>*FL?0%`}}@iS9!)B2{3 zNA5;RGZhH#vy&w1g}Hp+qV;)JC>^jqD&5uI?|2m9@6j{U830VRXug=Jn>Z5OVK}>Q zf17V)Th+Kp(U|uza58x}c-;e3j%YMq(&fx$1?Nq`KJ-4A<8^q+?k%JWkoe+79rW7IaC4k4SUO&To^_X zzHTpcUWRpkOZ=e%Rb~696GpJ!;%Jl|;~jL>zeUZ@S1PlAd337?2AbIhLY0CDcJIml zHRMm?9G2*&ooMPFwZ^EE!xhmDhs#cQc#c2q_*a_z#M@fB{S)s|0QEI8ZJ>62va)t#n#>Ay~qK(YIk`e0EPq;t>LcOffzU_+i zcy!8%P4lLZIeh;q{&{&mJ%-?E=JeLwpXt9933Bax*&AWght1ZAvB@Icx#`7`ErqB& zdvCMQ3IC)tt%_Qct~}|Q)oF+FvdV>zEj25sTD9%_?eTM6A24vE{`u?xx^yeJFkH5sItfzmD zh5fMsZ;a`A-?yl;UEfp{57pm3lJr3j^h!T@%Wdy)6c4^c-%SnwM`w&28ny8*MUKN$ z=>4QzCy|-0ri=Dgt`KT26L4T?cin2cRV>52E6o72_S#oboY!@!k$u7a^p94l<^-+i zTmuhd#$0Ss348C|wzrkJ868%FBTbR0pR=c*@|<@!s~;X;w_VHb({>;{!ypoc)csV` z`Agg{jsq_}2(yF!2>&y_ENR2mnORYg`9*1ijAd5Tio6w^cUY5mSPOo>j2-M9Q7s!7 zAfm`jkv|*Pz}E7Gh3}sOw+PYbDbWG^9%9INR59~5c+OF8Fw5}=;N_s=)WK`TDf~nH^kY!yoHp&VcNe3+HYGpJu z_R&ZKX(Xkkc4@0G;@xS|-p35n*PTWR`NYiOmDGVeDsGj9j%pFS8emM1jBvuT28b0Qf|_0#2P1r*i8E-U+8-zoG1QL~gQ z(Yt}W0*89K^o3URel&$w>gIu`Fp7TmSqXmr)sO9~qSGZqEfmoS=mAT`PtKyqK0hg-RS2*- zoDQ}TPDqUv{`KN+vtbUD;sBtO6CTML>nC&S1{kbOL9`?^cO@&^6(okxp~7 zv8i#h2~-=;k9X_t(!VfrRwz=e`=yFbItyP|?`C6m27DQX-tEaDSc~10nvdVyP860# z;f_nAomfr9UEYnj$UV_UamT{}X}EC*{&SY;yfiV(xHH(NJ5(bDpCkX_F1MM0qJZTrY! z`ULNkg}}4T((QxZ*UixOWl|WLwednGG0U3@_g>IMA9G*OR2{=&=e+eWq_IZ3l0C0S&>#E3`*8Jy$p zHrrCzRo?EM*$!0#JM{fy|D-;zu@Q71;D7Ny{cs=g6Uue^w_6{Two|0lkcs#6&kH1t z2~FNl5IcsviX52VIbqCM6V>yF9wnCK=%Kk?>(v*;{8j^_P{nq8ZkAgqO9KnmcuYrz zG*8CjHfy?u1x9(K0y_L+RC5L&Q6&e-;NiF&)t$p9@1m+Yz0eHRWYjl;>1 znR-}6=Rb}}lb}1lwJXB4ABk!|x;}QnZ@s?KkbdgVJ3w;FbMztKhX!qE(69FXeI2hH zseJ;6<22}ZY|&+h+r%6i>jsA7+FXkK?2L~y7mMRCso--alTUBiiwtUQ_xFUFJm)#= z#f-va7;+(-M?*Y9tI#H6D5Ij)wR=_;$W7I6s@J?>43!nODL(8`Ca(VsB+~f`vMN%p2ZAy76P1s^9S^k@;fG8bbICiR%|)O3;5ce# z^^jc?QfwF;mvv*;my0sJN8JRDOQLpvHaOBx8mhf+P%%zz`c9J{fW~*R4G7$hAlFwy#Y@CPXnU2UNKod+Kc1XRN6YxV>ltJ#J${KNKGoHvoE{(U z3T-j6%AtXdk*3XPhw24LU#!EEipF=BD&8%&(z>?m_)jtp?H(b_MOPeR?C|ha<6gs{ zi2m)IL&kSkCcgL$ccQtpI-)rdJjj>fHhod5jK;sa=)y+$OQli2>5!GY5LjJ|fmK9{fW1rZ?Dj3c`Fo*WB_=i3Jo5%<{*v$~ptyXxs# z1lQJC@My@hMk}N!AAg27eKsoqW)+Z32*l^oc#4tjUr)CS?fb^AKKo62wqcVd#YxD3 zQ)JfLUKJ|#V6Ij_LWkY(a|dH<2Q~9Ki`iKrJY1AUA~Lw&^`8jz?_lRlv7;POc{thll;>fKAv)Va1&Tr-bLY}rIBQCVpg0;+ah3)P zn;;4CX<&4hMpDS_fRubE5$L0)N#?!x_|L~M;}`xK;1_YJl`*lqQiu|YpPxS^rg%8>8D z=wG>rs+_{B$Jgz7yalK9zMsJrUcVHIRCAICkTq3SI^AWBSKy^ z(=FDg3F5q?Zks6{nm9I=qRVR7tBiVc)@r5n3ZzVRFR3GTo;K(tDva#ES>&r0Co8VB zqi0Gtwx2PY(Vgf0{&{_x`5{s9UL+LdxH%S)3hHR=pR2EAudU*pH*<+rVGA>NV0DWg zT{GJ?ccAx-ioLi)WnN5?v zdGt$WN~X_+eNgJVCNVcz|Ermfj>wDL`SL2CfHX3rKo*SV24!Y#^W zXtc+YENU|SB6K%#mzP6epH%?A|91o4Sz;EN-z4nT=Vd*AhLZOd>-1hjgMY0{_@u7G zZ0CexaDr^lqf!(Pzk;wzB>fWN_x>Wmd7CYaL`K2i8fOISiKg8+yV4C%tQvdUY1>tM zZFXIn>lWRPshLP~KQ_dgW10QL>uPh)TZ$g~VHOX|pT1qStZdAm5;k0UKo5zB(fEU< z%1JG`60eHY!fM~$ec+lOtCPf6V<#b!dN~PqQl_qpq(EEwUQ^9nsk1|YZ3I?=wo|a2=2oyQ zj7Qmc)toNC?b!O4YIGY^o?V!GN5X$J{TbiUgo`LyEPtOkvRa*&xT*sUU3|HmC&zxV z%5lajmL&NY3klhIbAP}oBAQh9a#-ffAP?Kdqf_&^r4>B^K9-ZB8GprtHp9jci>tT= zU2By|{*FAOTUYp>MnE)8@82@pE;3U)yd0K_Vy255W!X_p+PO)UMB@tn-K=m5xK+kf~r zp$h(lxMG+(UxNMHxzPpDH4TSqDPI+c-khFpX>)#@S>p|<>X9|>T>SQfd!~?1L@%aR zYDHGk5$bqk;_ORAvdQ^*>!re8(nQ?ByIWX=?y@yrP0NWrePQZaScn*BWQ~EZ3VWxk zy^RI))8js_2i1~R76yCGmE~qibHcfuyLnFRu;4+H=&McIcW>(}e@kewxYPEin#)%u zH9i+8__6Y@;Kb3UlLN?0?_^s4lN*wRTK0^MvJjPf`6u-=m5p~kmI}-%TnU4hn=*pX zEdTcrBz*=r_;6%L9#%lv$ME&YdA{c$_U{t_g!rQfp%fpdb{rPYp71$u$1TLglK@XdnbHh6z*?&}9Zg5NQek+(v`N6lFRM-2>IDWl% z%+gE=r`MkbTpyFBz7!bTJ7rK7+>5|E*B|-h>VF&c$APzYB%8Un_9XEqiOt%lzVCnE zIx_y{J3Z61>C4%5cYE~g|HJtmq+9%%11`=pZl1o7c|-7jjd^uIw3x0ldJPUS2Pb06 zivHXe|FKKBuSzs2S~)B~=3kirO+CtJ%U2bS(1kw6TUg&wYY*BLwk~Ct)reU6TS*hh ziq$=Z)wc$_HEg54~SUN`2oh)U9oH7j(V_9r^qL6iEjqa-y;~@g0 z>AtMDt!z>u25!Yn4|}c&a@yHX`JKcfZY7y^qBr+!JsS0%Y`8R~yq}K11wJuc7D!|* z5lHEigwENd%@yfyqy~JqJuu9=YL@SXR(kj$Q7gUSP+2>eT59?1sb(+2QXyuIFF40RZ?9 z+!b2FSU#Zpb~qquxO=EqSFTf;%GBKch~M{7*LhrecEdk`LSZFSjg9o}ja1n`S%9Aw zw?fnp+pw>~#rQ`&v++d)K|*Vw6e#>a+8_7maOb?9GQI*)sRNfvC7}uuz={7FM>?9C z$(_(ej>_9kVXL2T?NsI_z9QgUR-Zg7y8c3)N^3r)JMmEVStkDM^1D03dtT7p9Reqp z@7wF6XNNcNfna_Lue-vc(*9yMAPApv=F!BfZY*_E0H699SQ{=?dL1TwM0nm0*=>2l zpokDx1$B;I^OcH`TpVP)5H)a{%Nf3ZuD460@7>~^_4|PQOKP(Xs-RNu@z92&dtEmU zm8+`g-0}?Xq@#vsd&bzW()Jj~X>mPG43g{&6f+b2^Q2!fK@ZZbCofh+4*u|O0bv({ zWpfGVy**5Lb^s@WQoG~bo0WpbBzJa=ro-<`id^fjmlhq;mpIr=2YU|fNT?>lFRHP( z*4rF~4WH)f@UNa%q0f-tPhNPUjSeaDxGb_N8qdsV8vUHGGQhg(P-^9S=JV8b7lQ~7 zaP(34r!|AMP*5;|A4f%uQK{Q9S^7}Z#;~NRN3Yl$ifjnFr*!)C6VgQdTCIarMr>t?!4~-{dw>E``!2dcRtU~ zo_XfXoQX4Mo|$>(JHvt?Z3Gtd0H-Bes+u`?fXcu$*27T*cSZ2DEq~tuc01wYJ?B?C`x63$nhZC7jn;MO7FD{Y2;v{6nY#Stiy>X_ z?cGI!{u*inwBhgS<}x)|X0-k1N@-h5sX~a}n<)5UxOl^nzrJ(V3=&>Q(&`CQRyfQJ zTzA4CSL&@SP`0_hIS9h4=S%t1-a)s%kWh$pRz2=XU!fJqd&Qmu$Iys89*4Ynx4LbCqdrg3)A&BOh;M7LXAB`l*SZmYHYo6=W zX5hyQV0J3T%O&yNOb*SJ&~*OFcY3bCq74jVe@CktN^PSPq z!KfM0sum!O?ol~k8tz9*8#W*`G3;A^cvhsq53VnJ`iJaSO0+NOVLBKB2z_vM86(~N zI-Xpz;IAhDHZI5k2m`a}>fPx^pDuZcgiq3<)%$Jt=9^1JVf>D(zW` z|MRr}zCeAwIM(=@yW%&WA68D|jHNob_A51egNK6POD(LFYv-QJzLf}E{3qy8DfMV8 z)p%<`{&>Uf3`9bM6IF*X!$G=hM5_p*9Vx-9w-EgcH2IAX%YQbK#<$++b{f@&K2@5`_{Kr($u3<;UW@oW zLk|L5IK>G(=`*aA#K%%V31g^`}Hqne~T)ZS5h9{#f$^b(3Pw56A0 zZ}m%mbN}ZL#ghy50OBF=d&LtK1&)0uqbs2O`(AK0YNbf30J-jdM9EN5duP! z4i9R*_T(oqQA~%OfIzN*F6w0z9iT167m9x*s6U3~?fDmIRf{q+D42#rWbZ+-N}druy5zH<(fDg>cQaViJx0L)vF5SyPjo>_0z z2ilzIYtO%GskELkeKEgIfi^Er)RlBba`<`I=bBT~jr{$dw7IO}kerz(Yl)&a;r_)! zx!X1ApkrUk_tN9q7m&5sgiyw+SSp(@KQ>-gxx$4`Zr-b>!=L6?s?TqvC0QVS#`h61sr2t|1PFQ^rouW&*xkW>@1czoisR2ZIjMi~VYBgiWaGd}^iW52B z3VH{;P9cKVD1pl(;od|!q9od)J-M%~=Ryw`H+J9~W#_hPCC4_@=Rm4J)I)C>Z{qwn z>tbn3o1Z5MsIzN>@5mB2qCct=X}PKlLV<4(rnbd9V3j=-kIqO%X|J52&`Rv_(#U7sjIY+ys1_YUW;Ih?=qVc(nS{3vvE5qYDRY2R2DaoC0|8Uf? zao9nu({V?1BW}74#GCYmIvRIc2wr`28|bMNm_!Pp4FiD!GQcm%E!?p=fw>geJYzd8 z5DRL&VX%8b*VR%t@D@D>w$o6{fN1&|-NsHCD{_lZnl`bX!@enK4${wOD~$SPz$5RM zZSY60;RMrf6tHiy##Fifnh}6#p75U`DnKAJqfu|8l-sBR`kfpDKs$RNrt`g9>qSsI z@UW$b>!h3YDeapWfKa2jlb;$20&JiHkz?*o+xi^89Nt~x>5uy^wk_?V^dnvJ8a4}W zUx84S#t((VT;~0`pMEJfA=H1-&W)1T4J@DA8i`4v=axCYFpUo~?k89e25%@3M!pI@ z*+2g9P5Px_L8iSG)uAy>ynER2P#{_eXSa7}5lW9Qv)%S%ht>ZJW=gjb%RKsS4)>-z z6^}8nnzv+{C z4FAUkegxZvD1%#LqgGLm&o)AW8O6D}Nj@90tgs}pZnU6X_^MzVV=RkBZqjOqif6{0 zm{u`7&n(&n{Yrp06KYl5>4rh~ogzu~+pC=wHaYrR0;CqW=+^w(nq_0+)bl*|X*kV# z*<hbp1+ zGuF;Rq|y5KcmeMP$>IBnGPgezoss#Ki!Rs}pDKtq5i;)cLqZG0F~gX)N++{v_Ycn7 zsdV;aaxKilsfR@mfmqUy7h9v9o%Z z*bJ=AG#i`Q=xDruO}ADX(I&1fHX*^)wdyDYpvI>P(NG(Fix) ztne}yluFR{LC*tXNJp({yRquRsjD1A){J_ZL^7AZEZM`E+37g|;Rv8>>{PlopMIp7kvrionvaHY(%FXnMn9r^)XgUD~f(wQM-nGqAP$>UVXgE(9oE z`#Rd_81MOUr$SKk_->I!ld|54YFZ3rJH91n{Mg~9>1xHcB!07|ug5-md`n8Ry)vRD zDA&T9CKE3{*u{>z$-2jK_whvGgG*3sHuzO&w{jR+UD=Dv{TReoyT%opY)m_2w1a!% zd>?zTXJ~*AL?T8zu@&xgiwA=;Y=+wur@y$dWw#g3>b0uOAw_F;&%fG;Du|tg300eO8$P`Knx~>nXr1`lMh0%R4ys21lMVU>ptN`g79ln zQ2aXrG@>#vEIiI#mr~H^AUk?{K1^gA0N4=6U!eQ-G$9{&y@Pf#lvgV9FzEX9k7~2$ zkk0|ZA!D=>ZuQ=|dC{|@_FJogc(NeceZ24MZT06??@Tf%cZ3v}S!p(?y}1u{iHMDX#(*D}y;gw{Hh+nS)G@Rt@y)R-=y|sv zfCL~j1X4htqFkHnNKs537%V-|>1cV+uALxj5=$YkX0PQQ%961&jy(`6aoD0B#^u9h z%a|K~K(fRj`KZ0*{T>7owfQ1tI9iqS6s-0AKYcBDc*{Ta* zp+=(2^-R449_HUMozOf8LQ53T7LMh9cWghmu>Yp8@YL1jDVZ@e9T~;NjXz#kIiSbve)cXdIbA`wKK7U3kMmu~O+gJznfn1wO(z zh6cxU!ilyY<=_*`N#HKS$N`A`?&#mTYQQ}xdx`Q^ODn+2J^1&B{C7~`?}6bL3u)Ee zn$Q~-VGsJ2;x3ccZ+}2r9#^m`ALwyW+pTch!(y{!cDDQ#E7mN&W6AJJ4>Y~J4TlvUBt=s`GpsPxGP2&5-sFw5t05jb2lXKOfvU zkBngFYu$fga);7B83+5R=~Pi_(s|El@$h^`NfX~oYhR;;{oaJEp*tkd@nnl<0sMPC zrU@ua@2XZ>Ey^7|p61i`gAC}cJm}p9Gg0g{x_53RCTa*i0Nz#9)_QLb5d0`Zs5?oh zX%B5Nu}|$LKX>Bk&M5MS;Qg1s#NS&VrQ%n&_Ipx(<#t~LE;ci}-i9TftbCcUn*Ng~ zbnf1|+F0iW!_dJSuJt(mwF`+Lt9A7JV4`}Mx!GBf;LR{#2klD05kTs{#r zgVAv9e+x7GgnpVQO#uid6rU!FAP>g~_AZzAAcjii2zt9&uS_BFpzXzzb?gCVu zEA_3owuI)RE!5`M&={^EoPdG-EF6iO5H~I^zIdy`Y!!0tAG446bSxt zJ-*t#m*nu^YU_rBVIEAw4$cd4vG~soV3->{@PYu1m>Qy#$;4pg0r;Si&NMJmPbUik z*Ukt=h82%z>f`TCdj;HK44YdF4YX6^S{|URxB4Sd07yF@*z#%+gz_e(70w{{9OS|~ z1)#DELizzzU?BFgToe9_YQCE<&V4qFfDa(lO@$yx zvM9F)FtRw?QCCS*s{-x=Jv}5wN^pLNNvQm0$i5L<`bll<;G!t5VY!Hs&S*(eu0R zs8#~~jN7?HDVl4X{WTL;71tw21t>1iS7k)n&iT%rv^H?f8Pi9oP%Y>_!Xe5>-Lu;R zghuN~Ixy6hp@fd#Sy-6@0%g2e(73F8n&a_3wuOulgcg#lEy$QFp9f4tgrMK;p8Q#o z0eqH0Aoh7(eH%fLx>E&$acz7G7<}|hu=X^6KdJyO+t(NGqcsDg4Su>|2J>MO;Jv+u zw>OC@ThYV=E+1%xg}1Ib>qF1Y^YI{*(QzE;JSr?ilxZ=nqYz??X{MKsOVTQb-kjA5 zXno;BwX>L8^V)SR-0-|twUcJiKB^c6I`Q7p4^SWm{kCGQORZ{59I~8jMny-2l~oHb zk118v&OJCM0vUuE@(R&Utv^+;RMqF*tJBmqi`CNK0?k;mOV88z4wpX>fDjlMFxYME zmW;H%yC2UO7R#Z-wW-u@5tlTo^erT4nkvgpZQD6821qrq)S$=pdEy(CSi$Mrjr+#Z z=&LN}pPAp5dk2}+cp4(Q-3ze67-qUhM?nq1yQYAkRRjjnTeW)O`7E7S_2{pqvTrJ( zD0n6_#+jQr{-Mr{RarkUOAfk46gnT#?TV@t@IEn(DBj36`pg4h3t?NlGuX;^Y)ss_ z^zLcuyN1|k?SVhXQ~suktIDBwKaw=fIEvKSpxI_vRh>r!2=dT5_QhAh{$mM)ebw~6 zr6(LuH@5TS9Lw>3f-r2maIArznI@7KqDUu?I;A(BG4eI?Ef5MY5Y4eKTlrS^JwPZT zLo3GFcn<^zw}0S&ta2M=&YMgT)~X~OSHBYE4tLR3fJ zyvcR`^@x9RL&mZvqz&f)vtY6XihmK7H*UO&M<4_Jp6Arv&$wlJ%2u=4rAQIbu1=VQ z+{=zSxAKo?&BJMAAaQMH482Z0KI^N;uerza554?KJgynP%$)#w?~i*+zGc$V`1`wuXd(s^R4g?< zn7&3g(*A5XpY;6d*vl`jp5Kl@e|aPn3;q$nylu%`O*^u?4Ly|xn^KO>NqYpJi7A+X zK(DCsEApn(l722@i~ATs1D7;lj`yDsU!d_P(=%I!s`q}6!k zs=2k|*u3w9vQ)-ZhqKGpJVk@k6(tMn3%#;&I-*4NG?S*jZ4Qh7&}!^~bC{8;W64gI zPB8!Y6`bIigQz3*o^S~IL^gotrEffZJAz1(@{_5f={+ERdjF( zTt$>*uj3{<@-=c7%`+Axzill_4oI{Sy2IfY+#kPg3u3tXi>5rY9UcE3>h&Jt{2{K()Towy z|6~QZQY8;U(FkhS0pN(`Sv60CCLc^)wOU%rKY-e)tj3#%UpGDw7qY}||J}0K9F)ko zD`ChPqjEM}`g$iRu(i6c3m%;d&0Li;9g<{?vvy@JM$XQ4>~$W(%)2B4CV2|fCT?CC_Ck8$Ftf?iGm=IUU7N4-FWysoP}NXaoPPC;vWqI@ozg! zSorS_SvnAU!)S#Y2@72IP8!$GyFTV@IgEXhD4kd(EVvBl4r?^S=L^_8B`Hr8UNWBj zTAr!{nPP598uw9r9kwHu>ztJ5GGOrRbh|4g##9f;S1I+N4OLKE(*VkevHhlPa-cIc|uyZAcN}{uT6_8FX zfW^0Y&Xkq!Coh|){LazuKQ%YF;IV5+4XxUYI+#vA2k7ZRirHrJ$8#GwD3)$1cP73* zs4^2NBIqI{U7n;3VM>!1{f-%B`d6>j}zR*Y$vtbCIWbd^2|m zuz*#8(uzz?B?vfs=*R`)Oxp`rD$CUbX5s%fx5o2h4wqK)07=QS`{(B-{ix>9)&=?j zk?O1b(zRhdk>Ye!T2Tik3%IIFFuBt#5pZGwJRg`cxE(+&kr|=;LH9pzH(#~ODtk9wZODHGM?N_Bdvgr`Fl+q@S`R?#-D^iIWq?t!IowI<`r&N6GeI zolz8o{sq#rooxFfYyBr{`}(0})I;ekxY2CVAKdw)e_k)1sQ6ZoE5H88GW_#+ZRK?3 zft#Lm-xi@PlgD)^+&-^V)Wzl^b}>yN7?dRyQHJuHmeUyOppod{T7LJHhp*SyX5^2% z>tgSL{r(V!y(a~q#*$8MJ&)%&OZYq9x)58_z}l{^Ow2!&3;mq#w``@46}FqVe1L^Y z{1IMLS;)ToH(3-7AFfQEoY5Dasr2BU{sq!1D-#zeDl2<^LvlW&16T24r?dl(mm}wQD7-85ACdR)eTrKlgsFz7*m%U`~4nZff;;T2d0*zz4AefI6n4lu8>n@cUb# z4HibZ!c-DPlZW{!{i@%iOWFy)o*0B#5I%{u>k zi{W*3{4uCKsk(bhqi~JxlR}q0oy)AZfK*?$zqH_0SiCi98&xtmZe5}2k_U!jgy7Dh z#FDw)d)cI8+cbU?1LEUyBFCQSDs%KRa|OvN(x%Mc)KzcLR&Lw^PGo$&fxn#f_4cj+ zx67+bf&S-8>pPwUDV_uQLg%i3WYtmK)n)FefQ6VM&MQPF^Mmftthjp@9#@XeOc=pHizL{GZ z(3sg@!2^SNNUOhGLmC0#QT@S1CF{2&gVG~j18SH5=_2d8@-~J|gtYp*_qw|@rO8)# zjy@TtM7;(2x;rF>v5iL7YoX}daW_@-e_LpY%i%4xkF}@wIxFD-q}Z(>bK&O0`^!iH zJJIoPr$>`%3ut%j=N$-zx1RUpZIPy)j<2WG@ZL-0JsPyP@^6#`A-jSC}7O z&*8^bmSm+t&E6yhUF#MvWpb2%;}}lonEF8)#=icg6^l5}luW1+hTb+j=E_3`A^JsB zK@NU}o7Pd6yqBGazDRahwD8U@vIvXAaP(ue9=g_N0*B{@M|^lT+m%Jzi@D7#P;lRI`;%17ZsU?(vbbFD?6z*$+4!kb3uy`&dZEiusxfYj%g#-g_ zIef@^Cna|$6VMVnKiE55$;{sp{8rhqxK9W{giGg)E^6BHq??(Lp> z(Y#fY!Fvi7UdywV_Kev3Cjn`O4Mt^(FCvFaye*(D&)O_BEXQ?MA#YiT_`|9QC%~-E z&v4b6ndeQ?mOfgv1|iX6hNYu;Ei34!t3wQ3I}P_8T3Gk5llkA~zJoJ{?T^UsLT5iV zazOlQQ;Li${fvIy8n$GyQ|t8eij9LqEj^1bmIr~}1Lf>+ z3kOQnsn`~xiDxAQt1L|X1%)7+5rx&*_xooyr_srGs2<-ri}xZ;;Cp0d^T_hD%3x_K z-?DKmyDO16+iD7n68ExqPxsLuI`B*EnDzOAu(5sXZPfLFNsm=BV70~crI;z}ETMDw zMOa8Xml)5HFcsGx9vh%J#XSBjEO0OL%akUolx3QfWoZy+%Q!npwaJj9v?fBU9^Bmb zPW$U{O%K8&Vm*G+qfTzH?H>1MrpZMILCFI_Elg%Yqrf+;((o&Ol{wp;WP~_>?=zDj7^4_iB0Od;{RD^qc{6 z^92P=W4aec$d9{)8N(MAOBx0FG^^6K2!kjp+aE$@mHsBk0xQ+p&`T^H&*1!$Rn`7FHMsqa>d1|C zWTiUsY;4AeUyPPDOt&YbmA9ehnQiCY9)(6{R*UCdh4l}ncX1{8yCxZ#k4emv9*biL%4^{gv87$^KWx+p#(f) zN3|k0W)Lq*f|;Jg_k{ygi|nk)_-XL66bEt>kKbs+;&9&r*&3fZ7gcKljcuE8XU>s<%^B5w_fxLX8%ZjUlPiE2C3aPZ3 zhafmQ7p&@IS~5AMyM(#i%=BSdLFG(wF=5R1Wls!%$G1QlPh18>n@Ec12blhe+wjjE zoF1qbY~GK?RrQ?#ulQ@PgFQ;CM~o(-B8v1!fTaY95mA>G%?|c_e^9owxgDVMgXa<) zc-2*4@eV|GPXNCUoQ~E9?hk7+3#SR2fSp_`u{(EpR`Y!4Ufs)5HJc(jUF|Ihy0}5i z^m(?oV%_u5=frj}x!T}xHu{-c)S1e$p({J?q`etK#SPW{?5DLNk>xb3tu!s5O8(?a z6)4a>kCk~R^2I^Ji@!juII&MNpV+_$Cr5 zd&CR!-WTx%aYU@bDcQApDDn`3cYQpc&qhn!>8OTT8Xq^Lk4U zlNTFPSx;m}L3az3eBi_r_()&M7H5_(yE+d7p0UusZ(AY44r_MO&NqUeN8;5+x@DJEha`8F86s-GMah1 zS0~MyhW!I&Lr_z`dC(B%B9Wgl@ev2Pe^}x!x!RbCSC}oH#nOOP-gEJ zFi{y5BW_ni7oF0v1Ibl}@ov^ze$+_cg0MfngO6}+&aoPh;l%BLFoy4!&V=t$* zJ{uoa0&?etUW;tNiJ@9pRP?*{OedSoZvJ`IC8w3IO1}K8M2DZ2ck9(K4n69#On0i) z16CO+Bl)V!Ln^w1f;NNe;&t#GBFwixIzotTG~G_pHNjW5wKG#y7lF81{Kyn7b`B~<$J!8YYQ}3l_M(;9kG8yi_ZD^`})H-nHnDdTL8dx&g|VADZa{#Ll39Fy+|_`e@vLn+`Zbunal(Y6QMpHprvmy?2j$uN*Z>U zDK<nY zW?%4sv+6WZtb6@iW4bzY-jme;!NylU+|##M7iW6(@Y=oBu%+l-c?_ z!?168gi~79nis+zQzBEp)A*gVV`0rVRS{2ZIPPYPDfT_T%u|j*2}+0?W1OT!s4T7$ zX))yRv4Nz*2o@N7%E=MRDaEZ_DDC*ZGu+@x^vDb0iqYY1Cqc2+7u9JT_B`fdO~q)* z?{*h7pjYxW_Pw|s65QEU8D-rdA3G^wtIz+N@ZTIv3|MdBR`*YbNmpK`n$~s%h3a34@XrX|by|*x$MwQ6J(_k$xlaWuKCG}NN zk2F!nYMi1#?lYe!a}TGxMi#&D%I5TGw71#7UY-)3Y1?2XzoIKdXeMEE<-#+CjLA%^ zq2_xp^4IsM5*!7us6rB^n;s?Bg=*GD=T`$~s@A;xm%;%gPft3-2)a&RWK(af-X?eX zJUp`yiRbRW7_-aFiQtqjX~lLAj7iKQC}haYt})M~6dmNW>$Tt@5^d^;d*BtuL$5>R zfMp;t7R6}Zb^N7IU2n6>@-R2E;wF;Mhs>`C5xIrp;o4tf5)qG!3OLoK-lJ1W4sfOi z=d%v4`$e)5+#_{Rt9P(aaVCpWt6b-;%`ufJa^yo785jT|@$3Q&V~$lTKaWCf6^n(4 zu9W^I7r$79xaX<5QxQXpbPn>-{wjpEV;bwk5i|QwMF$)e24?r$Xa{QfKo6{NkG3fj z6+j>ng$>zg9EN?7>21E9R(M3!58?`AhSRsY-LNNJwJ&?(4Zj7e#lk`is1Fez`Z+c# z?`_#WBeUp`H(Y9Wzvsv9_)H{`%9bntKSW5_D7Mmg} zLSNNV4)*8yKx(gPK-XK;poq*_8AIg=^r|%t7z&yqxEJwQn~r?O8n#NSx6*n$-toyM z<*aeRnpb>;?}EZ!(x-Qll}?sQ)MPJcv|C)QodUf|humTh3@9-$h}mWyhsHAF%oWRK z;~7W_5SW9pp$oF*@miTBdDxZ;;BcucvI!wsWTQ>$Yl3@kwz$RuJz214f-F17JllIz|+e@greDz?E zeO0{KM?h8+`~25@{p1 zk+`HlJ6aZkXm;4D9R2&r?LlCPjx~_mY5?pM$APLON?~AXF2WNv8f-d6$)udpoJbwc z(CZ>-8!f0EKnMR|#t^Z_p>0>bD!{cl%;x7DJW0eAeaC0Yt*eMs)#OmmtANyx^r~)} zL^!SOUn&tDgC~&Gp)Jg6?@=?|nR-n!f3tjS{$8HdsPgbaqPexi@PK(r)nsOAbXSS$L0rC4dVKZLhg9xOMYY@2s3Y`j?FSVwpg;ynP+cKQ*pDv0t3d zTGuHegW_9sZ^tWsLne*cT<=js64#Ht=z7{X*9pE~Te`ZfIC*yT&%!c_7Nz zq&D9_{r=cxXGz;*IcRr?BELvAm)D=-Xlc*`A;!6;q7fZG=Cb*46!-+oAxY+6bsu~csflZ3Y0@Gv9+N7)RlqSvJmUVVqJ zuMhYpA2v|L#Hp{JUYs0UQU6Op;5x>TfqNBcSn3i(rCU7J^@*)=DSLaRU#`cRkcY{AvByAlADvx&+XRdf#<-R(_tiv-n%1iZ7t@jdTx<2KH<;Tj9W5FSdN)(D`!rlWBU6L%vZ%5DtDeR^ zgFr7>2^A1VAqu;TrR#R_7H+v`eEe(r+K;tCg$KvTwizC7gH*)#f62lx7P@!cikhPl z)7}Tm7{t)<0DQDvyDuN{9pkz)ZV9z>C2h^pno#))sw%7JZ^Hfy6!7)g)KR_6ORxi> ztrbc8Zs}wPYlLm#nQ`8(x@d0|I_6r5^5Rv;qJwQR@8Mq{8c44IhZTY01Nl*CI9?*@ z!*9XzopX%Xk^xn5PqfN$;+%`f!Cx0Gb6c!6R|)Kette$5O;01=`*UK_XS{6EQV;{y z;goH{U{p=}e(@dJ=V}3JV|x#sf+4Nr=Asd$;sac-{QDEP-h|GBXBHETdSb@eiz!#Q zK4f+p9>b;O7i*!uJ@O+XzDD?)jU}j99#b2;GuY%=oCv~nc)Bf3@kC9>JiOsw$_Hq) z??j;OEUgPHsx$9mGOEYm@K+!Fz=LC3cnq0s6(W5^VGh&)t{am|=d}g=L59it)p{Y})TM$22azrpY&x z#5zyUaj1w9-RU%@d4eF%lQf(S>E&wGS!H>Pk5Y@9+Dlai7>C7&i~HuFl_vW} z9C@-y<(5Qjilk90XNR6!M0Ucol;s1+TBEehO6Fmgsp7W!msg5IEs^Sm3B<#3u}4Yj zuJU7~rIU*XxmshU)H!a$*ljj~U*D&kbd>NbPH{jRYz?=K-PJd%D4K4A1(t0LlljK9xZI+bn zXXryUtdRQI2*dX@_C(i)9k!9=D_S_@*uf3Mq^N$m+7Bu0w`d72%TDv*A z2@i+%#fbgsL%&ps(tQI%T`~o-RYiir(6!f$X$aMJa%m`qIwMba`z{&bMD34B(_$3{ zGA~cf1XX79b`k@}`^{W?kRM@36JOfmGK;9y9v?(lC6Xc~abroOV!FJ174F$<#T{Le z8hAqJ_{zgYS8pQ-A7uDStUizW3nUcr{^X2M#rTyy=aN_t{(slZANMbzHfDr#6KwZ( z)D9xUcdra;Tfg)XgsM>ZxlGRz`JGhv6wA+7V#ov#YME#izrDzgB1HPn=#JEBGel;` zZU)Ife^IQ-v3@+cH!gqTpghff^qb51mM<$SQyOpoCWjawk&y!V18W zo85)N+}|Ip-)7Bw)V*{7E4E#3+S_E$yYijWPss94TP{bh1kf6SGB$vCd#|@cG4X3)^r(fmURHb4~gcdzXEpNx^9yk)I&5$q!m^T zn5^x0Gjm218J4!S+9>%7qGZ0$thGChNYj`?l-|dVpG}Q#pC$DWQ#vwbY#7NshG=L7 zSOugj{g}|F$J3kElayy0uF|R(_zG?J$na#`-D~a8I3Gb$tyeHE&K} zWvsb6h-91mP1SX+&cUz5duq<;OQ*IqY5{dGm!iDTmSSPT_6O$V%+-vnuhLK5_R*VdUg0K-N;0KR9mE=MTk` z>pAdU8TWuM>@QC6%CFo2rD906bRinnQ@MH|$EcfK-`c2$ulFUEi_bg^C|=w z<*9(vl?%6`pQO=qoCyARwf%Abnh!eM-*S5nVm2zd-^T2mR5jExw5>q&A)m|Ay)?Dv zU18+IWDuUciR$Bvk0jG(LVM$U&H%iI&f{p~(m)u-=U~mx$6LEr)>hTPxCqsJh@Lx_k`Ngz~_pLBnjqx+J-ukEZ%#4SZ?^d zB-2#2#Q9{I8n(SBFVhf=w}QW9pzUCfk2ge)-&c6QRrrjAr10c;e)6IM<=|AMR>(K! zYe+ztsRM?~@bg=hQnl7{+eLGU*7{3Vy_sm6kwQD5`xIE9TV+zs0@}VIINl-g=Rz?c zjVU!kR{U0mx?_yc5>N|8SlPyKz9F4CtwM+O(K)_)c*nZ4XsDfRKbK4tM0FH*SsQ7# zBqij(@MUCUb(HaqsNbSUNK8{T7?}YbHsdHC?%$e*UX1y`ee&rOb1gKWNij{uP4>^5 z9SpY1HvG4_#b&{ja)y0JL;3JU_D@uGV%Sn9o|fg9P4>6eQ@Anl>sT zC!#QR-__${?COT?J0WSQlsg?s%_cbBLCblxVrOH+e2l~hV)2a(bj$A(Xc`%h#KqZN z1}s7y&f*yOhLg0;Y^@3ZQ1HI=t5?OsBzQOk$e4Ouu_SU0cejx6yCJz&j_H=+x2=|> zGR3N!yN;n1;LV{qzX8m~LuOJ4FuW8gC;dfc#kFG(JL^~z!Qvs{3x|sa45Vwr^;RA; z0e3)y?^-DzP~aos#Qlj^x48xAr>h9wKLDnp0i2qzjf>3{w|-o$_=K-%J$X^#@}jir zC3jw`neYdSk5GIWl_p5&7e0KaIiqq?h6v{zahDok#&Cf*|NApFY4^|mN|n_w?5fDL zss+@1pqb`R6!$AX`;=zb=pBrye0)UF2gT{a^%4z>;KS$IyN&|!KF`FZsafK4dO}9u z<8kRLiT^~A7_XlrENN#p?7AFIES-jzf{a=aj263fcXb`peS`}-fN9~f4V)zz`(O3> z4<{Q&P|L%{-rkLDZD2on&5e)tm`)t1pOjfLv^GTtkqyDDa0wr3jpOu$Mt ziBvJU>kBzfN5U%s;?(UptR8(4=eJnO*~%cNa|(C;+W52c?HtGG0LMvHuT(vs`vFa^*#TzR7CZxk!<*k2)G#MIsB@F&{BCmJTY(g60 z>|CM-H`P!ldpl7&K|dJYQRJ44Gp6zv2ro>lUtlKfN5M$eW>%27-thfmHdFFAhJ160 zTxwB;j4;n|&FS`BKm}_}($QQxo`=1eyjaDr)`UY|L7HY2k{K3~yUpN(oF4!%sQLd< zRsZ+pEkcTzI(DS&5Q~P7gJqj@>0van6+aB@HhiFQT7RI=yLgE&M}(E8v0vr_=*76> zuVB!7`z_obbdTQc3<7JLJC-f(zNsiuI&pFZgIfc;sb2P&Ro-6a)xaOvM z=_UJ&tI1%z$zYA;(=H|fdsW-s?eZ#K?}9zov-5YR2kb`;f79hLTJiS2Hs8tGjz{hB z_3kT%0;tsUU->?S>ARtri{(GZh&wO`SQtd>WDU4@#|CUM*ee8>$UaO;lw30wZ|oP+ zp2J+xykQIri3iQCPtdR_#yOdcVl!oi7X+R!tv}YPfv@;4@=SQ*$=qU@aG>>KHD_0g z`;a&&e7FdSz{-*;ZL@!gL0xQucyZyCy3!yX!H!Kb!71hAohS39&0P{XLQ1FZJB!uf zh`7xda~oQ`F?84PC8ku5`-!lrhwcH5jUN^zif+1ORK-ziC9EKQ!j8k)<5wpB#a?#* zysW4KoOiz;qLV4`you#{BNIwg)6&EE^2!}`5JrLO8d$rht%%^H!EJxLeFwz3dB;8R z$>*WNr>CE?PF+jj1IUt`EI7EC`-_}1^$`EX`NBR$i-Jm`!g6ER{>s|Kn|#;r!ywc# zl|oi0nF%p9IhF~muvS~a0UVksaKo)cDZ{d_L>&4tg7> zo$wHMJ~{7O=Of92z&@p_a%zg_9 z$1$x80WU(@)H6>H#k6LC^DYCJ_0;)aM$BX2IYgV+h=)IO3IcTQqle+8W^=DY(w|5d z*A@rFs!&Ab0qXd_6qfS~pcxa_>&n?0L8b;UPMZOX)_cEDCJn5|a3<5*AMz4eW|V)i z4@y?UmVuJTVP;IFQvfiO#dO$B& ztS^DTOmbfViT-~W!n$6zz{o@Fk?wwW>NCq`{UMTbFS^bpFWw%1UEv?U=-oNiJzw=c z)-35;-HB5vx4<#>uEi0^xnGb4&Z5cmRTNaK@ zAk51g=sqcxPcj7Q%?rFYS2|il&p^EGp7m*YRk=D_e3rP`Q6#RS>`wrk?p?4`CR6>0 z<&;>5cOWK9p~HDFKS%e}NIt;CaQkJXObIqirvy@=RgrOhmL)z~zhHc24+j3A!XZh| zJ#P?B_J^QHdv)DBa^MTAy@t9Qpbu)*9d!T`qmB1whyk>{&29PyIxV#spjb%g8S@2k zqaB;1qOn)-Cj3ZwB%|WixHnQHoL>%Vx$wkFVroen4za0!2_0@E#pN5Jl*hCpg0j$- z#E1yc=qO19Ob{w`~u@#+9$KR&E_F)MP zuefjjx^oioXpEi4f={m$I=Ar{Vz%$LI?7f0&hhO3P=5&Y%j$`_h73yEM5w_s<(YHS zU6u#i;rB4{ROfc%wlnt9tV;5pa|6zY)WruumwjCJTHDkQGQ5}4X(W-UR}JfzNYvs{ z9){B9^MWERQl!3@chd!T)3rr!I!ptRkba)xV+OoFm}kOy8#X|hzA^Tg%zN`r#}DZl zty}~0g%_K1kHA9!!mP6GTk)uqk#22-(&N9dy!QI;`uQo+BSb(}rv-HySPl)n-F{&~g--3xuwrQ~;=O_!#>^)S>~w2(;Vga%7#<&p1Q_opj7 zki`9dTj?P@FvEz>6}LibMX0)ROFR%sMvR{CYDDVTqm|7n{(BHDj(#I4dM4pN9M=42m?5BwPB#KE;OS}IwHB~hZU#)++EIUY*h z94MBElXVOydm6;{jcOo1zXDp{JH%CD1pVgHYsYut;Y-GKLd|G z)yDk$6>JfEBgVZZ$0X1l*SSYB*JgA-I@Q+Ja&N*RJ2isvO?^9uOc@EQb#-qjBhxQ?AgNW6!Q5fC^U1y}x4U^1m)8cau&DQv|R1zWXWQ z$G1F05m@_d)%%`5(}fRN5L1uB(!bnUZgm#R1l%Ho@xTBipG$TL<7Fhjjs~)qX$42t z>$OfuNZs|_B6cyh6U!p@_A_6zgX;-43LUwx&YG;@i+@e>Wd7RdoF3`5*6Y=R<>w#! zx6fFx0f4ts-jPhd2;+9VXb6ybMEWkq0$^XHTrthUr zzW%q1sn1x(Q~`_^z%Z&`XG8vVGKzcVVikNaumpJQI^cDqd))J zzIlg)gg>Uje)aIEG}uF(3w!pD>O{g3Oe>^r&qK(QND{e3_CykYA%+t`aSsQC3{K;O zL(Sz7hh(b1M%g@bHWTR@gB?0GP8`q^H0Eh!yQ0i!O6w}s4d@Boee3=;Q7dj&o;~({ ziZw%?1dV?&D~?Bv2R!j*J}oQStIXbk&Z&C3kU;52M1vNwUQ?|7-tBX}*T?w?&9o#> z<&$8Rysx)f$E^$txVchxe(huP=N@jBJIA>bk#%!F$`Y$mhP5WUg41^V3^0SXhn8Pi zp5ZsWk^3HlHIBl2S@n!Ho_(Wuh1Aj0p$;JN{(zDN?2IOF<(CD(mN7FkQl{P{6CuG&Hv=e=dhcDMumkx2?{dmoN*+ zEry*!c%Iu~y|*#=)mKcKvAw^yoDJ4+91pFky9K;#9M8gjVRd8u`-S7!WBqO8s5Pp8 z;U9S@g|{KJDmX|$%d=qByx+LHON~-@1Y)k(>|zfwr1(=|)SosEeSTj0|AS4v*quDP z^bbIi_$`6_ZpMficp2*n^PFAGzkfgczjg?-4?G7m?~tn9%r1M>s}LRQKb(X{E1NnJ zlJ=GRKV$qikN@Qq|J=x0?&5eKP5#0u8l^FXpof@kncFFmQ zt^6$eA3pVb)cTE2c{}^GoQ9+Ia@0%RCdy~<$9IKi1if$>}YH6qe*eGHeotdjoHP9V3~6&CG_Bv(u^d8Y~X!4nzNK(*A4Di#;Qao?Cs+ za>881FazP9UFRzujccy<I{`v3!E3yCo zE|O!aYLLT6mI`)8B^=wDPnCa_g6B&*Ob#L*VsT;fCM3yhU;NGptFf^W<< zg#`3-f2nfzv^CZ`bZ;7cl=cnwHZr|99s|=(hMqPLd@bTo8O}8NH4LckW=|f|ju-#z zaPF`uPfbV0rT2?J7taVD%dE51MRKBWM%2v;hv-x$z(a;th0k@oqOMCDD`e4|Tk$B_ zBMy8~uDoGqm=w2Z4Yn=uT0yY39By-v9Q8YFpj7ZjO)x-^br{#5SC0C+gl1II;aLo9 zbW$vt3g0?PboHV4{_O4c$@|yJapCb_TXq@a_IP^Q`8{*JoO+E0zV{SIXb^roY!hV> z?(b^HSmfF9o_;f?l$htNx{%CZArj6EdFXrw{e|&64OY7t zcuTt?3-Ey6>0W1sC}T3vRLj;bEK;uty>+M@g{K|R1YX+bA( zDl^`SWkGRr&603isjW;YG1YgSX3$bm)O)I?9g(q>J6jInhU)feAZ-OB4MF@evxS?9 zxsf{?{$2HqA{hUgg`l%f7w`&R`J*F@-H>;7re;Z*iMarI8o#)LJrvGrV6tnvkwO9M zB$*845S?3i&zuA>+Diuoc@haghqb&TNAX70d%6;+>}pu#skvmMI92QuD_HE@V(UfJ zg!^ox#70wvA{>Swk;=};N?_)e8JIa+qJSfI`5Hx6<|>;GHRh|H<`{>4a|_L0BIw{%3i766qDtJ*#Bc)YDoc^SKtDwip}QRZ#EAAVk6Ckq zlCD6JdR`-kaJze1S(L|Z#aAVPg>CC7vgCpkTK$rU$^PB_9rU0FdF3E2rg#!kn_MMG zTFpf}s-=7LeuWO7GX!r}v#VaD)4!=|>-_m97BZ+(+#q(4g4sgVD$F=tVf*Mz=4Ml| z-<}Xgd`AX1jbMRep!dL7ts`2bS1d{MmMPVPDW=nXWi7c>xS*Hioc(vhxIgI9<1-Pa z13LSF09GDomKr?~MQVL-2~m3CNALPi1vK1@AJPTjQq~H>9|w@JB5k89lh&vb>5}rW z^fPl{zvjskN+q}=_W&5ibCdNgei1C&N0Wy}z_DS1BqDZ)G&s6<(A3Ep#-D~_x+QNe zuQHIrXo?37!7n6a=Agx>&rB=lBAY4Ufvjh|(AB)N=yv!T1xIGK#tveCc<` z%k)8{aTe{{+eZ}7iG-(Y2e~)N1x7eJ(w6zsjBuoC2`O-S4Ppgvp(^BRW_q#63+~dm z@k-9W_8YvKv1$?&mZx;fviB4C8ARb0_yDW6x{YX_DeldK0jVGUa>%vC`feFrbK(?z zq?1bLu(o|iCg^H=n0UnSv0g+BhXtR|roEXUCanbvr5E0!f>-1R@C}+X%*8QSSv?~eC0MPs zj5H&ZylxNr1NLd!ng9`<>bYMq!ao< zz=?|OapAR<)9vUIn--21{2GR!!4(U8vb7V%DC`30;DQu!QB2y~3QqHg;avTMQdkr@ zlQy-X%4u$%YK+1sf;OmXzUX;dAdD{8Lo*B_=fs4XNR zOvUHg(yycuw~gz4}lN%5MR1R*8(ZXv&IUi+1Ucrg63Bh*8wk zO@!cK4ZDhnC^EMNl}ByoDXZKrFPxqeC^+Ia(>v53Zh{GLv&c310J;76Lz#81@}NuM zNm+LOXJZ&f9L61y=A++Zu zntqZf$TUfrez*h|IwIp!%kIE=FnIc9R5=*Ym=;o=khma%ghCQ!9{oaluyk|YH>$zp zk38WtXar)4&7T}0JZRTRJ9<_gc(t+SAKl`vDb;Q{(L6C7SP$eWV(XOxmFu@^37(J` z)8&$sX*2S|1lwO>=ede6VZ0783U3%Kc4!~yNeg49uvL%7XFcEn;nR&ld;Wx`<0F(K zP>B^6(}J&e`bZP+67DF=4KUUS@;X#LA;E;Xr~B&VGlvW!KdW%ZF{Vp-m*d^GP{=G> zhTZY^iDjC-B9fvw)!Kf6+mmvIw$L+&5274t)6bUxJK2%C4&PCpX)+XciUu?<>G z`4PF3SdhpAfHkOJ+{5mCtg#jc|JYlnRPW2?jj(Ok>F7Cq4bIQpRuO)8V;|xy=G=0m z6Edo5PC8@Rs(@?cQN6%-J{kE|;&*Sud<#SnJbEBD69oIt9dH6)7K#0w+}63@k#AP5pKl6jjt#yh3t0!vEeECy5k{jOSK zOa-@g+bkycwZqZE_r<&(9MXR$SWI3BNsxTO!$P{ot@|TZ4B|xn9XFAL)nfeP z1@pk_U>IzoO}?-53);Hyr~=z7TMx^01Y!29H(>V%QB8;=h~o)K590FTNCzLbHNaGL zEv`fj4xkF_Xh}R>T4*0d+aiJuonW^seP$oZG8(aSM$i){rt^`v+iAHM%tf?3^+6AN zn?zQ(8kT-;F)Z%_V=gSpoFBN1C-sA-5K{s${9uYVjb3O(?_CubK zUiNbKMkKI~WaN`3j9Re67VcfwlNdCwNV7Ak+gQ0S)@ea8PK3=Fxn`YI{{w%CBkJ?@1!|t;5^C@6CQz2?fkaI zPV`Flm%X{EnCf@813l4Za(z6t@LTgYQVY{`&}+|-bi08HB_#E~S`UI{MRb z!#T;~anj_zy^ITDG9z@~^vOilMM(V(-5Iv`SAQ>*Wff`Sf3~)8Q*?qkqIkrVB6EFv zaW2SC$A-J*wS!M9h2JZZ2V%(u7wHa^p6w^3AJi_tVlGErEW^@xrI+YuF_sfMn9jQe z;Q7wV3)#$3&8%ST=!o}+z=*J|5rWjJu(qFW43<4ay}VTtn{^|CKI2CaLoJlHM&rnc zQ@C*04f7BCuJY;B_+UuAr0zzD-L!fug9AW#V=G{6u1%DJ=W}>ll5%7(>{KKvI#~f+JiUWX<0bQ2d3V)a87NSAupt- z$^WqbzLjG3lJi~8IGYhKqh>HvqxiAhWlk`@4^kKOm>)wrv2S$TNZjnAN|d4;C~f_Zc4qei$L$sI8XsV!Q4L!DYdYtm7M zau%0;d$X8McEBmcqM{nLZFhSpEBBNlaf^j=F-k(ji2sJ+BLoj{QS1BW;ho<3Cj++h;wGvx|0@63Z6 z2X#?wL`9b)0^(>B8LX`pp$J~e^)gmSJ`Z{Q!M`w(p5eL6Nocg{!BXAur!ts&m`9wH zG=(&iOdM+AzB3Ra>gq)dt=U)YoqK_W_?%qrH{x!h(8CDReC^&jbjp>bl;YQ52gSqW zL#hwlbEH|?P7}W{?GnoxMo7#LIZEV+9%Z=`l`ILVf3imH2%h=JHjy zQXi<}R@_G<)@c6v5NF-7WNJ4?`$0QD4QF9NSpn5X>XOQ(95QS4o=k4(L;=-W>Jqni zaoME*lBYzmsNe`tDhU0?tk666e9X>LYDoGwx?t4SvbL@zgD%UqQxE2+%2+cI?ngqt{Wc!0~aTq^XA9z zzHz+yG0c*!<@I-*)*~sH0S|`Ba%T08flfB ztF4PEaIl7T$RHT)j8}edmi;=7bv;1_%v^3>`;(x^v3_r+JW?iiqHH8F4a=$qA4dJt z>w>00^o^@~4odgvr(a%;QKlHZC*ePNYj-Hmly+BFSs!jzwUdj7mo{UfOMI=hN}0E^ z=8am=*M6qX-=i#?lhNr4B5Ods2hS*(Zq;hc9$FYo$Ed5c@7BHk;+3c^7P8HrlCm&~ zq-!5&9CNR8o?dRdx84bsYN>l0n*+ESlIBb1G#A4=7or z=m-4-hrCIFT!x5L?Zm)SBWl(vsSV3Z-WA#Gs_^WElI`lY1Ys9Qr3*B)l1qr77*lcc zBt(A6@&b**Et~Au_lUjhZs74%_JOs2kdY}5)?74cK|4ksG{bi<;T7Djn1u)~(+rW( zRFCT~Z!k6Yr5Ibp`dMr_CMl^eFob&rNkxhEJ3kMDHG zi>Y=1+f;c-)0<~LLPbY|;T(zPjj-C#^_%{cm1(|wQT!ra*)WwIa8x0_yKBiAFO0%q z_J?;=J0=X{eF{YH0yQ4~3D(te=16ao1wYS|nRC4ES{( zbDMqIx2MrUt<~9JC3pd5RRHL@*LV-hyJuyQ)$8pRm_lal+$`Jw9(b6CXNI-=ev$v^ zS@@p&4V{Z(S#f!`u~svk79{SRp=6w~J`+|fza4Xe!RchvpV2TJYMPq9G}IqeV!W~~ zlto~Hg71^B4EqjCDKWg=JN~&tI7A2Tl;Xp)aP^q)g43(eYSM2VrgVO&Uj7)EvV7ls zFUv#SfU?03djtGD^achP*%E{;(l+o}DxYWTvNT&~@&YS28vw4J`^l;GK=0_VhA?IG z+h|~55D$())Cu!=_&C1O)Mw`)8G1aYbP<$8#r9NfMw(HO&O`5BDts8TV)+dGb=jKB z7J%Q&4D~Ba#C<}F+|l24f3%8_D*3V+X$oOd$N+h#yE`W0aq}7jU#rB1EnPuW4pk;t zrpkkcKhu9F@FDFoNs@EWV+}7E2rc4pa8*4Hvt_PQD4?aQDo)-MqcbCoaqs==C@M%- z-fiH^Qh@nEy&2qkD=T@1y&rkT3JW<>tcAAQv6)>o@xuGq)vC`+DGd!h%t?FA6E`?u z=zJo&K7x$#wpUY|Pe{{Xq4fpEPvRP;mL-ck_EqGzzUPNd{aa%DyrlUw466dc-_qK- zH{*0@4uv;XxUY&@B2wAtIJCK5VpUJ>I?aFjZk-1$%C;G}6x6Kq5UPyWj8h2ZU}%-8 ze}-poc1w;h(#0>jnvjKI4&Wa=Zb@6FQZ^LRy_Ecz+`neoCz!ZWOgj{12(}$|q%g|E zFj7sp`$R?H71mJP4xjHk_1n1ZJG9#^dBuOW(~LH%D@~RrSKQ>`QNyV}g-$Y6}Xd5=+_bWelSA^-Hd} ztIRw^{0==LA~xk!lao_Hwco%@cun_aW2hy7$8cmzE|?#CQzefRw^A@CX2(i(Fxh7; z^-W4MHA-9I#F@nJ9v}3xm~X$n3Ya>HTwhkOw|Mr9xCuT1bN-B0z?TMdK^R4%QVjjo zK7G%sXZ=ITc7!w=6ot;h=98FO#pN+ed%Lh}6p|x5;UL_X?1-gOLA^+n^z2IurPKgT z_y$*BeH>g^cp|yv>-aaeFPJ@)oMvC zi=gRi7}FEd>)NK#i4Cm_f1g>_tfCfb=UKEUDPKT5uYd z{Ek|CrDYI6(mZ*BmsHBZLJD0_L__ownv1v|Y81J2?!axwa)r7lq{|hRTR}8Wo*4A1 zH^Ay=DdFppl^C7`DZ2C#AHXD-;Df?n;gq5wZkr2Pr~dedPTy-^bbOOJ!E=gZf@_~LZG6k5k^4iRz;%LS$9SwEWolX`loPXTMAQz7&0F&hq!nM%$n z8u{DWHd+tX|6mGmtGNfMhI=xbgtbagHwRPOtrccP-9pZ|CW=lHNW>W#N;H7Dy7&bGxEYZ0o^0Ky@&t zv3?f@bSW_L5NG_z9WAK?l6OME^C+)>znq|oE;t=Yz;QK66E-a zMJDe&7!MRUbqnG5Gb$ka4Trskg{f%lDx2K~;y@WfAvFdqOAO@ zH&InXT4tlq+r<(j19?cD?mx3Inn3Ee6y~|Ke+UrZY}d0se)#(CoB2hRrS9eOPWhUweqOSzCKDn9emo` z6JE%2*hqtNrHef34|c8vuM<3)x@Tf+CXBgkr0jk_Oy^@VWkTB7Yz@j475kC^w0LH) zn64Ee*b;g45L`)9kG{^kDACXrc4QNyjq|#h0pP6o7mJNl-8``c=e6J<`QiyWLiS9| zJ49YJca=^mQB>p&PI_VAr4A!oRz@eEWD_Pe2`5gdCr970zQ9NZ*TZ+_v@x%@;6tBQ zBj~EJtX^O=E#nIoIlm$?gB3Z-2eUTk!R<}*{p|F#_X?tq=0neR29*+^23g~eiAz5z zW+*m&yC?Ak-7A9|U-+#%IVw&iziD{w5BS0nLkuFmtPT+o0C+3^7RK1JH06@#1`LlJ zxiHNwDHRw>nMMCzSR-ta)Xx|*AhCP@g(a$)9gI$Qq?p}>1GZ`>B45YkK;epP>R`u! zAE`#Ys1isyIOpyY%}BHT+N3FX(u|aiP77=%J%}y-x}7hb5Nq!U z#1qmnjp=mDJ(+W($0g^QVJI;JEpOr+UW#+e(qpxKo$gW3zWe>hvw3=5;`Pv(;#L&d zFkBBJ;-wwLuv*`P_eYF!a*<`QQ%Bz&{!Sqb_D~#0+a6Yp@Nu-0Xsah{- zAY9Hv4md^@W3+!$Jr^!(MJweym4oJJGKfn=)nB4I3u=mCjqT2BetcDWq5_W0(PqfP2Hf7${ufWy`5Lc3Tsl%Y%$ zg`d)gxXi7(e54%kQg>yli-nQ0&cur5H+4?)xXCFt8vqOLJR6MGR3{ zHTv>2XR#x#UF7%29P8;Pq!{YFqT(vCWgYTD7l$QcmV)yo9q))?-jd}@jO3MdCD(#+ z8Oap>Vm5g>sakQqYPBb%GQui)pS)?vmqvS^Pf}AXZjz<=mf^Y=QpF)Ay20bD#aTvU z!tm)-c;0BOfC@()hZd+|Nk_Mc&{BCAl=H5nST{2wnAO%!Coy(3MmStFZqZxZ)m7Q1 zT@u~lEtH?SU4*e|CI3w&&KVX;jjBNm$!E#!95^YCadr|pr@B-Y0{{E->v_x30xIlv z%-Wh#GnBUWV{bk4L5VX9Qza*;g{+qLhDNdn%I_5!WYePy?R^Al)-KNx+e1bZvl8;f zEj1Z7{XrR!>Vd`uLpQr4*QmKCq#1)!Jl~S56tPZ9)@{R8P@3%U2(a8ORKD#hb(TUC zhX~DUl6AZ0Cf>dnt+==qJgltaK{?m?Ua}_e)xOG#>zj&BEyQUPuR4}gn5#K|>iqE5 zy=SY4|E4sqzo4N&oO$*!a?`R~Fk3Six1wFLqj=P=qLs31TtuN(zoWd|_h>5y^=5jx z9=*Ay71urn>dlf>(f(!gn8SD8C_ z0e;oCA{S(5e`3~df+sX-q8;YvEMTyeJn%_#q?obti_F+?YY@6V!p^ABf_;Lg`nYhj z&bv++&~Q1smc5^-gZLA#$J>LKyM|NA#h%kk)t;Iu#WPpBo#I3_y^2!@ex{O2Ri0HEvbu(unqfT|B1oUsFgFAI~bl? znXGg2lWn}N;|vfBNN~K)P%L9jfh$ld(EDb6QJ{waG$i$Oe;#pW|Gw9E48w@!9QjAk zneJ>!ZA3T-vu;zd7-%qBKflGqsJ5kdlUNJ{YhCs(Pc1(cX1vuMY zMPMap8S~KAABYD(A>m{VS4Z(E7T(mq!@ZO!S3YZC(gO2Te=mLc$b9W%3yzMKiY?c= z`elG~K_B^pf?_`si^$#rCRiFyl*5ST$t2>+SUQc_R7nIMVP%dU&DyqIvn06JOjhxEXfOPOk*)ND^t0~oaCDMTWIC%y9;cJMOlbFAV;`-L zI~gTbKkfi3K3l}u8{6>~&DJYNJXAa;(vNYfli?tYF>2ho(w`}*ahHCneGnnI(_sM5 zom+Y{!0d8Ic#tj#nul|;6WHG+JkG|V+L*9hc^xhC-jGp)*Xq~Ww&o|uZnjRA8?29? z`9)ea4Zyn#Uq2zK#S80{hokB%bafIZB>IbDcN21}o!{ro?N<+`l_?#2cWmIAPS@78{-7^1yhjwJgGQMWvFN5&bYUYY ztiNc12pFA`z)o`u##XSp_yTeRHbe5q$(tTC=PkKspSiXp!^qe`w_IlmKuS}l6bZA7 zH#^2u@svB$_lYz+3W%K~22}-%^+izx)Qv^GoG|`J%>f}T4Y_nMwg`=#bHZyAi2gUT zFcDxunK^31zzPBt8iXj7%9P9biIDwx9)*QvMYWkq?OXKJ}*{XEZw^*M6*Lj7swSCv0OpxIL@zRI?h!bC4V zsgsSi6*19iXeHq1i=6AuVk=`YDJ&V}kQoE2Do2xg=sX}4z9c~aP3SkgMxxH_f%*=G zk}e@-iX{^hK#U3wdnNtkOYJDZisWWme>3A_-yMO7A_=N~Jh1@D@%nT9{9lm3O5IY3 zR%Do^P7WV0yBwpfg?_4T48;ZIUq3mk&s&HpgAdRMGx#!2n6ajE04J4Yj;IAh&!$GB zCI6&$A{jpa7T2u*le#d`vyddT`bm}nu61qfm+zcfTIjm)V;*0=Tc1H;V+B#8y<1Oy zRWwN4#vzph<5|!Gp*$hY4}2c})ee#lcbJk@IGqjUbJH@z96u5E9edo^2DtO{mzq0W z$=CfVP|jy0=Ps1`I+X+IcwmYn3vkI4k)(;CjZQd7=q0z<$?LZQpc#;9)(H*2R?##k zFvGCKc^x#BbdV^yR4gQ9mW5%9sjNU>CKJXPq^{tpUhqp(-pIH}U?Yn_CmL@HT%;*a zq5b(XU1rWO2(fpnM?+hr+5CJaw2dD0L4i`5HnMeuy?cbrB9K3fJHy}LMb_UQeJv!i zn7BZj#W)@doRN#b$kt&@tTWx~!BRoENjz^u$p`0;Q7yv|g%F<0K_+EQDayL?m~mJ} z3}FBTeHBQQ2@B5xGb)YtpDS)|>^|LMa#uGwW*+cJN)4&eBiTf9M$IJatTGfro5Z+Nj~k&)-l>tMijm)czjVey*pY_5rig<-VOO9~%+{p6 zY$IF{PN?2-V4%vJ4ym+j2J}xNx)1;wuaQUBWbIcb;s}sLnm(_ckuFY!CnVk{ zq`^d-F0Y+olJ%Yw&heimA1@2k59ap%VwfaPI;t6satF5Dy&bI;ZaB8@o~kEomZ>U8 zlaY@u_MEEOLlrb%a~5u>z!dD;Gek@}n57bci1yQhDcE<|Qkm+A%aR~gDNHV4zGFmY zoL;lhnzJ@TTZEt{EAu1TM32vFasdI-VUABocHM=5wIGBPJIe4j7|@fgP>Pk{hXG|$ zU@yHot898ga1(CX|z4+Kpg_~a-t=hCe+v}^yte7gh>A@c$bEJ;5;;l3^T0& zkQYe)D|n{}$LYDjbeGq;AEB)(Apci@EX<_yG5;uo@2A;T+qLP!+YTfcG-OZJ3YAT-2PA2ZS8z!|TdEqb~Q>(0qWwrSTxtu#~Y<=$^?hOC$(AYfE zbojYI0ncx$jPfLMWOW1b8vyBFekUG;SA(K8n%HNDz|I;qlj((+8|BZPrS%lYK5uSl zKyO#xf|uKW;stK+L-idQu=E2sI+!mUXS&x8brO79Ps3tyC*nJ!*k?+Ntx;_@u!vf^ zQGWj;m+(&Oq=8#p{shPxY1bf0cmC~xR#k3k9KW<;=;qM=t86gL>RLp}$aiQeZH|Gb zoM6N&Jwr8SWRt4_!+3MP+S`QkP3-|acBn$eCR??SQ}o;VEb6e?=1&xegQj_Jmr_dK z$YG28=r4mxDT|NYR`wsf|DvPu=y)O8K8&r(p=B^a#L>5MXLod5V1+3 zg2+~MhG+KZv9SH@RP1(7!gf!3wQuoB^NN*S{xj<7Hm!5Y@K9tk30FFPO8!iE%i3DN zalz-dU5^pTRl%QlH3!Vpe`+E_+)E84v~uooECK!47P#5!c1`zdbT-Q6_@yr1nDETn zLX{f}(fz$yZ3!CDLQ~`WA?EP_?)ho-Tt|W*`(fSb_?MD9RR$n7fy;hVWxRI6mTlFd zLLpOSfy-XjES(g$13#zvX)H%^G!6y1Qg{`mPS7mz8aQmCF8_>9M@;9FXrB(-iNrGJ zAFt)sBi-u+vp1JKx3i9+;&eTkIoPiTxajph_-j?+Uo zz_L>S-N=vqpx`5A)$Vlr{UB>s9m;wi8{nS@0a#5zc#sxVZ3$7Jn!3V_SVcEevB%7& zi7L*UnQi;>fsSk@4+e#v%0I=dCc8fn7rS2SFs$f0$7FyZKVje@dXt=XV{c2Sa*PKC3sA_i&}pD0V>L)E-?5trg&%xB;4^iJ_B-BuGz#k>Wn zr59+Q1kMYq)Q|ILS#>^5ZWAcjSYaM38Ruy7m6#_BcCl$otMOGLhEqJ4koMsh*z~RM zw@S}z-sjY_cPK90&uvpN+&3#HAYb*0V*y034>SCtIdp%>w zYLPvRg}+EHaK6`%IzIvOUw+QExjjbr9LrHq&u@r9Y0RO?1Y&V9HAXr?d&&5lFCKF0 zdhxgGZ8TAfqbBfoVw&4UrD)cYZq`7wYxaSvzXlh~kDq^9Yx{{QKFH*>dC2XBSwywF z5PlQ4;(z=I73%aoqhlBwYqMqBZtXb)i0f)R{t;IS{~@kWKb-s)SDqm(-jwuQS$!mo z@aMzjT&4SvB5wCfoh|8syMJU=mr+$z(uR(MYaMorOiWPrwxO z?L+{-&5<={OKqg?Mp+rqNj7(=6*1ifsnL2UrdG4O*xOEKbVj>F&mHCAoONZFW9|~X zgbLHT$SYGF9vuFM;yU|}#f9R|B-}be>r~$lQrYbo@^25MHtxaUG!0MV>K{$p*d9GS z-@f&6eGvOskM(*2=zEhrsA-xnPhsi5^-A>y^7uZbl-RCAqTfwv@CW0Iv_Lbp(eg#i zN}u>v{FJ1C{O`^Wz!Jvoah%43X`CSbAwT@Le+rsmjo0f~qMycP_;)c={ny3pM_NrV zX;RmqSJU|M{oWjCv)%h_bTAC+MW>n{y_xKKwwV@3f@he&y*bE5jo&~i6)vM!!;}_B z`NiAbkq0eSg|SL&bTIb4m=Hrdxj8A$mp%hRT=PQ4sugAAt@3uo{>6tcce>u~*11K@=8AUHzQl__4Ii z{~G|nIeQq?#%;CT__%kh%*Mw0Gcv;Y_^_?4D96S`GSX>%os0t&DVHIldbZ@9@(4k) zW=g;_p|UB)q{!}wmLiCVDR84A!O)s?Re0cRKq!{y3pi#o+L_Yady}Hid@9|#p4Ng< zX5#C54Hh$EQDXRGjl1u(>Q4Q*S$$wZj!as(#<+PVY{Zl9H;O$8qQ19Tl7v9@8x0hh zC+&PL%p4R5+y3Hf&Ol9;J%^p3L5{4-AS?v#k9}Ej#wC1$%$1b-X)qSmLblppKrj3w zrSQ7RBSUPW=29vm3_3x)_hvXHhF*`)6tbl#0)w2p`9k|v@2vM+-3Z?r1*(FYxua71 z*B#M*?M_|09&+06_--CWv?^=TMXKEN9o=X4nxBw@b;jv4rDQ2J?Tdc4y2+Ur3iCqmNLk=B!gT z%qHkCio9N>J^WGD#G(O`8JbhH)~zyP2|^xW*Ui00k?35o{o(3=+zoZ>uxKp5rnt}4 zW8n(8rVTh@>i%&E6fHgG`XL_nzHxCtxaH1Ui1`ftu|s3dlJw|Uyq{i_TE%2UV{)7L z0MIqaeBz)nBY-EJqTJ1fcWacD>FI8~?d#HY#ztEdlvisSgaWf7Ner7ETvT81nH^igS(}MxEq7_I(Q!kZoczcAC&|m z#+vVy#V31VY?`(}u5_1>ZsIVr9b*TdGzw}<;9!PE{f@SSPs)FIpf#3GRNj3;lDGtM zjdqca9{h7C08L7Mz0G3NiETm^S1hP5zg20VLl7C6TygPiK}cm~t_I9rXapyXpP2Z@ z1r77E0_dlaOT|P=Mark$tlXcNl~xa5bo|8ox}W&9biw&R7CEi^CvNvm(wfILJ=OFL z$<^n6E`!5rpjMC^a>e>osvgLcyr(+ZBLgVv7w0ncbudZK_5b1REui8^qK46I1PJay z1_Wv99Uv514Gs%(O~zqxGrv7uD(~n zyW6IZ{dNh)CXES<-BU{Tv6y2rgB2?*we&}Y%e|%Q1XY#6PQpB$`s@r>XqE^S7f)9bQhh0@UkUMJ{K09QKd65PG1Iec#aep4@% zSlBpN#|`5!35f-pFS@eUNK^Gw(%YLv52jIoB8{FC>RK9PViN^O(@XY;65C--(|*a= zPV#q!nrKUM{|&^sp<>$NIE^&Q_EL?x&zFrbXRy$I2agZwp~NvGh0-r{B)M};eg{VU zv92`8YNXo{ajpT6v|e)m4zL??LH3f26tIh_E{Htu@8Bs&rv7#JD z@>J-@0~-QL9e)k9f!9wAXBw+j`VZA>z@?eB2x+1r(?u9tx>5`aEDE*H7#4sUV+lZO zGjzKtFm1dYpxDyZBHJpG$;DA+ur5-)=0)~L9KvTcT__PAnnmXHV4bxb`?{wN;9=4Cchirk?;AUGd=y3>`KQ10O`L>~k#s1| zLu04T`!uR4+~+(s81K<~ zU-(ETL@oF2$NV$jJyMbp2jOIBq&bKPf2Eb95ZMp+8%NC0pu`|cwl60=U?_*uS27vYOOkRN^%@|n zx^C0)2nd`XM~%0pJ*BOYv&>?d zlh2)goRv2|+D;ymg$n`1wfTPrdcMVfjlRv8mBCuwe>@ad z(F4i9ONcY{3=L?ba)u`Y4tW70#O4r=xw=IR4I%fYnvoTysw0=WdH7ieVN37p3~;T) zxl3y&sB#p|B!j3=pf=Le2JIM%Ze7K0WoB4_&HYgFz4X*mqPLXFRMw9nFQY+KBkj~+ z7do|Xah5wY<0A^$X{k~`o=C7AU~>UDQ1d({8>|xB$cHOS&uItPUH}f)IHy|f@Qjbx zWv4Al0qcUk6VY-LsttnLn6rCoJnEwG$JZAGbfv;AkL+;U5;f0p2gY)R(};t4U!XhM z13oSRI~tt*9d{7cM$kH1Zzz**=ixugd?+$GsLnz(13uST{C{=37a}ud&GBb0THZJ= zEE7d*e8!P$`H+}bkHD?)lo~tM3<7q;o?(?%7)PAY+c)igi05oOlt{HZ(>ovG`nEn)}hJ%|KM+z_*Co7iBwfO7@&z26pN;`Q1OooH)nWor5F3Nt(6Gwrc14HDo%Fn?D0H#n zTs)c2!RsQ%#7uR2?xkx?0BXPPnR%x`a}6N@GDzKCiK$ zIT?`NC-9%2mqWwlU;jrrbBi$d^0PP#2|f*nyXZ*NhIYDBZD+F#p z@Or;=SD+Yk^GVT9m8bkEQ26o$xZ@!!%piq~bgH>@3ddK~a0;%#)+xDuON%3wkSv@u zK@N8z3Q*&hSUURSI+*pRiH9>N1@EVCEE(8FL^zHqI_lHE>WZ2(qO?ov{30Rr#ZPmQ zG-z`#OSd*2%U@@24hu?Zb;1`@*#-oNnk%7C*viN7vPH3NQVs!;x@1uNPvZ!)VWx3# z>KnQ!`QE5AWHkhYYzSBP3?|2sdV;3%?x%9VPzH(VN(4;HRCQzvusr{$$Z+P&p!u8M4^kv8-vqRpQ>J8u?c3tPR0U9H zW31s2DIRjbT);0VMKrZRHZ{g$8veUkRQRmwub_jo5_!1)JQjy!=&2MNHQ)o(`Ju|c zsB-*Vp!ufvpv(5RpX^*fC)62PWr$vi?GB}OBR6U!JxK?){Ko!U<0(mqmct=^9Hp%9 z3v$3xfHN3txL@jDRLuL>g7Rj`>ebiLj@d)OsLPU9-$}JZjtZp)z%MQq=@JLo!>hLZ zTlesUbQ5r|6&!S-qaO*4R);Fz9c}qP!fvxlAiieaY2N(t;MxgFw$%WeF!R?ppt*)I zab1|fa~f*JxXsyXIelX?9Yypdeg3+O4zU~cI@y@D3)~fzY{rs8LZH6AUinYKNGyyf z{N|+fvq0Ce3m>6+W8-mhd~Yc6yf4Co({1hlP;j2~*5K;$^^%6b--QZWUGT4K9;yhmuv-xx2o$vjVWCd#f{%>4Kka{0>!A@S%TO6+e^ za;%CZe+OcGlwtZ6w&<2h>~1OVt%?xfU=ShgZ}5nNGxSa@4Hy5z{6)+lCZiu0i!-F3 zPe%m)(@GP^J@4*I$zHQJ7W&@dz`R{FTL)=b#NdwyMZFj?TLEN5(ASBSzL2WW*VzQk zEL!|4m8x5fMoQf?r;YNrePEEEXmL-k zI}BBIDFqpxZ-uO)!O(@wB9cFT+33bdGJA0G_FWv4#KqkmVO@Cx%EgiQ#v&ry@AbW# zA&*v?-#C+;rFXG78CcmEE^ZlvJg&a*BOYI8eB{^und<3q>qdZpV{jzqQ70*{SRayy zwEl@)qob6aQ94za5#}jACG+g^T9jx z2a*S3$uxS!zx1?QsQYZ)bC=1?{QT{Z#{*D4JQto90Ccq9o@6K>C^b$LgMTLMYPoJ<)5BcwX&|iz;bva1EGsU*gw3U*leg|dk$_7q(@VNY91ALgjnT!=Ht!dxY~y_qbFe;zw6 z#HFL|r!rR+?mjT$=vybZatt|>YunGciK1)wL5`0N@ITUuk6LYnpssE(5ttw0dA(bI zGN6z2ye6l<55&#qD|@umDs$OiDp)VYtOlc0PC^mVSA+CvU(lSBR#J)&F((9~wIk1)Sq!Y%oTZ)k{ zv@S-kGWaNclUvBhAl>PTuOkPk_q5I83`5~@S6}jfJVt*j7@O>QF69`VUH{& z4CR(tR7zzjTD0}$f^$p2)J5nogBFY*rB}=s)g#I*HYE^Tz0Q!LwU1ID35|sf`%<+C z^U3RSy6&As2*#DfzZjAx@l;OM|=yW-q;Q!nHh+x@|xGRZFdKMZ1HDN z=>qVY1fx2xMFQA6-p4U%UWLb{g4=M{I@5SngOMHA6an;IfxM3S2*&cvhV2mf4D6QY zUQ;H9Qwq{-eL{B~JGt4Cu4 zf4&4lo;lg>Vv}vuEP?T&KA}6CIOOCOLVDs|0@IDPDHo)WN_VkavtwlXU89l;r_Lm# zHy_suWX0!LIvY~jSb3X(x^M*=Hqdu3>%Vae<7!9lLcEi)s&8L1wHGHmpwc znKRO#J`U!m>u9S}#_(_y$x%DJydJx(TP(^_>@Dn!Vvc=^?OHEL)ijd`d^9pLW+=N!3)7sj1?j`?iaQwKSVr4Y%&yT?VAn+RBv8d=it`aXY6XbV0+RP9ZTD>1x;MQw64Yyu~1;o6k~mGhe?4d*VL|l zsK-#%-uwh}8m0{DaWJ7v2FuwbwL~}yCS#?ixpvP2t`Kh0i0N}@@<+)?fc}@;q7P~P z64I}N+p*IQe`O5ng;$D7VJQD`jtwo6q!2gtdoEgC zZPTTWfDIz#Ro#{SGvcZ??mzYh6zo_^t(;EYPvJmBc-dgG)yj(8k3ZH1qC48ws4?-WirYYoix@y5MQ7c;}~9nWtD8y@F$7n?O1VY{R%gj($(-Cpi- zXpnQmUxlKt%XT?wKt)ir3NGj+*7*E59CI$gpp}J@%OHc3jzbpAx{a!FQS!zsJ~Mh_ zwEO$xBl2(xwt+n{Q>V4Do#g%5buERj)xOEc-Tg<0mOAX(EmIl|1uoqKovQdfdC!tQN^8JZp4l_PkXX?ny9Xy12g6$3s?ssSo>6xF&qn15@Q*)@bS z@*5+-8(Y?rIUkETMv!-1n@Nc?G?oy|W9t5b3(=r_lI)PIVQ2OVYo+~!x^%XedKjID zpBug<>&^6L*96n_iM zbDtn1r_7;^(S1|_ews9(^vG8+HleH0T}(6slYW{AuJ(^>dBoEI@v>>y6MK*H#o`S0 zzNeCn!%7G0v&v^_6@zKo8l-a%V*EzniP4#4&@D(&Xi#ew1pmu_)yrt}i`TEe$wB zVjMF;71PL0@Z!q=cQwc(1HN5#W@7o#ipoITjEDO=dOEs-kMp*zuM|r?%}Q9iUOFEj z<#Ma=P%T#f4|fypcxG6JK)ExiW@8lNQeXU`$>2HrmUq%|>yW3x_Wz9_n8t!2Ep;5C zUwvg031166sdHL!Q91X%Vj{+70!nvE1_cN50=E;N3<43t4zLH3q9HY-WTH zIj2bBW6zg1a-JgwKL?>Y=BHqGGnDqI8mF^Jq4B;2-YdyW%PD4}!1wVkeAf`Cnrv;LfF6dX9rf%Bvn=Ek9Du4puGyVk`Kv_rRj#5&=vzHxU3T$5RO6u3^Kd> z{j;0>uV3|vO-ShBH<6IFRn_%>hI}bKn#Am94(F%#VrwpTN#J%qc1QAttx2ewg>D4l z(CHgZvnnYAP#UIEyx!&)pnV#l*R68KAi5{$hbY3w(~MgMUrtUv=F!v9sV`W}FCg;z zv!xj#bC0Y!`w-+HoOfvbEU>>EbsbQe?GgUg%c=e*xClnBhDHYkg>{*3TDc%xH*$B%jpV9?=X;qh|Fk%D zM^v=d5`AN+>*!fvQjt4k_X0zY+#*>%s6crMd~%n2<^b=;YvMPwt*Bb4HH~6%d-R72 zI;%^?-z9_4&7`s0Xtlh~DRhH5asY$THEl^{LI_AMV8WAlCG`>4|fdowY4Py{@`H8uZK^xbqwR0z2eeb zYE9%cxPAq)8U$kn8x zUm%RLjEG|`n+2XNlKBCdkE%|mthdcL>bfoJ`xlon z`<$MTj9CA8z#krK4vigP^j@_7bg7xSMp@;u9~+#9O*#s7Z;|H=)vcTT^7SB&*1Gv{ z;HqY`V>uwqV4i?J7)y`po6^M3+mUlSue)4`>r9fJbO7kH+~${JUjRmq54lB})KGZV z8b`#pF2w)B`$thulr~j^ET*_XB_$pqSENwcT8-#$||!ZfaSKF z6o(Eyb&WG~cCsCkfgL<%=Uu5h90$NGw{@jB7Jv!kLv@j6aD~iA!^KfIJs1hm?(90i zu<;>{NaGr3yzHb>K#%1%qv$hhX8sc=m10oHtKxk0(;Kep1~Z(45&b^KHbHp2L&Axt zc+-He%OLy0$o0?*)xPA{I<#L4PkleZ1}Q!@<{1<%Db-}C#bvaI22T3v^_bOm*QeUi z;)KVLu6e5j#|4cWhtL3p8cy~o@TXBeWXxMddL)`b%{42QB-3U)8ZC|kUZ=98OO<9i zdkFWd-+HlM*j=L|zS87`g_&4+Cf@<4T zp1B*TY0&7lPA@N0|H%C=; zEOGoDjvJ-F&vx3LRmvR>l2ThGO!Ae#9Lns`WMl@E)UZEiZ=~p7sGdHq8Xts7fL1j! zG<1O4mSO;nbdzsRW%S>*Ew{`v1Jlm7FyyVDm2R13$0r1-Nj?`hH8Sg^jVzaU4ks;I zH78j}=Z;}yn6w8Zy6RDeu3#Vkp-Ecx!g}Y)kIPPu85Rh5DOeJ>7k3g`8w~@OTWXQ;bKH#9UItY_uWJY5`R^b({P1}UG177#H zCEE*pQqxR^KZRChl75x0sh;<;Y34oMpDiqX3!?r2Phh3?%-kZP-vJt(61khA-7dQRe?`SaIu&C9$M~DMz{GcV{nGDU)mVf-6F)8fDMIPQ zG=y2}7Z4Aj*D@FNa&6?Tr{}hZmRQO5{tmhra3)J{dWLsVHS-vVj5B`+??)om@op37 zd911lQZp)vZ{yEYY$(dvY)N0u5v2F84y2pjM)Mycen{9Ft|8Q_ z0ufH)!CnUC6y)}wOipH?l#2{+L zK`2{w&%x7{n9`+AK0y7a@ZGA%tZN<PicSZ!b|f{!L6@WPkG5BHBzkl+@XFoQAUUwCcy$xk z7~TefjtVy6@w3P{M%`!M{a<}U_%PQLDQy#i4}zP(g?#1NI`B5YHHpmeo63p*oTiG+ zUVe9dHGUtNExdSBc_{QpQ^Ry`8Mjk0M9+Fd^XU?=KUfg<;e{5_hZ%$l?zbgquWbI{ zjKZ{+UG-203<9CfblWOJ*~wBIk!DDw4by}`d6(0r8_9nVj01L=w5&m!p+O7(Ivr+4t{;2PqUIu+~Z{s7n@9+cLWZfO`Po~*o(eJd8r&2$(}sqRCWf)FVc5*Z1`tsf5v!mmljhk{%*Wv8iX zb$880HDT)cZ4$@chH`>yDDem1HbFir^ClZ_E$Z5+ovj@b8yRG%`9obAPnuUAr^u`M z2*1!v@7X%*jh>|9h&*(7JR@mZ)>JGWr@=t`+~FrwG-J$%jvZ$+S*74V6E=6hbTmgwW-5|iA=B$t81-!J)9w3l3*H(xvN z|F=+S$e#Qgb&z`HdSh6zJiXaod`8UIa@T!zEpf!2j~e5xazwkA8ey@7qhs7^jPz zYN$vA+KucVm?0Q7r*5Drv6hDD@0u81ZUp#bJY(1o_|&~p z-%uE(J9wClsaVOkFUAW`rH-($Dv+3JJF(Nk*)c&Q+mP!Aey zy?*6sNIAR?Zk)sEZr=}@g_ zY(m#`0hE0L>W_h&-*35#437B6r$5$)fD!dFv@{D{w(`kgied?A!OTE(9WbC~6KK^B zo{WyDW)j`G3R|2a$d?Ck^-%r}$V&;k>2Ij1bP5G*ud*k}KAU2BwH6;=u5F!myv-;H zndvCgsD|wYC(a_*A6We4Kc?7}GewPmt*~|41=|aF!>PfI5yQ~J`cn!imJ&5Z*g$1K zu%aRAlCYBR`g9%?%N2txUcnO}TjMvJ<-y7K1W`&x*{fh4#h&B&TuC>(xB^dl@Bsrp z1q5?J{gtgGsf!(OwipuKmZC1c+;n;<{k#-njo*jlM4kdz=>+Gg?LNaHCg+RIHpMKu zkduw?tH0=FjKXdO^1hx(tF4smd#OlTLj#;(bB-5iYY=#`3t2JLgJYasyHgt!QPU@$ zK92C}lDsO$Bgw1>Hr$gRr-Bt6%E*>=zTXKy9_{kr|C{~S>tfd5YP3qX;RWto3 za3Bhv<-)f1un_kfwG@Y%^Y4J%tEdOa^two0U_fOk9`HLrA}y*_lp|w^SNdPlgC=7rSY|u=2{O`XGnKjoP4RN_KJ)@OMC*VcjTo$lzP#3K&)yf~)@q&_?kL zvQLhv5z;x042$5)1!c%%NXM!(@X>i3;X~w*$EL3euYLP0CXz;hglIw9-kMw-ftiFI%x}M!O%6u-H`vLds>8tl}>!-2$*JBtJ{YGY`}* zs|EJUuV~ikeQMoh%RHFYA4y;Vhc`Af7nUs5KXa$2^6BXjP4f0`EVpS&7VCsQ)QX=b zdMwC8;>O)GHf5|n2M#;`yyB=-r?d)=IOA{I&lL=&D}UqaDmxgY9zM7v4*Zm7rJ06x zwDb72X9pRnqg1!G!7Ag4ad5{g+(p#VDt6u;<8kVJ{t#Z7% zA{s-Kb`~=c#rW@FY-AymVRT7vtYFLQzcRQg>TnS@*Bj=1*|RT`HO(kSclAdPzfO9`UKqw8b#%8ES>lTn-jn z`I)CnZrwF)o+?fm(N&}v(jpifB7)6#5 zvPK&~47rkkIUU0^=vHjbD*h=EfT>s8*^g`v_NF(jn-VNYRQlAMrU9-}eqyxNX?WU{>xx~deM zg5YcrJUMElv?xQ^VEwTu)&P>Ir5{%qiH@v0<2|1!`||>tGlbe%Rm&P{prFgy)2?t2 zS;}WAHUxJ{+s!~7l~sYsD%WppPJJi8_T{7;k=3h!nS6w-ip7(O-V`mmCOk7;SW`VY z!?-o}VM#2XbLS0LWnKb`n2~!%#`L3mV;r3{qOh*=d*iH63Mp1bw)%Wm-^q~-!uoL2 zq65I9n5+1*3#IeX!7&3!QdHzHWmU4zJmJ=fk@F=bjO7M>dsH2#LKOVHtqzc|g(=(rw9pw7_cmQ9gwx4m8 zUrBP9On-gbDLfq09ycu?cPpR4|E>~nDBN(qd=ujCdc(<-S;uCm0Zcqeiehzr<@t?8 zTmTk>3Vf|~+j|9%8_c+d#b9u3zU<@14fIT;AT{d`@Wv$56j~R%{B&iQ3?|>!3o#mM z9sfS)&FOC8`;GLaKwmVfup8){Og$9);q+fN@zZiHQbUcHaQEcZ`T15X)B}2ybw7nc zgadk}q7J;C7IA-R(+`Msn~h=_PBv;{Q*q;VC!;-TE`nB!a%_sheT0n~EamXVQyP*g z!A}l1A+VtbM;D|GuMp;BnDqz4D~05&d_>rcy~)4wlja61j}rcD8j|?h*1AGytR}&y zzuL+b+aO(%teBBv+V)c$!%1*~$DCT)%J*mc6aY}p-q-inK@^AA(<9F7YjC4Q7}=99 zcz!R75aE6@6*|SBgXsyy`{al zRCq*wqmSj*X?EgFq=635({gJoJ23%{5-#B8=*J9{x2#}xbQ182AtJq6Lj{og2e3H? zJ^uWKX{elLuly@3>648QEgsLJ9v*!(pvySwT6xN_$2!mx4+?B)s{vn*!}g?DO9B6) zsKX2@B{zM%>su^bvvPf6Xdsgttw!h``m;CX`UeJam;HH!3wD~fw{iiRl#sq`mL znT}_7qWDlqUGN1yF?o(@#SD24EMA$7N>Zy2Nue9_I%9PuMFP4Vw<&|!0l<)tE62t? z+AQAe3}t`IdL6epgW1!75o<#h2s?5337#wKvCr=Q*E|)6+o4dADSV2m{mU8hRr}AC5{8A5 z^u8PZvMnC@Ck|LN`KP41QiuuTAIADa135~Bjdb#1mLHE7co`NlO55&4M}v>YG^k|L z=f;KNwx&FEgT_CO;@Z3d&Jc~9mH*TwO2Ou!YDYKeHQcIk^t1z@5aaMZhx-tyi$wUu zPYLQv+HeWV_<0@}v3(RGog8E5Ds#49g9G5bgM;E!LgU+N}4+$p@CkGPDggq_pqlh-VM>hjdlSV-u-h@82v>G+4uh4kfTe~CJ^wP#u- z7jr0o`XbDRjPLSxz+?18l7Y8Cnvn3zW-6)VT1(eg`6@1t($01|ylG zt-l*0voR1oteBeelJL~iw{Z`2{8~ne*VWnJp^1UKwx%b*_|{EhvE50|s`p(^23K5f ztjB~Bk<5mSfE6u->g!6tRNrkLN8VAE2O$5JY`m?AE0D-9rhpzb*Z?dM4%I4G25~>3 zxH_C<*v#`QW3a|#7CMcROagf^tBAn0WEBKvb1kr|@3hF%U=UHSQ(XRg!u4EhwSb<; zFM6ipn`S$y5xzCDGO>>^75a1UJC)}!yeaxn=^j}D?f|9OfG zqcc7xf!xb`gW0KWNmf-c?kAMWAi$S~8oqljBL=tc9{l63-TVzK?N4~vVxjVOu2@Pu zy0secUzF%wlwLipl`IrpV7J1#sAB{s#m@}buW_y#zeiEQW-RuEKxrtY?yz+*cM)Kw zz*#5rU-bRhwYT3=JRgC`XQyQEXh|32guyYl0@+4i8sN*;>-Oz6r%lUi zHz${<4E^bLmadjK%ns^r*57GHF2~V2#aaNa9Y|m_Z7UX=1NVure9h~89e+GnZViEF znvhw!2}$%CIEP2BmhHs_YwzYMkeDmB_RU3|V=Q^XSDwl>y~%+HtM4D{7)jxOc+M$ygrH2-)&!TjFxVDPde z{;Va%I{=>l3UKw#J4n8@?#F``)!n1TD$_#_`HgpwjYiYi#i8ru`ax`&^5Jy|Dr@%Q{?(_-DWyFED8odc8|@Tjksr#^8cVgbPIX2fV7QW z^|CE9|M&!S4K1Mdt6xwH6Xmz|vZ-i5$w-g!Q6k`mlz`<91aKtj+#BKXa>?3^ki}Fv z(~y+`&{6 zACF%0_cJYgg-6D}HF8T7L!z_;#i?{+zMywy7L`Cmcv;KNGjqJplScT+|GAco8OkKo z%!32YuY$?tg0NINqIo%<*yYaLJFqq`X*w?9a=dfap#AV$|L1U?Dvp^0LzmPY4YbuB@E6RRz8(Qq@HAj%8=l*YDe@zYEz7IdDt6(K%Y2CZ0x3Z+BNW`C6{KDsFF8kt?oUt#?p!P~(_siim7RfE$RrHOqOFx{|;Rf=2#?6|dik`P~lg zXq{P=o5hkJ4-E5XFExXq`!cD^SWR&^xYrq^=Fk;#))Np*fyT$Kah5&1dP;sth1S>M zIfKCO?=JU~uDqf~a-ZjEEG$Wa{9ekmXxt$8&_}_TczuZ6SKOTIT&ZP%in1&h0 zBIp@-+M_m<_3l39t$lKP%kXS%gwot&eDza~+K&f~WvB2^YBU~7Ep?CKLCBy#zv7nU zmY7}bMh@3PA~+kI#Vz6WP)JbU3jCvr)ggbGW+J0JCYomAbBL z)~I!Wr}c&x(?e`NjqH0(g}@eForXg-do>;x1YJX%J5P)CcQnVM4(}~Ib*#qnyS;DP z1CAX#VNU=56wVktdar@fqYh}Vo|R4-`{+E`-c>`l!Dbt(ML6bdQ0+28Vb$PK06Np3 z3bt+W8Va==IFanUD|ufC(wstjibC7+h>g_lc}&N{&kE>@jzaqg(zJc@bk$8kEyhm0 zwD3HicJ~S!Pn+P%CgBuW(=rE7(zLTqsh_#uyQ!bJvMveJj8Nq5Uato&dUS6@C^D-l zoD{jGf@An9__M-TU1uG`p`3K0O_*J+o=_^s^p4>IYATUV(TMQrXHVYOTK5 z5|aT9_RV5~WA%vtF9e~B*u0Ck@)fheUtxX;I)v-P@?1AdHr1x7XkU)Vg~#;WK9#zK zBC4L^=5>&LN<3{OT;Q!!zw$oO<&*&D*l+H^@itRKKjjX@g zIe;fAKOv8@)^T_TYr8|-e6VQpm{~J|VSJtRoPUxtxB}V9IR{8HyebdqNAt8r` z)+I%d<73tIABh-Rb3Yzr_U`$9BhYj^AwEBBcF|}C`{qrnYA7g>+O&VG~a)s_zbIg*k{1*)uJAbF^fk{||)G{TQ&_)+Ems%ck;x!$t*miq?H z-_76AcK$Z$#{+KLh!RemjRYO&E%l2--dbvRCU~lP!xODFrK@Y2vN_E!`fom6ZrL?= z9FXU|%cmy-NB1jiQfUQB?4E2u!%oBHZx9Or z4+Etqp+SCS)KB|yrw-u6R503xvem8xF?5P6Qy_jchrlNV9FtFvN5TLX}Rc;o0tAfKl;as($ zeD9VkTOfBOY*iLHh0)dK({jPrT>*wDBo7y2C~?jGepUzg^VJNKc1&f1x9N5ZjP|1^ zw1oU7DlkLL`tj16M?EWvy*22Q2>2I8>Tl{lo4oeG{Y>qcd%T!jOz71lI@U1RFB^xM z!0g?+bv(!+AIo2gZVk;*91STVi&sB7ghmY=dU@T}1FU{La0)O)yIDM39vQSN-_yL_ zs_+rkY1cb*t!z-yLp1(~|4Msc%U*eA2Ns@)F6AgO}xq0u*);gJg6{1nFn ze|nmJ5U6^E8ndago>0Zz+^RsR+8|l6BI**xnVjb#!O|0SLaVUF(?jre`-PK?`1iW9 zN*+^OYb&lrX+|ViV~=Amkt43GFK+TppNwgT+m}Cr`alux)Fm)8PkmRWI^^xj(-)__ zTi?G1ueCS2E;cUOf6e8wpijsN$qR>J$$=0;{a)m^JKcd~~+-N#+s^B%gqg`Li+Ao3ZL7k*e=-{!Dhw#`|y7iyV&yLIfc#tOAR6oMO(SAhi+uu!yT>zsHvj!V# zEEVf#gcsM`)0LC8=d;_NDk#CC2mUC_D8C3186rd+2?iLx6-y>;*junj~U;4(fBY+C4a$jAF z4Meo?N>W`!pJYb5^q{yYZ-k7XM6W+8dqz~tMBa1g7_0N49vv@7MZ$`?a}YR9t+qC_ z;A(gyh9ZPHQF2x3CWlf2v7O=5lRzM?Vlp-n7j4mWa*HbHA{RE$$l@shK zf;mV94WJgk%+!>aXVZ@y!P@d6T3}Nl>4AR%lY`72cy_{Gyjrc++0V6URgtyBUu;%V zeBtJ@439Uz73j9aI2JEEM8Ox|zGcOs30$yec$&_$-tgfzis+UF;>ql!#z8ONQpg}> z8YeT_%~rjJx;j*@^fZas$ac^iL& zYHNS`c7e9ezM0jt8z+<<>?#KpLO-!M;l9YunlM}FpltB-isa_>21BD<7<O+-4PL@xTPW&jgDYNCj7u2;fds z$}?B}QgLd}ws^Ta@L9%N4sUE)f|jig@%O_`YDxs@yD@Lk>N?3+CpOftOfQ@*f{|}h z%T~%X7Mu6F3##VjJb7k1Eh_D8OuG7QYCN}1FqK#w7?wosX-J;1r`!55G!D8@sc?@t zR|ds!kE_ka5Np$!a;Jowpej^{G%`LvenF3$+%jv9#K{nHZPmum$V0JCrEm-SDC>C) z&YFr=26irMla|wXT**AvDn#{62Al?Dy~p4gs`3z#(kS@Gwe^;`Sa$mB1t&+ywYe)n zNoip5&_f(aBMdZEJ*h}0D0_Vz*p$ElNB8H9vAw9GpW>9K-~ zQ?FjBP4`*dI=d+O3!RULtd}r~Jp72T*Ao{T7tpiDEk`pKvZB_G!L?h}MUZ%LbU z+Mu~LY2u2}06DK&O^nY@uUD~6kQqBdZhOw{MqqAla&2x<;vP*I%Wb3dOhF{m)d%2- zXAKqV;M@0@30@s655jJm?j9!wGn;hgQVA3?r7`%nYoF+Et**H9#tpynK(Z|t?!e-W zNRW<ag=dWVhQ~hxdyjb$;*u58A#uEUM;v+*bu96@*0rQRzlH z1QnE!?p~43U8IqaSGuJ`kX|~MSYU})nxz|dS&)C4^_f!Muk)P5UZ!6S&265sZRRa+5 z|9A-fzO3p^9d5z_oR)~#0_b^8(am)jmxjy+JDB5ab=Ux${IS-)68v#qmb*+oP*e_2 z+7IcDATCE9U1{2%P2k)l6v9L%6K2##$}l;bi0Ycw$@#2(8%d2tZO3s(y+ikL9FG8q z0*_}R1B-kZ@CFx=y9AyiIrRB6E{%n8Pno3A%v!%+2%yidy4zc8`mmaOxBL+!cdt6K z0P6yqxxcTy2Lvzc7Em8K?a#Wvh-W<5f;B8a$hof^fo70rM()I^(u)Advobs>?*iZO zduf6>^8h0ir=Z&TBapO~M$Fb0oz$~-$L30C)JUVK+!~fH*A*Hye`K&xPW~_;%;$sS z#=JYq0^AC7I~RnBL_n~Yce08Ss9S1}>&Q@_$#dHZ_d)_CpZq>0P!k;QOg)3lG2H!E z=Bu{ZqZ8Vw~5H7H`#KVR>t z@EvTtVjXNXa)`p5$Z{c$q37XeS3xJxF~;qgGdhuQ3~u^z-<8!A7Rl7ieB~=6;DZ5N z*{C!@#O1?NOr9VmAad2JjM^^O&PRDqfg8ENTU*t0jhFtln6lLI()P1f_O8-w_sP2oOF2EsO%i?W96jq{tG z2Uon8v>3(kWaD)=8+PleR_R4`&)kl}`9h3|+_roOaCjyUa8LD>t^NC&Y!^c9!91~% z(3)Q>U#t<0X@u#_*!8(eF!g%nj&ZrzL9FWuu-AQ^&wx#K0&NA3{=T5(ok=9Yl%rP* z@G~(|{FvwflU178th*dJ0^l|Q=#x|LUEO6X%=g@x7n^X>kA%o3?dpwR9V&l-EeDYp zzVjnLAp`lP`6<=_H3yS*{e$mN26O`0=E&D55!ji9q8wZES6pfbS7EGrK1uI_SVAeh z57>(|C*z2!_?f63Y9;4TlXWDv@Qjry=x1QPj6jer>mf$gISFVUgvW@|c4JO5#-pO6 z2JC)PJbE0iWnJ+RJv9fcy~crg9R&31D{dP|iJCxiH@CqbXG;-|YbNd}OG{+Uk=j@C zFQ_WFru@Q-I}39QtuiecjaVgFE~TQ)`MN~9@vMDzPOEuyXB?I(Z_?^J zCOWZE9?_T1cqDeh5Bm569etj;czQ|%cjo#X4M*)MZ zwl0gO#LluTGF>!mo81i3whSO!QL=>0MMNAKxDSmr4m4msemfeXZ{@aM%4kegB*r!; z) z9V(>~1jv1|r7Z3BuOn^S5ONp^3(w-1)P30lSvl>5VP!eD*q-Kg4tPyD+B+p)%qg5V zLY#L0(UBx)Gl?97*^Ttc~47|k&Ftr9vaL_%O=Ijc#il*bGi~_(-~j%Wgr4} zmB)f8J;Wbg59}!65fGrjthWrbUGNJQn@)H|)ZCS>on8)oqZ4T$nzkO`0 z>a-+K3+Zm)4z?{uBA>}~g2>QXEJR_t24w4_Mn=HJFW5hr3@Jz9Gg^a@p&elnoNY|A z(eZWeED-71t9#k-RJxV}_6ZY!J1TOom^JnKHnDjrqe${tAcc^3R4fkDXZnl*I9weqWO7R-;3g@uHt(2HqUY!Nb05-+qUIYsg2qaiNux07qmzr=$Fmu4V zLNp5^%ttnvSnfKt^RZYd0!P3qYtNe zXmNVgA8U1`O1_^Z3e*kZ{33Rn58P!S-8q?(MW&Iac|i_|;bx5%hAEt>j9#xMmG5|D zG;shgkY~O0zT4~EzKt{5EzU(6fH#1chMW&x4~TQsE9I&*P@%KC9pHYc4jn$F$x@Ey z2Lc8bE8uOhKc^J5(jrQPYv;1QFdV-!j(xR03Ra?McT;O{X zRIGYpSD=KcqG;(MJ@`@@kDP1kFh~ht%R$UU5-Ht7+0to+!wuuu(-l@p4x3~g@B0E0 zLzrljMCNbX*oH0@(XdR)1R8eBeOYLkFBNYsRbA56=V4u>9SH*6&41(cy>wVq`JK5z zUqA=@rL~jD5h{vX(DT z;G$%dgIhS^VbPUX@mWF|S$=e!e1J-&f5yJ>%z7_v=}S8vqX9Vm%WN5sP%p{|!z!x9 zh5Pa=?9Gu*JnS)z7!Hq%5*&bmIwa z5L=`SdBtGD7F)b(ZnMB0|-j1n> zk+qXF8j-FiQJYjD@PRuqqs(bQqb;2c zj^=xxCS3xGV@s%ME#bcbqEVu^yslU}d3xwbHCs<3=yV^x_k&`*m6zxTd6$`3E5$4w zo8rR)XPGZm{KwCC0b*dLV?4!+YSoPC8wWpwdX*B9V2O!u*wX4JnzkeFW3ILfi_^A=(8g1F_* zSHS6tDRrqDr4!GTVo0leIkbIJlv7f0GJPS!b8_8xhkH7V+T6jtBL!IvB`K}%nzh{- zz_!vp>mw3jv+%HISx4NSypQ4S{2cfl(Md>sv1Q6O5PZj{TN3Kh;!obCMde%r8*j(3 zzh%75_Y=x2#IR1$KCAfpuhB7HYo8`_KUdk2NHyM-d((Zy3Vm-cZyNmylz&rsVRXL@ zW!N)!r@^?6E2{ajqRy{L-$^Q$7rminJd4`rl0|7St5+Lpp9SE zVBY}*tx{UsE0c|I?m7UQr}&98G8ka z0b?hEWLWn7r#};7_nA4{oGg36beJs1s899eyWg6}0YO=(aW|fkqp?g;yR#TTM=6auL4t%b9M^wK~ zn^_he0lEZ^^+J|Zrg?F?sT}N~jcBk<+Vq=!WpasKO5ih+o+-Dm`A1E%d)|(y1@Q)E z_Gc^kM%QO6I^>JST9!MNY*@HZ{C#cH>}5yhxr5xXQ|ftydz-ryC;yqAUxYJf=F$}R zX&tdoo{9Tx6=IzN@%J^ou1BG!dfzVZ*TfxvP-QD!WqV-6Qm(=MEPhYl&p*notiB^H zD1m|^sz{|*`AKc;4Y96_ttseFQpm2uWmY>6bCY0Qxo1GcBcgAP;iMepGsC~ht!U4= zFPq@|I@1j?FJHuV-1;T?+!RT-K1RwbAQVDo!KRR_KaZuP`}>*|%lr>&c%&ci@Hd(( z@cmMFM`aU8$w^}tvb*hD4dQ~|^KYe&yKuS5J7d$y7xeD$Yn}YsCByyk>{W(d_nEuw zPji7W-E#Xw?SSlXMsHeBGO$;K(zfSm#_Q~bLiPFm#mLu4D6?lSMQF3DAb~N@n2%y4 z*r_Ho|4u=(kQQ(&mP;8oS4S5#z&+$(9M^U`R3KmC!_yVk)wTW%t1q%ap)>cN>ScB1 zC2JU4LZbOEst`XrT6kU*T&4ND?~0z{j`SsN?pnYhO>L^1d?93>xa(2)0$bbUp(n>r zV=E=CC4Kz`yF#db>;6BMFwxdrFdrPF{w}j;A12(@{j>}mo-LzwQFk1-vQ@9}scZyr z+)OV24unyH^ihAj6Y%0*XJ*j5)=okypJmgBA_JWIZI6-`7oR0`?5(Fp?eHsTf6S&= z`P!_?bHR73g~7>Htb%!_JI-MtyRy!zU-Q>mS7!Kv^|y3yU80iyyP>vO98}%)<0HdU!8T-yo1tW{g+AmQW@&7rqT(*L~%v( zA!@W6d(wT7`9MwBFT@(5|`T8A+F* ztavk}z?OcoU|mj3D3844Ye07CZdgd3UMR^*#{$0&BlKzfYWuAg^YjhjH(EkahZO>W zi&xHxM!_&0$)9LQDakJxSC>Y2AijW_yu zuk=v-7me(O2W&@Os)U`y4EqKvsf9@*P4+g=4-A|VuUj>=_jNR+5p2RaS2N0QGYSNJ zhgQwLd>`LxmfTk9w-|2|=G@g~8{hS40t3=TQit4SG9XRBHdFt#lf`jmn&foR@)69R1aerIT~2y3~cKp z1fH2zV^bzGznLZ~tPIs79!X8)i$;7gcgV0==d=2z2$5!sN@lZh@bPXZP~QAUB`Qej|#)=S-geg{02ubJ7bn&_`;3ok(?G;~+NBMCkrA75Guv3ZJ5# zXcMPLxp4lTnWi`n6v%Nd2Zf1dugrYYh-P?!ULY6O|_^8fU(9PdqJ}sEJ)^ zJWlR}&9Qk>>Q?$Zk-!rP=foCs(T=^6?@RPoYpq>QyaYbsT@FUZD8*8+i$^1)QJBV7 zx&$f3F7IW*P2~4vKVC1kq~eH0iqp2L!H!j9J0Hehg`Hx*hN^m^xn}=DpNEn=AD5DE zz%oI94dwMMdOkQI@k^0WTU8JF6=?oh4|Lka$vv`4)%D|UGCk>BBx+c3)N4OOzFwEK z+tbxA-*H^)Vrr&w8*K>@Ejb*FQYaIhY1z(NB+)W+KP=o-giz(9EKJ=WnpqIVuDWQl z5CS&)Fp2h++9P=8Bftw}Vb|g&pf&|vUx%l%zy4}@FIzc08Qg5NRnvD5!Scq$U~R1U zW`Y4r_-L|veggAbUZhvn>Ug+v9ohhDeb+DoDF2yL&(aoLGhir_pz+L0JMoZS+Kxmu zOhR&FFtS}z;$JL5PBOCAj%@c~sJhJ>rzHnGl~yNY8aP%dbsjqmVi2vsdePznc$Qm} zNGTl{Z?n~#QnF*Sq2`80B0;`(;~2p8 z$j(U~ZOP@jB0gII+}n?SQID~b6V4Ia0Jz+W>a6;=6%~zqCpc~!m=e8Fa?I*VU`>P6 z)MXq}B@9k~dYv~cUMXB((02CzU{@>rr{ORltvJYHLqUa(INUw|VOX|n zr<KZE@9!>7FJ06<$Sr)F_-IXzd8Yt?&%x?VF@9`ZA{lv>6|(ns#CztC(| z*PEK|*layae{ed@J8w?;_qAA;uA$f&@aV^eWXB3R?{z;bJup(uc0oFH<|)cnlK>OU z3Y^GM1Mt(%CV;cHK<%OVozC|a3klq~{`;oMhe!jwuH5#*y<#4C5`YaBms4Sk)UI!n z=i{Eo(_2|U^u*&~%7)DIrFvC{Tlp+LT!bG*A;^hie!gEck`BKfA{-480p8VARP$kL zHP>cc%S6JAV@|6^#GP_7UR}R5;(}lOXm1+dnM5_|1Yn{oAyPNLuGwis=4aw5by*F%< zf9Ne^`7u^$yQ6j4XtxIWh^-uMee7zT1zhdL0Wza0aa$HBS|0&T|)N5pTh&J5h1s*KPVV-Qr2*p06=aIp zCXUkkh}F-z0~?<<#Xhn2{-FF9+9J%3z~GmH9sX=e{1kE?nyXT6C=P!>na{+b)ol zmbIjD)y)q#VmEvRcy~ttmUsluT*dk&fiiVt?cr#e9OEKUq`aW~52(8;g9*1?n`7|c z*K{;3ci_O=Li>(1GBBMnk(;!@EB?wt84j5Q^*qglz8|WC5Yl5As|r`~Ikn*EsAy&+ z;NV!*E$Qy?0^_0~vwg~GvYH>2l>~_us6+UN$pmt{rxh{b=B{q0*7jaE&Afhm*>s3> zJ4jGCjI}42Z7_Y;I_!JAfzRVE5%`2$Wzpz~STHMA$>|P76=j9TUrdlK<-~H7`HNq% zqDD$|vlx_{V;g}-cG3;9$~1oPNR^KXAz-YC-r9*@X^BZgIm*D?(YM1vZgH#qHpbih zK_=8W!p4g0l8{j#>455+&WU!DjdRP&mP$~y5Yfk;3?}j8S&E$@a3-PpbdD;syT)^# zNSj^Dt7p8}at~~?@irg=HTd^6Rfj*UaRnMr*73D)3ktHTX5C`pEIz znQ!~(fyqyPwgoYc{Zbp*50s!}N-XAP&!AODRjtI-P014Nbt&Wagdf== zL;K&bK;CNQe$>az>+cm2>mNSce@&@ijYnTT74$!Z5%$;T^x z*i;j>(H3kX0Tzzq>Em`iEg)u)@EQemes(>LhX)8@i{N2RQ%?yQ44aE6fE zpPlocyrj+9zhFAG0A`fQ9EkUqHkaIApm$^AcPi{uzdb%G}bB*}g$^~v#xa_V1Y z$TRvgx+lI^q<&XL!pI}mOr^P@B=WR@MU7jJGr4wwSSz-3rOKHF!u~LC+S1p_c7N9N z+*~|40G*aJy_zoCz*?Q z{1k9%1nwRU9V%)eB3ff6<(8KXCMQ;U>sXZyONjdz=8UAIS;G=JakA-)yZhdNB^GA8 z7L-tCp|+fDyKWf~r)0%F&u&{@%?~p>G|;U6QWc9DwOmz%**q=dFEXPf+(1@!T0)<) zR1wj~y*!B+(NR*i?ULsrN>1ZU5?xR<*X;1CSX5km0g?8geI-MwGd(_8cwF(Zc1Bq# zW3ynv;!Y6aXy@FSYx^+GWIqx3aWuHOPcd$v*t@awwmRmhTvV%Ni4L1!H(Ic}c6}mV z6$o!Md?2Io(uHpu_-i8tTGdewg}kjlFap~Rnt(Y!c*eFe{_$3C?sfWga_f4>StXO`vGh{&y0o&CRPY#RDn-0~Yo|9C)y|X*`l|FHLAaeU8D!Vg@BKJvXKjan zq#05&cH>dc6Fc*oqf{=Tu_<=`S3KVaKBj+F*;Jom6V_u^v?Iuyo`X-_{1YbMI6Is* z!Err)Em49i0qdCK{<8=MA4m&N(=2;&$S+7CmL2TrTbz1MeP*Foe70S0q*8lAS`Kj7 zAr)hoGT@$g!?is=-HPh$hA+Y89I|wdI)6xqqsGW;7cjKbRyUbZtYuR-kkR?EA@TS2 zD{>PVTeX3cRC~yJzihm}IZ8htB{5E~5>9N*I(c%l z>^N7O-l0Sj$(Nn9+$_!+$Dd&r{QY}y_sueuWv(h(`_bu#BM&@P4+^!|Tt(TdIG?kQ zj)63?CwPvEB73nEi4X^N@%-mwlSK_!wqXX9@Hh5s!A`xspseT<79YVUT$<=d6rQY9 zQ%sqWtn8ClUb)X zHDa=#s(k`2M=yhs8~_zd(B%!fbjGc+t?@=cXx5G4ZnoBhGEm-!9SzbQ9TTu_dwl>_U zQZyQn7`;segP}L7G_^iIA72il(QC8P%~s)-zziVWKWTQp7EgD}(*(GTWww?)y2aft z^&drx8e2#tYKd)4C+DEsqRF$4#G|F1VJL~QAbWU(E~n~ur{h?N zZC&O*JXEK+Bucd|WEPAV=hW6@-T=gS`*?DYn8h+4#pVNmM_zT~9@OK$Z}3#JHhi1C z6^+~h`OXDApN%NE@U7;81A-msbJ6=;;FhnMOA@}$qMNCHE+xgfuepap^N6H=SO-Qb zd|}XNWje`Vh<(ar=dO4hfkS1!01?Hn^ue|EfwR7IZ8j@1w&(G zVCn3-UOIsyti?4^e*5*^-u5~o)3G|+ou0i@5NLx+xt>T`m6z^lT5+oTxv!4lii=ZC z&QH29hnr>7Sg$mCsJ-CTFD@sQPZHZ@QyjIZ)iHOz*Gc2i;a|5^yi+;0TcXagUvt#2 zgh)>7Y%20_Y>p+{%5xSA7314kn^?+w8ngW-avQM--@JuP8Q%t5&I&R*50klg-k7Q$L{O2RFmQ9TM8c$Hv#x# zw6U`XBVQ^k<`$t|;O;=i-o0d zzv2nbyv?ympaL zT?51n?(CNF#lSF}Mr}-VgZRw*4!d&EWNym}$HeLHn!$^jNeX2m!}&X)WgDGDnNaQsdzBgvWMBg=b&z@v31hNR zvn|iXQ@Jn$ZFc{j+U@#I#<1cd8CSo3-Hh;f{ou_$g{YU@B}7$#?Txq&k5bjtC;2$# ztOmfL>7vnRMK9N+d(!s4cMBEgD$|r6!<}=8A8RHyn}U+0Emq8ZDd=Obx7lbA#3mN9 z2g&Fj1)>cI;uz8*H4IscVMHT zNc9^%N(BYe_Vu-u6%|S$NN-xJjA`f^6P0?=s@DNgyLMIM$Mpux6{LTt-stFfryP(S zYU%9qvQ^0$J+`vOtkqot-*s^PhN0#CY~TEsdyb221srcY(^Udr?&&SgjlfxR?sP^J z`7qsXp*^}Ih4$npt1J+UzE#??JU8%DByOnlyvS9Q`1#%DPxsDyep*n&w~ z8WIm_OsYA1z7)DYopR74dw$mhBjF@q&F!XUrfhQD5S3qGZt^Q?VXCdFZ`Wp99D5w@ z>nmrHm;^cI!jmm)u-d|GnFQDiT|i{V;P`AnqK$hZ?)vQAcvG$(jSm~W%ADVW^X+>7 zAXd_PMXyEuF)KLOu9qO+$jPH2N1)GmWCl`h>sY_ErDnNvTZI<>Qf8ED)@xMD*xdh~ zTxksD2Rn!60e+9U?|~yF%+vyCni6J$`X}t5zxx$h8l%6Q(ub?0{9eYb@gJ2sSev?UvT5lm5Iye% z8LuVvMdDvdJ0(rOEsX(v4=e^Prf&{uYT5x~P-TBp8Dzr_I zz?r9>=EU1d?;>0a@-d?r+sKRb9g^wOdnuAF0X;-hs^+c#r z`8-xCaNNG5BFt;f_Bd`Vj`zG-|0}09tTT-;es65Lt)~R~;DYf3Tb@n;kva9t*~^Qe ztwq-&quloA#4I@u(vjuiIqele=E-xT^L%McrNK%e`QsxK1 z3}FYP5RZ!vkD8CSQW}3fnosa3aNT*pwr#)!7*VFr#acuZqh1V`Z#s+CEYD2!G#ck- zev$q4^Rp{d%3XW=f#}gFLj?P-#^X(Avo&fr1Mck)LMIPs7Et0R3SP4*UIf0nY7alN z|8PIitB2XEv2+>n5UDp0rst%n2BrJH$gVgo$6M}b#m95sz!pzjG zC+i)D|MLMv&{&P!=y6w1-0Y`v`bBiAKN!CKdN3Gewu_vgLVWI&j-01!JNG+Gjr=pY zXdtEMFL1eYo;^RYa3~KPcGdL=)4lkvYpyrMgxE7$q}3k~8WQVBYG2eBfFCR|wyNoz zYys4baQQbaS$l*84S?I#&@D(FK&URQIiv*}Ph-M#{xkY;=W9FjWeKyd`QO)GC3u&{ zx*5oyQDdQU$LVPjzi5;I^<;(C=Ndav-Q}pSxvW=>uSP5%ZZJQDDub_=m#}=TV9i)! zybzavzJA9j{?5ya;@+jI75bgI5`Z_us$XwtN-Jju;?A&C(Kt8R&l~&0cHU|;*Hh@t z6I$l&^}ny>2xR`?^!oz1y5rj1rZX-BIC_q)*7>8CC0p2Iej@RzqCPs5{CLV=rwPHN z|DJEwGuBSJdmV%b?Fxl`-))}AQ%%c$mdeSS$+Bo6}TKMnI&a_w*p1ltE`Y}u9&NXhvBJbi^iFSx{cPM zDq7$VAH(MgK9+hiME0Xt2VZ?O|0*CTamW07LUD6YalA=#X zXVWDY%?}0)D)-N%UC7Dn9uvjIMP{y(7F;-Uwk~mAN}HJ2ZEP3vrg5ybTx={SMQj=x z*+r^HGf8D{ll*kFTkIYizi5k-o6{xdQY}HTo(Xj@fjZC@r3ta{50AmJPbC{~`@D^1 zZIuQ%O!t3^bA9lXtP+UWOKp7<-d%LxCfRf#4xz@3VtAAAB*xzKMv&uH;m#X<5jlne+&^WmmHBJv~?xAS%)?~?0i_2&cf3CR_`*a(J#HC z^&LzUKrBDBIHZj>jf*Vp*!5u|ACK$-qFN>CoLJj1;j8b{sj-{oc_qT;VG8yY`Lt*6 zqHOIQq?Q9$H`jw(QuKWhvG`o0MHgl5;torDTrRT$*iv7C3tUEXNRL#l* zTsc#%?V2cDo7dllN>12w%Ua=!h*BmSSM}uS{jP0bHah5~`lj*Bv?8BDs_8kd$R{=1 zL19BA1Js1Vm-N=)5~MM&|oex2DEcy!YA<=09%6a)oEU?xP+PzIBV5w-|2vr(`dY z?KegT(tS|yqYe8jPM&hq3JzIp&4ZI7IOoU1nMFYA((J z?O82Y^r{JVTkm4*5CX{B>iN3Bu^Tugz{}j{BbFkDd`(7HfKnMX9zz<21H>g2hf3yf zs=F>Aahm*ph#>*R`Ucr&bBYNXj2}ASGRJ)I#mx*#Wp3z;vs~M+i5O{*S(5YSc&;ni zgc;@CRtX#nz+dT(7ns%EUukZ%E0z~((E#g`7CwnO!;LUiZ5WsY--X065NQol9l&u($N zCG3A+lR=jC%#Focn=r0BM}M@X`GJxesXv<0s16`y+Yn8+%D&HqQtQc}b%-3`hL50PY3j zCOiwEP*1!8wIW6oq}*fvE1)7_gj{cY3$DK7IbSr&B0|?+93oH*>P^&a8Z32@sOsP7 z7A&Fvca(I9t)8oAXwT$XJ-O#w@deKxK}#7l(f>2LC~%eieMAvL>{x8!Gg;|k_-(-c zmT)Qk|BMn?5+B3GX&2whf7PViw2pGvFUt_F?yB>g9HO0BGLp3aXI8s&T4HQ`F>-`? zRrXaXHMxbhc&z&W66K98E;FLik$H6BIS0$V(j%tgP}DjzD#ucn1CqF2Eo1)niLjNW zZaxM&LozwoL`~<3^ZruuXEFsx8O^R&37q(M#Ef-J_{eQjvE11><62QZ264~5$#JuL zSlWJL4}5Amdil+G6}@xxTS@2r#h+Zxytv2$)05uuM#g3*$tKU0rFJdbP)9chey4PN z%J5lCOi|?0I36SefhPmL2H>}Z-BsRs`QmpC(@+&xJO1CzF4Xe$jM29_h2U%4iq+Ei ziRBKC?b zJ3FU{UW#pvnEiD00ubuY?n8LTsvyH^ehFNT zt2w!hZQ>Jit)F?cR_;B*Ck(C_-LL_(X*5_@o4`Qx}p!!0V!+uo{vcHkknsqz$NIgV&~}(#YleP6O;SPg&(tT;p;j* zI!^DF?sp}cFZ?h_;&~lkmk#TBK!beGh6nXTuPesEKPU~~8K)~sD6TU%D-yJ`zxn-P z%|Gn&@P|8#kkyH2%JZ)=jfL>}iYujuHU9 zoeD+(%*j;6Wm3MTiDy}*UG_G)$o8gT11mRyfKHGojaYg-jU{L2-D6?h=?Mp7r(IgA zqX)o2lj&hMlOM=Ps*ntcf*?HDc(FUOFUXaUe)YYhmQS|l8nJ7z7wr$c4KFzF5C^nn z?G@s5R)~QM%h|%2oo%`!1|C3pUymL|Ca25KTB_EzZ6^Q554QrBBbSei>%qmk!YJcEfC#W73SOvmEZCB}95=MnUC zjz(lnnR|%X5q&0iwUti!S{Y`hXK_bX(j~q2{*P3|a@mU&e!$s~$##t>on<#YGz`irccUe!ggB-&I|o z6;MONe4i#ysgi*H)xS%Kb8XOSqog;XVD_vUqcki+9ZG$&-PdS81~$q7CGZ&J+bJ zQ0M~_aMFRy{s2k05Fh<;lM6m7$V=Z}?;^19&s_P331Vq+)ts7d-|35gdbdgtkOCtA z@6kq?YOAJFfihu8Du+NG%LEmf#>@Y2;P`yyrX|x`7kfuBaq{=h^xv1ozxB8T#5BHL z(HJOaN;8%hsgHJWqDmGU269dH&5kREr$yN%IdM#Dy4p)lBY?xTTdLBQBIsAP(`#(gJ6)%)x;0q)F+y2iovuD z_{7L}KQz9Tn@@)Wl5>{A@U76H*Z%C>7Bcq0mk(mQDP47T-x{B=W#EM3`@r>(vhclh z_snkqA|ov&Xq);S{3~MB+X3c>LUXu8*7BSHgg8JZOL5|Lfgi*W0<*2=YZ(Unt!xOi z{^U#rq#sMGt*pRndz));B;=&z%BH&HcmX`m^x{zi`Nc23HQ(YXs76HxPd>Wd_2ZJf zQbHg6oKA{Dwf*dz9hg%IU`}-y9?GrAZOZ}fV(BOhs+1Krfz7(MpU@D@Dru~ zZUKJO!RaFwB#BFXTAKRwyGlg}L-vT}n$|z6q((%!x(QQ8o2rWky@UWa0^Khd41Rav z!x6_LJak4cam50iscb$xbXsF&6Fqq?^}^qynegCN_TFFl+Ru6GY2mF*&Dc%A#M$?# zo=)Nyh!*Tt1kzGDA9Z$BTX0GUuI4+9w?t^3%K~SQWT@Wd49)TRrq>XZ8v9dl|2$@Q zDDEm`8dIr@bcy=-7GS8rRhpc0Jy&!WP-X3jMj@OC!efi5kki~wYC(v&)3ue38_9Rc zjbHxn(YvB-5F}u#MxaKhnd1HkMby6ovd6=tQ3Q{NDarl-f}5d=vP!ni>b9;D1MHt* zcZG5Y6JCyn(rR2k9%UCAB}{oacAM7phRS0W_$YU;*Y))HpDs4bt0TPwx$$YTd&e7E z2}a%Qi162Gb^q}~hS?I4e}&yeSaeSHTwUZgKP!0O`@*jG5#T*;c!gn;8LX>u735(Q z+g!jp2v~kwp}!dA8%prbb=rg%*z~rR#~d2SdxhOHil=#5Qrv58-&;`M(@L%})c9(0 zxqTRC_y!|jzxkq}Wxzji#%PVCysw#1?*gNe(6U_%AHuCF2ZDI$(6z>Bs#t9c?e|2% z=`k8G(7?lM!UM@D9(uXbZY-3S|3@$iERlDUNo4;BEH%yYl1$>yay1YK zB!8~b{L#;QBgL4&f}EQ6Y`>xc2;STa9yEPV-}YTR=q~|wel2U@)1V8#{_%A9-p#Vt zdBWx&l<@qKug$P-4)A0?iQAHjXu{CjzKh5Y{@TVWGCFWJU| z#(fF6>QVtwRjz*IJ2%?Y$o(*t&z<$Ai z<){%dg7;^fn>0LJBS;QYvpbxU@M8L!cJA?NS2e;wUpVm-tWze@QAzcUJG|1K>xMwC zPF(Yd2A6PRfPp@d$YzW^7a_co)m_&t*^Q9pO=&4D>aYH$BgMCR88MfZ?vl z%Vsp}YH>CAL;%tUSL2DBwlgA8E>8_#1699bCB-ZCq!IP4w3HO3tet9JoRheTE9i-^ zMdNaBV|8eYR^EhZ;&TQ0qzr}tuf00>>Z4hm98}LrD)9gbSZ58%@Q_br*-E{1?PTzf zAAr)!D|r7~jVd|(pxUZ6;Rjk-s|C^wAB`KP&MOKVu_GqG)Qf-kx*J(vKwzRf1BPiD zEY0I_Xh7WF%q!Ir;Sb1|J~&>n#%`&Wz#rW|ohsZNNVKid#b&hs=L}D9(}@sPRI{~A z?Z|tP<^hTZi_OA^T4Fz}{!j5l%!Lbkvd`5{4Vp^@b;*_Z_qA03^q0{zj`V2ced7{1 zdHc-6=Oz%5DM1&2(WCa6cRJGXLf%&gqT=RS)2^f}gbO%FlXA+4I*slP+ZF|XNx=tH zkvY7m8jPy01AXbVxIv|J^}7|TZ>X|>L^^p+k!WoQeeplCWkzG`w#xI5#3*lFfB9X3 zp6JG>$M2cRZ+?1xQ(L3*licAU{U0~}d`u-9xauUr&Ltxw#LV>1fQsq0oRIb_+1G&Y z5kngVIsZV)m>OLR7>5!`)u$%sP0y9d{l2oLWTS3gL;zDE$6J$^;=b2%Z}@WIbK6{3 z^4R__f0L+J`~5lKgYzL^r3{0eF;@-<>wGov^f|1yl%HyRTwxqlA+XyWsr7e*{5bYBE`caqXyo1Xj9E8$VmOCZ@;?Y$lS( zOl>Ch$ZX$FCjUoZY7l!1XK88L}WiV8w`e+I7P zXGir8wBi!}K^R*vO=%c2s4gGA6j~+w1(&#)!y??B%c6qDCt?h}9U;n=_yjdCIvK`H zt0`^nY*ry9ZU-5SBjb4d))MT~i6$YN+Cw!Y7Qu2z>J|jl7&4rTFb^agDr`k%_+%dpBpB`-02pwV5pd zT_f0A7kk@tkM!2R3fQdMtK@8Yyh>I{M0h#oyOyJZuj0==ymxqR(PlXn-{6c_HoW=} zWYSegZ}(>#WHn~FaFfQV7kYf zPHjC~b=Ze`^ScWlf-}ytMw9Ts1X7(s6H1?me|OB`55*m*5!0~OCi@l(BWIgOsM#+P8io+^q@996Ofz*p+S>A!-(ezFFS zKB6Q?u=0M=(}^3LidS=r{Yami=^9ej8G4g-LZ~qER7Z!GxsPiD=IHfNq9S#@NayJk7<+%w?%Q-2z6v7SQH@n z3Pw}>WtlY)T^mSXb#8V6U=E_S8Arca#Ca%J9iVCw{m4XT(Y7Za>1Y z+;YExNh=~DeX@|OspWzH~rXANuHd2p_5)LH{@Sgf^;8}mv?=Y zY$v9oz@GmK-cRU}8N8oJ{$~)#DX!?v*SmNg0a;3j7FYQ1fi^jAdlB<$GIh;4(?(eh|x_Iw>qrNiY`hgHL6al`xUeUrfW*F1F?Tg~t?A6yGFD)YG4>;~Fp z$C%%Z!uxn=W-QxrEky9tq5pfMARR8>e>W<7Oyh=cs4fYkn4pVqj7w2BgMz8;|MB(~ zU{QT*`!EKG2q=xifYLBDQX>jdL&wnF-65$MAkEN2BMn0f3|%56rF4g+w19|6`Ry5h z=RN0t&UvrvJKyy`-?iC`wby!{wbqW=d&TqIH3JK$GnOY(9t$(N6wvv@i%{}PGzXUSL0zI51N29g^?fFDZc8u~Mxe*w^ zB9@0>63SgF&$kx}A*~O-glDaRi2pSpAXdv-BfzhQC98ayBgS>Zf6AAQz_*o>1W^xl z2Mf?=6XDxRr10R!GqTq%;`>PE?Se&?dWd;=+!Iae0H98Oj-9&Q%jy5(;MHiR%4YjF z&Tv6X?EJgB;3Ix3cvHhtn&;CZrGQd|Hh-NT`%(V}hfz&mzjyGi5TRCMO|!xASHOtW z(y1zrt-M*v? z(F*K%=i_yYUBT~|D8)p?QYO;|=8;l5kFMs%=Q`yJnYd2n*nsRSbv?Y0NJ;xlF)z&iT%Omx^ zP(QkMC7~RN`G3t#XV~MSriBNiN^-dOZu*M1@t>+FCpx39FKV@334G7%_(%HcRhOK} zGY3C$OofbV^X4JTdo4EDh9ypnLBjZpu@wRAjMeXXeem#dDr!DRSWU)bYP#;%Ut03Z z+xSn8sDnL)bc&WSrf{y>8i#!)x|^@_ns$vY5dypPlIF;yK4qHTk+>MZ#|A0ZkmB_c zuHcweF-iWiuN^RMoOIEIJo?1i!z*OssfGmv!+?VQ&&JT$Dx`J#_v8OFMqlRO1$*%r z02%)`Tg^>ijC=Rki3T4E5M(7RTKCRG331F6Ew;4puZJl}POTm@yW}o%j zpo10TyAB3vc9-lImBIR2jR9sT7Tgl1ZG;;9nwpZ}_4m0EVo7E9n`*9is421g*HmK? zKaBV86$MWz|B;5UogviW6k}vW61;4)#IBtFnf(Y){b(seDz-Xs(*Kenb6kF;H+fwyxcAO9C>VU@2edj z&XlV6X&zg)UM;PY4r?v8iBxH3UiR#; zoCL~uf#tt(U@2aU@@+~HXH|_&>?vNSe8T6A5ijN*d?0>Zf^>w<+e9vTKlrCue2MycTM7Pg`9T_Q-%$@ZG!rQTWDZuMu z#zjYTZ?4wsT(MGe5p86YFsP#mYI>ovm&ae1qqH4n|Gdm{e|bN=o3=XTxhjDe1?1AO zVsxZTkBux3;QhQzxaT4Koty2PhWN$J4f*qS4!MSBVel-zOll-pC2Jzlk$ z-Xl4djU@@ypq$E8d^irA|Bum$MHwzaql*9PN9J?2W!GHWZ~-j@y{QV)(1L|d!uxkU zZ;q13lw9tFpWfP=Rsyiw9%x+>`2NO$;MVPsHzKL#Ya32*Z;{69kMy%E&%$Ey zXT^M+UBMq-xh22;poG4I`fY}snWL$%GK2k#&f4RUt9(whA%m<$!+lGuziQb=>HHO0 zoP4@iCdLOv3WDqpp8ser{Gr=9CQ+)2%CiN7)zOT?)Nu=z`Y)y-&`aq6{_r_-`R17~ z$R)CMmirE_iVEaUH92+u3|@iNUd7NFR{~q4Uv~^J68SrSdsI>R5JUF(yIfQ}xPx4#Kbj zC8dMT&M!TkUWA6Hs?`^HLy<` z#|ly#jXRBCsFR<}<>g1)4X9^gH4yZ@kjtir(8pg`&HJgy`xKaUO6utMUK-0nF9?~|Q{(|T<^H$0rm25xp79E=I;Xq$Auv zd4)!oeBk>6Sl^SzV>4xv&&N=YQ2d=k|qgtD2&&3j>7B3pUP9>MQ-Jc}AzGn&HQ$K`H8p;WFzUHr* z(D$e^C!8cDa=VY-c45V&2K2lsFwRyEQ^ck!ElW0y)sx0uV72=qwDqEIhPPqm3JoMZ zI|B{n2)vct!nG3P2C*U~y?R0o)nN;-gUd?a0nrfvffCRH^*68!7r`^0!A5@nSn=O;3$HosSHZXZ&vS zbkU!RIe0Eq1Gwyl^qc&hQ)}OpPJAI;2Pi`~ue+>Gi3cqGyFvqS16dn(Gje78579vi zZApA?r9JqiQgSi8C`Vg46@KBnANVmo4SqT5qhw;YgX>04bS7D~8XYX{vW=PILt|=f z$6nk%U~ySDgr5k3pOY~a8qjn-`IJ?oTNi3V+@TuGZR-#lD;mla7Oy2-7Gb)3G?Y}5 zLr!8HRptP&VlDYMrWUb=?l z4gSU<3e&HOd=Oy=ev>-2dfXZJBzR8LS4ys#gvqhTI>WP=_1;>rtgf=*+T^qpr(Q5# zuAw${u{%^>Qir+i)}#=dLkGu?8^Ismrq1rStWiV`dhyo;Dx#M+gUXdT5|<>R-tY-+ z?6PGG%0Eb`{LV=3nUs{wpV*oGuYF+rgkbnH{SOfn0Y=pbE+L-x0(P#t$CE@!El z5WD{W zJce{lq7tk0xej6@v(}#^R3b5~@&wDj5Bs)p>U;*V}c+K77rc$gVsrDV%^XD743GMLP|v zF%owqKVF4`Vl_eny{}+XW04XNqd4PAF>E`e>`(|VN2qxQhW`z(b23li8ACiVwt6+? zZU$`R$tOQg=Z#2xW2}a#FRFNF%DoGOX*rRdi(JC749h)i(JfqC2EJCNYrq;;EI^OG zG2)CC1DO#X==O2wuku+!3|0sR4Lbl-QUG(89>AwHcX_d6x$OnLRZ{y6@SzPC7m*ry z#{Tj?{Cu(c+$Qwi?T*agQ{>Yr#@SjcfUi{SnDsZ#gF&j<*tY25hepf_wV6ua9)751 z-03x0{_*(fWg_?FMPrSCoGc$XG&Bpb&=tbtUWw})nmr(MoukaiJfv!rc0h*MoK4rC zM;RL-KZh5T=}1d|Cy`1H$2W2`wEwDjCk30upl81v<OQNM zgq5N9sPyko-;sEUtuV-`B>eGF74bSl2p{<)Ki9cVwRMKqxAWh+uO()4B!`1L)g(8~ zw z&v%{^iLK$bh9sHCO{rvsn{=GoFHZ&H>s#6Un9h>=_G%h8rPOV3)sdTbclo>9ywtX0 zp(8T85n%4}KEGQ9vjNec?i{6X8D`W7Cco|QU8U2Hd&|t_NKeVk$0>%Kl_`m@W96aH zz4LXF)jkDXsp>;K{=Uv*GGeA}a&or`R(2RG-yH{ml$uk3 zY%=AIyLAul64lK51~LVdk_!>ws(;ma?Bo4hvQyGg%~Z_|Hs$Lahj(ltc{5%#&inxK z$nM9(KC;je?aBNw8%IUW0I>n&C6C6W>ZbYb%h~^)CyokjE{pNfC4ylu`dcC|6JPg- z)bnm9I`i7=3OmN}^T{cl;`5g#%BOy}3Qumt*hGJ!CeiaG-5YZrlh-C=N~h`l%G6;c zpwuF*++dNnAb>D*xs{Z@MCo_HO4i92oa1O6QOC=kl+`dvLHLFbv>2qt_n=^{!)qy2 zpn`R2kfXPZ{k$WQEmDYGnNnb?w4cmqj8ub(uK@`#^E!H=l7^>42=9%+R$?ctbYnIF zl8~uc0zEu>sKg9d-7Kzp_ANLU-g-}Rym>TqVMns8R9iKbd?W4ruvXIp6f1 zoGs7nlKclQKOJbx{yFjCyvH-!T&%ymnjg?weTXnz^ehu!Xw^ zCXdqu$5n5bm9K4bUrH!42AEv@BXeO}m>zSsRl zC$nUEe{}E)?(r6qA2=;`eI3z?{Di+gTW-w6Oxc|YtO1Zb_6|6GIXT7RHDD2vO0Rui?>AHr>Ti?Kt#Kk3%Wq!V-a za+GmAf-$Ka+~R8QSd2@otYz$i)t`%gO)o{%bFV*%*s$IZb|_|8@4nW~LD4$~r7u2g za9`5hoi#|a5B^EnSN)~2Q5;xtAi%e`m8Nti9-DG-5umuU# zj9-c35;T1Ir;V)F)2A-FX|P#f+3=(!Zg7JutXH$xFhv=W)L{ zt`=%-Cy&*9bUv%DJD*1`x>oO&Yp}uja~u^z*-OJs0@AQ2Cg z@2~cia?#PpI!bc3q(~;4uT)a~#+hIy*X{n1ko|5lku4WK5J?)UQd-*mt8rwxQsCgt zXU^VCHIjSC9@}YL-G-0bk0f13M2dx?Y$zexpYOgX|(x7`ihlYRfH5=}kWYC)fSXj89*Zd)RkXEV*Jo^ciOu*nCdR zKMqO>fatNjrY`z;9M;he_o>pYy4$R5iE!=M2$-rlv>r94HsiEsEY%Gm6H^Jhr8H>N zF1*s)U+_cm^tKYuipgw+hGtkp)WLvv@Xv@ul!to*?p%-0x@d>_Rt1;A+GfIc+Lh#Q zxS6eSRic|JX3eB$zJ($j#r)9Vg8|{-NyWtU5k+pTGsNtJS>`{x6b|Liga1fB^TevO$K_08Y_^d@CN$`h+ z4c`JZYJXzQ{j}kqHKP7#z}3|q>JVQ0h>y2wG<*i*YloH%gDV zqjNn!IP;{Xb1|BgJ8hh2<1ud=;-&LHd&QW1&4{Sjtz{;#C_RomR;RQ^p!irQSS zf8DZD9SdYN>|W4*$e(*88WQ5Vzeh*g(NXW?eVce`=}(fX!pK0nW7%fCg)E`%5(3PZ z8!X`39VX?TkB=56$*xBbb=o>JO1Y7c5-Q*OR~w;w}#s9^t_UtTo zM##ZAEw!roQ9dxu%y1UQ(z$vDLFz`PKs=~%L9Uu{^;?=q|5m}k?sq;D zw?XYIx{n-gAWRQR=>!8txf&ZZHaim_dD)$bB5^&Wy32}Sd5bVkH|8j&N=#B$@Qk%) ze<#`}@vnhSo9!Yny7@EEfnDSzsqbtX1$Ln|`(?vy54h-hrbw;dT}e`;NkF@`o^BHeBb9Wh*C6+ z%aO}2QTAgQ^yw=EEh-!u8jk=JxMR2(l31`2d<%AHKU}}Q)#$#qn97)PEq#N7Q$)qu zp@iGA5QB_8X_N)ici*S4`}1yJO&8{l0SZ#Q%LJ;SEeXf!lL%1)PCGK?!q|4nC5~Ap ze~=HxHW@EgO5+gQ_7nvr_f;Y8I;@Q3yg5z{s-_#8g+HPn8BZXOMh+IEvqXN-JzNR1>AaV6Cr1aTj%v zOF=Un9>T?4^&3axrL%>quK`3XnRV4&VR2O%Qs_R=ZAG`wol2}hzRZ4|s!xclch z^i8GSO^f6RaHsA%Lk}M%G_-O+Cd!=CgXB-5nMcldb|pmlrniWJ1WI4<7A)I#j0+T$ ztA@6xL`V+*l_cqnuvVzp|2XDE;fA)?Mkovy=?ikmfSo{yTsYcZD7=0~kVBH0T%_!v zm6jY1F|uzj(-kbit1+D@1&zP<2rBMBs`P5JzzlAe)m(<0^8GWmh)w|_h|a8q!cBmF zCxMO`@5kdLW-4ac@?y|d*2IQijq0sI8Lvi^Lvf%a&=G8ymmv5tE@h<2CALDc4S`kV z-Py_Nm?JliM5l~UYusA{ zur#cGcIR4%C=`rP!M003Z9M<+ro7E|ad0Dw?cwqeHgexvE!*Hj2T2n4Og=S_*Lb-@ zK^dNElUjH+!X*mXf79U7@Fki#t*j^^a_*x5hU&0rQJGbpZKoqUMLw6q{+j*BKIU{ls;s z8?7_V4nWQ#Cv~2^02nx5WPh#w2P5f^QL}NzMTcoLhZb9zevyA{Zbbjs-1J5qwq82^ zzqb?7iO;m|jc9n8EA~8WT_EM?2)@4l8^_N6Og%*1W#X~&%XPO6p52e@T{^V+#UTCc z0tm?bUjxF8PxRUVV=%L=Lw9iV{~E{)vB8`bO3OO;gQM;Ye*lR{$>Q5C5rOG*-w3$= zElJ8uhum1vu_(-ktV=DkI^dD3Y!v=MI~}K_&lgqj&6U@W4BuR&lPonig)_2N@u-E! z>k1$P|F__ER>?<~roB&@%d-^=Kazg>aX*-=y=WbubO!MDsugx9=bL%Xs(p4-{=t)o zv351O1aLw9zj}2;X4G0&lwbj!uSd9BG@{v{#BdLA!8#N^B>P4;}RxG5pHal0GL+? z=Fk5gs4OcK@6N_14VCctNegyb7{dO2U`9Hktrv@#n)kP>+yc)ITK4mFD;E zomH33`UhO@d~0)Qu{!gQz>+V{VyK!tcq7Q;b8@WME`W?92r< zERO~$s2lRRXBBEkjuLX^6>i!oTWPX#!Tlh7&-nG#-Y^x^A4wNA$t!T#Db_qRADOa{ z)NmPwBPghgw8t_isE6SJsqv)d3Hc>;GBC(clKoRF=Aw;o4I^f@$`oeX6m~mKzI^t& z4ODk{q>uK=LkM3^E-pGB^5bTNj&*pXE$V~ib*2U>H-ZXmq4bZ#A{sKde%fN>jkO9W zunkw#Hz!n&$$yF!Q3_2klkz3_hCNVu+nK*Glz4W8IVB=hv(63M?%>Ef6DfR|8zVd? z56{SO*Xx8Pq&KSGh?OhbJODfco`T0Vpp`=&4$13Z`2fGMbVEsRE8_2qz_$fRasQK7FQiFAh1yZnOpX8GUS9>iu5xEcR0%d#2S-2LAv6phI#SzJ&3Yqs3JKn3I0p z^~Rz}0_0aSM12Ga0HdsOor}ge;K6pJ){@%+EIQ6-0XVZeKN`~9xK87;iy(yCw;x+< zeJ3MOUJskrond+L5yZ`dN?Ac&ECq=+7rR~tiKS0UihKK5T_9zznzt66Bf2QEy2C0j6Acpepx zK_L`EC~TId)r9btLU0~Rs9eqdYZo3Xhq|U9vqYGDG32;(nF5>jfYx_ote(9_B@bn< z5ZJqZr?;F3w0uDORJ&;M0%Y?WN4W|)tE%2}i}a?S2106=_hri0tkS{r2A%C4A=a2M^kVh#d2O9v&AjVjJQKQH4A517;$e-y}Z(*I@&9^OSwc)_{V4C}X z;AG$?rOysOz7(n6{0|%$Gg}`g=(N1deU!U0lj|U|E-pJn-Q4A!`$S&?x_de1a{jpm z*WaD*%8xcxHGv;jkH2pk;~c!F6S>xdkG)_{!74RVW9sm2YT1{VXR*Q=<0z7=@t=@ z?a%}Q8(eW6qC#Bp8KMe*SDr5*n2lr)Bei0bc0XRw@c@^vLyZR9o08YO`TF)OkaIm8 z4N*ECM=`a%4Ve~#YwqVHa^nX^k^Yc-eSiODOQdmrO* zRbt@g4PGYj{}H9SFaHGLtGF3kpceN-9e3}|#I90oQJf$@tUzvD4peka>%M*7t$Dfm zvEJ*McE^@QT44l}BM3}0|0#%Ix+&Zcbbp=vHGc#}y!U)Y!mM`_03kXp^fsV{O2?`- z9DwT%wA(C8Og6ci?#kX376w_t=UjSU2qtjV6C#s}i*@}HA9YhCA7H(dd|UHHgMg7? zCfO)>7Av7|9$r)w!M}G?ydIB+2{4Qre zFWw!!aX@saa7s^`Offdre~!TIeIZTao+~?`s2>>tG1I11!4>(U5ipvdtfT}V|Dq9v zEAp;Mp%xWCZ;Z(ZpLUwq$4KkO1nlvq5f@@AY`3J87^)m3% zhuNEq~+8UZncFo9}u9IM_QW$6uX<=66kw0sM`W zHl>|{*_O1;+C`}rFCm|#Xwv;2|8WzB#@fYY<41~}ep_=HQ~fo!Ntc!l$tD3tc&nec zN%tfB>+5DFLZ2)4g!x1-8YIeX8)*3?-6s9gr9TeJb4G$04kFD2MGg__ZieJ}H6)Vmqqd#Z7 z06Yuofa~6=v-!@E1`kWu2?Te}I5c*vtrbscN_I8=coCUU}ed0X*Pe23wt65sBFKE@PhLJ4h+Q+fA z5A%mvS?Z%y-N9L_P?-VuwB$8eiN5kc3pAgi+t}Oyfe^k81)7ix!Q1(9tM9PIs8c6hv2?r0jwAsDtps&B@-$pw< zv!(1Bxbj^Tvp7IkEB{?!>8rc8_B!?oO0ieZ6Z&{J1i`yh02b&7PIj$9U8KTS6du{h zBG?W*^v(o20!W?jr)P{TRNmv^05_MqA6EAF_59mthyJsXY~U6x0=H=URQm$Jv$xh6 z1>ivkSN&r6){ISZMr|#VqUn} zjQj^^GaS1qf-p?SePq?D>AQwL9)%l6qOnK25j3=l{VCwu0A!wAd2L>kq@WPJ> z{VIIrQ`EH#ohWYrGRF36P~cmwU|E9&xBEuQNMOj4HQmUd{1wgkb|~tZBy<=5tEE)N zFtN8ZbpFoQr+=cLRlE5FE34Vs?~U}IHbO*mzDM`v$G) zN-5YgFzTJ#1o6Z69$mJ^z9^AKVzKw~N1_iWy^?4ZCc7|PwfeUy1Az$Yk^OB@V0RYp zmh%aka=s;Xt@b`+s(O^20yLj(Jt6nsMzcXB+J%4JjlBzw7drz0-tFeD(NCKeMb70g zFq~nuzqKyS%p#o0O*)!#)Ip=KwppGrj7irDT}UUl(o3$d?BE{Zh>cTnaP}>fR&D!L zIfLaHW_S6KYFZrTRd(Lwatj6D$mN#9$knvQoBI_D$>oaUyr!IjZw(MB(GQpBN{%hV1P|4 zCTOiovBS19g6LhlZrTWl98SL^Ji7PERYoH3lq(5hCihXw3zW}c(mbO8c9*YstQ z#?+g+7S37EKewZgl1%oS%8r-(7a7F{bA4P0bex}cJbMT@Mm5t@>4eVKF}6`vTRi|E z+Qph~9qFll-)!5(^>2`WX47zQSJ;1g_V^6%Vj2M7U;E!T8LslVzg=PLzxp{ofIaKI z&AC(H{nH%#!=;57epxOtZrL%+y^SkANc6|M0bG`vxEir`wS9q#7}M9$(4!%9$9kCgnC-@pj#0uYFYP`-7(QF(`^)# z*>x4!5vM~eZ^(Gi#l=z8m^N{!Kiume%Unh&nKh! z59P9rs(d@*tTXEO3Gd+jf(T8|HJ>joV4{`R&8~Qw@WxY(SG=xx#5nI0!s8|Uz0|jy z3WMNFnm$^TvHFwksD1lKplG~YOoWi7sg*(|ac`9MsZND#Utb6Z1YC-8ugd;PFzKsq<<*$w zn6rs}$c?L6Bv!WbKq|p2G$m5t1%0aXOyk+zP@wbkGdjf&AlAUE}gRm#>f+tHm?6^c4tUp z!$~|b$18OaT>4jmH%pc+RCKVF_i3#$Xoku3`QS3XQm4DwJfX~PJCYc>h06r|Qgcfs zR_0QAE(@Gn%9qO_`599FA8!vH0@h?3r2!uz*Cl)SDWNfy13^)ioF4S%n>W3Q&E9zs z1tvR!J4~gw?MJK1yfo7Kzzb^Yhe7eAW#DDCMoaBaE^HpYlD)A8(9Z1VoyAcuY=Vzw zHeblz{SrTrA7#bq!ETYF2=3EeSMT9ZSo4-co+QakRvh)vMHE0!#-sN-o?qd?DRis}dGriZs74 z-JGkXirWf3b77RaRy~^>!HMte`g(w;MBxU$&(a{>&A{7bWh~G|=NaplU&+fCfmRC9 z)iH{FcBlVUWi;$(r)p@wJK}CKV&(KWqV;t}!N+gaa~wO^@7Xke=E|_uA?lK$B%x(bvU`Mf?_U5iclDSi+(&kR9mm%57++Zr~%O%IQ#5snE zNw7J`j|bvM^+0YQiD=GNDIcI8IEt1-vVG{cRc z%H2k8Pv_14hud#|Xbxrs#JceZPRvYa5O>9B5Z0+B7ZgtG#{ zLD+~h2)K7KB(VBe#fa&JZk&!a`$iQnO|EQv1<+0aAwac1stZ;kb zbz1W{)89CKcqlFgPinM$F(Y-rJhPbGxuIRQ7qfs+O`SrxYXn)ilQ%l6Ykg(%BOgUe ztfmgBffd1cF&C-9f%Q0xjLkiIS6x7k)4ZZ)p?Zs;s$gtdgp`u**kZ0BfTFk%=_7Fq%rOUtFWv zf|B0YQBW9L)#jX$SlryA(Ewh=2`f>W>7+XeOI(nSOfQgHrA(d9aELhM2*79qddvK( zODTg=sX9YI#!sh z+$%r#@P#Y*3p)>aiX(UfI^>b{RpL`%#*K%!3Owrt9&XiII$e_Zy~>U1v$xOyC69F% z2=Wynl#JJYW&Jdcm>uXFB7mg?at{2IE-=5aBJM-D9pZ>&3pqlZz^CLdF=w{6?bl%m|emwAzVz z1EG9w1s_NT=R701KYR=~V$eGhkPWxpm{_ z6rS(YvmP`n{`#2QCl`h`8_@k8)qO)j%MYv;DU@ga(itB1S=u8CLYkhpDmy|NNL{5h z`X$qbaCqCYK0J&BAHNdhPHt{Kk$+$aOnirGWPaDa|NH@7YB+;9;$t5=52m^5qjC7$ zd1rq;CuK#YXve8rNl9SDOysG=^$vp~Y2E%bo9~)yQ{8d}_xe|cwTsS&gXN~emlt&r z>apnj;y7eP!6oONU8&o@apKorQUWBL%Of6I11kp!BzZoednuZSb9#s=ZSX_JM*Hqx zN}|Xb#pL{ZTQ)50PvY*sD{Kkpq)Mwz1R#ak=PBnEH+UCUFVl8spA;5-LLM_*bl0lq z)M~%_*0pC=rxl;%JSFr@upEB7Bd+!BjN@_n>GgXd?6;kNQa2ebdU;*6&fB4loWHfH z78aj9ZsME))??~+(_V6fRXtmJU?}#mi#j#}@s-Tu!3Uf5bE9XRN7b;I>LufQ=lsB( zpgbuMe6G!L;QRCNEq!ZIG|YU;Z{}o@$HCG1tr+H=yTE+D*7ZI(kCSMNPe`%)+&s>M z#+U~2WGiZoYcmN->!?WS#6KGmRmR*y%@0>z=9QbZpEawc=KcAL^fjv5jaykTsK;Vh ziRI+4yKqyC3UN_ZS7)Z$d(c#k_DbRTPJExUnA+hweD{Hp7^SAg(DeDZux-mBSJ0=8 z3~EZPYktq=;!qY5J^RDi_P84~X?&V}PCVC%>R!E@CAu@ekm;qdEs}h`!9SoQ6g)jr?+_Pso{#S2zB!tvG zi!>_7{+_(@!$;7^8%8p>@*k%~U_6Ms@&PiYj5Y@}`-A2f`Q-Rs1M~V9aV!8UE~io< ze*9DpPl!`x^=c%uy#8HKsinQUltbE6b#5j3Mmnw@@6++?{-@sRw+}; za7YK<+SYOPCQDylL| zsTL7oUA3t7vP7eRhBrW8L$P33=gl% zN0sP90!U(#ro=d!Z6Hxv;9{tzJ1Wn>G*8ubjlNGRA%q>TmP-((GuvX5c&-1+n$Kag zuL*oDsH2}IBvJS_CpVq0CMLlrva{jo9FnZFV70Q>BCdxyN)en0<VkNMNP?^s+^X0_gG8Isuow4KrOdEFM~tcxNnbmB>nd2z<_RuirQ<_sXHR zNCe;T$)(Tm_*)}ohARYmq{53Y;z3?<7`=83`B9LCY86SQ1M0wtRyk2@JaZ z6uRw)!{yF60NSNi6ckVSE4^_lfcpDj9G;!@XwrPaD6xv<;YoG_0-WvZ6GWN7^Aawq zsSbvlgzmMHbKAc)Phtn4IXL-xGJdRUO(iO7X7`i6Rhy=aXF=Ai$ZFsJ?(ANr!68c~ z@%hk?0u;xTP+R|L*D*P#bN1~*RmOlz2$NB$eRSzFiV>cce$sh47Fu3qUMYsy+(==A zP*N*qnfx#-r>WO$#=o?uw;LyKmZol`i55VZVMdnrWYP@kbhcK>Dsj@ApAOIL$m8`O zz;*hPWYEmv{H4PU# z9&>s6V1(}M)U2q3>h~v|mR=&}S^%48a4lSj-YR_YM#o~%!ke0tiFaw_jPJL%KUm0W z^NR2H&D(mh`ewsJSW@&XH!BRfuFHiwD<=vcqee()Ee^Y@D9wk1X6$^#ZHVYQWeM0b ze`V|A>sOPT<2|^9N4_NKVB7Kvo%^{orM!kmQR9+=sc39cOw9*}?X_gY=8;0(ePS#CFlu@@cd(i~v%Z9aIoH-=2a`;rSdF^6cA-XKMNW=oPJrc&8jR_*n$n&RGC1 z(fsJbZGV|tR*KlyI;9R|So#YNF896;4TpgE!_dfpv2{jEf zx7V@R!}|;ZPb%>!nIuhJ-P@NKHLC=r%;U0h0%8t+&*0!+xpw7>_!Zzm)pNYrc(2#p zMNVRw@UIj zC{fFK41fyfFL{9jOq5Ouapjx)=q#-#G6DE~uw-NUax&PakQ>ra#=wAe|Hd zZxz4NSgWopFLZgB_)1okez3(;)ORQgFK=|+NM*qj7bjJxC#^+`oNqMh%sz_u;HfE& zkr1yziu@hZM#uXRMT~@0l9UN29gb)s7%v?NO(e-{f~3~R9c9Grq@aSya(D=CMZMDE zZEox{4tI2G_*DmSMmR}JO73oz;+vJlEsf+Cf>s)n5#cLNbzG?u zmRyyxRgxJ_MTNT*lIt~Z;u_g^9a+Bi9BT3NzQ1T?XcU7HSO@rAI^-F5D19L)j18c4 zI=~(I?2zLWjku}j{7VN~U@Ipk9yd~*^0XwdkaB6w;WrM6Nv(Leh;5#Dqjs>0I-1`d znH`^yOCD%v_g;#A2BaNqtg)wm*H^W{mU&HV)hy;#V!8nTy?G8P9(GA~UxlIGK6tnU zx{ioEcW1P4;V3%4s@d;(*>G`&KYi!*I=%_^*R(Acn(7#DWsuSh68qp3m zHF0p5za)1Jd|y2#8lDL}7uw{PxNo!Pv`%iEzWb9}fl0mN!*UwpruF6C)(0NEMIASd zFGjOnUuL)VAm`VB5e+W6It;93#N^tOQiQ&f7*Jef5y5Q&+B(rhS5|q!gek`NvlZ#8 z%n!ILf+Q5daV;U8#rK1`5DU4ct}~HNH6I+VIE_w$gzk6GtjWC}uEd{t`+k;-hQz@G zhY}H1X6G57Jd`@@;Yqyt_2HV>Bx@0wVR)-h6WAtPd zU;ncgZG2n~`b;b;POGdz#^7MMX7I>e#yX)7CG}*Wgf&VjMHS)T>cL)!7$zfY1u(?Dj0 z`F4_T{Erq4UtLV{RAw{ffg-2Q!F%?l;=lBU35Np0&gRFveL7p8E6Q1 zjdu!0uKYXe_Ehmh`H{|b+t9X^)c2Clp3vN}(3rtZM4-^tS8QjhB9HZj3ecn)uV`Wn zcsa}=A4>>@ctF#E`Xk9hu(5T4>)lt$l5r#gdUtUg^+4%E_cjVOUwndxw%D;VdTF4WyULpV0@;=tIViqU$Bg|&WQsb*FJK&2Ii)pD^kRsJV7c>%Nc$9KE?GY zj~xa6*0V7lj|z8m#d7jb>N#iNf5d$DplAL|3dT9=5rUvkqVg&&Lyf|XPJ3a(%8d=m z9QFrXsJ$l%-t2V889|zl2faR|aP+@wOqH!!y_yxf=X1TVZ-1G)XyL1dWqHlo3d$zZ#)S_|`t=oO#0pB^*V10zM?Axs{(olm`bV=@;{ z`*Gd5548!F$Vr$=IR;Z6?`2uC=D^&0TkEXoV z`t8X|Rc3j_boPCJKi+VKbOaiBpwLKT(~D=6aU}!yD$M_AXozKX?+nH8Qe3}frQf^u z=T9ZQAXP>~g#&A~tpZe9K z=B|-7_Mi1xZL7HSw&c>+(YHyXgoVwdImjg%aprZiJ7Vtb zvdV40>GQ66Uu61h!H`@H8r`5$=<5c-H=-%mE@23#;w!(=U86@%$XDBCKoS!D@u@Qi z{c=Q{IoVAwgd*8ZYY&vynj0wEJ}18bW;Yb{xsW=`kCZbs*3_<_lBuIXcamx4hIQ!eifX3vhwCBde4;UT)R-YXSwE~vp7Z3 zsM|@Ao!!g_eXrB~sJM)Q>zUzWoArN_FczoaViwg^R9Q48%q!#sGRdEI^c?8t3*QSL zXRklkM33#7af$l-@F%9DNnfzRHJaQ`$|NaJk_9hfGhcOW8MnXhDx}+PVrDI`%r6iVaqyF!XvyBj3@ z4tn43`~05o{ZCv|hh^r9nQP+A3H^ymZqxU6AN{Qx3*};P4_PU;a}7O>C%H@Y4kT%y z$kL7nkqhsCSy2<# z)g8?*_6uGg&2#iDkqzF{w4V3aw*sPeH$htmzuu<~^)q7EN!?N(8)N~;c+AdxjQrQe z3_fJumF&-o%HCZrY*8d=?hP%4y{@g}**jw1Tdt@HDzJZPqx91Er%bIb^O7zdnJ;W& zm{*6TjBjzQ3>2ASYDbvRpuSJa>vHtXgbZwYq@C0OQaiuG^p)}S(0u!T zr%oL6LD>8R#g;_oJJVp-oEyfwBxmoi-D;56&RNl&)yf^gEzj2WN`BA0TX8dFVZBLo zi8p(eDb!YRd9v=-$|v#iHxf*K%odTVeQY3#0G~G1?M@3)K*Fp)R(Fr<;)t6}k;V4P zp46~JoAn|+a*R?r{{$`kMNmbmV?A4yN@>yb(_^yvNFnVNY>`gg6u5N548KF|{awxr z>5ik{Na&d!=BMwaU-r!|c;w!`dVko&Lh<40Lf6k+53h0cy&T;6`834|SLM3ulwsmM zA35Js_;CBrN}Z2nz(jA0H9?pL=S%q{?|d^v@4o8b_alX|B#t`LcFhyV21QCI1{8;N zat)^;8EP|HI)~JUImDaGo}_Z)FBhwx@-YT!Wj^E$Mjye91?>ZmmZB0?{uI0A1SiU|4skV2| zF3pq&;SIlqga{F3Vj}|uiZb}rtQ6toU-o((pB_kfV}5_PKu!38M28weN?LX04sK>+ zjLieSY{r528u@&-gbvyDP&4|ogICSF7dy5P8GETg5t=O5$NI~>?=WJ4a>n0CAh=In zi=c3I2vXDxXNX^oSSD=2~HB~+#(Xycqa?nq#e*{DF?CqakD$A_Z^#Cf$LD0O(`R@xis8vWWR92RqAS1%#S(rbx zXuNGh|0o)keqHfQ?siHFm{;fl^JuFdFy)2xmgX}$kJ60m@w?{4t!H@01hoA$!tD>y z46liP%_DfBW8~-J2~&wUJUk#f%YNzhmAl+JEJo3%wyCnnSXc8^Bl|Rdgu{18zkVZ? zYUn>yR@EcB0DJmyj~Y0;R-CBMKS=#rXZ*e9N5ES*H(?)U0fV-1%>je%V&HtMeI3s? zSZ)izh4zq;=LF;S)x3NdRGF}T3(CJ)M42ea+|QXq2OmB8-a2euiF*inddc$m1OC!G z3+ktvU!5LMT5{bGWh%Fz5%V$RQJUD=6U)e;$D^XF`F!a3PcC^q10~C2SRTOf z_J@#UA%gKT4vs}pa!eZqYfRHMRZUe+j-Fbcra~knw2XZB@+~3xLMuLG8ubHsxuBb# zvG#XL4UshY>%RM3LiMBWiwgJooV--Ma-_s%UgrI(X#Mrz`vDP79$ zq^oUtznycI2e!R)E1SbSjziLa%npE#2 zF-eMM+D`mF`3b6_A+(1X=eUD5_&`xZDBL)$!S=o;j0vj=SBQmGGrLZKppV}lfh$1W zME`gKI3boHmM!kM`V%D5s>IYc|9V7fdOzjBlrIS=M!P^pqr1{sHy8-Apl=NO=i;R( z-kQ(EvZR)sKt!TWB<4k|WJdMPJ-Y^;{*oEMcDu@((xq~7I|p{w2XU+GU-rEiDv|3m zSnq1SDXhA)_5`+9|EBV0bTWbGygWQz`tJXcJgB>}N+=iz5K&)C19;6=G7lamHApxn zb@<*;+XIWSeqPrx^<%D7{;RGp!?4=AKF$Y`jqAjaZ~J(t3on9PVVsR<-WRBz>r0|u zp*Q|}m!0Rppb?q;*QtkF7|!@Pa=x6d3|ao;*44UO+iIR>he!|W^0IL0w#`?LIAME? zN%aeFXK3{+J!o?wm{562jNxm41S=Z5YB@mH$vN?zC#q52Kxtr{jwe@qiB0xx30Z-P z_TkYtx5)|=z6)4k%b@U*)oZ74Hw`#WUCN@6?aEBKG?WYsPwtZ&gd)s15~Yc|QI0wr z%aY?13-Oab7MGb%y`L`QS*dl7c1s7p;EJ2A>uCU2vmokrXpf`|r8yToY_yUb*?KE8 z9Y=MH@JxSJezg|6s%BvvO0WTcj72mD1z zK6^4SOg?iI(`>5FMk{UiGw|ao1$egw`+6qx3>ATQ!~1rt&+IxG-|CUc+67vwI2Ora zY2=CZW@o|5YFoZZ$)aakus~v-7pVEdfzrG@N^4ttTU@m_>s+z_JX%H1uRb0Dh$QdR z+(jo+c`!n`w~EnQo&M_@NUinf?BXnSmR~8}L|#j86(2u12Zb>;mZ)>%$@%X8KU;F+ z!R2;1+M3J~=z2L%)Mvuaz||XNKl2An=5wgpMc(!BJZs7Vt!wvz!h)Z-kf;-iYpz># zeFta(S#Jp?PhyE|N5h*qUm1g-33t9qVhOtd%?^dGU;~EjE7G+88!6o7#xNiY!-n}? zRa1kN*r=}W&U7U?O}9v6!QVuVe}m-mIl9hHm!HSF^qP6i4}|yJeewCe*x8;zjU#r6 zWs>ely`WT6BOgcpS1zQ7KN#JB%>RT@`L3@69{4huS-nAG@yPu6>%8a>OxF{AaJHwW z(SIHr@j7JEgapEJwxwCLMxj69F|T^I_czL3Fp|7S^y~C|7wRGg&sflhM>R)yd4lEc zfEs8sU7`G_QktH_uc*}78M|CjCJXIW{1n1#4yiP|P=STa5*0jS#3{x2$^vH9NQai| z7)= z3V@{iB5IW1KxwoZZ?gBC{1XL(K~1UB)nc9$FL2bwsm$5PK2Lt#K!^Ayvshx;=QC2i zX9YwJ)6+4uK+vu5k6!3G6SMc7Puui4*sucyi(!zcmGO(84RV{~q~Qj-c0e+V=?nXO zy1?OYC|L8PxsPp-S6m5;#(;5ca_b5h;oje@%oi{kzQ?xG=k?~J{`CC-=q(mLey6T~ z^r@{YApF6{@3t~7D33lJbCw0gK9e5RS%zdiFHi^gI|AVEbZcA6Z&LF2Y=6X}hNY1& z)?4+KU;VVSnhs}{3gOo|3IWpKJQT)>musNW)}P4|_Cc|Dol*brw@Q{TTv-v{*a7o9 zQaLP!IUR>(ru=E2bVor)i{+6$q(PUUv~v;?9zwz+L)`vw6=gOfRPsowy z^Opsm%DA3|wphj7EdA@%C*Ex84+F0z;0_VLBZ*CAh2{h&SS+ur4UiZjEs;x#eS8Mc z8(P@bUj=MVpWzD7A??pIe!Hi-;}3?b)_;9|{-;C!BG)6q8XwHC)pLnj5joLTaO+PU zbVF;F(9&w6`=~1q01q|qUX%Ci1~~dqID7XW+`Tm@bGh3*&qU9>arJIl@LQP0TD)qd5s$Xc!(bpV!Cinx-& ziAszN`uY#}#D{n0%%@uZ)4lEVk9tD>OqrnHA+iP%%KP%ay&k1J#{0E?n{epJ`_h3Y z7GzG&xoXGA|LsS|hue_e+f*;49U{LvQ+@r#mBF$=x0Ry47u31M0xfu3WkFO#HIW6W z4B~t7ZCQqq>6h8lCS?Y+K-_lJ~fb#M_ zsVRp4lBQ_4S!OdYCv%ATT~f-DGxS-4UG4u1dPk~`jg4}TNIw`@e&Kub;7JnOFQ}+? zuM>Hygo!PDN{baFAd)_aR5~S?xh?SpH_dUJ@n)>C!;)LyVt%=bv4rcuK~-g3mRd_o zYUp7o^UB@S1z~}5w^yGS8|4KH2C!kU(Dp4KHueB+sa7PXktbKh07ftdMM(?sEs?dk z7BPY01#%Mz%h4)PQ;FA#mmc_R8EKm^lNliV1BO+chCcqdE-*PI7 zDQ3i8)ZLn9EFQRC85EA|k^_Sx(^+D$Z{xz6cZI6GYyC~hq$MlXg%5m8k6n3`V41$e zbHOh63J2%}fJ?X3y_VE2dEKemmn-UnMw=5X5bI1@Ny*of6{7qXMV6U^6z==9uONM# z-wFqpD+Gi3_nA1qW%I3*+50{6Ec{=30Hnz>#}s^xu+qhB9jW;IpDx!}(&W>peSdOP zWsJ(&tar3GL=VTcAdo$5i6HPoW=ejpIj9x|N>YGz6kctXe6Dxe99*t4b5QtL^`j-cx3b0J^B8hAx&nUIV;wZc z&M~e4CkE{;%%;aHt;k zB6i7Y{vE*Bs>d|g^UHVXSGoBnpE1j-25RToYM{egWJ%`K0E4S<)VU@|rFaiz$v&vo zp6f7kB}|zGyE*_-5dXC4UQ5bYjo~kWP7wx9Y|`Nl0B5WK&KzOs!7(fty6Nx7t!?nT z?XZU)&0$c_7T%kzxyu<~yn)r%-^0=xr+ZjUq<*>%Uf_<2K@_nLt;%6H%Iu)=rD!Ys zTs}zfN)}tp+?2hf~TATFZYIv3+)LR3EKg|2-{ZzZXW-CK9)-W;^kvVO-{pCzShm|#=W z?OR3O_-Oj)pz#W*w+;WJYcNm&H_M>9Y}zkVln;31YRGuXnt*RZWe*@+oSeJ|bL>?P zzV8tewv`7v`UHx$H+$KgltbblwrcZwln@ zG%zI#BlrvJc1TOkH&%}Ij5Y3NZ@lH2iY=BgqnI;*0}%vHU1U$))yCU5XCM4q5lIU4 z1QnfgyWO9l6gdeB#O(53=e12mK(Gh=mrU)ACokn(4Vj^dJ@95}AQTJ+*Lm;v`nDL4 z_% zpug09*o{B^$Ugg>Juifa1)T^MEZ&iBl*&^TM;<3N+QCJVDrBVsta!a#Opw8_?H>tO z^qK{jv+NEjJuo~Z4x2Rbsh<*M6C0JPn}KojGPr_AjL0d*tzMh>$3=&^@p2{m5rp`% z2;&3TUxcCTHJPvPbDdfX@|OQ}{QFexh7V#MW!g+yrtuy!aVsZX5+>#h+@KAB*6)x_ z^Gi4hrYGeNdQ&o_V_R~9Q^17RlZR=27cxcJl7=ge`kPWiU> zH_}AkzmKp+B$mJZm>W#r{7_W1a{+zn1KaE$2s>;vW8NAC`O-2BZ1WVK_&hp+Ot>apQY20mG%S3-nXJ-Vze8EC z%0J*e6y*j9gH~7~W69&BPeYxVNd?^6lDtvY>kbeN-4_&zJ58}C)U)1J#lLj=!yT@Q z>$mJUguDKycuOR*xWJDi`}K%7%EsjkfD_nt(me#Qv2S+|TekzU$YePZ83C{RxJEM2j-S#J!n!|S3O`Q9tL?eAoH@7oIxx1~4j?d{k7?GqwV zECoG84)Z>cH3Va1KMSZ*mn5^GKEOJNtEfzpsd%foJKJ||3%%y$xJ^p%9_UZ_+RssM zi@|z3;m*g@)1t40c5MG=^}jlVA)nR{ml$%UiCm&#{~v*{`f&wQ=U}ftWKzn~MVCcX z&b_(BW9&2s*!wN|KN7{URY#vx`q1xqE)CF^y!Y?n|8T@OKJm=EnVfO{0cyI)ukkW2 zLV7tq{T}Lv#D!w@N-JkqE?$(ijT53UTO7CRJi#`Y$jY4Zyo(;{g@K*(ShLHE=G_lp z*#FFcS-%{j{5S@_(sGx%>So5SSNzOkUZlZN&Zh$5H`k22bG5&lV`5hwY0eP)IpH4= zP5F_J=%5AG)~Jp6*D7)7*y6}=wu~TuEnI`34Ql*v0FqLf)R*4f%`4buNQh$TD=P3p z|1Xc0UzdSMpaIm3|C7YAjpK#g52Fzj)adB0S5*84=oPPjmYiDp9J&JAD#tgCa&Obn z%ez^As_&DJ>s_`|P&m_h*#Cz#_<7dn@%j*Ush#suH;uVb6s2}#_}e|+&G~fKwE=yl zChpkHGAEyTd*#1OzaIk^<4cJ=ghdXCTw3fx5~ot^FI+F6T>Rxe=_ z6AfuN`3#mXZ7rk%DlV%95{yag9KP?NosJ%mU}u_N-N;2@_`bHM9ywl^)c~3$f@7$- zvY&l`MUJ|-u(MyMxgU1!?G-2CH8V5!{>NNb$u-3aZ|9GTKF#3I6U&QLZr@n+Sl*liSfAjg>|DPX^SHAMT}Q>k5>PWgdAi$ za4Jg1`rJijavW6i%C_{JwX$&I5<6-~ZqVoB)&G@9toHC2}1o_dU$G~Yt;Jo6)N5E-C)yMdh>;x~VUUP`45pup!d1r_Wa)=z^vOg40FQbn5 z+9L7s)1M=X5C0xftOC~Cw~_z}2RoP7YvpJPmY(}Je9;(k$`VEfA2|w049}l)l2dNV z6>DXS7sN%(g|vN;+&9IqjaOPXy20qUB=PUk?wIb7t0n*PqrNI|vX4m%W_?^0pqFDs zu)W~|`~Xb1ry@%j&1taYtV;5vJjmTYrI+D|xtp#YnqXGP+#9$WC?9dPcoiIf75sE* z&3BcV@d*0<1V@>JF()MkG^i-l|25U9KPy$62#Sq8IW|p0!lhQk(A)2+5X}H~Y8;HF z;!g6795t~^J>VZTQK|?5jpu-!UhSG1+pd~SB}Bse`_*bw;ucMzb0PgUKjH2jKZ~@- zTqIorOIrad-f;sH!F_SV3GAc$0w`+mA{_K?^R9*UXI`~y|mng^u7m9NR1kp*J;=xc6|~E z3P|{*yE77XRauQF*qtcY4WA2zQLj=SGj=K-N{^7p>Eb?>m9=LQSF*`c^^3X8u#YwS zK9xzsN4`M zp#a1<-dlyIx#16Di0Os4_No1`sx-w|6i!s*-BlbcQZIzAZ_=FTW_vC;ygJZDe8`EV zGVEpT3HhRfz-9GcmSduL9XJZlc^#+;R!6uAyJav)zC-=B4Ncr*wWQghe0i*|(h0tZ zv+|lO#@ND?co?Om?0FsN2oiZ67-^ohtretVr>O~D2Mp!Ca1#LY6 zmTZrp*1V*nl5A3ULab*}cVv{Bb#nA7Ye7r-M(PXbVMs=hqpGf~5(`eUS_K`~L`1%H zhb#l@ljB#bNz%tO(pi@ua@Z3R5>kScQFzQuUCfN4TLu!QQWAb<9Xguc8wNwiwZ6E6 z698+5qG4~L{yOL{pnf8^Piy-1ssPn0MQJdj&)W@podC|($b(zbFOH5^I^P#aS)H^X z@}T!Nx2TwiS*y(UH9zWg|A}-Zpe%7lyH?YDRGDIn-*~WQq*kzZ1HY6m( zXC3zcF^Wqr_jNalKA4?qFg;4@WT?L54}NUl_G#6=?c6WOd?g>$%y~~p z0<{wsSTGT@4m#k;fZA3Qk#b}*;_2N<Df|^CcriFJy!{?)qIDzY6G`C=E zTN}Q2aPM!L7wQ9vXj6eFcq`wd_HdRci(e7O|9JQU3R{8(x=98)N{=9u*z+8?HH*hJ zgm2sEUnS8sNg24ajsNsoxqUs=IHx>z)n<5McXlPe5p-!5bjdxml0TBeP`zOUW3*3M z5e$PdUAMyMrLUk)nIku?1Yh&y7KleI*zj?VN-e;;Gybd=U6YXk6*UJq+_3GMQIjrk04MM_jLzOaY!@;GLaq z2Cw}Z04NZo-29gUYOa@iB@5j*f(vjV>l;zvwYSsOfK&KupW%`pm<7K)q}v{6#^;y3 zlF}={(<@J`43w!<3(e~Mri)mIo;<5};mG*Fkiitm_K>HWet>)3o9-ZP>n2IGDPTLa z`B$J%s&vb(q|Xplb7R`}Q0wi^?dzY719kuk4~HCqOMh3^_*ElGaAP~__kn$jwil1e zrp9--nQJx{rij82nR({hOdp(;+m^t#Hcd3-{seR>>~|zeN*7S_ATBnu{&PKHZOow8 zIgFa?o~)^bo2u_?c!lGX!RkLV!MMKOc(I+i$0q6Vb79Y12{?|(1%vqY21ZCNX*YXX zJGU(PXv>3A0?-QxMX_O;3DufE(<}S960qAWTIFP$kZnE_W^ePT{-hMQ%m8=@Gl!pF z?}*1MOi-voDzz`CfWcZNlSE1scel=?+<&mzZ7sZS20G=U7g&FRS-xA(+CF#a2=AN$ zX8a5b&E{oBt-9X)*eJdm0jW>Dei1!&6MBKY#;R~ngbun7&;N);kkX}Yt>fB1vZ3a7 zFB*VgivZt$_#-{MN4@w~pJ`AL?Q`S9F|TTq>JqW=8bjw4F`@>HT|r{vic@%gKR7xP zOtoX7!N5ubGZ3btC7)8G%DTuXSvXRoR=p1tx6n%mrTV4QUjl!Lf=XL#yKOGZci99o zfd)&_u>0zAYzhL&*JJzo^(9DJbz;me^uy(BnS2J^Pr< z=G*I{)H6k-`!72E)uGbre3^`G<$v<^e~l1nG+L71rt>Y z(2*hs#J>33Cj${Jp%)lh%tYvGtV-*D8?(5&e;>JT@V{CA^v_Ant#bf2UFgn0y)IZS zKZ$e}XKnS77xf=HG+{IR6hwKaZ3oM^ZwFy%e@OSetG{k)bOh=Z#CpkB zpI`!wjFmL{L)*W3081vA(!lO{)JW;}IvA)cM;V>ZHG#q1@WC~z56tJMyz4WziCXW9 zyEYxNi<*w#NQSlsizUEb(`{eLZu?bm+PM+<1TXyDD2GGn&IA|>v`Qxls1Tk>x#qU% z>2?18+oWW<<{EzLWb%)-%8WV!0+tE$LAf#H-^f{;bGMvtW#mh?1@wGWM|XaO`Zq%0 zMGS`ejv<8!jj7q^9!XQ%c4m2vsR-0wwdK2yQOfa|6bV%KG%-NqCSY^&dGwEch%oQv z(W1Q<+0Mfm7YKYE@n~vd$V^i>Mfcpb4Ta*TgK9&zq*%l$ZPxKyK&%t%A@KBy`CzrD4FrBYGW z;dH(dHw%$nzDMYpvT9S_Ht&S6^sk)UDMlqwW!f-S7k1<^Cp8DB42a8_lSSs0DWC}I zIfKgL`%O7X{10=uu}pu_{$pQlcdAyS0<2f-(7@A}=$wtfpE~pIs@T2_H=DSAF;s9a zoTj;JUC$!N%_)zWX_lhB9yq*Bj+OooAmbSlP-|%V7y(E&&^Lq>)LmMdwo3%x}7`D-e4W@xPH=-Zy5RQ0x!2 zT216tPCME8f3x9?N_%2Qm)DS*ee99cvkgQm_*GPw(Lm`BCy@z^c{tNul85PI86=!5qze0C! zrO&-HJw;VZ2h|EXmQ?HSMGfe^c)K=HxnMLoJv2SB^%c)?pE=KE{N+0*72eH-Oz+TN zyqhT+G}icK_k9Feo8=T#Ursi4FfVU+JMlk)3hSs$p@Ahne!-+{+k`sPt)y&m5`$s< zbylBOaj;O@s|1ndwYK_j?c>%oh;AtmwjJg5a<7;HcL!A4S$c%-DXTUDyi_GDeV&th zL**DdW_nZV1wDP_!ucUeH{nA`Ufb2`4_Jemqf9+gbd!ufe99R zX!_8bH1W!7+5J>e244es3LA$A@@_77<}DI-+P4`Nrk;C1-dGK$e(=-6nFC8`Q2Dk?>TKw0DCTio10LNTmz z13cQ=Fbg%Y1FS&Q2llB3pEt6B%~A4^(4nS#B&?wF0irQHKP2HPjJbOXMB%dBru9Yn zWJ1|B6&1Itbw8W=y*KBq4U(JR+g7s;8r1RXJlkNKBP$9%iQ5lNBongKhs!rl7=pQG z>($Dl!V7z3CQUMTI8@IP=zsO9W~Uix{(4Z5;4RClcf8MVmb26VK6EH72wu#zA1AlH9x+IAIwVbHQW5CC*m&AwAf*g!|-gHpjm(eG1thiz&t3iM^pH)Jn5%7+V5 zWA+b=q+{0HD>7Bi$VmfoH^+r#q|>(rxUp5G;R22g6vKF7hMN zXO@mX^Fd(Dnh_Q#Zy1D0fYMdQH`TcX2r-pl5^7-saxoPWX!%KT776B-(Db}m|7-{; z5LHtj+ZFvv%5rS#cQ6OYpkc+;S^g!3Gbjy`Cd89Qp#wd-eIR1AaqLn|w|bp*uXF)@ zmmA1QAD+RfF_Ox2-^j>VB5*_5ycDNEG)X_;knm#Wl*L zfkvfXv61V}zZl5MzIg&!|HrOQ=r$G-4$|S@qo>V06jlDNgwwU`H&Pt0C&iW1f5qq} z%s{`9a)waV{q(ZzEVlpCiQ7k4dKZOrJ>;aZ0n&){kv#MFmJlG}XWoi+ zq$QVZ3gAE(^FQ3vu@JtJ8_H&2*~|#%ewpkcG6n|P)5njIA0wdvNu|MSXOy=*A*AMhqoVe?bxGo#k#A}zcqAdAe8sC#L{FB_6@(Czc<q z$i|$p&#}j0PYf#TVeE|(nDtPxYH_v45)G+;W9C?JD8c#iukJV%8g(nnqP3DK`a9P9&&FkvAirvS=UQ zxa%6Hk1};%Ut>SsYDYk(aS|6@`gk0ZS-z2V8t0{Xq$#;bO>HE{;NzpV!>!EOK-&ev z9H4MHgiUTPPrfFz9E<63%&07#9o;ym8eSBpfPeAG1VXy%t)TAfg|68AS2F_|K{V96 zIp*$$fwEkW<*O5EQ-o{C1omrno~XjD7ICt)Cbhna^jW%l>*aP1Xs5m6jDr&3_fk%2 znn+;r%)kByo1mj7X=nyzD_I1_J!c7E>{XNE*iYGvOL#+x%@c=5nChpwwftX&Mo=U_ zA-``AD~5yGD~LW!QXQT8J?EJ?|mQ=2lGvA!|!WwSPU9 zZ0;hw{7#6%2HGPv?MX}n`;1dBpdZ0c>8e<|2=M&SdsU}#{7Gpr2ni~q6SB`I+MdQl zoAXRNrBtPieaaz+V)2bMOphn$2?7^O7gh$ug~+j%@Mys+$!%Czb|%C*(dffnt?W8! z>To|f9rO|ouxR!k<&(I35^az`vGbIl6xSxPi)==xIiT9uy^4xiCwcozVxF7(@>(+S zr(sRswhCYGYS!C-wcA7=iku!q6rL1f(zh`9cF|t#(0+_X7 z@UPS)tjeBssZ*RH=jxTof!A`=IH%*+K^&Vugvu&;BH|?*GU*LtOvRs|WUsC17#S|f z5PXtfkcxZPolt1WHjg!zt&g2f>uA#xQ-?!#Me~s%&eNMX))rB%LZ-NT`Yw8_ERU|O_L>WfiL(O_VMd8%=qw4=kK`+U;pXw!ma!f4FQw79 zQ&rx3nTQyCwBK^X5Nif9))DiPOMmbN&+GMy3uf$EhW`LYi(HWVW11w06h^fo!3y`J zeO4=!hD0OJoLQY?(6H5LbuVJf_?FWhJtr<7Uyw;9aSlS5f;JmC^*zKe4r)7F#3ZJb z}|4%&8Ex1MkHc3f%7Ba-iX1N7L*oeCmWPp6(o zXhL$|uDxxr$C7GvU#}6ex(HJoEDr3Cp);xYMrl=Uty23wB^$hT9En=77pN!3mHD1o zo%VbQh0&9+Pg{Y?{_BYHH)r^Kwl}eSHNwA)-bblbJ<)MvEl|5hhCOP(UtUjx;oP38 z4UfTr(K-9*=d|@!mEvVyH>INw%I+RS&HTBVdj0hhwIzFWEKYmKY9C*syA^)ymuTjE z$YIAqt}1G=1ZJ0qgTAe2H&HC%0wZ(vJz36c$*7Cl7k>exX znZq$tZ6{9F>|;F4@v#q`0+7}#jOD0e#wJLs^|kCL8i7pb$JPWz{k37J-~_nsJE6}R z;J4XY3W$VP^=jUmtWS0vP@HJpYR#uu5Q654tIUpV_~rQxZ5bK0ou@26h{$Xn9>-)x zJmr`q2+Qw#v+_0R1{F6XrRnOR;W56@n{>ocI?MJq@uO-@RB&&fL5KcY6Y0AuYgd*Q7on&Zq!9Vu+@ z&thRBPh%Q?9+=@K78H^}o|yQaqs=H^6;}05n{R!rI(vNi(mZ)M!N8BlA-ZG>2lgrl zMoY^=q(bX!V~i7%k`_yYvzL}z#;E$_%ucfh&oQq|KF=_Q>sg4vHkzg}#!MPvsMAkG zxh`U)9wNcH)Swn4r#@%(-WQ+K)|%KxiQ*+pR`3-KVK9@pZ(Iq;isy-Spf)bw$tM!| zICI2vUtSbuety19ix|h~ZHmC0LPg-?+^PgZfBL&-e1my)t2 zpN@zz@O_%>EN*^686K2#9&eC=YX!{B8M-b125q@OrHSD}{KvVJk@JjVcGk~0+e4Np z^$OkZV%|z^^ik8!r-kG8K~EJz#=pfR>3>IPh+ovy6Ta={qRaV|QNDxY$~$<98hDLE z3$i_}4-Iqpd{d;d%puTw{`U3hO)EG%5g}pB3VJJw|I;;buA^vdGmF+>2|4bI>_G&K z@}2GQhW~=0;iRtOY7zezSlx1)%~yT37{>=sAgz%T+UC#SeiTvCg9F$4fAsN8DIcg9bGc930^+4LyV4>=%)RWDKC3G%ZoAKM8d9fq2%j~;FVyxa>@5LYE8~Iky z?_JUa-fyI56|95&N<}kmd;|l{$Q&+~>tqyJ&VBG>mCh^{MmhT+L{10cnr3 zRY^CHQAu1mrK`JW8pqh+a+}FVy&xc<76s``%@OgsR1h3;Q>gF zRiyEGy(f+H)6QTLfyCzR7m4Ck{45GzH5~V?{V(UpeWUj)b0R(?kDK+rQ*QQxF-S4Q zl5P;by<(!2vval^zJ866*T|U}ecc%4(TP-B1s7=a#ax4YE!OWOj;z+Dn?r@WKM{A( z#)IeQ=^yfu#Cd4^=qp-b{QP8&68&1dv_C&fgFD2E#Esu2;qB}^Os*QHJdk7JKY*9A zoQs-NLDXi}Y%z0stRexzU>4}hYk-B85fz1*q_XPj5cjphso;U*jm)c^x%5;^4`WJn zNocd?48_XV)*=l9K$_?OGE94%;Mzwo8aY z6I8O8!Il1*M=c+fsaI!Tu8Dd&{~$!-fTMM-jp>{CfR*pEdapo$Ll|$!D-Z-5j6##O zrkm>iWb4)Ph<=(R4akM26YHm{?CBSLHTm5BN+uKEZ}tjPgT348>>-V>YaCTyyj6Q; zqMQ8Ymhne0iZ7-A>O!2lUz`7z{1QqT$1-ldEO@paLO_jYlL_y5x3{jvW)=#TUc#lL zszsRvVfqo$i2GygFcfeOM|=xO)<^n84k~^zP=c$(P%7>M_@J=L ziWlPqK7CVy^uQVxJ_UnuOjTV%P@1U--vlL|d+2S1=2Z5s8l+dJ$SX(xn{t#tK`Tz4 zc6;#8IF4~U^5&fKB;-Pq&2Ps~ic$ToJeE_W-NuhuQC;lx>*7>XDTq$u zr#`Qj&ZFokZdO9I3Ej7xt3N7oMB;}HRr(LFXo8##wXLSILV6O!gn|N3U0Y3?a<_Ns z%v7`GY->9CQFIk_q653?$iGc;vexr!V(BO2#jbuCA7q#*HafOCLFn>h$SRX3efcrA zhv84BB|Xql`Dsm@GLYZoG1sU$@5S;^s`x{FZIBgp)9l6fN;xY|p3D;d)(2l-d`VFo zWiKCyo6(=Pi1|hJ!?bQFT7&m&#M&*XsU>hKFjk^3R*qfF{S9rc7fu!nrMXJF98ApL z^c7H?pAd%pDKo`Ek-c=`ldNB%750y&>E5mvll>fOlH&KsN!>?)GP7jJeLO& z?dKhARAt_W1Z1NBMCf@y#r$$@G)2D>U#?qx-3iiG_@p5s+)Tq?Dre&IywnCCgM@A4Q%$|v`vh)h!r6G_R*T0C;=F1`&J(p% zV9SXTniF1{_;M*fV~*Od(SzB(cx8Udt3eF!wo!!^8P&MGmK%J{Zsk?q%g|FSJ2Zj^ zVw!5`eJ-M^RUw2{XQY5Y66N@KT`r|o?v$^)zAlBx-)$VA_~|f+VAC8;9mV&2n9=vo-FyP(S4#47FRXA zlvuqJ%+ij`mt!o!dEy3ANNfIkB>YoSL#1D_Xjb7YFBAvAIYJ55+oJN+wra%LPKo2A zlDum9;f+4T`)|s`7NpLTCAx)=F(r+rY%pqK=H7hWC5I;{vBQQ$)KjJjDc;jOeJhSf zj~Dsor4Gj)P0fs!{2H|>*3?1~zhPEg!Ju4W8R#2FxL8iqt6lKyQwsvf0V(?Sfc}Fq zOj)^UM@o9qHA_DY%NaUmg=!lTLWRoDD`?mmhbq%UHHmmL{@EH69DKD;w}Pod4wix` z+NaJREBC17J7+qDridg;WtuYYi+tMDTnbf)E=jVd-Ve!{>!U!#GzVIVlXkODA_Q zovD@_uMamU=#F!sqo3)ikC6x>2|_)Uz-)?G!N_W=B(F0)jU<@%+g2fclK>q@2x_E@ z+_k#cRLFd`9x_yj51@ebhG;0J;{qH-=t&|Z+7f% zRyVqrka*T#-TpJlaq3z5isswwEIHl*44;_0dijc+N^s$(bXwLLUD^=*R!LD#kx?=W z$F)+mMX*kf$o>RX@dXGp$!-E@mDMsb?`C7_iiHK_n5<+ z3{-xuT>8y&!r8BxXB#fX$7KIVA)=ZVF`~u|e1HmqwfQ>X89Y*HTgH_DtNoO?Qb=#e zqyw>tQuXPDAj>7y&KLQ>S(V(tmt*GUB-Y)Ef-B_RT}e1ASx=^*$foJg5oS4eQhv5e!6n1!g08+OVBzX~W90?_#^OAE>>@g)J3I69Pt$EFQF| zXM7|bb7f}Kx7Db@k$O;Hto&fwuNo|G!-t3Ps)k>ANQ!doOHYY8=pun>H}(XI^@==_ zij}8iP}FI%b>tOuUWy|l!F1GY{Ao#4x{Qtu9*;I~`_H*+X2*kR#?HU=yHiPYG~@ca z`DVPjQTMbA)MS4**#A=Cq{~FMPwsnVL@!yQMLUcyQameVL6QcI8>6XutR>=l`G$y-Py-FJcLF(Imah$QIE=Gx0 zulw905K+vk z{|~B2);JycCHii>{N&k)>6iytFY&OfGnkLZSSsZ=-bD!hERU%*RnPdIM~~~AJv!}* z!+N`cCA(mwAw-}jyK(sMUTV0MRFE${6Y?+yAcCLTa=O;=jo=NQu$8NnQ8A@N53^uV2Bmg`n!FAqyjt8T zgTRYN4E%$;0|ynA^z)Sr18Cc(&Do(}Q!<%LYmrW7m61LUWW#Se7rOwBgO6suG=rTn z+li!vJW+cyyd!*>P=)N^7^Hzz%zl^e6Bq+F@?o-24WabGYosw~9WtKh=j}{r;*553p!qe_zT~#-;%j zD^dLno_nSsEd-7`}lUq)v zjzV~tMM)#fUdNMNSHK|4T?%nwh?g+_niVH&CvAqxh~%d#{|9ovWntG#vtPRZ;)yuB z*rnZGd(nHvc%ui4gDGi_^G9-Rc7Hmz>u+oRzaay`iNuTmbBlvzkz!E|m{nc+Gv4)3 ziucshqbLM>>^Q$+X1^YMX)8+I&NeIinRA@X*TA>W=s)BAHg5q4iSB{B@O#=&j-Ude z&WiS&5#0?0Pq6NFce>Bgfj)Rw^-QMd$iF|p*D%fdw{l)L&Hq~W!*@>H zBCfa_s%Wzv*^|As9mW8FLc`97EPj+r3bf5@7xTqEx1XaBAZCPh=`fQ!)7`xSbRRc7 z2QA)7O2QRiLv3l2*RdyIW*TE#TGc>_5U$7n9kK4Y2#)UQ(RuRFXis z%o@Uu2(%lk{7rA%UqFlufeXh`H9hlG{r4jO+dD$2EQR=MqcDz6kmafnXZ&zSoxl3T zS~<9tSNpBF)X!9z;`^j+sPzNa-Zr&=qa`dT)puB5RdMej#XW;{7h>V#kRXzKBCng+ znCbE<)o^eO%ZjJciYcb81H48qGTISrA;1;s8g)VXV(&$Sa$MIg(mjy`X@pR2tUiE3 z+10m8=|y{aOZh%gd908+Ur7EM7{Jsg`$PKO4Otv&0c?o8G-w%<7cK{OTY{AJZisaP zUZvJG8f;RjFCw3E-REpvZv_w4dCz+owJ;^udBEIQXAZd#?yPzri}Qew8>L{N%5*N_=d=wL<$U zD3v^NWte^4y38~pT*r0cx%9^g(r-DxdY3hCRezTU4lSLQJ#tnxAh5^L^R4(*Jg+uZ z&r@*)iL=2lr}jUjT7)g|PR5nNlc`W&h2NJEJv|_M_RkYCl5*~rtm$jK5xZ9Y6ydQN z-3hfEvt7EB>o`C0TcQ^bF}!#Eq<1`2tH8`(^?w+<3K~s(<_Vj*p+L;w3SUZ8GZ`e-(qhkUwRDKgI|;5uhT+rB*4roTe0uo@L%m z{;~d$yLE`~Fz)?J1K*L%E8L!2O)p(@muzKXViFd#L)2OJ%-s?_(rL%}=Dmv#_h3y6 zqsM8LFTv%|JubmhnqM4zbrQ2BCk%v2?SN>#jI(#v+ons&1e9_lM_L2Gn$G=yPL#^= z>x;~Cbf?L`2taQx6hjEN5xiBXvS_g{QcexJ$ys5yas#lRwKv?uT;QcBD}35`%iC4U zT%q-@q}ehH(=e|}bpnP*(lIfdgQh_^c!9x;w$3TnIV(q5y=cVOAyI})t9z-`z{)wI zq9_d!w|&a7^wj2(Zd*%GsH(|Jfc2YojPo+dr?nu3yb(R4X_fp?)5@NPD*PgFe4uwktTRa(vHTn+Kvb23Pr6^>xB9+LGfYk zHk?#3g_mW^(lAxgJu%W&l8P%ZO|E7~yh#sto9*_>oP7`r-#c>GWs=<#-WVU^XHPb7 zH6sgmSG*#5cJO%!v6o!O$*afE&2xgjvEDK#7PV}W7l{*uLy~~+6Gz+=tuN7I2#1g* zk}w2sSdz~}hJrv>Gp}d4-S1O#ay-a|jy<@kF-&_}6ho*hCG?fEotUOcb1Jcmn?u(l zhuI%~+pYcQpf2!!>&*&$510plWxAP29~+sTi?jZH$`uibz@ytyo0iT ztf8)cmh8qkmw}^*=B542w@%Ad4s!#Xtsu?sZ%|a>i9i=;iE~`K%EZ}>Bi#f zgaO;jTz>2Hbp`SID1s#M)jq^CLJXhFUAqenbZ4DQCv_9Rpz8Pu^>`E?B4YPLE6Gm7 zJA6RbSBd30@Ww^da49{%5fvC7o|IenYi*QkAr9xk@8p&v`H1!P5)PXiM@vI4qNIx2 z0A8#mj`V)`*jTo%Q6ZT9bWOry67+6=?Q5>#(Rz$CMG~^mU@-CSm7Jt3N8GoD5^K@c zgelS`Oh_c0pfKH0@eml++AnXN<+u~K{`J!+iVv8lopFE`gWzVvWR-nc56)J^$DcRYej?moCzDSc9z>eG*sWJFF2*X>vn(`sa>7* z>NV#U<3(u-L)4hHnQfFOcuhqJ_esR1s!r>9or0<{y4Qh8d~psxu~NWmgPdkRd1!aE zSUk0)L9Ut*&I)RmDM?>-0i4#r^{gBk#!G{r>OG0P#_V!l3VplU@t>P*^@hn$1xO*Q zyz_qQdV+%1by{DEYy#44s3KPAhE~M2;eFL@?WlBf0aX812O1CeO|CPoVY6ZKC^>hY zKiP-#9> z0y)}mKIrw&CabF+Q}V82+hyz&Xe*Ex>C4k_@tq7V3|8#*9nR$2XeZQRI3|TObe3NoZxVG~DEawZSrTiQ^JT2_GYV-#j zoXh!dG>Wa+kb@2tuBmfhqWDH%P4rJKSteIeP$C0y&rqhwvUw^7?KfZk^PtJ!1{Y>S zMsEGK#C;#+rx<-Y&X?8KZnyPvk@;_wVu~R<3%ej6mQe;Kg`daVnT(bq^_!4Rf;?Z0 z(y4YGX>)_Khu9aX^xtL`>kbbt!CT;UD&kAj#!a*tC?Cyitd=fv*xVS|lw2WVf;8zv zgCDb&%gv#IX2ST%XpBkP;|l-bTvk}QNzYrkPk;-x+uB`y;4Ry=a%!QguBg(1dW?Zrk+P_KXRtu#+qXPK-EYN4BmpSfdXTjN>}o#5S@ri;o{z8U<1FQP zphk26gnh1}w!o1=?jB?P6nnI>%sEW|g@#Se?jp0HllCSN`xjG^_N`AxLYL-y%VpW& zEdsMS7L|L~)j-#fv;JmCV)1pFHluB5^W5yxy3YWK=qGBopLT>Dg1_Cx z#%T=Z-e{TzdmT?kjoEyQ}cuEDlb|W(bv`hPGqtOoDL!A zg4wAqLn7>ftp*8Necx^3XYvF%S(!!aGFG|f3)1YU$nFOh)4(-IoD z>D1O}s^pO;xNHYAd?R1YD0&Lw*{1JPlLUG6<%%E9I54|BKSpU>ldC~QQ$PyoF7$-- z&xqBaQ9AU_=m=(3Cww!lLaEik1!h3u`;V7CesPs$nzj)3wRG20bWRDD&xtgF=VyMh zEoD4*x5Jd}iKYd6?+y~wotWJA2m?0Ky!cT=dZ6=b2QT}oM8d-a^V2wuoKDI_R(@6Y z_=iC&92rs+X?E|&buqTg7qZ84r56)y6!d#c-R7&6ejA|dk}~lD0FvBLZdz>qj2w+> zJ_|4f4?af1gWAYnt12<`$ynWbhf=4rq%wO4zb3W+DD796RHs{impO06VO%m6+ zQMC3N5+!et^o}@3cT`(eYs-~zNZe5)xar-XX}UB)MDr@yn)3+e_*976yiq4D_W(%| zfEBWj%$;SR%{4GAI*=mVYNYG-U9AkBWRlv*j^u;3BL%0m?JC!joyjv|FMA-UPjfg& zaq={CdC-*DD@CfeIj{qhg}hf|upNOZHHj82KXQGpi{$)=YmH{$p|ydwxd&J;C5lG> zc9Ym1Oi4YMP4#Y9s)-obXKI5Gne}EyPvg&ji)XRPUfkUQsg1LkG|>-o(M}N=x-E)p ziPFcHnk`g&W;nRrjF-u*v14D%$>Gs7n(;DzL$sY)>$w*RG*`s}1{LEDrwC-=Vwql9z&eM+LAlT}+L&8FEuyHqfLi0o2Nt--Hgs8wQErgaVf)FcY zQf(Yn7Av6NinPDFuwIyh9m5@og;-4P!We&4+s9v_NuADw<@z5NtE` zcmiEAa3Pu%n4*+oj}2B}j)3Wo+rcL&q=-K*5lC(Ytk{{xHlr>(O@Y<~dB~k>t1$Bc z8)AI*vWQBI1)#o|qP^*Q!Qo+({@yBf*# zFsR9BOmOjLEdTh2tL(SPc}sC!jI#u(K%X}=&Hu8KT9N%`sK)cj(zRyMY=fFZSRaj@Q>Oh2<3 z<0iVKH05?R6}2&>>s;=?qeUDZEOlvoSx*a~1AFLidI8rUrap^|)W<>6d2hB@!(l`f z6YZ&ThRCfUUYSbse7#U_>St-DlltsS%BekyA@!UU=8V^4wfCa0hN{LXrnbOe_b7oC zO```=3Zk|a>badrt0Arf!?*e7_`$&zTCEZ5*iwl7D*(NgGD1hG!5bQF?5V3R>~D&i zxuH7Oei^SBUkRH$Tc^F%T9^3slh9{?V^_^feRMx0InzGk#S4Kl3xL2w1~AN@2`dw@ z5`=D_9i)hi4aafP9H}eN;WVw@I5JxS}iIoX+T5H=aQ!CgVq2{fdbS4vqe?n;&iJXu*d7P2V}aznl%$Zvz)zLM>B=K473sH0^YCI} z*qPz6dX2Z$Tbcf_>eiG*p(+Mas?Kv@ggaE zH5yV8Y|xZcs;mMy%vg%P?VShYKMyeL7hiwsV{&InQ$=0m;JI>+HoDn@5LWIbsWtH% z*QS)t?3}X1c-4d9;>4RzTA)k6#Alu&+d_M$nj;J9AyXL3{19<6M}Cx*;cKC?{n7_V zeDvdezo5W;i8t0_eVXa}H3a5(3#&A zGhec^MzL?8bywqPeA=J(scw^UbCA-C2&tDV`CHO*5qK2+5F1W+4r5>4HqSY~<4AJz zk0>ap#(W~0anORW1FhXoEfwR)(Hs_19vl7`Hxd-=;dE|oY zr-m~kwMMybu*kdDyNZj`8#TBHx_o~!2L}tOBp=0iO)= zTgGiTZZR^+HT#b`Rx_ATI1>~Io?BSfucSFN@;VaCesI8NiacKjM0+KpE-N-W}s!e=TpMi zHEW||Z_`}p=;$a!TBzMp$45!1^z?x!aj*B8SVnppQqm}gPt-|f*2F9luq$dWcB;94 z+K~G;B&TWrV)nCM@n(#3kA zH9pOv+Uh+$Jhq9jkF&nVB9pv{7F1DiIDZ{1B%KV4eyJgo>^T9uxJ;v67$M=C(e>>R zL=qymFrhU?xsp%jIWV@q=Sj0v!htLPn`f?I($d;3$L8vhTPD@aieJjMS@YLjR zHVcR=eUW0OzKLL$K*T85<0i%h6s17{N?Xzi49kHhUZ zVwy}O-VG6YaT7lQFZtyVX_!h#l%#?6HQMtA^f)xj>N5DpvdRt#fH+RcK*xjl*PZF5 zZU+PdNQF(~+m;=+RvUwM%MQ=NKhx`5y)12P4aHY*$^=ZP9?W)d8uK*%3{n^Uyfwwl zDLF%obB~s44baHGNT|3i6tk!oC*HcJytv!KJiKWPEuHh%FtldXTZR=%nqua%yYzm@ zS({BOO5yHsl?xoSL9)JvTh%4DJF8X-cxpW^VUSk2qGtr}1T3lO?=KDv7ICvy@Kd7t zP;RCP1Tv@L0&1zY4n2J1tyd&_*=aLAYzNZ2?*`W~L3*R8I2azm-mKf;Ip3S@J|vRT zc7&QQ#{KnJAqXxiR%E#DYTI=TWuQwjsTX%ju9Ky@4Ja{)5#ROR{Vo$Ap=0kzOO~FZ zI^brcdA%;X5iuO$QHmFB)y|lc<0KBaw}zl2@dy1`WY?y-4^2rRE9~3 zy8HU$Tayq~(D%OX&OPHqgeX<2;og;p^x8pj-QDD9W?0kH?4S;bspEL`T2 z7Y!N|V~!9Hkac|6E0idvwiNaxSUR0?7=L~%2KaTsZP_3Q~^^C z_EIP<<12|(GPr0Q*hP>`grURhkJ*d+SM577NBhSHE}q|xnBpy)E22$qWm(8(Wlvht zKvyvQp$YZlh$nkwkv~Z|IoJmpGy68?bE9(Jnb}9Xp<=Mut{#5W2&hmW6ZO2ZhJA8Z zGO!KvsHOLfI4E_;hb~SYE+!%Jz61n;_G0mYW?dQsnG5GY0_6-sotz&PxSP<`E5+mn z$Y7Be(QIQKo~Is!0Jc#7kUeG%GSo&8F*#ZeKLk@YD6R`vMs%a@UdCmde~$V*PM3cV z&&jBD5c=nBrCsAN;k^ZW)*i!W7hH)!A?+rMcuiN#xdeHdd2)aWl`bO-Kw|q={!Y$rk^&X!7Bz|So znd(QQsH6KE9)B61Rj117@rySs-*jNEJ$_ukGUVhdjwfK>2y(20WH+9Gcn2*zC{h_# z00|RC>4Ec0rvr0|>Mnf1Fz4X!uKT1crs>K<PjFZ@h- z*8;BG$j?Tx!FG1u4`z9(x#cANj^0n2)jJToUK+Q|V^v?kaxlqv(Mx(=O;Vm#(;V%F zW^bQScSo7}+hnyFJ*tx7Ndq`4cUM#vbznvpuYy#;czy|vSnmoH<9}T}pQ;w1GeTf) z#=h3Owr1__G0+HFxo@EhtJvz$R*dHFtoOYc|GAzG1>_`hB^i9LB^BlgS=#sTA`dV9Rt0BnfpBF3)Z`sn7h7N z%@IE;226eQY>G$(Hz5VLbUgD4znyidb7(7X{Q)ke%xmTI z(d)8ElYj_ND>|&1BC_zDzUMVfo}x6lsO}`G^tRCqRSRi%>N5Ix6|nAVjeLUI?uAT- zz9PDT`>6utFCkJp6_*c;!r4sZ{A9?VD3D*x*|#_OJION&J0l0ey^4?hVqt@ITLEce zHD(*S+$`ZB!v>>n_!%sD^obU#&XHp-UVvdo_N5{)1gr%hUdN>c4Q1!~<7;hGchGJT z2+w+I-*)7W#=D|AvsA|mpEv2?X_q`C6%b%GLOZ>1Ig%F^m#9M7Af?`h)r-5XH6|3z z=bS|cekR^Z-Fhxle-tOM+Q5)ALo1i}|Ex9`d7Z@i2Yh8dE|Xu5qU@0= z4|T-H)=Zm6BO9*BPBf_XBMHNpeFG?bSU=zrYOfgT!}OI;=Ttvfw63h{j#|CdTzkq% z^GB3o26LdveZ8I5Y^=ro%Yl0~H>1IT6N7O}r0dl4ni3ib*!>&JJrNjcu=rFJY{R*L zhZvGSsae%Z2+UD?4v`|0G=8NXb#SmoB!U6+ak_Arm|_Lt^V$SD>zu^=6w7W#*v(4f zj|9A8lT36b-dzZi9CWH)C)0aLr-IdRvn8F7RR?mEhr*rP{Mfx4u;$tf56N1eZQ3A>3J6z4eO|BLDJr8T%_X(?48a z>ugtlf7*+)8`!vc@IHRtlORV@Nh$mrd}CFfT> zpPHW2f~J5RL015!LgqA3EOjyO{;WPC*2H%uT)7-$540o-%6=_9v#37V_*U7A{h6`E8RIi$=h;A$7kxJl{ z1l;VGt821l!Ottz7pj?1(atsHtJ}iz5R`#inZAT1G30fkn#WM`X$8$i&FlVON;Bla zTtR5-tV>jWYZ+;WF)|IZJbBPL>5opk4qaR^h9{gZWjF8`;W!xl{^ zA0p?s(oAROs5&-2j~065y*811C8I~>5hr>2zTp?yz3n(AzllMyJd{WOg+Bz7lIPC3%8iFxG!HR30`dS5)NvZWr! z+Z#CX&z(|fhmyM9R_2jH|9rk2O#1476?0=dVk+ulICtvk;}9&zr3AaX{f6=?+OYcO*4}!4 z_Tduv@?IrQJjA`g?f#O%^P4r!c)MDD`kuWn?syQlPjs^#gE0tI8FzYHJ&*@g=-Jx( z5X(vZN5uSXvQaQ5D;oL?r#m(D$2tZm30~;I*6i41(VxFS39WY!5ybq^^XE1keDt~} zz+vvTI_-ccJkCIh{HsvRXeDF~@U54o331NQ#3P$5xWjN10g^3YkM0Z$zsSU<3vY>x zx9~W&x9G@kNqZ_Q)WDw>Z|79g^t{1u;(7SD{i4Il_a6&5Ui-^FzrPHd2TzSw8ShB5 zECqIgb6#FqowhJuszoWoFvY^43xA6~8 zz@Lcwk?JYNkWx?6Sta2i&K-B;0XTEH{Of!3>)hoWCAx2~&AnKC8C{@8he&*N?=!?9 z=R4;6O>hvag}}j}m%-f!O%{hAYu8QI_*$P5mMPWkCZltvi+tw?`^W~939}_3<|xh z;|mieG4;MMBwF@P*8F+SA>C=4k54*j(i_^Y`1gNWt&j?DFU*kVWB#Sv�`5q-ZW< zJbFV5=+nNuvc{i?fm6arzee?W7ac@;e*Uq|KxCc1+c+1h2a>YqSV{u|T+;GpGxgU~ zkFo|KRqe4B&$OqWCFL86KqIy^I$TxVI`-HE`-|Cb&jTHC68gSw$1yzpv(R|Fopxy- z=+d(lTe511)7q%arl0BMOgBHym#<57J1d{2^o(*U^WFOZ7LUQF&GF<4;-b@?PraUK zziHVeb!gjUKVZ8^zN2X*44kKUMp6^7RdSM6H5q6HSudTm>VF6}eelGij#dHJ?ujvU+73)LRl|D0h}GL0ugxYElGSdB&FKU=kZH}nEOI(sq*M_7zA%d@#C)`x3_`Q<=vM10&}hQTXFZ>iB% z$)k)qhZTTe^KOoRgRf-rOT91GJ$e(9V&`i%Smsgn0^eZn8YWdK>5JXJU;SAr9=ovo zMy1U-6)~8)t!(;-=d-feyS87RrxjcrI$rUqH&ILJ42}HUj?{c<1*sEzuRP=k0;zxw zuxUS5XD}C5jQ!+64!Hnc>;3V4AT6rQ6g(XhO`#)-dR*hd97WrR;NYNY3h@fdlK6t- zV*CGcKfRX=@VME>2HviE=3fkN_Hi^dCWUTR$&!lp#iJ0X^wN{C$b%|Ni;d2#=%+Tq zJM>4PJ4@#gvi#cd>^GEOhgU)&X0Q)j8Yg6h3c8Ox!wDUR)BW@XnfxzsHnCcsGp@#P zXEN=YdfW5n0k@)&R-hY8ML^Uw|3I^Y+Yg_Xj$yr~p7Jgfr|{O%jMzaH`$37`h-7uw zrdVEYd%1%_#5x2X!RCPv$ISWu(Ao%Jt5LQ#|M8C{-hLtfdEmc;T5PIVZr@tATn^Q| zjR@S+{lyH=OhcP2Rv{@LV?5a@%FoPZPlZfDcjFkJ!}?uf7PHoe#@5V~rs?bk9nwH& zUiI1^y#%81n!>^$m`)PnPNa_>p6XL69xcfI8P{iy6L($X_LYpyQ()+2M3ei|@8mbu zCh7B)@ic0|32?0BNG|d zY&9Akk4L^b6aGe&7rcq6)ScjQKR0Z^Hj3*0B7v@|h`7(`Oa~B*2V}KK>>iY{oCw56 zg70@myp;!vK~_o=?NC&3XknK_w?Q14jKJg2vk_%>ZDU02=USWwb7b%9*foz+o0ESzUPr6=q-@H`|}5=e^sD5f~n96Y_VUHcgqFyp1k-;~A_ zL5cIImpmTWyi3*xE^Lwgy`Gy}Z|B)@%QAea+Lat;G?;u=`6!s@gR}1Y=GfCUl|0vS zGnXO-HQ%ZFuPDLCHJasmX06G@pvmI>{SEPG8B#Pe@MSqgLlX@qTv)+{l>533k$E6 zf5ZhdKOL!B2RBXDo|ZcZ{Z4b$kVeJUQpA6G;VfI&>s6MQ3Cp6>P=2nZCfdlwuf`P8 z^kn4aSMrZb1OS|QHXW>{$N!+elNX$&e3%Tl#|lX^)WnD#PVv&KNcVttfU?-jqtY2I z(w?uYmenF`T&h#XD4UU?Lw@vI3w=xaYhbWAjg6cAp2YWivG7NVG0O|v+$55S2xe(a zHVK!(E?pmG`Ncgs7M|f^^08s|p(=J(mZ1aC+s=#-Qx6t_jbsTSDp${f+gg_i`_xsy z`g9^9idaHfg(AhiIr$?U0dqRmrKT3B*SQF>knAh#1#(00r*hs^W!yV)DJfOstX>H0 z8=RO@`b_&Di5$g(ETd^%KS7&E5fZ;D1tk(<&D}_t`O&MUQ0b)5m&_~11na8qoi;9wc_j_FO2qb;^aRHFzeP&eFN-Ff9@0CEMEAeKZwcuTzHSh%`-^y~G|W z*Mzq_@CMS+QCD_Y4)WxuI6$07D9{H7A1+{;8jOo;53YLuXYWp_0Tjk$)zWhaYB`8> z+PHlt^ch^TsVkc@u(0(RwugibHLK6|!NRbVNe$LhbVz@f#m6Bv@@eYc=gt1$Su|M-2Ykz>+|48x|kS3YTfEV5lgAK<@4bC;+Fc>qJWO3 z9^;%0cblpo<{)bU%gJ=8oyeti@Nj{?j0opJNQNObZ|lrR$AOqA5Z1;4x&|6 zmQt6#3NVAR!~j$t^f+)Jz*e%H^ty^hL`>5Bj$#b7=?8|dWUx;M;@l?*8Js_*N?<8m zr@CuBOtVW2h+oDjJWX}YB?Q(F%d5I44!)oPGJoe3xk4(9G06%#me7gv&l+hNFX(~! z<%LrZC<{2SB*od*<{rYmOs=eXa~iX}`tt@(DBcze#wpybn^m%6x|0Rvm&lFz%L9_w z#o*D}5#{=%IG^HvobMTFWe%#3i}P)*IEzJne_O%f<p~RdsK0tMvhFn;d zZM}aKG~IM={d8)iMd$*wcCsGgxS!*1_wg^zCFECCm{PxJB3XkJZ_dLUFRY9VuUcb!HcAPVLeWeEG6bRYFr{e%g8-(Vf7W0glSy6UBeYM6O3iH8fR zsbo#ZmsX6FWo27$jHL!9Hw)R8Ab(qy z@fX7`Eu|b$Q$bSyFdv2(%*Ppk-pvBz3ln-N^O7Lnf4A$Jm+)EMc2@dpV<1%N>?X<` z+!IlAiuZZ)>|F8o{Y8>sknFZCG)J*;E>S*+iI10E@}vGAiTR^>nOgdHG%s;jH4=)guu_8$HWY zI<3!YAX8}3VK5;D%Oq~79GMJr&QN7`V<>Wz37)6*q~iOr@P~GgknpT5-1`yE07j9- zN9%XO*bOmznXndVwdGDE`wYyv#B1LDSru+rDGBmdf|Zhck;#mT{s(0%=YiIQ`Rr}8 z-s0~QNbXOdR;7fBD+(TBYD?!T*fcM!uKDR`_0YA;8lIc`dz(Q&zNw~9w;m1(08go( zP&;_yO@~?CnrjvOa57R*{2j#1mItxBsn~k*4r0vco7Nguq2wX+ZOCRIaxZ0vqULb^ z1+R#Qrbb1i-A6p(UXlKq8Z$EIShwl&pvvhBJ>)g>Z2E?w59d}(l*cn<`ieJ2dP+lb zCf)u{6mmQ^0Bl=C(+^&4;cW{o&SrjWF(?O%`f)@CE%q90-;o-SIZy$o3T9oX^Yv^(L55)KSy!h7CGTX44<;Bx28UlN+7!emulN!iCd6l-&73wC# z6ZzgCKjz&f|NhWyLNdCMNCPo1k!|Gn0B98JL)g7m0MU&Tfku1V)n`x9IYuyR2vEdV z8hVEtpd$q=zWLlt^D3bWnWp684I8P}0<9g#KZb2Q}A6zJlJj_En|HLEAF|MHCQg(5BS&*TX;Mb>*J1ri)=^Nwo-I7an z@J0R~+{0`)X`vyV@O#A@8_C(iwGb{Sw-nVJs&jP^Kbp;|6DeBfnI2qoXPV2QqLMXV%-qi=!EDbyBYV?JosN}<# za)9hin?_6`J3A5Hotd9?tv!#b$Tij?7?(ABz(^&Bd(%v2+Z$+A|2V;rPED5kVPx~J2QwgoCbO{=(qHI0 zdz@-^v?W_isMA7`P`F!bvE7rWGsf7FeG#KPsYHWO+1>EEDr;=@Z^aVyMtV*?aD+viHx=dY*dpt zJeC*lkCDjgma~DF<;~%`Na1d}LF$2oQSt?# z`bM4wY7y)N0&FQ1=w#(NzlX^BFXV)vM}FrO&mH`CQBgh8zLE;lKlfS_mQ{_Z$KZuq{>kWk!T@$3Sg~US^l+Lu}P14Ije=S6l z$1!swsnIp~0y1^0xISEcoa~3Z%r7(IfGDJQ!kCH~&37I8_bh1HH_y`2vtH(`VvSK` zzSivc!~VFW$Unsw)%Cd2#r?Qk&~NRQVxvH0B?}PNAe|HKOJcYquy}l#)ps*sspsN! zuvjT)88nZ^@LSZB;@ z06d$fh_Yz6L9#Hjb_zy{c28XMAbcN_6a%Hl>3KdX)3In?efY&$G6hiG{fo|P(%0Jq z+}G>}4vH2}qy=Rz^^VlI#;dtDb~eHu5(|9M@#`{tFt!C?G_M(J|1yrik&{&?*10zjM0vZ zI`nz<$M`rkjj+a|20IgrKp>}=hk(_r+f$xvpbvXPgiRXoL?Nr?-9U3~T2~U$n1@mt zRIR<@Vc;1?a*0V`JX0xA#HF%wb5aeW=Go>iEtm)UC{~F zZ9^dkR6T${hFP_Guty>#+tl{dAWhqGR8$%R4v$YzP@5~Lkix!|x)WN77mBOQ@7x^_T_WWIn z!Q{D_3$czAElz#8I-T&cGomlPCi@v|u2(b@aUH)8BgQ1a107I1uE1Yq9)VlL$mVH8 z)62@^B3)^B1elZuQ5_*~bWCnU(Z2sEnEj8i^W!WxNV^T6l?Dh|WmCTLp9iaq8cu+e z1wktV!kDjM?tD@{bZfuI!Y;sNH^bunR?NHjY~UgNCncPmBj?i9j{ph(d*olH|bHwLtFm{d5M1e6D_6R=Py00K+@I7^^Dj=H6Lw_ zXd~{c7`}28j{ao-V;>|QWE14TlY{#BleLs%)#Ze!5MjP1PD|t`fx>1-R9D%!e6bqi z$r_=b&XV96Z`h`_h98)=xVCjXnV1mC(@)}6O6k7M>_*4cq|Yvh#QJ=WkTAuGiA7)w z9ta$E`e9>_#hcGp)QF6La3&i{+6Z!iWb1wNaLhQz?sus<6K!}M$h9{W!z zBx}_*7Hwm#$MBr-8j#0ReW5t@x)$BcCBcj>ve{9vi5sF)Vl3}VBzxfuQ6ayp*K1mD ztdA-5N`G(L2pj`@@U)3{%Il2rb9`-$A8J4L=oi7<v!@TF^>U( zONAPbA*Q^QB^D9(%#ZZKWmI7`Z#rWhA5Mp$evvafDEsVvrrX=D>@QRGz{>8zpF+PH z{Y)SguEah|-KhOYb7x+tfGsNUfXes$y=JOWmKl?P<*^I0t;1q++2*cP#~}&FJvZ$~ zU+FV1`O8+^z8swxzh!mk{#N_yv5simFQJTsD^2D%Ep*();H5LhjaIWW^rP{qcLqE# zRn4wi0k!h+H4(a-iKwnd=Gr+J$(t8osxT)C?i7Xw3kt{<+-9TuQMU{;QsLfQ!wg$u zqqj{jR3gV&pmlTdlIFz8r_uR?b)l$KvvLZ`-peGK0bbVc^rSZt{81cA zFn78n`lUJ8^#f6SJz_GFsvkBYS{5$uYMW3$o-|J1JlTt~(q}{MDlMH1fF?8z2>RqdLCXJ>H>D&o@<6vO2nkcyjU z!PkpiqVV8%blX;nnQKf#SK4a*daSCeo)HAti}C50XDu9@ZP*>{MCLVOE2Vpf|Ki(B z*u3R}8qG4pJA~32rh185V4YH=3bS6d_4lk`S8d%X7?)kLEFzTeA~d(tR5FC%mSD;Dm2Iz;m5gf;=4^k>CRNLs&wxSMKJNV^>M_MC`uU`G=`1qLDv;BDTfLD^K*Rvd{zQhUZ zxLBX_V=qJgQ5oQ-PW^q(cx&P{;ax+A(l2A`!8xh)Fi#l!+Q(_q!7sBsfH8A{_j4`M z>YtUW{T?y{AGIaczed)KAZ3|%)Ffz2wLd@nb|CK57ABoYhkX)Nm;Q8D%qT1xZhQ@k zS_opHeX7Y06|GZ(t{?R50omAXj}QW*ex|&M8NOM!erjJLq=jz1skkDV<6`J)sO#p4Ke^aU>LD^;A;okxJ z$Mc=|{rMXuIh9p!dJdw#{ntqcMWrWz1FADv#%jt93gJF@pf|=Gb7WwUp#u9%Uw9K| zHMl_5=_g+vJbWw37ug}0VhcSBcZ4Yj*4_37aF?1@;07G}Gl$10>H$!)qN=w7gEiIiQ?Qig$ z^SkwF>S>d5ijQowh#IWO9D+Wuj8e-vXnpa$$7uH zinL`alV`|fWoeUQ(z?tE0gi4l)$*2_n-STqlVtxsN7(%p*)1l0&ZRoX8>i9*lyGJ> zLER%`+p|dZl$KG$VKw3Kq~sx8i`L~ZCH1^5l6<-ehsHRhrTTjzhQjINBCF86TA|{7 z#wXEvT*p%5F@}<3v2tB>xuj9pj&9TjND*3*?XWJi)-t2snIOtf*P{fT)v5OT()lbk zVpJTvIaxIW+3!5Z62Bbo{FJOlP0sV9Z+FvZUJHAvMldnvg-dOS}@to0mr2>HH? zm~Yb?0}xb1%DO7yLqbt6oe zLH1y^A?gg7=+){`JnoOE@bx*hBJ`nG1VQlq&*AP-^F?4xc#3nGrBuwUul$dqZO(5h zXtW01H`32FX6%(;5i9RRRv-?_W~{Y zuT*Nd1_mzmc=QGs)_M}2*SN?~!5syHWm*}m>6c>e+x8Cg%nYo|k=K;)a@@&RrhL_& zdlt0x>3a5zYd5s`H~=`@r>LVOSfT!9@>Z3tuv4_+M#6y!jUVRA%Bnu!{zEbg()EI3 z{=$g6s@LHOW8CPdWDG0X0(fmUMjLA7o4g7$QM|~8;zC8@`WK28Pz+0*9Hma`c3_-Ra^Q%F>U_N_Aaol-UG6NWI`l$6|VWTcX^RSlpDbq4x{LssV~_ zj;Lsq4q!X{DnH*rtLp%Y6&j&?l*H8)LTBF4ucTDk8Q&82MMd+OMLz9fe810=Dse#? z;_1ve;^)WC%Xa;vNEJY@g}{BBryfEa)omEd?LU9k2YzOkd8i2bgjCW=v*rmw$7D@b zXy%%;PINX?0o;Evd!v$(sx+8@UvoG{D{M1B`%>tn2;IR0D6mc!oD*IDV+k5VmDbA=z**(t*^3YvW8zflWBG&IjA{iFoz}s^)cWZibs)* zPe44jBGZEgqb*(xnWa+ zl4Mp)Jvc(QiRuu#(Ek}Z4iMLgjc67(nAD!Rm80JV#BuU26G&!{D-7r=wjItCt!dzd z7^A0>Ki~k@fQp_jJIm4cc$&kLgyGY&1)?MP{j3%s(a^nMA+It6_rpkO#At`+iO$750 zRJT;$Lw=!Ti-<0BT;%`9bc!U9hF*@>=6f| zQXHg%F0??RP93F#5+W%0+aE6!L7r5z^f}aA$ZnPEEO!{auuygOdNvTt8c;FI_Fzmz zm&Ld{KUS}ts&UK;6Nko}c1) zxIu9=BBr)PvBVj-x-&7TtptY*~kSO*k#<`v4BAl_Px zDx8P92=|rsPZ~>9uYy^Fm(-vH27`85cO0ZBeVF3e^Sm_d8X40C$aVB&oL z7@_wSQ1{v80pVn7BbLZsp_M>zg4Nr`=EO39v8~T0uNop$P0qpK);jn211zP6fx)0A zO^H)E+ep4ad4{A)wylT$H8=&9IMuSDLCPSw6wJzRsKy)^+X*@2x46Wn7dFLoIZ$CEWnHXw9sazeOz9op|;OVe9R{=Aq}D);yO_Sil;#U z&`rBQi+&jAw=kTYm#4raEple}b;d{UGoeV^3~aksli?AWC0ggn1a+<1PZJ3FD_*W` zwJ*W4>27~EW6o0}f4Ph--V<~LE*;om`Wk+40~-eniYim+#V_S*E_4&zM!e3+-yNfs zw-SkUUo*Bo+ld_`)%5e9_`e%G`7FjY2QAgQ_ZXt%A=;20OM-s_u%Wm#4K5L!2+`>_ z)EF0zIYatGuQ+|@L2qj$*MtCpP~R1NujUJ6i~VvBRuMg;DtWwWkaZQ zD!SNawBxR7j`8r9@w|pOaw!$ihv+(WOMqAQh-PUKL8PcIeSVBN+&1sYeVFMJcP{Cq ztKkg*k#VoCqRylz7id`EyYw%VX=$flD88C&YkR_iaMfQZ>(sgXNW+8*amnUTCLW|8 zvaQ}3(e^63aeSnp-_@g0?fAr}i!vM|1m-j&)Q$!{u|@Pa4t4LQDEsK#_q|khTlYYs zh(02~zVzDsLP%j!V8Gk=xUoycmXB)~G%@s&bioU*`;+-&-**#c>i|TGx~@-hD_QpW zN=rR(mF(p5s6FamEAbh{^NhXOiBnYBU@u>w==?1DsUgO3_)SecIg^GyV0I}$NuC3~ zL2kYsr?2|e()ClkT9zIMs<7#O?T=+K>Dab;$@?^VYE>S~2PuN<{IDl+sM0 zA;4CrJ?!Heo0^T}i4*n%1Nny7G^<%u1}GXXIOKgCuR%rf_@xi!-+p36xjwxkl%*K% zc#5>Op^Vg^$2oKgT{jsMk z93O-_$_SZ55=gHUw4aU7SCv$bxHm-Nr#^#M=~g89gR zMFI-8H|MI$$|nT4l#kWlH!>6i%gyrJX2`szMNh~l(R-jpOvr@CA~ygx1OIdgl~G?W zTU@j&d8RU(!1gnn2ir-}l6)hP3c|>L!YJ9vwfxY6b$Uc_kTcUM`3*eij6UvI=G_n^ z;(qEFRyN8C{w|~`iyt0x}I+s>KOtOV%{)LJTM1ZXviC=#eO#NAbZq%Po zHJO+mZ`3?2Gzt{&O`tRNJG@c~B73qlQAhy)K~;5yI#@I9F49F1j;=DOhEc6wHCG+; zz2S6GfW=_0DfPl_Wq{BKYe**mpxs=_Ljo=%2BmSEBRTS_q**sA6acc2`%9E)&jo0Jr~h~AjaI^BeL ziG#Q)p%5IAj1RLlgge^~eEyNNxY$;Eoi|C4dIu-AhN61dOGeoQ(kMP|?mkO>sr{nI z@?=u^BH9XxM)~2*)4j#bSaTY)CEQjzrG`(v<8}!0PWM>sK1`>}opTqKZAYlvqH$~b zizxId>9h}2li3R#^`PPJ@dZ-r&Qy)W4{^O83m1$MT3mMxm%Fy0W@#4E7&)>aCD$t^ zsfNmS8#VP=VM%ZQ{JE3NI@6<@gQs1hx{&SIRWGut4oiMw9N#Q8(pAwY%-PuMu};+~ zvbtR152{So^w+78NI)P` z8-Z?F;kx)uTE@(cq-cWhyTX?~7}9bx7?MQ_3SdVw^u@d5b5}+Rd?$aOGtk>V>9+ST z7zI#Vbsh(K-0pg4$A>3G!R;*9WhDu|0V6evA54;`r9iqwo7D5# zxiMtk9f%keGO2(&G@9_OQ>T_@wnGvo$GeSpFq$+vp7zu00DCZtU95_>bT?;_TFG2* z;RZzu7`#K)2Z4+gKjh02yYE_<%vH&y>N3<=H|(DYh9`sRA1OKknsPdxawVAw7=pTl zR*IVJY(1!O9U)9cv{Crh>72uLaQe?zK~E+;+>F?iKtsv_8_XixClNFml0(b zv9=0S*uwAOvMq`_fxy&?Tpm{K6cxP{PKhq0jmfQcFACLg{01pJWr9GoVqLV17?rxH z{i`MNb*a`c(31JJKwq8d4U}0s36IJFMJS=9wi0=!;9Rb}Riu4C7=yZrIhh)_BvRDs z&4L!p6?b=SU?Izm+qPT}9a{5fPA(g%e1JBpS4R#tcd3dY@S|6pp}Not)b)5+`>sJ6 z%SxWMkY$lMJ6^w{-D0FRNS>*wxp7%uFpdeElRA*9p1^xi18I~8@+pv{(aQL%#v@7n zMSTn}+_mh9c-S4)V6iT*NF``<5#V<87mQOPba=;yo#JBraWJl0gkO>rlq*RLd9IC_SbXJksOik^g-|)%gF8m_mneL%GLW#@}7ohc{ zwuJf$IB!>c$eyXDA!fxfZi;ai9N!REC=-Ha_aud@UVSFojXWy4YWLXaVGr(G5TGQ{ z)uAdhb+kGnRRsX&?T$t!tU<YT`gx_LyLO^z9XV2&9XW2WZoyqR0dL!@Kw>=*#H z1eS?e+j;AWrpDRRnFs}*&)0qW*2R}nH)Wqo@!A3o+t2OO>GAE`-R+CtizG@tC(hM{ zxzrfxWcSgIZ=0Hl5M#1}Dt0=AE$s=O(xZlW)KTg><>jM(h zI?ATVU$k^`npb48`YMr18H~A6u3Bn0TyjJiMHfECSWJ z{eM@I{Vy#Mk$Es{erx~T&B}l81BuSYcdS|37Lcb*2Ov|;SA>j_I(MgW{wv{956-iF z3Xfww&65=P$N2v0Dl9@MzW61Y@o5~wj_2EOU2KgMxSl=_re56Y@KQ>W8*vdIEjK4o z*E0$E*6wcPEuf5tAz&mTpwaeRk@_hXajiFMb}#a~NjmMtT&NjgzzSWOgWx{-5!Y^R zv7;I;lqf0D)Ie7~+8|-x8$Z@V8eVGPJTU-!4@OtMSHd<@otIp_XH)s z$Fpsk)+=vH>DT?}h)o;?97YiEsHW&A53sfGuX!1N3b;O4jSRwS>W~<*np!|h`R8DNBMCa9$PlS|!NSpuw)%4Vs%0qY`w6Z`{ZUwtXKf>9YS~9`A zX{59L{;M3LSnG%W#-&zLxvRx$KSAsP7z<&XGlDYfnhd(qy7(GyS_%wS@$gU>RJfVU=PBpP8xqt%yiCV+qptD_%@r%2KFNasL$HR}G^!}vZ7H(WLLjc1rUtQ0>cxY%AHo;qY&B=o-IEktU`Yc%xi!2crg}y&*-T{*3M#qfnS`rolOR-*CY)#tu)Qh|bbq1Y6=7UpA zdinI@3S?Hc8erW1ZhAx22PXL3h9wHRxZI>4#}W!80}O$&XX^aaVB5tcm_wtq32A(s z{NvAO5hknQ#txp5;FYr--yg=)HTWy)A-e|SSX)g@gsr*xxVcI}`mg|Nyfa}D zra;wN36#b~?o{Bt^xr%A2A501HuTS3)e5ANVxYD0;jBDu7au00RQ#Q6yD#!P{z7Zo z!3v=&RS@X!^*j;0AICfO)NkY#-Tmv?==2QIz9~gqM#NJyRfxO#sgbMEy1rR%e5u}B zap^F#A>IBfh&WKtWZf{KBJZ}}bCYRni&)QgxKazklWV9JqQPjLEY|n&WLVuVo?Xnngdi2Z3f2bxn?c~u+NIS-{8UxQBq0XzZ8iwUnK5|b*T&X zEjS-C?lRMavd866xAQwqjbbVF15VgNLpRBmRK(W4@39^}59cPQ$*IYCyX+UdFss%j zU@qV#x01?!)U3Q921;5@M|hL`boC?eqq!FDSw{#@+?2B`V|gQA1_`E`2C|=IXO5li zeN3NK6h-StS31S;u{lbgYR1F0!H*5VkYr?5rfW#9$E?}))JA_TA}lp30YRieYW57j zX(ZVcnZK)qpU@u2Z$em6glK&52Uc{rH01LG)q#}JyW-;mo<7II9>%yxMY^*Dn664o zeC9X~nL|B3L3kxA+Yan`A}vp{O<_9U@)rGk$(TH*p7Q{f$|T>aEr2TMwc~1J*{X30 z4YC5R6Zd0)e)du6Gl=9}{eI%?6^EKKwx(@+nR#fpTviFRojlnIKd^)ko#JSaXh&*4 zv1PR9>A5L!mgL*RySb=^mX0=(5ed>5Z_+**A#&gUAj849c*Xh0$gr4IRDRJS#xL*g7ntHP)nJbeh|jI>&($X3?BXhcH&L_PQG3??m>s$H3%N?*(I> zmL1MWjdv{C!(kXxV;TE_4YT+idw|vBeVY_bOjWqFzJ68vtGV7}T5$V2HKwkHw3;Fo zFN*BvR@#fc`$lzHUo#Y0l~nj1YUqy#lB(t@Ht4OaUEgMcDfNs-(~oK7hEuL{{`Wy9 zwGI@fWL*F!Kyz>(Y=VnP4@8bS(VXgky&>ZIM+U`$DEuaF48y;n#e6}HGJ8Yzdt>w^ zo2hSD(myizCjUb`<8`ELD~O}?slc?eMG=kM6!&oijs9>SPlS^gBZ7ATIsE?y%3iCQ zV^eDdi6++w)m6UTDf%!|z3~f$@qZpR1K|B}WwkuLGp)=VZ&%si+c9p`D}>-x%x;dK zvr3~u49QsYC%#~Qn zbo>4{iCH4e#i(uskjFT3@3B#1n~(ViW{H@1iWTV>1ByZkI176zvpdx1t)_ZE8={m!-M5$)a_)Lttp|f3qVo?Y;5uZdLtp^rs zF#MBh>kCOMgCWw#lDh&h9ql5`*r;wgrwg!cp(PafO5_a_3o>+T1+Eyv`Eg0T|JU*!5989SU-Oj;G&a&lD@r7P*{B2O1@uH~JI3;n?I7~%b-Mf-Aep8(T$8( zF=o*0HO!IaETI%ctN!_LaBqj8O4v{G&l_oau5tSJ8#&0!xB-AbnOIqp4K%ojd%S5X z7@(mRDT=-BPM+Vm>38@%exXQPcHU~+G$yU!u&fnAU__M)@h^Q-39skV2H{V<45o)k1Z(gR+0ePZ~G(Qu>k3x)i2Srly{@FNd@Px}F(==b(?JZy(>7CY&} zZv=!|VKmt+3FN)>*H@N)_WAZ*nJ|^8IU6RqoA`UB`3a&AYQ@T6i7IMs9toMzjxC&A zE8pX*rE6Mi%Ch4k-`iDni;k%1JV{MH4%iF>g&gU*vz-hTOc8%P%|A09*NwujkdNfioU;Vu68a-GK}oPJ z6H;EGrx_cOrb(I^3?DesCM6L-3aNL{HEqmx^Mj&|2diUz@CLK8`J~9VzWF0v)B621 zS*GdYb+{LoyRRYWNwX~Uj4fx4K)l&i6#u;MWg_RH75I%Ns3#SSaAJCeSsTWwh3CON z_#oMQ7#k>hKcXR-wQ2JaOP&s-OF{Yb2VdR#ck?RcSnU4y7nM?3FFdJVA9z)KZl^k6 zkRfz`ak%W(JBDPJO1)(3T#aEXQ4yVVZcj}S_j&JCNV=9z(b0h$Q#EE4?a@E=TG}cN zBCL7~O<8~aqvxW|F5Yj7a%1L`#xE3ozZCNets95`v$C#0z>j8rk&xgyS|Y-S?C*eD zm496Xca5uNCNa|Iec6b^!td>;{oOk(nWg6OnK%n!C(PTCM@COseqN>->1S$U%`by> zTJTB$Vr6A%nl`p#J>b|e{Q&6kUZUXFG~gv;z*TR=`J!g4be7X9MXt~>*m8g`L$*m)DGUr8a>Ww(>+Lfqo z!RszOzk9rJd)tovd^%=+O|=I&l+AU`2&WsB7wk||IvT88AMPiO_%Z3K_#k;xmoKt; z=Rqlk+&!LtGQJzmYP+$Am@iT$da+qp@xdVxDZYs_^;_=;rFRXMq-^F+y9Ylj@-6E! z#_C7KWxkatPT+%HnwT%a3G)i+vy*6u<7qL}SpLL&G%n6N8%Vad&Ou1Hs6?C?>)`h6 zm<=~|zTs>KjfdwefJ9IZKjAXjJVxpO5iizDl&3t@=x z$W6ig^+4{*$TR#l{O1x}9VcZVVKI@zY`8hLuI-aHJkpq{l(={wvo3`i840HmBoCfC zkZ9V$3hdwLv?r4vU(bpzPkB3&7h!v>ktz+7)qI#zH%j6mUeM+^tKFi0J2An3NN-c zOXU-srEkK#@4JffV>440{@@)tgB5GnFd(+nV(Zy+H;AOUq;9fd@{W${YpLzu;c_Xd z=Hn^RiVK{B3XGWJOS7Sq_n#JI=4s8jR7+FoWfMP6%W#lpa+hy8@ zW8A;TK&(H#FjoTE{iwOBZ0cA~T$n(bWc{*71^*>S!rYznvbbE8d$@j&rhULw`@BW; zsWOhCh=ZTa#I1^Z2R|J8S|`4TNq*ZM+fG(2lB^nLHwzP%9ewv@v}t|JIW31$RPG|{ z7fR1xF{YLDDZiBg=(#ZIxZ&(_h^a0%ngyz_24p~a{DGkB?Sd-BX3EEt!5)u|)s_>7 zMURF9k$JatTj61T6kgZcV@RN!%Nqgi9rOrl@0soYLGU(66EjtUBdW=8rN?XV>_IstC~ zM_{;mb(RvM<}u9za!~T!SU+qsoNb~vBGvx~fj*3hO$~y7Y4v`8X>L;Xzt8rMW#}Sl z_Dqq_ha^UMafcvPlS^~%S<5R_4G}Kd?J?N+d5w&BV0LqDOCYl9M9YDk+I{faR=It< z*T)|>ejp!{Z1Y{CS`giih;7e#Rgm4S>ipv8b&A2ATNQJ*+mAO(v~K+67qU*Czfj`% zCi9!U5%tJQz$5p*qPS9;##X;R;gqRsr2LWbe&}v7aS7-;zI6K!^t+0$JlFeDCrX%v zl52Qo{%9L}NU=iKsAfrMJJ}vgl4hI-Qz1-V>m_@tX!DW-&oy0}gju_KZM-k8u{|^X zev|=0Ke7lk^Dd8W#uc$_I9g)R{F=q)C1=Iy7fqx^Lp^{le4&fhKbh}&;iIEPg6lvnivVE$^n@Y<0a^DERmCDF}qc&$Vh> zuN}d5>i@Q^_A5tjZabJ3IG+0p-~Wv+S<~!p{+?IVZtgngUyEmF(CTJ&w~W6*xI^^o zt_Mcus&-l2Na^XKoygdFZ^A>QWqS@+u?H=tUY5HyuOENzn2|ppInyA71uDxBA!X*z zkPZjNny|d5GlV`h5LV+hb(1S`X;=dk3h9GIRgCC_ak<{ao{O9E7t!ZMhq}JLk@?xE zXdG)aoHTrjGPiXq^`qV@m-nUcs|2*~DI`H?ukAY(YLcY? zegKv`G|R7?!T0KL;(2V1Q_{9u=cCp?T-JUqj`=Zv5PPtG*;d~v&P67kT|GbBXYuoqPVwi`W#N%CDzO&@lgi7;`Inz<49q?1a zW=2rvSX@FDZ8=?^MG0E&F3DE!lfr-nqIyJZTu)o@d-b<#6IaFfDn+i#d?u`A=Ekrz z)?FRS8hd%TQN31Ce3%jwf7PKRZfu?RLOT7E$uw`wBH@&gJ|KBwlJ|nSL9|90acoAk z=Rl&HbOe>WlRG2yOSJ>#DlcC_&jK0Qc%D<|Ug|(x=fHgHQmj?-H)-pZM@rghOz}R2 zg^C~hO00Kkm5A#kO8R9WbZY^1j1DB+@>RrD$=yPd5 zDTRv7HjfMQd2L|2fn~%=X-J${mKqQ>Vz-J zNCKQt1srd7+_mYx(?`O6n7A8nwJA zX8n-N0pm74r?1$}Tn@?>N~%@CyO3%hHQE7AJ{hgs5#d#|HnUQ3bUY9FZ#eAqu{xU@ zI3HfkXt~m+C-G~0DpvJPq>y;ltZjvn%KLkc=%3i~MW?;6Q1`no{)OTyidSr7j8;sE z-avdtv#jys!GaG4uVCOb=b^2+32pLo2LQO5PU`MUsd?oxKZOb87@a@@WtzYno?ub9 zuh8SXf3F!cH6_W`YTS={9*tUd`qk{SmB^Vde;xVwtHoVza%xJX*p|ti*cOtouSD3= z`mE3cuFBC@jWl|XkyuIOZZZBb^|OnCcO|QZu47Sic{j&*CSg>Qjk=4zoyg?5P3AID zYrE94hKtfqj!moo8NA|Y1s2B|;CI|^kB>EsZYI=v|MQIUVGFb007YaQJIV7|VKU5i z_Cr^&FVD+(q?0hU(I6|AHV-U5EobO~1Qb2Y%vGeiX+t{!&18+5Br!Qdqt=wkjhxM9 zw14WJB%`1Vmh);2!yANSsWS%8j0cn_!O1?cNQ0Cq)m+;oAHOet zDL!okx$KTdyo?;O;yDu&Ea0~5m*;0e5t!gfX>^j%k5ZsE7_?_W!A1Ua6=0BS;B)Vo zSZsF>Pef!ClZk2a`Igw+XwO_$R{wZIRjZ+|50DHtM535&%k+;&zzWIO!F-0fHm!0q zp3zl5s8=lvy4rJD)m4T||6Bb(nhlF6D*D!*E8wy!TE~4EoiQGMn``ldMy!zJMs9dp z#t+QK<8rNQM5S|1b5nUc{R_p$uTE@^axe0%=gRoQaN8*&o;=5s`?4})BK)?*;)jV? zG0Bbg@OHPKCR?53p=*t93B0_|)G-BmE1TO<$K5yi@;F>;5BgClvN)z{uMflg<9Cti zegDDtBL(>u~%hD^VN(KD|VK|=y?dnhS6uk!WzVewk(?zOp0Nm7qU0nANl)9iv4FB(7~{LC2lOaz@gga({kfy+qtEyzW} zAuxWGOfXcVIfl>o!#U5W1^G}BRh#Q!-Xr%3Frw}09!pG-z53ew&Q*v{&SHVMr;@>0Iy%XoeRp{wgNhircJ&ukk{hYZcNIPkB6?>GmTg1waksi+XA-T2ihtR?N3^+xVl}TWro6q20W?jr< zu*@r`c9yfbBuVV1n5(c?1+RD~*N0IafZmENV>jT%z&fO*#TXeKCGqM~j`xPLnt135 zA3ju4d3kiwl9B$bF>V%2^&^)`0XsB`TY}VTEZ$)vrBK4)K`e$u7dT6T=5Iee1Qr9l zzN)Wi>)pfs`K5T7k-oj3U>1O^#eX2lMIR9#?*|?nSl8sgUG@t8GyKTZzWEe#?8Cdr z4-qk(tV@;CLz}O?WbUeyJ!*WaF95@R=d7hiTtoK&+_d}~bk~eMh4wv-1Ei?&q@-d`8LA`{ z*DX>XG^7u)t(q~IC6!9OM8pZ4$lK?d6NI;OKSp_n8g0~9YtPG5e?K$-C7votfG(eB zp9?n@RVND9cpJ*&0UD0CirTx}_Wax&dX6~)x#l>D(tF?Uz>G9%=#+=V#j|BUea;{x z0p`z6EFEe~W2N)Z07};?_ecU1gjpS!^~a#3@J1}2)e=R~{3o~}nhL?An8^yn0m16i zz>~Fw&;TCz(83?9V82>`QYhTQ`LCsV41w6tP8nnIhN1Hp>-~EPv58QnXKjqSV-ot* zny~)0fE~6)?|@_~(%r}v`Zw?+JWzfi!e2kQjPHr#^9D%;AS%cSi$%vXlg$Gm!$e{` zf>$;_I16Q8_=UpBY39!EXLP97UXS#Cd5wD%y_ZvYF1D<7F8ptjL5F0V#5!Fbc}<3l zw3RbvPCsdQMIp%o++5}1{^QY{6nr@pJLwDyP;GP}_;Lb_ndC^5+cWs#By1Z_+b z9;$xUWJ0zTZdRqpv$@@w!MYQxmOxa;=h_Z=n%f94gknN$6X)hxvv^YQqEUoXiGcJv zgKa2WaxDa;1WfE=AXIWpHoXiyK(g0m?;-U)YG>8H*td5486tV>PQf)7qR9tjHR-3=%XnI+LXOD+&&nJ!Ko6^eO(I3{{%Dx2dk|We6?acY;M}8eauRl_ii6JGl zwXgg!+!Y18Ir4m>3G?5o>lgw)evp{6qvW}I7lagV70l8M8;d#fhhiQ8=~Cj97A9P} zQj2-f6H{vB6pQqy*U9MY9!p*Cp750y}e z=&pAh^F)to3ZwFOal#;AFLofCJ7;mF1ta1Gwk(64%tIhaKi4-|ru3gQhgvOLH9N4H z&NNgC9V3MaN7ZW<_xM)ibe%x6xOhrI4qMiQ@B}AyTiGt%1t6`9j7@zxg4K&lqg&g` za9Lr;_D5g%_*{7ts~4MtP11#sexig6{P3$>m=}&1U>0%7*`=s?3o2=R(V1Hk25}ro zGh|J7onGIXd_;w^(VhXPwy@SdK}MXe}ppk1d4=pLU+)i3@u@(Yk$R0?OY zmuRqWMT*oD$0?)-a(u@A6r+b&d7eW5>?CEH1*jWI$_JUr8a&62`Mae4eWCXo?PQjMx<~T>_H63g1Lg0Z_ie73TRbpi>%TOU`_P!O>-B6d_Jp_6iN3t zP6Ue3G1unG?5I?mL8~$aW;{lUVr$x4j`6d^itrN|E zim14DnasLu;fX+y6^-Cp_+|`uy4;Y*-;&hZ*80pw6YRv{r;0Z78#04KrvEW|qyrB9 zTBg3Nht1?R8a2vU?UY2uFBttFqYxt?bDkqSLmg>dxA{Ad(DV#@5*ghIjp-*-8`wZZ z%tdt7Yc4*84Vhl=uU#DKS_(62O4EP)E2)#yFV5=LbRA3MWEB~TIDM$*&X4~pM;{{{@@^~Y-q~WAfnV2FuqdDfYt`-&Juz6yj&;8 zQt|obeD$d@$xnkcV`7>cyQnTOc|(Ht(BlXz>0(kf9C5Q9W!tt2&EtjEqjMpxF$kHa z16ndb4|FS}Fv9qt1}Q;V@1A(hZXrO}X!=^B&kYp~M59K9wgUN~(S#MVzi&@mvth35 z1xD)1v)7Gk%h;R-u?lL=gHm1~Vqo*2d8=`z7zW+?r5aviLlvo5;xA#%S`yqB#^ceC z)tf(lTJLPEal|o{GuHG;D)f@iomVe_f#b482OrXWyQDDN(Dk`&nE6_R%1!YVs3U%! z;G^D@irD`_PWi|HDH0wlHE*LK`(hAldeWG}i+Edu<2VHjM#y5%E9kw0(bM*mCY#^` z;9XV(a@A`r;{t2NPUybI5Kvs+3%BfMhwvs_Z!LxKUVT$4QUDO<#h3Z>=IBsA`NkQOZBGyKcIuj8P?nR%fev&esUlp_X+X z7!$YI3$rD4g(?6Ei?dg)Ucd5xvB8VSHoz(`Z>FCAgb7pF;>Eu{uM&`-SL}x$yn06i zKR(vNCTjV8-=(pbu7nc2r)KW2#jhbRv~(J?mHa3MbKa3h?G8cW2IoBPC$s+=9ERU} zMIoci{J*@$B7jT1-P~(SFF=?=dwvc^RK*?P+;w^Qn%^p`khkb8FR9V-v9sBZ!}eHimmgG$6m`!Omw`uA z>0?U+QjXZ`Z-tD3F5DHZg4VL0&J z*|2Sxf#&prpdwW#*;nnB7`~oy?hH31*YzWyiKDzGoZH)IbXFBfYx~6CT$BHygkc-< z`LP$U1!s$*Hbp1*I=8;Jxl^$5siNZae(^0g?~h>d^~(y@{t471EZ^7!G5LPtx*gZr zeoP0g?HskcJZ0{BxQrgm)`R9WwenI)TOs7L2YogimAv#`ncNKn3$hkI{6Is8E3sS9 z^9NH@&z7LcX=(+TZN1yKf~CU#bneD}6b;}KJ=-2qc%@7FjV4myzVaLwdW- zZb$~l32!YO{A&IO^>CAk;%jLRq>A1(Q9|P(|U@Pm_k94KY zf~M+uxF8@c4N0jymtT%XY#Kp)4=zIPtYu$4{RbZ;vh6e0r-ZA&y=r*Ya*Sqj>#`9~ zy^Bn7LeFKPL)s8QULiUKh?rN|eME@WxYlKvV8p5UsZn+QZZqr51W#YoN(?$MMyY3M zJ)|E*`RH{k7lMp6dBk9D5U*~bDy~|PCu){%^pfYX!+pB7XgsJ;Zp~>7rP*X}ACZQN zr*T_y%q#uf*bOLF@#*`E-ch`&-g7a2dQ4}Zt#_qRVJ{tH)r5E-_>2U&bOddj+muSm z+Ue;GPR?sS9_x+n2=D39@)T0u6BC(FHkMO^5u_Ity^JbE8mV?hhH?WYoZ;Qv#D7L4 z9D~mSaa3ztNqHS*BJ`tkv}i3SdjFAmOoxHGT^)Yci{boY1Zg9UOqtwhaSV^u+CMS@ zF@g6Sbh`0=@XS0`y^yT^MGds;jh=*@$RP=b%E_|oBG)I=Kh-!?2%NECYKUBAvl@ zdZEYmDS%a=#*E-wh?eQ=PJ#NeSZDG$;mW`D1>$|l8wtf)L|u{5jcCJ2h)wORB9m`+ zV91pnVi=?;5_J%TP^i#b81%f10NbfxrOSCR)yT=|++X`^2*10NJSh`ehvYG?wv;ulwQJzLW-6s$?XS<$%ccBWgx`?V6u_ z2V5*OYQ>AGip_)LIBs)7cqZ@H(fn=BG07=TWE|^P{h0()(Y1Kc>=7mj3NV;cZqV@Y z@4@^raw3sAn0Y{zez%H6@81?iDxjwG$N@&49UJzlJe!t)it_^-leq4Pb&BinxvKtN zoRHJufz=L(yiz4~($5EqeSJ!V+OMQ>!sHKRzW;rCU@Pc=_0DGxXVj)+zRX`J8~<9f zKiiopl;u==GT+p6ALLHv@VA8r0tc0dv<Fcl>R@ z(pkkDoafnL$s zs{h`GKaLDuEV#@#Lwc?_lV|GP-+GWrEg4^S30!Ik*Ohh&*ux$@jP0Pw8)`?~`62%g z^f*d?Q4$~4vwsuh0%wBb9TH+RYi;j6EN1ISse4v-)4k8-4aehI1yiDZK)MOO=SA|J z-@Y#YSz!n)V$v*ShNb)E>qjwW)$(h?KI5)>w;sX^AjRULDP6RdMA( z&yZFnp{U#7)GtzG_nQo@ypM?hYdx*Mb$k)xUtsRT$z+=e- zuh8NutI4CMo3CXj96tdvpxWjlP{VfW$)kFo>VRJ;m*qDld|Ki4dr~l@(p8XK*DsW3 zf#K%LHRV7GLTSseE=At^S0{n4g26R9nIZUTI_PC;io`El9P?1U&7K!{w2+I-7l0h zu~EwP`s=OR55G`YKW{1{&3SyQdY-~>(KzDusZq@sv}I-Jg!K7f_^8KEG_Se!GjY8=PA` zzoCrpk;ikl&c9Hkbnrgoe(KJ2B5};q`tju^H7Vymj^nR3SYBUeZE>>PUhaT1X2{b1 z#=LPgSX(k1k7E_i?7G68Os$H)O}pAVmOT%Vw|029RAJlLdtp>fzyS!9XXMnM+Ag}} z@P5H{Nv_~N*I1b(CRPadPP<=?@x)Y6MwOIOE@*PqO-(W(#|g6t2*AEQwY;I`@r@=G zlZ&9tm5D^q$Zp}}oPHR+l%fMlKg3=nQjb%|YyJ<#f7H9D@ZXWfNW~N_Z#Z*KyCd(C z4_Wc?1q{KeDoVS*ppC`4gKcS%uV#$9S93N#5eL2b zBXeT@NSGsWYVs>c@4N6P`7QF(AC>wP^)S(stk)9-YGcx_{XIqPMU95Hf?@0&LF&il z4XP}a?)zyr+$^0(dP4-e6C;h^xc>MQFd8!Y zk=*vGydO8RP%Fd6*kG06IP3UEJs3(z=I|{gv$P#)6n#G2W8J04k#jSC3d4K|%37h} z9soGg_SjrcVi%~&i#Z~3JzH%gKIz-}xMniUT*gcFYTUgjw6b$SvRL#;g0tgKc{s}i0H=5@j1`fZw>G-HKP&XF}n!sl%W?(VdoQCv< zf{X0)8UBv+#aJBO)-ZPzP^Jxr0lHSKe0Bb;-s8x9H_Es~8#WDzzz1u&ZkWC<$THT8 z36~Bhu9y$DMq~tN_&$dtHN*;SP4g|;Ocd=)XcYhY+;7)t2|rMp48OORG+{9n)~ zeBS5%ul23oAuHGR%->hA% zv6jJD9nwc))VT*#uf$liRqU^< zs8k;0KfW2g$Rdo(zaOIDNv4k&*8s0iFNCjJ0rMsv`&Od3+H+|wa?p@_MqSIKV{LWH zp!vJ3YSivODeQw}{FTydQC0pJS5G{)03MApV9;qg2Y@*i!2L&v`R1{p1wYWggtJ~&0#ucsl)B? zJ8G^DY8KfH6lo_K_V|88|JfcPQ5W%+bh;Lf5vF0(D5W_;B)JBoO=TVWs+T1$|XudXN+ zT)9<){gD}x@peB9HxR}FVi&U@%1?1_MYdJF5Ymy#*_O+(IGBSfp?zZB{-c+N>s6`3 zNB*Pd=`g^;mzlBg2g+iL=+X0i!DcZG3Zt9De=H`PH4e6L9yP?ga_kj@QOI@K3ZOFD zopC0csrJ%k_twSGY2gj+O1xExt{`}@AUgckm!e1AjKPd(9;^E~@#>e=*_3!^2rf(Y zFf_Kfrk5^9UrFVjIxB#1BP!0849WS>E+iyK()0N=0Z`3b&J+JZ6WvH)Gc}sG>7%{r z@TFiU;?)1yQWp>*;?=~|B}e0%xV`XaYU_a2^VD`A_KavWJ5ExZeS0JVN7Ey==*HAy zU_#N1DYk8mPNHhTn9e~=W)1j>&rZk30bses^PR2vsW^A7Z{pFFp5IZ!mMk*e&99P9 zQN=l0gDl4AM6`nyW~nEf=zhamykj%Y2lKOt;a*3Y%r4AW_0G}@KiszgkCl~dF-_1E@+V&htM-ISs1i7+~-N~f%d<2jk-Oc+v9EItBXG4D-d04 zAzWokduvt49b!`v(JeODg@u}r6%qQy$GmW~u|dlnd>c0w8yVpz6jPz}fOSVXmkmR~ zV?5C*0D#zmXfY7B)7oM&{NAobbp5!k#5r9^zg6lK{4&e}eQU^R=jxwazu8oJe~ z7|$GSo0~Bhhd!F2L&73eg3_h=8Y#; zr|V?J>KU#nBluGq>>Ac_Mw|)M6HQY_pFi%|$F0#I+IxwSDr_(~L0g$s?SbY9@^-pI zY#_wbmgqZ|pj|fxZqIRr;SQH-%WXk$kmA9X_=fq3$9T3Vd&%FCi!~e2g zY4IK@p;EWf-AQi>S=&I}N%~Y~wV2DH+%F;+Dv*0SQf8$hV_HR7$X{#%N~p=+(K(cT-hV4{=TuadA{C?Xz zFX4SR<*0qep0LWrOYyq>A1G=OJ_;G_Ws`_FYk%Q<7W^leDtJRRSS5{wJ=uZD0`{Z9 z7E^QgcHShvjl<*?blGsyw9vVG9Gx1Ebz34@lZu$4Gla6w1UHU(^yQ+S%9nz=7!JL^ zekz)P5CRpz+KvZV^-;ZOe0o|hQ@ck|iCJA&Ef5_Q_Fbp@)^jV?KF*BH!;FR_o);I`p6$m?y|l~C^Qp6#GhL_N;yPNU1#IS9oGy73tq z`bt%U^`ZYn+}B!yg9Ex&Y>TNyj0XouFHehoa0gjgIPUwsl57j3*p8&`7x%Z2^9%l> zP$Ag5L-dmV#5i+mlC+|Rg8?}_s-lJV)&tM;mfNa*L~(i~;%7qzcCCO0SzhKY2|IP* z>}{ROnNkaT?HC_oJ40o*566KLPSlw&JU?VEb zAOq+I5q)7qCnXtx+E8h&O*4Q|!|u7@FqJM@)&Hn%4+4~m+)0#7hkwMb9pl&?! z{c|Y|N^d}d9@qT&v?@jqe`wEbn1^vO7OV=vzW!Qm^PK40D70;d9W!wjNiB4v{aK6} z4RK6fqCO!Ll5g-YEM09W(k^dZsY5N67y{cSGOfrWn9F%494?XF)+_$Z)>u*HdY+Ws zNx}f`!JysIs>O<)tKQdP>IN@yPmNEcE*z_y)^!pAN}L~yI{DdQR8gsL(>^%M>WuY_ zX);yN%t0G$*)FTQi)0R5S&^|TBMj|iS(zVK<#REb(U}k0(o=*nQ|Mzk5^W){kMa1V zib4#tl&YtAT@H%aES%h>T z5h^mOWc?#_p?542It5;w^+>Cd7-Gti7^qW=PUPkky^~Buf#vL2VJ>OT>_lx?XOva}*Gy<=S^BnWz`RBX zMZPD)G8hkR?k?$5^=_Oqq2ta0zfQd&{|bfaM2eoT6jTs^v<^xS(e1(#^@6d0wP~JD zrz8Uu;vgid2sYPV(T+mc5#N*9^96pM5! z?ifaE5xG{v(<0-dg%c4?&%s8Y>4}F4vPIUE{dmH#PfGZRu*0kFEn^M>Dr=HMQv+nh zjpA}1ae^;)(UeCPHsXVw$KH#IN&>SVFNPInXkWJq_m{r z+Dz+H=-i4DB+r~^`@iUZ>9O9n5P11k@#~=M%R}!;U8FO8F<}v{Uo?jjxoraC#4|AK zlt|Q6mMBc%?2AAw-)Q>>0Fk~C^Q5tR^~txkf(O<-J*aNz#61Q2K8`M}2j5GK&lH1E z+%}HdB3YkBy#sYzj@UwD;UYc`uXj9sw9spjQXrzTF*3(*MKTs;I7!_5x6u0i%Qt0}q>(SG@Q+XzR zZ`FMRe%74XtQg_fgPK73E?HB>T#U(0rjy|ZO2L!&Z#?NH;@`eI{ec3L#4B(mJ8vPv zrBBE4%|F0R0RndlSwsU4C14FL^P9VEu{KkD(aKnQ>EJjcH_F|V3D?D2A2%hc0`C4X zT&FdVY)%mL!k-kJ_0GIKolQk}C$$}Tr!uJU?U5EbjLS9byaJa_z)qO5>L$GtvehlK^pv5k42-&*ym`cpMzDaY|*@Q*H z+TYsW$o>fo3gUVxr!ay!g4zB=7d$S|EYOS!*EsNp)^4gdY?hRIfJFWa?K$1 zj#(f>pjoh>;RAi0dMbb0jnEt_dq+AP*FR#Z^A|$*D%B2%570Fyw5#zk(|G{`89ICk ze~{&-R2VnS#sI}qf1-O^I)rOE%9U4qKb61zX}6h#RVWNZu>e|+99FEw?&E7$7@$N}md3mzA-O1OS2ZuQ68)}7|4sayYXE@W2`ojF zs|g+y)Hc#LiMh1S=9Cn>E$J_XUZzGA0$7C}_WLp_V0QjKxRY!Y{-IaKU!30GXX3Oe{hFC)AyVdobSLhVQNoDV7m`4J0zG@y zs8K`er<{fKgcM|YJv__MgV?X~oe83gR8X$6*lVJ&`UTX%|X&uS3n*k^a-2iB}~ve*4zbw}#N!os|}Yx z%ZI_AT)tuK-c!P5#pETH$_3HG)#JnMBI-(H2lDjndjwTl2DJy zWEa8^orO%Po{)*H%*(D6HtPA|!_Q6A4-{b6M0biiVt1Yj`r^vsWEXR>D^s}qjBFZb z;oTVK(zAXN=Zj|A<<#T{HAf~^;dG5cWFqgMV$jx`81USws3OgPu<~Muu(w4G`4O~! zEgN`|6ZnYNOD>2)kvBN@ZCrx0=-^1Y9AHEv-|!04-XR@d^gTFMsu;6!#=W(C%7QX~ z9Et|&(gmQ=`n+NTZp)bFSY^k%ei7NgK#jJ+H@&AM4(%?#w9}hY(Db3)ZyCZ=3Qw;P zfxgU!9ATg_b*Kf1M%@5s?_hH9i?qDM-A@0YDG`}#MG_~I;^*Gutl=C zn2#3*OpQMOn|sWBLv@dI8%s=nv14S1cZQy%YrVyVuxq&BtR zvdZ@;fU^yaP81A>qA9XzTO`1PzT4>PXz}j==-AlRJs>vqpcI2cxVM^ekZF;vmi8l6 z>_^VxbZ=kWOucrrrF+X=hcl<(iDT&Zvw%n28F;Ngk2I+-coI|0QK>HmQ`E4>RkC6W zSS;0^ALiJ__Z9~sg@tZcs()-EaslObR^DK+*SA7!-O`NH&RV%z?*D?QyFxwug>!ff zZTCmIK4Asboi36ZmD4{uK_9WlJ++P`$Rd9-+Xh%E-QmZ zR}e(;TH_JGLO$9pM15vbqJyq*5ao9-3-su&o?Pp}uR{QwCjNcj%geN%xX1lIc399# zC=_l76q%O9Hpy6^$?+2PrVmucec;u@-f6C=peB4T5oKUxgbAPrc2H=V`Oa~3bJ#s* zG{b`T)$RHv)sv?|f%|3JwJ{@M2@F)&Ln)@Gv2Ibc8z?;(+hT05Yl4}B~#U6`TP#VQT<5UDgMuj|3Udz0h2u^>SA}+ z1%}k9!Ysn+RO2{(l7z-5DM2K-XXM$Yzw`Zd?CQ(KD{utbnX~t09R%}KILSH^H*e4J z6q6gH>80LKw|H7AM&B2O!&z}(B*+t@b&$lx*9Lno8 z!ng%$ggc6(H@11uqCeaC`3CvD(@+gY~Js!iV{@F69{TgEW^v?$)-_9 zW;@q-j-9k+ZGm+R%0jFQrWvurLx$LLDetvNUg;3>^jTND=vrCZ8POjF1 zqkUMXF~%*46cPP{uhNpn9%S{AJEV1g78Is`>j7D(xz`#!MiUoU)XL;iSyf@qy4bHE z;_GyKUu7(9GP?eVMdC%`+DY&al$D4>gLz)#^QcSPq&48(arA>eSD&^Q2l6Myj`8m< z6@H)y_8pP{kae_cHI1fCCk*GGgnOU-Kw}J{WBSy4`tYxx> zOVW)wpig=zNeS!+xYrHm0mqT~uwEVU^96d^{rTpfK!Asl2uk+s+k6m`$65A!H#cD-1>AHxuC?@~e^$3eE(Zm>w%Of_ ztU0RLViRH7coQ2hfJyC+<%sEN*qP)h$cteeyNoIQPXxj~cGjlAmQKyf^_U}@hw#!kZ zpwbvVO0gGDf^oPfL@#AgH)dD!)kR0KSa*veb0)_Nqtv!un76k};tEXXo3yE@bFql2 zpefbZJr?E#anzeKP1r2xRR))~)~^(-frsLT8qpUYm!?`}(ViiaZJE#OL#fp&j zePeu8K~aFaG^?4lQoCO$iS7^|OB#?f$svJ2rLYVVnC(eLH?bajFd<2F^fe|0K@H&cH|=j)szyA@1Bv6hZaE0Z3CK}_12Dq~ zuRa0U^!egOIz!|`z0jXO!j zIw^~q?^GR7CwnsyCNSqm>G&v2IITTt&`f26H|Hz)h&Qfv0|qLU%KDjB=r9+Gr_(6@PO|6l}l zCmOv;b|8gLeQYi(J@AqQRn``@&Y7KKa}<%Ih=)`28Gxrc!5DYfV8zknTEpNx#d5!E_TH@Jm*k0s#V>JQjVg z>Mj3S6tFV1W6A*x`8NE`9BmH`34e|QPMgvAdOe*A*Tu>X=Fj`~uMG>X0;~xkt}w28#|t&3^g+=;p?u^;wk+K z|H^`AETAg*5S~JRqW{Ix4KRDVM{@5~z_$#;HpXOs%$+A*B{Ri`W;KRu9WHf!;j5lW zG~ED_Kr8?yz+OLr;~r3%lnWxaoN-VeOIPYlOWvR1c4eqGT9bEb=nY?dmO#^d#3vrG zWA2GF)44-d6BX*K+a47{e-QKa(`X!NR;>9wc~{Xc8UVJ-lRr6Xrr+gX*+&{-AgpSa zsIMTZM+^*oak0&BJ5T9yf(oKTI8t$!bBDf(`{K+w)hs@K81xRiIQhZX`9R!v{xm!T zv_Oyymd4w}Ce*Gti#({6H$fqf0q@Z!QW?(mxVS_B&nYuf+CDID!fmLE##&M87|V}e zU{lAWK3}ggJ&ww`J#H=H?r;*m6AYF_N(-RZ7K7qgcEl= z)WNK|g}hLo<&Rm~es*~~H&6|;i)zB+m{y2c!-jMyZgq&I?spcESbt1q*6}cn;CZNF zf|eL%ko5lROxe+tWp$r*{afazt;~deK{bDzz**Bsd94ZhYOYd9BeFMofJ= z!^NEFjA26t`>ZPmAS@Sin+5 zeR2Tm)+*HQytw@frXyzY$)IGFzD`C#oqO1gb-9;^nGd~&(j+zp@47*6iwWFHp}g7u zauAwR=iXddBKyQ+3&{UB;f=H*6^||doNeoC1O@086`4?BH4m6j^);OoAUR_v)U1u=W4f?S-~kS!Nx@S5B>vN@$P@rz2HN*85Albr%;1IWtC)Cj?ePsP*G@PgXsazDprx{ za7-#m<-27XpTy;%a=Gtc_44nLFsl%YCuH<#~)Oys>dVmp(`vWA7`@*(elKwarEAh?yeSIh(~` zl02%z{RF&8>%iP-;r`(L$Ks2wVV;uJ4W2DNS4%k)X}vv_O(_I4T{WPf_ElqSD$N{u zEFvmlG)Aexc|oDBNIdG&()g{*gWjIQ9r=?4-{+k_Q24nAYG~IiPL3GPdz&OzUL+va zXCe;y&d>gDwDhg;l*+e>GCkX4ATJ9>V-!E5KjCYl7N6{#&i?891wc{&*pbH8)Z|8s z+YG@@P3_wc+%5z!0pCRFuPoU!-P3ujPi_#h<+|tSSrPA|I~2@+6Qj0S9So^mZlcXx zm)|%U%$dzzs+*c;lvr~1)Gkd&_&$*Dj~F`4)L3aZb#T;E1BA%MQD>*^@(k#bsy&2_w=m7M))6y-e;T4XL{?Mcd*ifq3!&> z5`QsWaf+VYU@TqShR#oWD%sp zM&N<94HBt4mUd5+YjwIYJl@w_;>(xmY0-q(IbEvA)75Q?No`tPz12yNO&iw{W(60O z;OgelY=Zsb=Xq_swtA3g|4)$j7X-pE@GDYx?7^h_TIT4E8l0vO6D19PqJsMAB1)Ql zAda#8#{?^OcUM{F&*{BFi^wiySJET94Gq3kNZ9${g!iFCxud`tTWw|kK5RY5(vp)R z?yGI0;TL0+h_LAF&6LN;;!nCB6*B{OMmixOWeCRr>em}-ISQ%e1leRcJM-9LLvCss zj4G;FCzC1~%C3;&8Jyf`Hh#I>gkHRrB8*_EDjEL#^>xb`3lpe*gk{b+)MDXmn7{Wm zT)BrWIA?K|An0WBlRYS>jZ$M2%kd>&2X0x~adIYBX4XP8@#0T5i zc1kPipE7*)*|?w;yW2kdX&9j_cEDw*ZIg0U0z4V81Ve6?`}bBIujc>i@!8@WK1G7} zeoem67J@6MCj2|Pz$Vd_4Kak24HrCBip}@gF8{owI~mEqyyiRmyL)9xJ%FBgdwiz| zkA!iAJ6Zrc3^ns%-XPmw34)4pNOFx+ZMlGEOcf=%+$;4oO(@A<2`v3E;+#M(5`t(M zJKWzts9A$1?h4jmMbYKoTD)pH4b&d2Aq`)2hg!=ruK&AF0ql@|0SK^|IknAKlF2HN zfND{a$OWihZOwiqNJJ(T>XI+;p|50=U!MBZT-<9*5mM2tlwCfPru!O!jO;>i&B>n5 zWqA_SudmXyz7su>dW&b8R|hZh6kA*qq|@fKAZbbsd@UYisKcQ6 zoAouhyfb^huUZdfWG;DH;(Xr*Y?hPEF@Kv~VTZ|Up!=FFkP?@)nE9KdHf~KE*9ykB z(`z2}8Svp-IM&NKb`Y7Z4#b@@3{T)K{8|qD{@!j-T$0KGYJR>n|Ca#Ml!wlAx_G$P z)|?+?yPx>pDQeIuTg9z~TVcP{+73fElm8r4$4{KaW>{Y1TZ8vU#%F=~Fxnge-*}vC zb|SttgKEo^HdK;CpkV0Kgj;0Tfb|pe9`Z37fMAsQwHXL?GanJWnXPKNUu7M=`?I6w ziYjY#GWJ!V9ZGjdw5c1AN}tRKwdl6BRjK{Q%?ficDS2)cHyP%)keD}w6A>@U(bH{?-@$3QD@U{GIT%PeuAUS zxx%z=>s(sof=u=MPXMEDwNW8OLR*yFdICn}&<5y`_fW>SOxZ2Dknu3Gq8ntO)#fbI zEtT5#>-_&9dnm`Xi@=_0ovW+z7JqvMo^%0#QT^|Un>qzaNAK&nk=#pt+-4wS8PD7M zFIabImjGC*2W{t>)h|L$_i|MEK)>8#183f;C`Vm4cqP6Xm@nQ^4UAG6l$ z#F)hw{L-8l2tlEfOvtv?6SW{~CGaB;+jq24Odpm=J=ug7g^U=NTPyQA9F4N+J^B%m z9yT^c+e85<#2~%VwW**XB( z`OEggu~!I5q-%v8S;WZ7(dcY+#uGU+qRJTmB0GV~X}dK=7B<-Bs1GgQ^0|Y+JZK@t|OifMH$NSLb9hT<2umI8M^t>_7jg;t@__ zJ+F~^u6J7&ssbEO{QktnG<_<2Nn13t4L~ydTmgX=L@bZz0F-(9_kH_kk7pjJJPVt8 zOt#_i61BiWc8keB!YDU=*JcftB*!q9nGK zfF`&cZ*P(nD=H^!LgW)pKY_-)onjEn^z$qp*h@h*GT~4iODPkO#t2&ZaW`M+-f9$& zV}>~*J4Y*}Wu8g@SQ4kG4eg%ts(LRUCL+@Qk!#uQOd|jF#URn37#G9OQY{YbJcgTX z5mr5{TwbnW)$m*~l>WRQC?`)YS9}=7Uod>>XqLpdAAhVK?vAIv!aA?Try;>6d0X{C zw0f0xY~C99&3IelI4@IMzfe{d^)io%j*ytjKoo#J`8xQ{DD z$X~MTyDC;w-8L>?)3Zrly0d{6*t4ByOOF1}7;tlMeCB!T)ORrtH34vlx&o9`-v_}+ z=H8uk;l!83S?UcSlPAM(R$Z(I*@_J#)a~Z+`D)Wf8BfgC29|3s)DNxpXD;(j%8qwl zX!=Z^dp^jQw7(R8J4P$?1BGT_Ga~6l`1&{bzNQY>(+qDHh&&0;Y)?UZ;c&rCySCVc zhKqkeaWw-_T(tT{in21$WqC@Gq4sbA=gcr_bPzGQc2-8(uk6pd50I!w0b}X;S-_4F zYn3ovB13=$F6V#nW-wKq@=iB{l$Q+jG3(1Js@g>AiWa^t$f`Y37f;5MZXI(+x}R_+ zFqxs|wurpTU8nFBiA9=Dua;u*$iqN`TgHD>&JMGy!%W6kyCbvB1qo+z0n*0+X{uBB zDnQ!p)hrG_^!Ee-Eot-+9wz?<>4J!FKETlpVJ5>7JN`SjZE zgY;H7`^(k@S3RA?wwjOn`9~MZf&T4mn_Q4pj{yTur^6S?*C){H9if6IJq@V4^zP<| zWYg}?A$=MxG&2(O>(?hr*bkHqUg6IT;118*!Y9|KmfM>j3iX9P0%+hOKhLA9y$;n$qgf!matpzsOQ+Zj zA11|aeD1>lUcRi_fM5>8PSe-DCTz1k7`pzIL9WIzbyiSn!Rv)Xn zf}RcTt(BET#Ix<;`q(~PiPbT&HYFPdGlRYER4ixZ{4?`VB+u#EDrJKH2PPou!b$-B zBRIISaFO9b90}kf%kCwY1z%zD6ChoEt!tOR57ww6KPnuj_O4(9AUgVRS+i;H*^F12 ze8k&)iQu(q3=RvL4P(iW#22FAE>DQ288=cOrdViV=Nfr@_n1W(V};aGiWYiQpkhYB z++xpe_X-h*w%wXzv&K%PT;)1U8}G1bVqQr|@EI{E z{~A`!hEvy|pmB%H$Xa(WtF3*N0(H-*gH!HY$SRvE(YeG8^ei$^4{nW5_efDq)vV`+ zSfS9wu7lNAU`_;2T1u<-JwF#`du;tZcX%o`X{+5^4jMK|CEW7+A_cSCwtDXJf6wE$ zXEdR?&%lto31g4yt)98m8Su`w^j$h%ot%0r&MmncvxOFb@O^h355sr6D3Mk9uOqXL81boysI?+$h>@2)+rnFo~=DuSgU^xBZP5PG{{+-x*FfP~PYd zLjlx7u9*)-htjAw2JzjXx5NZ)reN8UcwFm-gp0T5{+k5geL2py?XeYLZ|eAeYwJd# z)=a|xh00T8UEVd^&9vkRpLPYHjuI|TKFX}=5ED;!{qGA|a7}c@yAcD&JNh}X4K)3t zLo>@M|9K&=*XoTe@(qICh~mnVklaIl9cQ#WL?lQJ;PG96VTtw?s3Bv9p2|=i07WR> zEb?3PoI42Yhj33urM2MXgG>i26nRA8{#D&xnAY^BeZ;29mXcKTw!u}oQwy7QB!ZO4 z+Nj#TJvx?zYo}3w13=0dEa~)K9G|r>_s{uWii1}<72`&kUvkw)dqBAKryEB^@ceyZ zWlp*{C?5aaY9*-2>ukrn)rpB@d0Lo{gqbtF2Pody+gv!-i&$)t&*v}14F z44Lgyh!7|&*z^M3T4y3h2;`sjD()t|3N7hUK|Z*2ekSXRrL{#tkGK*O2mEZL3H=`C zL=KZdEJR;#Ey*saVmx_oBuCZ8UX=)yajMa$F0v*?mHft-5P^!S0>03+k|S>(5P$RP z23tljZ-8a_*oD01u5R^us06u|RhZa|HOFf*jEViZDZHzWkUwj#2k}x6Q;rK>vl^H^ zI06QPVzr8)R|B{#)d2uGS(2h^mcUtHw>!&z6@}qY2NvIre~1kMMd_AZ59q&oDF_G= z)nFYP(B1-d7moGROxZuUaC-l>9g-63l{1rt={gDQ?YG zp8k483X-Ep%>vIkXR>a&y;(nVy%EgR4&hKlqLK;Ph{!$Z9-B|f8SDot!dxW%%y-Lk z35dQ_tcbxqyQz{z=@NFshwkOdeg!jYu}jGvl_PsHW8}Z;_QK z%;ro6-FYC!Nr8osW@;Xkno{pe5{U@m##3N)KC+Jcc4`%3)Q?D%slzn@xs?;E$M8^{ zokiJBy2fA{%w+SrDmva@jD)w>l$`pW0}CYp zQKT!7d7TFLedy)@4x$udZgFT99&0>86c@S~mouoXonEY-uW;1v^;YoYU#HggbS|HO z>EqUs`_@AfdN1lOO-;M^j4uDR^xk{c{4SeTzj}!(Uv*Y)1=?Uj$*luyYY8*H`Tv$- zau9Z>ud(5&E*I8JZqE=fdLKqs$RN9gIpp2^TZZW8LWtZ@4`|Zt>3_BLv*CyOVY4du z6V020vk+D3|0aBHLP@LhZ^FQF!gN5o>pEaZ`gcb5NWV6zF7tD75uO&47Ud$q`&<3r z8FYt?O@|SYQ+nE)9Ce?yJi{M94S^`XeZMKsD!Gf)+he%eP27ETr;*~h$cv31C>RtE zuv6pBnt_l53m1FE^iHqNEUA~tYYSg}^z0!GCWx+xQz43in1U5QQ^`}0fTWOY`PxjJ zIQJ2}2d+21@*OyUJYC;-OG(Mu4Jh9$SKE@M%U@xVxb zW2fI44^Q&aV3!UJ@&NCy1jr(6drFZhQWa*Da-Hc?{%@j`;mO0(%G0*`!RA>k z-W1N?=bboZ83=})PVaTrw)QxRosM(!q$N7F}{^zFw24h+WkOu%&~?4ZsYF;O+y}uva9(f9?~1KVQR7K?3e>9p{V}wHF+`d1nP}{ zWjT=WG{Xip-yoYZaI#b!#;JQt29-k%Go}8Eki82q!bl&frZQ_nqn;hVj-NN&>l&yD ziqouM%>Uj5n;d&iKEvs1mP1-4bll`;3YY7PvbqHl&O$%0jE ziwDrb?&meG;XfrnTc$P?8kIF&?4uZ?7`yQ>KLdRUaIkilbAME;YB<5hq{W+}o}xy? zkI9@Lx*56|LNXpVr62u*eDx3>_9g17ju<=to>8$%XMbLQp0Hw19lA_@-Ow+l{=E9- zftB7!*;^Uca^fJI^61=H;3&^lQA8u(lL!$-nL!Gv2-< zct{VOP`|Kb_55gkQ!cqGa&iBzP^aTMUjQI+;G7~^*sq_`YfX|lb8&8E&0=>+*8w}B zZn4a|zyx_s01}7it`uc`lryNua1lEWo`A;(K;ncbekPpF?9I#BF_{B8F zdxJtE@<2ljRjTtQQFtD(Z&e?=ZY%v{jm*V@lkSmbF+}DcLNI{@g{V!M3%c4EJ#)P| zE*^eop?vZ}heXeCcJW$%_1PAn6HZQ}OYu7)HHBE&j!%TqRJy?`81W9Pqf61QX*4DO zkXEs;kDL4qvC9VUEv2{4S(j1^)P#qE`JcT0$LJZRP3`nynb-IxStk8#|GH8pw7hk~ z8DUVHwG_#|5k#VJKZpzYNNz!U?aMzTmPk0-|5%=ts{XDI&6|90ru;5N{+hb$F}V}tQqa~o+moN|T8R1gC}lt(4p6Zc#4ZbIR07&I zW|{lV<75xM%lB;Mb;{?w*F$kFDu9AB*I^pWWs7Qkje-MW7osvLO+vOY9{u*5L?b0anlQmjP^A6i%keYF{f5^@P5*DB_={brCR_J~05ePk2 z`H7@;`8SeQDiBd5{MRelh>vdpB(2M;=il`e?KRWuh0&lf1B4lmCFFu6_9+<;TmO|6 zI@|WIywOdJKTc*~kCLuiF_cNYNXgqz_+J+S1hoPd%J!oAO-8_k@!GlvFniPtjaS^J z$jHShrbEe=XRZ}6OJBV*cW1&;Ok4aB8V10{E-kWwUdhRV&K5#u=8oZc47=iU)tRIrx znWOO?CzDX2xrvvsXxg*owmkY4xoCQr)H&Nz!Z6maPd!dLWwee-<8!&v_n1B+MN)!Bqf2Kf^y7JXeWV0YsUa z1|k>cUiYPDu6cY&!SO9+3m=(LPd1lkXyKM8FkhChSZabOf2g=xsZIjY^izXY!db7P`{?wcMn8vT= zh>tB=E(r}qEPZgEgIX;23MP2X_v((FrfF-=ouL4Lopp}{1vQs zuf$^i(;&~3kBnP<@>10oTs~^9XEuHEwz+C zmm5fn@DUlG=Er^<`SSVtrcBd)_6T=SHAyJM#S8!3XZBjR>~N)9#zsV05e&!rs>v!|Sq>x+_(+ATKwA z^EjL|z)TzJWrFvZKSGM+DNdWHPt@Uy{p8EM9;+WH6a$+9b#&}+oULm!Lm{PJ$~Fm; z%C2AQo{^Ufn*G+zu!^Z}(NCW(W#R=sAOdu?!xa>aB&?MlEZM=a|y0)EiZHxFpwGq3wrGpBL*ij?Cf(@26z zUJod(@C!g!56iJwadxbfsZpEb>Lj9{Z1Q|(|6vxfRw82Gv-Z9|yUeBM>tN)xvHwC37-twk*kOI{^P53G* zie|>H=7lzGUbw%t*WH~--13ZgG|{Eam+Swj98^VP%i7lJDrKix%Wvx3c6a5>58D6h zXf`mM&Zr3|qn4e{_2p-Y`_F@DX=gUQ*Qt7mYgYT*@-fGM9=FImwuOL(RHpNRpsEu8 z^VsMOce_Ph=aD3jb(&y9oGxB6chK>yhh~?9!@`<{la6f2K z)F`-DAZUVKL7X(Lo@>kj5J{(Z5ZB_Bk*RG)a;;$q?6_vCu8R?6=OML3l|^oDWiM9& zO%DP5_#o!sy3}6;l;BytXxqq?4O`OE;F!c;rL^swO4%Td7921QL?h-NZ?3VJz}C97 ztlUBP-hbW&BurDz75l_}eKt%yd1K^V3O=tv>6kc zjt9?4_&5_rB@tLGV0Uo0dgw|wDK|j3^&dt4rR73Iwk5Z4dg&^I)#LLh*A~O>02)jI zWk69p#0xOW0D8{5Y|q4aN7K(JnXA-a2m*xDg-!yh>P#y@jU1K==vIZi05$ag zEW*%+HA7X)so5Wbp~7aSZzeK?Si^!D{!Gv>MMr`v0@{aRor=={x;qR2cP9ss^UVO@ z{C{Ob^@d-93>`;|EI7*{0cx6n!2DMNU)|}QtD9H8y7~WE#3=$BPO7R%yC5A=ol+y~ zOf$7e$+d`JRXe}TatsPEaJ{|XPoi-_>p?0qIDchd%<+6EfSQ?lnVM_gscdxv_EeQ3 zqHZ@(0tkJ2@U>;eHDKUYKq|Ry7!W~)k^QBKZO;f0Sm+3dE97hR262S~?DP`_n1ekK zXMS!Dh?7bS?)uhpsrqV4f8jp4WwrdnZO(nIA(+|EV9|zv%^wgs`e!8dPt-v4QU0A% z&YXeiIA=*@5MlX(-hJtB3_i+i%keM5oeDmjy|;kl;1L#?#&@rKSXcss+_&@i#8_t`_D7Y2gDD;VkCCEg_IAQSpTC)X(w(6MXr3& zeUx-3Zv`2;?V2fWu+jfXjK8A)x6I3}05YhYbbzhO6wPsexXynbjRoa7<7Cu|vST!5 z6w|Ar+Xw9)CuZV4B`J+M{qr1Fj&b}|J>+fh*=#=Aocc{1Pwo8R`2XnaSI=Cv>K==f zBVV-`*XUs!-RsIz`V!@*(FOMG;!6UYhcZz8wKo|=M8YV+agmxYEdMO3Q%*oqlAdH- zhBL zwwA9arMjVPI~cp~f7Rb$RUd&f6wLlzMoT4K-VMmBa$L*tP+^7;(qq1FcGX8fBQzun zHKh}vzUr4AX^kbG&in46F#svh2)*vNm5ExNL<6cm3;IipNFrc=^Rtg@uHYNePVK+w zNt6^6XR%&A$1;>0Ec}zGP)faxP)i_yW;p)Oo`Q&}UjHfjRhcYGdi9?o1!re5?f}4* zy}>(6EOy$$ftqlyU{s7fwSkF%HJaDu~;l%6$-e2N0WTG zs;_V0DXhDynNy7)pjZTA!YWOLbHWX0ffn2Vuy$vqmJJC*qt@;14bf;;NQK76~=51)+M@AITeJm?V7`W1Kj#dXLx zIbx@H*HEXrw&npu2F0R?@mb@sD}V$3mRtFLhls)6Q{6>a^@Q<(R5b<=a0%GuaQ`E! zG9VWaR9R=K83+b&@qfsB>!>)I=Y5m}_uw9aEU>ssu;8$`hsA=kxVt21umpE^*IitK z23Xu3f`=f1-~@6v@B8`6@1A?^|M%QK*d6A~R98>;R@YQlJ$&wQm?L;#p)Q*a=sRd|_VViSvuJI68YS4+(I7cg zBPc&l(m}9JLoW4TCB^o>=t%H30RbwA{uzH)lg>bZ}YVL zZ3mG$|KB@}7FglV;R`KM(yG!DyyxCb$ntvo)Ytfupye59b|KU3QxSOZ!ul|*AgX@$ zXbgTo=t;{wTXnrR<)_fjj(`{2DR#5G|7^0Y*{K4_I>49}njr9H#ok22;1a}G>7w+y z>*GK96=qW2k=sz_oXw!NDJ+LP%=4_Q6H=u|}Z7|Q8(`FjDv|01sWQodLNtXttJ{a1)oE3`Jxa)G5xBdJyQ zzZ%wMWO4df;9KwWj~f4q!9=((|F*%Brcu;NLU~m}ZJ{{ZGC_b)=hYfmU%e5bOZorS zC4^W=*2)*rHXq_fgifr5ghKz_U)lC>nlI(#gC!*6J~N2k4*YFqfrD*3b8!B3SWaL0 ze?-$)yo&ojdQ|yQas9s@fgIrxvG?=t!jbzfjw+gTXYGxSj~W!;bMsFXJSOi3f@Qr` z{vc&5T>k&$KUQg_uSU{TN(Me5&eR=m9JMic-8g|o_qsncN6pw^HqL*LdQ*w1cjl2b zhM}W9b0zu{B0)JUHOo>bmF!IqST7sy>`5V1YsAgCC<935%hC5@L7D< z2{N&lq&{Y~f@HtrYo8&w8dVVP5`x^D1yTA;>yL252oT)!+zo@_F5mngR{J*7iVj6~ zB+dSboh|LhPG>zy^$2lS#WrS*@(2iUu^ji;wdy$GM6Hp8)y`(h7D~?An{*qkb7N~k z0;lf{a{SsOla8oF466x#$mMk!(GL7$&8t3{J&E)oOk&?xN~)cQ`mVvx?c7CS{n-}b z8=Ry8T`wjv`+05xe;y0x0~|K(hZ@Ni&pi|Mm?-Yeh>t4ftY$1J$F}`EPIkN^cz?)`|Uj6hD)kq>}Ia-#THj z;0%#mVE>OKEiRT5jDPzP26O}h;mL8QlagjxcmWheLQ}l*MndXGZFY7xq{TDW4YKrY zbuU=**abeNlX&Pt1+nHD+6l8uX`FN)PGQ9mB^I!cmfN*Ti*4iT%=e~j)6soemh1*Z z1yq1hMZr^21^8=eR{=&kI6K2#f%Qx0(eH%7EP6e;t{P)*XIgrB#SUjUwkRD-x}(pK zQRzHH!#h?;GHQTso{cooZ5{1>I(TpIO1qy=|5P82Z@}>ur1L|7lUbK3s`OE4yXV&H z`oc!@tx;%Pm(7aeBpZJw%TK2*?Bqn9hm=wdP5_0Fo!Ol)8|HzLh0+rp67whmlKcA~ zq-y(GCWo)pUytu1sD{`m^S!>B&MzpCZ1&^&SY1Af247V%Tba&XwONdDWTdKbmryL^ zG~mHSerB4hLFg|CVz9ae5v=oke~`*(T%Q~>{vgeMYF+rvnSKHPd=z38^dc&6S+D+5 zwjQ>Tfm`xLZ#B7awne4|Q%iNXT!(-(#29Rp31x!8sgCLjfX@N_DQ=#QWt7`Z`@^4} zkrs^|mXtIoYplLr%kFw{Ts+KbH69O&R%iKi;Qt3n>%hocbn`y#^-kNyAEddqMLZ$K z@B&WC%3sl@DuTazWe3KZygfxUUOl(hWn#8l(K_w`d8QGv6a(oCv*TBr(0=!`v2z65 zCBugm-ofH^2-#$#)0H_5nlp{{203{NOti5xkPt0eut?f_kI=pZGme3qHi5D3uV1*k zZRMHaRNOd)3x;7;{uI?@W|Dj}W4QaGh150-d*Dp2&9PNr7H=IrY(pR9^7gt2o$NJy zQL7<0<8dS1Y6}?qMpxL}&)qRqr=VXYLEv>wHhix3Nc%2DCe=+krl#>FxH()E>DvgA zUX&s(t7fy7eR9?)^LkjtKR7j#w%y-&C&#L^n&k%@k{-71E5B}c+?w^^^i)vzxwd30 zSGMvX5q+np4pJwN1=j-ck^bT^K5(RNDqmPh4GJfCOzVuKHBYDmp7r%xd1IECn@kx> zg}>IVMuQf9ypcZk5H1pn4vC4#5PFcyF~_v5AyNdzlo)7h6=aR9WCaD7AtMWoW0xO-_r=cJcWa06`I6@w@2v2m&(xD44|+3(Cxaz28zs-L2hA8doB* zy*!_ZnUa?PR~`|KE?-@VilP#qTOo~IwMtD-6`C+g;UC8Zb*O5<$c#&8J}thYiOT}y zRB$KOF^|%^#1N{=$Cmr)w7>lD3`ndvjD`uZs@f~%GM`nsBoXEy36*P&1qEHJMGt(W zkMSX!ziILVhK$q$JV)i*W<*!ll}Il7G!kTsH8AG3ew5~b#6pcv2^cqr7P1=#QGul< zm>?auOnYw?L(a}0O?aZh#prplW{dv}uqS&l+id(9FDTl%m#IEa4&W^C(#+rZm0L;7!0n#?};4WW7 z|M@c8Tgk1g>$QGK!Ev5y?lK~Ofj3b~cU9>(&I?=l!-eAOi4;)4K)RFD z>5eKR9p)Oc(7U7RJ55gGl6#vuSrE_4LHIs!GYlqPVfWpJB} zZlNER1guB7tK{dsE!dY!)~{&FRmLkg*vc!Bt5RHLn~a+@ljS-bs{}xrrYpLBL^wTw z#OWm6>*hM#hQk-BHRwcED~iNkVEf&RGd&&$6>MqCW2@b6U7IAKRB9NN;dCF^=lyqokr zT^e^r8@2>*XiRCy9IA7ttlrq!{c;(e4964Uj4xZy(8N#@CEi_K4e#DzvvY0XVbbjK zOYN%a=m>})<6aAk=8Wkxa0aT;9DM){Rav2yl;~4KN>w z>~tteih7B8%0=|oCra2p&Z-*rOcLv4r#L#C)Uux}nWAxfJ1rCE~f zTcX4!dvLjS0NF z;|by(0KCk#iZb;=4rekQMVGddbm$V7umP^6L-{%a;8n(*mQr z{dAD)to+*{n@!QFmJ^XkeeORgM*5Xsof!SX~T%fl&T%@ZUlw|lfv$)^|mDMwY z?=DQ%nmb03P6RwGQW=>G_1?UCm-InPFdJ>L!yb8kIfk(G2CK3%oI9s2$R4h%Q)S7? z{UXju#d#Q9IK$$J2J22ht8H7)eaH8^krg6vVdbG%_%Z5wW;;4FrJcIAb;>4!5pbYy zV~_|_o<&M*(NW^x<*s6t_cVV8kRnpELCIX4MRi|@z1&@%IDb32SFQe=JU38V(Yz#5 zKaXjVF3Pk1Xj!jHvC6Jssg`j>0FZ1YMj&ZJKjo*T$VUD!o|h~tg>>SGsNt!y0=pyX zlC~2sD(&+)4@U@pA!&%M7?ojE8{Qsdd;J*h2$Y?*A*D(zP9x+3wC`$mB{qD{Za$!f zVbt9Jj@R2#+f^T)XfMl71WW5QY0DU;JT%w@P&V!3WR?!zJMlC4Oquu*f@63o@1s@u?q`5oX(0E(+;>sEJyRbQH-L8s%ac zt8960XKo%eo#ChVRr-{Nb)+P?YYAbIvwVNhIU%t0a|Jq#TX`?|O63O2buSMvSBqJW zz7ML|5MjMtcF|(4u9C7^jv~Y#QPk>9TcJfIW!cJ^tRh8|746JU=31V*m857GUq*a&T6-Xmoq<?;_ zMeM38mGR32qUMw)iHA=mKPplo9vH-jh;6ZAM_>(zEl6`vmvy8|?aL!G*U?PIa7O9X zn6THUgJ+QvcoTx1M=U6|$%1drg8hS~GINKG6pJ9D_9BH&tC>|zNwoyAdj3U{z{42b zL5&Sd=(BabmQ4(wIFH$A?~w)?a+U|&ZprOeteo{T6lyz=GRGjRS@{epR|OxL>uLm% z9f_S-MYyZ~^yBGPvrR5o?;`bbnPHCl13goX4Uz9&9pSR5jKh5(frw!B7_{V`-pg{| z_o3pU=+}Mbq}mS3{j|+l8vZ%S6(Hx^WD~BdEUE*0mUe=G#GW5c0LxCt@1+LXUF;|JcZ7`y$ z&MHHeL*cYHy%A8tzABdZ;`elCNpJwcPY>{~lgV}yhy3Oxx#?I61Y?e;Y$SH>a*Y0$ zCt{h|CQ$pV#b^-}h3lJKZC)bQA0&+4@g$xt8Xk#RupGQK`R6)!OAvovif;=@Z?2<7 zO9&T;+Ts|aJ3#&2@Js07=TP7xr{3GtD7|$~n+xP;N&@IC^z0H0fo%Q%;FV5(bH9s@ zG5o^HI833}f_ac|8m`uzxNrE;QT)DzjtDn{iqg2(7gNi*OHf6MWn4GMnk`sxsdUwFA+RpU5(?x_&j#r+ZY3F1IS)6s1HqSBB?kM$#t~Pd-~dOHYP2NV$Dcg zT&&u&L8LC-`}pb0#&M}}P7|5+y4w$k8%*Qky6v@tfJWE2F8wG(^0A%RN3r626qN}2 z9urNsmJs@WAc3V4pwn(NGYKSBY7BFG?xLbDZoAQh?F~*+n;yxVT`jKG!)4-7=r_Wk z(D9=yr6dDG=rF0?zhg(P&Gh8BLuWeVqP^eu(-y&`aG8yJG^}cJ=wD$QArMCJivU8jk`|dSoy%J zce)g{3|DNf_V?&a*#PKQJxe2rO1C$H_jieFmQQ6U4U~a1VCDKV3dJ(bH5EKnHQr5i z5NZNT@YwbhW1|*zW#e&j;4OQMLAFV8DhN3(rFOW4HVAW9-TdfcMmt8=U5`8^u`N#a zi^~p)N}7)N`MFVVhu>S?!ki+nlu(;d_{g|YaaN44d1xpLk2N@HR%;duG6#e>gHvZn zz|7uG%)3jaBKy%qK2v%COY7R5RkoUb%@o;zD`=-lp>Ib;A#6Dbwi}Ve{`R&lL8rL| zsOy-70jZ05v4T2R!tjcK@BDk3D79MI#X|Q-zxiA42ivcEeoE`ntqd>ji;wlx&inbTmgk}<4@2#+rtsYOmG_5$P*%?qSYzBH_Qs>{(OD|+($)3zmr@Lf9Mcf#j zw9;)DM>;AgwY^(d>GUkC+~VLH0Iv4@s!rPU$m*;1$b&_W7*1;}UA}?p*!eYoE6-Lt z-tu&vB72N<>ChSy>rNqG6?#>z3py4ul(D`_$c=hCHO!`zDoc9GSqv9blPZ3!I%Ukx z8aa}@63bZQW}r!VnGRY!8^R%O7=xf5fhL>J8_oX+11T8|Mb7MZ=@1wEhyr>8VE!O<%yMD0E?zdISk342ymxNoB1}x z8j`GUq|Dou!R`oK5zSJszxslq_dNoYFN% zG-Tg-wOA027N4X>3_8kd$-%(h5eZEPvTG`B3h=3q8QTIUzJ zpwEWdin31Xt7S-foy{!ZsHlVJ|;i@kTD0VctA!4*}jo9r|#`@8d>Xhwx9yjMrJSDY4i6U zO|(C;&|#8jrKDpJO%@~Iw(b8Q;mq{hhd(xTJ)T~pjH}fTO02$@F0K|GHO!n90J)=g zGZuI3XOZ15&wsBLrDN6G9WTC5(zOEDQC8gt6f*A0dl?qC%xB-*uO1v|5JsgD#Yn#& z)p5g9c~WcwY%Zv(*IunyvC(&jeOeU}tD9=pdA2GllL~-$%?cQ#MNR0`jppfrMWb5t zGa5f}f9aXm?~BRT8-ZW5ui!RkEG)0tfrrcL#Cs2T7Z@1Oc5F?~WM5;+eEzdbry5=pCVY$KmC@8lrT|{ZTjT z>D>x^-Lhq3;bE)oDYkunN*q!8Q8(2XD3ewHw;>4P<6oiwH2n84uYR1OL*BHET=K%5 z?5X2W0n{~Zuac9$YB8hUaa5frJF@c2tNJ_-`d0-*#8qyi$>=BTb#GqYy6)uMg|iK_ zh<`nV-ZNd(oK?7AAOAs;N36s1bAON$9;Hwuw)p>=B z-|r;B)~A#i8J%EZE@Sm}!~*c*g9mSQ6{MYzUmez$~Dz!2EUS%=$c%zj$PuDH_9s?eMjRlTj4y?cq3OLA>WH>#T*Gd zq?JPRbFHx3$JW!{eIIe6$3@$sc2rGo~6Z_P5WY(u!`DBD!^Cxw#GvxEwa!8DC%dgNI`PU47e5~eRacrX5k5PV>Rk8 zKO7<-;!ngA@Anqx-q{G5CQ7_+u|~ue37zSH%PdZdw$9QiR%7jRUnpMOb%uK?3^aex zNCltkQ$J$o==PhIon4Cr#iU)~iv+2| zUT&(-iAvj23RAUz;u(Hz7GAE~reP{3YLX}-5plUYbILnavAOZ$v59|ugt04D?@^HD z<)>N57;n{Q0rs7bdfzY^rPJfHbowh)FXCrqn3J7{YU2fD9?q%nN!usJa#ja_7lCWX61=c$6gFqUtPXj(iXZy)3nWar2(g4 zVLp-oyoWSk44%HYCa2TXk_Igs(ws8a{&WEKmqFxxj-(zc*v=;wkgwTkUW61fbw z9%OY^6^1tDWH_Y{y)}%gwF_Gy0}7Mlq7fq&=m8H^=-`B=wBxXL;06nG79= zAaS-9u%&+)nyx3>EKj|(+ATzGI}nHhi%C={LSBH-H*B~aOTKr8n@bK$triy|mgZSEqg zsdRsj0z^11gRO`Nq9<7n^CgIho6r8YTRWPUklU3W(%yDCVS+h6AoDTkNOdYZ znJVWaAK_THy!rO?if6b590ekX!4Tz5K`Ks=##=Kr*G<(2nK;%btpow*gYSY~pE7N` z8`b*hLz9KWTElG+92Fr}dl~Pt_n;_?jbH6V0z7{#jcIfluRP*G90T&T(<9!jBEQPL)zIZHh~q7f!LX=S$t4Ie^NKFns}And^dW?uzV7hNlq%( zb%*OYXe_lVFxf-Iv}sR#s4-}%II`O>_B$kmDZ`mOv$U--PtM$x4-VQ-e=7iT`l0?9 z3mnWUSF8{X0TA&QWAj?SgA&C&PZ>}1tLS9_TGs#r2hyl3xh%)>0_ff5z4xdf!ZQht zCR^0p+tpziKt@V?3`J@-kx3M~RYYz9p)F`w;{YfBXFpM6X^j|LnlJ*VkTSuOMJA4f z_2PNs0y(%D^oBW+QOxw593+8f(AuUqxUToen{%}LJ&bc2TuXE>KUF&4rSm1z)_h&5 z{fF)nD{Xd?hRk;@Bu0IaNb9oHl+KKBK`d==z}(fAPVDNiQP+GX+L!SV`!6CV%qWX7 zT&wEuE{|tw8>cGPMf-^7<5RffO5eWWr-c^X$xAFI&R@tV7O!`IKAR zHkec68F7{zSkgd`5B?l%Sc$g+EG4s-OC3M-ZmDD<2pv0CT5EKBK(qGzB9^a@aN^Sg zJ6apSOjo>j-0JF`;q9sW)&d_VlsZu?c6XC(l(Q?G4FW#*q!kJXw5N1V@w?&)5^f6( z9Siqen#tlC@Ob-u5L@o;W8D?74^CGdzcD2gfDd!JT|A~dYJ52&p#_7Su;HqP+Q7Zdh`-j?< zaU#A(JB&b7Vn+Sf@59H{**UiE6Ld;5Sm>6QVe1g8j5%)!aQ^2V)G135XWiwIO z(kBc)J0(YCPF7P=r#s%H=i|`RioNGxU-}cp75bMMKI@QVlJYYQ|LsJ^Ju2}N)b-IS zaUTZCwQreae$?2PtUL=v!cqC#30~i%^4H11x*;39&eYGtpDk#dp3MLj?S$Os$ao-y z-kxnmZRk9$seJTFfZ`p66K(bYZL^k2W@jVq7H3UUD+Y(5ay}=3@A=!y(2-2kl%E;Z zT0gvXfI=Yk6G!dfhTEd8RiDK(Bd!vmD9NTp$~v9aa|>fx7vlNPu+~4p=3+f`NDF zFski(1-~YF?p3Bf2LtuG!*@>ydHJ4ooZ1<8y;1gA6Xz1c#imf4NoCtLiccOn^=8Wz zDG^H$q2wa$w0Z-b4$KE6lhUJPcX<&a&Q5v5*4?@(bI%wx9UdMQt;9;=6uq|D-59D5 zo4tgo-n~(x_nF2}ap5gutae`PWjZ`C9yR#$_ZQ*KHAkhCIUpVk6z2+dBuI@{AM2ZrDULypKO*t#cNJ8mO8wbl@cB~(S4P@xjqg3Crv zi6xtWZwB2lhN46TAViYLn*X9zIBT7ySovi5mOX>%WJPHM=C^gaC>4<$xz0`HC z5)v(E%Yvf_8)j0{`_z0z;xV(mGRtH_dUF%sISwdK07^biOQaFUS7L-$zQ?lvTmTS~ z!ar{t6Sv_y{&+Z#54YQw*X8X9*Li3&SM$XMW;+bD4%WR^Y{!?!k6~{f7%X-J%9`LV za+j_Y>DrYMcHSl)N$u~4ny<#0b09!Xj5~Zz2|AWk)kezcE{Qb~jELtIZST3nyFJp}{gO!NAk{MGi zVaa)CQpv|5+R7AwoIB$p34DD-I{H@Yi}E|G&V;iNrF<%D+>D>bjq(mYimWD2Mp0Bh zZLoi2O1srOGh53xDL2qUCaV5!NYN^PDK12R%4Ej&2D_c4rH8O~oHmxjJNl}PkdRA* zL((Z|X3Z2tr-j4gWC)Jkzmf+ZT)w=CT64DZw-$2ZRmjhzf zgnuWWreP<6&1Ok~Lw87tUKB`RD5Xr6$$Z`sQSK6MuPoJITyxtL#D!Qp&;PS_4(9QG zuc}jS-k!KqWJ0b|du=rq+hm-2w-Kky3WI}kqNH#cMw%^jB!k=xM0EXr^utwMqC^fT zaX7B~PLaW8zK~@A$|uEa>%ckTh?UN)a!mN!3|F)m8IXqPWz=M>k!}fhRe4I93`BJ= z^DzOthE?cGjJ7fd%dm6+GmLQt!BCx zD@#M$F}AhMWw~Q_)M6VB+AhTcWg-mlkqyc_K`Bp4M3MX7rAp&uE6d!^B?}Olt7OX` zJ%wsaQsO=%pqm9OXG}?f&OZau`d@qiI)(f&(>GR*3v2z*4$OUqmVeTxIu@|A!qj!( zhw@7qN3h0?vItVXC2_0zDN4k0Wic9*vydwdd7I)j2iB+a+0jT5%FJ3Xbu(jazZe46 z^7b?a#IKH$SgTlkrs}%fy9mUr;2*;@j|^=Y<#HrA`{Pw|T19gH#c%B~h?<{nHUtY+JHWv@&>t z;hr9>Sy(vN0f7Jv4pDgr;~vg>Nfn}C#~6q=a)oY2L)GhMAM9IsQ!YGSLAk-{8T)R=d(rKW}V=6DHI)=au6X)a>W*U z{8n(E`D)Yz_{w?h#awYa>QO*Z^5}Mpi};SaG4>a&ZRy!6YZ94uu7ptib@Ds8IPc>m zkXr(3cxHxv3}pM@COzTYdB9w|m!XhZ_~gNyDZtgFkAD7KKQNSZ@ih1)<_2HTQluCp z5Hm`Yqv^sipEavWe^2dgT4E{^|ESaM>N>X+3A!DR+x>{HSaNdgQVTF?#8lQ$g6XE- z+{w|r)S$-tTyOK0SSu%#vrK=V>JQRQjI?fco$z$KkeZ+DK-dEw+QP}hA>}R`4~j1x zOB$4tlZ}^pfe5}d6120@YpW(`yG>PyxBUs~F>qD6?c7Nc<&b*=(4FBHvR!{}ZeS+V zqT$sBA>5nw9r2`a^M+(i$x}(26?v|vLldAnkh7?as2gJD93fOH*7nb1h_u{}eO*GL zYtRHe1fmR6U3Y_(t(ixJIS=_jU%;5j<5PDDTk213FQUdx++=DW@ z&Y8kWKefOARk^PD^gX0V;`r@vyeIOeJ~{02uPmo22y}G$pEYygaofc7Gip=5%+pI& z_W)6c?%JV%@E_?gZE?0~9Hj;$Wng>OP#F@865lTWn>SkqYw+1hbq`8+0;y@pO}g_? z(p0&hQyv@lIa!aDNY}YAJmi8JqQo#lRI$Qad_NLi(B$c^jH45neNdSDRxnEaHVQ`L zwf(F_M5-5TwH|ic_ZkP}x^^AQx1)5eOZY?(G^gSt;30buod@BGRw~H!;P+{plnt;f zM&=PLl5=`RshvzsrV*b|^_HKi^c#-=zgq;Qc6LYupS__|c~6S!53EXt04mN`PKwjNvm^i#G-__&Cv+nF`%yyIY1_UNNit>{+tK`j zloNpqz)|mJZlqrPF5@Sjzf!NjHvD=PKO#kLtXg@|XHVk2nP0_x)Xls;hHWNpfvI|` zXvrpTxz;g^M(Q;w(Cf_KXBR6A~E`>vS1 zIaB*OLiHGtD}qe)#CI!?)(LqzS^&2;tKVrkcQB!%L_D#!kio~Btr&(u%4-xmX3PP= zqq^zF4(^=hSXVE~R{Xja5cm#T-I6TyQix zo&>^o*M%^@9|s?OZV0ac9D?RrAlC{viZVEYsRnNfH9ov~6~c<=;sUODU3|m4@J#fb zj-M>}MBhMuUiH!A-=E^jj6>!N1rumrub{uvY3slyPybv!txo$@A?&?$J%9DVM0=?wel{~9NteFP6 zq5y(i!rjrjG496E#q08C=(D8GAM{mt)a!F&ac&1%5EX!N;~t#6eVuML1wKquMYVy7 zj?_XQ4$+u$MUa~)H(wDjY;S%XC<{@LRXwZev6Y?t#I__0?q)r!H(y2!Zm;N&^XiL( z*t!Dlxp{#qM?B_{q;*`0nL5F3>6wCN9IIHP?oPYZjPl1!RR8VebD^$K(jJnY!S_D* zflJs;yo;uirIzcOKO#$yzr#8yl)xE2R{3l6OaN;=V>gvzz@QCKpiHVilkxNW>w2(i zz7-~(FJZoVzFYybX|i&`iZg+F|9f~+UAiX_Ih3oNxG7$eyQD1SgTxMEe?dLWO6$2% z=}-FZk1i_1^}ko~;tQd3Q~Ng;3gf)pPE@{Cyv8kt8d^F8$>ffz`J#e%bj4NNrEY;Q zqTupNM%hh$7gSUhUWfTnBPZa8N#2;gd?dHQvTp>_qz>}};aE-`TZ@~d0sYnmD$J?L z3~9ItU(UY$^Rtr$=i+bB&;Bewyw?p%u(}0Yer)_r*4ZD#pBpaol4K%1_@f`a`yhUE zy4l>`fXegQ=NR9MmHCru)3jFKKFl8qYa=)%cz|EsSGXtRn3r{wZxzJLGngT;zvh(I zk%#r?C}lcXXZ}EQuW&7$&odnXboFb;N{w=`Jc?Dc$ z^hRNvu-i$(7BMf8hSC0Src>rgu+EB{Y-PI+GJ&l>fTFL$}u{0Z@Xo3I_{|D4!# zp-hhpxaJ1EjzFB|J-)gYk5xnoxsv3%j8Q8{dkh~V`(vsC!~h?)g?Khwjk@i^$X6DdEQ3d z5kwh;%+E05PLjZ8-Q4SG23y>TS-#iTabYAmkB5#em~5Fotrrn9?v;dyYy?J)XHbb) z(SKY2Csa+J?%=!`gwfEcn)gn2hFr<8hPa(j@O$3$Tb-C0YsfP>#|pL%(249$lb3H& zQlY0vLBXU^nZ)9|Ur{7(no{3-)9>*M$uoU)Nj>tVeL2fELpB>VYtques*=_%v$LcD zZNx4uv8EITW^SW&Ez@;Y{To%466`R>^<#-di5D3xFoY=sa1rfOn( zp0mK|`1?b{?#TC8gNcZ2H5DS0zoW_2X%MvHj|wl1%5d1UQaEw)HS=;fuFN)Lof4_h zw(&WYN~~vJr1JG8;r%we@N~&L6Q>hvr}R*U{6uK6Ne(B$Hd--O60XPWXi0M75iCh| z{k}{(?>^7%kR>?jN{$G3Zh~*#zgQ)rU^3S9ojWChA-24KeOJ;}-GEqn!^Y1d0CAb7 zS<=X6QlqHO`w{yb6c%j!C2VgIu}wG2t%v|~!=K;$hxXgaLhQHumA(mItiYSG;>2GV zcvt$*OGIjoSsI5()L(tMYIeJGT!V)u6 zrXRZP`3l$O#-;2q0qeb?j>b}z40oJ{|F+>gF(^WU3vFrmQX4$yR%MK3R#DI^l~a_%okxoT$Ve`$-t&c`(DsM;|Us%HlEPMI0rwxB~e14p)WgegRB<6v7xPp z@__~MGG=w+EK6_aK;etB+XJEluE~Fouu4}Q0xL>ucZDmy#X_*%iMn*ik16?0=v{sD zKCY`3STIaq|8iKyL9aPff>RQ|aQ) zzo51EBbREbU0X-!&O-S>@ixWA_%w>lm&jcPxX76lKM^rj;bPmO2BPFGr~Q0<7+a0C?G4;uE$EVrZFXU zjmz>df@z*0&CBhRA5H)8O(|CX&;)0jXOet%k3aK<{o>|A$KeF_Cr>_&= zDK881)v4E+6Ib<$-bu2gI8DEG^Vj2T(ZO{qFMjwLyp4^@N`K8`Ypc^-)0qX~M%hcR z_L+<_fnMx?;4oX>*RA}Kt){(aVz#92ieNMs-BG^QCNd%Dv75hz#Gf~ebc`$=Pqe2qL)P;l=pqnc@xwNHfI z3R`r-G0FxbOhyOul1FR2=5^tT|E^4UWa3T#%LQkeil^OKQ6=8IulVr+;GoEHW3aHv zw3FvUKDWS;;(2BLTG#CvVFfHoFru~1S64(w@lqRSGE`p{EP>>&DY~UUP4vWd8-agc z;cP|odB|2?j)M=ED8m3%K?T#F{;OuT$@d}cVHAE?elv`bFmxkl_+Pit&j9dsPd^Uv zMz75w*;Vc|nUlna+mwJOSjmFhM{lP$(wJkX(_9_U*Emh(-^by^5FE|%&t{fm(82j& z!*;Dd&)*{il6KucZu;}iz{wzN=8p%8F$@DM*C+-H$=Q84KU8DJZ8~HUJ6{bMO6aF( zJ~53dsr^u!EtbwrG^peip3NZrG&LAI+E}-jS9;(l{yYA$_Q><>caoT9_QhqOdK{qx zy8GrJ)^*3H<}}S~Jy!Z^W!@1)+W-69fr5eX`~y(9ER8znEXicd8#QW+MhD@8Rl|4b zn2MN+W9Vi{Xo$R(pEv;>>a&78&oo}F5U!y6B}gqzUn9EoiWQIExKxbtxeIX(osvjz zbOSq_*|Yh}2{J9`>{f93`*KVOrXw=177Oc|+T7&ET7r0#f^#v=T zTx#6#uYC3{sBT`*i1=catE0o(x0K4#P(?LV##RZB=<{E1srm-xmrVgzZ=R~^W*XlO zsfev&)Nl#sCp;UXkr~f@gqhIlS>J!tPMkQR_>}^BUSY6F)o~u322JArke(*}*g)O^ zUQpILu#CC7(oV#7%deULPHQq^nr1!!v9E!lEg1p3dCr_{YCJ}e#;Z}uGUbb#w~)=Tx}8gBTO`@d0Af@ zGd$UuNt5?Fb9#GL+=iTn11-8P?E=MnDAlD4|0Y#faLpS$H5iS}7L;l?`_bV?tXr~m z`jQcg<-JMI_2*=_*St=FzM`{Zbu+VCTz>$ra?f=DTW6IpK=MU=Syr(JRpp% zlu-u;RW|jdr-KZjU=l6u42u&STA~0dCBv3l;3ZZH?efNEtlvph-9fwcS{*wR{8ojy z^plF)wc=ytP2XQadkxvY52rvG9HNz=rDle^oQFnSEQrhV1(CXM@5{Zat!U$63P@FRGVp|LOYR`k80i@~>8_REc}f;OUBAT-i`akRB(}67u2EJ!ZGUTrU^iH=k)D8WSVA0q}>3> zujipC!n;LBNr_6O4OXSIU79JHpHg5AM>IR`%J5x1#eh{@^U}BeBq?OXZpd0<4Rn89 z={_I2Du%7m=(M2l>p|z7lapHz2j;xG8g-S+KDvvu7d}5z_8d;@GVk|NopmL80oi)h zB$n!$!mB}o{=C8$88@4nlWt}{Lu*uDE6MZHmfCXAU+Xp_ws%W3p^i95q9N40c&yTL zOexZETXn1m#1ObOC+?qGX;J!u#O7_nl~qw-w#}?>#euU9bXacDA!@~Y_#+PR= zZ&P+ePRJegl!foU=Hh&CJ10SBd8znS-4IPDZ#r#0X*fF`?*0oaFUg$++W@R5c$$<< zZtiBRpW#af;jwSuU3^ma?RUHP&>a?`f49U(o{F5u!o_G`(Sx7mXFZ5R-c5Ts)OWh93NX;8Ym z8zg1u4i%)4lx~n3x}{428Kk>Qx$$G`xxepy|GIwz%-(CSeXL_`&fe!< zYc)tki0y8ZV6q)b%j^Jx$2RwNw|*GA7T>LIUndvl*b=oadN{3w`cqmDUrz0eA01PK z;wy}HZ#UI|)I@dS zOZL5l4@l3dzbr<4FlMKD>3?Kf69LRFL-$XxbNHmCzRpLoplMa8T7Co}3!Q#% z#SBU4pSqC>4AA$MNW^H6pCMWX!=;N&*gIHY){bXU?_11;6!#2or&WyS&>`4d~H3IJw z6NM$s3=xsw)q&3SxELX*&~%$;X8@h6WK>>CM*J(|hhQ07?A&5!o&*B)ugUSgUz1`! zj(98Nm5-jqy#?b{F@=KTkPp*82K_#|1?q6?eE`dq_tAa2j&6A{*jxHk?WuIp;+0SZ zuK-yQkl<7GoB!u<{JPkqs!i6~()P3#Txd6b2?}HR*k?4Stp^;94kX;b=CMMAS#A|#qa!8d&*Nk+Xm}jyh}e(4C{k>Mf&KmA=Cj9E7_(g6ThGJs@{(P3ovu5-l#>ox zerw$xnQ>RQ1P4?KF#YUxrkt|{k_|Lk$Hvh33J{auODC%HVYdVVnv&SpSJ$SGgV4;G z6nlj@=05X~xiV6hsdazy%Frv>Rh}&_V5W*o z5%SrIQmd70L0dc!Mm+(}%L=KCxP?G1BYp z4V7tDJAbb(z{mJrZA(`bv$)I|`^U(2+?U)eD4%2a;~ICSsckEGmr;2x!LdhuU9>j-^J{`1b9JI@x{4)3rP)2G+i~N%H^F@P2NKgoV(5y$vd;Q={-gH3*S%O zcDwotGsN&I@Dhbh6ORt(wteZ0x{m?^*arsuwn>8`@h3-HhR_$#1mI_jSLiPKQR$PVqq$$cAZXeM)m*+`+$)ucU zcBkvibgQPH9GMMe*7;S5%=qUjfiH`N28a1@J^g8vKQK5Jp7y-7+>ZC&9aEjx>j zf>rbQUVI_uy(GV?gp8Cx(!dItApg)9HaWQH~YUtF; zcKBjjzF<1qTsi#s1dhqbf=>Eix<`}Gzs=8)wbSfDxW|sBWNasf80o#}lVA|!dqAA=k(ll!hp0c8_GrL3mT#~PJL^od=TtUg65nN44I|EFc@DImeuWX( zBLsOLT`H_i(@Ica5w4JEmJf2$fT-n)@Afw?sU0)O#>6wgs92*M#Ql;IJWqn?E=x z_3mMB* zPaqcNed48tH^_#GfV7P9dKYoEr!|jDy#bEeKlgREjn1-aCe1ce&uw#R|51{aqHZ64 zthBVSD_FLC9;c_&#bl$Xc>1Fjt?faf(u=pszXYprYUoq0yrq(QW}Z2RXj^X}1r+$o zm`TWj_SFls)|pV5B?7kO!X_@KDlFF3wGn@h*gE-jQrq=}p6|sIrJQ`r-U@s~#|)y@ zO+Y*hrb0mr4qvDJacX)=e!IajH$|q1z_}(V7PZxo*MQriY^CV&f9uDt5YSc)s|mi zkoqV};kipSl1J4k_Z{gOjPW|HL~%1H&tIY_ilFvK8T~Sxo-FZBHc$?;8nR!DV`Lg> z8b15s9p#;Dltl+$Ys3xkvH%W1gwOq(PVi^P-5CmFDhA4gUxFBdV{-&%yD-*OIH;2p zUJt=Ayo%Pp1%)`)$m^qg0qayG!x}4Q^ue`JmK-4U{Ur;PBJiXLxQkmT*`JM;%T&5& zMNSy?!$)bccU+>K-yDSKdrmoYzAj%KckTqzz*g8_jX19@wV&iLjCETib)Ij8O=8Fr z5h>dhGg~@rb2jGbiFtGpL_`;U1y&b<65-bGCSX=Yk;wIXc7(_a6bBF3IY;A?pt`{R zKVIfCf^#KVpuGAS#-rjYfg$@b_Y4a>#Kkx>)^Mg(@8M9@vyD04SX#6%kzG6vhDtD2 z<4-6O!>V8t1=5;R^AvtC0}M+aY<4ld1e$rXIWL7Y-G3oK`JkUg)bArSmWB>F!}Lt= zRqbcNJ6wuk{BsqcPu_}V3+6o&+qI+PfeqYgOgorix^C*F^kD4LQ~7RuSC1Y;d+(;C zdx`cMB4j>VE_-OkN#Ng!@ur zc6!7ua_aLL&AjxiEIt2_et9CNp-W&?4K;7Ha0T`o?B>!iD8qB+iR9RFegvo)BP_W2dfe(5;V>doO7>HdY|+bn16E} zg;rxr1jNpw9#4qBRBLo^^iJ}Ibx&hjDrTW|ajrA08`#c&qQyNm-W`15UEe$VaENww z!RUvkL4lcjo9>cY1E%OB-RT1|L%j@r^i*AY8Q$_l`xcBD18`wpv}2I6YWzr24-m0R zQj48>yF6S|gIm#2;VT9fB*g6fNGRl`zcotFoGb0IomaeQQm5u{&&%)G!*FTx~| zPs%rZ8x&$a0$m(vguN}>tec)pCx*s2?K{b{i+0+zgZncC$|6~gWYC~^280(8mLz%^ z@rqnfMP_)h+}6V4`=}TtYEg)~}GbUZEUQs>BV*;cH< zbv2!(nWZ27SBL_2HEYf9Iq!K=LTC&5L`d|Zm#A?Zu{BtQqaJ>$lj}a*twdbB6_tl4wLTA5i>+ zc*gm&O}ubsvdumGl5?aupCIS!k@!}sT6k`@8=E&^UN5fM-QkIyzNUG``L(!h9(xrl zshL2+=9HGd<)J7Z%t)DlVg7U9Zb9p@Fic@0>5H8xj4a-(Mxa#1wJGZ!(u^ZdRF)`Z zPk*Um<4{ReV#@m2FJjUbM=1u4dbb>c+D8@+?)@LbMeD)R$>tSA*;~5AwF*#qPy<*Z z%@3LjGLk%q!PyOHhYgjwe%VP?{JB`mxCqiNv=GYeh`&g z^nI{iU8Gk&F0dygzE3Xjr=R}7$n#FOUw(V>W&%gAiSuF?y?fm2mbe2hF_^~Et=tx# zDaV1q^wOEqea}C2&!FCIi8jcexsX8#jP@GZMq_G~CmBiD zrNSk#o|LLfS8THOW-5yHsu%?fZcBBrnm!-Lym<7-*6!4%vXvp|EuK_@D0XYTqzL3V zvm=dXfj}#Q-TrDgRFfS{Tu%McCo3s*E!;0^cH}}k3tkkD-d{G~O)^flC2Ozoq{hb` zcxGrXW-n)*ObAa=r}iqHi;iG`Q`4TVh943(B{K2iNV{z{xy1(ujE}WWqpJzqxc{tJ zF*#L{9W(QR#@TjcYXgrv@~SajFUh%t+a3{25p<@1HJF|0G~gl#P4go|n2T}Rb}T}e z7s~jFR6H%MSgj?f{FxNUccb6jU5}zZzq;n2HIHs>;3=Fo&G&+knjyL9=?lm6IXU@S z23Wxmnh=`zgVvD3yqeetbd4pea!mw(iH6W5cXifuF33@SoD;Hr+9S&k{fq8y&K`xw5#Ghu-Me%R~LXmUL z?rpu?7ECX23#Owp_|jV^*W0MT0IdP#iR`{6E;DRYOYEH;=MmheHK0anf*Ih@<{EZY z$SCZ<_};X9(okx7XL33_DO+^`#&=R&flcbNY?0Q#rkYMeGV$6KqtQYti#I=V5Oa2l zG&P?q9Ulr^%bR;N7rb}>R6tJSb8)Dq5oJ)`nm)X@ks)v>^ynHpzh-r}>iWT1qK>oz z+^1QlR}K43%vi5%<4Shi{Y3(EsQ3yM#e2i(PV(->lBb;2%*nxawRz>k3<9A++kQJ` zWQC&KeJsXj<;#kUYE5#!W*c)yU%rIpKW2y^Z^t&R=Y7D@D+M~VMVzUa`f4T zWjAnaStBjt%JD6C_^Uy(iZuX`3Mzu7qezDPY_h>rn{f^Dx9a%)H&gi2V_zdcd@8WRbUvDXRQ!XlA^FImCs_9XnJr3tdPVYI~7dr^#-?T;bwJIHpJdu2E5)YHMW<|iM%}aGG z5$*)&SvxBj1b~1l?DsR<>06c9^ z#GC5|xz(dX*Qac|uvAIsl>sx765+mGxDZAExq|E9Vxh zFN7qc%qPV>sY{)zEw2EBYar(f4o|MOXBSt4MjaI!m$H?CnH`q8aLVO-tg6de`atl{ zXSCa4eVkb#DUync_a4@!mDV=FtTi-Ywy^~D%et!HAxQ+&7GIxsnu+4GGl@bjW&E7| zCqWe}D#@Ky*wqEL(M_Pzs{)BRi|UDx2x2w(d7yDC#KMs zZ(k8Dt1tVYZmV12@E}{&?bn9Y_d`qZkyO^_6I#c@Zu&(y#JX4PgZea~2ho)K*$EZbr16nbGBnOSv?v1tRe&g!)aMDP*#O z0#S8%`9yfYES0sRfFGCMx$>mSlvJ1p%oC*_H zYZO`7Wl&gzOqqUTfrx+#J9TF8YC>ukk!vaaxgV)H408TKQ;U|c8q5raJ>T}#7#mWy zTMKB-Bq}VAwdu3bo!>laSG`R1jQ<%vgN4yG6HpsEPu6fKH{)h9c9>AcF9uz;9X$&^ z|3qb@!}zqyriAx8L-;}&joLBrsd#xitA80!q*I7z4Ou~weY2)lSRI z_hJk=p9M<9e;cee>HpF4r7jX~1gxQ&gpQ&+;?+fEDLdM(N3Q-NbEGdknNMd9x4bTG z{p?osp*WU*Tq6|e3i4H7!5Y*k&&4r1a}umB(P{Ti^h4uGyY&%_uehNFn@xc=R5L!8 zLBWn-P_CSW2=+MszP(R))V;l`u|_o)xtcPWrEWxeX<5k^R0PDK8s9P-&1J}E0? z=w&IOogiN9twFI2!lI!G7GFrBVgnLA2io|Wcu%1by>ApZRPO4qy4OTK-b`$te!*5C z&ZjuD$%VR6vO%8~f;~N0pPd}=(|6q;r-ZTtC}0fIduHd=;R?Fa3`OfPIpId<9T=UpafVzx5{$mxe1Y-o z+(5Vs4&=ul6-qa;@J8@GAl(7<@lDr-My}#=N>gi4s#p*YO z%+NxtQBV(u;XQlrsKP?wzFr{iq=61em2)&!;mk9(=sZX-HYZdfH=n*B*rv-+I|-Tp zOB6P6Rc(p!_{9olcPsaij_G(dFibl#7~`e13!f#dVDDzleJNp`Ub!Iuy|T=?@% zgvGRE&Ccpwy+=!EDXC37P_MG@WvoVHR8IM6B0oJ*J*Dnd>HIBxd_&%HV~TRTwS3y(kNRp`_wydBN~_96 z3`O1qe#~QXQX%Ws_0S6wm7rMHyoCf&wR$An zneRd%nH6V!AS^CB&5GERx{l$4D@i81`=5|+qGY+A-_tLuL8{76A!=&e9B{slr(Ak>d3V+qU}WQ@8zl#&Yfw?k6hWZ(-x;VsTw4Jau^ zfld`A=g3nACBBNf7uG0m%9=weSNEHXXN@6gmb6}W!@DPV0laZpd1QL@!uE=e++X+| zBLCG9#cY_f)Ks1_fH%O;%98sV7~3zW2BEPy6P_e2W(P(C7Ukb3XgV1LfT%YDAc~ZQ zWO$O6qiD@D<#cWsa-VTy2r8;kiunuU;7KB`R*h4$k()~3n_r(?|-LI zP5IvW8f{OylA0)&w*qt9-aL^5nP%3|e2hO!U%Uln0M8l&&w2loX>$cJ%b5!!FWY30 zks{&rmWIWK#87l~#K-;M$H|H;77l@Nuw(1{95@y0l7au5BW{$u$p3 zvhwUsrRcIsm>-zg^cwFcvOf??D$si;v_f&gjr!{#hS<6c^(EFX&kOJe(!?{h znc%ir&Z0e0a;y#4mo49XY=Kk31*sadaNLD7SBR(jARmuuX>UJxZm;CQT(A{WIjUC< z5zoCvKQb@w;La*^X)$4Dg+RnuO41+loSh}xrKEJ+V@>D5y9H}yCr`OpnoY6ujx>PD zsr--KQ)`zIgLI0xooro*?FXPoOYQGbfC%a2E?o;GT3^oJ*@pyt?-S~J_K!Rm;4&7% z@7L{*SJH&)fAVUtypswz2U6%53yK`H^THpkq_KDiG97m4+pu1M$$?qm^YCW?g8p)T zuCcw(iQDqe*}dLsfek`wHEu?p0bBLM?o}myLh5JFFQ)lFmP#eE1<332*rHb^>^+xx zr0|J#{Qf54ZrzpbYiYS@;S@!&D@wH-O<>Xowk%K@`oI}TH z;WW*+Q=mmNNe`nSajvDrQBvS-$NCIs2^Q}2fs;3H58tiiPuSP7l=Saw6uuO5ES!$P}oQCaD z|Hh*pv4gT;cV;$OkEt_4BIx0VK6S=V6LsUzcydK;WzmGeLIlN`{fMV zB#@ptxl?m(*-1-C9Gn+JW{b3~`V^65glo4o{$5Ay;zN!cY+{e0b8G?+Rv_`wO5XFs_Ex%Pu{ZK`Hxco-{vHy)4a?D?{QW-8zYxfm zuDh0i?LbtJu71fq41M{bH!hs*oP5c-ZRuqYXvO$#H`?;+fyd3p`vAlS;WH}CH}`Ws zGN_WzErX|s1dRlWa==3=J3n4Py7WOt@x8!;_fnYwG zG4%0HpZRm~PT{0#!z5(TDGy75HC1kWuQkst z6KlyN^jxJq>;eXYqQp5=GbiJUZ}6^$aIEpNplqjdh}f4`~A)@czdpl z+9j<~4XxT9xpY5E5NeqcA4~1_gxN-vuAqGs^HsBUV9InaB&%2q)HhIgR88@idT?1_ z$hO#NC^5OlU=kU!8a25wOtHwg5sNW)j3wm$oY`1?kG^^zuOrgRSMdjV7&+e(zN$UV znO3q)5K}cTF;66RA>;tzz7Ug)W7=8YGczFkQ=r0FVRGQAY!Ub6HykVMsgMw-$|4Im zq}q1W=1BZByL$Lx)a>h%Ql1kbEr~J#^0Li|(}w1mUuu$OeY=6z z$h4I_=p_xcKMHyN>=aFN>WHI3qdxrKt9_%1+THPdKn3&$E;hL zrAJz}r(R-qw<-=~XSTSNYh^hFTq;zJ}pScvA1Pi zR{5aw<|e>+2X{0?7XjVV6AyL`EQXn8E@3dXyH75Cp!!tnY9^?+yZ&UPca6@{_SOA- z#Log)KY=xb+ND^b1*6orso*8faa}p*aV8+dg}Q5ZdF6Et6h<_fuuHMVsaGfOvdjn6 zN`4}iy-O+23FC>P$1lTmtn}0>zRg;X1LF@XY?40=x zWB64&r(s1{BxX?wZ{~t%_FsIM<7mrpI3$B7akEapfV*@XeX#X9s=5?gT+;nzfDtTe-}Q6^eR^y$FdH+udg*RLu;N&dzNL9~gF*TW3GeX4KQ;^GD=d@)_9Y3z{NV zCE4o89qTUkwd+@tYh)ixP4Qf`>m))SRHuEl<-E5Y>h;JUEv73+BX04_q=CB|Mx3Z> z8cd4KZ=po(%S%0lm<;W4Jbd~jYxbrzhlsMt*pKSD%dC`Z7-ZBG!^l0e+6jYia6}7N z*?zVb=x!K3IoT;}megG(Q#Olbvk-A;c$2sP%2m8S&5+~A&L?^lAd~C<`wac#$G~4= zVD&O)EIzH(F>H2?$Ia*5EtZixEqSDRH_9D&+tt=2_1TfKXkjmAg51uy9X-T3lJ=+` zVJ6@!?3fKHv^;Kp5Is%(2=>l~$AJe3o$CGE&FE$g^qbvJ?7ucsR4w`@1gXvK_8))vosSepOcw~KZnb;nRF*26r+npfccgM|yx6wM;K z#Q?p=77>xvmUo?YYBbg@GyZB^Vg)_3o6}TBesOw9dDw4M(dfFhtyB$8*)|Q%X*B^9 z`B`;bMoNkIt$a|&i-vd7LRC=ws zE4XPZP11BuY|e!bI7zjKTV5Lp5OM>&qmU3QDuy}6-9dy!k`HRe#Gi27I#J{7Uo+> z(^i3+sGhWeMl8KgKBFk~#vC9GL_IK@4}kTK^EsjQZBPR=FlI{;^R!h7{xgDdRAa#? zCrX1ZzI=4@fHcsBpzc+s7JXQxVd-BS!X==eFk~F3ubms+~yXw9v1?!=GfLQL(?n&OP6?tHe1LINnfZO2g7*+R&FNz=8y6bppCFwSe?|Z&&{uhGAWTlX3=B_t)l`X1S z;=2bs$G{4(GxX;h)ck@o@@IG=R{KEj+2jO5*V-^^AXa)7glkA9P6 z@zw7viv2v479-#+z{* zd_eLVKxpYyl5QVFY7_wR29(&*Yd+WPjdc!hS8C*%P`ekRej-_GFyZl@aZ&pM(f}X0 zjPV*u_6!+Un){kaEgf&2%Nvxz8-tZlSltwf4p2up|RNP1PrAd@Xj*n zshVywq6VtJ1|hoAItFlB@^4Lx)ZE*f1Vun@(>78L_{KyLlp4w1RC+OwGvd;e1V$OE zM$N698}`ZSrVfaS;-jprW5^vDg)syibL~e#lGkhrPKIbrK!F{oW0jN=$ohusqM_FG z-ia`D$6_#`F?iJwO5{-{DA@@-`jb4-ga}V@47s*0Jq-U&4<#jRs;Twe2+vjom3NYD zdo@n1epDQ#k7#kiALE+2sU0l=5De+eJncI@RU#1@4PfG$D3pHk54edHtI7T=s#`CN z>oA7JiyA1w$Vi0xm3NeDm}R&Kx)Ci^g`4E>Wb9!o$P*1!9P9)sV8sIUqUf1k#?iw_?xobdp#!<24G{bdeapuFc@ahf~yiVDTpDDS}J) z2Lm@`*9kW(Rn{-L6&LdD7V^d{pIHvQ8V>8!@7A5bVRt{Uhx<=N+%Lt@=#AqX#I(nM zwLiZJC}f^yK9k>Sk^WuGLi;lE9jPvd)a{q9YGvkWj>wLPB+D9#wYDs~`!-_+=c8%> z7H-gx5Ld$7tYDX-dGg!LkYT>3cl2+4Q8EAmW-qGcbgl68)+8oLQ=-b?M`U4SQj`4Z}nMM_YD?(7Fez& ztZy*^FV}TdE2>QxT}I9CFv})sprFUPoX@LKzq?!48I?h1VK8cQ8rtO4$6!x}X7#5L ziRaqAhpz9RD4oKM{3DJvu2l}*SF8n;37Go~MOn>yKlr@pow5j1)-{)lsoJj{E44%y z@Q!?!bCb zRlc=MX|F^kU3Fmm$oP&0B<17_{;`pd(fqIplibq&{eJs7GFNANt5ChzPv_wC$HAE7 zj%bG&D}f74bnl-1kRtiFEWT7B2fW_jgX=f)(U-&NkFITvGIi`tdoiMu9){A^)m2Qp zMLu88`gG)O?S1ysd^e)*&}+4ZKX=|Un41+X=U)h@a$)l?=-*42?AEopa<@#z6up9? z*LmX##t(pbv=i)2Chc)xgiWf}X|9s}3o7GeJf2eOHuGx~Dj(hB{vSl)wsH(Gt}b-4 zN?6hccEL!WbFezpg0G{9^CS7vI7lbbVLO7li`P{$!8DCOM~;1fLgiLb{zj4Zk0LfC zkT<(h>`geMtsY@+?g+QNQ)Iz(OAmjJ#X0?UMCx2l(f@;}Ij4w-jDe?h0&k{R!cMY# zKzo|6myC|j)NF&vCT%SuHm$PCIe}Zz3I2~JeF^DOr=`x4sFX>)|6l*BZib}$W|Zh| zMkyT_rQf2PfkTH9LKF{d`u(3jnE!P1Z00HaZT5e6+i;dHw_m(e3&B)CdN&2{C;vK~ zlT60xOeGy@2F2}*c)}N*E5*V2Ze22({Obr+GDzH+5=bPh;!ykq+7OU5ou5j^d2DKt zVPiAeACcl&`B90RTY6ox8^EDTqWVi-ap)-=;46!98(Qn#G+ing^|7fL&@Hw#D^OO^k+h5YC3$q3lcZz!e=ZUyX4K2|;ShT~$ga$lVgu;5 z0bG#*A8>E0e;l)jG_x7*i%7X~j7^NS*=o61*Wp9<%eK@_&)2U~THKl(^983yGi*%R z>#9`a0`S&)n}%;C5T1jvNlxZIsKQ=x<*f+SI+@2BYU0 zyis6;8x`bP%6?Czhoo?Bqyl>320Itl+ImULte}xy=pZ*BMM*hNWW0kQ@4|PPNHnJZ z$9UuaczU{#7UT$7$Yz+{ZG`)2IPFvkA3r{*WmG`V5ha2XN z3#l;rCGi}NZaw@ysAQw%_lydoUchFMz&Pp9YICpLA)Li+6yEN!wO6(HdPYcYFc??~ z353eF>_3s|8wgqryE65G#y98bYK||EnJ}qpUDe;}n{x zvc$~J9~aoe5J02U%wVix1}ZyltRoVDoUvl4A$rLj^A6iWdPKdbKuYSd^WnRZ!8VeW zn6Y^QMi)Wh`U~Ea2!Sh4+7a>ZNTwjjD<60#ZlyRf1cI>dojr=E`&F8>d&Va>C3g+k zP2;n75b&7v1h|-?8euE;ukHITN%|hOEOO=?txenc=4z%($u{9?Vfb?CzjXJdL}09x zxRdg2agXqGxTzA~+iT!LA1Jwsj?XwvzcZG0n}C#$5QT3 zdnp#Wo(r@Z4WM1L>kYiSXxAE~6=>80{?r-`yxXr8{1&7#osp`kPaX>0~I*IWNjw zYzPM;{2GQUwfMOiHY7WNrs8xmLgy@zOdyakULcSe`v`8@OZuM4La_2b7ajR3mJN}! zYkz%tt103bqKRYzRgXu+L4@C*Et4%9 z3pxY4zapW>W8m5RO@{QqeKhN^z&CFlfXkeR`$E`6yvw*t&$=+1wUuyQA4akRmoAtY z4>1MVCdn{=NlW1`I9#thbR1N~?<7%MyF6-Y$!p0m;iHnwkRHlO1)*MK3#6C5h+KQw z|8AdKa42#q;aDc}Tl_KPed=2Ci0S&pTW+d)l65JEr~Q#j)-lXHGjXC6b~>FjVwUq^ zkAW5I_boEXyA9evO2QFznUb+XR5F{mxhH5NuaKe7kT^4Z|7v)$J*lG5#{m4_1;3n1 zxjr&^mDBj1{h7+VR8mI4eomi~y1U9cU6DY+G5$Ekm}kiLe^gAu$6i5Fg+x7Iq~gW> zcp^DCOeQ35bN!!-IK}XK1ILf9rtA?yMN~@L??x}PIe!)Wsl3YOTqua(DWXyUT7Fmj z>NGsH3JG}G!;6*=5ViN;t4q1O$`;5O$DMyv5V?6O?k-hC)%Z+>VtD)AKO2J#T(uIc zff4>z+c+-}y?E61wwL6KVN0tqP4CS7*MkewPiuY&JEGPJ1k2@ynoE(ud#D1xw6U<> z5d$qZHM~miLasq#VqMJ!s#^)hY>O%+?z38A5ef&%Z-lUjiTE;Y)&9$ZNkR9Ym5WDk z#+D{G0O#Psn1%Lwb!`xlMbl@J`gdDMq)6Fgn_K`cQQe=Y>AscWRBe~zlzw$Za|*dn z*?9Ng##`e5X<&V)FrvOEqI_TX$(2=O@o?U$IJ89tL$0_vwWbf432+xta7O$4_JN(n zG7J{Svb@`LShS2nTMRC5j00Uc1Dn)i6LrBDP%Rmar*}eL3WIN`tK1-iA<1udy;RdR z){OEjUD4e_Ffu3Zw!d&&*B`~>i@AV(fSH)IT8#`q3(Nr3((q2sx!p{O04)4^N-%6* zVP^ftBA|?D$8c1fK`!l1sX~T(=TjY?Wku-3?G_HsW$ww{AQtD=)fiD+y+f}bUw}bk z@DdVTc^H%Kh#EO+#cMl)XArVctM5eoSF(+pvX({m@Af||O?zGKML8MV zv)-Bik#B6+xfM9EBE!fz7=w~o+~zms%&kjedv=u7=Anrs>!Z@g_@Nqm)Y_vlyZgIj zi|=%Kz*$FUdl~jDS0{=7a>mRCHt~=9O16T|0qZi?8GAeFyL$ zFTg{|vHgGu|I4GFRN=rZWc!AH```mEcsr115%o*#>&+tl;$b)eB^whzZ zSjtrBo5nCYSxm$UTGS!hX4vD<4Ut_bo0@1Zp6_&e`jzHCiVeT+b~CRnvFDs54PVMz zUej`V$AVa8B7$?_??+6fw*IGq6-*$=3#I7YDJ~O8oYc~HXDutM-#0QaaGYyxY+PyZ za&_%;dbPb1xq3Q2Nzwhk)EshHCYXw!*QZgF!?_E4F{WN>=h?% z^$>lUI+RuNWzI*H5%)_>MRrJrUcv7wVnF2*&iVXrEi9a=g{2P53koMMVJ1OFH~fTd z_`SX1_v?n=HNemL65zLS`d;7inJMU@Hc;;>6Zr5)Sx$j=8`b!<sbZ`4Xn6wX&f1>}YXWEQS_H?1sXWlant-9X zq}-IWB#1lD;?xC%e5M%LU~-qeFl|g*i`;fj5GWkb-p}XzM>YPaV`)RZ5m7d2(D(E~ zC>q@`$n^-G{-!eDhzUB``;Q96P^Uq9&Kp#_JHPNV!tm;kw7Z7HItU6+Sg(VOHJdjP!60HPj8R_e9 zrN$Y%@I#=F$^g(~D39G`LN#vTcMFD`Fueu4>z4)6m75lH+X_FEwh-0%^wx%;UfcIffuR-dr;QeCg4KLH2y-4CD5iLW}b^)(?@Jt3%f@ZWKZ8eCv4G z3cH4HsiAWlNq{2TS5{sT?Ssf&4GaiRkg}=bhTsiqU|@YsSo-&}sQ_f&T+EiO+&}8F zW199GD#1RUpOFUS;GFx;eK#$tMfN7ez%A_`Ots!8XC?(AqCHsqCnJB9+^#|m6@N^S z)h|bv9w<=#u@qlWvUm{T79eMe@!kVgSPmynUU7bivEaEtQffXWP*h~KeImi@n@JYm8lulJ^qYNX1->Yey8$yXCdk9 z(Z9xQXGe8QA-aSb+?lE07u#_&U^kc=Rt!9ms>VZ)956RpZwB{fpgM|0$KvM|>&Y9w zxU2ng=yDa>c)*?b*cF=f6&71_>8m*49(Le!m5MrpAsGEVFn2quy;`2iiQcU3YJ11u za%!NI!Jc7CrBv-fKNZdQI6I2gMJUU1nSo?uvcOc#<3p9<=?3_-itpHj4sO#K^d0O# zoYz^SB@+F;{D|P*S(UY||Jse5!Ob+9wXebBf`&syhliG8&FT-OgQ7TSAom{hpmsiG zyf-`*Id0iw<&ZMWbMR{_MCD%y=+7w9b=J!6Y2(4Yh3&TU3+h`aMsH4KB0;lAvd z;+D7|WJzCKpn#9e?Gw6h+?b^TqK1dCXUkj0{wV}le);{0B@k|8o7XOgAi4d765Yzb zXc19K-h3#LLm&S4t^*cst^?``hcDmroO@+qCL}6sXt?=R&xSrQI$U`%ee9TNbJ1Gw zh##@Mpw?r+fW5wsIOF7o-)!Ic;QbT@McQZlScLn8QT*JTamj-$3VbN z(}AS@?)Utr|0bHg5&c#L{$$C#GM`inr47@e;%zsTZFPZa4MZ?p48x50Ig2+WJ=<#*HM$M+;v1waBilH!FP zO4i@|OADPGw+vm7{k9P_gs>1^DB*Eq20(?>++ju2mm^#IPw11bEWZ9@I{yF+lR-I+rJ)fP>lZ+8UUU%y-jo1I{X&`?|%8xFUNVKj$6I<^e)rq z@_C8ShAHSX+}u-Bxy=aHY))WK^^)q}i$wMlO;Z-SrL5J!o9z_i+un}zhWP;g|E2K9 z)Ne|`a2`^R$g?DUyXvwzXp8pftIE7nQ-y7K>UacCG00w%bD`vK;x>t$neDT`QZRad z6~3JR?>*U*LcEsB+bP%zHrM}3#KGvWM}kMxgsG|${V>Po^>Smuq%%=j0SD1j%*j=~ z?QW6%pfzQC+<4h5sk+9L)nZRJ?60wOEHbmqX#0@2oJr_%%rGiTT2AO{1jv(0GE#aN zcv!^gY$_RPV>#R(0phF7k>X~EqoMO~BMq>?O_DVJx3BxgfC8r_(EYb6r{y|4f_&=l zDBU2=BgqKC=u&B=`^J^z(f|9WLW$Z2zncs=X8iskMG-Kx>_N7`0p9LmdssS$puV~l z0cxOx|6=^u-YTECMx9a0(8J;VvbHTXJQ9K&g~8i3WT)d+k8?cj8iT2pIp zdKA|VPRsFb5SNqa0d<7M)FRIOSZv7*ZjKf;AV$rM@SYPY(d^rkjL?q;Wb|q*bn}2R zP>t%QSDv&Ja~I*iF1Upcwxi*( z(qped&!jV|`D_Zf0Kkwp$8a^7;EMws$k}g6AmA!E4r1>ReiTr-6<`wdhTXYn7Dtb< zFQz5JzxRj30d@HF5~mD*e^%=qB)tSzm*BI%Ie{x*CvHR7O@`tmiY`NX_DN-S(W^C$ zq%K9Q;EItD_qB4J5tGRj^g+&U7Uv%4ZaBU7sQ2!@D)%*jf@7{0)6+7(9KP`Nr%3~* ze|=pU1pV&=K6_ewMX}#suTBG*PH&QuE@>+bIp1X`4Nwuc&ao)do(5ik2wjUvX)pyU zRt<`VAIeynMIiI+`S$4`w~Jv{Y?zb-2}u`T=zh{@-9UuEo6TZtOtK->{HS=WY*WQW ze$Kt)>>vDdr#DT2hR?mP)CZkw1(bGw^mFWJUV3zrJJJiR1G+gt7gn%{sf!58DiU|` zjSavQ@lGkQdgHc?2=PQt&on7xqO^yn11ICz4Ux|J_)&*{Krqbs!8iTl1$`vP|B$$L ztp?_=eiO>1VWo_%(Jl@}Bg?z9Nk1E+v%oGzE{nbN$2U(546-Mewu&E)T+qetK0v~Hi|jt>>G*wIUa;qYjObRNUUSu^2w!Vrn9B1xLMB4 zl*OewVTpWy%~1U2tqd-lx!8}kJlyM@@r?Sfow8KvQ=w`h1D`ePJ;BcSELlOgRgM-u zlFxOib#{yk!BZ!|Sk{DLIcC*Oq&IFHT( z9Ght0vHab6Fj1@{$x&w+8@S%)EGyZMEw_)gP<-T2?9J?rd4yVBy%vy5vLTZKOeO9 zh$aDf;c&ReRVjrRm&et%x!xrwvRdY9*|tqt>kl;4Jg7E5uSkM3HhihoZwzc z)dF1ULvyE@ir$9QmY!Qvf~Cdv@1pdWnZj&N%uhu$69R{p3Hpyx&p-Zj*-I$K*i5l_ zw&4|Z$7SO|z-KdyXmN{E9F9%jUXk8z@7=pwSsCo_r2wy6?#uh_*sJ&pWLRH*ZQMB< zzq{&B&FR|9*zFrl=3SFT_j!Zb;*RS$nItW#w^@}i*^k(0){c$uK$=HZ2FGUcTJX!8 zz(Y0tR;BNTcwv)q(Ew%aV~bX%iIW+c9?)(A&1cLLdC)CM#2w{Od)R^xtZ9ka!De)0oa6sit>xdZcK2XQclUQ z&-DCSbJExtZ#5G6T!(a`1fqqi@$?_pC3u*xV+$@ZLWSwi;0%#FJotb#p4 zMdJG!)PkWb%&RJ+qVdp%-TRDpCPwMORh*INJ?xqhJlfPJ#zjhC5-9 zoGI8N%iHjpq+@J$xO6!ms9Kv_)+%}S!}*Rf7aj3V$=Z78PBBG_YhPC+8iFZY1Gl~t zGb>ScavGuaK>9Y57&8YKpRH=H0=o(*=avnAs@OZdqvGT^(VsENZnx$T=epg6@w!`w zMcjz$mqPJw&H^S}lc;YiSe%r-Rw^+Vf>~X9fV!@QNd%eTU^@vGJ6OTzuWfHf>XP-- zZ=6}p9=r6HX2CbrA z(q0@G*8ZazanLa|Ez=6lPSKSSm9jQI(Cva3jhSgqX0S9Um`@oU!9J1fPu{@XJvq{t z@vtVlgo#oc;q7u@U{UUaXfKppflAhQs=bQ zp(K&kib52HatcMrn4pePk|+jiymb}9)D|yj04q;i(g=d;zG-07txS+T1Ytmj7GQn| z>nLB!mo1qrBQB-_k%2I{nzXt5gEe!tiCfakIt}H9RpbNeifVQKsfZ;H( zZBSqqJ{4r^u^@XCR_^pB7Q669-7?lIRpe0F_)VwuPoapUFlAjiq-M>CqAn7!^YrE5 zL{J_fF}oSaQ6%cyr7BTlhaBQ=?G-E89Y&-nH;f^3+S{Q@ma*CvNSgE+dpV*2DOgU$ zNC9+PQkY8N9Osf=oF&#pR4_vTtD#8Yd52+bh-r7Gx?s8k?ynrzZJ1U5!+ zOq`a&*aNLtwqNG8FwxBLol`tb!Va%{CAHp^7m3&^>}}iAJ!B!1!|Do>H6xT&31C$9 zYW%iDxTAbgT5UXqn7>xjj(bY`549Cz)8cE-q32I3?MKTd=e#tG!aA z(>?H^yFrY&vuLZksqVW1Lu#xGdBcb8zs$|YH+fccNrqJkEqB{*eAWxN=5BlwX?im#o-4$@60QDNA^i~=VuY!n0(SIwEy}dv;*1WwPp$1*D>MtU#)jL@bdU9LRd@&_Cv#6ao7va1n3V-Hpv4JKnrghBD z2i+sxUOp32n|3QNA0cdM>M!+aX!+1?nbGi;K=7-BU8(L!iI+1AlzvXK{FFD892&hw zlrSkTespXcpD2Wr5T|ngV*l9+7kvTFkgP3lanzC!;I4byQZN5v2?(T`5C*Dr>%7M7 zPty}tx?*6Z3*BU~g_7obJ1dR?QCdDVG*no+!Zztt0%Ji1oO6nSOb=bA#PY3yxGl*= zHb?nbjh31&{42C6(aAu_+x)Rtg($VgF-R{;%q#72aVE7&q&;aP_AN}yC zuu(g*HE%1NUc1WnY6Uz^b}1kZA@p?eD;hi091U0p(p=nJ`I1v^|dbw2gsGwq&(&2gv59d^EQf-0c45i!{FO{QZ#i@V(65}5z697cjtEsjP zGnnkO1Q&0S8V)T} zZ&BmEH&Y$K>=Xd$m{0(D+)g@k&+l=kmnn}o9CO^D;!rZY0I`C?By={EZA!qOg_=ng zFb#=`gRz2f`Z*72wRWs!%L=0M;Had5jBHN&tnqOiKy{x8vgg$>%_v(_{;E*z+e^cD zRKcSRYMN@dZi(|Lm+1<0XFi517bRrc!{{*T9nljik_==^8x6Jc$TO2Qw491Mw3_Id z_cSJ<0can<2`NZi?OhhmWy(p;$9p)dYb#$u5XWoFSUnfP)pqww0}C65Rm%3M{J%K9 zQ=Cmrl*A`)1|TaJm<*)2Q~vTBpV|AcsX{=!U~@UWQz2vdNB&I0%@e_zEz1ftoE0pE z!anr|2D#5!%eh!7WaTQToOx!nBKOZnztP;B}NgPn^5QwH7293c! z%d3t~|6*;$7%f*eS9w&PK2iMH`FUX-JSEd^8rBG&n1fVq2c(tE zQn!C1_~2Kasw%o#4?mV7#G0O#)}YcUpm*axQp<_5$anLRWja=2{>+>?V1k>do64a% z^e)V1p1ZO?zD@*_+#lbicv)M7cd@h7q~j%MXpT^=p++}|xmoK%8(lvGTcdxJu5MX) z41QN$weq$5XecepiAm0W*dHi&zWClcYiH^BUwlWv&d~lC$@QK#X4aQ#VCB`;$f9t3 zK4kx=2tSz@Wuq2@lTDcPTe_V>FCI$I@Cz*t*>F;nym(u;7xi*6Et`5U!7`Bi@kOZU zz)e4zkgB>}+;lOc<+Ug;xOL>i0shZYQ({lT2l_pO^NxP}#Io~bOv0(y>TLlrrp$O*EN*MTfNYQc&GHBr{`q8Azv*BD$*Z;ouDm0q^cy+fm54V^a1SOQqckyuGhsy9&{y!C9b4AU7URu*gz8{tT=SHBTw@n@(t~c25oNa7 z!iXim#4sNw(PoWdr{LC%w_5g)Z8j6S;};$ylq=fl*b>w^GcGfdrkTwwEXAIg9ztNn zw~!}iAm+&aO)88our3SNv4Jnu!`QyvA=OV|=@K-hXAAz&;{#3U-E;ZQ?C60_Xl`V? zXCu?@`G&$La~7jTNi@Gucj=_GFS5V##=MJuuZk>l!EG5z_ z^9yeXB)Ge6VK(kFodj}^?pckTsR$-pS%|DG$?CScr&%u4nECXL}oTN!2^Vq zCGxO^y}Csx5X%bMX>jr{{nkE*vXi;wecSkyX=r$iis!;gQECg_UPSEtxUqEb@0(__ zX75ToM34tNml@NA@>jHPeNao_=w)*L=pU$egG*sQW)3)cSdlV=&|#lXJafJ_{7bpWCFPM~s9v z#Wz)Dh)Ga)W^An9EKmp%aYTRYqE%Y~BO0NUhxaRBj5CTxZVL3fQuNvoWGK2)yocts zrwGd;5B`!noR%6=`b!vpJ?2X|o^4#sdhzEU_G!sl^9D+kpU5K4aKt6JQ<6niOM zr@$p;IzN+A+btg)?GDof_GZMUw9}%dK6S4s@KD-Be@K!T*YD>(UUnpRXofhwyJ8p` z7b;gNx%RHL<+sxS@d6n&NzsE0M_w7V>+l|X5w^seS=-wx!>TFvDsM1VA&`Z1*=uG8 zKgKAMLv>s$T)sAfym9)IXKo)hf}K_J()8{$Du%QNmo^J;R;{UR1fjo+J-%9mL}XE0 zYA%PLT=2C6^n=ZUzsN5M3Eyrpl5$c$3pVCi8Coo-7S~XUcfs6K(HVZGcBV zTr>#*Y-G`gB`?X3oU(k?3i!lvM|1=uDaDZ)W+Z!Ji|{$BBA!fWmQx$J56gJdcF&9y z>**gTbgw>M-*d}{o@d?p8MwGCiFwg$@CQoAFY|VhmJlu;EOFi+EI&Vg@Kt}kE`C3C zkpuI{DmWxmd&qf*n?G@<>dTSBr$Ul3?~B=MuM}vS9evw2nUcAmiwz3lNWv$MFJFnv zGqz2}f!kG^h!^A{kUWcD05 zf#y#aawnNaLA(aeKrde_>}2f#N^l>s;ke@NFmwDNgNYdm=}hXqHeEF%c(|o#Zf1R)+|V6Y979uLcWHlrEB-MecM``G zCu;-!e(K>w6NZl3_k{}?DiBs0or=jjiIJw2)yL>_YA@tQMUxEyOoarEN8FxPIC{Kd z+TOQ}2gILW%q|r$&nHnk%9cVQGQpw;(C0L*`i_^9F?L(gIhfU(aX5Gp~1T!)-~4Eyl1Cpy0|OD zsbv&Su-*=+t1<^=%#Bx}lGpQY37uv*OcQR|0ly@&@pd@;t4D)o&bF^_E!-OI5eVy5 znIUgRl_nX>;xlYZ*F!7^_`t{ydq_M#ZKTDSSAlJ%l1S?|Fd0s?o!*nfLL6^ZE@X}_1LBlO+ zQ_>#GL@JiHKS4Q!>k03A&%#_G79+OMSoj(nStm&#=*+mr>6nN4#+NspS+(|qv!xg# zJLqw>e6@VG+hPgFFY3?lP6MA{5djgO5pi`!hsdQ*9Stl(;z#p*+J%IZ#Bal50uKx` z5MCU5O;POkC>WqaiOk&NnO~tfYB_C<{-4QQOlpD{iZ*Vmgqq-&~eP3+b z*6i2NOTtElrL=}PGo%i$$pCP{%z)YjpcBahP8WgoXkl_7uqPp#Z*|(7FBkCU2T<) z0@6We2G6FBxc6SsF}u=V?wOT+7*CIWd54Ym3r=17DF;V2L848W0J*jrvQ8d&nFv~v znd(QoPu4BXq;d20jm3IO!r($2dq+z(zi_Hl7-3W3fJubtoO3a~)QhafV_uUa%b-SX zPl=wP(?IUytU(1fY{wwPdjcXL`7hh0Wf}|dTEfUbE{T(t1StACB~X?Sosf*${d%#~ zCqZ+BNK17{h>MT=MtqZ|mor}hxNyxIa#44p%PBuRRL`BPTfWf%z0mHau^UNp?4=rl z8j_$3c#7o%xjj>90 z7ugfxqiM}LBc=HmPKx9d4hI6GF!b195!vNnEy(zCVN~j1T${PcZJRJG%{QeHHwQGA zk&kHt8x59~AvY@_rbZ$N=?>f6Q-TlXlAGT)oRz>VCUZsa=;gQ1Q_|~>^rozb!SU3o z3}u3XVP9Ljf&@N93wE#OCCitEje%1x0l^|5rBT-9+~`9;^GJDXKR-f5 zl%>B^oW#XTirE&`wW4Y;#Bo!(`41EVO$P%vcHEaR@bM}WEs%|ugumSukeVdEhwf+x z{{~kGgB>j?pcWnI`M$w$ZH{f=GzC8&r&D>T>}fVTMR$x-@DbjdfLC%`NKEtAZl(H5 z>XZxmlDoz4-hQo-jzwPcE0W|?V3z@Xu}_CC)J{*SGC|I zxRp-nIu>ay$iWO!lEH5EdVIwRW~#bO*z14py;}eq5C*uYX}gf+{tPz@HL;rDt>cts62a;qS9T7KKUFudD)v02D741Mej7+r~;Cp4NDcBBZi4xs?ZQZ#u z{FPBntQHs|mon*;HXVyVGtR9*s_03_X7)P89Rvqn<{tJSEei-J|D2&6X%I5)xUS%3~ z{3Y3bER*$bW_81Tk6xQ73UMT;HY^RApwVT?=xcy$XSiXsjFTd{@B~m934$P#)m7>f0mg@7XkE#f9Rz$c0MK5Uh zN|!|k^r}2Ig&&j=IqEy3??#U?YYP>!3)d?$_)jLsChIC}N^>v{Es%zk8*$uOZ4K8H ztF{^HLQRh5+O~qdoRSca55woM>E zCd0(ygT+Y;{ZdOt2zND#wMR;(*abWA{GSuP%WN=AG%I9EZ{RQy6@TyJ^3_+kgSF)X zn*DU;zAqK;Qsfo2$-AhUK&a zEIr?jF!os)5WfPL2ABgK zcAC;2!V?Kce^=yvWlN+7|K{n;pJKcLF*eU?(&u#euT_JXQ3XkcP-_--n_91z#1||A zpN?%XPpAZ-9Ik9GE`<^qDqa~&pSMWXbcms$q_d@pG#*$eo(`%AEbXRQOu3W;}$ z7r0hsQK}^kqof}JPSm<*9N%KhymdaxF*uIu3O7U8`Jb-if#!Fq7((7vh=yU#@o{f- zSS=(91>Lg;_U3lA8&hd&)%fd{PjT}2KSI>q^17o$%;9jvY0|40lN;#x{3w+ z&qT1PGPS5GkWSif0lZH9(;sRUVb*x$$#Q*MQGfo+!oS@k9{};pmw)MCn>-dp__a3g zp|GIJZhx5HqOB0;hfs=CHdZ?+=d=&(G3mPfYnSYWFrKVRfB^{w!a_}g5}@;_$Wp|u-x zIA*-xRm3wf)+O2bL1qC1Rtv;o0FL8hg}#jR^cw43!7_c!2a7bp&o}grE&{NJN}|Lc zfjDlFFDVWSYt}^}9P+KG*p?LK79D#3{;-Gm;#|j{{psEtTKw(j`!iruRY;8SQW8>Z zc;%}&O85RuzG79_45j-kBfE}-=1;F-^3S{&mG<%kqaQPpCzfO^nK9czolX?vCgfq% zA7$eMAP{(W28TZTY7}GEi)pd_XM-%sPW^F%(_)zVh6f4yeL#e)aa;jhFj~e?J5Q-d zAIlJdK|DnH04bFZ#uUE5$z^3<7G-gzgsxu8^LV-cTD&o~0Z3>B6`KjgJ|1a-(l%Qz zSZFP%)J`L(NNHM(@&R4$tysW7 zK5_^&wvaO1>RDr{^mOX)rIE%fO(YT5HO?v40+WEb_=ZfA;lEdD44wyjnSl%#QsKW@ zCm?WVJE}eO`HjEIsDyVX>9&>3U)l;Rk0}irS8wdFfZ7WHFal1a+tGXVxCap%jz^@+ zCGt-L8hVFs?-$LSh8yatlu$X19M1gw`pCR!lH`JVK3g4T#7*)>CO6OgJn5ySU1~8g zDBo@)&w}(@?TboV4U0!v=K1_S5z(LHT;JPl9>8e3RNmOI_#+)iU~VF%sWL&kD9U`r zlD{;tWHII!F!}QIIBgS0D4F5?VaB!pAdrz5$Up}mm}6hXNkLT{g<+9CO&e0al1%tiHa?-QD*{9 zW0#1AW&A~_w1Yi@Ov$)R$=F6QaG+}J77Wyew|zB+xu>pGe>_VGI31b7Nl?@HqK?W{ zYp;eER3X}pClPrFhF1Bj;UukDpZMiUlp%z9=*T(1$^H2YfYWd-0t?iZ--@Z_UJCp& zxt}Y#pUd_j28Dy@L{#$~01C-S+F>VAjXfaL@?bBry_(kl7ZET(M4`rZP-UC$r{6NB zci7P27#I%WCRLggGef{hwz{NeN38_P(WJ;1_hA(Gp;wEFRj{~fHqC3fpUPRaX=k=v z$@ijwID%v_O zg}mvLS^mtqvRGMT?cruh^^SpGMppQNsRQs&82yG?oI)KdkIE=*DMO}-O!20K)mO&I z2i%TX-Rf#Y$G_9$j`JOdzHli=LN}0!7B^L6XZ{IF^avTm`quMDW z{ZBcop!fbqoM*jwyht=%S<#k%OIh^_p!)ApR{gZU*EGL+d|&g3R^a7eh;gbG!WN$} z5d5llSeWDj!9%OH96-odU2eV@R(yBip5J@C@mD`TpMRGfw9;nh%h9$axr__?{pr5` zn9BemOcCwgIy9p#CsqS)+!Cp-t0U37#$Nnet;zpjx_tF16viN~BAEfTQ{DXg1FHiR z3{xuQ%L9)5KMy2zMJr%YZnP{rm%SfklrEWM7sQIXk^gUv+yc?`GLoU4z(u^3geI$R zd`D#eg#4uS_UB)apD;Ez{tYQ%9mDy*K}sMw(>l4@nyT?W`Oesw#PsRJYuu|Oplpj) zcc{;XkM|ruAh;r~DvHo&QxLTG;>2}A)l1}zXaPv|Ev}M4dc>WU+V5oE0dTyo_O-U~{3hUGY(!=|c z=lBXgsUkH#)CRH{_=*7Hss|u|o`*0!pyF^()DthNN)GUV`yVK{KzQu^9su-1{r%h= z&CCxWj|cw`Tv<~qy%k*xu*^4fwKesE%~cX*01($g;gJ00GFufn`Rsa;Lv7xtNZU>e zG%l8XF}K%-D`nAsMS`ZaQFkjWNYR@R$iEK`$+2;F(DOk6*zM2+0G8XY_|t+c9fAwW zx{(1<8p;y%uZ^ty*3T`>UDO!N1(^d2iEkl?Ma=Vs7u6-hIBW0M@1{B4)XQXa^f6-B<1)%#{0NVQc`-spKI*PdmkwySZBpsY5Kmr#d z+i1CMY|bgDfEE78m6BDM=@+bZW^+FDAWH<>TwkG<#kLf7t{s+tSfN6qxP2x4m21;U ziF5X>FYee-L`Hr?)0SG2nsBh(vpylC7dBoFef_&wTCW(}62}OxU^c2X0lApobvQhWjdcnR~Vm?>I;rvo^KA`bVzob#ngM%!<{^&kc_*zIhtB z0RG5?{nUIB7&NxLYzU#4=~`@jtFjLyD25{xVW1#0M(NR?>9XN!Wy#@012)l663bU> znPDfa^H^+K7^1XNOw|}x8K$cM>I{VI_=-k}Z>hBy8oQJjuLPm5g2+7cRZC{ru`)Cb zt8IACq36%A1qq6cVGC#;3!BK?)&)wB9%ZdE5c?PgS4qzLGB97Y#ErbJ4Y$QXxbB(a zEG&`IqZ_dqKK9KjG(H)DqH8--C3qO|=^xWUTElcDmO-VB#8e-FXG%c7i|B?v=1Ndu zGx2jb-b)UaVR)hzq9;c(=SCz_XZJiV3B2*L3&!pNRu;>xTOMZjL@uq0wx6>KQ3%9V zc0Y$VTpCN3lvijfrYe4O7=^7Ld4a?a?LjWg;v3zJZL2%chi*w!jtlb*_;8}mZ_%R@ zf1m*67hh&iG~9YeD?vJw(Zd5N=N4g<8OSb4K{X7z6T=l?m|rvsO0LV+J8yJU+e zlmF>n`EMwBB|vHm0J*R8_6cwgmGgV*N#_0Ef%ab_x?ZNDet>?rcX-~@Na*x=fxA4& z@+bl-+XdUI#aA1fsdF_L<=c3&RyFn>@1tihRl0AUFl`*w9VGxYL9u6=(d->T&Bqsu zkL4ZiG5pz~pqqx}v;llig07ITfiDfgeP`it+#v>%plyk|U3J;>7ywp4_R!_d zuIU`?bsL(IEqGJv#;vZmJ>F{VbpJHq_W0dl^zNsb1)4GB??C#3y0Mo?RaVQ1v&lWz ze++{I&S!w`>Qy)91rX6Uowaq_t?$eU9gqCw`Qg{!U|!JvOHh8te;1U`CCM~aABL~l={JrwOa22+9dm`b@ZkB+oc!^(1r(YSnw*S3-CBG(F zvsR)8sPvn!SKHPA#J~4HfB7itrr&E4IVG7j%;%Ag9L38dNCp776u=h5eDgwqE|EW& z?HB=!v;{h*$fU{Kpn|v`TF>diLGCCb;!mz1=6#pXWXPnYPQ)1jj5HRM&q{tY7 zJQJN0DhX$k+7jruaPCRKS2`&&N|586LPuP;$mqbr*1&9W5o{s-h8GZK8)X4|K&+s< ztk(|=Nn-@K3i!{Sx5Ozit2QaG{8dgN2x%=~)2$m64g^g^#D+KeH<7wAkn**K5qFn} z!z&9oPfC6pNh~Xi{+0mRV|jMD4CFBPXIibl#h9wFfkPHm?I#n@ew*_sFv5|o`ajCa zUP7(|X`~{TEqLleX^5tOAe~^Xe}6T6j;$E5Iz*oM7wM&$dYbBY-UWUn_{9G%K}6Ld zh7&lk;ysEeavABP;&y;Dl-<56v~fW4s*mxtZ@#6Eg{6;KlkWR)`TulqqGVzn*p-El2B7^UKk=kFP)dx-+V^57WT%oh8X~)W9^ksxXfP_mS zX@@}Rlm(i&>uKS~n@^^$y82s=`d1l3(*8hk{cV`?e7CLAyhs1wIdHzJGI0&{7_j!- zy0UNLHWlLtaN-(XtUJP5U)X`gB@e)xd4qjBNJv2ZC>~!8~Q&?uZ<9e%a)?rjONkq4nXW!xY$(f194$C0BMeTe=$Fkir6jxpSoyyZavEZ z$pvrY#>Yxu3jP4NHl|V~_eZnbt<}ZHY$~`=7nEVGI8k z#B{)Im<_lMDN~Ffo3Ha!Z?B~qJbZk|mu9}jb@w5Z@Ll;cr;~o=`a#CF7$M++)-(^m zl6x8gz{ouxRfQ#t5)`WQDkAY%MMh($+$;nn7Z+;*jPiUW(zzTDfRQdc0AQrwluD7S zqGtougKe`hmq`O=J}DQOEl^TK%lz9vtY2>%+h%XmxaK{L_2>`#1&sJ7)@!Q*@kEo( z>YLS=>rq5M0;>eo&+%HO?hlG@$pgV@hyPPvDB&X7~-uBW@@-{c|$C z++)EK072t&DhW%V=QWk%dh6sS<<70&YV!5}ZYrCJf;V#Uz$AqoHfD)}0e7JZf;C*X zUK6=NX!DIO&B^FbTxhi2q4I7t6FEGO$M27lb1bQPE4cp&{5j^T2!LCcZI}{$ya@o9 z$M7fvE{(M`BvLEIbzH|)6DM2lUcjHX$&BAY9FqSQ#PR>bZPF_|D=xSS&D}r0{32D| zOfHSBqVkw#eEkoUKTxXfkE+`uVH%FZ-#;iYy8_?iczayl?f=ZXr>S=59j`Cy@`LFK z=ZC-3NZo#?k=p!g8tGrNNdLKDUNq+3$zRL);A~L~wQTXL=hB3|t#+clUN1YpZx-!O zqf<_{_7Z4TtHq;BPIi2H2RzF&QijBIUB8y@#b=Zm}XCJ%b*K0?EW^8tWue|THI z3eUKnHq7el@Ie%3mzarWVe%eFnPAp>I;_0}AvuGj!_BG5AM2Y8%aqM!cK`w!lwSnz=8C%IuCNZA(@V^*w? z{m6HJe+28MdPykGI$oa(7U2#B?2QYGaame(l!`vU%Rj}=J0cy1Qbd&1P%GBTY_^mELG zB}5r4*G$5A+#?ephw)~{;-uJEh_clqwCD@AphT*rAY1C){^#QkoRy7q#w+*%A+xn! zwu%4);J?OJmg!9C>E*A(8aJxFR;nfEO9J*`1T+D%%@RS0&qy+g@bLrceLH{Kb$l@k z;4x6KIjYPbV*vV)YV{4VaDNO_b<={ZEJ5KOR`6nSJ3NHUt!fUF~?snt!-69aoQ2ABXB7r$^Hvg+ziDuR9Kz zX9UO5lZqe%Tww3lxHcEqfI_D;X6n zE)O3I#lJnx-xH6h3L#_O{^>SZLHp{7_5JP40WI$bLw}%zCx2(PX4J3rUnYr-BzlSV zgq)*RcRyTWeEHZjG^}{}PZvG_zwCMqzx)RZM))HDe_3Sm7VRs~gC}6Apu$QSHlOdlDfCe#(@DmZKxI=dWCWh!U*04@f+s+FdJ>_YHfJ*$=JlG>nr5Tbq9(Ya=sNI^rpQsYyiaDF*cw940EjuWY) zyBO2O#08O02-P6G$Mc~Ptw$*tXZ1%cOzU;l#@5PgO0H}vEtnnl2?3c!&f!lH5@@ZA z{9y>XDFvzr_3~Q++}LE~0fuwL5bp3pZ(oXL$FOADL@i6<#8C@2XWS~h*>qBv-Ipc^ z*@&{>%ksw0&v76sO7x*i!@)V3!#=)3Rf<+bAep>y;VxaaLo+|VUxZf5ci*)p^8I4h zfBq}qK;AFCRC9IZHQXwv&05!r_1VDC-Kc_4?~WKIm*ei((Uvyh%)wlm;b?a+2C=DC z*xrC^jM&!)u20F%qc?u>?2wL*^N{_b$n|VI#?loMgSY{X&A38!CRoA{rRBL|s$^Wz zzYn5JJEfYwe25LO#O;?n^uhO+UwO^h3Ne?IsS#I=`{lqTI# zSTc!=I#Id_~lK!iMHiTIZN!43B)urOEyRUXakLW8IT_7 z1*dMli6=Zw<+`f5q)BoRqy}mZtm4dVmB$`uPxr%5HE30ikUA^;?1o1(Xj~4tXFDcR zAYW%k&}{)_#siPVX*Zo%xsX3L+cCB(AR{f)eP$Dk&E1>XV5H{pT|Cl;8)q}HUu`H>w#IT^kptV)oA$#nE*EB)1BQjdBBV67tb&oqR4ns|c z^HfGF$zcoHBk_)`+Z)KB(gR}a1E3AZKVrLA1UOYYier^HnefF_Ap!zC{O6hQFPuyS z>3**=lO5;N*+FgZbB9RF@7!dI5RKg$&j>Z~Vq%wtgqGmzMeFAbp--rG^nUzOXkaB) z#*tk{Wj9x6T~Q!b>pGgBBYTxZDaVID7f_Tvr-l{$jN7{ENOeLqV3QoOiP+=g5(cmdk&v3?FVk0cpwe32)f(UlH7Q9vcYFqO!mz_FS@=zI6+yX zR~;D{HRPf5D=Ma({~Q*`xon)2e&YhP{>7R6x|n7tRcvf}L%O~72he=C4j0!w!7~a2 z$BdGVVATUX%VA@#&Gm`Kh38*CPi{pCfeUkWtmu3YD;pc~sFV}|+7VxFe48Mv^OY$7a(v@%eWkoZ5(lg`6-qNGoQ3$Pp zWImcu=fAITQcWPQT=F)tNT8X+W zamHxL6-s)vcuN$cEiL}uWEkkt_V+S)Ux>K-im zaz1^5V882_@K9buoUpQQJ1n7ZSTv8H+T80ypD_FCy{N{whR0ZHMU}Y1PZ3@s)n6aD zJ^lmbss%W9iO!>UezES*jQmg^mPr1Rjqu_*yU%v=aMNECaQObnZ+#l}RmPdf&FFD+ zc~Iuy*Y+=Oj$_ihqJAy@mrWnabb<(!qf@=9;04&@O>imA5K?slQZ*}5^?ju3Y^3V@ zVP7A>p8^(cBQ5-wVd|Uy8k@EH?IG#x8|)KGQ{Mn2T3r?n0}N53b*(Vdg>h2U(}M`zIk_4TXl%iLaAdqdb;1K zn)9B;=$rtQ)sQ4>=gGj@U5r0ap8nV%H3=U+!7Szx>q6Z6zEP^I*yGRyN^tf&?QLx4 zIwu;Hj5Se|Ra~HdF0Oz*WM%2RP7+e-(;qatvdrmZY=|zSrqm<}IepZ7oCn1|@WM}| z$w86&14UjLvP%AO)V%t54WwC|$N4>s?2WC|X49-r5@+xTj%lfDGAU7;F7z>df2NX= zy)ro7vzM7!b(l9J>{bwyLX?EHYTY{^NkfObrlNV@UVSM-u5a~1Rn2?pg6J-K%ln_+?JhL2Vug@u7*{EKaf z5VPcph{2>Rn9>t85s#0oBC(InNZntvfVDO9y;;t7#OOI2mW^Nx&oPNnIO1*>pY|pn(s5LP#s)V}XjY8+Y%Xm2sZpJxPf*D+rn^sI)m|3r`?qF`@5p1tW;sWYD=7 z#$@P0g5Q~CYpoAl(zwWy6qn1BSe6sW3_QwSB6b&V&ZwDo`@VRZl+wVGDq@P!64;{c zv@mm*+f+EBznW9X6jfig5PPU$XP$T@xV_y+;fa(|l5r2rqqPXXF;|Yh1f7t^BULv< z#z*D`lQrkSvPWw$!?zTDsGUdI81~x;l^)jo+uiElV)ihuJ^~Zb9E@T2sqj&3UD zHyj#4d~CzArPYtqK4pF&tn*Bp=WL~m-^AyDDx9h{NIQzYT_lseNdY45?M(GHsK?z} zkqzc9A!z&@X?VPRURADLRr=&cQF2(x8v4r@v*#ZmQuWFB2e#_5dq1_lDo(#7Ad<-| z0&2*2Fj^*y!XFIRR1qXeDWA5re3eI>9xKWjlb#0JG)VN{s|5<)*Hi~Kjh_+CgxJ64 z$OoQaLr5;(a9x1HedH>L~E)! z2ObpoN+Yb9J#x56nj z76}H5N}GOZ@9 zNEZvY;%{xw*Qq(^k-MR2soP5HYp@kM>oGG}o12*R>qNHIrc4oc?^6t`8bTIKX~3nO z>cb?iGJ231xs-u!CxNc9*DxVFb@9=H(s`wZ@>@c^jmQyMHJ`ox7%j-dhd^@8WKifbrhIN zuYrV$^C;sDjfkex`kHd9Kvi;B{xNe52CblRTlmq#F^)!DI@Q7C(j8e{IL!F^LzK~VriY>vCG&OW_Qn} zXx{F6AYA$Ph6d_N-4iNS*vPYs8%&h|!1NvPYTGi*eN9v92cFXCEM2VZ43JqMm!PWJ zah|l!9t}->{aJSDNkiLo-yFSc&I_X?$5H|7T!R(LQ|~A(opCk5m z7bm^kPe`%)ir_Xe>(-IyBW~M<9 z7$m2lONDrBw&Xe8VKd~|M+6&x-l-&)O{tb%KuJhH zX1lZ zu!kRpZfsNn4zK2s(oeQC`%I2gQbPJJiOSXrLrGPAtnPbNnZulnd(Oy1Ki=M+ab8N3 z_eKSRgF?6kbVQwm^PAZoJhh%t;Ui7H!78%Kx8DmtqhT9x?-sYCW+nNo@VloalaHz$i1|!UFKnC-_NC1R!~M}r);Rcn&{nk1{vI4f)4IJ zBtUQ-+$FfXdxE>WyAJLaAh^2)5AG0x1qdOR_kZf%I#u`F4|hIw^>kPD-nC~<@2A)L zJqiM;dIXbQ+|lB%JS_}tRuib$of*tzl{26VcEyNXA2j!`K2uMDza_)!_OFW86xvm? zKk;f_prz)S3r)1OW1a*@efZ+4r|}x|I^-_Cl`-;xcc-g;OG}MfozVSdIdiXfpdRxZ zRDjQB9gnanDV1(B>gMoc)tkikRsccxW;T@dW;S$SHXR8`C$sFl0;4{JacEq`c~br# zfxrJ782J1D2N-zgYr=?~abcb(Q^B=YNPC8|tR9W)>(0ns8 z`n2|@b`e+q=3Z&3@lu98jk-S<skwth#DZ-waxJq696fp4Tie(lDA3sW(X2S+w~%n1~d4;0q>$`}g?X zOaV1go5xKCTB@F_TXCDPx-#-?@3nGA-W<(W{VDdpE^I=A5@BG;v-0wdEUeJb%{qB) zuZR=e=gwhT3VBrz@IQdhACyi)a`dii0Gn>Oit+eVZ&n0XnOD-M%O6izDJ_T3^h%2v z_c{-lU5nnw*l$(bwH&0D5#pYOo{!}ZQBjq2-C%{F0G~ln&0Yw zZ8|6v9ktq(Oa)D+o3m4vn|Z=hM>fBvhgko-k?bBRUH1OY<>ivWloB|=aA<$*K&0t@ zW?5~}4H(+0%)vAnZ!JD;9zH)^DgB&%>F6*xJ)@(}FRzl*@3j5h$wD@U)Z{*V&8B>* zHLvKivt`XcKu8Mv+CF2ZJC3*G9Qxc@;(SJ?W#))no&OLy$2ufVJ%)++>u)=C=v|cV zixw?K&g$3iu*0YgZkNQRmL?~&fvPgH!a0UX(#jW-&oHce3NWeD!8n|&0pUq?jDzl6 z<)iSZS6Me+^pbCadTJcopXIvi%{WzF`ShB&ro7}S+r?}1GmMPSUw;1XLCH-8qi*Tj z&V)INvwCHT(=|#hPc!-9I9CfK6Q?%SWE@$b!6Z`5Nx|B}fbqjY<{LW689@tPmaovn zMiZU-_38E`H>NuFuTkUa^dzk$kF^iD8cw7k(0$LZ^VM{(m{9*uxzLZQs$X-!Y)6SY zJ|!}`s6un}d8 zPon2Wg0<4Z^zqc_V4=JQ8pTm^ByJYiTSbve-g*IW8UP%iY zfiP)>=h6&cUlL)Zlby_^5IucI2S)Y!e}Emp100(lt?AIR*4%x{33C*hpx%?6aG21^ zk2gfxV~pOA_e)J^e+6)SL6}RBuPLKh-9?C2e?h}kV&o%AlrlnoWJ@>6177e*x*LlV zbt+3>M^zs{1*7s&e9(6#B@RgIGY{q2wEWvhs2%(}Q|%p5Vp;^Fm)acdqU`4i!~-@& zW+ZzxHuFoww$G>uA)8*l`u+AU8!xK5Su(XKK7hn5ODVaP@9SKzhOI)I`nHbZrsC2z zkb|~cFYBs*fR3S-C;_!=8)n*z%V+9NAIs!PMtpf}Zj)wz?k<|y3~N`YvY^9GBe9f{|FeH}kZ^MeJ1YKS zlk?tIT*)OHh9w{tKyR3@(;!4!5CZ3K@eV{+9%4K_T|(OKdE$mC+u99`azcKiRUBK1CFu&_*iNOsryfO!{EyH%VCRU z9=h^`i-ZeTiQA{L(WX&^VIv4p+gmK`*kd7w1COW;f(97w#nqZIAO(_{e_g1(5Z=O-gB^4k#2_< z)!M?MY-W~%#jt5pAQ1jzI;)J|66El+j>)hxC-xJt*8BT+K57x}_z=R6QLPuq|Mwll zo0@saD-c$TXbYr=i9n0F&hj!81b>Cid!^G8P(28v*X^Z)i_tWG7TRto<>z@o#vvMi?udV@r$s#3K<^I@Fa`!oa8?lh`- zxes$0b;A5D*%casiIjA-YD!^DIsLxF>a1MM(-P1^{^%>9>=z|t+{y8-DpGN9=|J*; z9L+D-Bx(MaPaGwXb*3FI|2h|&H-b$U)4a~o4fjd9oy!M*2B%l4gJ)@@e}LkA$4P;I zfUSyuLjZPcAOHx5fB=FA{?7*h0OA0sxx`dWuwAZyQ3WLxHO_O2_N(j~I|mm2TL&Nm z-^PEWehWj>EAS{0%C7?LZg8Q3M{lQRL-UEo#$|%3;i9g5W2!~@$dyhtTT%!XGh*K; z)L1gUa?826&_tbHz_MfJO@HN;YYf@h3NEMT#F-MkyDf9R#`TGq=CnFA2g}|<0^(I) zroiUgXMNKL2n>F&kC(qVlT!Suu~uZU^L~n66V1eMuOYH65!}6a{iR;}h2yCYaz2%4 z!2dKMHr~USeZ&8n`Elx$|HVNOrvrdnYXU36q^_h}4v|Od|2aDCEz`7PCRy`M8+X|~ zwBAwE<+u2b(5v{{ws|-T2k2 zON+mFlyt0;Iw>6xkPupGm1gsc!Uifgx`lPAo{I8A`r+pnU>x$@>cqkwY(@2O^+-TH zO=JOM>jr=2b|z>}H%ShgT%pgTtP(_3ahzO2Ivh5oU3nv~CaM{=D+twyM%btmO^p21 z%5QesnoP+ns@CBWsW_c+++S+LQND$Coyf1fefO6I<1Gs}osL3NglOSGEA(7Z>wA-i zw{E;fTM&v->ZY1l0sdmwBHmES=9r&~Deoz(h>CwVQ|hdPxg+_=Vw@Bi*mtXP7cj>;AiCuVVJ+&UqN^52#M;Pc6NigcO-D(a0z& zCElkbc_{J*wGSR9R6}Li@JdC}%R~$|YthYlQ=pHRWc6l+4Gq?8m4vn+A zsRV18_ z=qSAuUCw0633FT1;ugmm?lfC7t(=YBUP`By!LxicRpnDkat*1xW-#hm3zO#bOJ2NF z8g*>FUFWJW(@YUKInvWw+h|cDg^5Lag_#GPFGnSq0X2L7RsDO;kU8U$XG&6=;JOON;DyT z-!V;StMj0u6M}XTvpbQFSp9gPYl7W_y(LP?=%1?V4n#dA@!v&NfevCsnQ>4HlnuHZ z#fG|Of0VXVbs^=5MxdGLF7vPy(vV8$tnDWeH9oQ@?ztKoC+>Ie<_*DT$2P|KibYojM@6- zH}0rptrEnZas@xHd9xA?2bVP0u#ksQv>ID4XEKbRlXL$Agl>*VU()4l={N>5v0iX( z#yv>b>=LbDA{2empu$G~G$9%&eIT1gJlubI#L@6!n5H+he(H&_It2wvN~}b2usXt_ ze07BGCihFFk}AfV@P@oT8GQ>DIA|<5CR2;>u5by31o!2EnZhZ1Au9|h*My74kcapP zr^dieytTV%J0kFO4b_hahT?1f1DuuX7~gOOl}^|j*v!Ne1wWliYlNEAvU?=(BP7i* z_py9YH4mFmem|UQ5FvJ|e(AEy#x@f+zf_c^lKCNBJV%X!o&3LrOJO4jUK1?vR-X_X zxf_wrKUwT!T=-R1x$u98N52s(g~=33R_iGAS9_z8|6rc+f)o7YI~P~#JfMb_y${4f zr)4rqBZXsQfkzo+?)_FaM+$%I3@(4i^gg`oRm@2X6%%d&!Fui3xmrGVmWre`LNd2m zhGWUz&OSv#;lwoI?hwAx7;}?CVoJhNQiQ1@-<#EmT$YphIG~E3Ly?C zl0M!T%{`1-asln5HaRgPm6rKRY3EH|eVtDtx)Iu( zr#Wr3%;w=Gi=lN}oxr4ni6|-|9=LqPYO)fFca%n0YL!;>FQW zX`K}F_e+SXkF`#GLRxE~28dRvx21R;)YM3vxt*0SvG`FX5^rfNLsw+-`v-vtjk&-c zR)e^31OM@!>fQ%H&1VGgBLQ(6M|)tH;icsy8Vq8@2XlqPaA*Hv$TufN)<4s}(8zSd{ zZtqgR0VcmMi+^>s&A@>}A(Ha7Dz)@^8kBn!>cXL-!f6iWK|}^i;4fwu%u*l8m*SVp zx*n}i#J6G;z)5AB>b6||V5mV(L1}ALYJ$D9eyAdhl{+Ahg2zaS0|DS^-ekXNm&7Jm zjg!26u4hpm0+lXagg?Pt{8z;USg|4PY-1WBV&)53umg}pnQ|- zDe&_PvW5Jn%#dqgiJv)MnY}jIFs49c^W9HH)9l;#991d-<*B`zG$^#Dtk0Q3C!7}y zN#?M4SG6u#lcZMGDySmYdnHinZM{DnNj_~}74I5*+)2=39{~Q5dlfCjCf7-AUUVwt zk<%6xM;#|Yd*T`r1GCXcfE3KxdL4*rI~MUiUrHHYi!vGCl)Xp2BM`*#QV|j%Mv}CV zp!?-K2NEr+d3+CwriVfjr5r|)TlF#u1-LSp`?WY}Kiw*A;J?Uk4Lp|55U<5jBvSY< zN|^1^&!L9)asd&bk-aexjk6#~p#PH3EJD|pcuXA{n`HkWuCrsYxs%n#s+->DMU`X{ zFBG`bJcWIn%VPcIjC~Ro9sv~9=fYNZKRp7s;;#N^84*8f`P`7;Y_q2=segc zKz{qqYs=?RXlQVcgysJLO`oUn=))a+)aWoA@166`etZ9kQcb~6zp^@hL&1g6lbHCQ zD0DBxMD)DlB+$BK66xyn->cItXUvy*b8O2+bdO^q9|}bp*g|};(MrP(tlSL_6eP(l zT9Ic5`FwGo`=GRcoDOJhu+kgMp*zUt6-*&1_=#=B8c2IAj@P8v;Sy%@tDnMTa0~_Z zE+5yVHM)I7nLJ%SXM``M9WA^)nQ|{iUiyRUu;UVzB-myRf){cuk3P)7Px)Rr%UK9e zqhx8^B#0XFi-#T+Lu+HK`v+h*8T35QU-enSO0-WL zFO52;fXgyd{4rPY5^>vs^{pLoCuKeHPkt*~2E6+dC9bp9-P!ZC`w8iF6N35F0#cPo zDyW>NFrrCIKw9WMdoQJ;@*sw`k5Ng~?a2M9hez1*1jsHDQg%!GXl;=~@%DOS@O0gG zO_@PA<;5%eseqXQP&`PE57Ema_FC)BP0t4 z&V6w=bNGDe&GRD!hda6S5XDSIiS8zyeN*;eug6C#3P*~mv072tjm$iumY2zyfF4T= zB&=xr9S*{p$eqVWiDa-CdYVFd%<~ z4P^k?_Y^vD(_nH#4>lp3AmgjBVIq@v$&(x(F^*y&ZPH~%JNbcT{$XX?<(KS<4lLOr z0A@!(vvJ8`zrfBAv=k;dptFRRV6>>^9dDXD3=I1R7z3Ev!lz<)0(DHkdRIMHjCq5` zZ@2=cDH?;_y|MakriMIfO&syovhkYxVq+o_M2!LNIWhhr_d6~`YKyTi?3y3t0H%j3 z&ZK<;Mjj&`QwUSUzfMRY^Nc3xEck$*q-aEhT9tu%|2tfO#F8O|p{JMMOupxwJFvx$FKd6!V zRyNoRNxmx&CmA99JD%0Z$NnQaFKusQiVxA~y_<*-#myi#w4?*dCJ2RDQvVUr?T)YX zwOI7lG-+W2$&fX3IjmQ#@q@s<-p(ioExg60r?ru_Bu0KTP7j&O@9q z6;UDM5lPvtD7^f~kOp4#^5RjkVR#}hN0Rm|J6mP? zb2P>XvGm26JMxk9tQx>QXr)A&-cLv4Abq)5W5mTxrYZT*$6iu&n+8&6=rH!An3rC* z%}5F+!3M7??ul~5h*SN(t@LU^3?;KNO30!?E`$VmXJ*INed|5Yop=k@rlG}?EUwvb zF*ro{ZJEl#OnBg_)Ngv(Nc~b!@foxceBW=#ai$<5t@KiHSejhAE9jmHs0eHAmB()Y z%jaluHUI1{A>aNUBBZNe!p?Qma7#d+o2*@z)1PWEz~oP6**t^V|3{8?k}w4oDQ6R+ za}+4Nx?B&i>S}6i{;7PkR^l~uNMV6JLQbVVgk7uutk-TWui`{f4{W&zQhpMpwuQX_tL}t$X}9flGeBsP?%pvmVt3ArPrty zWk=|vyIwNN8heF(L(2IZ7Y<)9>4Q3_7BpEoO$7BiD3Q&XOFl$tvlA|s-no|ml0byR zB#V_q=OoL7^*tsVcw+*iPa=!`l3Y^vo7d?!0z}g|ubX(p^v5s73Ab;BqS7OS?28-hI!uQ}2NcqTPiS4FNUxJ} zb`z&Wpi;)IH5GD^w`%dhtk~PDNH#8VBZrUp)4)KacND5L@sdM>Qx&pB-mM>ySmvW0 zIQgDJ#({abD$1HBT8LSytjxm|@0eeR*B~-yVVOvtx1?*=4jg`#Nrb@pjn!D!m_XZ0} z407H!e_!7fIIX5<>M)z!VAoT`KXiD<+QW?gB?KTYzWL759H!qrM%+3s|g7iSIn`cppKurE_uJDWH8*RL+YV4tPM z=PcyJarc~0l|`=ZJ#1#vjCUb|CBiSb{}a42;hDG_9!nz!&4VXEvIFJa=(z)U#Eym! zI{~VTfptk`O0>29UfF_+yeTGg{ReLzk>;NUXs4weM!c{E`W{r4T#sI5Qnaf3`O&ZS z*m8$mKer&>W~6?>c?Bpa=F#=`e3r6uJ5or2|6y8YHB)WrSA!SIpPQ$8anZ=Dz%;3Z z^oJ6D3QP9;%ksHia;xd2yR9Vs_eb|b_)pp>lt^>ThIK{D@^~w8ME?LQ@ax6?W?@mla75!J z9^^nF*yXc%1T!JR4nR~xT8X?bWRn?-OouA!VaEVJ*7;Nr>pd^|_*I{us;jIsFI*yW zvY`6SDVl%MrFH{(;*zEi42UrjgSpSy)s)6_oM>~DU<-hQ>+>dMw~NTg?Xgw(qQ6qp z?`54DN}eN?=^*4%*nEU=ueo{xy0f{I*1?AB7b7n`_>y_;KTWe5$D64CC2LNf z(X+%R8oH8QNi{K_pn5uR_}LvS6UIHfBepcZS>58q{jj|3L797YTTNpz+9~=qqu@{j z-4DJ6oSTy;{CYy7L{K6A%3tmfQ9F15HUUXQL}1hv z0Y-Hxw&WkRuU~RQzsz}B^9f}yie}w}AsWmj-O39S%hWi&N_B_`5_dn9b%%$7c;{<8rp`G`a6@O8D9&2{ zjF5NG+LH*!0DPnbeoq0;7UGVqa3$-DX=Lw0wu!Gm)uyx zmADZ8TCy_$zTX=f*|m~3wU&TU*RvdO?WZc>zka62_*UM6A0nM1Ie|AubzePmGI(TMG8`a)L?l;*OhP}NBp3tDb{@AVZsT*$|OYt~*C_m-qX7ykuUC)BcqaWFqn zOvikaEb<`ks8R3;ctPnaUFPC)9@)8^>;$CiC#gkiFy^mTP7Aifj$MgB#MZF(b~}lK zCsEL*=e0CwC5ZneZnHD!$mh@fhKj>tcAFLW=!`{gO@#{wSnNX#3B>_yk&v?z$RbOd z=_&K96=n_h`cv!eU`e{b{=kf%p+ePls>y^qLacb+vh-|8u;CriQzBV zB42e2t$6y1%VKrY=2QN58!#jl3Z3}Z6osYw7Cars+KFcnoxv2*3l7+~sZ{7eM8dW% zzJQ5jfxr02I%OgJOi|NQQyV~=9{CzCjv7RpY*|} zHHq(V+{L*x1N}sQ0mFLu%?THxqmQ)5koT5Q;d@{6a8t?t4IR*mn!KM%>$F-B(vBMf zby2iH#`oY3nJ>gZXOd#@sW1dSXs>S@LsDCJ9 z^U#1Xk$$SFbD149)b+3N$D^=h)6gns%{K$Ifj=hHd_pA;_e^PvckkQsakx4fLEsyk zCu{rj^?(Ud$e>}je-fWXG<2M6&%^`a?TQm$fRfJbSMt&UJ7ks;;@qj`u$$sTD%J|> z$A*bh-!FKEu<9Tp6}UcD2|!6d0tJfsPV#{Tl!luzLx2O#(Z@7XR|=K$7y?AvTjCj= z2F03uLH@48MUB<0xb2cEx%Qs&scfg8V6Pz_9|1+T0cQzY!Jb+8cPUm-#D|g&-t`*oUCgN0f>UX>zszjV!L7}xUse$Rv;J=H2 z0xEQ9Gd0e9S;Q!nWVNg+gXa>zp!ka10DN^q(yiKKj&ZSmU$`_|oH{{6xL0*o!lCBpuUrF(2>XXNbUo{fq% z$16I^BxfYjk)*z3|9i_d*`Cr4Jl_CQ6_44t{b%8IW>fMY!dbyX@s)o71;=&4=j%>q z%3$f^;hqV1hD@3C3_c$ZUC-?=+sK#G`)kyuw;^NM=?t79*2#^v_T~W8f|_hQ6P#nV zsy#I?P{5LWH(}IK2-rb(A3pYgjiY9s704a-BTFj`P1gN42%=hn%3Um3-5h0_1+s&_ zaYwTtF&Q~J`vWEdAye~xcNm58HUUO(kKjQMBO2pfbavu=}TRhlDmSF&DCOVK&bR>b{4y=J^gt` zGDjHZnPNfi%RYwaUtI-$brn~|ab&a94$OWna!!VnocEzHcI%|#&S3peD3^_Y6(>Qa zs|5zCk-xut*<`BEN@cVWfBrI~8wn|HiOtr%cBXrY&JKT#84^0X<`X+qf4Xcvcj16$ zO=kGxqxWvV@WggGaX+t9+PHNR#SBvLqWXII6q@yI~n}+3hbUN7U}IeiR;%GK_J;6Y`!@v6&1KsHBKZWw@eG zKbuFXc(nYu>D*U zt(m393gJ`>!*{KPqxb#d4r?gL@!eo^%eD)JRJWQTN#V7=+w88!P_t4~n*r@Ge4WN# z#Zx_0-SB|rlcHkjaZgZ(69@9XcucXu%AkBS^c0`scBX!VbqWWW#cuZ>AfdwoS!1AK zLF-lH0Gl-=GL5FGC^ui0yf_D>=DVsPJln;NeF7!(<~R!Gbo^>?i|Vb|^n39jfpKq6 zKz30jl6!Nt6Dvs$F!PD*YEjaVRQdZS9-~MFTtq%?0*k({XG6lF&!=kKE9^+4aBx&n z+H&Y8D7PBhxpWs%rxXqhD*#8d>rxKe7LE=zP`bi5R!R2iMwtWSQc_+;M08q-#cRtV zl#wRo$nhfx#sEnlI{o)~bzm#mWA%Kk0P|@2X12E5v*y^H_-ljaj>n}C$kydFYMxys z-WRJrFu}RtNm=Noct}2P4Uc&K2hprvtCv*@aKeZV%R~DBD7?X7WeD;?yeOMWVQ`br z>#wV7&-Lw)KGxPPNm=1s35v`9^a-ti9rOxcsO2vUR3UlLD@Br(m`QrX>__lFHKsB% zK}MOme3V8v$@dVpO|`_6X<$2)k7DgUPo>QdF67FX_A=V-$HpgPsUhhGg{~h4y#Mnf z9VcrDUWaw=$Cza=U@Q2&=`=A24PcBY>(STMQ0tg7tLN|&FHz;r_jG1Rf-q)`1@D7w zx0g@voOP6JwCIgbTkV}a8;IYK#B z!0*PeNE_>)4#SH$;yh+z3^qng$KiP7X0*Cj0T+;LyquthxB}>HBBZsh^?eXO)y-WL zX%Vt9nI7ztQYv$+8}GPqVK;Fas>)hVg(a(=;xhNO3`^hEIOeDq+3toScg2o?f?~dv z;a^1sJ$u0h_G;SR_fdG9WOsw@?|Ou!>mWf5btAhR%+a9jDSUDl_*q+VDpeenVIN5| zkmR<&?p*8QFD%gSiTtAD@RU~DR=)!PsF>7>l_{LfBkytN+*fUbyNPqL3D`hhXefjU zSiN?j0vGD~A57b6v1&BneeuTWo+FcuK%qGv#*ToUUB93#xbLjt#vf3r-N+;tNy%nr zP>atbiS|(=AFHm$pxofe>U(O2jQIjadm!VP=N(U^7%X9E0A@(mSQ>@U1 zRfdNu<&>eGFsMs9A5!4=>D}A!K+ZxCzF46yCv)0N(+${(R8!ef(E^xihF%UAzhlwr zvg{>tSiwBm4EW3>hnqQOan|7x7l6ozSz8j58264R8C4P%=bY34r9RnM} zyW3q2o1^v$Y@%FA+Ii<0M~M`sq63jNiY7#7KY(AmgURaYXVQLU?6dx%V^D@wnkFQQW)*ZZ``TPq@kIKcz^!kSO3agVG8_l;y00ZrJT$vgcpJ7nf=7ZC=uUqvG0TpDTCfC<&t5=_NR?Y z7*d5$QO0A1;c;ad-$yKzEq<5>KPY2-Se@uV2?`$^-;S~R3ZK+6uk}?x_qRQ3dy@_2 z<9Xn*Oyq_CD2Rc6IT;b&qsveRrZ_NZ%#kQ22qZ`0;kPl-J?xFOE)2@3*{p!+78aJ> zP%R!t3l}!hD4X`#lk0)r|(G|tyzOL&af zcik{zBWs4l0!U>Cxf80zfT&Jm?PH5XneBJz_koKY!ae$9(TdzdwVoR?;zn}Q)v4uT zo6o9YI*XfS+fdmlE(Oc$)dyTTBzF$(OvH|UNOZ9Ku0++zwP5lPwyb97x$9Kr4!Vmg z-Vp))1zmTa<&4G6^?y+rp|4Rzgz5J?pJ`*NV?5+j^enbo$!JNrBNjG#p)+sy+ce`a zV9(k_`_GVKW)sjsa*aCLPn`1Y#4Qow{fT@Q`$={Um)-`{jdXTwZI#7fH&BQ}M0NLq zfHt1%0}sDEa8Kr3_TUkpjmKskI#OJd!m`3GCiLBItP+qBmDf_;wb*1AS_nV&dC6LH zDR1txJo@LZa>OWB`*i>6db>(x6jr&|)~VYSJ4DF_Ry@uX=ZV;&Z`TtKrF#yzbB-xd zBi@6&&Mk!-w?Q`^Fa?5>#9J&L4v@;07B$*U#Y3(RxtRK{nh%<*t~1q5iZ3YISvioE zAHSn>0AbiG_?=ISHe&aR@oFcitpe-Alp8??vj(W3gYE>o# zZsuTh_J)~84Z$0Bv?fR*dELS%6dp}+Xpr;wD*}qJxmM=^(Kyux>!Cb4`pm>npI{2E z1L~HA zDx_mj0dGa){8GdE9b@xTs^6+OzfIcwP;!sF-PznbQ3o_?WYTbtEww+B7afbO*H(c= zJ@BXZ;DkZL*zf@B=%-KGLswhHBP6+=mGvFPn)sX;+w$kW-sx?L;UJ~{VXD%TKW`qW zQ!A`J*II&B$^)h!SQQB^1Z|XD5iFwa`BKss%8jLY;$+=Ma3#Q9cww#kH(Zfqb>SWP z8VmVk^GU+tNu4TLxfn^0C=B}~AK0ifnB7E!^0j8%cHA8B^Us#whZ?VS{{Ems2agFt z^l*UuR=FqC=^%9wo>4!P>`WGHN&c!qv$06BfG^yJ8J0 zbQYzNJ%r|Gi3xZ|)wL)gp}6$@v^OOzH0<920aV(ooyf#v;V=R@;Tp2Sj1)(fD)P~5 zRh&)ko>oNMlP2-%AsjnHyyeFdd8Ba$;uL2O5KHUy;y4w#HgskK(VTPQuuAn(@H#Mz zt%w3Ma$nb6kb{e}m!lqVl?VsQDsJeQv5aREFn7*sY~-E{K<*(Y;hiSa*nA9|%%m%M zWpu?LMYCfViG`TypbtlnCu1oAi04pnljU{IIFf9?FGl+G?-)$r%O6_S_cprKkYgpT zjSu2T_g}vkWXXl{iqUV{xw~OM3s)kSl468}1QptsV>CMgLkfGnT}t|r;@U$8?m-Bw z8hxoX7v&V1org+EV?mZ#9T*isSRG5D$NX3&bb|F1Vk*((6Ds1zW`U)VsEX*CevELg z1T%1!!3-qSc>M#<-o%l6|8uVf`-A4g@abs~vn`8S%rdz5o_-DN3z9<8!dB7_SaR`N zDu9t$X{idyk5G=7bES$!@gW)992*(B>rA6J+vx(mscAN1Jo`WoUa#sjg)!XD{^Svp z3q@XwOzS1xgkP>ked%sW$*Hr06gbLZiEbQU``i*|Qb!@eOd1o2VHP_K6=9Tg2KXc- z6gn^GVDJ}@Mz3mFLuq2(D$Tx$NiB=bP|nZmP?E2>O-h(|?ltE3C`2hLZ0-?-#LnL; zy^vFhJnz2GsI(Pl-jwwZvu#e51QJEDn&bTsj;h*B5;vFO35;qU9=9fHvZR7O3ErKq|U>8V>5Njj}O($bHS z{?d$_(ne`VE<6+JA4*gcbIg1Bw0>bF)r~c)^mRT0xI>DSJiaKuDYlJ=qPdKU@*QV3 z+)%T2xOXCzPo#*P+yqYoPxC$y!3RvtG(*gfJc7R0UBa5Q#3d8f&Tw!|BY2O$yF&Fy zNba~(es1E#aH3x+9UKEAH_WkZdQ6cv^C6UGbv0c_AU z{$Q>v!2VqS!gE<>nXL_We~0zvE?6+C0fd+InGxI&ENf0q zWNad8f~E_hitaiTn=lr(J#_W**ron!5p~MPZ(4{LgH~cBro1enN{?p~b1a0sZ=6d; zNF*U1`;mipV3_COR~je#AXnJjT+Ohz%7hR}K}M_)2qLQ&)=FZYb6~Ggir5h`q5C&~4r@8w$)^p(Ou@jX%(nT87out0$i-K?Wg6U_k@ill zZ+|23p|JE*$AMvF(dd--DIkdcOXgBXHzauYHx{H?u)kEBA8-74fX`GPCucm4o;I9g zA}j}aexTG5+mtxBnSU2ap{FqWb1hL3e=J6El$fL>Q-a251{%NcM^5jexyYINmv2Cl z6O#`v4gHDK@siaS)m!}`ay=gKcooP%z_kIP124;m*Mfl_6(#FC%7mbW$w$!r&TYBz4Kn>d2)na%sD)=s4OE z`gE2mjZEJc2%FoA&9K; z8^9i)HcO^nJD&!Qbef)YFeFH@UAq8rl_kuDiS#~>tt@_QAC*EcW`M2n@Mb7TVH+?M z@OXBBx-y~Cq|3}|JQR#!j3^!4s1WWRghGFx_Gw_A8;@R29OH+m~Z= z)q52MV_2PN9i8QQ!#ANSNos%MEHqt=w!&YogGW*#!Nus6^&+jeBj({GhN$VTjF<|o z?PzTTOKRVTY#-GFjax%}1eaLX-M>v|fST66!0h%^<$yHQb_m<%$C4q5Qm^(Mfd0Gz zy~jB4f?P;gO}Z0HmQJpzLg~?@@3@OGe%8fkX=3P-ON6cTDs-2sbC6OA+q?r8vhF-} zQ+ntXnBXeuRuh0&mQ6hvp8S9;5X3nl0E3Po%-* zq#OHI04M#JdyJ{!R*roVi%P~|t8I=Dn_~J`8BpullFO(;{LXR;xlP-_)+VS?fuG~} zSl5viF4I=@F|D#cv-^V@LX-~*;tcV#5Qh`Z8Xr4$`i%g8M>D3-`taR=(rW%k)j3126r>f8>@v_p7iP;T--+$YYG@%nV0(Q9zMx4P%|y1$Lc+@DvfUvyY=lO>z6D+GZG~WxD{&`F4gzJQ-pQ88XS3! zxdzV9^@3W=gxp-d9&l6lGKH)wAcwoKN&3g zwj}$Ly2;O!8R9aTVT0%8CaH3c8!CsyAG51VLmdk>^6qA~AU)^2$My}F+Cz1C5~n@C znNnD)Iz0~DwiY~b5i3q%a7d;PzKJ#)ihWqG3m+G)0LAj9Q|j^}tbApeex3vv%5D#VnI5!C@sx1RYK$f_kFZbf*Od1q|?l7MrIwXyhil`r@Y z!1ZaL8CWgG4XW5jN#wItfma_j(fI|a2Ii;-k|yk{F!x!`<4UN$*deqrW>HtpUKP+^ z6}*Vi(67qT-+jQ9Cd$wkp0d^$eQ!FTYa-kDh44-nm}!KGR*|JGfuAzWgAYLA z-e7wzo>nSgPvxG-|9(a27u5oDa|@BxxhOk3*K_k zrVZNb!k!`)jFrhVoHrvw7!nJO8!J=qX>({|1-8Q{SRrV(ak)g^mg_=48fQjkzkO+) z#YG=32_SRMGcyG__s>Vxb|{ZQC26?22{`liYGX*qG%uoPON-9(8kj{=W^0)~=-V1M1G}!o{ z**My0q;FwF)EN^?m8%IFNH+7M<^m=HwZ($L6RDHQh2_4rX9TWAgDX@{mb_Fu zjc|#XKJ!MNxaCg8l04C6!)1@G#k!{J zt9eQgwDcx=)IkJ1*_o7e zq^)cYtCROs(T+C#s*>`)%}w^8EHyCj*oV6aOc9Qo^7xGc;WJ@HyKg^gIvGcE$!fOhtU>B)_t%0 zUTz#-^;O`BrS$|CiAu$=D>~x*Z9rLt#?t9JXXTe3&mL(|;R@zm{;5`SXq8!B=0t@U>usG$nT@hc;@3B7 z7rv;td5QHu$t^4XBEbizA=7-4h#y)4cdg=lnq=@Z89|cxer-rd9$`I(In?rn-YUCx z^f-9mY~)Id-5JjRwipAt5IQSZ8e8I5hqe)lVyj4eJyGO|3@ID1ML3PKp`Gk5qqIF; zKO06v0di24w~Mxwvf^-6AL~?-__;a(<6i8FBKO$QRGu8vWqp<-`G|6zpZ2u34!Lb& zCU{B?&3xDO!#}i~kqWtFfZ;nt{Vug94Qe9^gw28uqi`8UwNYtVgH>)Q&ama&?tOcs zBd26wXzA|V_AYe@Dj03ovM4SOxLaLWeTNN&=7|>t6qQ@7 zWnwu-Iw`7oiplgis~uU06+({&X>`93wM~!PLO<4o=RnE2YI;=|jxO}16ErMSlkXxl z(?QJC$xj^BJ39j>p2WR=D-m7+{w6;+PI!Hu_+~se?|(QslsoACJX&a>=WQBq7AvjC z>h;^KJ5Yz8Z-K%19QG-s8$@qrLY8D$KoOmaK-Cx*aA-$QJ~5xqS;)Y+8B9GE#9yx> z3l}r0k%Ko?D7X=VO6qC+)7nx3Jm#qP;G&0nA%ZTlcyNVawB7e>XrgMnZ!xJ!1gBRx zX>x?=IxVHH*^fCu2ka96F&Nb+Q#bei<;F6VHkhK3D z7G|8l?uF!BOGfZ74yPVTG&s}tw`2T`VSDd_-5xkc?N|)S;BB~P4Chdm3gLxdVN)Hb zo~wvLT`7*=Nv89_dEdnXt$^=tUyyXEAjZ$uh4AF)w$z8E)_MZniVS*?hd zm)|hH6YoO#OdH?;I@>Ek@HvB;J_^)w0FxB9u!#AJYxC-IdRiQ;P&wZJ z21pif?XR>k)t&GC26U{7w@$40J5(M|lsN#Y4Cwd`&w<1W=LU?B+3lQ2{2;I`8GKaxE z{Y%xKW>HHOi4!a~oeUCW`*0wRN5R3SmL0}gg=pb?=41Zt6`q%r9QvmGtn+2-mZlFg z==rFR_>jONIEdT$@7=KzTcxKdx;Ul~Z)*aUTl}yIm*>J64_VDiuts;Kq-Yso&9hUc z9Nu2dZH;`dSCM{eh7Sd2*?J*A86~(Uh(<71PCm%Vi9HmKj9-&L(?(4%r3;w7d6`HWJ>NR|tGxG^Q>i5rxAt2&V zq=bDAhYjnB_hqqsR!oR=KuU!6Eu=JlTO-tW0()b=gI$t&&t7T+r5H{{&vXUk#Q_s= zCd67yMa8pk$;UXTbCuDQRx;q;XV1$oCC*7cDfrEp#eC*pv0u$ys5Ry$Tvb^Wo5bDG zkaIPh7L=c_m@~5m*>cf3Lu8IfHy08cacf6waZh2&uR=FmathJzip=WCg85j7Y!_|hLhU55Jd>JD%)DIf z!c4V+6qN^QTNAkn?pPRB?%Y%(T`;`}Y}mxrD_44H##JyDXw1 zh=F4yU9c$T){8tD&&bC{^mRCt6Y-#@r6tu;a^y|HQOUVw_@tyxV(B3SlSsCK&lfQ} zU*A^Eqi*g}Re%;Vs6;*#7xf6wDy==Uce^NBIW$=L`jLeC+V_A=!9~EQ_bPp_7+FBW zj#d@F(1L-yhObTp%fM#rxQW~?ibxzoPqq=2ni|*q+OUlV%qxxUo}`jqGe{t^QkN=x zVbvnM?dhk@i$QJ)ZIBc;2o_)-W4YuL3tc5nQfN;YG`8lnwmH^GFfowf9i|Q8sjqNsm12jb7ViYLz-&0Bf_iPeM} zaS{y(cp*$l1%xiJ9dHSWMEfq~leiC!Z+lNsW$+PiW2*DO+R*SjrnHUg(Ji{~y)?(I zYFD~6k?mAN2=7sCcbHq$BBi4r(g(u4^QEJbHK=b`wK{Em>oHc?2Mf*L*Fy}Jo~A`o z#mLzyrU;25s5dn&B34r0f1MHPsqJA)OdV7)JKyYt(!YVDV(cUrkWE4$@Z9XmAJ5EU z@X`w|amJQUr_UeUT2to|1L%!F`^J>En3WQ-t9HwR)y)S{la`URsFlVegd?tz&+~H~ z+c^ZJwhlEbD2j(&4|KTP-0eBkfqH~DrQIcS$tn&FSY8;=-Qhx|#jND7{x@*SdG;xz z@O-35bD`?aN*2Rv%pr&?im~0GSUN#R0=TKn%M0@egPY@V9K_wU`}XhZG>5XqIQb_| zhV}9^^Y!$wlo%Sd9m0&9Z>izCt*6_Dl3*{iC>TeWYH_i#;y`v=yByD*#%G~jTVl&S zuOSSycTGPW%uZIK;kXgwc|lQShHqqwyCW~^AbBJ6WfVHF>h$o-f5|M?&^JiKISLc4 z`SI}mVsNKb&Rz54o@gQx%R=7#3HofW0RLP^Ag!DnA;mvb;QT?I8{n;itl`>uXjr)1 zeIZ6YT$NlBapLe7rT!X&Y4Ewi`Gd}JI}1#pc06PYA=t}zU8XvPs}0EOaR>XL(vU~9 z=KmZNnM{5!>iOTtbfi>Tdab83@`oV{isT+0O8=V>+t5KH-~C}fN^r39dA{5MMsWR+ zr}c;Jr;4m^s<;+G(=W1vJFnrqLb8!eLA!&P#ncKgoQj_fYci{7Ly^$+nr8hwC5s*l zSf2xqAB=z9PE>yxEMVP6zBdY3G(J@K<*JL*qLDBt@PM>^q(81nQ0JkW2tXgUwoQp$ zldDz`NHqmNWHkUfScD4dIqhgZSorT40AytLG>bBYfm|b5g6x=nL|2^zqim7}zgp-p zIMoZ*$%#SCjXOLpjITSp=Z8F*r;YRv%?VL>SWaP}I1;@&UIjf=L1z>k( z&7+yBl7k%J_<*8Z*j~l@ViSR6_{T{f?;EnR$Xs=Y?)NWh=Lg8*ZYPeOU7Hy*Q7UJ@ zr!Ofl@FTF_Up#v3>WD6}3b|F?bORs5McdyqUTSaFb?dq}9Lg5BI$V!=)E>SEfA`ut zTaujZxp=?vz5T0g*~Rz(baUTYH*lk;{k{F=a^78CkFIyah0KzC>a*j!Hs?S3z^R^$ z{}iVUj-MR=XsQ|Ur>O1ozSR#6zkruLTYB?KMc@W!%%Oit>$=ye)y4PnuBM9&6D8HN z8rk$k)x^jtkyOgYO>aht{&)o5KYhOG`+l6H-WehY?8Uw6daCT3>iwpxvHKfb%8P8(Sb2(hAR_Ekoy~{3thZW^ZBQ%^sJA<57ADar=kU(fS*}FCbzK?I8257}_-hmZoo) z8{+UjS^cNM2^%D=qX9X*4kAsND9+6EroJpc>vjo%d`EA&{09T{I^)_V+83Q5*{HFH z{TmUcio=Ob+^azT;dt}DUw>auEi<~?U2HVrR&8{@y>rW)W;H6c^rml4qE+AfpZ-Dc zd3j_QU;d&9R_ERx4%ddGmrx&^!k-b>#t-)3rHYG8aDs+=JZ!t4M^GYnHY==;ShoX4KW3hEMTBAK6*-3zSi z%4wT>*iF)LK-t-O>&Hq=y0M*0lA0=dad?z4qf`j#=R+^$dL$i$GUxq>cvfdmv5l@) zUf^9M)tpvncCpWu4KeFl09|yJ%QYbp8p#)adnY>2jHf~mVA53lsCctA!Ell z#LRFxnH}Ym>8Hs@ChOG6ecoiNBu2UrmVNBzk0uv{q0<+M9=j|GV~tb@?v+J9jWd08 zeavvaa)e=P&e|3mFi#4d4SoA&o`UDo&OPBb1T8_ho7;B7F`QqH*@+eY4Zykp*jFtv zWuxIb)CJ}{>l&$5F~Q9Ksf`j5+8dpUm;uJWPQ6#IPcBD_(+=n+5rBg_$R=;trmLr` zrCmUs*kNlWQr1M1wUB$N}LTI?mCgJ<$ zdRH2%9PB<-vWUMzhM!3a?f_*6gD-4jiFzmrhT5#dQ|^LY?nh5L%7r#n(wPogmb@Nh z|8){2+^(3zSA?#6AM@)UTSk7gxj`^hr(+i;Z${=l_gB6Js@@R?lDxUD_EPq@vu&ET z9!V1>D9z<#_H79&r>lpReFB#Kf}o00J_*;N;_C!7^=-ra69miUBDdYJ|K>{4VZ!>) zz3x!)Y>0Yohz=S+GtL+SSE0#=4`Ar~R}P zj?MrT{HW=zV8|hq>|kh|n^qH=kIk=9no{B9=ism?L!i>OLEGc)^a5&BcJi~=B2~FW z^*?u5gorZB3#JpfSo-Fa{#91Z2M|c2#tFhOWyYF?UMdg4T8}k(VJ_>MhN&w(bvaz7$+^Zn*)QS+ow`c>@j#YebK!+O#CL<< zeW;U(jP4$h&&blXeP7`+mh6|-cl-D4?=}i9RDM={P^0xgA~3!lbkokiFg(Xy{td`I z`wf`!3OWK0+yYG=8XBKhbzdDIuY5dVYh$YuSj+IBd(d`+_Q2{?=i`IUZ-5Eo*YiWe z7ys&)psDTiL)oTC*0RNf)GCo@UD;QMGKk_lq)>=M))`Mop`y1yvWJGG=ec^88QChF zNp^jkg`UEK&nm}N9*f+SP{-GhSh%^J_nPBURDNZ0JMN$W{ zEx(p)HV3&IrfX5&RV|;G*2U#h%}`Aa?J95#$It0)Rr=_-*bjo&inG~oX{nBJ(vE@~ z9N>(T8Q>UF22h*SFDuKeDl{?Y85QKox$vOJFY2!>%2$?@_EZJ~ zy&0G7KWxBiW=LUU?VEj4{aJxY=voZ;$jZo)iXoRC&eD#KRo>`raE|*emc8Nw)XxhQ z=BM6z#L7D}F2(3Ju%yImOKApP8f-Ibnhe(FYz0BaML)X81jNFjx7~LX zB7Wv2pYAUbA=2A;GTc7gyWapS3*+beK+?V{{8oRmV22vJ3DDYe_PMtxI+-!qF*&*9 z5mL~_SSOl6WTAY=D<1mn*OMQ4T~A39@m?>PV6;M!F|)Jg|25vqGJ=}Ss_d*2(JC6W`o6XA~z@=(Sfo} zG45)ifMAV{os;$L*n_#S!9~h0p)G-b!YwUOFX(Q z^z2q~U0mRl9+_}SsmPs;i3)NNFuwHOuc(EB5HxqSE=uQxmPsPh364Z4se7swLVnP<^FZ#xKF~oT0Dv&gy zinP<>#sYHD_yc@`3WKvpG4MtYOaqwk)bOGT%?}`xA|>YIEXx}`oeL1mW7Ql9xsw&7 z)^WaRxcC9Vx^8^;Jf@xi2r^76gnu6U0OQ)e)t|A|pKi*Z^`agAidGbR`Bede&@)_- z!ERJuk2M!9%G*osIi~yN=>bO~j?kgBigdAZP?#AQ$S@zT0I6Dp6ps2>t7I|lSn5S1vc~eyNOYWp#dS#5n!Hnl za}GsT3H#P3P16ojpFjBt-@sMhDOzBMU7>$U$O?lHAGG?FPm(|&gT5LA%i7XIi}7;P zm(kkZPVLjqP#;bsk-usASl1#ZCEaA5y4o_5{E70g>CXssdo=qHY+(%9bn!O;8MGO*I#F|@>t3& z>{2_9_+ue?x-P>~wsN_3c`Bj1lC{qwktnobW87 zc~$Y%p#pLTFF@M&)1Vp8fg`ZVEs*V@R792R7$2TJxzJ{p#h~4OxlSiTNAFF+{3wet zGWO@Idc9`zu{b<7NmQmpx2F#5>{ORiAnx7&ii)k7af?@Ld}vd;BFb=A?+S<)jC`vz zO{pbRZ%3xiSaY4!R<(GR^DsZ6b(WWwD^lbMj3nMc0JvT}k zsvWBDvdurJo%Y`y{-u`l(TU!2p`tK(RTIsK*Pc1KDz#}3-cNsl^KM@JGGHCQCs?K1 zQjwuR-jr((dI6XrtXCL`&4z*Q^q!sw*gcAhcvf7)Ig`jPj6I4}_`)M_L>Ekmhc0i{ zuT#vz&I|^CJD^mMZ=z>RIeN9une*8gq1={>k4cnAp8h&Ges#||m2}bR>V(c}hg!k2 zPYaj{B3i!K(|PE5OFY_W+72cr9Fi1PIFj0G%Q})A{j}F=%UV2Igv+0!=~U<>(%RvG z$$T$iT304=KLj56V=_pY)|b74wI=G3yxOz_1wu353x>v~K1pQsQ)-0uqIW9hENbs|*2&TQWauzLgr_wJa*&)o`zwZFv(v7sXc}(fgv!pptPagNUXha7Tce=-DQ=)MQOZ3k zfl^{R`nmm9v-L1?;de1iX*t;M%3HHj{XT4n^vGiOA-PbrWvF5+&q$~cXqn@bRY7BbV|1&)? zaa6Cfj3h|ZgJ07$8iaEz-w%h|oZoQ}-*-)DDQ6G3vgsh2Z6m`Q5E%-bQb6z2ez`>9Ox z&b-iUOwz3mdr-Z6CU~%quYO2~B6I1cnV6FQ-J4CfD^Ft4H?8+>=R~4!R3W{=@xy(< zNN8$bhC6`HKC_ z0MeU>sINDe(9j61tbtb^+u+ z*1po0slLbH78vJVNQ;AUF0uZ>c6)|6^O?(8TR2~fBaLW5^ldA@OR^K+mNFr z=ujoGA!p{YtM<^3C(^uYbaAlu(kD0*1@*V%hlfKcSVnLNOg%-A&FC3?^UePS(&35A zNn+EeFGCH6qsbS4W=Nw|%HgzJ&4fYCNsNF5#X90qxq8+Ef^ZSAXnJRAhkY^Ux)8id zNOuj_H!D}sQRP(}1$g)kla6MK=G(0z9^tVzQ)f!q)RN%gxt%EvPueyn6>hOR$%M>^ z9I25l1cV(N%zm(2MBNh-6Dhm&1!?!+z34ox-Wh&3Stf$~U4(ecFr(-iFvcvtIq23 zouP^DvQvPGcKcnx{~*s~>72~wzb7~mP<(9<{w78@wE_H_|+4XL;jvYAV#vNyD zQ}(}VVhXdDxe+sjO3&;p{5XXZUN`PuAyvpM!Sd4#BT!PryIh8n_PY#(k1g>|r5r+K z!6j#Bl^CSA7py+fgl0lWD=^#})<5dQsS$K{E{ncYPm4BVp`n-xA1ff7v*zd{MDb-J z(dV#G;88E*Lj98v;eT^b@{u4%%Zx+{&JN-EIH6!F$>*0}c$r;ndk@w@vjuPb+>nhz ze@m_@pH#(NabaU0w%YiD`TjHHa`Sjk{KS6hX0?mf6k$~Os_>6r72=0y>#d9&$h?6 zU$d9rB{c+L04tqtIN)kYKGhHozU?YQg-tipmNE=J6t8$=8qAD~5X(LpyD;b6o$6=p z-mj=D_8VZ^-?dpzyiHR!(5^9vRFmo5<+e}6jflspj>Q!{+gym`8wv4w4SBq-ekk)l z-R|D|{mc)@+j$mtP+@8;dF=X}_tbxQAFz7xY%j-;jFG@GLtCDAa>+wX4*l)HOmrr> zvPQ>h{!83M*lt%iB^vM)q|!p}Qz*nc)LeVXMoKK7MwyDK!)SVP?Q?ZSNttuGJn$+v zAY?4`ULAwQ9$#wDi~eK^P=ka<&OJ?7QytHVihRN1otb3mS-eiTl8?lTNKeuTtf*70 zREM3G!ysC9MH?+PW6v(+ma76sz8xA)BR0_;xcEb#Yp~J*Ygf3ZrYyvNNK2x|R zyuD8*Kk`JR721d(kmQ0|je&-$4`H7NQzVjy04$+V`J=FK*xXVV&Aj%BHdv2p*F*ZI z1IZjzOST}LMisX_)XP8Ec-<1Cu~h zN2|(Tdij&+e8naV^*KbmP>MLzhT+j~PJ6q;0c>7kl%8voH%XiR6vLl)YHXL<>C|U7uc=P_3Y;CpYzoG0rP<6899b4j^Ph zKLd+7$iU*RAqPZ+KXkMT8CQ(Qa@sG#J9X+W*Rd= zGyB*R;T&&X&^ps0xO|y!FGX`T4-0^`-hBL>^PkkW~((?Cv7=L!+L+^UY@xzpag5Dk7hE_P>eP!#1)xVz_{6 zf_c|u-7fAJU>5+{+?9U4^pH17gH=Z}nf;xAp61Wj^ga5X&w;K7iiesD5tpb=6)jB6 zM;#dayUZR|hLS4k9w9~~_97h(@~&C!vxL~bbgRg%8F(If*<0i52Afn}j!aG>UfXAo z#!{XOFXNDUsoKFDaQ`y#236OdF$>`dxcV>n3W+O9qL&pbn;mvFj1f+Oh$zoPKM{gR z4W~rg6Z#hn_xuNP+W|w)RE~3u>LO$B{{>K1YGw$yBV0yj$GU>u%d-t$9(4ukVPsom zumlSZJ$3(r8Yqh)^6w8gb62nwL@RhJ8iE_2a}+M(7z(ZT>9QUQzurJbFuA>!vYI_f z?a!;?pOZ{hz}D^jrY>I@NvNTi?71{70meUsG!B

_jj@0Qr9$&EH_WSyvFxZH`C zuwJYPO5xS*p1@3P%I!+*LMPCC%tzjbrw8+SwXtAxy*8C*%|~=gWR@K=r&DH?<01k3 zMpbTxJxr@E+9WfllL|SKMzJhRHWX6k+=xw1J#X)%>FKbz2m)tGK6}+Y=(bkuavS+{ znxk=~m_rQ((AOPl&E-8!ghva(gd$2toSvMLwopKZL1j?wHjC=(OH;?IhIBU%=@Zl^!O9w`aTqeB^mY@pH8el zlLSP~hz(I$+K!9dN4b4(^^%aXmEbED%o+2Ny#bd!0M(xxs$bO#;Zff}UD*A2H^5Q* zK5B2syBm*;+>Iz%Qasc>}Ju|%YboPNe$}bMYva5KU7QonsTm&2z|fa1PVHVS#_0M`MO?XR~WJiMm+&$DwDQK9I;?!_ezzCDu)N z6ulZ&^KmccyaPcJh_S`e-iORk*Q%I zHxHq^eGRpwgVm*Qh7jVr{Sj>0al_*`yRuk~6&$FtP@**~>fnvc?e~oOCXP1yh=dux z%Hm4{Npxe}@0-cnkO$e3vpB4R2lrWx$(|H?(~tR|w7oD4oUhgas}uG*od?J=(TT$wk_p|*-X*PAJAh0aEignm9mHs_LG2>)skO^#!*{N zeq>(%COsXbHP+3&_Bd?VnOqtRobGNx{rz5@tFs-C4=Z0y1izdg8wU8F#AO>n$MfTO zPz-_+`@NbWY{7?>2@}CJ2>Y%7xkY1vqt|V*1$!3)wzCvJ8PDGm^LYQo`@d@QDcMF; zF$0t&WV!#+{=X7`3XaIk$9yXGESFFPIX;SdwVu4Jd@lQCD3Y}GKS*-^ZJZ?FZvb!Y zk{#Wrut|e{o*qFz@#zh{}BxRGpmx;CemN(PNaO+{ie=N4+;=$?sh514VpNnwv$h&3Uam*GgmJnx} zn~1<1f42lLlym#$}T5(#E zXt>=c+nvJ1E5~Jin?GC(3lq*g=KGxe_&}d!?%?|mxNDYb5g1UpnchQkANfFEW$p$b z3NOTv1$t$EUm+T=3Fq9h5)~TtH1s11e`lO5KZ=~cKY0RuExYUtOagEHO;!!P<#dzb zEJOYe4N~yAHVsNJEIEG&&ta0^`a!Cx$=RML#r4tTEd9!sG@)#GlEenMvT^*> zj_aEc40>AK*n3EyFoJuEV|=)0vGIzrwEXM)WhA%SeP|I@1d)5y_N&KjwO`v2Y3{ge5nTB>x`NFUmz;=i?n zf0&5rhAPY<7Jssr4yaiLFBSd)!e|Z1yx}dP?A5aNpP%fs#%|KFr7|l#_PI(U;48Fo z&>`$$=kQ$|>T0}Rb(o{XpuR47g~TjPOB&k!SZpKV;>jYveIbW=_ANtL7@FB@i8#)v zQdPWOE0vmzCcgecdzgqM3QW9SWjOA*-j+d*qCQ&@s`}i!c8z&_XiR(o64fF_tOnTV zb$u2F7QGDYP+K>Vpw)?`Pquv)Nlnm$ANh+!{qwr8GL~oTl6H_a*$8iU_RypVuoI0` ziwYLmS(PzX8e%43_x$fOcS@wjX+apzZ%#1u2;D>iUozQgSRi_KEWH_421Mm;Uvk|T zpjc_m#?HP#lL)Fo_UY$AfG3vEpIGqLO%IcU6Wn(G(P$0O#EDL)GQx)uoHLQTV3`Jq zo>OL_xrp>fb@cQLdBz`Aw(rH5)b3TtvP=9NhZ^8JtTWT{c#XL)IZN9kFc>{IC{poC zh^WXy{Jo#)1fvr6$M|96O?X$MininE@a*(Q%XuIUGoU7h)0-63!_R?)FAedpt1>|j zRibUK&F=gFEZjBQ5c$?#MknKlm${P?TAod`MOu zX#Wim)O)RH9Bt?;1#fNoZ56_KvD8&M=zItM;IeZP=ie>|Lq3XGO&SPfLXgo$KchwA zzp`hdZ2H1@PliC2b2GoNNgvt3-`kIlUAaG5BtDLkVadS_nLZ%F)@hEFZ;3t=pcc28 z4aoxM&6oK2*JWur+w4*V|Jr`b-WYVxAN&xWbetRFwT#3?LbipqyY4X~B9iaw`|Y-f zUdXMNKq~Y%fSx7b(|$DK8!;w5N4s>TEc~Eri0O3yY{T`z$gS(+`G{QB=8T|nuWS~? zaSxTYXHO71J{ty}wj)Pyt~NfRIOWmG%93mzCO2Kr{9tz?J;d1hl++XyMSA_0^HxLHLp@J0LU+{LTM^2TR375YjLS@vDc^mLke|o6O6krAuc3aA< zRMWaJ*ZK$<&^z-zZDdjIPDHX=d*yymJGf|5A~?G(AAa&6nx7H#;EyHXRAF5f`szpT z27|qj<*u#=-rsCYy4EZ_9fT%SdeWGwy#U*eWv9gHm!OteQV^hX;B4p)+KTOzZvKo{ z*LV!XB6t(9yNLYjRfL=cNwGm6dW?e?7o$C4TrEo{X5Ch81kWjrY^~4l?tHb;_+J}~ zw{CNKkstkO5LJu@QW*#^S_E8gZ@-cVl@~)|_LpMzR@=HHW{7~0~{#D2r|0%rL=xB=H0JSQ}rF4 zZgGxdR+2Rp=(qc16bSydyk4{3WoU^OzY^_t@Rm?utx{)=ef&lrjinX2b5p`{@i&0ICY{E;3`ij! zvc_2xM!kmxi_Q~V&iM9LpejQ_YpjfeN^ex3dlomgtJrY1Wr`{Jw0v*IT>6SRA6YDf z>t{D-W?)MC(o#p93o0+fHAt@|v-D<}AafMN{&c7bhdS3>QsNcNr^k@S6O7`m;>xt* zS|0J`*qV437qRSljo1p7fX|#oJlmKeVIX{Y2y@WDp;=?J+Q{jB`#%7*bS`&lj);31<^`L{mjVo~I%g+op*j{oy_wpWLbUAdv!A z%$%{fIQDQ~per1=k(kx8Uh>dxl)Un;E2hyc${-BCIs90k#WeN13>BxE#I<-ct=>v= z0Zwqfd#L<_+MM$Q)Kmz!%acvHX}sOSls5`}D7iO*CoD@X_ z4l2ib7AN<5T6%X768&6xVSf%@cTq<*exYy}!NF%crd-HrSkAECT*8!2wzrJ6}X3#n%)h8cL>sy@M( zFu;+S0@vVRevOx(c)F|rf^G?})~KdA6hF|$o|{4N1lw+kIEVlS(Kx;sQJDTgy5Wok z`!@h6qG_rq8=xDYj(d;_N1)`K6Z>oV{j!mp1qZj-C2R%o4tOW)`3YkW^xA#iU)O)? zo(FC4?yTo5*rA);m1DpDqZf&+`<=5{{!$a7a<!qSomFA43ic#fa*@rIRGZ&8+)*32m7bxXw9>}s4=DKVtOj}uk! zX*%u7E!>Zq$80ePO)JSzR}YoAhwO_rd9NHh;>_-0?k){k)Z@muQ>NLQ~MV$+U%RvIob=`<-55{qM6B&|kUMxKn3qG>VAWvmm>JGMMe7*o0CbQV z`V?PB)Pm*(@&XYvb85JJ`I9R~Oi%}M0NHwE!u(uL_zhsg-L#Y$`WB$j>DYY4mH+1U z+W2++aW|Hcf2GX2Z$;7pSvn7wGHqzGmKPc!GtZc!+P!ZM6MhcJ&GNrU}) z=x1!4Z7m_GGJByv!Z~lTZ@!U)Z=1>;k9r<-d`lA`d%V=3cq(aVQf>Nv#8FB>V6HGY z7PUULV~wfmydk9){L-tL<&lcJRQx!{UgrkuT0n%BDaB`zx)tju6_!zuM5FO;fx1{q zq}GmfaDEN>ONI$SW^0oTmO3tO8T5$-(~gU-tt$gHU&PbS_lIPqD1`nhlUB zZ*d0B^YkK?B3>6J=&;IAYnJe{AFB#Xs#nRI;%(N)9k|Xyk)ZgyqZkKi@%v@C@}{!= z`Jm3S&?xQSITKKjC6Z3=4ZeKYf&{#kS$ZMV2QZI@W#WX`cG)WU@BnV9}hM8b2S6r2RMNqj1 zdqB9T5}(h@q>7M(%Mo4~Q}M#uhJgl&xl`0b8R;rLo?VjCtAqKY#Q=d6PlP9Rj(>T? zRc@Brk`(cpDS1h9U?wLTJOG8iFd&tntA*y`D5kmxHDAyv-w(8ne&V~@4;i0KJ~&_}DYqsm_e9LS zgC6n0r5a`TQ&L(XUw?^k7FA1ZLbNho~R06uSEhK2QZ zGTVt#rO?vqUNrD_MUd4oYqzH9~DhOJ}%^RqBp} z4yKelSTcnVQ4;s`f#=9(-=YkRDVws~54Wj7E&Y|ROiN#hqG_8Ft5KDykC3nRT43BL zT`&d*Ai`0SCtyRWIBpPO*fB5Zrjm=Hg4=EK*N>h&)4?JI_~JCnV$-F`$<|egN+y7C zT)|awsO%$9$*1I^Uis}^RX+exAy3V57OB5O1lapWK9mTK=&7!_{0KMQa?4GmbTO#{&iQj6)5{5=0^R zs1R%s(bi~L(bCtaVIU6EV=xu)fK^2;z+K#@IpUA=({;^!M-1y^`a&HRCiM_wiC{sj zABl&Xox1vvs42zAa!YL>{CF`6a=o>ChPK`a{+bw?SQlk6cRTHHJMGl8uPVqvQkC4c z)-ZO+n;V%gM=D>&jB>zaVQw$dj86r*ZsD9-1-NP5^NGryhWZ*)jCFE!R!?7X#n(q{fD z4KFgwZv&GR<5qYdxwloVf+H?4rCL{VS5S*;W24lsQWsO)vHj8k!5 zi7I?3IT}R#N^BwF*3l2|=%PC&-xxoRk^mj?l`UrSlRIdP6*`m`2CYy_BDzay_{Tto z0#~T|eztIBhZ;s27yYzC$mr^^!A*2Ya2tp9=FE}p4goCC-i60}3+Bdpyz4xorl-U` zGydi%*IIPA{bbtM?JWV^g@PFxU5E~I9JamGXQY%7wYy?>H|<_ulr-I}`1q#1I!+(? z&qPL_#17oF6FMw?m$I(+dTf7$OpfUs9%3v&I(7;AlnN4+U8R~qgv;X4dWatd4bAqUYJ3yT}%N=JMG2xlJ z&wJ3OmSr1Yga;tZb4@^qx&J+QTO;Rn%F1kTcZb~V`^G~MuA@Mq7RW`&VSn*3t4;`DWz>M-s)<4`hl9@P73MupH zhd+EnG z#nT^gr$;hR?Jen+(euuW#o?gFSV@JiE3ch*1E3d|^B}`V)@uG@UdrXxt6zQJKdj`1 zaZ{l)=OZpzGi_Wpr^Gu+LL%?zBlFZq_L!h|kg-yF zyx)ZwAUIe+*JQ2^L!5lM21_!oxNIfbj$vp~A=T+KtZVQU5_x*|gjkf1&bU-bd=+J; zbx0a$(9Xh=RaXj^Wl=dbTbXxDAlgsa(#%CE=1_;Q7LiLl@`_`pL|-9Tkvxtk*4CHu zHZ-VPD_)4W7^RVs#qz)(?ztAx!fNK%OU;SsM0l`r#6Id7&s)P`MUKf5zpyR7RbmVi zD}R^RTfoB{O2e3pD zXuN;*mK$2QM7$T-x>?nH9P}NQzykXT=Zfq+3=^Y;%dVmQ4(MbX$uWGRB~o9mH{UTD(R#zfIDL(tX)Xnpt9 zucJ-o=#9Z?F62U%Io*mdFngNm)CTLZ^9|MsbPjU}SsB^os=r+__zi$8SNCjEG-JFe ziy#g+3XG?S8huoZo6J{F$DG6I4ZdOc_~x!dAi7#AO$Cc+P8;UCJ}0gJb(!nNR?AzM z2W6k|9DI~3Mac*IK3Etku@r>vM1=AOoS`T+OsceRTGa5&*dp$WhkDO#Tn#>xZ<(tB z++A3?kByLJV_w(&^t4u!QxE+YD#@C0ndusw)t@Ac<{>MqO+N+~qoBcmS;10`yP8yg z+7P5IB)57O`-?RNO7hiq(6*>@4hhbr9r-(TMlj+EF_3M9vIIGt z4(tfH=NSrtG(uBZ6=It`;d5i`@|0x*)}K)H9=S{B#-7&{?L*IdBGIAcxGEd=`p}k> z7aWN9jLo4UochG3vZ8pk>ZeZH{Uj9hgS);fY97#@K(YSly+zchBU_oW!GP(586OeM z2YSD1Lp<5t4a2P;wFz0~C8PA|lq7w}qr7jLpIONfFpKjGwk3EpO-Yb>Wb}^6YU}z! z6nH6q&cR!Fm7ZAOimER4UA(e04%h^l(lk742f(8sJp8F97r(Ds8y5Iq%9PJu3ow(O z|5VDKeC&2+3#6w?JfHP0G^YR5U%kSDK8L{v>$AUV^UsA?0T3{g9YV_ZQ=eKuujQcuE%zkzGr>f7TDGmiF)#6vwEO~oEKmGa zrQFpP+c5T$i9I(RhFQf;N*FAGyVB3oRXv}PK+=rcYg*~&yXvOLgy|wEbL}ERB`bhH@lnC79Ya{T3vw24LBx;ke!@)^u)GK(6McIY}+|Jm3`aWI|-Ew+>}g1M3}>9#bxYCW}J zu$)%(+?Iir&3vL^o__$L$s`qNHYIrFAr;2|`9p{;y4K3)GLf5Vz2nukw&i~;iX6`| z-rfZxjk3Qqdma`FC9Zv_W`8NH9Q@H)&{6&7Ro%SMw7i(@v*`C;4as(M{)oNGg)tE)00KIq)f; zk7_W7?2o(`15mR~lDzw~KqRVFsWmS}0>?}4n*Xy&L8P1B;*}^{Vf5P5qO zlEt=rJ3Tmna{j8hUj+wl5XwLmD4#uiF1vh)+@1v$CjIMa`2-|wq-|)pnf<>tl@anz5eS{KBbS99_bZu!WF|l({t;3=q z9PPVFP}zW@`B;{gLf8r|?XrX&h5T&XcLWqUcf)y$NwDbJcj*du(!YlIgAPikxUo~r z4rZy2WuDa4zL5$z)S#F9t->+dRj%h^_NMt9Q%$C?M?nLrH^-b`XN|5Z^pjjr;+52y zpE0TqEVC(@Azh?CgQ#t9OM9-0KYdG$alZf|JG&96Uo$D&6*ChR=S=c1F|+%tLZgym zH5G%T=vO%e_Z`u&`~LZC z4@R|n;;^{_-=IrP^T!J4i^8_rw|Ti|{T!rM`_QrSs&D(Aml~w=F!#R9UBO(v0tyEk zfWpCE0_Sd^=sL`-1%e00m7Zz8>s3(boeVklJB~^0>M0 z9Y-o7CfH2gG4~qAQFl3_24{ozWLxCrJ*w(gdsDu!MHWLUCyZ~0o2QM3(f=+kn~{f4 zrV;9&e}kRnS+|ZLpLnPYS28=s#o=Y&X=7WY_oQXuhiC7@BX@Sf<;xjQDrzeeV@ekt z?HV7pMHVnJ3bCIoiU+T#USP^r49{_sD?XJMR3_DGn>`Pa(;mv|rh~~HjYWWw;r3rY z*}-kpCG0v4R)R;&%AYQ!O^sxjX{hN*p>pn7)gwn;A4fHlyST*t+Mc# zpdV6SF{uK_&VLsyzJ7Tj%Fe! zFZfiJylcw8)MVc4bDecKEV*u1oc5@VvM@5jFp~R|#B*lT{kwwOcEx#*`if6Eq=}v$ ze_L2kyg%omzWZ6n<c7-<-YbZ*gD`-ocNEYiAYcF0_1`t;eJYFIl`+lg z1*gdh&HV+I{^`(^eyK^k*QYw`P+W4&tvD4_r^}2}e10uV6a@RfRMd9ExJyg6%kj(z z>nJpdCPxRBeTP2B8q)+)eIXAg|0JOdZgDAH1S^fYM=nkL;YHAvKpxl|Tj6iqekw6hHox(6(TxjBEbt3~Fw{a$eR1?dKI*o_gB~yQ_pA-0^ zL)r~j$kHDtF$jckaemPhT7;kiAJ?1W-N5nG6j~9QEkf5iR)`Y=QMumK_}BQClc5i} zBDo^rX%#HUE#@I%}B_3nvC~G`r8#rC=oI%u4{4XUM0o~zgqDF5CFWChdvVT zmFf37L%AP}rz?a<6?OSq9jNqE1S2e-yalV!bK|%CxF5Qd+jcmqB4rF9$d?*6%A-K6 z>bNA2Xp1~tAE*>SX!(WFqhS^zjafK{#|P4+e6ILkIb=7dx%VHppzLf61p!dn_Sqbec+w7arOCO2RWolnPz!6YNz&|nf7gd< zR=2ZQD#p>y4OAWt(6%jn_1vM5n=H)8Yl{Q$LSi@K55ZKIeldz6bi4kb_5S7?rnRn6 ztwF8w#R)x1PP+ex;gsv93|7&-vO_<6@YB@kgKg2#YL4TYl z)hyipa#!DN1@GrI!H+)xn5G&Pukoz@KY*kqzLEaY7maF<;6$&J_xrim^TNj!$;+D9 zZGEC29fIFoeO@B=lB!xfraDvTqyPJ0OwW5w&$hnr50EGGyW|?QhAzL^%rFF_+sf;C z?c<8!<$sO3eY~8%-y`rjmzC-n7Ec}%A0?fsfPHxxo({oUw z2s*=zK=~wIm0*7V&;mQmA20`-)D-l)EQ;%;(?d*_!*}Fa2KT#xyp+Yuaoy{}?4qf- z=R$3s3v0`6@~TMGJi%8nz0A`K3xPi#Q>KM^=`fH7ypQ8XasmDqOBde^{*6~48FUBAU z{8p$4HMa(+u^^zTTh}Met7wXEweZ?2J=c9Mvq3uqV&x(UrHOWL33loOG^j|FWF*cG zyX7jy!VSzqKtdF2!^+KGr-chaTu_#6D9T;Be_z3zlm$EM@A9G5OtWe!NC8U z47siUk0{_HhC+nF@|JYoyPsVyZCCba`qaGZzl0T9unn2gYpg7{{Ag~zudc1O|88r0 zt}VITBG9zpKYTm{f4}MW?}DdyK(winK(p1k{_he9?u-og_>xoO=PRVviozKrI0U7P z8YFYnyeF6Ghp*M(A7?FIMax}o8kR0#mu3qz9ZY8x5dI|LAmu?&XF9P|}L4x`<-37{Z^<|tD8pH1u#4g#+@^el>GX|3&rT*d;r+@9Ma!S4_PzE0S z=^slfQ|xFai=h1~J9E{FYl_`n&H@H@1OWtnR+VD`X_KKnPV?bm>BP z&z=yjrrM>8D#O>Y)Q&f@a@+kCP%UL_DUr*Qqo-ix?(X&V2`8u+Tg$f@ zGGQLcXkU~K@7I&d2Q44Q=->2JuFDWG)_H#bi#kgx11$OsMD**f&;-kYf*kjDkNr*o zU?UmalZjVN>vZ`234&8BXmr^>oj`o}q7o^Bw3iN?eG(Wn-05|RvRY(0{mV@n`XD^ zmJoRHemARqUT3V{=0!I=DG#-3qF`xwvCeAEv zKd?<%8~wUxN@20eg#S2vYT}L6H%PnMSR)l>vhVY%)n3TXj7M$s*?yc$_3wcxILL^B z*uD86nBdqR?*Vx0sN_teE{E-z3h_QY0?`|Zhvb*;EnHZHMJ>ZCRaV3Eg>=s$f8#^_ z&+j!kA1jKWq^+Ljh3)mz%ipe#4;|lL#Y{BO$s*aorC0?K7vx%g+SgqgiyRIX(0ToQ zS>9f5`DkvvukNX~f46lw*Y(vQHNm^4#&)kRTyBZgS^sQnb^ouQ;r{2kC^|vd(H;{X z;>WGk9 zRY2MR>p%2_UJZGP;1lg>cq;ZXjSoF3Bg`*{Q18VG*kL0rt^^cQ&;$eCYD)D@!Y zx1#dzG6y!4E+0eV0Bg=I5AxN41jLE*9|E4GowPf2ZH8D1f1>m6^0l3fW*^%M5yX!E zU*i7uchesI1rXn)6?uFnb;y*#B8@_L*>qsB93F)#kp@gVOPqTS=oub7n3r+!zxo zxQA-^%}hOv3=5P!W6_%K*RTSnmGTKa8B(_036^sDcs!h!($Y`5xS)3+Fhl~ltf2OS z#)Wq^x8gTn;@(@KiqlktAyT)105?b+Exu)I>e|euBxk+{OYi%`36k8BOn?N6?x5{} z%$!*#%D^6w#Xul}a2PAf0jyhY>6;a*OM?M~q(r*1t0m9xH9Lp|k}(n(LqTCGRfiEz zOL&VQaI(A|G!)AZI6Cz&gb9oMESWWqt`=WGID)08gh36M#}SX$XQY$GK8}Kw4?|PJ z5e;*KWw>%httvq5It{foD-`24`tjtBu=@g9_^94#LFH(GAv9*=tJIYq<7D zYAv7zKXF1Pf%*{xj+NBstM|YTLL&bSroqB8>Vs8ntFv7H0Qh1)D3KsGLUvXCZR@c+ zml>K;PkIbpVS4}+(Aaul?Tu0(v+vFR@fozW&?oeSo|a{ZNSCZvai~e#u0CF)?p&d? z`&#IIBhP8IR%j8O@gFNptj&aXExE3gnF%b?EQn*2(5Ijd%6N+vA59;%t&=3-TAq0UTwZN2E!ebldhtyz9wvGqA`?r~##?s5CM z-ND3*i}-dOv#p=|KbhMgF7ZJ>oJeKVYj@N@7BiCvos-6b)aR1V zAw?w^NXnmcuRmd55;o`S3mR~ua7Et)Sxvg85*a;L5mtHJMcbCx?im!ixlXn1M7wd<6(?BbW2^i~2)QV8G^w`KG1 zA2O%B*XLjNFSe1st&9-azORTZ#7h3?z72u;#uNxD7N>(jTi~iGwDbeHbdF=|1qi

b>R_CuZdO}N24a4%b+bDXUr6?V!fvVKLYwsl`vo3* zhqT5qBiWOM){edY*!}<76ulfcdQx%vy%YZcur`R)|3{;H5vc3{efg9}wr%J`CSL3B z-~X>oZpCmYMD{^tGv+7#h0)YCbC-b=ST+Rf3$vQUBE8>o2{L0b$Q*%uA8 zDwkO3^UhbqkqRT=bJ|^38Q8vlkcFT+*0!dFh{g!=Wzipa*VfgEnNY+PjI&nQJIUKlBlOVJCi3_$Zk z!OlqQFool&`zAY75=OAYtTNeUU6aecht98KWgR1BYbpZkWL+2s1B5G?*eIN~^9M@F!ZD~ZM*nj|1gXXsTVpi)L4Qdy6&_TX@w%$Db zmznJEk)Jc9==HC4V|6yX=A|{ZH{qcKD*ib{)`aitGONjSD`7L3do_2!@)MoYIY`en z;^#pLs!H@fhKfMwfijq~_5bX@knuDX`}zDIEhp&Ee$bwv1IuK3=XO)P6GHp}FnTZx z4mV2jB^p!g*%Qxp^IjhV`Tqb0!m0(PG`*>FuU;6oFjo7!&v$A-Ov&!>{oNz6>$Z8D zBdY?oFqMR#$v`2-r_{1deKwn0_C}d8;k1RF{r{g2&EbsH<;V5w5!KTwWPlk74Qcqz ze%cZfZ}>=aN()MS0sSd5LvVq=1G+V*;uUM}gW|Jx{iWdU7&}egW?{Z_&88H)gdvo4 zj%N2)n_r7S(%V{#nkiAUs__KWPM*p}O^b4I>^%oD5?)DOP|9#wzLX2kO5?uQ-`{1x z70`H5G4#}8T)3I1+O|4k;CyKsi@QNKtd)c? ztUA|B>1ULJF~m;C1*Jj?7e zqyZ_ZN`p`W6M`~1{~D7(YIc7c@8tMuLo2>2fCU-`6>_qE0o5gdd3_)Ga2$z%fX}2>=MD^)^!JM#HdXWQ ze5W61;Z|h}j+W3NHU*6RCp4Xq)ZxR{iCOuPRx42pNbYq|R_OcaFt{c1cr4IZS++s^ z7MX4qV=|Blqt_+rj+={tQ43;?I}~M_G{o|-0cG^-t=J^NZn|hHXoSxQk%0GoVsPbpkB? zWTO(1i-3Lbj(lN3`|JZcSJHUmK{@OT0yIO!+*8bw{E$7n(!S!xA9Jsz;gVGi!v}`f z|3WB&{=ZfNU2q7j=zEIP2>YSqHp&`I_Yfz?N>?BNV@3B%JEpr=TjOq!qgk*I>_^gt z&N-JAg^hsMxenIC{r&~N&w@Vmg+Sk})@QBjQ@p=qly$4ME;q`FyrhMf15haDS56Px z6t3AG_nl6U4k6v=`?1X*=lRwHh^G1qXOGiBtiKp;GY3^ztT~_9yv{g1Ahtg|UNLq@ ze!gwzL$J1NaQ=2VH!f4Huu)5iNC~#Xs#9yC)#3i6sx(*g-6C#LUZ%KMQ6?u^QKE&; zEcVvM8h1PQ`*+#LBn)nX@T<{s~ z`yyF4(e@RhgH-WO9Ivbpw_7#*>C5#ap`dk5p!YV)lKQF>3RFfg9Y#s=i z@t93TX*RW)0J|jyPFu%vNBn4)bITZAn^iB3`TTrs@}|f+8sYfi`gd3!mBmwQPr*UK zerA5TOHDsump44=T1cpf{3))tP?8*&@%uvCSQR2G4=%Od0iJx|8@C(9+;pK1iu>hY z!MFizwa71Vi>@w>cKSp_9r$l@fw=q{t#Dl*&euAlfmm9~lZm37p%Kd-V9+$((4mKO zxmZ=jbdcN&_2h&;2y)##fl?4Sek%dl%Z^P$lBZHpBUn}M1oiBiX+ePy5zE0s(C}Ht=n*Q^z&uOqAXwN^n=96FrqDVKsmN9fw-S=psrPtBJ{|V?9Y`C z!N9(b&;Ia%5CA4Yz|--ur0oFnBUSJ6Bq*!Jg99zGs;dXvf@$FCqHSGv$k5t{Y`M~s zAaFuibJWt&c8$_#acDW6~yMd}e%$rvswZKC&h{F(qCjwT!i9nWi- zuyjvq6h8(}l`4r^eD!w$prsX==-9xkB<5?6r%1sB6SA%_(5GUER=5~r+^_FtMFoi# zB-Dqokm=0c@G!?n)5$R{3o-HpCyOsip7Ma+*d+{A;i~+N^PO-ceL*!&6`!Z~v@-^& zedNdNJ=)31PbFsUB{|;Ojs8X#<>-|}X$oKY%kiqS{3bV=esZT(*~#=JGMY8B zx@10U7zORUgaLRayY=5*58dAkK=6g1f`hPP z&xZQVhbqy7bq={FYg>~Kz;x#eO8Na@oncZ#R+_zwSY67$3TFTmWl(m7j0n)4u2duR z0g)Y9pFQC9-Y3Sp`e9`}IG5a=`qlHq+nfA@g+x3&M;7_d^w%P^1{-NKUS(ZF@qgq( z=9sk9pO_?L$w{Yu9yS!bX_xLkXI=7j!uL{_?ZZ?8@_({b7qq(lUdsdn8Gzx=)DNzP zWbjE;CZJZYza$0q2H!C;EI>lOO9c5Cy+oKikC>QSR+v$=H%ta26!9~1iP(nO(r!c~ zJk$d}*n5A&5etVmunQ;0rp^l|sKJ}tSQOU6Lu8C%!p|`x$8?u6B60#QDF3=2#2f%% z^;@d)7;84mWA^#+vOKS9oWJz}dsyvqd;fOG{NDJm+|k(gaj8K7&P`{!L~1mfz3?DM zgowfN-2;st{df=D7$Uxw74=6BA|v9CX(Ee{t}5*yOBrmTZV@H(Jb^}tbdWRIbZQ?! zx2!XfS|M{5OQeylhLp!rM+BhQnht#*<3T$XVvanEO)XXS1{f;upGujiG9+AA3gX6w z#I@IuDm1syF2u{&lwL}eF{m>(8l7gqoXR_3W`a3aq5~%{FgXrA8X${}KjTKkKCcSH zGGBueM;`-ot%axIU=m}`_kxLR3;e-_(^*Eoep2TYhtuD1Yr`kudwOF zVj&Q}v>36#tNH^bQzfa(x?!@zOKI=~Z{8@uV%fgf1y=WldBB6DvbUuS5fYNsT&#-R z8SG?=k-9kR=Zue5RS~`%;3JrW_j&A==Ly$D^`y>&yn$*x5 zNo1X|p@e!ZhXTS>37VP-Sf+BUtjHWJ1iqY930JYPa*=-mPv1q-r@4sCxb&~2v_DaK zjW{4J`J4tQs%uglShQAj$xJO~q}P^vtz%GboMf#>C^;gJk?%@cw^5RlJO`I#fu2t5| zBLTv6+MM`#4In}vY&3AH#NYCJ5_U(EDNTLJ0i;#8x6pFJ)jJjlQwEM>ZKKm%JPS}N zVo@>Vp40achgm9ehmqea8OA(=84BPJ6gxL^BCMO!7LaAW{ykll?|?mnH3fRaW_&l& zJJVoO;IKunI4qbCnPiq_5-oU`xN;S`WMwkP;Jg;)w~nTSyV*Dk#3(WxVYEm*95gm@ zp&f0D{SW~}!;1d8iDefX^Z-zLITobxlX+!oPaUT38c}EX%6;&OfPy-FSebTAxajdN z0h{9kOw=I_tyHTVk5J~O!h(=`nO|v84@S>dY3mp1+hc_C@Rc-+svA3E(w+T~YgbCJ zNzN30JGQd*G@`~r7`vP))WlJ$Qjl96lv!N4hVr+EFn1KTU};^f^}o)k(kDwSSRu9w z#|2lBcPtwXouDm!RJJ`l;71ov5j3YpJgX5~4}B6(FaS z&3&V9G=fC+bR1i0nw8Upyw)C?YtBPWf*a-);nB~&Fup7Diadm(89I1G|!T+pYmyZ=ei^Nsl8U)0yLqekC1PBeBJtY$mo0Hk;7z?*&D?LwRx}oAJHr zT)V7ld|DqVq|1rRsrc;Z>+SYsNMUb@>d42hyuN2)xX;@DP&)}e^8pxU2IrYbO|$%T zP~ocuP+g`vegk?$w1Xd%A})0+ALA%wWg0fM!P!AO_jsmryi)TME-`4E)=_X|h~nIh z@z1#gF(!qhCTlhVte&taqQM1)(1JHq*>VY^Gr0U!H>K8qK^7+Wjw&9xodTefFVl5X z;|jY2SB>xGuPXEv>K*hUiS1znavg3w4c@T3N-DNEGfA;r?)g%T$2vIsM!7>Nl@?Nq zyc}^Ly%hOaoA*jK`m<7{0IQS90wgvyY@+g{7}_^vFvOe?8DX5d9Ua6{6IJ^Y^{>DY zv`MA#iHoK{E$MPr1_*5}@f2gW;-*WIM@b0=Sw$S-N^xmZT8d{1^aK?8!Y;0PUMR6Q zXfJ#hDlVf#N3`WkTfGKbrw&f^8y{uJO^KT9jm+p=qF8hoVNz@y&}1O< zn70Lm;EUrbE>gfSCPfz~BUxIRNRO=)bYPH`Rv&d|wKaAIjuVgyw)k@~3fo zJN8fX7=6s*zZ{WIs<>Ma?P-rBx}?YJU`G)B@DcnzilCku&0ueAQ4V zembCX4J09=4X$j4C*je44u;Y&b`(lvUa!Jlj7>>|e-g(=fsc@XO9oHW11e{<#Q`mizowTh2Q4EVt7~Fdc+SYNui}wBgOel)ZDs7KuiG6X{9A z8!aUcAsxY?v~FUZpaV%UIbkdc@;fPx@^e`4J~87ibX7qAcws!$T4RvS3A`HEy7V6lY*V}2?|$;7_lr| zhlUs}=C>({{Kaj!V`r0FQM%qd0)>p^?er=vL*SJ&$`G;N;R5Qv zsQAC<st~n#V19{O@LZ={GA)i$IAC=t-Ka@EF=n*(LD_qO@Pz${TTlE# z;_dtKy}G0iikPvJdf}ZA!s;+&hL_@I1Q*k`?n&9-vzMwKka`HZSgvzBjBDE>`45?& zmFC}wq1^G@@w{I+mbS+ev7J20e;Pp5lggdA6GU|G7b>)AWht|O+iQVPVRkLVx)UVK zNf34Y;~Ga$BXd-RgoLzEZed^Y_$vm;31U}{v0{;>Uu@v4f73gS7*z>= zvB67jx~~qcwy(6;wOLbz#%&G6xF;1=b;?1pbfQXu*L9cN*9U!RFyD0JS z?ub6Wu61Q*(9B>REg8$38m6XU>E|0~;=A`HxC=zp)OzlhQoDgi8bEdzCM#8YI!yN> zs4j|4!+$!8r=j+BJS8(C^>(7#furrd`ub*FCgKdsAoTjO@Q*cU7akH`kvZdv|S zBm|Qc<;*W7U3O>_%P&iGIBqvJNDbNW6`>Tz$M&t#l&q(g2DNv8l4zb`iZs=a7M?Ei zBOaKRZ|tQCPUF3Oh?EgJ9#0hm2KIgzl9am$hbPGdN= z0>ZLAK@VzHMoK2fXt|B1nx>c<4U*1=ty7-gm7D_sa#;;Xc4QKY%4uc&&LYktg1PXH zm#0Lf8axzkrbzR5mO+RsikmEx#in=T7CDjuv}`cDgf@!(a7E@e$~(A1U*Ev3mS^R9 zSP9D63$Kzv(Z|rV>|HmMLRF~0iNI86Dd>y0{bRzAK2a@wwOG2gB1Ugd6ACBAPHcN{ z`8W6Z7IWDQ5JEt-CrVnxEY9CR${cXH!L*uj~Q@0*ZCOIB-?yEsQg)4vU_yv1sTEXMW8>GDDN;G$tscW5|ijmEU&MyFat+$ zTYYsIs_=7tO5Fq@J!H{T1hVRM0P`F5-wQOq3MPB>HtZy1f<%X}6odGYAdO_o$Zf1` zQ2TSF%D}KV;dTc-po%E%N|owBREFz{C;A-8*Bp7V=cx~KG-}TBNrMUj>C&e}yErv? zfb0%&LZ)RQs8iSU`$2d;c_{P`0C|X= z@zP()067d+P-QH>mm(rYkQ-s4++)C3@(5qHsYbIGzWMyahJf6M-=yv*QBzutQl?;m z$X!>)l_bTd>f9AA?D~9{7o>{Q{<|HYtS2?Ypr&%iVL$VI;o-Y45hDsbQ%zMSGrz`y z%Sk~l?p=Rv9;N(q%=sV&a>d2cEiMpej>L-!kB6yNE75pfO2z~VG<)^0b-&fvvE;Mj z+zs@EMGdWw2x(TS+Klpm0a+Xhuvp?~+k=k}^lTB0J;UWSl z999Hc@Xj)lDe`qU6?ReHIJ!uOnLGD9S7uPRdJN~HdU6%&1;o)sgssm+{OgOP(W^Qf z;lzcB(uVIM48Z8IqTcdAx60iW<$%rr%gPNkMM+cZ4(Q=8dr{(%mrTenjX|Qt!xE6u zw2{lz8A9m`=@S(!a8I$u0J08(N!NiW%G1z}wb(OvL zT<9Q6;{l#mi2tFoDW-Bg+}KT&1le?M&o5NQq%%X?t2jSuePuYjk;I&MCuNXqUkj8L_%c!L=`dQSRmp&;V7 zjKx*Jt``w4wv}>PHH1)3RSY$Yj z8Z)kA@FEQAM9CoBXVbsuNg5F4@a$0CSDv{)w1LT&?Tr@1PKYduhZ6H9bHY`s>ydB{ z4`IS@CeM!l;qAVh=u{i=b&*y2g@)oB*J*@qY1jY9SpTV0lF&xV+>qM;5;DRwCuW9(u)b*tm0E^3R?I>YFZ3l?KGX}Bl(wL4Xt6Ygg0k?1DCv`Hzpz@Z9 zR%Zz$GYzgK)#mz^*Jpu(oMx@}8apf&tK*S3+ywzK50_Ekn^5?f9BppVf_PN>mL8eX z#bEy9_zZ_`SI1ubs=EVs*Q@1ZfjV0OaNPIZz6cU%U(*dkIP!#F2wFpW`OUsjyD0x6 z(um3b!S*ft8M-Ge*ir;}hYUjMg$hiCwb=9ds$mdMLTq6Ni7d(xwGKn&c$gbEYl@ml zMr#m%{0b4BM=9L)iACu9L%)diM<#s(PU@i_2j+*i%NnDU4irkB$94UFpBkkZu|4o+)$zqJlQFM>oNu^FG+bu>eGSYsQ3(@?Cm$ zxZ-6K%TTR_EBbFUxR?&zWK}EA@TJhajvY*O@>T3RbVVer?B*-;;Q;MsibNp;QukD7Fyzbk)$`XniOhj}<0B4&Ez|(hO}kH^3PAE72B- z_|ZS|Q0w%#4=b(qwG4dnY7m46OmMP#3ig=hquo&RgCj$^40piho60hKS<}>2>_Sjq zQxaPQGIZa*@CCBa9tDd}0_42t)zM7A@V;;V-;sUMyIp1nn6O}MsEBFE$>SH7AB=Vk zE(Y-(Ex0gk6UQ%9{5^x()zo0WIEP0B_v;?H6vHSx5MH$$M|bAV>0qU)8lO0N5CBDb z_5OGJ3;8XuMQ~^=cw3DAT}Q>e!hWcgE>@XR`<9d0Rvn47E>#LPGys2AuKR1R14kNb zWL9c6aBxXmJb`=(bFLl8YPm&BX5?=ryJ(2g`T*Y262C@R@Lk?gFwISpIvjk$2pCh5 z2Oe4IDVEjv@Bv3wWtsV;`fLX0?hd@i2(n3fcrmW;T+Wq(Z9LEnqY|~hW2s({$?j&O z+m0aZehV=Whvl2P^$Lr&C)?%*vg#4N=rcob*lSIm8`M=#kB92s@ec?xXC_gAJ@t{l zTm^?a5$x0Qtrmzx!EnH5Y-1t^5k(|!Pry4L3dwk(_HMM;E;95D4kw;d4~0l%tSt@= zkA{;;plRUv6*xPZY}pG(NXbqBZ*$aY&3;78(z%KdzA^!5tr7D-t7OKLjCj&@m3f~c z#(3b@vL$F>4SWvA3x-{j(*A`RJOOJFoVOr4Nn&XcoPLdl6}<#`Oy;Z|c+>(9nFKJP${R9!sXa-Ra zO*yzqv40t;u-^YLEt?BI5(rD=d%TApi=f@@V+m2U4ay$0_23mu$|^={V9TdVSE zDX?LT``|4sCgUybS_R%r%X|7pk#o~c6Yb$$rUJX>S%XpqJc3LI`CU7z5o8*K^Y|xN5Pnd=`l*(Wl zK~}0_;D<6w;KOi6L(UU)XujO!AC@~qcvT>-hX%5O`AtgORmWh$+etpTjFmeU&!;mI z1P?Zz^2i;k zCvLwBQ2{Q24X!nT*LR}! z4K_%iKr-Tbh=Cc^MS_Qx3|1Li}GTKKGa zUh-14U(k_FbAs66(&56uM=VJTvHgkTbm-s(3`1x9y=izbSy1c;NZ!e!u(-vn5?eYS zvdK>TS_%kCWXXm;2|~B^V6*zig$|%YHkZOCM4}quG=IAQlWmN>zDmss9POj| z;sE;`FW+_wB53$z|ELr9Ai7lcR)^R>L#arc_<8#@TDpREjEz6^tblLL@0fGQ{5DT} zXFZGAggG-AIPWFJkLaq8`P~i5j-^*u7z~RX#Gc*nT*gPlJRl(C{e~Ut7VQgdR>jI- z{Pv*^mav)gt5H(X2sH+LjatFSRgPvU3CK$)M6=`W#)@-iYe5SJ&8f`!&sO1xy7 zD*2I^7Oabb?DC~@LqpFfDgp3fEJt4&x=tp*u>w_`b(IuJ8&d=73#{oEzrjG{jm7i$ zs4?`iNAk}Ry2X5f0ElBZIzLoukE=9+%T9#Sd9xqcor&FuO7`;o48yNe{Oo}S1o7uE zvGU&+1zhiuzB;7h=mcVeTha)q2U^w;uOVAfxTf9|M#9OzGB!?H2|1y;*;P2S^aEj% za@+(ILedO_IbCYPV3I0F^QT~NV7KE56##=MsQSfeGn=Y80^fk%`G+FMw6`!i^W^K$ zzd@xV1b}MEVk4^4L;fdmOjOCemdhg^`(_*70SL_?5!&wWoYC9SIGtB~lKczofX|x8 zB11Zp+A7d@TmnaFFBZh1OVrqyPN`BTEZt32>LutV#1o;yDuu(s_NN_{Wi|!gBiVKL0zIz2+0AFjls7N98@mOBDhxxG;~gFeMuo}sqB@()H7 z8F#bZ-@7rTRaV8<5(Z?eN3WP@n|*OX<=mNEkRKXoE`j-S464x0JK}QEpOro5SzB;^ z0hv#lLDwhPHBL!u{>B3#lC}uoUv4^`l?SH@5mhdvPExQaXA8yY5`kZOH7fQzAN1>! z@%gE;$A#r1#%*3{tTWsZtbm&J(_H%_dyT$CfjG4d~ z-0)hn%{vG3*Z7%)&<>Xd6^-q{n&Ve}JpT3&{)5{>*|a)0K*;I*r83_Nghv)K z#(gG!w}UTyn(XG`)WNtfSSS+mlJoUez0f&tHBZh8GQ8Jai-4~1hV~Bl8JqWTMEixX z0?$R0b*+nSs()R(4t#f_x4T6(-|7wz9DofTJXXE4os!Pu(a|LG)PJlwwR79X0rnI` zd9s_fdJ|jI__o>}^rCoLBMHQ>-Cy?c20A`28sf9Nbt*q;-YlD~+cO@M#;sd=`jv@nbAV3Wv zuKis9o4TyYAAr#3`MdmUhN#dy$eaDS=3Vle0kra+J6ezn4c4csQI`H{1Lar2qX3nS zmro78Y@6^+npd})^Ii!V*{`Rdfc5a+j|hvZV$U`fUhgCrfjZAwKZz-@3;rjXa)-(p zX8|h}$VqU2{_~&eKh@y+{aE=~7Vo18G+nzDz$L<`^w_6nW9^=ya2=Bd6TBA(-Z{ua z;}KXa6ML2Llam*9<&4Qd;D@~77)VC^aA8yy`PygM-+tRVPg|J1_#aCM2+i~$buIl= z^RkBqBkBGYIwmuMWfpP?^Pu<1-|cz_U|1UQpZ9Tht-t@SIzqt(`PU~kvng_7?NyXO zULk=EIMNHH4sJ}ly*PQ-qBawNAj(Q5nLCVUhI{*(xmS=oKTfe zEn1c_5augNjXHZqlXNG#{cDnN?g3S*KtaG!OMROO(9}H;nm5ySi@Kj{eOqb`de6dA zD1P2q z2_0J!f@JVmFnQD}bkX&5PNfm70Q4|Q8Z#ulRQ#lBTRa%-)g@#$$FObVZo3JbwJ~z5 zrpN3`g-eL7&zti0=@;D!6UNIA&~h0g zMd(;m0mz3;^V{)rAaoH^j0q!ncvXs~kWS1F1zh(fIl;B5fLm#?wDMF0$4gFyxBw90 z&x}`#`ofqVf&F)9>dF*tWLmLMDFTTi$<(+K6GjQW)WI_L`tIAaZ)$-9Nipk~e*h;c z@S_P;f!yB3MWOO>N*LVJHp;*0l7{@+w31_7XNO*~3KejO>OZB#7r-SpM;=MWEo019 z5vm9jD<-alC(EHtE_ZM?J86q;#IDBcG;li;ZHTv;Vax#;UpHuHid|-p9I}~Q!>dP) zzVE3IUH1(g-q(E9gpkBP6sAdnNd`<5@{zwqj2?B-$$)5{EZ_g6`L{OExsw?pz^Zx zTdi$6U)@K14s?29<4?44g&yc*0|$ab`ykN0r@|PkI}`~!a~{^5 zV&!-U-6qtOZPdr53=$8-8ZPN0h`Tbl{q{ygfv-z2+zoo`{gpB6H;R*LqR2I0-@p*2 z@4QES$2x5mz|l6_rD9@=X2Fq=;1StGK8X#(;arnhi`{tlS^XEcAla2c@~G%k8;1EO zkWbV>2~NS$aQ*+G?Jc9?>b5oE!mV(Z;O=fAxVw81G`LHE;O_3ho#5^Shu}_-;4Z-} zK;AAw&bhZo-~Rf~ce3UPwdQ)JY^brT_LBLqVSkw3Njy^SufYmM;B;lWXGf+q9k|M& zA|Or>gS`v#DbjU7BZ;^G-&X6CMNwEZ@Dc3lG4D%yg5Rn43nnUI)%u_nI;iwJHc1I9 zQ72kkoAdGx(>!v~A{3%Pg&}C?e&amN0z1(%ZaW{G}G6B4Ss zRgL11tBcJ2nwN7xm?l8s%{Gar&z&Vr?Pj-x?3w~mNDG)RbK7x=-KQlt48V}87m7Bg z3qI>2)brqbHSG8Hj|N8t+@CQg)JtED^L_<2(-+cXUavHP$JNRi@ik4sxu z*M#uwO@zWqr>>{n--_U$YTa^>MT;-b!@ry2ZLxoBP=Em7Kn+2C=v_L}oV_CvyF$YI zaN=Zb-HO^A{AP)^%TajD4HD{VpBgoQlDEhHJXhE#j$U5`ny*;UD3`s>sO};AD+$#u znxWmMUNz${k|I7fvap+HDFHH?^UXxWuE@$Sh7ErFGFcN4OwO@Ha>YuDXzAxla+u3` zMppQgt;%_^ff?>aB-`7{>Kq$P;}pqHY!XHrF$BNm-$U|RKd=cP@oY&@mBX)GYtN#< z7v?73;2@NjAO%9*Ixb{Q%W-6$Hu=^?}OT zeZmu;ku8WDQkfUQOOa{|(XvZ~#L-6LnsN)J*i9w7sN1n9-r4Ho!%EEbvPTm4>&X3cArSm@x?Cm34~`|LCepi zmn3CJLFyrE_M;A(Tj)&>&X5q>a1uZ?Sc7&@Nt5doHq)`sVnhvq>SOEt2Y`b;CL~6X zYzLFzC-cigNf3jyloY&iZn-J2?|X$-{znNzn)`!(1%^g`y3xuE4zwzx>W3Ogm2{16 znE7EA)krLjIgtM1HL123*E;IwQzlUZvh=sQu%#+jee&?Ab2b7FKG3s@dj1l-B%v>v zFRR>d52N;dtaIGi-}XCSdpP&8cds~6Q~x0$oh5*y1Q_c1$kAiug(0T|u~{P&^X3@~ zn#*ugkcecaOLeptQu}LV5U65r*DeDgGU_0k7l`=}AUTmF1U}QhyJwOv zx2{Ke5+A96z$H|a+6W50A?_#8|PeSbB)ho>1UX91&(D|KmEdT#-~QM0CGCyBe>VEHIoS2%}A*b zAWNdME*HBSH?_87j*UbmZ2Y(&9jMsh9KNKGj;RnZCXv4Dq&Tp~=;1*ZC=0V|!=9PK zsSa0~r?rkvE@!9lx+K3UYYaZYG34{()equ^(6p4q$vhzfW#HBpxNwHySpB%={j&Lx zjbYAG$qRSMZR5jKCVc!pYt_3~O+FA0`R0`|PPa`SxRl4ut-y~+xse1~f3l0fX$U8F>0DMB| ztQ_@(nxs%ZmyVa^{_ZLp4Qe3|$#5l7cupT#J$$!^bnwDZ|ooNgG&Cb&LPU#3g3E;;26-HxGFc2^RrUell(GY{!kjy<^0@VPBBpw z0%)Ma$$8WHzHHQDG6WkCOu`pn#7n*HOK){6L>XJ?bAJSdO)h{Tc;a6&lqp8qlPybs+gBEx73si1{(ag*0$089uz}=^2KQ_y~RGCoG<>+vGXKa)qZ3#Bs#>$ zgQ&(;B4RW^24MF&ya5zy&oHfNGyb`uI}k2N`rg5vaystZ`0ANfxqpK_AwOp_3A}(i zpd%#h*jRXFiM@2bNrnHVl6Q9`O4BOL@DkTwd76ULw|mQS%7+G(J9;56g&RM%NC|#f zJSTbcy~P$ke*fm)RM8p-N*`Njqa5wB8MyKjxUOR|@G2KYx$Ux4rgD~^Xk?Ug1c%tnGHUn0WzWc(Arc<`KdEH{|~JbJ=BIw9kox zb3s>?82Wqc9uCM402^Gg+c`=mWTIPjYGnaolf*+_hA1ENTlA|UC~vpWk=R;JWf2lC z>_LTIom^Q-<5P@o1dsL0Gpt5a2b|q8p@f)3hB!IRPv5%Z_*pG!#3G5%Y$@$F3BS9} zof*g%o(L~w`z3_=*ANLTld&z!XUA4WzqN5?-81T!S%}5;=L^ijcw*A_!!+1za}-om zFutcofl_5hCgj}4g1^ZY({zt1Iwzr@Tl!e?-BV@ME-LQ`n5_?wVYY<-1hFd$-S>EJ zKYOF4IK2r6AZ^Ks!IRUE=zkB4lnUtCG^}(g$M@nmanJPi5K9(d#25OUsX7S{-n2lp zcN3ypd%jtcWs+LK3(L+jR@ng*=lEWNh~Ws1U;JBuos@@@XJopDcakQ`JpNc)R=bG7 zkoR+`7*B#gCYr@)jo(qogWj;IdE&r>2l_-!V7aVhx9@g3u0&LxsV6}QPw7?~P8nZ# zWbqx1)bki3nYJ=(au?<=_*iD)T3w!|r*ypZkz0D$Z?QAXOGlSN-Kz0LDeel(R5K!y zWA#0;$V|#?YKb4&WmP8n`{tNq(P2QBcR>-gqGJtenE zqB<*5GID79rFQNr=)`LVO<+$y9QEEy#RD|Ls>M)D(YtH;M_S;AHh0uj>}CC_etg`Q zd6Os78&@ccLwwsn0T0!t+(jhLSJxCcT#OKiAW5`&qkoKM?dKrB*?usMwLDTH z{q>QuW;Y&DciAxzgCOxdMy{-7BWYvjDUU;H%HlvsJq`ZdhunhM`?%oGDWBA}2MrzY zc#T7*ToreQ#?0j;ZGr>oY99FSsh~!a2A358LB(RMaYM8_8qLNy<9L0a^W+<-XRrs3 zx;3UH5vBa@u*wygZ z5j0(h%=}*m-5-DdP=&>lp2$5yhe78_d1kCEH_<4_=;j=fDfi}L$@y10N}8M=N~e@yB6T7zY+;Xij9 z1WtRzE5`nMW3gp&2T_Pep8{pc)`|Q5*Y}<8j&bZ6Qx=f}MC|R^Ej|vWkVm_^*LVI; zyXD?=1z`};Y4a(;R}?qTp+t&4v=G&J(NB0j_;)eRBQ@K^c(JhLbM8p)rW=>%-+fT9 zYYQX7p70h)uoz6VeTb}MV@YxX_byIIO1rFk!C&>L#GnFbX}5&9(Ou2CX9uFRD|gq= zzppslw}`uwKbds*VW>MWYp_Jt!i@9k@|$SS#0N)3G`ugaF>`unMeKml`K_7ojqk() zZ<5VQco1j-mh=-&?gI7B4dh<5IEJgJOiviErVrZiDXm-?ejT2bd(|~upF@Zw`1C?= zuy~x|*&uCNlq^)dQ_*?hMm)L9#2I{LCjwP=py#+@Mjz>bl5C4%b;Q_5UZpv)0A3ou zce30<2#KkLT9*uimHMsf79rb!(J;XQ$nZ`v7FeM&xr7INIm<=_5T*i54S{+{3r7ED z^eCE>yrJE4xXh2F=2XeEv@K!v78~dks?Vpwedj;Z;1Nf{s4)`+;PX-`T^Wh%8yOOJ0Fd@Xkim#-{+tpXaH&Wu z9ennt;#GJ`hE~Njt(e(Xv!XS)2cu29LL4t2T{zv>+ByW@D<+S)%)A0J(UrZx zrzZV1HVJ6bCzI$B7fnG7Q_2-niKVBbo<7_iiQ1Bn0&+q5?YQSSgSCnMRReEM*L%)7 zNN8dWpT!@#jWY01w_;&g@Km=SsA;#L0UCz^8bQ_;IK_ zsA*RMWZ91s^$`*Ujxm2qZ~7OrjobF)UlGpt@HAmPuuR>Ywg5aXqy4M~yVHdvX^9X`EB zgf7vB*}Yme7F9bMPl1S(v@AR$oNW&&m*QD1lz}E4?%6+SIjEVPeKw2g*_DP#Po-}o zURkK@Ez7tG2XEZJZv6Cl4u*E(BWGC@YssLlban^)XE=vJ-Eb=5qf|MN=EO87m^wg2 z;Ay1gPWOb{vd8gq3!@Zb)XXy&L9ExrRa>l0!cZ}3Hsx;7KZ3Fr=>vZwefM)%DIv~}Rx5c?$y0nbstc%_fEvCYD8_tEl735$Lr4>B_s~i@4Oa;po z5Wi7+EzKA@89k~6Zy#Lt%10t25XsUbf_mR*(8&zXL!}8H0hHF0HyD{al!cFdrVr?- z^%zBL6U1zeQ~oAVBMgs-pVFEQ5Mz9A)wLl9Rt|F&1E68>Qwl!xdAK++9t-pFWVOOU za;0D>GC|qvN*ncUPV$Twm1++}=qqHB&;t?V@a4~BE)k(QEPH#MRl}6oSC8=sr{kb0 z0!4O_tBvyx$QZ+ES2*=*MSrP}z$>iQOhf~@9r@QHy^Ta?x?}n2#I--r&J}0_<=p#X z)Zrn-kQ(9Yu~yjG?F>h(P6}rVFiEY!v?Ktu(3#x+k?zQrsS}iG`$ZZoyhM?zLw+`P z*mmV45rEjCL@vH2e3z?TF`^O$6!ozYrUE8=CR-Smu{=U6vIq_*eCa!?s-^fuXIM2Y zGwNi#bSO}1NE{$YFbV%YC^3?BWg)%A={twQm8ED1lj@eu#)E1Z0vS@q{5AXs{LeCp zpwe)Om`Ej%IbVr1hQc?_Ooan{z<@~liy@Wqdb#tibLE#xp4${%dSE@(ZuMek1cN0jZORd ztH_-!qxP3g?KoAJq0gh3blwlPm!rDCsVFFbkj)T8&v~idg`fQcU-?inNthf+de^+~ zZT=dn_YhQ{4M5o6u%%c%MIePJoHWLKzc&y*IiSaie8gzXM(S$Aul@=WaaZlO_d=Z; zDtzm14QSwljl144rX0{h|6ORF%8)nO@M&z~I7pK$6SfWlP!N_yFo; zM=Jj0e$1Y8A+yC^2}|5Oegt0!hVetVMl+}z`uf287e1;Ua&okW)l0sk8zu%tEjlB~ zJCva(E79!=e1Yp>Q#dDRMwDF$GP0$}+Ku|4@t6oXY}}#8{$>>(ErDuhjv_=uo)=jk zF&qdFYfF)BW5RBY9f>mma*;_V8EulRm3C9CD>V=gVC?8~$YY_8D@GtXW8$E93i(3D z$T*M|@<|OVWRw)$FRE%zp0ubit|f&9wg^X$9eQQS8-7H$J1+z?*~{h#Nh`W_VnoMz zBzYq)EkIWGs@G1{Z{W;pB#}{v=nFcuQj)HSpP^7ATBW%xf{r0%o|3(chzzw^B82r^ zYt((@Is*rkr|tRzMT{`B6urJ)1>qG29Pg+se@JxNM`bbg&xhD7u8jQBK*fZS(GQNX(21_KbzR~YRQ;b1TspAoG)O2V^cq3A8&O~X)yyFan_Z}pG(eQ%oo_gPl zJ5SqWm~AY@uru>;E^&0CyryZ?2mBlRk%SV|9zLqQBarEIqitnn9Ff+SMGsW<_e`ez z#CQosXFVpAf$q`RsR*Iw8i^dMq6O2Z5`duXLVwQ&x^?DA&L=Ax-3v;`J93;V>pf8d z$UGOEW19^e(e9fgv<3_Ew`8^{oFdmuW!Si2)NY!{qu7GBN}w>3V5z918qlIb8Res= z-3ySlg_~e`Q$CFA3~%l7emQF1{2Hzs{-M_1mi9}}VW7O!cc(S7y*r@Dd>^JoC(ekR z9{Cteicp@tZ(2vU`Zf4w|A~st8}@!yF@mQ=%9|D*ppCl9HV&?|qlrQe@xbTNO;t{_ zF~By{1f_&Rc&t6JJaBI4yzM_agvGiBRWlDu`)Tgz!`4>s^1$T0+TCv#Op!ddfv+?M z(|weIRS`p$G<hSIDVcN934al2Dc!1Vfi3}*Fcv^Lf=@fGRX-dCuSiLPDzp! zR`4P{{Km$I;G{z!93-36ZIC3<4q;=-D)D4UkE5cJ-<`4*-dOeO=}I;%4^IMdkf{X9 zRWw%_UlM3XEpx|0`SGdhS$L39xMuV!D{eZY`FAD1OK;d2k4qhy78Zs8moU%2FVh>lGQ(&av7sf(zMau`h|&x=-&aMr z6jh}*UbOXav_^;lbMeS;kYDzju1f~d%}-{0FdAf)E-G*>Obc+R#pxtQ^ApBGXiGHo z?{n$oQ;nkICh~*-rYAmrAdH2(_}f9Aai&a-lG@ZQg7Gsv@r1#7RvpZ_y)T_2Z4wH5yNxtEmQ>8?`+MY%lxe#8Z*<#=xPUk?2>da3h2i|ZD3A3C7u!Or^X_FyU`Sc zaieu5=Iz2jTiOg9?LNWz_O_7IC1#P_Q~*eBEg@7m^`23Yk7j4NCOrPQkiBRIl2ZD7 zM63!_LKoHT5ts_wq+)DC^iANpLT_`%l%dG=dPFMH5x5cI$^&Gy$yXi>DSOuGG=F{vUJq6limCb!gwA_UFnO zhb2sE9b$<}N?!CsNU6?lU=D8@?r#r!oG+0Cdbw5Btn>sVmtI$3J!`S}2i_xO(`lLb ze0hghFRH2ESdLcaH(RLbu$D4>*up3i9dTkm59cPDY};FEE0POoCkAQa309#d-_6lOcj zLUlg^XMB8bJK>Gk4=RXU6p64+F?nb=D2iY=o}=6Vi(pdYppbDQHVjLOKdOR>Z*^d> zTsZVjg+2pfA4Df1j&LK`f4@69T;n@jaqe7Zv#?Di^kkiMpe`7xBEu<~sUi*bp$h4M zCr7vlP8c1a)#(Cv*#`BXEe}9FunvS7XTznBqj0HkW_9X%P)WM3U=C4qSH7$;QEXkU zVSj!=;eqp4g4cVX3CSdnfEU>!^E6NpU|$@XC`+}l&&*I zqP;9f9*bW$4f7b}uNOZ1;Ex7}>bQ|s=DK0#G-l*r>*T6EQXNKl;v|UWDVUaxs*&6p z{D}an>F4kFN$)f9XHAe7Wz+Yb;kHd=R3GrLu4Sm}dDT-_i#X%<8g-q{^@%0e2^}HC zCU|q1TYJJssw|Df-ZdZ$*sJ74;n$O#0hM}!O!^K5F^GlLmHnzZ zG}g9hr2Cx8sIQ24F)S{@vWFaWYpTEcqnJ6tsa`7mK#ZAqdk|;;7yK`wOsC-!$lxI3 z+|mZs-560zQFq}uJQ!dWZ5pti+hvK9JSwpBYfz=qN0sOEyBGj8#yvuCJg2AfGv7Hm zpj32TliFn@>pFEBRrP8!cGN#M6w*X{j%T1G$gx&ei93~00K%b@flzaEqUJe0`n0?I zN$_mz+@!*GsB*cjvqSmFlc9KjX#L=j$NgFX6>jZ#=p=atTsuvO>ug=X`yW6z1yuMv zSGD1mYdg4-WA~8CVysFtiRZob`k#t1ouT3WQlR;*>1kK@o=;`nT)J_0J(Db-U0Y`Z zel6-&@tDLp8usi6%Ftxw5(00gYWF(IGS85V?UtA4yN<+~gcA4~1F9W5V$Zj-JCIcU zx5N8zQ@3jz^q#aCb?=?Mu|qco3w@x5M+t5Ku+1@3#46uP8A7gx;Ts4hR^>nC3^j!% zkXeTFZyy%aI5RDTD2|gVOz5Oeb4MKfnmQjK#CtIMkdd@mXL>R#GCYg8qb<@cotl#nj~RU4W9pDAZ;2sZAvb6zV=byR zPgf-Wl$}g^<#1=fn^XqeP#ByTAU;YDOqrOIWxnnf6Wv`&{-kTaX&E$9?sP6;SYpQa zmW@z7ZbT(+6-oR0iC4a4Z{KBcqIvgwFicl7HIyS$)4Cs|-5$G^UGEst5@&{FmrcJ> zp1Ox!NpKjMA4}Xl+T`0Fe_F)(3Ts~|zRtHus^cblxcoRgxvc1qiDHVlsGIT3K?Uq> zLS`;LBXB-4pEnB+D785kUKJgNkjMpAory^#%LmWN(~tB*0XpW`(~k@Vx9VyJ>~L9z zA!KCs{I6p&fk3@)1zOT{;U2Em1-DvKCuQg;7~FxfsnSuR9*a^ZRk2G_$1UR#ukb?- zRbQ(U7&Io7##e{@w2VIV(<0jVXCRU~aC{k7!$1qpvaAGC&sSq{J5!t2SE__$eoFi? zx#(mTuKTZOT~;XAt46Ux_J2puz1s~ZiAbIfCI1^-wcjtr@fEG~sqQngE5Dc?uo;{c zSeJGB9oBKCU<<3iqwO^xmCtqku>(q#+y63G4oGM1q5iE=f4J7<{dNusFg$8LfZ6Q` zr*yF5nc3^(o3(;mpPwnZlQAO`xr@=sAhMQi8F9@3Cx_zwX)38B0mO(4W; z%vN|HcS2ieAVb$N-U9rd#Z2A)><)7F60#h?V0Dj;SnlV}8g4~UfC3}mn^mqj zU$}PgZbB)Fu|?d`U9r+DeL*8wlL!8BRx~foEI|g}1C#3`_;7mw$jU&@Um5Iwg7wep}&@STW30q5Sr?C30AI>)?&8hSfbF27a z_LFV@1S9^0^)TP$0XSl7_Bc6v!nMgx$>W<={8~J$yflYLO^ly zEYWV0Kfr>DiQOw`0WGb{0jj?Pi*R02lzKxDtJL`GxO9qZl=z~}Gg#(SYla7|N_*?l zt($y~n|uzsIk+olB`*CsUrlAkh=3fbT~d&gS5-@oU@4|I84t@i&1nP6XVD<^E2%5L zs8?T8L70rjfkCW)O(h_D9epy|YSF-??dCth(un#VoohwN6)YJ27!-RkB|&@7*nbo* z{0MUD;-AOE4ql_=Iuxh$sm&F^vGBi!iql&)d4;+P7(t_7i6#=E|INa|-*k{IpRQy1 z$6o-N_zLS>Cz;6T#B1E(lza{+lI0TuH)O~l^&m#YE|>s4Z18fS`t5JcAHAt*Ca4r* z#;dx*thPY@@)#KF_kxxz`4K-E!>|7VkWofp5K*iNK8CQc43PCD-kUuLpl2!I5S%+w z{K^1SkZE3M6}p|Ht0VA)GtYdB9Zrm9QS=7D?FR)@vi)K+-@Y?|kf`#8O9mL(FZui= zy4XxSD(S2hCe7_1uzY8Y{h`icUnIlA0sS{{j8`lt^2d__&^5Z{rd>;`IpU3ygv;vv z>xUEC2Eu~9PKaWhGNk*k*e7Mi7F+b-(ntE=6vxhm__I7k58zWkTLD$18>W$&rh|x%gUGx3J0#r$&0}I54}pSPq*KF z`s z{{t)vo>jgff9t!74Y_krKVGPEDo{P}{;4o4;C`|Ft&I|v`M3uQ8;_PTg60zNuYVA- zE#JuDOYXuPcs*$of85cX=~(dD9Zz2AXf>1LE7-_h-No&qDA_Z<+)>Iq-3^2rM+% z4yk?c^{-epmcSmJg02q1qA*0a*`1q%!NB98<@F0J z$}6V9E3F1Ko)BW%+cJ|H*{zXv^I}uykDa-?8D^fd8^p zb0Fim5aYq@3%q-cB$^n_W1!CA5NVEUd;EGO`gAa-TBv;lU@QkLDOGRJABL zp7lbM@S19f5#cr?2+rH~F4z*aWbp!k^6uL!SZE@H3&|4PsgqXr5Yz9J3*`05vhyp5 zAjkK7ywneMe~aZ>5BwCuN|YgqEjyZ`M~v=tdE*;DP*qz(;(zJ$-t}IHIv}x97O#tm z_MKuvO?TO}lyK!5lB?cGN~PRO(@m+ge`oF8W4Fj0oP`V^*J(_%6Ly&UFKu8Mf8vA|2$-t#h_g&d@sOlhdp2<}v#8E6FPuUkf}-Ry!b zzG&TAv(O{7$iuvjsY7%sjB&0-!$q#-(7!@TS9RUT1&^f6kI7yvUYr^;vb~kpysLge zsCHJNp9+Ua)&C0~tqfcjoU<2D{eJrrtn+vz=-r^H&~m)aJ;J{<;ik?V(|?`G+IX3M zQl_%#HT^pBH)5oF(LaC@9;_H#9^1bck&EfApAnD1F^T#&PWk7!h1U$kVJM9yT2&h3 z7$g4Li!ylbncJ2cI5@Pn+x`thr2m1u7*hVKZ=|`?<0bdBZlt!m%1K3y*29qyImj(= zG0XW3Ix1B7g#;q?&~t6h(#r~aW=_l+8Ua@9Ok|!OtGJ|lT+Fo1pAO^K4!yy__h{B8 zPT`L2A2WjqXxF{wGCq*~LL*44aZnTnZRl;}Otoo#jXNN?_!K(?ondsJfvY=(D`KF#zhUHerUY_9zEO1XAY0Jz*7CDiV2LiXUH9y4B{v-{ zX;@PeQqVSoP!0^uuq^MlM2eyFy4O3UV_-oTrKfGbdbd=7*TeDbK!BE`*oI+GATC@< zoc?hjcCmsPWuG((OC5OjA$DlCL@|rs<#0A%3Q->I5y#}*@jWcJjdhb?`TtAGj5#oU zxMRX;y+k0P;nCuTUkK%&YPW}P-5)g?LS)5@VRxcRvGkjEB| zhY;TxkD>YhZ1mI7b?r{c-6AH@L(wd~%H}L}R8}kn=E74CHj41S)I1FavDIYaSLNM-#C`@P z4N&?vjFHO8erOB*;fmUZwm{pPU7;3Ed^0C_TI_bQ6 zlbi>L!52C8ApGqT_J21e8BvgF_*Y>0E~$jSff`73&cNryvFemg!5JQrxKLvf+3xWq@x$-186V&m2V!(?H!uy?&WV4MdAmU&x3-0PU z&{75!4|Rf)q3Gxwb(`-eNowiQ#qsi2i~;G*#fD8-t!6=`(Vrx6#6@3&`?}@s1ohC0 zE=xSKuvU2=z4az#m#sbYzGR0?3y)VsUTaUs8MR~2=PVyhp2B6QGCoa1_SX0umkTw5 zb^iQQE~RzKVnST60366}BeuVInSk&~j@EQpr zaP`A5^#FIQ%Y_;~(z5k}f4*FqTO%TVUzT8F2r@VoZ~~&dAUM8tE3&XOb6oH8S4`(n zA2J01pIzYZxJ_0j{UKl6KX0UO9di zZOPa$@A@{j=bFzqwoHDXt^Fr%*ZL4#z z<{9ljUt1C@-@7PpE%in9xrEv#uC?U_&V4QWaE5 zBI=bt{{t}jPu$VpC`Wf zOjfz$Y@;2&Tjh}AT%Fec5IZnv7s7(Gx?$s(9I!Ic{QOB1%#&laesik#yR%d}$6&tg zaJf0_t>#sU)dv=X`LD;z|H6C|IgsaIxn*_3q;;%Z>(Xdn90_D9?595h1Z_2YK4_l3 zryo>J%#i&c=YOHs=~e!*I$0JxB_UyVa|wp4*Y)C0cxVY&hWrZ)!BaV{{Sk8h?|AlA z_~#j4+U>Wye~sCk1#QG%y#6@eRV)ls$iWVAp}_`61v}6f3n84h?n8P>;>Di;LgspY zx^=$LzcN|q^!m5$HEGU6rEX&6^}?$&?PZJ`Bnu~Nm%n|{T8TDYtHLDfmporQ7P;Nw z4g5{(c^in>l>CkN?@*ivUFN2y?0*^Sgga6d{uGh?_$Lo=fR+^`!j*9Bxc9eFRvMHv zfx#dFn`I!S zsQ+Uiv%MAcmH$!7NRUoVS?#8qEC!S}QB!v%C@4?%XS= zirvirLisJL>EvtlqcMOO@9R);^$M*17vUyGqUKjc;Hf(#+f*5Q52gX=Z?vpwYN)Mz zb?|(r)}IShGuwg1BIKS7mr%&B($5roZV2K$w(2(>taUG>l`SjwIv|aqkwxHuVAK={ zv++y0DuKWjr(NJ(vD@JVak#haZPA?nfbNZgi{@|YkPN>ke!O(Ns~(-dalalHBII56 zV=|Z?Dz>=bbx&yH4khgC4;MgDx;^-htmL5o46c6eMw>lR?`O?Ra6grbmqOpTL{*@I zA?HW3I$!%DaLy1zes|u2fE7bK67)vEl@Dultt#gSNT{vN*VX0s(Lw6C>9emZ6T;D2 z+gDdtpQEGa`T56kq$+A9G_ydfCG|he2R=`RGG5l^i>P2=%Z&lKl_*AA`Op7T;=+in z#vj*R16=qPLEd1QA}403e&b3RUe&w~vhQkJ_jc|2b?klJGkaI+`IT$_(YpM0p#VJD z4IX>%O&j+;`D+f5vAl>cy;92skhKp;IS_4^vknsKic-jQ!g~+%vqFyVA3*KT-z>2F zgR7oD%F9;SHTUpG+Nj&8L)xeiTg$O|s=}+n z3xc8FGUQw5TR!Wj)+@_!{sCl^B`DAur9&QqFhQ6qU%8?!?N(FUUpjIsT8ZFez2JJj zKEu>=q#uU)7s1|V0vS#CA&nqkUeBVplP+Z0jVK(0y;?Ir@sq#6v4Q~6&Kn`2jCEmTl#{^MNieDOOjRy)M=`s zWjuet5@FpHczs*#L(5<(B<&tc{Cz2d%nRhDn|2WN(oc$QmBh)xsO3-=8?V+uSJGpk zvI!x&KBpt5M%_GGEytJ=FjrjcCBdKPVnC(^s)+$2zp|*zZvpo`WJXeaAhzBxCB@p+R-Hr-kN>wCdm4ewPE3KOVg3tNQV~r_Kh4h)5h0vmim)NIqc{g&f`|2!>sG zrzsc75l!k+?X$-gng}~08@EQdx9NVr2m+owZJx%r{5|(D%594l2U=VVNTdrik)V3L z36Cw(ST%oH&X__Cb^&oT&fM-AaK*WAxIF04l#U4zeos?zykTU7UObssU`JQ7DxH$Q zWst&1Lyx3!xrDLOAtL)lIQTs{Jw+g_p2X2?!h<`8(4lP`Wzp-EqL|f_Tf(TCcg#Y# zpe{^91VrIAS(r0$NrfAI$l(<*5Mg&+ly2K-9_1QGMB1rdo}}H<7p`;~u5!mBdhj=i zftC-bgCVA3goe3a%3P!_KQ%ykvU^UkGUYLfshNMvCl)BKx!Kt40{965AEB_mWi*v` zb_r%x<30R_v4~o)+5Xp~PfJlD&lBc$vLLWJIWf+Nl^cTzIh|>llDebCBP@O#9lk(1;lhB__N ziUJQ!HsV^)mv#-eFQ*#I<&9G?7e`xhs9}qyRo_4f?Nn~EU?^Q86?fG-sH@!@)jJ|k zhZ=>=xp-!MuwB6|7jr z>tfRb7SQaWF$S5k23}*ZP$5Tg2x{B+L4v~+$R=ZQHAV%Zm+{07Hnn(Xvnh7NW308c zKZ4YrtIB>as)HHPzxX!J3oRezoVqIA`W-(Xmw#=XUtb>ZxYLg7@DBdy`ad_+1o-r= z=?PxtO@bpZ>z=FcZQIvJ(EI1A>fdwEES>Mvu0WX2RpGxE#la!+eH*8RmMQs)qqD3? z68(_G8L5*0ue_Ak=Ho8l=SYr6(Z1ZZd;v=&s0f={PgoMw46&ov+i1nu>M@u+wrHqR zruOHHN}7E1ci)r+tCppfrTUO6^c>0xdeuzTOx{#yt>WUIsC7emLn%Du_dqb^WYtV= zSO^+DSs_e0So;N30k6E9RRAgi6{#ST8pqLVm%d<108C*c3@>IQMga(=Y@a@~%eKoV znmdQ78vj%wIHfxw+a^u7V204rjcYSTvBOz{{W~ii#_s8TzWS8*I;M>T3RlRNAN*ZX1SthJ4yF!4Z>x09>=!v)DvGJiB?% z;}}rrCMlkMKVG=AEpTryoPrG2XkFi7&pL0PUAuhl@T3k6JagT?l|AGEPYh?=on?bJ z4sD?O_a9{});RM2IIgzdFeyDADqV@~^aAf`A@Z)iGwAE@E~q-%UUcQXMix51>qK2q zq!!Wt2T-J=@4>nmgM6_N(sbp!6+Xzurwz$RyHN8T{faCd3t>WLdL`&R4w}KDYR?d$namy7XSyRz_@5!H2{ie z$9AJX(fD3otM#{Qm^Zu?fIwmDT<~^A{i|j>#{bqd6Mx+--@J90CK47!D3E=&2Xl1l zHI@m9wnG;C@(^mPLo^5cq2ZxpRv)4dly!)+*JF{DD7402DgQ*EWtR)N0G>{upkqY% z_ChVKJ$4?fQ{Sza3}%LQ!Kg|yQ}Q#x`O~3BFy<(6U38s66!G??H^@fIvTi$WJ3R1M zLnH_tr?&kF`UnEHGr$iQlZ9}4E`pJ>g{&0(cVIP%AFc)wL!6?*3&x6js zs=X>EIV(sxJFkQOXWP#<^6|hg)-Q3>NaHmd*iUSzV~(qi3*&CnXQwYZ(q~`7wF*FB z6|?@xZhXNRD}?cBfdfr-BOXoX<7osqRe=r4cPNbrW|l5Gl3v2Ws&%6)LLo05zg(R- z<%Z~e@>=6sgW*9dn4ToGr9mG_{eq-$@jB=oHEkS<^rM5reK=eo-F`n6j0DlUn|JG5 z=K(!hQ7zCtT%L>Gvl%h2YwblJt^OO+cKZ#Vto=YyHeT|-$+!PEE^f7mCbsP}4R8hT z#g|pEKh%7G*!})+^Zg+MwDvewp!}w zY5l|S2PPr}8+P1O?pHF{;C^Ut0=Jl#|C7|xJq~((t2!dF0@|J0o$9~bDe$88SXqN7 z&*_9I5cg{70H!Zf@p&RTW)0S$R#2rae|^2%kd7fdl=3W^EV)>ymq#VaF#in^SBsm}N%!MXsaZ%UeEm{uTeM*kP+ljs(0E!3gs`!B4F<|x&^xh<9l{nga42nTr3~^ zD|o{N?B?n&8d;c`nRvOL^acB3)Fxw97DL6p`jHc;`&AxD^6b@o>F>EZ=VucErs^u~ zC46CBE678jycgM7pxM1GlFK*N4*w9}Coe-PH;xEBQ86M(Y_dTxFeWa!4g;>k)^_2r zNfNl5KRzZrDH6FWhePT|#)_nWQ=N2i-7ZK6D+T^m_8*y@V{&hssD2WJQoFOx&OlBK zr9zEY){hWXUt2q)0&>gr7g_d$9P4%*>!w7AOyBlc>5(=Ugw|{_&icHQGI+!MWcOBj zN6fWic26zuNPv1!qnS8Ks*!QY{2y-9&NfG_V1MZ$zH4-z4n#BCIU zxtaN7A5uaP97L7L#0eU~+6`WnQfH{8Q+(V33ogK~2B$BypT5O9( zLFw0^deCpcuq*JvvxPGR_ne{;?-;dda@W+I>Kg+xyxc?%Jh#$3sS$K9{ONr8mK+A+qz9_F{%@R?cDaQ%j@TRl zd0Y{aAMrhrA>#EUr~*GjVtPW1jceYh@_wKGKSCP(UCljJ&gr*`-Mr`YcC{jQCv%0IT4y@vDDvR(e-60;s?YG z;{PJ=t;4En+WqlON_T^FcbA*)?iM6ON(li4q`SL8y1SLm4blyYG}6)_Ac()UjmPJG z&U?PsIq!9yzkYk>T6<01pSfpd&8#(Z&pj_%Bsq)$($ele(GVydCgXTkl?U}$yumHh zE!25vFzyrF=CXCr=<5rUo20hkC~bkZkRA|4r$cpvgk4#F z&JdalqQNlDsgh^U?V#}n1Vhq{&5X@tdgC8<9$$G$mX%YpOOHe+HmL|u+6uXRYz+-l zY!SK=7*#^HJs3zLCldaw{>cXpd5=Tlo^K?nz$r(JAr{iREhkyG7J9X@t|`MeNxt|F zoWT=|%a}`)r_vuJ0Rbr-0@`EUOLr4yT~ zbG{^Zv({ieaa#ryj_K@}7~RhUfqdWPWFclcd)a6a9Ki3V5lC`{;Zx}#M}2R_bU#Pd1UxCUa?%dM1(12VlS@_<|-*~3^@lstI)Cy`m~~q@dTe3 z#LJ9-Cu0fh`c+w82ecH>n`nPhu!xmzU5LI4?FwsBcb_>ZmxY&Uh-@g91`Z4FyjRGm z!h_|We^0=$A#Mt;Vs|}0w2D`^t?4;c;`wILwXjWrOv_Gy5_R zCFt!Cs_`I!#7(sWV)hDoG)M(S5)$2JeC7C^_FX%rZew+yQjgWFm4R~gr_Jz`} z@0>bPLP?-KXRP7PU@!+wjB)+?zJO27POc)ZF~k_6+lGtE2xEM#@)G|gzDSdznYeaM zfZf~jx1~8rq{MnodcYHf#So!oL)Xo>sI#cULythPTRch>xX>Ej8Xn8YPHW=t2Lo0- zWIklJ&tR9;fX31u9fo!JkdLLZ4<)2r&zcu+_Be1JN~|uEiEuYJTAm>a3ixCpb@E4o zD|O@z>6v3HfgR{dQBi|}BI8t&A3xMldK*ti?WZyH`A4qG!;-n7zU3y$1e)kiu>sga z=7((MaZ#k2HS|mP&^!vnOBbkP2pNY#8C1oBh2K~SA68AnG6=aNKQjtsDyo_Ia8Z$W zJ;k0lJJQ33JD4@vgqVaE~K9b6ztB_WHf+;^wW&WN}#+Gq*1OJs((_suJ^dW9KkiNs<1jR zNC8r7l^5{N4&j&#i_|@X4&@@tHC$J@x6<=k2|NTc^&L(<zH2H(CIiwjzpdL^A*-fMGb(X6rpPH9W}aukB}K%HdiI__&PudJ=~^wh$BKH# zY(TNSHrT|BYp?|CabLmxHUb(!7IR~Lrla1iHA{hB8fvXiwzCUiZ`dJ z201(&-3sZwlx}xSJ{%vreiUV|fBAp%DY>pO>*}+1$hMM&BUEaQGm-9bq^nXPuQZpZ zP6XloqBGAkkbZh%_|lhj7H==};l#G=8lZXS7=mX}(HWZcaF6x8;t_{pG_1oDg$N`O>P$o=!6B%vnRTc0CIL+~9`K3YpD3_gzbd5N7G; z0sV6Uv4T;wp^w_tkepjWlJ=K4dXw!ZiJ9qq)FAE+lY4$`F@&s`kosCG2SJ zX_kygVWf_E5owKa#Ya$UN6bmSxA}xA+%F%jKZx>KhCA2XK+sG+HUL9kT1o95(2ea1 z`21fiBGBE})4L;IUA|s@7KJ3AI7lQhBH!!AhQl+_5F;8a}s zOvFwedXm#q5j(SeBTU2h9&^2OP2>$U0bq4Hi`Or04`-L4sj3d71|J~2koj#8Xr3tC zL52e95cy_+Bdhj&MhPMGbVe_*8;#YjtBP;ukO?w3^^3X%^tA5mN`c7=5(8aa{9jQ` zlJtl85(S>KKm(wDZLk^h3468u2o(zeC0DVn>c`~?*38QuV>$q>d96EZIo!?Bq+iJ` zkQu8uyD|+)e#Wh^q-@u(Bbjsk75peyd27N;XRAiVBl`@Iw zwGtt9`74(Qi0!Wp3B#y84cBB;{L}H{bF_xF3fFvUy18)H1=i`<{Fx(N&Nf%NXulU7Xxs=)mQ?;wph{dC`I6 z>6S@8ATxZb0u1z$ZYFV8VRO711B-VkVW2#rL4s0Arx-XHWh#agitLynx3BlJ36(m( z1#T0(uY$bnXa*Lh-j=_e(*=OZQ~FH zj*PpiR@EJ3QswT**WTY~^~7@~^jvF8B1OZqckkOmeA|Ps!#CQrbx)l$uEQk>;BX}y)nwu@6K6BH**)lK>etfY;=aJQ zOb{A}GbdYIW!$K<6M^^@0finz56WAeMZaY>RalMREVN@ZZQS$66aQXTX2ZZIDaINLnUeUZUKYhfEK{VtLS zlSUC~2|EeK8}n~%Eav}W=5dx!@M2W#_)6s*r7#njNucmn+;2!|;@TB{xX2;_DOC8* zEXI-llCftcujUvc;>IPzCp_;pk&YoCSwJ08&OP0)i#H67 zuB~~3&At{)yOmsr=_xeVJ6*L)Ihc_O`HlZX8atEMJ)!ATa%YJ+!rS%3pa`NSrva*wWPqP zwB6OypQQY_p0u_d*Wx$UQ!Htd6$ES$vC z!TUDI;jgu(LP&8CwZ{S~_qb{z-=V?EKeR^4NQ+-Rw-}0Ar%v7I7wl-@W9MZF?dN-Y zUbM)w(FDo>CRL~2+$D9M1ENZW4&PXOb~PDn4EYu>ewl{Q2&52-e|?4H3{ea88CC3g zS>}N+f^zoH)qE7P5`suE-eFme^sFF>YM*1Iu60)xyZ3^x?JwpVkGW=CH)fsukyYBn54mW8L3~8V9-l=n@knkoK5qky&QA?Q6qfEkdmlBYe zA*B#NtnRSUllJR_(2|5;m0e!vq7F@&S){`3#b0Y1;m7&CC4z~>cMRvL#CA8k!{_#Y zUxG@jVT02uVAraxk=Bl6@%)%+XxHGzY=Dr?o;ZPfHgtB8%;E*Zt+n%w`l?(C91H>x zst!G#vi3myD|JkNZ3)6%%-5I|yx8Pd^sm3fkmXj=$hpSnI7QN@`Ap@-kJz7MVJ~S2 zryUSeD_EfbGh2p-iL5!#243OiwTk{RiA=s!cKN_EUFX8i%7MJ_itrAws;5+Erg;)} z*AvBR&w%(#kT;UrpBKecY3P@%F>Y!tT8X^@m&Ni=+2dD9-Cx21@!pWQY2jY|hsS3{ zfn>DwEyYW&YAx7^6Tud@IaC34V?QHn5WEeY-U|Xv{TBLFGpi!}B{5uJ@Dj5fhRm3O z8IS^|nz_hV0yR)a$s!E_w3zl>`U5|Q1p?B>1PJG%zr>tJmi2z6EuvI2I#Sm3H-G+H z+L(dUoOp5#lwcFf4I^aIpKngL)V5Ypm8k zuo%=Hw=!65bx$jZXK#TioLWzOop9`T-IL6AeS}e zV9$ZeQ9_Os+Mo1Bh%~T+EqDXBpattVZ zr2VxnHaE6{i`1hYLFv#WF=k_Td;!5AumZO>iFhuw1G!R7LhN0KpJEsTV|P~#dLZwF zwe+DzSvg}UVeC|A`lQs3XtOR!N!bn|=kkov7I79Mj#=xr?nIveBijN>2L=@;2UXm- zA=iwgOdmTD(#_}trtx&skP}rRKSbAF8Cu!;r?d42iI!VWHyn7*eDN72R2^f{Ht@)@cg2>i){;IVb7oT-8kY0j>qiR47NF-r2bdF&dn zKSZ(rW50MW>Dns>AEbfJlN*W~2=%7y!TYM6s-3)|Bm^4i}teA0dk zOAFKTy&C#t(e>8TOv$*U!V(e<|9KZZ7xiX)?Jchf9tUkJ<4NO5S%qk3mzx|@dL1`A zXreXw>vvF6`AqNeOLALXiTRfB*-n{x-!1WlR)({*Fgj@I5Il*~iv13ogm83!CisLEyDSQw!6_NZeR}{bQFLF( zW-#`QFI*3N3dnFLe0f`85_M)ui3r`~!@LY6&hz0fba9w+P58R!owO+Lm z?UOp~zo@EqfG59)k)7AV?&&obL80{b6;a2&YOCw20B@q%|Ebc1Dv_63F%4HfYsdFB zN`$UZ$cjR$m1GO^H7850XBun``PfL3qu{$&9~@@iv>XcmSJO&QL=()+Gh8*Q323f{ zPL+Qwf1_t#wPf7V>iS|&Y+FknHye*gDHk&EOB!VoGeBs}tnz~`0lv_Q0s<=?v~m14 zdEuWbkgfIfeOM~QlIAFe_AxhILr$wS)01vu{&g(I(gX{8hwn6>TpCiyWc*#%1e)R( zpY$DpwJ4MS&1&+ETF4zm!A|M7LNqy|fakF^JzcLl6yr)Xj2e_a$T&BBJG?17O*ZiL zz15H;KxB~&A{w0i=zMHqv`jeK_bwjiVuCN&H@+#Usj~6M`NJP|7n){a>5eB+PeP%; zFv?z4SSHi-_nJ}sNez{S&sc1X6c==-nvxpw*s~}tYmJ0Mf}0u}D=VMbl9POx8*6gK zM1N+jTq-z}syp{Hz|Kx`g@~Yq|F;RFC$hTh1HMEmMqgqU0+t}^y{srRDh#UKTQ8Qk zUX1By%-KVl*eAxlyMWd7<*eeuHlFxewd19g78v>XG!Hvne>eA#LT3rHtYKVm&r2t% zL^kL)WEP+$VA2{LSoxHiWAkTfC1O&bPgpw#Y)PT7q$u-Urrn}w5?g8Mg#tF`Ml=uy-ag!%4^D~D+g1&v8I`5k|x^I5cDz+VI3a)xXG=A6` zG6mQu)u*)+yScDdzhI@fUUB`5S$L(16%$VIAgB?5T@+2Mt9i26I-Ay8zQ&78$w70~5?m zqB!U`9`MNr{Y->@X+pmqHEAvBHZ`X53&~RLYD=un^c?;&UMy`%%Zi6|FLB}<&}R|B zJl=W>733E$%Z_xG%iHxlsxFN8T4#SaOBB_W-69GCFoh4b;lAp5RoxhWX+03UOIj>% z)_a~yG1Yp&#J=VcophgRSXrK>6MrTA2=a2h^~Ct9h<}NFnLWj4O_o;h*}_UbQp}nZ zw$VkaQ3Jo)lp3A=+yQ2M?JYVIEo~-n(#t1AL)}6PAOZPYd5S;>vAf0CO#ZJ9y^RR??z-@>It106z&*`+ecW#4NC)nw5ywb zc?Ufee@Mujb!{kZt^5R?y;T>zgbuL^XN1J1Rv--6=Q`>0fq)b6Ofq6w7Qpk^KEs@+ z_?F~2$CF=ch3)PJfvCtB==h?$G5^c5O-H?;9Oq|UdM_J z4vu+>+kLdcIQ3-<>l0uD4HhA!D%*rm z=vlinV`yp!Vrf zj>6tEldDb8Q?AXEBDX5V3+2MT+|{T|YiL!Uea>umt^34+jb6S&;{+8BZ>tQ<@d<{l z{BBQK`AWtw+*t>?zUv*ylJ4Fc?c(BqKN5pRCHoro3)YBs`tR^%7WClYyT_ms+M#Hnod$(w=67?L7mrxS@mC3_4I0d|q zFfa*l#67RtfnP>QLYXQZOgP{B{n~kpssZEZ*84AnKfC^NN?KcH-8$(~gF9Jx zpg%6H(S{QnL`tFK57j77p2Qu$YB3>O?y$p5!5fafEnI^BJ|=dsR#m&Nn!jdhA_r5T z>0W4#uM<74;OFS0@nTx+48lva@Il%C@<#%Q#kKv11W=;O&Cl`9>z4#jcKZICe`pAr z`|mUaRaJ>7nXuQF1N)FH?l{RKd3Ow{@CWo!G$d_^(&hMzHTUNdM)v07<`r0HajO}N zJcy$(yvfL!lX}rFrhT;g0YTETJ>h?vuGrK#h~iuPnXk;QKC(&kU(2g6fK4Xlw#yg5 zF5k-XSF^wl=cdXd-P!iNg*WX+TBk1sD0pG$uGnzUbJRcR?+?B4dZl|*ZMy_)O6h1X ze>o5Ap!rg{VlvwbY)1JjNS)_{V~mL~f*BXEbEOoG&}6n92yTA`Y#IiT4ipP5=6`K# zDfgLgltq{aP`?yRpvH4-55Vm@IS;*=5%{-;qe@RRnDjx?5f}}Q=GgR#6F7UAD`pj$ z39mEKl1^C^jhZ(Bzinsv&1=$`00=IBf~Aq;Rpqwq1>jd8POETTl-jJ}W9L5EuGAYH zxkCo`cjk3WKakk4Y4ochuvMz@P=DK$-fmq4| zV*tS1S-!`^stsNl2Vnt#$a#GNv)X}@hIpJ0Gmt=B&&5G=DAeAT{S-AEa4RQmOnSuA zP_g(3KT$~FnfpcHvfmmSYBUzLYe26&ZukJ+H?cEC_@~z%9;|Y9ItNlS_p;7UrshP? z`zi$*XSO_`=s+yZ{>Pz-iBx}$hj(N?{^rv1s9hyri1H0pIahDh9ny0Ka# zf;vVjS|!N&>v>Ln%X!WjBcB7EAdRcie(#ynB}Mk&VKxIrN-xDJ*J9#x{P{*rg=^ut zOb_~K402pdxDbq}+<|*&iY5&uZYTI)<*13 z)2*hw--G~S?l)v-wyX3h$I(1(rGGwr%nUl|u4=5Qt}sOxjlay&rcLP!PJ%m!>B;m= zTG<_`cH%v{5L;2hiG7=(n5lw?Ur_{TcZLnZs+ zY8 zc>%p~+*PHWcBe5zhpr4zGQut=Afu%XLvtu?ppE0^>(~yZ9N1442_PNet#vc?kMiZp z=r2H#n-<%8NsM{r#|)vX3TThjZSr8c(xLkcfA?j51!2+fD9=Wf{ARul(!gFky6GhWSK{}zNDm8Kwdaxej+#Biyb$xVe`Ihxt z7fK^bh%c;V;SXuXbsE>n3Hy_?fYrY?b`T{2yT17*9C}HB!)gA9dbmDJH+~1Lt*?7F zj<+!1DT~%!9hcyn@>ugoE|1QZ=(v^zSrx8CK4&^QE+2gxL(lX8x(lZKQuo`1h4G`7 zzi*8xztqz^tg2p8;b7u)OGRXzS3AgMudZ|z9oq9snz>;q{i3J!xvKbsv07bAw&hQ6 zA})lO0Zb=3R*YF47w@m=OWv-NH~aaQ^KwU3F)PLt9W7#Bfd9_7i^-cEjz@Dj>MFrr zf?X~tu=aMwm`%0hoq#ebAna$7Jo1=!7tw_suf~ZRTkG>{fJ!N&D%LZ^%;%wwHQQGm zH%CzGh9QDy05d76-&}8U);pHn3wAlS(5DE7=i_yw&PMyVnwI>?e`X?Hl)TgGmy)@O z14X4(m}uJ{%Fp^$-pEO`_CM=fJvRQerRF~_`E}lFHR*gU-(8hlhH_oi^ii;WC$S-LikY3^^kW*! zu6L!PDFl9et+Jf?&!dI}7#IKg3&FQ;(vh-@th17Fn2cdu$@h->itGfv3hb=prk1Sa zo!UoaI&%`xzu8?W_I~~K#Wm)`#%RETn@@l&*%laeuN~eR4I}zGGle zuw}dzP($*-k{^T&aQATrXDO@(lRb3G3z}fcEbucC^CZhLV8b(An-ejOB_hRqWwS#JpT@_+=E@oABM5!fUSQi)uK5dQk%vnOp&b~hS>XQ?cU;0^C*5hPcJFhQ1^9YegyGb?h!-5T&< zosw&E83^1C%nQ}x6y@^+`Wl<-Qal)O1RtCjj)mz*!N0rHBQ=1>b9Q%XjKrg}{7@R% z9PozJ2vLi4EW9ro6K9T~(3}^G20(jEqyYg|WR2IOb!l+l+)*aBWKN>nbI`H7n;j&X z7SAfy(QKci9W=^>^fEADysq{gSanrVWvIJ#`un5+v!QlFsg34T3tp>~HwQ+0S)CYG zN{Hfg%o|Zf$2R(-$jg|3(9b&enT54oN_L@NL~@G2_X;!OLl5JNBQPT7nl3q9TxD<* z$$LAq%HZ65^~nJL!uOpzg#WrrQ)AR@uvlH(Ww0x{YX zChtbK_Fsn|2aj$0)@S|A!m*u?^iPVq?$>nT&4M>miOM_Hvs)(`K$oa}){|wQmP3 zc#~l~Dn0~L5C-6$+{D~7s>{oZl9OM^%gfn)CRbf3BO|>N9xg(zxZgaZt=9<*PBYya zCKt!*duJ8c8YVaTCHXrn6-yeDd%#S|oy>#J|DdCV-Zp(VLfRwmuj+ImXuuY5+S z^Mf&^Udzb1l4@(KUqID(L1biJJvwX;4lSub!8A(G%WH`)rf@h0CVmG@A1heqW^*NA zwxHhm=#b>@l7)sVFaZh*itZUxo(VVPYwk~l>$9#fpu?3{ofBAP9|nw-0P55E zqph@~!@lL0e3g&3_oP*W^~;U737DG_B09z`RxQ=DI{9|CJ)SK}k}D67oHrlFwsieS z)bsj>pXW7*3VtAd3W8i(-e$)3V!hSL7!ZEevuAmO|1tak@?t$U#yThNfWq}-Oba@8 zAFn}+z0f&-{YJ`|1e+te$ERTlXwK=3?$>A!=LArSO*4!#$zG?Y#D4SE0> zjm8R)u0M;RdDU3yk{4bbw*{^e96JpD!Yoa7ULEx`ADF05I_bmghuL_xK0rwPwef2? zELKo*ow??Et8lAuf~;~_#l#FTG{9&yh>jjhiJGOpyuJ(_K8?R>tZFQWLm$D~Nu)a* z8oS`BU?1d3aL&-@z0Z5gPlcE`Vc&&A@1H6~O3k$d+!#xI%MX2z?sI&}2H~@eW`tKD zQC1L**&#=JgNfp{YYe9sbgPFWR9386>PB0p&F!dOs-9zUWTT(9cK97c5zZ5_NanuZ z%X@xHIQW&uXxUAqc&%%5^&#z@OS}SYy!_4hJ;*9#`|8{Fujf6~^_AYEH=*~%WLzFz z-BjL3JR&d6HRdqEK>4Lg_J2RUcfty2HaWg3!WPPcH)FvG;_f`*3tjeT*dOXlL z!RMB6ym(oSXpceWCkSx~#&y(${5KgUr2eHs?tk4mwJ8aIslF8?U?%6vux)&#E90@Ic%F4$taqe1i;rNS z{rKRJFDpi0-_GdT)-$W!PI{JKg?{){H=>$qQz!QP8(-|HXnIm3wWqLgnWylmOCrtczSZtn$3?1||RTmnwD*_Xw8ty4OaV_h>5@?5AI3G%OVz@Gv_h#l6ig|nA9^HT^Ug?`nLCIxQ&w#oPiv)r-L=zx&m}dx zM@4ezap*z$x#%ppHLEo%*+&u7e*Y_J5%ueF<8jq~acy8PlypOyUGskPp0{@}i6%^- zUi(Xtmm*vzSShJcoI*ZiaRO@tm{M9pD20lvd@8}om7L@RUMOV*_SjxUS}yz}V}PCl zjUUrfJuM%W(#}NTwn~JRtAMjkl}5ap39Y^Ad+PIwj+x5j+FAro{bH0U6M}6s$73Fr zmisjzn zoQgnyg)gOd%J^G|&Z%dstvIKVL}1D{8{#q+%TtD0?TfI>*a4X8zEgj z$Hnkm=ipn8dwpS&TkkvZ=jTpfn0)}9FG|r9yop8EgcQV&&bZmc&aiEsHV;-fvS%=N zxk_jsNpA_s8#OyF5~ZoX@I!Q?nG;POxq|V|@$QSiGkFuk`nHvO^l^M;+Vc*Sx=SWqT|K8-g#-d0&X{6yaE8s8b z07kolysVKWy>;r32~L^Qrq!SdA)rcr(&|N({LDD0s%PHJuBP$5;b*oN1g)r#`YK<1 zHI)l>>Yp@1d9T~kgkMt`^pwTymjgNcM=^^^tSONurFvY&nnUh+=n@oWy1jBht5raA z!!~S6QG@O0i`zT7QAI0mBEEO8#=W6}w8#R7A?eiu6YWGE#84i_Hp`UOIUGlcvtMvu zHu*(S%%Gn{#%tMmSRx}l_be^VTGsft3*SM)nEdfeoJ8!9;!nt7FVQXB>^_hpJzv&p zhgGE{vhPkcAu3M`$I{H{KoBJgiYaam7v+5W&H}u_J+`Qh?fRtJY>~``q+PuPJlMrb z-Y@uxd@tVdw6~&kzy+b`SYj*DVdoxGj|y3QfrR=i%Dbk*C+Zk`M6c#3Pz9YRHQA}> z{iDytdxm?}{S+u}4%V}%QdSR%r!NR4> zC}M(nBOj*66SN#@8;LFjMf-c1Y9J!-PE9H4`{>hHHRXi}i+vu%?YzuC--lPG9TZJV z49Xk`APg{MT6TaX46e`aPJPT)ty>v(!VSN^g{QW~r6?h^kgfYJh^0e~ zI=w#R7}uj|3GEanLn)NU9XJIvcm&>RoIX3b>y2iRmLQwD0qXkc7qM%2t1<-^wui8y zjfohXowE#EGM5F8LR=+ZHZAD}Hyr|%tWB`O3c^<3HxGk(v(k%;ig;&5@YG~*C#BCd ze%&xCNud%o$2HwP_3Sop!hg#AUZXJO{OHnlD4*3#)(H13$6_b*X)ro{RL% z!N=4eQCHlBvACAi6@V`H4MRHevW)>>)k$r0pT~8!bZYX-L zi$)4I_@4MkUotbsLv`%k7B^$G<&_=tPD{m8+7*dGCA8*B(llz#sK`~pWq-QQtKe^Y z{Zp98^qbcWFbwixq1SaTQOsDpccF;d#k)A{Aa1&61I4?>ujG&gJ=8O1M74AU9T7gM zFX4=E!SBH4w=;JUH}g#K}RtvlS_vgd+){Gy}xtPt-4qrgkOz0tOkODexvj`mSETTbTPx#Jr81JB`SD|I~B3#bYDLa`7fbf-p`ea`n|668yE>ENfE3sUS1+yOSEWTy(e~JqQ@ale)y*TV>qmcfQqQAbvj8Ej3KN%-RMH| z+7P@X9H`J4yG|N+OEl^84J<}jn!HVodSC@166ti%IH-JRYv$Wi>aEhL?DVK1^n&$O zTF_~JbFx*U+>(6IL4Cu{;+iOlquyrTGlZ~u6n^aR0KYKYWmCjEt8Trg7$t;^ngOqZdzXAV~(;!c1t$&#)HvSY4Q&q?EoevB9-b+Ugv#v`CbHT<&HKF;>?yWk~xJcBx5s)$>eQSU|)^ zF9GiL0$kwI3x&#+Y(%AlN1~(sa|S7409ko7(Hh zciYXjsmCFBtz12zLz@y`GDq+%W4e&LR`*Q0En%2x@fVRVxkH+9I>bDX^8J*3Ac?Kr zHNejsnl6B*5&CWMro=o*inj>qOsy=w81=C}??OV@PG*Tinvr1{X+gH)jIL5ip@*X= zl^%EU77^FM#vq{iaK85*2ZjE8_)Cu4w=^f`QQE1<1y1Ibd}r6+L3o)SPhyA-0M)y(u3e7h2dZ3*F(g^v6AnA!4zc5jb~CX~?JcC7|I}gR3QZ z`jThlQ^WRHI82;*1?N=WAPYD%mINto%$BfOTCjHNWox^zwnh)yVp+fWEv`VcYoOWv z`wW$Qjac2LhB*oaS!rffa^A z>9lC$+?o0$2?>x&Oy8ccQ0C<}_SF)`$G&&~kF6UPu+GJwxqFmKi6rUlRM&1)n$xn$@f(ykTq|rp)pDG2Qyzj6=XYUVJ(K6=3v1G-%Ouc%46t>4C&jW)7ncf_?%AR zOqc4{o$=AJ3D1Udh?N!6Z9l5Yc;@p=ZZ#%{m3w$YT3QnR@p0tH$P6LBfXl}7bxBV~ zzi9De|Hi1M5O$Rc1v>Q4;P8y3`Ot1XA zi}K0tSyjxpsSoC9q3em!GIic2htS`FWBQ;Trm#WkF|9x5#9)4sFykJ6y%FDVzIu+8 z_-U34_BEDMI}55-Y*M`u?-M#W0&7m}4U_5cTO?MLAZhPN70F!SbqcFQn;9Sb+Ej%p zOO9(KPdb&W*=p}hL)Q;Cp-I%~L*Wr&sX^(i&e?a+VnGs1pBx?fTd_9M^p_lAb2C^H zz_gMH@R5*08#_mN5tv45vJ%%&m^CJjBkFsK+14CvpiY>&`JI-k0z!<&z#_8)l6l z(%jBog~Y*kDTLR5h@@kC%|7Hg)yE^inN#<;Oz?wFvv$c=%^|r)$>(-!p-`>v~2 zM%3qxp9PeZJ_c1Of#SyuoX$bv&Lq2!sNQL^GN3cBkwjvJh4+#K6MH3dMqZ}BJmsJa zX5hxYf0Z*CAr{^-gRS`0c7o^mIWzls%^7Zi)iY}>nD^BWZPcvq;nY0ec&z~YjEFcC z;ZVp)ousOksk`798%`JD*~4kl$-sNBMQ8R%KZ*AARy5FWa57-0Nx5;_#K-qHSmMiY z*gK7AKvh`^e$zgWOK64E5_DR{-la|pNGm+ zyAR8kclJHpKJc!I7QSK}Mgi$FQcu4}*r7o3@Y_3lS*wUcMlK&%pwxi=K@@{g*gmP< zLCH7W#&r%ZL`}pP7Khm4@o*A~3A-ff{ngp5J73}PE*tb-A-^+x&s}0FubhYOafPds zbOh45(*8`7_*@^5(`wtRWo%dJz%W0>+nf5(NzRwClLL!+yZ$9H&vfDa zy;9H8fTdtH(b)8Q6$sXKIup(ato+Vc`(gB(qPnqaUk0~mm-?~CE_7(@NuRGAeAZz0 zzGS1ns)hiXxh7nem^GL2*6bZc%qlq0EV44wDH@-^_gR{l&>)_B@g2BfWO#{ZMDro< z!Fd?i(%#R<2hR5lGxFnqP0E`bcFk}#-G=u>88l8V=;0DoUS?a`nkCCcs^H~VdPNt2 zZXezWlOw3gsszZ3_i)8$AdD~ev+CmbY(-)1ssqNPr=pTBwVOgO?aYT2C^7=W_Ml^qJo&{qO;HsRc)X7bvu zhbWAX^o(kAr}Rm&v>`A_(b_`nbW~B*pl*K(p?pXjZIAigGx+Y)M9P#ndEXFjp-Cs4 zP9_HpP6MM#$xRbFwG*o)X+#6neeISdU~O7|P}R0T=iN6`M_kpN3fk|W;~@;%shY-n z9@2d=69`@Dsxzz(;-)mfa=`nkW=(Fx=Qs&NIiu>ktI1Ic=|Mx0eK}7i+_mMtgXUg- zUjU)O!GK^vaKK>|SZCwl|6kbQ|C~)4wf`(F2wJch|6N+%e-!)l&x&Zjh5E<*NTu?x zX-EInIS$_V8SQ7nA6K|>@M#F^#u(=`I(QmJaAQmmxao}o_m1)QVgm7bHzEb6Rl&V# z;EgEoMl?W>_pR^74@5lRdJ;o4t;X3K$$O?Mcs9;?76G0HQmKN^&;?iJdC#K2r=x-! zYP`LFLp&ZPgC-y2U6-fL700U`XT(*5hf(GP`Jr)-lby=(ngsTg?b0dP(v-dQh50|BoVnNcf}%y}5>I1ZYM zfyV#4tx<%pi$oHaqXY$m>t(>85%}15WSmhT=RV%LXn4mlM#%xZ(;lqT0M^rL9{>^- ziMH)`dNM3I(Bv>HPY7cd0$5#+=(#vq!Vo<6An2r**Do4A4pZ<yWrQx82_KhjM#GU$Kk+MZTJmI zqO)Gcyg_iSJnO_bqyH!x)i~{0G@4@W#Z;4_NR7nHGF?>K!` z7^hXh&sB+VkjW<1@VFuHFc87NY7NNHf3>y$&ht++Xy+l|{2LScf0qX>WTzNwo8742(V(@MNJ>P)XZf0$0cUIehWYakFbkglhuM_Eh`?zI5<2u>}Z{bQgod&v^U1t*k$W#L!`Y}N&q zl)&Q&Ao{NbB}qm-^7(K)ubS*qEqTD#=IRz1{>%E4$5yw|*1b?g&wvFT{S)};CvwNl zn%Am0+1WUj#=s>7AJW+{EmjQpObY22PCv9koedFjs`38UmU6OeQayB=4LM7dg69e) z!8S!fVw(T9GT&C%b%yzOz%Q?2I{)te4k`w?{%PB`vN(YebqoH@caXzXuNqe`8t>T( z`0VI+`HzFQ0TzaNr~efD@1W7jQ$Wo5+9$&^7;7Vt{ltzjo>}P8xUiwEF2=Ygl!XZ8 z>3y-gQL%wr*CRsj-qxK<`B37ZxR0l?|s2D+?Mf9@Q!>l zx{9Te6!sEqay-Uy^{R7`V_c}F7x`w>C;vO>R`ZlJVCn76`0rxg^BV&nTU>|%2g2Ru ztAX=-Kp=!R2iCffe?i~?04>!1iFtNM_6Fi9H|iNF8&f+>hz{Cu?)F;SRpU2Ax4LgS zK<1r)l=OA{JLp6aa;sKg;UM~()d+aX`kN!*R0#ld|6MN{zRu6WIs4HX|Aq_#o?hSy zg!~uekx#wKSY2pDU#8E^E;pY>f&Fm=)>VO>fGG!G=DaRfJio?t=^vqQ=Y=mEUr$q9td_CxZiQMCwi?0Y`6vLwyHb6F7_{mJ9e-fhv_)Jk<|@Z z?ctJH+`*x7t>6rhJ^w-V2Y0=yf*ZF`XUspDr8A#I!`ejC*JXS)cIbU{1vz@>VJCEg z3uCIoxsiD@UIJ8YJy5k;=MU{gCvJ{@qfSUTuXt@7h+Hlc02MlYm3?yrwqvbJ`NI_c z<;qhi<{p%Uxd0aTnTqTF46!9r5}^B2OVZoGge+Tr)LEyn#&Q-H{*C`!0m}* z*OlDZgI|wq$ITuLq86I)Q6cZN!p%70m!qs2&Yn|0ICQZ;v}r*gC1@4|r~hQ+N9Yhi zDW?M>r4<1-ihiuMG>!y3HX$*bN<(?L&|ma_G?}OIH%Bl)TRQvQQ^4cj9RPKo>W{`D zgw8wfLxqB%H>?D{Z2PNtPksH7IG`w}m_M-oKoedocATx=RpCay+qswY{4p-X?UkBn z`QF^wC&S;#1p!X0=8qzs|9WhWIzC}N|9JPQ2cM5q2k3`v=E=M|jG5$Ei+-m8pAD-D zZrwti|K%fzFDyoIBVO3tq|gYn{*^b;$_Kj4!hY5UTh{i=Up##W0af~0>MzEB+%w~G9iY@%4%C@Rz6`t1*kJka3d`$HE5%8Ga{R6TC=wdy%fR(5Nt%ZjKu7 zHdJzbfX{ijKbYN-oC7}@WYqe>?_WVoj&ZQ+U|6L&Uh{J%#j2~rZ~pJaX;IYuVnyYk zwFiAnfgTv>+5Z(57!Z`eXBoh_3%vFd^`bra)^Wvqn|U^F^@l$90GZ=YpwpP2=)bNf z9^cA0MbB0#KDblFI$y1KlLg9qpC}R*uXezyy}BG0%|A-~4*F54|9}PXD(Z&BGR8%2 z?xa-sfKjg({p}f=Q~%#yyByFs|8*SrjolkfT(80__5X4A)&Ws9-{UZyOGqfONFyL! zF1R#;bV*r+bVwRVHz=JVNGtG&5{m+g)KXGQgMd;Jiqx)zAo!bI@QKgI=llD<@B81& z9qyexXXebAGiT1s-Fx-Yxp&1Aqgz`CFTRduCtXQJ1WB*M(Z~a=^u=~6PMQ0AbKmZ- zfH8Q;gl?e$YYZOCJL`?coy|lHzW}4)8yw0+$5nv+lfu|JDv0JMUZWsW*2L{HJa;Ah zMq%$R|KcuF%>F|>Vz}xa=+_d!<%j=4W6b(XpthF2rVrHVt(RH z{|DCn>4qRI32N5R-7rN!%c1S6E_eS}T4gY&{X$F+w3uW6F<%`gq0^SHvNlh#G?#HO zKwzhc#8io-RjGr|UQTSjf`WT@z!6<(SP6{PQ2vQnVILK+UH)5ka6 z4A}V_fQ`wAX*lc!nQb{? zT9glSKOg3fluy7nF9(4oQT?y#u}<&*2g-(>^$>_y>;6?tpXm2)p{Iv^p6!i90P(LP za7x0@MXyV6(9ZT4qz^nv2!f3_1> zhvmj2G5OaW@Mifx;LraZ9axb4y}{k2JNrR3--2#W@SA>EJ)E9BVBJLMvDNql?f)&? zjsF@wINWRLe_{R0clhI%N{^G|ltO<)$p1G#925Kp+x)$Ah3YcaukPJ0cN~o4PYH?P zjE6omJ$Ky03Gg8uvH7OG?kUf)?7RPDu7bZ%{^i?QANpIpF(_!uKZ1Xw3pLS8{h;FA zAhU|AJ5d7L(+h{A-ys8d16o052lbc)ra~GXL^rgui6{&Q0lg-Dj2=ge*Wicp|NGjX zp2pgHumq<8sQxkWp(Fl6j*gU?$3dM};lNP@bPBAcm*Lok5eGdcQ>f6NLUu>K9*6%B z+_U~0^K^g;r9W7NN$3d~(@1S5{Dlh(s*B+mAdq+(5KIRxup2fjLFhH%Dew4=AuusA z{-Fq|p;-R;Un74o64c>OR-s(Ms&cpr{5R@-@MGc!I{?8DB?OIxVj?(>DE15GUsOO- z175(BxBx9|VKR(^Z0gevUL>k$!NaG4c!%&9w;eZo180{zrUj5{!-1g9KpWY`$*_M` zU>_{W)`qZ#&12T`z|r*=Zg~Em!9Qnatv{Ou7P0)^y+>Ff2%rdcLlLOpg015 z;`J$BJKX!PafYo(^9P0$+!UdTE%z6dkMN`aLN4Zw^Uv%38V$Yhrf!HN3fOM3wpe7h zP+>RxC=X!TqvDu8I9d6XAiaO$H01lW4 zus{osS`-7pbYclS(+%H1!j1+2x+P~kGC#%wT6BG$LNC$z|?kJHw3V`Vc`g+ z;Gh9%a>SFrrUV`|65c|HC5|yJ11|GFQG@#9l8_hR=x%5TP!K^65)ok|J78F|vp$G{ zVLGG$Dq4|1Z%49;Gpp6|;b5u|-6htzC3{4G7j^IbD!n}LL3@IT68pqlL8e~J^ zgh4ltQ3r>S5}C2S&`bn>OFSMriX*ujniori4ksGwgZFnpk`v%uA;0ottAs5$+k!Xr z4r-$UA;ggv-NiJc51o$|QwIFp1_K9XLV(JnCswF|t2QXr{>y`c5w?iY;?W|&$c_;07AS~yppEc3Q0HFwoQ4#p4?5Qk-;9HA z;8Kk`;9ec@6wskaAhqEV-M`E&QNP3p3N{bKJ$9GjCs6aZvEgJ2&Nqr>B=4`v`91PCU8iXG7tG6aGsFonRp!5z3BxOXcQ z>-n1#|4Ux}ir2?_HvS|L=&K*9S_uqqUp5((Kn@jr2aRse^~1MV|4SU;NZS4=mK z|GR-Y3{3)^i0w571ZXMfUwxcffVJ2`3%h>hPXHcl?_>J-F}*y%Qw7R)V8x?%D8E+( zY>rB5^MTmBKL0E=9BBOX!MKhC`N^QzQaEP_?)tzlGz&a5(D^R-5YX_tAr$~}w~!dP zgMB#wAEaWX5GW{wm0G(K%?CM7C}6q=)kC1QaC zW(EbZ9^kOdgY#?vA=3d1OMnmPiy2`fd5p2d;h&3Q*n9^AshB>V`3}HlFl(%+6#K=H zAP?*|jlmX*$Xg$}F)hC_$cBi68OMN^vmK=fgN}fM-OD2n_TC1=RUY1waD*MegjZg` z;u7zJf?mGSB85Q;iK%0Y%tbO7>jR8ZC#WI6AP^0q;5&Okq+3oqDERf<&p+5l4ler_ zqD~d>1<`B;fkg}uWr?JJ!-(SFJdz=*Ad<%$1?Q%LBZ#nKttw2&8?Xl%|A0QcDuN84 zn9<_^MD~n@Gy^!t4mJRYhE$gUI=w@r<6PXs@wnQ7`>^~RXrc2t?#t3KKjuj}jQ`4y zEr>wzA5AFpC>}|W)D0QJWjM-)^|t|0@C!vdgGvX~`_JI9C#ZNM$TZg<(*_nU%?Fkv z#a-~tdm!>Bp4nCd;I>~g)}K(Y+zH?U&KRP>MMB50v+l}f957wp?Z;~H{TtYOGq&I3 z7*xUjFtrpi{uh8|;1OyQ@@nI+$|_5Kz)<;ssZyKol{1e`yU@M%uh(miEA}ftF6c!C z1yse4;Sj@OhGUxgfH5!(g+l$r!xT4D!{+bjB4Yv4v#p*Tsh%h#1kr;v(IJpbm>vd1 z0=CTm(4Bd|o3JrSHI!!R(nXo8PClxU$`E>Hw zUVU^xD}ho-4Q4G|HIbKPZrIDsM>;OXIev)af1mU4Sq@{NsdSjpLn=& z{a=VT)v?>qn=M)jG&J|h!Qg$hzy5Z^J03P4hbgP-7Ao6x|B2Tl*#TB{ph7{Pe}@(> zGw&tiuqL<|b$2q(d_1lWlT%wdAq^tuP0IfZBMt&&Fa%n}@D0dwY_JQZ*nsJ7_dx1^ zjS+n~dQ7zA;M=Kyj=qMyyjQ=4@IQ1ZFc<%}l9+fXdL4~R0u~xM$X!gIzEs_Oyf2ak zuxz3`JxUgNolT}>{#D8Em;w{))h^@}$6wVgU)jxEdG>Vi%FR~}t~cAtY+8{&evc6x zaA={}!cqXM^%a4&uB89EP*u0^k9=lbsj(kmp8#4}HZ6^@JpbD|{}27S7xsIU0GYlV z>GN-3%xnU&QU=EHEJnYo`$h4_4U|8k2dxSb>ybtm3!!NQeg{EFy3k9;rtjW-l;g)Z zvXQt<#QnP22`LbHV2+0)M$k_M!??gCG#6@GfGfRHY=^lGg};Nb#ImW# z@M%d;(#HQrtG{Jit#}lqKY?{q@zmyI`EUB-xZ?XAdst{vDJU}1gWJ;)XQ@<4X;n|H z2l@wnkl9}e!FOleU&k`kpBBBpw-`MwhY^*TUk~_+cfEe4r+%`xc00q}xSV1`lmva} z-?5_lKnZNJsy=LC4*Vijc_-epd$&G&yx*M{-VP42Q(WH))`ou~V)*fQbjSWa|7R^f zt|E1JUJ9c3Pm8;sEWCW{KDU-*56ClS^^1(I1SLlmxi<`m<07t+J?I!+LJ= zcYlv*JTGqODiItX*e}YbRelFTUc8W7B0%8eSgT{#IN62RhYvJgZnBCOa*2B@{#IBX z0Xd{T2R!*eDeG>lKtk>kGj~8eTQ6*Zt%~(##+{T$--r*v2J3Gp;d#o6^F(QxCU@+G zL=t&VHVgP=WDh(bt2j?mv7P(xwEf}kje^^BKrn6mtU&C!3s->2aYqq5N`K->&O^kL zY0mAh;4C#jJ;W|6hrqOrr4dx32ZkBaJ}0TqJ$flnrC0bv@6 zb_*zFj+har!(8NnCD=Cj4Zh=+zsZHC=)k~HK1{dwG;V%wZ#)_wD^W?HbLrcX+NT31 zv3e8JfIfE7V2sn=fEEVUeCS>{V*Kr0)@S=)BCoI8RP&l{t!%$pOM5fGI#;!Zi=XZ}uJ62hEP_PoaSJ%Xct9EdvK0n$f{SSWgpnmtr~qN8012dV`$~Kwy*r zP^_uboQxvQp4oFD#&#?}PTK|@yYjcq6f7?)d}vT7 z0oj>`d5u6aj`W?DW-z~A(8n%rKv&`lOKs9%*T;n`AN^PQ0GK7j_2of{pLl_d0$>nPF%m+>K);>^eh4ql z0HlwD1dcFv0TM&e32+W#5F?gHBxT&hfJwl1Jeab68IS%IEvWtlJAZG$7R#aJ!f+v4 zWg=StbwaR+fPVwbMdJ_PVER(AD+(kW-3LR1RbmHRk^*RrLWR0wbEHtbE*|SXGP*8) z12v(cRw+puh{QjZXa7Xq2O0yLZ$5v&5sv0S0?u9WLSy$z=a|6ZwGynh3-%O$YzASU z*z2fY=M%7NVq+}NgY&ZdgWdc|D54o1A%X}vzOE3GL`ob5NqWFQ2Wkl%>VI;TAhqw+ z-g3Mo%D7`0Av$KqQ5P&{{?9*hk*teki(lwV#Y{#VkpDx-)#y1+lC8CVhIJz zT|$!o+amu`pZ^ENfsqfGZJ@s+`~P>~|G$UuvlUO|1opiLB0^GPeC!`~D#JTL%PDTH zY?$PkUpX+lvB@E(6w+*9<&l>N-jg81C&b5Bqy=ko)4b_Jkhcva!acXk7?$rV9JxH1b~);v6+qtIr-Zf9y~pMKfY_b1*5;;ziDrLF!)2QLV} znRIJuK`ww7)rgzQtE;atmakbv!ST0V2Tevz)sE`WBF{#z*jxRvB-}0Lv$P+V>VDa&XxaxpV|a$msSJ1)QVp(+Ib;Lh+% zKO|V_4zKfC=+4+ZC2|#bn!|!`T27Qn1)9bZkcA->m~$bhNULakM6ik~LcwX_G^xdc zg60A}9;^5u0bP{6zwv?;N)w$6X}SdRX!Cz{KuXSLZuAf^duR>$)?$m_))8yF66FnNS(JBDt*wUtP}B9!IB~<`Jl(R;sq1RB-R|+9g;cuc>qKq(Z!KyY zy?lP#DtKu0@`9ZA6ei2vJm`Yx+l(! zD43%kNEB+BYCsFn(PcBY7H{k4e#=$-%x?KDk|*CgfV(IlBCNFaYp#N?el9#ucxNe9 zn3po^bEv);+wJ4nVpNVnYpRI*Q<@az;W z;Kf}zh9?Jnp4<)eFrZES`07G_1)S^B1mc5?n8@n{i^>Six1!&4V}-1$re)5ax21Pe zG8r{UyFAoBU*coM?8(JMIeu5|YwhdF3?F`kb zln1x2Xa=278onNhuz`~3j7NM~kl`3~l9fCC(>BOrz1ONDrr)p<<#E>vHanFIkq=Ayc4sG&`pew_w%&AGRN$w*^Giq?+fM;( zpLh0SX20Fh?-LX%FX0do)zGXvP0DpsWec5^#K$Y2Lr#+%ri=%iso2xE{+4O|@h~^&(go(FT%T>xF27R?@(vSGy@RT;FeZ5g`dRyGO&Aal^6J_5MsTb)kU5M5ex3i`|M4TRj8Qf_SyvV4g zW}#KU^r`K+L2{lzIhoOG&bdZ+JByXo`Pl{eph*K&zZ$orfCWfS*_eCYD@ zR5qW-e&55?;cv!vGRAf+1o!kH2WP?&+1Sfr@Q*QZ@@jkzY-j-n{3tfCKG z(Rb^``uC(84xjCkH|V0>liJ~y*%aYqMEFMWh7?+|T3qo4ly=a3I!*ⅆ^Mbat(e; zUT6VBs#ox3!!CmN8vI_dQZcC|Gc#?aniqMsv_*{gyTIMz6cBb1-biz;{XOx%OO&|` zr;Voq&eZNH6Q5T))IP;2Av?+sN=bEoahv|t%IEVp8ocKoRYuHJJ zSefxvzNJD2B2lNhUmDoGCdp^7v{KI2Pd?P)ymYR226 zRo9^iej7kl%^i*>K6Vuj#lx4a{=?6`XnR`o+NShdg@Zyo%C~RMyxP-u`<65LS&&-m z&Cnan+gEF}JMn63?2Bw|%^zUri`z|=z3NJDZ+nDKlkTb$>T};)s9{p>EDAEdu9G3M zn-P*o^|I;xa2o4nzufDI4EE22OS65lzH1kAl$Bo@QmCQjO9}kg$T|q%@2wFd%_ZJ-5EG#pgTi!nFYDo zF6B61dHPD(_^s57q3VQD%$|cD(Y%m#Lb(@YF$NrTr5^2`P9a522CDXVb&D;kSLjq6 z4>mqk!}*5FCT?ZvRp#;KW*bS3d}wu5Bx6nc`Z4@!)Xw`OUq-@ziq?`IMMUhc(tbaAiX3xPt3+5gA zUgl%{J}Y5nm|X&*5TE&Eiz6{4R5DA>FOBA9{Ant zpk}RirxJ$JBaWug6HK>(ktmsC5(k#f|?KQ|KQzHRuaw8ep!&?K}WHa zrRGzZN=yoy%M&etKDMeMdQR;j@obFBIM2R*#L>(Y;V{N5%uDjk^qXued+p~3MeOkg zqr|#-D?M59WmC6o^1MGfMlvEU<<&l{%g1-5;@pjCvE;OoOr^f<*i?EiTPhp|e<3BS z(&v4)joy7L)NP4B-K5X`H6manjMD1$!H%z^-0Hax77br|cT7*MMk)5)dQ>4Mg^Z+* zntAlJz{Oe)Vajq<>cY({UH@@)pVH6!am{ey!9Ao`FbkTmqilRDDufYA9&y2L@9ZmX zW5VF3QV3PkOi7vGf~VQ??6rLQt0~%2EAmB{oIzhCpV6efhGoNz+5*JnE5Gux9ZQR~ zhKl&t_8%XEw>zka3ysjeG5Gv!PaUNwr5%tOJ&uf73M>@&U)-`7-gSJ;rc%9{{)H7O zbXD!85`#qWcnL)tq|zXOE1U(H6CG~f`|Z>jScnsUyDgPoID*42L($b=%zmMZ{$6d= z7x}xeBBG{6yFMZhuby0HTD%3JF+|z!J4bSqtfk9?u;NekD=g4m-Oq1fPZ_y|&#D(- zWCLi!g>Hsj$bK@Pgwmk%`z-%0XF@}$Yf-yvlfS|4Mjy@EU2SUo!v2=W-PU;n)#bc@ zp!#!}Fu=t?pK}Ri{`qmIvsvkGcO8ecl{=mL=dI>?jyczx? zB;)Jjr@lJv5}RE{W)#-t-qC48@BPW@sP$uSL0wurvge+-WcX{{5u#Dm@Bx63H=vcI7^YSaXMYctGTR+wP1ak|14W??t4KsZVjTzT2f z!X5aa#Pw`%x~?m)!~8V0@^$he)RiQI@%STx2^dD`?!Q)QkDcSoOpYf~v#4|-yA2to z%KN(Y0MFPr+i}{)n1efm2o)BTfx#P|w2j#8j zuK2*-zwo1Dcr-S^+zIO9%C-WYOeT-!i`b?jxz4NxTTd~7p1cb zjFSq+qyEKZ0mh?LNmNV77Zj>mXC?IzcEgw~k>c$4x8QS4sVGUkT2AtOy?hIsLiKbH z3L=wNnQZ0~N%^{uObs_WcWwP8@LEk?^&~WKP#JcNy1m}(t+BLtwX#C>US|@#FHLYc zAjQmmS-qi0#of~=hUkID+Z(kOM9#EVvgwBS>#Pvhg{`X9CS10^>wQHr&cjc|t`QBc zBHw6uTV49lLfx(PQM_CWdS>fwN(Wg#s@lBVW zzEVnt^w|^>B!cQ0&*gaIoTopdr>xaw0nZ9|`PGA*?Zf7P8h^E|OZZpcc!)^N$g-jW zgB%(a5!1QjVfRajcMmX+d|=s6yGv!RX;ytvk$BnjV^Pztx;JdgP%rR`wtw$s-RX=#6v`#EBu(IOE(_MDR)J~nl_!zy~#7G=lfKmbiZLa zBvAYr_ZEBC^?R0Yxj3(~ct2n}d(*c$a;!#Zj>5vOnE#^gz)ttfaC@24= z>ISbD7jXk%Fd91msvEo~zyJ`YrKi{aU(0(V^-!+a%#oej*+!E3d&G}fB8GgwxjCCdvVjjbLBY<#CV>VoOcPfNJ}|>9-S== zGT5MO*VGWR39gO%7g%3d`j5P$Kt@&1SleCOsN?cu4&dY!K;Jo+68X6B)c2gU4cJaCK6*x{=&!#^555x?S`O%h&U74E zOGBxeD}Y$XWGMjgcSb)b`*aHpoPk+4rc#C@nLU7WmapB_zTMQ0rcX56Jr)K2S9R8I zKfQC_;LYO_OY&TpCCNF9iU!XH$(O8;OCR?Q_boY}I&aW$qQT?!ySPzTTnY6h=W`va zaVpcwyAwK>GL^5TWiYP>IK14-85^aNAX!Os=%CEUGKA*>yWv*unH2^7QNstU#d2l4 zGo#?Qw##Z!-YfE`vK@tr)x>j`$F5TM^1tX>&B2WNp)4E}9oF!(WB27|APm`y??>Gp z+rJ+@%}e5Yut!c?wyXazS6soyP)xEI6UP_r1ZDG>4|&1DT{OVP`Sgr2=~dW_OJwwC zD()hw)-Tr2)>21ZLcWWgriE^w>C{im<)o4;k(?y7SJIv%`u+}+vcwg`K{--JVvUco z@&ESyGp7H&V9DDGO-jNRH@?ABj^PW2P2j5$^e&@r_j=3CmDAW8?vlY)74Ct|%Qhw~ zqL@%#tr(*x4uYm$d0HV}c~>THpqYMAUMXHuEO;0S-EC&f(_!< zGIDJ2Rw$wjKUjo^5}U|trS*gE(Y)6?+bRpuO|wIeK{*wP`Leq9tI`=Ekx7G0Ids25 zgDe^sZSHEvH)hyC(d(-$mY1hzjHwh?+w9yIAd$0eG*Kl{{OnLM{wdFNyb5q6$ z<>zq`q4^}JjDvhYzxfbsum56*@_xV$MFU}_IMO2P>&d52vv0rZS-T=x=X&xA<*aaM zn_Xp{>nm)yZ=s@rZd}Nek}dDy$7`r4hu6F7e3IxpUbEq=znRqM9@fE)+`0%ojKL#;0?Rl zcdEsKGHx@oebHMsyKm9^d}hn8&a6MTN9{Co_Du)LI&7qDHf^`e2Nw;+VVVbIpPV`p z_SC7Ft8Zm#SRCI{(^MmW9GDnW=`cSX_7V;MxSaKY-mk#jM|UlHPv*Ja#hDC&)Gj-T zhl~+9-p~gN=j8a7)TWlR9@Bdl@K6zES{MZdiH%XiJe!beK9MPEQt{Nao=s!Rj`4|u zj?Kf(Ac5uT$r<$62JU!3rrP0x^_x5l~?1x#`T^hUe- z!DpK2XUI2*b_rNFyTxQ53ISTAcU#TePcJ)JvZNrs@OMWiK^#09e=TZp3D;ncx@^(# zzEiC8l5stt^^bcN??^^$0LrM(0`UdIQVtvG4P)sNbqn(FQk9qbe8{XFMP?R5{G77) zKs|!^>3fuQ5O3hhYtHhj2{IvKXW?gF64DRMs=r3NjQOf1>Iy>sl z7QnX&3`c)HxR?wu-&)$wJ5B0M+#Y(B(2j?2Y1+zSBh4OPhA(y~FUiQ!$X53IHVeOc zDSHk@RObZ#92vt1O=Ss;ShI^;xMxpnMDMEka}o4YD>0Cs%lsZ$>j_v1$5xz09$2<& zRK*foo%$(`*wkA;&iME%b}$SCFX!D;C?mZv65=%KKf`kCw1=?Vy6l?%dcR%TxySaH zvX8g2`VhXOJd9(I>bHEzqH&)r>J;liS2^`Ot516Vnl;P1jVf+?^zQvxf#DFs48j}= zEb;9s9YVXwADvRYukRFern)g@!t&j!=%?)rv_$)p;L%8p#UaT$ou><8%sM38_|aWA zyfSsW?C%IBe;?6#~qx0+<9yJS(na!lhmVwVtmdeFEwX4sVz?fi8 zj4AnHa;8O>or6GH_EX|6D{&@>rjDjg*#b6CYI`e5$u&n5A15uNsQ6`%wGQ**|zqpk?METQ-I-E3h*we7G+A*CRBQ+jm>>p-U7LHe+2Nr>a@NqJf)l zzv1X<^~DNHPmgs)vhADtyO9SkF5Vk5Zjq3L5ijhgs!Ae=)7v7$I(s(*Ka`bCfBJCC ze&yM@YZkA+^p2VGo0P+;rV6mTD4e39!FAH4;k$)h{)P(*{ODW;U;X>={6W@!QWBOY zSmVw}Fch8rL{`fybwQwuRvLv#du&y6~r@?F$P?wM9mg6 zWh=y1>Lx53Qx97EUtVvVb3Pc=8NNK$7v?{l`snl77x(Uyo}6eL9%WjnWl((G_=@|& z)dr=&#j%9!iA_w_bp>yij=q!3j~niaB>eG*vL54o#^oS&_dQ?m-IU_wzdgBvO+04Y zL+bL=Th!I)yh51R1bT7^fEVhv-SV z4Z6L)Ks)h51LiC21GePpRy4rpYS_7m$YhDfDR zyF%_B$PI;m_R5&dMQxiJD}3|TeW)S192iEGpDp0)v2MN-6t!4-RxUN|v6fQZ#Zz=d zT4{7ocpU_=QQ1Y5sUe<%L|Uma0%v74EryHA7PyGUM>RX&FP(ZrJUnwdC}!%%#b9-V zGu(aO%-pHblY(EoUahQ#kfh-!AMV@9B15RqwUV5X1qVfBr}&LI%tSEN4pmLhcVu)*(0vPRxJ$&Sv#N7NDBCX4J6*Z6Y-wCY;g7QIpLq1Ib#EBoQ2MkSpTV>e zJ@!Cn!qlXgjENvmiqi3{;C`Tl^S0F)6@pJ{x>g1Uypbnw7&_l4eqn5Kb;N|(Y7=e~ zSvI=&h3W42PPShohQRQwtXkOuEibJICti#JSsH&#c7dr^G;$BVGZ_6r9GC7)!pbnURsLi`1l#yKpsiL~Tk#v#A!uIMyGA3QQ z&w}Jh;>9P*T*-i3b=@FSSay}cz?=Ebj1I#YH24tV$&B{k`!d6#x>_-;am?96-eMEA z*-u`Zq}N>uau0QBS?CT6uHPtI<1#c6pf-yjg{1~lwLb3}QByA;w6T6R@e@ycXHqdY z8cPH~Epj!o2qi&21G41k%r0IB{*defq5={_ir`B+fL13Yk7?!W)Ho|i9fta$>$LPL z{;a#a%KH7a_R~+(4=*FKvEIadsN2N)yOPPO+KI}o^N->E*tq;jhGkaL%Z#&-h0h_gt*E`-_B|% z`>;)4D%5$aoG{_XKHaD^O6gH2!BYbPM#gTfSRV7&H{aG0f|?%|JcLv z$lwb|_f2Vsa)W*f=r`$ugF%A(IbZZ|oq3(|L36E0(eAw1t0xCn+|q2tYU$e}HnO5` za@I!9u|^A%_Y`e5ldECp2_>;t`%|qbp|b8T;2a{S%IJk2X%jHbhNsn59Y zX4WVa?(I(~w0H(w-o1lBepz>ji7bC+%97L5*m8wePokLFAT0#Dj#9sLD-G5aSygdK z3TVG_RvT$-c6t)|WhJ*qwn)R-lRc=f<<2=*g|AA+9>%hXywQKnWcz0mcZ~2B-_}ar zYAnGce4|Qu&m~8DCglc6J5?+zlZ+UpP?i&Ht(Fsaq#&Av@_CrE{YX0Dn=ja}e|@42 zSL#bfs6RkrG9J&=uj7s4msHq%B(6V|L)o`+{Z6n8-td*MlN-VyjdCwjchq1&0fv~T$}zLf?H0-U|V@S zzhhU^@eje;6uZj7f0Mfr-aoS&^q^@atU#8#%C`6FVgKfpo%?qW+96Q_L1&5@QwA;H zeHQ)nEdzZ|qXZ^DO8nY#Vso?cZ~H-CRtzViR_=dbof=y|eeK=Agx#m-yKCO(B=#OsZoB-* zYLL9sesi~Xn)%H*^43k)fcvM#2DiWB-Sy*Xo9^5rsq*Il>k$7RMxpUL@(o;iF~GyWJ^ zSf__csLLUC`pUh2h&cFz?QBqC`@yv(D#h}hFWYZC))Hu*HQZ%VG>j~FY1(_!c-WIq zJbYw7l^?<6eEE*L;xl1FkMw+@q*yR_pz+|gvZcMxQNr8I{h!76+ zmswSvUBWnU&d5}5;nXF>LBa?)_lxfwb`OGqkt!Z?Hsa2);St~y5*(jn1IO6FK{jxX z&01WU!_ZSqDG3~9+c2;SY2KXub&ieT=p5Tj8I{J5%VgY#{Go!4bDk|31mN^5bv#e) zqlDc{C)#T#a(C4zE3_+}{rLCS$kyWAkwhMf4L9CYD?{Da@2;M*jL+wTKVogFR1)iYHxlfrtgj%a$fP%DuI!y z1Eu@Ki#!4{ug(Z@-WzXkJw+k|6Fetg*~(iHC+K*EDQ$Q-d084^&DQDnMIbv~^ChqK zGfoyd)fh@^r}SmtcV6G<)cZ%`#tXH^y_barC#PD^>cbDBvDC)8^OIi(b4VJRjEy2N znJkIR$=3ttV2nv|zWeceMzJoaDJOG*6c)@5m(7Gk zszCQKPH{ebI$E-j^ZMnSLT&u-Hf(yN5NDbb5jxAZyW)K9gWL4R3|1N>+`1R*L*InE zMZDo`Rf^L|T;-9?fA+lMc}bj7`W0(8iSd(H9yhuM)Qq^E4>O%4Gj`S5w{m;8uMK#bE` zg^u;tkj9DRQn5SoZXHJ zb{bBnsNUv$s_A4hyc5IS6I(3hbsif>>r5BzRvB!?L@=l@HS($zd>$cjf77m7x8M2O&B5BrPtwp}n$P9YTSHf>t>8S+7 zFNB>ZJ|V-$#~>Hi`b{!9iLLP2x^Rk1Xyl_y$}gJ;bp<&l7?qr!sIgvU6t6D%sF}id z#=XXkCx(t?>dq%Au8SAB43wO-I8-Wv@0&kRHW}FFf5a&5SwzmyQjyHUNkto}$ukys z&a5@DZR%zn8)2Rhmomdkw+5ml@?xpv#V|sGqzlwFO`1Hz1{k?MM492e?X^QEaEU#uNYM?EmmHtcZ`>&<3={PJq=B^Z-Sq} z5{E>po;cK!wsMxr&M0kcUm03iUO7asjf{ zF0!KGSP3H@FSDgjT9?YiTN%DCo}q12b4PH;rN4V^vkq!Q}S?k(>Bu03vhvC?VDm11_)h zTeb8x@h1y8Im@KOcOpDhX;C{vMBp(Q&l?aHPfb%}p^aw|XNoxa;SKM22G`XXFE+|M zt0YN|E|>aezS!rR3=C`tYnUc2_wt1}Iz4@&YrCCL|3}e;WMu}0w{+lgBx z>85ZZl!|Z!Q5?nGZKPdkwBvBACQSSk#{+A=DXP-zp^nLs>R*45o2r^5cIjH|GHR(P zTx5*7laE%~qhoj8?|SCv1?FXrEM2*CHA0=X66DoGh2C!@m?5Fwh@vQIyru!wLT|5k z0s?&Z*>5fk1N(Ba>;6un{0;s5XXT;JKE=5^%zeI85J}DaYX4zhD%&VGqBU3EXH*k> zy`C>c+#sLg0vn5bZ{J%tz2|QHj$6B=8vYY+)mXpytwUA7f=RI)-6!oN5fxd}LOmbe zPC&tMujiVSOc3x&FL5;+q|qK&-S~$%oj0X z)FqH;dvfEP%ybBm<|K&e=p__?QLVtw8wZ3JyRXTGmwaG*$C}>fKRllFK8`mRjRksQ zxo*Z$O#y<$a=F*wqEbES1NK!y`zLj&sq@P&LoU-OJIcwqUnprSE$Vl)Pbza%zwdHs zqyw&B|3{~`1mdhj6#?6FidW~7t@mTQ5$d?LD_Fw2NiC0WNixT~Ui2y%e$lze{V?8| zCx3gjxBEP}qQu1aB!PS$n+vO84JvA*s#u#_Fr$_!sgJSKGhkW3^UQ*K=t+h_!P!}?5d!Ity2iTMYs zFJ%e~3yeQ9RPbIRhf)as32kpVCK7z0R1!}yT-pLr1Oz-KA4}hjWb$by z+h+32OH-$78Q5joTcq112-=$U1h^ z{2#Wy1FEU!={pTbs7VMNAqk;M2?z*?N+7h*LN5Z+1q1{H1q&vjh8lX8UZsN+MGaND zG_flnMX`aPq7U+X(f9v;J$KL9+1%uw**i18yLWeXCgwkqDYgyN(a}-0what6wlG<| zUrnE4tQlnT61CxL*dqvM#Ku3SY}yrkAdsiz<9ARjg$6<>>wAEcRJh+yCA#n?;rq|LshJcY*8)mJe^% z=xsl_&r=5c!-G5g7PY>|L{j5gu5m81?hXWUo8@4kv+4Sw^kNHM8veJxSkYvIRB55D zsrxG1ldJy%xc|4BwK&6yh&$m7A(Te_3rL83dnocGWB;7{>6Jh2SaEAa$gKzeQLMju z#Zs&Wmegvi1J9bo>0`#2UaG>5oCqk3=5fiJkHGhHtAM$fgWLFicA6Du1w4s}g+H$E zS_>+?F!qQR^S_tj5twu5nTrwBKDiCQ{m$bfh4&Swcl<7-Nbr~TM7#Z!<^QcXLvn=L ztg}Rai`38oOp9A(=$~G*zYoyBKTx^I_4UX9s=~%?@A1*~51?B4n2%> zgnA6Ubv7tT5O1jGCJL@*Slj~1uuMxbBU@0_$$B^F?SNYpHcJrgh&wqqZ#fmI$7;l@Eo~ownoVhxcA?Op;DK(PPd|T&^{(VkWA3F_}Q%ZKZp8&8;IVht`;W-JT?G%$;-1~K82A<9p^%5nPMO>-;BF#6li zs=C)lRWf>qcYX-Agf6wlw4{k3rEG{vlK)>L^q>7D(^ucc)n`QtSzW4JjktOR`Io74 z_PEpu*YS%Y-8awPe&h3BD4MB89RF`v{YzD6nQTy-Yb1$JlwH$nHsn*jQ)lSXkW5nES^x{wJb2Jg3g~%xR3y z&(6EGLya%yUS2T}>8m*+F%%Xyk4p)$FNn)_u`BT?AlKGJ>Ri42>^}^dXR7@-V%LwxFe4;q>J%dFI zB+<*vw(isAXVvooQa*D?Y=F%0A1%cOsAqcxNIf6)Hv4lK`U!eAR}j}kAvTj}h-CyBqP;I61K$vTyI0{Z_U2Az;c>{1PaxKFw)e1KHGFX8NqI#I z#^9>nCYZcTcDnyvN73QA?1%Nu8s2j&x~12K9{**W#V+x$GS2(ctLH7~)#;W)V!veT z?-`Yx-xaiaG%k2m-8IWpre!vj9So!}HR=OdOlR9~0(S=f9J{RJYO=Xl^2cui$ChK4 zsR{a0L6@Ia+sK27e;8&T{!1ZDs>Ki{;;-V~KbkvfxTT=&@5LJ3^SwDw|BDSZ0{pCN zd*Hd7t^e%L*9g>2G5%DhW;~U-i@P>1sYcxD(lN70t+nPZjns*G@_#9K=pO(c{%GR( zhi~gEliwQ#l!<+R$F&B*jaH*p3IBZ?`r*y z?YXP>QdS?&Q5K8a?sxfAaT#5QbM@e{2bE{2 z--4n+1$VK|+P@5l2{HDYhe@VxJ%;KrN8;v2YRkk_Qf4EbjI@64>QR__DqGhUm$P5f z&Kl9cMdRN~^WKS>l1Sh);HZe59e5`*t+Zb!mGXIVPW9tCtAV8>o_`8bsp;@N9&>0_ zcx=P`XZ9(dM}=y3W(U8Ye>pBiT5Aza=W%tBuiE*&bM0h6Ra<}0r=2UxH}eJSte*cf zt=YvAfSUI?(>`LEn)}14{PtbxkJrYpN~*aJX><-&2K%#4`;5-ydtGq^syRe=Dx-3C8+%ikY&E<+dlsJ8qYsv;FwM3o8Di-H_8qSZv2`0hXVnusecH| z^#AL_n)wG{R0-B_PuAx3FlxoX9BTofo$Rp^N-6)lg`M^LKU>%j{MP*phbB^Grdf1| z(U6e$G-;$L`L!u)1%SWb+b>#iiYO70)e>XG0pVfNvvM;=O>=teS%I$h7rK;=Ib*mW z7-VX^)iB8{T3A~q4We>JOYIZVj)65TmN2D3o-lwzVSTJ$-+mLX!Pl+(!rZc)9woKl zudnDQD~eyN9jHco_bSf3W9@C<`4OXn>&ir#^=sle%jF~Q7VuN`r*nAf2*iA?V4(_r z;&}*;fO#t7aeBF(zTq z8=QdM@b;XexsIObw473N5wb%ro?p^pC(0PnMjz}X>Xm~H-4}*kh^u&k z0gZc^4adyy0NH3}S6+?UpF6z!KnN_w%a{e8RS=TqysH(g-x~x6>5+xo!6!aK002}g zvB6X^FB$*xx4mCCCr_3L#=~2OcJZgq3n@ba3bhGYpEp66+8G}Uum?XA6U4?d>2|hbAo7+;f zcn;Wo@%#ndj~Pp;Gr03Q`R0U@+1+T&L1w?WbVwwB(u)?$r3!CTI1S(-4f-Xny~k6Q z&JLziq{e=}`A7Lc)3H=Asfs_O#T|9sU3^o~q!w~C7V;COWvD1U<1JQoP;sp^g`UGr zAQud=D{L4>a#%mnOlcTQGS}FI-WtCm(G(mI4?h9`gKlHc;ku2Dz)3b*q3KueV!3Y} zDT71~ZVfS@V&zjzc{z6vhXHx{cVZjUC29xCMpb8mW1Dp~D=X}M3S#-XZ5 zLur_*oR1q%4|{>mZFc~MPx5pZzah)eL&(8b5;>nvh(f%M=iRY=F?O!hiLO|-TIICm zaWWlM*&rvzUXcsgYv5vOg-D{pb))5rgsN2Gm6%S8Ti*F9A%!IGRcmNiSj?!Hc0!gCkSL$_ZThn@XJIu-al% z;Uw2rx2AGUxT|B^y&d(t4N#>|V~gx=CJvtii-qRL%VfypKz^9&C#!YjQLXcq`G$imuiKo=H%_VgDY9~QP#(--4-6PCvn3=gm639xMi zj7P$_lHfSll7R5{a@*H~>VXVRdk|H?a^K;-I^b2UYdwZqY>vZnphwzTgCC zVp4gy9=-c3^ka@BIe@b*FQkWD^Dlst=D-{2)%VFIRZ}bOXOv=586)bsP*L6m^j)6D zqcvpnvO?vEQOOe`nQPv`8DpOf&&mA@ko;PK&QZ~=^LlvsGF8u>^0>xMZ?!FrFeP?` zgWqS#PD#wFpq{YTJKpPCU`Wk!Xd|f0$z}jN)UtsJOwOFKD{1MC1LK&~1UZb{AQT3b zmqQhJex2S*%DJ7|7fQLJ81az6-qc^e{wSPchQ-S4@i*4_b`zx*Z@ug+!Adc3RbD(s z5pjO61o`BnXQh8aOd-}lJ4PhnR~?fW{}|&S_&ikeQxp#zoNd#VaBcC$h=PdGr1|`g zUTAWJfj}}_RvZWZ6fuBeF|WUJK|!o|^73x*mI zMw5#3BD|GIMdbk>q~GHH>LZ!=?u|jox#)JoOt=IA!<5u7;=rIooYbYM-l9oHpRxPi zrQAC*A(&V5;RG2i-`bHRuG%vnEK5G1!F=CSoaZX;)IB-YqGWM|$d?P(8GOk|<{p8b z&kNF^2WeH;agTrHzc?P}Lr?&x3qrC%PSXvaAer#XeyU>XcZ*xG@$*AJm(!LAr|;C)b|mnbWc*w+|8~${d zWK!w_c=&+#I@s5o=8!o(p!ArX?AY?izRCIp&-3&0a;kV!JqLgsy85lX$7B#_pmX!qcy!XsdL9LnXH>Gs9YboG#fcDYEyF=c9CJ?>YF>2pRrI`R1 zQXDn3sD2I3+P_<`6n)JH*?^OrrO%v z_LJdS47To`%F|TPuO_Hl-A&15y)-kCYrYDDcaA=H@#QKo4KKUM#ED>lNgEeM(~}&I z>1^Umm9bOO2CM`WNq}$_?pwLaLa0?(5iDV_V&;SQ?|Sl1h+e90e^su3cmAI1RCjzo z7|7XZSBDOfetTSI$X4yh8jL(7#Ypv%-Cp3smu1J#21rZmk0G)f^gTrdxn%UI2*orc z#v*>O=M4IyY;Fm!q4wM(8d?Mlyls(N1(@V6 zY&T_jE_d3P$`1E_q~Ufvn0S+mu2jH7Nm20VR{GLrD8u=T^9(JeH8%H=hL!VKK+Vsp zF;D+Rz6e`qGRCL5T$nZ99XGrRu|yUNtAL!(+#>B4zqLVY`rhY6!zGh|p3Zs^ud5q7 z4r2YeT&0D8mA%g@&|Z=jq4z3E^4QpB`TB?bElR^b>M01{DZR!PZ|uwwg$$H89*+m3 zApl`c$oopg!3hl*Taf9%j8-d}MBf-}_4LxsV++$imbwM$XY8T%*p9RE9QMie55yud zmM1$N#BzV+w^VP}AJt)7m|MIF<|XI7ZktbS$&wyr`EX|I+ACgAi=A1Q0pJS2(RJ`6 zK8&(9m6LkvGLd6?XSzFHTAxZQeo^kG)|Cpf6fa^Qd+#Z|SSh5~SY|l$DImYSa@FxB zaSV8b{}KhDDIW|8ZTTgQ;?An!hAcHm9o*uZvSo`m)HD^Jg@3XXPI_MFZhR-elPHMt zW@0pQPyRR;Z}|;HQN>e;46*l)JTr_q1j^{9D3c;jMEWRw(e7)$xNOQkU8XuIlInNg9CJ4ss{V>$X1qoVK0GJqPmb6e4LW>mS z(GT+tX=Ou*b$pST0x^Aa@fWy%JXcX>3DbZq(U2&oqVaapt)28%{JDe8&wbFfemqyg zqcs_ql`h|OrZfdPWF7@r@iL9^?NQfU_I@s_&c-gHsMqy)2q*?d4rX0~-Dyxt+vj@R zMz~ZuWkZd(Q~P=Gz(`8*z3j*GSWv-7!nQ%(sBz|4!-?DE-K$$#;$zUP95&e_f*bUK z^0*Kw$2plA>)&nK<#wD16ur4I(56XRJ>bTedLKF9C7AnCc=Chm+@M@$8$Omq!x~yg zle0_!A?1aZwOOD!TDaF|LsK*UU*gyud}Hbcu_&EFy?i!VqsfOcBu_&qc=O4KK$i=e z*Yfh?LaI7b*vAJu%#vY&JmqP6IgmD?+Ti3k-h_L@dVt0`s5hqIEdobq&WQsyo^PH1;_8yfc}pIodTDp$fF zeNNUSG0j16x#)Hmu)F=x3)YpeXlm(EC9E9j)BM=~&F|`AhhvbodJw19B{+uAfyKu>Tg|dal2$DscT+K!d%Sjt{#al8OmGJ;Rgkm+_U64vqUlMIrjY!lx838NAVoCpuQ%{Z)xbV!}Wh+f@O~TM>T{e^N#y^Cbk~->)@OvM|Zf2xPYVzm?Dv=UzBC z8K?`kAZ4gaU1QkVoc%D~aS#w7`l|0Pg-)nt#wIe#k05|omQ$8#qQ*t&g zI0U_wM)NP0lwv=tGVamW_?@$!drC`qU+b!s6k`-F3R zp5hnlkv+>G(1k5C=Jz!Zrx07ykgsza7X3 zilSY+b+Drs;{3ZdVekUqm`#eE&CNc&HP;~~-^5dvql^1vEJ0BwBZNPh*N<=#wr7`t z_$3Uv=SP)SNOw4l0@8`;Jv$3U{pNzlYD3L@6?e6awod4BpUf)rs~}phhw{u@x$syu z=V*%_bu_#-8SCI2>Y+sCyPadwL@&F3GoNh1SoCu4L zF^24GREMt;?V@p)v$+QW3Fm>=#9;v}W)JX1Qy*UFo=m<;?MXaE?IGrSy)CO;naiHpEDKcd-`f83QfY< zji^?$Hj2=W`AXjxL{a1Ya~E^0wLa<^g=U~JIp;VDRMoDrIdE^?zNg#XM`}Yzy|CD_ zlke2z{-&X}FT>Uwjg_bS8c$#K;k2-^ZETgee&IsgF_9m>&p@iO?Zt7lA++*qr$8ru zOJ|)}!-c~06V3yE!kiq_(g5lhV6S>w#TptYbruttCM9P1^jQByQIKLEc72czkcq+< zw=F1Ejmdp9vle>j0;Ck8iSaoin_7A7dgb~#io)gj-DzYxZJ?bd2OmQtg}67f0&Qj} zkBdwFo*X}B+Jfsy*nkt8k|vK0tGC_4c~|nIvI1h!;A&-h82h@U;Qike(eUJ_1EJm% zzVYK$d;L3d&iOSsvJyhLxUk%eQXW|NEIVy+!tTxB(0Mt2L{mzJJSMjg&X%K7|D>P) z`U}Lpz(CKKWCl!)OT=&B;|ea%>q&XMDQ$9fPF5+IWUzGX*>v8 zju9ygdzG1E^b7=X=0$PEiYSAfqosrB9W&~8enoD z&b9TGUx_lvcwC=4)wGuw!j?#r9(nHzD)xNLZPNroT?>hiOWz^$iRXLh*U`~BhN5yr zs`bW?<0|?gp9W@D$DZhsP}>kWaddeS8Q#=v`^offd6ITjwh^6s^2{cYO%H)O_bfo> z^AJbL1+u*S<%IJ#QZg8;N~WC7m&HG|x-1ZxORTHCjkFeF|r2eXRj1ocy5{#-%Ri)GYu_a)p|8L``Ox#lCYg&@!=r zFI_boz?vzD9A|?J*&03_SGl44K3QGUBU2=2IgXSX4Xtl(7ro3wawZHa9c{x)3DGUi zzu2MSCICv8CRPs@DFB2Eo8S{UGrdRl1OY~tFM15*+f$11q{zzMw?I-V&w1+6)l-=qH}f77NrM#Ph;}9hky{lr zg6)|uv2E-9)H<{a+zy@-VZKT5EUeWeVc`O8xo6ne4C!9Z^~h`@=aFYb%rdO| zw=2c1Fl7mzR;deDnp285t%1gSToQ8dCb8?Em#oLu%5WIU9lN0}4;>1yxG%={bG1l_ zhk}=uk~xwk&O=|Qw8!A{$^i==}1*{|=oa6@@+doDJJr<8@J#@)ozK5pYu zB|1u~GE{SFQ&Xk`^Zq+i1~ZLVYC53m{%ti-JazM(pO7?AB9MIl(K8BQLa)u$yt++s zV!BUVF2yf+Imb&2lo@(FkDSZ{u7ciaaYS-Uft-gN8c}!1E=DGt{%bN+#f5i|ND@tf zY&ml7hJ&8;X?cj*5=q+3oRf`c0%17n!(TS+6IF>rir!~5Y89_EndF%|sz~u`Ug~^4 z^5|v4-g~w}cx>Cya5BI38bww!+WZ{r-Raw7NgniFTvBE{mr*>*BGwS}G5)H+N$`a~ z@BC;2v)o2PhRaWgM|Mk0;}ztLQWj2(f49mQbiDjp>O!}Bu^A9rZjru+46VNE^lRD# zH3$s-GAOtTAjc&5*os}&`XD2TkPLhKz8CyhReVVQsTm4&8w0KA7cxySImae{)@i2~ zK&HGyTBHd}-A{52!(~}Nt+1iHV)Z0^c8M4_lf%oqjNoF+-=`)aO z>PBlb7c|+4rz3N>Rx^a*KOXiu^v)FbdQjbrtu@0M{hxo zHomHjfR&6;3UKHFDK`1t?0?wqcU=)Sxg09Jy$7@gi%O?>)}uuN+(q z;CpLL-l*lZR(yj6U#PtF1;F027dfm+#(+R2ry~kiu0%I-uRQ(CdeJa)f>#o&b2g%0 zIh);z#D3njSJ*Yw0>J55n~3ok7vx6jZ2VBIh*;cfN&yoNr0DWG7mu<>Pd_Ru?BjOJ zys{`%>Pw!?8K=9G*GE5|u=qi`7g4s)#SguHT+UzE9|&LvH6NcSNP50ZgAftOaufX% zt4`6Pg43Yi2^->?6lN^EjzxH6$=88}-$B^~2QRGlSZ7e{-k7HZC5x zA#3Z6hkEUKLD<-swsiCj*FA_hl{J5Kf7?eJUm;VqjYQ_~nLD{O=ls$V;TN4{*q%VF zuh?G)Yt>ws;V2DaB2@Oj22WHENNMnpZ-%>^Z(*OI`z?PZs9KB`xL(f7Pz*1>0Jp)m zf#nBmGEOZcNw0zLnIFqj9QZNi%mzoOUzGdJz(k!(Cz8Mn8_K}=e4KPnk#epNuOu2g zANNaSrBpn3LqH*RLjc#jERh@QIk_FT#L9^lrSh@t_7M}`J;vC}a5Q6{bN<)K<9qd| zM{lH44;tn#Gqy~@#o{F*?0wU*CKN$-Y{hmtTh}A;GW91?3YHjP^tI;zW1jkAg_U@= z;!MpI#AdnXy_=59amwP-=%7`vg9h#N;p5*($D>r7&-dTO&HLkoJu+`8Yk94;+tcnJ zyV>?CH@vYYO4q8rx)HR2{eYY{-7>TM&hyMAfHcFyX!lYisz&E0u#H|A5}*qc?W;Xv zKjICoRKH}6KtXh=Nq|}I#H;!cTQP1;k`qBaAWQ_9mT-EeXrZcv*oO+*Zn7fK%+Dp| z<*5c)J4Kcgx46}x%w!-xc;|%dL^#|UTwc~KkM*gEW2aj@UxsQ5vyxqvLPuW*Xx4qECL~#k!n=B959m9#t zSRh8b;@4wyl~g%SRe++*90z-#NeLu7Iz&nHAy9m7LsVw#%aEffsfRKsz*)Smpz?{rP)h640Cp zxA|yGS;JZBq+?VrwW(`EMGYaeJoq@kXj#6>H3vT$m(gJHqz1O;u?EhS>=6Xkw`OTu z_dWGY0yu$YqoDub9R}Yj63@il)KYd`=gBR$F>P$dDc@@97w#^olR{r28BTxUoJ53> zc|bLdz7YQe*kNZ}`R$lbj3CVIcMdCeCSTc1-Wfq_-VQ0jNGrR?I;_0~K=0Q9ufW(| zUiJ3Fq~x>mmIL``wl6jExhRh(<(Lae{5Xu_F4vyTJ2qY2vUfmG^2k6eTZ@vBp8otsU)VX2IQloK)SPk`4&c*tDdphZ(5 zLmbewtL&G!F!`H@GVyl+A%>76_|Eb{lWOnH6Jl@zobcA}gs?Bj)#xjk{pD9*1O0)F zVLufiTx^o8&fwhnVO$gP#f}{J!OJR)j7fiI88{b?R%iH3xBe7}V8E+s zsT?2e*KqcJnGSlrdOLca4I5OwqBFeMbMsMX2=)E`+Ws~hbBNz(^!8D!Bta>eK33Wj zW4X6m1c-FaS(hVXK|sJ$-D7~%PAu*@DOXn(DOB+_z2`7M zJMb1sgo;r{9g&hf@=p#cr^hWz5aE{3?1KX~wW@M8lC`v@9uj`5m&=_2eJRYALcksd zlkl&25T)+}rlYotvNg;5^qj7OYS=On$Sl_?=ZG`%g#!lcJA`=>vp(0`#g~AUh9HHG z8J?Q@T?$9bCV{U!{Dr5>4?G_XR5t@C78162zF`z|kK|$E=U1!IGjDCq;9I@e^-3zA zmYaf2L;V18{VxV2GCd(}f}Q|#O&4<~gjgn2|NQa!ipu5XqSDMbutzV1hLpmR6;5f1 zQ8th1*nTdQlbP4^&9MQ;xN76gD`rH~IIid*%OMOP20dnmS>(=RsxyLXnbpdc5e2dz zN2Ya}(?Dj1#-lY(EqGR%lvC}{IHAxtiqgloFen%&FR;m5JUba;gz{ylE|6*Kbis$ssOr$0?ClCQ zR72_t>BAX^_ph>EQI)GfTk$T={%q}jvmJ`#h9eN~JEXmgLgPSQ!37atH!&WD0O)P? z*LUpCAObadfVc_2b;-49jm9Fk_!^72N`{d@1ke6C)oHomi=J6qsqvhTJHLz*Oy$bG z5-z4HQiI#bD`BB;@*C8UmI_F?(2J~}gDO`{+rz+kuOYddu4HUt-F*%MS(wd9Nez06 zC@GySo>2R&(u$!ix2#E{3r%0+ZqGzWL|m_wM~}8KcwP+1dn2o_F^jq22^J))3nDkq zzJr!K>E415xb;Bvw^ab%XGp*L;DY!zUz2}KWFT$#X{Iy-rrZ^4IM9~r{|5E&K*m^& zeI?=BP2Dp>;WQiNiGb{&JdT17S1f$8sAzy26wpXHY9pjP4ch4lhu_nz7UiaQ~l|L|VA4@r{&!PBS<|d$YLEkimqp*f7X2;a@skYb8GU;B$Onm^(p(emzrO4lu zw#GGE8N|Ld1?X>mISmSPa-(LVT!~?}JGUrK2bYG=7u&To?&g*37NppgDvbznTV|tl zJ=*U!bH6&acF!VW>i|ce8aLB!Gl=;)HFJg2`rTNfH!P*n+^3fFe7CZ2sjBs>wB~c< z4~1drB2UlCBc7hYf<4tBO6@!|UoUVdq;9>9fzhPds7SpO=S$m9a)_h)_EbdySlz8~ z8u(1A&tPWPDIvKUD5aVzuV;Px=(rC>SX(YmjH)X(siKzXdDJC!NARNnD~Q3v%>EH^ zDMtvNPB%wIw-x&9v&J$-8SE}`MUi;L*CKfygRAap4@QW?Nwks>a;;l}mF>^m z8%58th2x(mj%Oh2l*6A_$(Ts08p^TO8HJ5H-7J|gdVE&Q-OTOBYnf6_T$NhJ@y@oD zgXvrb0y9Nhm=C@tNxHN(7?3bnt7>Q9ZqDUH)$El1abY=ksr%Lzp1D(^ohBda&dDH6 zRKI~h0pPJX1k_oSLqjGupb_^))Yl56xqNLow$%8e_Z97yNJ@hEtgzx5ygGU-R+AhE zSL^-wkaLgcLyUpkNFPO=p(dP*nqqiyzwi?4^au?{i25Na@=*_+c)qzdHb3{5Mb=C~ z+2RKEg$u4tdc2%oZ!!VNgSp|G8?XGNb zdZ^U|Pbkns>tuF^cpFUlPEQnAkdW>5lqZT;|L#mzK69-{0F$LFeJ^;7j_p_FQ{WUg z&jES>03#xG)QQoxSS_$YM@q`OiPQfAl3kwei`Cc37N2*H*y*yG80V)5g3^MYvU-jf zGP8`sm|2XIzNik~FnALS+ni6j_r2|f+?pI~RaIOaoots{Qyv-=1;R#+b?9@K^EjP1 z3Lh=i5mO`aALYt{lPuS6=-H=5=Cw>$&87|c5V#e?%BZAI`h65w?f&yVjtM^LksNbuuvuy8holdfiRo6X zy$XU8oo4>yQ$57U<6;Cc3c%#Vdz(`%{Y=LD;{=o*eaa?}VZ@SVrOxlSMOk*;^FK(T zu{-N4u_e71Rg5l*?Ob5L&V7GOijFu*7QqIK*@bsPp(+ZD@@Fe&v3oBm&w1w;Pns=Sf z%#2kWw7*g`TPrB20wU=TU<{yw_>1Ev8<9~WTgCm3eC+RD?2BNUPjRIr)hNq(l}nUY zGPKP@Dh=0J0pqm6nHus*lq9Ah{gkE&#oQ^&SHyKv!rL-@L{H5^Od?FLkRnV`1nS{& zQ|#gckIf=$Dr478f4dBrdc9k&vQRL{2uQ<~}a>~-~ zP0gVzR>o|f_-B&BTp?}6-R*66LQ61Ccd66WLh=QC8@hZ$iJFIkKj$6RiKDMaIoOB{ z_Hl~?{YxRje3o8&VHZkXA}581Tz&rT02^d^{MApB=-yp9ZZ5VKRw~(-uxCS!=Ly+oj0rg5CN!{U6XYud2AOBT|CY zY<+lbjjhpAp=A&v@BH}kRcb(<{&fLE#k&ZL4JAYvNH&RxuJj?YL3V_k*c!Uy@8lH} zirj@{A~x`gm5DGUtfEWBD4ljB2Z{Bul*PAo&h^~Es;J)&ib`-Du**}z3HaT$B;h{` zb5$-{#a4Y`E_A1q3^|-zE0w#@6`uby@yAiJm5Iy6yqDAOvj*p(yp>YS$h@Die1e}B zUc?v{t~evZ(q;pXH!dO1x@Mqkr^s^c9tvq6ti-KHk9-e?D84KOq{x*#ZvJGgKk}|sN+U&m=AM# z1lirggqqKFbFyU{`L3n+gI*JJzDWO$&}Y{hbYzOqe`)F5JQOGZu!+D0p-oZ%L7+`+ zj-IQJq~Y`Pc)>V@&w5QVivI%M4r0SzF^tHa0<2JDd9uJ9XAb@-6+{DZZHFVXn+zbl zz*%<$=9n(h*r?J-ityq3aQH<*Rb_t^Q(lFdRFldUGK(BL18FY@jVpZJKhO$SDLra% zM=M|DidE;jM}l4eFL7A<$d!B^(`u}PS{Q~aJ>`g`2E^;3CrVz7X5e$mP((w5X#NT6 z%305Gf1fK|-qUqgN>E>WOrjmHJW`bqx&#Lswmq2i;5R7#7zpc z#+Tse^1Q)&L25EjMB0hL9KesU)mLZ101znU%~f=V<6Q=~JZba6^*K7@lw=Nb8f{WffR-TfPEK^!|?3axX*qJ0$XuB-9Iu5m9%_Q#45_`h2TZa!pKWNh~612@ab49lVJd; z9V+*YQ`*p|(w8?HmpQ99b>i_d9PAm+hc!G(mNFbhr7;bzrcAr5Wg7W3xk^*Hq0+V# zL?TmU9H2jXwd;w`%oVt~95kH1Px5q4&~t%WSiQTtFC4Vzwhz0mOu()BCf-~`S%S1y z3RZH|?BowlY>-nkuTFp9>f~C#@O`uN_~j|~P#K1vxx%p0$XHIfJDFFXc`%@nG5mhN z&)fK`H?5sLh55rFC<<9AF0`BtsEIx?pASgE8WzEVYa%PP6Rq8oKs^`=mG0O0iWy8Co4@ati7HvJ>#KgD`#` z)z6?5tpDOr`(V(Fv~fEc*aihWUG@QT;z+x4XZE1(-s%8)A?y`QMe~5qwJrf}8o@6@ zOT8|j0dN8UR8flzrg^0aYJgw>%%dm*0D_>5Uchr851D4l2z(^+y&WJ3$PWR2rF3A~ z^gSRGOgk8YbeT;#_<;KY+x-FxWllMO=$fzl5+JE)n=j(Lsxp$-X1_TxTMGd>445|= zY{DRSV5tMH6{q!*IMB5wU>AKg$XMoCm=ajAT`TW^_k0O3$$O%p{+kndgZnVWO#JT0 z?{`+@eM|Pe?@~K97F`|8NzcbPh@)D7s7~S6ivw7z$>D(zRd3NOOTix)6g%O_MPu!{oSn zO_WI=$yzZ_7CDFa5zowSGlP7vkw~4Lvf7gO$P|+C75-YV((rX}4aPrJUI7QbKD%La zsLC`~pqaq~=l`(64(bUW8XKG{7+zB~xk-E%C^Tych@@14^aMr<6|s zu4cny(Ag_Y>D|~j44ic#B%(qes3jfOF}&)wZxQ;rRnzbMseyJw=k4i&Vn`e2?v>E% zz`T{~kQG4_;j6(a%@CUIPljC|<1jR1^!@&3hUrc{vh8B?!Mf$L?gzEEUatW~@`&Y8 z2PKx*??;zl=1C6LboD(`mnq7xO4&9i3#DPkNP zoL^B-*BHBwCGjQmF`*R}$cNmYxDmlx@RUxEB#Y9nQ0dp#xSz!p1Xm5p!@5|jS&mtg zBv_;R6l^I;REG#^Tcf>%8h2Ao(Spue3iA#FkDE9AM@Qs;4rsH-RZVh$0c)t3Ie5G@ z!&*E|LDu?sIG(lAhQj)5gH5gtY?AJg6?+!2VdjHTI^N`Xp&iUx^0GQRZe>C#;s+FRaf~P*XSlsVD>uFQ7a)@gw!)RVKcyc;ce^F`{NmV3c#yg9 z)yZe>O}duc{;7#ky>-&W+0#%96m-W&{9nNDUT5ox=I)cMbQ`i?KmCNCC+Kfp>cnd* z?Q=>}Sed$gKrBD~n&0Zl4(>Oa_F2zviH)OWb~4otv^8g8&EHGxh>>fQXTCZkN$ox2 z0NRFh1y5-dELtEk)rqU4BZ6oWkB8uM$!4s3uJ7LIAgNv4rdf1MXc2VvbhZTNHP%#r z2Ag(GE<8(sm{pySX67U@{8$E|stf3pa}EELPuDB7+q?uiA@Wq!gPBzaSkE)b;NY2E zj(VjH#t-iKG#F8xE$vg9Un6cRs~XC*E7c4yoF?m6Kx?f_DDi@=$#D+>^#*{7K^Mb2 ziy<{0-ef_#%CfkR*LH5&;sf9>;Ix5Qg_CeGql>+n&0!eX9t&;vOvRPHQo2t&>(kAiMnc>{VJb7Ax1;%rF?Qqu1PAmuXb#Zr<| zGEWl-xpGL=io>?1qN*~EM3ar}Z^<0_+(8)b0Sw#vJy@noP@xBui?keuw9_%Uv*!ji zL)|H)sM|fa{E`;Ltiow|Aem0)U2D>!b()m}MiLP~54)vtHo9C=XS`3*qa*KV0MZ%s zj(R~FWBlTn2To2T6bn%h>#>M}WK@vW&HnK|J4VZ|Q9PtXCGNZ(u*zl)&8N>a#yLT5 zGfAeOl=Dd3OPD+?0G$LVfZkQ|_>r5hFb6y`NIBnXsc!eI5Qj`0qYtJgWZe?A;yTE~ z74nel7hO8OG76tbp6stE^5-$6Uhe7)RvnClo%!Zz&wW>l%j$p7^&L=6MBTbc2$(>C z5PA_pAP}k$I--UU5JEBZA|PErkgnK62qhE^y(j{qO0P;&=}l0iNL7lW6bm3ID&mWO zx$oY0|F`Cw$(nsi)|@$KPTTw2%LT>6-k4MpM@643P81`_Gnh_cM+hb3dQiL>34)){ zDueg0x^#L{q}+EnnbT)=TK3I|tVbkh-goFfFbP7iitUcvZu zCzfd5h3tZ#BhJ|ZBQK>`FX%$$Cm&@m3P#D;=YDb821y;5-Hulmto?CC90YDyFkfWh zLK2NYN=UcH^MYD~{2(r|ei!;nSra-o0kjESA`^HTG-Qk-K2{a#$U(x?-O*7VfXv=H zr~^_v<{KCI4dfXFLeV(Ncp*nnb}NON9s%S2Dn%0LGP%^e$|SM@81P;Y?ZQ}m;_{>PQ>&8QcQQr8qyRJKhBpFHNd_sd zU^4OEy&k$-8jMeb7hU&b4QcEhVC{D~=vUOS6^GH|erMLfH-UHh@{h=-GHpZWG1~}x zaH~s&`e~zb#mhI)GqLfRBXr|#;BmS~4zvkj&3jiy6UCrOtFc>)Y24~0gB)P*lmrnQ zW6_d;GzY(KK!yV-y7S`&79rrQo2oo)WpJiZIvv2SXgRk@yObh3RJ<6Iolm=;f} zv9a@$>w$H4Hqw(EuqT8TAk{}(q-&VbiKvEKBgaRu_zC|dX>559{yTn2rN67+0Nj8CqAbGP$W@-(wg^X?0mfd!5+(@+evf_ zmq3!amC0STL=?3H&p>+thrOYL9Dz0Fcn`D3^p4b-h!AzR*M{6asF8_iwVMHNP5ys zL0EwZZ=;JU$^n=iFIHNpD&sonY7(H^lB;dBr5$%&(S&vMzx_di5WT6l=2Pn!S!ro7 z(*kDvSlJ2Mxo)z&;;+(-v7!iAj^ITV_=at?r3TDuwO7m!1R z3xI?9$ZrS?N2sh!_Tyyk*N_9K&iI>irM#9B@t#{By3T%Jes1Kdmb-Ye3{_rQv#a9| zvX#X+6JhBCu~dJeqclCfB|XczLP^jsE{ZRP^0j-8FmN!c#K0%BIWK4 z{`9Ly=LA7ip$~ZV=f7o`}1I(Z%EElx?USbYDc*Rm zI{R)_k-`ZwJomf-$fhW2*utoL!YG7GMqy(bkR=r1J8s4-Kn&*dj8Xx+oUoi^3)NQ( z5(TfLGP+ScrV7rv$bhkEcLks4kKqDWOgKB4TVX>-A0ecjV(?!X4~JzGuy|TK5KKgl zp>RXribV-k)~Qj#_0Q9iYCHVZA_h zK!N~M=k{Wb*sK#{p3cwAb5)o_H#R1rVz#_p|CDXUaZf z*z5Q@fiZwvr?y$FMQ{|l$rDDX-$qlZ5xEPoE@fO+>!EuF2!=@a_FCbW?!1!(=b0)F zL9D)qqShVzjZwy~^1)`$mqA%)U6v%odQ<=u&iSgBSCN2jFE1*bt=qs@#taBNBQfUz zU=?a*^84I!xOO5t_gV;mGz+<31Mhb9sR+Eg%9Px=X^*6mPyq)X3K z>D-_kK!I}ZWl4FBU=9c7?EH!U;duuW857>0q|hCJs+r(1D+r|vr`!igp|lhKWiL!U~EkGufT@_s_eOC&4L%<%Bio1-QLTN2+rcV#Q0+7+UW5032D0j^jm8Gd2BoG88_{MieP zKC5PpitC$)bpTQ)qSe;6YCs4-NRvLZHnvg+fG0T{S}C#c7(b8hD#u5;dBrG-DWx98 z8zHluli41D**nb;-5$HNOM0wb%gUgxE6aTK0whiQ5&<)gN|2lTB{OV%g>jmNhjpF# zMXG=dM{pM}en&gWs327!I!MqlxX;K%eG?}FwwP>q_N(w%46D-W&#oWuG3UD{?!&{X zm)G;|%{LjiBiGNXp5~9t84^ym)MP!yFA?NiUPEOyyc~^`wkW@pJFXGM!8$3A8dlMz znh9f_40$5KF%M({d~LFW_wAep6aD#l24T_o66o{h5jhEIsa!V^wIMxg18=sY#=(bb zErgn!jJZl>{BElNje>Q5qU-aSfmi%%>#=W)&I|3TNPkZjtAPoY6 zx7g447moR;+?zCD+`{>&s86U$DrmYlE(zSX!XsU?uKfW_z^lOzI+WxS9>j4!4wo3! zVDqv*ZgS=$X-^p8s{#bzuTAS8jK5X8q+<3QrBOUcmE!o~a^$P={BmeJU@UeQWbfzC zA1Zl~IguN?r4B*hnK~a2cXN7!IP@KkY5uxUaOmrq2%MNKlY|$Z{$)kNjy{mo z-c#ZfHAN|yX88#-QKso7LSK{prsA#ba3XckYf84u1dsZZK2 zj;*^)f2rx=n!DFrq4)grs`uv$65K8y0|@BT7vXf4P_r7$#gkgyZ>{S(TZr)iMz;@; zN2@?&<8nk941IGe)YmD&H3`cHYDiEx(b2&D{U!(Qn#m`3k|fQWfi9O4Mbd(8IM~S* zVrl^DJ>g8i#pW9WVIY8kve;TT#5jrebB^X3Oe+CX5!4Nv3o8+R>~5VtLg}~-ToP4& z8V8W)#RQvtvf)T-YV`7gSM#xbG%>X5WVP$GO+CCAMyT(YY04_47O)ND7(I~8SYZs; zs@Lj6aH66FsnVmgw@G0CqStK&^P?|cb>5Q|OYOh^`4T)#3-QwOp~RDo3o`CXoq`it z=5^PQ zStE-RvdSla9I79Y@2FGbx0HvfOcux`B7oSyR%0&e z6m?AUp(({mh(!p|IkW1zv{j^YM|k4fmgHW?)WoAmVQsPd{d&T{SO@PI4FMrmO;(1( zwxxka{D)Y_Q%#6*_Dkb6x141tr%2~-pM<(|K+VnOHTY~&EjZGptjOl?ps?$PUO-;X zd(@j^yTN2Hp=YBpK%asYs~%4}!IYZfTs?#$`D=XIEN6Dz+FpPmvQl`V14zJX94pW$ zWGdIU1OO2pybCGpSGAY#?iI!eMU14av%CU8GK-LYb>BQ_LW4!;DPxde%s!*vwlp)w z@c;#2H<3gGIzpxlN2(B%@<9(*IL)jf{c(bfw2;>e@Fz_VCAcoqhBhc;mY``xD}dB+ zSgNe?W_CfFJSP)((yyTsFvyf=5V4LH3 zaH_D+?TMwdxt60|VB~4u+Dq5S0b>I{;=mtQwYq=+p!X8J7igp4+-g_atdk8_raJeT z+V|)`amc}Uc80k3DMUdJCy$nvtDJ^r=%91THbHz#6R>t()3XoZv+Dfj7Wm}$yGe5E zilGS>k}XP+ZjTmB5?F^_qR_gXZZqmP!gT1iQ&%?{2k>t_LSi|)`|E^_^GSgof0wj3 zCKFbDMs#$rh!iE*Nj_ZwnU&1Y=I4t6^&p=9|ElX5O#Bs|Q_H=HO^9A%zX<=8opjOftr%ve^_*a?qmPz(^!b(Q=;G z1a<@=kI4jw#lKc^xXY8ti?J1Z`%X#EOWD!+WfR_%c<@>cE8S3yAkD0{WdfCP^)pwg z<>YK^*EQC)r`Z~ma)M#0ET-B8G3SmgVXj>`CfcN~jJwMo4?r1?5-{_L?fS{R>g9Zb zGSrmS^T>KsOg-B`$3r_aC;#c=rR79O3N^GLyFXXr3P>N`6~Q1`^;yxa2*a;xa+>`o znxSC+uq+N!eDTznKA|(-=p`1bJ_2xq@rLP_w!O^}y66S05Q%emczxBon|;5A2l(K zbZ`&ofNNvt>ve#wF3qb|EZ6Xg7gEb~9tKx3VpM zD;ksG!zCZb7Rl}l?1)ToG{)MHVugQ!0gMz-Ng^HuZVt~0|BwRkIjh2}ztX=F8t^Yy z0L(A2h*h4*$WWcXTMx3RJUZ=W1GNQE3+uL!BY?D{uSY@nVm#Q17vMr&3V%e;b-%oRBlp{|!l|BOOVdwMl6Ri02CJ$e8YtHnRoXVvg3!vflGrf*G{5 z(q+U9c<5Io*2E{ttV;XqD@z)$_Er$ua|EbacO||KYvGI3F+O|ai~Qq@e0SXF{gL5~ zM1T_MCA`%|X`2b+WQzQys*zDf+u0Q+?$nwjPbnFRJ`gg*2-RfnGf{AJi@zW>;XGt= zR`Q=|j943XL=`7W*Le7${#?Q5;G4n<68-_er=F+i+8ViX&*< z06sW;ajalz5L6c<3NP=&;O=53$89pc1FLm?1jl+}?YW}tPF%T-bmx@2Ad6vsyYMa9 z{HV8#mhZw{l5G8$>pW`G=GpYup=6IstHG{Eh1%E(`nzMx#S++t+yx5 z*eZ&HAJc4Nyg1L@@TB-(S&D9=I$@4iHq5zubQY~&@}N~% z#p%^vLC_H!n+lt=QRH@ye-?OLFUk*Olb|p%B48$veJkUiZz`!y9+z%%_3{OeP8F(_ z$a+u$6LBxa39VvQArNWvg9j{(L)*HzN@6awI>bSt;0Y7*TGvbIX!HHhu|eaWC^}vx z{$;Y|^0)@Fc~cAi`_ldEXU^q+0v`_UU%uhbhD7#8$pX{?O7SY16Bdr41wt+3BW;&O zuJ6&d`9A4ahVs47`^&FVh2mMY0SW*!pn)yRpn$t~0PIyohP%MmE1!tt zrGVmxd2^aaX8!=@V$qJqKrx8wa>Ej@E-6%6f5{TWrexY^au4oSg)ii@DYzj&?9=kv zR{;mP3@3re09Kvv9Xj64tHz(@3gN6|bv=(uAksoOA`oi|Rg&DsmU>B3F{-KtIZDyq zsfv%-ik`3ITb*KxwHwH+2Y1BV)GZYYrOOGKD952|;cFlY=C1n6a)Q&4(Dm|%DLY*l z3=jvGw9pt*SA~r@DD{&GlRPW{0KjvP0FWTlU4Wa!XWMjuvsY_EK%SePVhY=2aK;?q zeochf2AfJj@NL}DA45TPmg<^UF4&njMuGKHDfL={*GVO(3W zK%)V>EMc>Yoxi|6o{3>H&kxYM{#BsqYaUWgc4DCwDjeAuJ?ovE8ZAtv+7-oFl{uSo zX}9tZ!_(`iI5bzIVPSU~nXLDKJL1X3aZmxGj!(`+Ba{nJ5)1-m3|sRU(Ivg48z?yz z`kO<6l)}9n0k114h$Ngqk zL1yiJE!K9+NYk}#QUx;W3SnhuG_U4 zH`NsEGf_yx0)F4@7>Ig*Wc}-Jdj33VUq|QB8tX{=Ct}s3>pCy5>|gwio=tGL(or`Gqx7OwVBu3c;SRV&drZ8rmb|h ztUh(-LAA<`2J}iveIyS@<$d0j&0rs#QG=xnmH~+UWZHgI^^5uZ*243TQSm7`_NuK$ zVzqgpnOuRacO(mrp{3^vEVRD$B7Tx&RH*UGBA2MB;TaCuXK|4!ub>T&oa>*$%pJEPWeN0+z1BC7J36OYBGT7h7j~mBQI=>1$h(11 zk<1JbT1Rn;#g;r=;8LtH3rE6LTY?B7 z&b&suUbPnv#cq`SVs0Y@Ot&+O=LENAvB3J__P5350j594GFZWn9M*FjQkY@}kIB{N z!YgyV#=Tt}KTY}eU43rdTKstAa}R$mnI|3rGJi+}=7;PV2)%o~=kmY691Ed=%6CI- zoS8dA%&1?URxp9f`gK_0c1-hd>h)9Gf5XNqk5GVK&eJf8C(xc2xeV<67tBy5FZNs( z{{`YaC=}LVS7}mAdtBLi_#aVT)WGF{-L{{W&uDa6-TRNQN0>&loqHXE2$W1|zI=J= zKSJrK-!rh_!qyB1CH0z1N~*fw5J6v$83tBja?p~Q(a_h7$1t&e^)?oz5590U+*9$iv<=19wqr6q`Y6`8 z%f${JB~$$K;ziC*gi4aqhSrrMVqUZ8i;TgUPdKNfC`pHEX{uGin_wSpr~njkk^U|J zt9fMn(zBE?vEz+KWg8Z(Gs4NSp!B&H?f6VwVD&Y8mSwZwQmmoOS7m;l&l&`v4<_AJ zv@z~W=-NdcuR;9jOK^al%D0qJ%Q@FW%hboU`yt=*KP^3vYIOZ_pgo-U^ltN){4WxL z%qm&?)2c?b&9#BsNT`&lcKmK*vWg{)FG{y`#*225kS46qzWI9Z^Uv=$u046Y==!Dh z^7Q{BxHjFs{(3iAJsUwddqw-z^^?Kp&M(Fn#ij^?m5XJjFqa26kfVFIaHk@kb$TY) ztIY2NRH&PExCGsT9ayLm&%o~w>|ZRbAEN{=e@u`fD;}{Cs~w4X4U$~<99?v zVTx3ew+CNT>pp=7azrq-N{R2JRgWou3eD5oOke+CRV##? zU=DskKXAzhU&PW`7iR`OdjZ6G%s{7ZsVVhI3X^a9M{_I%^p-pap^0NahJ!=`)llva z;75*I7Cr}Cl)E@dP~hg7Pu!Onw%nsXHw<@%(SW z18>MV@)|wnqSG3W$^7e0JV?)RzxVM~=0&->slVm!YUF%bz6-~zJK?}%vQTrTC@=Je z*v1h6#p9yJ&~DgY0p8!_?&gX_n{eAS$2vvb+N{1S+tpcrtR@0uwjp*#ob6FNI0VZY>QUYPIA_r;S`aP0Xk~oR{x{pPQ_cAj zXBT_ylRJA1PV$_%v-R4*^N#J%{?!-9UT@4F##{g#-I{@EI{DK%!VX*I#8$8O@fm#FWSS8A`xqz}0}MM6r!x0+`pB4#AQoupq@sv1kD zM58hI>1_Ug2qPj|J|In45aSAUCpYU(uGxx1q<$AGc4&WfEjGM#xAT#7T+ym2^ts<I-n&#H2BZZ9<(3JD_A!b% z5W`CBs|5jAv(o3-)gByc`UpXJ@==qfGZ^!-IKipeK5>c;&+bs3IzDIPxBcOBAWUDB zURPQx_M@v^+7yxQB#~GA;w|67hotuK6AR}hLIwrl-Kq0a%NCyPj{1!viHV)m-i$M6 z<9}`{h_jOL)c69F#CYY2Va1vi;F zY|Xke5y5=@jH}0o1Ql>)PVV`nPIe2LM?w0({sf;TU&pQ&{CDF)zUpTe> z3(QHQOEhoI1U>y8m`Zn(qBy&9s0lgojo7(s0%zD9%^9tGSN_Lgl{xd3f4Lug)-$&1 zgZV9vfji=Qk=x7KAKS#tC!U>qZEx;&9{c>9)0c$HeaBbwwk`I{u3JBmTjbNKXAx%G zZhCm;4wq&my@%mt8jl_puNmac-ip^AcU8;I-m(gGvlJ^j_w(elrLz$y7H)jGQ~t&E z^-}S=&c0|k=OnNKHmSa6{f1IlfD_Ty)42jC6uNObJvX<+WC+&@%JH5K80sNZwvoL_ z$}d-lsK8ota7`R6=fmj7sQMODtwtgpFi5wCJJ_*JYMN{yz|NK&Fs)`}oL)%7QNuKU z$#v~Fj9a_}qg9j=QSUnhIfcoMiexhIqP+ZX@J(vlXk|AGc%*J(P5YA2!_j%BTwY0O zFpPkRJ`v0Ut5{ma)hTial&jK1W49SQP#J|8?De%4vs>ylF;}9e>OsKBlAB5@-D=VCIMesrR)g$n_gI&_Iu$D{(b)1(R+GX)scf7vyC5kY;w9>d$@cPq zUo;*pfv$GXSYaOMU#$28=!u=XgW#*P^CeYeWaY0oPG8>YTJn8y+e1&{+UaH8n*#@V zTpPa*@4&lr`(*FJ?--LdFNv7N+=IGKzw$JY&i(@^{P=|Rk@f3=H~T+t?|pj1ZjqKB z^$7QxxrBHhm%sa#W$24VtI(r{D=uxY&w-_ z_a%h!46hML>%n55BLO6+)5W);;W`S1AoMG zv34ClV=O-&2;;gtSj$>9Q#O8R#h`>G0rm$dXX_lAIH7k^7AQ?tUJy3k}?@o z1+rvuX4T*!7-IU9V}_-Lhq=x5&UjT4k{1wppiWj3L#}MFNr+eL9p)#9_ZH#kP(cU9 z<z-y~`vUh44Fn=yAE+M4#EMw=y?MEBkzddB7GTiE)6{4pX`KdH8NeY{hp`HiZoUQ6C_*=8Wy zgn?3N6M~cN`%3uwF9-Gb(RX%eh6!)OB5k7jqLZS4Sq$8L+*tn2gNxIaoJLjKpByB; zc;8z}Y`}4exbl-dDfl56pJyJY!mSJFkeKu?>$;%$LK&AP8V{ePQZW~o^75M9eiU7c;s*G9iaQ5FV(YUSmrc(j8--$Wud zvR4*3SXaGs%GhDTtS3KFKE1Nd1t`|pcn(1=`Ab;r$?$OQdyP=V@3I?dL9CUm%WxW1 z4uw>}B!c*<1w~-_2b%ua9w0Kw_f7K2u25>7-IHJyV?K3LP1jI8CLE)XzENlBF4PBY+mpDpqop*1OX=Fe{3UMY>%7_5S6q|d> zxR+kcT1->Sy(p*&&J@&z5n1kqx)=c2Cspc4Ez|paQ9G;?Ogblxj0_Botw~_rUEAZx zW!^HYYhTNwMq+nUhV)6M-SY|3_UC)0hv%-Zv7w+-kty;Z6r!6Y@@NZ#z_wK(=P7j) z_%<<>su<7EDKZ|o{B5xQ9uX;%&)#>cI9AT$?ZDAC@y|-fXTZb{54xGuZ%CXe&5&VX z0lD;H8 zI&SE#Q@!O2zyb!4mf?w){AOgJB1Z7t`8_JmW{ZCdPg9)J*wT+?#ite@8)-Iw{!#ud zXI|GT&&i=|LvdSUh4>hDJ`4v8Dw%%dsnGjXkO)=x!oJ@` znMIy%c=vKDLuMtg_Vb&S)z{9bM7P{VU=8eZ68# zfreh2QPKQ&uF-l$($_4lRt6--UJVTwKCR98oBp;3*X|2t8*MYDCxyQ;PdD=Laezje zAEc!2!1#L2Vp$VyzmGpg*{?&8)>lItj7KgzlJIgK3?i~HF9+s7Al?Qlo%PW^9k@n2 zap&p8sSk0+Q$n!O`osoRW9r!JdRY;*zF1g!jng|CC8XK);Cw)j7>s?Ao&Fhyja7TZ z+_hA9Zn!7abe?uzlJ;GYa}J8~S+JA4WhGSk=-letZ_H7CT1YGDKFG@W&PnjI>-o3k z@MjUB3HIK{vU%sDCjI+jf#D8lQw|k_Sva@SgX)8FrCmh?;n8cXVE`b3u(YgDE7DEE z?>sH;iQ-EILzKAT#Z>BuJT#MnoJuRI7ZuGu)t(>yCK9x@_vsHHRRX`*!+P!f{(*9I zv-#z+w{NF4wRb!6x63!q9^UJGFF^X4A{xkVuOrrk758#XKK0{Ez=6qa$ZjB$Ws*Dc z>X#*+SC_|D^A(MOd*>(O^O&v!&_bvU9Oo{&a3_+hYdHZSM0$W~^pg!biyL_WT-D z8Hh|UgQ*g%hIeI2O=kxgl0+v>xt{405nq$C(s|rh+4jtP4Rc48GU|p}sT}FRmw{i( zx9J8Uh3q@jjBn7vZcou^+#JozHl;mv)*HvU6vWt9#xqzzG?Q3iUZc5~*85}ev0e&h zoi9mk*~cHw_zG?kh*v50Eqcd;mOD*6+Frc@-3lwJ_`Kw`H9wFv=cJE+)&njl`$vAf zgAZ+hsUp474CS*bSYY}pAN>91_gSB1{q?mp^dRPQaW_npsn*fw`XzGlVqE*}(SFJi zM{dLL!wMJXW+}gX!dX^GPJx*l1kf2QG&?JR?@Dp^AL!7JBH|0fEA3F@sOhbVoG8b+EsG@0P?hw$W=X7r zH$CqxrwJN~#raO>{sH8D%BQJzi6=k4c0TG0x61)Y_xaD$-OhV91KZ~;K+vnckdNj- zeyk+e$MO#Lg1IJgBKuJ%mK-coTas__abvi_zO0Cg-h0e9=MlEBQbZc}`gqh0zg}*+ zs!b8;3*C;RF0ul z69h#`ZS_3LiVFos#>x&nP3~=nUA0qFi;uJ_ zytU?nYtN5tkB+XxLoH?vNn}h(${aC*`l)lc?mhOTtn3;A+3EY+GyLkX;&=qflTRCe z`bW1)dOWmENks*UmM*;4{1VCy*Q=ApzXJ$#MoJF%%6AECgR8RPEOK*0IZgJi*QZ^t zzkInI6)%Q5>Tbvy#HUE$16`45S=~A{L&DQdV~3UrOoyEF@kK@-6i!-JfzfS&{az}H zhTs<0p?xR6(p~EMO>VDB>yS zKcB8H!Xu{eZei~+ahcCw6FzfS{71wZUEq%ml4w_aFI=e87k)Eg?Kl5RCg=XH{pq!u zJ1<=%7;UEv)e;|Cbem7)?EkjY2e*&DGfil)KH0~}D^%58{CYb}d-k1~-2C3Bp9>8j z6A10v11>kvI&@g3{pRDRj!tJCMCA5pHH)=DlN7R-fl0ch^cl5I+CLI>B+V$Q>DU1Q zu@hHf_x@(FczAo(jLXiyfyVcZxC*FGEu21DTF6WC6LcwoODJhmJ8SV*|Z+Y<*M_>B> zpur$jdxa1$B+H*c%W^f0U|i`_vkxnjH8ai>drvDLk-BW=R`)@J^D$>9*@+`l)Q(9f z4M=9%0lN$i!jTL5>i&kivz15al~puFYlCsePq@K1yN!AAP_Pi&ic{7_w+ki?4s#Uk z+5YnEsMJ{LG>gAQ{FWWi;H|ViSgCT$IsOkIyy@4~Z#TR}b~I`2I>nPv0B)82(v52; zYx-s0wfZpCbx)WIIXTvzpA>vMrH>Rws-*dU{Q2#MDIx{*Ed5c*eZY+g%gh33x|*uw ze%m$f_wET|69bQ#BSOV(EgK*(0Z*kiosAfzCSC8$G0bMEmDNT=XgKpP29qg+a8_=) zch5P1f^VVrnOh#H=j~4tpYR%;f=x)s{+*kO-%Z&M-hKQ1YX2WV zIo<2&_Obm#&>Wa59uOedw!(5oRBodw%kumK>PIA-2%PQ)!cI=B*FIqeDGLnu9x za7|I(5WrA%W2KM%ON(Ib#uKBv_#y^EREqR8pOlq`1L5_av=nGw@G$A;l##rg@ zu@a)2&Tup_y&raBNO!cP2X>-Ak@9KBNKW)6zadYS=)t4vr&==`!{f3R!wQLQ&~Wm8HAo9K7n>h}CC@Rtk4JPjejz z9DAlC%v=07PF*(q4Xxkt4}pa)`|6jnsDZCE~}7S*9Hf2B|N|0YuSM;szMXIO%j%qRxF|4+fc z5~Z&cOzVg_KK!Y}|NNH1y+eDQgPO1(&L4JKU#y1Z4}Un7-j#A{|8P@NLMrbc_Zj~y znkq6s#k_4_fjey(v+uyyo&60n+dHm5qs$H*pwJ zPA>tVR%KELv&->c&-%8LMD2JW?4z+i^NTn-SffP>j4hAgf;1bYIZIH@o1*H<)kxMx zVZ86JhLIaI-$p+VQFegQk$c&hq70d8W#?yVl9fh2T72&J#_R)&y$Z_puTIh1a-`kc z%<#x@EvpyiU_(p4V?Ma2V2ej;U~g$tI3H*UFvTR?OVld5P~KH8d0f(gALZu;nTFE4 zrDIKe%6%@9c_{dW(eUI4FI@tR$!(fC%zZKl4BN}wgTN2gPlW{&{h^pK#M468wJmQC z;fn8KH%lIo7Z=Od=R05%`A5RL?R~o3SZVqy>_oxhs7?OQwyda+S#Xa6)-oG6?m6#$>6}(7{@_iskNS(avbsYbp`sDJYx9Ml->Nh%b->--!rKZ)cCn2a9&wO$oDW0* zdbE3Y9MfbxC*{x-Lxhwa(th=6j45NO3Ke-f7@4k)vuQo0 zv1t204x9Skg2IccUHa4lgX**dlOrpC{ofXj8j~`=@Op0pqZ+GIOZvIH{^1oDuXev! zFtU#1NNLUQ3p^j z;#Uc}Z`RH~K9SYczJR&JBZwf&>r4(<{zlgehdKYN8t$>AcvC!oJMWdH8>AF)RD%m$ z4}cVt)QmUfIy}-1PAgE-1Eqn0UQln8B!+l|K|@~&Nz2qx&Tfu18qwExo^_DB*|`_L z{K%dUH@g1Fn8$GNuHRUE9qrv60Wu_%>`r7e7JhpL;*p!q(;|7MdW4gQ8kYy_E`MA1 z#>E98=>I0kdoOiMnP{GGtN%>J7X9GrG#fcVZVki>VV^}5u%(U{BlR21{Ec7L(0k?Q zh7600iR@+uhgVpyNvkN4yj}x^O(e%y=|l4Dzut%}SbD!zwZ;WVQYk`V{Ae$Scoto&n7 zSUiKBIGYQCwQQ)@5I^nPv{z{szO^9(8)qXqfB(Fl*WURe=Cumh3(7N?aWe5@Q(*Wj z#qN(ulSFoYDHb+Z3TqI1`1(Y2>@EPw0$g7Icq%6r^R;{XHM3KdRqtlUX%Z~ug|atQ zs>F@S#_8O4#*hZ_S7I~H^kAo@$qg1juDMr~NLbwbn@4e~td^&}IN&lzMhY|H3%mds z!M2Aiw-P@gfaF--Rzvxj0x&yj?ii|1cz~ZgX#$?Lvc60$hU=hCK5f0THhTJ{=1uXk zW}m|tI4SA=$=3RTH-{46Z+V&TiVc+mZ^UNUq;w^lSu zthZT=EE;tI6VLFaSBapRT!0TEWGlWWM=rEH#38*rh`R54r)cGC9T%O?l6oM*O|1B7 zP^X#W(fx&%-)c%--W`0`FG&)g;e^o-0@VNlVYuA;4rtEJ{DYH~%=3YAdt$7VKaAi{ zRCE3VSl;``#eQ#A?cSN!=u6ga^`iEt_HX<>Z62MYndX!=_~D+d|KF%My+Ywx!xSHd ze`ok6ONM)=Fjxo8|Kd=>mU-G#Sj?ucu%!D$Z!a#$sC-z*4IX-gBiUy1 zw0gJCmufSQ^rL?ByPtl__EN?>=|5sF{XTv(9)&;J<-{qNG>!)KrUrDUjUJfk{&QT_ zsdt_^`=I&!0ZesL$+08v3N)|jOlb{qYNg2RbTl9q)CO!i4XxgHd!boBhde?)6P1i_ z3Nj6a@k3d=JWS9&ai?mb!0)9n$UTaWWe~E*YIE3yy$(vqqnS`WxYOOw)~wPSZeKL%b6cxX6-^>@%?N{i5)k7 zxsKGQyuy3S{HVvJhF)oA2V~eWM?t@E?))E2_2UymzJCe>pAo|u7sr3CuJttSN{qz^ z4lF+2v^>)ziV`?xbTvLYBh~2O#xF|@ucjI3Z0}fSYE!)6+~0(8z9~vqnjW|O`gS3m zOHmrHm|fI@ic7vLqN;XX?8dJ@fX$pX$LZ{Dbivi_JMyV65ziNH{4y*^*A9nYVACoJ zQfw)A>I7G9fB!U}2$9HiooU#A`(>^~0XQxR6C&p6R()12>=J(aWr5J{>8aE`Ma?@=NL`%=O^<~~&(nzUt$3lF zrZI4QTSQ$gNvgXz&QbNzs{6#uEj5_EZrs4%wYKQ6R`_jxavev#k&_#sC6(9V>shg{ zv@{~f0=>IP!DMW0srv8ZX)8L&W%Rs4NDV?^)!cq|$8}h_cr_ze z_+oK3nhfBw6?~PXUd1+dCj<8QdJ~-$NP^y#XiZtZ`HH#KBIJ9D)v1{|%eBd!B5f12 zElriUFpMt#7E!fEm!}qsWkC~<*8jG=zb!7d6QvD0`>fy33g4jLsJB_SJ`?^?SK{F4 zaq+x676WCTW3GS!fa#UU!V*eDRO9=af#P$nG~QF0dH**j&Ca?`pPg+dQ$259pZ_NR z=Rc4kc1dhO%6|ZD55AScda{sx6@q?C?fLhR8GN9QbC(Ott29x@*RH$u zq4fE~^eoDaI|nzlkJpA(eA!n&b!Y8YQl^i0 z95CYmjvo0CIr2&-=njN=!(o(AclFr$5P`T(;}p~xgq!x_nbz}O=lYlfqZQy3IeJpM zIOS-rsH5u#l-sSHCh?r#8`7Spd66`rf-P(9*Q;WP!|QL%9GZ)ppI!L_h|Yg!E_R6~ zDRm>+zyeTIYzXQO>o+u-N$T|9IY2X)*-3@?+cQ?LHj>mKPOUay7Vx9h=RgD+%XMi8 z97)JUHn*MS3{^-WUVkNEd~q8BABjxP5Rmp(s!Uioxs?C@04r?(%@dH&=c z$?CzljZSGTdDWg@(&pQ~aV-b}QLb%YIrj0?^}-5elG6erk>;Ed9|vZeq~ydKds*u7 z^!Av>23+4z4Y(IIn&9Dsf6pFrc)}?s%WugH3BX8orF#Aj)Ha`dgYb+!|HS8PjsdA* zqo!Jmzr&bXY_%b=r55RPU((bL-zK~A-$C6tQGUdxz(}+2ay55Jh=DRX)@mhOq3wSj za=wZQ>MLU2%Ch)&#{WF-g9eKdgP-GPYSwDj3U2m6zx_L@T&LXY(|l9O@s-TnC)zgo%06(60ciJ25_ewN-S&$AYL>qf zhx3qAfuz3Ih+G{$2wQb=Pm32Y~FVx)+NfJhVtrcr^Hv6_?0*Y5j z_KHpI=1{E4Hw}4~Zn7gvCs?ug_DPaP8jM8kwg;3QS+`2hsXSaHT9x)C+I7af>gti( z3*c}nZC*>jLfMVi`-tfS!zBN@m#67_!= zdke5Qf^2Pg(BLw-B*5VA8l1u1o#4S;gZtp_?oMzgxI+R2cL|dXEW>eD5-{mJFssbU=zGp` z%j5gE1r**g!GY)LT+=C60a8_-a4XGX7NfZtmy+4*hm0ChcBK*G0yd+~NxLvlQkD^` z@>zQscZZlL{ty0UCnQY_xwk;ZBZdtw}RRbUo8Z z1&vnOUOkN_t0VQ+LYd&(M=>2r!9$nwmf;bPMPp@@3S##S42&tJLKdGx4u+E&X8S4o zoD&bRo!8K~+<$}j#Y@ak4Nayq1b$>Jc-0h1hIk^AG%`vw>hRy7kkp~(OCawMqf3qt zFunCne{UExVg>=r_jzS*A(60zifuC>TlY?VWgJj|`G$ABWsjL#30~xqUMY7=)Qf50 zFjV-;JoB2X2s+EkomLmzJs6tx%b77}&$Y}ap7WWB#Yeldk6T_}&wG5YjeGAJwgXly zL{FLuJy+j#v)IuMiET3t9l~83C*3tn1guz!9%J}D*%IGj0by<|z^Og`866SxK7^Ej zuXEN5;k>kFWK;B}c1Q}2;>?Ttyocmk1t~n`#f(qoyFyVPL3d8^FNO~%IcVi{3qfe# z|BQb61&7bVVG{lZ6^abXl^!S(rQ}>7=|hoI=_h`JTCgd;E$#$Hx9qmmUzFhqf-&Du zJTz)&h!_#Zd^ogFaK+jaFF8KH`b1W{Ge@gugpkBC+>bZgn3OSrOV9l?|HTqzlapy9+HfmmB&&gg`yuD?M(cWmtIE>~)ZL#AoyjmR}f(iWdH zS+~qltWU&xa`~8L4417o&&Dn-J>n$Su&G>F_qw_14MIlt;rpT)L^^o|P!N(;&9SBO z<6KzE2AY3^f)DSJhHR)4r8kB*fNVxMAk%C8OwJ>S;7vJ@(G^>C`o`J+!^O=2)vpvB zmG-ia=2N>^12Yrr(OKR}j`Ez`V6E)zQA?SZDb)J=Y-0Q{A(d?wW8z=iulrA*5;eD$ zNHjAl-gnU$xG9lwh2mUP_Fo3AMkU_bIrl@?F7=nM=RVPIhQGUURS$G+@9b@&Bu{}# zR2W}-E{wp7r!7FwNEQz&11Q}qOYku#d?}N-iCdLI%!goBG=r?W)%|QAhHyzq&KOoMT8GqkIEP@8y>yU^d+ z{h&BB+WF8sgw{_vQS#&35eI}kp@y=_F2;O%cS%O;s+$S36+fZrL=xn zVrw_gJ%07F=!=a1`f%FWZ~G$UUFd$i8WAkTa-J+Dop2~T8Yfy|;5&Uk4KWGwr@%2F zLd@iU6UvHrtxTJ65I@4m+IaK@+%G|3Hg-{9h6$>`@TdmMYTI$KVL`qVd9{!F~C8LJPlv!mwG%X{Aqs~iG=V0BL*ekUe^^_pXk$Awru}9IgCZ$QhZ;$ecIx!%2!m`1e1GMd6CVm(` zQbsWvnipbkD0IzU366v_nD7%0#%C-4$S|6t5FA__?pLy8nWhOuU(Nc&m_k?-t*n|m z)|s3mAlxV_C-rFV5l)Ls>5G+3%+MaC>!qm?f@l}8F)gZ{?MM?Y_X;YaZFKcsTd+Dk zB4R31Svhvufwn|4owfyf`oNX!l0}`9QJ$uDbeJSjf~N6>dcXlUS%e40=)VIm(QnUD z^IG>LcF2v&nEvX&LQYpQZa3{0`Z}(fWRHh0a*pFAqXz#J)&MZ!%tR`a4Uys(Yg?iq zSfar9#T6{$#>xmvTYKw|4jY|1PK*KYH8e2S&3fKg{R)7D-p2v7I=$+lD;n-bT`hkA z8Z>GoMsC?HN@m3AFVz~VZUEGVOeM?v*vd&?ywRp~Nv9^3fr&~%`&1Tmvh+SmMK#XA z^VOLrrM;0)LPG+xS`&whoi%50;!mQZr1Ekpazu+OKtn9EoLC~FwVGfEwDEd2aPAep z!j{FYq;8k~?KwOeb8Tg{NxoGu3qeB3IF9#nIq(jaWuLfO;bEP(!n-CasCxQ2CP!Qn z-YnEulsWm{a{9?@ke5DFTZJMpTf|z#$UXcD({4gb=|*}K8als*231mS7cK~KCtn|p zPJ6;-LB@c5ZjqKtH(NMCe|u5zF*mI z5Drn&@ZiR~+RMBu^OU;-yeVml26qg|vc2_Sv3pNq@2|AYBswNH4J z(lWeuz;;&k?!~Ls2w>p?mI@HD3RF5JeMo?_vtKk)k|$342$I>9)a-Jv5gSsBUtVU0 z$et}jxHktrJWaK7Ry(^1#+bR_c{-ve`>hnG&Q;tOdq75yxe*8<#_o1BDMR`4)}5!o z!!T`94K5f2s>VqIA!~Y3Bvn~p3~@PYz9k5^-p=j)jOsK8+G9|NM0nf$`O2C9pp=y5 z_FEW-IApcx-jT<$+VYlml)vvY0geNTC?SKDfJ;!l&?Fb`FJE7cAR-&HFiDdno}0evT6x|T))uhdkMrQ;X+=o%0O zuk~tDN90YBUn<@#(97nw;M1pB-L!B0#oYRf5wfHTLS~;7YrUFk79X;Bbta|&K&{Aa z`8O!ysR20NE@sOi6i4MVOKqi$mA;7J!-+tvA@R5A790 zEZ#zo>OwMZw#G9DgS0GE+qJ@)(|1cC3m-!{zr(uqJ!+awoWk(W)U>*4+m2+Jzr(%=dBxQtF(#I%7n!pjnvpVFS7F8{5r zg)GM?Aln=jK1k^ch`&!kl7jPMz@^7zk0iMntXAx^yw!EO)k1kMuK1)I*m~R<%0OTN zAFSG&v~Pr~2v42uC;+gQ?3p5QZ6eFbBe9aPDhpJsMW1rO2;`=u4H?Uywjq*2(v3}_d^+h@z zqp!Py>3AZMfj>UIYA?=nVsR9N7j8q)7B@8wwW!WS+^{}qy%NDyX_nL2FJ|I0Pi3^EE-A7eYRz1uPiB4jPv{#ROkd)NBlTMk%VAQ3c1DlvwD;gOTr#}ZUlFRT%K1- zy-hR~)#L|rlaWT8r$iKID$M&|l4!}-&n8L|rphB+!cpnKBUgTC6?s?Wx$;3!{us!X z4Kl6;yqYWJK%-GG{wD-9bu=_fr4xPf++I`p_0i8F3gupE$7KU}1`HEuruyqzZV69c z06B3E@A|}o))9yX?#XZ1g^!C?&y*sB=Z#i=vBiV2OHLbd3kP*g);=r?>gsHj7dYWj zz8gG|l`wUoL>5QY4rV0C;k%Y1;%tPRVV?zt>>;=z%TBIOR!<}hx7PM3VeYGPsVD#hk^h(WvBbl}T&h6U5n@RW) z4muj*AZHM%)skT_fL^yci@hZJOg)S=vak99zs`1 zV1T4saO7E}fTYBEEdy@dsaZpch@G~uQ_46K za*!FNpH7Pd^ow~<&%L2>;L2{f z6+nO|J-D-4&y3DCj6MC3{2L@cUn23%IQ_l*4HFPX_iCK!c%yZEK>op1OjJ#!PCrIiB?<{~;yGhOox!jaZz6 zlxnJG2qb|AoTi{qhd=#MT_Jop1_`9b^@VmQe~?&68-pfJz`(Yxa5xRGT{?Iu&9TAC z+`*elo~EvZVBi|@Mu|x}qEQ+SNp<5%c6Bff)l~98G?c`W4g&&*gTz09zMK>)0W$%< zxCAtnH+T>(ITbA(@F1`Sp~+c;4dLX@bal*760mU{G$hQ67SQT#Ug_O`MUQt z^xOBV?BAgDloEUIfl^GUZV+>S`p(qFq4$@bpOO_Q?rK|-T;y642-KIRy%QXBn;&z}m&RN|CdAWrG@f(+13`qGg)W(1hMTe8NsLBTc=L|1&lE zGCduJOZW?S;5IaT-@)o5P8*FcE5=n(4>n4}n191K=_?|srQaE=G`B3N z;0hj;oEDmy7i^WpKy3ugB(YIOy?w<1p4w%UDx)MUQ#t}6aAPf(O`%R~gCtV}iId^f z{pL3t2XqL?@MR*Lu2a*sry_qcT4(9bvNR1~OskZ%rR*RN)(x^v;;Q;Qsm!&08lDjf z0$q#8UqUgu!u#lntQ$F5jt}Y>WaYGBv*{KzC<@JTM{RTyW?SuxlHr7&9>4adz)#(1 zrYL`d>|!S6s4$bQ@ipXI1>%cw&*@$-X|=P(O(V**6~wwev*u;NeH&%Mo7LE2rJR8l znv5Tfh>o}(Cq9PFC?`SBwHsGf^(C*}$ec)*CUOFg`e^R3PWpfl_dT#8b|Xr16A&p! z;n`CU7akDV-Es~ficpU-g?Ed2y6>l9A>CiuYiXCuo@Z7}Xp~a+f!a%jo=b%ubx!HZieni$Y{ z)F#E=hbIs?s`W1C58ox?rEGjA=m$KbvEP0p`k_7waMG z6TC`Uxfr_LX0~btQVcaS4_6^CVjPrlW<|Nx9|P*KE>Cl=Q5!TycJN(O7vUu=U-C{p zh7}8qS zCmLq&0JdLgq+I*g6Gd$JflT^Wr?0!ayjYHA7OcF+>N&B%7HDx#c&j=-uni4~I*Kp^ zBbL95kf2cDfJfye_xE8;QD;Nbx8b!iiQ56lf1<)qi!6~_p=`&21Q-o)7(6PX!Y>wD z&16(L%VVl($W4R|lbR1xZ7=i?4@lmvHw_4!M_(+U;t*LHj-H&<>6G-n(@#7DdNAT&W~xrGb05; zj!G~zMCpjlBJmC5KyV)!)fgm%9VK}ue<6 z_|0R;VASBo&+#PeX&gIxOQ`bH#fD$q(;ocHT|)L@XLcZ}so z7d!pxPgBbBMA{lOrXvrV5Oo=8WLRj zr}$A17)XcDZ;~@Y1cmPll#qm@^5xeyzWKkXirov)pwyH}3u_pu zG}mlmTunPw@S=SV&X1P6FM>=!CHg;`N|N{*=A!x9eU6J|Y25NV_zf!ARFCp_T7ZY~ z*--dZ6w{}T>O9a_km{`~)%zkvf4AW@X1?U9(BMWy#dVa(M)ij+X3J?j`IKf&Vp0~+ zQ8Xj%-aOC!a*PWS8gJbcz73aDAW9w#jfF{r{<2v-RlaeQ{4N-6FT0&>IYOAEteq#@ z8KvbFn&gc-B^zUy&Z6WYyaeNzR`pcfcCtc&g#=Q{W^-4LgOu=`5=uD;$L&%`6k&@d zK6g|!G?9d8h9>f?ShvaF4Q@k1FjO4@e4WJUHDcEE4w*bYsd>P+6`m_597a}SyaZwF zZPEosPS7|Bl^T&+e4cP>T0fK^gC{BF92}wkclA-(l(PB`?}KSt;rU~SEcq-aLgP+6 zt7iS!xYsm>6F&R-2%1T3Ghg>#$5jzdeDGs>-}nBm?yxLWo|+a?kD9`@&R|%_BK^s2 z_wDQPp)O9|l^}xu2jHi;C>dPBLBnOQ(A;~kL|^>q`L^{Nlqywp@!^c(!=;t6tdr>I zKKFQkH81(X9n{`boC@W|@)PVMI;MD;)_dTt3IMXG&t~mvVdW0HRAx}axrqhmO8wSB4OiJs8eUFm{ z#CM30Pma2_OanC85&XxS=x)OmM*!MwqJ+Lv(0xkv1)u?k+{d z@Gv09c{m746@i{hm%_!6#Hj$!O1)L_K?b%WpU?(JC-V{?t@R!kP01Qw5tXGJLp~C7 zLR_vH9>vVYMotD~qbHaIuc)S!k`}8KNr?kWi?8p*TEHR5&g^LUbf~z* z#7W{Q@*LrGQrc|u8m^{woCVNa3(wD1skzN4sQ3rwjC(z+Ym_u^WgziGxO+7tGMWe> zv?}|-FF@>0TI=6>7tVtjU2z>XI@-D_6yq$>!@kKw&J`wUquE4w`ZZ=zM(IVdn57o(!}p2458tJaoSzORKSXr=t*;-=bCEwY2}ue=6Mg}wG`SX4rWa6 z{pGVn9D3t1K=Qv;BfkAw(FmAZt&T*Jc#Xql2_Yj`O#kwl`8 zj&GUHPJ%~Beyys@ClLuY0B8N*P`N&isDL0`_`da2Zff@UwSoPHGzX9z-Ak|^7qiPf zSNJ9vv`3#^zlbFl&p0+qBX6h954{3`VwBWSHn_{iD{(;ZQOsymW}MIGl1vL<5ePwe z_d!TKA|)0mltcM&bL3gk`~3SLguxgdj$Q=rHdIvSOt1xpS{+ym&o8E!jpK_|uUa_o zTqvk$$E%0z7(j%wtsoQ~cBsHe16h=1^GMO1>x|TBo>z&|oJv>u87x5scB>a7Lzm#f zqRyvczQ(ROb*>>C&Y6+q{w%AY!xvsHNU*x2y}F~te`<`=$i|%6X7Ij);!%Jq)0`+* zt;RtZ-sJH)2WSSszbdYy!k0`V8%YL_a$j)8B0!VHD-5(GXH)(%novz|B1>)>>DWBh z0S~1E97&V6xW7j%vhvf(QIZxXzV{>F6|A4s6$E-9L*x6v?c`ZslP6kE^Ba^v38D`U z){}vt6#mGJE{WeCi~BqR0y!<7+v%y6z=CaQIW9poCFoz})BKPj_qIS?qdLRJ%xu?k zM6mw}Prwp;46ee&09sLGYjJqg_q~{&fCm%C#9P`ls;Z=Y!$tmi%x_WwI6KgPZb+Ek zE=UMO;i*rMMn3^>cTEZYtH6_JK3&9^wv?mC*;nq|s1nF~`Ra)2KMRQTy+0l<#dYkj z(r2dWt8Xu2Vi-z7<2GeBxM-g@!No6LjCQZ2*6DY45PCicJl2E7Yf!iLQ-k~%AF?GD#n z6zyB{q#(XkxFwkXu@^{Tq+DiSGk9FYFp3+4^R`53N=b6PAo|BBsJDKfotJ6WL|l%x zi3=$T!&v@p^C+LHu+Uf&K7R6;U#KfYDVSM>K%nzfPD#n&hZ$;!2vw>{QnDpcrTH2B zM`S(gd@vYVX`Vzd8)!5>%pLorC)X3q5v|d=A{%;iAa{XS&K*W95Jne_X%MP7;L#sB z0Znm2EH>bxjc$j_B;yz|#gQVy&2*T)V30l(%P2mQpTv(iB4Q0p?r=yv@ zE=D=?x)$|gjlQ9;=5v1J>9$B;{MGQLY3?tXJI zJfZmWmZ$WKAGXXY;Pv3PoP#!h`l}^#^E^DM=b=Mwq?iqQ@m*@48%-v6p~HvcrZ1x(C_Et6N zSUs*#SMB*3IqBkvnRt#G6}b?O?k!SO5Z~r%Y^WML{H`q$KI6zV)L6DAyXT2%TnWgE z-HSGjI9e((BSG?YH6}*<6BHRvfyi*2I#m)&yv|WsL8K5WHx$+3b8Q_t4%4JH98_3UZG2dEh8tJX5_gK z^GZX?a8|7c!eY&Y+!dvG97JR_z;cCr(#QT!6ke9t&JZ^+9L6jR&O(08^eHIkzd{65 z4^cDQ|F8Uz->=-2pDQo!0n8I(+IqAEsMvG(&?;bzHM@xnoeGXlpm0o*dgw>Y;-=hzCw|Or}tK$wk^@+1)c!E{wNeBgNd+oP*;d%I- zq|b0k7uU6b^j2g3e}qkDD0`0K&eBu`x!}4W55$;A$@+44YnCajg`W$;rnJ=dW4B+1 zxtzjtB9}1o_CJ+NL2=9FpOO(Gx)Jy`DpXBtF6$pd&63HFWUkpdg=kh=S>TPNc?L#2rziwXqsdPM+XfZN+T9j2OBCLJKr~ge!XcN zoaT!^a+a)`Q5>zaMR}b~#MNSuei%mC@Sx!%az>%ziL>cp;D;h&3?Kw*(TkXQ$#6h% z*(lQ-aB6>Y8PDJmjY@Q*GBI$ZtLHBeVY@t8$hDYdi$mt2))m6@kBKcB!wy2tb+7l1 z_ZzGHgqYqBi@hp$ny&aEB>Z0Z&nIZqBV8x_?We2w47biay$)}~)oY){?9!}`=y;Na z=n5M8T9ZdfJYtgJjnYQG{o}uaRKnHp=Fy)KWc$X^5oc4%!}Y`bFa^DZb7bM3`Try8 z`TcN3&}W)FJY7!OKJQV#{k|)co7eujZRPzCVht|LCa6 zU8i46T^}ku!FCPbCzqwus0|tEXt;kJsN2uH^QX%1OW#kmz0pn6JSmvZn8AQRnspB# zGWn+tmDqce&s7G+ub)E5<)x8&5jYuUy_Nkc`T+53kCkrwE#C8-IG)+g(Gs$UPXvQ1 zV@(nNr@r8g+CB{Cd<$FmGE}&OgCxuU5I!9xG7>A-oOKqNNN*T2A}gaDP`; z@j|RQ`wb);hwZ+4K8k@4w)RnRt{<^&LSs{<#rXA^JgkRCA>y+`m938MuD+#^@JZXR z+O}UcUo6d=3s$<*5wvvwIa_0pH_4J$tY zRI4Q6X`z})DhL&XSS5M*uGrscUb!o^H{l^o=kR=r5x4R~8G(3KZFfnonS*sC$7p9+ z*TE(uMY!868wRw!TCCjH+PoYIWCaX3sAy3|ooN%C_3q7N#hbL;O{4H|cPtDT6-u&q zD#C{cjG`VL2*>1g(&d_eqnWh0xP^ zSj_bALO^MDb^12AKnIx{L+$7pA#kU3?pNNq;UT&7c={WZx41p}0fYRJPgt9J`Muyp z&-?3NbA0z?^~fqVcTD$V8CMpcq=4B5jt!58of|5Opi+mc=KLUOPk08~Pa$H1B0dc7 zs}}NK5krY!IuHkWE@odsGAT@qii0IwV@b?Dm?u%Dp0qDUbw91zjChNJkInfc14hbc zZP;w4B(`7m)B`=ie0q$ZuTv#YBON)>=i(FP-?}9i!Idptz?+MwZ8ERw zL{J@IuTP`UL|TrFDpXFBfWN|lQ}EQ@RSxOpZ(6IpjQkC9TiOf(O}0tWmCKw-y#z(E z!EwlsV)W{o&gn@ARZ~Y+<>wB`rIzB#$v8Ponoo?**Gt<*4J(Dw1at5xFP9JQ>5WLF zI5KY9W2`Dy$&o0n09|h;oU0*%YV{^89pa=yt#KXlxmXa?3OnfGo+i}>{}Z?mCqS6t zE(jr^yd#{6zVkPzHy|pk#T6GT2|=w0^d?nfZz*C($x?sGQgM`$pq18WN?v>Gaxaot zaS~4W2z6;4^PUA|U_*Imd3}Kzvv3ka{ixvS^aGF*t&ft7YGQ>|HluV}{jloGqOYBB z+cSe`69RK)bm{`S@3WJLYgjT}f24`5(z@}_bP}0m(@Xdtv;QRT!h;5l}9X-qc?}q%OS;p^=Uw z{Qe-B!!mj|8#I&aR0u@pAyQv_b1dNZQI7G5Uo@@JgO~a>4HGjf!gmS8gkIo{^2n}JF;~;jh<=j%%dq* z$=-~sb$b1~sEwTX;5(Xfn=cIFX{IgADIsa|Y`F-Y9;3%Aqrsaclxxv3Pl7k%GoWM~ z>q$=96=x?wkN3z z-7EY56&Y^?_jYr}!Gys_dLeJSOZWeUVCs@D4WT!GQ5hRw?2bCJYeVze?(Y=q|NKRU zWYLo&lM07GUniuMGs^#s6Z*2&Yr;K%!*FE5vS66U^u<)@aMgfHsy|GxLswOZsT^3D$c9=&NGlFoyD0bV zJJ(>gx+qb4EGZ31`g*K{#jKaU&*-rP5yLpglbogSB|v)dvQG3^sv1=0TgC{l)D6nu zN1SQl&(CXzwVT-B@KCrsMy{4>ej?*%C+is;F`qFU#G=4f66&FOijJ?jX5OaRd_xQm zF~T$vYGWlqL}C=Tve7_>*0~zAi68Ew5UWkR#~atFjUfhGOPkqe?Kg=oDSPDRNx?vMH@6XQz}8p4)W{aH*$hpG7XCzA*j!M7PS%-jmD&mC(fSR3+G$WzW-bTHtJ z@YP0UA&AwcynvO#`v&=Foz~>hRO%o2>MdPp&E_zXMwcN&nQ)~m%jxc{2kYJuLsq$l zL4B;0lkO|aiBWRM$PMK(hK@STl3FYXTq~z;u$USzSX9vYjty>djXIN13CN*sZ%*84 z#^kev(ildhw?CQ8-yzQXzKkSK^Ln*WtB#fz5CJjGc=$s|reoiX2H^(<`VknIGe@cY zU1IT$Q;pn(^|?ZxA~Zgw((!1RviO#BiPMKsO*dMcA+#bt+E*^oj!bXd z#lGd3N7u#AWWI>%aN6qeI$PV_NfK{-s~NL7D!$!A!YTfoc*tW>kDay zHGcWEf<7tDXJ&aX)*txm?)qXF5uLG9{4LS(#5V3;!mECbBHV>VizVa@rK)sH+M)Dk{I+gz1w zVcuGuIwc;KTx!mc3d0*_ojPvhOWPzm*^E}=li(!!XO)8nZRW4}6Xs}=v)qO4b-LZd zA{WVB^ie=6U(_FjMT9&)+kH60!_B%7RYfhlm4Aqz=TG|DaCQA(QPz08MqN7%voHCS z>LGm68}JvO5DxBEeEtUu3J#ahHlIZ1a+=7?|9@lmXP0=jBc;C8Ec{XV-idwx3;0pr z?pECImWzj0eS4Kp88W(c{<9P}LMQL)TaNZHFN{^!y9?Vq#q5p)mll83R^RDr{%EK? zlYG002O;eRV)Uq2J0eU9cQl0$r`;$YDCB;FGHG`UfrR7FQ0R4_N$(TiaW$mF>P~u9 z{)a|n=!_$_^9?lV&lQQWr5T8<@Z$}CrclN<1$ax>JC?ommkAPBR>5}7u^I>xx_IyX zpo|sP|1~Ffwbw!ytMH$!TD_g?6o=9V%O@&!$6umA6fbsz_RT7(qLLdP>4@VAJcG7y z@8wL*S9|oSSlDAMWk7hQAI8BJx$z50OWc9v!XVOC`jy?ia$XdUXv6J!N7u7tbfR4? zv08Lx7!PtI*Ip8tP7p7liy2;4fv_uLS*wn|C1F-AZiqg(QM{+}nZSGM>folxD`wXZ zarp*r$Kf#_@{i5+GT%77ts5~k5q4dO_cv?Lt3ZiiBvT@ms3Y#XaVY?)wZ*Z_?xA|F zJl9b@u1`Uv)d2}y4cPXtaH6z!dA5uvKpgc(O8-6Av3n{b%2^-HCxzz7{`^M_62x-mi{TqB0)<@*VAc3pM#kT#;V8Ng-f1dHOE2P zix0Kcztl2$6w;AO3keZiXPV4`@T^C9Rcz(ie`;2}&*LhWu-Q+y(#wK#6V(0h#?<-w ztrP&Q*aN++5MX(5de!wF_Tzs?j>XNJ+E&pjjr8VnR-1hZAdu5dtmYs{d*Y!s`Ij#B zZG-H0qr~GbiX|zIQ2k#kVomm#o;x)s{n{&mwGF>?jc*%zz8iZVcV8^wPSGAz-3QiB z>5qLFo_qlJcl{VNCixADrb*Oe(R60|gq;K6iKQ(nCws0esdDfD+FubxTZr}djoD9% z74II>^2+8$FXXHu5G;$0+N;yiY|Truz4se}f0ln73SCq-3Rh3I95jae3P7rm7KX+) zx2r~pglu0V3cyBu8-vW~@3w7STONx15&pAgAiV?U-xj(S-nE&~izi=_(oVu#Eb=85 z{1)LgkNuX$w)fz?CrU@y#WCQ&&r$FKusju~?7jHg4r+t|Z=cqEOG-)%oC$zZcmsnE z0)>#!bBUr3y}z<@HFq0i#RDY5ED-C9x7`RwX&7#v$}ck#L1teJ{$8oIdqdK~JE4BE zK0V+|7S#6VNIbw9hH~R6TQ%&;VDtX9RqbU)OWQpc2SW)p@tmXJk@F~M?fp82_Ad4B z32DNoFW|Jy{DOf}UO&n1*ya2UlBUP;IiN*NuiR7mJ2-~Ti=&Op^4^K{`PE{H8i_$5 z5!;_AVVokoc4pf8<)%Ve}=($#E_NbGyGgrcEoka26iIM*)f5fRC-LD zF5^pZp((;EB*PRQ;@hYVJCT3JU`wIqb=1kYY4&&ZXJ1wY)$l&wXjN;JE2J|?``e62 zekQWLZJe35XAdGuSq_{N+vj;Uxtshg;9jn;nIbY>-{~@&AC1v8@pXt|z4Peu^@`!n z_{M|4UCyr;PV9aIT{LZNKBR?*>hm2}$AbQLbti^5?0(%Mn%+_U!Yn`hw&``(3qB_P z!2aG)+Y^#{-EJ;|0M()qyBUT{?Iyfq|8$=P|QJ)~9OTTv_-0uwpyi68#N*tiy zVH%YNWj*V^;&1k$37`m2!6$HJZSGqrW+CTveA#$|-@SZq>>S|PcOTIIA+hWA(vSV? zR{1l-4tPJExxJUA&L0-tZkF_yJ1$MGJziB@U

pI-N>ylxYDM{!DVVyx?pK!0&M zf^=3~75x1Y7)mUuJ0lCHA*_8d{Fu`5^ZN?3F*-Ctnb)|=`UJR?-wo#8?f&`Tw5nKy zRJ2&~eXmf;a6k>{nvb1^$27@ZA&3JKmagbUVOa1F2_&fZGytJjCN%_^)E8;_1;@TM z&Xpwm$30`og83c%z%d+I0j32q5ZTk$JecyN zONV6aRV-5VJz7d7)1x5(suOtZmgdZ*ZvZepi~C7VWf!(dhks3R z6(_Ohpf4I;VB&hx1DE=EwbovsCzPXfP%C{3yH6J}09t&8z9Lw6Ki>yXjygx`Fobs( z8JyD44`g3Q^aLiJFvM$Hyv9^hgf>D zb)O%fA9H9qGUX-9VrXtHZ7sbEUn}OwisY^I*7epD(((LQ^cy7W+;HA-UM5~jR*(+p zEJ8yZb-G{~bUc;bYA!-Yr5i0dHi-})2Q@hqcc7IBb9Rw6iaq}dU_t|4edY6G>ry1- zg_NK4Jyjqc;zi|oV zpWt7R8JqnCf`i*5cs9JNabkGo+4p_<`sxQT6)Aeuue%;h*}ag=G_|~+>8Q%d9t8LQ z_$$2^$%6HDXqmpc1P0+;wu2lxLU*1&a1_88yL4-0m!|?^sn8>_ zOlve;`lxTictd=EPFw2j*pTg`RGaoKz6v{+(EA?y9$RJ85y6EyiGJ3lt+uUJ2PQuB z&&EIlNn~yW!A548GT-hM zY+6OL=mUn18ki4f-j3kHvtn1dfsBN+hO^>Q zwkok*+z?t5)zT36hgAuRPKyL=QBDKqK>a*fo3BA5|Hd8sPbT4?J?msWT8>Fua*ZB@ zJ$}nBDz#Mz2I;04C8KgRPo7G0bH5K_ve;V&+{mCkl((wXuY?Y%EnHr_`gD5U7PCs4;F9537T}lw<^}H19V5ZBQ*CHN z3L+<^Z|t0Uwj1AacgEHOcP#tTH@00QYRO)3Ip&9r5PC+zNB-U8H8*SF#OCo;jfkwE z(&+!wmaskwS#%DRc!(SO$Kw^EAH6>3{6=I*@aJ>lWqXx2rwDp?&s>WMC$|PXsyd5Z zPBTz1J|tUHv&@!0JpDs){xib{cGy4AYjGP2th`l1ERL#AVHWvI13VoB`p|>Ts@Zw8fPGzv7q;vIdV(Go?%Tg82SK_301VFAa9$KN%$2 zU3=U{-i!R)6L;U9r{PZ>u02zAAHREYyP~}{FnOEHDL8txKw(8)I%g}QT25Q~tx};< zKFGeZIluBp&&Q~mo;PaWDb|tONWWty`KD>D+hbzLyQjNR9J=pJ9rdSgFt_V+-Zz+c zO?_m}`}4+#`C>{|_e+{(OybSqC(6Vv?%mFyQTvP2K~sRY0h$$f(_KDAZ8wc3Z-x+O zZ+bs>C?VA?jR&InC((6;35!WF2T)$}sEe&axmtM|-+@~d?+%oqQ-AY3!8ptUaj>ppty zd~ks@G5YeNXxMDAn`o;&!bH(=9~y5By7r00ae$53Lk_ky+&3*TgB}Uv?t1Yx3^j-C zXa&bJiX0r2iIZC_@c?G=M5f!wmSv})5ut~l3H(1zD9O(xOFX!HKB}@vyElrbnrdVT z>xD?cWl5NIoQKR}TStZ}gwgIx29jy2zrnL7b}-Fz-A6wL zHiNTay5kEEg9HG+tf#spd<;ptYVH$0q!@6XatPXLeEx!p2F!1meBR|O#y0Dq!oT3D z{f!P!2cX*cGuxq>tJ0Kf%0c>0-Z1on z+0-@1U*?8A3zM13jd{MGa5;_G*k_`+<)kk*{?qtMkD-Oh-IpMkB0Y!E=N%wlXio_k)IoR+0Zll1fC>!p!jY@iGf!%tG-Nzbv73r(mYpAKgw{w z#~gpPoQ2WYj?WOUFtbJ=ms18GVjm3xO3zveSl5$4#F~TRGSyu5} z{dlY1FVMx9r10rT$mo53AxnvVBIh8DL1PtCoF)`TYCt}vgZ_^z=KB;A5Q!^Vm}ukdaK@vdM_r{%R!{aS7w<$J%}31HJ61GScPh@ zW@vXPmZmcNQ${>-8N$Fj6~vtOpv!n)l+Frv7Dpr$fTh_u+|ZCw|C~ycHn43<0M#M( z)>`rxpY%pJcktQ6`(iKZ(N^Ql+8RVV?Ut4uy@p@KA}UC;s6~DE-8o{hjQ%1&$Z`m` zQ$cPGo~XH7>xcls&F|OhA z6>MSP`Othr2wt46HkN`KX)ms63aVE3E*kA{8uD+@Q3$rlpoL|a+f0lPgIZBva25j| z)JvrW$@p?q;HK!QhaP2;)VdZD6Mi&de<}Ya8>$`Q8@SIH#sq5+IM3j+0U`_uovA+(!Vo)EwYf%FGzkiz1aL3g^;4yAoMHLr>RK9l7+>!;jS_w-ah4 z;4-;nh<8KW(ooWkK@9H(sPHJs;PPc&|CGGAcxe&bGblI7oEh%WINe5qf-Ka(xKFTuZdTG!iyHpyg@vtBrp&7(#;eW}`9i@D3eJTdeO#cKgk%^2}f|mGEGyl=GL`ol%7H zTEUS?g3B(OP<3glXBukF&8(@eh?@hgla3GUuQ+kuz%hJg!FwCM`pgVVZx8POW9dAj zn)tr2pM(SwY7%-;0t5`bS5X24Na%!KrAzN3U`s-g-aCl&4gvxKqEe+JT~ts&x~K@K zpvd>hZ$1CDChumg%!|4A&N*lA{Xw^I489pq?oUI-kB$8PKHy-85+u4|rl!QOlAu5r zBBt8i%wo+r5|fwC#;6YnZ%ESgbj`KMHGu-TlP)aSlsh;rdq?u~OZ|>i5C9%PkWy-A z!mglzbLMzZt2lH3tUg|~a)xtya(GQegAYYjcjbn(YW9`%k)nWB5O$ZFN5n<{H&do+ zz|8sy!;g}JtHK-xC?0etn|M8dn^3Xgc z=WPtc&Duu}ei@mVv(z1|o_ct=do@CW)K~N>RyCIhiX+Glc)adom06KeU&cVElF=_iQ3xo}l5E(?~@)_^lUT5Ge9luLDtFr-F z`rqznic>38Y1v3V1#4&kM<=x&ogbEosy-f}kS##_^CZZI+!gSkV21KN75a=w-8J;o z*hg|1tt&p-`>q!UEkV2nTp5X{heP96jHN4#dh2jD;|FFL(LcWqj*GqTrDR3>TSIi7 zl$k_fm#5_%T0p z028mZ{;c(UEhAftp)BDpUJ+}&sXL|(54_xva2;AtY@6u%*6(MkO;}-9rvuQEotYrd zvOs(v&37tt69Y)5`n(WECZWddr&a*jLF7m|Zg+6w%aNvbur+aDJ^y5OT(qXTv>TqJpM2rP7h3 z>kqo|nVW(x&?|CrZmZAMhS={8_*28L{7AIR*-Y6wcqdAB(je?ZI-#j|I|axwnUHYg z?Y0#Lx}Kd}Y9RT#U4gV&BQLi%2~a0VWHJd0O|89EB)90AMT@#wN2$h4MT29PBI&h9 zX}t&onud-vVm$M1aUM_5j9SDOT_6++#BAG6TJC-HPE+KEC2d-1|b0qQ`HGh6>#du?7R8Yq%1 zV9L<-Hh-EKC3whLH0S#2&uG;~77Y)q0Nt8sTl&;QbAKzl0?8PnR3*Rxzhy18e4l&x6Zc`P^WT5HqixK0!!!xk=0N`e@;kbHhl<+v(S8fNLEpVGmC<~O!3Y;R%aG>warfTexb<7X zIhEQ>31a$O>CeR2X58m7a@1@{GdT?bQ*L`G(8dNe(-h`ohg7Tb45}G~#f~@JnfSI3 z3Fd+Hi+JaAE=rh${@9r$ku2h1{e%4(zI6Mw!r0Y1>VHdEro;`h!`WN4w01H{B*4tM z43j|)%~h5I80?F_)ArA_Wx0z3+Qbw4=i7i~JRAGI#}NaPj`nXFEB{c8P=?r0VJy3( zL|I#BvTUG8N*@5t5H*15<|oInVBFq5f!%aGln1(IE6Yv&9pKLkc>wwm8cXaS&Vp~L zKOMMvV}C&t+bQo`ZqUMfbz=I+bNiSFhkXmlNIbo2kYjqC*SKj^LF9Wdr@hScF6ZN+$ZA)+0pw z5Iq2s&nRQsy^Bjg$ceNM#W;U-2zdB1eQ46i#A&&dt64x+>euyHrE7}%*w$QH2&iw! zj3#`4K^GB#al)TJVvY6^`5G}*A_L>7l(_ckd-U6vJ7u*4^2HH|Cg!B{$w5Yr84lWD z^4|?Y_*NH$DD4mia#F};x)LWXMKi+tN>n|Pe9iltj8P^`_-NIl)6bYju+*hC`Gg5$ z0XB`t3L6CBP1j!hPCmiBx|7MB4j=(#iE+A-#9_u{!qCw_Fj;1W?V^|;v?B&U=<*Gb z!l4O?ZpIN{YoTUXI)GEM34iBcR?aRkrVoz1$@GYW45AkLmq!dsEIU;0)?$T(p6s~_ ze6XAOiOKGOlN!p5R|h?PxU`RS6#^^@VjpWPYYKFmIUW1p|M|^m{;hMpwk%_d6bh2y zML*?&=1Qc}3(@s$h%nS1iSZ3`B~|IS!8qh`PqgwT z4~lvWYyB&8PiHwZ4p->()6_{ReSh8C&{J?S>4kQgNS`nxTnHoQ#C=KJFEKr-tg)5> zteFg_-(`U7cNMEVuj`$B8~yV&==opm?#Ok|2RW2XM6ezuVB0HNivQ82>>s|Re%cYD zE>i0)x`ZqAl%L>lsp1fVu;ybK8fuXXiDM3vit`WhV$JcJomeIPLOpVPR;knsmmS2s zP|DUyxMZC`!2$qVCI{;v+GX{U_6!NqzO_>(@*|^EJ6hFSf5Bk0-1<#Au4o0_1$F^H z)LL|2mB7HkK)IVstYKU#usO>hzgy{UIp9&>E9u6ziNbz{W)=py4iYx8{>_Z0 z^a33tYbL^lw+-|~`RECNfm&A#ypqadlwl}Onox{|bh!JAX_+!II^P;IqdmQU3v7RC zJ6I>AeR(njcW;4)IXoq0d*l%lX+*w(i1(rL&zGOaO6NX)L|t?Rc6saJjw+(bGOKTW zoctbIkR$_ONP5*4|K(TQ7mY>v zn1_LjY8Sg!zCyeX&WerSXo&Od%#;*NUo>o_>8Rf$JV*WHFqE}NfSGUG8MFO zsC=9obQ%ig>tOxVfW)z5Qoi%oH_ zn1cBaJ>QK}Ksvqf(x$|4PTI*?0I=*2WH&U(y6pRd~(8&O_ItuO0-cAOhuGLYl8eLFzwNSyTPowi6FBZ zB@$OuP}?2_0vSbC{7u&>evKgDdG!k{i2ibz7AmDyhzkUep*tTr4$U0A)~GnI(CeWk$b)AKLFp9 z%JHio;^GC9^e;^xztv4=oBqd?3VM2d!42`p+xZmJsg?oXHbfT_{+fU?4|O>f2G&=>AqSAts(ZCTdWPmvo6@jsP@&( zSFs@v26|i^QB^;Qk3A%_f12k%DLR!;Dz#7wAaPZJA&^McyWYJjZS%mivQ3)6>}ug@+uap+}Xem`d0D0bia^H;*W56&63sgK9W15C`kdxpbYEyK%v{}@exQEI z0>(yoMH*cesd#>7+ECL|W%29(T2Ia9XAU(--6mTmH;<|of17^NwtawE_*E-u-9%US z)9dx+XBF^k^nm`$F#NsTVz5XWlw0{0W&pZ>|A1l2+>iUq*&iQSw$C$4>p%(a@KP7r zzZaM7d$MEB1*MFxig^)34t@L2PeYwf=0o%cq428S3eRR`5$?6Z+bn%h%kN%1Bm(IVu zCq{IRJ%^dc%*B4#o$1j?;ksO5qj~heN=3Tdz1BHP^I^=az1g6)-TS<6Tqm!_!&UER zlC40wuT@mIeqO4@+x$Lvc0pHW1%D@87Lbk>WD2~-Gnz4M%hgmMMb130EHd6EJqLJKqBXPU%&D&k-n)iN#{gm|d z0E0&tp7hdB{$V6e2cFT2 zv+v8>kS?sM@BYvoWj5L41*E2+E8i`wi6LSZ%3LR?D!Tjgk4?^-lOFFG-I@B)KO`KC z1AS}$4^X)7W(3lUR)9yG)J9qfPFtpel9Lr_x&AJF^V|6q%o);17TJk4Y4{Jo5?00F zeP@Tiua5Rjm#*iI_)p-BN5EnCn~Z;8Z@RzMIdLkD;|c+#6vq3uB&PQSMrF?& z{kEQ$;2`$ph9s7sO`9M0M>bl=S_@l-{sY+Qi8_8-up&Man|6E2vtsn_pQufEn0KF` z*Wt74233RtjW_QKLKBrE?}^h$WX6*_^E6~m;Tr#_1x;>gR#a9g#Y%s5J- z<~Xx*IO7~n-@VOVK&Yp^a}LiH30yK>#wrvIR#`|Vlpvzp#CttUD^){-H9<}F>DrKPX9#%|BZTfITw{H4n+Hnr1H z*S^yEkE`@$LT2WR6J}%_5BZ4dlyE5SGZM?0F#DGOnq_?|YM)#>vs4dMBFTjM>UZt9Gg43^gaF zBwao8pMF$2L@N9q@x~0P=<$5CrxBN{hS`gi)~gq>E2_hepC1jk6t$W`s1Y~_wEOw) z7>iQz_dUY*BOU|fk?33tA_95_EL`esaw~dptrR;+pRlOR4*wHG3pW(u97YOr+Io zkE4H}+lIXDlLb~z99eD|pS2?%y+K>VjB)7-u!SqcEK^uaA4ovTj;U1Vf0y`{ryTtM z&WRx}l|VO?=9*mZ>}kOemrOX=jumrRa&{xG(fk!!T8C5PCVkhOcTrXqK z{>h_J+PHRBIGkoo+W=$xt#<$B1;z59{wZnD#Tb8?Kx!IVWDPU{`a#?y z{nn|7AHw0-9&v$m3eBfk_A??ZZ;q*qr)ws%Q2f7&PiaKlI6J|i>|9`jj-8(%##ipLJ%(PyN!~oXATQ}$#=>Ca1T+y-` zG?6g?nfb+d0jt6!KaGbTi&>6q7*3d6^VuzXS{M%MjF6WInQI)d&lBg~W&1-{WHWTc zo@H`_zJ-0X$usi(9q0^1)$7V|;;tVSv@g(ouS$b>JUR zrmAS7MpqdwKSnGfT}8K#u7-v&p_9utc02itc_5uc_rKCy#*q-LCzs z?A6NWbX3XKkihkJfj1zN?`y9(q3B}!IDwGDp(qIZP*mLyPz!QN_;Psl9mwuuO*8*7 zN-cmCeUznwulf{DHHY@UQ@chgRgT@Xj7w^QO|M-KoHrkJ^?zF7!;$?o;Ny>g4!~?Gfb=>jqJDnVHY>9T9y(0@4WuH1ac!ii-oC0 zSCgDY8>YMg751T<3{wvlEf$9Q^+%kX_q6Capu>@Vy|X;ji1fJv>&GC1?VC4y1eUnH z8(;t26t|n=!6m>d`sYWLDquoSHWA_OCpL8zo#g1~wvj)j0F=!ATa=nh94bzR)mQ)sSp#^gu6m261+rD&+#sgZMYn5>qcaB?j4;;(4 zkL}8I1{QZXMSm!UjuKUX@K=bCDK9hpWO$un6Qg{fohnPQT@9YMgz;!P{;44Py?*Y2 z+-tO%@YgfDl}C{?Np)wYSm)6?tRN*uh`}twwDWs%rKev;Edfmsu50A}^c;wARr)9N zgK^vxr6b*qH_bpVR3bLgzX7p3sUQ6wX&8Sli`S&-O3SZ`**u6Aa%agqsS-EHJI2p2 zywsZa5?qoqKoX>QgYD={Uek#+&06Z$*FRx+E_F{su#X`d=g}h)J>IPvqfAXUH`e7a z!cs^1KD)<$C^tmtIOA&;kIss?3!((ETYzk^1g?@C&}! zKHadahB238aAt=4hqTD=9ohl6RtIFzavM4661qNpnWkhb29Lw)hl1ZU6St3*&+ST$ z4V-%sB$6!ypI1jMa)a9nVDJq+?-)l_32kHEJ8_|mwfS*yz!1Tm@J&>vl1wT2kk z(Q`}`@rLU0gPQd%b+8TC*oYU$8;wdN%>H1PA@QHk_fTO*H3EB$52J#Q>k~9-X|CN2 zcu?L=muji@0ZOMxBB~BL@{$?*+Cs{FAUv^9pmzYLOF%S@pxS(-Sxa`M56hD^n>o{> zKB#DLzHzZDPNHF|?FfkA()B{e;k4-`bW3jL$nY5UaDkSFwI3j4({pFXd@=(ggq*Af zLk;Yn%t^#P8eK2?p*`FU;)~a|)ff9!7$ysvXjdu=hYO`Mn0~x1Y{O8Q6mnBtSWDIu z%XNHJbe?o+X2t>N+(T-AO?%@E{ER?CyA@jB^>(D!R7=&-T2Y}fO}?a%sUzU{>BsG8bgp{~(}yxiUawUMd8OT49HD#nr!kYW@PkTB*&RqedX58>@oeU+6QWqA$TxU4$}#2%4nV>Vbj0c2KH-eJ(cCa zFlqH{tfVB*7J{^#b9Um(vsWfK1g_-H)V84yy5w30VQli)HC>Uq!dQ5B4jTxiZRq#j z|7H|T$kl2=2f3#aH`M`zDi*O8#_dA`22aBcC@!zWV~XCQ*y}5;P7IxyLKA3D;yOMB z{^eEk4>2wauL>+DK=xMY!rCMSzfxXlUxq_<-#s%Qju+eRb5H1Ku>Pg5R9kQ&UFDyE zXr#XKUYm+-`UKt_TBnJw7?HrxnqNT(HzQ<=b5ALzJfJL@RE@SdNd4m&)E_9*xh>EZ zP5XeY#n6*0Td2ZF*-RkTnHD4)jDfq>-oDPfVJbjl*+0_QH`D33Z=44na`B1AB&O*8 z2SDd=G1;ILoy$K`ZZPuLIFK9ZM^`VjQ8%!Tbw@7P7$d6#AP}f4q+(P87x0&l$RH8! zd{u}7lS-SDa<`7xQ{}x#k78_6#Qn|pqxDgjvrn@A0lXJ)CDh%Um{=;`5SQC`O+vW z_+G2V8!PbJ4f2h&Pg*3@q>T--fj9&E@51sJYcsjuam0!Qgj8{R;Q#;|j2#+GA%65= zxNmZF@ga7+~}*y4TL1ZbIycAZiTCm%-vkHH#@I;y(pU*I$j4qJz$URad)Ue!N!MS zLW3PRS2lxsrm;OmQB2C3`*5x(=f1nDikU<%Unww5QGy@=5Id?a+ru0vC%%4&9!KD3 zU7QTr3Mi0%-R;-Dtc5s2*NVd%=~gL_`x|bjmcK8B66mO!kJk6BxQZRSHs0|wtcRz_ zyXQK5H_JFT%L)VV3P^RHvM72__G8o5*+MJ9cQplDSg*Pck_}sHA`jtFxg2MRiv{DD z^H=?9fpl)EUW51s5_;`hTb~>}&;VbCTmB0L6saLH zDfKU^kJ$U^p@I=3)(6}kc;1mwoTD=o7)E<>>!zRmyFFT)kM;|f%7)nX zO{{Mb{QR42iiDYfb5%mYN;+3S)5?xtn!l}&#Tm#fNMit3{Cnzq>KAVR9eRELiX_5d zR^RaLOwpUQ+B@}r=k(X1eY%jm*#m#QoMZus1?XmgEsjyaR7)tMd?2rHGz6f+F8Jl5 zT`I~E72lH#`hwikylnrJB!CABKn2Z;5qBQus^*!?nT_wf31udD_rXB7*m*bC+vvG%i;^Cf zNg2j9IeQ?zb|vt2p9y^tIhWms zq4ht&sM$6P+U7ri*S7?Md+@~rAHvQ8_DBmu_flteyBOuCO?c3E~^7u)xc z#kaFJW9U4@Lr7W)b_HO;HDpvxRvlOlSQZU5YvdD5WgC+5Y{9I2jt zKNUo{qhy&N6 zhsyd32}~v&`S;&^xz^xy;prMMyxO@#1MbBfYQwSqfe_@~n9qxRCGAueGe8EyxWgSv zLrp)ivrjp)WWQ%e-sI_?;ML+BZ^f!h{oK*DxNdgZ%(&Px9x!=XH>c!^Gv+v&?&NP7 zVkDpCbqnO^+3Na|Qq$9ScYc;svLdjN5qIoB-!}~LEZ5}rs$X7oJ{&K@a@;88xm(~v z^;C#lohttikn0MK43(T(gq%SyGrs4wQ~Ie%AE3`;I;h6hJ2R~6QLLeOYjt7I&z(wo z0HY)^V%TvS^0UfmtUogdn3U^v!EzkyN*wMngKK_b zr?zb)QU~5&5fZwDG!c3Mvl4JI+7>dqS-2FZ-r$KqdCh&5rk(mn7sXOc<74CcP=fkp zmec~sFMG`OiULRylJ#_Uba2kOL`7Do&)JbM2>HUO2dA}AwgUf~MnS4+TXMh4eXhIN zZT8oq!nfJJJz(d0dGFyb0mt8UqViQWH1ko!KXrj^(U`ehtABJ=*y|_`pMAIEAhs^- z+JqNLF9+|g9xnv#FU>3>&EHuo6fJZnz{j14G$L>yD+BGYI8DQD;l@+XFW)8${@BXvmdM-HM zUa>^fMm}u1)zrQTuiNfFt;K7!Xc3mqbyK{{-^JNn85OS#9628eHmPUkf4Vtp0KV?j z>*X045XmXHt@ULf+%zy5pv)E_EfS0}PuepCYEM1<{80N7C2D}x@QUIb@w4=|C5 zVFF-`lBBGIEhd=}YW98fv5DQt_v_YYp*%mJ-KE8QHjDrpTFZ>lr+LlaN-@}=l-bL| z=veTO3|1)YiQn4vvJt3^F)@ZsMibYTD4Y$Bb!RY1s0&k(s%0bX-9;?M-$igsdlF{y zhUUx?XvdhYqAhB(II^f^>m|=kc;fERJ7>Qdv(FCt*b1^PI9Fn7-n3R0{6DXiv1qvcfhhkjxAiV{Wv4`LkcyYY{Q}esi&+;8r}pA} z!9sdmjQ3$5S+bg&o#lyVRx$}GH;L|N_v%b9{05kMJM^=%T*5* zn=y_5t3R`F0gr?;UXIcHBom(NI#(C>7+$FjrM!2^&Q_C1-UuNJ#3qF5KY|t%ATogY zLyWJ-UKqMmL(Vj18+w14wg|Kt`w?Duo_*#>SE~1G<g$G68y z?H-tc19e0DFPAc)u+7!(O!+KJPl(BK%H*~XdB_hkfuee?_BqLJy41W zTy=^Iv1=UKV~CQ3j)}5l0B#jR`s1P4m@cX=keh)DwdgHl6rv$0u*M;V0J^)~a-}Nx z3eV+HJDL3Gu~JyG|1ytuOJOdPreFf)x?Q;kfL_AMk~R&s@HuKPq&)b=ShP1*&^l>MKKvQV*khYaZgiP9vu9772 zA_r~Yk8BJ_K>?6Wq={aYfYl{sibJ4?aIBTtA~CM4y%v||9Ur@Z?p1kF|KV?RrL=QQ z�hnQ!m)8ev>PT5MWL~SjHsKE*LJBsph)za4B(-hF0NNx-fIo4g$_mE4~1vyAEjS z)0&{8y$!!&O76xgVRRWf<+aCg^4Yh&54FP%D^|Q{vj@*GYjkG6f&elK6Sr`&6oX_X zgJfY~JX_T6dor#8Da_SfdDERpHBKqYm|q)jG5k<#OEKm(yh{vAIsFg7O03-udJytOM zUt}OI;sonBQThc9$l#ySW#9X+_w{Zb!CD%^G<$~mowVTCPPw<$Fm?+MTmhXG=YWva}e zGB|}aYq*TBT7r%ItTw`C|J3R&sRA*85H6U$0WqYz_J~*fOIi~Zcz{V0{#X4;U5IkS zt9^qOmC$!ALktbT>5MG#TWh7V4XwHYz~PSvirl(-#TVjJfzJsX!m`^itffGt60t&9 z9QQ)+zF;RxyAAqhDGZ~3Hc<3^I6VAH)X2xO8_tiNCHS22q2;pMyhNEZh&Xj3pl@39 zZm=|DUa#s-q?RdQ{nmeg!_RSg8Jj%uoc(a#VTHuq7NUFq?>6DAEE7(NhN!%?xn1=m zBR1Ia)DnQ1MGFMj)DwpA$9xFmh@7L&y1xqcNT_ggKm&CR1MvThpI5+{IV4|;J+gz ze&~|KCtW|U>Q!P(nE*d>HDypmtUAgz9ye~sXec1afK_@EwXi?qk9V~P_6u!!o1XlN zUKs$kaGh`SCIe(7)wD(_1C$CZ7yo$H-Q)~+T9~uQ$&H0^KMO${rl%j(-A618b`|lo z^*r;p8Q315Jih40O%T(8ju!g+4NvwX5viB4d51Ychz^jAFg6cku1?*HtF&O;F>54n z6NSRp@Fdw;TiODb-HCd>CeW5XCJ4#Ag1l;qgyK^&n<3PLmQw2Ig${l5HgVrHtPViL zW%&REg@P3i5s(#L%hSgL8~ZZ=209AkjX46%yK zkW(!eHf9st1x4z?u9U7xYY!;*RxRuJv-nBUlVJ`t=kzE+zQ6YQ4wL-=F zimFf(xb|^l%4Pwdz^*@7j06AF-ZrstU$t%^>r+XenpR zHhq2I*ZpjA5e70&*!vYZt?EWoGv~D53Ip|z#lAxuVC7%101lQ3qjIewlh3KK@GhzD&w29b^rMfpC0Rwpmh z_*+M?2|bH5dd~9MUx+1#@p0)U+B?Hx_EYTS=_|ao70n|-aq(O0fl8|U^`-j##D~?x zHp_l)gCLIy@@?ON(MRK5r4B~>UqaM|#F34m{{hD1n7@PLMXJ8mBDf;55zbv@Y&a~+ ziEaX+i(A1zKaO8zv$hKFvVSikM*rF>T2h&;JITPmXj=Sf2LXGvX3MH& zjAAbf3JLU{{O!=io-G3yq-XPvm5l8lOrP>W*+gyAvr}RYFtW^at+gC>XZ)i%rhG~J zt(C{y>I;Csio4=sQ+;vBf`n6C0>~D5Jdhu>v023cpl=qeDM1OimVJ{|Rn~I6Q_{u7 zKqI4H-#$dU*^cejVoVVkO2}_-$JD)%3UTl2JjxtQ${Yb<;B^Kr~f31C-c3dA2bR)%BxH#TvXv|8 z8i0w`sq*h4w9hc%7Md`;1B%ec3PE)-X>AVz0?! zqzba?0)To{!MKQ|P7YJ4j678K>z{88#XMjoXbxiHo9k}6f6=q~=0v>G(u?coJ)psI z)RM1f1&)!%MW&0*BskzRL^OF@Zo52N@&krcmPNmYj49lVmnA>9si<~s01+UQAB*vm z3grtdu{q}xXshJQ(74?Qkm(^*D)4u|fKLQ)^20!cP!a2Tp=_l2vp*PuThQT(t>hnc z;&kf5KD=Q4ey8f=(iQ{PW)US78;5`;4Z|_)Y6B|wn`qHkX9-7HowKp9VYgs0? z64;1xCC?nqYS$~3#Ve}Ud z-?>U|zaYS`P_lq-51+^kXuTZoRYA>yL2}E`$$4~!flRZ4qZo(>(8>W)S+6rL1`yBd z$&m2}RwauJ(Ts@d#TTqT84bSWi#U#d(D-@$^Tm_2sIcQIOZz8`^H0XjVbAqJ<30n% zsws9QliI<(9aNI0m%1O;kjB|L#x*}DhvmnMSP|6*y2Z+!qaazX`7p=4ozR8y=g1WG zm?H)Lsy_YtHLB3G3qs3*%~i~NPoLOTK}J2Zi|$Iwgni6c_FmF|ZVKru#Q{qp)C6XU zNDOe6Is4-pYxK#Q`uVkYS(_wNxjTEgNo2X@n)C(V_G{KnihhFq*RS@+j;Q8Ty=E`C z?|;^O{je$-M44I75Ws5AKec!73=d^jLzIyJD%^~`EV-LY~=;hQY&VI*9@8mi4T4ro z=pWzsw+2M&35P_cfu7ENs&ZCVE#@O4lUKt&91j@XmQICYb6U*er>$?Fq-Lu6;>XyL zA`a)6k47gA8Ms=8>xNY`@HQEA9v8GLns-W!GC)N=NHg{X zIahSWu0n6tKc|kRXX^Yf2AxL923^Wh=R*apNV9KTmu|5xMp0=H#r3_E77SFYlHsX| zzJ`Y|OjQ#$14PM+`n7PkGH5yc#<_zm?=Ie$#+8A9$6h&c;I`_-$j|JeO0KvTWI{|D z2<=GgP*V&C+<(&kIeO~HFr&xL+GH_JE`x38*vIzIjSI=_@3CGL#;lunrY}`kUxB|5 z5m2*0E1LhA6G&`a_JJQ>`ewOh=f^LzV49R~TIp9CAk<{iQ;L)?6Ff;VDATzUCM3e; zHMbIShw_0Z$%s{?>59SBT0RfGnPT0jQuwV`(fV(BkhI1Jaa-8EPofe!FfvVBTr8+L zS;ReN5QWlFQ~MJw>dRUKgWk81v=6aFb0wmL?a>v2La7)lE4HF#77wN_h)sGbNb;5n z8-t&BJV(6v$IGBFC_71PM?m5HFP@Ky=%;9B$|fi{<&l<`Z}YWJB*1ba&w-jPC=} zSZmLkVN*NwkDZ!PsVJatuB7JKkoEGX;HE6LggRy(M*k|?krMH%BJU$S7;Pd?g=ur| zN*VCgrV82*`?R*EF%Aqrd>pT(&ei+z`f>kv4dZWtd|gg~@6O1MR`j?A>Iudp{(^tD zRc_g_0rD!!QorH(`m)h-Lo8Ju2f%8!_Ls)u`=T#qTNS71cWhAmaX!OaMj`piSPoUf zKNXQU*-;rYum>Yw@{CuM6>sx;g}Z?uPe?kp-K$#XO#S*7`)o>lg(=lglXe^<3mhO1 z7^rdChI^?tg-$dXN4w|Z6y`?cy~>R8Qob}E{0Dd)&G&6sCAk2gQ-!4(^$c)u#HdwG zJg2@zNl^#hWLGybjM=P(?V6!O8@a#%MuEjTWMuI)^$@4*EBTuaRG!?~;0R*HPl!vy ztZ0UWU^b9ea%GGj2#X*fBd=N^v$@~|GzUR&MH4hSxBuNtI7wde3-fd+?XuL=LA*$l zuYXQl4s^T^J370^CQ6ebDv@3x)0}AvVx!uhl$!(%Z79J6 zDpwPQF0FErblC`-v!YK&h6K>U8NuzS%;IIAwzHOv9GmvJ!RgmCr0L24({RfTgH%Jw zvJA@)Cy+WWF5S!Yw~f6#5qHWLv01buw?z_V-Y*%Br=k|8P@kST@H#C*lG+cg00vij zjC3251YzMnEWwEj4-MSxGHtQ}xgeWD|2Y&&n;`rSOwmP5cltl&$kvpRYNjn8i8Js- zvgO;KPH33Y1wX(uO?9dwdv(!Qz5+j zA>kkomQVB>t?Fc+9Of!^fRt8xO$mG$Vu!Z%Cjz&xqMps0a?^Ycf!Ra;r*J)_Rmd>P z;7@=eJLzdc8{P~FP2J_+|M8njg1dXapt>c-?P>OwzhsazO`scq#4s{3G3*3?}m4WnmM_k!GNLDv6Rjk%U@%`7$pRSieH}oeaqBb8n6N8M=zjAM@mQD|+#oJ} z15Ul8fiCOwOPk>73_&fg{~0hY%j7b2&|dx5YyE*)^{A4F7I{n(u0?!`4&?8S&{!QQ z+gkfQvaeO_u`^)w`QLqmgzhP-4$mvE@%x#T*x;`0zXo2WG-jWj{io**?V%b_AfVpy z>!5PcXA&#GK*Dz|cE*)3(%ZpBwFiu%nUimGDIypI8#F3Ps}kuuDbO zn*J>S9(huv8>D1q4|TjrBCG#Yxg65Bo0)y$g_H?FL8Arz90F&wCV^fQksU_!_D2PV zD#nC09cRhy{Q6o8>z#G9FKEPT& zzA_+%cQyKV@;AN43Opo z;XLs&)hS&~EN{|WJZI|TBAA+qKcE z%vbnXoZZ(RSUuz3KMlA5YFDca){X}~nY}wzzn{Dwu_m)lpe_E2`6%oW?uk%-Vt=|Z zASA5B4hy5+IZwu&*)4m&Z&reYA01>doH}FbqlPgW6_|>)aZ9*VTj~-cQtd=I6#h#wAzz(H&zDPzJ&@CrJjx4v zpqAqd$Il5g>LGevu!|AVGWoAUjW+NIH?iuT(ADn} zd;WqwRAk#JALIm^5p1qwooXvHz24&fBu9cn7EyTaE0nQfUA#BzZgh?CpiJdQp^dAl z88DjEmoG?evL__%8Nt zh#J;*gpwWvR;%R242z+C4fA_cNC8C?1cuF!cv~dyfb{NpTO_I;>P)-?O_W9Ei&Dq3 zE1hW2$R`r0XO3U;mU7mV@DwhoQMiwB%YuiSOp8Hs7%0{^vT<+9UgFh^tQZJAO@z*V z-jp9|$xD5~d-~er{IH42a#>-R zs!r0;YAKsMyj-85s(P|KV)~^#=~P0gEw?$j3d~{8m*7kh11!S%e+7&Bz377)nur_T z-*39%FqcY-NfVJzW2>zxf1GQ{pIt^ClF__&6JKSYtO8sWByUKblImm~UYaj0aaqM%PNZYz#(hJb6CG$6_%AK_Udy zGRqz0fdor5Ux+AOJUOoA0E!2qMx}g|)5a!DqQ;v7!A_RZG&&pu^g?S9AQ0yqO|sPwQRP{j$Emi*W7W@>wn%MJa6F}_iE|S6V23k-eqm# zGhy8rg7|*%L)q_%@OP8Occz|#Q4tAJ1EGPs9&;htSzCOjFAF3kKC*ZZ)^2@!{jGP# zcl)!%-5ipmes){Ff-JoK=Td?Z7J_WW_4h-aElH;nfR<{C&*BTdnnBrF{VW8xegpV6 z$$HO$>3Dp+OLP+2_#VtMv6Z-++dzZj?vXf2yWu{;;L1Wwz~_D0n`78#Cl_41nS49L zu@2%ULgmePTtwWvn+3hRnWYZDQn<)%&jOrw_MQPwnZ7D{YbEcLPuo`P3zbXut}w+%g9CDBjB zx0M#}2WE*&6QiDT7AXzW4~Ij+ zC3Bj!M7lMsHsna+RMd$c8k#lOSA)LuwRe%o2l!l(C>FiKuF(??BZQYH_RoNk^kd!{ zbE^y%D(qO>Q2_VKlc&Xk{75RchhTRY=7^ok;}=lXI`}z0XLTT@219Xl`Wce9z1x z$%dihmY^Z=YgWzdV;`6mltU$I3?nJz+)J3zHstpzG0-ORsZV?yhw-vcxR=N)j!H8n zjA&(hiQ|o>K=Z-DMgv%|e5u_x>j(!*Q6|t15yhrfvcgh`7$IP%(%EFZ z)tOm}hEpI~1Xf^@%NTmT1zX}|T^jaj?hOL-L=q8;vzOwmU^%}uI2c@_Z3<_FD9|Gj zPiW{@c~u5QEu}S?i8eIqPzF7P2Yy(r zwl$?|lMrpED)<#;r=wp&qn2qmSbZH;lQ|h?N;{&Ys(Sz*RJzk?V6S!5rY7z{g5^PR zz65GTjC7HaG+#KmKtw8w7`UrWw2rEgQJ4a&q>(D=-3HGFBC1y~+u*)PT2_QSilEq) z`R-wbstyFu!`x2yadQNG|MvOOgfBd0gq4vH3}AEDflk(a-T=MF9goD{k>*_{_-OcU z-nMIV!~5dVZQQ#%{$*cXig0Xl_D}I*mI6u&)Cj;j7on?6WnfvkF$aP5gj!s8azK{$ zrW5=5$x(sSWj{nJ-|WpsOi8Uh6|+swFokT26UC{P)5tmki?hl~b!m_NC)$kRrj@2f zOtJC9q1dKpnmm`HmQzTJld zv&>nM3kAUNbSliL1O#)<-x7lJh)|(7wv7Gb`R+tU(73D0auO?4X1i#kpm&1=?w5UN z*cvsqQC`VWhlnj@06KtNz3#i=7AtZTNS6}&(W-9*?oH}IBV1avU3YE-LY9IEuFSm$ z6c7&|kD`;g=EKu5NYgzdmmr2LqQWnpdn5fc7u?>w>50}(hkSI%HItZ$|J4rEONP2q z7JxLcbih0ti&wxQvJYQKnoC*fQ)!z3G?+#Q{PVQfkjBu zNE?3(0Pt~PfB?S-nYD?AT%GY5(2`?bT3x(*hQYrB9D)W?pnfCtZ*?x&`Jr!^ z_fk1j=cLbsLwLPqrjGJ;0vBTxU-vW{uG>@GP*s?{*+oR7&3`=$1*{k@N*Cw8md3gQ zxQJ~J|0)l;p(`R8)iI4Qm0UDkRlmmUW5gfa`T~|^OPu&Xi}1uiA0~^9V29+h8V65aPMcO zDkqhEkx6u#P@(vji3{;#GEXwA-s7Dx@mIvMJK`5KKD&C8()D;ELx=H*lau1??#!CWz3_jHQm3X)vSe_sqfnuz#%Yiw|p$OFnI9i$7( zy=?@$pa5Plizvpt;7Eh73}u~?mvE%vu}rjV(k|(GV-8bWXJDW1hB9f`7vVQ3PoT6= z!K2^Ey??GsIL_@8GNT&TFf;5(%tI`+VMp0392(>dplfq}yyB7Fz{u+C++kk{0mza; zqa!^?!W3wWH8fPFm?9~5dr{FN92#`RXH#=J#Wqu6neL6;}g&)f)@eY;wp+jCUzS|rXbNo{BoirxxhM;7T}Fz%Xjm&cnuLGmd&Y{PUiva)&n zDjLDNpdUKI3`f2danFK%XEc-DxJ@^+_b2y@3H-^g*Hkwklz=~vZ^;LZG&=R7-Z3F} z#ty$DaIqOTiasc@0OYnF@$A#PxV@1*!kyrqksot(h2t!Q+%|}ZMoQ8`qW1~U_6=G= zva#+oen=P1uPXar?Oi}lSI0-#7twRIP2mIg0zrS~UPvCo9cScIJc~9QQP_lL7 zx7$do*Ms{%t82ZENcIbkzmoCiJnPL^4tpJLY`I$3*qx$2vW@@BaJ_eVp*=J_Iw#Wc z{AhP!{4LT!1lB<{Ay9T#dz?4@WASqp&@u$up;(h!VH94`2~`?!*K-~t0${{`12E~V z(PgA`n;;h_toq%s7vQ=(+Cv%}>}lCXR<@@z+1lh?CspsS(AW zN!DnH-+SZdDdLnUZe>Q|BJ1E0(*L)UTaD-f!v|7}xRo1tXy3Z9hKvERwpkAyYzef0 z#yZWnEj7J5z?z?+m9RbRrEr0(yO8vj$Mv?PcoTjP-w+vnXDuUpNk}Z+JqK9`20#sB zIS7_tX56Q;izo2_rqPm6DV zWU>@+ir{Ga4R{Q$>uk5Q3%ik%F+`kP-5v3)*7stR!AkJdZ^pm`kmU+TBUXjpWZGFb z(L4_y|JdLofsc{Md1v!H7q5+UCO$wFd@ub+Tp*Q&L4MX$^~Y=j8KVmBl~CtLHIDBlG&K?@sXmas>U%{wO66!M)FdDrdJ>^-EV|j%l7#8XG^N z{t*EJezF`cl9dLnSVZ)LbnXUhQg@*V4p*rcpU+B13bmufsbrgm(K@`p9;*7$RDcGE z;}`+X1-eh5bi(Oz(Hu@gY+@UXj|HRH8vR=q-KRM3XY2C&jKf*elQ(ZU8tsf*;WtvM zN0f}>895V)a+CHx8iFCwP3+Ap4Y3_J&`PmVGel>Cw+N16d}d=h!d^;_xC}zZ5#ZgA zY~-q`ZkDce2CViB*NX4A3iB(;AqflwTMzoi28pUhtNV$T_uU%{Djmtkx`b)@@+sBy z*pX7q(Nu$$lS9SZCjKQK05+$nTqTMoQs&)!Q(lwQIiM}2`QgfeQ!?8Sp34Q)et>3bsHnrT_~+tQ1+7F@;9-W@Psrszm>FCo3`lcTxwb%$zJ< za1H|-LU&wxx*tQu>_h3rY)Og>Dy-cyYe$Vp5ij~F?448C=??l?9oyAb%dCxjN_#XX z2&D7C3@mketz2d|Jje&z@)8%h%rk4~osKcp;0B$1f6LtGcJpGicd_=@)GONU?ZOAswfy%|dRY@xq2@9>0T3C-^(#a!b{OIun)o;|-Us;~i27kFLi@5k7Pt2%i~I z=WLN}zh-BB?E294RCxd0n$~f(N_i{;uMV9Iib&C(MzN`oKkvl6-~a7x;>urgIH&)< z*w}2f$1NFo)(ArQNtsm0jyph~mkr#5HmCGkeHJb(3i3WwKg=Iak|;sv$i;7}(pssc zFh4Tc*nMxW6Bmv!wFg!pC#ybo+RR?#BQp)Mw|NHOZi2H&NSPBSQh0h%DNdl=5Re8{ z)&e*No8an?mL7#NK?PI0fXJxCOkt<*5Q5&T=`G!dKI{OuN~F3Y34)J9GY38@d1{Q3 zsMsPhwO<7iJpc&1_cBg$Sn*d-Ah4A$YdDzWl+X15QHSI?C7p62;fv4b_KhGubK`FUEEuGsxmXB|8e@>`(0OwKx> zsA(ARCR??o0Z7Sq)MF@$FocQ)EMh)ft?MDrCaU>Jg!86r101j9lYU4|hrD1Dt}`pg z)Qz|oKE<+JY9$$+DhD^%P#xv38K}rCpA7Jt(PR}`M`~E{JV`_# zU0ZOpJO5csk~16Yjbsr{U?l}6mP$x;i&M|ZKe8kBLRSVn3OEz)fz|u+^a4qBr>AO`ozkQhzMW1T4HioGHjT@M8!pC_!VhQ|L#B! zZq`9X^BW*AC_UfUN9WzS`9j@Q^hj<8cHMjyEagqXS8QRoKTa@qjK#2zn312@W-sr|!b>zSn71wk6`)j!O{mMn+QOG%DY%;ue55j%e{*#8 zrTbI^?r}S=yUK)k>a$7G^`-P$?!gosQ@j%+1)8~A(kjX0BHk3|x5KLw@Z#~2HN)Wg zZASh2EW3Z>3to)`W2|g02vu-nTEZ{i5f3zyW;N$DYRQa7b}{3pN2$t|58R|kT*js7 z1`nn@rlVg9qC0&WY0)b~b@iGc-g3Q7T>3ovg&eq6_5!Md8Kea!M4p=Eh_xZZdqC)> zTvvZw96FCRZiJ!Qu3b;0cMf21wJzuE6Kep8ET)bn5&# zr+GZHSPiK>XS+*yH6_Qm>lEuRmv;0+OW4QfyhgNpiz=xEfw|fU?z6J^5=C?;` zA3C!Z#E;(xOIx=cbe`XoS`-*A8YUYOjC4W2mRX_K?KccdelydA9*h_*ap{mNNE5}E zlEbJFX*8gt)qVp`TAyl5sn~1=+pb$tl1B87zGIH0w`MET=Mg3I?%YT=`;l^*RBMx^ zE4O!W?!fP(2umw;bGW#i$?*`kummCp6B&HZJC@Q617tT}{oQvn8b5x+=<%VI%b2svCf*o#MJrZqOUX7o*>MJ>>i zxXi`eNWN`JHdy>#ML6h4gAX@DE{EIiyzF^W)G zUCWT<&n?
PN`_zj?ue>=5S*jx{x=sN_H+Z{_L8-!D%_zN8;C~Un)PN~>WV>L?A zWUG!7*^db@*kIBs5kw|AZB#4$GwodSn+xo>Bqj<(1)>1O9dov4{&2GXZu(~m#)_ck zRkH&jHCrCGVy_|hsUA5ah+^N@3M;<~G~MWh;{T>*!JJU=|L){p+vglp{l=;4s)+{!f{&^@-m|;`GF9Gc6H)%Ljd)XAZ1*Nmat>7 zYMAvBLd2^>rSHl+sJ+n{m7e`BIG%Y5O$roTuCOUr#K< z!kCXP@;Z_agsI}0MDD7S1MAl7z2?)? zuirvY2aFKt>=iD0Ew&y+YF_?|b|igoH5S|>k?GhO+SRR3{ZNiE_FMenkFUS9Q<{>n zf{$F4@T-lyP=}V?R_oOhR&K{2mSt6J!*W|}_WYYKhd(N(;nP&(R4!6X>fv@w##hAp z&EL%T&vLJ%&H<)UR~J}B4j~GTdydzuTR*({(lu|SPZ1@mxG|4*_`^yfAi9)>3eYd} zX0?e+4#_ni+>R&7-{LAMop5Xr)}YWgA){9>o z#%Kq@_V%i(f{MUSNdV=g{>2PO^6Ia2=U&!S^sDo_{--umP0RSPwB&CtD?002q$1hc zG;%$I=d|_S-SEdd=SzgtCXVfDr|9|Otx=Kl^2hfkjS&9QcfrQtV&;W!zx993W7Esa zobx99%H&uOciW_FU#59k;;s3xt{O}DbIRrE+*7JkGYP$<^(E~m#<+Di1Z+SJ$75;T z8(Td7k2~uVfaU3)4EbQxmnwSSUAONq?CT;rV^cWJDA%66KBCUKIaAFeKm*oWc^YnZ zYwkzgfw4a>km1aoQ7h~z7YW-nz<1G9tDG94QyGD0Pt4Lz*(37tDuP97eni>C)K-(M zX>XX+^qh5pUXQQTd==yMfx|kCwt~YX{(Vn$hPW=buwQQej`!OniCNjm+cbviX(qoTlB#{6;MVQT-V_c@sKWV+!utVC{f{6_);E{6Oj7({$2C=j^CyVv}%H7s6 z&nlNrfNiBpo&rbn$JMQI!H_DcMW}@c&d^|HP!tiE+>(=c4j+O5iWzQOrU&I z&q<5GwMzA(?^%)8SWOz=e~$`(q)CW#zWEbdn9K4`Utx+|P54W%s05tNMSyK{GBOyheoxM z{QYl0*4%FZ_5bHX6_Ct{3(AngL?ArQQ+)Z~Piy3pWriCBynjqk zv_;GQI?G>t@V|mvjfDB)g6VV;R&Mlb{>O*Pms}@){>9e3CZQ>v#pYonrPeP#o3WfC zuo1&Dl{v8OW1B1(txU6qGr062W? z6(Ib?flcM{(L~v}n><$(_aC$;4f!6b^CwY8=4CWpktn_nBKm=Dej$$*C(+Kos!;~D zz|_$Le@22q)5DY;y6{^F&}jR*#iwpFKu%m*RhxWqZYnL9qt@t^AflQ&vSa5QSC!%t z#%PX-tC5skA5ax18JK;}^@hr`8KZ_C$3=#$8#5=H8z|<2pmMkr6Q3TJMi;G^Ztu!b zgG3wqzIxqtrox4}1`c<%tr5SDWvi@8a7+V7Hu_EU|`(G#kp3qx}%r$@C4Do7U=bZp(u6^iA0m0 zFF&IOU|^)Xft8tNSBahL8O4;LrZS#CPYDJ|yDzJDY;}e@ay(`flTUBQ+QbEx z%Aje8Cepw9JaY^Y1k@0Zt`)Qo*hiX}O^UgQCrB)I39BR z%;Cu#le`D-tSb$n3V?m8(0!R0qDz_X;9hpyke-!>=2=n!JfDRNByRaWxJL^OUH24o zFKoo~l3Va!PjY=-6sb;G_2zQ$UHZGxbcXiF*lNbWg;WA1=fpnk$UJNIh=5FlHKNY( z8zS6c9LyUwdSLIYUdxL^q)38DaFBP*em_w{%bIJ3UQyVFo?3UO+%&g;#q zU)D*%j(^?N@E!wGh}b8}6tNq3q8H>p4YHCOOJC}_Udys${CIgiJFJzGwf}SXVcPXT z+YYM>tFcSB`kn2=OVaUMMn~-Tm?K)}zXgrZK06M%eel8j0#Azl;z2}0*~cS$oc(qq zYcY_g#%p`#B&gSCKJX~Ozs^dOF$A~B`6)E={dSPX%R9H3R-erFi@27Q)9^~a&o~oz z#)_t)^5SqFUjUVU@VAntA}b)mx{LE1-&~&->3ZTlqWNO9v`b#sqoRo2@#{e>i|6x7 zIxrvPH=rc}`|$OzQYOlZuofmy3Tj9Z4extJTW=ohXhaToKXbg*T7^!y)#{~7XGs6$ zl#Eul&VZ9H_ZeHmmmBh{YG&pLV_+BF3lbI|N{(dyq5+lK7c0J?ASF7R*aM;Y4D*Uz zOVi`O&~CMcX4tP9YdiRT)m_fANGbCV9QTk_k4;ln*2Lnk#((nj;R zJ~!ky(ym1Ot;T|R66|afg9`kv1#nXxpye9W-iek zy=-G!&R45Zbiqw|Fkrq$b*GK*RIp1Um9mf5q|sF`N#qfqS#xpp8Ue4#^S$ph;SD6f zB$x73}UJnTjcF0fRSf z;}uIQdnS)98?~0o8q(3Js;AV*aWhQri<^t?`p-G*MS7sDjTL3%Z)D>vAFdq62Mm&z zFZ#fR2?jyPyWsR!ReyLp9w;0X4F2I$p$kyjr4aPswh;?-yLx*V=sp3HtPG-xQ6P3N z$E)Uw?d3of(j@#F3QzhhdbM!6>IlbFAH4ZP5T{E&!ivmvhA0kzm#2-dlsK&hPzX1+ z+=v?h5p7+ce?xT`&&5lN+B-6QvD%Zx8&Z8GCiIyeuTtQjf#1@NvXAqXoHb%*C_{Yx zqlB*++EOKfvH6H17EHZ7|A*GnG^BW9W2MGo0UZB#9Ev&8@(}eO)UPkY3~?WW7CCuv zJK5u}sgQIjyswBCvz^%LHGo!}z%T?col z9qPonJvqb)zknzZ|H41Dh!wa#-Riya_#lnW{Zj3rfYTX;->u((bSjRm67|*FDVs4F z$*v-)=g?lL2ep54D9>IBv%uYrZ@&Rw!Ismms$$Tn?POu5oGb)CsoK}R3)D#GD<-?b zg-HTl9=W{82t#R~A+iTln%^;r@0omtmT1?&s_dpE(qGSnWHoV9GMT+nreO&)%#j71 zKIewScA^<0mk-EtwMrzhJZDQG+K?iAv=0oq9gH-n>k)uLHSDlf@TQh~`Mli0pt zDt7QByo~qTV3ViME2iKPl^_hgSr_pxwvqRd#MW2e&S8XbQ0#WH*iujv5SjMglr}x{ zJx~ZIM?Zh5M>9Tq_*c;X45@s~J#9 zbt8dZAUPM!#U4G*tqI64{n%%$yuO`4=E5~dO<#VM4hWT4J?jqF?*7WUSM2r0$AGfn zm)Xg>&Pi?T*e>qG{q}BO&{%^ElyzS|Q|O^vcFe0-bOz$>Qx5Zt+L48w5B7jh2GgIp z!6yV0TJkY03aQ1zI?hVk;K901CKXdQ#*`@BgDM?353=TSmyCkWoA`wX^T*M}W8ber$2~LnB^AjU)@M^4*uMD8F`NnXmt@?fS4yD!z&h zq+gly$A?whqOWpDAE--zuoUt=%3qB+P}CqhQn#k&8y>Xf8h8sE*o!JnjLxUI1IwK$ z$M}{EMCKt5?KG4M@23&ciI0cUN#1$;bx}Y$5ne?MkPXtXql3zp?h6mt(G5}6NuqOZ zt#MX?DT8W>PV1gdRY+a>^^oHAoLA{(mbvUcidt^ks8sv0X^!7D?~%(J4&Aw==2Fwk ze1g?>j7UpHMQfF_5an*Ip)J^Al0M&CJ31vK2f6_5hQ7V^iQ-jWxyoHcE{tp#JZE^5 zbDN0;!eRbY6wxn3kt8(o@hIxk6W<2q@3=*QZqZ0Iz2JD=m#;V!AuD)`M*?CSSUr-x zPyW_bn$DOvQU>G~jN+1tcRi@?dQc@kDh#%j$y$3kzxHwtRYfB^{4S*Bh1Ec=$2-dp zhd5}9zi%40{j_;esm;HPY;^MfOI6A_p$-Fj?dVasnzHgd{P35}KyKge#+Y%)PaF&e z0qN#@rLCRT;hODyf%Ppo5I>`Se#1k(kYOn=LG-W8+1mv~Ag^JW%!ZzUjGTWesjQ+N z;N#T8p`tM=j(}-eag7|sgbU9q4s|U zY{;}3srXkmXGXRv+G{fBEX)0~W9Q-u4*BGmg2*8&4@RpPd!nd^i$DJk**J!HDTnWU zji!5A*5sGJYW@aDe+47PmzgI~aEf{jj51GJ*k@vJa*`$ZdBDz$JwG}tn=bg^zFz8T zNP3-SNB2ydEl_rkHEg5hQw1pr(>=M)jB62(MvqTDaOqOY9xFfI5asXe?qLmdb>Mk_ zPQKrZJ7=QuoLASy{l(*{Y0BjH`at5Tt#2VG6(+BZ`C`(N3j(kkBg zM|a%0Etc}&`UUDKeeiMebbS^{ILZ6UnPiy9p?7* zHV)1ZHV8#SEn2U407XJQ`&e}Lu}C1pOlO8<-a^OSJsk2|wLp*r-(*X8NdmI2@H-AA z5S_+wo4bKOm& z#1y+o#>*&S$vl7LyO?o9%6yHuZl6P%I-q~`MAO6pk!o9?zeeOj;(AD{|2DkXd>gtm zaitLsSUQzRl?30+2{Ry4W$V?+-`enG*$Bq)nGa-2cYF5!Rhd`Cr3(eO7Ar{hFagE> zN_90G$+d3V!;K#Y=f&mD0bUUoD>MV0+TWqbh!Jc!!5YIK#hKlu`cJC*0XaoCb%hQE zj0EVP0nTpR-w($YB>dXaDS2#ZOP5I#_PK}x=~t53l)6%26ag488un0HZcoUH`Yu5v zdO6R%8umn$^KLq6;R0U1-j4-&-cbc51tWQOC5phXW$n8Y{6mfi(;PU~*txo?GGiEZ zpioa{d5vk22XevBh)>E=A~YC~_9D!3eZWKDlFKmrbMl9XfYK7PRjb-(KSCHCRRX!U zOvi>eM;xQy$jtlwT2G$)I&uES?>-|_WDu=NRJsK^gdSY5a^%B8htN%uO z*X4C3X;p)d${1TLR{7=(&K1<`A#Q2;RQo#zqt&g7#+ZEo_OQ1HB9DwyG3F0P-#XV~ zj7Yqfail7_@l}Hy`^?elvDtLD`*DJ(O1pVI@x&)SW=5H--JYXlA9v37w-Yb;8^1`9 z6l>eXWwAgGj34}3{NFIeWB?@~Ocsj+k*DM?`u+M3OnqIC`KD>PPsY<0`r*If^8Iu2 zCon;2`Q;(X`VHJN3jmdC-`Oey&2*-{78UVOIlTGP2OXIrtW;8+eBFxh<`g6S{6=Oe z1OBs#!O1Kp`*~v>F$&&`-+(t1ivc`kp9ep7?{!>i-Co~V?VtAoIW?Z&d~hc)G3}lV z?UPS+E!{oBC0?(VBs#2K6KLah`fful1ARM&E^i5NB+_0t4kL6j+x-oQ(M^E?cFe@N zC?aE=2KiQHR-$EC2D$*eRZdaBv5RjNkKC3FqUE~ozw<2D0qKp*Ms*8r`!Bsj>d^c! zG7NUpaJy+Pp0?rc7j2j3aZ=^1|5iLXVQS z2>+~{^bIz6#4_+02W3JO#f}o9O7 z9gl;TxCZcc=3=j2b{ zIGM`->G4zuqg~!sb~%$Jz4KR4{oaZ7fbBneh9*;(D=h{M|`TB3bYXXf&&5x>g^vg4oZhe^+@%*M+QSm^3yoTj+QTy#hF!HoC zX0Veo`0;5!ndSxi>_Cg;_JTGCA&b{Z^YW8!CIgD^mghY&=F-8o&vyBoE~oyD+b0QK z#$lI~M0LmN658y!>0VWlCr-01vWyA6b-QYS*rb82)!2(Z(k`n>*yMe?>th@3>l9nE zwWHsp?MBN8RW7N7pebbMf}{1DGsR+>A?QAzmVm6AY7fh_W}dp}H&?D09C8O_F%nBZ z;}e!ayA6RFIrG)^6&|GLr@b>4{J)C<{~#Spa7%=?gr2Y~<6XR^E!DW({42BWO`I)Q z3up4hnZH-g49?H4SZ}JuXX9pi2%*A74Pyk-MLvCI3gj!(P(Nf#kj+I0Le-ub~8h*BijpbL&*i*TuQ% zDT9Cc&&p8$cTizv<90k))WFYcbE;J{L|4B2kL9eXd=D>xMW8v!G0wsneE;9UWVw_>wOIzh#<1X$|J z`frz=DpT#!;T?EeIFvX+2uSrF;-|k5X!=`|uLj{(`K;FoXjrZ>mc5UIt|*HTUI|ap zZ5+7n_+ki@=Ey9OoJsh8go8LBhd`i>c4b|l>7UB z(~OIM`1CiqAks@uFyUo{{C{QUD$E+AYx>mR+g zfocuSe>F0|Oi4&it0~L#-okaiVt9GIKkGOjI&J{%{SO8gVoe6?DIoYU_x`nb9K2Td zZYy;NZ{dI?%#(Xm`5O?&V6Hb>yas;cKY@I8Cs_Y94tae8e7d?S+E z1Ri{zU$5>q4z2rT`6GPcOBoCoVhC4NS!LLLTbahy$foY#Zz)U3Qtjq0K!;mYm)&+N zyX_3)#VtT)I`sE&NR_g$da<$&$C*lj#kd%gos%7SdpJZ2Cz24DaOxt`G2@zkgcCR@ z-BrNLwS9<`i`@IC2Hnp^L>Vw`2|UF?cKX^V3MiT1GA}#4k^PsYI5wd~4&}o2t70eQ zJv;m5KmB;=HBcfM<4+4&tx<83b(rrslmnZp1q+@hTuH(~{B|a#b$+};9Sys{4(^`A zVcKbvi*%_(xIsjm{V#nPQEV?TPBijzCas}~9_Chb8wXeI8Zi^xmUZDwAd{B0ZAS_6(&<(f`?uK-#5TUA54O7MQ= z@Apgzr8XL_<1-SzxW6iwsyD`2<>sC^>tT|R8D!})qx;wLap-tuZ7-y%H6Eizr}Slt zS#*k71bl>s^axLCANsXU;6O@i2!<`;;xjE)1dbn{pkK(X`H8zwc(w-h5*@Y`K0ov= z^ZR}zxYIT|@%sI%8n4ZnQypbm;$DDIwRQ2d9b43jK>lq-WU!4>`~=Lo5u(zdrn8@& z1RzGmPLBx|*RF$M#zrT9J{F(l4d~E59N|xi;(6FO2wI{s%JB>la8v2}P>?&%SSviBq zsT6^&y!4aPu)`9mPlJA4=Uq7$i+4s|y}4k_4XNGG_$ou_RsHEWUBGGU?7{p(>p6j+ z=J`)TTJ`0Pk&oT)dU@(T8Jn!XD%&{@->&++@$rL2_wgu&?8`)fs0UU|bV-`;5y8=o zw5QCZ2E@ib#Rfj~1I>{vchz#NeI7e2GYU?9+1h->RE zUV^o3AG4PdN3miBKK+I%B2QCExr{nU$u#;xj^W3wM~nsw{>xgGisY=eiL1ZgEW!vMNN#_~jckkJ6RjWn9S`CT;vhs>;+Z-EumYC! zR1?J^LMs7!rQw8^Y>BqJ0KY9sgkw9zS<iluSQQhA9Ft$ch<%uF0J|iwz4i*7i$LjQz zEqPgJHDb_RN_Eh{LOFER+eK|Ox|Vw@j>O6%h$uQZV=+l@_^^6qw7$ZA5k)a|xh^T* zh9J7#PcHU@Mot^Nu`p3gM6r`Z0SDU*re^lC28k{upF$lOB6>p*ynypYd8fLOyOHCF znQ`-pOU+YmmW~ij-~4_iruL53eTF#5LlDL!(2;`4a31EDN&);$0Pd*uLlI$PgjRHX z#&$dABVPe15T*zvv(521lz{2>Tb@+6JgKjmnp$0RIty6*BFWGr{gEqBbSx-ezy(jsiL zfd07rHMA;h6$wk|t?3li1Z4bEKcdf>@2hW3>kXnIx?j@U@vr`+C4cF5T<;CBtYX|$ zteUp?HTkTFoMMDtHsE3AJHWR#Z0i!I4tfi)h^oL}MJG+O`7{P_5h?RzAnp!fU)-kLeg!2+rl3>V5IzFi~iixbFxAZpkg{R$a z*OYU4Fck?^vDIvaXr_q7cRtV98lU}CFS$YjgNDkYXA1&yN;v8|0|K5BMEBtxsd@X( zDDw}m+RbfpszMveiyZp9sUxl93*d3^R-c;U(QB@7Zd`v`WwPGZq5%OFmoe{3Lj%J| zXrt}3gi3Ek5}uq7qdH4i3UYS~$_QeZt=&2AD2Yty4kKCF5SgYkvma&#NzWOVsbie? z;^g2~SUPq60|{U>M&j#P4c$AmdRla8x^-5N9sQb*ac(?Em<_RN(E7oYAajFQAzh8P zh7x7_G-j0}utqQeD@KH0oe(5u04)d!62DZ5Ib-;F9~TkEiasFP>aYnQOu7gie}yA; zn2d=m{DtpCKLBmW7p2(w@j{pFUb19V&N+Y*Y*xg_do(vk0f;*@4a*0vR6nVS%d{TI zxGm&@l2}IRdFgyf;bs|h7y0Wa75NZix&Y5%wDc3iX2^d+aKGX8O*QcPWu<>?Yk(O~L+eQ_{mM2(xl717iPNveMhJ31@S&eHt1n5QdVfQ4!4D7Lx1vaA_zroF4~3-YOCuz5j%}Rt5k^7Mw5aL5U@hjsAP*+(xR~Yw8cK^1TTnsU1-5 zOV}dQI@>0iFYAt0@MdNgNe?HMvWHoncvUxBhOJJ0KR~$1B{Vdgckb2Uh=~^RnqFN| z&|HJ`idZb`x96vecKhv^NM@SwJYsl-bNrF2H1b(D4=Mbl2vZ|fXfc`ZJr`(=a|s_6 z8b}^gm&1uZ7!zv}&%<6H>f3WC=hBIo;^&sFqd^s`mg1N>MJXrWWlww%Cv&=sj-d)S zDa2wf_&KPdplgApsdV!Bk&Lppgd4ego-<2!I>_8qE~0U|7ESQ2u>aFx0r0GCjju;o za5UFvQEf>aw#J*$A%Y|dZ}cY zH+~`Jo9f*<6%JyJR_^+zca}fKmnMT!EQbXRTQlr56NU2RML3SdUKT1~D{N{hdqc># zi=(u~Fx5itj(l;Su+oeZ%-OO4vJ`MsypM;B%QPE|b^sB<(QCiwGE-o%5ibSHM zwPAGIz=4v-#yv0VRW0^6A?4)4XZ6~wY#-Tp6z+ptNXTKADZzh;oFzO!4j!v0jt889QMFFuf1uy9!^k5b9{#`eceiIqg{Up$=@r z@RU2W!EXGL2z#To`Ca}2VHH|TqMVOTR2Iru0V!Q7NTy8Sc)Rg`RtoRUdeJg(n_N93}YWSi8mf3+Bhy6_+H5mZ<$? zQ<@gQP>#!hHsBeH#w%%5mzbsZ7(j#ySt_s!1`o`Twy-J>lW|ZE*Sc*F5oa4~ElI@) z@bzt%M!ds|L1%D^#R^|cZjY|XPO`~|7?9e}NW3&&5jAEa=YUhv_`-BZ)Z9sGS?$ta zp3#KWxDn^sN~!uFgbVCRVRETVq>EupVCDxNv`9~U5QPjoMM8|BZW_x4XZlsqQw2}+ zW=Gal+-)N-CaeRUOkp&TTQ7SqGL8kUq*N0J6=Ya)h{s-Yg5d=@ z?2L5*DQX!ztymn;zRiV_non1Tz{EeceU-zGyYi=1_|B=_K~b8`Mdq6pd&0{4-H$-cK;dm zq3dj!+Lu&8_DJp!Xv^B_^KakObw^&P$1C}~&Lxj&Z1U-RC$YAgCmU0Q^?t5OKL~=C zGW>Tvf9v#(JttJaS^53T8J zOK`jaw#MpN8XOj%zYDL-Sbu1i{jFD~uc5o1!ZqY|jc5S(i{GEeufNY@FQnuxRKJbv z7IU2ktO2FPQkcvObB1cadhwk8?zwz|bco@S%Wae4tRPFfFs?IkW4TkpqQ6x$JnI(l z7p=|}6bjpv_*?fIjyAa0BY*c%YFOUi{h4{_LK2$07xD-gJ>RuZ4Y_fmU+MiO9Lh$s zZ{M^6&PtP5Pm5gL4GyVO8d5hyOY!0{eWHrwtnkRw0j!w_F_I2yxE`0Mby5feCPyKvKkN<#9uk>c6o)lkg1*C_RvdEyx)ueZE zRP(S3|H3&v661hT!j{+A#KgCOsex#^j{A%4M(_TB>dmOIIOVIyTg7$(q=8d#yUL;?6OukjC-Y?C>6PrKQ_Zu#qscX>%F6 z^%<~a?cA!?b6dqB5xjk^x*jW!zDlp(rwfKetO5f7;Do<1c+VGxFe+$AgS6f1E{-J? zvOPI1qZpJ&r2b=A4{)RUu(|w_#sH9e5Pm8rB@t773QU3N zTN$|9LZ!y3g2KH79Hfi_l$O~aB9&goF$!hlRR&xwtd26FM{SiwvSA|QUMS7cRX;9T zrS6Y^>$h?xDdT~O3AgW}@2;{WtqpZ*gbuD^<4BZ0M_5>j9N~}+Cx|VR;mi4g0xG;7 zrCYP%1XIVM{aW45*9Jar|=-pP~~ z&FbBc*M9IJkR3563pjNZuj3>kjne^&6B$T}Rpcfw<3jA?$S?NREW(in#A*mC%f-@e zmT68frT2aV-WN>FT3#X2v$L zi=Hg|q$t`eXU;N=!_#dp43Wd+&ve0QI+DITCk@zQV=C;Mgvpe!+0x;jZ>>nQHYLe+ ztD1Z40xGHn@cDD=7XtPwyzBF=rC-D`w?oqTVg!%?XDLVLVdIuWVQNemAaEuez zCvLu(j0wV0L!)M=j8GG;&VDz-WF-m84<>5dj5&pO=tcr#w~9n>84sa#el zeHGdW+yjbbd*vvW=GZDqn2$Lsm>^6{%XE#U)s5}0^oiM;>uO+n<|bE1usUutZyX^e zWsSrhUZj(*#G{^W!Pnf8JsbJn2fmp`tr>GglTHfW65)a3a?t4S3H0Rd38G50yqZ?^ znP?I-YP-$@x;^36PH3{2EZS|Tt#Sle1^IZaSuDH^i8X^ntIxo3lodx0jHS(0RF>ht zN6=`AEW9&<@$#+eAfpY;Oi5V<5!|XCLV`wPr)pu*hLqzZ#Ks?is>auI@zrf)uDa-Xm zHqQMA8#XK8hx#uovr(i>=AU@Ol8FHy8SxZ0SDeJ8|6HZTQ{b^K^tjF~eR!P~##I6E zm;769qrh30Pxk>`XFgd+lh8~IjCPUxReB(zbtMwWyJArGh!u;ksa&SQILHKD1seO7 z6^y$L3*j+qmf%Ga*a+6E^T`Rl(L>GD86gTpb|M1rnbk3eZOaQOw-H8i+*JDG?y_$s1g-?Ia7DaqC;oyTsKkp3b4$A*@~;csIM*~KVS=PrK)Uye zu6dQ@#}5RGg5UT^~zhUWE8idq;KQuDJ&4}u`-2$nr2bqR;5 z7O65%_w_X}bK+nft-Z_6K^A}I5dD~=jaOY($QhAbOdlL$hu?!2s$8n6OwvM)ck;;@ zyFvlt!F&)2<&q!K?BPW}j$m8oOTUtOO0%bL{dgq{al_nazKO zny(&=$zZ3=@Ek)iey*zCq(FOK{$>m zYQNlX+J6YWqmfs&yznOQZpHVoRWgh>sz>N6!p}g&(qYJaUTeSjxhWhkmAmzmFsHq~ z4qDF$7Wczae3F+N#5?*j3d8Q)S z?v15ML&{0Q@Kx^C&nv!U-0}&+xe}A!Z85Q8sQp51^Et&4bl$Q6M>egr`=i^a>*8%0 zH{#E+Euo}mh;;Paq3qlM^ZTeU>%>J#{=ge`Q`1=ZmZO@y8|bb#!FM-WrG@q11HpH? z)XB6s_t@=j8f0D;aDOJ9;!qS`7L;*5Yz;8S@7WvVC8M|*DGw7{C9{mP96E~yRLID0 zru8$J4lSf^+Lu_}ip}=muTFE!6Dk%KzOTZ5{+X^gB$!8yeLed{=Uno)azkMzjMEQLyG_M@T zTFzwMgS_VhjL%YnlqwhTLj}EC98AIj%FX3KcD;>vjt-vT;*m(*e$RHy_2vO+aMFy0 z9Shu^qO8PGY#5}rYpC{RfS0X*1XmAi*G^-s6ZMk}JyhzI@>%<1wZP%2(kJQ7_=9+u zNZ+uyq+&6ce3{fYG@j8?fiAQ|QuSCH=z|IQUf-aw~uk$s$7Lx_u`7;zu(jYm_20E~dNVP-OP?)E_>+U%S`vc0ZV1kznaa)=w z#RyUBEOBj-cBzSkA~BI5LVYu~VLoDKP7Ee1uGGY^OUnk9RFdvS*td}wIeD>XA6N|x z11RUm&DFy(9+wmJLxIb4}T zsg8wG3xCz)XlpX)+SKCmx=VRmNj{w4^z*kk0Lde^&8NlK?X}%+UihB%1CVpl$9nIv zhroZ0)!pZVdU+0@vZLq$A!qeb&q|fSm(KJxTS_^yM8CX50SgZgt}py?03J^Bs(~#c z=~X#wmD;PzJGjd`xST~8k$R8lnlYJh2DonRNehlI`q3I`2LnK?x5{U}Eh0%rl4{;G z;e1u`LkYA%VaMYF6Y|ZrfX>Z6B&Xq5R}*eDCQf+-e#**d?PV@~3_#I?M)ZZ@(ybod zZwJ$q>=aB(b?9mCX}1*+rgNNi@fWk3`1%}5l{0G;09X@;c=r^#C0U7|d*6<$!ZmS$ zl8p_xivo~BluC)y=Ux7@3dZ;X=>E+JD1u0TMG}V^7Ao*+{dFda2Tw^&`9mZ@p(Gp; zTN$^fOpSvTvxn^k=dn5Gv3fw{7lo&?z;!U`Bx@mZk0-^vN7-F>rbvc3rXCtIO9-bH z3Y~(aM^&{m57LJ9h~_|AGmAkt%&J?wYP4u?f0XpkD(%(QN;~QF>PhX7HvF2_eUxNS zFlK`FKE2_b#eLag=lbX;Fx7#_o<+74GEOMVptZ8`t-9Rpf|dLX6Caz1TfQ1pTSJZw zd8mU9j#=(8dM!H>%kBam-31(`HGTs z?N%_E-P=g=WiTvSL7qjKID|F}%^v)A;q^t{?2k8`b=zLjAl8l3FgnwoxVLAh%VIgE zh_Y8ls3G&zHdcSNgN#M)1|!=_g?MK#O!NcU9pAlvaHr02tK#PW1<(75YhFzt)8HTj z%R5VL%^g?s!T&0gI%gsQeo&JBd;vMNWz=~^2(<0PtrT`nnIGTC!(i@rPUI<;is1?8 zYchDBj4ZDH1v)Ik+W1BCoKK(fT%7J{eNMd=P#}e=3grK5T*~}MLJ>jV2VG8*(`5fa z5y0R%Z?s~ufBp6`>i;3;74H#1CvYXhi*FPDrmPocYuVwH%!cd#2|s8qeID@LLH;;8N?`ZvZ(yQ=^jNp% zm(q~uAky1@8Y#^G7nysZ-~EOh$CbcWQ}8NdTI9cpFj9xJIKs($Qpgg=0vwvO1-*G> zgYMBG+G#96dmAJq_`x91{RntnD$5q7P+!yJKS_WB+ku8nhTodnen-aq6K>QmBlGW- z5%e!{Wi~?=ycn}LK!acVLNRvOC>fZv`560#HE#BCCQqyrFbhO0e`pn1^GZF36L%r; z0{b+)`eQS>ND%Ui&Mx&aR`(OXzpFbk&OIS5AkV`*s33JOP&HS1q|k1%9wSs%$|o0( zHJh_YonE0uwJC=LamK9HlhMJ6@jfWnu5w~#q*Fx1Ctg9we~`zI4;5g!p{Zhbb)|`= zY{csPdPb&jCyqlY?1P@Lq_v@&6bu4d5YAhL+Qg#c%2(#lL~(>r>%WD(T{>iesD!7< z!IvlAmFO zTuJY!C8d7&6%p5A7q6?>93=?OpHUd)_7L5s_uesne-&ygV@ zZ*#A}$BB~cdQj!-6)4pDse;v;dnQCme?aLGqvP5as*y?Bn~|S?B15ia9}=~PtV#!j z;DjbXJkG0bAB4K->dd<>al0JUHBhNUh+qh{iCw0#(1=^3Sz?$_*-L|rEHq;Us~Wag z7JgQLk^hqrzD_RJ;v0>mag}jttaLf3e$D-taQaj?v`OAbZpUz&Ley)R^1zA@#i$PH z$&ttX^BxFe7_Qx098%bWiT-aP1ZUe$3p4pRGU5N|stA0DVv)7rxR3{PMXjQk$)pz)H#!#JnBPTb%)`yY)rB$8k|b&3 z9gP@18Ce31>JGTGcyg4S-ZOZ~~}e0;h~rzdwOL0li@tXXvRq^&pqR+9awQ)m_>`*%eM%Kz3iP8~ftM zq$)=?+}P{pmHg~>I?hC?W0NAR8c^oA@-_5r{)Xn{I+sd5(@h19RP|7>QU&S2bc5o; zjk1CZuT*)rFQ_HdQeguzu5#8XaQ(wnkwokPgzSR)%@!p-nwA0)weQo0iVk0%!WgQ) zd&DouMeYu)dr>@83hR~|78|8)`+jotU50k!59s0L0P5Su*G@nDC5i?Ov5prM^Cy0_ z`2?@*Hb(Z=C)T^{Ta#o*zBy;%V+|ldBZ%DD!D(8wFAEZFt_ ziNi%oW1rR~SXFAPOXuMFuPdT2K=%8kt$JB@{$E^}$^W|x^B>S%L2~>%gEAxnnYN!x zf+i=eyHMNMMNhdUQNLTU|9~pKdXP9Q_MHWrJv-ALi2fb&jZ^{&QDV+Lh+gqNu0J!6 zxJ30lnYWVJugmva&~80oU-2h9Y{R)ZP%il%riQV&$qq+rqgcxK`*-j)Hv%Q6^S_s3 zbTH7NSR6js@%L_%Fn^93H*(-^%=kxvI=`^<1j%=0Jsb1%%R)&O?@Ls{f1-rx)yQER z=NWJeHT+~cQpjIsze477hO`w)z#9zV6OMcGLi08<`}II~Ct432R{ZZBw$YrKCtRZ5 zKlv{QB=P3A_ZmdLy~KoqEYcN+)+PqNGMx7M`OSZZXBhZ}kfuVNb>4r-QaRShtI;)^ z;3euIu&Ve#sCGtAE8MxO_1%vkd=4W{&0nM!4>JtCBf#BmI5`m%wY zn(%sQ(JAvgW$3X29~R3dJ(Y}2|L;V6&d+9Z?CB`oyi)bFNTiofVgjiKnm7~|!<{Dd_B!zf05+tZJW(24D~tr(Y|!$C z{bg>CLD<)5ir^VG>=Q(xLL@FZ2e3iTGD|IYy%e@RD?+B7w%h8G&MI z2kw4pA}mMa%n0}C6L6?|G)kOBGw9M!3SF1C#vw=tLMnS#L{F(h_s_Kh7ct$J-9so` zA2$#onz9}kT&lhZFg(kXm)~baq|K_KVf=6|(SBK)eo8UPK7kkxny88<+v7J(+Um*| z>!7FNPu-dg@Pg}|$o&l2ioWP%FkGn&+I5^tLVOH&Dt-hHp+SDmig>9#qA_}HC%wZE zd zyFT^%@g$@^GY*U|o;_BsHY~-Lhn_4FAVMMYePc?rZ>Tp3eWmety+ocYeU`L?<8l35 z+c3GP9JMG;eD=!$iTf$ah;RxQ`VSC6O`|%`s~*S#kxdD5p4@jVuQAx2A0W~gT*-6g}g#IT-*swJFj4)+ljHr=jEDmP;%-L{d549=^o zidp><$F1iHaor-LTjzR3wGf?&BiZLZdC8NkVT{%HT|Nl_X0|^UA{$2^exuttJQEM6 z+|EDu6RuR)I8g%GK;VfCwB7E@=pmJpeOqy-*lnkm%}m_8vqTa{zV6r9FazZp=nI$TV;?j-;Li?0C0tY90Q+&5})xz zuHJ1YY4{|=VY$Z7kL5V%Q}VHI9qjA$W|e`eb@QKnvGtp}mf_;FLj1);=+CnUXvJs47z-=8w^#`GX zW|li2sT_7Wjv+1Y?@s#Gebf!M#WH_WdIvtHx-=@_=CSd)EY}IgIq4@D!eZKfUqi1s z;Naab!gB*p3sGvj-xsfS6T%%#&n5^~#9F+sM7ceFthW&yhD6Y@IOz`Pf3)JK=7|D1T|P+?H=(b@R=oaFG0ClDF6tk=s`yd=Jyku}`JKhJ*? zA4T-mrTXxd-ye{f^!&f4X1#h8<4Xy}@5LE@w1Z}UKOKxq6HxMI_fGM{R3)l|Ztglw z=~PI5QJ@65nQ_L}Z=WyfL>Gtm!$-2-so?#=Q{!!3Hd^BB#^nvCMHK7$`57x zqfP^Pf9-{P7ZHIt@LxkeJ*7sg&By0p^p8&!9rN4k2XIH4R@57g*Z{k#3$j+flHq$k>;NTLPvVGbKL=16GhMN) zFbAEIVQb@lAg=$^ z;z}Mh>u*0>Pd5ReMvXb2*Xbq}o#a*Z|7{bG4a1692cPkm7YO8_hK_@e*5Yvm;}<-G zw4jtnd`k^9;Bx-(K&h6X-$wl!>c2T=PQN=gl$wXHQ{be=jItcIh_Bt3S6|syKqZ&bfN3*%=02 zcnJw03Qr%o=AbWRkFd`py>tKLVcW;UhQe?SAPpH$-5m>DY% z{Z=0OMyliIjaOOVS$}K&=F#fQr~^+Yi@|_D~B}h37pmPA)KarlIT;6r;zw7v%5URv%$I0Q{OMcr`f>iL`8C`XJam`!x zCd8pxtPDU3k5{OlmjDVr*#sxA@&X$DXEP6o2+33)Kbe30 zWENkErxkIIj%NDnwm}|K!J5cn>tU;d+B93pd4kmhW#}p8bKBSk@OYn zQCydZVf@FjkMWK%b`zV?;=aFE@hrNImZha6uBh>6>Qug>Ok13Hsx$v%6RnD!^sf5b zOCs@+Z|6@eJ0PRY@lPeIstyl77ruRoN$(Km7Ak|&jC{(HLWfB97F<6wyLEgczE|)E zB%C4kFg~eSR|I8on->CNzu;qVtNnKGz$a5z07KQCeWKAompz-0+4iIvI?+b0clpvUgS zBF%#Iq_$b#Nc?WROZ`8KqiKecD&8Y0FJkav#*!?MCHCKdvWsajGo;;0iA|nGy^xp7 zP$jj^`ClbJ?Ui-V=I zBV^vvlxN#0gxUZ2_|V|rK|c;9PKkHH`gjjHu#vd-`CawzBMGT!nM`T^vsJmG0Wy#A5Y|DB+PzZ)pr7l*wPO3s`@Z8;M zt*EExcyn>2t7mBcfMzt!X!V;Nr|oX)@`_MIDvuc zb?)o=1NvU+I@|0Rg87$bUfHWzL+)vH`_e`wHt5=^k|kT~O*crUm&%_f@Lyi`pB`SQj7+PM)=k;?mz2mFq7Xq9z1rFD5PPEtI<<@6 zBFU#2k)$te4l~>KBr!t-jd=GsY^8qaWDzkJd@N4ZrI_2k(;igHXkxqi@H86wF_0WT zVUoB9aO>LhCe;!>vUnE+s#vVCCHX8P6%5RXPt=MDLRgAiXA+7SjqNai6g%lRrTH>9 z2j^>%kmN))q*`d3F8lN88e8LrP?1@N|33(m!{1RALj@Aixw4og*EZC zwR1NE%g&lO=Q+&fmS{s$cmuvoAS8gq{mSbXGUnsBp9_{}LhKP>2&C)p@cJ=~VOgI$ zNGb~qM);=zYseN@WrP3drN~v1G1~#m!ln8efRlD`ZXRegvO|E=y!68upbP-TqYMuB zUb!{kr`$%+&B*?GK6yy8eadNbf*zZ~26PF)bIk&f*mI@ZM#kVz6O2xkV31MzKTHW9 zD@ZoQMlBS00Pz?{ZyRqrZ!~vI=~hx2UV+O7`|DKRoKDa>Gi1TQW$5?j#W89BgZ>4q zkfVcb=4zbliC1Ji5EUFfgBGf^>I_JjkmPj`)_(Cs1mlN^DAm2Iz+@U(5(+EQi|xtB zB7@w?x@y)RJ&h?_FuieO%>%{Re$6x1vGtOj3NzH*2e=YWWrN>*-SnKgaG2r~qZ}+|998trAS3|N9=b#`f zk@A}4v!|>e##pG!MYDMRD*YGR)1w1>2L3w``iBuo>WfFgW&GdTiEV1;sglclVu7W^}n4pJ!y^??^+1 zq>X|!2j6dOaC}B5O~XAqmVWeOe{K-f#4C0G?Au!0hn(5>XjCq5sPr+Y zNME8UM4If7yu>?Eq~-@E;Hp^W!>$Hbeb*z=*T71YL8xFe1ayNVKm4z}5tCjb-EwtV zMqNM>ip5D)KB?Y_?~<6@>*o7!I}Mf}Rvq%snJQ);0&HmvRo{~40tUXcZT&Dm{)IaK zjiBGC{$bw#t4xDu47jA-;hjofx$A#bHCJs|Yue-tY_cc=>xc{1Z|3cO^7ryBCV?lE z*UDW+rO5Ia;!}G!+P&2#N3pp+<;MXKcT1fnRzrBAEo-ph>Dq#{Rn^nKX>nw@k-^P! zQ*z)CE8P9HH~`%EaDteG(a+|Oj{zw5CUsF5^62F4JeT*SnC}EYE?m)~9;>l7&Y3+Og>2#U>-k#?gT|OvR%(0~Te_QyVzN!Q4L#{5NRxXt z15dhMD+`*F9RtuSw9Uc{zjwQ%HHGJEq?=sbWuph!xiQs2{w#f*r9;j2;(aMB%00y8 zFPmIj=yV#21nl1V=u_1FzS=jq#~ag4OjRd4E>MUw@Sy}pm+$8GEZMEbzg1#0g2nY< zVOKQdBOO9gwbl&04fFXa?OhbRxWt~deJI5~K-sw}V^uY8r>c53-#?Vq>YrgX{OJ;|qR_;vz(Qs-Y z)Kue2*At_hV(KAzLnfwYW_T94=*eV6bdw)@|Gz-OSWC>x zPg=BhAQ7?|gY(>Gwcbs`|GW_nOa7bZ1gMuj;6YxR@b$KZcm6PFJ^l}9e}=au>@Yql z!~34K>+7;rB*HzFQ(4JgjAesy{nD0*NO~C3SPyUwjZw?v0E?dHTkp@0)Uk}D?z;H7 zJdyHb0dhs8Qu}Nw&NK%@A>>k5LhwvkE){bxdcOSDR~c&#ggK1~oWBdg2lBg-W40$x zAD)HOKHdvo=%ifIcxi$6xqj$3qg_OtpOD~shdG0GFJHFt(It@eW!>^fmA4aOYJx#Q zfW$O%_9G-u&MgE#OM5Dy9s;&_%0@_Wiu(X>RA7Sc^q71ry~bmCh8iPvA|8 zjIc|b7`u0IXXNF~_M?-%Ns^@g8e_SSbDmHzlvAueQ)At7Oa~7xnpgJvZF= z$gg%*U_P7XII_QU4qC``RQUO$Y*osnI_Ajn>HU6lcjGrKqO z3soWE4!^-8m>-7=Q;aghlZPD8j^5($G7VaW#yc6w%7QII-b0Gz^RmTBO0wT#J!cDk zwcal$LtXKf)kq;4ho(RtZb7q^KhgEU?={8yI9;I^!JMxic%c=CW`7X5(&fCW)k1H? z3sV)|(sJ#h95ifsex8o$sY~+tb1ElZDG*{oi8zwwpnv8qFB>62uRG;_9a1x_lJ{99 zjW3SD9TO+<(Xt3Xwe=RFJ;SMt!wq-6pjkfh?m&e=Z4)NzH;iNX!4pX=mT#B{yC7&m z?1MhX*yNK89OQUI=>3?Fz4jg0F>^5`tj1>>$i5Qj1{MG?LVZmfFWELmXg^<;m>touuxrZWR2=K zkNbPK(io;Wari^>6Htl^U}$IN@4S{Ov?n8W8vxJL2v<;TSWN zbRx_oJmo_y)2peIZ|Up^uu2SuoDebo5hlH+2sQHK|jC&h7*APsqR z<34zZ~(qBFtV8^gL6?rmIW0v}48*V!UYNC*q#8Y(=K(szyC^qZJ6v zq^O%_T%&ax1)NzRF=0}g=QK)!HFlG3u+k=ueimEcMO98KdU3DlJ2ji?eJ{>#Rz^H$ z*E3BHxdXgRH{BEg=mz*()KM(r-qAwxoA%U9Ee-;+hA`F?5;A5@`2>&?p_4=4X{E<8 zNC9_fw}fz$VyC;pP}fQk)5t{OB zpHzJ9f_!PQ;6R#CGH|)v>8jLH<#n(vtX`i);-PCKJgd$Zb&}(&5K4i`wTy4}=BLJAj(?t^-GQlEbFxcDcifRx60R zQHCefWf}u`TP{Yo^`aHjpr;&hC9{)Z$~>IY-h)yT!nT|GU^*0qMJ%pSy7fwSDIm+1 z6KcN7FF~>wSeo43k00ahi{RO2N;gHP=oI6FX@`5qip6ulFY2gDbIqv789%_fAd;KN>Ns z{2xV8Osqo{{N?@aH?L#8N7H>G_7JKjX1Omva6dUtq02m)ZWi zTfW+2hrEUwpGPQjH}C-te@_Urb>vh-U%vEDwF$~N;N^(vX?jKkm0RG+G<;P`Gm3=f zm9X-?79CjT6c`=-M2#fHmLYB-p^GziC9v*58xxv~6=03yu$99({LZU!!g1ISQ^!vY z6z}*-w58O>Gfe4uhvaehs$Bi!7S~#IMd$;9ijv?j!nMV(^)al=m$gmRFa_HPt{CI6 zdm>1%hN)OZ5gOso#y%UF#(}BpI5ska)9>_Vq`N9m|5{)AVpjE~np=;4=444}^svcs zmd2QCs6kGVwglmNrJq8M&1g`+Zdp)kg@+kW_vq&}>)nIN#3D9T)_!u6qF2-buHe%V z#0nT&Swb$%GD*^uo?-&sH2~9)r!)mESqfmxi!X8Z=Y3H`e)x)1AxL*X&cB ztE+?Y+hyggIksOJX*gGb7N`pmpujGHSxe=e6uq`+Kq$YcOCBc+eGQq7EApa` zcfGFhZdCmSESxORh0yi1Pz^eW|1v$);{}-cyAveWTRu8D){zd^hAUpEyXYuos}6!; zT{7hNAPk8oMr98yv3<$Af=X%zc#Kg6wu`oBzQ|35txONAZG-fDtua4T3aCTD4|(c7 zp(V3{cu7cLp$FDDeJq1hpAfV(XduD?RWuS9^SZh%MjTlwS^2d#0X`ZLJA0`HeE{u( z&r@G3ZPomp&Zaw`3m$=$RNBuK2aUDWkSr3Nm1Y;;m+y}x9dEV}pn|I@CeO~KFfCGpN>A z$!9}LFW-G9nmC{rh5Hc5($-n7ufCfPUwC`Q#VpK7YHgewb3PCx z0J0m-4y}XLl#!gWC^3sZTEKvK!=_o36Y+~?+{C?$NK!oFBOVD97fTDHe+(rj&m_TD z5Vr7Kv`*7%1;KP>7h`8phltU^A{Tq+`oPWN?C|;1ue%fJse~!V|_C?>L_oLC4LVTB3Zv!LBa~(!zqA* zW})x}TX#eqIjB7(?d;dDs4rjc2il&wi>dy;|AY&`yV#l43wLg_ybJT!9V(?tynojBsv%1uTIi zg%E;73=SI>BrHpmWF`y*J7`210&R@V3o!F_8iZ3#uyi(%^ro+5XoVYrA-o4#Bc|+^ ziM_AnQA@Z0Ju;DCz$_05q-7@j1K02UphI&bAO;!|wJM1Kh?go{{{X?9;=%F4LiuvxMVzU6GfFakM+@Ha6?!AFc9O|u-S}04 zTEH>lNuuNp?EHEB&K+}Oftu<({{Sz&tM8rG`2#F5#W>1tgya-06SQv{;7%R?2IBpAc%ugT<(NyeFDYk%F~ctoh3^N5nF|X zr@XR>5Sk(Y2^bq90NgN*GZewrM#-9tBF;L6b%Nz8RcDS*I)O)Yx`Km&Uz8LHJot;TY{@;idmQ>s~1voweu296@hfKF@RZtYKH?- zGRo*TISPfb1w={&o??ujGFv!e5Im`BNuV@Jv>H}2ZFoUrTU0%Z!rR1hi_L%owbYqMaF33YY>* zFqy7q_yJ-eg@px%f7fa5iVR=f1>SbaGv&Lrp&ngJv{S1KYNij1L zeg^cXxMv?=V9eGo>a)T(!wpC0vyO0RdqA=_EHRaJRCLQ1d7_Pg+ky z>s*)gp)?+CD<$Q67Dvd6-fvX^fCFRl8vq+MbW8q}439ffdLLRJLF-7s_pF8OD0m^u zkGSxA7Znu>-D{GgU;?=X1z3;KO1U-2E09+ppqUQ@npYsML0p1>EJx*LjT+$$qzwe5 zz=(oc&1Vw>RsR6$QsEde42X~{S%VF#ypd=s*;4SpN`Ni@0J#JbMZ9!ATWRAy~l&tzx=ouByv(Li(R~No@J>}q(5pz1D#I)0HCb+ zrwxu;Fc2Dl17afZC*sIsWx!-GbQ6TbAO1FgM3K>@o$Y4cITBo` z+Zg1j!UpH-0h9*qP6k`CQs!PZhuY8(5Eu{qr?~KY7lcVQ=tdB6EbmMQE>SZ#g-!2z zhu6}*q-A$!US2R}Gvb1aVzSL4w3v2b62M1iP%nwS}eHvI)ldT)Hi z^BedTO$t@Ti*U7|2@Dq?c&9$R;U!|fT&T0_BoPr1f-OrU;)8sQ8P>YdaeN>n#51I)fdoPtr z0|=J{U9p1h3Uj%Q?dF!QM2oROC}9FX32|H_1h~9C(5Qa~FUn#qn%1)a0PkUw4AuKN zVnd9?&62?=%_s#%+$GT#B`(nsgMtx@AcRSPh;V`Yx-dDM4?OIPwn-d{3QE$2Y!fPo z%~GUUVv{87UllD%FbENfHV7<;*^qSZGV|8c%xs%bP#{#sLWpdlM-uJilVVAy5mxzG zHcOd!xSWJVeP)47P@a)cL7GNo1R~@h$p|i9^gaEl6&NuDl1N3$wv393IR$}LKm}U4 zgWDwx2ur|4?o^EhBOu5aOd&AHA`0Fh9pwwC(6MQcJ!LWs!m2_HnW-ASU|0+qV_H^} zF^l#l-QwkHK|3J;V62ks#eYtx>EujILOIZeNvC!WKwv;XfkQVvH(wHlKl!Itksz$0 z03{M6BmV%wPL|q5MFq4>KZgR-Ypd<)yxUnq$`h0Lg_I6$Z#SyG-j-G-)_xeFwyT1Q zpH=?=m-ikIVyu4>vSqv;X|3BkHX0|Y?KD(76@VKhxi>2{%PfmCpM+XpMe-7ifb~re z3>pAH0fIE4P77et4q!a|27Tgy{{WClAWlm3qYKPRAT~uQwAuS$1?qnz!2_aL<|Ym- zKxDy|wRisj)ffxGT%-)3&=%19AUK#Ff#(zi9x$)?d&rK& z)-v8&8TK3klE{iC)i;`vBe!%H`qR^Bq2TKQ%+y`l5CGVyB4r1htk_bCvWikhZh{uYAEM}Hk5ThJ@xXu4_jpl2==RkL10P`vOw26`Wf;OH<-FDexP?!jtaM;!WK3#}zY?Hk z{{Yr3CV2o#*kqe!ZcvIsk_FU&)C8gt30?^c=G8(-w6awQ-T?JKl|WM(2tz4G3teCr zh*mvUehy;WOH#%)O@(C{7KIaLP*WV{qG4@TZvl!HLX0VY{{RRECU@4SpPfkzKeQ6T zSI=FV2U~YP`KE0-+PTl=bDzrRKb3EgrBMJRi~{9@$x`9>>;&CP@&MA8eE>&^OpYE@ zieE4aB{HtAs+(Lc=glD5d$d z1q)@jI43!G!h>rF*tZp^o>72{mTFkQsm*e`bwTz)((tz69(SpmSr*9@7=m0x-o*^L z094(r-@)I7z={|q_9(J4va=(&5J3onB2pG`5StipIrF08K*nAYqHhEa0)T-BAQOY& zN57rIVpAkhmJO*xNb6vtm*j(RVA*iY@7PI6xYCa}HFQfm@%O1+d%b?I>wZ0dJkJs0JdZOu0O?byW}* z6$nHHJV3w3LCPQ+YU-mOV9)|TbSLQOXldzA9zTBU1mj6YMuw9N45v;rpE-5<>}dvu zGi+zhvYcaOWo2YyKhJjVJo7o$b4QC%(E#tErK6{#qd#|w;ncbRc>HPtojpM_bR0uN zbrnQ?mWt*q)vp%NB@hVsWK>6=_aB3XnwE<0*zptez-12L%AfI%s~{?B;1-S@|1}0; zq@e;)pP@Me+$tsOpV$BYcR*`I5PFIvzN3C}^x3H%%#3O*Qc3=~54z*9_&4H%e@;&5 z)oou*kp#W_eafA9VUHk2?epgvb^cc8ckuF`6Drl2A4dI8w11yboqJnUfpMdODE_%< z2=N+ulZF7@_tap#s{0xtN8M(y%VfOxhBm)nPaRK*UjN$;6dia^Fsz~FfJh3y|> zg&%kpzJQmVMM#r%?3J>oFj%wVI?tV^Zhmp68tsati%9b0MC}NNnwiXsu-v(g&00D6 zxaEcJscVH%)$0N#ILOUZ>Ja$sRIIi-6$R@^*2-v#%cG@)r)hBr$n)9HHR zn5Xm$Br(uoYn@p|XfJo+XkNrPcSt(9PnZ=sgV)3gv#G>ZhVE_!y6);)w=Zfl;t)OX zC0N3miSO-8ux$DW@LVyH|^!~klBW! zNaYWkgrMn#!y$X~#uo=%vKpz92);D^GIHRw?riPF$A_EaZSd`zr!7DICo1~uwaCyu zk&G~_OAc*W2Skc`p+sS;g{M7qsf-Y$4AUWM$>Kv&OUl9ztx9Y@mUuo| z$n-AD=+!kDo>DD!UW;@agQ=UtiZ)30b$J1al`y?%RhRpHQ$a>=rc7~buJ0G%KI0$P z#!f#sZ-g)m?cAPO;0q4PtNTQME~6uI39kNO6kUwzRzX7OWj3x$cXIJK6o?H$D5 zh1bU9!quQ#LXwa=V=mg=;IbG2=*Gpua@XK+S)xv7xqg5jp+e%_%ck7S`9`lW4~n)H z<5o%q>NdWh-+TXEEWwv*HtuE8utBcKKSmXk}-3r?BG&iD1LB))Ki9hMb=y)ELJl_qTgjI=4{;+0T_5 zsNPR3$S~1W7?ZZ9#?_86 z+9oC~59Di)cDuQ`pR|Oq`PWanekZL>4Y;OtNXC;2nWPqFgVRFf$UUu4a=CZcbb2;R zRB%;%=$udLq>%k>Si&_!%w*eam*7rHUjFtYXE;7v=779(wIsFc*?I+PzI`is3Rg3? z?%WoJJ20ipw<0(tdR$N299FNN$V|`2Rd#N5;a7cpbG}c$Um&C0YV{ixP<}yrX;HVz ztD19fm=l+*O(ZV5zk|V3)pUA7owvppbLsFDUS)?wKlA&kR6Xu$n+M0+uxn(6l>Lw` z6@0F*q#x|TZFys;boF*kqdW0&v~z8<=gyAlE{;$zdqDeOFmPko2j%KH(7MShL!N5P zz2$59ykvHDbJD8pQT5(0kXZffiadH&{t+dyEh1H|;CNSRt`s?$#70QizNXT$eMygm zAD)&AI{kP@V8}c7-8lKcX|ugg-fH!DqUU6YFnmu}?K{iMMafHjp#K2@GW{9SE>VxIHMc!2 z-}*q}g9hCEF=-|)<=Y5en{SmV~9kiw*1wbjVtUg;B}aB%yg|HuX&9hScQ1W4Sl z$J2K~G(88GPcz{UzRRRvr7+3|@0dSrf(6}@57|zw4kWH#dgwvX)n`f%8ea6z*}_CL5~P7n)nFc z9|H)X?Xm>CcU{8FPE*$!m4zDCcS+JE|+SVscejd)fldjga%vvx? z+Oq!~_34zYaYY-}W4P*jyK_L~@%W+-UX*oGQLKICd}UR9l{NZ6ktL&GOqlc|nu*LA z)=X&>-_d2R#c=5^9@>orG>Vg3)6-428`BO1+J@GHtyaZvl7c+w{-KxuhjkRDG@s>4}=+`vO?_Jt{>;m=Jr?ykahgSDEIe1;o#-+QqKK4>XWZ zR7?b@jxX~I$E!p$_p|;>l?jFJ35v5DAkb%}Nrs&mLnqj!W0TuXqP6bsw60-HSyZ<` z7YZY9u<4iz( zbV60g(#h`OHLH?hl*)SYc(blMa`VWJKzjB0bsJta7Gs-f31fr{$s6n#S&O+f6RNH; z5bHJ-K7_qa$87upneTJ9b<4App8zuYUkuus zGX6AMk9MOfDWm#OSZZRPxF0A50NpX68`AUcYW=l_*^)V}X*|&(4nDmbSjb6=9-d64 z(<2aZ1{LEU1aBEtNnv}-SNWtDMCX$iL`M|F%o^s8Mo7j|6+F~_0;0~ZWlv7}S{@bJwC~LUo_toG$uTMx01+e%GFV_&=ItQd4(HGAKTs{%+*t{bZ zC(fWztsagXZ0HKD@d4cj4UfH8AkWfvNf=djV!-y{2|1fIil9m(d$wy4uVpLR3JgmV zgfvdNePwwFx@cYYZM%Mm_+Si`;B1~hCTQ5ud@_NkNC!I2UiG7nyjV9tkU-b(7iI4@ zWA{>qG!wUiL&}%Szc0;CK8?QjzmO&f#8Aj*X=B$?@+Q?eMnt-#f~}cvB97{P1q#{v zA|Px`-=VXz;4Sy<_|YP|O4?~YJi<%e=aZ_%|CO^6TwX+CdLEZI<%o%Vs7o?5unh5yv5N6PMP5B5a9= zPbHBHQqs*vszXjp)GoAF+xWk_TD&r>uyhuTj5t@VEf`@pLfwY3_Y5r)9DSoxq0XLm zp;Lo6fW2WrGAi^*Su8vDy4t>6-lf19%dAymB=D)4cI3SDFVNZTdUp}GvdkZx#Pr&J znflVLc6d=8-ir7^X)7QqZUNyIT>QVXTu~GfCn-S?9U~OKV}UEyt#P+Z27ainrd@n# z9N(0ADng7J&Oesl;avE>Qb@ma%4s65z~MuuWWj_#(W%diCD+P4Z2f_xGqkwC$t!0_ z-PDfdR>_BI!?D|!seSV4h>z&|^KMP~sg%02&N_DvaXXO8j|H&08am<-K|5(_Z@(R| z-?z1@OKx%BG9ctFN+i08FEo_3YdhBlsU8$@hZ$DXQKB!7`j7%eiZ{`%J$62+7pH~# ze$=ZFo40)~c%}_3G^EQ4xo8uu-^NWoXoW@Co9o*fePau|2Hm(RK3L$K&z3)SvNA?^ zNPz7Ye~%*@eCN#QKn+MI9=KU0yGhqemt32IM8K)+tdid(NWtHpIF=X<8EG22vi@*L zO#ti63s>k`e72pLRSG;V+L@x_Br z|GB|anX{Rycvi7}N>x1B=IK95?ZaoGFW@ROF=CM_>Sztt2Hi5OFhq|QJ!JdrWTGQ{ zo>Iw{@@O3lN`%adDcsPhVP?zdF+cyG&y+HPpB8NyXNG5llML~#ybM{g(ZuBvcTg@zPfXK z&mF{w_7nbs(E|S|YM(Pd_@H?lFGoKLK_Z31K{_wY4M$s1!-~i9XNClDTXy3n29J7r z|5^smZsjg(xFl2Y-9V1TR0a&VUY2*DNQZt*_3AD5(iD*R1n7^ayrX)bUGs5Z z^g$bTcyg%X7L_cBRh-Sz7nBM*rCnVF<)vFSd;FT+p6s*sLuKf~9OL zL!O)~a4z7^uY7VW6sw*;2`j@G_|;6J-?4n4GJcjoA^9n>yrVLTAB07&3AM9*!aoH4 zX7|9(Dl{sdC^^cMPIPW!Y+EZ&L`!H*2aV*e_#X1K&sQ3%Ps}WB@lePEkw+ z(y(}5$?(KV4dmGNjG|*T)yc{aQvqjq6{!nR+hF2)Mha=|Q~L*yz({`gN5gW^ErZe= zs}6Eas^{29MhHZaLt0bv3t3d;wxZ4x<9nh*7=K(@`ov1)6VWBm$zi8b0-+crQKTd_ z_6_G>qy;d6yaV^TdUoLHQU_|9J=^= zLCJ@$^CONef7EOZwuBk!uPmi@(Tt0C@2mHslA3SGSO|XW5jSQootn=O2*~))TIbfd zQ_dw7Bq`g+h-V@mQN1tB|Cmr<3R243ROtzTZ{%dUl>>nZbPB#KwIM?#dq=Ol$2lP$ zMX+=(kbqB80K16vBN>|vA;6x%*Nibx43LwyouZd*;ECzuV;0ILaQgOwS%R>>ixC)11o{FVC}Wa|3qrG3+F)JmO_f zf2xGLd69&gqnWdq4{F#nwvWCZ8eU|iktndLLEgy1**_efh|E^GlR5(J9Tt4P5h3 z39OH`AvxF!x#kwIg?VkEqU00g4jCtLN#ur>e~);R~gv&&|53E93Rx zS9D$TvGHhgR36@J6c+V7D1Q{Xd3(Zz=L4_P>-_l%3GIXnkJGuAN7*ap#EMxbmd*%DFc2b5~-B=dapIH`&Ov;*7bk zo8@K2qxCb9n(RC_Vjosgb8Nv=i6J=m4e>h-y}7DoEafgv+phX(4vZ9^7>i~jX5{=! zgP|5aSM>^$>Nan0TfI`k%8T*AD`}C>J-J(Hr!KzELw>Tt)rJ@(cKTV?lgudVVSX6| z1=FwOeV6*Ge1E?j3@_6SU305&clJTN>`G*8pg{lWw`LnW)!7@3k^yy7o55#SU4G^^ z?(dOVbfRv)-DI*NaWB<`uy5b$-5rh{G|r!;6v<)9FEv&1+Lr*$L!*rx^@CckLl_+Savs7x$fOo zuii`VD!Z{jk|-7SH9rh_M_grO@xKXX1dF~Aee<`Ob5VLoB z4%Si>yvEIQlU=mM3yF-=fLUKyQ<63pbU5uOgCl*Gv=m8PLFamQ?H!oISL*_}G8Xr^ zk{}@p+2kD~N2?}H>DP8SyJPD10WKh+__ZwC+%Xv5)~aNXgEhOq;tP86P0JvMWz+nr zBCgweQ@O2U6sd&`lTFsGNOsMRDY@KCe-V3GpR|nmsrN~{(Mv_Cgp8HlsEHq%i(@@| z>3Gi93yG5Tgf`iQz=3m=1xzkxBsbxJjD`)uRj$F+i5g*0!L!@#KbHmc6f7Mfc#q*|uYC8rO4#lGGoRe2-xB4qeD3UZQcRx!1ui6lbT* z|L9djwafccHiFvWt~Ot&Q!GF6RYb@hysVtrPNp1q;rh$DOR0xym8~V7eCnlk_S&7! z*u}pm|Bz~fq$m9dPVcF2p2wT~2+kZ>9vKdDC$4|7G{1ECJfv$81Qe0~Zcvj#MU!~0 z-5BL2GPX;=km6VdW@lV6aWYiK5Z{A)e0_t$g|if&)7_yB+= zv}mA`w-1i}0wwkykqiH3h{h*<5kLPTlqU|SbzSte=__q@v?KjR8)H#q%=vCNQ4kiH zdG>7_vva4OpR*}1_>N;2xo-IJuJ%FDFA#p8%R`w!5$P9isZVM3rijRHx!>3d3a)L$ z7L9J~y4KmsY#h7t8O(P{(18?|N?}U-{{6>+JNHmvW3n>7A!X(ePrjtOle=3h-*|sL z*oHzk{*fjm4%xf=UR*^BcRmuOYin9Dg)ggJ zHY*+L>_KRI+GP5V-i)!K!%+ikK2>RL^X4o$wh{uKug{>&vK-<(k#A5JY62*QxM~xc zOW$tE-0L~=7QCekjnfBwp?OssqXUe;e~TsDC; zh|CapVn%p0^$2$}QuGaKgpi7mLQ_O^88)EjkD4;_1{ zy{9Dor^0=j93wqSkU29Kj|UJBLqx6&dZik7eMd;i)D9dJ^^z5&gRYPQ4=jronbI4c z1#y;2_bZ&ODzVr}V?mqh;_7J(g-Jy1){A(94J;WRn2 z1v_#NwKcdO_8Jo1cL*mBBgXk1_O}}xDUL)fM7?okzd`k~KMZJib}dgm|MIhcMw?-ov=wI@hwUUNJ2Yv{7ec#jppbHg$XD^|Qh1ZIV?k(Q~6sloJFx z^;st(y&A|>kVK^Vg`CSNh7v~tGKoQ~Fc2ioa(&$%IW$-*^!`li$u7}$fRB`nwitRn z6XsuiWQ8qez#EPf*^6342+r~wjusvrNn9v&+pEH39qOZk>x4axL+uNSGlK5T-dJ^D zNXRQ__PZzoZ@gwOX{BJvaSR$by0R~F=SRo~#0M_eZd7~7yXj2|2w=(new>LJxN`0l z`_=g~$b^^F#515aqbD|V2JeziKxddE+YE1HnUr>@8t6;GMy|F-+~~X}0H(DGTT4IvL^5?^p)p49h84 ztg@9=33MOW%%l<9AlY1BId^(=QquG-mX6<^omSe+*=enBR`Ss{NbXE2#FiGsySYj{ zz$-I9lRRd0`fiiB$yPGIf_YVShF4*{54Ag^)*v#W>#{?)x>w>Y*Ioi7l7p-{3N7o! zU9-K0W^S<;bfgA^)Gv(0c)FE2WPkRnVdEk8)isI<`iBZ{A!Dq!ow%tyHzz@Z_>X@+=B-oc>#FKrMCkBMvV4C1 ztAGcZXH2^9=@*!OBseojtloF<%O6D*UYG2;LZ97G|3vv*RGXoll|`$RahVq-VQW73 z%7BT<>ChPUJg19F%J9lQZ)Y(xWRZ=L2StSoui8hm*BJnCB09Ys;Kpt??-Obfn#*w> zt09~%z$WUg$yVHBFs=P6@*_9Bte*r+5Ig^iYH>|NZ}n(D?enA4;oL}W)s9xPsvBnO ze)&WmxwcLBK=;UH(K4=zoy;D1u`fvno3bD9v(WFIF`?NQ{|ltP!WBXcu^uIR7rhuE z_vCW5i=$S{57XZj=Gnjr+M`u+J9_W%nGg3HzGo!HpPRnZ^)syYXW&6a^gB7O#pAdl zhA)jt$~+x6CY~qtz|^q@saIpqcd;#mTW7{0y9tBH52oi_ScM$b z+iHA89a!OF1pkMb9MZ)qWKE0t=P%1VmgetDo)33}p>k6cONV(rKrMy~)JLyHoQ-#l zH5*uW9_f%^8HT6_;#rTTmz3NN@6Ii1RG zKW`7npcLI)$2I!lcfYMVBEPKG3hNpm=!wD*(598yX{pjVT=C{<#ztrm`tPL#>519GN3-pzS}lGAZeZKbODL zMB!zRUbG-&d2oa+AUFb#JA5BO<&nP+lNxrq1Jf0PCU!%LHgeO}YG9CBC9$Xjx6p6G zDIpIg7SWFEDJvuDAEx!OQdU>|>zR~hS_7zQZXJO&y|lgK+&kgo&)PE33*kxmrD2Muu^eE`g~LRQSQ4M{3_!m0w5e^ey<*LDCxUgAm4wV(8Npx?oW?k@Se5WO zRL<*ed)xbDO_?SeNBXS(xdLwR2`}hL=eO_pQ8{Qe`?hUuS(fKFy?`#sGa1(h-#R=# zI-Z!TX8;yE(dKC%3SkDgV%}TFbp=Y$mxm-gXfbV`w<5|hRmj)=k%RCoTlS>%NtfPs)8>x_so!tgFV3 zm!ZSd|IlWqEbhTv*%Lt{@dzY$hq|hsJxl0TQpTcslZU-*J*GfDl7P2T+TRLZFPhjF zAGgve)KTM9Sg?#GPQ6vID#9OVcp$D0N##CP)w@LBbdwCxM}9>>W7z~o?t=tz|P$?dm~yirW(Qdy3NC+$fhPiz*J!NicNj! z`4Y>|P_bxWYJSllDu%?6!Ct7VdhrrWcXd3Hl0U~^$KLEvy9!ODM&&;`lat++Y1bNs zd@q#mc|wN`qHR#^1?3luOgJ;w^+$0=XUM98!=>JeKAW}R@YDLnAdOW=mV31tE01wv zl431`sGLHWp{uskF=Ih`e@s%FsOO)rDRRl09Wflj;2fxcm^%x|_4Z7$Yo#^|`{odeta$BJ6 zjWCSL#THj6ErT)`=y?=u4y=DD4Lf|uV8GsY;iJhi_G(E zua0A!r4OGu`;$xw4r!hm{kk0_rR0%%bx88d-)IWLq?8~qV)oV48c>vQJeYwq>|K{Z zpIokSr{n0Xd64SV;ohWKe%Wqzf+}M67FJ3y&W`&{vP6Yem2@|E2b|tZO+ySfJXhlk zxR|dv*yF#s7+Y(=iE9oOF%9C+Iys*MrXlcBZ)?>UTF)e~YnAQIxEvB7^5C3nq>Zwkb?Z6KB1|x3E%?&9%+$q1_n5%UVY&9(Yf|zSOYJf}KMw1q>ShpA z$7cHzZFRki%RVCf)9rQN5=8E_d19o!pw-mIw$dn-F=O*K&$lIU4mm>-FtY+@*W{b( z5q5pjI!1x@TTCwlr^ojiO`H6Ji$bX|YG{NEe&}o3K&j{wDO%CLQ6bT*VcU};e{JnE zp>=^XjXgcgaHlM6h{Scb+;2D7kFcivnJEC+6kz_w-2fr4E1_J@LR6f`p`l zut%<1WK%0cy-@5hdm_jOC;n0&Y-`F4xN2)eo2O?emhT3U@x4b{**OTkbjQ7V)pmV| zc~*t_Y~lDOh7`KmhPxXX2(i9CMj5Ga1K3H}32Z>UUE1+FwisC(kO}`{`L+H3ZdV zo*n>p1ZK}!md*I3H`c1u5$N_wWRaJS6(J6rB3@SM!g<)04Nm%{G)K(+opc6r#b%Aj z1Tm+lK47c88`9NWFSrCvgoQs|t9V%#Js=k_K_axuKi}P0y!%zMrZ5|~qgyjOfUhmz z9zaYBZ*d7vrwmI>Ox4WxP9(z8K=C6up0~d7!#~h=ZsOp)3T?CD{IL`VEoPh1TRJT9 zX}T+%0rLlR)Arws;Im{9#OlnE^8V-e%atAJK7|pqW~J>#6C~aWYf4kM+WvNB95u4dR*uk9R}proQV_j)yEAS}(I~ zdg;d&y_IQukR1V!Rqk&Vb1;z6&d|9(;mF+D>5EWew1{je0RTS#H|fl{In2ex_nv4b zl~6&KD>I^R5Q3K@LrwnW!+BU;sqyWKh5XHfOqdD;ot&2V_?aFWFTXY_F*NHDd$Ha0 zntr){>;;RlNZ)>?#VQ*8&{da0D&EkeW(&kEW(aD($C zT+J$%$5O1{bg^~3*F%@}yYo8^%r~m;hiwtN4?g8AL4JW^s~;a*-QITgwd-_}R8zom zUUV8g`BwDtg$uDRkG3st%#IhAGXk2Ve{qai2 zosPf{GP1CN-5sydU=|8{|Lu=EA3iP3q0lK^T|b3)o~8HR1~$!Y0>A08C}?WnOAWDj z5lxI;A~y%xPSaq3-}&XNG*U(wZ&DbcW-D!i6=nzDML^@5d&xTD&sqam2TP8;kb!me za%=eZ$!~E%0jvZO<<4($_I~fW;`*48c^^V}MfBv|vsk3XI=DMtYRE9pRi7KuS9~EM ze6v!xT~#mMRsWTeN6E5tc5}=(y*QWRFdnE$(L+`7@Zzxdq7D_YD4N{(k?#9msQ(^6 z6L_-z#80PcG3+0Yc_KlY9!}>5Hdltc0|ha4`1`5kOmj*sPqcj}$rF2mP84?gntLa8 zjd|Rj{0j3ib8FdYP%fT7&5*VtOA|9wi=mx#J8LrXNYID(9eE-e`bSNF;nXM9XDCZ( zN5&knsIoY$;8#(b;p&p}N>T1OUjm_K&_ zD1c`8%Xu%$3RY)!B^P6NHi3F(>XF|FV?^Ldv`&Q5&>(qFWoVKiEMrk@u3GU#3sx#I z@jGOPPT`&3VaEJnh63*`tn4mhoIU?oB~)Gb-o)(9guHUT{7cB0t7HX;4IoKd(Zti5 zFWU4t6S+Oe5EkS&7vth*k}v?nNDF2xirUmatG4F?jH9b}vx;t#R5~>;x6#k5qdQjT zH+t3*zSt1(aj{+*CyAZ*&L8^udG&U*=w#|Jff&@kZL#sj2$J35Tg`d&GO6wWk++nU z`(76Q)-EQCIv|;%6+4De@?x7m!Ty>-mt8&%|2=8VSD0u={K=={S|f%98jcm>7%=*58V^3k)O{%C-v zm3{3i2y{FV@#@dS(eqz;lX1eD36tFB(P=9PyK14HJlqp;ru!R!I7{_#;Je?1?BU?+ zcJ{{}%vCebMg8SJK8reMk8`(Nxe*=u)p`+;3vQCnZjmr*vG;ibAPUwoL1Ut^Yy0@B zv3VH+%o!F>LXMCj$INn1h{mB%xe6j*1J>z)ZNG|JOd?zKyD|dirHx`P=Fq{`(gk%d zRU$aFLOVnI{;K0y_6$;%QES0}kiTL!H{3T}w!T3cO~*OmONTNmmAcU@1ODF~5*zUr zR_HzXtObO?&U&B@K5HX~yg^zYBidK}h<0$43$uDa#MkqFfjIc2#H@5N*t9EoT%M~A z<)!L!27H+1AMp7s@@sL}J79OnX_wP!#!V(#GQvc#~vIffGAvVj=72RMRT9^ji!FgsQE>j0@?9IwrX~^Gh~|HpQ;}Ey=1U6`Z6sL z2BU9&ZFte9aEE9a?(f|*;T9Y~4lC#ztAvs0?Rf2C<{|+w;+&#{$^s16hAi(BK2p+` zz#S3u%`;9dy8{YgIzl|g#V-J$a~Am`6msk+^y2T<;>gH0TM5a$0<$4=eeE`tqs-)i{z!%4Z#{IM58j!0vU6zg zFNgWBNB=eT2=JFCcAEW{Vr?h}etk|k~rV>2j|8xEdM9}q%a2H!x} z`^n-wd6;{f?q-`#+trQgBrC1PO!8}0UE-3>=mX7|TVmY#kNEO0Pyy-0(bJqAFi8L? z9Kk+ejo+~EvrT>YvPq#`5%NX&_8B0^(0O#q*N3kpX3TqZiq-bocRqTpYA^^ndsI&! zGl-dM=Dzs)sGiPyr-Vv8Z_aL;d84y3ujPug$(F=XH4W^QGf|HHOVB3DegBMB4mV`9 z|C>*Ybg?}6;}f?gz|ayG6fb?}rwc>=(Ir2BfjSPKlAmxXZz}#*3IFTZxWoee)Z2vN zS$3qe1J16~_Qr~gW(XlUJf5=+c;0Bc zkjoG{Bct<2+)edtXH;7QxC0&OLQU>)oG3PCJmN31L*+1FbwcrTVa-qtvGa-7W3Pyc zhjwr;fzkJr4Zdo|+y>KPg>UA3baA9+G=I_Nkt-DO3ai8N8UPNlOe#;$Ml3@O0k ztX^eom$}2Y1ppg7x!VSmn|emAY{O}$9CkVHV9phIPTl1Oqdhsttlgq)t6qv+Chyk= zt)*Fs*WEmmG!4c86vud1aJWp0wZ!!{7-`+%QQdj1e@K`!3_!v{q4HS3+2enp*PhS-_8@=oa6&rfwUkwOu_opa_4 z`nXX3&ADnF@uy2(TED^6vnHTEhMC9|%lccYTQzdb`II>$vLCkh0Ja8KS;aT4?=~wTzJta&jU$q4weU zued%&mV!K8N`&a)2}d_&~o`ZAwQH0vH4$SdCC(;YF4!Lfi91 zuy#mjvm|x4Wh^FmhqmBY)U!K`j${sC*lUW0652-?R>>pbv?hRI)r9h)GQzCLZx<6Y zJZG8$mD*V}`pGzAXZ_)Fpuu zZlwcO3`58gJ;9pDIjH@GK;9Ft6AO>GN1JR9uB!r4_Ch)`0AY4PEJdT38NfnKK$Lav zPj;0QNdCiHD)(9!xbl8E@cUA>{WD#Qq@dN9Lr#*^C64X!Fq{b0wE1p5m}IJD07ZK7 z^qV>uwKo5MK8%m-{~#*>Uu+TBnbv>0vH$LGkGV(bzMc5RK7?m3THuMW03Sr^SKxSL8Ji?w?z2ui#=Ql+xHu-j> z0f~#;Ys@C^4|eSydR4veE54s>`YgYP7C9s3_1! z8Vx=BhcB`JLq#{=pll0ZfaRO{Zb6Pg6Hmn1Yb(dHpm}&al5b(JiN6MTXaHe>xH~*w zl-OoSpK(SjvAmFRfm(OKd6|KQMFEZn;Gm5FHrD2!95m0xMB)Ex;4b$6lht@4QS#m0 z|GPrsUSm#~8NP}HV*6Wp)0rEX76xZ`p+2V|^(#JVjkSzw*+<%G#&kP%v8jF%4aG`o zNPmoNRfjgQIGcnnO=M}RJJ3u1)Kn?F@=AklxZg8w^ggnU`-2US6~}R=Uhn#kgRCXm zJyPB55^_M1>``LYNf3<#j%=6_(mEbALl=4{qq9-|#;epmOOkU%DuuQn}wxbSIdK9Kv|#Y90qi(z74O`FHu zZ`!|2`H1$3s-@>e9kDN*{p(ffx;@?P{i;gxO9Mmz|X^tVEXZC_Z>H2gR* zjJxJ(!R)2^Cw;~h!jfBCYkt#7TowYD2O6bV9U|L(a$2oD23u4%>_TJZ`__5VT(#s) zGF(QFUezA??PaA{hqOyYOXY>0o6#;{xDdUk1qAP+Y3L#^zFB9N6ZR5mXqzwONr!$c`Yxw6fxgS}_f2e)usT&SAe#Tqq5sv`O)vmTM7J3M+R2yy9<`=6&o#o6Gl}U6@R$ z;w&5%k7#9gvTz3vyDW@@Rd`w}j;LLym;VIih0&pVRpEj7iGV1gVa3)y=#=N*s^FAw zyJd$%b}ee?V;|NfDaRZ6Ev`*t_0ZB(gZgwTr`8s8rF_^0$52T`@VQ)BECb|mmwN;{!}6dtfZhuH zQQAcbi0*+i^lmxM1@;{F?BYN$pp6?6+>amxh6gq94#?pLM3#d1J+c*OoFeR7-jMa5 zgkpv1B$RH8gO z?A3*}#Poos-=~~>6fIvK5Nmv8#gGNRgH5!)q434TlgLc_P#LkAKa4^s{?tPL z>Rql_QgE5vj_lOctOF^GP3+dDsGnvHYQW8BJbAPvK?E%En^Hx&F=GCiU5#4(8>S92 zF#j&>2vnFB_*o$q3N>$FtHTJ6KyO6*wn0UrdcNcNR&!b=>+>}c{F-Of1{sdz-{yfG z-pz#*u`9(J+nlsPv^F_d$QRF78Z?510xm>jHLt(h7*6Z`!U$m%yi{yj16`zB)>rYGw>%x@#MZmY`oR2YN0qLQ|u2! zvYjO<$V0A2t!0)@R~9cCm-b)GDdq0uKz{wxhWt-@BrhL4LNH>5&)(=nh4K6oQaUmN z@CG1DK$-jq~rT2EqO0*|7w|&a2L-M&6)&XMp(`>+DW)*81Jy zD!V|2Zt?T)hNuu&U16P65C7ITfJzGG#^TK?!!C7(vPuLu-%=SH_BO5A)~Xohl>tc8 zxh&_&Cg*vZ`tGrutl1kAq;$vuJ<)&T3x(gz>g&$fYnL^6-Pqz`=w{DJig7lfq)0xc z>0kiws4KBv=&he^8Qu4|60vb(E#0A^s%U=|9)Oq>gWp>Bq|DC_LL3u_OWRE1Yx^ru z@?uVrb*aDpOyAx3#X#}qHWNLFreLa=`SrVjda`oA@dUQ{vz|PY;^-bvIzOL%(98` zD0otv7XEHX0wjZ2{hwyCoy2MN-@RmjV-VAO6ULqMU@nRgfFOT(x{HZtLchwm>H!P` zK)P$j*!*@XfE5Ai-p@U4X1Fv6A&tV6?m=>@l@}@B+@)j}Q@9L=R*h2#LzSxm`@G&Y z6O9&&H3qwb`@}M-o*m-@8h|@0)5(`_ly;Y0bR{BFmP_Yj*;_ za_oNnJ9Prz81dDR8t^u&oArjDR+iC+iD5Nz%3e3bN}kj&8NmtPOym}St%4dVG%R)1 zjPXPZ!{vw$qgnn^#g{iVf_uOi7M11U)x3H6W?C0sj#QD*fF0hnIuZY2hk2j@=6Vc1 zWxcXtb$(qI&3Svw_X|`;ux_8*aumItF(>~EWQjkwwU@v`Oxx%!ZjgYDI0N0;6wKq) zJCctcXhsv{-Wt<8$0L)Ab6%0~%IK}fj{d!_;^d#&jx2X53h(xddiPiG`@L(@n68x0 zZ8M?tDhjdpv_5>lS~E@x{OK?L_y5IeLt^}rd=Q6fUd^|^Qq=!tTY%_CqW^wiwE0&_ z$gv`nZlgr>26XjRx}XagRgJ{&(`pQhgt{JP(EV>j1uN@8Iyfz^A=$e`i}7~n|{HUqIVhLN-`(6odIGNCWL&;sPU1m+=OF3Tof@MoLaULcA66TWeK9F;;xjg7BI=#7`Vi5skcf39Vr zk{->guc?13&I^EV@e(&C<)?MijfV2l>iW%+ilS`ZIHOdVBct`%g)fGMM*%xF{*}q( z>wt1xE*AKf#83I-ri32{A&sLSCzh%pw;zhiD*S`t1E1<&uP`C6p%PXi0}O?Fezsp4 z`LO}CwHF(_7f1FLD@CkCS`_zph!9fY9Adg^Pb-NR`jRo}=<(%$=;*IbMoG*})<*-y z?on;OzG1Z)ErRAX+dq45G-R4@pm{&$Cp5(X@W;7s#cx4ogKSgRmcrTKzg|w>@wk{B za91u`Poz$bD3U4I`hTc<%dob(Ze5tRlolwZE$+n~iaQi{g1ft0ffO&r-Q6{~Yl0Ma z4{pWXT|!U#yzf3|@3Z&0zTaPd%p~iYnR8@~HD!$Zo@11Pb;d!_!KdahD<7U#f%p7# z3Ct6MXsuJDj|ERo438?`^-*>g%AKCoJESzuqGfr9iSdfl9^P&|?dwgg{ zXXAfNdw7TDPNeSv%si)d_j3Sl7Z0Crdc#-kQ7Ut;_j8Sel?N_MqH%U@^s2|-jgY%C z8kxRqf#d(WwNl{V7orKrQHOpPWp5_-jl;kd0qjliWt)XT{|%h3VP5sEc7FKsgKRSVU0B`ER=X=fb~Ti)>C| zjb(+jEeyHeeCjA$-c1|ZYGs(=ypzF9-i)XX?Zn|O-u{#sy1aLXF0oZN9etC{arLcN zU}?m??_j(lmt+5#>uA-%{i1zo=egL%%VMcMng9K8!a?0Ma0}KUr_`H;G;VnYs|dUi z%4!cepLO%Vl(C=VS+h!R+U6s|+TG#PcPJb8mAF9HzI+@vIb4-IOxuCO$Cv+AvAXa9 z3Ri4N;b76BbkdtkC zv<;DC3_j;gz^PC04aPrS_QGDTO3~IR1+H+0dE#a*rx5;;Lc%3d_?s;cF zu2t7D-Tq_t=sNkHbDc@`VAGis~aD>NAnb>_B6ff__!wH{(s;c5#MonWA?9dDvOIvfvIvaS;vxdv`}6Pn(XoSTH$9ngqFed^B{!-w?ESPe z@DL{(I^OCFFp@PQC9eCxJ+=Dbq#3x=0;SJ5T7i`eBXGC)(m>|Y_s!`1)l-LjN5z&m z7v=PWZ_&x$@f>~!(wKK9^6i-Ckl)|`f5R{%)EQUP4)u`M2l{+-hvDiyc7kUixtF&8 zlP>=^?)9H}SnQQE<`rr_B+QW3swP#=3J-NiOo1@iB$EZrqm) zq&_8vg_+~GitVv%0kZbuen^gR!00cc5z^<_2!B%GJ9YXW@S)fY-!lH!(fD_{{AZe_ zDl%6FagUC5Xw}VHbOaq;r}M%1_R#Eyb$sLPF=rI&Z}9C#PQdsT6!+~x6(XmNV{1_d z-wtQzHc7L`J$%_n@aB%&gJ$k=D%*WC;J+4%hCjxd`JN;WCvcD(BL0sx$ko|)k>!&e z;E7>vu{3!6^v%X+cT(+0W9|3F6&z2H`1bhhD;PPQu;M=X*BZ(DYmNLKDqqo`7SPHx zkGtoM=04$$67+s!*5{T5AaQ&%`6y$pYupx?akH_X4Zzx5?cPkj(Zgx^?$9F02Awe& z`-|_$hGl?lyqFEQx0PG6(Faeii0qhiuQj4#@owDBH^Fs+RT+nTqP{6m2R0&G=X4K4 zWW9(DtAFhHK8Zffw0PB-3hRoJqTfS1N~W__4#;B>FaEzy`MhikJK91``=$FILkQ2(HV zP)yraHC%(EX~0OW%rIvmU2Try2NxZpAGpF{}zJagbro8o9kLQR_#P z=qHwjMJB71tx@u#)R>`*goaxOj?0tr*+xQm9Hh$MFebhy0Imm4yv6YzB%Az->+r8o5RQc zrN^7YhL7nS49Ybh8iPKM1-pw1R%N*@!4hEOYz9tzJ#)sr zH_Q0+T?gG6$gBw%9ca04F}69!Y8KsZma=D|mjXB}aKt^e>d44ya)BIZ@|q@!|J+o|uBmx{ z(>^0x*HR^vp?tdfTMyd3?FNf9xGqTyseFSw5$HO*E3y5FQQ~7-^*)`I@bKz?1jqcp2kgJMPySrE{~|0$?0AI9Cw_%trI{KWp#QeR&{@G8R?=Ak#l995;ne}>&mT(hAX?g`En$I5u?FKjQc5Aw0 zsEYbbQ0h`f#to2>LwS*uj&nytgB~Dbx%jzsIF;jD?L``=6n*V)aLballEkF6vc?8@ zRBVKWH`kgSTdijSIZ{;^ea)piba3f3 zD9^`^MdUzLsh;(X*y_ZC-`k$DJ#~QhUedGhgzc&+<5r-@X7FTc8{+MyXL=02FQzxo zkGo{ny%BXyKeLgu7}cdY<$*#B_j~S zT=_=dr!X3o#j7+gvQx&iH8-|oUdn7l`11<~^?dtjaASM!8D}~=H!_}nQIFWG~ zbV$kRZPjys5@E`WbcQDz8mTjVFz(`<`{40%Wa^L{dy*=nMujKzoeZkAO?6BXVj!QI zx_xiF?z~~k7IF9H!ef@rewNw(11zvzr)1lQ`kc6b((7{KxoK|3-f;!?F35FI$%&n@ zMlQ%!mL*2V*p!tvHiof=6%TV@0DHUKpLXmDAMM6lkOY2C2fL%d6TR?08?<{Co?%uM z|2A!h2JWkIw-~y+fi?BOs;2iQPF)}!8q1>uwR1_IqNH{lwM4YYX|_rE*PR=Y5J)$( zMmKpL+g%Tcbq>;C-*`@fGw%#~JhN^CPJPDbr|rIFWrZRvF5uznsN!yiX;^WLDBT zvRb8s*8h~Qt+9aza}Fgg;n}k)(Mog$x#N_(v;Wo1vk_b~nSV9&OmcS8Kk;-qDcCVr z7sL!IHzlOTL!#YNin0DoArA-BY&A7Tbd=@8&Xa`$t;BoE7RnXA$BsH&5fOV66a4aK z8wSiC!Z?#j1&JTMxAhabAsVBTr}vfF!va-l7bgBU$LQkIletR`dV&$#G^^IeYo@9* z@zlub1A@I0N_g0%nq!PL`aDNY;~m=_ir096y6ucp$N}6f@^j~LfNT1S0+g@kJW!*Q zQUk|0N5Jp&^&K4m?`mk42V*m8(i+{5x>?5E6z z&d?b}YZh`^mv;-|>vkdP<;yrk5s7L>D1#M){6B&#{PeqqiX3^n<^w%Z)pGYf8oS9z zEz(qkH*-94aL|%$*aHpYZolU<-(Z%~}$peUS0blr<@&2h=fQyzi9T0^-A!*y&nF_Yxm0V z-&g))wlYi{Q!=hKmk>j&bOcep+rl&kvdG5C6wcrua%3bY)0D>6Rw^av#`YvMl8G4` zU+J1*#mwQl=+dQ^7U_kf@`(3pH9clHF{<=%~dpm6cDUA2LHJFye_q4bxQ8~o!bf9 z!$i#rcZKh1c}m78)jw4ziW9U`bmC-`yGoUu#qE@6Wl>s$_uwc{5Ti*359W)MHaZz5 zF$=tZ8sBKmXO*ebOxV~=%%Gm`z096mNmb^(#ri0j zHCx$USb2%2eY2XAO$Xq-?cjePwwHFV^d<8~g&)n7|8{lu;DG<&UCAtYK*E;|I8SQT zypEy|c(9$b;dcTxpQ)FoGk?9WD#Npx`^jC)s2rA zJZS(<`bRqyp2i!vh!ZEDetYtH`I>q<@OAH@=l!F}zqXwy4#xr>sop`Y=7!%| z?T8to?s@v>qF)Krh=u8vSjOearqv(A+2nOOKd6>&i1Ad&dzLqdw@^vVipg+c`Kw0Q zzb+1E4ODO{lKWJbbWN`%tgZeMU1g%o>Xj2eezMk}dc#$*=8*CNZ6C+dTP!E1*L)b& zD)K{zcHwb0t9_|DA~wbYf*5qrd4>sics zWW6$v#PS@|PI?{(lvzbqLp6^KEW6LP5m-mGee_z2%Q9pPj0MWeaPlEt*SxU&uihHt zBN5|u{&*F2@rjgjb2!H?)t;5i$tvyUgZvK->s2_F*z{NVb}PNs*M%2DEr9)v^OgPr zXP)B;K*CkV_5G#r(PQe=TE@dQL8Jlg{!H>%BXM}2%qjKFLd7`Ft+(=bt|lt^Yo*cY zT_G1cFAsi;eInwsd;GQ59$>ULMmazFekAtQedhCid`R2k$o+KVl>r(LbAHPjrXU32 zTu-R=aQx*l?2V(>{+>YLbnwM#UNJG(n#Y^WETTnElCSQ!k)CfHzX$$7F!Z~aYqfv) z`j6uN9;smAgWIJRv^Dy*pWqLo#RqE}?lGmWY=Jx!HL2k&e6-9KGHPu+I?i+Rl0n#- zDWcqQl3yyd%0epNEsALm7^W0SlUYQhgeSEi=UGGq5GTg4(AV;mSB|=y#fEA-W(F{5 zBz+%7RA8<9=wX(`+Q}MGULIawRNkVlq}5a2hhO3>BT^T-K*-{}2V4~N7c4yFj9&|N zPcR=Dco&C}YDhNE{)G*g78z>f3+!fNNct z#K!Vpn&nX5N)q2nj2X$N4Kj>kaeE4AT!D#$US=GrrCB3Ite0 zUGzB@IW`z%?&hQ)DVUVV$>OmwtUH^F%O)JlkGq89o*+#8d-lSwDE=O8vGU{JU;ob= z){%y^=_d&Ga2(_?vJ+DUi;p#kk|07XVwygc%$fn$4`*7_#jmr>SpAI2Hqk2&urjciw^iJ|-IgV|2kuaV zI2f=Q{SO<#4MQ;{lzWR~<(19_R94=RwY@*}v`aa@lc>isc;%z&`Pb<{ST~iPuC;GxjSx$by?5oZjDofm}0o}OunFZZ%@uVuU zeY1AP-D>f%TQ(1R+x;f>;kveuav8E7H^cm0wB-{w3pT;y)|(;yPFNdmnUch964QyN zY*X+cS^wug{`XHGQog28I8tGY)R@vHzf5ole0gUhA7g`-V&hbYX6`0j&S_N9H^Py+ zfKu>S{DEPNOQ$7v)*MIL;P=>KBmyR_nLKpZ>q3vUT(OKVml#eEBck{T2If z1y13zqydhaMWp|^9j7JYA3T>iXzzi!d1GhnK}YO;TkIE0NFYANne?v{4!{2U-TeQ- z6(M-K3*HY1FW^}^Ucystyh8k&76aiW`~u=@7NiftA3wdpz$E?5%Ep03M$Rtsg;Nom zf|3izkm{?bcnmxn##?yp7q4IbL4e`L8HI8fUR#_v97rp(G^_auUCUH!Z5HGCmWusa ztygu9h=fT~;>N3oeKLkuQG{0sg;yyThgT6Q73&(;rsI(?I2feqTdh}&C-7C4@ry(= zgup!%C{^mp)0rL0(;Cu$5UTNjae2(X`y+a>b~1XCz6#agmlk`!Vcjs^u0ygP5X$F* z!d2^t4z8?r;r9>=RR;z1h5<}-0PXU%#*GCjjV z39wxpR=rLdhq>;U!nXR}^9d`!Y6e$eI6mQHC(7YBADa2NG~UX5d}ZRW_pzG!OC*4x zeX{QY=lBmofpQ2_Eq;m+D=2t`lo}pBCr<@cXqUoh+LCsqkB6@#KXBJoxyJ27v!JSD z!dIlewFxpKeJlJdF4q8r9liEUw(I9msn3uI==rrPqM#LOD**N>Maw6*2^)?R`Ha7Dtm${pzUt^r$Heu!;1jL}pP_^_{HJjE^c82QChX9eD07N$yM<*)I zRrjZ&Yh~A{s+JDLi7v8;N2JetVJs9~-Axm|KRvSgtxNFGWC`3-b0dUh><3=_@XDDF zjQo*7@;k2LAyBLCe&ulJn+OAgZ>b(ZzSz41sYj?Y&1BplwslG%h;;4_F#uOUS+Y0H*1QM}5C0e0c^?*@mb|{VWWHn6k5k z&;l9|+)#VoSBGe(NN@W7)NVT9FdyH&tzw9LTU{cg2<8IN;6Y<=i?VJ22dGWwIm~ilQ?P^_8N$BKFM0OPGb&1#R3f>JYAvsN_q7R|{CHoK$4lp_#xmfF z{b1b$$`6W7sL717EX2$IgD_L-69v7v>0|Yyl}B)UH=`Fd6Esd3`MKm(X_UgvViUo; zqpm)R=xYEPrO7o)|H1g2F-~MpO1gDbd7e+4`$kIY#B)h%y|G%*DVoN&@~WK2CKc3CJ>iChey$mkAQ*D>rY@1D}fY0`3T6f#58{5nib$Q5kxHxB8W23A{%E;3_@w7g4Chx8)vh^%I&uX zYHCD;Sa35zw9&fDU8PB#h40-j6B^%x%9T+{7Swu5D9M2G7q@=E(J@UUwAPK3MW6uv z`*>SI&3&Hg4l@y7EMnH>+g$$|&?P=&$&OKEX1K9M=)!!es_;(LXXppWXU^Ct79M^O zmfBkL7$2Xjxi95SR_e-Aywa;*LP^4wK`KX_D!1;XpN?b9Kh1YW{#Yw@JD5_Afr#iy z!oY%K^S||KQM^i3j+|luR^$iy@f}v=mk1>z`n*6gWM2Le0i^7jqDv%!l0u`%(C~j$ zxB&0FDaIOC4>7`8s%L!ANlg#NiM-TAProNPgEja~WcaslQbcUEL=v@GnhHy^^9vvO2>cjc^FWhQ*`Y3`=E&4JF4iD`f`HV^F^Wr6Q1b~f!_F`WiJ zg`#2zxn(CvNBHn1TCC3=I)q-|hM1l+6|Gj_r12f&^losLS4<@|T>CrZPNh|*YbZV8 z(EV=IsV^!4&D4|ZzSHa>X-Q`|l2i5^mJvzAp#I5BBST6#Z`?YE#oje`%RB=rzT`~c zi;o%np!YMYi07vfwfp3B3b#sXbNJiVmxl#ETTjBYoojAogud`Z{B5{y_sSp{cih8Yea zlNgpN>IPs+iP=k(a6gHXL(fK1W6Kq_*#^w|nF+O;!Ba*SR?EoMw_dq40dq;lty5V1 zT~XUp+8$ZA%5OY@?`odu*k_aggd5_!w6p_+pV=Id5Wb7g+5VN%>ZvwtW0wNMy}ZD_ zKO2-R=^hLj!Fh#KVjR%BUPR^-aX4W1kk`7dJabPj??|qcE=9kW=@qSjsbQ_vvt?*y z$fJ>9ot4=;pqU)M2r0m#sEJY4NU#IwEDtWVGN*&hF)u zkgFI1-2>>{1Ermv;<7+(P`}j?c?L5Ve3@cvvf@ud9MW=M*I%UflO&cr+VQ!*b=n5dEH0lKs>xxl1$P)nq{(*x$6#9EwpYRmyv>W;0YK zjwK5}a%r8W&0B(nAotNrt;>F73?NMeFrcg8ca@tISEW`zx>TH05B=50^28w|Q7&EZ zL$07G3^aUE3O$G3$Y6b(YHkJD7F32Tk)Q#GY(-ph$#cTW>W0cxdes3gmq=0El*N2< zSWZtsY%(u8DUiOJHvFftsx7PF`YUK6YnjWp0AH-{>6Mv7alBcl|LV6_wO{}EfTp-i zv!qWrd_8+#9U0g|p{(4mvaqdIl3!xE%A2xIb&0YhsfQP>zZTTAl^|tWgiN#%R`!oz zYP`+${vsa+l_-m{Dw(4;dQLA&rFQqX)vD+!WU-Wp4$FuH7{$izqH@Sr)*3)F>egmT z8w%+l4Q5Q+v8(upA#14v&(_j}_zVObQ^^V-LXG zCb+iG@??f`Y$Mc6@#nvO<^Q(2Nk2r0Z|2v`rZnQ09>czAIVYyd5|GaZ71l10shg6* z>4&U?J3hW(I^NCO=NPEcYq^>qebgk*X*nw_d~{Z=eY9dYDPpZ4N_LUGijxar!1ggM zwNrq|%3Emu`(BD(@`wO*%q&C9u3@kUd>26>Xw~oF$Zcd9h;xv!R?N=+K{-hYGR=ZV z(90tM8=c%WdkARFHFc??7ykPD7Xv)!+yRy@qdum~Sts5r0$JLw6}=h$lb)@v)|0YE z5lKQr)(L)pG!prUc|?Mm(&Dc2jR^8r(QWdIHf5z&Pd&{C&m^4k> zdDeCNG-+LpR(8>=_}bl3-k`NA*dMD0_1g(y{6RR+O4EifL`vU%*H^megQCyDiWgNv zmS-l9fNmcA;nGAoe~yk$4Tr1t+{ZWmqjQpXqHNlS_E=|&d#$*I$&;4ok2t-)N0F$o z>}SQ{Y6a!ko@HXefblQ9uUPcrKDq53BnjT)_HeTK2wq-{ds_t zBB(H8RiI@3Br-7A!k1?O%vze+>#8Vq0@)0Wc+8NI#xF(=HuM7<@!iyMWjv}X?8tn& zoXdEbCHZonmM+FAR!m^(9T^|q4ZG1v2t-0!M)H%>cQ66}?;j&LwgHh4X_Ytb?~Dng ze9nVSXtytZg#N4VNJ9V50db$=*c=zrh<_Uy&4+F~nPaP|G&7arA+t!L;@vi9#5G7^ zAWi;S$>gQ>i5T5&pTU5{1+;GO^dWCKF+(Yp-&6TPB6G3lNBw>ft?=E{*I`EWY@GAS zjB}H+!nX~SpKsEN-rpmUq7}sFMG1};#SgvY;)>FU?5@|`M#iOxo$NE^bFWrbj@(%H z)rh1Nmjrono?Y)cVIyfLUAT}*NT`=ZoMW;sl~3y0q-~VgoN#@Gaz5nbERGuo$>46h z6&o?ZHtrZ2=~^Qii#{i?51SI&J3wn-f@gmY$R(F|pHLOT^MOb`oPiQF5W^W1RwWD|yzr zn2KF+hojo5uTTQ_pw}#;U;4;vd}1LiEJ~ie(u;?vhwT?!qH)3@mK0IN^O|nxQ?%q( z;(MW@3X`_xzkaPz;2n(<`-G6i+=$&Y52~rM0*R zIc1Z(YxYy}m?5D*d^vLx#DDl+ThTnUnKc3gbHnCj?VDxUdetz;$=x7B42&uS;Ft%R z?wZPv&|&csSNr-JbwGu~4C>bfx`v9S3Ad(o*4FD8Wo3Y*{>8N97r8aDF~dm2-<}`& zE6KW+hB0KOb;6|f7>iV}rW_N?#QlqhK6cn7iW#qe9>{#4f&uTSKvrt5%|mq7BKe>~ zFksiCdA{KNIXr-8X(TcW3tR_$9i5cq7_&_1g=r?9JWb@&JMX< z3p?`Uh2`^w??CMTC&@pHoBf->=1xL>_;aBLEH{&}1qSzKS4kaX?VddOE;d22O$ zK5S7~Q3#$^j9)B0IK5J}5C;}E^FdJg5}y^3B7CnyI@F6AKWAYVSbwJ}G3CvN$P%tR zo6xz3H*!C6fwireRXv8g79pR`K!8ryIs$7xZY3LhYmF>A&wLHh*|0S{F%~^DvPB*g z!;eM|LB)^80sPH{O}!lA%P696;?_jbejRXra-{l$;B*ds&W&e_z*JI^){^r6wT32qBdI{Fi?Os(Tnu?NFmXGf*zFt#6j5MWu@>a*7z2BX zi@*D%`#FAAouiT?Xad0~YHT}szy|Msd9h8sHfkn})q~aWh5|loU1z-z{5?l*i zYbengI4sF8Ra2~2Kafqd+pyuFa4U_>O^86iCHM^=Z5;Nc^euPp4C;q=c3%#THP+g3aB1^h!g+kzjaCh|Kx1!b5PgP0^SW44V z@0>_+xUi2M0^w3FjU)k|D#;fxIH(%Iu(|9@TM( zIC#h3q-u0NLqOv{9K{qTt+G6>(51)${#*)Q@KX55?+Rh=QV=Ahr$my(ogZ{YpnTj; zHXt{;<&)xb{dBUW+uJ??ufs@&Z(ZyafKNa8w(g4Jhoh1>v8Y9p%%2k0l-Iw0l5win zQ3~F+dc3|dq7 zqyWR~;2GJPgw(Ti5n(1h=mc^-C=~f|7Nv#i?B-VlZzMoOupLF~$6ielDe#`}VnDWvnoM7$bbRAOKAIhDMtP z^dZFLZs$NIJC|#Rc4wa)>JR4e8gkb(W|qbOVCWEL#V7N{ax=VTu6}9|N+*>OU5x-^ zuu2E@0amX)qKb|g;}xbC(y&dw%MLNEu?l~PMdL<&(s+zsSNBF^lwVvgIVe7fvzBLc zC@3tgo}HY+Nuw;X%c_D;j}?nQlY&NN9eiQtTKqYnr8R7#J95H6MdY&6#F> zD$eC6Rt2j_*`#@{c~rk{zorI?9(+>Yd!Pk9iQJp0mmQ;>2>CBc=NxW}PmJ(pBocY0 z1^Ep5<`9VhqN`j?Ebj=@W`mUzA(i(Dugv(Bl`%p zbsu9oMw>d;H4o2$%cX~Q6A{uC;(UI2LK1chEUZ&{ z{o1F3X5Z)s<$+i}@p{q8rP&G2-KmSCHC-mX)Qt^sWD)%s2>Oua!FN0ckG9nwTUgOx znQM79e?R*;{@ie)BC;Y!6s4Kon&8cUZ-Pfh7F+mX|6TS6NNs>u@z;6MARz`dE{Oom zubDxL0gVu=Fx1v8*IPiD_(RY=KG!h@&IOea6Px*mk4E^F(E#e{;KmHI=n=~;E7A3I zNG`fop&L0F%$Pex$&yF|K0X6sen2|J3Y&>qb*Kt5&cV0Pb`%G(3eVWWX z+El-Vho&xXnXmTy5A}&JrnGRHN{b`R;%NEi>OVbkMM}U!r@LHbfiuB~&gizJOZ7xJ zChouGX+)>E?&Kgy;Dy(HKlh#fAjoRb99WfVJAnx#FfG2=rLjTo|@iAT)9ZIRejrO|^C7jT) z@ptGm!R5}fPy^5Q)I7-GUEr=lDSaa41=pS{`toJ>{Y{CRC8bYHI8Yol8Q_?{9$gUE zJu%!GnE!rK8R8z_Ax;Ac=}2J#vOOWmy%fsDw4D^aU*t5$7CAm}HCwJw1=|nJrqP<< zhvYd=Eru(5hfDg}^L%M6^8o`DaQ+|=CWw)xSD8E)l*>sE29{)oeDB9JX;bNVOw~+XN&6ZQ;_s2Zt@i9`mX_2hV?3!c z3^isRr<&m74Sg6qC9Gq+Y3k@MrU)~fyc%_edibkg2|VMPwgN2jEA#M|j={TFjS^2d zfOoFi9y5&M{0;Jqujk-&SD|r+HoaXZ=K$dQQ*O`JUHb}xdhz3T9f-aq)Q~Z;aSGw> z{(7QSU4fO_zNrXj_b}pE!7W(6Mqz$&QnCu%8RH?x2Uyt}o*N`dQoQw9Od%7zSv7Hd z<3RPvm+q|}q%j`Su76Veppx(_Af?~kmp|Usl`oW^apuPURHz4sS)<3BROsXD{CLG{ zZ=w9PqI3`_U3=LTDHBybTl%>$ePx%8giB6w*s%o4$ucdElw|{>ilaSFUJ?CPY8^D; zb_=T30N6qf(i5Q8@uJGz#FWS$*Nwap83 zxYa*5oFrIwF(dr(+HQv{N-#@KZ&jY(sY$Q!L(Ko7`+!MN1ie)^`rdyg`)*--o&m z8ZyP66Xmm8U+)lA*KMe?br zZ<8zXiB%WHo07C>wrVi%P&Ob}Lz_j3%F3*6+e~-}NN-X}Ge4yZB`QF$?6=Y>85!BF zSJ*heM&<&c z;n-(P*-Zl^V7UsbCVK|Uho`kmf=jmcX%}sa!P@kp0`_1iXzRd>yFC4I{Db-22sx>P9ewhG2^nyqa=1dA1T5u z6I`;8{e9&Ze1xB(YZwR!3#GGhNi2aLvmy^-qaA|}UJu>Aj*E*MR`Ds8(DVP$Ecwa* zTJ}3Zo?p6^7#QcE9}gHb7j?hct~#ZxV^t6}lYjLCsw%K1JGo}<+RLbC);MnO!ASn$ z?Y<57JGVfcoh2q5b6%c~Yb&sn0xa*OQx?&I&@&jIL;?`9;{B@ZlabCPZC*@ed4RXV z=GJ(%!Vk8OM(Un=g+3W;Bs>Ale6JtPq1?LPY#+DOfC-&(CE^c@%1Spuih2Fii+W|o zkfkV2Aog-YMLG1N1DC=*ultfTi8E}3C0&XRXly_ate#epQ0j-cT$3-V%vf26@wQlU z5cSw4X?qL3-ie-7oIl>SX;^k?6V0W_4%R~OeD^S8Fwlq&ZTf%5E#3ZC*apg#5;Yp zwumek0nyQ?kJJdo2-n`HDk*8SA`({ed26e_3tpzLbZKZJUhDpqXGdvu{KN=&tJ-aw zSW?Q6$q3)mm-a3n_?+|`rRD#|fqj3tPY-u1mWD=cd<+l4(MZcKC_Gc?(AM`bU*Ipbn4U65|ueC#J1x9s={2`SsjA_# zc9^;-8uV!CpK^bGYf#oX*JNQ|&ba&~Jw>me z7==6E(xg1QaUIl`b#6I@jZ*eu&tSCJUtKk+sLCgSWyq|c*t6~ZT$uF1-XR@QQ53Rv z9y(0HcVuJ*-x#KCdV}lnmwk##SUH&(+~UpjYQbBL z|Lr>&BV7U|U7R72AqEL`a;Z_ztkDL#cYtu&jo6yJ=}9a)PY+&n=-+y$mX^EG*v8}| z#rU%4iE;4B0akEtB+ta>Dukw2mGoAKUqo%XB?$+83CkS> zX6kMniJApczgBQ3f? z={5>WqQbeOc(7R7dwyeILGSu~d<3NqCl3C^GXKm?T~x^O?WE!N+hK#!@>!di@2$P6 z^GI*laA)F_rNHLZ+qRFEl3x~|XaCJ6ebzP7D?dGW$26VznzHd@A!sJ?d}5(Kvk2Gv zx;-nY(N5U0ZBdtr6Q7e5GW@OFoKAW%!-bDaVmY`~JAOK_w5`$&&W(_;@aeRlEat*3 zS;%O|-ajz(>3CPMLl9b5t0qE8Fhonzz5H>g|0(?iwG;#41bcw=k)tk{yTyc==Q+OtSzx*RrmO2Fr zBNW8V1gkuDo_%n?0Fr2#`Le&rNB;>Z5?f9F-DT)(C^$s|6$;8-NLhpRcaplHoBU#M zc;|8qZ6{b+b+=b15En>7|LHpt@m95#-w!a6z8<*Lp7_vtCWUU`df@GlEMgN_8Ndid z*)KX|zV@hGvCV6g^#M=sAL{kkCn1AET)4Vj@~)M14k=MOU}7G!tg%n`!? z<4G zbsg5I^Ld){x#17OzEHTL4(&$Vpv#kT{2v5BOoj-{F>(eN6>&xggX=NPZI57h`BCb0 zl?6yM-s(e~ecV3}eClOQc|cT3)V%l}~naoV-dw&O5 z&jeRypEPkD43}`=32g#s zGA|@Y?nwTOWMK0_2+_O89)@opP591b`wYsZwEukQ>=TKKA;}hxIEi>Mjpv2`^Um$s zs3tHea|zbNc;A#dT$7$J4UMiFtge)#vPDP$JTDWy(^ z39G|geVVd`hSx;il>GoND)8ZAk3b-^M5twECkEcZriwjy%#9^`3U<0^|X zi}2kS9q*|3(m^H553+R8xosrMC_gOakRyE19n=8RP2V4+DdO0L$pr-1Nf$`A^U~L! zDn59n6f$n`1Iz&$AH>B{NhJ7@>LqZot}<%(Fxu>`p@ zm(mw^-!hc*Xwc+)(xRqD&gjc!Qun3uNe460Y0XKcVq^n^o1 z&Hh#LMO<%x;Ua=F>Y7qnSx)ht1=svrB(H8-T2$qO8Y3DK>y;ASO4Sg=K4yPEb(_Xs zEHrDG(N2nLRn@mfOzGc%bgMWtzlg|vmTq)e@t{7OUh&NPcX7!S5Do!}=po|(g-_t6 zGx160zf1XL$(dhd5vkGytv`FGbIDHR{wWVe^mwIrCD*pEC?}G4cubwK->!bKTt>fS z9?9Z@A24Hm8omU&Qi)yM%Wa+B$MS`X^c5V^cL1e?H%Uf1ScAwST_|7!p5Sy6hl9^2 zhZ6~7B3P<|;I)DSispGxY4ksqy|q>@Ah?{uF>$@VO8EEqRAFA_R$h=*y8?GF+Z8uFDSAr@=??v#F0m z@;&_RL}cRjI=MXjk~N4m!sFA^5KRVpF7=7%%FO(OOMJa2_==J~29eBIfLM#_W!x8I zJ!Gk|O+?cd{}P2wdJ*<)dF!$#liwT<{L(`KV6Zp+D_RNh@7z~2t}ysrR5#v}r%osF zmG~Py~7UtZpHSUGK(0b`q;N2V@OYU&WQk^+ViZkl*?HqZ#^vSTmTY8 zwM+|`PPl`MHucji*T)>+vl^oT+D9{7blxQH)ILGM9aW-wXy_h$GMY~V^)t-5{!igW zT|>6ixAFn|)!D$l-V=Xg;Q%Q|@x@6W^ZUW+Eyuk==HB4<*^EE;rn{$y%=oi35!h`> ziX9-}ZOrhNhy3G<&HIFtK0QxW`FrUJGhU+530gsbeQq5@mKMJEwG+1Z(KcZ##YutU zlSu$d!v*e5KD+!u$jIa+T=Q6de;3&^B}c-pMZC>NFhL=mP9I93SA=8PW3^F-tQxVqta`7I zPCKzJDl02cdtPYPsI*_UQaf$mG?Qred7GE06uq?38*qVMXLLg-I2x?W>T;)$jl3fG zf)6Ad?6Swg25AkLjO2EcIH?R$>Jy>|Gq-UK?O&`hM6J3Q0HVL8Ik=2Y7yw;%PfqL1 zVDZgY4ykL4OuGik7W2-JC}pK#VOeyKYawvP(5Nf!KBI|%*F#D=jBDMd#fnz8-Zfcq ziWrYYFpgUHB{krO!u}ab;ZlyvX&&_-gb|e%URKM#3_?+9jE<0#;BBBn164tX)vsxs zJSaA`cGTT0fkGyf3nu-JL}p))acGxN9wikP2zfDcAUd5WVr*1*Je=>eKV19}D+K2U zvU>Y7SDUUaa$(}dKqNTbqDY{L@c)Sc`bF(>_geLMI~u^w!k28tC&9jP#PoRSh(iFou%LD2bKS(` zT`rMo5`6@Y0Tw(dl8!v>UD5?d(XqUH&XmxNXk|wtZ|n3k%eS*6(Ad<)z~q|{te!<1 zIYq3}UW#K`$sDbr;SX+T@*6>6(oCu~%1XLcm82SSSoN#K>VuTA&ZkXG`;T*M$#oD> zO!6fN^}NavE!2BYPkKs=N-&bZ4aE!f=+0tcnm4;DX=UZ;xpPVy{hGq!MxoR?KfiFg zergUX=9JS-Ag3ywlq+W8EciGP*LZ`mHm@2^QX}x=KHPW zOGxF_=?wO1llsBwNoL#sBJE4Sp={r`y_FV{Hk2hrk-cOIA=#1``@STkVNlFawkeb% z6h+b4*TL9FjqpZVNMfurhA71_M2uyOzWW*K?fw1!-~afI<2!Q9q-J^U>%Q*uI6ImZ{U()DB?7~!gN{;)Up8ty z_}n&S4i;jBA|HZJy)yKPC*EQuX0B z3DF#CW=u_#ZUeJ02(4kM5 zT}D$MGM6?Ms%f;}4o&AOcX{}e?X3EpcYW$vY#wYv7zgx{(6FkYxY1pW9L*%rm3HM? zP5fwW{nRSel*$bKJcWJ!VbmGd|d*Q+nf3EM?c++UB4OQdGJ&SjX#NL{Y9ZrA5B#1K+v zJVEA|^o$WxvN@kNCE~L0C)>HBLw!RC?jRu0L4nR27*$&7QR@42Riao@tlz z=0H~EWt!Qtqm`GGdOeq>H@25Q*$kXCl(zy@%X9Er5Dyu+`A~eQ=5RjUQFVFC}+;8>Vqht>B20*uX&F&oMx?We8dMeer5)yI_glyq0HH_42YexBrjvv$dNvPtd%V8@@ydC@CQjeX!V= zz|$Dj?W&I%`uT1rJ+xzu=(#>j20`uxUa@_^D#B8qWR|!q@p`yyx@qSvYPM66KmHOldbP4Aw(79?49d%jlHk?v0cDB(bTt;&qP`}xMV zEE*6R9v8UBb!138qIMLZzs<7^_lnLEMnM5)+fV_YH4H4UiHB3El4qD62XTar@9a?z zSDJ({=A){Je4`&yf^CW>qncmsvW;}TZofoVlS?XvA8Xe^pcflDOC3Cfp<%zWdh(XMoa)$~vnqcwyuG$mUl5m0TI+7Nu>l~8!PTKxbuAppBD zk+*HI(iV0&P}p3~QTX^-!Z9u$SW%=ur6cfAYkebzz6my|;Z+KE3+w}*z9Enm@+HHj z`WTeqaZv*DByudw+BkL8IF(0z{O&6st9OM`nUAm{4pP;J3DGs9UL`4M-wS9R9eZf4 zMczqLhr0VR(+DXklKgJ{f&Gc~P067Vvg0qhX@_bj*GSCY`N%N~=AcdB^X5)7~5l|eaVd7nEfbnyxwBG-j;EHPo>V_yIui}f%*YL!Tl^$oK}fruTG)5tPOwB&e_|fmZKAXS)eFc2 z?Fubmd>Ip(I5oZyN)h<0SHHXc(qhzAQeq`xN3jGEfOP-&_R3+5CDQX+D*i9}-tlx2 zz=0EJ*4ro3)kBL20cCTJ0jORNSnUc=HjjJD!<+(K;1p7Yl>W89-aAza*0ep4_rr%q zZH240KFfzCLuMcgyi=mpa8(ZDLl8E7Woo?a0;?JFf+E_n2J{->E+QxjtwiI~gf{l> z{+yk&&3N&F^gwhdj8UD}5Tr{wwi-O7tockvnT~nFT;+Wq$9$OP5T^pW`5~)oz9(kl zlV8ln@(Y&s470hTQ5Ut#4>cY>sT0NHckF&tQd^{P>c!A^3xNyhOVnl)a4idD91W8y zUwg-_%5UeW#ukb1D!MoO$|S+=1!cqywcTBT`(@~!rR# z1gy59{=@1Nv)KPso1m7_fwmX>$xGfVNl#C5eCj8eQ|)kK2}F+o%`33X4lCgJ>V zlX|VlQ_A*MxGG2Lgm0A+mHa3~sYL~qo04$56-x_TD4yYHGF0bsP@G%vqVPng42^Il zO_;QkT{4j5D{bYm<{4oi@NuJm1~wEV3VxK5!IhJHJd##x_KAq`$_7) zQ*l!FL1V~+fH!LU@>^YqqpHxKuj6F1LgOsE*8{8=;o;Ik996N1vxpkK<31CBOZ(>E z8EETcQ48_nLu>f+451o|(fEMjKvK2$ZX{uL2oW=1=P)j2pch^-loay^cw^sZlB^nn zP@=2$`gL`h0P`@AM-AoL%stUw=xHNPaT+mmRt>f1Yjy*DFZw)r&K=@LF%( z_(t0nqmxlPFe>GO9rOx>+GJlKWhl_bs1wz}w0BoUOp@aQL}y=ZavXp1(VduKjM;!3 zGtW2MBO=^+i*QDw+wrU^5zgTn{b9On7=UwqvuI$uFalmST?_ng(y!@SVhyv&!{tNQ z(%Sme)3#C3b8?BZ%E!aB>@#U!c6o!aJCeT46`LxcvGD>!~;Fu>Yi9;TFoG8~GleS{9Z4%WERb?N$>Ku0m8Z-Q{n#f|I zP@7;bP1qZIWf-(>p9s(PuaZ;}^fsS)apr}kdAHIXB7WJKLv?QE9a>9CU#^maIrGASgFvMZCv=fB*Qm=sJ z4+eBZ?c}>57h}dQ;b3Jiv$C>q)Zh8Ao5txID?bI_qpF@dj-1u!1h7oFp?NePNCWj2 zzNE~HC`-~r8lUWHYcfb3a202}==2lu*+F?9nlOd43?=7FttG0*Vr{XD+*1k0mofTrYwXml4GA4gKGgPT@-+6hNVD2qlN>Vs9AyR>+%J+;D+qn%ip>% zp@VH0R4ot9D_W3jrRV%)+nfgif%lL@2IHIZLGjYrckm%EgTc$oc2ivsOq!3jD988H zLds03{C2)Oq^me#aFW{Ydib)DfkA$i5SA#VL5eaqrP>)Dh^Xqr;`jsN#;-mn4--C; zO1|3`NFhgsLc5}#Nn3HKt8mVb$aX-B?AMO~%AxuiPe| zlAW*T?!9jbRdz6I$}jS`Cl+j!R%69(am@33+9i@tj!U&>U;+p{eXV^;Av*?VTO=>h zA{IBlirD%QgbKB}%XJ?|oTg?0=8u=%WzxkLEH@7ey7s-Rt6j$NvPM+vEMVpCK=sC8Cp<(l>Fj{=JlqA7^H z*@I52*_c!{tsmYiKt8Yy8C>G5H;Pj8WSs2m8jsiIs)O7t3Rs__#Ih&2iWt9vhyD@m zYO8t&K7v>l=@Z2tMZ*BHS&wJF_Wvnt( zZ|8Q2s~b)At0O+U7AaSd#j^`cLczT&E)Jhn37~d;RFu)kV4TCRu9&KE>ZtEi0s8M^AsT%g&wK#K}HB2eYx~34& z&WGc8YB9h|t{@vRLbZ}`5_ z!op?mI0iIm@r@zI`zc!+3>O=QnYaR7rVDt{dAZJEm+1OS#4GCF6(4w;HPWID3Z=D9 zSyZJZq~&mSU3Yux%v2<``~@mFhq;dRl7}}7gd7aT-PU+L#te4U;a;rqnx~;iOdID? zFI?|LDBp@)b@wyp-mQcL*%=yAWW0G<=YTyMK`g+R^$T<%I&vMimbVg45*obb{z54@ zekL(DIqoEYE0z0rtJf=&Z)|cR8$8a;DSLnmJiDA&na4MFp6Q4QAc%$42?Y=WSdaS{ z0@cfO!es)GW`i9A2rglDxLW5T&V0Uc5>v;y_(5Ozj7XE8=bIo+R1;(7Y2XX9vOPT}W$SI_Bp3|NQWx1q>faOKHS>|qwa^oEO$1FVj|md> z_I>yf!ATXeiP>_iq6(Lyj-Ns~ISfZ?Fpo>DMq_;u{K1x{RJkaVfw*?Jye{<0xmF}6 zrSuv~!Q2xabDny+T3z`koBQ`v?#MoAv6n9L(pO>xMJ)L^f~=!yXM2xIWfVKVGR1A# zO(?(=jtDxh4i=!V1uJdaLb2bJ6?x&n;hbCc@GMC=>0nu8hoWZQU5vG!nXL7xf_iGd z1KwJBsJbX{`j&>u-o?{n_)Ruv?kcw&P2PY(@6vlFc%Uh~?+ToFfsWuc7ik4bLz?v| zoV`>MEKXkz{6uaoyUBZUli^LBhC5jf9JOJ%yiZqxcW}GBoa|FrLhuHJOfHS`b%S@w zJbJI!My8`0*}lUP;6Zf*jeQ?BZ9^=dO;Ef~FjOUzWsqgHqSP=;Mi z*rH`f^VV63@7cPMN6Y`tdNZ@v@TM_jJIH2;lM3Tyz}OW}fZ?WFzJYi7y_n;}#=__Db?xJ_;JlhLy_ilIPqj!7U*i_pWdPE|=6< zL#dvoUqN@(bUq9bz~4LM)00lJlVY5sfK$J(1u++f_8PW&!*;ffc*F=J$c^!p*GzX- zt4GW(=3oBFhJRtX^HMjO&KEO8P6MLLMbc_NlXo%IqtR{AC>%1U$eXRnTa)s0lY;VU zx>VqHaQL93uE93)JeSNVg?TO8jSX7a7%F_)*~f%wrngaa2lu6ya?*BIFU9-r>mBES zoXNjs>sGH+sic;#40^Q>^M_P^>)X-qH#K&OGy+os(-@Q;P2L#cu!9r$AY@PbE4KYV zzW-<$t$6M>@u2ndgLf{Q4Y%$>K(>SoT7cFl`t7ra$IA@p(AQ{olQGW`I2Y(`>psiT zTSXhayx)LY(qQfFk|$tAa*2O;OE-@sV;fjy?kv+WE--i8Ndcue(xh>l;OJeFZwn4x zO%G2sewkIO^!5|+<$-i!%9(NZpirfcp0z!(LAAL>y6=Q~hV$$i%suPmNmQhDU_ zrsLl7q)Yrl@?CbBv_yvxGyLW24FcVEW`^@dL#M4g_o28RE3yMIvo;qC8>^WU-d6UH zeajn%c*>QX3zh;_=cSk8trrp}fn6%UJ!)4| z0*kY)@zDfa68WmC!yJA+vw+ZpVHeWNO$*5A^6$3V5ONS!mi1=MUU#qTQ~diAU68ja z!Z>O(Mv2q!Lpz4$c6e*)bgl=mI4xc<8JlcAnaZ;71`k_ylEYIPZi3kW{#tn_NuBVoDK?~#Dv`~tTltrsdQzdeK;1-K+|k|^Mk zqwAxPV5SK4=GT*ji__V?2crkj)R;53FBC-Mr*(afiybej`)FdgT2OU#{s<1v_t`?u zD&CBgtvRooX*yWov-`OsB|I*BDqwiV*xZ zjppP(=KbwxuWBOgT-(BN7{KydyeIcgK#R!mhR!d^e;EMu!|^p;>0(ZP)cl$8FhMb+7Y4wcvq40i(m!9+ej!{bVzv6mz^p zu6Ta1?cz%+Cr|Q+|3K%23>3MSKG)sXB=GH#Ux6A=a4_w|z`vbB%?a+7`=W1CBX@-P z*w?OV&BYuZeEj!`u>6j;LeC2^ZzIeuyL2fFub6Q}H670?(aze0;>Uh}C5dgAIV+=> zrEhHA{gPUUtiSKAsMr%VqZYU-!TysiS_>X~isTbzW|L|{u9|;R%u6xNBbyP?Bln%b zH`FJ7G(j8um+Q%W=BC8=3*nP=5%|t`Y>hkrvIqBO#4p2-_H90>9l;8gfZ-X?rR$+( z;s2oxLU9!%i~&9O663KrhUt0`n7aILaj>2Mmj5UASC)g-0ZcIa9D%({O5)~%11TXy z$v1g9RK;L>n>lQAPHsVi^SmIxXUamUOw#pfLNu2Kj(?WyQ}SXEE{${<(N$AFqJuI4 z=bcyBNW9RmX|=vq*K{G9xQtIA7f2@;_;#Jik?qKz+1#&Q-#a2yPpF>&+CjMe=Irb) zWUo}UOj7L1xm%=ORNf0yirikjsw8^%n53c3dmjFliR(G#9q2D9VPzKj*4($9U#O>U zE(Il$=$>otrr+Rjqimr-kb~YBZxL>iwnS|-fMt$I$ehD{HaKHusl8pH`-;^Y$8$U% z$6AQkO455Dc(u0`Oh#cD}r8fb$(4`xH{-bi>&jE!+>WvUj*Ew z|3}L=3=|r^_%~g`-jm8KAF7E9n$XG;E<=yPd4G+G@Yy3nJyfmE?C(^O3w>>6{<}Z{mf#so8$%$ zv_;rmPs!uv+%_CnyQ)`dhi;u<$BA82z+>Xym024dQ5WLEsh$i`y*bfjG!D)TXkNuV zLd%&*s*cI-0^WpkSxT0{q#o5M`{weY3Ez-%I{$l5@wW_+Q2htGHt6;iG%c7pzY@s) z2k?NE0Sd@fe)DLTN((>$4)~xo?`=^kir21kqC|9_WmL8pVAK`e-K!!Ph@G&PeIn)o zCMcb!3oxhV6bo6E&t0E8kek^RKz_=@kL@?I9l5UpZ*^QH^{5|r;)CrvWyMfHYd33> z!@Fo=Pr4(p-s~{fdUNN@?{5~IB+x##8pM1biyjQzr`F8zhNXVz_iS32bZA*v5^%#Z zl#btcosHe=)^P=J)R>HAv7O=OLK$a|AWd$p`Xy{v3HV57YLN=Dzw}|vScp7xbivfr zMhW3ubReAQlkFAC9#=9+>WPl;(8_E0idybH_F~!IxP@pVIdL!}ziYu}eC~Bb)|}nN z=x1IJLLH?CFJ*Mqz-dX${ep8zasR<`caR3}=U0JS8bqQ;$iwRZU1oDWMo<*(GgZ=o z7XB=xrU=3$|0i3*K3rGRWbh6MZ9@QogkS)oJL~unqNJ;#(w*)I7cSmL-)r+}mn(~M z1KuyYW%DyuFlPC$+CR+kYUE7UVxVh*n(?FR{+xo`+k~wOMsfE?sydpT?Y(jcZ*G=W zcRKE~nA!CGCtK#ElefYybA5eyHeNlHkTSlB?bxmMb9O$}78ZMZ#@K~x?P;flWR_HB9+#WL1DC zRaMpV^4RzUHvG2Cem!JQXaLOA@LMyFzEl$VvEOsVWWUyYl-iB}p;Bb073qbdO3HXu zLZtS@cyei;*1@-$pHfG>)3+r;K@qaKP>^yk;fnJcaPMotwzon4xtSow?JAVI2{g*U z{$+130m^5X0e5tDc-G7%WxGJr9hr9%aVU>oJN*ejhIAvhu9L^RS55*lga>I-gmEf+ z4Lc@Xlbe+v89x`SR}oCw=?3znMi$S69yWsbu*ScP!eNzPM|F>|e^A+cdFWKp(=8#T z;~fqqyBu1QA|<$e?4_U~;_TO#Yqc{vw*w__PW8so?WE>qUwp)?=_L}K4va6;)V*su z`^Wb^#uOctz#(^0Wav_-_=7slU$a%wlXHksU$>k-e&1N@&N-96ti2G{!QLso?6^vN;{I6>@LDLcoLI9}qNNW*Z84W4pv^snmaFlso8# z9*>HZXVe%1`ajt`z4+}2KiPKCmQ{zms;DLExiSG~W#uU$Cx}1U=8bQ6Ao-Z0!{;lw z@|&DXaunwH!xwT2B8(SrDQ;KIe#NQruo_nUgrHQuO3VViD+&R3YKE$kBj|2-+ORa(by_${I`kkn^olub-jBS*@;xq>N3*;(v z$W{8w&PIWN1qQQjKugbampwcUo4aXregW5k@toqC@iqzl{D|AF;*IO8{`)Onh`6r% z(x2!G5%~s|L0ZL|+{{1UhsEj!s2#~S9{4Z=mp-H$$D-kDTpUUm1rM}^JK#CNcf-=m znNF;P{$Ji!M~xy54C!zTI;R%q{_5i_!eE&;Pdy^ zrk3k)1>f6qsJ*VkG{L!qMYO)D;nl3{0%0D-BbJAkdkv_mTv?VyA}VvuLWe&O2SkLQ za}*uEC8Wz!3HEK4E8gaySr$uT}~J~S5f2Atb36Nr@>;ICj9X!(HKmQxJ!Uy*EvK=8{n3|Gry&CI)8N|$0Q)?uT1A1dj&Y9^Q$ZC3l>VbM$4{?Q3*(0 z);Br)@M|!)&+&2(9X0KEskIvLwZ;mb#LM>0mJB~qez}$hUux)uztpKJ&basz z9yg79p;PwpjY~%ozg>(B`&OG%Ml)h#BZWQ$rIeV{Jz7GoH+8CS;S+*+Bg@G}3N>DA z-~RwioIQJ*`C7j9Ut-dZW$Htx`oBA>6IaxHhB$|nggI-gV+Rtq=m5D)H*RnIN3FUC zP%io(CJDV4It1m3mW~bz$EqzGS4)zj?a#w20PgCa`ZL~cnz;GTA_4#oq=kGV2=J5p zKoSVd0q9kUfnP$08^mNR00wDZ2aRa6HhdHbfCMgu3)s?@#aF z2!SC056Ek{{`4A1TjGBIM?9QwJ8Y(v#i2~Jr}q2)WtnrBkV+BniuIf4FLXA3fpE&U zlo6L1o0#9*JM$ta*5cm8Pqxlg@Q=!hi&G?=eV;>}9@Z+NR7Q=eXjTO)X!%z3rFK&q zc!5~p#u|)X;dEX>_knCH%v*_2J*?ds!@??cN;*>Y_~NwgpVQ4!)_q(SJ^wAmTjwjtyY6a^Ne zAx;UU-fM>MFS9)lG0`={7iMxij%X6NVq}y-vrwFUwdW1@EqZwp2cQnmli0Ag`&^q-?O2#qbFgx z-E!|?p0Y5}y)*BD?h1GyWZYyRA}+@LKk0{wdeUds6hy<{f=seZc)P@pmXO&PDh-0a zuYx#xH6@c!kR@QELT$JE^_@~o*i!YAEz0*UK13w|SNATWe~nbd_hT6MxpfR9p-Anj zE8v`Y#42up1{TH+K*?L(xAXx0y)UpX(!=F@%jIBnQ_bA= zTKli2QF+M_%ayF3ys%4^L5S$eF3A>wJ&zFd3LY z)xfaog{>tIt9q7W=JjD=Lt)+IY4S9w)*6O_GJS}|2pml3v{+@rg&zVU^BX!4>l*JE zB&gR&v&3}h>iph)@87%lWtYna>4U#t{OQU+GxWboW@-bU;fORo0Y&19T8Nt?uSvWo z2$>hkQrV-zVy6sis;&Y3T9(~#c4ps#=v_6WcGQ9!U=O>4#UGqVEZXR??gwU z$cJA3MenLE1@l44eu6k%1Cq}i>uHbPegiC|$!*WDXZlddp;<_V2FkSqlsy|kP9y`V zfMCWwE&?HN3Bp30t%iU|x?uRaF0}Sq=ncQzAg4nYu~hDVPm{~VobMzoFW90L z?LAOyFWW?V2Lw)_$DeiFPe|A`)*QFX6?mMv>z!lg&c|s+L#VjYOnRTX={ZVLBFx(|F(s|O^LFy|^;d@z+g@@;veH2T0-9mI?F7I~2e*VB_rEzTNd+)H~S zDoa1({jn`M_|OA_;GK%YlZTB7YR}$Go~8uUJCKOFuGMuFk3c6WA$ASDRTwcSIXQAA z<%wD_azU_`aiYYWF`5%*0tC~wS`L`Cf2|^S`E}tCiid8CU;LmSRz?IQSyn9kbJ{@W z3Km`ZsjTJNUn=y6wuizt{y^jb4PAB(tuq@}Oq0vv{bwoxFWTf-BvqNsDj^1*P=0(5 z)4LVaUFUWpPft{m^K}fO3hw!SeA(-oxn=CaL@V+*L(r={*`u<-bD~>`TketePd4=+ zR?;Gz&RSGrXSWnZRn`lr?{dAbGe{}247Og7ImkS1>Uz`7hz7+u+dWW#f}%f^@gYFC z17Z-Mp)7?BWPE-SpuPpX5+sNapg@5UB5Lc{3#sfNF#TW0=X1t;gnKcdFFo6xGoS)w z`wyc!DsBWFhamm=ll`?77Xo4YEs{4;l` zy_MQv$3zppQ`2%!h+O118WDSmb_s+l?mma*Jjb;6((hvgUCw$z;S!1!+#OF@shx4H zdhN2N6ZYLtHW?I%GW|c^1}8&U9wfFP<`WC5Wgyy&LJnzHx-g-?5>7FVs#J@<*Rh^+ zBi&kyklWV^zaJq1&j;j}K)eH0U)GufGX|A+0FL|`)Te(58$gMLs*&?7S`1atKqJh0 z0LC*NJ#1x^%6_u#cni%L()Z?fy&FL|3_F-Gz}c58l`TX&K$h3kpcP2HdjfUg?n!x? zPx#v>Nct|H1ig2k16*F_$=-nyCso8*b!RiHclzG=PJQ1@XrQycQ(QDq&oXVWzFXHT zng)WyV*tPe zu9bA$77@|w`0`KU{Rh6tLaEz-$PxdV!8!S)0w6+26_AVYw5q5{nQl%bl+bRiwJ){o z?|uRb`LW9<8mt^K=E_=YYR3UeUE4uA;U3}mBr6KI`@WiYHtrtkrib9_cBMayCjg22 zFms=M;SOtn=pf2Q`#$h&7%CB--JpEQ_e7sjNA2ABWLrnD7%z?Wk-k?0sc*sR`_C*F zg3o_1J}5$G{|*U|D#pT=Km7Ao`F;m|sAz;QP4!m!>i!3s)|SdC!mzHT;dr1AzxC2%)z+;#lg2Z3;w7MN&qqp zq8;MYPn11?OQ9faI0@Wzs_O3CUxCQQ3t>05W1ghm)Z0!tS>-1V-~zXWee&!-;KCAB z{MWn{cKhw<%v<-h_`vqxBHaI)Q$i!OeHuu0y)FS3p=Q}6`2iAv!(^p$-|Ed4m_MU# z&gu;iA3Ji>hZf(TW?U>Wl(ipdiqC$imgBH0>ahYSmzN%OM1eU`#ijqIT=3aG78U@?RdXueYKg9sL3uj9_S^G*y%A*^xAY3qQpYS~nt%`O#=AL^0hOwFr@)>&-T1?~LBFSb{sC zck4CSbv6gqjItV;kjOh;(L{U+G$wp5gqI#{bAYz9DOVj19qf8l<7zu zmRBB=JVb|@hE^@wx`#%M(stG8kEjUb{f&>OXXAgN-{5w_sa zjJb*M53ecatJWS=Xncutmwa&jP$4}*Hl~4+Y+Rr{z$azNe94+XmtS!ZkV&p9z&F86 zfSBP85qy@KHkycoT7`~8?g!aT*?&8x0THw!fFK1q<}ppitF^INRy>pj-?^TJ2K#o; z$Upf)uydU)aURYw^?(ZxDp9_t?ryM<>1L&(;&h{grJdhy9h>)I-#v1Ow%d%`3giWG zgFQyGDZWq2#jYdZ@ofT-R4RIOhyp5w+lGo_kVcQ*js~7k6R54<6!dIEh#%)8>`%6^ zD)d`u4Xty^*k7frOh_giSW|rP)*y3RQ(Hkb0BV9_fh`n>e+vcSkU|6o1_%sw_IW;& zdc=#F@>Q@Xx#f4i17OxAo$q{}zuT+52@LENTB&_%gtEn>F>p-orq8`TX_kV3=$Cu2 zQ{t8nmw;tNKy+QP^3m(mPY+xT{l3L zE|@uhA-eswOh0kL&3u>x+~NT8N_NvO9wTK|`U8i%0z{>LdHqVbOkwapKL zl(krI&H1#dTMPO0NOce9f?=Wf{)1g7%-`sq=<0mW6s=Cg$UhG$`6m&_Gwp}otqQK1 zE+wv%qzQO5ic=#45WE~lwLvu=T1JUrCL_;P4QFT03i+3Mve9b8C{YP6A}Z!}Dm*CAW(ig4(2e@BXgE@h>vt>cXx$oo`= z@!jYvcA8&!JekQXg%eVhekq)wylM<8K>$(HqoCTt+T=d?)fEAa8w~z`7H(^qFKxn} zrntWyuln584CCxH#mrWB%S=`5@+^;zBhU<`bOUme9DmhbLFzHM3j4m~*4`aI%AFe6 z)SxGfB4*&~_O(8*2-oyjx;EaytJuZOU}2pp*iV25B))S3^lq#+osoy{ctZ+J9wh`OEjm{#p~3ynbi+ zP{0Xbc6*gX|TMe%` zE@gHqJi6&aRFc|GpplzMs{9nM0Bl0t%JJ$TBOYX7ahCPD@73gDRZnjpy9HR-MxVTe z%)IX-H%a8La4vYsT%BTLjKNqmXInJ>235v%(G@;hCFC|2c$+14)&q9Rl~!j^ue%O8 zlHCkhfqwlHFm65=i51+|3Pw<))qnOTK(#V}s#;Kg(m$P6e`HR-n=duHr+h|P_K!Yx zxgR!HqFDv)-a`t5&1WqYN^O`4EE_~~fDOJr z-PWs*8c5RImKi>LuFuC>Bwey|6t6L0w?K0BCd2?6C!`W`qi{vBGpi{N9MRfOAY!a; zeE*75jXmmRMTURtMK89zE++WI80XuOHWqZ&3wLc=wr-B{ZItvmPC{7hrmCB7n%iR^ za=KfnJ^z&2rBDfM_0I8Bgi?tSDqin2lCZR{HHxHfa@_0PmCk9oMmj8u=nkkA{7}tL zHi3Ma0gK2)v&u;N|I49wh9_2{H3;jscU+ECi(sm-f7%MYzr zYRBvtU6x8^xGMX(`n<|}{0DF{TJ~FT*?72&u2cWo zMhwGRTkx3k_(J?Ln#F*AuKQz&9mXP-*&R9ptly|K) z4`}L;qe>rq=dwZO78Ke9cpLiutL*@&GEhUufXL#oi+C=kE=^J5tOeZIW!59wU!o-T)nXO5i@(O8gxxQbXC6P85z#&u3Z@Tou`6^ z>Hh~Yon?iS|8Qx*I6)l~$eotcQyfh{e!v*r>y-k>c{0>+VHw(ext;7S-XX}V0IJYX zE82Hj-{;A(?@o=Pq_)XdObbNJ7wIdk^5WK@a@k#KxYEvH+{K*YO2@$;y4?)jp-k{k zA03wk2-Gp~Sl<-ogG=2MyDX+xy*W3=_*jFLG(l-$;v+s}c1_oPE(pmor_;CY)bYM@ zro_48^P+v&5Sn2q6*$e(cX0o%!~z-`2GNvX)q_1CEBaOH7Z|NxZHk-vc-z->%SO;` z0Ms<7AhV`cNG>d}yDUE?uFm_BS-QLRE`t-0K=Yr|+^+_b^#bQ2Ufa3{n&1%5d)N9| zg*}NaMFDE^N1P?Je80ciU$wL@*m1eD1@qzC!p&UDoS$6FqKJBo5$5|+i!^kX|LkLe z1V3>|@M8^;T)#fVpKQ?nfwk5Ca!S1N0>~l6`zp(&%xgPV1plci{OUCYFWz>i=p@1p zoTvUOU5`Rv{Z*a||5cv*&)Hq;efwu_`C9W=qZV&38W75Y#e8s!P41is1N>T-dfB(h zyPicTchrtMU+9j1(&Ci#cSa)CATAt1{;K{9DD=MDh_6p50&h-vzxMR6*BUE zW9i9R++(@A_ezK8xY8Qt7Da4+G zC3Mq&tdR=!wt*O#V-Mc~Kr~!y<6ff{+Q0X9L^Iw$wA2oHC|>Qr@q%N)=7_0J8Z=ZcR!*O7q!3_?uS_hd!xourX-Nu^_aR(Tni#t_1iuaIj12gwg?ZOHm3HqQf z52&bSAxWh*>-`hpm@+IaASJG(RB@wv%I70Nh24s~+BfH=rKouuSeY73@)G^C9M-)Z z$a@E&eIKmGDZYm1s>F2Bi@l5;hnIPOvei&Et17BPo7ZZTYiw_zUT5i9sCzIV+e~9t z4QHh_X{{W9t*2Ib(C~?4ah&|fhE@xA(+wahwdvsT^wH6a(F7IUw zoT+KAg{`)X{>HUjP;EIIafT#`z{CEQ(smw#yOmU4CL`$Tnm)J}u7&yZ0C^ z^;Md8dtG9Ug(&H0V&c}NxG|{KfAgiLPG?3~aqvl$E4kF_$Xns6hGQXbD~!6;kd8sS zk{z)w{ExlU?_Iv64)-ti7RYxe&aKtjM3dM0!=Q{FAgBAA89WnI!05MpnBcZVxR-0S z8VB?DCtm%+2)=Xe+A#AFmF^=eSV_60aVizfQTU@cV7%bVagS>^8+Kp1WI6O=tk@Zz zlqK8hHdD>13HD%VtyiziLCN)(d=N7dBWg-+fi+Y@o`!e+p6Gptf;U^xRR%o2k64vIkX3{cs_l% zOX8=Yq6n03)cbS#V`|6fH5}$+>#_y!-%NqpT9;R!D_ygZ$Be?ytMjoxA{OnNLyDw> z-sJwZbm2KD#mj^0##WFf7%q~3#o$HUOID!sicy|45Ps*$xQe~|`so(_c)Dy!!<8eT ztMHn1b^nehJZCcU;2F(aJ9#mC|~hGg;Oc6L&56vnljyFe9-wVFD<-B8;`OsLippj%6P}5tt<@uUkV92zR%^?z1%%qFp$k8Cy$R0K3(}cEM-B z92+v~uHp_V2mNrcZMb;OFjvkS|IJWnd-d^6%2K(~W%>a*%3bJklsDt1+#%*gpwQ^f zwz?Hh$F*7?jny9xh;B3DI*(k=;<(!j$t}&A6m|_2R`(~5k@jgYXX_3FNaUut;}2!` zUb&|c$0lP=e;sJW_mk}rWT3N;is*YnO3ja-usF}w_1V$4xZ>s3CYSu%_RTMF7T_q+ zZI(4g2HLh)YlIw~KP41-|7H5`MX9#Ka1^g+Y7z(Z{?$XQ-5xKMP2Ft%G75iVTs+(n zv;7I-_=T7%h|b>SdpmTTc^s^Hh4zk(^Lxyctq>Y*;h>%A&kZ%-fXNy+t;r(spM4mi zVAFskg)t>nW_W|~9LsoNEJPhMobzWpat+oX|r#iV`OA6iu;8a3kqc;p^QMlHYz@Re#N)y>cJ^{6Wj1 zWq>m1|3v#h)etQZNS?m&%_~Tv*#z7VrUrf_I{I8orv>WK8|#88%f0f|v!!jsWXb3y zxrPS!eE;Tj`r-I5)s?ZEa~NjTVL#bSSDU+Qrq;B&2vcURUHu5w^Hez0b|SGlwpP>f zX?$zh{&nOh8>83GpUQcn;8w%VOJj|o2-LeV!Y-e7uUHeM9nt6bAba_#-6};YkX_K1 zOX};{*N*=x$o^UIT6d2|#2yCf{OT)Ej`j1&G3r$I`S8_D;Mjt)F*_17 z#Fw6?minWn+y^MZ;4iXY&l~kVPxeP$^r?Gan_7^FnsR$Z(Rrd$V@yAN{wX;>_@J1v zH}U!XJt9iu0~EE?iUIFMbZ|g-L|3EP!3P%l-^%yMI2nB_m+#aYbO$Z|L&r`2~>9aIuVc; zM3qo@F;jgYB_$2_!!H;mNXj!9Bo~erO`a zVf99z%X^jM4mb~N_MPq3xFN1fIeD1zY9U1nQn;39??c|FO4m2_U%#*?P+3a}GG9ij zh?ryBMj0+I^4`Oa^A8Q(sYn!6SaEOd7YZah-gS z`}jj-9mhG}*6~EMtURYwtf$i*IP*&`RmP};M}CfCoo57F3YD`!Har-=Pef^+%gv+D z)mi818cvU^Hsr#<(@tWTNB)AR<#;mdu{Yo7T%GFN7~DGFDSZHLPGkh319D+TwDJ3$ z;^571;i_v-K_cL{cZKTN8T&+3(JfsCg^A|i)NV=l;!Ik>q@BirCAqE-z1*T%#4=Ps zj-Qvh=(0$jQ!;k0jyE@d96o7Pusnd0(LBd}#QxAvwg-~#DdTyB#*_v4DT%F$GxAny+YlWkLk3ZJ;~?`>1*8wSV9)K&A(ka^kh((WDc zdc3Bek}7={ICYs`4D{Os;*(nqM^fpBZ{dsc;&4*2;n5t|=H|D{Tbb$G?jLVCW6}>l z_%KnaJ;^&=n(R;X2`I^zXmdg`97l!3BYv_U zGGUg?XfwJ|OSkYKwP+;o-rRRzx#Fb1V4xnj~9%NTqkXQThZppJROK3Bh}7VA{ZBKS^@{eP-WEzvgSO6~2F_ z^0POlpe<*tks5qYGQJ5gX-6aGG?(PsgK^ItD65J)YhT!fj zE&+lq5W)hBLvSa!OMu`KBzPd-kl(%cyZ`szIq$r)9HytIx~f{Hy1J&i`e&=3B%$T2 z)d5K-z$M&xPJ6CiJifWvI^b^3q=nadH{-TsVU{o-ya#uEAG^=L2N&nH+CKy4R^0Ji z`!~xRgcldM%dc%Kd(Rd+%v)f|v>96PpSev-xj=BRhJ>HW`u7uQk+-;^R3m8PP4LJywkTZ^e=jxcLqTFD-{G{vc@cZt?C^ zD};&)v7`gP)WAm77H8$>*YN8FNkXPB3t*cQdXi}h%gJCSfhFa8+xI#u18^(ZlpFG- z!;x(C`_RPSBPCr{@~MMo3l=w9DX6-pj?KVfn_(?c=~RhVLiioqnjO*_32q%jnW|@H z8UU&{Pz$8PS5v&!21{3LasiIAR|I2OBX8K~7BZg>vPOW& zZ04#pTX6PT2YsuBv(XmEVZmj1=T(WNu1#@OGUhXUBb%ym5dWfXYZ?C>w6yTkuhbOV z_#nkn33;i4*s^0CyH}>s!-_<)3D`A^^NEMAJYR@WIL3MMFcwD@zcV$CRaZf}PN~JP zng@=nF|-~Q{C1J;WN3X;gwtsRUxpujdB5IH`aT-171*pvwIqpZ-R+>~81opLH11e0 z!pyj>g#Zh8`5w4KQG)neUlJbo z6#~1Gfmf@xF3?XsT5s@kWV@lbX6l&09C`k4*dRhT zOd-B~92-vweAEI|X0)R`oT6>4glubX8!y4$pMteRV^+sGpnBS6e{KKG0kB!SdcCq7 z(rN|l+5y4}Qc4OwiXl&S!N~@>Azt-xg=ZS3a0bfDLe8zZ`hnXR`h_GK>RAh$;R=}5 zLZ)ii^D%vu!LzWGt#Q~ErXwSHs`Gd{)#tpGfa%ywg@bGO;L|vJgo-kUc z*wLv5=nKh#fS(rGPp-@~{3~Ou6l>`hfVY*Gy};r|W;m6-Nu^0;fbzlCWx>cbw}ZJs zHGj)7|8XvS@i+V*<0CvEHDXvE3;9158hn6t{6RRA!qacj>X7>}oq8Jxn%p-qd?@`i zWe5Fev!~dDslLuHT9VcN^?}B~tWLi=(Zo@knn~kkF;GLpPDweuy0CuD&L0fd8gy>a zGCFFG(VH;3kdH7Lx1-qB8ks(Mb5Vueud`?D=sfLF-`pG@qA`9QXWRUyp1QqZZ*kZw z>X#O9whDB&{>4?Usk~Xs+;xL(k=V1#1P#MuiqG*20UagY(` zTgKwv5lJ$tEn#UPcndlWo)J(1f<({7TtE^H_tblvWhF$Qj<+RgozUG2s>O7zi9G{R zjQP*tH~Ubb>5dskKpxZmmB-k1MbrA2oX6`H>?Ozp`Vh5a0@?PalpU!uamSBAn)D{^ZL7vBv8rNntT?`A?bD!SmWJcwlA7eMIzfm&7->JLk~Ar3VhP`%5WI_bmxa zN$A7ENrUQ?s`N&Emy+SLB7>#M*?Tb;irj$8GMTLn?TvWW3Zangtx|#XtB2d7bMLfP zIe6Pr zJAg$c570o@!<+OYi%RjLq+h)GH6r3zNQl|R z`K()Z4y$@w;QF)$YBT_2|8G$2ULdgMW^IQ&qkqm=1Qrxl=-NHv0IN z<-F`4eCcGCQ~~vYL>|tSnGJj>u5}!yqu!b-a9FUoPS{b8z;)MiC~_WReG$m!*UV-5 zEbi%4!xvs4TFH*(jeP2J4}OjsTE{rSzA(6WjDtQ2FsC#CdS3%attH@;8H1gw=TP`{ ziyGBjwg#Y%jMjNTVkQsZ;E&QWM=*QAPlW|$h++xJcQIeU=~XiECbg#ypFWC8}p z*I%*k6<(g@QmKV=IkP13I?xU0E68VbTQ4s)`MeG(YI*P2k+vcwl-61|`tu9?aBx$d z@?+HPjeSisl7k+J8dkK_+bC{6mIshVIb~TaP$CP9UN(&0Lzc7_*k3I2n5e`5gD_0= z8Xz-O(j|WK_|onDY)}u#Y`e5tHe3-Uxl#{ucGn}hdJNV|NsS5|(qQ8fg@a$v#SX`} zgtI@|jR5dwKcM*!b8iLSN7NbrC_35hRir`aRp>6PeSbXXlXskiJrt}{FxmpoVC>%cD^t4M8i zv=-)=n_?=)*#8G%*?m5mkTRs?S{aDu=zsL>>l%{N5?uFjO7pyQ%x(MsCwAb}3~Tuo zAY$Vq!a!&d_1TV-w)SICQp0-x`j)>MXXumdLEzhOlffPrQR(`#>=yOiY=tbYhXbdj zceR^leZ6sj0_I;;Bh+Bk6Lw3RBmyJ7w6Oj1DJjz@VWMrKM58ak%EFLWyudr?*62{Zv7mkipYpFPls|*_hO0ygk`p-KuBig^!e{a%2S6=WOl6N@)+YhM z{7Wo{fYrxZye>B6w;{`e&JY*mak2WZLcIL*z`V{YkqJ!H3BW}CSO8GQ2ij}zK2pza z@owwOpdGL->Ao)Gm>QdHlUMWrCjBO0aVPc*)W3y#L$cCf@lOcKZbl1li-D8jT4>_4 z6jNY!@T2!qd7K_Xp!j<=CG!p9@ZK~)PCc%E^&1?UNS1*uMRdT%H4^!7U>@2Z!6~Gk z*0UG^?4>}VskgT1pBXcO877Hw{z@DWTl4=coF7$itCdWk=s?W?jNO$TfTt#Zgey^r zRd;dOPwk5e*iSz()_8u*QuLGSjHB^$1=QMD ztD{!*(Z)rV=>?;6niVuYj9jXFo$g03xQjg8mXz&|2v~$h6EB-REo{&qvtO#TWKzWH z^Tp5$%BZv~ki!T7fVdREh^C-A6P4~q$GG3~XoT|O;RUQuD{n0>BiX(}%xg{^8@1 zjKhcQQa`$XsQ<+rn_U6q%l!}X9%h!S!%dHttVi1yfX5uLiM1^J7e%*!QS^)gxR3+5 z_*>~0|5fS#5CdwhIt4TaE#uxJb5Vbpll#kDCBPhh!(;t&mL9nI}9~;_U>TMBOa55vTUI)!e!nEB>7(=8S>|tHWAL_W{kr26NHxoMwSf9>C5? zyR*5p*7_gS3_h4s)oOW}zS?_Qlf1ti+9;U{tyN`4{<+HipRNVgs&d2QFyANc)9k|& zFkcdn-ArFWb(yg?LN0~DUH!R2%iRTE5|gC#PnN(F1t3vqxknBpqJSJA!T8^g3Lb$4 z|5D_V8!XdLF9He)U_P=!$XH-y5|>_R@t4jomA%KO7LV7=|I?NIU-|3@6F}*_uv~x; zQ2G(Sg1X#T8$H*n!QITc|DA<@aToST`jPQC0Q(4kN@1kSheehnDL3~} z9WP@#SU!RT3R@%qfU<>=<_05x1ij6}Wt4e_!J?7b=qL=wQ z3-Q9zZ2mWys?Fq_>{k-CW2T7GX9jAQ%K zb$!e&7q{o#Ft~HV9%$?Vb&qM*^}EC z(JYU39~vyA$hp+Q`_W&&S=d6YYe22X3*PS_QX&(8NaYSfY#YDIxuKEEeND9&&rIz{ zje4gj8L3vKLdM1vie=$c2X^tqE1PCwBA&2GjLBAUpAn4>w4{zVMPjpGTr-DJq+U#b z7H5Y9cz&Xvk7_uk)fUeP#RK0TZ_oZ95M-^}yuV*r;=gHGAO3yK>St}$+%AKArT9(T zB>Wl?OW5XplXrzgz>k^@YJulG$CswHLZF(LVPRF|bp1$p<@JX)Xo;?_#5f;R==yZY zg6)O$k!K@*MSq@DW8AGgbl=caD4>Y5icCCj@eXUW7k%|Bo|Lt6cioO?c@wW$b~TDx z-Wk?Q&ZIy8kSO~R&;BJuBOV7#c3Dq2*ol$(-N`0gm-=;GgY^r9GK(O^PY7_DXmyQB@IMW$aAyh0Ale)F%InzX z$jV!r_u1o9gtcM~SjB2*IpKkTlh2K7akMk(D9WeYa+CLvcd*^r0@CPs6-K5qv_fC% z8I0%jBV~ZIE(IRh33rK|u;)P3vOSJ}{ZV{_wdM6_Gn?2m;nYSud_VoH&qJPRloFo% zxnGzU0x6U~Y>kO4w%S#+>d*}vpz%yyJ)h6N!L(UmiDvv}oujA_Z2sE@(K3EQvt*Zh z1H5Ek`$dQ;=tPP;P~N1THSrDqDnHZBc*r_S_S@#DIwvE&<(&0TP!CG|;3BM$!$NZb zVYL@zh8YUePX=f%(>C-{9C7j_|(Dof!G<*!oofgtD5Fq4NXF*Xq(ZyC&Fi{ z-#*4P$Vw18()OXre`QVHPRz9lP@+r2nzOJf!C?rEPgJY>@D_Qlezd1%EIQ6hfAu(a zs5Ejo7pQ9T^C+l?@%#v@cCShrEq5+Lgg0w_u8o^7H)5Ucas&IaE7EHDd!)FeMA>?t zMDhNPWf2>Y^{pP)5ntx-?wjW@=S5$IPZR;q$!`LO0-#}n5hC3JI`zu(LcF=+2}16Q z=eIx%P)g$z3o$Ff;iyIaez^p~i>xmb0eK@s9s<~$q4jQeQ4;{D{?#R7c$)vq{NN>J&amdm<@@O`(vCw7%_H2b@;YU z7SQR}yfZSfR=-p%2<6z6Bb}6D-jtEk%~9Jd%zsRUui{svb0frp={lTHOHgMXZ65?CqME;oNN0=6Mfr<0XI$~1IsQn6)B|&`WiIKqr?Q!yCTh=KrQQ7leASU#xnCtfe=89TYNMic?+1O<=HsX~4Lj zVvAI(OWnS?>v#wZr(43FAosMbd7^PGWnh?0_oOr1qvomFX_X?aLfY?pb zNUT3r`^!|b&TOD0-r6t{SsHLzz|mVFmHr{n#2V)&UH(0+d@Aec=1nNiQYW-;c{tpX zLhi#;SEz1)(AyK0idJ&F4>37?9d%j9@uW6HsL9e{y2MYS1;9tVrUU4ElzR`D%~W$Q zW9jMgUuAX=2JB50;;drTwT;*cM-x90hN6TN(F|2yfrN=gug|u#=yv#prB#FtY~x53 zikd}OgQl0-LWf`HB|l>y{f?p5H>$(fdCD>!wHrHCg<~7KnEjs$y{=B=*(E~64ALI1 z{9O+|B7?aiWj6g<(l5#G35LJ#cIsNFkD;6_C%319R9I@290-I*thWMQ#9?&~g3**q z%7)A5r&OJ{bNp*{vvEI~LU8B!pFI?19Cus=_W#`aoHs(=tjone`SL?tX3vj|5$tZi z-<;z^RjM;8Lq&?77AM#Y!j|Y~m+LYkdm~$4-CVCqk_W>yL-O`8AQJO2%Ws(nkUe@T z_-BT+*+lGX#6H*=6(u9De`r74EKqgACv6c==U>(>(b!;>-z(Exew&g0@`;6%P`*yt|c!d^8XtKPy?0F=dFOfl`K?kO9H(jB32P$y*63VVLv;x%Nt!a1~KD(=)#q=t_nP3RJAlF zyibUaJLJSUg-}LawZ`rkK55p>kYUXP@ucD?uHO~CEwHNx6j@zz%bXcN5tzb@WSKd)9p0FFZbMVG;U#CP^DJKPnhDPcpt6azIU9ywgFDKAgo_aXq zl8Qy2sCwF1;2CKJ+N11eONh>U-TQ>x=%8MU;h}D7*;^-X@&;*4yBO}T>}yLo-cY{Cz(vl_^Lws4 z$0t&_Rm4J@WSTUNK{^7PZ#>@N{t{rXuu0WFvQ3!=)sZk&V{X^(-$Xlr16U z+(J~e&`!kD_u-&DQD={z1UVKmh`c}}jcpb>D}p=69=vibsZ**&Lsw8{O0lq&>cq7^ z8t`si{D5m)A9gLK_5}_VU)kFf%b4~XQCQ~%DrPO}P}DI;$9I!AaCs$U)%N^BxNg|- z3oI>dlzO(ddpGrzFN_GeZ83+*@?;*}K+);zfuS zUphi#sRga(fc$Z*cfnHjN+IcU6ks2!dq-|R!$&X-fa?Bq=1;8NSJ$v|al5zW5n^*c zo4pHms98g?sS&KAR~P`cjv*>5IOSJA5Q}Ovcc6fzv7H)bKe6#))PlJj#54}$kU#Q$ zr&uXh8zs4>vDz`xeE?7JW~b>sBXqhd8>Na}8{5o#1&Js^BX2C(jaPp8VH77SX34Xd z3J9o3+PwKaAsA7?w*BdjO;~rYf<4T6$Fo?3WGnfplPd*leAfbcH>Nz}IS__{b!=QY ze^~klpt!8SE@RlI*R0FwNd@=eqw#?8l$4zv!R&Ycx8Mpd^|gl0i{pzX&0;gXG=?a1 zHgdh8J{{he8+Hpwl;~&T*h)*(Oucn;HiABoYme4y4J3h|%0qgX007E5iy7)4G}D_E zh^@5J;cbY}j$w$cRHoo!sphWP4%y+Ss$sV@XO`lBgGmaYW)>yk-+{zd?Fhv5KH}5x z38mHuWQa|!Vb`pI#8z4p^wmHB4hZ_R@^pA>HUg7N_OeRz53>%DFcp-&bxx|*-&n!p zxLWRKWmyT9@AT^CE`Pv}n6&EJD{R*t`ux8|8g?~=nV*8c+9NpSb&-a#$*nFy)J(ic z()Pp4t{JaS2y#kDn<#$~wCFQ>g2_`|Wd%@dK87KuK1fS%=T?m>f)BH@Q!MSh*y2WK z2*UcV?4m-sNX6u4#AhSvqA}0uzemLNT@@a3K*$P{cBBK+&L7e`s zjvc(DRl?U}sqI%WWfLZ7JYil4{~!`!y||8A|Mh+Ff*=%b z_-aM2h5grPz{iC{7mt#t!>akG&8G#7L#bLuvNyqbBhKERUFYjJ%11=uTTSaUxM zXA*(=KRH4EgV0s2RE$vq(Lgp@9k@hq>lNI;UrO;Hb&yh_mUnc!CwzF`?Bpbl!&?vbP+U5m-HH~qS>h-c!1n_f2WT7P7I zI*ncSKEfG5%;JXQw0a7!SbBED?^oN`y9~AHa}?VGJ{4=gM-Mihee%Y!+P+k*5fGX? z&dDxoG4s_sV{;I@Wx98J+5KfjqmkP1dyZ4R&afjbC^Cl^(L5pZN%V=yUZ8IEMr6brp!IlvG}s&i=PFL1gO{oe zgU;~pst!(7szPKS<)3B=6Ai6~72>!r1<_ggnrBv6jW@OU*$)38IF%+Zwa2gVGwUqp z1x>r2gBPOK!w+RhXk!Q*6keBLAS0F*p^s(g^%SAf#H*`v>txX!o@=EVsW4fr^5og( zD0V)R?>dQ!o!e5Kz0yZVXyCJEyHCNab@vWp~4q3s55eN~}Nlr^E9DtmMC{7s=aYH&Qb_YPy}iqhU>M^;Vo zVAOGmr`9kJozX4j;zac@0W72hVI$zwiw0!cqaczu1GCIAPYsN;kZt*hx)v_i&62_`9hJRS=WT-U!gA*ev`3q1(}I*r zT)mQeYKSANB7g3!Y{Qr01?je;%tB3c7@|noUx`;L6Z8N{RAS&x-i!@jFYSd<4RO?C zD!QkJ{Z*40aMfXme1waEVG;m&KnlQDsSH5$M$kMmYia&ikRb};smz^(mQ1Z(!&&AV zJ0nM;!^`lxPKukezZJ~>rcs;MY4tWw*+w(K&+tmdT_;;W=|%E9nmxv(nlQwAF9Ma; z%3}1VCv$sIO>3jux+#L#aEO_&SnKT4Q4JW$b84r@nY#VBX5}+sA7L5q)TKK(S()$K z82xX?OGPN<=uTNt(eIA18nONxj{4orQD>+<%ddti+#L)Etx|wYuVaNZUACHB$y@+@ zLPclPfZg-mt8V_MS0(|WzWtufU*Xt%j$alXGMF;H<#R0v!8pyIf!5z9=POjP@AD@F zv-`b#M^3aP;Dv6Fd(H9-s`--=eZ)^phIEDF`BDCfR(V?lW)IQ6v;7xgtM)x$R|6oT zf*YKBgZWMj8`9s#{ssW006-1=q;(;nH53(Z0x_)@mn__3LWdT0whhRekYw7AZ$c-+ zKKeppCiSFUxUh*9=1Z186M(x?=v_j=;3XX3{Z zCZ5rmr;Xdh(vzXvY5UO2&lYO78&onMW|gRg8eiq(eBpvYM#ap_16}SRI!zF&emm@0 zdsRaQ)gwY2owic@27bZu8>D4PrDABkflA@|gj00d(jpN{G`;$_bK$KCKr-3hHMMoMnejoEHj> zHS^ifDu?P8XO$f`>zb0GE}6x*n93jaQOjz%^}YZ*FT+V6=(YI-7_@i~CnjmBZ-f*% z2>o!>ju$(mI{ERiN)csE-vY`AJ_yz?vzwChn&vK0QBKulAIgVaZin&pvkxsUe7*4F8$ zv7-?e40L>%5-PNNBM6~%HYeNd->9#B57l06r9;K)XEEXDSBJH_y~5cScee$ttIiVs zr~T0xQ(*^IxAY{_ze+vmr=L>Wqv2qDG8^ROvC4VkPt%!mrl(JqRLY4!H}c77zBj9! zEY@S}$Rqfu2&BD0!ndA2ytN;G?sM_d>V&}pf5Ug7J+il6f@Scia~c<9{sH|ba_H|D z11~#MmEW0R;_dlO|DrVy#t`<`zzdg`Lgy9T@oZElYZ?`h{HEo8Y5e;Sf{<=;yA_tk zk6d%kPm6}#=>kf9w)LZ{gi?jYZ~0fBx4lze*gEE1Z4of&v+Li$$2_)nYhHaWX%`kd z@lLGEf6DWneb8#lZ|E$c_SRiLie{flU#+*?`|>p2A*a$-5vU-c!pHX~sluw;=lie^ z5b+4mh2`RC?Sv!U3ZblLZsnQIebyp>9O#I?jC#<^KJ+xEaz!EpWIf?>_D@YgB9JI0 z2rEb&XYwdv?+r?Q0P&j9ZO~!5XLG0L@H;W;>4B5mdkN60vt$_}qxbfC@zV3jc1*il zw)UnA>`Ubq-{Rdh29mAt6;=Gewc0a<3+Vur` zD?nYk$GcdJ4ldRKz2>NRI`s3iJqF@6s6c3e^T;KS$7`X`3t}Mh<#*d7?GY+ki2BGC5c5v8+)_poG#xvy@gqqDr%oRm!qFi6@*d~KfJ zDA8g0W&%WC?TX(2tz;Yin$dmX@yyZlaP?OI+a#}r6q4;5Vl8-p$|JcW-}|^n_Wunw z%1h{1v@Pki{^3~<=N<3M23~*M!wRp@=aMDPLiw7Wt=47ptB?~z!-;1XqoxdvZ6A*( z-MWz>N)_tzLRO1H7Ic%(K9t6C(u%b>_eUC=ajgm=#i&Qhjag5RZlC+C+mIk(tQ$gPs{?cjwkCpWr|jjR|;%aQewKdD=))&dZw-Dv~?%z zW*uopM4;KMiKXR3S_VdMjrqLw`j~Z|rf7sBOHEZa2l~8~493G|D@! z)spu6y~(Ig1e>A-t3-{P2YNhKJ*wDiZO2(Rm0$G`^BV1{>#8XMn_9`C?~759{5ieL zsc42=_Bu71j)Irkuoa!<@3+fsK$AF@ox_V6DLivp z6+ummiK$m8VALq{jS(oAHOTH`bDF5o&n|s}jG}I}5GOfSVav~3V>3Im7)z17kBvh z_zXJi>#^zg}p118lOjKa`A9AK)n7`)|xP zWH|OuHNUtqJSFe?Vo6gfgn{TPpNI06FBvZ(Tx&zo(~h+ ze)e+9i2w1ItJHRkYR{jXB4ffqsjGXd<~wt38yV@W9$ClHj=}KV*d{mXUqf76ANaL* zr&n|8ndhr$UYqXoH^M1!qbpWLK!=RB(|Q3E`N@kWX%~a8;1>9!AoUHJXX!7*C))tD z&hJ6HM4J=n5$LJby@(Xt1q)xZxLaSViV%&{;C(W4{hb*32Z5!8q5S+C)Y(qfFON#C+A_WCgpZyOxl>A6oF@;v)_w z@GUzRgFrzrDwr@guD3sT^ij#q{)Ws9`dG$_5&S>@g`r4;!2kWE@|gcbl0WnwwyDN# zR{?T6M{)3zx_DAO>PRUexM1{sWJV#!QonO0%m&r}r$FdPRV2>$o9em`zx-3A69L>r z%&bE$I_9|>P(z2Uh?2@|$*Pae2)h;XX;rD=(-;^fIHWiJU5{K8ORH-uo{mxYK;dk! zN)}9qv`tPbxPSA{KBB=0hBHffURO2Z+_x)X2xLuROigwfRDRr?FdYLkvtqairxF9K z>kxd&M+vQDkcOa0za&Nh7So{BfF#nxcY=Mtb*61dgrK26sVdv%b@FZ=yc{Y#Sg}{c zt(2BXpq|&u@kUb)GuCA=L%Z?cf;kXOE&D~Jqoy#YfDaRN8GEjS`j@J{sWri((gw_; zqoB$6g?GBXm2$rziejE4zqMcTp1)Ki8G&WSp> za-~1u)g@8~5TLeI0sA%YEwx&#`>#N_8Z`mzKE(AAztnHHpig-#Zi&nlY$sb@J;Ep1J$f}63@>Zh#`#??gUhYn`Hw*%5>woH?d?@(=%2gL zPMeFa61YwkTd9?JEPUtC>ErG{&Av+hf}?d_M-+JA|jJhpaufb zundx^5*|D~_c{1hrTe>Gnqu%NF&p5 zg9v%~UTV;p0-g|W-)ROB^SdV1xP0hbKl$qdf%uOLMB&2g9YhWMZ>i@W)L#aRoaMQo zYf#<1+8I(8SE7jz?4CgqG+8H;ymjp|3uwv`m_r&yhpSJY^N22geS0vl6*|mxFwzMX z+zw^W4-hFYT}s;z6R&^jLC2tNlQY`jKv7xnjJ*nq?FwhNiBfaWH`zrzphIhJ$0E?M zF{p-dxVt_LT}PF_V&>1)hMbs7&xjNrXa^v2&|L>HSYm=IVWPw}_Hs;Uum=#zKrm-@ z2y!VuhwePxqzG@QoPNN#J*385ekz;90D>|rhwtGpY-d)b@>zYoT%-oQuSg*n6)=|+ z(F|~EtlNh(N^za!jWox2e#2XTQI6h2&=CfCZ55d_kTJnsn&dzOag-TTi(U+!s)W$m zh8rjPP_UJ=e3J1&u3lF-(4jjV8jRQyk2GM;CGzLO`;L(qOBk-5(IEH!liFaZ#TgMZ zq{0JJ($bzWKlRlUn*vJ{T3t;gWF|bN2NL7yG&(o>k|>~jj45(Sipq|yY^;ZR$zT!J zw$MEUBYrgZQlSFLO^GpkWP#^1$!)MWyBMRR+TbV$x1aGZEw*luLxD)aDb#BLn>f+Q zVtWe|Dvid^P1Alb(kg&v1H)P$HYs-*kv#KVvJPR zgdz-m1c#4hPFnw|lACK82-EbXy(#j0u$4j%v#d-Af~T$S8%t&_Iq5L;QCuXjS?52L zEI|=n*jS`0R%W>u33!L4rxbeCCs&s3Wm@mXk!Qp4i~<*Q)>)=E#N9L{-)rllwl;bu zd3;3cDPKVi-BE&LQ83|^TFfbg0VlCwTywuxi<>SFVH4J3owq@BXGxuloDrFFunj#U z3Y`P`euH1PvpMyxu~Kf=7(Q=Rh#c~Upk~$|gw!j*qfbDM_SMTy`E_GMYy<_OkEiHk zDRDf7J{U6gg?rzwb-P2&-{??5n-j@RbR%b51cpwv1f0vfEdDYoML5>ruzW1f zv;eZ1)UVv!$yOYo(O;Xer`;=s1b{YOM33e)O|<4uTPigH4ewF3@u-OXyBJt)R1ah z9hsr5hIoHAr`nJDq^!?8U)@Y6W_#HW%PYH){)3R6y2$VfI?;Gn!tdQQ0egy1;Z-{) zVLC23{hPI)F?3fq;zu094VdpBe@pCXZW8|DH;D-D-8?^cT*>9`LZ73N=}u1VMYdPW zJEfUR0W%a;ZXcNdr}gLDN3tJ}1Dd+TcXHf9fyC0d_$*ogdsh|J4`CxlFo&mg+O0T%n0<@cKS~fLNYFCb@b?7zPnDs6gglmsnbvP{jW& z>oG$klo?b+_6SM#Z{Wx5E+Xca%o%p8?k;1O`omXO zWS?)syVPHF-?)p|eQ;;FDF^cu5-2nx`$?n-!`nJ*g6^0();{?(7<$O8!H^NfXE-aq zNaX7bhe8#{k4`0W7C;wtYs;!*8Zs*Cpg$=-YFD;F!udKN?#N+9BCN0}@{oB-90|RK<@* zI7y$buW$XU*w@4O`8WlEdm&E3Y5Dh`oEqX)zXS9rd11wuw)3?k-K3KCF}KHBT)lJO9c1X_(yT90_gz6|dq^Tu81tiyZ4TSC8$IjC zv&-Q$g}55BI1zQ4<`w5f)FvGEP5LeB9#6v}m)xGTW?Io~JD(IiG8#Wr>@GS)jyS)M z4Z6{i&vd}4E-x@Yes=)JupdzZ*Py4vVv-=5eWDQPDf!taxpI8A9_w#)VjNC*DNrvg zDyx9(%3NnpusJKZ9OGTm2UZXa1@kS)@C>J#YN1pLy+kgSx^J4?C{sSQ{;hPJEhlBE z%XE%Km|6;p528KAH`d=A&-YYIX0fV(e+@K%FXZz*6H3<~3*L?#gMaU2RV6Z)o|M*O z6NE_1F6ZY68PZ$0+_P1o0=lSK%%TA-M^)w0{dM*|0| z&}+Wp&ei3PJjx0ylTi$Op96XjTUdD364_wTaT88`>8a=xbo~2ublo3>ybHGVXU(W} z@eQN}rgv8`_V^*6Z(f`1QhAS18yvHnty%t8wJ)iI4~TJZT_;9J&UUkso1EYM?N}vwft+_Nc=y`M8s7mcP9`dK z)^bJQyZDL7sB%AsB}KvC=fO;E1Df@7w757Ol`x8MD%)bJzsGIZ5a zeL$3^FR|9L*}j;<`J09FvUzzsPK{av_q3U|B<+~DtwOeOIU3sdTRvImog7NRmG^p4H{oK(mpLpi= z#^NPr)CI8=jVc^*_qgmfBy(1Nl9uqK#6`^Yl_njiY@vDMi2y(NcSwd&)fj=Kwr6_n zo)I6_;OODEL%LkphWa){d4SL{h@rVkWyl&ksdJ$j*(p-?@KZK2)@Gk^McA)#ioDQH zSnm4kWzH85w&34!+t2GT-VaA#C>8J3VU`=#B9Z zo**J2KYfbw6ay6%5#X z6YD20_+_=+lT!rDLkerVPUR*A!5*pI42dN27NKFEHqQPk{Z9~{N+ANVnqtilVvN{DTK|>lQ8&^j?9q!|tUeVvgk)9O=!!A$ z;F4mhu2tg8bGg?T(eqBE+7MpWnztfD!p^rn(l2&}E`Iv@qIq6F=zSd`9@K4N1&!S}JoI!3tKs+$9fSLSP^VNU0; z)9zGb9gO|5Apc8gPS%iQb9`kk><5i2522NJ9Hm*YnlgHgQte)rc#F?7z&%S5KNg&k zBUwN!pb;`d%pfH{t2^q>6F?(*yyALJzenYBjZBn>iQNd=6PdL=zc=^ zYLR1Dxk2$t*^%aa>Fo=%X^ zS8(PB{gYQs`*FLaXe%V(w*^VjR78BwFfzK)mM;6cT(--Wgbq9QV~gWz z9UtLfO{ssWK#9~KT)rqi0`A>%ChI{W!_a^DxjxaHvnf0Pam9~n6hZqc|MYZ-g|cvn ztg)JYqEMnHp65$)*^+J|P{;4Xv(YYb6V`ORJRW+??yTlm$c`J2Js~)Hz_yPC(Cehp z51?ElE88`pJC9yb0gK|#6VMboiWw0F<)`|`$h@xpe9U&KP;)S)9^h4TC6jA{_A*Re zB_Z9*LrO!uu$nh4&%&I@@>9b)3F!igrm__C_=T7pgALH=c#7N2RXr4Lu|#zHUa=r~ z!l+mqv%=xQP8ktuy6HtOMDZd-ETiv2MKD3vxXP;boW{tqy7?>x0d@bkW`7W%ywMq5 z!ORu5_xar)9fS#2$fw%bNFlA}neq4o3qS+sBaO`%M)H4t*G5Fy&-_V4IHNC|mA$M} zHr5q1Vvw!njlcBjYvfGAZ2Z9dW#23vD|>qa&loGmSq)E|)+#-cUUgYci&+VG7jv~^ z$MM566t9IOiHTe>U`s4l1>bS}Jn%-a{%zH<87~8u7oG~gOrt~;oe7D^GS-PnTtNf9 zI2cqM&n`<&=;2Ahp?3W_TA>=$^NT(5^63-#_yDPwB6hv((h4WSn#{M)(xiTBJb!UU zx?Hxyz!^Ag#ImD`mm!*%rqPX&q3rh7SZa|eELBAM9RgGuh>5>K9WYK3;==^AVm+2= z9gVJ4>&AYAH!JH$ER%Aj7J6^C_8FbJSZs&Z+@G0pj0dY*kx;U;Q7@rg;m&I)@mo8B(*t8|>7v@-EwwL9n-rI5lFCivpBF;qxH%3<6T zY52IvBwI$X!jK|!r^l&uTkeNG?=wMAZ|_fk$8LPY!rg7Ny_MSuZ3H=!UQ`%2NH?HV zRN;{X2L3>7Z)$AC3XvF(hxhYXr{JR0w=gU!X_5((`LDCXd*}EXR^)wqF{?0dd0ufj zZ$N_D3*wqM>aUztP9K1WcP_7&pU5d@lB#& zZ-rBgFf+6UipeIRHQp#@>=2V6hIssNO6iAjX6mxdXLZLYB}YYhAX4aUVpvYwl@PWM ztv9{)#LI+i6o1m(@abh>0(`->9nowFS#2FxJ5{6o98JrUOLPcObaIGyNE1KSoxcu7 zD-)+HgiZ{L1iA;hqr8}=iri-`IbNIQFLj!Qe5_g7+o{l`5FmyvJHCCfm$I~*ML98KDdjbFbmEbx8nV{S%4Yf5GV`ln? zqdE3f@r?Ppn%9<#mu*i(ubWq*zX)I1xsZg+wVs}X)xOz%-8S@ zuzXqaVji87P=jTY6_12239sg(V4UJ zv^_?gQ3cd^6*zJVt6O(sC&BCzeG!(E0`llMfA8MSO|AL)7F6KVp0}ZeYApdplz+Qa zaBs=dzo-DJPQl2KO_69ncVJcUbv!SNJ!SFBwIZ_v|8YfiQ{*ny2V?a_w+z^A!L8JW zUz|&MhNEkrKCnHLOolGi@TnR~=f@z4`Va-TM3!W&|8Hk0uuN2g=_M5y>Cul#ROHEf|$SY83a&@@M?D{7aOV$4ohD!c=Jj; zfNQD`@oX@;bL2SKB;_gC3h{=kb*+-B=kYp(^YF1ro0<48wK>wn6yOlN%hPJWLYIK?^`SEo+?yziK(aOg0%IUr@0u zlGnpkQ*gLC5hW%rWX+i~Yo!nDHr<07hjO$_kOdY|pM@u*B|8%L+e=g@u zCHcyw;YyEz{R0fXKBt@SX(+A@jC?;94xw4$don|HpK}_sPy7!@>-sCYRT6h#V}0{c zQby}XRe&vN>kpnkddO{FiYzzPVR@1tLG&_lp;P~NdA}%e$^1VYI*>OD({ju@>BV>L zA{M?!p)+WNcsQw^FnB^2rC$={V5TV&_#OZ2{|*)TN}*aw@2}aN#kb`F3kZn=NNS$0 zMrXahQ#(dY_|36f0K%!HnP{j$=;$9#q@TXZjbA^G`p%bM_rl9p$UI+nuDClUUfJ>H zO`1MH@Ay(f`Osy%!EdxOPI_w&hnVQi5U_gizOj^;OvE#0Wbg%5RI@-vl?YE$?03)C32SmM?XWOqE8jD;9K2@X#7U86_tjay8{@U}Q6&HS z=PQq-&LYVlcfXSVoqdr1jnK)T=j-_Td#WAkpTUvZK}pPFQJ)2)dC=U&PW1E&MEE56 z)$9Ijb$4b@cCq;HYh%u_Bhbsl6J8FShFnB*`pVE~uvN0Ym1f^GhE0Bz1!O*WSDlog zwpjzWuvg17c>sQe`-I~)6X=5B$AVdjhq_G*Uxk~Od_fsurjGvqJrBfdRZD1%{SIXb zbzGk*%3zvDM}JdMjJ5)iNHnHaX54{ z{rwNZm&OGc)zVN?IaBKgQ7mC4$iH3jp0{Dpu>$9lun@~e(gjrg-S2%TV~kUzq!thV zw|jD4XrEbCs6yrm58XZF?AeFOZfJpKU9d=FwWCZ)IeOS-Z_fDj0a5@-g`ACbZK^cs(zQ)yVN#XMNgxg zZZ~H8ZYKNY36DUr>QF^s6sDqQ!5oz~Rc(tT14^vEB;r3g3N5zYoZA1Mrehik%oHp< zIS%`-=DuWBra~Pw(@3k~cc*y2ljY+wea{hxnn_i_mbGrh7TNSi$K*JeSoP0TUZWot z(T@&R=REUu z%Bpwk{TL;I7Ek+{P?df_r30T$_eaZE-Om2=gL7%psk-rkezvkwDR?3M#UDJspPol*N zqT|an9?UF&dqQO|O?pWIlc6G`(o)KV ze2=GoQw2C(^)G}MZ1f{k;e1TZ#uv)Y2AbB(grDfg_Zi_ybR}_-8lJA-tl+|GeG?MYD7y0F5OS#6WgoyE z?-Iwkz82j+Ok||iB5SaXBN3bUTff_>lKsn~|ArmXCCc;R+0OwgX% z(*5AH>;9$~H1eQk-;brcV?~+hr{cNadb&#cZp=%wzj;iGE zjN#fZdgn0?pa)~ATNc#pWAtLSkj}bWVR!d7g)3dKvt{%oMZ?pK$+A8^w78ZfB~;;; zX@_8LnH9a=BUjm2^jWa=5(~H~qaEQY9D4kqInN-qrORkB2NkFPy%p$}HO{ zlBpz0lsTw)GLE)M3K?RI6zS^gi7bvP)8Qsbar*kf8CT z23GTTfI+P0TE)+__v$+57WyS1cf;f#PRAo#uz*;XN9&iQWL|Amj`f~lZQUi|#(h$=P z?pUT3yfTT@(;`FmWJ$B*TJ;{k6F3wJt2FBuLCXvy@T$ky5^*B;2`g$$o6Bq5O~F>< zt-XB&DLSvK8>wI)aaGky!z0QlNl^RH!-7RC4HNV+t}VXjNQ3_>u%;dmALzcVKg^90 z5kIVBds?>TaGU@8bLS%(x?8q<@rdXD3t&5Gr_cQ;fM05|Uigu*CU5i#*WhU++=93a z%^enwgIZ!c#9@E-wft{z6Uuf_7+AmOx@I0QR5m)4lq+MNgompLU+s_6_F&^h`An_! zUM}jicDXecn(|``{IB}>4kC6HO6dD-wcG*emvaxSbcW}3XlFZyl9HlF(XU?q-h5(+ z5e)j}+Cf*!jREqnNf+Q+O|2C$ejlFKFB1>{)Q&=Wr4|9&f{b(9rT#-vtpS%w*4MN% zC}leeGKGD|hNi>d?9Uris#ct%1aGglOCk$`brQIKU#a!%MSmsVg|n@Up7EFZ4S(3l zt>eeG%CAFKD+=+%d#9f3N|!4Q<{67nxf%Jtw=gK}v5&X97#6b&RTK^=M)qP9FmI+x z>1H~_%EEb2k~VSn6dJ*pa$Q!~yjG&IBZ;Cn2NICzRA{aiF<36PZ z_SLJgg_{)FYO8)QoahkJ->Sc?-V`nobqdcq?@-CDz8f*#zYhFCRw@7COqBcs%&cSM z9}e-9?Db1wY=3yQPL*bwg8BA+$gGha%4nbBd2Zumx_FAH-J$LAz$m@pg7KAgh2HCI z=mWW?Me^r_+H#_7KM)c3r_RYi*O8V zl>O347iW$M(eaQR(a<1gxRyVN%oy+>kt{pB>TXm`^H^wXzBqQ|vun9sAA8R}(TS+W zKJZUBck25Qro!>WAsOFQl1aZHq0zzIS!5jScD2m*bq~LUU=hhGy$;OBq-_!wU7@F* zggV%dlr6ZpkEsMd8~R+PWR#C>99lgkLG>ZAW}r106n@w_*~T*L`T=B$it7x zIu3e&Svc?6=`8%{m8Qi{&|@glcAi0H-f`;aR7AlxJTs5yx*vy>Z;wyzTi&3ARx@uU zJV^w*=G=+H%R&8_WTxQPmG_m1esie|KHBMylO{A`d!F_L;05lzvrSI zZ5`IcHPjUJOZuQQYZc2smaaU02_;CONudE^-BmEDY!UxQCKsA6npE}g>rnKIuMX~I z@z}}(=ezpxl8vN%>XaC|`O@l(g=*yRlgb}CTF=v7)z!y&7Q`o)r(4+ zL99i{XT&ZoSOQWJ?d>T`@Zi^ot6iJ)!e1sd@v#}YLrF{s*0fAWG`7#3JkvP|zHwjb zsek$lyf~WR8}FBmz$2%IDA-WBTlk1;WcU-$JOgV0f=~d!QOuYgMnvS}{|Jjf!4ToX z6}m2iT^H^}K#Pt4TMbUlcUi0b!mMgC1?YDLka_X?5;yf6AcJ zeY_|zExICnqxA5X1tamOXn3H0S;YR26u~H@EeTQ;bpR&L`-5U!Rn?zCpz{Tb%&(>+ zc87h}E}SC|a{H8}O1FG~Qp$rJwT5aBto4;NG8#3x085&?d5Eas;J48CQ0sZR)UJ6K zxz{|;fd66a(^LPX*Zhb)iQY z-L#wkej8*`O_haRA{%TaAjThxA=q^+$Pwsk*Pd6LCMM~k%hQs+JfbM^2((OMxX-Ei zD};eI)A>H!9HVUw-C6&KlWXS3`|BT$S6zy8kUFe6JUTn@`x(ob?Sx%XpNHN>YrDLoDJxo zw;5?x`Y6JDS%RM5qzFt z>!jrZ@KrJ87HuZZ`{eg13;a<^K`Qq{I@GWsVWO_dH%42luVJfLfxb}oa)$br_g6L3 zQl6WE-1)($-k3v%<1J}($5U}}Zb<0SzLg1B(kWcm(yQ;+d3Cy#3=@55M}hS+0X-_^ zTxlQe1|KJCwIFtMgal9NLIMpwk=H-x@Qi1jol4@Y&9RNB#wn~IRdn4d&E2JxO!6<} zlnDZl!^l@h89b!nz%iX56PBVrK1=8TPP+);kuFq|c=JbbNIxVXPm@w%Q2g3h%2V;H z8Zyz2wK(p}Te~{1a6WcT2~X(7c9-k!$_NQ?C=ytmm9$MEmhiipV;qsV#8P<3jA;Da zk^nCyEyC?jHXc3+&(w93Wz$q-e0h;H6vWE)@!c&;{kOh`Mj|@HK3;WOwX<-30+v3vFp&E>1ABQ4&q|`fVkq-wU!n%Y1I6G7i}Y1!;=ITPjTRr^fpu8AmKmonKLj4a`Mfi^$e z&&YJ}1N}#V9!iuiB;rYtx#YI@C%>6uDL;C!;}C8AcnW97N5Eg`>{)VW3Z0n$xI{tdEp}57(g}_1|!C2sdi`; zc(W+-cDjgFavWY2y~B&bqF})hxu{z=-A#aH<2_R7%BD>Ihhy-xj4t|G@*~Bh9d$1G z1v3#O|1IRDZ>R>#M&)fDc(~8F{a~>9ak9PbfR^d)s?ct@#Hj6fWCg&lU9W06(a5&L z&@`4x;{iRnKVR4&kHbKWI{G($)E&x-&_rKEaAZS*C!+WiBv$P8&a781V5-!ObP(^f zg!gwy@y~LEe-X29aqQVm?V>%W9#uD|m}@MfVWk8zrYx@G;m8vG@(gbtO6MK!7C(t< zdx1aRW6Q0$7{9HFD7&aGYOgNF2PiL|{)PZ&%Kgv>rWDNAnhdlSNT}_nN~_7TLF%M6 zwJ#rs5*ad<*h*MfYQj(ZAGDvhxsg{lDwUQ}9z?ekR zHLCuuLUt>)WLW-#_SP1Ilt7aSlPJst0yqmyBYCW6hqG2mz>*7CC3rAxd)4Da z`6d?~fEUWF?Y~>Zs=MZ!e_h)7#)4=MH^{DiSoAHKTf{fISzgb+_4pH+0-jvpZW(G0 z?+|w#mz|XW)|GG+@s}zz6S7^7Q$^^!ECQzxDPt#_)E_FvRlc||0YVUp)!`LEj?Q$c zf)=Q=RUjJIP7FB63=8>Q#1c$i_{%~tZLpmWD2~hk*1Y2cuelsWPx=RQIGPPF7aJ?zXi`e*px+zvT6q#42KmcG+?Jj=0J3NdI zhuOqXbwU4p&hi_JhXD%=KuQMuk7mVQ!9ZT-b5e06sq~9Gs#e8@#S1M;5xNz(3|Dc- z{wc@%;Qii)Ah1yV?`)e~oJtB4=cP46ffUIAavrZw?UTNK#`-xU<$siH@Ji|8y}O(m z`TQwGAa9|Bcudt`v2J2*Ajy1q!8;G!O3zls??8OqATmGhC>E}a$LVvpEEU<#n}_>z zU5I7rpPgh=d2Z0roWqTg!Rv(h_O*nG0`?aK$%xt#I=#Hn`dr-bsk8klq3!(q@)S%k}xm* z`55YBR_Xaiu{IyclgSFMJ{qS?M3!DuSb>_t4nyC>`P3$H3mKLcKj)4_DTYJ3GppX~ zb3`Q`b??~NW}3i-K4Rf0JekkL1h@OcKI_o^!U=HWU5MBN8suT<9o@o<{b`yQdx14x zzBg7XL98rqcz;*W+ACJY0UdNV*kWC}j}RaI-vlwVEwM1hP)!Psw3Ry?TxPHKKQGA} zd7O_kpFsBv%T(ay^>CbWxPn++K=P9#Np2x04vO2Sy;2`NGyWe+Ut(Co&GNF!T3=PK zo&0;4Qz8)yncVsaFk{Gs2U8!HWH!I;GL5qC^~-T*#WVi=*gRabNvdKWpN67C&8pgH zt7@9+M<MH#%$#{GG!Ev*esFTsy zUMEOEcWJ(o*4}Xbt^N^YRev}*Xv){T)tNof=vTW6e1JAoaECU^D}jDO@Mphh86?3p zuQZeyen8SLH6T{1;8a+~c<>N1|JqTy!F6lB?9S|JV2J(nJobJ(&ee*&NlIuCQ1pZp z`F0T^z_dH8uL_dMWr=;%zK*LCOZiw0qnhhas7R}$0I1dtvUYiQE&ofQR2cY7p8Kg; zEjgY@8?jSTH5Z zW~P8+uH7X!C_J_{J&-wih%Y0ZFu2iK#K#RP*F!43)G?M)H2c2c;V>sFv~eXb1l!9t zSg?(~J>p3kOmS-N+f@$@uc~4uC^z&`JN46}NVt8JnHH7jIemJWa^akfeWTfcj?ZsR z_DnZekX^Z&mh@{e7e`iduFuyecb5s7T7v?_ws77Z#CjA<=58Wq;X#=`G09_MV`pFt zk7BTsVA@Hzr5hoyG{%r&fY=e|<6FDfYN@{g!ig2lnW&rEbv`y0g-wyciWqSvUXXvC~TPc+xX`Rz$~q(YP? z#GG%aP}4o|+229^iD-r-#Va)k%VYzwWr&_gg{o88YKZkHX9q|wO7wYsJPqoVc<*WM zHub(>-u#PFkTXQ-Da-NS{M!~k**Lz7EX~I3e>khE7OG+;)10EU*z`@p)5Yh+h5*#U zS}38d9zI}YrZIo=qj|{5-+5jsVq`S8=pvAjh|aO05!DH=UX8})8_F8};z4d#-L3;n zewkA9c)L5H!%grgr-jjU@-R4w+)ke6pww9-0$~xFDFkq&1B&^6}djr5mjFkehzl>mD0v82TLNX`OlMGTqndUuXGisG&@;3*7hP-V@$_#9&t5 z2QDDW1p*yz$#l7zdpRoU%h#OG4KIgv(<$aM*cpiqi4)*9MI0Xwg_G;@o1G&{zp4H+Kf>Bu=f4 zRP?XTjkN#aq;`7-ATm*rU^7*eFc31)H>In?Q?t`849x8otUPo+q~wtR<@l*qH<8ST z(%#hWovAM4@ye4z%9F*ljaxK93semOEp!8NY`_2s<>MC#vB53yH;&MQibi{q$5krx zIAl2WiTzv7aay|TxT<$%!z%cL1gB7!9YHMxN_sWSWdfp{YVPNZ9tOD7;Srtnd1=9A z+D&^Mx6>B_oieQqBHYEM6TzzyIu1Q8I&`ZQa_h9QST4PDl$3hed5SiCYqF<-PQe`v z#e6|{%|^n$u?Vg`N|LTXPqcDn=DTcu-dU(G<}}VD)gs_!e6g`z%uW8XO;LKyQ+wC- zbGlNH#F_6D5qmeM5loPKd?Gj6Q0hgk4tY^i!6W~2x5H4&?MF#GJjt{O(oaVRA+M}* zt#anmFqsO!>b*ZW1R--B=@jKlx#>u>i1Y0#hYLUGBpb@YpE&SPDGIZRqYjyt@w^ps;yeua)GDaY z7%EOnIOQ!)K7Qm_8=->4>>YXElypj<{=genoAlncC{`cq5QF@Gfso!Vkx!-Q4(XUf z@XEOY$iK#R$rIr@&u^1pszOfAnlM&IW^G>#8H7&ISJn=kqdWt|fXB=MhDFM+sbl$c zV%#6l+^DoDW>&qiTDI$~VteuA`?6lIiX~xPULVI_Ht673`PDLw#}5>B$-u2Lv&XFN zQlN4e%K%Im8L}k1HNS9C{&p-P%~x%RoUM_r=&Fbf;Vxuy zA-tO)x%XD&I*%AF+Ya6+A`EkI&JF9JTN_*i!>9?$98J0-C zvCwx@g0~oJIolG%KRQl5YBipEdGTR(oBbIgERVE&e&G-{uzZb|!)G2amSy#jsS z&-;qwAC7o95tDiklgE+l*sU96M7YBsY1lamqIDI1S3sJS@vce89mtVx(xs^T)zjFv z7tB``(bl2CeqP;epT)vAqp3Yt3a--4d?+8w>qR{Om@Ejd@|K7Wrr2!wqjSpmX0BWC z&diXsWj)}_J5Kqv%ohY?F$w|`c`vWzNkC%PHb^X!NlW$#Aeu&BTfEuRgz>I%1t3>~RA zM|7-Y{0F|xh?o807;R}1TkFjCJEfP?L`^qU!1s#u$4up}?d%-Nwvh|4au<%1tKOyF z3O@C3im@5HM0T016^o*B(R{|~yvG>M_pEx^Eke}x$06CY9c(WR`oqad?w&vpNEq^$ zK+F6HV2#z4&qH%~P@4j*R&=CtW>Uh#sa%-eon$a(Tu7>2$*%lLo0Ioh9lhcZ^} ztp}l0i%X);kZj)_WUx}J4iSQ8@B&=O`MSW@(<^#Zzz>zsRea66icw3a2iD` z9#Xx7tHS@|*xwd*B2p;Bdz+7d&yMFn7iLn||FVKr^<_Tj1j+79E7vEr+b^F+31+Ro zE4jBIaoKp3RfIR@QbPP!GPtno`TDBkZ^Zky&WjU-MZyzKt`(5 zY(tD<)5X*9S%x%ysEv1D-&n;AWr;NjgC2t-f$r`vYDlV+pC$(bI2qMmBf0;rXqZrKhu7mIZ31d+i0p#VsNaW)u>d%e#xV254s#| zMlQ=$iT@cC1Iqs}BX}||F8s&%&ex+&!)RW=Y*lGfgZ;HL= z6(g1b%2AHSCY#D&LRK*8Rq)FX_8~Eh8Z=!BEK7nGF%H_wzcCcn27sOxi(tnY9oe0k znXn@Aru!rCMYD6rf}4c%wd_t>dkXwj-IE3 z`Y)lLp|)IPml0{yKlr?P2Oov)o~&0*tRZ;jo`0a6!i-$77?*pklh-d)r{X_z3+oe@ zYOP_3XcxUZNv8be5F9*h;ym&E%ZI@lvI-=Qj;?ogy$v_z8MmWPuGiH z5#yLX5Nt(!UCXHDXW1}=Q^j6>ng*fNuJv`jR6cOk7F-apYW^&T;Ep*p{d}y10xdMU z(S5p!U|t6?&0w|E9^%|@iYp9~osEvXr?M21)e1cftqaTj(~^bEU(=MYQWockAet!? zF#-DwTQ zjg;0_FcRK*l>EkIxrPU+TJ$rBnm`K|uWiYUErUfjRpHr;MiTjLaI=z#p*fU1+s!MxODF@J4CeOv~m|kZaa}8g8&X6^d{j$Iaxa zG!%C50u>)NACvmE`zkY@!jX&Foyvz+(LzW0QByFN9DmzGr(mYniO}MOwz?AvpRU@F!*>{3+kog z&{di@5L{@rOiQb{hBY9(KIKb>UFz>mNg##8CvAOIayubVMVGduXAFelbp1_ur%xy( zID7(WOaTl|>4^V zK_Ni8iC$GzF>bMj?o#V%`I~(in$u9O*G!`6a4d-kTqtIzB;J=Ln78cZeC4B~adywa zns?jeg=Bn@;6bN@=!jGuJ94N*45=-B2{AIr5x~a|(8#@y$8d7*-KQjSlHlNeLvj~$ z(=i8fbz#tBl8G*oS{SjK0brpmxB9hvId9AES3H9~RB{v-RxgO7zp zhEC?5nz@nAH@mCk7+hRi0Ip)_nU0pJ~ZI!2;9_Q>>7>t)&+2ceJK1Pim%lxp3v6MWtz zE3zH52yC^4YTlqwgZCj`K44)4nYxv2bFfP6QB>5QEW~Lm{=glFWd7w!(isGPy)u$mw>dVknOxq&Uz@+x2RYDO0KopOs{R0W!G?GWqkPAbVsf zZ^=HH5kq&Zg=Zxq^pCK;}>G{{b@WMaE)N1x)NHdc<4A=lb&bOF!q} zWn@m8`^pE~kt@{VZDn3e=^AL0L>0}jf`%aMFvj@nI4(P_I>wdkG)7vAH&AqDo)G#I zTM(V@m~++Bq{U$p!5w|ScU%Zku7Z{nmRv)*AgbIajTFbIR&i}w&O+j#|B-Yad; z#Q!Vtcd%3;SGu8lu2tdYqJAsyOO3Vij562XZDs}v1|7<+_AC7%wHH0mB{A`0v2e(Z z(q=fV?iT1&NsX(J!(xmkO2j)Y1@AS?Ui&LA{`x~0=57Pd-^QPS^waoY2gZ_K1zNd) z^gb81yI_PvZ7)uDQYvZ*$0AY$r%bwm1>Jak&~c29dVhW~l}6dPKvwQvv%)H{&GQHt zR2WdP%?dzB2DF^(=_shmqhEZ|?IGbCqbRQt&U{aK=YAh5(oMYvd=v}_=4KOM&*ZGn zvWSWpBmUOcr)Q~~ji4E~E6#9^eygQoIj%%SFcv9f_r30Aae0ZKx{Q+*#pJW~@tN5N zQF-$WDQ_ayVvCq5lfuI-+|M(y3P}()v&n~#$OdAAyG=mrv?NxD2zXSeURtYiM08GqfzFW#YfWly3prIBV*5vsWFe5=@z2MqDj^}p`)h#XuKUh@*F*+P zLk8#+5U8_>1D;NMk5sPXS@@A;O7T(c+bYTUXr;!e`iwd3lr7hiB;>WC(<{2d6B$3T zQepraua?j~&<&IsL<3(AKilOabBZ(@1~>bFp}xB8=&X#E+>;~Wh(>{}0@^ags)7mk zk8!%}AxzY2<%Bq%T?Uu7^>sa+Bt<%y>yZ@7b-q!iu~NTnVHdh5bY;Rn4qYcARyOAz<;(&{B#4JYgIF8>1m8VH;~ zvG(QqPXh7wlgK*d5nr$KdaO(02)M9Qgo}+t6riKx(jP9F;fk@fy$=vtZu4VKXap_@ zM}k9)8R3Xb@FMMWgByV>d!|Xz1_#2k98Xy(zk275HU3+Q_$(muOobP^$xv#-!uK&; zKze(QNDYKN9p*~#V^o5@(A4tnnxryjYFuZ5f-Bme7UX|4o`-~wOKQd;<+RVjt+a

;UxH+#qdu%RfH#cpDW^(YITd-&>~&b7|A%nMQZtNQeU09QZSlhI(C%uCDu?e zFCUqDRhPyJ>V_YKyq+G`%ur!wV)Y56W0EEC(Ww|o4W6*fYB6PoKOh$F0?Bl6o7f6O zetOOl658IIZ}}6J>G%S9w^se(lp5x!ZdQke9ld#P@MN1uAL_NR=T=D4- zsPxWdppem#Sa%U*Nbn9soRNx5P01or&fA-QXr(acCl+>pjBDz`O=ZAw>QBKPqxK4Y zz)o89YF+o7tmAGN)%a3`Qd zRgoWpfBLB$+)Hqxcm+01z(FhGddw+1%fTx*b8p?%8&|Fs%WXC6ALM2XiJ{)>Ouo&6 zv(kPpi;>4F&?T9i`m?+gY=SJ)bxkJkFv@8Y^u?DWKCEAA2K(^xu`!@IFJ&1C+X?u% zaURJ{;?DB@@e^s9uxD8AUyDeN3Tl^~0;o-GidaGS=V)+>%B>`38;sR0 z+*@(9`b=doc6DzLgcZH3I9+_920JW%^bQ{I&%No7m-6 z;EzJ|qMm7?l*BCtiImQ9uI8E`5}Jd)d6L-tfmWyr)N7J#HL9BZ2kFa+c(v--@ooK} ze>grd#$QB^L%8NpTPE@b8erIVC{`h3EqOZ|R%mRpwX-W3CDBV`?i~mF#Ddn0OW#v=_3R6uM=0K#YDh)l_n$X}^w!PlYCu8($Zf40XF1_}>FK<~% z#4j`{kHi;tv>Ez)3!QGpN)`!Eeq2L=(9^7ET8@5Xtawd5azSmXB?Vj+Ucc-wbgurq zlznz)hD7cXr{E3)9ttE1jK#<1i z!tE(4F>k@6kE$e5HJ;U$6RnepBJ8`i6iv&~6p;DY8>q)7VR=hObzdi&DGcCkrM^RZ#G|`kaFi&}9 zu~)V3Iio32KG;|}6R=Qs~2AZ53;rnbdFW_CqO8eLK(j=GoC)YB~6YJXFZN}}(tZbbf3 z65te9=<4f9eEiN%q-=d$c3dcn7d+Nx0Q$wkVac}rz}z$-E78(|vDbwFLVpgmZ+{vy z(K<#xD^T$Ivq)?(q;MQnXn$JnMIj7)DnYO$G>q|ZP%rSF{vB$~jeauGpx?RkT_!eX z{1N-Jv7W@6?OZ+Ms~B+c&P|ZBB=1_@gSxRI|zMU zUL@#fuuvt&;#2mUaw#9(r~!|kv?3|f68Bb)7rP=j}6 zf19!xEy?|Dsps+!$D1)oM;-I47zg{}%FQ)wZhr)<3BCgMxAGD=%gQ37t9sM(2GOyt zh!h4-osB^gX-l`~X}#O3WkBuT32J}dgbot~gBnKHCjh)>O;U>SHo$|zqF8Wa19cC$ zdC!8h_1Q(Ncb#SOBYr-Pj+l~Pf#}EuXnU5eh}YjFK%WFX+uSSs#oxJdef0@Tf*MUF z)DY^a@6IXq4gUPmFzG;;t)bvM`^^W1CNeAu3cabg$|cfx!T2h^!O}iF-EHQKJpeLC zE+Hx?9}S5rsQUeQg&GNOmpqjcK(XxY)ZH>A78PB%x#K)YyGtG)hb>r|WP6kW*Rbuy zOHCQJ=r@r!qCnMBxw%J0ZB3f5n?Yq_QWWl?j2*g1^WI+uiHUN}Us+nrCo>UHXHxkK zV`GE)<+)O&pBBV<#q0Rs;vc>o??m!7vaAL#-Ovs?l_99(uxZliTy1RBqDA3q#g#|* zJDyHfKFtOo?xVos^K?OYOGf?gZ7g}*Pa&#xI|2x!nFnw|B0dh~>`zGLdDJz)>^y;2 zrsRb!7B%|Gd>`3Ilk$aFi%cLnek#!q4Z$xa|9OY(zhl9b-`UMix430M{Llw!8*kf3BPA;ut(ZhWN~ zUw=XAy5}(j4u^|Y73np*Qo+VyUGl2clnqfR#}n0GK7tnTN^(bzOqO^%wMx7F1qTvP zZsS-wje$EkpMetOWPrh0OuI|zcSIjWqoMdA70d7Mor^w2UJ3OdD3SKIYTx>J;m=6g zj92>^2gro0%H&ayRE|vo53X}o{euM&mOz9LyYy%8uCkh}rj7vTvyai-xHyW_*wJrd zZ8h&B7Ig>k4ed-z(^tf!P;gCd%us%FdC=IpW`sE8v>!5`MWKjO27%hy#6$w8!WR_2 z^(`c5hK$&oI9EF7b~gtRAhvw-i%FBXE{(?mqet454@Ei#spK~OK z8=^jq9+JBRM!Cp1e%0q*2Z@L4CiHfvM5z|#d9XxB%uOcC#`2emQJ&9y)(@RV?>G}< z_104N=JT%pfJPNM7*a3~&tK^iHdDSuS67o`fNmYY)^ZXp(hnJ+izD)?4mUpW9`o!S z8pco~L~IzfEI$Z2WBVCd-xfr=V6UQZ=?4af=P3p;Bbvp-T}YNYx?KMbbM)OHD(HyY zbhqrs)jj5>(b?~*?W2bCfP=|y+yE#cLRTl5WG|U7q#-G-6!2lwoOT_Zv=iltnix2g7MlmvaWccm#x4)_;2kpf9U>aG^CsJbIX0p4S+do z8tdHZ>oF*KuhjnmA~EdP4MYFCIR_~LMRoIFTl<7?rG9eD27P|{gP6ET;YYr^7$)4x zFux>QOOTvma-g1a(Uw>dN*x!=M0naCO5DTmz#IM70(ej$cwcT||Ao{bAwThTnDX~L zl433#)qF}}9Oq!I_AcA2T1^k~6GFig${MOYu@{(S!Db^APb4U(p6PvAK%t&3J9#$c zdNfx8sy9-1gLS&T%(X@v;V5Cm?0l87Wsb~|+bP4^C$IaR(%+Rs%Wj{1$#?GGUmcKx#_0wEbPF*r-^N=1d7B&LHY-*FPL> zV|Lr1S*AP+vwm&@xcE!I9*qYjPv4L0_*2^1sg=caQyKBpEHD9EOUI9!>)Y};-}>V< ziSWr31__M4;grH7dEybR@gUR$<0RoVru#?Hno6i(-g27lkFo0GZXR$5B8!x3Ej`~O zA*MiL{44u;OgDp7o<~T6yd0G~J@}MB8xMY53myU3cDi-P0RZ4ZO|1{jEiKDC2q*5m=2D<^h{Bm-hDI=11ezJ$VynwngyDy6igH}sGm;<{tpAf={d)Em5oX2uy(}-fsG70d+!OUjDvL?@O!-$D z6z?|D2Fp8=a{+R=F>;)l<^N?PA(8mIZkZVo!rl08B?eY+`28x5qJfW3q;SdpcgRa| zMd@h@i9z)6Ai>&XX6q+cJIAp};^9;w*bd2V7%R4QEdKSANbo!~1tDXmFMmR}Hub8^`gxPE>qz>!#8Fk^!PMOt*K;b(EJpJ8Tshs@ z*G_iYBNlw?L-o&zeUol23JM>uXf#I~`c)r(*m&pw{&JLq{6KUZ&_tP7sGHEul;K+D z>aAuK#5gv3)tUN-f~ugXdaNkC;-UW9`z_$v9N{9LUTo+&mP90F;vx9rE!y8tX1#IE z{CsNY>7I06-`Hdp3!K-ia7c#;);GCLfUIf0reiXwQ1vk$_x_AP`< z?GakuoHQn?guHZPYZs2JRUPDm@uX*0Rq_a~2Y-B`XVZnFTS1&Vn8iNUrN?9JTK$Mt z%Hb1nk$ms@qq9L`+gjZwY;N16n+@2<)HucV)m%f*@RpBi51C8Q%1WW`21_)S#TYeuI{Sq~K?dBD|C;ksG#w_AgJn@S4QJx%@X$RjYg* z%0H=YnT0<~_+7S+Lxj|N$04K2I~agJjT$jt$}Nhwi!*vbxWd5`{)=_Ch9TU6!wfJ* z?PRSDSje4DWg1JIv;=fv&%ErT;zO5BC*CKCbuji?=o^QUhjInV&-e1|70&RnRYCf4>X%~NtC0Jew!i5?~$ul{Rxw9rc$Dn`MNdemedJ#XM%B7 z)jLKD6|5T{yRa(Um%8HtX~=HA?uzFa>Qy*;aTnulO)n)-NO};8T0lk0EwZgWiqDT* zSk~$k;|g*9hx6j=bjiMMPOIECv7#bnoP1nt7!hYkRqF}$9!(st*j>LE^Cw=g)pRi5 zh@PKcvp7GeKz-ngJQ+4`vRn>eCQc8qFm?QytcH{*?}B*0aa=@drfk@Cx8XmWD>oPd z@IhH@P!?dWtmYJ&=6DWVuCkM(x>j1MS!8vBOYu`zFGj@&f;lUeiWCB;yv3Nnp~D510<+`?E=9u~AAsBi8>%*jvX%)qZ=! zC?O3Jl7dJMNH@~Wzz_pNhkzhMBOncmghNPo!w^G*LnBH{N;fE7(hbV}Y`yREJNG%y z^M2kx2li+0>$=vuR({u7xP?&nNQ98i?!!5nBFW{B`q$1DI8%AiEBr{Id96izy(Mp! z6&T>fH|^C00`k%C1lM@9s?VUx`t#h6P9l*xZ@Qbs>SCd&{(*M3WoSv<;NRw3?t2k4 znhf*fp95SK%JRIrUU%$&$_pS|s9&2oaTSY!g>hR-D?syn&Mhy>Y_EA@BXI^4sN)-rS z+??W)9CJ#XcqVuEK6qbV$JkA3g!X#&LP_F9u0zUmajgoqgCsuWqn>iR4QIxggv+aP}YWm6j zQ?gWxqhDOM=c~E_|Ags!pP&Ag!|%bur67eueAvB;Xr;(`9d_- zsMjGYMzmih16T(Z156j~GTLL+Fthg*JC&LPOosv=Q}_8Eu-yI1{-BV=lHc13M|0g(IfwV-)Qa@+ot-f+h;+JWk% zU5Mcek#C%fwKO9K^)lRQk1if1!r3Xk;0ONdr?E@TC+b!H#zkL1;lILt$t zCRO2SVFVLXFeK;M(_TkHUbEs=>R>S$3sEh=8qG<@TMqVQap4irVB zgX?_c()^)T0h18kv)WrfD-~stSDQt#-%@VjgkOq-417TJWPBkNw3;tfvBuV#ASBL@ zW#+H7etrS4g^YW9thdjD_Cf}vHTs=HN?OuyR>7K!1d6eWwOpT%q}Tnd<+nmsk{3=J zTEZ-X>L0AyRxp8GTHh+X$WirB{&IJXN|Tb_M~-taJNJ~BGMrM&gm@<3D@NKC+urB& zVUAeem-w!RZQf~%#l0Q!Q1`G=$WdnqE1uBi;#8c6ZR$iKC7b?x^Gd2=`M3a{7=>pP zqnu0b(S8LVasfAmwZmp)4wHvAV);08uNYUOQytdL8A6`IVOcaUomfsWuU3lKr1BLsPn#eektMiJF$_cMWYXGxDbVIk(uO(9Y~Rgd%2Z;oVjrHizoL? zYV9~I)y++rJ_xFAUpf{A?PL?R{>EmNzieaw}nR*Dq29Jt=EwVsuT|)9=7G+ z3j?O2t~H-UOU&lSmbg8un9eM`$P|Wt%2p^4vN;G+^>59uv&y$b4OtHkC)yK4aok0| z(8z0grR!GIeOJX8+l@KL(f;y|?nH(LUCU_p7xpNN>qg>W)mUl>EFL zp9>&tTRYfG$`o~3>6TQ-(5ACeGrKr2Ot0#>U{%?Ns?xoQSfbUNJhF=Cg6x8$7Lg@d zNHI7{zun@Yt;hcA$VajC!fH3pLS~bKBp)OP#rP7FCaWkuOg{aT;rM4CDkP7ccEy+L zh=8@#>hrqDHhAV+a9<7wE@$Dg#y!Wt*AI0W2pA4K2%Jz>8_(pFQ3hdFas1_SN^tNy zC zJmP$L9kWE0v&A^-X&$cTin5gZ%Tj1P*jyWvP~hIxG6`T0^%vW`<|sXZ0iNTiL1c+2 z<5amoo@^$lT+){SznjIHglXd}Y9+I#Ss91kLqxH(8$U++4};~sn^CRMa1blYQ3_;w z-Vo(McW5NIPL&*`*&k4&iORu_@qU)Me0pCAR0DmIdX&L>olC~4N$MO-wkZ9Oh6d!r zsA_rBu$VVnikhI#5BuD8^GpJp_c;l1^YN&$V-aXW%gnXteG`H??7m!V!`ISTN^vK3jIPvIt@jZFp!4SJPVY0Azws7 z(1le#%PoezPXW{C3(MLgc8Qe_W58YFUVVu3^MMAE7NmFgzTywUXs*ec?==U6cT8*e z1@u72R(&xvwH8Gd89q;h_*zSvs$S+;=)wP_`lmRPSSF}PmBFtLEe{g$@F}|FZCE$P zR(BWWO&&K|cPwJ9CGMT4dZUhG^W^jbSMM4_T_vwW<()bF71jGSWXe)gh2T`0Q$pdk zwv=wk-Fl_^G#Rg4S&KtWs37Y;FYq-oM@$fu9NHkEli>bh7mjD@cAQ#hd5aV`UV&C9 z`Y76eciWPlE4?KI@6`r@Gsh^?+TzDetjvNtfk&v$5Voe-yur-9@11KOJvg*sGL0Rq zh3uGL%UDdcFC&ddhd;h@>iP!e$*fsT!K6*nGJiCJ!i6U6g-CZIJd+{{;bIKYT=K4@ zA%((V@3D`rO7LBVEtaBeGdJT`mEZ3O~EQGZ7VnTJO;sFf@^+f1fqz=l#(WFtU3fbRPIZw3B%CscNhPt(RQJN~D9p9}St==YkxFv?v?`{i;U zR{Psu&RACNXGmjO_lAC11A0=uVEi#2n!q6Vyp^k&W`8TMsa|0(&-Zhvjm28WI-BB$ zCm^BbGE}jL)<8l4w*(LIUl>1?gD2F{bNW}{%3Ok(VrjQ>XMHo`sN{L8g5huNT^~3H z)ZDJ0Fjez9F^f$JvUT#17!hGuA3E%j1B0}b<8NeI<=M?`NZPK1bc4?(-hxSVfoVUwqw#HM^E(?wXGX-MA zg6up4Z&{gg2`S3JYQe;h+ESr5FR0U?N~-YBrGCuH{qfKgV|kd@(}pAq4-NHeat^RD ztRoQv3Qr9mDbQYM$2-Ry-J=6ZRgcM++*$3dT!SI@qkn)}rlzH}LGfa*I(EutPIjUj zh*hMwjN9|HS3{WqWdF47Lm?+*Qpq`>wmK37BLy`t%9z$TBr+Rb9)z%Prmm=ct%PEq zo2JpNGRCQ0=#yfTg%lxFn8pE_G^XUjI!PIIx9Wb@$6%t!EIT6bIcS;%KV^eKtS4b||C`D; zTI|l+c^8hHD@(M)sL%NEvl5uMYr~=2db&tBg8eLxNW@&$?OR{bIp=CfAW=YmkmDbs zD~9jh&K1ac>)hCfOR}-w&iFH{^}{Lh1Wx*=YIonh5K!rg;oxaavrWI=#k-3+b^&86 zO^oFL8G7^&=38Xy&~Bq2ffAN!d(Z^Z z8$iiNfY_-={=%@$iCD4zF?opibi{tHJF>S-=g)`L39Iz>L~90l=tXdB;t!uKp-q>M zB5=*-HhW7&(pTR8K=L#}?`u>fj3R~u4(Hkiop=|dj^x614Cj}7?)$bZNGzJ*Hy?k+ zMBgJ5Zx1++U+8RFey3GRNK&Dh-fE_tFbT4kSS;(v{Lw5Zb+?m1UQO&<*xCjgJ(z5N z*S_O|q5eO3bUvG5OFV?Qmv!ErYH4c1T1f zywaMP(5E9D`M7#a{~DSr5Q4%%iBtdb_~}Hb$r4KhMW%!J%iVY(W_`T8B%Vy`U~G(A zwxc(^B9tCr;N0~}}o@#D1lhDi+9^!4O+BHIm(VwjySd?s8 zpE0l~+*CMm^EGk*60oRx3$R$ywaFu_*mAMMlTap;QVPC$mX7$jN&h97+ZN8Q{G&TW zmS&aphvJ6iQl{4(1s%dk6^`ZBZ$^(};~xUE5C{a$+_$T`C*WKatj*F&O0RMgsNUOlT6MrY6`xgV zuYB_azigZqbI9U=)U zZS1(9Rm(5$HQ8Cufz4z;xoo_7r|U6Fsyb9G9fe*=Z9;*%=$DCha9iP7D5bgLXscwW z7gL@BFB-(9g0WCD(Q@2MUf1nzBn~>0z^wC<4#KnflQGHRbUF#{yT~#huLJZ22*ZE^KoE{)re2(vn@(u;{3u5jO4kbPjNI<$%9>u&t&x8<>ym#k0|LOy>m%;h=u;>EX zGmi@VYEp8mnqr`mEK|#~!n(KoqgkS*=Da}+VRIfzEx!Ja%H$oOuAi17tBb!9uZ<^dr7}MC}r1A{{4ZlR{7(dF$!^N`zWM;E-6Ja6u0_L z(uap7T993uBLa?>TiP~o^r_qmceD3tcYa4D+lY^6B8`b(j=tPLE*$t5REq#Uj8Q9Z z02Fd%II+CK^u8S=Q2PDk=#0Y@H}_lhwH2q2vzqBG*N8}D#_kmDi4>;em{`aXgLV>f zk0qFiOA|~MfsH{ zI-0S~^yA0?^Yb zvb6wb+@v$yF%Tj>K-bKra%WrW9KQY0TAFE0Bty%P`$0_ZP|^5^loa8&3%Jkft#;gv zHt@&PsxBa(*&Hr`Ge-QVeOAq~K@Rklm@j3vvckw5*lEdM7-TWQd9?+yr2KqqOVLa( z$BvtuQsSEHypEP{4b$Q>2wb3FUS0P5(!|3AobnfiIioN4vSyK^&s~M~${*W1$bAKG z)vb%=853g_zmCyLds&`t+Jtb;d=1n^dq|$gIGuxyeI{U7nPvSf#q?^%)bYd=#e#~R z#@c^js3e;hkC&gRI8vd#zMr1{h;~hF7i|frcDo9AD5F;NPz^RvWEv%M6@)tqbr{uM zmWU&pwwlJyY$x=h+FHwSHFQ=BeI(G&yJEK-C+0Kc|3mIukXUGd_jPZW(`# z$6CrmLqoMY8Bc_`WTmXUF)tJc;Fhy$C~MDIuWpLev{Ca&p`Y_=YG7)m<8@gvKkv_@ z?zQ9+oK8XU}$lXE^> znvt8%>P`q8V{8L$XMt!=Y2rUX|W2zyPrVvhWaB_?w&pp0lB(Z73l z)b1+cr%EYc2)UQoErMX}-BkrLtAf?~U z)GZNPq4>^|@_L4(O-R9OY@Q%T);O6nE{pe-!r1V(5w_3{K7|y;n*Zt{or`KeJ*`m`aCG zG0@vntNZ??Uh$_&Unfe_eZ9rwXpH)xDxycx5W&5R#hCVx5y6p8k<9ba1KYPZxc6HE6o)SV3<%GvLDd(BPK6 zgH@)d;ZP4Qa~X6-PL^X86aJN`7#Tcouiest?+O0cAA1XQzbbMqn)E$}2!yg4l=)}o z@ia91%%tEJTJG=C9ubgx>!+SZY@?O9#+u_4-#;{Ye_g8jP_+C-*AweK-Ty&>?YCz% zmNMvP#@vkj7Vpl$N1nt!9whQc#{#7!Cwt+a)tbt>#_@qNz3|owGG*)0lJ75n4a^!` zblf*y#L(|We?#guac^B`G+7Zjtdpog@^5$uecq4Wg?;Ks(PiOd938K~1m@qiQIA$i z?rr|Wtpq?T4oa7=h28HmeW3ud}VWmYz-~!U4rSm*ajSN&d&w(t`)xTs-?lLvIIK2 zvHf0F+h=QCuO~HCiGf0+3A>{bs*z=Gcx@^7w%P6%3bnTTZeZ_{pv%9u33jV<>PJzc#>U&W4`iUD*L z?wG`jG2XN;R@#gVG{%c{+83k6rm<bYy8wevQ|$;v^cv3AY^!u$RnUqE$x)_A36 zEdVeDg4lS)E;8A4zxCGxnN<pyd?Hfm?@tB)>ze2K-ixBCpUa&B z8E7#RUwJ^&P(`WGP+4OX;~tI685V2L#Dlp|XkHv?;t)irBa!p&q$<8ziP&3OR9d~^ z%cSNyZJ@~2TymK<07 zMftK2ki9504yuISjr%ghOU+ri~gmkb{v8bi{q z($1#Dc6t`Br{fY2G^b=p*AgkR>T2bMl#EcO|NZvASIh*gI5e_PdqvvUNlLVB&=6=i zJM%S@tIN^4^X}*=eoJ?hcf~*AFRMVsZsQlbT2m)q#IK{uF^ND+5_(U^shxloU=;N) zBll+MVncedH=4?A`*G+wM=0se1GEb(CrV3?d%O=ifeht*or$xUFSLG!^6(UDec2sA zftm#wkhM%qw&vSO<{Q7-8?v{Din-SG?mL*~H7>e`K!lk&CYoTJtNp0Wj<4MU?;m`D zTosRJ*nN+?RwfJio)|&=CWDhR)^nf=kXz5apS*J4Nm|*VKi)C76Y&~a8hII6`Ru#= z-Inc9n^hs3GrJb`l(K`E^Eg0TUUs!s>|Q5gnHO@Y+Km6_pJDSA&Ma&-+sdvr)%U#& zEZSt|RDLn85^4nUaZZXQEl$2r`1WGeMO}DmSJ?PdTfOAPDp{6Q|2C?>&FywfW((-t ztM?^?cgeBA%fRd1-jDFKibX4YuKwzqV7oyFC^nB-aT^<}%g$Lzi*FJ#2pw}R)ARQy zh}FZ-E*7TDTfAN~dia9;bn2;m_bdh8ECQYXD#yia^8dmp*qQWgaMAz!dB#2F+n6H+ z++q5RyRc

mM2#dDq5kaG(+sfWfA?^3KJH2GH~liXx~67E*^fgg|A4^WXvh$_HHy z9`kZKNOAdtNR{C%ACrpQ@+;-oadJ8r^L#D%@Px(?Yqv<@&aH;tRQQGbLx^()lYI3# z0&IGk+^xp^BK#mMC)KH!tZv~>vs`lNMiCA+A^ze4KW~*&_wKZtM6=~x9oD|$UDq%r zc`bh$L<0}=Be9ZW(g(NiGpt}Q?oDJekqA|hTMmV1u5c^9n|o% z>AC9mbq-6ho=Tu8v2^luX|3FPiYOMw4jwU4GLvWXp~6U4NU2`FkD%k~C+wb%ci9>1 zo@@L7{(jG&4bqd7;AxAPX2N1e)b)Iv3en&}+<)Yh0y zE-Sr8Wj?k1{Ao!}_WmB7^oS3&prEs<<1_J+q9KJS_(8o^cfdoAsIK}LUF)9KKs*g5#zhHq zz+V{tATSe%I^NQqYu0q76Y^*ytl{G+8o` z**ZTeFlruK*`)cdSGclOo4ZXgWu8UA$oHb3w+yw#l+Oi+Xbdzp?yXT1J-vd%k}20G z%DN}S*GHr%B$ID7kQi+<)1SNyYjoXti-!J4KG=E0<;>&P+J@YWvOAd-OJAJ#_6kEb z@!UkK*y)$EOhPHOFsbTFD`DkQ1RoR03?M72GwX zM7dbvlxZM4=?1*~Kf-Ce6yTKT8c($+LDT`^q&uYJ{gdIk3LdLKerE{us3O6Wa@jj) zk?g>XcA|}paz^ggC`X}{UO(Qs$};af<2)>ka*=%8=20WA6MO9Hg<|AIpx%@G3!BG4 zktZ#>o=ipBt;i;Q)`m$>>t#0+0x!BrO>t>56Vc*#;rDnCdkz}MUywSRm=4*qz;Isr zbyR>hDZUnjw<1qIm)yK2rB#5d%ZbWkbi@_YI*HQBCnq9Zsg{d2QhlAFYL9mQYa5}p z`(a~TaCkE^S&(7IC!Ed`Ep%NcH!-p((-$pu-gjA5ATHu|+Ax*3Oow;DCWKBKN(o<* z*^5zNj8&rIpYLPz`Ed|UQXnXASNHxTrbz@&NmAxDCmT&4*J1W)HaycN{{*GR;pp2N zLUuL{;QucYMZYrgd`S)HFaYr!kyUW9kRxtB770ZyOl%@6;f8nmsaE(e<-ZbnTh!M( z*kq{>O%xsboC5wP{T4%1DOsFrN%|4RUl@ZLA|K3*DE4Lk_Awvcw?hYD_0C07AI|)R z5eDen-^%qvA+=*u0h?Aot8|4sqxb7w@~GGV7-z)873$E+6pJw!|sd?bqZ50`H~ z##R2b;dWh(e9dUpY;fnPOba{=5bHdNJTOVnnU{>)=C z)HN@YTKrsoHgyf>=-Qz4ehB)#rwRnhq7UUvKdmhkr(ObbS8ElA2ioA`1mK^<#(5Vx zrk2bmZ8;m0$J=YyGEkEmX3L!ye$1vqiFN&Mu9!cDAHLE?LFzXuk3+d$|2iseo0XyL zG*7K#HZeg%UjBtK^dTntwBGqQdcfE)==qFFwQ^Pp*roYQ9rDYDQZI12_AY)^BV?2Q zx?enX?88-FL?QdF&OXl2yKU^6PyY-k`vri$Qt8nOm+K_|LeNmC|H))+$D4-)o zfM`-sCt!R577^Ni8PfQAh4KwJR4(|FlCD{F`5UVBLb8XqM}FXoj?`2q9(3zSUWfzT z=+M$H0fQqxRF`_5ASVUSooFWRq3V>81aIt<2-grqJ=vvPsf%k3@VZs|&xP{2?e0gu zyxvJ#HqS%`fGEUTBK{rXQhhT*Du{vC)U5+8st9^xX;4{~$%;#*9g9ICblMS}T~$qJz4J!FVJ4-HnfT;7m7#1LVg5RIp8-v>KlSMH9K=Co`X#}%Yyhw-8EMFOvN%~LjkY;6xKW-}ZckGhPUl-vJuqr7oDCSx>GCl$lBO)` z!Gf$C{_mCceE4UjmOzm@Fx7z-fc=&xr>L`z!xqm96bEp?2O3UdI4e>PV2g1c92awS zyXP3OERCmIzDBB3BbBF( zFh|eOeTn$Do5Xe(pUPF6i}apyX#Dmv|K<6t=GX zYoMJuSOo~i{OXej_td(DCHi{*jgptg zzh%oQ4SLSE-45-U6J1DZ$|}m1s_D!e9KMYU9*@xDl)%y4)_ALF9-wlg+m87mD}J9F z?6mD^0q?R+*d#DpmpIATZ))|m|Jwmn+A}IhAsU}2n}3c?7eT-4QlRyr%J%H0j+v1o zCnGVIa8UH+&RK-eJSHNVj_>w8fzPZ5xBV0^neI0tHSi;W0nf(8opIWu;5%4Rt58PA zmS2m@ezW8P&PlAr^V}r`688yJ2Pzya(?>PqOx0xk&2W-g{GdJLK`){V?-o2 z_Z-Q+b;MT_&ED96iL}5j3w1LvGeX5O?sct&`dD& zCCiN3qdPskOmb=)+@ntlkuBRF9-GFOnpQF95~a#jDSC7RLb!iOgDHJ2{#A@CqR8gd zKJH1;kVWQ|WKD^tYx2CDTQ@BhQZH3wIJs*h?OBk+g{(XoSu6*c*ngk)Kc3US-I-FG z_&rQ_nsEu6!!5Yy9=JU%^j&Hig_%k_HzZSIc~UdZ-81fGhGiIpC`U6%hUM z@ABb$(M&826gw=N55{dQ^kwj5t`p)eHPUXrs?|xh)K0%sGt!w8`pPjj_gEVxfS*i` zXII*w`So^{R#_b!d|IA>tB0=9!4tk)x;G3AY(E6O&4DHi?b+=0d)2J`+98(Z+#q;#O&MF>yzX4G3!q2+kfow5f@1P@JCh9%%(2c z$$MdxZc?eZl~UOF6oFE~2Gd!UZ=158LbiT(aNVx@nkP3H^$clp+sR2alDZ6edNVH7JQgw7ldTiE{38sL^qq7y- z_!v(8NxF{RFgT+#HKv2ek65sLj2J}Q73b}K{}NoB>K40pofy1=+sf&7AIYXpjAx^Y z6<$dfGD)p3l-;cFr;0)gC1;J=i>_}Gt>uU*M=M6{M*!7}CqpYcEil_8Cp4&+^2v8F zlvAw*Klw=`n+z2a>8dCcmrATDo~$22Fj>jz9Q|l1KSu&!;PKd1%a(zAYwnk14sg^o=GehA8IpAG7*t zW79|-;h7re#XFup4c4`&%lk`@Um3Z}rPU3T)Mty+Dwer7%Yo9U4jec&k6xqG0ktzL z&T0ErxAggfY{gchy}K%l#%q_*iL6zEw1Cg^7xnW3b%G<5(#XIAJ`Wru{iC6j=oQweb0K3^P77nv9>Nr(}3KbS+OwdPbqV zF~!6jWDJqp(gau935(TJm}oHwA@qwqf^`LW&58B0R`HA>&^&G5yZ?%j43xSi1jpxu zsPSfSTzX#1`AVo&@G^^Cr8b%oNrciLiQOj~)Y>{nOTN{*m#E%84b!}v=7G-5kd+H` z(|R0%gOZV5hNI^-k$~4BLb#^^DSKt~#^=R_Pq#-xTjZ0VQ)vI!o;GQ1ptFoRH(~r6 z&6mFVx9qdc3?SQS-igIe*d%lQ z#Qfx>^b@Rcgk>)kh&sNJ#ZEZkN`|}Ry|5X3g2~8E?*-EH0*{YVcu9dY4q5|VT~;J{K!8WR;A(3pXX|EHr)qr+<^B2a%6aTWerz*rJ*aCZ_;BIG* zNe9Wk-PC_u{-~yUU7xAZ59Zejz6xM18R^xK$&WT<@4t*5UyCzwKTi2&wfaSlS`yw3GVr@mM9gMjKh5>WPr!VI!5+ z7Kqb+3cm}}9#dbRN;i2_#-CM?)gI$$BYK-DemP))np`f2VB0etv9L+e+3t1eIlt?7`6bpWoD(Ii5H`8> z`-hztUW{_y$8w-H>rFx3b+ln+g)k*5stC|e!Kkbk#b)M*bCp7Wej_Xe*a8?XKlxBY zSwB5+Kq&<2rT6FtZFlBG)urz_)3z&safQe|sxDk7kAzF|vY4C1UsnD7 zt@NKSD$YE8C&se^w?sd*R>ce*ZqxnAVx?U_`C$7hz=$u#piX6y6!DScduBRyVa*j= zEAxR2-yfC3CW-gL1Jp zA3w+_z4z{|!A9b!One*YFl+`cYV5>>NB_E}#|(15paxDDq-NC`5?gdA<-+i1G8XrW zwRyItK=Y zsAng>k>&c3#MF}|>sI$2S;mb2WN>Ieqoas!$PMrAYJce|HoaCJ&oU-^d_<6+wqYO#N5{KRsLJG z3X9?D(t?ap&+OK&7&fjhdMeE*D!GhnpkD>Ne|F~pg4-(64J5Q;dFG(&kta4q>1Js* z^(R0g2Lz2LdA8bIIg6xJ0aal2>HeQU_QnfD36S>WA^ZE@%h=#jUMvvwbuRP(}0_2`;XiIUomx9;hO<%t!jO?(V_I4 z-f?B%A*Du7XN^IgRn@@`q4%-M<;*m2rJxtr;9tDqLxcaHOB@DPh*2>%T2QjgJRE2IDA zF8Fr`GEZ{-j&n^)t=~Zh3%`RZGX!~!QyV%-JwqEy3|eQ0C0is8yuBSgkP+w00bm09 z|3#%g@y$HB3qx9AnAI!5JJO1|a5Bfz-> z4zs8ItNogF(lBG1iWf}xj4K>hdkt?s83i@^_Yx!je_tBY2P7PwUkr|KHG6K1j}8=Z zKZq3kK%9Px=+V?mw($X(H{ULw(0VHWO~&7h@h4C6KdgPfem=R|#&2B#LVur<_uNEO zZRM~Qm)_em@?G993eX=#yAX`Tm`pm2zlr!eH2i<@1E6QJk0a_AcO?5BHnMM(D*1Xm zl=~AE*j{$rND@QmY}GE2zDn8%WR(1U^wRp-t;W)C|4AbJCxc$xn%B+B{H$KrT#N+i zx&tXLj!rXX=EBaur;X$fnP?{zUNJba;IM*~laAe0(a|jWXaw*<8o&p@rmhL+{_NpD z*KCU+VM)80T_`_CalA-H@9zyJ7KotT4Pxead9MPclo^Hj0K5pWn)g!t|4{FDWb;2M zL{BXrd9W&`Ub^2Z{q6f=eJ3gWGr+K^&3defVNLwC#ObR<@l8U!N;QGoRo`3OLnHom z_!gbp*8eL}G?447<)fe{M&ofimRCsMFLzCU)|=}jr71=D+YDRwlfPjM_mCHf>J*i4 zby9Yzlks6^IO8z)&MmhoU8Iu;Nh)#Pc!x0rw9f?##M|Ab+^OE8*82N}zv(o3dgpKF z{r7DglNy;U1^jT}5EX0iAK451vL<6yKCZ(>NxljUC-v<=h1r)s3>sUnd|sOov-!tHUu>Ot}1$tpk~!S(+n4F9*lOBLFovP^G1B9+5*dxAszW|K0diT_ zeP}>56}9sV`{@@h%ux1Tkq&mkCYxWQr&FL4oeURR-=J~n3;b@u3SQLQ~z z;xyKGk`^u=rj4meHElM4Q7oHtE4S`vlhUQZdFxC*Gq}z}*j6DCtjw!cUb`iD#kVr5T7?> zC+;ku$t3;Y?3&u$nEKEF7>7RJ2ZrkQKMeKZZ1U*sNNa1EX72wYQ(2?og+IUqy&}p# zL|=zqM6&+W7fRQ^uAIBkTT~U^qSFvRMl4ENMl_}`&|9v9WZ4Ndz@rn4DJs;to>&*Z zm{hB@|LX|&wnlxWQ1!w^$5W*`depW|S=74%a{`vXCNtglP~i_vvJ)gO)){1QQxp<3 zKRk(?>zR{D?2kR22&*@@H)9?7XxS6Hw6W>WHX_-q@B3O~@uTAxr{~(2!ZEBC<&-2jsAV*3 z0i){&-<|jmDV7_|V>-?I71o3AQeFkZ`vPd^FrC!vcFj7f2Z*Lt)t&IzCGL=&XzmMh zB;_1;rW+)NElhVh&OtrZf7-(ZM%>7*6LtV3Ch*5|$@yXO)&^ts0%WllQ_4OoA4*g+D zv98wn1C7G&Bt4O89IW@+D3I4Qk2B8L;S#Ar^QNhIb1{E8oCxbCbWL3kUFeE2e(=x- z;DFTQKOBHVw(?RWusKlbx8(b+gJgfpbO5)b3^MgGtX#HUJjX(68*YZt{mlsAfZ|2R z>Dl63C->&DPDSk4w^1VcE3c+V`NJ4Z-f_s6x?Ml6e_zais5ATbC9pR9a+}6ozbGi# z#Zvwk2Ka^L=*3m1F{2WCu9%)Gakr)9zR2lDHng&~4pfwmr`E{<Znr!O@jrDC$;;*b!00L^RoVc@)_j@BY3pM` za&5VDhb-Zwu1E`=_pU7g08IumOZe2-*}4IjP=wCMvX5&-S{*DMxMg0kd0!^Fa#ySV z!a(;HN#Ms0fP{d?QbR>t76BYuOIUM$Xto1PnZ9OLJ zJ55m*T@dPL$uk7^u=}28rqKdx*`$ROGwSt^jJ10jmBFQc6DjR)BH{0c`IQ5>Vc$3Z z%c_bJ|NejrxUNBRU-81eqZ2$5ddSofqKV{0Pb<`#9q#mO8;?`Vzte1Q ze}_~ybT`Y@4et^L=37$zQo@cYxF_4~n_%CpEASm?CV|vDEI~fySw(rE=FhZTe~pIl zs;aext(*TuHZgqj7esnE_S?KLnJum5EAmKARW6U|8z}H=9Z|_|TrtzHt~|C(L->RY zls~F(j_N%zbVA~@;x*rU&gQ0zr&r)y(KA~FyH%;3{aS7Dqw_<$K+q_ga3Q_|qIIqt zTbeKJqm)EiOPbJ<#hnlyNL^te$6FtxXI4-c7+;s0wN`}|_f&a79o$Cgqr=Tp;Ef*B z){?+|>2et8ubSeO>h6AUsbD`&YkQ-rtzTrQ&9xwXza+>;fkG^EjJjgy zoOKy$?@wZDrvp3qZoSgGxQ}zcD~g>5k8z~fbVh2qUeI=!W{mk>&28u^k7di^>b;%G zbK03u>5&3P16HTWzH(tZVo(e?;zDMP;W#Z;Fd+s8gA=TkwSB${A$0fV#Vz zQS2EN(Eq9zFnRxAGROHstGuUk|9~+W4Leh#M*-w_OK6PODg9yRD;E6}J<4Yo2KJ&P zwKS|OvUz4a*VU~CzKpFf4beO~E~pZYGFLgwZ~3^8(#h>_HVKRa%3lm%S#Rl-J`gec zIDwdFiTQ;+#LT#6g*$@VPibmjd`{)NcsDcc7V0&7o!=vA{ojlEmec!ZEg1~QkpoKW z3ftTf_)uIYQHK$z&%36q1qp%1Aozu{zD7o)DJ0?$03I1{dWOm9$S6pshSL=A||Ha<(Uk*^*y z)(rgP|F{C%sV-kvtztNw`k{yAzoXG?y)B^$GG${1kp(BO9Cac!Ry+8GvmVbYK zn|O;oIuY#wDK6w&@rs>Gl!~nzcXBB`Zd(7HQ`9Yxp7`}mN+jdAox)eS^`e=vCWd+& zFsZ93d&!Tk@&F5y_fb~jSTZ&6S?jDG)}N%mDG!`>?Ta(tKhQtJDqNtDh_cel(E5!0 zG~n^ayS}#L&JU8*wMun}L~2*=I+S-_o%hrms8=e*Zqw#Q9PdTb*IT&dnu(s-CXpgO zsr@u~1Tk@<>(}7V7t&^oV&mtMnSD1^d>H(EXcHmIH6RwGvW*o2r!hMir(aQPf^l0y z@YR>{_44(w14hbs|X*6L|Y3on#p|qc86B(Y{<^1 zK>3@wW41DF!suTZ+{leXIH+gqdc*)(G%2t`YU(pd--0DBxIJw5kv7T9iO9CZ5_)ap z4xR3)L&cQ!r1eA_uv1>h+b)a~jPAOLjz0O!#6f@OQP+OV7{SyaxXgbu3m5SU?eRU% zZrXfZq!F5Z>8{!jpKVxr=I|gKA|KAcH#iIFCGH=Su0|k0qTsVP98Ps*Ya`%0AGI9C zk?B7l1zI{eReT)PY&TXcpHG}B&GXnGdhDX2I?Sb_ruL(G*Kfz*#4tY2Jj3UuZdo&% z-s{5LtBZH(uiU{OC1c9bR{#aT4|&}mTnPZkPUOUM5Rs8pq6aT#ay387VY`E41rf~$TD`4C_)GD z*Ryd<&G_Fj0)4KICqdl~SyR-%6o2vf;*5{>uuFFb;abd>eP>o7b}l6giC`7GLr%Le z^{sFEvrKf%g9dqybxF(vq@C|C#lW$=gFBasvww7{^82H*{c6Ry+b=S8KLGp%OYvd4 zD>5Qj=7ezj-f7u*rAT~)L^cJ0$oAGMFb>C_5Z(ouf@q_k^rz>f!lk2j!}1R%e()oC zUU>NN;@=dgBp>5Fx>yqapyEAw_*zUk=cwuyDa^{t(=>abJOJvPCy(%~UgbLYY)a^P`dOsU zvic^W)F?7ow2t6XfkYiXo2E#XYvjU`e$k=BU%Zhc5A$a2nlP9@!G!;@w|=Ygyc_DH z_o--Ozq}X{?W)?FACvXuU;}Csw2~f1xaVtv*1RrLWqhP5SXULk?!w_3>ZQP`gJRxb z1E~1>s9WX2e%^|RFIZW$6Z}T1y|LS+yHT76oqzHS>uqfg1svBW1|a}aF=F~@IsY}{ z+BTU^3v7wDs;fuFJm)lbx&w4@(ZgxfR(zKzYvj~X2$MCO_>qtugG^npz*Z@^<+dbV z-=+$05Odp{uF46bq6f|qjx&z~OGI*eg;AGXbRsluSQPd&Md*wdTQRi7v=!5z&l5kG z$S*A+;0w25TreE8B@sInQWq{{jeF~Co`*S)mo{dDf_osxbjl}!vS%c8HF8Df87FOA zu(8flpOh>F&}p=sQQ{|!KOwTyEH6D|n&>Q0WTjP65p%IO6)U1W4f&8?7+NnFA{=c4i}2*&))hdjO`>QkM1=FCy_Oy z6k>a%82;WmekHZSoR+>6ky@_V@)<@yyUi?BmS+;v<5oCPjacmXJ80soGC5j(heNk$ z|AYf`)cI|LZqhruPdI|NUH1&8475+DJByWDx--#O{vyd zDFus=V^bJ3?Ti3lN5DT`1Z_O5weGY9d$2@^0nFXhYV~(lB5X$EN`B7k^<)1dJdHTC z>_;%r<)jVi&v@#trD$lv8;%a&7YUrv8S5NwHpd7!QR>#E@Fl({?(6=zqe(69XQ>io;Efmtn9qHj*pTfLfvM_xcLvZyA{l zv5vT1kbl1)CP^S42R$GYTt(envN_GYCpka_r1KCH?9=I#sds$>@kJ3b40;p8Xi2qD^=ZG2Hr?sc^P8l|G`Xm|+snD84p&_F40;5LHH`~yr{ zLK>oo@&9LKVM%Mx!G72B#_-PWlcIBzcgks9oU1v!d?)M$ZVxyXA=U9x)DLv=-Mx35Tyx;LFJ zUzbS?Iho@Yn6&;vtJx2;>tmw3739%|g2?~vb9=Fulo&!)0(;_~QSsJ7sFXnO79?V7 zY7V%H?L0YO>|xHbUIO%APg2lYiHC7hTtF?&PO&Kd22ZKa&=~X>%_OTX;J#4WW@gg) zbo8m{OGExRrDDq`lJKWlvHwQ;{{eRI?7Qlge1+Um!Q}6oWZs4GXDReJHk@3p2?N(n!C9^eh?fLG)!flkN`%x%CsL2NIszdzJ7wFp zf%D*^CH`wVZ54a!sYikj!c$0j!<=gVD%$nrCN&bt^D73_{fnutuxIp!_*zVEk@Bvi zYm`_Uihjg3 zMZYumZGF?yzdlCZo~)8*DL77*zpAej>usNY(jQnByZPWF>&$(La5nw_Wmyau^(Bon z&PEjk%C~9AoPt6*gH&i}i8G6_ptBDtfPN|jFE49XbNu66X6@}0u25*PH#>D69mqqY zM5F41{0O>5lu6HDawP@K&cc3zgmQ|$eo<6%qm&=+waSK~lKp-gMwRLsd^PyzVQd8l z$^bA4nNQD?j885;rlA4#0d}!IGj@9lLe5YXU^~tmyDf+~ZI1dPx|%x9+bB4W%bjgQ zTxEmuQJ@f^J@G}Gagx3Gza6j3)~;nQY>s0eCqK9IUdsmVj|&v~tx`>X{l6W@arvtx&X9%*DEr)r76RK)s|G7`<-nc4lV4mRXpU81g z{pG)g{l~ojb=IWrh;zZ8coQ0+=Ni6t@kvruJA7i2Sko>t9jH z|90!&AN|t_#1XxA85V2VFYtAuge_XjxBe^U!0bP^_em^WOZgc;aet-y&+Y$q__|#n z-_N8z?yvhlo$~vq8~@rt3%{1z6`QU;(;;PC zDS!&91B$C2-PE$8%vej^{m$%70WB|TmvO=^s)FQHajw*k--5*UiX$zY(O^u=53o17 ziJQ2l-xn#H^`t-4sJ9?r&2ab7ykWhZVwrM7OeRZMz^4Vny+ynPql)itlyd|g%d>); zaKOgppH3E>;*<{xDrOx)_R+eDPha(FrV_uSH`w&i^_m3Hlm{}*tZnv|dySMo5i(xe7Vr4+kQ!_{-N_Iu4D<0*OIlm@qTzD33%vn@WnR#N;8pCR*=HvN3 zeOEBF>yo7C)N$854Xd0=8^_-(sul=?V5E8N629by62Fx+mcQ;YG%N@%yRD%k=QJi8 zL{wjtrIY@JGP_BRTaSphIC;oX{eI?#PU4L8o$W>2iyoS~b8prGH)US>|8~3doxMp=JpJ?f?7&nA`#Z0kNuU|W zFweiS;<^Q4xyoDFsc)w*h2bA&=HN%s%}W|4lB~^Bg`<(_ltt=Iz-0o41(UyMpvAh?E%D+y+ z2A%s%1Xqd9D*Mg?9P_b0a26}(i=j5RsHc%2a`h>rqUY`Z#NqD-!>2Vp%-;^=(~X9{ z^aU7x^}RHkwzH!-Zjoy&Twt4Io&7UkiI(N3f9+mmFjuQ>AVOC60J{mmQ~D^Z|Nh|W z^3@*vfC{HgRs#;U4ZWfbXrw`Xd`1#C*es`|ElS(sSKOc7X8-X*EoC=wAIXf}D0&Ek zl`m{`_!Ww+N!6d1IJS7}gir9-!e^2m7T?F6FM7I}Aup}4Pw?aDw&sSZLCg@TT@ux9 zJcxVKZ}2jf&g=_sXfLJ2Y<1nEhdserczA}7q0p^lJcT}(NDfkH;)iS4*Q9<3jzxJTzNEHNMh+U->L$rhs{|pm!<=eHJliR^R;4>XUIy+XN;cfKO|g}w z$tVl3h%>z{Fm`huEs5e+{lc0iSxk)otDbK!tpV8NKm1`%<+zo_MR1zh?Dy^-V>()H z;~a{qcqyqGplP;xkvBfjFJ3UL z_OtMQPJi+GhkeeZO7V=!uKNl?2WJJ@HM_1p1Sr*6B?*Wg9oCDP*5bc()QKjpI$AGqsv|J~H{lkK@ z32`NaO&yeTj4-weZ_UC6>Wp+3_PT}6VuU6SN8zXNl$!@m=$>MKTvoF8yPicp^?Oin zG7)ESDiw+5Y+_i33YoYNd+fR-SI9-o7nXI1MU0$xR6iK@$8nf-yUd4AG{{|mSEP#B z@EF8_nK?zCX*FA#NUi3Q&<9bf$VUHGKvb8cxD}NB1nsRlYR+LWyNwsa7H*uQt-C)Q zXY|Lo)Av96Nl^EP3pXcuZT)Yg)$cXP1ei(}obnU}m4|AR~&HXf6XZfGsvxNk2EesTiwT|&K z<8?mGH2opsjT0u0GZA#@=hLd88t2q$C>&iSxKX>>ziFdH)-kGGwZCnX=K;fgxn*m0 zRI)k?mA-s{ie~!Z9Vc}EL=J*zAe4Zc2!5NgqA>i^;h3?D@hP2Op&Sc5d}I*wgCHyg z1!Hz@Gefrdvu?!SB0ebEGZLAv;E8dOYHR?x%4-$BFo#J)IgdEH9HodBFNCTv5Z6|t zDdf7i^Yf&t>9uUegyZXSie%UC2+95SI;E4+>k_bdVN=jeUGhMcAf3`3!$m2R#q&+@IfesQn9QdB5Wk|^kMBv#K! z;lNR>zy=T24hT`l^^u;&9p#^HzrWV4Fn15!pFBIOiMYfB4G=_s zgoODQ$|DGZZdb}d{idswdJnvx8wuCmKFyeBWSM{xvK5QHR~F%aS0S^shFJnPj(FxE@OLrjwJV$fbjdb%?q zys2^>i*WWc9qq%(tJjT_BJ6fBk$ z$mQzCl-^aK zs>eD&EDEgUP`ZMliI=T7rewmi!@{gV{8ct3)^@!IgU?Q{a0V-iI=D=1@!p(QJXSxg zrvh?~teZP5T{K!ia#JZg@h7r)3nyT2g+b<^L~eLc8S|Ptkj~HYBvkcRMn;?x>m%wj z5(6S}>Jbx|Qd3=UJ~pd#tC3PHgn28e+B6CQ&*2rEAE)7_fRG0BL{8b$}But*S!gftk z5+a4=(CvdohRDP(TcQE4o3-;mBuPQA4YO5dfv5POT|*`I*u*TXX-Y9n_g_VX(f8QI zLy9eLjE7~J#tB4s+DZ7=(A|I06-$=WLnSj|<~&E`deK_rq@xe50Z1X-bP3#KJ4Hs| zpS7K}kcQiBIG`H9&PW)ly(K|o@kV~Ispsk2$E$4fYPy{U65)vg8lm8aWV1kDnKz;@ z--NLMuP1tnlM2ak*!-~7a-Em4^AOr43`w~ccr~g9*evIIm;sg#$q2#%&!DG;amaM^ znBd$%0Ga{e(zMRXHug_3_6Wws?iN)@*oX=rjk({Tu3W6Z8}`BVfd$`L2pA+kczU=MJ-Klsw*$6-?09@ zwUc)Y;BhE%sQ2IujwCMaVRj8U&*bRXdmYAtvVa3%SLBNF)PhkF+pxlWzM{7ca@r3IeV=H z+ETnRw>x`%YqrM9HZ}1Qx$$+6+3;lvkfN?6=7m)ORhp2B>dZ@Tb?zG(ol^Yx{+($x zpM%phSR*rS~+y@<^WtkhCuSg05n3dC2%YjXI4NS8_;NbqJ{r4^A=8IezXu|WJL z^cm@7k}OFGR3ta!0~>m0mad)Min`?aONglhr#ULYYMy&&KQ;7e$jy9C!K<#2nXH$W{LH zz(XbPPjv9EQW6eq{cJrOJ)GfB(Agxl++4&-?kmKNQ~&^31gA9Yc2)e=N^L_EtORo% zenv`J=|GQR_(PP!`hCZN!J)?}o}N#jc0s*;gQp@cWK0R-9q^i0v~{ z(r+&tSwJbyYf~-6JmJk!XrR3!ervd#L(?3m_7JND!5K*_+;vR2rGc+fGU4iE_sl;e z_#x-e#AH7s4PViWlg3)lJR>ESev}Aw&Hrm1&zFtif!G>++I<{L%boq?* zlk@Bp?jpA91nV*y=Sxn;RTSTpBs2;ep-#O16WT*MvVdu>^-Df^hmQvJL~G}$ zrXMk6G4jiQB!|}g4WY6yFc;En_`ccaiN#;a=(OQykxe@um%-J{+^lte@r)G1EV4hf ztcUYWPSmQwSCH2)31)KOHvdG5$h;q3i{q%{ANM|WIrb;MLKA!ARj4p60~zf}X$q58 z66X>@od*a|GBLwV;o3Lhy67?^$(UV=_!AO#X?t1|N1$6l+ugG9@ytZ@WeG?Qoi(k-`QF zknA1A^H$UGA33$+mt#5cp?m}Ro`g6BoY<|r4~WDxXDlMMPv+Q87&us+D~KhAZAvWx z*@tcP5TbmE`@bRhQ+DCpa9Rg0+J>M46qP_EJNEQ6X!!U~l)Kpl$nXsdCQ-LIGU(EYcp z6uH(?ekW?T2L%{u<;ST9vIxgg+xy8e{M7 zX|shfsZhGHe<{dRFfa{!f8A1-oc=D3+~Czc*F*N^NPT4Ew%S55qmO@mQCkNF5Ss5! zKFJcz0e^TDr_d)fW#BJ9WafSCoHX^Ji7AXp;>MN-}sTJdpUc$9+IxFY849)a}8*O4W%Ir#K{>Y8+&$wi7#SuM%{2 zreR0gxFExP_nuo$>R?9l`xTRJl0(XiuCP;=!VkPwxBGyF2<&CO0O~(H#ztM&W&E#r z1H2cLNoYc&rgB{zu+!$Blco9%CA1s zyv}1oDdA7sSly`IfrMi+}g2(i7|#+jk3iQ zznuOC3u%5y&fG0rnc{|1I`3|pSw~alhaM;eC+(Kw2ISDmzB-)TB}|7c&Wot8d>Jg| zV8rapr9pqfz~&pVNs9Eo^~LBMa(RjFSHt%y(~O6#edcemgLXJpo>2nB)PmES9(-sI zj&c~5OC;HQDfiNdv1O1l@bs*JQB z6}T5dL=&^iwE=@Fb<-f}AK*qaefSROEuvC0M;lRhgm9$K$H@XI^CjAPS?v!vUCkej zaWobH32zijbyz}@C+5=UZ=hoe;mfwo-)KfdlqvzUbu8gkFxuL>=w!rHZ!yub%KH@) zE@oxU4-U4<7q*=%cY2&?7*dm0!VHMx!~!xdq%RZ@u7>Rz*gz(fdWIZeuX&UlRDoVN z>=Kfakhh1arV)JYYdRhe1)eL zl@P2_@EHfc?=1a@W;FwK0O^SIL-{?a4BBCi^$*4(UxnZ->%Yoh`FJtS$}YI4`PFhm zNVYdAS$=LN9b95@MvZUd*Stf1FpKah#U9^o*rI7uhi{P+D-xB&vml0;G^wJmgeYlv z+DoxQO=HPLDcih?{Y4CNgMN9i_pp<;PPT}J^pv7&I817o$VK#vn7xouz2m%uFo`nB z4WpuP&$4b#uiLjMhA-+LGR2S`O69?{NWR+6S`+Fzm87#1B*wjvP;Svn^4DSYqCCJTCkDoP%a2&$Jwhx3;}*CHFe&hO@fAr@MuCmc=wO(!`e*xW)E7y~iV_SG0y z1XmV8aK|LbMNsxubul!#kMF}1ClP2K)p~_TP(k7RX03FK2qd!Mlu(cav^q_=ovX<3 zuF1)coUuca>;;Jpw66R#8&5rn3^`-Kb4VTSFxWNEc3HU>_eQmLZ*Kj~t*z^wbhPo# z-Gbfc8EL|^Kjmzo2#KL7$GbCxEY!_up-m{UO$d2WVy1A%oByC6J6>#7@s@1(>gt4> zp05pGw~qOK)P zW80_hHgK7-bmY_U#&_kHmiqY9zEUnD7?cNF^pI8$b3`(|roo>4T!;81lhB$rWFUtV zL{1C-qFH1~Z0{?$<0`l`_CgH^Sn!#mt7 z#B)p)<)a?@mt;UA+O|axI>5%@F(N-7E(1@9COVI4@@N0yBh>&7GgjINQ9h%%CNus7ckbnBexK!~ zD9u0qZCjI4#hI_o+9M}zM>XGdQF&17_>cvDvJ)k;ZAjiJZwwks}# zJ2$(>;qqRQ>fxxXVIGTN*e)IXyu*S3hrtN-#9Is6#2{u>?F!NvMa$Pom0}?~7DV&t z#96-{{IR(ww_Cqah!)UcI`&omZYNRlKX%$brn(b}42>51ZjsfIg0herYO?9N!i^s# zY|_tVb4V#t*fqk+lN3WMOouVA7p$m!J`*I4!QXJ|l)RD}sXjbnUamKI0wl$_E_%Dd zHr|oXsY2rYq}vh08j!K6js8u$Jp#JEEP}%kIz;Gcw2Q4B5yRMyp{d;f2asAg!2Dd!@ZsRUc*?Vd*z}6Q9Bc|P_VmAk@@U92R@Aw zVcD5x6AKG)!R6i;kdKp+tu@G(eN|<=t&idW-kC=$wu_RvO#*x#a;)U#8pJZv1436O z98JFh-JEstUXE2xuox9_3w`N^on_G2xd?o-5a$LJeH+P3UGrRua2j@dJOH+vhX|m? zjPV*-#tuy|()T1@l}|q-p~4hRjsgw3rhG`RY|4~VRu~nJ!fBv!!cAp8=+R{`RqA7* zl8!H3DWKY?>u02(Oa~nKN);)6NOW&eTop{Nf*-~L}Ir{H2=@0 z;8XAyY})`-fUOqg{)}Xq6!p1R>8G58?h@c9UXBu} zM!X{KbdRjGFMjI6+-o|U%91-qF&P6b&rl{5ag1{EY`+A?8G2JmtaGvSA{;4);*3xq z*R!>o4JG&?d|~sOdUI5_Dva3CBOVI41aT` zB7&>EO7Bqt7WR!GY78|yDEb^G?3_&~X@>e*1uP~vNQUo(X2&1#Fzs?Ty1sKsO$F!2 zZiTThgFEmp#;D42mA+RIZ;brzaHO(>k97>o_o~Xh9ai|B0%vZ(P@m$x^iiuMJcqY1 z8$_fPaQ`J|d$F@5d#0A$f~=u5^7-r_$*RY-&n`fK-OH%-D4OsG#W8J~%bbh8Ykh3} zyU_ISORO2220wBV3NJrNE&`j&5qxvJD&y}G45-I8T`8#foZDwVmAc&xg=I;aGk6>Y zF|^Vow6hcP#pxZUvlpp5YRI~&Fn7Jl6Etwm)})2r zvK(4P)uCpLa*X(uC&^Fk@VB~5Ccro$2dM%o&MPp&6vt6?0;RJ)Q3V`==94M|=VFtQ zBq!d~`4E7B74J@Pt~b8f!a9mu>e!j2Fm4vna7UDX64ZJAG-=a8$1 zQFViHz3!IE9rqH(<_c5?o-Vm2e}E?B@J4)s^`2_oT#&h0h9ELs7h^{hn_j{bs(^=^@p-s_NxrM&9dBz_H7ybcssBZ0vis~#tF z<#OTN(;^4(_G8pSmsYBUtCF{mlnd#Sde=^7OuIW}^|VOd!oeiEF{O)U29ZjrBVl%l zTGG^Tn;NX}1J2QFxRdQ6AS`*0%8L-KFj-V}rkN3qbu7;P_j~H48;1+KwyYHVsgxyd5c)p^ZjD)k7bq7$`A2QQ^i$Y9IyGGOQ(6)>EZ zqQS#@ko59!EKip@HUij(W=IL*Q*9l|5qJz@qJwAMkoHb6O#(mxX%b&!}eu%bcS+rN_-iUhYxwO}6ooV$q;a`nq;?x25T6CfiDP1&wN&hPgfm@bTK6rVG7U|L=E7p-Du{IgRW|tg~ne zjAC@#R>a*{I`ooGgk1&rIecze7^;u>Vf2KPu@l!;B{o1BQxA&}_ee{lulSjSyEOWf z6fR*Ef8kUms4z+Ou=>QrivqHrv%e=cx#Nb&lTONyMJ>K%pxf>X?W}N2GD@o$$ug&8 zNwM6bOuC7-D73CF?`vf$Dok6NrH)p`HgFu~yaF*518rd#HKDfdWpi`6X*)6*B$p25@1S6Yg=1FU?qK%lc$Bh`TUp%w!5>jcQPq-@b}6GsRTRZPE+b4ZJu395%4%EOhxqIA;j;GBg~Z{{8B$FV{8| zOw4&+=|%a={zY8Zw|yPaLXkHLJiNM-`_=Ui52b2RG0FY1AxksxSl4Ycb;zy|9Z7Gm zdR}&`0#lfHG+tG-&@bcV$rb>6Gi%0P0Ix);Di9M2c zJZd;b-}mketN^njK^v=M)%TDd+9%uwbP#`AkC z592daB2nz1hO3yk=;m-C0^-idYz{Wj#VNJM#1heH{l#Qz^n{JUO~s80!V-Q+>5QK? zD_&F!Eu0u@G1$XuZRyTAn>B@t1niOIlUYKweQ2z*w`lQFJ^KHJZG^n}Oe0rFC5hL{nplNwO`9K4*g2%*w35p^&^|kTd2SE_y~XR zPT=4HOI~rOyioT9V5_mjE;zIqQ^6Ncgb678yrw*VVmOL>GT@Y5-xS`naSo>xJ$YO&*hf zxzD0>+I4c!VT6|!Zu>-WY+jERE1!9j_4nHkB0#hHXi+PE8pg-eY>4NV$H7lZ@zUQ-plF6h0ANrhTJ7E?J@Nz; z`2E&h=sK9ceSoz)@z+M}Ve)YeHFubw+u+UO$Ig6;*|!8$e$|vCyq7a-+^LTKy>|8> zeN_&(prtjOU(IS)(mF!zUY%=<)EAD*KuIF|D2lEXvU^e^IgVd}(vkcNksWtMw?7km z82qs<-{8rTNzNajBZ2^c8?n#Gurl)o=AY~$^o-_Hdo-ga6VvvX5-%KB68v0ihXJZi zdC5qU{3}3)JkxvH7~6N>8O`}zp%FqfHQN>1qi=T-v~okDdTfGY+?I6?`2=sfkrgIx z$lFAvpBVb(Y@`m+XraBPuqFO{kB0%y}>0{x`Dc!_A7Xr0;D=gl!-pThHwY^NcL>OMz#3{VWXLLPd z?o+z*gc)~~T@u_GVOnsrJXs}iRKOs7JZVFXW5GRVXiT%&+OrnTdK8D~$C7=%*4e29 z&m-oXL#+hL{n?K{`e9$tqbSc)gwk8as*|= zzGPizDM^xch$=VbD-nlV-YuB2XJHekS;wD8B;CDV?SF|#y6MG^tFrIzv#bMI7*R+$ zwD~ZiT1baSDxGDFGs06(%z&AULPXNpJiAWF$Pl3yPCVAb0@BY&_-HY>!;C{XZ*Q{$ zN3mMpv$Tijysys4UdrlaY;jNo<4`s)(hJT`*m|D(DLo;D<+c$!!bY)lk_RRjz$YZ~ zbtP8#iEH5w-N^ndg$*UYovnD!P(q z-=DNQ!Z6J5j82)0u;n=FD06%{p3w@!jmM0Orf-_zM|d!JJK;)=p47~>A>lNw>7GHN z=841y-hgI;D%Bb$60ts%#{zWkT*`6zU(}n8Q$SVs-&W%8&q$yziP1**j6B9P$ntW( z8N7+wv_!(sAA?^E<=&=F-zkxnsMDRSmr>;qw3FU59NLJ-2m(5RyXhfp-??3So>Yj% z8gkBfa*Fv?Q^Rkm9qE6~j1iZ%(DrJX?}G1q)4X2DRHgx%S;WY9zjpHga|QkKq~IHM zY0pT!)UXikHxwa)%lqK16}N09@<~+3#mJ4QaLZ6$jg8Ikjikq{c5;O-f~Kh!1-sl- z7`t4Jt<6P3KTs^L#MNoLON^9Crnm1~ zJDz7GL6%rXhhg@|JQW`+Npt?r-4R~dRqpiwK^o1~WZm@hW+KLmstO~33OrB=qM!sa zq=;TA`Sbhbpvw55O9x23B0^o_j3H9JfT&0zvgP{_CUei|fJxj(0jh^p_9#~>VpgL( z-U6opQPsX_RST}^y(!V;5$U{a{cJqpvdM$y0@W+jEWfqQ95879E21i<9 z5|P>ZE2oLUf~ycYdSFhos;ub>39CkjG*~~A6|KEvrCO|{k6_hmjdI-vGUp3 z6olPYq3$^D0zVJh4Iq#FW*#Ew40$}H$nz=g1>vvhwMTj!R zO)+pXLy`{uDXSk$;|V6?8w)GTdl^Aa2o8}SB*vHbXA*=N${-0_)=;uyH*%Eu5?_t; z3$7G25o#DcC$7dJ-NX}a<;4wqb|Y2!iEbiC*L$_z`9gUD!)Xs;91J z(${#WLMR+4GlEV);h^L#x37|>f3OsFP@ja@`{)t8qJ?!jc=m^Jg&F5#%EjberulPY zD%K#>dh=ISjgXeKu-HR^R{l-FlNHIZAu3t?xOBtRJx>yT@a>BAG~Q zLTIP7H`5gb+6ohK?8SzhX9l7iw${G8umXbb^^C+ad41t&(;< zSdYdG$Bd#`Ld<&g+QciBVx~!k-}=$d7^6vlUo&8{`=}~2rlFU;Tymy=#FM*jE>!3l$!#C} zVhCL>U$Nbm_%rujr4N4~rIOWfFCIQOkWXm`}z<+`Ev{CLRON-L)K1<>d4v>)2U)Av% zqC&N{QT7E6I0_$u0B^M9XSy@L&9s}xvCaiIU?r4o#SF&|s9*r!)R z4Pp7|u${~H2nKaZWrD%_-p$5`KSDq)Q0ZH1fuHl;RL4_H>;)A{cZ5&{H%uPTq{!bl z=HqsZvzE4@_=LMl>6W?dq*~kOsY0O!ZcF|9(9~qrqw(1W9}7Oia25p_D`*Cx!UqIu zLX^T5B@mKLE(as`cS5G0j=+e$c1L{4b{>bZl?adXvKEa88xs0m&wu~MC zQRDfYG7y4KXGd&kB3==uc9e>F2_n^!^&61fEfHj66 z%Fek0drJdoI@Ld-pwc~oL-sHUMd`+#yLJVjD39#qx~Hb9`qc(PFJMKx#ZD`Q{Z-|A zZxa=^5=I-VeKLsfx(0fmXqUM7;Wu&5illm^B(u8^ME(G5l};D6)!(16(D^F^OS$Y- z{fM4tqgb2$`W0HJrEKswMh3aQtMdP>i=ypFNNtVn=rheZ? z;QJ4COBj?H3l_LQCEM@yy;EvlLHP!{AdtiCBXL8nwvZ*F_jR!p>U0F68uQZUcYJ5@yVaswLZ5%k-{;kf)gmRPsw zao7c>STa6gv63y=#^JWNfJld^HBM1dXbr}LD?mAQI0){a!1zTh>60x`$;RnI9t_>> zcnXUMwTrv<1@Y*YN2YQQg&tN#b!NQm$eLDX06!x+jnIa3dnb6ku@I@?(sAH#%~e-y zABR=VPX>=<8EU}WTpeWj4Oz=|L0VLs9SZd;+36P!jOO1T2nW)f9ptZaE%FnH$|ev;#QFP3SzE{ROqb} z(TTHvI$DgN47iHGiMR!+i(}UiBmb9SEHYD8j;6&-F6#aDF<)U+S6m+egrPYZR#xKL zNhrPPTfOmp*Euvpi&G0bNw#5G{D-_z4^u1jB8teq@s68sF}u3oKoU}X@q2H?KqRWk z`ShE8Ho1}ZOU;QX{SmSXe$$eU55O3G00}#PfM@>G}wZY^; zw@|S=!A%~xLL`?BoQqh5GZHLcf>_s8R!)dxsrO|*&?d;L zU17zjNzx~)H?x;Nb~2|u{s>k#KapMT!q>Zs_kw5#&V;BM^`oi>6rK5NcamsABi3Sd z&3=IA%surPY2QFSnsDmVj^bgziN{{rz_K(0j7AI0^2AFKA#;MoPx#S#`{-(>zYSQu zgena5nUIDbNbk^zLK}wsRf|R?*Y{hrwFG)zjq&N>M#5B$f}5*@M}H0DV%L8oRfzNw zz?ah?O|oR=xdsSHVYWHY=oT^dtflQUZfM&X1w?t~;zfVAT==L`O$MPDv`JO1el37{ zK7vNh4fiHBP|?Cd-5R5U@4-;Lv(qbf5Q}StLDp!$*w*aQEPrQ}h9dB&>B2Ko4*FD= zd@7`ONOH{iKF(F4O_{{1QnYs?;-xvW#=urYnRGMf0oF%ua#Q4L*xGp#AIZZm(-Rgh zDyT$`BX?Rx@yyjgl1AT;mPv{Z!8kw5uuP$OOnC!uEUwNF@-dLl2U40R|^ie3k zjZ`tcHkgXtoa3lw!q%r{NTXZ3`XoO6Tg)?(guWtA!`!|t)6M0Ft&G&4@OR{G6vzs! zQoXDOtn-S)okiso!`kIjHJvM92#e?o(u*~aCeTVz()MOx>yn| zQnhk6hGq0^Kr$R*(>sR4LNB39270WgH=r?KR<(dIUibA(0&c)JlJ-OYACb52M|W?_ zNBI@{i#K7$kwfVittzT7hs|@*Ko#KI8Jm-8^_m@}_y%>-Eqj zDm%a(M*-)P9GBi}PF`w?BB@dgg~q+0=1qkcLrIljZ%-Z94i<)dc7i=$PGEm;Nw@E zQQA9MetE_Z`q*R#tk-eT8r`??O%tEH93$RVX)uEcWjU1~LGjDcClWGPQrgdi3=l7= zrPFOuJgHH30iL=en+cAm0PA3YWNRXH1(J^0`~C`BhNmyLKzrnQ6d)eG)7Y%egvG`? zdo3c-g30&;E+>h(Ho)B=*(1w_b*F{Cvu*N)r{9u{S-;;d6bjoLaE~+ryoc#h-S8}v zOfp|raJ89+La5wh*W`7BVeut4)RBKvgrG}41j(9R-1G>cB}Q!OXt!d_M3>IT!0M9| zmh^2bi|M-YqzZb*8#}=JQO}NmREC*IGJ)i!*vf^dZ+&N-++3E~Wiz#$+D@Gw&q%v- zbCTd@Rirvs*dqNK?LPaXr-+1Ot18+fai|NYo`C_rD(0qMfit}fdwLr{k!d18xBBC@ z5ll)SO;_#!t|lZ12@Nc@f*g6H8V(_OBGr%xH$fBpcr zC`w3^eCgrk#^y`pk^OnfjUIQCQ$|(K*AMa*c(@>I-0YkLAdPE^Cr-W-$>6t&vMGF^ zS76{Bmd7y?-hHz}{|jx*sdk~gRmIvRjONBFf(NF7*Z{LJ@WA-b6T5A#lsyddM>pA< zWfr=2IiQr)p1;nYqFB#(6cZElcCFlZ<$CPjE8C|VZ79_H=&xVUX)XENa57@)YJ(7Z20vM)<9{chI z0jV^TGYoFEIJG&{tlAz%p*NVmi!JIUm%JY_RwSua-ja%>N3r2zW2gnyoz;&k?4sad zu$^h*3lD)5*8+#9T$3T=FA+H!t4lmxBi6AE3*R#}mD6*=0|N|0!7)? z&M%0%+QZ$NAhA8&4cXCHt)7LqSV|OLBa~|nJc>d0@1ekac2F&rpj(G1QZM8y-H;MS z-C3_TAa=hh{_DfW(G&^Uc}j`9!qlP&UK@=XjOsi-Ea6I(iN_}8^P}7*8}%Zhnn+=X zmO57X? z-j2Snad%*F!oWg`xL|#~%0D-+A(c|4QsR4nALB1kyCa3`s5S26>cZnxHQfrf1egoX z>JI~A1g{2gIv-#q#WIWQxu?o8Iw-ynrB=7j8e!zv~XrR1QU5{Vk|bz z$ZZeMx^_zb2+~@KEKPT&Y0JGSF@I&>p{PwFIn!@Uc7k5DHioq z_{Q97lW%%&xuGQCXgc|41#sxA2A7XW`iw{1a$eqh-M)=jf4lv|li;{cNin3eJIVI= zuB=(;3{kAT=T)d(5N$N|nrBL2?_4Dl$=>U@`64LryK9e{B?gb+nJ%br7*awJKNii5V& z-daqb>|HSMyR^1jK&(C#nF`yuf{?e)5h*9dIH;;%yd;=kFwl@#C^4=YM%XT zS(TmlJGiVDV4*Yi=AnFyw8SeBHe_3gSN)%7>f*E)RC&SRv<h9$AF1Jv)qWs>LlKZmfd?m-dimW8sAUFjW=otE>cA7; z!Ah%4%n3akowsg=(#jVk>)Ys=#(J3tq-zgRQBevWI#!VPdR|@In5ZL;HPBM{oQEuw zHCvRr9_ebp?CnIJOKK;r-%HoyYDX#XA^(l@aOronjToNCt+qjH#b>laUrDr6<0rSIqYu|>h48x>fen_0!b`a8=kRWrG*;;Qp-h^@9l+6O2tW<#11*{sNF zj~uj=xTZiE@|r!85Jshw1k;rK#58oGJi$ij?G83%Sg=9CP7;Q+q1KkR;Qm9~PF z|3I)tMsTv|t^C@T2_z3LX10tp#ERgOQ1o&PEqD(RPPcVkyy#bez`$t@<}bE11lSYZ zVy40e+Jg3zpD>rzX0d^y^kr}shN0$>CjN>Y;$@>Ycri8!*mai5Iu=^`Fd!paVUlfIDNFs*R5-zs!nnX1ut@3nQ zUvsUh>*g7~=b|^C$%2tuV}E}wEerNHtoa?6f1li_K=WyZ%jx6wSSSHj!@@(p?clVRQ;&$dp-u&$#!eGBZNaMj z$#=w#C&?SCO078=3aPo_Y~Q{MBc~RlOX38FW+3!!vdjq-h>oms5E(lA!wNcgAdWQG zi?u9&bCTR-(M;)=kk?KTWQ0&(SpDbvOniOR(&)Wm({`LrziXwQn%CER>UnhzryMo1 zXTIyLS$-3ximc(03KY;6+;Q&z=snss&Yh~cF}CF4fJBrxl43SeQtXLgzZK_5Qa;oR zs*}K(oLGfNx$kVK2dOdY_PWa<;*+5mwiVL)f$#eF)@zbi=!|Zc5)_<2VH| z@*=NvlAaPcrn*KR$mkYBv@4pOmMSez=Va$d(QwBG5ndC-NbD_pd-GW{t*}WSSHmdu z`G-H;M?rOxf$)Y+{-z|aGHCt?cqjv`0iK`iKztF;`^8_8x= zYvbL1FMjNikrx5g3JJ!8fvx%#!9S&G@?BOI^_IjKfEbLeQji5UE8RCu1Jh3%p=Kwk zv&Bz?d>hUi7-HvJYYUx9wC!vp3k@b*`;&LA>tB_26(hAfl1Y)NlYA$MZ+YqE2f9A7 zrjYygkC2h?aPI#5BsG{!>?{27!U^i3o>7r;`sYkOQ#$#xVB(swH^FHJot8ve6U_=+ zrKxC6Y~{;SurZHji2C^r_}#Xd^3SrMXwdJ|Leet(F6fa(u$7xGo*)1z-iPt98m~V| zN1JI_A9F&x3vR1O*;(nc(2AZWzc~pt{)|LM9bX~>S}5)w%AsyL{r&Qe#fAcV}@2P-2nPO06!A2fw^!R)}-%a!#1^=&g#Hf9y-Q^g|K32on zCn2KBxHxjepni)N#piHqb$;8=KT}|iV-=;TsUPrB@*EzII?*AT#7tvt>UJ>&Zi#62 z(a`&psoe&_gmBJdW|leQDz-%2d@`?`lKB{0*;XAZYg&@|i&=$@i( zZO#L7KQnw<0y<2HQxG?`Jyo-%lr?to)XkC-UgkF`ty-2J1Wg;lpK58AUp zgs1)>{LsP?Na4@kIAm@(sKDu>3T{aoQaI7%(pM%d+PD4mHb6F%`w75-rEeIp1M+s(?qv15Dw0fN! zKzz_mJu7;Wyh!6Gu6U`fh?i(J3b>i5;kI9wOaJG&0I`cQcH0~Ldq=ZkTN^)$jo8%Y zgUIi%2ff1HnI5z@CXIxud(Zr6fBX?n#mGt}6(KjM)^-y(#Q|W|$VCzcx|B%7-ke-w zk{q-OBNLmrMxXUb23&q40B^}pvVyuIEUw*twCU8`P2`?DB0Va-~sPnSFceKau zH2*#^olft_%`xzyHd^c*;8!zFh%*xVTu4(WsePC@L1C|~$80%zHLQzsvcNpxp}lR( z%)f=J`FH5wCxDqBsdLe*doZd-L^Oqk?S@lbRa+=(z(9P>V%K2fNc^7kmj}v}O0}TX zwVB+)3^ZlE*=B<_w7R}u&u~s0vZlCJ8)8-T8~GSD{P`{2(=pGfp0`iDC04f1D#lNW zyq{BsQVSJr2|6xse%4Nv=;i4hI%D~v;;y--8W~S(jU|U;F{?YU0ZbYL`3YyJTzcK> z2a(ix0&mIKv@*VE6o>1tWO&$p)LRYZjj<5R(|B;)5X0BNMh>ViDOqFM2iSCYbT{1@ zso< z%rhYY;aU9eNCN=pB%=V)?#_ihKZ}}UyVGyn6+!G2DM-JdtyAai<;Y!fa5t?X`DRg2 zlUgrBe(0bHc6s{a;KZu&)A{rK=a(i)11RLA@}-nQBeo2qigwMDlo$pPR-NLrw+dTv zJ&Wr-X`-ynD53R&l*ArTL%6e>Op6~iQ7~8wLjN#9(iJT3UvMDGJTkY zvtpsKeneXea0~><2t^uXXeU!M_TDo{9$^~(=F|yUb-BF}?Mi}P1<0ibTz#N)`E?Zs z^latnjB^an;ev`$Ik9$eA!DSGtNVO05_?eY&=!x&EUq{}&8NQefq)B2tpU1B8nYV6 z^q+=?ee2Sei_qhsegV$YJFpO=Pa&9(wCRtfCs`6bIbp>1Fh5K$Mooz;_Vn0loI0^oG10N#yTWgj9nFb zC{x5uG~GYY7lZ4}0tmoJ&e{Q>5Uri!pjq4p@rWFnnqZ{H*8`mzl5BExO3@m{W=+SV z0cOkkTUZ!IX1nb>S6!Qz2Gv|IaPW|f}KmF99PjeyISX!dYcO`yNfCM(6{4nJ$t#3 zq-K(4f7*U?Zw8IN@YHEY#;dAM5&)w-la& z$+bQq<**wEYCv9RO%Jt2S4@R^*-P%TFr;OF+f9QW6IgU76vl+1PezV|^och&?bSb~ zRX_k-XMuP$L9#fJo*Io0Hb{@r+;v4Eq%(VH(qLlYIsq)vlN^&i%WqJH$>MCP7x5H$ zm16Kabft(&$-7GuDcart(v26r`FZLk1xY+MMCMtXUvr%U>|T=qzV$Adf(F($=jvkL zbxh~>r@mo3bJ&bHDdOUaZY6>=ic4_61?xlU6f{8+w)3sejFR^05m_)7!3_QjIU4&) zT@ll^g@2!fr;OwR`K>)kBQ3Nbt$jf(3e6>pjdF7*+?C+C!chubzU$MLQB39p@RR+- zc=i)aq+!vb8k1w2>}6eIxlX)%(P(W4ApD@9<=oNB=G_Q8q}=pNW9Uukj5k zoFP?i-Vp^G2i0wD_;dxCzgtxcS>4^%L}b^aWrMfAI!x6*CdCni_egzyEbt=+WPM)l zY7&w#kbsVApDPVVLOj(Td`r|HUk@Auq8gJ+P1PlkFHkg-S8sgZ7P zqf>ltkR|JPB89cxqkP*FEo5Xi99>A7T!*49JyYZ)cpEt3%$o`rOC3XyW5+w+n{5+u zjeQ_qW;KZGnklNR)>Ss_AF_Y_>J-jKUZNi*P0raSeC>TLEXmWi{!9xG$ak%ry*HON zPEU0b5mz6MHkMZ-?>|$wMR`ktAhlB<|BAKvtN@cw^<~IEK{GRwv&q2+p9`42DC^X^!f33aS0d9WJ3-2H?j32 z<0HNxUidVZwP3#YhC(iKSxN=%+jI)Gzh8g=FWK3NW> z8^GVKG7x%onyTe)4kP!aoK|%qkhH=SRq&~!>(`s&wKCd=8TjaRX2me z$j`6SCRvr#(9;6_2zrEPBJE6StuzuEj8xqr+8l$`2s^`FXb)dCaj}fc{MhL7NB;m9 zWS@`w+D59SLXzR-+!hM%-z62b)Bae0P*&47tDPhTCcXH;DSDb%G|X~X;)7{k0ianm zou->;S@9ITBxC9Ni>eYEM7t5ZwaB(JrS|RP``q;unRV8${%8IlR52#Q+HN_rdc&sn zsDxoiXZPPuj}o7oQt+*|3k3Wk{M`0pZ>UpK6^+rv{$DIzUZ z*!cf_#L+IS%l?CPJXaA(NnEJwGUD1GIeB#k6&HC9@OI1}URhdat6LwwR5$vRYMy(h zVmZ7PN7<}tGxa<%{Ig`~#RjzD2jb2j%Qulb6cPVZx!?Z->nhXowDj1b}f(f zqz3cvBk7@tziw&Aj@!{%FZ1iS>tC#WQ1pOY+eoith5+;}Vp3xNgu)E#DmZ?BedH;m zu^=jQmHz4fj1c%wDX%y4Ehw&J%u6%Jp`5f?kzR728scxXynbP|3 zp`dh0qWyUq&+uo=dmyhyL(Us$PC|uUX7xZ@vkKp{9y0Vf?dxH(y2Qq*ZO8T-wknNi zbTK{|pMNl7L0q7u0gY_hDEui(wW#`yg+9*3rFjK?d=KZ7V{1@!ppfcjW==~N&dq% zx%VhQ#38H0HFo?_C}zS&2ZiH?Gi;&uZTzX!MjsD4Qc_pQ{~HY3d1cp9>u2LDb8)B5 zv|p2dFeQvvxa~Lm&u^bZtE_Rl$^SR=f-ra)lH3zSz7FVQ5j3T95Tq{g|Dg3}INwT3 z%#3k8Me#>dqW{SzcI>O6##OcVgsNlFG466?VYzl&>+_Ep>Wdl&6*<}t`lPYv1 zR)kEP&NcD2zL5U6`-y64REo9$CmGN_Rd@8;x+@(JpEyOJRg8R-r01840-Vmc1xirU ztCk%Fj`x&u-hWFLN+(iEQ)y!|oUfzx!c&ij3Hly16d|^Emf0GhIBL)AX;QG0FhL<8 zx}#*(*jg^88sC;BHrSs5)vVno!kwtOb;X^)AB=Q}0%)#+B@4aWo)cAY^H)`RIcCxV z{}E-Z8n;IfPZh+%m*ZD2RJB<*J|#K8*n;YQlC{xjB@dhLZCO7O1kQ;T+5E2m3sVSS zkMDS4-V_u8`{fCC^&&-9(y;_E>GKYS?a1J4l7NnXej!54>)E*WavU3@oe1a6nt+GV z?jbF{&%RAarNN)7_5l=kl5Wj=a~L&PB2(yI=(8Co_u;%yu=88m{$Z#CL|N;^n*;(U zjYSX1{liF)N^p*}9T1!wv%MMQY+Xr;2(y#T*-J?mtR~hH1gp!Uis#yOTu)+X{lQ>( zC(XtFOL@1Qavmw8e8x1$S#Vz*=g^zj{mPoAE~spFI;YEhwXKUKhKLZ4vG$C8#rSeI z?2=Ki@RBi}WfwcVM0lgsWPmA+Kx*IRPtiE5g&7tM-^ylw%;i8s+T>{Ory5@7B?d|O zf`;9e2W6aF7Asy5Ely48Uiu+)Pj-d>){RpM_)MJ!&Eo945LStDBrcU!7ZKgi``@H=8$HIz;oN3vsQa0S1lned2w_ zNjg%bxA zx4dV64yW)tXKs|81PPacW1!=x)$ORrE)y||gI-7i3;vi{n|T9UsajiXZ-9!Dhsj$=gK$HL{t}0LqmgR0!+gT~%+wN}>9Zjh zC&e)_qUV~(Y-xf;C*^%DigH=CJN)$NnbP|t?9OLB^A*E~B2l^Q-~N&(ymNa`Fqf{6 zb^xoY=U!BSFTgz4FQJ=t?B9t;bl$Oxg->n`@A+sG_?Ihfnqf7Am$JEN?D51q?|_c@KXOci3}V;U_efyH#+G=nw)*u35x6azZz+P z(n(n_irj7u;_gr`#~iVlEl`)_C8U>gT@y2=s8ZL)#PcfN=hF}ROs0RnekWaft2N`; z8a+XfkTSz39qlL$2^6SkTHDo}^W;Yyg6w9FC5WzG;5GNr^Nh5BS(p`C!Soo&-8g@G zcjd5>#;v-H8@aH5elWL|kB>WDlNVX-be4E?k`%PJ=GdnvZ4Spg?I4}{7Su0*a~<^c zw;X{gEy{?$+Ybb{wzw}m9AshfBbI{x#riEB;i9)}_EYH{_qLwWgD~n(!MdK2nnZ6T zB3w=Fp!*#bbt?IqGtUxIjr5QJ7o-ND(2rXco`aB4lBH*9t~mUWMKFXBam+LyHosGA z$1b$9YV|K~jUCQR+WT=t9H~5wFm;iq!El{>Ho`B&u=Ch^2ZAGJQPP|=mFI5z#vkGt z@Dg0zLLh*p{`RVsiZf3pT?1WM_F35nhBwG+?0-l{JaYY3Zl5y7bc*?dnBFm(p#a*4 zS`7hj4keVoMdBT^d$!FKZ#(p(J4lEKS7tVA`b4ET9oQlT`65^)*&M%(BnR*eRC*+< zx*1zzrI!lYtwroeA6M=-qKRDv=>ASAF}lkHC0S&w6C!UDFxjZU%ymgpn7 zsG;zGh#oR-vQnHSs2A?!ISII)6H1FMI@hiCK@M(gM!pE-G+>(I?PQBuDsubp zS-KJ^$ml#=(M;}A_G4Y3BYX+&8*O0}t3ls@0R(I(&v2>cY&pLG$jWlPzKh&9M1zv^ zd$!qMk@r#oq73W&oc$Sp1PaR4@cvAXjCi&|qYfjEL7mK;tuWNh*?Q}TcPmox_iM4m z#_vf6DLLVpbN9}E?kACCdTD-w44u;mEzYoOiH=_BMqniv7V`4>d7oLe&NQnTcWBO= zDR8EIX}5Hct(S+XLet#OwH5u?4Pkudsa@6IdMiTasK4)RnyO|sYL_4|HGrKU= zXsZEes(E(OvcJGB`BWA4JS&eP4k@C}svRqKOkouDlBn|=9G3cAw|H*}W0olb8zb%8 zl50s7pA*mr;kwLDJ~|87vth$&#E#vTC#4|N)S6`fIj-jdRaAebQ%;LMa91T3m;-Nk zoH|fq*(?UkY!(ZXO;M`jb>QL67p@5cm-`7fM1tI0E~tH++>Nq-96CRFtA#C6;2S&_ zADrMx#6nv7A;;MByTsmzu0(Q^(>yZjyZGrG9Z!S(hO%UR<53Z^C4vL&-@%K4WMYE8 zKIf_5GlYnx=IqmA0C6C=mfO~aZ)Yk_3W}Bcshn7GYPOh;- zhXcQ*E@OGcU#F7Q&n6`@<$T5_B3Qg1PZhEKJc0Wz3bdfxpqug`^g zc|Y@xk7qJf3EL(QD`?B-;U9wBTkgryRSn;;8cbK#KXH zI@~#H;NACFdX>=5A8J!S_nD1X0IAn6`U`XES=|{rpf7R~9EATMdL#X)K1Ka0p$myH z6>`H`{c#J(<$+Oo_(J8_EF`OdZb8=W7o>6cc5C~`r)-iADVCWzUb_mdX7_mp?9tp3 zf(+#4yU>T7McZGvdhH%Nf*;HpoW3iI%Th{ornr~}vK=tJxiuUC)@S-$r!lh%JZ(t} z7Nd(js}&$Oc#?b#OFaV`rEUgpSiLJdMbyg4i=VD_?>*BjREbuSja*%zGmCU4Nhj%B z?18TfCX7*>J32~y{T`hTrw?oMe3F|ti~>P*S4465M)B^FUz^jZlN7K8Lbwt0c2|zw z%8?HDd((DbrP%;8-znPqf>rA&tWCc`*gT43$>5QQP&zy6_jVt!&`_Bw$*OE*&UqNX zcHhO%R(b+%ow~k@}J^O*vl=zI&T&NEtu;K07979QJhU`{OCEkmlz{gUkUj@=$QKHpkXnn?Z|VyqeSM(Rrv+B6`Xo&Td zMr9wxpS?mL97#0Et_nojG*vos_KUuV1ow2j zEw0=Qp_yuCW|J_>^X|jON59ylPjD3$$(xp(tKfr<2{6CJ(7Uk1)u zB^_=~v(Sh%pv6xWo1e90RKLJlaLHP31wLpFDxtenOI!y@PIDk$j%-ECakCcK7v#4X z($BQm@1wv}w4_G=wr21k-z`FdK_dY-*Rv)LK!%VqWhU+C5Xj>-no#po;gZg$PTB1d z+v%1VQpI(H`B7VW^%diJ<8>G;NtNyq8XY)=AlE9j)%l9fS5((%C14 zT4{M?E!x7s&A&AL=K_?ccbD_2j7kC+@t@OFS@0BLN)UEAOnS+X0&9<^d)bFhv`~9UC`QCPJW-6Jnj&;GCt2Xq@pM5VHvQ`&OiJZ7f zEbVMEfTzxsw{EceUvYE>-jYa8brD9lgP~>4y~SfO*IqY+ArO`myDM4{+y5j z9R=TyGq^-z`k>R^*5EVM+}dT40J>FJGcyyhuX(r*wVPMTBe^AvE1Tz*Zs%Jx&W zdKT?%f!Fa<|A%L3wp0cRb4Ntr1}0KBV;@5ykyb5RDH7Aep@(G6ld-tqjtWh+EP4w4 z)U$kLDG+IW-^RB1_Abe{7EEn3c_eC(;=48VfxdR5CWWymr`{5vw3l4(6jAu1Fuszg z>9w@hpr2Qygzy)y!oSo+T8m9rzbz-fn9MtCY8srRrM3zbEt9=bD%op+a}x;K9C5~{ z5wFo8wOn#rdVyA(E26>mrP+jpG^pQ?N&T$E*rPY@pS|ug3TDSxHlL%6oD}>4;2K*O znhY^^vJXm?NF~u`A0E_+*we)$IBuiIv{7daqXdnQ!tzx5Q=8||o|4UqH-SbVs^U*P zT~$ivgQw-oM7;}c)Om%id|abBzhIpi)bK%s?5e;UzxA*LhP3;{@ArxR{D7Zx=eJ^* zBuDh-+vzV_BIR`E6c`U*oCi!NY-U4^dS0rL_;X!ouR4{g(j=FfqOa%z2#6W(59`*_&CuqhcuXLj6<~ z(RU`vPBd7O94C+qJ8gs{I`rgP(DCfeU`-}5V_zf2>@{7ojyWbxS=yB4Wt!>&a>4h& zr)1;wAyp)Z*Y&bmvat#k%9FRb;k~Rq*0Xy?5Zfhcm8SakFmu}mjyVr>X=%Q7xk~h4 zu9&SlT{?=Ff+PLX8pUBDaVEVteGea-5x-$FZk?VHyl`@A1}P|{;m_aJQ@eBJMR>sy zw$z`;&Hnjj#OzsMR$AX6;Z5C`AD*&}`9G?THG;*f@GgcrBcBFpMvY`&C#REi)CkMY zZ3i2msqs+4Il@*ei3rXIjoEfH=Eq9u()-#Gt1|{|=M)HPhq)%HNx{eF4r9_wA!Bn3 zrZf^BZyhvRav*_xhZb>~3fyHa49pSD-z4WwT-Z!CIcxkaFo}(AmbUjR-*>bUVW`AG zbq%8tr<3hTII;BJiQw7HhC{xW8F*A|8B1b}Sd>kMgbY0%%8qoai7T{?IOy!8I(&!( z(X*7z%LP)FddIn=^fjjHzL3L(q?zsiC|}Et>ui*fRz}Y??LOf30wZ(SZW;F0Lpwcu zwM=G`I2R4AUh}sh_?HED#2zEGqciyDHh9TN*fYMvfFQ7RGn~+FEs5BEo24WpfAgsv zsSnaLaw~y2@{G|pRI{dJ*@Z-vDf0+P!G*?P zz2wSYUxm&+?nf;Y7I1)34}k>So2uA5WaSfAd;n%=AT^_mOxHp1`Xg} zk|ZTjU0za`)!6My!t-YcJVVR|3_Dm$@LWl@@D0!4I2YSf?w#CIl1!;Bx(-4F4;j-; zWG!v7>}=+QEbCNLHnk}0vpYF#h7(tW)+t#loXrNeou+?;m@*Xw@EA^Jv6*Oj9zuZj zpN-3XiAccPT9&G5+c8Uh#)@v-+d2BFp-C|-hAVGYjCZk3gh#n%N(JG9_|6$T0Lu5$ zh5V<8dT60}YyDzi;_VcVGbYZ96m2?NFxpAWw>(N54WqfCCP#Con4|+qu%9W=>#?&g zx)Hc5hoeIghqebH$vtyJQ_pCb9a=|@FY%B@VhY2OJX56NZw<;Z!jhes)t%^ch|Y3+`$Q{AX!Azu~j z>w9sgM(m@W^%+G$sQnoeYDMDI-jizzo0KY~r*fGrU3!EKfms%8M@b`?2xI}H4+^hl zqa>vZ3FGp*ARHzX+*(4@fg5+U(- zW(#|_CQ-LIU6Gz&jl~pUERHsaOgEUFiry=&1I%=Tx>w<(SpRAohNL3F9g)GnxV=G##1qIC z$}!uCiq0o7BJSQ9nvV`W<_^`9YP$Hv59Rf((9+nKuvgat1Uj{r6eZk3SE$%Fj3!Vk zUR{5bQH?1>eW70>|+U5>`qBl8Y zo+$__br9h|QMNr<@3jn7r^jo>BOBi8O473y?Yq6cV%XEKJ_RIX=y3cCr>RHj7*YjwdIOPSy>&9<=xRIY}IuDr}nx9QopszCP3NUvjzXL zVf5nI^kkJt10%K}BVT38&m`v7E`v?wK3D3#5Q4N|s&T8_7$T$qIo;3KmjbLyFND4b zDwy)R_fQb2;9Fh{xT#>(CthRZ2-b+nhl08bWs141^`fse;ed5Bqf3o5JNKU+5C_uc zAmwTGQtwns15OVTKmIr3tBOqew!5it6B6WBL_KJW3O4vmH))Mfi@Ra`<|jvzJr zrZ`^UjK$1SLr&N*yiZr}Oea%#UI1bgi6mSC_Wy7$O!mvR#8b8(Bux9splGnnuMMBCLyVi&?NB)>>wKEIBf>7n!m+d6C>O!GdM7QN+gi+#F4Pt- z_ebEG!{^fRLW1_MiLmO~g&V!#@xxx`EkhtrQtvuQdML}URWb#`!QpZ*eo1A6Re@w1 z7OINp&V$dSpBF^#)6!)3r4F71m+UM7x3f2(DP^sVIft?2BmuWM_(hh$FuROfTT% zsG@TyiPne?r3EC^Rl@CgnCOo_E80$`xb&40 zS&|j8zfj%;{pVyOlo33oUV=sOwv@iP!Uz9#I0dDabIwClbI1t=_J zx}ph7ew?^Fst-W7(4KiI^^L;j#>2;j&n%+5XB!b59n^UzJ*YiYFSS#eE?O zGn@6h%F$!9q_?mwQ4*gw%}-gdX~)tM7F$Xg$L9VJOlHu}F*8!0-EcN_w7DOE!)H}B z8&Ccba8+BSu|LJXGz!N-clXwobq_+u7S4-35@uQ#oJ^(&Ul;c!a5_!@oJ=CNNO~Uh zyM(s%S(%#bEhIferRHanzFhn8->ODdhq(M2p#U7rUza3PIfMo);MU9q23!-6MWV{(*r&6 zR{eJ`9vZ7NsrEA|;k_x=>}p0a*zB>vBo(@4jy4HU8WZ8>Y$zP(ikB+kuxj4$*A8R|uVZm#U9`@LJdWaB2g-#08GxH~ zwi*EY3zJJ=tCwqGzmS-DlbW*Vx!ERQEzjPf#)>ZqC9|b9cYjKQ7hd#2+r<<%`Hw!A zGtvv|&A(5;+HKE}ngEe`_Qhx_n{OBj^1TXTgL<~Q3^?U7LYr4Yukx} zowJ6@H`hMZ^fKo-n*xy6!Y|#7W)QO{MJ_c@T=)x;wn%%%fjbf$SAue33NpPruj==S`7q zrh`X#@l1rvl=I(4u=>rOgX#2I6@GHIl`(_l`T6NkhrkH(MK~Gw)d1ecq`7m&@5XI) zS)cIX@=LtSnVC%gnGNmS;$G?(8l=&9ZSS6=#;c>6NMB`*54l#x#FWv!9FF7!H4$<|OF6$0UmmI$j zm%7HgFMS?|-!Z?Z6MH;F41`A^uN$NW?=xEO_=#FEIPfc4{! zIWGomeZB33Y2aUmtmDkPFxy=vH9Gvqp0ku42A`kFCDy``Ro$5SIsAwI%m$P0d)C&% zVA)7==k)vEC&yWYiD`n{eXK*>nG<#>CGw$jn<(2;GXD% zIo%)j8&Pkig*Her+k({kqvgUzh$3tlYU_%P?quH~FeML5b9b_;qmrR^`8W*lw$&WX z8%a3_X@)Zb>uz?%!;KK6q1g8Re}eN=TJw1vXtA?3zbAJC$-UJni$G~2Q+*!6|H`{OugfaSHx>cD~3Hm&UIJB zQF|W-!vAvdQpG1k*{lhZw!pEd?a8{;t)kDJn?4^~2dZ}WxT;@>Qf9Xk>KP}_7(2Yj zC$Ty-2#Rr6lgj?0{OE(Zr%zu(er3n|7*1K_g}J*A1L?4YhkZ%$32;?3?(sA23Qo38 z;cBrbOL$f(wvXl+IzxBC#Dnbh3N3*l0*g2mCJ~ib(f;<>T-m~ovMMfiMg`g+!yq!7 z`35=UKKv~XIRl41J# zP!x4+#KLy-D&u1MVrnD))dg&YQoA>p3UY0GFqV&&) z4mVaxV~j>CG7=lJJQxg|1dqjzV#1E~!mZq{>D#H5l(%l~8^V9CVVhWXU9wOsNicTG zlpQe!C#-01cuhIP=OzX0S;Wjez#hdwJ*lS~<7jg~KC0-c@Tmc!Tl~SZ1g*`VL{XOB zZeMQco%}KLbty_kZI6)m_gsPmJ$F%Q^}F!}Ip;U%-<8~r4h0XwdTxgTPzNa`rNsw)g7(DK$eRb`iXV{Ve*YM=u~c))x5neL^RHEO1&jC8XZXzjv!9?AZtvxrj2Qi9aNUf?ZmbI#z+$cFHYXXJ2WFoxMd!smVWrb}jU&64a)#Jj= z@=28G?2pHC=IhE&)r^8!My#dF3-pxKxVlb!bxPPF9Ff;=plDq-`z4$wb}ZxqanTVRTxAlh;X57pi0oiIQ>W;m*7rdo zlKIUrEcw7RAmiV=A%zRstgG_VHqpK6+c)?F9ZoE+*yaOE@@8wgxPL?`(?TQveWIz2 zG048f60}i8J7auIou3QFmQ)-beHQtwv$Z|;EJ{04ZT{CHv2AB6Tf(DP2bL5AI~LhG zmR-3fd*rx{x)rUV?R5j7&1Ehz)QOJih_&u9MpdGZJi8K$zO0*@bj@ile;TUQX*r`G z@`(157`*^-mMtbRhcn4oX9HqWWw%7|g>$w|Wue!Dzrew9eWDH%(g`b`E&e4@SA`p= zcx!^|c^LwqlXV*M=%bIgh01yVbUvu=%%5xh^gL7G3)$Zce|vj~)pu(_2rTX9WB z#3Oax-QL@JGktfP05O6uj6m3qpW^%f-sU9Ivo+d^+4jIBV)u|6y|e0Z97M5m?eiIjbJqdACz+p%#KDzwRQ>mk61LA3WImB++=es~Y-qW6xY)vi1h3KD+(;qn}xYa9_YC-4JdC->zJcXxMU*gnKsw*R^5s&=e$>n zsIC|Zrj_EyAqZUFn&&e6;3&$E1uSpESll9?*Gwk_@RGKK)%ElKC1{8~4^%3vTIRX} zi_-gMJ6$=IG#+cDz1o^gNT;5F8?#UgYQ>a z+?0FOd7Uy(S6vnk$DM1{bklrjzrTK|)RjZ9xlai=NvQCZ)bq|F7`=)HQ;rcuM81&ceK6v6epSeo*A z;A@N=UAgYEedz;4DN!f?v+`j7xAI6O-sArN@f-a4RBahjH2;|iGyL^ocgCjFl-_My z?Ux5oOPfqBeVtI41P$n2E(heV1d6kWA(ld9xSQZx8B8NGm|zfX$#^PiviQZ|O} z0Z@if%`q1<7|Xrn7GnwzKWqup``>9@%bW8{{7>taiY{dO3g1&w+B8S|gy>>cW_h8F zI)&S)CH+8*a9bB$>Db#^PC(0y%bdJ-BsL8g+OVR-x=N;=6)L99m6@gPTv37*&6rjn zK6%$AsLIE*uD#_Y8WQ$P8kJt5iV@>ja|3V!M(kfP=bV6f%fUiF;ZTLdO_%@EKDDlY zpM30AcvV)sM+S9VM04nY;UWJxNOAM_&)vhwc-kzYS=%$z>{j?vg9tr&zYVCn9;Q6- zOdsOEj@6Y+agoh7Tsp8(>T+Ho*HRO!wTj!WiEWFZoV7V zcFcKd#V`Gy=UJF=yI2}!_*ju4AT)CQ10u6ML2@sl#&I10=`q*2LJCVNhl40_weX)i zc=KZbX!=q|Zqxtizuep9(k%1)_ersq2Wx`{^Swmif_%8K5X;LWY7a=}&C`nAGCDRt7Hv%zo%(F4=B{Vk)YW_e{KRPnRg__rBi z`!DmVJ1n5VUgQN$$pcZ8!x26KW>Hb8YsA>k%w7I|ajj>NbKB93&Q)U6o}T3UXrHMH zaI^`D%zNq+GmH%|_4#2I5(tZ-izrpdD9^hKQL5}FJ@-QwZtxi83*rHcCBFY}r#N&k z`(M{B^RM1h^$WOiab|XwlX;za6hZ#(Eizr5@(T0N@1I-ik z<=YRDKENiNbFVaqe{XhR%q-c-XCL3?0h+l(q)ob-eBFO+is4_i5im!W1SK(6T5?A- zESx%-9|873)!eO>c*VJ8`(I!rC<91YFG{fajO1$QY5g0}q@E1x|G1qt9~K=StH)<#g*eVt$RkslL1|?r ztd}mUxx+vW^4}J7j8)R_$scG7{vT^+0aj(V?Rx_hShR$IbR(eTA_Qc?qPrwSV9`oQ zD*e&jh#=jdpv01v?hqECOF#rc1*Jixz4HaXz4tlyoV%ZM?|mNQF_+GFjxoph|9@k? z78S}L*s4A$Tx4WCm&(JhLPhLJ(=L!LnP;wYD+b&!_U=SRtJ&`R4M)BEhr{suF9O1- zPUQR(rLt?2PF#8B-p(x?sf!f~Cu*c)GP`(5?+R4XVZ{^MmYKMzH{`QXqw-gc|a!(;R z2RU^_IuB8A*FAYWie(GoHIw#-;mf7Y*VmlkDF_jpPJ6|f=`#PYj+&ns5QJC*qf{<_ zUa*rqyL@_`6=WR$gm{-6S6qdBNdHqM8Q0G7qSh}i(w#_ysoH&YK6iJTjK z@h^tE#^_e}byJ-~`my2fjD>|exC_5?b58H$3)gzknWCf(2j(m?pB?xT4`{hgE!X%z z{djOqtjR@9@_^KvOLSIg-^{_3m$|b?Gu>^A&+dL1mnMETakq`?i~9#8m&3`rVDcY* zSkXL7{U=`G3>_C@)M#$t&1pq3v2YW-Y;+lf^Q3#<4oF~CJeS;#d+pibp*{JKXH>KR zZ--ygn~Pflx6snngx6Lr9cKgnO6=Lza^c*aF(JM{XSU|bdFfEgVPS7n%s27ssLe6a z^GF(bA)0o6i_W;*X9G;=BskIA{RsU+rmPAN`xA8dxgP>B#4gA#^%xU`4)E68ry=j z#^MuJ$1W9#45V}4^>8U9ib~m^7r{y*ZD*#85J!n*E>eRCK<7S<=-Sa5ev|knM1|sW zc-Bd1b!b(p<o*%>Di>HiaWNE!Hhz=))^QQHCEDKKEbTnYFJkK|1_gUZUDP3Ognn?K?jICB~y| zrp)bx?Z4>5nAx_}E3?bh8x<201+|(lV#MOZXFdc`D)h0IZuGgrsA}nmBHm-Gd1!kM zV*lDF%DIr&(rMe42GoNOtbW5?u?vfuE6UU?k0@1C`}Y1Ly1s1EPZrHp z0wUCs-S8Rnbu=dp$a59d1(xt)W1t}b&yg_S)o-)hgPQ$-eYGNo&q z5hJ5a4bjPt5w6ZoRt8Gh*M{HmkKq_^#PG&T2djwa4)4k&9R~ja7wmh@8k|+kFT3 z`v?tCpA0wyr8O;d7;9+askS`9c^+I^Ozw!%gEw-Dn9QqM5LiZLHbcwDkFrwi23Y-!fw?%t?GN{NtMQYRFh;qphs!=g~Ll}OfDvcxj{JfI; z(wU338O!s;W>|A)0(1{EQc>#NQPEY{QlHDGnZncgn|jbL4I0Qb-%a`zZs>ptuC&rX ziZ_pA3Y(XmIAFgQizqF>c?|5Amb2TATxH3ukVUm$=)O$aV`-ke_ zh-7$!07dpVhFrE&(7{ejC6Q6!`M`B~U%k#$O;1*ljscOd=MJ_b-wWs-7NA(&*mE7; zIZfZ}r_gAMFE!EpbrC079&t@ z?Kq!ZS(f7hpFW8w(5yQ`|L0CHOhNP9qAeMRa9up%ExY@}#j{_GYBq{H-l(VwXs~q# zqBLb?Jx$t5cd5vr4!SW5xmx_KPX78DkK3H5S}UIQ_pE$5?re578dllPkc7XeX`Z_7 zfSb>qu7#05OpV61T=P52`{u1Ua652*8$Ze*D!P}|I1J@{pubR1SvONqTUO~v1=q69 zNO}t#+E?z<(Lwsl%H7HxNPn@$RvPdv$ZFNaqE?=RN7?u&L;cNQE51p4C1@m}7RWRT zv|~z)eTZ!Y+q^Kt}_v78tV~BM|`;+z~ygBCXgo1MwBiq zUD&Y9J;XJvSuHN+-jQnKSi0re6!vT@a+D?S0uNJ|vbDG~t4O5I7<=6KjsByT%5&@N zJvvEBDO){s<%1HR12*FDeF@wWiwiTQ#aI>YM2_Hr3VeqdU#BOJt!w?ID!T@y!=RIu zP3%2mg`%yFkf@iIyxqb^bC&M{+H0k2I@Ed^#l>DUxY@n8%vD1MDse|Q1Z<7#0|38* zm(u>8W1fALtUaM1tI12lamKen6?|XKpP%-SQavMWbOf5<4E{-XLL2?{7h-Sq7?!8NXcd7S4Bsn zqM%y8^(xG1(23IRP!b)`4Sz9sR%3xKNp_z9AM~n7NdiZ6zNW;R`xW!es|A&`tsN=$(FRC&k42-O)a?2>MoKNLJ1C2wa#&%8}9Z(#l z+R}+FpYg^iVhhc!3)EiW?N)5|+}X`Qd#v4Xgwt4KxQPn1EvtbQr zGPwnmvHyIad%PGw0@xd(uUpXy+ZI5wcBDd2%QYd{KUu7R%6UW2G?kJSXg>uY_$1-m ze|o1#b-s?k?zim(8;T!7I-aKuBX>zO?B}(@qHB9lAghQQFJ&Tn6ImwKQqu$wAiKReHq%|cJUz_J>lB-OD#Cd z`qd#%dGE^>>nfrS5rR}W5<*_hQ~U@ddbJRIicN@@E%=aNJ)&{ZCtynVy#F(9-b1=5 zRe=vio{+>2SX(Gx0Kum-0NiA7^!*=o&v&XHHg(=rcBL}viq0| zcN6pT*1;nwbRUtetTY&D-j$Hf_~4A&LA9B;OEd+7j4JDECvZKc_^$ukem3ndN79$U z7A>w+dQ}EHe+h}h7HToH${+ignf>C}$<`EA5F$x0 ze7`X^ilAqsP6a_<%*o6V-I{-~qPCp|5hR#kC`m%TLu&Kwsy5GSCkdPplUJmy$-C;+*2^>ToyX@VDl!Y`Fw3hf{D?&*fo^3`;DCam zJrpSDV{uyR3m;y0W>29#o>gC$6hJXH~>PD25-F$IjZ8ITe;BZ1;jXHiiv%(2h8s3Df|CMw6S z`aCv_^;XT$W&WCsHG}bmpjoj_H90+aO58J4Wp}Du@>tS&4vmqk_WM+MSauQcu13XQ zUd!7D*X(o@G|~h_;_4p-!o|k!RUY6K?Iw(Otnu*339hQR*T(5{4+*hlF`0b3s=$52 zazIg?PNn7RM@^uK*-j&Y)JaKYn|qovYk^h>a%MSvyP)*VYyB5%!UmHDKaU})@BP&) zI=YP8Wh6{GiW~j`xZaX0am8XRV@2PoaPCU34G6R!zuIAGzB2e(=JiRa1l#(Lq|XZ9 zU!^L!4CPFLD>7*XG2bkc56za4oNR(Iy*Jq}JsISfkp+Kc3$t$zgCkGk2C!gffV=; zp>?Fs_K?(_I0}Ztg(gKPs1C6c6;Vai*jK(-bC@`ND1uck*2sI948 z1dN%i?kYD=7kHtk^nV>OA3)FZG!sQv9Z*m|qHluKaP66o&rd@HeN6Co$Mf@;q`k)U zpa(4`m2+^_F%4I(GyUk!UKDf;6sO{;C6m_Fz7FdK!iSu-CF(Bba+$1|3NtMzg^V?X z3Kcp{YvXq?!0s|qkU>CRS9m=zsML~L@Za|3F{kN@(e0K0y5di+q=;|#-G9nS_mf$( z`wc`;%+B%)%5dGC$}!~;J*(D7S7~1M`St`%y9n-xuJI2LeBA;aN>JII;_QQf;*usL z$NY+p6R)0WT(=xnQWUZVY62u2s4Is{fOf4jt0Lh2gqolm8C112G2Jfb^mzqraR;(l zM^WIR{ZH(#p9($nx!+bVCfGOfR-UY%usnn0AG_LZoZ2s`hy_QOdX%sy>67^1=X5iG zQ51 z=`vy^!&Z6v@nZ`R&WVCLoYVCi%w{eiTwx-2s32|kr(D5d)5*D}(OGA%)B307z<#}a zS`ic}IhGc(HXsLhN}gqAunr(IlUP(i*1SgdAc?a3-Ro4Q(YU~yr-VIz$3IZu)jVC| zg`S>i+IZ4TM*uKJP;9=&mNxOtL39TBXwX2{$P$`p4Tp)=)ItYqCY(B}nQvc-FGe;UOBAbVvqcCFAra zxd>WIa8N0HY)`+fQx%!`#fL1F7kh1@R1MTAq;Xyu@gb?J8#2=-f}2ji&-LGGZ&C-m zLJK&tXm8;>7y_Fla9YmZ!onH%Gozlwm3@yRZYWruj~jr8YBLtA(2?vJr7GRikw4_A zbyom1Ou&wt#>`d(0q1O;Ucj4*SBf3PiP34C(1#w~r-K&b?9clRH^xG%@%UR0_WjAI60dHkcab$a9fnU zIa+0v>jpgq%-aYBccZ9~)!~5yr|fYn^VRrpFVO`1q9swx$un+nHC`I^mB;U;LfR_T z_9%_kdADvZeC(imB9XHv?(ktEf+=reONrhmdRl$N{=#=E--zTuZ{R*V@KQ$a@sl1@ znCy&xr}l;lq%*sHN*id7AZ@FuwsfdjmDljil{V4#rIJm|0M$3#SpmK!Tf_4G6J+i1 zSA^_pM`n8h3(>+AqBcaRH%s}A{&o=0GO0Q&PW3T3R}(*+5gY%poQ$5<`2wlmrv{>| z(CTyR=U);FkzW*W6`Y?qZ@yRq!6}<44K=~ z4&sFit)G@`_wH_NVTmj=y?n@1sNf*9309T@DWnw<3wz8GWrQ|~{knm$o@;Qf+gl6< z^7s^QrWgh|is+f$rSN9=UYYY2Dng?Uyu5?dcz)VdN2Y^zpzwuO2lQy*%ygs*1&ZwO z6gM0Y4XM^quyEig!la3Hoe;pn3FzxcX#MHY5j$>&wCjJn$u`t*AHAU}mggh@c>i&Ioq)L~rO8ypOK z+FhbOBEb+PV2(1io^ubNcH5cwXJzNW|7Sf$-k&BABOAlq`cn^(Y@Pni)nhdFh9mHq z$EouI*ux7;n+m!a}DT4+?uXy(zJ35d9H ztkb(QxC8X6wAb&y@IHY8=7jWXx9)uLArvk8`RqDp5;|+9Tm8+Ew%kfYwCsfnZsigj zpTtC>FwUU7+jm`(?Vd0hbpEx}UErUvA3}V!e9mf&0*2@BD%Cz`#a_tXcb_p(hnyi@ z?#4kfvrO6Ng*nyG5~u-;6b$Q>D?QyEz;8wpQ;C~Y_XprmU4yk}@^+08s#%k6URInw zjvtr(kHyc|kH~*p{M-v)F+5Yy>LJ+$_IOm*Mw@P@7S9=UrvDdJ$nSFysbzi;hL(?$i9yVaMeWV-!EGbM+G3BZ6B@!?+x_SV zSH&*o3uw&~M}^nR6(qe4{9MKRh|q@JHK=oxM>O@D?a{=bbO&TSpl8gN0?thMGVHJ9 z;xJ7+G8QtMptOXJD(4Gy$#}bBS7H`5H=X<&VazUnA<}C3so>18%Ur-t9L^tYm2)JX z|C3*jOjI$|L5 z*LHdclG^(ljB^4Sm6tOvZgVPR!-EKVPq*~_y6Ml!7puY_pO<_8gp8<-a1TL- z_1%Do;^anrA7S@yV?!PpbB+q;$HIJ~Tr>hz8|5q!PFwAy%m%5O=%l3whV3?hhFAmj zpA}Ysw7aPXr~|-{A~y?cr@!<;Fv6;yswao&!Ke0L#NdqZ_EAB#ovC7y5?6lIns+a0 z85LC48wica-83SUw7?wGceQ9U(2#PDrq0aS|9$pGg{Qb`2a@PRpdfVo1_7oZXvOo) zZv-a~@EaLAEl;;cbSgs<*dDx1R1n+5Lj{mA2->LA*tV~0crOG<5V5|iR_C0X@F@}Gv8=zqXgovrP$Ho6TVF}6Ln@qS4=@emH`@{mS3IJ3g zy2ARAbrWW=p5 zxHyXzaoc^3oE-`Jm4j(_7NMm9X;!U}0q>-x1(i-{{YC^qu z^A)H)dGzA+{OJYfPxRmAS{L7fqD260*jDMRGgMKY=Gy{@>6yI*md7r)i7`*s*5y+m z*e#kCO?UmLB%-o@>c3b=8$YppUy^IGPXyW_;ip$>9>)iKrljzq(IBu}P8qYhJfSnj zKNhszjGts=Kbnw8VH{Lw^MSo^5+1^W10}E;=Q*Mf@8-C`&3tYL&|w03jR1xl@Z4mN z)w)?Kg}x6$Hy!M4R0#z&;l>ib01KG?Q5vsZ5}Z8XH@aCoEcXxoGAj4uvN5umMvj(C zP{lp08fCK)KA^RnC|~#j`V6NarQ4pOqiY!DM_3_ z$^imd7^}`mC$MSw7PJFsL;t}!oslN$P;Uevrkl1)h;xee%>B`pE>l#CC=RyBu~CpvvbT&u3~3yWUEQ8N~s zShKarE8=bt@C3?}torSAsQ9EvLD9i*9qNMW!`~4m2~f7K7z>!RD@~*c74^rk`X~1* z2;U`?uvbtA_@&M3qy~kK8i4s71MW5*6*P}6W9S&sgj(>}oqsu6x17=sGKgU8O|ZI9%lGn3710K3^YMAb${2QGB{X3>`uyFj(%Ws3bAzSmu@l>0GxN z%i7m=^pj@%T0HyaIkl8)SATr<50PPK?Mm3xrcviyrWn{Lv3sq+Oe}%j5vG+m>(YRz#1 zK$T!40ly^v)w>8fpln!&M4(ym_w4f%@Su_@>J+^ZuA~RBm)t{wa ze!(CSaASkG5XId@pQ3n56?w}zLO5OfNhfvg$>1g3x82B&$~~}5fID@j#e)-tIFC(n zz@hBpCj_*J_D#qd1pQI<&<;>Do2wvd^q(xR7Tww}DBiTs8b)wp-MK{E@F-v^;zwCV z9xpIS!54tKdm!JSi=T)hdsTtR?-jkmMt&VIGU~BeICyH;s!SAe5-A#(oqP&xvd3kz z*m+EGOP=IbeHNGlKDb@!jO>gSzghDgjMRmK8Kd2sZ@by{K55!FDIP|DQbBk%V05b> zLAkV2mP&|uF`3PcOE;L4P`*;}J#qZp`rbva6?n$L%LDMA0)b|S8A{DGoVJ828}*N9 z9^OHgptS`6SP)|&y=(n%#$98|fp5hW=m}3zo3O`|pL(jA^8sv=;qJT!de7}WxgCrw z;AtVw} zMdDF{>PB7A>aTK9F5WR7`3pAE7Wh*0cNJJmqhrNT36%1wzV?y(V{%5FfqVvWPThIj z8xuOr!fZrEmt?G49!&#YjA8ym6^lgD_k8B;V~zY>$&K!4{7?FR5OhfzXC%5ALzEed z=^%gR$|f?cQuXs9IJg{%nxX}R8k+OTUWgJDna#_L>we5nez?mGr{VzsYTgQc2GmHO zvVmBe@fiu~GH3RyZ&X0)$!|B@x$iiaSLfmh_MIqTlrvOX|JJ>eAFfi;0rZq_u@cJb zkni|OQF~Z@Km*nTP^Q^eXbgJe)2{a)&=}l~+*UTW0RMy;zF!v#-a~DapYMK`eB?X? z(pB%tpFA_jps*|Aapjq`(s3RWxS?)@Hd5bu6gf_8{LVp0t`GAaZqFm#w~Q@aNm(fn z1wd|R6~kw>4LM>>r#c<$8;QvbUaJ@apSbYPI7HEx4A|wrq(XZgoK=cDkN8SWm;N^5 z%CU~`91G44>s-)E=K|0=afSX=PlyM2WF;2#i-;1`m+(RegyioIs&b|!2yF(^n>HwS z$3eNfX1b>`OBBkUV*!e;sTJTNZD{$x1VF@~cj;dDt`?N-z7fv#nFae|d@nQQA0Y1f zM(B6HIf+ihVUeN*!fbwmaXCM)&^PA-a09eI4&oid)FYlN*4{+(@Qm?r!`R1$bgW zjTHSZy5YdhH+L#W!hjqJ=JoeL!2DSlF3PLQ)_Mzp!lmqUsM(C%rRUbL2JXTz(aLe* zBf?uDwZD3DhQ8?=Lj2B&T);fz=(i)p$O8sphHJ|50_wbEJYpQ0$Vn+f6{LUqs2($C7`X@pkE!m9Ri18Z=5(?#@tPex?V-Dv&9S{p7) z7mL*EmvM1}8bH~6C2WOXUOLyK(0L_*^Im8+fHU<^<=BjX3yyhUbFBl87eL+s&nAF2 z7uXta=ri{Js1b-A8C2&OlyiV&Wzv&>0?sT@42mv866U}4`vwCz{BjC><|rJpadG6g zG$Yv#FYs)2ROYFSd!YC_rTgQ*R`vf~7QNwck6&^-VJ*k|ZHW;T!QvmW3h z&>Tl1#cqDNX~m^6+1NnQ%6E^5bK_HAbLo9DrN1ifa}JV<_~hpxo5^(FVWdmyYcON? zXNg$_oL3)G-ZtqX-=}6{7VcM5{94A&VNBJM8Zi%}Pk0mhjh@B%;vhEg6U$;yt`Af! zvtJGCXg8sYKrE6Tv}1nBSO!?`OLWnp;qgOZnw&w=NFy}%Y=o(@9OSbAONPhl zq}E%;((exl*}Y`eZ{dsrP)u$!s56nErQfBCV)bhNa7L4#kONMWV+|5Y(FR7qmER?V zK#G({@CUI+KlqMvRF|L(pQR6>pF? z)1ie!%+I-nmXXftB@O;QM}oMW{vOwFpck|*p~keE^s<7g94F#eB692GAF;af;UYP< zQPC|y3}fPSu8yVqEXxE=pVehwQ(BHMJ`y4kcXn|YF(%g-t|GCux*L-oX6C@DWqTvG z``eIe-%3>2ynoX`;**_P-bT1_vWkVfd z)VbXoROJG6?3Ut>Oznu_mZ1=oGD^E2W~>h(1H z!VgPzBZgS1UUAQc#-|MTs}}M$iL*LnAD=pz0^XFVc0l=){wqO0zL1`Aa^@5EMB~ln zU>McWW6(D%^ zL0duJ8MGYG99G@msSBbGq(SoQwJWUwU?&iXajVW+4)e541-ePF8#I@^{BaFU*-NIp zrE`ABUblU>d?aQ3`&|2XlYNpJF#HgjgR?MwiO{g_k@soxcchn!{q3PIB?UHlXU!N zArz~_BH?-gGPfgq*NZn8SOo8O&*Y%?-_|s77iO>4KVPWg#yciibR`|<(d!fU5fn2$ zx=}@sM!ze%<`=!A>>E~*O7n*JvSq3neuBArtCdo3*e>Fw{tRtfp~?*~I%KQ3u2^dn zA+60)Zze~w=gqG=uGyFJle?&O1H)D1Y{%AIoA8x?s$_t@?&Q72*ysL83G&UzEmnA_ zV%kQnhr*+4!gSdo0Q|aR@35G;Z~Y=__Z`5Sr52!R1?w#IU8^l5%ckMKL^A1{O9row z>B6)JexEC6p)!c7vHSlqlK7t*qyI3C2z+So!2r(j1B^jV-`qh!o{yF|Y%~q$7eg0# zK}JjZck<|Y`0er!9ApCZ1Kq#R)$i}J^NrS)_BngWy)F50z%89<_ABrKkKD&Qd!Y!~ z{0DNs!s9l;Twhija^k8j-!xUjAQ@G3@9BVFJ7vaB{u-CDf3a0d@5`-;73j0#G2M6022W^PiYT4r}mvV6iY#u-T`C(qUx|Zsi4X=5#CN3_*<(3tw!>`uonqaw|99)0|hCLiT55`?0WI9<#j;HX&etED+FxjS)!fyL(!)TY1#-2u<{9vcJ zkp<6TR-Vp`Uh{>TZ613W=4;u9#mDC2e!&9M=MVc_2ReAU6ma^+J?@6>Ds^6*98~_g z!f_??>;8GQU(&{&gv>GTe%xV7|Jb4)Xc$DkWZ_z#Zb+S{Z4=yQ>@fbXK!1|EExLRh zY=7s|M)-0*fVQEM4V1o1kmI@n+$l)oPZs1?P~LR=Y0Zt@Nh#a%b1vB7Sw9PLC9^U_ z;>k1NMzUMQc03<%@VPUmGc46A8<1xW!O!F2S$P=vv4g5epxz8$%1Y{PzMn^Dqz%5vX~@d@A( z7Hv^!>z~@4FMtSoCy#45?((Gt7D~GCC8U3-19Q?GDC;Z-QzzN&SKpn9pG?K|2EwIg z=R&IGZXk7-K?T-Kjc3w+)Kc8OWv5yH79Oc&LzVRIR;kWRh&OLwJynKdqKQny+fy;& z|80QJqqS*G^cVjT`+q;Z=l5teV_!PWL;8L3AAU1{L>r<3WdeJ0On;cw?CAKhw5~IL zwlCLE)UxcTDVPjDQa zS#X?21CJe2%LvOQq#+fTKIh8KNR412=Yhv0FR75m$;-`@&X+d$DWTTBgz<0sCJS2a zF7*xk5_Iwr2Fbkm2IAqFF3Pxz0)O5#95S+c9&Y*+TyfC`vUluR+ov6~WgIXp#|3ap69L@5&^18jc=KS_KT0$A&M`FMq@RB^ zEnY-|O;p)J9B0LnC_FVZ)j`Xchr$@PLTPXFD$$xOLXgEG{0T*8K>=!7IisjQ)1R1J zgORo6g>B8v89BGfUAdkO@&`|wIS$w-x+6c{3y;4D-ZqG2OIjelUzNh|_2Y2><+mka zVj4p3hjQpej_8=Hj*}`>W(Cr;{pZWo=2E`MOv%U!3Erd`%U8KKy_B9HQcstpUfXgv z+7JKc!Lk2l+iM1rqyUwH;*iNm80$WxaHKz$9rs3lOV`Ihmi(28pN?H)d<{plKJi>D zDElmi3)1#z0`1uPCVJ2F#$opBM7%%YpSuwNh-9Bh(CWz%6E!pnXCoF|rn*eez?g*4 z**4E9V7?`4?E8(Cm9Bxj`TqhZ%SL5orsS98+Ls!~w3*gVwWq&071*zyN0z{!fb$O3No2l@i-q8?E{mVm& z1XUa0mPL6AAbqHSrn_8rv(cu1E5!ZLG>f3Kk^OpQYiw3?VGDCe_eVtsF?pg2^PvQZ z4EHn8DY#5}!YgOA>AGQK{gFIk3|&Ouc#L4*X*dZjBFU&T(~C;Z_sZJQK1wY_^WIWZ zxH76F`fMm9_J3=DI?_q2#l5v>q5|nXNHkv_XXL0!#K1U^+p+%~jLR8a(0uvx6;v4o zn*EC36n^(i!iHZ{$I^cR;}F`;Jm)S-YpCt6#kSQYa+#f~48hBj?4N{`b)i1V;@w&C zWA;u_a!Bn8xEkA-wh`9xd5}9)<~Dk3xIj>kFi)nJlgVt(gYqJ=&;5(MozCZ_Zg3Mh z=t}O&*=Q$w;d{_9E?3-LQN2&`IUD!0Kx2^CnwSh$){3C^xNTqK-7yMRD~gP5S12RS6Hpw1B_@k*Mv@jzv3PDu&gh+UX0QKk4XH zU(}T3KZ>{<5t6d)28O*$P%TjMs~#o_4GpD-{Bju1X(VsJz3&b8{uGBSeqg-+f%kIk ze<=OURKk`+l`I}nM7@r&!C5x2_y;C_>vLon+%|0?_c7pAIEHjH66w`_IeU;#_|I|3#M~)A}~Mb6E;~u>PT82iW8P!bOhnU9uAB85l9eORhU@#n&(CmmN(GN$OITPXz9Ep|rYaTeC@7NrM!c@6r+qV?pgzC~sQ zGZRTR(-cw@@-^X0V+fvUI~)@*wy8SPS?GAEO_wI4^jt6{*CML34;*+!zZ~ze(q!DR zGt6<$m;cU22Zp;ookR+BHKwSBhf^VTt#prGXBDaSXzg))6w2yBP}U%dpYrO3w^$RK z-^?t(!e9B4HqBAR@oInYeKm)#^6$K=GANeQolHa8bsB`Xuh>WIwqaCdts`Sia4gfG z1lF^WIZOq69*J;iI6pyBE9YQRy&kA$jjI(tplnrs@I`vLt;O27p~J{|KN&p#Q;Ipa zfOi#Cmpf3&@F#R%D4aNzSP!qojVyon+x*!G$7eZp+F_%|(YW9B?p}wY`b@3o z9bokj#|B8o(h5(7OE1`^z6riba-Q=S*bH9(#E=1};RBQI7$8ng9n-B}f9oedwI9M_ zY&7G~ld*&HA@&D(hJ=O2{Xb;YDmZDwQlTMjW1Twyf}OF$MEf z`P=puN%iDc_Itr>m|3$^*(9-yDz~VZ!Sx0k&py3OiEK zZhM*ETTaQ%3jb%Yx9P95V`ltHDg+@f)`!Xmk5-y>G7zOprh`m3pKf=Hf)$PH%QVR148Bb79jTm1>s8r?H@Y;dK0NhU@IEN|S1%QqV9^dkXJ0jl!Abv=1^ z{LH-at2Cl!Tl4k`@^FY(7 zO!2%uav=MNbRqrR)qkq(FG#my4X=Y5hj{$hFFL|s!^>+v zosAxQ+?hG$GVcL}8x5$l$GoIKSRt-)T7_?VNZqhbKh{7N`{3%l+Q?aw$m#g%B)$DE zxju`V<5sQitCqqh>T!bbN?YT2RbM(1?u2RicjFpLLXP;A&%d~_zb=aYr;$)L6rT8> z0f`;a&b=AX*^Rx@+!kln>W;TJg0qpHq*H7Bb#`k>+6YQ1(wTGCTc)i2$Vss!?I16w ze-z4WKRZh+9syDS4N$Bqa6Q1$~=>g;7cmlCK+8mfNdH%mHvVcCPdd$*BC84jD%X820*uOL#P%VpkEm zrwcVq9||}cFuAZ{$U`OY**-LP`1qryDz+kL=9{T!h6=gVewrwqqESXs)xB`ty}P}U zd~a#1ar*5WS+6eb5pd>mR+B7~CVB20GU}KnC@3?(vjT(wMA%@fTVnio#0Bs5n53KX zI=BmQy?cwKE{sJxKpwenm22mL)Y(-ShS#Y!b4yLymb-*w(flAG1b-;1gNDJ-rGm02 z`*}U1?`)IqKQpF`@d}85iK_ie5monD8 z-VIZB1?W;fuikiW^7S%uG~2yHDt&tZ<6%~gtVAhta(msGL;OSFT&0gK-AM8jp8WAA z;&yibfS;5k;J%_zxmrl*Jqv20uvP8T+PKOIq*RY7$I)qAwJFL={i+d6v(9vGf^flb zf@|@YuLnAj^I8=XD&03p!?POaJ3s$Ws64hsF#zN3ho%dI;1tl;d;3*(O@&_*I*^FM z9L2d~Pj%&JlBJ(#x2u2(%x>idn3E=vrDaLUeO(-O&FY4cQS+6qFDUtpm>=eyjywta zXK+T4OY+cST`arYLF4H?SF$$jFcvF@4<8nt5AU2$X-f~}UdHbT{fVDk&N%t)zhyQS zm<`Yvh;kZhQi?K=pLRn9fs>JZFi=~rhK~zfOkB)xK^LD{ms6aHnVVK!lT{`+JQ3Uq zGMN!>mw4&3^)(7W3`52jE4N_|AKDxg$3AZ8_PU-8?S9HX?1oxSoqHAEm07MyrH3$OoIuyyesC-I6a{v zlh}tD&}`TB^cJ|P=XvTg2eg?LC*>WR5@W@Rwp85Yo(mik z?4fBD(^ufEL-&qjIX+__nZC7nckOPpTf;j+qxvc&Kye?i%ES& zUhK17v&cs9?K@zOQIc9@UIJJ&eT7NlSkl1xZaG^(8TNPclfmO(!Y*)uM!of-Pd*8Ijg>|5E3 ziQ%e37HphblH66?^!e8=zEL9Qz~DL87P{Puv5+ebGEKOry1v!7l4L}&lJJ3Ji+hX1 zf;^{7SDR zGq-)=1y8GNxeUw#B78oRu!SeFA2TRzu&*#%y;bhgWh%^`?>A*KsGd;eVaMjx_|CXh zd#F#lNpkA7UcT=3T+EB=hucfnzM(abIE$N34gqZ3pYnd6Q;?~qh~F7E!;pC5WP3_{ z(@W}~lqmMsXaH|pY3(YjI78ehDAV`H@&$W*maR_0%$8)*WHvMpB9GgnzO)q1uD24{3Iy+m?ULRHz4mwb}N$rd}=8}z>>=Ea}y`I zFLpTHzDpNxCde0|OQKE;;tVTTS^BKWQV`%?$;T2axG=5(8Mq=wk!45_1r~*$bLjD-<)As-?F6yP* zVBq1$L)jzJf(G)=n%qi?ab)|TQ0r){%(Yhs88Tx_p7P2+Bx2so5Fk!+_v3Jx1GId@ zabj9aWx)DA-MrX&KvM{QkOWuV3w|!Dg?yUqcAUHiuD+}ekg>5p@z(^Cps3-a<3!uT zeVNwdG^w2Kgl}4SlC9)A%IV2f_AZovniOM0{)2n`2gcxAc3S^-W+^%@Lu2i$=sUXt zcobb!JxD4CEsPWy3*o3d?`}vn`%v`J2Mwka#w^L z;xUP}_5~}oF8oB+1VK&yOEzs<-JMnyiFuD)g&8!SDT@n`+?<<2&iY{dIjh-D%3HD$ zcurnKS`jGStf`JP7J7UuVj-B_vREbWCRYZ{;={S81gWXGgG}b7Q?Q`B)qy(`sWn z6L@3!W6=feVD;il&FWW1(YFNJ+1(Ipq>FyjUP$U>juIC-!}T@`s}y;BTH}$OxrZzG zN8@_2#%i+F)$_dW+P$&v@*rI3fnJ`KDeoOs_{A(1}f37vI5%mMyK6tW;)BF0*TX-OeA` znvZ6b5aGwvS1xwjp)K>Q^DC&?X?a^O6y*>4#V>r{F?!EYb0j8Rbwn@k_d33|%BFHq9RFbJ`&j}vHB(^eqAPANncP9O=SK;3vd|S4O#Q+}LzK9TZY_HRKWO_o~cDT?+l&r+)bf3oN+|fhx@tFv1q6 zVzir0rd1zTmd=t(?hMn#^jq9VA*a|Fw%N!@?6x`(bgUhuh3t62?vW8JWLG(HZr=-k zs3%GKC{k52!r<*&w%Z0>sd{`}>7z4jNw-p15T%BJ0%*VRDn|8{R|kWmC(;M7hI&18 z#e!0`QGSmq+7Owwqa^i=jRaf8rO!fF1SJJt}iRR zeC2IoQV~{>Rb4$w6k|d42xI=sE+Kq1U%RLx5<#y@9{f`#E|ndo(_Y_fS8ft3%kA=9 z=V$PTJ{1a{-{+{+p{rn%#@cPDp6*h)qm9P4MZ3bsovD)XhExb%enqtz3iRL%THL!S zdi83DTKj~9ut{4rTRoM)s1Y5GlVG#`QnCO$LI2vqA)ibRX9=y3{2;0-b`^e8oqQ72 zX_&{~i;z{&x_zq-YcRrNY!nI87&3I^%}HEb{6%U9K4GC}M&&Y-^^DK>(wlV7pVo;- zeqC9TAzA!~FbPApH~bvlKKSTg4Lz<6Pvy z_DKjWM{@H5@G`x+Am3j|7R$`L4ibUbP_FUsaS1JX3xS^UBdU_o55os%hNX>L(wwH0Cor;K&?Jq^0gry93 zN+uZmTdx?bBA}xZjV9K247X=BTxGPj4oQQ@n{SuYf z4$?oQ&JqQW>GHKGQd?ENb(e9tbB{Z&6eAaGDlK3249Q}I?8su^Pf%7*{FHvSsr8W4 zJRL7Ka>@buFj~n{_(V{+d1EUx_g_g>F&y@oQq$gf0qQuZqUfHKE_NxyB?N!9u znu#7(rf<`Q51rado@$`p_{hQ+q9&g~qIgJoWs>2RG{_j*^o_uy_7@Z5=W!P!@43H5 zd?jVvFmaY|?FF>8qp20}B17y2Zp7x&i>YXYKH}FJXGP7jS3H*SpN*w4n;HCgEgr&o z?(BPORmK^LP%G`zopY);Hy74r9>>eV<3=u5> zj4DhoNLt2t8wbh+p0OdI4m_-0`M&feBRGOqXH_P~r?9q$16A{z%+W;9dKt#&Hclm< zZaE9KJgfHi=F$O5eP4_R-Zu_2qO6%!TT-a3#Z9h>vewvmA;0ey6BGYQ+lUoPi=oIU zWeb(;&ZX$PClf2HqgiAU^DiwV!9-&y*!Ecp78Z z{tzVh16JP5@1Z~hlf!|5oE^Qc+QfPtKbVC8e)%nU?cN0hHk4BL6DU7bq;kybN_JDK zaJr@V<;pVV&-W?FPw0}OXlGVsseat-6d119*-aW~{khIdVGvoS$qE=>(cZwB&I>X! z3TVW<;Q4N!Ef#gaNnWu2=%ymK6^pr6gtI(znMrc^z|U8R}VI$avT6 zZgF<&FLM49(=EgfZhmfbah%CNc!zA56#0v5889Ylm-Gy!YASA!WRKVmOTqGatM2<3 zFi>7;zGop9zQ@x3wnJ{UB^;S|it>jkY0GR+8pNdhu zB}w7*aqd03>4K7kPd?hLEcGTUwlq3nerL7H5G!;Q8F6sfs}pRAY~nLIVY1bm%ggK`VR>RVgIo%(`k$Eu|L1X!{Kg}297e?ReT1pCQ=^>C8 zYzIhKjEP60#ReEZ|9p}xYJe3Y=u5#scbl7&mh}G~8xa&oE0m*51$PrBixDVd1XgB) z@$PMxSeFjYI@Rf!Z?d5Dyd8Q!wBLw74t9XfbvbXHg(kyFQ*y!er zVQ`gN#-jEPrqV(i%DvgdT~uLW6EZSCFy!~2U{#p&W3EMjpwo8pg8S}AcTqvPTk)TA z?|_wJlLT$4LW^Gf)dI*ZVm@+|T|e_y>}-D?J))`B*PsdpYU@5Fb8S&zuRt+w@J5An zu9_$$ZvnrlHCn0Mn^yx!Df0vGQx{8LD}z$$GVT1GQ8*{xg&R|pf46oc@PN&y6UqKq zaBYhwE@9M*Au*S`mEBPQd=8zHqj3BMP>MK=I?cdD-lC*Flllx&Y=`~PjN!oS6DxzG zz4XD`!(J_{M_u9Wh-s}}w0;OEg}?(|Yl%Yp{ls>Z@ayCAPbI3Kn_zi+PvgAfN_cwv z&*O=RX4MH-;z8Zm_k)-$Ej)G-Q)=0Kc`h{qYP^vof$her!JK{wnq9^ezGM+i4Nx`b zker>%HOCVhg0Uoyml%uo~ zvlRA0S}eieQqd13vK1;&ld*u63VPdGl?pm%2>wm)u#A0UY|7~{`0+(C8L1fPdpPsP zyGgdGICAD35*2FcXBL$!YH9V_4Q;(^9N9KmE}@RN!b9VUDQ$XqIA^mpO7b2SbXq2M zBU;IgA$U=b1UHN*sN{z*QOPAO^vXD^$~Y$O1giW;++^Tcv0Jfu9r)G~h5oopwV}&5 zrV662D{hlg%!ML^h88!gGfQdOxx)mO&WOZe&RRo_qOs#;_JZ@?&01D=gcN(aSru%N zG>SXH&{?rBGMo|R>qaXn0u@e`u2-6~-H$R)K4JscTiFFpAqd7RnCQu}BUH?~@dNzQ zT4&@U@3ebK)&g)k!@n{2w#XfHqX^E+w$U*FRXlLCvbjJ^Qf8QZB85HY`LWsVw%(Nv$_;t8yNJoZ2QOti=eL*8#|m4`VI^Vn3t*Qu zHy?RSES&p@MQ`*RLp6+7Mh9LVmyGCJgO^ya%CZ*2`U3dz>BuljOB^M`0wL}SwI5I1 z&>6ds--#6;s;J~o%G4tuyUb}g4G%^CV}X3oK-pIM$QFqX-F4FXF8~&&GoxizWvVWv zWo;J|i&*#UYFXG7^(Adf^tl*j1h~5mSEj*tvPj~IgKp~^PW}8Wviy`L7qZJm$7rnF{R}oFE;1) zSY@;6SZp)nEg~F;$SX1?3|&oa%JyQGRo&Jr<%=0kIv8HFnx*@Gg5!u5(jJx|hpR zm|_{9(IyS75XA}DI36%0;?avk@4Q0WZ3**SQ^}szO~E%wMFXf@WOXD`(SB0TS%tqb zO*lAe52B*=T)vMZ;!H>8x{lpkSD>DwA- zEXyxK>ICu23sK9xGDx3BimUVbG^_MnRW07DN66($7LMg$c_Fz3LFDv(#r1n{ED_e0 z%E*%+fxIexLoA z?pP+^haM~Q{S+30mssr}OcGVthN))ASk8EbSQ#eMC4qxv1oD36N-}uJjIB9k$yi0} zjzW;|-h&FUI?xqle3NsspAR;&y=Q!EOb zV@x&JF&iG`6G8+w!BKt$s$>}0Qj_;vIYZ=vwaRmG`R1#~X4L5jr?2~n;HWK=4gGTR zL%sJGzt%QD)Q8FX=|}>L`1F)CmB1UG25?}1C>(UnREt-?0Uk0uFjD7-INM+PJ{0LA zhhwuAg_$h}QZY7E2Id2I6RXcbjn%z=EscpaST2OBKN~TUje!n7LLe6h=KL@(>`+-PJDMPSzL&7mj4fW2$ohmxPEsEf zC60~V8_d|3%t#=-?MR8d+?yYUxKtG17ud2#S@W76!xOSH>r0v^oZQHAkeDp<1A&!R zv1?(&Ip$t2Mr;W`izRT6zP$8KGd6uNj8Ri>L58s( zArNdIwZ$$ywb`nPEkm0%SNTct$;|aLTHg1dGCGuIviqcRKB*nyt*py-;9uNYvNu=lA zq5Fi1vGAZib2t# z)iZLe^1T~9qosKuK(^>>RDl&d+>_jz^xNk?v_$fg%ET}+MhwkF@^PvNxSA%4$rZUv zpYFZNG)e3FgaQV=+VaCs&P<04mii0?L0c&;89j73juB0%*vP?_nqwsxI4brgoPj}8 z9e&hM**K?^or&b@l<&|D<9oAl+%UAHTEFZaQhyRLLKee5N?Y8K9Tj!46C2do%w%AT zNKiD|L_Pd|D`s#dQl)#xLX`dl)$XoVq34dN?lL?aZl5*z^S!0e zHtn~4R1Y`?{s4V3giGoRG>xnVh%Nq2Y06SUMT*%u8Fg9HPXnr%_%`Xvs#eu5KfcB^ zq>nZ=NAbKI?YwKgiCZ`_rJKM-SGx=w~&%-;@>L>yc!bnoK0EOdP- z<+Wy6o3D9>ofG?7p?`jL=LbI`2+Pc-GZ>$`=M#WuxaSXSn3G=JJ@AIFh({hRTKsvC zphKmO?}G$0`UpmIGGgS(;oPxW!r2wsozZdZrY==9)#Z|xM zUkxa`#V}Y~5`on@rR?Zx_CaY_TNZ(keI&96e}Re<0iWkV?Tin5{P2T35S-cQ0~Lh% za-H)dJi+tt(y%J0E*+5OVE*B1N2yt}a}o*ocFPa)YZ?Z>Q?Pa*j5PU2MgS)W{0*G- z?#~irt7``@8m%~9Bo^P|%e_dn(YVF^^8@9}8NVunwOH`AKCrEi1R>!6P{3Or#jcr0 zg2FW!-xmuNJL74A!Vo1D*SQ3Fy4tuG?pXU0t_;$vYow{T~%0A-0{D;{Cn+Ud( zApH*`VRpcpUNShqs()Jz{0BN;Uo`(|q~Ey@Ou`E&1X#~7sQ>-}9_d9n*rbnt;Dg6&s*oK45haU~FXcc=8&_^XiQkV7*fsTHjpsP~#0lsZ zK(ynQ=@;OM<}lE-@@9M!;y-1g4gY!a`~BKEb&Fk5|@g(tzG9A(dxl>?|1d!~AD()+2jNaZMF z{FW#gPrww1Rhq`@VnFb`N04KPk^SFoerbuySs-s_q;q_aj3Nskk%1pb#9wIsVW{Ur z?)Akq&{+QIB30Pyi&qu!+s{=&@`B|!@KD%3mRJNd;A@EeyT$*T)PGhYl>TWE1Td?y zy#T!pNfo;2ZxdI0JfAnrG!Q&*NE`~B|i?ha@2$r}}Zz>8}Z{+GIQm^j4x=`fVh;F+WB-*%cJ#QK@ zh!H$T|6Xh_*eJ}>^FM8VQMdfaGw|3_cT5?Bq+Z^=W5D!d+;#{wy%G;VNMH~$~P8R!(rZxks7Mw7u$_uCIR?8UzjpEJBp83i+Z~#B8S1q}c>@6GO_PuOC&f_A#@%aP>k0s>4WARUwz0nuFUI&}) z7bj0m2CXd>LXIb11EY-tiu{Y9m^2XE#Z0?rpKjZ}@I_fG&2Z}lJ_Er{>w&dMs|3El zVB7x|ey8zx$|A_BC|CvF?kMDhg=)|l2!gm6ac~+;J}<_)7;Zhs^15IV^qp-jwr6%c z5tIT6_JHVhKnZToz2^h}EAB5q^uSd?i=iL^5bf+}{p3(xj*;#~zt`5h*ZEN^E|%9> zrU>{MydVw;zXim+I2i0Zig&Svburj+A$92_C3xQUKc&a=T#O)`o2M_1wtkaOx!8I= zd(e7%!f#Pg)O6gcqbNww`DMp|`#j%#XDIm$Y9pY#wW;0be4wOa#LGI}>wMYkKY-yI z2C=zUHF867?V1;geP;`~7~%M$z^Lu@{d2wcl-`wstHU?n2%{?StI_7KyzXPYCUHK* z=ak-!y2HeHlk`{hBz?YN$66?wygT_1%kkDCj`#61z-vGVe~XlB;f8q69Whrua&T|n z?Mgmdg@ldqY?6E)^bQ&SX!HvXco{$_j<>M)SMhCB&9&F-Q%FB{4gE2ghxtqo-nUtR zIRD`PKWo6tDAHZv7J8=WenUktToK^O`NfPLf+&Pg8zAKD$oOfpHk?B z<`ZD_G5Ao`z-epbEt{I@1(7CDnTYW6-n%^j^3C+0GXJbWH^xU1phEnI(>qb&`4LpN z_YjrU-$Oo{hr%afPxx5`3k4DgIk;7d>vCc0##PJQiS#uaR%{Hs25K@fGZyiJ`%l<5 zp;wIRir)RP&aJncvNV%?I#Wmw^SD48=xx3b#CYy4?*@CJXH{72x7V#uBj=(f2_ z@dD*pdrIFWbAiMH!e(NWOKAI<*SeasiNnVHABU-Y`KcyWfyEd@*Kjs9>H9se0``#W z${g%=;H#br5rBd;2*_3tFP-7v(z#OIF3k%5)RP8-amh3tS*)nPP4&&jR{{=M)j0W0 zw$HCeM=+s$N{DxcpHh*HK6?ixq1nHYO|MOIY!u6ogfTNK9){ybXvxrOnx_9Z$Qe^y>EXXV(jU? z&NW=GpA0+)smpZ~!})EHx^M-^MH7!#-%26}ZdSvE_SnthJ5jM?_|Tw6VRqT>zNFe3 ziloi7KH(+;XAcVJ<}fRme9tRG2rg9^26L07hxRrRaU>98Hik+n=3kM)Ifz=$In{Q# zb3Rh(5Y|~NqiZ8+KWmo(UCh8sUo@8@TjCUZ_oMo>m1J3}R)?t?N{VWKSMTDS=x6*K z!+Va33$DO7s{SX{LC35omD{0K2_99z{=CtrofFRk8}0|>#|sbxyu&Ym znrC`Z^P6Lyx17^S0RHy-%9GMD^)YA-{0(mnY4L11f7SN~8g+5`N}|Vk&>Z@hwqs2Z zVO;qgd-P+-@Lt=E@CbY^KT2*^}_yl6mF(8i*f zK^)tceD4P#X^jY`%fDLlZScVqtlwqd-Ao(ajsn>VW=y^FoKKNrV%4Fu#% zL)kn0k!to~=xIA0 zYLM{r9>f+MWfNa}IjVv$64ZtgtJq9+9T^ArX7}w7oWJ#hMVYin>9wJxG`hF~j_(`Md94gsn|1E+Jqs1gxwN%ltq3MK!OgTojeJVt__a^Ok!jy zs{21fF8gng@AN<74_YjBD8;lF!0~}PUHz(N>H+n6BebPEC`zs~zIBDu|5wO?e?juw z&pjM&Tt3CqHrw`r$Z{Zb=dmhYo6%-L?>4=oKgf7J=5`oW(XI3svk*M|0ubx1)s*f& z{}(@J%lsGVL+A2U<8rh63>C%bUD9L(Xu*2x?TnfXE|}d~IbA`o_)7Ni0L^i0oe$loCY2Qtb7kqNwQnxT1--mm|In9wv#zm3$n($s$_8`54{Xf!qfbXS2?r6gF zu1`D=f%0u}trg}>lGt0Em2RF>Hu+H<-c<0LmO6NKZU%VTa{9*v#SL7xzX0&IhaaziFG5WAhG^aHc0R>KjcX-+~j)tA+bI9I~n>rm-JWNuinoV zx=9}GepAUyG}!3)H*WcZ&i`_A?N6?;k1upy8Q}Rh$rm6NUL!D_dm&D{_yj*5<@BhM zl~F+PSkb^(V!or_kA8icpYnsH`)}OxUqSyy#Mq$(t%X(AE%FyAU_FEPDLzj^(;rF= zimCOsw{C5{e?X&O&BcdnOr(@dhM+;md{{jF-+mrM8@Sh-FI8^g#fkGL+`nsZ-j8wgZogO%{>$nYEx(5?8VEuzqm+3@c4CU16TY@O7njK zkao*HJ(8xZ1y)8t>c7qck6SeU7`LcH+uI~9_mez3_)M?;W87l+dG=q$zjUJSw@3=7 z?{S*V2l>|D>9@U_sZ**ibfGQX-_Ly`P%}}7CuMkum8qIot)URP`!a4}SP6o3Z)O~K z2?x)|Ca}@D`eWdd9GmnnuK15$v_&cDCwo~hBnCvU%5xEvsu5v7m01e;eroj376?_2 zy}ZP}GBbjp0QUr}*v<4*J{|11Ip-$b_A4{$qi0)`ll~@)89I94H}GABzuW#Ya5W?Vv#raOOFj@L^rjOH&KR(hmr{lw zR-rr+a+gf`^p7P~oR*l0>z(S$qHLz?hjuy^{$=xTEe!B(B~ojhO0dv9!RQHzHUKO_ z9>~}Jb-GpJB-2-?_op9?I3$m$UR@`%#y~$I6%0GSor8)R1u;<1@Q@sX#*hOf5b>V-lVOHezTVKwR&vevl*~*b!kx@A0kP)75XMHLz3S%- z0K0ir2CHKv#&)-KVtQp%>>9xjzg!g;N`@yQJ}tuLFB5G z)Edh)RhuW<>hp$Bbf(VaTK5k{H0-@A8`+UU8keCJRYW57r;S761>^D!^wqHiKy))7 zZLytJ^;3^I0L8Bd*8@8=ZWf1iB-B?S3Ru;FWZi*If^iCZE~UpPr1ha;gG!sN$*gOKO~Z6{yJE5U6>}Da#euD6jroTMAwb z2|5f3o`V?Az-frEW`M9-mauw&uo6Ug4kE1jA?P$D=mmp=gD6#$(LjZE~tk%M1$Ro7|UIO&`9GbPrKpTRe+;a-7DiJ<)g zoY_|5QZzjV)Q1YSJUi3*j?kAtnq_|1p-tn;9I1?WHqZybGr|&X>(zG}PB1_H>?YHQnLQiM5ul~w`o;nH=nFOcDrt@y z8KcE+Awk3L^Sl!@o4f|kM000{A@*TO>GzBAs*_&;;JmV9MNF@dk$cze&gdEzE9l_n zdAL^L1C27UYx*5VU^_kL}^F;@4fuB=VBBALfIFgWr7TbZRaP~dN+?}iwMqU;9^bp@Y zJ!HC$+tTvC%}G7)5V?B|*>b&Zdt$)%vD@OxV5cFNbn-(Fbk#wtkg}88F97MQ1>QrS z`M}aew%M;&hffQ5O%CHcRA-RK8LYmQs1cd7nYAN7k3#L``#B!%4Q^It#U&a_Tz&xz ze==;hJ8qYG8)4Nq`q?onChSy1zA~q_-6aNsHi`C$mf0zVWp7J@0tq=R>U;XO*(j#d zP74$kV|bKRz-x)a&YitEK;%IC{`d`v=98q4yuy#c^-YwWXKgLH)`~(Q(~pi3LW!+{U0@iflR<{4a^{dOWXB>a>i^X6ZDqea}Ca(CxJ( zWOqOQ0>lx56lWE`s8$A-)vzSq6L(;mN~ZB&>c@^5mgZXN0i~FqM0`Wx+L8uN8Z2Qi zGuT%W`xln3SFqGH2SG#td%E|{8*Z<5md^H`B}*FicR=6d@rlgkUnOEPub$K*V}InT zb2&>Q#56>{ddMirIqPV)8{Y>`UGc`|@=EJf=<*dEQJEb z)9f{}Bt-gooFD5_$cLGg)7f58HW>w>naXAXseND-#ViEj;r0ln?3e*{pQ1WN2)U}5 zsdK;gIIG}h0lHw|W2Gyf{wsH&T^aMf5yG7LJ~zxSKwT+MR;huMAk2?3`@fuSlLqU!-eoHP#|tt4pmONePp0C)4;s>mH(3e@*#<<#BQ z^SjM$`!L>>*7^tKcr9oMbcgdtqNgkrxbd|#d0n7S9D{qQZJPr z4sU`7DSco2R&mZZA!xfTDJ`W3{I$*uL!OB-_O7bPP`d9ZDnwuvugzdorsc1b)gX6} zw^<0zwkA{3i7h3{?)x-SPUh|&-#~;;t!MdD!AG>ZQs72g+l#E;F{EdH>TB|lZ?xMy zkx4#oGBSZ>lOBLA{L1rDofZhF-M=`33JzrqLMOq29Xgxr=fY$)$6|0r<+0uglPlSb zVQLfkzELkm-z_o<-lWt!3zz#5qX9;?`8>Hvz_y_xQLk&~A*iB3rZNVC_u|1-_yj{j zI)QZSXyTy@Ypav*^-}f2bgRNwp?B&Vlu&L&H?41DOO-;9r9IH|6e1f+J18u4I&NyY zb6#cJVp(VE6kIp)lYM8z6RV3|af_ zjsD`{+fn(^ab|%t-9BQs1x{ydN!@y*&P+{O?yaemIHjFm0E3;5;NpM=VzwW;aS#U% zXi9C#7bNzl0$~0a!6y(~&tkzzq-*@xud#M>fM0lq=XeVKx#lcN_jzuX1>R6-_z2@i z(!fWAM9uRjpE# z#?J-NryspKU?;`N-f@QMrb$XJnXbo3+tzI1o_ZWTUj{ov&WAGb1Z;R%66-tj0@^t(RfLXoE5am$tX%pJ|R`6`*mvjUdf?rBE+@p(TC8_u5*{LOP zJ0P{Hmb2CO;}; z19V`m`%&Fx*dB>;PnP<-MjoCeu}kM4Kgp(H0YLaCM83YOwD~}&`W@l)EaX8a!jiga z<@K^ApVbY-)ga&Rs(XTbD@aATw|LFo59{ng1|5e=iK`h-uZuhKSoxc?L96FVoZmOt%&NBl^DC1m(2=5`*ly$tZts23ERfTVdK2@TA z5{tgn1$?uo%Iun(7htmS-7Zz#H0f8H46tusz!fw|Nt_H>5up1~aM4Uk{NO>NY1zQ& zaOy+KCD*B!;(gg$T38N_ZR(rZTW#ZHW5)H3B0wNgj}C`JPTJdZK0Kgq`$3-R|_pZljb>|CVu7(C92#! zDGx`qK;?iqemc0HC>3YA5Nf~g!Wh9mm|Srr=6-tgG3TZsc35u5y)zv@l48qp-#l3% zTUe^7iyps+IzCnun+u@?WVzNd5`CJ7}UOAtMxRzcn! z#oa()_l!HRFU*Ryj4*EaHP_ zozaZ-vs_O;wxx$NC|R0p4pg4aNjVT+#Kyxe2{UfVUt#O&LdaN(Hr#|@LDbv@N?K4$pMR533B7EGxTap}Lhx2Ond?lhOu}vtP%6sNiXEhpkiP0A%h)mR z<+3PVx2IH71K(sV)X->~gq)D$GpO)Fcg0R$k`RW<7|&eb)lq$XuA)f9PY(2QyHaYi z-})@we|7)lKFEYuV9$hKwygr1K*U*Xv6l6cw}bEZ5t$3Q^SK&r^|JVJQus4jFs8#( z{LE$e)cZ;hXVHHprlZZOBe(@{|ISBoRTDsH6>qk2%b8zJz2e*9{v|`tlE>wCDqq(V z{0fPO7_abi-Mt`9&bT|A$Hh;bEp+5Q?q)vgtd`RgI3?zhy`O@!n=n-66>T#?P*dEhTPBkxX*&y!Oo#0r-|O&`_B8s3TcMA+2Db+~J-s3_x56i@2J7b! zsUrTGbZcHUmX2(+@P&Qj+4aT%)I30jcNJ4hzo5tnyMlo`C#FX=Y~Ss}0&)f4)7^)M zXCfZB;mez}0g8+#Gx>``61-q~9q^JBD_99ItBsyG_**xn*$H_fnYG#M#LJ}f=3l(` zt)loody;kNpK*P+N_ILMSGsy4s}{7ot!|JLesU24KTwbEn)ID|N4Mra5<-Cyv{&7J zI*d+0jf!J3c(EYtnb6MO!hF>(!I75Aooiw@WGlbu;qE;49SN>vJVQajedW3%r-MKv zJdJ8)4adzEE{cC9(-^xDHO(#Ylr+rZT0=5_D?fkhih1hLN>Z2FX?fg^1}N zH%v4oW#X&|YjM`)MeYuk9qYgJSMiTaGchQeUb+kR|s0mFxyhl;J1XW~JbgnU&#%tO> z@5Rx8E38&F!xY)tMBrm~R_NwONtL4z{prW-+ApIg zc={G!vmwu2S2P-NY_O0XFRgD}4^_=w0y!yHQk*}ux{iBwjNM@Q;qW6!(fEa|9xX$pSb%&g7T3A$TRYw=U(*@2Hi?7l!=Eepn^XEP;(cRNh#@Nu!CQI zVag(JqFiD!n%}}e!MIaD2#;>{|D{vL{p;w6$$5R43yB!n(qi5LwQ#ijNOR#ss z%^=yxmD8{ei}^*e9j*Ze2-m>!ArWUK&X>i${rtL!i-X0Ii-+K8-XEuIHW}gSj7(kZ zkr$P`^=Ld8NY8m?K(de=xU-DSq1ZO0Rd%fhDfh4;EXpGQ_M5sS_7A3}BEOj_o;F`;?)5)6U zZLbm@w52m@@Q~0)PjDso9`4UG?533v=xYCOrXff$4bxGglMV6v1)(_Q}Vjz87&puSGIKDZ=9> zEI8-{y8z=TZD2_@^a0hW3~x0cn3W$d0!dc+WVjrKe1;VB#HzUy=t0y|KeviMlP*GF z1#ZYuP8YLrzHkC7(c0JgF=rf8#OERG6SVi0X~00ets|tN_x5K(JKt`Bv2e*P6Ym>b z?`>PL$Ft|pM_m1FK@GNj6W1v~M}Oq<&xfQ_;3@TlLLG5` z1V|_^V_XGJg@>6;;Cv)+1GQ=z2dhDMSkR@d<-V?o6zDit{z=V79i1U|8j*BcNuxIP zUAd<3Peq#?rWHv0#(-^Bb<5BOk-g++L}+&1$8HlFGXV}Cgw}g?XaYx@MH}n#p}2|h zk%z|YbQf-^R5FrX3C<21?OQo=S-g>CkM$kTyvy0JBg__UM8Y`RMcw4Xmw$BNe^T5B z*QmUt5ZyK+C%-F!pM9j)*{A~T=8JJ`3)<+A>|(>dc10t(X4*|lpr^~H!}e3Alp3C} z6YyL$T2%XZ{+8V3NXxxzGnQc@0cw&(eS&Vabak_si2KhG^SAyCVaB+ay@68F?1b#o z5n6#tdr-;n$*LJ)@2|vE7m#C`f_sOsaf6Kq%st$z?DWtgm_y5uDz+o)gI1Wsv6J4N zkN8`7llNdUY6_zZzHP~-fe$UJYX)5bk+$s7U{=;{d`UPHksalM4_}pIpWW%y$#R) z@9b5eNc_MlIoKSd`l6q|Cr29B; zWwwRFwhEWRQh~{xi=rOxb+6N6BkSg5yJdmnny#UV-8$p3``R1k*{W!VZzWYeF>BG0 z9#gc*cJB_J#@D=;k%4+E41SX({?x%i5zTSd1rJMh$yANN8W3ShMt%e%-egIi2LCpg z+WWHy1y5)GgM{90>s3a=eTwDIU8D)oYnaj8Z%rkqtNsF>mbwyl-VYx;>C4;MhZ7fB zBI(~Qsvm#tv0Ov#Qu*T2N{Gj@^>x?S32Bz|iY#o{qR10-QuybqsMWX}U9)bZY#e;_ zv0f)!Z|%)-@LEw@r6ju^(#TqZIk$1Y_T#?jevuA9j0!Pfua*pEFk4aIUOTD}EX2nT z?W$fUg4UxD?z&DzvLww5^-2 zfW?HBtq?%tR2)#D>KQkb^4jXuBVdt`95v4ny|GGY%l4YFigDinO&L_SpnD>U{>a{| zkSOJ4o4&KyPKDDhK+v9)qO?jBwfb)O!L3&!h~}(EN@UmscWR534;i4>w%82ieItUd^f%q8B@YoQq)mi1i0>XfBur+BAoM5K6mp ztX*h}nUu?McS~h$^CUQsBR9oDgCBRi8o{fseKY=AuZH1m=M1A;`7$Soex;>TO*Ktk zC1;KC*N;?tNBIQ$Zirqps%2w3QmJO`8&N^1pTn8rDqpkO5V!QCiCg)pGJHZZF+`PC z^g`)eNvedIa>$MfV1^X>PlZxap6_9`XxC z5sK;%m}A$W2=oGpzHBN;;8g;Sp#3JdmE*$bk1(Yob@cQyxbWn$4tk4PG%7fdoQ{k3 z5aJTCkB;DDNLHs7n)f>}t@A#$44{&P(h^o-_F#??|KuSUUCi)B>*Xj=8ZCEb%Q9w? zzYi2r)+p$yXM(;m*f;uZ|3qL-%{mcuGS%A5`2g6*dN9+G6imQrY*;olD&9f}@falM zDW%XH7f6v!2zJWo{x<%h;pfeEZZUL{7hD-?FZwJW) zTy9s2AFZ_Y2a-5g&ak}^WC-Ys>j@Fx7sWRZHIu~S6F1R;F8t_`Y|7Bu@?4(yJ%h19`NphLnu|TgG3XVlT$>5N9l^sg&qH|Y`hHl%M zELYoV&}Dd}QJ(&I8*J zHf*-oNY1?Q5(L3UeP0y}>k3eDdfk3La8hKK{|m{{u#NrQYVZYLILXp~yIy~JhU&S& zr{@BEzGo96BFI~Qd{s8S&n4sxpRfL8`Yy3-JKPC2L-%%?nrq7sQgVO9-ak4E<3IZW zu;G7D$rnyGq`gFr(sZN`;teER-CqCjYPO`DA#mlek@S!hEJRpVFD{T2oWcHMj2FoU zlY;YRs=XqV?!l82N$mZrRz-A)Eh3~BdipB`a{P|0-#!m^yht{7OJlrAS}^6}mx@kv zfsXzFi18(O`q4)Nv)~YI#Q&fWakj4tCYWa3AeZ$8J5B9O7)d{V#%@9MmU`@7GJcBY>E`3Qbio*kz17Y=NfbPWVKaDePnjTwH$MCuBYOH(nnY7wUqxiXVUf5b`WGJv&YL6lMgOBmcld(>;aqRq zNh@gPe-p#M1l~439vJ%8k3g zx2?K*a{QbeFo2IT5+p8#a5cxoW|n`%2uPSphQ$2*i}I(ViP$he4573D^#0G$`{(`t zB3JAm`q34>W~9dDs~-qjyAT$VUI7SFWJ_mO_)g}*(`NhPm{RbfhaYjx5U>E;iIQq-PznT089H!)`+O$_cbf@HJ2R%k! zV0g`FcrKs6OAW1beGAedFnhf!7oODcl1(D|1eaf+(`s_T#Fy+C8*C|r_h|_b0~&Ii zI07WV5-Ob=F7&6SaQag~H8;uy2k>Svq@25d!Ti6Q0{>2)&U<1qmy!ykA6>JkNCj))_l!~-v>*#UD#@%zpnf8+&V&>sl-x4`}z zLURAD6^aGxX1=^xedIK75^pdVb(d?Mk0?eQ;j*wGfi5J3aRqb@US)ID{Y zzm6w&ZNSE0CU|A|wBhQQ6gFTUZF7#w$^GItSvasp0u;tjL)K1xoc;CYUwkXe52w=n zV_s=ew2Y&FbUX`@I&I`n^~@%^5#=c(9Cv8M!hF2$=^eT5IIeN?N9)N3{vt^D-=^`u zP2L=rDe-zKgqFr&z#9HqoQUh--l@p#YG|x!8n|?MWpeZ(HgPT+>KxYSC9(XsFHmky zjuX)`hrtF58>x+duXvSYRlmq*j-v+S%Jj{V&sfw;@zR^Dx4xwXPS>9hA-r0)bIy$R zJ?2L}1G1lO?wrs!K&?QIKH$Ug-f>*>l%dpH;*XGu8lStIqPfWSVTZ2CssBaGKWJZs zp5K~&Y#CU zj>el0m_vS}kxAUgX#r&g)IQNEPNYa!tzHLl<7>pBuEVcpM=2q!MK#k)@c3FMi^4+k zXQ&}UFI2n+S4d6Qoc{qt0E&obpnuA;8*Pu`#t`vuoih>rU=-gcqTQZda{R7}cwgF_ zSbYv45lEfPp^I# zii&#vr)r~g4`9*(ln$ra}%Kt(lqfkMbhyoIgIl}?&RSTh)i%_U8lSuTAKb; z-+B`Yw-B>3e^0FB1{rmw;^6Iiod@A~1;tM>KH;kBs>_8B_DG>`xHxrt$o{?MqjY2% zVfeAx#sl**%%vx|u{+rU@o8fo5<5V=-{w8W)rbAJnj=uBd?$bJ0QBu4aW~@l%i%D ziz90$2HBS`IQwE+^i{xl5kzooQ8}* zNOg|IHs!eEQ9|oFI;+KJvpBP}ObB})paw5s=bV$HAWj_Lf#+s)ksxE^Eahius|VHD zX?T+baM?`94oE3Jssz;zCEO8WJ*wFPc*TQcMI4Bh0N5J9gPc$x8I7VpfFQTV$*grz zpnCM|%>(*BT{SR*g)K0TP%pC>$ zSn=cDZ1p6U;^kauol}8#>4yVY0v$Zi=Z&HIFnE+Pv1i8y+-75+zFX=Lw9R-d*Fy$# zqYaQ#29g9Rj_AX@X>Un0+eXGZLqU*aTPN582V$jyc}!s^xTu*0+p705JZP(J3!m(3 z0#6=C%5QIhm<*tsWLUc5$sWss>PiD5`Jm5wxW@{9K+k2L$27kTP8yz;#abG^Y^pfe zg~i%NWjf{~2|ajb+A82EW6*phyCwT_HIz%scodmNPFCM~cnN283nx?Dhv@yMKC%1p z4>2-kkG^}=-gQ`XZ#R323wcbed+`nh6$kqkMaQ>)xH9 zPpYhEypdq&@@;faY2`=*zdN*@y|TA%j3OaxESp>wtO5fpwfjn5@f*)tz^f^@gvAF| z-4sPz_vvR1OvhGyzg(Zt0QdFrtGVB|;|{A&Haw=qA_fI%Q46c ztdCjjUg=wpQMW&`$B%JVsosT}(P^e$BK`7?r-{j8COy!^z}1SK%B$b(lDv>{3cFQ3{riz<=GPxm=Na-comM57C;V5-wN+S2Hq8_fud-QQP(SO z10_xdF2_prid z;DREvDMHK`9hjC=g>g{C6jNbp`sK7xv|o`n)}IcdQ+tuujy?f3`P&s1vSv73XB@OH037Lg3*Y5 zX?2Q(`FNiCzU=ivx&`SS6o^z&S)u>feR=2Z%QXvKnUN=!*MG)sT_#0X(0wc7;k|z zDw#Lo*v!}=*b|_lZ|_e_aQDIZex~dig-j_2Y*TE|Bc+*Rr5Ut{rL2{~m@=AWhr9Y5v}Y7yyq+6bPx?bHZ$ZoB#7{F%e#aka`1OW!{R?%o3;8O+o1YY^)Q2axIe9TUc*tJ`-ew;`MW2n8X}fFY*zhEeP%zYcX+Aaaaqv8_=j|-G9(!t~ zpHQ(lP&Zgdd_l2ToII#z*9oyuc#H8^txU-t%NvXt-R7ZViKwe2b1dB0Pi1apLo&xE(UY|~xjiOynUJjXq1}=mssyHUjGq$Zl z9-VOV^nRTKzO3k|B9)X@qc5BrCi*`AfVc-X`lG?!(z8&N3gc*Uo9{I|oUmEkO|H3{ zp=B_9i;O+r8INeBR+fk12I=-1HyUX>x>i(+ON~I}Gxqu_>IPhhTZN%88_mqzspw?v zdXYBfxrjk^x4IZ6|fdt=%;hr55^Q zBr9vgZ(5By#d(AHh{rPWJYsH4`ncBJZ}VBCu+ZP?xL5}n;Pov%ksh7zkTpj5sg4)> z`DEEMPRo2GdY-LH*;)ED$G6y}zWDB9Chr1ktp#tO0D)SkMOhS)R#15>z*Vj&fmX)2 zVpcoM_ae{fw+1AsP<~OAvXJ<5~WmB&s4W zs`=zwkzg>JCMf;z6P8DGLviTIM%FNi?l3apjGUMJHj=@Jp+@cq%E$9?*T9Btj%d3< zRs1F&`PVbd71bWzio$1gfO2vw<3d&xdu}ruV1MKn(&+UEDd`TNCaW3BA5&uPlo1E+ zlghK$64V~@b`dCLPdtBC)RaCgQ~6-4_=8#$o! zLprtMihCXOGLZ0X!!wv3o~oFfio;glzaWIk19Z&03Q^>qB?`c(A56{@`s7q&+Whld zzY}bfS(zxF2IQKhHS?$Pcse{rV*0*Wkf%+qGej@#4?ZeH(Fc7 zu&OtiNurvVOxbDFDb{2|(W`N zEA{?ZxnSnpbU_Mn%;V$$f_xvuh#l;LUKskQANm|AWE%8tNDvBGqZ8^nMqYas{Y3q2 z)kXNynQbzQmCs?B`elv@tz^Vvc7aT7S70yS#qd$QT&XWMu5?0r*z%91-HkX4i|yBAc?SA4gTIe@#D;+aW& zp1Q{WwppsZdDn&5g@2zR`*U0@vOVW!YKvj*aDi^wYA2;J6e~6zwXo}WeliJx!}Vm` z_&MDx=cs18NcnVqPN!;?L4QA-!ot)Rd6AN~k2Y3W9eU z-6-B4Sw{-4JTuVmSw<`{#_Lc<#57-q-y-4oo{Ujr*8}}L$f+WWEgh!sH{t=UTqZKe zh|r(tLg~47^2lh)iMW2=HJGP_W@Og5k0eI!BVUH7@HyGFlMP0<8{PVw!i-9M3qZOf#T=c9L%jq^^8?dFa-zT$gpvAoI|z@Xg9& zsZx4+5}_|Uycro^zr7rve4(4>w|2MKc{YQdW0p9fS%nUS5Ycd ze6`*92ZS@C>Nas6i~JEQT5cb2j7`G>Ad^Dv8Y4M-;cc^$Hg2^{n!R)XjJuJ6qlAjL z0H?$Nn_yf-P}=5h!J)HNBR@5gcZ3ctMoHxSr>p0)aPl!@Eul3=w#sq5c2+$gsGqJ6 zhwBSUU8_73N|n~NkZpIN=hncVxgy!Q$F&2vD6UqwT$9pVPWAfrOnqo zxIe`qNv#|VVqp=EFE#QD^kQ~tK1Y_PpoI#XZ~7;*z_;w7tdlkHuk5gvhHk@$?j&?) zjn>s~hLS0Mt5Xk`$;8Q|l1cB$#%9a@?C28Wo_WFk(p<1aH_vjTZh+m`(T9*1hTegmz?H4&a)Q~#$r~vTnsL2 zE$yj-h2lFpyXLnvw-cXWLW#HVGNyc>3>+Fehc_nHv7e(oBh>YuFa75*-icn&=!qdD z6vzeOW=UqAqN8hqS)9(2)f!O5n8|G2Q-Zm*c)ENi`|TOQEnzK&WJChu*-6+U(XjU*&Os8GfEIP zN0W9yYt<4mnwhtH3GG5RSu-#3^pe}6(L&dn^bUx(9XEqbLlMT#cZJ3IY4e1!hXVCI z=@)&GM5oW;@EJQRRF{y+M>2C#xZPe=?>f3VKS(?;$or^!%pUbU$L-`;JSt(yENLh- zDB0`67in$_(PMG`2IUq_KV^`rz#}Jqk+cz|1Rcx<&BZDzsvM-{Z>E#q5&Z^AR+2}YhK?JaQ^tcHH*!1*Z zIXSSY3OxnV5U3#M<$@^wz~F$qp#ffRqTm7#+y84i=;2)qZfsLGVOC> z`=5|<68!v-rxT=zsNbMhm^+`xmY?=J{9ya`IO6c85{84Fwf_+MfuQ zKAZ8Ksv_5yehaXwzJNblvR!A*hxL|!K$Zr1s_AE5<1<}HMaF^opM(iS)c_*{k&!c~ zCVqdO2n;wMOTztTT4GV;36>b#;OE@AL*#j1>oIo6Ts>K`Vvl`E7Qcyjxr(n=fY5Rp z#^*GfHg*}KSkaSjOG5kbpfo@R&;lc%y+!R@l_VJFAEb>Q-QuGxn_Mu?jgeyBHs|4W zMJ`R&@%sG+oE!G%BrR+HfF+g7AOaA7lJKxUnPhD~$X*r1BYDbrqABKm=D*k%l;0%A z)%#{*ByADP!?(EEwM!(dYYq5xudK5!R%Y( zs=;A$X{CI0&YhrnU)-upyiV^9E03xk(=d*1&UI^M?%9?N$&|=hk5p@B7K}2Cxq|)X zzYB_wXZQt$TuX#KPEUa>6N`sJiM>+nLTfFu*6P~kSnhx~-XX#^Vb|&~#s0vIZfq|Q zQVb4R`@ttbADBsO8cOUu_UaD}Bn{(0JBka+F^62XZa=Z8)_F#fun~gsS*)?5ww;3i z7#%_RJq*(`7*)rO@MXfr(r1&6p-6S+3*swo(-C%q$`kY~gr6FdyVd+}n;7KL1-umE zd&pjK18n}6(iVNHr~;y9F}hJ=A2|0PNvQ1{qPI8>#Tu22&mNyq0IbRaipn_@`igTS zPG@8PcnWqt?sR*Cmr_(lMvvY9PX6b`fXsF>x&OSoF}Y}%T&CL#J&UX+s}=nlneDZFT-h>0 zYLrP&`~Q#p@2MSOT|g-*M~~!@#L3RcM|5fI9o{$t90JUK<2DTiXb*$;8bp$@&pTSS`JbmR;@zUN;u`q2L^y2EvDjMNJ&)kFrUfG^X(!UzSl zt_w6o2GIK&;E${>)QGZ^7%1ncYm|ND%!XVI1D2&?(%nvHaq@OwOw;X#uYyQSCZbl0P);9B^Cw70GKdoCBEJ z@Crj#2uYA-zhz&Fw`*{jRVM1>AF>K~;Ts$(Iu~gPnEMb%!c_lKcr+N7e@raDY69@? z3zNYAFP+4uS*=*TPh)<{7j)o;-xQIaK{iXez(T}+jOOD>Rgf6PJB34Ti|!iO#u&b0 zVOA23XjEo29Ok}WO(P+(!1MsplJC;zmAD%^(J(f>s)qj|mqOIo*0C>VdL2>WZ)Sfb z4P@--TctaGa3U?tovMtY2}iPTurP=J#@;xGXaE({vMLH-d*k&9)i4)?u>muH5B_*is@Fq`|F(d}zL{$I~70`CBwaV=9+ z(iBjU)Ev|JTJP5E86g+VqK#2PqcjQPtW)?>b79YN$SniN|J!&Fx}Z^5-25_Rh>X}Y zDqhk)B8G3i={vUeH~mhO-2r0LFusqTN`BuTak+XXmwc@*`Jau)1VjzSlmR=*;y@vSk zePhmFa%U{=iFkQ98H+835`=v_7c=j;rGtdzVHb*Z`0)pCa>{8Oal%otL6j&aXSw4C zr^s7}{M_c}U39;$Zr_aKH(@*XP;lJXEx#KYMl5nntJwiLXBr zmw$Td6YzfxVg3Od&8e@6yJ9}YW*h;{TV(timi)3kgbq>-Lz1zLMEsM-EBdV=i&|aQ z6nXlDA?N@8x&i#y8KD}X|0-)yi%mc1>~?^dDZ>0ZcRyjX99sMDzja>wW@rp|A1=1g*i;(7gKSf)>VE)OGxm7CqKV=dw)eN?Rq~r@q~9A z2mO4o8QE8vm}`Bnz6HSeA?Fhjk~Zy3PJ%DdmQU_oV6*_!(3$7E@?Y0}Z=w!goc1VJ zZgS#py1V=@4|5V1a{}%d7AeCUug>=My_7 z_Rl5YzQMmKLIySvAp;1J0S^lY`gi;o z7(J%u`ZIQn4U8PepHcNps%)Nfnt;FljgSEY`WqqRnOMZ^!a`T?JP0=2bbK=N>yrR$esP#bS{pSXihEdCt=(E_ni(#xNx3T6yaUc+;LgVy82DNajUjsgq@8`POj!CEz3! zm{EVYa$-c70U{sH`Za)EF(_hR|HcH!p8kX(47F$&Zhv6Z|Hz>v`oP1KmL)UPdk%-|7K$9mSns+PpC4Gt8Y3?zze^`p~yX=!K^6Bs9i;aRpe&>!L3?_^g@^omZQgDxDsMDK#AFv{g#g-PtmMjY(!voxc0*FH#^ zptho9%v;=}s3kkc<^KbMQoTdS4%4nltLW*K5o+3GI);gY)mZ<*YLt0Je1g8f7JiVl zfAPDt9dUL>>xU{BM%YczeH7GL1B_AP0R5T!U}c1@samIYC%n0qNGNMz|M;J64eb^xfCvm@m2O#IV}xJ%VJH9L%`#Ehi2001I^A zj>LSbun&tod-}WVGqAHg(>2u2O9~4Q#%B(hxE()Rw!m&lq|r2gWsadcgb{JlpH^j0 zdT0DT8SaV?ud{gQ{>s5+neqm~TSvxx@&#_`xpla@14D`pi|M?wF1z#mIDbvezPPCg zUY*l==?q8W8VkzchD~&zRqa$9NTUO#E9_%FbPvTAhfY7Sfegc*3o?yDJE=QylONaN zH>!qwUnr)YSx&icO_t5858H4a8s}{qQa&G6ng$RCh+%M*ysAj>OYB-d(_@?YiCac^ zCxH=1gS1F*2N^j+1w0+?04()9&T0aubbyfJj!YzoHAginl{I~^gG+^sf6-BJ0%n1z zl~qIE3`RStZA?T_Q}Xt^g|mfSo$ot(*;EyQ4RpACe?Gq9BxGprvI4?Vu!d2tOkB_B z=9EtT6#qhK6Z3#mzoEO^*JAJIOKM^47QEGbs}{qK8;7GTrnc=xuK6 zEwcjB?~oUA7vsWrHPA&KPURvP;kJSWy`V z@h38Ib5DU)rt2EltJ3?j{qhW!!EsAp{d#JPFwo`%mz{IMputg%WBUna37Ayxof}Ae z4qsrwOPoZupcKh^*?9dn;;^RyX4Clifn3r^yW}oVsn<9qAA4(tSTD191?aPL-S>g) zyl`lt!;OZtS_xM^Wd0$vg?O4gk3JL=l64HF7$=nv>yXb&QHlJP=Zj_c!Sx1}kJz#; ziRZ}3mAUukgUV*;Gg@a#MY%*_GslBu?N446K`0767D{cC4t>nzGp<0n zFa7elVC7>l5fbWt|H~xA@hrnA<_DjC-AqVs?U7ej$9w6Mr#aCs&A|{;oz+B}4)iYE zfk2q;M`AKiGY<@-hLd7dAPJcr50wr<#CTp?m{OQ}`~5O)-Qg^cH`el%ZLhj{mQ?Xo z_?JT+%tQ3v&k+8Xc&Njtf~Fb@SN@biGXi3?(|u&BSe&epr&7>Y85*zphrPY2(LBVZ zy0^kYRgBJpSq@nSDHTE3WhW6A#Y!Z7=CNwLml^8PW+!m(S=8dDPz0q=~OHOJ8{ch=LQ zEST7$IqCnNu|^;OMukeqUpwn||uDULhcTt1OwY1J?QZ9*8& zCvHQmIjD^S7;7)m40P?ODA0^`aic!&7>kLaCW~1)Mjc7y4CZZ6W?>e$qX|e&UKueu zKn8x0ArPZ51KXv(zvAX}bTgjm;O_x zqM)*gG4tZ4daD#BYgxBYcL}Sh;TP{|1q%`|914ebQ}0Vb@9yqBhLtv}G-~nuLg>sMZm}qsQpz2(?^AHv1j6}s?6WdhDTH>2EGyITGd(AgT7W4c zOGLqjpVL6#J|z@M?xgMe2b!L%hWpp9YE?P%%kj4T$M*s2W@IlSFi63el@lm(1%Tm+ za5;S57crAFf*sI#EW`Xlg@F65tmz&T^$V(YMJA5;fIf~gI!q0O+2Q<<74t|rX6#nX zIKOsSG^lIkO>fOr4fsG%Dd9z?#%dSez2eL5!O}-Cukuw4p84-{=f2)0s(ir`1h$iR z6-5{eoC}Aiq#jMCZp3TVqJ#Y5q7)e26C??SOreaAO7-;g@-Ew9z8^7eGO%`=vo#)QRo z%ks3mG+tOJGbX5VCqPdk-8FbK9ahlQTYi%Q{D~{RCj3IH)5&gP|1n~ z8%o^gNX!BThBzC08{I|}U>AnSrgl(*P*`SuCH*2u!9mnZmlGEahAJWRPN*@^>FKQ6 zH##?+?}ElFgLNWm0B`djdND!j!lZ>&3Tp4aFOk@NY)hyB6SU76`RVo>j#PuRa~Dl` zVkM1#DMyOq7x2z|p!D$bVpE$UW#r?&a-Tz+>&$PY7QpE)*~84y#FzXUyZRE zE4n8+O3EutSM2{gC=e$9TLndtO6WL0j(-c*x0eF2F+VVemB@Rd0k;}pMv-}?mcy9tI@D57yR5OsVJ=)gdD;D*Q4~AvVilzv z>xUH1kdPSpRzaYizt3@sv6ek$^G4UwSZj*+({m-l1+?mjgDDpE;MUTSeKgV5muTAx zgTmlNwTnsPj5IkAeNKQmr+7&8Adu=CG5d5O%_J{QEZ=i#9XSDFkFCyWB^QE;{p38e z+4Ox}vY)Ea(MOqdL9q7o#FvryZwiwWQF@||iC?eMeLZ-TMD;D2ZmPO*`1>@;-%^Uu z2BJ|2Q%Z|@G8n#b54{Hp!Z7~<(P8{R z5Q8U0)lN+@J~ilqZmh>`_&GNh%zXJyE*O%-G))@yyM5#=ZiW_5$L7QfK@DFtX9ync z!oZH9NT+F`W=IoBbA|rlqPHk`WzutyLpzsJ$N=*ARgG|!qB~PMgHT%e9nw&Vw>S{z zX@E%-f1c7us(v{384icDg;3-{0gDt1iy=2(Crh+H4XuxYy;9lU1p&q`+E@OPAojM( zx(G9OJJ)N3vbud2hAnGRNz`&=XXSKYd?YJpm=m%97a~5K{xdN-im$e-Ad3oDk5$Q1 zpPZK5EA@etHn1$TD4i`iz4KM{l(G8MqEF#jtKf+N&L?@3DGrJ61iK24sV1K3Z z6F)^%p9FSgC5pEtZY^@lf*mEkx+cHNJTzi`XX%EB>E(sMtJjWuK;s-HddoRb4lFwEl(Ned@gTt9(5T?bopWCz&fOO&d8U=BviM8wjzYDrq6xjveNFMv;7pL?^ z`NMrh?Zmd0><_F#T}l6CJ#uBq&$(+Q_ef1V*fx%eF*X%Pugj8rEhS9@w<_B|wFH{P z?8yFFUUpg*oY3$<8Jpab`izg_DPCnIlFLd{C8Osu0H~5j`xX5TQ#_)k|m!602t0LaSx1AI8 zX@((4Omie|_&8L$AYBbwA>u&pe2|;vecgNBIF&#5O8Q zU;=GqTO95cB?8D$`eOm|H|j$O7Hf?vlOgKMNXiI@VYk#&P6A$SJPHi{T1BN3>#hqQ zoVQ)*S*mk`I+i#{6JZreoRB-uN;BZ3SBk9`&icv&^-503J}6{m`$c#*t)!j3y};Z#j01_Z zB)(a`LmN;@K*VP>(%UjV#|#!vRWz-PU)HE~^vcSU^_lh&A0O=*3_vxng3x-xTbJ4$ z&Je>ViJ01zjR`RDG!T0IrZ7j#t~l+hrER+W-Yh%DZH(m^_{O?c;*b@4x(O z(%C3@*%|yi_KN;0)wuio!ypA-^{4q=TX1#meLm?~T+XjqzRL5pA-kQ<%_%O6&5rA1 z7Nn;!bNS1C>^}OkaEaIct0Y%*chQFzuXOZv$F_(X>?>Kcgr*dKMtu@;OzWu!=p5;GNc9%V|c*{OMrcPaN{5!>b{ zXbmwKwf~j)(z$Kd7W;R>wrLc-B`1+q&rmilb8kZ54=EMY644Lh_EX(5$)1wMfwk~0 zo}B{8YAy7>SZV5HyK7HaC}GO-FErO-Vohha;R%ZND>xLS13JH|@PJM2C?AMP7pwx> zPUxof_ElAOJ`*=P-z;4hh;8kNWez;2+?u`f;oEP5W`UO4amWW4?u>icNoOBdD7m!1 zuxZCk_lu{@(1SwoXSdUBV>RSW$O%*TCg?#AU(qx-RaMLp!#Cc-c^138&~t^mCl{Wj%%Zr>T=z#z%SK-$(XJ62^T4w%ZkCQrCM0+)FFnnr}2J=z2f%g zGk=>^aE|AfYf3}DvA~=JIS=ngafmjVk zQR^?d5_WQ$W^u10OyGZYz`{e|T|KdCu|Y8YvvT%=Bu8Wk&6A{=O;eN{I~wJ}TA70V zjM*L=N{*Rb(jR_^R_T8D%yJKS$crbz8GWkqWWptg!JI^#xMXs5fA{-ENeevgvfS?R z-DJ6*3}t`^8wAP!USPi2RwT_~3~NGVJDnVtK*-CvpOM-%$#E{g_3IAnY2Ac_qVuXT8 zJkp_4gw-&E+69MLwaZ_$@Rl$n_RGd^*rj_y+=#zb>%74KUUP1?ZMsig4oi!ZmE&_$ zRGRs6v>DY0)SPqrzAz?|6F3wjP}s6|ax5GFeaT8MeMNW*xP?0kdT-Q;f_>+$u>iz2 zv(Y1Be^MaAzsUR2G6HY%2k`EfBxvpMBOc5YdX<^4BkBtuZ=avMEll^a3{4?>EuONJ zkXVs6_|wazgq>K`foM8*9YI=q)gwjVtLD6#z%nv06$ruuR*3S7D{A+xA~gY)uOW7D zU?phOiR}-I*1Rhj=@NOtIRvHTqq0t%Gk_y}Ma*s@JXU9Uq!EXIg+vb{t3wj+HaTMBpaj13O zA$lG)F4Y9LtOCQnSTMVV4qk#kjj2}5f4U7P09iVyIPqqVRXxPG#NRa+s(yTu0^fpH z7k#$Hi+*{)g&Aq6Waic*rN45OnmAOV? zD~%Q=LHH&#ZKjZ=Gi=z)<-Pze!<$16R&Q@{eEHX1@U2R_IR{H`HBf<(F!kC^p58>6 zYL5)3lcsZl^^&{BpTDb~E2_r~WrC7j)32784y@j;srt8Jo^p{=@2c_D7*yd|?i1A6 zt~0#p)1Po%c|8DOM9w$SnO*Pl%wVS*?9x#w%MV+^;gn*~5y{PGRELC5aY=5@AWn`b z^0_CKO-TEzGw_OxPt`McEl8287mO}Wbior3LeP$}42lltDhK=ysLD~gH>}@Mx0BTN zEuoZk7X;fHJ`Kj1Bb(wZvZ5!e5?N~LPrNgY3@k=clz>9iS=w{?y}wbB%W zk>igk;*@1&*gpY9V6sewG2H!a$G%@AF*p9>p#bOW6jgSF(#QE`@?`x1HW9v<0Z7b( za}EC%m}xt+-4pB;GdfJ3%ai8~yC9rDd3{lgTe-}yZ%f}eMk!9RFOO1&MDv-dxv7Y1 z?s59gM-VNNoq+x@O=P0z(;6|KO@L|hd)Q+||r*JV^3*-8#Jstjs ztBGrl%J|+ij?{vP(Q~0xzV}Gke(vkI@?>?33z0h6WuuCxT2*uu&n9Xz z^kr*xd@b{H<}Oh*`W$&S4jD-PBCr#@y}uaHIVV06gRD^lz;jhNdEC=i?jl-I?%Ij) zhU^{$*xX~l#KIOK6({9Yl0qr!6=v23so~m-Zb@T?v1A3_45seMn(t7t0g=}BzE~4l zP{0Hb)sdGbNh>jYgI=64HQ`BKO+(XjtYgDkvf%b5?~U4Bm~^F6RnBqEh;8KY>Hf{7 zTu76;1$oMg4-)9h{hMd%ZMj0=9;0xS*@vh=;u)ku7y|OxsWCL(mT1gjbYYPlB8d;OhZi+V zyk|%>?P8!1ZrSq)g);7}eu@V8{`MPKTpVkBFglBNcv8lvk+LS6D(RC9ewG{?vx~iU z6!mP{wh1UschCcN5f|cK08H^Ge77Cw$48N&$N8z1?m*kIovGG>7ZX)UZS!tBM_hkY zA(`Cuu6!&dBjn=74M&_fSfYd{aV6qj`}*Us#zKzuhs+S(3$jNgL>IfKni8>VikE1# zWXKnTA9!WSClAy}(Xo}2ugE_&ym%3SDmYjT0%m(QQVa}Xhl9s0x~IDx#f2k(OwVmK zf2}YNjYw&5RcR<&k+ZoHmDlIA!#;9z!uL3I@uV+wLLwhWOjALo0+(x$_h%6{Dv=8an-}$?a(PaanS{2Ypky_!R>Jtr%8iRo z1MvhPZ{)n+Ye6NXx`)+}Br!349O;WXJutsFygNZ2a)VLA(|r53ba(pRCh3uK8Gli9JOwGN}e_p>*#hR#|{;cFz5OB(zZ1nQOrXh zAm(g^ygvqI>*w%mQY!3KX)JU{Dv?u9i~z8s7afwy<%4Aqcm_+Q4mqo|s~9fIJkaI` zoVaD-#@ZdxLDc|U+yQ4CVFE$fmX0!%OaNqZeY)?7@p0?5l>X&|AVrCp*oQRkD^&=Z zwzaKF`2j(n!9jrPD#$;LStfx?svu_~AqDc|SbWmQE^Z4uw1Bl#km4I|N@J=x=XcFs zFY<@Z@6W&T(;X(bq-L$%Og9L7_stSRPqdI-Q6m#2luZj!=R_e{X=_0`x;q3$CAEE0 zF=d~+a^aVFd?QTgF65uCsR(nAKfLR}Z~aCic6gBcp^+1m4-aX28Ga>?HE5j~p{MVD zz?i;Hd(yq}5iU-OI$YS#f@c6CqEd%p#D&#!WKBB`!jbs$#s^4^4}_7pTXgLD zOuAMFw*ZkJZ{tnv)~j+;5X$n12paEFv^Xc03_HM?L z9vFYs_m}LjYgOZoDh$PTVxI^4?&3-pbtXKoeD5ZBn~{Y(Q2;a((^rbEQ-yMe*trGa zsuWriGT@dgi>V{gezkM_Dy%?WiIip<#Yn$683TD3F@}~jwis4P`}wDO)U(B;I+!;V z6Zj5&cqZX3JrDwBm&Lr!QjCZ}5V-oDLZx|@(Vq94`UcUs%=O$}H`%}YU#Y~WzBNmt zE#N+ggv?2YF@ivl2`Oes6l}V45jIuEn%rAecuY9CkJUvcxBg%lFh2a}8%~&y zm4Wu&1$)1kPIobs!$B%Az)a-`RroKHS$3Q1vw;=2?tBJeRp(;9<0>X4a`a68WTH|z z_6$^tw+274)876LCl||oJ_>$1b~^)y?#UzV0@LflJ>MxXu{6~5I>jU}``+e#U=`zw zYWjwoil|Ga~UXC=M2UVPoB9tBhQ4o79m|~uIE%u#RuzQJ( z7qVHW<%0UxK6!=DHly~G3mHp}q z=7P+VE6Kvke%xbO!ZAH!kq!KH+%Y9=Prfp5y%RYbqgzm)j z>f^#0xA#}{#d<7-jLr$)xFUNALeXQE;2;T*9e-asbXy_jOH86}WKi4zcQf4&?iXb1 z{E5b=2aO*lo>xL+0hZ7IfKX)mvID{{xypWxoZKSA$9*00BXOKY|78Dy83& z26e!&0LiVv5~vwKET)_vmf@}S2$t3XcjacJ{5y9YR9%Vk4ia4h6e1_$ShHfy-R)xT_(q)ROU!j;=eIn9R%T@LPmD;&ns`@u- z_Cx)YRvTTfSNKs-u4<0YvBM8jJB-GVFJaEPhn%)Tu&Cb9wN~YVb=ztb@PvZ^i&Duh zgti0O$de@(ox~0)o=qjDoT^Spjl_NJZt5+aZ@2)l76$ng!8%>Wr{$GTU_T3pJU252#6nZAX@8!guX%b9Ee6_QFL1V58=9zq zx;QR@Wp#YfKxD3;lHrQ%sO~%7%ueH_JBmop!idE{K(Mlfj(XI%i7%5%>g6rIZtsZ5 z1i5w<;e@{N_xAx@4f9QIB1Jn~!66z=By>Ml^tzTycq3Kz=8KZs2>hxiEj9YM!o$7 zJPuX#uJD;xwm3BdO z08ku%CG{u}>u?J>-9@Q85sK+uB@6-#MFG7dmDq7Dj@M6270{x)N0FyZXalY(dqvy8 zfNaO-()YTHl3*;`qt);H#XohUB)W9E{4iaxYz3)VlQpQ$fTMulV`CSHAk47Py`G%+ z;#iY@DBp0Lkjq)2uXE7#Q)kEYKgiGU8|-%eK`|t78;Us{@=%-OC*gwrNhnP?w~ib{ z2;^uPD#=By)}12q%ajdLI|U$10}gg)>%CCmiUCbrS8BXxIP|}oycb`)#5Q=^wz|K> z+hIXVQNRSM&HzK}Yc0N9;wrZEq@w|~RHG|IFPJdW>r}Qa{H8HM@$02Y9O-O?><-W@ zGSGX$x#BJQ^nQoez7AcNI7Z0V{QEXn)0Nh0DONv}|WEHvo<0AOiEJ&)vp z)041VsfRBV3U9X#?7q(mfeuP^TrEbo0Hu9J3IIU6q?dJ1JjMcHWPk;R!MFFAJctBc z8L`x-(0r{tO&`s+_OK$0O&t41)qBbZ0iJN*c2L(2fU*Rv^7~5%CKB*zEj5?ga^?^Q z4S*Y^zHF5&!_P%z!jA9m4?F^!Txs#>HzUjEIetHznpRLEWm~o@%v$%XUA(`T14_Qg zS_;PxXaQ?TXw_)$>WnG>!ZmSM5$3#IXq%@CMwhtDm00u-Au1`nU>H=oV zrr*e!C>F5k3dMq{2O6$9Vw5*BFj3-*eW4|hw9u^kOb!BrPfi#>(9l zFe#X2bgRwl*@8(ZTv#bL+%?CDga&ld)`yXJ<|GAv?W`j_2m<=(Wc)w?00x8kAA+dR z8)1{;2CN;%#}$~?uSBdYUV3lZSx~@Qw?Wn8ya2+OiW+UxADLzQ`^=@Cfhjm(z+E9^ z4}lF-I-@PCve>a!5y`vLbg=~F+0XN)7Sa@8ise$Npq*~&^8kf#022s$wfKVEDu9j` zqim!6sHIfAIlFwvpgnf6f?H#+B(Pwk5i*rj^zA85Vydz20Yh%qSKs-oCaP6yYqEiT z;9`sjT;StW6$lL?hTEapMj$|RQG5@AsQ$&kOe;O5xarTDbN zLD1?I0pt0Ls;Nj(PedxOtj5_*?t_>MXp{hGDZFRi0H%~Lg)Qd4v>D%?_RB?Av=AXqv@-{V5gbK;nN2CK00N>tYQ5!~gD9n{9-H=*nNq^ORX1yxWm=hsNu1V~tTm36vfXN^RM_wi zyouL|ku<`x{(qYxElQ;o-J{+mG^t}|hBZa|#0Zk`gKNXGf*wd}E`nIMjgQQw;kr~v z*GvRSh@k)ka`O_7$gFP#+g|dnCsR|fQ91r$OL0iT9veHbmuL%H-rlf|+aL&VT5j8Y z;1E`@VwtrQN1C^AXRvp7%je8I3?FKa$aIqX!NX zQDx;b90ijvbZX)3a4dqDWM8#t9S{J!OP2)<5Qc*;-;o{KBo~@hdJfm-S$eWev{N;Y z6FXd>R@!j?08n*XBJRJLC9#!X(1P|DQ23cWc-+}gv;{^Eh^f#O7U@C}=OSd4uXaSJ z1poR6T_Oe8w6Kc^K0cR7+wCOO$(K7v2@rVz(h3C{J<#lc7)&o#0GNQQ?0uXj)+? zL);hQCfETBxXS9oj?$K?=#4y>{&yOptT2}QEgtZi)Ru??N1Hw|G8s(fq*tsOcZCaW z*2WIvU*pl;x?qh1%0*}2m%u%(6oIw4l8j_s)mHSyx0F4Wg^p2IsWGYY5V{$6ToSxH zB`Ks--YCC~XnB}Ag>r;4-dB~cCt9KLF~Uh=>LJ$b+(#)*dfC&5h*(vurz6=hq;(73 zZykJ))3^gzwVd;rT!{_?4F>go{{YMpPzqui!c2<+9W@VXJltq}dG}TzP;%8ege=FCJnkF?Y$y@)7p zxW+iDln^c!lMXPycvlsK3rsco76^hiKh3oPAWM@(-G+VTp|N8NK=o*DRQgk6Gru<0 zPLT(hm8pg4n)Za?w;9#Gk20?qlxc>}FxI1AX@IU!g@Wox@FsaP9Kr`bjJ zn06(zLa#Z`ytk*(KO={4f9j3ImAl)c`UB(zl3UeHOs&P`0 zeFYU%f*AuSQ};A%o%&Qvg~ixuR%PwCN)!J8=FYkmHn^vX59Uw=#M%xS=HMjRTFwkQ zDWd$tmzncDrI?Kjl}|STLM@an^5mx!%hm>rY+z9k#5JjJoYT+pS|WteA?uSg zqSQf(cGOnWo--bhj4WtF#bECn8!Oe_xZw8mGfta_c0AmrZRlZT*KCuW#4S=;q#&XQ zuvozD$G4>}ePyNwf)0mbHC*4UwWS3XowZRPFC0o}gJw!mrSyJ@y9%&G=ROhUwA~Wp zW=v$T0dW*0jX=rHF^lTTmjQGFh@zJ43pnN_U_T{Obq#7*#k0PT;eKUA0NV%pS?@3G zuy7#KJ;(Vf^pyjg9pJBco#SvcnnI}lkdt1q7DREop-}4RCKXxm+IaL;ATq+W+r}Vw zi*8lG-yLAEa*ux%;4$C@aBoUqEdF2;3M@{4k+}~fEWo`c@5G^4Qato~*^J_-P0OL?DriVxkQ@3SD}>i6LHS8WJ3!q|#p5~dKnjI-Xv$YQAz4G1)GBXv zl&Op^8WhGV)`P{%-I6O!u|Ii~0T6l)y-pbD#HgQCSkwX?k| zn}?wIJ~00rlOc7O^=7*!|+fVM^Ewv zfCU9BW}9!?Q7kHOIXJw)kwAv9E-O!W7cBjSpjwIM=r6PBvtU|rz zmSJpmVZvmHge^d0xn8snC-SU>z=wqF&cFuj1rLEbWvXdel%=FphTCvjUl>M)P{^%C zPOmsSN{Q#-DWi%%zf9;0iIXAJ?-<2d$`!}XH_3@j+En9_J|i3>Ln6>Ooe$(GfiUdg zFPB6g@(6Xu*BIE8h z`4XYUqLpd!N0EYPAB^l2RX8Fl(ksN666{QmM6niK#6iIR*z{?U3Bj`E6N*cE91B;>NhktPOl+*HL6*JU7L!Xyhs1uZEnbU&EH0E8;lhVWYn6}C`AfI3#+$DuZ` zMMq0AOZbio$%Z$`=3!8jH>X}9m6r5CBCa#TE>W;~0QhN+E4sfrY3 ztYc6XVOawJeh;_tx5Fo5Hx;pEizw>=R9^|Z40Zpqq>_0YXcmJ;$kJ_8U<(-W#xUTtD=mo!mAT7h%sHE$2}Bxhc0L8J9;U) zq!1qLUnL1++<@3qq%72QIpSSlhyYewr-8YS@ER)%7rM5hK%`5X_!_3-3K|5fzevy@ zVjq@P9WXuN3GXq&1b+O?kN+7S&dYK@#3UujokB!sjMEiYR18wk-#TNY|1L8T8R zaE2~3n*pkCUFFt!OG>KN>Fz`nI|xHf$c27Cll2XlEzEh~zRV!pSVYaGV=MTD0%HZi zwHK9Fv?WAxDX4O~*QG2pDmz1dJHxsx%SD{yl2JP-mV5B&^D8QV0fm)bWu&m3WaHGQ z^Zab24B!i6wzaYjBe0Yp28PEaxH*=)<6|n?aA0NzdO^YHU`!btx@&p3d1{Lt1srGk zi?WD-;JbT_^_dWHp?3S3rfj3}6l$HCv(y;?7OxV5&1ta458>LV3JZVstj&dM87K)x z>l}j>w)?R;0ePnuUmVu^SUTiqG*0FHPQmPTOZiH$!k`e4wHg0D%+fb%TD0bOV-cI7{;-TlM8Wyn?hWf$BzJyW;kH ziI`U2u~gLt^TXW!7mIPxZ3V-F^j3B?#JdY*6)(v6iG+C+iptX}54ucJRY5EJc7GxqqEQOBbsWO1L`k$_@e|u*FJrE&w@HcquQB5qT(_95>^G+6nfd(W2lQ zIt>6`%N2U54bW^cY*N4~s9vI!G6*S3`U#86rUq@=jykHR=3m1PXC-zhH(g>Zwk6T3 z%GC*7VfQc3-!|G-aOjxq@CrykRYifa0nwK$ZXDQEyIk9Abz^iL*oubQ-Asl?VRj2x zrVMi8g0{NoatV;2uQkrIdW#LS0wvK>t z+5>YiF$kxdu1=%VL|hOJ*;sd&?%Ks_pAnb|G%l^(3(uGpG@@ALuMQxgl8Qg66XFxI zmjwowr+>oKidMp?zS)`3$`KKP=r{!T{|lF`tRYQ8m+gi)?i@E zKo@*7zaE6^Us8+Mtfss`TFjY7j%km?-x-bgd;b6fbnTFL2P%lGm1nrQCaQ$@mu{A+ zMIh$`tM4kxXb29^D}NrQY_MprKJ~;7g-fWcRgLQNHyH6PGOZmsPkEByFf9jf6={7R zMgWil@~^NPR1B~sP$%$5g2ySKqN7qI=okn=CcL3TP7p`tW0PbvFayG6iHsN zU*P3XTJ$0{ywKRli$Dges6InVgj>-juyiGdjJhJI(kN6Q$+>P|2Ek(sg6}B@f*H{T zM4drwIb1|3o7W@hi%Lp`u9ho@1=@!!{2-XgAp;u4f z1}^cuCf$l+K&Kk4lN<(WGvk%~KQG05w$Cvrww51yT)o9s1`S9RkdI?e>N? z5UW};tTkbTpmhqp2h1!JAXs7-S`VM#Mwt2QDOu^SX+_nwS!~18Uf!8=5Yo1|j6v<`dQQx0v3`klxF<%{h0e~aJE>y+B zWjdE$&;+2g268=Zv$NaO0SQKc<{-7;dP)bee)Ax#6}1U?(HB~BRy=0&cgI1v0004i zIQ1K3E$LZ0SNSf9LQz;AX&a!D^(vhz$3g?jaK=)rmDYfWmttj36%Ld-%hrStC0Ah+ z<|VX=dIVLP!Xu9G7Hp~R3Aiu%whv@P&}I4)YFD;oB*#&mlDARSZN^ARuz^C zdm=#ert&?3PW`0AoSZ9=uC#l?lMpnZwT!KQQnOjqqcvHoAUTs12MAIt~BhK!kfLx!+`pmetNgHA}LSw?*@We^MSo=$%-hD)|yVmk|- z-ammzn=G!dItR-D8*r3f0bJmRq-F%f0(Z6SN(s42wwqgW=hT?eth1~Y%K<{7DX2yo z^V`rE1r(1nyr@v9nhH^076~s0C0As|v}hX$nVoD+M%wE|t>pR4{ub2mlp# z$_00Tm_RfgJ!|>U{*`eo#jeY4ZE;oT?1&rAwoEk*_{oR0nTOM`#CJkkQ|4~u=Kn-&W2Dve=%p$ZbSYK^Zi@L0eDn75l;UB?W} zxHG)@^xU}GWrFI`uvSlam{lUmtlxJb=h4amYUm(UjDi}trZ7AAh%Z~vX|Y)q_>F)7 zK?P`4Z0n=D=|E6O7VqWraaT>2KT0h`=d|o(jQ})9g+>*{$7I2{MGe=kG01;sd2voM zC{?{&-$a?R;8q>8ymQoj7>Fj(ngiXJm<2}A+OI9Kgk*Y}Dk}+#@dGG?#M=mT2ftt8 zS{07cx~t-^Qt=Q92v=t~$Ehr>MoMh-bjyn(qiCfZnJVQUbkRb^o-kU@P2Hq;_Ds=~07k9x*?Lfn~htChe7!v%B9>cw|# zy%1tA6k7&BRE9X#C5OWDp{lCQL##@}3aNV&3d7k1xaWqZLagU`esvS7V$?7O*#j>H z1yo@=c(c-}bi>wS9P+ZJGy^y93^O+gjG+oCVFJ}$EvhY%xai93OCI5C(uT1x12|oikV(&gF#!2#QrNwqHNN;$>#de|xxK$$BkOUU~JX zxlu8^6ri-i_TpPnPb-z#9QR$(VnU7sNa^%s%E(N&VO>6uiVR6gtpVMFR@R|1UzMLG>@)e6X_$qF#RM?GS%j--Vg-irXPzNPPGV-o=u z4XW2f4(hG~(&4fEd;pZUrzQ+ zU!Gw=FiHimq7=HkTpA7o6U

QiGb}B%Cq0OiKcVqW4CGF8P~}U3qd&m4BFwVJ-uVVVVoxKZ`9A z!%m5LukKVx4p7$H?#J*HyD*luqp5vvqFc?laI+U}VeU&jL#9JPZg-)E~b=o*uRmQ1qPWS0#;jm#q4QP7{5PM1Uz za;J6`?+gG60H7e&tR|;NHE5_jFyxNyqU+YAjM5D@U6Fv~NGj&wwmHE-X`;3vQ&0gC zD6++mO#c81rlsW6q1kJ^;*AY10n~Fz!wisQa4@vsa$$bt$_f^?V{%TYhFcpbJhGu2 zU60_@sfxmP9UeX44ZYIcFDK9NRQp?$cY(%Qe=()lz^Ssl0w0smtwl;K6`E$2Ys9M% z3`K!cpj^II(*ftxrN71To~#~cPHVg&GY$r=_gBZR;vpQ>LR;}i8PnQf7Z((qo5@JT zv{4q7!S3QD*mVmp9WY;MSfnPObbXK$8i1f6T}cUs>Cm+4WMtE5 zg_^oGe}BOUXrWn?2?UhbXu2`A_=17k)}0S(mG7{NEepAp&=|>9U-t;hq_WCkvnT*y z*`)rFXt)J6L?}ye>b9iB$&c;t{212IW}|p?POyMd;|BEKb}#sB_bQ`cfz9FU_LOdx za+ECMkaJ#zZeXEs+0lNZGI|Y*S6vIgXoN?&9bpRREn^?WTT8$lx^q#6I#N~G>a1hG zqZ|#1YLjexls4F1KpGK~Mfm6iwySUeDj2;Q#yr&6WeA@N-bVeZVB8A<-)LtI7**+3 z=%d;r1I0jCXG+z0mt7jLq|jG-yU+1)w2`ZPQwELF z2*wjxY$LaA*K^T8;}dMXE4tlg6vqH4JRijE+ZGEeJW*KU2i9(jxOx5<1(y-F+Br@y zysj7wSm|^VuOS}Rc%X(IM=R|YX%>PS4tG(t^iePR{wBg@6#^I{G~Oe1a7e?pmg~07 zltcW21X!3vc7FztuO7FU#{*7!%5DX_b6{6!wZ4+>tp$m6;)*M@%!ckJ8Y{7ctrCLF zEl%ZM{eKj#L9b{#=CLiH(^lmW=DsiZ1m=;Xc0u(d((wiA22PNJC_iYj2AZyWl9$9? zf?uy#YI-iKcL9v2qh^nT@vq}&bO(30*;$z;5G@Va#qaoA94lysM$D4gw+uvJfw3|< zj88U!%~y=(tB*(R2m>_3uOa7%DgcKOd27${$uUm3!P~^kAewt84?n?dM(9I$jBTN- z<|EM2u-wDR8?Gbs8j1*W9&MugL*&CW0c#p*&qIV00t&zD_=Hz{P}5EE0yKh%e`vp8exps_A!nt8)!Z zFwhU+GhLtKM`#}oeV3ROrUQ4GPA?v600AA=5yk)mm?fe)5 z*`-6ZTkU+*t@Htfxz`A&@l3SU7Bbm3(p5RRq`|0yE1)bV6;9JB0I5JRl*E+W$jqYV z&>PJ`_OsrhMTd>iKGjPtO323yX*p%OR0TQaBJ4Y8o<~RGh zmW#Wv%9Z~B8%;(y7Ph&Lz}}0hgm`(3{20OIw#Dug>H*0vlo2qY5jH-qC__~=oH6Zo z8}(O{MVTy@?MmXBNCc`Qh7^>o;f}d?gDGiEzc2Alc7lBn=RKjwnEaRo9{Tra0+9!mf($K!Sjv{w8<0II`{M6#{)i00LEAabLtWri`Q( z!NK^rHyqdn+Vq2gjLh^kmsdujSPeq8wgH=?x%`_H<#>zkuCU)EAPBxg_=_H2%?$9s zOz%>%lxVcr7cGe9SsHh6=M6n8-Ym=E*-Mi}?UR81H!8-9qS>xx@{x36(({kvR>~Dk zH20aUezm}*T4kHgsP_IgPEk&P?7pxnC=dm( z8+hOGa0EG_lKP!UJ27~~K+y^5Yxx#R zfVASf@t*LD%zy&6b>9B~i}r9`fHkJJ$yJ1a85grCDzM6?rH!Zvz@~Sr*Eb8NMiGzk z=gzIm(Gm^UbPS_MNXmi2;b8E90;wX%z8O-OAOz~EO9^dMv8YDCwg%1^+63F!9;F$s zU~tq5hJhD=j<=buBs>T3%L&=kl+pCnV1nu1Z$+JLl~p?YH#4TvZeLWbzSHI(d_l)=IKjC>0cD~n$a0t5Q(Klrx98dGT5r{C}WHM zL98m);uY|XIVYwJcus;0Es6HE<|A5?S*A+5{=bWq0@ct*nafiE2DIOF=@nHW19CJ`e_M*XnS!LH3k@-+`6RhO z9Rb~4CBB;ly-jhS>-f<5q^Udy-lgvv0pv1py)L{nt^#npHs5T7>u%h>$ z<)kvT?v%#sTBCcDp{MIM4`84GKZr9Tlvb9fIz8BbR(s4RtG z%I>7#tk-|!T(mit-P??Ii^0JZGWE~#s$vAPfEI>cK%}s>6HtQby8wHg;q{aOPL$|S z4ylH+s|Sz^R}ZsAMzynB<(i6Z z5fmqYrR|uc%;EAspXBK^vrjn(Rx1-XYi8HtcK!$GYB&wEFMQZQP%(nD?=KFv*m^m) z72XoG=z&Ym1(?;8uqggpkSc{aZ}pf)Te<65{sD1h3SB@=)x4`!ZC94rb&D8DRK8ad zFwuMp>;$8n8DHd&nwYYR=R8E~t4so}rFpG-V=@OcOy}pW3XSy%)L&H}Ex3?LZckS2Pui^l>vAd8} zY*ufWEm$ITB_F=(p6$wpQx`D_I5|bS{#S>r$K&>KIUR1-0=4h_0TKXOyh~LblE{e0 z2X)>zspgEqJ@}1=6ijS!Nyc2QzyJjXgZXC3YVa2Co?#g-%Xly6_#}#GHq8UdkfX9K z!dq`UfIw>%P0TbGo|7cd>l`c+#6;tN%UaMX?W|*%?uS4qVztLO{xvy_pe5N)W8MN` z0MHSmfY@`8A!S9ue&v~A2WN5iKa&rj)umSLj(C@3Q;%8GzOVR^Od!xSC1ninn1Dwk z4So0*44JYQMMN#7)!Z29A?^Hf*l>ZNcG709Ft&pf5S#ZUT7UynBYaeAg8py#l3RNK zqk|Ux5v~hGrnv`%xk@4h6?4@JySs8iF`%RQXp${jE1Tyxm<{EpD!can0-JnQq+QTE zh$+OexMUqo1`H;zaEBm0gWU*m%Ciiv;>xj5`mHNIR?Drbp~_5LtEB*QT43#4QaVzJ z@rPCLjT*?IXp<~z+m{TJMYB;8WR9TGurRp(ZL_9bP~`pU%AI(chj!JYLIQE$QMdhUR|4xSHldl{OmSr8>KWUsR5D4hBTw6NpK`)RGhpokj1YP$(2oR{?ym6gY27 zg&aG&7m(b`78MAO=cSK*jS&&q74f)7~Hafz_3No*EP5J(6Kv!OVxpP9ohXz0)#qThfFnA9O(#-q+04$6~grb;EI^rIx$Us)<{sg%L*#T}-kwyhK z=nsR|(Rf1>V`^AN#DuZqiUcq$;1K*Ji(~KVGYWv^9vqe3>ey5SRulniasUF$0+gLy z!{V+1I4NPew4y8CZj8&F1TYbQkr3ys1Ww4X9rWQvu;ICzst^AFAhZBlLgjl4YPliE zI*kFg0>ZU8O95`&`;enf)+A6hef;sImPPx#PJ|v1BYS*&m_a3>KUZ)%p{%AD7MhhSw*;ee-g*u z?ON-+X+c7)DxOC;>FIf|Mo&8BSS%J~)tpLc0XPDP2yt$)%J;Q*BSafSKfU2{36k$?;dof3yyUgxUEfi?#w)Uaszm>oU$bz{A(yMTQ zP|z)*UaeKLb|ET>eqM!UsG_?-U$)?eOq6kbQYLq7Uo>@lt!&TBcjTgrRc^kqR zr6|R&Pjlv;AYFu^$4lM&zvG*z)rH}$@o+|C2dq3-TQ~e1;b80sBY4hhwA3^fBmgap z;Aa>jBeaW!x=mrl?hS@zAkzZ80HcQicfF|Xhu5jqQEZ#RS4i^^VB-htiJ7_Tt};2f|x?7~ubGJyr- zjxpL3rc@MUy!Fbau(N-}t#H_AHQUEtVF8%J(4||~)Le4Dr2Gu7R=5&DK}OqG7f|^y z=dRs75VeFFgVtRyFc)NXB^1C+m{d3lt^ms3BM~!7E3kfD?06NnBtG7T1AG;EO`B6$&*@`;L9#@Q7M~rp~wh_4F1ugFC^ana`bxwvMK3l0lo1SKPzWXjNgf3*i;iULpf!vNEBtVyI0?( z1=wm8yEwbS{{X+lTIm&XbgoEcp%&jGs=hDO-2sp-XyA=B_{+M$-yuil8D-1j7(Lw{ z^1!fBLz4$g6HXe%ts7y$G6jh6#E!}sDd*&VP(=r>9>mlR1QdnWl3%=4QgYBQ1+Zh0K#eZ8=Lt_EbH`7Rqs%4F zwgqucn7FI4t!fImhy*DU2`tGGW>-cR*W0{TK%;4`l7KRS%tuOc#w*dSof-u78)Tb~ z4GuH?MMP^1F+p`+y!x@!s_H4C>x|C=aBa)J>j~&L(#$+5Uoc@Rni;^J_IvtUkjgQ@ z(~L`$fni&5Q$w$B;sIHcJ8G75(usTyN}#B~aq55;*wWZ2PEMN=c82f)2Pc8EEYpS( z!v!)BlmhX1QEtjy~6LeM z`^vvRn3cP%W;C~_S^Jo)nn$J(4;z?w`NSE3A`TZFE8L&@F#xIvrW#SGPC1Dmhu%*@WQvSgmFHLrGE_d{Sc-)Vg@+jpsND!-N?Ul))O zuZZO;00#-)k>J1{5e7iLN>1di^C#8bGtlCV@@4pfL=_9Tw~NK!ym(#z09&2OQ$QfZ zTnCy$k18|_^@?({GY3WSe=$P(B%-b)D)PUpo#`8+$z(c2qPXu#29Ntp<#>!6bUO-a z?^h^x7mIV|_%ZD^y-(I>CCb@E55b$D)c=J)=cl~a6BpiW*5OYc1h^(z6K=+!2rqqlTeQk;_ zeyWULjmM(AvEGt}UW&IDlH!4vv5*B+8(I$CI3DzX0p$$J@%r4eYXu8*#pjukLd6hT zi@r5eYuD-KcOP;k4Mi3xmrFxBGdBX5>7xGk-69 z&iC!+C6zF$-l*?>-eGMHC>P1!F0i8m13H3Ms3XQ{=R2-_47Ae58mRU zHa~r0cPVicD4i8(#WNnS!4E4JxM@^SH))(Xvgj-6Sr^>sPVejHTP*`o-2FVtvwx?V zXgohzo%m<2KV~9lNThrxxfo1tuYZatKjeH7HF@;7cr`xs{mjUCq$k9SQ>7J&@CdY9 zf+fzF?LH($j)` zcGZj#7iiXCD>fdE^I4Tq+kZ#u7YSgi#zn<>fKJYa){Y2zXft|HW#IX_!f0so1$m!M zW{pfK-JDAlcp;$ffo(w2!yACB7nLei>Br_81ZjG8n-m+z0b2^sb_H8Xi)Pwz7Yuel z3)MWuUUL*0#=A@(0xcHn#9~jpip9YYpP$VAuF;CUkv1^*b(V4rG!SZ@XURnj0@c)4 zfh_L}DU1o4?<$}JwL^jHXW*GH!4>`2DX-pnPArPw0?*gxdFN4H<2#=Wg9p=Db&@v##^_g%DthW z^XYRQ2q^oZ5Ly*&c80|*sSU+%TY`#Qc!+kOaVHhZ8NTsZl9ItREpT%P1QlgUwz=;V zowd-O!icGEyhb-Oe`K&-L@C@~%+sbSRjp*^v{BbkQW7Ntg*?xaf{VvkFDp%!%_cBA zs29amq2NwcBP#%`1!^m|VIM7(ySj>y1)$LLL1^e@qy)a)B?ABe0X*a4Djq4TpA>{H zxL8rMd{_4xQi(~t-#0!v3I{5iT?e~FK~_PrWw!Iv+F($W=$1*?5c4Qk=47uS4HAWZ zYw8(iYEQX9oCaWT)jpXOh1}y0J!O-8Ju7ImNjyTLt0_5_G+3PxV7Mp?SQ@NA8l*s5 zBnIV^KIhGTn1;Ikl?pb)McMSAqihxFQl)ZI)(|XETQG7OZ4FLUj$uVngw@COSt7>D zqEZX67SKW2SvA+8z~ndX`$|15Y40)iJrz#!oE;ztC5K%jxm2*de3Jq|dV-N5) z;iE1dwNW4p3@%&+5jb2P6x&^;iab<7QVx}7J($6~+b~^zr|x=V3pH7n>#xMq3`?R7 zdI@N!2lsKccA-{?V%@|c)e)vnE*qc8L*_U{xkAuOV@)kzGQ=wjZ$!cLnXJryAeH2P zZX;HfWs6po;^O;LImIAg_LQp>ljXc0QK%;=n!s(p=ZmIO9H3#CRPm4a0b8iKI)+3Yw-e)bf)g9F^H zLKf!AjTXm#Guk0?cH87(fIPH11+no^xyIIUFM2vn7pr@2y@&vZ8UAsB5C}YmSHY8)Yb~ zS|$b0ELIxg+2jj4rF4J~Nud^l2$b2@J8y>B^yXKdBlbCm$Vz?qAg!Pk0k_7u{7d)d*kys&3i`L6d#H~i%T*EayMJna^gJasKvdpblPcwHP1%MlSUs!H1 z!Bhhs?&gT7Kp03{i#=K*SX>|(agLhyfY?UCmTz2uI)qqpxq@|_KLi59kC~0r>Q0-l zmKyg>JfhQvp>8Jp!6Kt;tYeD5N)Q8!7jc2zA9CgawdfM(3X;tL4ydWX=L9sG<~p;h z6#+FG*!atP2UP21BcZv_6`(Xm7fW^5j)cd+#0uyoWDBSYv0U5&;$c+Jm#_C2w^T&j zrCT1sv#+$ZV~{C>oOFO-n>Mbp%yXM(!yyaX1m>_eF{hFZ-0>GR*E45?t>24^2~i3b z6|9p+F*N!wSn7VPSQXD1W-=u~cmo*~2IGJhVaLCT?+G};Rh zJ!irCzw!VT(G^PV^eY}O?iZEY9ZEHs+qt*HE#0v@v*J$p z)*+4Nd~L@ax+sPYyY$p`wq37{7aMF`Le^fXYTp_nYmP{ut*&^Mt4+q%93{};wOi2* zFr`{;6n0x-=Vm0jx;ddU)>02Qp72$-Z!ryvmWo)cw+ppQu|J{4Tt0?#%oMfU&g9OE zMOSKzS-@w!Fx)N%_`kw4Q>@WcFXjeR$}(C3))l1c1t{9TyhonZmAcywFMo)|0iH7p z4Qf-F9T=h*aZE3y7gr=yfxUQ$;3B6+hcR`35CatgH! zmAW6AhFuu0*39LMir*66lCc;S5Y}DGdVI=5U1iJu)`_@_u#FW!&glEgV_d%GG70z* z@-46$h-4Rt5QTAt#J<(cXTgj|266}jHYfq`+wg@U9~$FN~FD!@b1&5WuLm zJ9gF8pkzxS#MrAy2m`Id(U^;ijjKc;IROlcB0>9tRRT!?2I&!sMp}yswOuwTX0C2;*hP~n5WulR!vaBQ7EJbR)WVeqozV`I`Er)ze?# zPWg9O<+28RU*cd%0%olY01CaJcAlHYEUfGxi?i2YsaDthnu_AFfl=#N1~xwY-tNfvU7Cgn_$uD6e zwAc)P+)kbjVi-JT8q^v};mlQodhr=)0)TRw&pH19B~7g!V36?=&_!u_2*k+@g`$FD zm_#Q+%<&yvRq@&j(p4e8)}`{feZyQ|DZItTxd1RxRgQ5H z-FU>hh9D`LFB#2ti8mCqD72=ixmKFkiDtvS#`_zLXk`E*mg>V%x+97iYAUXE=-|;SQtN)Fn?MtlBtQZuqGpboIm-b&Aelw{F;-95Hm(0Oyz?5N~uVJYURtkz!Co zeWC%AJm7_6ag7VoI}}2w%{8>JRphS`5{9*ySZXIv1Vw_nXY&gQZk|qJwH4DmLd;Zo z_;VHDt3D<$N(f%NCTl(=SFg@ez=MN8JOo>KAuGYv{{VDz*C~@s;mT#G_e2<|TTtP+ z{{XblPe~%7j%Br|_&8%atQ*X}(Y#?QtP<#{&w05yvQ=cQ+`foo#lsd6o71oD8n%gL z*@ChR1u$L%{jPFl3f7Dlz-lGcgS@w-PL7<$L6ofUxn;c~bn*3Ny(PURy(PURy(PUR zy(PURy(PURy+7e`hL2jH(7ukDi)i%t)35~-?tBJWI1EmE$D6b%I95F#PAWSNr=kYB zi(T!`eHKSnMmF;FusOmigUE_9yC_e1hz)aCFxO~B%o3{W`HWY>+(uN=F!y2dYK}n7 z-lNFsRGkdRK0B-iq8{@C?sVx;qU2i(t@(wV;DX9;+8c?1-NR=}NIixvcD|yadcnhd zIvr*gfCT3RW)YKCc&?=@5qWGuFBR1)8y1f0ipacT;u-dFFr5DYP;WhT=}C7k1z%Vn zz!7zIEDY&&HmkNOmMT|R{BGKmSGMyhmihsTws=zzV3J2)Eju49C{B;_0=Ku)@ZevV zdT4Z7^Ah97P~Gx=7`q7zDwED>@hk(PADNKIUc%Jw6V4m9x4FRbUeH`Vc7>pI#6~j^ zCbQ|GCV*jJ2`S1f{{XRG5MBq%x_;|8RuK&E_QV_ z0baPv%H7)wgc`x-9aiqJjo)O^n?lT#t1&)pY4KA5Od16c$m*3>Tp9zQv(n8(^^8}- z-=){yQ7)?pJ%taKSeKT*IlsVY_nZVTrP=2II>AMk3wgq}F1`uh6B~i6(>2IJRgefF zo)LawBa)7=;nm*7zi&hqlp6&q*NEmb65Ewp&7i@lO{Ziv#Rl_*b(w<{_$;o42!IB$ zLXp?uHM|ny>0S2lOVlZ{Ue{F=DX|TU47U1Geaog9wcEs1uu&;SjuudkD{+=vd=-U` z+OH8lMMtij4`6T1%$+R5dGu*RkCY1Y4YNVQ{$N*_>6VGn30g-964{plwhDx(DAqnoXIAy-jY<#R1ol9TE1|@(1U@!#l z9MoG_Ju%IrQElKQ4`N+h8ab#W1fhLB; zRy`>_Lp`CK{yi#wE~fJH$E67Dx|_A;CB)#|Qwhb?X-=7DB9c}yaSH0&ZkRhOcH=+| z6a+MYgul4SAHfBGo|%^a0HA?hWzOKH1r^MHb1|8=0ODRW)&BsG*KO1a^sPqS zLr2T$dK|Dm$mCgAg?z_EM9H=W8YzjqHA&H6rW3w+Dd#> zoI!_Lh*3qzbbaHmH4KEu^fL`rXaEgYOU?G1QdP=?pprt}2SI%R$5;x{0e}h*`s%R& zt*riLii=BiKMYw|h!+|T<>FBP0I8Wgsw~}t%}c`A2rQ{i!S|O0v0S6KY{GJK^jB5z zD$7(0BDq&8_sk)BJbHO)o=jMrf7HIFmS#A%ii8{Y&Dy$N11LF^)v_S!9!lj=3q*4| zRs1Mh)wDT6Y@g81x3=!tM#KP7K%_Ns*BxRAvkvg5h2LT(D&~NSg>S=jc^f@7xJeMI zyRm?@IsF|Awr9>m_dRBS@<%)VGab>whBaHU@NEiLW<}*Siq5q>Fr}r%S=?QSx)UjC z0B{RcEuw?_pr|qI{>KH%>#ximeGs9QT;HqF}35Y>9oRqP8t?W`}UY*F6%Hmp?Q>8h%;1q>>lwQkxCvQ zVfKAdAKqpwM*`zPj0klooPwp$_aoLiNn;Gq+0zyFfaff~Wd>$Iv}ir8edbVoqNqS! zq`*~XFfGVBg@n_R*Hp!d*`c#r@%^C2CXPK(grwyy{Kv>d<|zy7O~shFm6MP)YHsRW zQoYo$mO1AeLsEtQC0g$#;sV7>4zGL;BL}VqwJRjmymT%q`DD<~WE%F~ECnoq%HYrC zlrhR`+KsC7xsdzMspXwa9*0MzcR6N9jFX9IX?~-%#W?gx0YL)%&bD=xWRw;p;qKA- znbBp~=6&2u3xK18f@GJYk4g)nnRQ-6k4lTC=4y7Q6Pb{3KQQC1dPAD6&^l`jV;kJO zZi5xZr7+lR+OGI$ub3vGQ;6!hp*`8c){s}fq|JAN`=4fDbQ{%5j>v@PZ9C!64kgS7 zxItbyb(_>d9Ef;^ynP}!5(K2?c*Jc&L86qrrX?b)DPRt*V=PUTHoO%Y;`d%7vCEXk ztYG3UPA9QNh6$MbN*{4aBW^z!<9E*dO4WMS_mmyVAq(9=`|b{{Y7=63P5}Ef&je zdlOHVAS_5jL!PN%iy(m{FpED%(PrOP4{+wcD?>}eORH>5k z{TcD&!yt3YGl}{zWXNK+aXND?ZmW)3&e7RX_$2&QAA_X@D^v zmaPni4%|gosY5jv2DhX=6n4EvvIcb)g>@`hxZAgAznU&9S*=F-VR$eEsIk-lU;v;d z6_{1a*RUhigbi(pQ(yqQzbaSt;!SE%V7g0`>-++25dka{Tt@SS!j%9m4|SCn;yueN zJs?nJqZR;FjPkLgadF}Vve-(S=`id702&BVvKC)M94pIMmJk6*Sj*AWx|>x-uUxr? z2|6i2fI9@Msc;}fE8*C&JWqij@XA2i9qrPm+ zT9%$6KnOHU9jH{7U^(O?!HjCk&E4y5(HrwgSN_=%IxY47*!W7||4pqGe);7G9HX0&Od*19&X5&|-=k=00vnAyKCd z)l8V2DcR0$qp<562>@JLrx_8##iVHlF2eQHzLWw-2uFds+Mv8p(1K|YK#UDpcXF;( zRaI40h6eH}xPz+I2P#EQVQ?Y$jV_fElX$ii_D&TH!AoYuo+S= zZkGXF%3A=Wc&n!PNLm8%KM{yf;bQdel})yW?C2WV^@b(-5K!nC=~T4*6I^RL?=4Y< zK`4x;-bpXECTJ>;}=q;o?uwh+V@b3q?PG`@L#;D zFO1KxDlf!)mR5R9yd?7y#I_%EeqYvRBJ=<8q zHmKd|R79(Plr>O=ajGAo-6|NP0S4{=0Ix|_FaFq-&w}B^&4+5R|03*=MR>dV{M#bwmmx1-7gcg}wD*t^0at4LrlIT`v=TqQA7q z#QG~~u6<>+GYYFEJ1$jGN1$#PCJ-zfiP9D-&;=r!08-n1(*QwU8oRIs3YLUcu6z2P z)9CvAAFKZW9cxd2Tti0vG;QOBndkQCrFnOKkrk9GBF!mv- zXqu025K<;6$Gd$&MDH>WR%g9^M3jI4C7C@iuX%okjr>($aevW2vYYvl)Uk!<{ATd1 zKdTz9dGinH@r2&d{-WWR%sM5i4|qbVdiOQnF>`52Fm?h?Tt4!EWhZuVX9m5clg9)Z zez!YoQu1(3*?`jzXlhsHBg77%ek`{9Zc@5#1Om1w`NnTbq3kduDJjlQ3%O_ggpE+U z)T|-{u|37jzTXo8eOs5P$;mQBct)XFJD95oZUgSapD9sYC1Jg@PnXbyg^8qqt<6MdV3eX z*9{bl0cFzNP(LwOt}$9KB<_{qnC`z&(cp+x>d)+<&m57^u`B%g8?zd(&!0l;$1p`> z1R+BVg_vKuxl~v0(HBCivwimZ>TnHJ zjd_&PyfGpTw^Uto5h{UT0n6sE+Ge9h4=_{lF3y&|yu`I>w)xoc6}=Yj#fmyGiYIxzsCSWTH}&tYN|Z#qbbk z`rM*4q`aC5@ri}gF4xBqKOrktpxJ*Ad{9EaJ|o4rSz7?m&sa0wSRrVN0T>sAk4q)C zsb9)3wlz>tR6m^nvkTj{58LT75Q7g4%yR=CXEjAaw(11HQI|r{2#0T02`X;iM{Rg%X7g3j{cS1 zRR~~?nM+&I40)UPsfad^D^ub|gmRF^C&rY8=Ccg9)@3RwdD|_%vu5Q7-#_FzU9-!% zh8z82Rd7%ZqR;`6^804-ru~={RmU=m z3uA5kpnjs(XB zDYUxsi6#NiQQ}y;ce$^EXc?u3w|AC#_i~^c%kKphcRpmrjqD&%SMTU(G2$jYED|a) zg1YlAn&2wURXkSz023h{lz*_nHckqapYk6oQe9FtUmoyPijdbeTXe=QRVtxgD&J{tTfO>i0lHs#W%8FN zM~u)0`ib%;0B)2!z@v9NzGc+wNBayLnm48tXG%ZVP{BH6gBmlqp+E_xKv-WZf6qWF^hT=u{{WFWfKgFtD$11DuB8;7 zB8K5(VM7)RBHf#Jfm=?3XfGur}6Q1Z=Y?y~DT7E5$3t8mu;(dwqTn)&BsE@kYH@HwL{K z*p}C#1iFni_Vi(9G14?wbvF>Fr%7O4(e3C+cr_JMwtX(29`P=&!+vMfv3at;F}`+efcz_oMtQiXT$mKD}&{Ll10F;mqIv)la7 z-9pqb-(ox2ZH;bikm>er449 zcld;$XX+_6YuK6C0N5S7C}rx90YSEx^`BC~&mWlG*g1HL7Vwplh`6!K+EpLNTWe26`bAa1x9Np$k!8FJ zu72uyAB3>HTUS4IEgn_;OoP)GKXo{dF)MV9#{qLw0u74fmr&SvkT_`01&|_4gE(3t$PvX2FgOadlMvrfyjq$Sn6J)8P_Ouj8oDx z9g8@Gr>3J8_%3nnE}QB3jV0i#^XX%YZy%WNmJU!9S+J}UfpK?Fv|tINdjaHLVxr(% zpSqZC(KmEcnqlIvWij}9P)qKb_fq9{vn9pie(F?u-T0SP>5G;7Qk3qtS$Cu=#!b`P zbckb^R4Uq>ofzn2+F564Pu)sL*3dLvr2W*W%>e}#?tbcGO$qM6P34_tNN6Dl1(F#n zfHuP+U_*k!gk(8w8V~bY8A-P7UR8O8Fh(HFssMl2r9#{NP%9r%muxzNBEUKrCv-Pl>6}%cNT5| z2U$k@$7%WIhXEGEW|yt;2g+hpMYL&_uk@Lge`nPl)|tD-Z{;hO<|5z|W5mrItT4bS z`HL&84>`k>P|UE97Ka#RaKec5af}jX1WQUV$74}fqT;TvkpLMA_S({6a`A|8?*W$) zEN%svWOT!)scUL7)<%dOVj6{s04~NE5ntcuWVI^FXpEvmK~Ql5fsMqxAk=?k`P#kbVw_!p{=w)!R8Hciwe*PR)AbWj1?RKC<|*> zRzcFZ#WP1dY6r1`3=v+hi`$A?!NB#+{id|LQ7Ig}O5e{BK(`%Y4Z2K;ZXqyhxgUZS zjn?>k)c^rYYOGH7Cd5XW=D+0gyhjD`D1K(N$`>;lgDlU5>=yL1{{SUL^i~cYB}@E8 z;AYvt=t|em5%>oB=6nf9>UgrZFHE7`)Ug~10E4g=hpvq!O~_O_j_#Ly%-v`+p3;?R z=d{3&GLy@g?oMSkcD5^5yysBqKQnKsFZ8{|^}5ko?<)&f*c{fNF4?EN9w&efIf50r zsK|W7Pk}g|D?}8J6Oa0zm80_%^B}wjGLi>aFu{F7wY7-n5>ydt!K_r!g`)K>ccplW zrZbol+*4DMD@C)-U%@_O7DiQi+v%BOsZeuD!3vOcrvpzJT^&eM34GgE<)Lz6*&XYp&$|JxdpYG|^&T;U1A2&_Hzbw-& zo_LO)XZ`7>-vq()hcn?#cHMTE_@}^{?Y|M)%>@%g!gcr~wD^M2)}<=^_?ZjJ6W+yU zUVBXJ@{-X6u}Y+;b>Ru*C0jp1itQ}Y?T7PaSiBp1&5mR2Og8KC)(U!=e=`37bj;iD z-XA3___G7hLYk)+tha@AY>B^bsToM2)I_= zIf13jRadpqNv+S-9)=hxTNz^@j+tqU2r9g?w(q|Dy;$g7tE%;kX56Z1R!cjGq`yG= zkRVe68m9IL{4HIMw<51qV>wGDtuo#yv%X^q2+Apek!PKh2EX9YfUgC%>k;XtwupgJ z#T@CVKORs9ksMOB39+9spb-njv6P9{Fq2;;GKCS#5&{EvoQ;cRK=pU9?e0n+BYvK? zJ)qB5X9m0cp8yRl)-OA==8)Q;w4|1jJ8Z_J=vpsOShh#SgqvS z&SpD#nT~ijACO-2Eeal@jR)n9l@78mJ%ZCD9lg3dG_wJ?({3vs0}}>zO4MzR*8xL5 z0V!4QeTkBMM&DK?C2`MUWS3G5qR7!LS?pIg5yaJ)VUqF8L9DH+Wx~%oD=Pf6zvL)> zB`D%S+77TV9A8qbUJiT19Ei0jYP*Fwm(YQ=TSvH{hg6$P$AA5EG2-BwJU)xWOnoc<8njQTU;vWBBB#Wf+_Enh zd&@qzrHcJQPV|9w#~l7icU=)#>66Bf<3 zN6ZQ)@UFJ_@&b!mlI$p5jQh^RS9SH`tnn1EzbqVk0DaOQ;R{6XkHrEQ)daN$&CV|a zHFWBZ=PB_Pa_vi!@j-o<$Zpd0ORS70D4DnC^t3xJq4XGjBl~Wt(DwfT7*1VQf<0zf zZDAOBTvcs#2h6l_kyL9%TbS?6IyxTG!&%Ca4Ic)m4)){0diPx(x-T%A?(QZ_QzdFhVbGt+@b(D4E13#9uV5y(b3q9U_u1i z_FgsfG7(QIs~iJ>?0W3VOzGnXb8Ng`cyac|i)BWapHKRPhuI$z2EOFq>6x93smOcqoQ5qGv zgccQE0rZ(8x(YLw9$^?Y0<(=*5dpn*N2{0WaXLl|dJJ|seepgdR&J=6C!D^EG2&bv z?7?)zVmz~;0ZGbVv?wdoEj$qX50`!=lbs^BjmN?lI=VOK#ttOZ0iZfmTXsHK@@IW2 zeZ{q!HQs%>NaZEk_~H_#!k_~1Th+u!6iT@=(DW{zBd8P*D)+uUBTzAg&9Cu4W4xdj zip_L`YP`>iK?>m9FK*fPlA^dbn0MO?$b}LYTUP+L1=<;H9nvjmbW4D;xoBW#p0h;x z=M@6H&PP_|Zu)^(`kem&-lZxwz^R9dQG;(-6&*_jHp|fTF}DIEN^PsdMoI7^A?rX( zK|U$y_Bq=m02YRItCns$3#?**8MU~EP!$D9rCl{N`cev3iChlOgPPJ9Kflwf2_KmC_t=3g9BhJ*}8R! zSxt&NGO#$03pz7U^F%OIb$`lIom%#Rw`A>%!Cx_Scla(CBC*o*#D3uPQT{$;>b*xF(3n~UH& z!bk|SlfU0+VjFqPDz1yF1hJ0FA+A7tdmlhsLc09tGfaz;kpp*yG|TMbkSoejuazh;;%=Psq2O=>Bju7T(DB^X2@0w z*e^Bf6}33#FflQzbRlgRSi@!vq(F`_V45mwq>84~SyPzih+wGUlE_HlW=$dGO-iSX zELfpL&lPl5DT?>pv`k;r(c)Y~ie@J~*k{T>Vw@cVP$jGo^x6kLWd-zi5PZovvL=p)iE$d;?)i0(S=dvfjF=;^%#GPj>alZY*C z%D_@4%3h|s%>Y{wdToEy>t0twr>E5PVaMQCMoTc$?JksKN=%+8JQJF|TD_l3@P4oS zO3`d>KaWKLT|rX03rNrj#TAKEiru(IEJG0QAtm%Pb z@o?tl`y&V|(DMh8GwI=dLqX5>#lQOrE?MGZ;k`@sT~U2L;IG=;*5#_Z5Gbg`g37Bl z8scB(Cay6)_a>^b2ms)k>V_~(o373JyPBC?;?*L*iAd~Pn$fAPrWggCA>#)>}F1l;9Jkro=(7kt$u;zD_`qNB1CX@RhQUzLT#7G_V?*uBP zr5sy~N3p^Ep_QAS%44HSOnUK5zwr*~Ddd$~&vZ5>8K*9_i;pFG>Hh9wJ zHXwnPYlE2KyD+!JR+q}AiCA9?iSH{@UeHm#?99V?V-Lw-Zg!C_igFxi?=e_q3UITo zKA_pO&v;$rvc(v!ed1JU()WV1W{Y~xW{Kj7kAP=+aSs5mJBEkmYFSS35XZDiVF0yT zwCOG6EV?sS;#NBngku?CUGBHmJ|?Wg1rcan9iYp+pZOM+;!vGc`Xb8Y;!y7yS(X+> zD+vz#JpLf&D8Z_qF_AZ@rZ`~NF;Q)YqrR#;U4Y-EYVwiWya6-jR^Kn$Bap8MQ?bPzN=MLV};#gzK z+0q#8C~ar>^;EbITn`7L<;krLDu#%U7kACe7P7yoKXK{xeJ*2P2Q$F;8R*2zgPlQ6 z0tO6L_v!F{ul!AoHP5Y%gMPS!N2gheG)-P-67OTsaWh;7KiRLObljm%n1Z|!!E<=u z!3`cU7K>&70Dq_$=@2c~iJVsV801dXT&9JxAP4vwO|IK@J?Lc+%|7NWI<>x)*&!2aGuO+j1#6`>SVw-e^rE}(ol8E zW}br!uaCZB2kfF-64exep0KKETC4{yiw!D(8TQN-W{@(i!1@rQXLau%^K~eBs)HKu zX|^JY#|oBjjQN~_Fy>GztMde)ZqVZxLPjMCtL?e2>)Fjy2KQ(w;7EMP0HrGb0C3OR z!gZN+zoh_h>l1>=P9Tk;HUpqNp$EhF3O0O7N&yK6EanNUP^eqB7KX7KO3g4#=AY3B zJYT^I@T?I4w!E<%E9ue&V@@IjmBGtegLEd5-!KST0_ zxWx_8ejz-9{LCLLKrV~%DT|Bz!uHq(Z7Q~hhm6z%OB#Wvh|q-SH`s(4FbzQV$b_I6 zMOG^iy`nG>tL70E#vmy15HPO2h=OaUw5R^e1_c6FQ-=uvw{lYT1XLIeo8J=KyF{Zf zsrz-qj(GB(5~yXZxDkX zWi$#1G-3d>gG^o_YJ~wGhIFXRuysI3x-YyVTFZ%1bYB?e1DBk=S#X&ZhzbR|#3mip z!1kd8kQ#2TX*HVPNFodrX5o7>>jBFgsl)=cfMI(iV4{W+D;h$lh=D~Ndl3XxVgfKL z9pDpOl9#A~9J28vKmZU(f>aQ3!cb*NIM$LY{@YCq=kqvp86anXJ88PLiD=eknw!NR8uOG~hU)q{8;BxIY!i z16cYP%5;IJkE3R!PLiD^I!bhu=_%4vq@l^^1u ze8OPi<`fNM%&x~hWxlxUFzMG)WnUs_4!R)LTIi3=anUZ8^GT8o>q)mXaF=<05ZF01 zl-X;dR(Etv%|C=F{{RS1CZ+Ea3l3h0kElU=oFkPSrzH0mSUsgt zdP_<>#uB3+GwB5h{{Ud0WJg#(T1RAcNn1Eu_ z4onQ~3RI)&6*0gqw%@q)ApPlQwKWPJZrHrGb(jFxr?x**ij-4PwYvOT$##bb#tP~& zLJ)UUe=*rCfk>@*azmj7kyW^GkLqzBnJPcRR$QHNKg+(}M*B{%yM z@1*F6uV#@Ss+22rnL_1*sUKGkq!U-dJ|$&Qc~|>|ki9gINk@bz2yu2tEIM*Sf1v@$ z#x$Jet`GydEOhUpJKE%q{>j8yoJ_1PcXFZXy^weqlrOBp!@iEO4e%he>%uuv$mE{l z>ja#t9aFo~RAb)AvxLV)NWuJ*R{0{dRchM7;$dmmQevkaNna;G6_T?lFzMu)J!7ob z9rYrw4!S0-$5LejzOuPobV8F4Bv$tX0#ssVL7<(U}a6n{$~Ugkn85nDCI zCLi=eG&;SJhYz&|yE-K%9bS@?x#WQC9ungmbVXsk@Qi63l(@`HC<1{fj_n>uCi*4o z62#k>!?v13H|T{KKQW83$vx504I|S@$XGZ-9`1>er+sCtdT@%mqs3{`S)-S!5O=ze z3kXPDF%AH+bWiL=Sd>X+Y&{Q z>K7~>I!Z-4cuxa#M5}>8y65vT)%Yif4oC%-!(`UKB6tI$a2*i*92_V1G?+?j4Uudd zoRjnjL02K{lx=idoEwKJbm9Kb(m&ara7xu(%KZJQ-+mwLC}peC7c*`WzxpQJaB7nn zUUGL|guPX+iC06Xl6E6pe=|&R)^>hnwk3ht6bxXkz30I6x$zHD7hkq6_1y=&M8-_Z zf%yp1TTY-S!#)qy^`QwC4pPy=xs;7tM2ozFxPV}J>s8#{u|=#1n*J(=3l}25*M=nt z6|oy0cZ*<%RgDSk{{Z}7ccZ3i^!U?PiSIruG)-Rf z4;u|l6X4CBC&ctxYqoqTZ*y~e+| zb~ig{)4`u{1@YXsM*;VM+|Jp+A9E>ov%1uoYX1PXe$O*~bp20{HPiK-l;4d$aRek! zwd60$1&3(mvbSMt#J)~KS3YVERwBcXO4UkLWJqbVj*yeMK#2Z5D)mg@S!X{atSlF+ zY!&TF{Q5D|P>^EAqQ_fJ987rs0P1I>PozvU_PrCqxPJpI;nDRq^{?(4@kd%b#|3X# z%PruVYq&D-o(?C@X8LHF{E&V})9p76^|_|7>BYI&Z~L9JY2d@;A-+4F&8LfVAvNfz z>Q7E@%l^~DsBghD)5O=ULDSUjHcVEGPlYpD(sf3lSk|USew3q{pL%kxr=oQB&l`++ zhigpJso6E^H#Hb|8N0%L*ek_5T@$p^J~Zrj`kHkEIi|aVEonSkCLK#>VWNDdvx^p4ADP{SnE^dzJ~hL%hOMSd}sKz@xH&n zJ|XIExJpE%mlBzCOIuPi$c^AY!KJF)HD_d>2kQR-4}#Z~0=ZWKC=Pc~Cy>c9-V8wR zk;XPzIrOAZ6e>TK5{s7tU8WW`h-@|)c(s*;tbu5xz}O9f2G#&o0%M;YD5@htSW}`< z7G>R_7KSYeW7Al)h)|$zmR<7?5e7B}sT@)bf;h;+jML<<`0wI8V{0upe zWyKkXVY?WrP*W~Z(Z@&uurj{7IE!0 zPk|O*CnKc#lTRn=d@1&j=a^(K9fdc`*=l+FK~pkRd&sg>F`E-iH?f+nf|;@202(U4P)2pZum0?%!fyoZ{ST{ zXU=}-^v;u*Io_|Yd9R$g=LGj6-PJo>Gwdg0mTY5`esXH}pH$L6>if@#l%J12r`-5V zueClaQ+6{`;S+a^<8&sZSkDi(Vf)%Yn~BT986(HueSTxDLHsUdD?57Bf3l|D?D!LX zV+qmq<}CfUJ~In>T!;tN!A29J6>*{9mciajA!doi_~lQ{UM?G-`G} zrd|f4gHk>d!RSJKtjATyr|G)S>W65*15;P7`42;X9iX?3 zPDbHF7RgRq&q-2`Vwt**s58kv57qu3GCI&?t5|K;quYf*S{K_r%mEGrJbD(Um-isRWSg|o9u=Aq0P4W150UskhjX?FnF&1w zz0g1)g$iZ!YWuMWg)4bBLlYFEWgMvYiPn7Isu`d}3%PuG)v=TR0FeEjj-OlOO%s{% zCWQM~e;hUKG>_9$BY^*7e#>qQ|vR z{Z3*(04j@B)uZ@NgY|uC>DML7d-A*EP(!B*tP58cUV+udunw~D!`+FgBS4$t5D!x# zD4Gie&I0p>3t7Qfg5vsealv)QRwzT+m$jn~h$%}8Nc!u<@K@$<<$#kFRx+h8lB^(& zlHgP$*6Y+b!f;AbOA>1_N~UN6+PLqGV4xN%?UX2_qK^x@AlzKK9_(Z}GEU$JiUDgj zBoW^-$Qw=U;9GbA7<8cKEqLWwD~o{-JXd4;+2qx0KCkKt?&#LfBKQ>ozUZ&Tb-P!3 zjn=9$QGR*VE2^>{N`Rn}>1EbXM1mX?&v-smAGrP`4G=uf4RFmSaLz*y=3!P2XoTkJ4*NqouPSX9Rh>k#qM= zUM7CT-{AY25kZ#M6U$^mXNPA3Ak&u0n5G}O`ei;g14#9xs%c$(np&1n>jYrore z0r`aKm^OkxoJt*rY^BTw#oW$|&ZGW8L9W@04!(`Ij-u^~x>%ZX-Y3K--48BVpR+aj zhnF#XZ~C7yA9?a-kgAW>#b{)@L{toQeTEFddB|%J!g|P_(g}9d2-LI75Sl_e8-tpAs*?PngVKjF0mZ34B64yeGn%4$;MQO+!&p<0-BT zlJS|35Bo#1u$Mduqlw~wN%9$UpO4_x@x4BrewKbOtoR0|Wmoc;i; z4E#WxOvPc>8HeSVsmyQU&xbRoIGT8me^cie^jh54^)r4GCDNh9`daZabq&oQtw7Jo zKNHBxS@t1&Avu}j+6(5xG9FW7yiIX`&8WhQSnw764EmD~(##D%?9JpT#vU1Nu0N^n z5qmns`#)2E0&ecZ9#(JSOTK7)lOONxG4exyKi{Bf=MS6vmIrzLlkT5;Has)#m;93# zh9j>{_nwF2&0fUU;g>vf<&ov_Gw@CRGk=0;9jElLvOfIwbv_{n=1=>RZg~Eto=Cxc!}akuL!;&x^0N2hPnEg${{VCP=gEdW z{{Yuz=40zV2d~ui+Fd({w4l^& z%!^c%qk;Y)E>apwe~v=1`s#w@vZwDK+yj$XC7{@i+{hQ7T~J(>Q~-rW)L4&1g_1#M zY$&>E{7>FLxV2mT02h@|4*reD9e5q*yT-i-oDnTZUJ(+0@%_f5lFUIs=tBFE5kM3b zqrd{f)|VaMlxQ}(LZ?p=kFifVPFJcNjaphoJ?VkPA13iG+RM54 zDLv)k?})LNOjx(>^ECNQo*}N!^%m}{F!9ej5dN1lXQ1)xepA{^-nm$}G2~4%&SvcJ zPqrOSS@8``4|(^`>6rN;-_Q3o^P2fXt<0dWl+<4g`-rph&8qlyJN+iC&&^y%QPy!f zh|TexZxbxs6Hk=*1o>py;(TepHH^=}%+Z{BJfFleZhh?M&Ks3@`-__T?svHEIUf*n zaZ^n0@igcAPTALFpq3aU0${=_!-7=0HasDn7BeHn#CwbraJjk*pMHEO{{RU+{XU!E zlT2@hv(9IKg!`d;IcLN*Ej%;Tr4wqzq~d+^Z*0@f8u{bQ)6RE#!{v|jCJ1(6pDe@P z&Ar-&!g}zjV@5HF)^jyJ5%^e5qe2@CEoj$C)bLGy8S!Vd#55c#hdpo|FDP>@XTeFy zOF?Gu2zLq&Q&t*9URp8^VzwYq_#iDUTxNSyStsfB05-f^RELo9h!(2gGPfQcQ@hEz zzJD^|x*00ZVgrnTysa4o4T;;+*i$v;Sp8?oeGR;NH@rRx@TYn7{YSw1Pk}#i>Ggen zkERwLC{<^XAd3{J4Kn8={9YK`y=7F}U)%4AQ(D}iNP*%|B)9}^X|d2kaiYQE9u-RIdA+G zK}~S|`%adj#GEkB-o(VjkGt$n6Yu$0n4X~NlMss@>#0R!4OZ5E!@(D~%D@?m_l^QR z*~K~q`<);HlYyfw%Xj<$pO&ty6F|2APvVdv;-JOz~b~{yiNi zduiE3mB=6_w3+bvL(*BLjq@?=iR;*!zp(Ia1bMDBT-};~DvGRoD2^R&~ewP`|oWBxDuPI!xD0u!vw0Ycs zhtfU8wf_wJ+A!a!zCP+rWLg!wZtCi5+?qe9M>$KyiTv&mJ83$x2d{EuyCUP8&K|iV zDArX{V^Hr2BckBs0`o&{mjN+KLC?Wg$TB92wJvHI7yOk7YFYYzR1&~dz{+1pvz8we zjI?-dDhMIZsoY$ gN(IK;*KcE!NhH|oSM+P-O2*2OzVVgKS_wzO;mf1RX7n~^DwK#L+c#85tVVJ z%r37_6PGf*bnN&?|FbpN^{aW#>W4Rfyg$EB;uyfnOHy`pe3Kl!HmESD9rDExm-s{E zi`cM%q4LkzR$psK9;L8GX6|rIcyjM!6+R{}0=qgf4JzW=m4&*q3TkP|PEF72eWH@r zSK0=$cM`IL0XYMx9CrQqBOf^e!sh2o`@>9AZeG=@Qsu^TpC>Br`#y?7mreHhNdG21c)LxE=Z#z}C}Z1YjYNZ`qHDYLp9`PM zcN9|_?Zxlk+kCr5AcYZJ+29w&&}Y1qI~YvsXh-;^pOiUbVjW-x@3F~F)JyRw4nt_}5|DZ3qJlkXB7io>eIdp&?6#)ne z?eEAlh(mS@rRR@$oN53*`@dNgOb<`K60s5WVBS}kiTZ@>*^gZu6qS}4&f=!~>78H| zX}&BfERJ!?kQ*=%$@)C_)ZW0mY0h(P?f0CZOfdsVE!cqq>kQnDqkQl48P7_sRZ#Yx913;*fJ7-B@byvIQ8F9~kHB12eVTQ%Bt) zp&FT%QYWNa>#`S~P?B6ThJGnVQ?LL*(S0+rLaO~(d)^NAnY*!jDnyR2?6XJsP5`XW zGtI}K<_D)4=d7Puccf%~};0s0QZd!JH|F`&_<<2Y- z^<-E=$twhtQ8z3?MmXQHokhQBl~pN0(#s;`H{#1QrtI&>%RIw)4x3)JrTF)~l9}Ml zC-V)e!+uiEr7~u-d>LBmg_-wffSOUzDTz*BhyJH}#yp(H?iu27`&R`eeLccpvo2TRUDU477PrEa9g$6{31LiX<= zt$XxRpB252V&Srbe=#mZb_<6S6IPich@Tmj8fq`?%E#m0sFqgQ@<9qZ*S(v~|P@^k-YEi`p<@AV;I;AUg zS)XiEbU4|2Xj6wU8yw5NfGEwlxa1ReMamORp~psthG%}V)nE25b3?E5D?cfl3Q(-5 zPxjebjq=oy;7R4=7?2RXr-NP=`~r7o%O?J*(XA74_yF{xUnAu+8snvTxt-n}CzCtM zKAoIvumME@da8*RYvyWCoUKU?Ves00xc~X^#l!{C-sFrXa5Gk=7bcNSD;I53_kCZK z$j#LYWPue`q3MzB%L&o!as(D>1)eECr>!c=H_}LClDo-DnTv>m(TRyR8c%wC$y;cL z8v?eu2fRk@;)_IPT&<7M54B(BI_gV1l}yPak}2Fu;_c9A5}|tDcIt|Js~ZKVOsMMwbN;OU{_GJN7yjA7j}DZ( znee}ti}`+9IC%XRBkC5WqCU2B(&CM<>;U#c3=p+bKpX>~%(%o|4H;O`F??Pre-KD% zBkw8r`1&Ie0^1=sO>fp5HPB<~} zRWDV~(OU|O1wd14Q&|2yZU@sKUD`&SvrT7R zxPx*0K&!K8$bCc}XKv2eZpL6S`@j+$&N4%hFBQIMu06ndJNZJWN7_G!2aT)gGKvti6yx5N zrBC8kk9!-`R9%ozB}$NumL_x$^GC^bY-0|q1nD1(sPZ&@9og9Ixcr*T*R@|4^EXd}ny3N7mNrJ}o_5@lmOg^|D=gzx}i8 z$TCemS#XZfM^Z9?{+Kq?>Rn&&iri8HFQaoA%M(39VkddEq`^=o0wW?qhbFSqci+Bk zo~+67md-ki)vb{Jwv0}f5)As+J>5NXeHqM-L7{9+zQ=={5 zXdKCW&EcpVEr#)PV`RGUc{KL>?ej`IhbOkA24QXW+a`KE2S)GT>l!}PO@-MWj-s>A zevXg$IG6InG*LRIMspA2mqN2IS3iF+-6<2P%s6tSvbDzQ)N}H(dspD#<;;5I zgK#oge?{pNpVZrT>nQfiAo9ERHW%_NbadygooErWs|ru->m}b;t!2tflWO#@|M?Te zmq0ZsqnGQ`8-w9QCN@xYNpkJBwg2PCzYRSc%6o2m=b}|)OcF+J3#yAa=`x*A$@4k^ za^{T(Qr^u_UeX4uEbDyWS*MAVUm=bx^&@0H?+Cm;v$$6C{|6}~YA?~9xdAtWUg^~O zcU*>>JH(SdP2SShscsCDEg@OS9uJ@ppj;g9qlMTt?T5&bPINPD{QkaX`Qmax{iui; znf^&upPAB=$Qe%$2S|{gGP)t?O+fRVh{r1ds?uR^5D(x>|6z0)Wuf?5O zV(C93S_??C^NLiB7lGoj5<90M|gtFB8grC>vABQ=+cas1puePklR|@NAYe&VyN;#lf*2i|}6m<@W)lqy$5ZXyNwb7XThmC(PjI{h5!|jv!XSQb; zInwG60As=r3=3{EdxSMWfs1ym-_$a=jIC&KN zOR_*cw>=C-Lw#b|jpXPdUT>H0Am1>LXPEEuYj|9As_}Mnl`uD-tv1joh_pW^trF=i z0*GBrT2VU)S&6-@6nE(HhCbM#`opzea>rdbM4CvD{qamR4K1SINy2^@u|Ocrjq&#}EuL=GlS5(&w^X4dzUl2V5a(tmSSB*(T2F!q}tXkd;PF}SEJ8#b)SD^4fO#mpkb$@kInVge7r#9UMA``{#e@0Wi;1;wsLD?u(LH%gY8UKk4bkg17F2OiH<($HW~UE_nOT$AN?ohZ(L zp6ANfnG@1IA_`k0Cy)Rc;O?ukGK}6v-UYi2=GR z^m%xFBtV^Y;cys92>1hYpf)v)Ke#%NZ8k3t?*}(WbOmf7d zu7q}Z@-k{-@4GLvy$&w*44oW_WpPf8Ix@L~lVEMx4nm$zb48cO3Vd&&MTXKPCDg4^ zZ^@VXqLZ~sq#u7FD|?$b9!z8!I8%(v@~*`Vvsm$_e)H@*Pi1QGLM?BS`alB15w16G zgsSwWT7S;FM^2I^8Pqw{B`V(2b$NXaCqMQc>#RiXPg1{H#HIL*jiW)q&da@^@EY%n zGp?!EhDY)X>bK9$j0h=9{F^79@KpX9_#>$pVaGLQ^cQ2V28niEWWea4TZfzoo2W<_ zHNA#u*za!9m9(1tH1psQ6NOP&Ucn3ML8EY)(g~64aG9!!v!)0rbI{Sah{VEeQ-%!u z67(}*fA$!0kIt1m?}?S=m_~dFhe*hidPZ3 zal<*FZQ-Hm)FM9U3aI{of}D6&XP^YH39bN#ujE*rW1E?k@@M0Z1zU%1P;oi*k`=q) z6-$@TA&4JR(n?VkgA~A3$!wQfrR}Uh`0NG#p73cUGBN;so_-KPsBjysMee&QVyqtbQMQgjU4#Q?*oqUtP63Ko#D!xR^HXG zxGRN=$FjXKp*85F0icTaY4}H6{+iN>#QL>gLcgZH(O*3sZ(0WAa`svRmZfh&#&0b=PVc^DN6h zzn@j%;leOtXvIWQ8%#bnEMTvyzKIr0isckk4Bf;?c#qKqdh^a%K?H|qfG<}y&iyY& zo^0O0?xeiMQ|ISkqInm7ZL25DnG-X9yg7AWdGC{kmYr)^a4PZs{e4#5`->H&X%HV{qL(#&oaSLx5*^E{_YNB6^!&2$gq2EYbDVGiad zcI_#chw_vww;Ngx>_d{R@q|E7Hf1N$i=5jQ8N8B_F@r*4y`?8&C{P$OJwNzG1>LD+ z_8FRyM`?Es1U^>lp1NCm|JL3RceZ?}Fj{)u$46S7rLp4)&~k0VB0$YyP~_U#@r1}L zsHhH7+J%tWSNDhafDuCY&#?vMg(Txy>5-BdH63Et5u&HaNP8KkGgE_m0P}O{CNzu& z+DM-fHwXp3D#)(^zM&2@s52vm2YsJPzb=W~l^WrEOV0aW{xOxQ`3nc?RL$jP8#g;SjqnT78gvN3Vw=*nt1ynUMIbLW=#bo}Jo= zpXwY?Ly=DQPo4ZK7<9TAS4y&C%BPI0nz^D)c|W^(7_?7?CP!2zgCoQN)-gnRMN z!m?c2+495a$;AIVQ7@q?N`4f*lXR`0+R2Q+5p=(hg86E4)au*RpN51Zu3Ck+PZGlC zgT5+E*pa6?PDIF|Gj9JruFkAaFcXdeH(hT8V;TM1RWaB&$m`RZUoBP^k~7$8 zewP;9RGnenA!?OT=zq0+9oBy4Tma6T8DYNhO-eMwgXB4#c|;_*A?vn9j~6@{>>7fd z);-oHk)0Yct7wgQlXN4Fvo~IV?zotaghKPVSzUV2{w;0z!)xRb%uuGQC!Eh`iKK(y z^|uijon|_TagN?E6-?~m&{u^kIZA5LLL|x0lTUS?eg9!DwlK`6PQP?M%tSF*R&(Q8I`sz2 zo2MqKQqR(u_*c-=a{R@>Ag%xYGVplxBlefnrckchb}c`SoLS0y~Gg2j6NWNjS)9XW}H$S zO#Uu6$cP_GZu{o6>T%;CKp_4Lpi*aSQu?)PB|W2=Ww1ws2|Xs`(~ty=oA;Csb_--b zBPp3!0lbpBXYPd$pNpU17kr&rJvU>xxd)1mLrBbG(6r)0H<#xL`xb0l2~T+#Y9q;| zNhb9EF9zBF9aApG6|lg`*k({y&NrseuHABAcAYV)S=Cs24Jn-%|JlL*W6a;`BE!_L zt3-QIp$&FOTX$wrf|jc3eT1*9p8pHe9!bJ@T$X6$7BnjW-VqDS<*=d??|%{H_@Qf@ zca%_9=XVRo2&oVtt%&g0oT!NJNZ^}-$EqqH5uT&efO_iHWizSc~uvAC=SK z%UT?fil%%(N!qz|0h4lv4=noWuM~SP-wdgDuBZ1yexB11?QzC?wyzS*w!1x;PI;TS zo5)a_`R77^Y__5~O12{g&Hz%@w_pF+<~i)PM6%l!wqCUo`rjlTpgB0G5yoTfx+gGy ze*F-8Be}5_;0rGUte*kx#M~nHx|LutS5worz%Yo%^KDPx%w`VwuId9GzxIRzk;7M4CiC}0{Sp+GY#suSMbU0ooMZq#!{$C^375D8*(OKu zU@Xw^2U)dqjy9&Yy>mpz1q^$9t6SFV z@+M!2__HqJBRaTfX&)%^L4}_yXP5crVKZQp%bwL zT!SIJ^K1S~oDbx$wm;VS+0C9eC7!TX4w(VH(T)~^XTyb(oFra6L96V}Mjd^%UVz50 zdtt*GHC<~#tIu&98R%B%O5HL+ch8eC9(i2qMiPZqyKnK*`Xle2$FvJjZRa=bxppBE z-uq-wA#^(W*Y$9zbUdb1Z99WjIi|LPm-9`_4#-9A?QU93&~alJIvNIT0g6nm`6Hu; zDfF6)R?UMqa{zE2HarIbzW)TFt8gC^yz$}QR}|Ee2ycwYZPeQUOf+>YF~*L}iyVU& z6d>NR86fISL(VG=Ib-!Yjpgaho1L5Ri(R|#6NRXWZ3Q+|4zkk$DB$b zKtrhp_)#3;@ed>az|lg*_-1GLI%wXRm5Yu;jiyag1TkOuLrdWv0Gcu$6`rV)^Z)UVAY1KDcy0OVudJT${;Z-l| zf1IUj&u{FzV;G;aDN&)go+OkQ2bzyhQpc)3Wl+>@{#saxbxRslSWs^?eXS#X!Pt_! zplN2GyE5r3&wnmNhtuwb4ULydCU|7uSvWs?B0gS4EV9s|PT{PFeiWOn=4Y$Q#7~*d z6cV5*VYb*jYm6-&e);j?KJ_5ttxJ^}-0k0T;PiUvLrx0?zZ7RTxC5r>BNlQ(;Z>V) zQ3UTE38;_KU-SBn#6yB6JhK;Nf!O~~sQ7}s`s&+oE4A8Bxk|YcW)JiVlLd;GqxSXL z`*IpJ?Pm0IgMaQl!hC5C!{Te!RC#m3kqw$eX&TLffXeM>;5q8UqWM*rg|tl7dmr$!s6V3dce%hxMrZb*RG=vHI9k7p1uAT#kNpM9 z1c$kW3T(69&)2-dts?mmM|;0Ovg;tx{ z#7iw#>c(S+WdATRnGD|%6pxrccpnC-??_IHwU`O`!ZUAI<2myJ2!p<6#Tit^y}j-) zSqjib)AN!eAnUt8nH!kLs&MNhs!U^c7!<0p1c)7lCRoV;=OT`5ADlS_9({;s_Pw}{ ztHVETN-90rd;2hu;a{_&aS_9*GLu$wxq$i;lpw8&%CNe9Yw=WKQZ&M>$c$LKWHMp| z8L0`yW)8-3EwcTX;UCmAx20897vMiEFdb;zwZ%E0lGsI(nS3|*YlG43iXvwtqce5w zUNaz(?2*igx0G$#N|zYF(S~eKwv{JloPO~_@X}~ z#X`Us2U|~_k)^^ee@xm3Mloh5_xLe8zR-%O{b!&-3WMKVQP&<3D7~BJ2qwmyMDLJ? zwzxrPAod&*;4}k=R-3^89azRS%T>JmHQv_L7Ttvk_<%aS^rgkeLx;##NZW74W0f@hjqcn$#+x=#DqN22jH|{4pOBTl(Kv`gXxm?H&5T^oEw%1S^7@{)-8;F)0a$3!COMQJLY^_c5w?!P5pJFBmHmy6w#ELH~r9 z&6v0;TUR?i+Bv(C}DddIWgW z37hXnKU_EwylMGfDKhj!1ptJ`X{?SXqKHAuI|^`Q-86)&qSMm$+^N}9&SffD?mM;7 z5T0oVaNE~?=sxoJjN}D)wQU#D=YMuTEK~3moFTVxsT*%N0eqtUzbxM?Vo+peKS=WZ zbiVua<{@ZJ%omMtwLR7W#!@jYj4X&g!7o2;geca!-js~~LboPUFjw<58ReIA>dP>) zS%j*EFzMPsAze)RwrRe(R}c3r{Zm>eHd<-;yiqaIaJ06DFTACxU*z-_tr`dQ3t55Q zNOQ3i<`WxeuNqrgx4YGwSBORO>df#JNM=~HRzJdUxJj2I_|;v9FDMu=1Sd+?A}pqG zHXN@jb)x5g`YhAh39bGvU7CYOUW3^9##R9&%~RY~>srLIg6}Uhpl+f)e=)@1-CxA3 zM6W>cV$!5mzsJP3;&%g6OSZKZ^_@{1z-_C6;e+}A0?+KqSI`5YkBjN!YU~*K*TvRYPy4O$fWZ5h@ zzY2L_@(d~Sm4eMtCX+q_Z*c(Oyt21L^uZgO@7MYeZOIU&%^@D|)@?|_JpGU3j_~Ve z$QQF@C0>)|K7xZ!`hP~nzy)`SUvH^I`uc>Q9xKE)=QL-Mdi>(BG0}901neI^?MDcj z|0tA7r@k7;J?)x2)b(8z-a4a@ZCa2EX-shn-|G>yN>OtB8A5Brf2ABw*=YVmjy%MMad>?*{i@14tPK z-`N=-hD?Rw?($=g^N@1hW|_Fw#IKgWTE~CYr28Z zh#VF4z~kT#p|QJ4%Qn?Ei^}gO5%UfI;c05_EB>w`Q6|uRQr*JPz zzRNXTRi1QTYJrxmYZvBAyh`gUDGi^}6RH=+zXtwdfCoT7wjD&>n(ggQgcv>yJrhq0 znlxv(PV+$@PX9qjW-&G}iq%(U(-t?^)_RZ~$xlC4GJL05PC4dWHS{-kneK0Z7OpbC zi&9h2j2=CjFwh!JyUqw2hDVrQB7ieFGhJ?zZUsvO4Y9~C+HwRUXW@_lWSUZn1dLv$ z8QfX;K}JI&A#fU9R8X^T`ShAit~J2+jD(uDYTZW;XQta?j2Sn(Gw~1^fks0l?L&7I z+xHqmYQ7xkS!uQa$z0$mGpo~zW=00uuo>hyGlD1U+zKAf-dsJqVVQ{l1S$sn%UZPuX7X{%h;bH9kh0*{4o8B-4W`{KC`$%LTS8&Ocv#6$5cl1*J}g#4^OBJ z%UR{_WR6W-riEPt!nWjVAvqbxRVePbTeLmG)tNG|M z!?8StCHsKAYUO_;>dN6^M$Pd!ge1 z@V{c!|6T$;*U2WX95Gq(sCcw`ArvFOT7JLLVJ_$iZ3g+C806VsrTNU-zvrF({lZuq z^TqnK+$17|>5GOJGEBywv)`u7-gdel8k1GUp|&pg(J1?o9e>BhR>i$Ky1nZ!Mxm?# z@?O;|n#*maM@%otTg(BX%7~zdwhX@-_r7)G{)?gFqCo$A9VVpnLd@iZ@b%{}EHF*U zh^mEG(^MOg6})1dIT$OX4_wD>zutu;uQDia;>*1srL@QBa3pl|lrc@v1`PVTP>F!( zhYrMs#6_ay7N!Wtd-J;n%(KT!LLM?zE32H+&c#WE)TH5Qt(C4a9;D+d5ttf59+C!H z8A8kR?C&_QYG@B~=pdZ&QFDt0>+~FL&!znXC`K2=rZ)}E@8A|~hhC&^E68t3|AAf< z9md{HW@j#^-oqobmV?3Ys0zanc+tFLS{W_uG1~pN8!KN`rXF+8t%4q6x?uuR-Z{t? zr#&N{PKRq(J4`y;3m&8Yb8jp4qP>1{+7&?TW>3&ANjwiUUojl_saKau8R=5K{(8NM zHplg5WD^aGE~a8fI4R^|wph3#KlZW&UDx2cVT%JuwqhBqw3v zBV1hU8Hg7j#Xh(vDRxHd#7QE!Q?zIMivUH!c)%NY;sucM_O*SDv)B;EML5=P&2aYL zO-R#YN?yK~ObRsF;)6?kQ`yxiRN!d2o>t(Qd8p;fxk#!&Ot z0{8}782bRn3H2FC!SrFur0w`{{**sLaD=kd5EB)kk z>Bk1)G3+MsS10gpHJ21$b3=Rsb;)Limrln+w|xv$gN=J11eN+TJ2#trkFcY_T@r~SXA&hv0xQFkl>bcGgcEZIoLkwYOm7iBl`%H zoVMF|elKotuXRAPVsC%3%(CsWKGmiOu|>jxZJ?-&3>m+BGn1Ae{N0{*TUw_l_8DzK zsQ8-Ew<~vp#v6hviM&e?yF3rz^HG??6S;EJug8W8F2A<3nn6+9wF#^%V`MZ&lh#}2 z)>?a^&t~f_xoW&e%=3bq(X?T=1A}YJnSrn*5aDe&KLC<@C~XR-rD(970{W z^t1Gn9|zM|oXr3@nO6TRn|Q(jS(ei^2Iruu5Yxq+Nv)3yqGE5ubtr_IQj=jg(3`oU zQ;Tb4diU>5$ZVk2mB68yN7}n8qp()HE4~M$l1hr*s>Ja*LY4~HD96yVllK7v5Ma2* zTOI61B~tw?6=_xikq&_mHAxPc0<6KSwCf{70g-|zn3_ocVE0(pL+NatS=!)OdELlAN@@ntchLA8ttxKy9eRzO*lH=ysRQ3VDFv5igx4Jlz#BwqqXz2`R^vhoi+v*w8BWn72t_X}o z&VspFC#a0|nj7@T3@Kx+A`N#pJ)Qm1{tUAujN0*7{TIA59i=XhfneoM3rk+t#*(|O zLL{W|5Sl6o>=dRYv$ZJJ53!ZKUY)xL1E~7;Vw+92b4imcBtWml_CgBr* zMB4fc2W$1!|Ee~lijJEqgO{wG{G4xl+t3gqD?wzn|8eL>+%y5$bE^iIhv`CGPqH@jxsEhYrV3BoY4X=9uD=FF8uT$z}i~+02 zU38PzWq9XN~E zUpkw3g0z6mqBZ>z!j?_?`C`Y3sB-@6I%(Cy8h;9s0bdl1Hu$db!a-bYE9&N<^tVju zJ(~+wz^E~S^Tw9|!M_+)A-*1kxneCRj~BCvN2xev*2J~}$aGDaW5x2A)5)9=pmkWq zs%TfsfYxA)2L9s*8$@`(_+N}~mdD+mJFRDpR$41xfCei>-o(LzFR)34Z=6U1LFk>rl;tDtDj zYyu!ET0{)!9PhD^czehCcot)B_sm8V%rz4!exG};*`~ek7jlvAs}Jr5*4(JyWJp5q zf}S=Hp-+t$H`%}Z9!RF*XR|y^veP&l?o<|qBa)i@m;oyAUT6RxF?U0$j>yIiVGQIy zSOw`!iup5~x@W2QOPB+I_drU-lF}x)d+kvHer%8G87hMG)doUWcT^xApnjc3cSZt* z6j=HI(4pU=vAM~Iw5LAXpkd4csf-TfnoJXd)dfa4)ynDr{hshD8K^*+&Xo0G3$QX$6Ym@GvEwM}L|wtsq$HJUKhcFQ{$K;$Q;z0iH8QM8-udpI zc$UT{ZAEbnzZ3<=d5d}&q^1y_5SS18Z;}=SqoCj+Q8T%P>zIG?q*1DyljYSrzLhfu z%R8brf%a=saWzqsuUumGe9~=Jpb&UF=O71?VqN39)FzPFCT4x?&jU-M#}OFlx=?~x zFrJz5XhFF3p+YjEpwQLZJjP1Uw~jBhw?iNS?uzbi;;(B=Hk(|RB*F!J#(>F1RsR}O zgNU6??OhEB-4(^rPbd@l?AC9XI}0!4Pw(`0F;NoU&emFnx;{!;VqiYvcIjm{0c%gd4kbZzo zwTW??T2&{fZB4sFQ-#1O+ydbQn&ufl(Ztbn{d6oAkdwN?+GnI9S90vy+>k^sEjm{?WVxJR0{o=0$Lzr11j{>s$oib}?$2D5GGOhu? zC%ZcR6iJP>k-q8;2Sw>SrZP1ziabS{VLN2bCr+PoFetRH=^8XB^g@B&>FEKM{)OUa z_k%ih`zfNn^AGvB{fO*up|m$l)@y!=2U!Pi4uETrqU5g9HiwC=8}aR})ToQi>{=w+ z^74JIm#u<~I1sR`XOMG4buyHgJGJS442At3x;ZN)!IdpYM^Tu)-u6_N;hP&zEZdWf z1;v0ukxt~iqN&#!SweVGvu6YbjO89;&f1u;I*cWyGtqw zx)$kAQ{vJBg7YZS*EbzqI1os40wf&#NwK62WEiaYa7zJ(a&~K+z~5_TUKryDX@fFv z35sqHD{l(WvDByq^Wj*w!j^e=j{iP}Kz#E^ugaEwru)nmLupd1yx5wR(cM|s^!4DDGY=&X6Tu>a}$5&g|5tc_VO^Ign5*q~}7nAe5T zCd-a5g#A{@_@xlvb+Eg*kZyKga0*eoBm46gBxK161#fY_^GjPY4tL} zdk9eWWwVNb#!2Z1ZhK8%wKWb7&0NrKh+_Rp`{_UMl;}cX&7p+&(MKC7$YqfDYg)Vo zPjQU08Ark+%VL7Y`dbAW?jE1dmF&sY!c7btA^o~)Jo_;#zsZKixhLa_FzOALj9=Rr z_?wwLHhea0sF;m6@dN3IQ_QB0O_!_MhXqR;XSRf_4R7qIyj4)^wT z*ml2KH?g8hd>z}Hlk_3RoBe4Vd)Dibtk#W`?{h;nMbtGJRzg+1+FO zZjvs?-4AsX8K9yk%h0qW?9CU00$%i)2xSl{J4qmGU%D&ee`Bo?Cv$|pau-vkBdVfF z&{3RqCO|X}_LVAY)woJN-6J!|i2UP$v*au^q$HH6o|`r|qylhb)X2*K4L#EV(Dk7U zPhS3UxQTQpBUVy=)%EtNhO)Cnd(J zDi1qgz?-=5uew0ON&RA;Y=phbT2O8DW)8MxJ+)_b$}-QSk3SY3`TpwkMPp47?e5Gp z8<>3Ldrc8t5PS48@ zAJn)z3A7t7#83)77*vW9=@@nK*TPHCL-_+Um@nc0v{O0mj&e7YXNxS6B9|0)^LpC|0t{U^xj4C)19GS+)dojpmiUT(xm*>JN zF?+yfb$hC*uC&J^%j7@XzW+9^S&_HIMz^hH1T>JWvx~SVuT(?=;tw>?^hB|Qxb0ml z!w+%Ee?Q&Nswv<@%jVMZp_lI@uN4itaqWUJbCa1{Y&~fKLCg{_$%{yFUjGN8$cu+9 zNOp1ul1vuFi2@0voav6?ZPoU@R__{t1!`UI>bvbVbU+=lj5M<9o}Q1B?w!!JYj(%9 zy)A&D%R5kJs4N4Tr?jq96s*~k#8I4_qhJO#h#y`sT|F0#7OUfX{l`Nxm{xK5C}o@+ z9)7iC`n6C@9r`3`Oeahaj?ls=}7VMw#K4W4AqGjA}gcU!lp1~x%;8@`qRm`Uio+|v|QAWjyAHlC`cv-MML^sC4E?~jDa6f&u6BGL5UTa2fGbY?D-vLJ&2fy|mJP6*76SSK{jjlSyNVt|-~Pbe`ad_122*v_Fg1!Y zRTLc6dl+w3eiIQEYuPSFw>SErj0y^Z3ovDE(*Djise3-Lj-DdAjshmGsGi>}{p(s4 z%#kznIIF3M?rE&f)bGuT2ZpUhCZ@3oQPor>pmq3*Bwy$~P-aZ-r-cd@dGkm95Q>I> zaTalvPg-jZFk!2~#!?n$Bni+W@7Ysg&B}lFaWr1xd?JFSpYg^3(Z!`vp!6-E^!Hrq zau})MN|AEQ5CXI{Zx;c3)F!rtV7I#B`V0kt$H%}AqLcMN_08?<5}`cbw(sZig+ z)~WYR_r6&74u+bxbMT`VWlhBXh76GZ6IgG9xTE!J`Z3edXD^TYS05Q7~V@37qb@+;k%C`Y*;G zl!p@RLIeeE$=AB{+$eOBe%f+0#=zKClz5!ctfu85`HQ7;A}V_=l}#;(6tMN2Ga>Dr zI>DwGv5yowKMx{3rQ9Sgxa7XMMltXyBtR$Gx=%vE=>Vn0w2pIJTI8-mbiR?{oJ#eeUhi zcXan?J}?-3fvQ^1^Q<-hbN;5W49auYez>u=L$-E6!a=t}%FG#CL=TR)-c_PpR?lOd zvG{|c*aU*cv`1y%>^IHxBGBTYcy%-$=b`XVoNFWUVA_8t_x3`kM)1~hochkf+fVV4 zI52i?JsrdLFMyFhn|DhSw7fl`Bk|V=sK7g-f0X!?8QgYkof!TxzDqnyCGyQ(;?wWt zY>Cq1Z~rd_qYBGC!WqerLC@M`g@FXze1;e#9*u1vaSz1p;g|B4vEs{7z5X@t-7m%JxA7q29T~mXQMG}n zzBSQ5H+dD_VOMVy^#9RY5P_rgzi$cDcj+Q>^A(mfVlObaJ@7p&l;2}&N&x9Mh&z6^ z4E_b+&q-^XI`=B1Jds#sz%DrmE_!>0y<6csZ1Bu7hH7n6`*{cybr7=9pj@@cIHZ zy5x0pZQ83)+ey*!;(ZXtlpoeZyh_=@rCG3J4j}T&R|v-b_^+$Le$S}< z0SLtHFhDl+uY)2y>J#fo8y6|6L~ju*jF#2rjKfh9`;rGii)VhhGr5qRaZA;oJ4%h0 zXZqR9s`1kKH*cT7!&N~~h(DXA{0|oCnA^|MdcC3kP4(6wBrtRN++ySwvnl>Bm{*#0 zU&bSG0HyUn{8?;il3P# zY5Hax(0UyTOea1;_)=C;_gb2LZ-lN9nRxTL1DN09u_K*C9*vi)&~KLI(k8x1Lfb zN@Zf!a=D#C80piVCwNg!sxZznjK`<}%|1(J+O0(t1;Nd}mgm8HZ*!tTOt%O{>^e05 z#z|T35E!u5*+sy7^|MrC36S;2;tphHMxDxU6SK|xg+BxKgJf;Q5gw99 z7NLOfSvL_w0)5uwthxy;J$kfid2D5LiWKg&AB%SNzn@DsNIDBTr#JSDD(*LNcf4Va zYvMqOs^p_Aj6B@Jkni2@t)Y0^vh>D2_n0D~Ipif@jq5T`TMdaOP13^vi4j}9P%oJf zq8KQFm0FH(gfV`bVx2lE#_D4_u{v3#q_)iW;XHi#=z3kzR27$Ce+kv^uRaAkmT-?b zk7{%qNbQQr-BubsLwfEsS@Tl(CB zn(30f^nK{)x}w)_nh8=n=*kOTrouveToh>e3yoc|7tqLaL#*osw0`+;#Rs7>30HLkiWUXeMIz@Jn2e(6@pEBD zETy2uWcT`qsPiHnyjpmb3S@UJKii09KHX;;BmsmaT2p*!mlnPq_y1rOx~nt9uN`sRxXl{-BZ-Lyq zD&C5>uG}4ohf3Kt5JI8BEG-e&S!aR)C6MbObMZ-QPo4!%Xld!aW9$~XE=ge=JBYzc zdInCXc}rISuQHaIy~jT@zq&Vnpo|aRbKRuvs@qRV_yA)`$}R6EWzTs5Bc?qPxa1b3 zroht_Aw?|u8+l?c5ED^v-{$i(s}1Pi-2nukAk(*Pr&5fvIcX^ZG8_gP@74}@gV9L) z|0ClQ+GO#(L{n3m{`t8|4_W!QRDdM@y{7lZ6@DEcnOJmj^c*a zzpp{*(aVS%vQpuPgQA#t0#hb}#JJDRSK`INFZ_qxsrVPF{vUsa#NmGyq&Ju+P9s- zLG%{jG!&eD1iN!(g-2a5T#R0pzl{?Td!BCJ0#7(Pel|<(hX99pc}moZgLz~ZAH+uM z=daLq(`*q`6cElmaZf#6WJW{5T7?pb+4ZPieu%hn)$5f{Kt}l?9Od#CWzBAS&0O(M ze^I)J#Pz3dMa|pTPLWjVyX@g~j!CR8=-O&7h@eXXDt=HK&I6x+kbA44b-ir&`N({W zXON9tO-w098v!v&^e{Ubs(Y33;uXS6qcN3a2%gP8T7F_YHYqJ8)LP_L3hDUiyi0z8 z$PgVfXuMvMq?F`LpSzEubO&u zJVHVdYO?eJ_bz9M!}!|=a*bKiM-Ce(5DkyL>74_|(CIWLZ-HJr!!Fk0G!nccKs zzfFUD5r9Px9Sg1Q8>b8=dyf-o(a1}*I29p^|ia`j0frW)mF zmjREUPyLD!Di$8s$vEd)ACDwT5oRE)7e`hy z9o-@`DV#;{TxpdLt!?w*_3p3fb{k~6ttZ7$z(KiyeZV9gCkNT;X-Wq`ywYez(6V%F zyIFFX%(g6unk=~QdHDQ%6(TX!eGz*1|DV}Eu1KA|TLHa5P+Y|&RA$N7;o>CbUuS<+cI*tMXWgYbe$9Wcs;$=? zgGuR%>dt0$D7j)POKNsksM5hVRs)s&dwmuOR+4;$n4a}|C~=3W zo+Y7af+>#_V!|gbV zyvEH>af)?v;RyGUYgBBP<(xqLlDiViNo(<;NbY#~Tgsl0#+kn( zSyb2}nR+^u_rGf=$c6u~qCTtQMLAz(W%Qy==2qj_&FqJ=djSkb7dmT8Ih_#^c+M0! zM5p*p<+d^a5z;!b&z^n?8l$wXGqaCIYZl^gi<9ym>{{S-N-Zy^Nx2OQ;%_c7Zb7TK zXEtd5cY43&4Q@6wgm>m#KnjUXROKte8%&ExKY8Z{Sc!XCcclFR;2Q_d8!Xkkyz+(~p!K&| z>bVP|VFkBNF1B46sM)${bH@33#NxfRp}ZeUl80$(+5(2U@rUMD(D$7W5NtmUVntzu zlIL&J^%e{qX*M!@%~Z$VobF4dVnx0n_ok6)O4C%k2}jL{;S3|VR33c?3Mt(Uf82S9 zKOYsz&Hjk+Pu`*yFv4dhmJ12Ip&Py04T^DX6B6ScM@XT1-aZmdYg4bz z5FSS3AP670i;@n;OmzF%+e;-(fF?hBH$6Nre^A4al4%Rr^BGQj^+9i4vNgDRpLY{do-s7 z2siS{(XA&NF-_#y=+axAYeatP&!17SOAHlhXAt0lfm*cwrgGMIPTw()tfk3Yd zXqWblY;7R_3XNHJki}qY5Ka?qVb&LVlPn6DUWOpxgz|ORcho_q+*mSN~b z?d?A>ums5o%1aJH>mohrqI~Ram9A=^TZ9#5wkv$v=)YFHfEl%j*9i{he86Lx%vN8a z+IwdB^(Xbdv8fi<$8>sS>NB49na(Fnw&|qKX=ua#rZHk#)Jvj!$Kc0(d&H&pH8hd$ z*~T(@PosZYErQ|sK0x9THaT$P$KW}K_?-Pe>(>bZn}M7a4xFbj>(x}4LrNVpc#UlP zRydV9lt8e&bBf#)iS*5%Kg0FPCZ)1*(=Cx`Iquu_@0qrQ`S-rJBWuiG=xsRX-v?wb zeE+Oy=is4W%)I?-?uXH_wbypm8{8e8P4ebxF2%&3RKb<++EzxKn@xu03wHfRoY)Jm zsaEjGCfe}%=Mh$`hEJNhdScyP-uJ1EN<_0in~kQO2$n%u6HP7IG>C!U1{ z#9>K{8m=_##jtp+=S?Uw>8$oyIN=OweUWFf(LH>UmcnIXv{7pLrT!zOLji7vfTCi$ zsYiP1)~4Y?u+&J^{)fpTrV|7%-Ogy~YqiU>!S_DD9M~2+pa_N4e3Zvkam9+dUTe6XlY2tw~q^aki^swKkS4qu7ml0)6;^=Zhz>R-Vjrw9t z95tDoT(lx^?!M*r8B*L~lGT|aX4BdlOsGgfX{Vi6wJ7LC_$SSj5Gc$zk6zlwF zUH|QLfITH7G%f#iM($GVK&Oqx%ERJ#rV#tt4Y}ae%8AR$ClEbsvncJ?rtSE3#@93E zal~z_H~utnZF;X0OJMNUUJT}WtI|qdm*V$U<2&pR{2wJ`@?GA}(w_!^Tb`NybvjPO ze&|vm0G9jD5`h03upx-7@(mPj?0v|VW&N};bO2J=b=aEh*f>=Q`LY|YzBLtbmoKBQJpe11_lQR=$TeDnvPe6Gvq0r8I7 z-WI%#p-tr&6KcBR`l;nJp2oo)vma`Q;r+-Qhy(6fM=pyp!2jN&nY2m{#?suk=K_3U zY*qX8a)b6#nsZ;6BrhuM_wQKlotPS)@ zZqLe0Aqfg5B(m@M^^#xxO61hwms0;2@id7#4wtO*-|d0~Rmzf?JeA&gW0A#SlUkPO z6`i>en#%lYL#{X3@33+~wV_@O6_8fVq`%p!)a|WZZE5ahkbM!WKgNmiIQNOe*ihes zS0KbW6|2Kz)VQs{MOjNnN%e~0Ga1Ifx9ACQFz-whvT6D=u8JNm61C!PoZ6)+(?W3| zLqUz|B<4qJzOPjd2D!jDTqRF$hGkyc=@>MLs94!(CYN#9GvS9H;$jj|lT6G}E9icn z)gZv`MiaEaAdwtPSfq*|3(76j$*RUnz(%UvG z_=MY$LOlo&%i=X3?&DX;8mQhoA_hnIteq~zev@}M9#ISM{>Hj5t3nN9-6{ad@6%vP-*a5u3zHXI%Gk`kP!UcsCRh!I}ylW#?dre^nZ0mx_>34>g;Zx>FuhKDe z@<@l={w1PT3Q>6R=?tN@#Lg20@h}l=`W0BQ^i^owvOJAr!1*2!&-0yb%g0ct+Ap8~ zmYq-clc~#k7DRFvfBXH5QhW2+|(SnxO z*a&idP>?WgrMQ~7Vc5z#=7$QgyR)(L)AxbkFR_!Sg{k_eIUE}w{Yu@-c(kCPMu z5KKkdk*H<6>?`-6OTy9}P%iS4W*#@`a zsYfnndzAdGToBiCcTzJxqHpwo&>%jI#1a!vs^eQVdrS0hc}W%h(sk?2gE6J5II-Sf zQyY9-35B~`fp*wN!iz^cnsLoA$8a@ zDauNRnvtsN57kqhu=24=Q*@B@Ey08Mw`F1Lnt}`;HKB9sCPa_jWxSJG|GM$Eu>R1B zrpinV-dFsu@eNKz0Ye^s+#l*}ns)d8D3PuiRCj&9pSE-U>5KapTghei&HabP|Efd| z9^c3ry&%;*)n1O?Ki4Gx{y1WvFZChbm~fen)Q!~@KaI5N#ZH-^`XKr=VbauB8bQJ+ z{07heedC-j{$>2O;9mfJYl*Iz8(#dRw16VjrybfX%j1X}Z^p!%ezA7wUjW&yu4lIp zbjx2@Cu|Tr%lTUWsJc<&P3GI_@5L(9w$bT+oA>NOA3EwhGoZb6O~-?;1->C<{XO1qodG>I}s z3EAR_&n8CLsfuB?a|!PnFQw!`L>?KT6*MF+jyOf@`qcVi`War$G#=c%L}k+lx>#c2?dAu56RV`JsF_w$~8x03=T3}1nTm32iM)0jbP08tbC=W5M_;lkkh!{ll(e5auWOr;{ zZe|j}e^Sk!IkM`IrR|^o&=)5F#+B~VTBb$pbrs=wd*N~ZKpVJK1tUZ#bn0w~EwMX&UjjH%Kv7RjBfW3hWh1 zZs8XIq6lTnuWw0wMxg7Xjec+GlQeno_3pDv+m?vq1MA^cvH>GyyGoJK3gIV!(8p=7=f0_s?SYj;tN%13ROT-MX5bP*~Y^^~WKfQ$6!S4=c{3fOi_A1Af zi??2J`ykLuJ)y~N!3ODvXI~h{?>n4EhC0s()4Y5P78zE@jB%b4wp2bDz?e9lrmUhf zRB#7WBiC!;oN|I4q%b1JDQqShdPp$qbI7H-jnNY#Q`%j(F)=BA z9u8Zm#U)OV8mX=1Izor z0MXBljXHJ?AE5ZJCBAX}(rD-d^?C?f4GPPYi44A*AF#yu4bg-TY3@FXFP;%4o~_$e z7S=VF=P_l0@51iqW5V|f5$tkd$pJJCZTo=tI>Y-vmT9<3P5M(_gw(vR-mNP6P5e~q z%0f))Y%I>7l({6#Hn=WWf-HQ3AEleZ0?r%Gcpp7p1$8fx2!j?{JL$(?YcBtSAUlgMYu4+fP|W1t1SUDYwa+s}qT8U!b3G()fMs)Kz;Q zCfWq)Kk|_Hg~k@!Po@7?+$9yRvEsVhI+w=+BHIWQ<7tjo!VLHk~ z4YQ)Sa&ui6+o8EjBQqfCq-%+%duH~a^X2vr% z^8RcdU|sq(?FH?cj|Cb6UFXjCqfF=u+w0MmT87n6x&azI5(r{Dg*XC_QUjU2jNuZ0 zZNipQNyAutwOdsj-S+@7z+X=`)D0^qD-=8qMk1`><;XkNv=?REwEO5Qgcrb|uBTzV z|1K=$O09dpUWt6RIc-gY@bM1(tIT>Tep6eB=bhV$>dcY&cF~^QD#C?(2fdxTK0qQ~ z#qSe8BL0^=81c=P8G?%I(+_eBkeh)U?cKvWLm%TmFT=bfZvLy0uavF$Gkjk2wyP1* zS&E`+xBL#ID?z)V%%Z+7P?%{{duNQu%~H~*9hnLisp$fPt!cva6$EXO8B z{f{^$v0T(S1%lw_f{4yqjiZjV`mvoUD)IiV+bX>(o`87+twShBNOl{VV6=vHzd>ma z;?v{#2N3>%;=JbhLP&LGtd5Am+`l5B|7C$5C+3|}DYoeMyamA~4f z+_FhaeMZ!>%o5$jQEKBRiZ5Ie%|G&;dp2Vp=U;%22z5H3EFVE||Bg2=pndmmYGF88 z0eQRhSDV2Kr%ypCgl{jfo=MH>vx7263~(HhF#TJN>W ztwE?HhQw#sc>2i1%>%RRerEW+_|9qm(jbuF58YjKva@;|WZm!-E;e*>&-Xv*?h2f` zx^p#D9W`Vn+3e=8B1mVV!K)`$sylZnA_`8!4x2E)cU?#zbG`t638mr>3!~xa&_|yIt8#&^p zUHS*$&mwwc77@kXf=d5PY}QHltHeA+% z*)gSe+3hZ#rM%`dOqsXmz4&J_=d0I4H$r2y%yt`f$L7;1CfRff$4XnzNju?1;or88 zgk!zBi|;_x+Dget;TcLs)1Ha{*-ywk-7i6`lZ12k6YIBSx)`(RO?q)uDl;YcJVRj9 zJ|@6HdFngoYJcB(b^yyD{m`VxPW=1J+-tZ}JJ!uxT5Z4LX$kAr0K@`VmC%^{`)qI{ zju^ty3AAmFUn^GpK;u3;Bg}DNOcnaDf&1dk!e0Q!=SRqypz|QU=h0CrabnF-}T`qaad;wDC>6aQ~P+!0~# ze>P?zAXEbPh*DW4TEr1-qb|xLzhuRuD1xuAGV>pWpDF6Q?TY?WeXt+i>0&OtAS+%6 z)+*jjzaO3#Ltu|gF^2e3GK$tZC@)|*A5&8SlNA{%66MT)Df{Zu<+Y#EKU{}JwJS_t zcY6J_&6hs;=;i0Bcx(J7{%H5Jw5iwW3;{PU@b{Su-2FVeExMAC>2v{db=}pxft(Oa z7tu~Ddp*J0_Q77v{D@~0uzYLP^cS!1CSpv1vxtDnS%Le6-XUbQj|izP0pTV5htwCJ zom}qqBDqwStJ{tdoWNRi2Fz@G?HY+kjC{HgABnO=^EM?PpDolO28u@RpzLW3E}83_#P zhr*B72>_gV#be8Z{&9VXC`B})HB>mIVrmf)AN^O#6H>fK_n-Pf^;MCy?1TPM61x8) zwJrteRVHYn48@;5PrQpYuaQ6eB^NJA7;5=x*?-p*3TfXRJG^u7{tqTOZZ%N^)BevD zNX34(pWHolZGx*tjC^u-i?*(nuexfH5tjMU{|2QTVQlU?ld#S&p1Ykkj#u;pN=u4W z^R@F*=Z3YM5T-`1jT7WH_+ZtZ52}2=`AS*E=FZyY!@HLi9QkPuVbGTvAvcz#xNDjU zlv>DmNit;fv1#k6NPoGt>OL~b#|u&K;qWUnaAq0FUIor@m@ZwxH&%^QL||-;)9@vE zTBLY!4~B23_EG*eVj3$Z#>ZAR$?h*&goG`Wz`F_rJ0@8z+E0j$&_-yJ z(78%X@Xyzbpps;9?Xh2S^%veyE4Hp)q4YOJDJ^WI{RMy?JwYp{2LtlduI1tU8`)TU z5+kBxY$JVQPE&f<)Ne$D`un{ut)sqB8eH+w#LF7quovGYQf{6O(2**P+lTAkfQ!6M zH+_ftCbLi7u4ITS4kP<)ez4BZ)ZzQXbeAbR#;}iD7AQ_QfT^j zpUgI;;lHTM2UdzR&09yU56`fP`tZl^Wvw8}e0c-p2s&1r2ktA!f=#{eFjHo0@>uy_ z{CIVSRbSYkwAwCPHnXVLxtJV}wZJxWs`4&rT5W3i%z1@!k8rsixQDxketrvv(8kWW z%i-UNy*M585T`tZ5Czk4G1-8})nu5$bDdPVS$I}sctV`>6C;6;SYdB(vv({4>@VTFGk#GMd%GhR_eYRc^ z^7zL~6O~Kr-(;(M z((9xUg`>V4iaAHS;}>&Y{iD1+a6N^UjrjDmB8$h%@p6oEUJ*m!Wo(XHJ6|ZCkM$vA z@bqL%{+Y@OZ`bl}I&vjnZyDbR(a@Fmp^zFPS~BhL{YM@1ufVTJxVv;Xs+KRr`+fbl&x>3i zER7?MciJU){h zzj{z4*1S`*Ic`2B({;f&(mYxa9aINDRmP#595*-NVM`zW0vr?HKf%sl!&d-2vBpm@ zo$F#Rx2}0_CDc}qr}Lp4xRuT1!vVX59F!4`?=`L?QDnpbS>1rKH=O{PO~?0=_ThJ- zhK)}bualk+!MBx-yZFP$0=-i%E9Q-1n%fT-tnhf42WCa!F>dYriLNbZ5 z{DRg0FF?nz9CZ^S#83WNAX5d((o-~8al&Sxar8LaFn3>j{U|YB%@R7wuiB!OKXCs7m~?FN>BP@=d)Ibhw6_UapYTpwj~@e`ncqN3JtVv{ zGY=k2B#f`=1nD|AEK6qR;V`k2n>oDi6>ykEJ^#}p41O2KFARq{GkC^749d*)=bd$) z_X;I7)b7;Rd+d4HyR@wd{QA<7yE$wc%*^A_v#_W2Jr4GXCg28XBejcW>biS=FM9xf zthWSdAGcEt-1n$DDX;bo^v=Wrm=Cp=20(x>qCNZ0QwY=w_=-6RE!}n3ea-ZM1mnXf z`K<3T?9I}&ve0A7j&La4zO02CeMmjKi}6k_yvbz8)BhY?c84_3nmgT-6WMA2-E7Fd#1~KBPv|!u^Vj(Do%*oZQwA~ zrt)ljFIBo?nCol!L24|&gbvp0y}tmp7S~nR&@bd&)7W%fE5#Rxty{x%)m|9leGRZ= z=v-@uc4h=B_T42A zWM3X2x{A$`Ay32d$gCpxm2TAs%O%}_a$654MtqKXvf9gHZHJ(Ne25>pkAs0W~a zMEaEiC!eG51yNYgUPOKtgc;p2F(AQE^4_1Y`i^9Dnz!SS_~o#_H%=U%IuuTadYnM5 z{G`~%?@Xc&1DKe(W{%jrwYiZsZjOB1AsN z8lyjVS`Bn(&^=1p!y3;67%j@RA)O`&&VPzd$JxFKlg@`xWFF_t@R6?K0?iAbH+8N@ zXPHNoS3C}MQ!gwR`pP_{*FO*A%nMIB;SA`VOmM!jL>(JNbyCISR2gIfFraSYr#g!n zYqNh~@}IWeAPD>ZuIk%1H$~EqcKxYgQRu)ndewVr67zY2k9+fJ(!yfAhetBQKqUJN z*%cB)lz1@A71N{v&DT9onG7^Pfnx123L^ESgdf0!mpe|uTsT?lKCS9o!F&Lpb~fc6KnfIOCZ^&PD) zASW3c!v>I3DqYlOm?O^w47!(>&wjm zB{-6ZPZMuVg?1a@V8pAC0|k^Ve@?1BUO-A5Z1Brp8DUp}&(H7eM7mL{Y%iFpsw){J zIr@(zhzW4da5U#~Fi1BI=|4)bb_r}fW)s`yZo2S)340vr-??-Y(wgJu^jQ>8fLxV& ziUf2iju=j4GnT>?H$B7)g^o)@wH&PQ!gr9nzc()7j=l7i5f<&Su4sxZ?T82+f&tYD8;vwyvu(vaG@(Nts7NG1+mF|Mu@XCDs2B%17tGRx9MdP=?GrZ|X)3En~q&QQpY z^F#D~w5{S%?I5M*U98>;`$R*FSh=YmzSCbrHw`9gX}?%iHoITCW3Po$_PfFji+{{) z3fmv6KS{ZuMcuG_7Hr+{E@V!3CRko>-+wSIUUHMTtA4^6JI$X&GeIk3^9EdFZz4{H zMHcYfvOrT|a1uWj9T-fJh{Sb)rQ$g?*kKCTng*P(lE0nmLom$25e_~9YOZu7UVEy1q@Y_8}2CR-7sl7$ckj@#mO&3=gD`km)Wv9v*UJ`G2@jc9D>*#we1B0 zT{+#(g#$M`$K_Ni&Wa)%V4i4Jb}>elLXB4OM2O^6)RK){ZWB>=NZMpv^Q$qkkjL>> z#}G5Bs%ys!p%4SnM;~vjIkZr1>u;L=wNSkiZiTNf?Msm!^sRg_U>BDGj~~Kc{SlmB z4;4H^4z)PnAJ?@*JBrOoc>2WW$Ezl44-M4inoAoiBPbF3TD<$#y5$}A-afO6`8znz zIreTL-S2?SUnO%R8D`UjOXg|4CO$LH0*`GG zol&6r!3vW7GWC;mX^|K@Y6kc`cevf$(fq_YR;wD4zsCK{yqiP`Z^Ge4C;UT;63y~T zUBrXhPN*KfieoIhn*#)z^I1<5<%2BK92dbEgQLUOOG1KP$qp_8zgm>UHnq@4QVEx={47YDAw}oVSM!&3P9-*G zZ-8IZBR2CBBV~}IWg^|1v6UJ6l~eE7u?JX(_5u@8;M`miH*^mwVRXJJ&u=n#m4fi0 zu707pJvddgX)MyWMyVpN(;h%6udX3EdwbuWRuTMuryPEpXJ&;+YJn_W$B2XF)4!C# ztROo-uTz9dOuX6CyZ_=*TyV1G$Ad~CF1{Ait0Esc?ceX%1=^ytFGQY7$&Ac6tbeR% z%~Pu2!2&Eay;EPEUVw`EuV7YPx3Cu3A^IPvQ){N5` zHoSZ|bn7Q5b!AySYCH6TA}#r7d6$n`kK^p)0|odCPgG->C(aSID^1mdY0cdif7;)c zA5AWvq2CYpd=7&bfiUP0sdn#pc$*t6=v3+D2Poh4(4q%y!d0V-`3&6o3uGph) z4OG>R1qQOVo<+seeD%w3Ke!<-@3`is9EzVKB+?1(Lg`@dBPCi)1qju zOrT&0npdG@i0yNB3c<7c!prv#nY_l-!aSt!WQSnSbbW9@<&L^C-$H^(&%az$2L3#M zF)~7^B-TXg-mf<~GUs%XF5F+!B&XLE3chkIt{13V_`Lp0=~*SuB%uvj<(%3H@~4Hj zXjC70cyynsq86o@p+D!b?q?v2OBpHuG-DuaMT@#@oH?w0bSZ1~@Ns@_Vw^6@SkQ5& z<>?2$Y6?2HgdlH2ByxVa_?~?wqJ{a1(E)>371K|;&{Hxms^^FUWfee*ND8Vs!zjdg z4P6CyH;rcYrjP*0L0eBe)`)KDSs@S*H#FNsm3XU9^1ParZ%-~RkSaq&;(a?P5qvi`n`uFi={LCga|fO5}jl2YHndpD&q(n-gOUrLk!dYlPig zO03h_lOw7Zcwg_rb?IJ9V?RG1_n!Jy9L__t-Z%d4?7(qbj_2i!Azp{czJvOYixa|V zwOVRYYa=F@52A$GRH;?23e70564y$zeq@RDR>aD3Q;jXzk*VWLBIB`A<6Zy581tJ| z|0ZIm#3-;#Xg;p1K;XLuaNEV7SAx^s$C9RsUvjfRenv)Tei1BPU*iku#P!O*qa*a( zDJGG`H970=N|vdLk0c5FtkSsP}hILc*B58N21P^|z$Qruj72TwpNg`bnX{*Y=+xjFc22R90FyJJQ!_fe1RTNj z3~P^M6&eBQ?}A4>jy1}SjRGQ)t~9g?ruxe?RUrpFXCgT-wev$M_2hv^m_6#lMSX^w zbj@H_Qyznahj5O^=q?(C^WUNl>3;!!oK-A&2bViQTH}Y|-@g3?AXBo}pK+kcX0m)@ zr1fAw-0!VJ>g+I*D1)O3uB6LzKvmaSL%PlS^&q>8Z`7}kC?3BfFs zwJi>qx3VwjPuNzFY{9f-tG_%BzZ0PsVDmD7uGu7eC;) zto>{Wx5BHQsrKI96<@mqGZx*CFK>iQ?B0VJCw6x*0r}FKtfH+U6BAG(-&_@-9Al+m|KgQmi+iIE`hmaou}SYfG1|EIs;(~+70YoR{uY3v zSg#~o<_17w^QE{=7b$4yeMslGG3KInMdPTGmXt5k_|H+94^6{@k6erU=V1$sBohML z_$~K}B3-K~z@fINhe({A%zk`|v{d?g8{8FaU!G~+?dUZ(m<8?ARN}3&)eaO9oSA-E z8N>BHg_!%t@$VI{v<|i#w)-MDzFNhwQE5x++lSLDJt02BIqLw1y535){I!V(a5uI% zZ&|V`k}PsDanAgyAvRxW9?dgi<10XE&gg)1Ig84}Ta7gL34^^epaZD{DAfr2itGcL z1Z^YN(O$0Qsa4_jHvFk;3o#v2sffst02&v@O@J*L=kj>B$s*7hJSEX|6uhD| zL!VJk?9Z6RIyPk5`FG8@4As}@gr1LjuS2~C^)!qmn0N4PhJQ_zsx5pPQJ5*a{3TF3 zxoDJZL-ABJpizn=-ws*VH zC00=!LaW_&Pe!O}-7(TNv@x!mGNrYrh)51I%?QAL==UgQ>;CZKdVpzZ1QGXQ9O+$h zSI}Pof11fgBT z!JI777n#0WzfKvgZ~|$T9=W2Mih_H;uYefGAyc$R`hKK)2LRrfgXTgx!F#bb?Kg@CXnonIHR3fM5Pye!KyM|a`L zB*^Y~0kaUPnXqwNH_H)A=Gl^yNAqpF!az*8FqDNqwrj?V>Wv$I%V%1~hv-)0nU)`#900A5boZ=SiM_H{Fv% z!@i(ICr&=E5aT|hhf5S!;Y#*Y;n}R}&uj8D9l0YPqK!Z)Lt|$D!kwf?m3;PDL}x9pp_R614uPHP#t~}2{zR% z$`d9-sBXx0{F>((9YzQ?fwP={=|kpQ$!`3XK5>@to&Mg=&vqU$Be)tf_iWZn-)z6(2s@B_EgYuP zTI^m>mVceE<4OaI(urRRj`Ogu$213!U_@zGK|4M)9Hn($eX95%#O4!Gf4$CnBzny6 z=^K0s9}e`+hjL{grujskUGEm!iE_uL?bFj$P?1qoFI?Y(bBzRXgT3PLYDC@PZ#C_r zO3e%5ind2o+%Lw2twif%bhuqB>P9)#$i=L@UwvU=%sb3>A( z#&{R3V7JeSO+jcfL(v$X2abIo0;nkawaHEyB_W%i>(|qO9ce*AmaI?n)4y%HR4|pH zFfZY`!v!ja8`XRZs#}e#rZdupXJk{+uZ>&_t}w+{1-(Mn7|c&VJY69^&e}BeubR_l zB=URtYRGY%KUcH;i-{M!$!%@t1YA1`-^nGea@$Qa6k=yjN48EagB6|)Mjx;mG@`>& z=IPCrsDg$NHEsQqAr4H*%Epp}by_yAkzbp8OmHy`!nRu(C#4S!$Ry*~S{sZ%=s0o5kfnn)t~-J*pYVBss>tn|)CgffvNDOU5=_qNhumj#CDA zu;CvyR_b%Frit|C!_0pzgPx?Ll+EdC8~I7(vhyt|g;Y%#ju4-=5;l#R z3rP-Y$o_M7{%CFzWL6i)gaSl3b?Oo=I_z`*5z*jVbCgXCZr|imOQ4Rr;JYSCloFgd zw1UO&BzX(i0Y+NgXy1d_iApOEuCdSBq_T9AG7?$QUlUgo(Z8QLDH|?9N|q$We2{eu zHEE0GO6mZ%6Wh)9RGh*bsa!WHC;P^ z0Ys{FrN0USQX{m#kgAAWcVlC`lo%D%{tXEfOI>* z%k9=aKUU7by^x}jFF?tC(LUe+`To_WKrr9%E`%h%#I5eTSbyyOMC2}888fZXF~ukY zi4ih=DOFFCw@x{tSBwFC#63D>TFfzkC%9Zx(*Dps|8dEy*|i4Ff;L-tN%b?ypm{h) zm$w#a20mGnnTGFoilA-5_hM*}nRUM`Uz!{8_t}#N{bdW?x@GFmjkYc zInS!D3ew2U@r0snb@SpXJuOkT#w7V-^R}lg9^iDdr1kcYT*1$nybD6>F5>bVr}+2k1Ni&=h)3_Vl&5z6?^2Hg z`ToEDo2$%=I!7 z=I49M0m<|j+3!bBsF$K|8ZFqD9vX{%p5KVX!GE!ZeRn>v2!mds+0g z+%~oLF&A+&#i!;WT;pfiLo`fT7nR+oT_^|3QjLmj1?Rn7_NJS9hSW%1Gp@6n+mOx; zy#l#Z)x&MJEM0FsWlVW#;>^6DF$m38G3FpQx>p?tB1}!;iC$3r;?9$n-%B;~UtIh$ zUdp;Kwj@&ZTf(4}3_iw*$%~nE=RW%P8@qj>l?pCwj{q#QJEWIKe!-YYFBAi94`_5E z`_nq0^b7sW9VG;V^^mU6A?QQbJ@wM>ro$kR^&&;7awu=OXh z0L&YUPV1n{LYooDGMZ_yA(Qr1f#a4z>Ds>N`OOZQulSp+&?K>U4Jg4hVtnnk$PFQ1 zz5{xjMHD3NY?k*As55kqa^kl7XfDduLp&ePN&j5RpmNGmw)G$eD{3lNs64Hh`&1q` z;5kro>jlrPwd0sK>(rJqksl6`LzRdylBTVjRD_3oORsm(WQ}GE(JHj3YeDGZYZUQx zq@|}A4NKY>cfjo@N*wp^WVtUnm-`2~KG4@7*qy?5t60WJlA)npaUm~(Op+x5kr8^) zRb*$*wLK&m44VlPK{pIU$zo#(sgb)nsG_Cl-~QOvlKht%$qrEli8F=sQ`X>v4#}`l z*>1HWTbG#i8KlvdU(W;mV{B1Ge9{QMcYH;l(Z}^ z;;%}vf$1u=5#`YDm!-xjp7Q((CB+ z2h<@TGLa#t;=O&Yp&9Iuzi1!!xvzZd)5#6nY}_2%+gS4Wr$$IeLo?y%?#2&@M{EkA z9oF$FwxSP5{L5ZRRfRl{H*hwz{I1`JCb)x=&8lWzpZPmph_TPCE~_^vP;#boQ+}qb zdjC%KcDMsiI;~U9HFCwKdkJsn7hj6`I^!r%ALU*iox`5FMyWQGt3OMQ7vgp%avr{Z(!{;__*P7!2J@ukn@-TOVbGme^Dq`eB)R`jFt^zh!~6Bh2PRAu1h zPgF3ZlCC=gB2J^pKHf{o`(8j5Kj$|%n<*R9B5ggsk3seO&uG?vEIJBmO_Z?0H8^uA zevmKo5C0hD5e6a!s?W(Va`FPxJ+i5=Q=o!ba_Ks@aB50}PyWm);}IQM!H{GE!&@!o z=QBzmM+I>1<|g>g;F*Kv{V20%AEtx_EpyhMTf79dpb9s`Pb7_{oTh?fvO+kg z4R5jJjYvL6hLeWHUTcxl6uuwI8~;z@PylLQKuAMy?P-l$_He@7uAp19%p0}KT1Njl zWP2@KNr|G)DmG88pHT7{GP&tfk$-c(tpYvrui04+D2_2 zd#GVPiH*K5({ux_mQH`@EawVEO(cRptvl4m4{rXb1pnDX0$(sn#J$Vtmd zQY$pH6mOmqo|qlTJI%0x^s?Kns2Mj?t|bR`$nCHO;dZ%+vQ#s5NkJ=*9^am6*fjA; z!Ry0e5BLkp_3ZXngAZy^T1*cf2~)I4c6|KZ5xuQ$VE1oXegAMYCeW85 zhO)RrPcg$XpZYP0*8SA}hoxgU5G~Q}f=W#@Q?H@Cyb;@>%C-IvgO^L~)fuIk5l@Z!SHXCJo#xtYog8BBl5&L_$lf zG(t?A`!o@a?DSD>$G@4im422h-fY}jl|_Y+tR$CL5lxq#?4b#Ga8_?qwv{*WNCRdl4SodtXr4*2iNE zWQJu%wBKs!zYX&klPs3U%SAuUY(@|lSMVB{WOiTYcSc%25G3yo4DM~$F9HhqU`orq z5RH_1%v0vWOr_=fO#{Y$sRyY;HggEz|F_vq2Q90TWx4RU*nF4`(qr93T2Nn!=>B(R7vEUBVbSa#Hw{_b9igpmKM&(^cHIPpx%#{7mi)Py+10IZ?@#;5(n4UP5W#-fFYn)>d2V%fGe z^I^vEgsw-;A^m~DUoBi-OgI^TDQHu)|6Gry1AV>aDma1{R_x!=L^R0&DtcdoTNl4# zrhPY@P0-^D^j@*^6y%+?xkqb&*UkEUR>OO1c;1O6ykG|+d8KE(X$NhhoYw06Byx?< zgx$7*C`U{!!Awe>UVtdA)#sIi>D!-A*J+u|GuL?ZE7^|Y3Z;=omR_t98e(87!z*4D zu=-+=$7Vpf!(H(=q~^mUd1N6yF7N^KVRN-?KL$1XW1-cPXa-9k&jPD6*kv_8*yGGY zpv@iAGYaGei=Z;i8(wgAmibvp5^D|1ZFanod%GYVxk@Aedm(+djmOpw8RAot{Z7J| zs(=00s?aLGx;T_p7eql4=k^46=^+32{@c7d^Z(BwiXPWQHCS6+9W&t@qSouH`^V2_$%AQ>yo#<@KsFJ(P}6zR*4{%aM3wp69*=eo^3K@Qi6$h{$Szten>%#h9sTX!PqU=>X{}+M{{7J2kJ&pG zc1YzAiNru`hpkKprWl2u93C2=4pOGFGa%TdqDs`VJvM>J=FXRvTX!8Zp!9}DWr-fQ z8(*B|5TPmArQ)(TnQZ(Xg2B#pTa8f2?{|W|xLTR63c$5VfT3MXU?2`n7>ZD+7H}YT z+rzvcsJ=(`{3pbst?r?|-E*rq`NvyRT>0$mYL>zu-~;%}3w8Bg?cvBXeAwb~l*)OU z%pOMm`4>eAVJs<7xrLuCge7yD1_7mC6VVRRAetY~c~~KsJ^bnH1-?>|r-C|2>2JMxqSBnwTmiw$M#;f_YT)sahC z*nd;O@D-J^{gqcEDWd{KT^V_)Ao^uaceB9u;i68LJr!@+#$jVCAU!!{REtMD(<;wR zAJRgiEnKDRTwGBGWQElg92vZ3vihvV?v89N7Gl@$K9UmB_Z6<3l#Csu-Rj_j9u0qPJT}3u8^r6|D*0?FEKo*hR%4_cJy{;H!03J)Abze-<7?g_pI6r#g zrBk*EQoQ-a<&OWMkATEaAJgWR?^L5}bB~p0N=n74H~UP+2VZ?ibTw$VimUuUW({w8 zRenHZjy`+Vp&3=akD1uRG5lr^#NNl=fm8a)W22W_(Xbtu-*W3ozLM6i=rdjpgS7w$ zs|^OyLLvX6QP=%L;&|y*QG*HQHL#K6s+gsf*9XlUdUd<}@n&HMw(9Xy(X)S#75AqN zN|MZ@+Xmi_o9<0zh^1Rcyp9^Gr&@IMmyl~l99om9U+I0~Fr>htGvuE8v0Hxm)_9UC z6Z>XzWtxGGw9{$>;5K=#vla7wXe{9O^mnod;6MmUQDo*%w4b9oMJAzpIWF+#-XUm} zY1(_xqtHOt06txigk+0i-a*XEsK zzR-U@vPkDVdq?XhijWs~z2+_9hx0`|K@jFjB4qV3g1vYs{B&hOjF&~S{>oI-VsHQ+Lb)kp|j5J;R`s|s!U9H$`>yt zSBakWU+HTw>ETJ8-s$-r%29(iZD>MBo{M9JGF8Ez9=mc$T>^U}60E+ypE>EjpeS+K zJ+v-7r4XWw^Mp!d_Kj3MffK)bOL%0FrN1?+w>RENXhBoNX#*OiQ}=go3F&ovT{A+_eeB=C9Db_DaNS@> z;C{%9;w>os!in@IGmu?foJ+Z4#2`aoJ4l5rgioAJDM4VTXW3^6*_Qq*F6zZj_mW4~ z*wcQ=aeC6Z7}KTN*mo7$d>$uoOEV8v-xgs-KA%w5CWQoDqgNIh0imqMW$g*og6z%7 zp8K#`qvaU7Eb7dGUpp8&--TGF4zuO*Ehwc~k<}`mD}fDyDtCE1u(>&WOj%3vV$aUE*C|(V!db$m&daF;_EEE+s~Q{(54iEx=%=r8DxPaW0G%4jw1=S)NTcBro-nkk?8mpqdz`tb5vygSCXR=)E+EbE z@TvClUIfWE_pb~Ba*+0OQA5)T7FOeP9d}*+Y-N$_?}qN{d$`6lmRj)`R1b!I2kG_! zK0!uYAevpl@srpa%XJTO7X6^HeYlijntYx!li2dFp)FWkShb$oWitY$gSki|&{FHn ziKa&iLN zknqhvAlnwV8UyENs11&AIq20TXc;4+>7Fhlp&97h^|3|yfhdX0fXK8X=!Wa?uVZ)O zI^X~HYFbh4HeLloP^Zg&0uTZ_NzgeS&N36RIvz`|+Y_-m3LUvP$aVMt?4Y1DfOeGU z4WJp~q!bs`c#%{62V^fYvd>@msU9d#vJp0sn^QEarxkyJ1Hdxe79%v({7k7NKj*W6WA|@vZddo~yFmez^h)4t2S%OLi$?`_+L8UeFre5Sx<$bN4&4H7{svmK z4`PX|&jBj=!6&ID09C0ZwlO?$1aSNT@zYF&dF)rWh0U4iPc>{GhbCVX5h%{zj9(lS zZ9*rgn%w~Uy?@8Hxe)n_9H2$I%J`QN`ubVO5@RXK?}{C1C(4hlqTBiM%7E2FwVW;S zBvya%^ghDQK|s!cm9CH~h3OzP(~k*Px&rHET;L>x8vv*7{~Fq=9nSSXL-UK8Y$IG0 zLQk($3?~%;xW?=JMNr;t)ovrEmV8e?Y-k zh_37&+}!f_YXdM!9V7g;5P-ZJD-4VeG_J&|Ao(+4S9<$9>Zl{v$C~)yiB;&n)dt{W z8CFG=4jdgWp%*zPe?Uj@NQg3zN4_(Lry@rE3EqN zi=!(5s11X2)W#EK)0=g*t|q;IK)JxS4Y>agXmtdLv5@=&B5(kqCKqPV3)9APW|2Rj ZBAu{5pt7_-pbdXu`gRB`K9PSW{{uzC4|xCp literal 0 HcmV?d00001 diff --git a/en/2023.4/assets/freqai_pytorch-diagram.png b/en/2023.4/assets/freqai_pytorch-diagram.png new file mode 100644 index 0000000000000000000000000000000000000000..f48ebae2570b76ad7cea67c4084b350db3491894 GIT binary patch literal 18611 zcmch<2V7M9k}kRs1tV+`1x3KEARaYdbHDN_jt;TFs9uOUsNffu<~CnHp1$%rMME%Tm4iu%64RYRJ>2;G@E<>iCuCCOP? za>WnZxM)d`((_hJL4( zQv2V(f$sxaTH5UF>~M1(_}Z1B6sPDHS(u;C&dbB|;r;uob1vuAD>byWb1Eu?e*I3r z?1@Qg^)sdV!+OGSl5CO^za$D$C; z&URl%R8%xOCx_*}```_;%1K@3#>Bj*tA#1UiO4MtSTSFsDD$<-`pzvkZEfvcd36<) z5SaM(NcKQ)Z~C!Q7n+J#H8nIW_qNxF`xZpHb?)6u2@WPJ#b8>BSbKYV>Kt;;s3k@i z3$$v%qQJ`teW10ab1REBF@z8O8-VGMfIt!r)TBJ z;GmV?w+FCT!WRMtQ&gMs3ku5Me*JlR779he!om|tA$P92-k>ZpExccA9?#Xq%yF3v z7e0TV^`T7aJw3Iji90Jl9}O%nE@I%axw*NWw`UoqO)RNuD=J#a-bb2lyb;t9wZ*N^ zQyETe^>!C{h39ctjs@Pix+gZVE8Oi(=o=V#=URCL5#e>@Ip%x0NmFe$ahkz{jOnL> ze&5#ng!zC{Behv=Lk5avB{E%IU0b>J>9)9u?uyFF9k%G3(`K;wS+lW9YP#};q^Zn%|o#D-6_nd9~%N3L6vUM4v}SXD51dZnr6DS!7n3aVA&;+gmo0fzXYmcj%h&R4sy zJk$$^H6HJdk3K6n=aiTpP;cU@q!i@fjxFl6OTIACo0v5_9~r$lY$;V_p#^UcTfnX71rvs)CiNk6HZCy zcRomZLOVcNbWOv>)pd4rBBbA+De(gYAFrRMjAqJ$D;Fyi!dE+_$gLKK%WI|tTl;!@ zt-inGd~@gONxyGHm2L%TFSN7XcUo3eWoO?^@ZKhze0e*&prD`J``3#|PAWZg8U~o1 z15)*Qe{>dn55K?jK5%JQnK#Xvjmh;2FSS4oQgnMJb@}(zu-EpVe!kkwt0Al_AQIk( zf&H1Hk+a6}@+VoX#u_aPYAuE27pi8CBadHE(HS{Q*kmR(thi6&Kh*$>3#HzE!zA3^t)^# zL`kU%9!mHFG4A+ii^C|95+UjK4S^Ruem65I`4J_5-n?`S(XqJ>$I%t3v7ONk=XTMJ znZM2{cl71yX=-Ub6k4%~+Q!xH>FsUz*V6r>N&oihOn00^O;&`m zP_v>jJev*<5hOGigW;?G-NG`vu<+JgCp_)vH$MfTO*@-?VFf zpU*R>)FLIj9BQruM<05>j}N}B{`8}_?)h+UsMIDJ*2Yvw1Fj)LuQ_f#*OzB)W#!M5 zcIpCaD1)Z9_8b_prjAa``qEgz(2(x8up@}75Sn2epV{A8(l#FIK?w=wtm8S$&sV@NU&nhgB$xpJu{M~2=@Bb>*kk9EXt%xiDj z=*iEQ=nnDP4&16Y;pFAxi#>i0vJ^B3NgyF2zhAE}j^xxH?9}#8goxAmG>7TkpzI1~ zRA7)h*k3weH^d=`gf*Ibq%fD(Y4l^niS*SgSmCvGoP*7m z=f^5$-Upn#coRVw{aDb){i-{sV!1F>Dm4Ah^oY-Nm<(O#OlRsEZfu_~B#a1A-9}gH zS7%>ehUZeXcG7lFX4keptc_N_ewMbOJEIDG=$62v9d-l}!-}jZ(t6CKT=a5<&57kS zl@7jk1bpl#Y8YZ%O0rIC?vT$3|99qE|K@V=hyP#6+W(y!F!R$QzFlRejbx>zr7>j0 zw@g-NI=6cZyv>D9e|~*VSwIq>;Fd_Zg+vOpLb~Am2O_Vht{(BQEfzk2NIX+Y5;JIP zYs>hCs1C)JtDU@JA<%dXk!pl|SJ4kK4*e1H-a}SU2jevmunqQu)}3OC(GT%s2%=Ge!~|2c>uOx}IgzNInj zn2tq4dP|Eur1EA4%6x?)yT85-VeH}>f@m_?^e{&ajl8_PYPT8k`!ROlp!gCpdr3@) z9S&x{>?6LvIUyJA(hm87pj_Iikul^775E#qdK35RKU3@r>6<2b7CSLWI1b|QH41<3TW(vV&%h!j_tKM(p;Ro zSoZt(w_*3K?d%|8XJ=*Iv$3)H@jITOPX|*`p=)4Zu-7t6Al#FZk`kdSU=fov;TuGA z(PnOo#V2U)@AW5nS3cHa4$?p<2@90EtW*j<4ZAoFeWpz<7CkrN+PK6z5zINrRX#%{ zImBpyt@#mL-hr!^`x?b=aEYV8h?VXBM`MYYD09LVK8akT>{_tE?M}-cH@)+9$Wfc~7VkaM6%^Jhs{vS#CmFsC! zh50N-B!@~fG8%%VQ(+@q-Q3D({qGdns}#68i*=8V(uI8S7am-9?YjqK=15e%LAgYX z{>FS#c{b zN4898(^%b(Pz2N$*|%$Ok|a!*qzj8N#XPS0g@13v5Wu8^et6u7fkLjb)}qpYLWShh zIX$NGoUE)u#`*{&9A*KF#S)z;(WX6jY)~s}J*%$$)$Not^XqZb!Sz|0>IE)yedY#! zPuI$lVaSum<(Pfvr>-$8P1Y2JZCc=(3#5-K$w&Bu3vskAPC+&D?`oe|dmZJfwY4?k z@Ry1>ZcSFnO!w0C2t?q70{iw$qd~)kaAgD2!U10vy#3Y?yWz^!FxlD&M#$zdtDPrK z;}(5h@>$Z3ELSkv=VS+4$W~+3c;k;;+kIGk?Cqz*e}pVmdXqEslR5VSQ@kH@*yqBR>ilGLfR~;ogZie>bDK{ z+T3WlM5T{P{@WNr<+678N}XX0?u#5@hOO!9hDC?lWv1Kj-10a)K*L-x-ePz7glfh@ z%*}@6Zi=LkwQM7FEj&F-6}Y_T{37AQ){1+^8RXkCCn>|IusQ9NpkdAl-b)_&6> zJW{Bg(AVFOgH_$j-=9t;E1z}9f?TL6?D}ZUsbq$pzCQVItvuayh#_+H)%VndHoH%S zmsoXY#7FX@br$_ziB!>(-4=!OPVAHK(Mpm=xg3Qth#{Cnd3bp6xc@d|(4B-8TwJ$>M@=>`I3S5Da*%|7%0iIpiD`tmu@7JKTUA|sLNBq{ zmgmxfn{cXA@7MC*KiS*&v~Z6E;<=_}8MuC2kdIgx*VNwF9?`a~GI8shsQO&*K9{-k z#xUFUE!n8nY~8HL1a(?|)+g~L5h{Xat(lAK=+LY;U2#oK%@|w8;Cg_W`A%s-Kjv3< zjvf6xIrh}6VQuX1a)V)xQMvUC2K;z)u|*4oY6csYOvPHex7JDWPsz9ehd`a5wL8Zf z`*`tFycf>v7+;*wfb0IYo+=!3q1Pa3zv4-_o^EIqN^uZR<^JhzYUd`^>>~k@D;FKd zbnolw3IAY=#72KbXAz)l;_lxngR`shBieCs1D4!Il|vA z^s^IIui=@btF~puFbYKiYL1C^O4pk__Ar{+UM*I{nM(2ZpRZEYeySRMq#wA)b^~%n z!7j5wjNK`usH3MRO{e%_7Bx9jl)0SRy@rPr3(rk@A!)$bS1Kt)v2xO$RC{E9&EDRg z@A3&>ivI81km4p=^6>JeKV3YB-1-DYOI=b?Q6X~R^C8th)fQ|93=r~vh|N4Mvl`uh z5+Nr)3_Bx8D91!Y(A>tsaJEN>u`Lzeby9;w%7FA5-mU~Px>$aP#GjT3@={vuq;bbp zqGdj6ocXOjs~a)=4s^{okkb#)DhXcy13+Vo5-DQ*%hKAKi}uKGv#}blp-+yfFh>SC zzsrwbx3un%|M|_2ydSE|E{7o~o_L^gBAOG1eBXJ#uyLBh*vt-(O+XKCDEiPMiI|jZ z<{P^ZKBLP2KCW zuWVlG4;7mFiETBbnC#~(5@_mrnAILd?#}oeZ1^bo`HI8<1P3MV=_66#oP;Y-cH*^b zx8KIb#g%%jS%Y_~Uy}~(*x%dn5?41g%um{zx2Tv5S2#(@ShO6Gw0q}ZGvweyO^q1i zPQ@+TfiFzYUXj!1L%1Q>812NeSFft&zd3?Dw?>O(vwF;e_lenjgG!Ne8m&dEXt@Ky z;?>o=*At(NpJ+=Ec9T%k)zZ3-!|;LucLY8%le@j*Sf zw!S_KC1@|2vq5o5p@v8_Jb`-Y3QIxleg))4&R?~HIxo@6Pt>MOqTaxSiKuT(em@ZgDwDYeE?GD~6ap-5x$OPJsc+BNh zEMRcs?V(Z8(EzzB6jZOrdM^hUS$_TN#P2>cl+s^lQ z0cg+!r~)gxq0h_98y0w)i(GUy6u5+QKAmAZH#eKYuDNt8@e{~>HeaCS^V{fXo1qe` zOXA*D_LU2KV3FO%lZdZLJ&+3$Y8+L+o?l?oP2;ucn|Ws%d=HB9>Gq_8)Lq$NEuZP(9tjt6YdDvpu-tIKH2$5*7i6H*fnbeTk zP8s^%#gR&0yCEJxa%RT^IM^TcAih@*pX#%-zR=zBmc?Z{#$xvCDUKdJhwAjH)@YVX zL9sne`G>nW26xhq}VC?29nn zBnP!f$8fn5dJ{m*m{o7zlnV+9Lg(qdzmgOYO7>@}Oh_MhyGUxcinzhSGX25XO(%T3 zQ*acUUWP%XYgR^vbi4RL6;@#4wAZ3j61s?6Iu{XYlTU!oNQJIMu%o66-kk}_#N2F7 zK_5Jh7oucI|Hf9kkKw^~!9lhHM>P+5=lg*@f{IU9dbDA7sZIavxunfILrF_>bKOuT zY3k|apcVl;*~eeAe8#|QtrraFDcw~Pgfi?Q9AwnDA=N1{sXrof6CjaMam?e1d( zC@)b^#!N>`dhRL2(enZIJZ_7_mt_LfiU2=A@h@!y7fr3XK|s6EEX$A1L)M{o50jys zkZn^?t!1?TJDPhj74dx!rx7Cc1?FAyWZ8txs74=rELPMLV_&r_4X``soJ%brLP85A z-H-oZls90#fL6;+;UbG;wMOC)vNd1|_(V@-6y{l7aPZL^+rqkq*VvvyRKp?THU9Y6 z-e3RIqm;gUgPaHzF`bp^wr+Hrpk{wZMt*#LqxPUWidE-*Bb}!rz4vAcz`iLcs%DYk z5=H1hkilz!+J1T1o+zqY?NNGxUH`+_-lAK@kLRbHM_l>|Ei45Z04))I`35eGR_j3+ zL-GE#xUpSnbU$yhu&k99cqzgBKM;N)NDPVre=z~TPZwhw0|dh=YIEEc*dDgB;rCz+ zBhHo*Si6Cbk`P2Rq@-R*htTB}`0UM9KsE0SXk8rH86=brHs*%Qzjo<%CpT(}{mn-4 zSv1IA3lVt_u&w26PgZtr?j#n#P}*H(W+d{x*Uq}tRLfhTmG9gxKVMSKLeARtd#V+1 zb{Bwo!b44(1l9*kUFP!2PJh8;(3U!m(-s zm;2xp)(oN(9ddqx_`ZN0Prm>6vjlbiM9)P*r9=@OZ6mMJX>3Byy;SKKGV&YDG<>$I zc6NEFX#$XIvpzobvOkLSAW z$dienSOLfPgbrWH%z^?H;7PS%Ac-Ec%t1hS7#(c)m$T6!CqsbjReW4S%>l8^ra!;8 znJFn3s#`*KPJS2E6^eeOr%ymE1Q;rT&VBOi9zZ~y0H!pIoIq3sb{S41?><_z)AjcC z%@!x^<+7JgQdGmSX>DuU1N0lZtKp$XjydlxH_kyI?pa^JX+p$>6`l=OVDHQn-?JJ4 z>~y=;Xx~KW4np07hMi+2E^}HiBi4QO^MAM4udQ0XS0}2Lw}FgNMUM(_o;dpoxdK>VyY*x#Ay*~|Wi7NFFs`+9;CR(X0yoM7!&e*Io#sDJ=I7ES#}os0CVvqn4s{`dygAQOV%d2W#h7FH>|*n4_dODP z_O~*mh(n3dD1`#W^#N2ikYmOQxn!fZI`vi?9Oye3nW{Bd$QA8)iw^W-kaOveZTFRJ zx}PxqV{T!gx7~Z4n@zVw7Mv!aR&)nIX$+ctfyqc@T>^$X>O0u~ zDsUY2MkZPSZvjMS@N~HzhUIG_0~-)CLm+g**jGOjlUu^kTCw}Pb<02sLC+Pf0>eBR zI(LSBR(Cks(TvU-(A$h}aOoYmlrT_a9&mxfFc(hWoZMVRl)DMyAN}BejwzPxj6~Sr z#|JO0ZEQ+`b_}BCybaoeE|-3TOu2~b=*qo00V|i*)!(1FHrH?S^Xp69N>{tx^1___ z3VG9pKu%gZ->dyVlOOE1AM^nJi3Snyl&qp6HKVN-Z4E<1dm$0zR-XGi!f_<|{=#6f z3=|~W^A_Sc=pnyUvy*>`hNcuqNHnKoNJEZX6n4$EuU^xthRWIl4j7loFd1vefrXaq z&Y`)JRbw!1hSX6c6*bHg82gcIAld7U?sGX!D~>#FWfR(4?N;(5F%EsQ(#8*jN+!&E z4oz{P!iRX>>_k9bU9}%OhBThys7k&YaKQ>98_FlZ53{yb@N*whgSlug`AS+taRS`T zJ=90hSQ0OA4>c>GTC%{zx=?#K4hifl48&7_>+Wrg`S|QA#q=9`=0U2K3c)3gjQsN- z<5koI@R1I*mH|sWjB68tfQyqLTd+UImT&%7(Gz*~31A9fXsluFw?~$|8=JY17qkE| zrIhaM&W-WHwPc>7U$?H0J;swoZw`DP}tNO?NN(lWFzk^5Uo2ReY8qHBr9s`A?~ zfM0;4bVCAK3VtC9se7!5dr^MPwlx$R+Nsh(T(opZT8T|Rd!BBoJY>ZmT^EMZv(z(Z zhinRZqic7~4FiuR_bf-;LP!c=sAxvUl=QSH|I_%%pv!`4;*j*O%^4l&0&fS~E8hUw zCe?vX40yd<7i(Yw_*u5s7KoSK|)I4gYJSMIKM1XSY`F{S=rSn#%#B zD?&q_OzF78?)xR>Z^5&sK6t_!*ePk%R~6>xsV?9HHy+zqgCq5P2dYI%88qkaJg zgyz3Eu&PV@F56ow{YTI8D;I?f+^60`T;PUvq3n5ehNk4v&)a~fTY}TY z3OZ#Rh7=&cP~SC^7iYFREOim7M6Y!nfYfu!%CxBrpQNx?&U1PFj!5F6y+Gn0!)*~u zCP^`InQLXSa>*K$Cc0(z36Ln&&V1)???Eet-@hMsrx4vryWhY|o}Lu>yFn0<5Tk7U z)g(ZQG`%#r*e}9E{-)a;Vr6~&7BM0(`wxC5(3=_$*~iTWN4B%#CxCGNA1U)8P@+$2 zI6atFlA0<7Hdn-|0AHa@i6De>0Y#q<767suv+8^k>*(28YjD*a_K4OBay~-InT7kq z17b~gIBDC9>jfJ3mtH#B|7vPnXI|yoz$Kjv5e5wq^TytvjqHx$^9>CRC1S<)BSIi! z2$F^pePMgsLzwd9)HM+hLsV8);;oZ4%A^xC?0Zod*b!`{=z%Hve{gTub z_kvo`dQ<|@JK=)A8Kru|tiP{ssPI1PoD1VltKSt|V*wUQF;u{^LV+{x85p>)(MZ`K zh`t>Houq8y6?#TT?PacpZQ?;8fU0vEO*r+`zHB6z?G*PrHu{&TzOmC^>CouuJQmh$ zyx`n*+sewS+0(szfPjjiz8WX!3)w{kxH^+OuUiu!orD52w=c<#6%-g^zQ>fhs>au| zQ@NWa6+#y#H1YRYFBCXgSXhjX(Nxw+hVm3%tRM`5mdtObm-9kE zeaKy`$D0RZ%&#Vt3xjOvUOiOPZ6X^2zj!G!p2Z= zf7n9xMgFxWj}@=<8#C+60c@5DeibG(RTuGOCh;uQpH4kr(ZM4e!q@m z)HL9BC;LpRqN({7K(W4pHnq~6l|c}PxVX5i*+e)Gb+fR$8N~BFYMilculqEzEZq9Z zR1u@dzI`{;RW7=iBBd@oZB+ zP8T`7(heU1lX7!)ZD?l)mL4p+-ofUPefy&^5`282d43l#?HrC_@%mZ=(~UvgWcI#cH#4Gfr|t2E~HldAeGvz-iHq@>lqaR zwFT_XM11`zzi-Q*BJ3P2XtBjNl|@oiZ;X=0n->wN3>N?iZDzpP|NbO;V+>c%a$!SL zYQt4_a9!iQ;n1!x4%hQ;%7-5Btiy^Kt2DLmheOhvb( zu7ldEp#;va4!De2x4{y}Kk8kjV;qCwy_jlFjeoT+fG<($uy|NK8`j1r$Jg%Qvo5^& zY0GZOwh?1O(=YO9$9|k=<4ZfHdMPPd%tJtM@EOkG<2p-wxr~@jE;jnkhUhBxOQR0^ zzp21Rz6`Yqh6<9lT`?~we}v8Bket2rmk}-__K*g$2z4aW2Zp10T&@pg05_@p?{qCL z4cbT7amXza)m>Z#!H6v(mFkZHj2-O7wis1pt^>*6X&|Ap8|FiP@n62rwaR38eN7mC z=n?n}hw)x1vYJRf^h?ut^u3ZD;^`%29)GH$eUmK-CdS*&f=S+y?AVWM9?-I#h$RhI ztF^+69N0>aC)qLiJQ18RPEh2z6hwb#`2Qtoi~AxpQWXCLPHfWo`525v_WkF|E7l4Y zS6s;^VHV!zK$3vz8)72nb8Txd0sSKQg{Q$-tg@b3ZQ1pT_PJFiJfaI+f zYXKE^&*zTV6PWS7n;PH65diyRe=E?bd?q?r5#v8r z<-D$5U=_1E29wlZkx-lEZ*%YflF#j9eqLTFXcTu@EkhGHsS+?93nsXjUS4{0yV0uh zu`i$J2KF6I^wP7{DN-11%s$9mZAq!#=8Ta2!h8$7ocwg@GFCK1B6`Y1)jqIcLL z=A`b!P%Hm3`E>P+mAegsI}5wQO|G(~_O*6#&93depY3q<{7<;C$5=RyA4cTzE;{`1 z+VQJ8F){WLwvCz2nd#hPDroSNmEX%K&t&q;V#ohVscNPAvL9ljn#n4-%eP z!5Yqee;BvfcB)I}g<@;kL5AADnT6Lc#Wy_%g3lo6FZp7lOBIt!<<9Gk*CzapqvFcI zF-E^|89$#07T#VVm_)ADe+AY!;YFq=;%NWY(Z2Qj-CEfTDn+hxnfaK7z2^Z>9ipAd zeqeu4OuFnh_tBW4tn@I)@@whTnio$ytePphHPG>g|F(6`XgpqPH^=|akw0ri$L*dM z;)oqHzKD;fGGv?>j-qi+AlOD|tUSNG60QWpC((rA`=vxj3MI&F&nVt0qQ9*ikKcE~ z<_kpc9g0}T$u}z6Zvywf8HH!136kC!QTq`e_L%I+;_qRH9f(EDronfv64%7X+ENu% zRP-;2(csTl7SrL4XPy$d(U+)~<~aTF$i#)d0|>hF8~72@m^$k4-a+@cgdM@RaC~}` zPTKC=Eh^d@s<+Vti64f1owR^{1&eoetVOOh>21eT@qS7wUk=aBi|0CIbZ#!YQ{er+ zRJW{VCmGo;TD+->+OU=?a9`YyY(qoVc3W#k(2JXxFL8$v3yoUre8!?!ixE&4ieSvv znjXe6#6{=kYBbUoL|;#zt&O>Xacuk6AunIe=RBU&KbUICb4k48uWvtzNUb&)cY$pp za(Lk+WX3>ovEwI2D+p)ZrQTI4;&a1yuFMZ>;puuJ>f@h;vPHQ%_VX6*niUE|Mo%T) zgx-O}RjxN5`|OM0O2QLY0ZoIwC%ud_JTj~o0+J*ns^~wqXBaWC;0g@72r}5YiRyn+ zSR8WyiNcD$QdmwZlgYNX?K*BVGvcFc)VfOjv&>H2Zo_s+;58imjDJgZn4}@Y5XP3f zxVhoj*x2?O-q+cM&aVvGMwqQUucJ@1dttkz*l-xCd34}JzWE@^94(FjHQ=F@7srN7 zMLk)WzWg7FF!@fjN|g@2oV7gBNd8XDalB5llAizXf^K#3hVri4mC5(PUkvK`^4%@z z9twd(l0->K33OYp2)Euy17HRc(H^1`Sqwom{K^0Xk-R^Gn7wf%s}D@$A^%^pXy=1^ zbF_2v^OYfAoa6UujbekK-!(s2eCF;elh4b;G%(z60)A?eYO$nHbgk?jq}boW1UMzD ztoe;k>L1I^%j*ST)LbBTya>u@TJO!D+ds*DsFxZ+Jc1HTkTzNfiBR)Vhv3U^QR7}Z z=9h3{Q#7Dq!1&h9@txA3OMweq)N?$M-{bg0jcd{)-mz4}>U_Cn+c1 zsReQp)z1Kal!X=Bv88nzcFd3>dLzvo_-B;Cc(gieD-eE&@gFy%eug=cfu5MW1a!gj;f5Xr#&4YnU|;~60Z|Y21`5{#-)ai?A%fO~7k5BDg;snh zw-b(*w1+t6<@3OdF0Xd0v=*@rV)9UBk+yam*$D{=#sfuWgaBYkM~p>c7S#8(V|m^O zdt(QAfUox$n}K*2FoGSPhu^6eMk@Km+KO0RoSkQeo!WV+FZv>=5-#!+imajVAn=1b zjAJeG_XB7OPm4f3^hm`E{4F4X=Z5DX1 zu}IwmeWL*Q+$BzvJGE9QV9jSYlm_6(#WyD_{7AnyBNHp^X5Y1yv{}alGt56C zeH0!9({h|>Sn~|I{WqN`qnsbD>$}k>6KB9zuo|zCT_~JYvkiUr0J+zKwuUqGV*2Pu zQ2YvImo|U>3{NW?^Nx!KK{H7y@FI1M6;P7uM5RlS2{P={)S`UP8UA5ot~&@bAa{&2 z0A1&xAOKpPgBBKm`g}vLHBFHAs=$pZTUy1pPmIIkF9lK}zPq5*>V19o8p^VmhKOo8 z;jz7g7D;(P+hr|3XIrgvoLB`6ruB~GBdDFB)Q!jz9{J!^NTI8bhVq|(g> z6hswkeR`VkBOp8pZ?z3MX+iPnz5=~B)(880vjBL^_U5=WT(qzRu&x4FhSG;^mr%F@ zV`l>*y`fs4TEZiuCxF_?e~0m8Xy#}UBGr{UTRCcTQDza8z3O^;alr!P$GxVxlRD7~ z9mO3%Jy$h<9Jyr)WvCKLm@*Ke^z1udQW<_YaqjxeLfKd!YA`^VfuUFqe{%FdK`M21 zQvnmi5Z(LVKGq;qZ4Qb|02!h6)DWJPp&0WXW@7~~Lj^1*gB(ZoYIqhXYMXNU z`m|t&yOo8N#5cc?GsvC0uBZw^VG#H;SpA-C_5c!c4G?=>R2m3?nKhRg0;$k5ibBr~ zdjrpf((UCCpInru0k~(D-=`~0ZX4@XbM9aRC5Dt@dE|U{C+OV=&mzXED3FWN*HFCE zK5qImoa}l!pcxNp3mrg}h%#!cFFBq{uHf7G&$Ih7pbuFF6j~9)p(vBeV>}Af_Uhh} zj}iUBX5&ORK;x@FUXaTRZ9cpWTH6TNy*XH{4$vv*abpFpiFtvLf4j}WI~9f1#(vZ9 z3^bMirbFwqyX*nT(Mi}>dej;|+UcRlC1>O13BiMUDkGGc0seH@yPE?5a0AytbCheA z#Mkc6)vlJednxmeICuNH!%F?CQYx#e(ubfQ!B7n_vwO%{iCgF`A9>rb#Q-9Jk_g}n z1HSGkw+HP|zt3Mn5NXf@)plfS=77oa85G%_ZE>5_9Ut3*yLSjL zRiU0hb>k{Bh#IjzYOLb>PlD*hN1bmV2<0(6W>kFlC;3BSqCY^eraEr;bouVy$k<*2 zsC?VlNb(28cekEdfJuNJEb=3)Yy!}_m&BfaBB;V6Z-Vki zo)m`i6|$mFE>s9Y4i7lIQx7^FDEc9;AU_@e!#`B*S+)<9aU9t(iIeCs0Kkr5P=ccR z5%d#K(&NkI%XLpuio?+F0NiYaS~T!t_Ck~h@&{jfOa@(vCAc{{r%iNDR_L72*u4rm{5ja#TeQ$ycQ+u7@&rpg5qC^C& z$tL>5?^8nRq8G=!oF13!f%}yFnQM6SxYLLF7Ah zfxXrTF9|+QJ6vGvD05Ugy#0-S-xwA`40X;GVC^5zL2M*iCNWIjghP^X62;0Le0KB< zAH&fh0aIHB9#mddM#c~AFaS$MjsW(9 z`|K%%zFll57i|vN0|VfrJ$Z{E{?XSH?q1z4>gwvFeP@zurN~7!fWbGbC4TeydkiipC5x}5zbhU0D<0T1>@aX^c!OjR-}1*n^~@nI6rcm^)&K1^cIjI z&m+4B24>tFs1e^U4NK!pXj2Qk{*Rh5|MzYns3Z@zH&;DPZBxzJNAm4{oN!?2AoEmVU(P(x=$pUMBpYC5Nby0HfDLvU z{4Jy$8=OTiENN<*y1*%z4uZ!{JXbFiv>U$}%nhVdaOsiexY3t0=hCkxV7}iK52-gH z?mGuPgt81K3Pt=5qn8DuAnlD@Mm6v*dZbnnP`4SeUnXNRzzUMc9`;naFS^bxF0Qk2 zSkid1Q3ceGw?Z1a9rfZxhDwsXN|Kmpfw^oE4Gm5A_G!~R{)V9+)HY6jg?z&4B7H)C z$7hw`xyyTXIx6fKAIzJd0)z+)zA08_HZxOD)#6fAKQNHAn9Tf@el-qrQzPld`ImGz zBJvsqKkb}h@Si(gG`hHGKS7*ctPkx4Y|S9qCZ3i7o#d|BT&t^|dugFj+_8}hQU-=dxvyxG&)nF6^7B!MWPI5mA`3K|z-SM$fr*Xk60{i6$Z!X-{G z^qmq$A-jU4I&*|JHEnYnvbXVta68f6AsmPD2A;K54d`qJ#)Wf)TqYf1!3Dj)wa0E@ z3`~SQ@-jgYTSLc0vQd)kNMfbn#wxj!|4nCVR;B;Ub21KQiE+-P0&MiquCfXhDQ$VXa#!U zqIl&WAU6dCeXWb}NQDcrJ&O}oOd$G4f_da*;0q&n?9?paJ+)`9EZeF<7cU5icZ+JA zTA%-WKv(v=av;X*mu*b(A^@%y2IRHt>jB)9CkMl18TOYR30yuXGt%?S%fw(XMpa-= zV>-redl73i;)`nnE^gRZkR?mZ!4cxfuC~F`pbPJz)lAO5>kc%yhq@NgizUlLFz*)G@4>10>RscJ8E(MF1ryy_M=l4xi0$gDJ{&ut7? z{Z)h7B7lgA>7A?k=>6|(dlfp>*a}+c+f^P6iq)=NZ*dkLa$Ve38?Sl0nJbqqlolIp zt5eITjEU#AWNI<5IgE<&s1c>}fBY_1r(w7ovPzc(1<5lNDS=-c|78%d0?pQGBNPra3RP3wGagj}q+QUBzL{qolUvr|#WTp`=k$>CQV_6kxnmct zYJYbTv||=){mWF+i7Bxxlp*5qrLa3jV%Z+qhmp$S4 z^BQ}c!^oCxwBlT>vQF0GNPGXfi))Qp`-Nk6HWs4(WtaJ{7i#(gdrwvujV}myn86Hk(H=k{vyGaq(!6s4T($^WU+(o-;(O1lS#r%Lqk^IWpCp<*Y!Yz(!j|xfuW8HJm98&NaJc+wlt70~7}fbraf#&=bQ6 zEaT-H3B@RcpwoLz*3&Z&w-vv0mlznvg)eP_$3KyGVDoNxw=+_DB%h|RVBfYcdNOQr z0F2;B#r28H5(#=fTsjUJxK4~ye>B`z#3}rVev@;ut%C#k(+dKzL({*wN}XVaX$?M?@R7AND0AK##SqoS~TR(B&N4oKRu-|^mM(OhUV$aO%!W63MZW=~B9acU(ab8#87rn_FDTq!;a`F9%j_ zD^x5@R55(nt5V43Q`C15!gbnIH*be)#~7|TML9K2&95wSuzS1w3h1=Dwh?gQc}by> z`yw7LOzPjC{D$*zX_PYQ0!z_1F`m&T)wxmEZEtwT@7krH z1{|x7Ex1`sU|4gDyi2^I2+2`GYfO`k+U$80P`((80?OTI$^1w~o-Wi}HaP5zuuz_m zz8Xdp6n)=tyaDL45c zMsu;GAah(q>TJng(O|Hl7mtq7$)cd$@!u84ZgjiI&z6N`T2EYQEhoKhQMtWP;9ubD zb|ARN{*UOmfFXZB5Z-mg>t$io;hXQQsoi3uZW$yumFel}V^640((ax}(7=iJa&{N^ z0KTq7^XV8YLSTWnT-Msbsn<3D*5>LIN8T#GAt1i18(0_<8RD9f7p4LU2&xhFWafo~F5XvmWH$ zEnIHqj`NE-e*AMsUr!I^L08ClZ%0Qm;FB&AfZYKQst!VuXCDwm@>bvb$Vf%Cq2Au9 ztph0ov7wh3X))li0a%0_2_PXW5MyRY-98Czky?=Npp3P{Hh@??&?*8AoiuMwZ+%@J zqcLTZKL-D6HZ!FE9d2f)U%0K8Ieqz;`rglL*j^0CUO<9s@G4dEj~uPME~tv~>Gyy6 zEsH@uorkJE0mRDqCPnvXccbCH4T-7UHnq*YZ0NoCgR))kNvlj{M_*Q_Gy&Rh==>G^ z(d-9RD7Gbv##w~FIv#8{N^R4Y9Msi#KR`{B3++^g*9&rxg2C&*@wd-JGR1ZF^k_5* zpb!LFJu-pLw<;Ps4-+@I;NM8RO%tKXVv&C|}5PJG@NpuJWDy+s4oT EA8mju#{d8T literal 0 HcmV?d00001 diff --git a/en/2023.4/assets/freqai_weight-factor.jpg b/en/2023.4/assets/freqai_weight-factor.jpg new file mode 100644 index 0000000000000000000000000000000000000000..c7580787de79862f2af2539eb904682050614a53 GIT binary patch literal 189828 zcmdqJ2UL^G*EbqOMUi3w>7XK@RO!+Jiu5KRy@~WLEl9TkDkajTMtTh$LPwhPgx-4# zJwOPM0O7_{Jm>#@@4f4O-(7dzGR%6ODbGx1ezRxqJ$q(2A3vW5TvwJ?k_QkG0RTjV zKfw7EKo&rJ;leLZa2E;w(v?e>E?&GuNQc_1o8P z+@hkUrY57Lp}kE-dy|Tq>Q^O1#Dp>zFI~NK=_(cZHFB!|X?NZNxN(J;=+Z1P(LKP0 z8$`r6h|b#pjD$vBASU|N-oM>t5<(j<5M3eUG7!>#W&SJo0`WznOP5K`Cjn%{gre7p zuM>KDruBE>|Lboe#Ku5I@F5z#Wa)bB(@AM_*n>}O% zJZ}s>%P78{>%B5VK&G+e&&0oRzQ~x&xq}sW3BOTu$Du_D&GgCAV`J=Z0 z;449kR&>MB0%k0u(yuRMy@d7A-$nlq-xuot`XNSvsv28|`$1&oB`~ zX=7dDLK(^cl#Px4+cMxEJF2x4HGCFP06$>8w-;9j?4P>-yAYX9kk2-Kcis^893acG zf#_hV02ae>54r)=D26Q~t^HJ&oW5&)0_9s;0f z)DI>WaeSx0$aDo-sdNj^<6mor zV*WwWy2DH6z$A>Le00L}ON$4|1Wc>ww4rXm_2mnB*t=STmQfF9R%OKT536opqY8OR)6z|^Xu_fcResf6t z{Xcv9XRH~HMC)oD)FMg=UCTZEhQPk}*FHuDXHAGeJG+ zcbp4n2-vNJ`Hpy0p~KU=d|2fRU;aBv_(34XK5=E;DbWGi&(2)QS?~Y&Jx2GEEGXLU zc1T<|4Kt7S^%jr}06@IdWcFCWQ{4z|rQ0Cj{MIe>Ic3K8H2~4k(o$+w#j#&ln))<_ zk6!ZNh@DQ65%X19c2zUuUhJXB+dnG(kB8qE`E?u4(4t5FR;`*=$;W0EMhHh|WViPo z{&0@3^K{8x46HqUI;N-Xd7y)?V@SaF$iw=eE?G^{&1*OeFdE(=x9BfbI<$vs*(Xj0 zknXn@bN|U$yfMqgh2s0Itd$KT`)5>T>jqZbwQ@W*hmB)e(25Ky%|jmQSVri~S|slm zje`Ay6J$StQ86|WZA!Xe0Zp&gU#w1XKFd%}hnY`QA(vXLpDli@EQmVJ^zw_RsF?te z+?0^QhQQzDKseK{q@=u8Jp`_e=ImUNZ%7av9csO(YjaJ??LOoFE>4`{UC)ZLsdIIi zR(5BZ*1bX1Mo;7m&@!>Laa4Xps~cta*1FjEDro);xn_Z~*hC?6{d+>@IJVq={0CTs z$m7abU{I^^(BoZyFH;`&xZ`tx5+Fgn7Ec-8vh5-X`ML*INUB>nLxY5+nvz?WEJX9x z>`<0R2Zg3M@3YVvX5cd^K4pc@O_g$737Cl&WCq$|77YUc^x~NQ;-UAQDaOn5>P{w0NU| z)#5p(v#E;8Gmo=Lj4xe!YIxFVjc9}3hTGjy48C1XYRO@2P2aPjT@zb1-AZx&AG%hs zd@)h(r0f2QFN1zSTE-i5QFu#yvEvHHGe@mATX)L+VWj)RB!R>_WAzyWi0Fl>!Ov0O z{EOA1JHw%BG!FoP#b?{i76l)3|6+Ocq8U-*UmB+CLcnbjk0i{{SjSPlcD@PYb3gzS ziHw+}v~+jE-F)}nuLAYN~GIP%Qz+zu@~j!uWO{u(pHbBi&7nr{hUlfQn6;j*x92~c=3!WBn|0a zvfkxkKNRkt?9fDcjn8E=f;^1g2MJHLZ}yl#H}IuZI;XVpiQf}1Rt^{`K1Hd!lC2$> z)4lE>kVY{#VHOdZK%Eu$fPvHYA(O+jByL}C^iP7Jpz+Xj+oXN{5ZvD)rG^h;mU{cE zxvT=_@~X&jrr_qH5Ucj3-nZhvQXhZrq0kb{`wo2rdpIs+xCmGyB(ebgkH63e;leqj zzSWUtXf29eTQI!4iKuvjD2B)g__W8zh1~yD`VV(+uSzpGebM^RM$;LHcky~R(>|6l z5bjlXR{wZW3)%pY?&zNx$0YT&$FoylvHWLXyIou|Jcl?id}IZ*gyY?Em8ds5@T`{d zPXTT1gTzxa{oMVOoKd2B3E~5f2L3Ww6c z{SS<}YyV*F#P4sbQCao6a2d0G&-1pLW+Hq|j6|5}_~G7xZjI1^=td}N+#g;|lf)ng z?<}DZ-^s?^_5&5Te^pCxFOlvA-?o{d1tuMWHiv?Hb-)K0!x{BT=i6Jw7PpxLX3qh0 z(p~M*?o4m+r{iOB+_(wvBWwfnWxC4uj@s&FaM0pZjrNxC$`~#cbH|qrHZ+B8 zilFJ*4x?@gl{I2hn%?f-!N}Qb)1WGavnnpzb@OQ4NFM#M^XcMQc2u1m7p#Ho&5!?< z=Tci?&>pr{`vdfJeqZRPQ2u>v26p>|gxTg8Q=;3$ zLStJTpr7ys?p?Q;sKZp(7BR!jZD{+{Vyym3iG9l1V#i*Ym-Z?$-qWD;s5a4GFPd39 z9|70S7c=2=YS2Eq7&w3u6}s)Xtu&arJB{48K1DixT?0HbKimgvr5SbF=IZX|LLJWm z&ilidRvH*$L1{BDjkd_SNV@?yXqRa)C-;)*fb82@)|UJ)-UeWy%&or-k)IEeE!AG% za`5egj39Sx(BzK=j(o8DZOgv?lH+1Vjp(K3kkK7dx0QAn6r_rq6LIb;-j^DJwXGCV0Kb5UODEmg7d&UbJ7}cl{`%E{k$8u!u zRaZ))WTPHJxzQLH0Y~SvnO?cQHeL(^WInQ`4}@dy3vS^tGl}mYC~TTiqj_ zncOCz{z^i0!jKGrC4jQA{>iZ^Nm134PEan-eXl(pi;OyKybhdV~X{WgI@X_Df*=wbUR% z_L$ytgBlhMOUM97Ag0Q=KXo&rpIcu<%*{z*PZw@X1|HnL`(bqe{1^ae@$LaXz94m% z%5U%utMJGuYwiAb0ucfznSW-!vM8R=x|J6{W-*@G@v2|t>-;ZyLNbPE$SN*TK(B-X zuUA{a@A6B;fFZJP0sM9GTjF;W^4#QKN_AK5mw0;9okhU<#Gb8%IN1#53`BRF2zYx} z4(C{)ATtfT_{Uhuoen|kZwuJT{-Ts*pA*~zM_%)iLg$w2vcHs$ zyVu=W-)W=4GcNA+XBk>G>Rhi}TDhyVbx$j5Xxn=FysBoG1IzjAM8V=C!xg?MPR?GR zU~toK{;3I7Z*oC}dqvXWqd~@aRsLi=F(~3(czl9BAy(2X*OPAH9)Pn zZHoe(+F!!vFZoNPVJw}b#FS^I-=U98xfJ-4SHFO~M4?#1^;Yp9!9KafYr#zV#v6i(qzXfEeMh<0yXXr1ub$slv`dAW=THD&$LG-SnG5gvL84WdMjb8G6a8%^zvXc1Z-OkNt z-U9Io^5#P^2UT~cjkqUzgyy_{R!9d~rLx^@NbTwa?|Thpy*%*WAEQiwSEVp;&5hg1 z@vm?7E&5qa7KsBzt=7B1(!S`ep1_%~9=}~4Gh469Yf!f9q5pvM~INnv)bf2DxCXcn6idfipfC(8hvsrx8rl) zteR7k5YoHW*9T{^j@8^6q%ea3Q+Wdw)J?q_=cEOl5aYGsfelWIvZAx8SY-I3B13x?P`_M8a}ZsvopXKqPQwN^5g3(sK;i~=!^<`BC9~AjdKyk zb^m7_V!oTc^q|UNh7JxEjOTuUxej$Y!-KAtM)%8$q=}=qCoVft@L8a%dc^hMWEjs- z9$^iOb3mHKP2XxUeaF0=$;63?68=}2&Kt_xguuUO6pKD!yjbA$DYnSv&7l~BQHCDL zvwsOC;;T!@6v9||qYFi(e18Ew+T1_FkM#vQV-EqBDGfZytE6p4C9OO9%oK{n8ouoV z$izzNO#KOCfk=ZQ$EaP*_4n5>yjSYkHM$gb;pQXGecbPM_J-!i#oZC)u|Hi0))99b zl(O(@MkDd=^)n{W0559zoPG*@i4nc3pLL!^^hUe;jnuq6U%S>`xGySYgb=x!zY^GYVvb3Jb5%8 z)|hxtM8m*yi~l&gRuE)a*RZfpHZmPTn8x1Y?xn~>|1tWZK$VW+sBcmVTx=G_Koti> zZ6Wo6$FssWLC_5uNeK_ztuelGv*&L{Il0`Tr)p=`X!rP+7pa3rbSDixf{niu^zJ{+ zNz#u4f(j{g@_JiTe*pfrTf8JN!ElW_k|r4d88eU>gg*l?Yp6NO&r11)%!x@Qo&!XJ z3oM#Hj3tL4YW!(|TjYUB(rRW&5H*&hW6Fd_PAoEn$?s1qxVc4qBbNO0z`D_7tjMs- zjvRXps~>A=R)Io}xxDv~(dp#UHkhz#gX?OCc%>%!jo?63XSo`*kmV!GiQXJB0^Q=jB)1}Y z<(DBTu$Ctti#>X>-06&KSp!Xhr5`Somp_y;EX<>U7@$ACBIxsfnuab0RMg_BgCbFP z4*-CFYFob^^SI1?Djo4#-Y|>5Kbu*f;z#zeqqLG%bH(*8)+z zyCzhHkRfJnM7I7^)3E+=nU#}^GH0NwbZE`?Tl>jifO2WiOoZm9YApt)+Z*^ME+E8$ zD0bh&7Q=I;4P+4*e--WQ<~~9FZ4YRbQT}rVWrTbNS~f1!bMafXgh_|_7ndQq@(M!v zB8eP4Ndynp=wbf1-C*>W$hBE7acrHZKz?PiYMMvFyZDT@pr`%R4( zdvY@|tAZ>w*CAX}7GsMTcEgfWocC-iB%*vKWCeD_ZAp9RcX|$(Oz@|cf_Kdu3N1~d zed^DyZjLfbg-dlOo?wc270epQu*2a~N?Q}4KN$1h@=a!LE)_nLrehdl`)|x?OUmhF zNcyB}C~JOPmaerBB`{>6eTT5dDh-0D17JVAbjNRPB+XX9o)o~h?+hgRa|J5r_AzAc z)v(xcBDtgZ6!IQM>(C(+HEyLSuB2woMz*hEkW$J`0lgB_64!Dj^B*#XXd@bBH-ur* z9PkQ=IKTQ&H8K$|jJJeOxxM&_q-*UrkMfT4yxty9GglW&A@uR+ibP@njE%zD|1?E# z_t?;HBGX1!jd2h7v6wo}zl%BCz+;1H!$RkglAg;74w#_{?DiyChjXSkrEFDf-P`^4 z(~dNz&mo&hKzPRw+*;HzyVzqXl_BdFQn=R=GK1XNR>GMSAP2!F1zjOdXg|2PFXNiU zfw75VU>b71Dgc+?-vu?mm^ZGN_$s2U=`6oHGE0}5d>EEY_~^SI74L$YNH2Y7rl>gPf< z0?LX4r-F?Pfqg?n=*w4@onyCN_{8A!Y%bSsx>wg|pC+exK?1`S3PiR>O&DGuWW-Uu z*U>i8ev=xg`-H6!Xb@+0JE|v3PA#>6(X#1cB4+{Ot1C!C8$sR9Gmx%ce65#j8!d_9 z385JaWHd~gp4QPD9c(VO6q0g2ePrDpQgOg_Tq66i+oNe3FU*JS$`?z*Q^>)X1$X>aMZ>xy2>S#5H; z`g3F3L8lL899UN1s{-6iPEO&m6F6O7v{_&GUvp3PkBAt^stFc@aTk~}l$t@3)jME>VY6=$O>GL~*^W2YwtSDNJBF?>W`&uX7+h+YCl=9$>H&nO9$sVTZ}N{{ zYkqq<_{`O+-Zy~nLGt@AghBY5&0ujY1y{pkwI_3I3w!uwOjld@wIJX9ek;=L@y*4s ztXd%uhP>$|v9^V3PbJbUWl}aB<0AqWcVXnKWaaOoYh9h11{#@R0@gIw;S!#@Mz6aU zqr{B!@&t`azhLC8N*SrQ%Du0rj?8*VLu@>iLwwqomg-2XrM@%)O9fE#0_qKmD``wcQkV@Il{Vlzm8 zpu0{#+WImJeCmh~W~Rr#s1Nv}?%6kEJR$CRSyJYqMA_GgvaLglpTCCWZ_P#;|MnDb z?;%nXvL@~}zi>kg>d(yy_gZ`7E&f5>FLng{Yda0-PtPzZSv@h+Ve0nqz*!>(Rb*h> zUOnh`D|n!}#{pUtav7{_G*eLJ*zxk`o(hMYOYl1QXE&-KI$9&o+$-;_zz@5n;$|?p zDc98Vv7k%d$yhMXGT0vye`m#V<0zT33~?QGtu(J7VQgd-+ZcXE-j@M}jbnah??hbD z8|&!ldT`P_0C&Y>FjhauIB0CQY}dD1Q(ZlP$@OO}iTY>-Z1n!>bncNMwL5>G+YK>h zIFD~xdb09*lAFa~;$|4GBp{g?HSK#Q`6-ZCM1q+iL!hXdPuVY~5|k?HW7?k!cLb>t z3*l=w=?Y_=nt+Y=Q5BoB4Oq5VDkBU$>Y+My0K@QqXs09N7N*~A2!GCKvErB zir!{e*NEb71omAriZNIyT4ZNzn4m(PocHTAEyWC5{3c)GA(`Xc}%@ggqvjC zGE=j5(J|d=28Md9NEMZ~y1r(03Y_k0&KXKMHa`=E~>U=)r9THeq1>Lf>B(buH?h z)e5?mP8xK!C2g86v?alcO)Vh_<%rY+#cA*c@DsYn-!H_K4}D*=%NSfKZ`ls~CcmUk zltVRobtq-M<8~Zv!=3y++OisIi#&op>S*ZUjeo2Cde=AqjTYj2PU!1Yop(_(_fkkT z27}M`kRwcFQAD9UN}ki9JwY=m?~)!6H!GP^#ZHyK{EzAJz(F8GaJU=z?)e>?+Xw6X1|prx z5LbuG!;unR87WtgOU*CkgxN|zvRHTHwwx6RU_WZ9p_^?mNfJ@dx#>U6MTb-Zy+g|v z=PnzXdWMn>CLDg_bGBYdm5MwF5iMPRSv-dI^|@6Pnhab}$Y!i)C2qRELmcc|Kt=Cg z)>k69#X?!zk0;4M`fl|c$)FpqUliX_v~}@9H0a%Y{^oX&2~juZiO+Qmj`e4+53MEO z*tMWYZDGwvOaX!fT(~NcS+Sn1=DM8y_2Vzn|I2hmZWwHC6boBFCeQRFtm^;TBz{Sl z1*_WD>NZ7?ger;m_IK<2!whUfGkzZAKS8WuYa6y0Ce&{uv|rEfIByJ2b#-AmYmdDr zBzkEjSi{!+gbwM4iaq-<^swFeUWcXiGcj-Rt6rD=P1UEPm^!t>Zt;o=mc91msQst` zWgTr&_p3LRnq;ScG)?EKwp2T~NPXCPLx~;Kw;HCHf-O9=?&1ywt?=qjjS`n3J=A7D zBR-C(g+uim2a@tyQje62?CuzOA+l3~VM*UhsO;PDkFY>ANc#b9n6){1=0I7I?_`32OXmE=Ud{#X;cptb=_lMrInT!KgPwn+}O$ zwPF#@L-Nlz&X)XEObY!VMNaN$q!`4PfrBH^Lgj^5K>LBjs@myO|CYvcz^#nQyFi=8 z8xxxw_EWG_!LAMO{RW#p!OvS>oz_a(SYnh{rO@WY34FyQ(88^}J(?UsyV!O77n8ywrCD<-+nG^1Wvp) zi85JzWEySa-PE_U%a($?BPiAGd!gY&Pio|YojGx*Yn!TD`2^L!+>s3T1LTJ3$L3bpgsTOd(cTifeT*IA))pTd^RtT@y z;bZnO)`{gqUrW~Qk0^7g*`m~jDcYKWMVoTHrM{?^MTHeh7wXurg%agVD$Yl5$La#N z+&CaSkcfgVotwE8QCd1lAoqs(qj00(`jwIpS1Y7}@nf)irQ4(NiBI?IJel70Swv6b zx{nM|B2G7E;|_&U_AkS$_o;9)>#J-|H_zzL0bXX1CBv4f4T*b?a8Itu2yty2IQkqJ zwvNQrmf##GW!Xyr7H`J_+TF+{1k}8Q4=G9e=ZT=BeG*J%{)tN<>}scAC*UHx`x53j zF0SES2m!FX?Jr!WbvD&j-Hh{u=-)*8EBMc>7Z$;5!UZe#7JeR$cCa@hBq9SI8s1uo zat&rmN?v#MmEi4oq*`8ceKw{L7BtYxJdm%jbL|Ld#Bh<4Lu@v}*Tlm-ja{C4`*V< zbeAjPEmlct;vGHCxFvbEFF*E-PIl~6ZOYlXQj*9}H02n|XSZotZ}st_jiIevp@(dj zJVz31r2OB56>leI=FP5L;g+lYb6?sHVOpW=OcMPLlTCi)2TgyQr!J zG$9=9bO)ed&!T?2Qt3feRZDFtzTTIZu&;zxvkTi#i8KL*dHY**d{g)iVBL?qa=UV( zE^q*e{Q&@6=X)(N<~)~i`_3=qe(wkpktAzdXH6KZUKzXDRiFY!d`*{{e5ETuC}HU!fNqr?D)Ii zQ=Nn_uRw>N-23NG7J){6g_r(H3Q0TzbE|t!jfO=fL#aKw&Tc;z_k1~Mw3;V5QMXNR zGThraX_8NsOlCMZ{2HL^SLNKNl6bw$Vw$3>ZKzC?wBQwdV&l@x=3{3*h}jOjA;M!US5M#}OYIJNtD^zEL*f~h&9g(Lt?t03 zTQ*5B+=iQDx?9X<2l{em%_Iyw{cwOQeaX@08Ga|ior`g8`O4ir2m97^nFQOV9;PGf zYN1GP@(i5i^}h9uHNsv;9&cUUkIiu$q2NpO4}$$Z@4~v?QPE6V==DF~{CnZ?EAnq4 z6mDW@9MNP(10lv*lOjf9HpnMYOLg*+r;SwG30!K*?vveFaD&rtHpat?2+?ljNY2;1 zTHVVP$c_z5+tG;?_pN;G{f5Cf3F#FJ3TJwJ>c`&u`|~>2anR^ub<({@I1%ZVJqNbEjpk;4HNq%h;H@+Cq(7H|SWj+pa zt0v6)i6AKgH127oPIaDz_?_NYbOOJy%E)T0+WvN+`lj_4nG)Q8bExnK!0&`E$baXV zbZInAs%lTZt{IP6mS)tDr|FSpqf?!cgHe1se4nN-PKD0MZyUrM!*S!$>dVOM+bR9&E zS8!ir2?+$(_-(fLi0X5ws>J z8ru7MoG3)j+eghL99Hz!PrXZ*vt&b4>1OnT|fqzuk!n`Z_lsC`Z zoKo|Q^b-uqf^`1CXu+$Y8~)JTaq4>YncymE{vV0I9)BO?BJ3Ic-i^QLU0KR#Ov9o6XIa;f5o*e9vv9@GxQ23yzKRccC}K zxU0qcLe-^Q?kB}+2)=M`64L0Y@qeHpH^)$HnH^6$yhEiCda5a)qGa|)DhvVCpZFyP zBvTV_OmTn9PjyZaRY%QEr;tZFsyfXO#dY9=~6I=2Rn=sFfVG z%yQ{NqB3G8B|>>yY7BDM&boeJYep)d9#j41WwIwha}&#QM?F6W2!gQBi}g-IU11IIp*tYD`I`@Ymdg9+`L@E`^^ryce_4`*5K5{Cns|Q;69ZhY; z)$W4TgD~ueuXW# zT%!61%zx%Dsq-?l*In7U^B#+UJ&iGi_i|C858;#{k(x1-uMHb$*8s;q>N$DHj2G)V zxk`Zt1~f|?V@|HZ@lpXTEMQ4|yTtJvVk^h@xCgO?ZVL6o4|^VGAhzRuk3&0lh?|1< z@M~U8LBN15!gC+Li!xDiGXjEv5MB_{4uzJ!PLK0Z8^!f|4WHWROdQ^JXQH;AIo=sA9){6=LJ!U2K8Nf!2-F zlbv&bJQYj9+Vh?j`13iJchCGAA_f9o_!anW#RT@WZ8I$%Pyr=&pVOWLo`B)w(u$OQ zw9*=sJ%&vM9m`H;<2KSXYxr4~V3@RH4dQ15%NXYZ)+GlgXJVF877!#aojR9Bf`qL- z!EWQO2lxhjWcgo9xM-*a09^9^eIWQ>+xh<<*#Ft^zwbz_X_sBQqF3;Tudg!{c-=TI zt&A}z{r|i7<4i5KfBsj;3$EyZzW+v%T-(N!S{;)6*J;|{>Hp^gM2nA<-0-Gn0*-%| zxnydfZy%P4P_G)5s2>k%k>>|FMuT}Xx)_7(JyabR=8~oR-f8wtMN*$7N z!(W+Fn-)?f?XS`+koI?th|B!-1sT=$kV2FXvgsxDPp+lAbQ=eM=loN}-=TnjGh%G* zltcvu9{aG+vO!RRqJwc>ZAnYJm~?!TwyZO+9OE3IVEp&@|K-i;e})^z-s%vMcvKf% zrgfOyDZl*o_ctVn@4v(ULx2c+D>3}dbcuTdiXs5=T0z{OD1u3t5l>eAKYaREf?pB$ zyPI)u|M#;07t&b>*W5%2>l1=PPE2x@- z(y%?h$9n%IyYOp4(YHc`D{xl`H{*ydT{s7vkp?!)^shM?ATg|v;fO#ju`|X0w+GYm zm=vD>f1sWJ&w>f|f8E!##g7sE&hcSVj#XKrSjO>c|qf;9uY*T}f1E%m4qG;ZoT zb+j9G)#bPb7z&QrHoavxHZriCV2d^xvW$5g1Inu_JlvZn4o#g4xKdB%S_bN|F)l+!I6Su*r2cXdzXlgpC}40Qz(WaW>_-!LxA+u7wa zP8Rd=sv0~l#G$@_{~XsBNUwpfugof|nUyWREmEb&*UqZy;`yuFy>?OaEp2AaDeU${zZ}FI*I9&KE^5(b>K(qS zE;X^F-jRmY3OEPQQPL_Aekkwp2Avul&j>gu2!FhvKK3OC-lK2U>{}ejc6Q0XTDs#L z(5kLpnA(kmzo%)j`9WWe|IEUSxq*tyOxdz!P9840mWdH;6P4sQ*~Iu; z=exNXU#$Bl%I(hmQykA_#J<6Wkj;3;Ik7$YWN7Z-abtQ-R&>q)6Ss%0^TQk(G zE9+7R-vkx4oZ;%i{f#s8G?xE(QN?qcjZq8+(rRgW@4(3w;z!Q`1qPD_G}(o!)8nAn;{(1`ma5U2 zFV<4N`v1k>Reg4xQGqyP8`SdNB8@3bk5N>D8mX~wWhsssHphPUCf&;zxA|c>UH$@W zsP|v|Cq%Y3afR7AK^2gzP8%Fk|);rtWHZVTng3{7>s|vzFHbdL_7t~F*gO(Sb^>{f5 zJY$uy=d_w+OOJi1&hl}YK~wUst)OZUo0|hy{ll#VMhoBHq|#cLc5u-7A^RDVz@7C- z64g)QmNHX{R;*97l4@hcGTlmla;)ZI4qPTGIi=t=GgefeKl@1itTNPI(5Q6EYFVLW z%pi_==@v$z`L6rX&8R2jQhcU4UsogG4a*v7;7Lk@uQNPZVV4hu1h}c$rQo$sRa#Qf zdK4>F@g$s`D+Y`Otg)7Eg%LVmpDEqV&y?uyO8wq1>#x1ZW@>0S*7K@yNXgUP&gp^{ z*S5zPa7l@Fe^9al%)|O4awXH-ZbxQpK+N-kyH{ZpO#*U{bcpfG3lnw8?_E9Gz9-~>T1;BbJePfqMMfO(lk*shcNL4{pmK=@Q^P4 zp8kJ^;2hAFXEq-?FYJn9JqPfM5AxXa_&&RL3uWJ8E#1I;67cTKsA$D^c?{B|{D7)k zSw|yp`uW~$5WF7^Pp#9p`O#1X;xGnl7y6`F3~w^$=wrk?@^(fb&DhRJMmSiqsl7TP!nOSM0$8@@tmx)rS{5 zh!%FU+M|rnGs)LcxX~rg2=U-NQX%5!;Y>Zqk)?s2iZ`oLZ?G@);GXp$vs)$uf@diT z5_Fz{GhX5>Sl__+zg331xATuFh~F;EW#(?ibXPREWwi2fD3#?z3yW};PjIeIT0PBT zM_5?OE$}57D0@|u6{2`Jhx5F1<6*b#XG_&m;nsMooY*I?J)y2k=*>JmPd#&Sq%2X_ z&AhqVx+Faw{w+_h_w{BczR@nW>62x)ZnLMmhKXB)B z4b4?C`xqB#7NJi%ibbmWVB1XfsJc86w@5kCS_W_2^t9ZhTD9ipAydY$2n&MS(q$jTf!RLTrO@h#-P}rwAuH^oz>-&Iu5fm}lOrFJ8d-CAP&%$`N zd>n{Y@4?3h-wX#Qh<62C=LfgQ_Vj+w5U*$Yoh!nT=+cGOKsnCIX1CP) z9lg5G$9Rp7_==x6WmmZHLw*c0IhmgWme0~$tlbi}4uBAq*mFR1=`r6w(!3@29MBP8 z(bU#k9!_)<#*Qn@?DBhNj){bk^RegqRTUH%go()~X&HJB!{!|yUUdwty8|h+=d8JT z$8E)gp}3__^k&K?2Zyn=n+V{evT6@@+ zDfaBk!-o>Wg)e;keQJZ>Ar3P@9;R*q*9htz*Ho45%ks*Yw)c4DBQ4Kz2Vr=PE%o92 zIe_K3$cZ5vs#BCgJtFd;s;E&XO}@>Z$}`E-so7Bz|MwC;l0@lM>NMe52AF9I5+n0?eC0STfVu! zq%S_Kt^=d^yncbqj7Ktq>_Jk%6V~pRebLX+F(tUAF`Cn!3?EBnq0;APQL842BbQU9 z1YMNki@s(A#l17p#|($oMX2UI{L)Y-rf!C~_bQc~+jDNx5hczfLJ_vXDIxQC;a8BhPHYDDA>k37BVzWNt4APw6w>~#gbJ^SBy$5_WZCWoa8$Pq&Bs#*F!>waK{M5_62y2eh&jJO7-K2Z;Te?&tXlkZdU!;im8)}8~#QK!J@v^@`h$*0vrg16-;%DTBv@lk#?hYT!x+D_?u z_vB}~RFZphx*67mm7%gWS?croyl->8E`4ys;=I3N(<0*~uRs?m2 zpMv0OW#jKCf?}IQpD)G|^IZL;EeuAKghM_XsBecJd9j&(5!VhAx?LbVG;%MWXE}Cu zLOx=eb`Cyfx%4RlemBc>4Qd%9zJsu!w}Wf|O+zvrDi;1!|G(_#(dgI*<(N2rx-AqdssRU6eBn*a_H6sS_3ZoS?EF%NTN-^C z*=iV--HOpiZ1Saf)|iuENTjBFPLi7zwO(|N4x1<bouKUYBz>V}ovh>bioq zYkDaxU`HrB&#BlM=&y6L_>MBPX#jSoaJ;HMqx99;mr&ojsXWme%R`##k?y-3N7LS9 z*{v>@?qZ7@m9uotlS5-Do;=~PcEv9AN<0~Vyq1N592W6R5z7O~*zfVkEiQKdg z;oyC(@bIFymi2`PyB#E%;KRG|G%F$0^rfWn0`ueVZl}SZR>r)SxP25e8U`vrks|gS3@$s_-Mrk3XHo6l;xoLKWI|X$pUkX?In)86U^E z4)$)#hV$Qy<1SE2wApybI#llBQ(33>EOEcOTFydfct`RekEit&m6uTTr_XGb`v;@F zk3RK*KD@3JzdX8Cb9tsZ{Zo&8$`of>kvW#qn)y5=b) z66CJ!PvN&VPNzK`S=8v3F6HSsqKs8DCMx(eY<$4!&Cj>VWp9%`SZkcUAfH}Wkx*O< zEiduES%~(bgcx|_E7xq9BEPppX}_{BW(j3aZPc<_hqCh?WKP%= z9wH(e8*i-HKYXGR|L9&?1%>CsOIIg|+uEKI{ttqp3XeRSiqX&Md$)KKzT7b39Z)t_ zw%tq;$9N78Y~HfTJO{WD$Qm>_D=bLXl2Yg+Y3;ByZq#IdCMT$)W^62~KO!Xou3!3W zcTLn&d%}uLG#94*FfCV2RpM&jXlvUOxBXs66WtKy9wl?ADjuQ`@}bKnh~;z!+4|3d zP}j}gq2Yg{>YbzG>iR#>#GQt7d*8dx zS+mZX|K@x4{_c<&`Bc(rsFnA#P<|Z1N+o1}_@P&Ctr8DQ z!Yud5XR3dLzkA_3#B01z4D@hUO|YAr5)>$xL>mOR-<#z=J6UJ?a4NCd`ZJS4+baEu zRXcnKxnjSrxlT0CkCx7+N6&^w&-Q;m!4H4$gjT_yTmO728)$(k6%eGZ#D3P!z;Fg& z*f7~x0P8_%lhjAN`G{Xt6sd6%hMT&XWISD*td+;`T6rSJPMk-S3TCox9hnij{THCX zjw%~#;5VmUYIOIE6B)N$*n5uHk5>wg^#moL1GgkkZWsR2*uLPw?y_9H^d%l1#qkQo zn7f5`L;3M(rhyu1ZC9-%MO8KTLFSfcdt48rYWDBnM-KWOoiY|457nLvqA+GGTgap8 z8^NS|bDdefz%~*W$Fh#X_E410mkccp6R742CE&@gPPdY#d3`&@n@pSVN%h$!UblgDcgS4tB_gruEvC}7mYNca`-BYTr?Olq)P=(APf7Pu+x4RWx7L~gE8M}MM zPF^ltU0lA`Xe{P^s|y>$_uq`%pY^I)oNe6-^ht&$9U@7~CVQ9COPCgNl(JHm+WPZn zxhQc76@Skj4Fgho_-w3ERlS7~m!B@ENXH>tJUv6xf6T&n_>C7Xp-Aw!(Gbwm__}k_ zz*cKYxEWdcn5Li)IAR{Vat|1lJw%Z6b}@Lxgv)NgnYoj;QV+g8O4cPBZGX^ExJX(bnD|;6N3&j+PcQh_8GwpqPXyL z7I3Sn=$1iIg^YJCN|>{$(@Iam(EB9*=dlAX@otIaJI7A(-zq<={hD8%_)~g3q+8Z> zYWnye0qXM!%1@PBGSz1>@p)Mkvi;P1wcb$U%%^K5J(-?%IFq~?ys9HrnDpirvX>`I z(unn)wP67k6_0W4d#>^bJ0pcE3yj!5+6_zo%HflWa%m6WB8eFo?h!3aIg z&`^hMnPOv>74t5$vEr69*v)DG>;fKMys+og$JzW$-`hGHji$aIO<3kgxAibzw$oej z3PheW=5rGWSk@I|%K!HVi`RyJZCS9Y!c*gAVweO47(M^1nf@Mk|F+`*oubxu<^G`D z9x8`+BXdi)pXZI(p1(z$?s18-`tV3Y@qMg7#}H}3SwKOOc#XVI+tid;c*sP6aVJ!+6}fdyWn%)dR2HN#DX6h3DgAdniXMSxML zS0e6K?u?X5gLd2nLsC>bUdMlN&(jijx}v-C7JvoF%EXr4nnc>rF`?GuG%jt-k{;I~ za#Rn`L)dS$d!=zhJMnTgZ-Rp&b!F%9MORfO7XcFkCq{%PFQWGfH0}`V_k>!#$9)TH zECTfJ4L4xsgdLrm?bp{bKZptU`|L~P_`~V*WY%d8WOClK z8pcDVHyliMjkkxVM!g`NU*~utc9%H4AKnoXHT=>nt7N_SB29`ya}ysEj4W1JRmJKc z5kP6Ss@Rc*V|PqDjiQN>T_;SPJ5my+3!M3=7}IX$$3jshph2c_Z_k!?-B+?Te?du9(e^h$T9?k&tfa(PmsGryTwzV)1p;W4x{+QGau6@~73>{Uht5`fvrniKb+# zE!pK!g^I@(cg4YV-CkOc+H~uVp{V3gsP1x31C2+$S--i_z6zDKV`GE8HPyCIu{DSF z^z*N_!1ep1M`}~de}|nIu2+}J2Uh|kla|Kk>%XFWfUQZBbyvj|oW>?jEwF|}j- z;H@jLrT_0^1d8>DXZ=5)W{0<5+Aq;Zn=&lz+Ahf=%b3I-e?SEK{|WKDAbj}IDS&wT zTPSH+w5naGp}LmUSWbUg!{e{V*rco| z7_XxeOh1OGQ9<(~*d9SGtn0dRqo>*}Q%R6`6qh}xVQR&&``Bh~Lc^C-Y4H^p6sd9t z^a*8!l2XiBU#aBgy*(UiG6%e2mTFs{`O$D+3<2Vz@Odrlf_zEMIB3iDD6h;u3+a>= z3avleUOT6kPCsTb6?gv#WS}nJVXL22onO(5^tx6SO)u<3-ao5`dP4;EEk5X(#nEwE zcPj6Ztc~WD%o=YJ6PH>U2)Pmp_-0^*aszH}IZzlIj0y@mU+lS*u~OB8|GaTCrr%WSPX5S5f@Ftb>$n-~Gl zc@3n!BDQ7bzL+=Xp|&tchb@;c3z=TNgPF1nbahp5KICwwP|-{MSDL5Q@n-w==@sS4 zKdtrHGfK(-eJ>B#y-=?PrJ`iP4m;@i>$`&t!eadlcH()}b>DF6^UbaDI@^$~7#U-9 zr|Y^sRZvzf5IYc+?Tzo<_z(;`P+)dKV?wc4SV{a;=amb9ktIaz7DIH7hK@d^N+`M% zR3FGGWw)XxD+8{xvtE*fa5QTW?P6n?>RN30al+Y8dn8KeVz(`!>00plm7l zJS)-0ta$0 zapLDS*(R?#8@#Tp0RB43(xk(x%r=eDcR`^>G$b5ENi!YL;db1s!h*D) z?FAe)4{`dJ{W5iv03M7os1)_tBN884yr}swkm@9$a6d@i9uZS440~ePBA?C7kz#Qq z47&xC-6>6@GC$F*t*<;FRZkJ7Jm0}GD)5zKkb@os`n0=_k0hEQ$ZKw z;>l!z8e6{||12eNgTeT_KOSD%^t;}Vl(+QSy77x2RUVLaQ?Xf;0Z&YnKR0#t6>-}* z?DVTwBta?v5w>14dMV)dZ)yV|F-CSYyJP{!@8`2ri0BUBaWjFJl76$%f+yk_Cj4ICK9_5?(Z@) zMpi$QiN<|CoJK&A6G300pMQmye?B?XIgAG1f`!w|9DiJx@Qk_; z;%mbr8R^kwoPDo=p+sApODX}we?IX*DcJ9|J$4EpZ$@{nSFnA-Rki-3Qy(QMic>>D zhBy-wnTDN@=B5@CShkCP%OLf$m{k2P$_L&@Y25)@{GU&u&p`RU^2o{36zMGy5A>t8 zSNqd-B3C>ll!AjQKW?!uI-LVMR%vBo%WOUa#l@aM%j&5)$+B{|nLa_7$P5LJSVXb$$MNY_u!vTuq2`ffxRQhG-L=h zQD2#$jfLXcQ_!LVVDqD2=k~FC24U6eei=cI>90n0FjnoQmsVBM@oLdA72C3eF|!1G zsl<6iQP^w*Thx>t7!8v?#d?EvQcoA00&6M#3>h!xP#B?2K+AOVHng;4avqp6iHanp z+!SJJQKGQ01ZashK#TD60kk+Pg9Q2alueHwzqZ8~y(30$UUTZn6s;CVS%v{ zzeCeto4#z{zsbH_gV_+D=~C&33<)3s4u>vh3Z@`e**;2rQ=1l#R$AAp4YQLO@?Tfr z-9cZH{~fj-o{NW2V_@G^WK*p@Hu}wkZ5MIRbV2mfCfS)jHHN?12h$6fO0bd+(I;u~ z;510(*wB!%m4J}V2n>o~OgL$s94rX7P1nOL9hp&KC7V7ayMiSQ(0kuqCs8mS3I?M& zwmw48FCnV$+4QufD+4%_a!8&im^Gs#nXK~mttF5OQiROg$<>c3S+0F z#?cTJDs5YgLs6F12JTZpG!Es44Ft_-8b0{}SoMoKRTN?SJIrG)n6~;OBUV%c=oSi= z5~DN9W!Q7(hwA$3MEzOM1NLaWCKi(FX&&UaT2}mOCy{_!sG5eT%TQ4mJCBRw(^JA~ z=g#pVVEdBiwQU)O3+rX={(_~A?OH+YRSpZfxdo*_QP=ghqSk1AOqP0Au3k94=F#OJ zq3yB2C7aKTzxAv&s1?Ies$P2krJR3gs=p})$X#o`Q~kSCr>optDF#R|+`KKCLMgsW zMgQ%x!iws&lbG;5S9rfQ1C(~PN@xs+A~f67H$>~dT4SZ;3-2DBQy}0zrc3|1WMzuT z*pcfiw2xKzY?uwSU%xZ$H{L&4$=D-hgmwo+P}00bT_pHf763Cp^Du7)Q35yzjZ3Z4mW80+6dt+u`CQY$Zx_>#Z>JAn1oTFaLG_Exo2yoS4e{U_Da zU+SZD#mNeDr4`I@3&~1)Z-9Oe4m?kC>~rjaufeyd7c_!If<*R_Ebx~{__tYm*8KQv z2fpG-GPkYQenuSQ*6U3S9d(V38X6}o(8|cT@#yDOsa$p@*fS~sYpdMOX7rn-IZ0o| zMWuD|I1veVd#tpEMt%dvRjH4PQ@_KA5k5hE6|O`goSG3tPI0%smPAhNZ@_B|P(LAT zk-YI>606xw6a#_;HG_20!9&u`XcAG#It1RQ{ZZQzTa3X%wI&RPeW@$Yr&Z-Vws`rn z9BY?zi$3`vDc58s^$w&&pr3p4UMof;rIrR!dPTDNrI&R4B6wJfQM6BucBMNVn!0Am zZBZ(@8u|5`2CsK|F>YL?GqeL3#tbueCG2|=PMZybNx8irn~ea>#ht&FfVKkmZ7okd zzlkzV?7KwETqb_#*vwbD@+wd&=62b&Kzli(B($<<+eBXE3CLYmt3TBB1_SzR2&KWDCWq-OrT3_9hfk$p0mW> zzi)Dd&?S&Mbs}1A^w-fnqcFcuku=z#su3%{!%Oc^jUaV7M&_BZQzlL#|1|qnTH53` zRgx%Z{0+E`@5BXw5``E8Q2QndB9cW>loU`4uz(^t@B7$Q%s}ef9@X@>2zr=hIUcI= z{5NYclG;ZUrCUe~*vr1BG4Uu_(P*PioHX`26pq#h`mD^fX5=h(0A9fdn$gz|uPd2| zN!quXhqys4?363wM1uDCXTx7y`dN>f^(1Ctvf$=12LxI4rI!R|Xpr=tHWPjuK$c%o zwkv7acAudX`h4vb={G^2uo74uvvIa=$K1AEZt%-?HUvQOws6cmZ;Z4L_H;Jf<#2xs zt4=VkrG1||8QpK#)D8Sk?NP~Y!fwT!XBmRX!~D~ZU3sk$piKu{q6jDdKx!n>iJZwr zeWmBFAG)a=ERbJV?$_$t7BF@CzQ`9Dq*!=RO&)o?vP^6G2THbU<4ygS==h-IkghbR zw7z^!V{HVQk%hS+$z?@k^zcd&{9@K?xB{|+y(BKCD6-0 zp{_Z$;0}`G=+4@SNV$nJEEPB{7=~Hq6PM+Cv&p+bCSIC?s2d!i_LiO%D?UJ*MgvW| zDH|f^U}o>?&Tz{;jxV3gP*IUu;ikAcHTu21Av@w5I9_y4CBdz;ZkukRhTGVC`s8|^ zqg{=HSxsNOT;AcJWS0~0aBSZ^8dLWc`*a=yie5{^z*Oq=II7+sNu)Aj$xt5%+?FV5 z{c2#Rp*zJFRg{p0b4cMA9(gT6e;O&(fFv==PFWA4?JpSd$cC_Ub2;vTRFQH7SCNvHS+? z=aRP0p>qDFonq@A>*6-ui9t{{5>Ji&O3nH+$~EOh#z#pM@=ZDi|5m}SmmUY)Nh+(R#CGxhj@&vsH>I*=^E&*;VhkFvn4!t zFi%mM1M?Ju@2oV`Ay-Xp=8Y{Qc_qg5$Uhz0XMzp3Yme>jPr?G9c$t^wVlh?gj8tec z-k$vAw6BbKm%ZQu##H*^vto|ABrbl<=p}1G@T7kn1;FBWoY$v-fu2X!6X7M_!_JV5 zw_^Qs_1{qEwaW{Z)zJSGULqeo<6}ZEwu62T%;gxJ+Y`SjcFSA^%#K?+Fnx*S`|fkZ z{S=H}sX54OVhoPPVYXs{mRv$WST;<_3x*Lc$n7rotv3p~f&zWlT^^EVsWtarWEk-k8@3eToehU?_{!+^dHkgTS$cw?cVl)dGZOwI;HR zK;i>PM&2`EeaO##zMXLqybU=?XXc~XZ*jrY`*ycwqVuqH=& z3n|5ypKwG7+*sV86$o3a(FEolg%#YNtCF<{>&}dvrhQnVv zWV6%5W^KT+nL6O-vZmzJJR?F# ze1yRN(f~(4p(tf%C$bFW_q24Yk--9M^h}p zs?ngS+Ip3;aUBm}qK)mHN++Yw5Nh+^PO@RaI9F{Xq@u)2Vq?khJ?rED6p62k6E9CJ z#*$bg>oz~*Uc#t0i+3a#Zt*p{$(OD9LkyCrZoMn}27Tf}jQ5a7!zTEXrQ?VEEtZy3 zH z@q#GE@udZ!5||7aE?3}qkm{0bLXghTPW|LS2z z62T2T9%27dD>Ts$DFmF+Nkn+2e!+JcNTf(fFRzUv-fG!7>m>RVH#HQM_^n=Iq%0g= z%m$M{adWxBrm7ieS2brC;V+ud9GzVoCJT_2?on|hKRTa*J<3?5Bchb(8`Bx2%>p|h zG-Xn4HPo-K^K#0WTeCc0&m|$5JjpM{p7Sv^2&pF_GV^g&@d8NSXm9+~PCAD!t*@?#sHe%Z{7c z*7#Af_2`s4n%0ddoIHw2CF(OR{M>U)6Mwt1{jOf(wINl_xKurv>Z_JPSY-X7K^?A% zC^rR5dGCbmq?9$#s()YFo+$91wi%{$Jg>y9sW1lgk!dI;m1fW%5j2q#^^jD_SaQjn zk|G%h9J@Mma&TC$4!>B~z!8z9EIB{x0Vfo<+P<=Yo6CX*wZ?W`|6ai2PyNJXnQvkT^QM?j`~@kuWDBdO&zQ;|kq;;z!X+S>K7|NPr zH<_=o7@+g0lh@E*T_YC6PwKP3Flts(&<+i;8 zCst_WNq+ID)NM*54h=SoqgsMDrgXh|sTwu^JHxnWK^q!YQM=RXM4XDK>tvdW3NRTR z8YM`hu#`@@PGzWw@3G^29It|9LRw?BP&Y8T_ncYL<>lm0RmaQ2B2&eMSeO^5Fm2C!5FX z+gqD9gJ5XCx_isJF!&MCME*Vu#Psw`mVBP#7aR`pQGq#l(VABuff*t} z9I>t(n{}$O)_ZbzVHu`gFrYKH25k(Prhux?xhjOyF5HnAQpd_NWY_yR27nmrWls}5 zp{F)OrXV>_+A#AHmc!eE58gu=2J}8#^7bkGuo*cvR6{wAwjRW#@~$0h(QIj?$`^64 zB%JgeA%wy}0?+ll;&5n^snW{EH0^MiU{Xx~<1MLLC7{7mRbrBt`JQ+p49T5UD+bzW zUKWT;O~u$YURus?Ib(rO*_8kF6&DH2plaV(xB*i94=J^qW)b^*7wtWXs*8r%etc_I z7#mN98**ilu*ki$%=l&)z$f85suojc+ssSu45FvannyxQa8SZ&%f`$t49oV_8O&Qx zQJ-f7P3#O^*J7O9%TZd3Z?oFUk`odJO5OyS+{$YE%8Ucq|J^;b`JLcgafWJ25}X~T@u6VY~?B=v44a`T62!Ny@rLa zP9zB^G(mBzxX>$eS{g%Cq^PP{eY!|%?6>IS$(3@+w#C5k7T(hgxRdW*Lru_dAZ`Z0 zZd`FuA_3;rYfK z_Zq@W*qV&`9b1Ml$5ycy?5}r+C7(G&@0ZzLK>6@)@^>hX7J`UuRwte2`KCea!g>mQbyv5vo-h}h%in9Z8 zs7L)iszKTV(F&OOl)JzF_BX%PTC#+z9huRFbS{F5vFl6=LPX08?cWO$b}pW;Cpx`r zD;N&*)j3EO(di1cW$OK4CNUyI2v5s=ozR=XW9N_EB%!w8lN+96oJj;=p@=Y3!LLT|t~-K{Mw> z{0}&{a={GJvHGsk@vZK(wT^2P4D+jmt|LwT2<=eQ$0c3Jh#(2f?mhmIlPS(K>8XOo}UGQ9tV=eUdKWD!Rm~$yTnik!<@M zK~w^mlZh!|qTgR$g%P*do+% zRJQj8(iIOoF_~_OHi`P?{?S#LY@$@~_xv4ZQ)u3F9(}!P;sM-@XOpXW_6iruEGr6K zun%$oqW};g{SqisXxyin>blugN5gh$>g#{~P|tTJgjzp>=P+X#3ws%E`bvOu1NCV& zWi&~_3JA@_ft`ggdbeA;QX-DTLIUWc5elGyc+*!8%Nw%z^wea9CZr~eE7@;rfnXX; znt-fHP4g*F5oN-sHrzC2gu{y5qr*8NVjRkq0Nk@sDKw45bwc3sDdxt;Txe^X;Nuo7 zB0rt*q)3e0#%8%njVJ9WLx(klc>}?mgFASJ!J?00*5zwj!-mw0{9ZBXr^hh;!YNGc zASQ#t?&rAGns|H+muoOAUZB#zX+CDTfbz{G<8Qg8gn`C`MARTJvnv>Y^S83Z$WG@` zKPl{kHWo>(Gh4yxJjF`OJxs5vzE5%p#zV81CrZdR1ij0c*txjSQ;L(RndL*6701H+ z0Gwrlf|$`f&gCCRO~T_}mu)P5J`0`^NM8~?Icpseg#}%zPFAO94HK#PxA?(wHE?FX z;%R+M^Q-XbhAy9qJa7N<$`|>Ih9#9B3{SY7m+UOoY?x?UE#Ur(!a9Flu`oG1Z>XR9 zArzOKTwpFEm$35ERtA$cF{|4H3Te-xQBQh73%rrzwDdj?jD;^A(3A<0nz{vM<3GrKKX*;N3bUrvz+Jg)rod-t_iK$27 zLfJU7rmQWmf^GNQQ6dkeOpAWWRY67&Q4adFL9@2s$t^s=4dQ2vAqFh&*&?Xkrbh0g zqZxkR6ZK#Wye8>oN?OyU5WUo|9_0HJ4FL8HgeR4b=ueqnd54sbNHQu~Z+Gcd!8%kB zP}qWgxy`&c0z3hfK}D3Z5h^h|hYS0P2r!qRS=lNj7iR7&$$ImieET$cm8V{rlI~uC z80xn=mi_tuqUz+}hJOpLXnd_E$3(G@DTLZVGMeAPei?vrUr zBbw&6o(k~!*o60i7X-9u*6)tMxeDr^^12NM9O$Z)8GY%N6}A?(CP~SN(fv;Be&z+D zjhn0tA@xHh2Ma{Mk~Tf2*7_6bynY|Z4V5lQ*TqBkB!KtqWmYGwcjNIv;u=!BOJc4E zCUi$&v1uk^+Sk-hqi`GEJ$1ji&M~Q0BE$o7`m8`bFp0+b4y}M?5e!`iLGrJS9y-zX(b?WxduxQt!v^-2YmFeHW~Y{hum7BB4nx!?PUNA4;Wb;t@pGyf}7B zw}_*w90Vxsi&jR$@*V0a(!hDASQ|0MF;_D#1k&vS&|)L>YEh2QfxctE!&c#<)C>n| z#)K;)G4B_StHwt2Ix1wPOTuYkuRX3St{tsh-3iGB1*;!HeI|O6>%yQvf5&+fsa3=^ zE9X3P+}pC#5#?mK+e@@8=TvI@sRnBjAL5}JpuWahjpl9ihstg>(QD- zhjl8exm4&Vc%l4(F5{1Y0HD9toR#jNSGcFO_;^e(P=vvOMnF*eie8?SzB8se>Z+TtAuH3ZV~JD zQ#7pphWgF0$nDT%232uO>M%uTFMz)#bd0i)v3AZR}Q zSWKtt$K%QByHw_e>TJlsF|ecGKQb>>7EdHve?Pemix1!?(`j!&m{!fAv?`E6X^GIf zNHrrRP}q=B?3j4_vdn%!EXc2}v_a{Bp(xv$kC*A`NO`c&TV~NpBD#h}5S2|>rDJ!jHi!uZt!+Um)l z=%3-bZflQMK&2+wDpZRnsBrYZss)w>Cslbl=7-wZEYLT$=CQ^;u>I`{06HD1J6=7C zx=ErG&-COo3jhCBRQNYOjbV$__dtt34#nl24M0$^j9rmJK_q&qD|dTXxT<5wVWo9~ znMIoGmjZHHAzNbjuEmxTR(A!&ynRx>WdR#w_&RxM>sfx6YaUFg?L|9~T_M;68mq8XBT69Jvbo*Vc7PnFUsIc1UP#BUB}WQ^$yQE@|Rou1u>cP1C-Qt=slx`Womber#BLd&Mxp(;Jz z#W76PMnL+3dnRI7?YiK6AQ!hN+w&U*8w|=j6*laqI3tLjJI@RN&_{c>@U#mzc8lJf zVI@iVvtZKn@Sn8SGygaVf*t%dPcn=mfMP6~^Ck-)M>M%s|ZFp}%s8P42($&YVxgCVKAu)hs3xs#~iTA2c+^-|-zBY_ny z?B}#am)>5;v3!HFUsyA_)*+FoOMF=72e$po&BVq7 z7OS}=%_)ttrmc;BE+c=E94hG{J$RNkQsi?udypSmG}uZSv4Q2de%kYHC$5}q`*s)n zZ`-}|do)65o11{V8H+-7x}8Ob{p0nSL+Y&wMFp4}JLat)4OO-FWq5t8lIF5~BigT@ zi~nzfjM$a$on^-o*c#J6{H1?@X~sb)Pfa883zDo~G9N&pi-LZX_9a1cOyfHzy&Vh= zJvG>8=*$09>WY*xVW!te1Q&E3I^s<2VW1kOJ}B-)#Nsb3E8z~b6e|rG)lFZCB_&!sOwm9Xx9}@}7~)o)3GFH@9$e9kRdG+56i#P>YB{+k-$`g~Mdb8zn2Uz=#0=$}ASV&WY2 ztEPXnIWX#S$YEyz5TDtM|6ON0(x^yVbK*Ev0d4mZa!5s@q>i!b$K3ynGy;qSk~3EZ zMBu!{YKc2rjSmgK01Q1Xlr%~j1+~e99+n&`C*6o0>*rzd^O;WU%bMi5gKqrk5fnW+ zOH2g{rJGPukqJ+Fy1LyaymEltB8z20rWHM3mL-Q4F28es!ejZMXz*!`9Gd+XPBaN# z^~1WOh&#(dG%-X9iWa`DM?|^Et{2P}oST6z5B@p}8B2`BWcuAhhr)<8K)rhtkr`*J zx1&Q1=ZTNIo%53T=Lij%N*dC|CMkcw>$RZ%c#H?8Z z=mVMw3^$y)Pz|7&=n@M(X1$p->UqGxuD?0qXvz{HkXT$==8B6tTF+%Sk_UBEkHXaw zTqI7ygbjpe*Bx3D=@6xPMPSMC@oFw3y^9Dk|Ndo|H-tW&Pa*iSeVWYZPX;`Aa#7{Hk z%)1h1R4t>M`24#{@LW8>q6BhII3vm{Vu}Q0IZ?aUStd^M2Wy zC{`{N%B++lA4)_Yb?j=J`kI(4Ov6zT)|Uq^Qc|WdsRETGtn|1P>mn)$khzfVg z&)Dzv5$NkgU|1d&+pZzi`K0{%*738sxeWeQUO7jDz_Mr42Z4`$56&~%<=s02qNW_^ zYF9V_Eu9}?GAJ*xWuC7`1)hZmdz=6{K*$T7iBugOk6U?V_u?CmcyXk0e#C|hWOAFQ zx1iPHlh3>fI8j!-+c_s$V5Z6XT&I3SVUzVzSGPYeQ(m500R+1rw+>Eny*g*iMVUd;q3@tX_lt&}^2PiIO{3*6#ls%QArOG$Q5yoC4rtP65 znRy8m|M@f@L`C}d;v)5fFEj3sadzA-@B{YiA3k7(>MxD)uOTi&x#*jbiEnQbzxgCk zxEVQ$6@~pYEmbAU@#e;Ll1@#`2>EHxq>Kkv%!7ux&A%EEP*oYtz~tw%%TDxv;UfVl z{VFYmG>yR)ujQ_coCnl|DVNaif`}|LzMk!&2jE?z(CYRTi80>oNRCM-05~Hw zZUSM3`$1Xh{sl`_<;@k1n|5jJiQ#{KxFWI=8^ydqsDz&i^vn)c)~x^500z$3jh3onmY9B+2_qWWtmD z$+C0UA(6fVA4JktfOX`;`oSPH`$t*>R$<0Sp=WxTui6HIF#vOXT;bCHZ0`hn9FO|SnUN)G#e zlgV%s#Ze+R=IGm2$^s0X4kW(__UX4BqZwo?^Afp8Tn?Yu$hcrKGK=!9UGNc9*|&RG z8~juG9A6#rhp{3t%c!x#0}Fq2ZP^vqmn*%kEuTAJ*?D#54ta8l3SIVi;@{YU%B2$P z3)O~ge%3wKiT?4Xj#+r6QE(y-xUf276VnCrss9|#w)qgTpW3PTB$b0-Nm`xsa(rEq zkmF13r!W$#|DqC6@r50I0Fa4Fx+J;|^_*l8EyQ@oWA>{Hr8crWCcV} z9!`t}mR3GXX`tW3`-Y(`CK~a)D3wH*?&?}BEEIy zzDY@_g2l`%*Ukus14S7exm*>{a(%Wk)~=_8Qf%&Do!v6zvFdjQ`FN@WjcY$D5);fAV_+;`p1qF`$bVUXtM zFlXa8?b%{5D^U3;+%`XAU)7C)cJntOuvsaWCuXw@DagH%RiwjjQ8y%&a?JD#y4wF; zd9;DHEs;RFBCCGxg()G|m4d{*Ul_NE3&oL@AiMzco+MU5xhGtBh_+kEdPD_%)WnC` z+L|S(GWc^@ln?jRTn6kQqD8~u4E7t5Vv<54sYP~7=qIcHS*IY=>24wgHe^s*IqXGE zQhYx~dW9EdxYk=IdMjNW4$1x~)obeY*DJ;l;sG44%=cHR4WEyi;eHg0hc!i(t&q^- zHf05h8coI74}=193WLDXN9*1Bg9#<8n8@7jpH}eUb_M@g1ACl7b3$KDe|lRPuIgS1 z&MZD9CRiv?dl;=z@TW8LMb(oJyY=!*%MyWQ;JOd&J#`anEa_Xd5lh9mqIc$Ufw2Fog#Eq zr~jf`e9+t*tR5{=INIb|V>R^qjPr4^;??nrQk(O#wXVa%S3Yg%-gJvg4vtCnv1WVa ze*mb-Vu$@Yd-6+bLaif@48*sr%Yvw=p0Xl&?J{+eo{&*@0~iV;NRAbj&%|scp@l60 zsVS22af8?lRauUZcF@S(biY<+!TX^Z$jkM`z0P`)j><44&hk8nH!9^8SPmd3e1mHy zX$Sq<8%L8;zyVZ3)??Ax*+1~CN-U8rJQr}Dx1*kK0#up386xcWBV=p{&DNdVE?|dJ zto#Sis*p{~ggAHAHzHf_ex&q*7JhXUIf7|lh7Kth_Rp_ z?MAO=OlM)RN=e3(o{FQvL_WF>Zt5W)V9D(3f6qY-R-z79P55^BfEx<@eE_vmFj7(8 zEszJbU;axNiHqfDuA>x+HPbrEl9Vk?OYv&ajv3ux+!)2H>%6LIzb%^{eDLXVfo{eO z8tdbY!P+(SAyvi3)KANthyl|@y69)-m;71x9e2EG4`+fT*?(8>kKi1*9#xNS!SKzHN!nE5^b| z^!vdC_}`k7^x;EE8sruI$NG0r`ggYm1vtQN-3<}as<Qpp+uTTZ&5y6aoY((Bkgy?jGC=#e!4Z-6c3F zP>L3UYj7vH1c%=AJ>NOscYoZUcZ~bvj=@UyX73Eh+|$-v&zcV=bO))pcu<_m2x+Xu z0=-lrP&KCYgm48}ve9GKaelEo^i}$P2sjYg3?>L}UpnH)tDk~q3BLUR*_Aac`_x3+ zAYiLZ(SR+6uTb6*Pt`Oi#qp$JwR-Dc@S3xEABFc#-pB_6vo^r{b`muog)4B6^fM9m zVSkO2;HquuQJQMWz>ojT6Yufc(uEGw7)$ZNHrEhc;~n6?;bec&p0_06Bn^ojt~HSs zSARBWdW$}L9bU>tMR~a2licC!?cdv_{j3s@5Ei%W0j)*lL_xd>lux6ALRn+I$Ga#> zXTaD~scegjPR> zP>nTOaoz=S<_Ch@F}sBtb&jul_0sPX9xDE~ow;|4tZvp&QDKgBX{(nqQee}u8U#i% zVPgkT)k?p=n?k%I7Un|#5vPl9v!8))yH-hEBq4wd2`pL}piDmRnVeCGmGb=Me3F5; zv&u`TPh6Yy3OLXb?AW=Bg01qr5|sr99^PY$x1(S)O{c^Bl_@6((ut>Ut<0YOnd=}5 z2iL-c>*A)_NW>^)%?J=V?F+@%pjH9kJBj@2kMRp%xfvg%xv5Jg6|fBJ>o9`redi8f zMOYMc$}dr~!%gcol-Z%Oq9^%R0{V+RcE$er)wJ;o<|=r)QQA902lh{az<9RL2bT%Y zC4#64eYbs|{keWABWcCf5^y6l7d3x}qsLO3JovyW+mdt$w0Ep^31NN}A=3Ik!T!XV zwhS+q1;+BUW`B&hv_{IH@2Tj)MyYoZ5)8Fl-Ko)|#=|WrTx=8KWFE`(7IQuc@v_X=}}AWKF7&yv-_52t5!!Aw`8V zy}HclknT)~u8q?dqBfTHJaR}fw8hH|yT#|8nP!(3Stx}{M7f_FjqmB;dz9%}`0Pug zSJ`fbRL_r42SHLO4t_Gl&i7I}QwZt5j;AVazjmh9CSVGB{~d@Dk~JN*_EQe@@7}7M zLlSM-Oo$KYOOx|CvEayes=r5Gy^HvgZr2fPh%FF)gdKPGFUvj#yKi z6suD=hFwfI!uk)D$!YiSZYBol=j)3Hc*q1-=JQ|m#$=_;*n>0g--SVIzjMN=pS>p^ z2%l|b+Bg8b`n<&z*UlFc{t0Q11#q*+0!o)ixi(@)&`gzNQ`e6P3uF%|?ZMxOkGC=Q zUyS$n2>sCHeRJ>_p9?do=KI1x{!UDG;}$(`LOjKpoY}iA1$`-RqKR}Nl=shVESD4# zzs+dVLsc@bh90w>HL6d<(k=S?{KA|uTT{*$seU zD)M}EcmC}?P=_5Ajt~(LRGae%G9+_N*C(qwc;uF~06LG~UH!Kcd)(J8of<%ncp^KT z_eJ*^hrP?a`wg455TJc(kF4&^y%Yh5AXIwE=B+a*J`OAOrWOcxN-iXdu9Q;HXnf0m zyZLPT_h&ha_{GBO;HY1Co9`*-_NWY4p6uFAkXDp`HdxPk;+H=``=OEi(>ISIEu%Xb zqwdS-(^!2SL0s7aH+TKSp6*spH+$Ef335P?ho4!kej0Brc&O|EFR{RU!aEPh+i%oAstWJ=;`Jd05xg@3L z2B@RP1~BF;sxj}FOV+btdvswgjcnM{r>Tk~7bw5|^l&v-sG`CQxPS)~)zAN`B=q~L zUvqMZN<@>1glqg{#gb>+`81iZhq%tH#brLNs~>~XStIBP7zX!Jzj(j0w)6MP@kC;AuEg2#03Mi-r1U zlM=EiMtoYJ8B1r5j~`kP(%tIh25m2x6!;;TSFKMoDb4ZhQZF$3Pj2X?p$x^r_P5++ zuB_Y{|HioV-H7MLNd&9mgs6pC)JE}7as2G>S=@{~l%(8;F3=~{3#r?(XfKmCiQ7tP ztH*3qh!^BlB#{g_;Yjq~LiQ$gh%EGwO6w(RhsNp`Z{J^J+G1{`tXCC3i%r z5h{m`OLZQ@jt~(+{f+OzIrdLHYT>vO(@m5Mlf0CN_<6NICt0JZ&rEx2Z0NzdhM!2N z`;06bBF_l9iQ~lW-{${C!~ItwYC@4o9sj3AA?Wp?p5c{`%v?ZoF$wAU`oO`1M8`@y z#ABwvq2_np3Zx~l3~|qL+9yAqZbC0Vc(Pz8 zsWOjVJ0FtKi1on^Ut*PcY498M?yod>DGSN8sM{U!w;i587%9`I60UM_k@7g=?pGt2 zvn5vVW{u`p56cLY9k5E9(Wi8S5bRw@l2z|0C!wz`)cK45uUyDTHsxyg*GRV)QtI#K zBl`wUqHb|Ok@d`4?Ut{IRkPqfJD>leon`#*B<4}|e**Sjw4wisLd*h(?>TCG1)wL0 ztDcG#{Qn%*O~I!n5LX5Neg1!3`u`9AOGN*lkps@5dSivY!L_^3+VpX_=1;2sKK$Wm z>2pU;UeH?oVYTxb=3S&VF4CQLEgiuwBl#DNT)fovYD>n3f|Doi$N<&_VZ9(lHQ>{? zWtf=BZoQdySmrJ;#<9=>&8;b&lP$mXFl3<+-!fbvX|C3Lk%N1Y zZq@G*jO~}R!q+73Ef@_TzOq0VdWCkqTTBBw&^!- z1}ZAj(gS;X#m)6pwx$f1zN=S?2BKz(ox1||nRff-Q;32Rp!W>80bVsen4j_QCF?KR zSLdnvuq^SB%$ZlM0=&kHGSiAxb5s7x84(=Lg?SKkl;1w={%qw(TJ~E8x9W@2xnm z5r3fH@Ojm4p?WvGKp}+;fd{;PtG{S>_CJH=q&223rS(e|WF)96_0kA85+?d@9$?@e z!FLP2^J9~SQ2i+9^dGs!k?zZe0PdRI`bHA>^gFv4`^jnk-Pc8Ts^yigfF@p&wqnXZ zJ^n>_34|!(W&VK4r5?Bq2^oCv2PnvTnEtD#c}~!!^tB$rU$mO>dC4J%rZ$27sK@cY zXqQWczbtsp3XD|AzZSdjx~j@&6jqnGh);{`TqcKGic&YiaS(S>mjXK3bt#sD)8#Sc zQg(BXr=$e8YTt*4%{$a_70Q9KCin{Jjn~QJwi~}fs9Qs4`@X21#Eh}z#5(eK>F z+>~rRuMcHQkiY~SL0k#a9@RDwft8QyvZAUwcHdHpH+2Nl7mt(^>&$DA(}5Fa9!riL zrL>96>V|RfpUduA-9#BLw`r^s%CWQDkI|}xf1Ey@+zWWgRBoI)Pf>OIoRxXaeUSQz zK|;a35|m!@I!|03L$sbTq)R>%pq_~rEE%OUT_Ar{MegV>%-9bRNmSlUZ#$B2ab%$k z`I^}0(l`q~9yiZHeiN1hZweYz*)nZ;FV?S{1GL*mcqtY&5Zaj0pSR&75$a*{u?r5t z)y*z#O9N;;eow3fezhzOZGdRern@6F{d8p@6*7!BWF)J^B%doQen%A)iN@D^$dT^w zwwDZQ29##tI&}Bqvi?OYH%sml@Kv~=^*tU{uxAA<-JJ^dG%8IvAJS>|gDiWMc>DO_ z2RN?Ynl#&6=#~rXVbWP1ybJ0yw+_~xlNnf%VX=&xdG@K-k9Wxi2DJK77)kXTCK#ij zsZ3J1VwRnK<5-?n%zWT=;IS*p*zW)qtR1mA01{O3(gdM@mUx+d2 zJ935U*~#rQWyp8R_4keUnL{gp1Z&ro2O{C?+1HDFRSmlgVR=VQ{0`t;jaW?-pUFDO zf4#jXlMS8#9Fe?86JIt47#J!*~uVh zS2DBS6rPP5ue@T>OI9RYZ#oaR{);x{k#yT)+y}Qq^#*Oq`-`Sj&Q-AprwH-sya6`K zYr+EjA{&_MF_+`!Pg(N4%JJsetpw6}mu8Pb2RI3rFnW$+Qjo{YF=Q>k+nh5SzzeTI zjCefpto|Y&sR5r>O9b`q!@k&)Z*&5S0X~y2~JLOHQe{iP2;3B}dm(v(_4ixbKIUFI{rX~GEd?Kx0uL09sw zOsrqd@(Ry?ziW!qlVNNXFTt?Zy8Mj8=W_E(OM7E_{PUK}&ICOS0IDG54XeXnGzvA& zyQ-zzq%nj=dW3WGO8aY03%Z75Dyt|oE2Y!7K6fPL{tfrRqxJA#exeuSz3*Y|WRTJ` z)v9`sJ}`{q2uFaAS*o##W~2{fiZvo_rXk?}0l9uO(1+eAuQ%fy_!kY5it4$T7RjV^0Xd@K0*%87z` zu5ith;qJR$;jTrb0UmkA-HiH=zi5$_-36+?x2oh80{R+w|)ac z$s<1fu(|u`SJq+b^+LYf;XPYyqjV@6Re(doLE!_ge>AQWj$j%;>DD^Hr}jbZ4e!ej zVLvXiSp|o_TVgs;7>>bMee|O1L=TQpq%0pYCijOawGN_goxS#5+=$+1}OYNT116q74`*ly&Z=%^a}8=LzqjBKv=^x^a}+=6jSfcdCTMdOIzzK zQ_r)U6(faq*+UCJp+HA4;o>6)WSze?9FZS0?J&Xe(P~QAoZvn>1x0a`@hCYcf3KDf z^pKcM0JRXVN58h~5Pw?Xz9*j^Mm1D|TeT>h83NS;Rp4JdS-Bt*{RD1INCD?UbYict zTD;gv3Ld(TsAaItqZe*)+o8JZ9nYu`d&|{XyHEKagUQQ1=g6_2j4Z*_2W|Qx>yqu6$Rl~@ zIEylj7|KWK39iJst`AZH9;mL7Qh4&RRURm`jb{qzqvb986lEz$oE-EV8aSP>9t(K6Ei$J zyyF1>&R`f;+c~*`-wyM8{H-jXTpJ0=d5ZxFi3Szd%mWz@rJn~&M6TeKD@n(M6zqAk zc^TRMw2q~aEX=8y7k-%mo4$$E1JG~i_Q0EVMFB!d*~4-wufJ$eQH#y;K0Au`5#hX1 z0U;XRhG7vk3VelVXg}5cW~q^%c<)z5Q;_>b*_ONPg%tskYKh8fyELyyNo^eXrf5MJ z-9^yUgrO^)4L7m)v0#;$uSugM1ID=br&9Bl_%W2qaZFm}ojbe`iBTNuy!1z&dZRr_ z{i&*vHI?ZuI{S|H&t#|LFWMI`$J#8g!XfU-yZ!#GA}t>zfT!9I-D(r*KF@vR+haP# z_>6PQa!9!GZi8RDjb|{VOt!XMBjkd{afc)72)u6ZI=|Yf6SVi2daQ-0o+ll)3swdC zRQ4{@aHR>d_knTymlJ95WQr??10s_;b8WN-(bk`kot|qJU`2kQPP%Gc%sY@!T1t}5 zEF9u&oG-W7p&37CzKBh-bcxP1bTe9B+DnxFI5JkhFSK!n3sC&*nm{o@JvrB)I%6tA z+}?U8!_6yVECW{z+?V@SduP35z0zv%gUOhNMCg8}B2W7LC&Dm@`x97hvCm4O+Mz4H z@ojm;3hRZ9pYIL`ksm3bb=0h+jIPPx2dA?f9psxM`Z@+?_>IxR(qQ`A^k}uL*4bk< zARTB}=2(F*U!G*pZnW%lohK_zSg-vdYvf$Hht1?dsp;*ph+%Ak&)%kcg3x5UQ=7@XoPqWpX?H68&D#BWp`${kT$kPQ>Fa zGi8*W#2&U(A(m1?0R~t*(GGv7os`RWkjx z{8ojbXY3(WlvGMdlnk3=hrcKnnN^?r4%Lp%<#a%DHsheMl$Exn(+)G_Or3H|quS!; zSe;mm;$5`Pd8z+RG7o*fWicSzSKw)Bsc-D~-eoK7cv=*ldEPmqU*6x&4$n^r7mdiX z?~LL1cyOU3Src}d_C}LA@gPt{=8**%bUc8G}3j`bA#e(W6V?*vrGRm;}+Ag#hz{^ zNTPZI!V=kGYiRUOK>JK;Jp{FR<{RS`ki4(-lH_$WkIEHDiL$E8PgNb>Jor-!+;3s( z6#bx&ie>!(v<}~=;2$xWsmq|AGFc71gOQIN$4_R;_MS=doqOHgH#3)KUTrfwgZQ^; zphM}&ko6@80T@z9e9J#pz9#eNLG2}1)(^6h_(>7TChx%M-Fu_!sSC^AzaWr^G22IeKrNIzJ1goQ{+mHo#lWJovcL4!Wq*oxtM90AK zihx}M;eoD`|IIjIh#j~iz;{>fP?`6)-OO3^4|k&)QX7T3{EKgtMOgr*w15&-pZgyL zTV)%X(vDkXfMHadHNS6k9|(t+#4D@tZ7OY+!rOQ}vfcJp3*VIX0fUDX)ni24R@{7t zT-OTB$HT>1c{G`O9gh5`w<0ewdzi%g4^z{6#}7 z`jau~E9iSfrVfDGX`Te92;4PkD9q`$|3!=Px8+YL8zrtzTjIk%v{e|ww~q_L72x|R zFnQtRD06mxx5Jz`gv3oIzKgEHzQE>i=?nNGfd{#Nqq{`x}u^YKy-gbw92-*hK|o6Vu%-o)e>lurc78-$EoX37IU^f?X`R=@g1X>S%wtCF zFY}RB&C#B`u^6;Q{Ascx{mm zXU^rK&Y_Ls3Zxz%34(?o6)t=&=FQwnT0og<ae5Z6tHZ=8I$+uTf z44s!%^gRxUQ<*mB&u~8;X>eKxuU4z{Lu5x??9MWo=Rp-C!mRQQgXfy-mJG`m8=u-G zF%IAp4K2pOP=_M7EF0fEc{vAN+vXQC&?J^i32W%l8IE9%y!>I&2GA#d2-AR_;=={P z(R<8Lm?yS5*GNR|rfheCq9WIe)uz1BhxaeqkKMeY$Q+Vr(0PgaGJuveHr(5u{oFry zSs2H9qr`p>FX}pSWnlD7ScZ%)sF{w0d7sV0;)&NIdX=c8`}|(a&b>h5^4x;Kxsir+ zlKZL2DQGu+tlPaU`KyHEe1FtXNy%`O0*AmT=6rKTP~UHikf6sQ%!aJF8z)s<^XGl1 z%{Lm2B#Px3B(?L45bTYuZS;y)A*F9Zo!i7G!@D_KKS!B-|FU$AGSdV;_NMCc)5S$! z1swjes&;94-$y&&r*7KnhnL4M8|r?dzcw9oGG(hyx-bI5nA5&QGckvF+jaO1Wg%TR zv;pRww4I@`N*pK88>U3Cb=};&3dm+zA8mmYycEk(+ds6swYn=Sw=$T1P8+-U6ntwg zOqi+R+!aP`b-Iz_hb042igHsm^ zNiNO_c<_j{4i1XXuWp#%o5}w?H`3gx-n@=K-VxGG`Y5RYkN+kl&J4FjRYK%S5{iE5 zROyYJ@2zI%uZ{FajUG*u?fvG<)N{{9?*~X?_fuoj!3N07-@^v~(s2rWYX7}=Y-yPa1i7`{A@DK~_{jT;Pa8c3bwQ!B1=dst#Gu!&ik-F>= zD5X*s<>%#WT}26DRuNo&rn}RxNY@wgAuVi+@^E*ktl-R<{Y{C0Hg9BIh_>dz{PPLj z%4$wcP+5J}Cx%r(R%=Ug0F%015RtdD&_a?I2byaloUXb3k4nPNY@!TDvLmvZjdGN5 zeL8}}g0bcA*gei8RA#@vX<6oO#&9Vwv0HRltAUEk5({^!sMm#Uc)}J*Y^-lgjTVK? zC5hgVYO$$m)B`LwNV8geuii?tE0R~P9mVLkPwJ3yG%D)M0C`?^JS_VpxjDqoFEn@_ zZBr6WFL_uHq7qO1MhsNMw*cgH{t31T@!`hjxOOjP{*CS>Fp3`2p%mE0Mu|oJ;e~%6 zmNe`DYc$*=zU&x@<kJHQ0iNDaUvv4^4oe&3W?!R-;Zbq;H`u?eGQ z<^Bo)sQ}Je5W%P65}gBPx)(IbokBJmvdmCTKur096!PiuGnOL0&v1cYHFB$#z!}l*1r{yN6Zp~ytA9qe>|r$UK(A547J5ph#M^*2=dfO{PeY@ z=h!8~FlR^SPbJBEf2|_vDBZ=5qkUSDupAV4uH1{q>KXWnPLJAQlK1B1Sch%$!T8#N z+J?zAP(NBLN2AMu8>Rbc+5Xxycs(PML~s8+h3Q^Q{NG39GY-WyNdQ#XK33|c0y^}u zsVY(d9i{9vre+YAD2jEu&D;`jRD&YGf+=q;4HxF}RsLt_gwXgxJP(tqJx?A`;1(uWCP_iWsUx-LiC$~*n zqbh~X_Aze-Gu@1A2T2meKp$45M>6kxPpiec1N<(W6J-PiKizw|Q zn#n${x>k1C%G+nsQFLnv<|TAl4ISD_THDalob&8rPw z0zxIN4`QQ{W+!lS9-floS=z+Y_Z}#kcj5Nd&jc zLR-@E$wf51vQBO-1Wz_0dRZjgYNRwn!sga#0@M=OS4f(;x)&u$g(Rv`A!kv70qSh3 zQugIPIu_6b{-UXeBHN#ALs-fl8+bxoola+RbBl_5qx%Cy zyc3xwYnQA$Krb|0+6s*GIVU5I#2MKlms9G@kpNS@C2WOai!90qGUt-HqO*8>L%AY z8FO-it<%sQg|G|D0VUl(kJFzHW1(kRpzGM$tt|+$S^I?OlHw4k}~L z>~K|G^iAtpEe&XhcQ3h9vI_cFt2cL%t*{fh(r|Si9;ips%oYfrj@z`9M_Sf#2GQ^3 z(m>yPEYkYZIyt&&NJbfM8`=7U!J#7N!03i?>bAb%cbCH7JF#E?)j61Y zcYZ5u4RJ_aso=eECE%$D)~TRM4>bVwREf*G&(F*6_zJYhwx*}UJKn~+V-}t7re5O?5rUi;&103R-)2fonJ~_p;u$p(#SJ<^gKB*gV|cr z*=1yFFuVo$W-}*ozPh>!n$mLZRl)T|@HnxxU|X}WY~vQn>SGVuTZO(&E+&z100+4X zzoE7-O#&@cPjKr-b&$m4^7Dl|uu7LFVj52ohKWa4AG2)Y&Fp)pcrY9^K&@Y8QiQ9p z`_#F#jaOC-lLy0qX{5(2r`tIF#7?ed70hL=nt@Q}tXqgeK;SXUwn~om$UpxUH~6sy zuC(!XPlR0xn_2f;4AsCEU#49kX?D>0R8>um5+V3}Gfdv}wA4Akpn5fp(kY6J4M*mh z%pt^jkyCqNu$5PjN?7(YL0{h1BYGyZOruiU(<$n#6~Icj|h zw$+sS(NUocl{)k)hJb;s|AJidzaV$W!eknR3TF7v-`mM|pc1{}iuv`1-#1KoX6$X< zzob3zK4T)*7*lE4*23smkIEyyfY%Y^P?>CI!Xp~9V$;GTp6WJGCShK_1rPD(oT2BZ z-IiQmbW*R7al#~d$z;RKWvK#(&ZiRXtix>h>{OJ>XNtj+8J7PZi3;kGv;~iLjeKt( zB6i3B#c8ReVE!d_z-MSgVyWbcbqt_HZ%tzN>6ux&f_uiYdNd{CTp~9b{B_Mhuv=># zMQ8Q?L^aCH33pnBp5%3UyQaS(x>+I42>>>Y*E}4Rk zSHPJ$O_1|tDzEhPx*h`!HOmy> z6S!W9OMbom2hn&Ob=zH?Ar@Dj>x1CLoN$%h(4l(c(>vlz~5jx<`eyoOp{HzrNp7jzk<#Qa6mj2T6cfs+3t z;L3g89E{zYqTAm`LVD^Xa)qG_bEkX#>qX@yC2&vN3DJMI!wzaaYwkmr_p!Xr3}Ogw ze}e<;Lt2>1@Yo)Lndiv4kEaA>e1^dspizqldMgOr^}sr{IsTlc#Ep zcE0VsdK6;>cKs9T&wW9e3+Kci$*>b@92OQUe%@y{1>c>QOfEtle$nbr5#P>x=)9SY z9Lr(+z%2aatMoJAi!VV{0!=UNFek?&VN-*H^{`I*{6R+6w+G#l@~?IAR#U36c^4tA zMVs#qQMU0l+rhw~Z|%0!nl>cS5q;B4?0h6@;>|E~s^KbdCRSbK1O0gGPaAS6<{RFV zK|><>fos&f#-X|-34n>3?ug3b{maorUQ82qH>o!g3$(P+Fj3y{6K1#JGP>f zZ4}rr6-LKhxve_%9-v##8PcV+A*5T!YVw zLsRm+SPLkzeUKVte5C^F}U?=fMDabrVZt%4|Z;F8s-@UTC5Yr6B zDp;*}nyMN>QpHlOAcHGXI@iE+WP(Ho+ZNMqn2zeopoF3U&X!sRYW67>#vPr~Ud0n* zRG0r-iY)0qPhrx%_;k0|60x;e(EeZ`YNaBJjF3i6S%}$Qy4W`a9ow1AW)Kc>qWqSD zHxHXRZqpqC zA9Rd9SmGDw%f^%(%z~1{CFrFMB{3HRJ!)jTj-;ytRAnYv%{I#p9`mu%Taz|tP`apl zY7padW4-7r{5*{zA`G6Ao$1f?M^=s}?20a#87XPo7Q`xB9zc)fRi`)f%GOnPiv7Aq zlJsn%b*ahswUNsEzL?c9Kk!FUa@3>vYtPH}iyASd%Agog6h-KjUwixwhL~ zGwsnv83(xR19ZS>C8;}aQWyGK z0x3o}s56z&jQFT8``EqeD0BPQ6eb?4^!EYU!tm;UH${eF$Lyp0fpEdTDdvH%=~#n1 z_gUELBF|N6*?PE<_qSD6%0nwO9`De(N{(4XW6}1$>LPPShQ%!ZjrW8PQ>gC_(WP|; zUFsK!o@EG5_*o;zLG9c4J)S|^wn66FaTSEE?qXIl6Fkrls#PP!znH`!l;W!aVv=s4 z(4DcrW~XkvbJhN*opd0UD_~{muC+;TUFhCgxOEp@Qts3r{VAM@k^si%hu!+s{`E80 zRnL#N%$-lJ8O&}iOkI{clx(%jgp0{8GfC2CUFVl3>3rYn1$y@|JF&?OPmWo(?BEJ7 zsK}CrbFcx0|4ZuTWsy|P?C-q1NVF$>Q^L&p@~+x=p7rr^nqp+$;J76l*9;jB$wrtj)QXI9A2#(mmW!|DAzg2)Cw+DTV$**(YN=!pZQ82Yy7W^z z?VrLKzbVW8b%MRc;@D=7cvi(F>EhIaT&>X*QPo+7Ys9KOv{oW;$H%jkVIXa}{E}Gn zAW|;Qh&q9$tWGcMZ;P2`*Q`j#sWE|};VGbl!779dHqWuooOCK#JJ!g|YMsB}X%KJP z%D}&rZ41<87&eAl4q72qToSdXUJlmU`c?1q7K>Br;xzfVC(!8wg z9y0~i++Lq%%-Rd_-6p8@;KNgr-)Bbqm0%Q%eqd)bQda$n-{U7Q^U%(+2mFeLS)Ivq|aGbHqk_w5e~UsnEE-`8V|qp`Vf}6-00DIyEt>H1Ch>Z<~*qVASvc?2-0ST zB%L1A(tYCGcfy_(9kxg4=w{VX#0JCE+f~u_SpjN?WoQUoacJ{fp`X$awZpqo&Z_z6 zNJ7b`Rl*^)755pW31psT@Ym5IG2#lP;p-5cA7_amG|~bi!J_JYdP&6>x-Za!&Lyeg zfs)LVezOLMUhBlpguu(y%9=!NTVIX9owJFSYn0m%P8SpxMj4osu!LmP$oDN4KRG~K zosPU*W)g3Ja-D_kdw9aJ(z!UJc(ttlf7`s?b~EW~+q@|lC(7;G6waA1chW+%-)|K% zR$x$l`gj1C{>lG?@`5!aX9kLr4+JB-f%ya`XGcvhE41bu+;@BWV$v0M@C6=m2OPAu zw}Mn49j3Z~>O;-8o>Wm^;|^u~aYe(Ab-ah{KfRdhimsEa>S8n}mw5P#mVOaMz|c?K zTR7Ejf`3MOdVHn3tP!Ph*g9(T#G9AV1?<{3D3@0pYwTc2zN*{+h>r+a)#wAh9@|-%VEXzR) z7MdYhOBMSU=%l#)C9laT0+G0ZGc`xJU~l5?(fwCXAz3nqZ4aKgzU>0ZyVjR{u$AHL zD=A08+tV)uRCjM>-zLvI^PeCu%kau zFYFlY5@%%WF<5ydNb94V6sn6Xrb3|J3K>TU-a)7@J@>7@zBi}JW0==HhaSWPJm!IJ zStaAcGcQs4VghvdC$Pbb%4srrQ|)>}Cx`kReYugcrx!_BCbRBOA%j_iT|Ra=NFO9m z)nTI*(h;5x%Y{<*yTrP-t1>aCm7@gR<@5>@f@t%w_0(~*G%mbkdys86QgC`Z&FD@+<`stVPjE$a4wb;OtS%wY@nsE*#)WP4 zvI?=JL)bAIl!6fM8v79GAYbUXD8}O1_z8a#QI7iOjwP8@2!2B8-ptv?noQCIKLWeh zkhYCqu(UNa+p`LwR?{5toED7T%dA#PAJjY(Qb3V{ELQEI{)2X)6Q3M2vuRrvzJjax z{ott&&O{OLV-8x)3{VyAP`jzRWn_j4OZQn>r5Q6lRYvqRe#xN;m zMfDCVFaH>t4YKB&=Q&27Ca3}pIj_i;!LI#%eyFpZI0q~&omE2oz9wnm`~r=X;^{~| zrTJ|8h%{>vZ%(7sdf$&vI*{~&a)Or)js0py9yRwtSCZZ%np7@9f6)x5ts1fPc3Rk< zWtE}Iyi|%;gFDc%M!TyfYiQO% z&)pq94NK0jX~QbB=^#F-apT~sgHJCfyLlhR9L-}ZkHFaWb$}GjwAU5Z+Qa&*kcy6s z&x+~Nyh|5s5odPa(#(7-+Mn`};MU;^i`XR~C!OZhlSLwdo$4+D{U^Ae9=%3`hjpth z*~x`L?NF3L;6)?GavE@&6r3cdOSW~^L!4F&F$f$wKr^xCYH4<}LG`@+n zLHNB6j&Z+$fZ^WbsXPNDHkr1G-P$tzgPMC`#Y78`?|s9f)(LoVSSIdju$fJ1fd|{? zV$1|fV(F$=8c~cwA&nMfyxGJ_hdcC^3rk_XlRNRZqwv}tjju=774NQmHS7Op2+A~l zaQ$6jawRAOML{8xy=i(*S?*DINcHOgg5Q;!BB6pB9yGtGS%nBzV~MlFPb4Ep#jojl zE!Xc~Hk|rn>wF0aoi3ET7PaPle5s=4UV$67GbH!@pq=p9t!;K3-baNC$Z)7olGXvr0L(2x|D{ zgpvH;Mknh0{}^#XU69xTlK3033B@*GO44CZ3pP{cn8mGYs@+=xG1X_+UCQ6h4^tld6eJ0tz4e@pG`al4<@gdNm`=ul4JJaTEVR z2tWNVgz#zi884@8WFpIs`FI(|PBMdnhXNiEg_#rNCSXG`IsTB?@W)ya33{Ym9S6)LZXl(`s zC1gnO+?Y{!&;>uK?p01k*y_8yc$z!Nk-UD%!TNU7bMNa$m6{xFNsEZ@Uo?WqO5Wjs zvJQzNKlt>O`{x#{IM|w(SLkUy4!H7J?B5#D|Fi8 ztZR8jm%z}}-|uKMs`~IFLK_iKf6Ai$&U2j$wz4Vv`pCU9rF)7bR{ioX+99rF`CY)} zw->G-ZwEhkXx~TVmihcCNZ(y@BhH^eF(IP>lggDr#N{gEDSs=Hw5p0%2=>s14Fp0}R$84uR~tM)BWwc;{fVwjd?B zz!t1?SH5VGx*V>8V#0mQS$~jwwowND+Y}v1K>adw>nov8OV}jrI8Nyo-J6yvIa}lb zz)-(bw%LP6r!BkZ}b* zfk+-xtx2QjlOzC9NWi&2-zvBcH=q{!9iS2k-I|+MGyS+neAHbs7fHhVoy-OaQA?mL z7Mx-=3UU;t_l6*=lJ})gO2SBgOBH=atg~E_bj59n%#5vKsX*9s*!zv0)xuI+eK(WC z9->X69iJs9+r>pi+s0=Y>n4_D=Z)!|(7ifKsH%hu%sAA3Uw8tXBe3r!#*dU(&Vv=0 zZ}@Xt@1Pb_zbn^|$$0{@`n3_t;#s6*Y?cvn=9v-<+mi%o}d8>0r z80pEH8&W+=8QdCV4H*Hy5y}*G(rJ(9OvvQ7BLImg?b*!|0pKSM+slq* z#}l{1JO^lr`#FBEI)ZtlWJl2SOy%^gDNrp{tb6v+Ga%K%AurMmHf^6x4}75?Zot?o z)SZ8KLb0^jaFju&>*)jQE`LmLXqu0eAdD3uev8kark@F#?+=mGK39o~A)Ibo-3AL; zR_=?Deur9?Mma4ASq_RnC!9xj!W+w6e=>BZI+Wt|*V z<}z36F#Tgpq>*$L-v>XfDSg*%7anUYM z-Umu-SuIvk#ET4+7M7*Rj!%)q%n+pNU|2rD+cP#HIHPyXAi$u?q~dc1?Wo(Xn}D~& zUcBL24Jtl=(J;~zSFRa2M`rc|^4%&Oo=>z_RGB(ZQ!K2Yj15=!wbeDlo;O3JQDVok|;#c7TTy_>2Ey;5i z{}4egfGUj2J}!vBNb?OZ#~+=_t@P#KTZcjtE1Sa3Z|e7sbd;_XT_(cYe3D$BWGa`}rvKaSQVo)OFHEtHUi@i+PW8r>AoE&hWuO5k?Ic7po2 z>G630=R{rG5KQm@e4$4bgu<4PfGtAVJO()Sr;1C>wB+ftp72Z)Y|^y=+6_VJu6vnx z&v^(IdAyi0q|Fg&&p40B@9vX!n+T;!EnJ*0FLZrKlWy% zlq9b=gVJG`-sR4*fy$F}j-90JNi%6Uu3e;W%EXCG9lY)JdIEN{ucm}2E6bz`V?Qnx zG;VE|GP{*|UVk*%7^^6UwALUTase0wCsWlK3;yU8Sx+2rD(6ngE}hio_}ZOBcB<22 zeogyJ+K{;CuR3u+rvm>k_TDq9sjpl3#R8%0N5*2uP7$LJ5%~ozQy+MFA-R zArv7L=@6Pg=uMc9P9rZLT%v`pr3A zY4C>2EaN1HLIYx)YLy9Y&m;p}0}<`VeTgPa<|Wj>F#y<){tYJcF3r-io%d!e*5?V8 za!>p<)F=|H6RFc!Wiy-iL4IJvZ0$(PnuI5La7DOfqex4&AeZ@F6Vq=UQ3TL*g8G7s?pZe-Fz65E|on#GWy12&b%|pcuA2k zJI=K3WN+uJh{^Mi5I0y-;FvHMXI+o4GLqi>7p658O?rI4WlrI0^`l6xWKg<}XJPh( zj?$DTEY)~)RF#Oq!xpFXe!_7*)5>JbUQEafMaSHj7aIFADlg@GE3{W){G>67qeSvr0(%N? zl0SGk3CgO=zqYyJb3z}9Un&!joqMy3)z5P=F4@OikNte~YGduIw28RX-WlnwuEIy` zSJR^+X_z6?DW~GDYg~DSfw!FdNv(x~eC<=e`H-OKN%H>TEIpM;6ZH4p4#tr?h8pjk z$YayByz`&sge`P^xANW%*}Ypeq0Tq1_kq1!%9 zWF#;}wJ<~LA{;Q9B&=SC zZr^M#c~JhL(i^?;kX|*G?c40mZ$M=IQ%y-yBu!aJE8knMSqrtY^=SmgkgvhbzrkdLv9GlW!FPnGwdK&EYw8FTb&AUFD8U9SV zGA?ocb5E)A-S}AT!H@BIw7qS2{**HvA}{pgt4Hk+9i55~3ydRz9L30RB1x7h`DgRF zZz~UKGJb&4OJ>_e!q&f55dO4j*)}G2;%MXJTYn6D`G~th(zvjYjCR`9{Pz@cxx`XU>>Ku_z{_ zufLyRTWrIzhW;8?YS7is&y*n%7CBvsX~j+k#e}ddu_j|KZ;o-y0EuHBjl{*~2e$?# zf9TO<@6{sVk%FC9cx8RmnW&FOQ=Yj<>8!(9U!9F>n9t5ic@+#L{K)Xms9PB6mfpJ7 zLrvR7wum(~=>}`%7M_d`wfZzK%@q^pY${?1)xwXuV8dK6x6%|cuX@to_7&N^^W{vj zOOXZ>*e%-1c|DHBcN4vl^E#!1ZuZE4!c@CBBwY%0_}Sgj4pqLb9|l-3Y#G#ZtfSXM z$~)O2k#o>Qh(2Q*n+Fw$X5A)Od#58>&=a;{PLYQ;alRn}=v8yD37xGjK!qEB@XCoR zGfRA%rTyJ6bA2k>74!C7?dQ-px*?<41Ei3P>mPFyerFtu6fYf4p^^GRX_4v?Z?92! zgFcpP?drUV?DX9Whjpsm+uRbGeTHQc10O!MdijS0#CRF zGYqN#`N(&u-i!5|Z*&%_tDiw%ZBQqrG~e+bfw5n7esXqxEHO-p=3R%4vS)p?e&=_n zkwXe91Pw(1e!U!Gg$*I(va^r5xO|s7iG2#O!t|w`rG0DNmf?$O!y}gZ$qtIRkNDkR zOC605I0yMflPLKuB42H6JN*)vIifU^3)b$pGwL z%!gV5d2T?oTD+HExN!!#4EwNOJ;ykIo{%1SU08d$`HN8P)WrY68!SR|B)*yl8|x?d(>{0)4ROl=<@s@#sMI@ z|3VICd?JDmJ&A8tWx|^Wh|~hx46R$(@z5{XE;=p=ZU~AM?0y;*(fY#Ha|-MFJK3V2CrA(B zBuh@}oo8gXx!g@Sn+Wk2JWaXP%8(yJ2UF;!ZVs2QU38v$0wAhA_MGos9B=aP75f7P zu8n88VMBlL?Dn}u61PO>0cn(Z7_O4%Qd-QqvClYVnwS$YK6VhhP83YTbE;dsxYlc07RLSUs%uuuasz&bnfbr zehk}vbuj>R6Ho1K7jpY~H(f1CNo(6eWP%eQeB|NfpaM`_gnN57-N&7$4_k?ZSemPw zZd%d(iHu1K}%AJ>N&uzgY~=7+ooHQSyGyH}Fip{HBifc?pX z-^S@DP0<&n-8#s`M4_K?sfU`qiEam5836~3Qwm(#9b21Wd*MVOih0&tv8wmmTY6R> z=Gr2+^>6X{DQ^4s0d{FSo(&6L0S&oh!-3y8+gU@2d}UzT`IT6ljL_mz0p5IWyucSi0sgK zE9lHPnm~ZIjqxbml;q`_j(bK}Gmo46xdH*C)ow(JnAdfB!5whj3*N|V9s;-)bMSMm3oWn6NJ0Y+|_@eF_f1iH)I;Ak!7GbC2ny; zz7FTuEY{Bp-99P4q^kf8UK!eX@jcDCj$*hBm7-lQiE*OmZDFT5{)~ss(Ea-NdfSJT zy*NsY-IUXWzb{|uO_of8PEB^aiW=mN2PTn=v5E1)3SPzIyfG$))SgE_yVM*!s_Dzl zPB>8&Ad&lxLvdy#ycD;DQO(Tj^LrH{5TlDaQX9C!C)cqT-y!*$&(`UO5|RxEkC;y~ zXbOy`Wc~u>bQwNjV#&WFA5c(dijIOwJ5tqo3J;;I&w*&QGGv-eDx2+a= zjrp{Z$(j300I{r`$6#OPciFu#BGJaKb3nI*T^R9K%5f>tLyYe6Z&lN(2J33K|59va z1u^SU_y<_Y`r?*~X*oH%lEW)hY{d!tn(st4nTY)8VpcH z+k6DV#eXFF+!p#%!_tVyZn4Rx#wjhS_9tO%o%!EZd-%}8_i16^6dXE!_OBi3y!b=p zkQ90MxrtZkXnBWnU3c)_>-GH9qsF5lVsNF-Gc9OED)K^e5rxVx2|6d+9Pd!moV+D9 z_00DeUr&20WnQ8oe2>Obuyw2Gqsks%cF7O$fl8d~b^LnwjQpUAw^3=4j~HRs5K9)z zR(ckB*VoKr!8cOx|ABNo{u|PvF4r^&!SHD1WJl9QK+n=s%@EnkW6>c*PR3rZn&INt z%Qpws#J%$zB*NMlS$}-~gIB7$;XNfqeym8)^!Zno$fmdKJ-hVlZ1a6lo;@(qLbSRR z_tP{Em9kC6e(BqWqGbzxECsL2?_P)!*9|3QajwhGSqME211}gBN;@x?cI*8kVxt4K z5@_k}SZZ&)oCsDCO?YokH-x{*k?}qBI{V0QJAc~F)>Ib%bE~GQuK$;NToS%+mNyfs z2)}qFr|_i5DA(bqqEZqQ_6^gVZj*@C1JqnOH^Cy@(Kr=AU{AnlZ(K~d6-d|8O@6Aw z7T(Cb(d}G-gnNE zgSn-PVi*h3ZmHuD`$KW@#svy2Ws)}tq{&y;`Ll}{(X32^PTCSQ*cez`5X?(DlImqb zXVMwJ+O1M`G@l4o$kzvNd?%q6hB3po)Nu=8B^srOgu<&%G7%)k{a>=tFsJX_FMSUuL)e~R zcMvo(8P`QzsaCBy`UmgZD7rbDpH<6tS(XVla+o-qx77AN5OXZdzc#;dlJ)Tmx1g!k zFX(KG0MVC1%Pg$aM0{ew>Jssl@~YHC7!<^>W4lD{yNrx4=52YxFYB97otPNfr(l?8 zw5J&BW1H4y!q%tYe{cr4Es5JlZnZ1RX;3qLI$q8z3V&3Ghu*%5CTf=9KAQ#!b(Yump493T+(Ewlaz2GSuyQ5=PwR!;}^vaxo^Yok(+P$paUa@ z3-SkvcM7^mLq6*3LoVq?ED5qNH!ki!6SmDN@=g4eQs1y3m#*?^_bfzB)O)FaI-pQ@ zgnM<3i|2}Y5bF-6+j_j=)nN7=8@wpJF+f~F|D`Z;aLegkcURsTr=QrZ5Q=v;)k|9*&@{%$DRBt;YMyB3cdo!y?uk0_5U&)N#d($FB; z?y@Tn;bxM#Ct;T}1*`4cDhkKLVXF~suamt>I~FjqO}D=@$62}JsebV&d5(=_e{<7| zXB=tMvBa0Q=X zfZqRyv=s$e!7k|jMcN)$*D5Q$VC(UzPzLw*VVkPDD-?!7@MkS&k+M|zR%udn#u9Cc zfS`(x+`Gofi1IRXs`Y-QKg(D}JLdCqin%Pgq#aAStgQxI%&l2i_n7BNN`id8R9#|F;-CR0u zd74ME_1NJ#e6}RI5fhuqQA>9-@IbkCt)ONWsn$fbRk&pbh$E%=tI~+-+ZYwv4=`+& z4uwdAJ*f{{$4^5TW7&ik?*myE_(a<9ES(r%lKH*U)t6Fu6vZ|wGwo0cf1T8x8xh2A zGJ2NyB}!I-O_v0$_PlW)oiZda0dq_>>+>}B)`0sl=!l!s{$Fwnm4}RSr4O&jCeR3& z&MX4aHXogkdR~3Tc(4DU_t}4x`**ST4Zr^O6mt$OCf%eT@avX-kwn(UbV2;Ljfyj; zaH(;1g+IpL28elyQ25S--a39Vq%2bs8q4~PS~W6DoN zWqYOx?+?wJiU4h277@`rRX*7UO`Fe#4ZpfJuw^6*b=Jst2Jvop(Q+=*eO2a=+cDnz zj82OEaI!@i7`7{=xdX>>o5CKAbL=kGDbJ1sVWR zYNQe|c9zpH^zE%8XxLg-KrJ)C=hpB~C>TVIVIJr`wF+Pk-Q?Ok>AH8YzU08LD`!)R z6Xmva&V~s|YIlVF_2?(mX{hgazW&91JmMJLlm8O1;_udy#(l|0?6sma6#`FHu{q1SL_YjA3uW2h2nn2mbb4L7A5EZTK6jaA*aQvR^CnRPxqKt zmiGE|uR<6wt`1(1*aQ1tzwX(rPuC0vo2hE)-;YaQ4k<{K?td!u+L&Q*jUABsg4Qy1 zGkaCqQOQ}_BzsD&sr7dNJZp4Mlbw+)YS2K|`fZD~4OdnjZ-$&5u@A@c9u}kOdul|H zA0=IghyD?i?jJhJ=G_U*i`LNdjzEK`4fh0!MWphrphP_9yv=2svE;&&+G= zteZ>T>2DbUez15M5Kd~LC<$`W?WL$RNG%=OmuW4_<@u4ZdE7m3lg$ae`Aw1v(cgrE zR!1PQJ!Vk)&B5Ujca6a&zv;b=U40P*AU(X=$g-^)5q3DjL4Blz_)tP-nEOO-i&w#JlSip`o2&iqQQ1C4iVQPQ*OKD^e&eyn@V`hQN*sYr>}QP!1x8ZyTECB zAZ3l7M(ut!gG0*<~KE4OO#^In?T%L)Y^#W687O-*LYH^F$;&#W4i2T6NXE;vbkCucdL~ zJuf+Tcn^}7F_|Ztm4^xB$NV?_`dcjAzG1`rsZ-}4Jk7)BXd?X!MDb_uk|cKMn;m1x zL2Q^^e1Oqu!C2>m`=j`Z5_d;CydK0b-a-msVX1yFqK} zcLt?csKRm-OekhJD8ZZDWBUz%#?0^I=LFAvYQbxC1`}tpzkVkTKZ%9 zN2{IH{3Fta#;Orf^XRSfK@yO%5@CH4#W7P}W-j3;)nmQf3oP8$sj!}VS};r7GvY8# zc&9>0amJOI(KGO*@cTK*22yJzJk9=GbOx>R$$i7k79su{R!A=LFvKiI`4Xdzw3Ut_G-gUfy%j-OXlt!7)EtwQw$0Aw`JJW)pJ+E zolT2eAl)PykPQ!J?jea<>Brst6|j3P5wAhC&-u}wpHLj&zh^@XJGi?($aBG zc2Np-HR)Q8+(UkjEoa%BQTy!Gr|5lenR~G5jrP3`AJ7;8vkG7z7Denc;p6O=+G)eRiLtrGP_>2T)WEHI04OQXF>W6;DujzSiRnkolp)-WzX51FZRdzMx!}Bij{bh7& zpKmkSRgyGBL>hA7fK(VfYxG`9T0pTap_LTMQR4EspQf=C8_61l6yNAfij@&t>r)>x zI|W5d8+cqGL?g#LYcS}vy`lRn|Jh%sfGhKs$~o8T-*s*wzuGZ^)=jb*o6x9de_aeG^?E;5eV@erj5{$GC*5}@ zuMxO-4w<12T|%G6 zj^`+e7P!PE^y`s$_;RBapy>fO+g?aVJRZ-}gMSL>HMXM{05$l_huE}Fk(U*F*Cm=T zpClk}S4`VWM>dF7UgY?=)6&&%#|{p+kUzr(-1&w7;|H`(cY)&2SX9e-TuxfUV70(Q zdam8xlX$AIuB>Ox;`p-*`>i|HfR$#zxuA9M^UZ|X0%u+M03DX?#Y$0H#;Q>yVrvd1 z*g`9Y`{MNx&4H|w3{)K$7J|v|Wv^Fb)u%sXRqv6?3?$;VtNr`?;$w&&tIZTd&iO`qozXo>+K z&D(;-Fzb4*^qL)&6gGq^+W-#E{WXIIKB%t&EX{jr<-^%0%yiK}WgD z0y8L=bs3dm5bzgJ`0zOCIe-4cIh6Y!yZUq+n z&Iti+#cCFNk7s@It5E72T)#GNdF;1l>pyHsx8V{sVjten;|57%Dz`k>pUspmeB8cX zVRf)me;SsVLzgX^YVS(&>5d=w=xxW&k*OATy`7H|Z|!@u_l+f$8g33e7wKIr&1YCGIqsN}-j=1Aj^X89vAQvpp(j++ z@IbRI|22mq6Bf%^AcGmP+W6vK+Ceb(ttz8=we%ADWhgE1P!@9Cm}#lSk?6s%`Erkp zYO^rtlWB^q)%d2%F`&BgUPlL{BBj+x^{h-H)-CDFtkE%ZB*{zY*&#oP7Hf>n%_J;Q4f(m&q)c`<6Wm z`gC9n4esaVZLFKh?0neUgJ012P2XN1?KXTzQ%I@biL((TKg+ zwY(v+Thfq}RFg1`Fr{|O9?fA8+S+qVdB+y+AX{&!&D zrvf_nDW5Ql$mzazOME6M@BTKvYGjl}Ncg#)g-1g65&@M$sAc8P7usNRS48`PD67}I z9O^gAivJM6H}D8#Zvv5Vh;8$G6Jx_K(}=bR#pp%(RCD%~mtG|)&c}Ug;EHZx40G3i z=r4P36~%WPOw|iE`EkXaP}DSXX2`1Jll!y$f4~0!y=F%Kw<7A&#Ryw{#|Qtai_3x0<#rh)uuwt08-$Ad+NbI&v=AotM~Yjl`1PtBMGi z!ZQW$Wbjum3I57luyR%lxJdlA|C#3go5u7+SPr(J6Db)5h zD&NsU{(ElB6dnyvWqe`k*uFuEGT>RKJ78!i@V2^M@Y$Z{1Q`nu`&qOrUaL7bs250M~c-SLps*8&*!^S4zFi9?WUo5!8!b zg{!FaEU7`uxwJs}VL8!lvrf{CTJ?>_VbQ6~X-iM%g$MMHdCWRND;~EIM0H`wi*1SH zI@r4&;XSp2(Cyj22NIxK5W&>)GckNKeJCQYEH;?(P zuRe*|QEA}QQ8J6tCN)-(GG1q&o37fc#we~2>vr`fNKgNgAV!{A( zqC;b5G)ntW3>IG{Fp_b=RQYY_yx?e-W-hN#SrgRRVjM5@@>9WdF>2?Q__Imwh9)Ws z-sWx$@AgmHTZ1W67m*rQkf7@8jif}}$S@-37n1PUD07-+5UakkTCx2HFM~LtS>RoI z4o0Ar1=o*GUt%;MfneTcq`ZdIA6wFs-|FT{&+u*gYDFetX1T=*IX^oCRP_3qQ?@eF#v5h(DoQ$EkD;29E!oGUjj(EITbq!1NFOY*stz7mE6RQbmRPZR z5Ai(tEe~sOBc`h@SmR2w^Lg7fB5aT5Ks^=hVRpf%gzynA_!@7rV8^vyCg()$YLX9- z-W;c}jFZ-vELWu2;9@qIG1<{X8VRRKhiC_>b*&Va&49h8k33KfAC=0Mp$cYQ75#tHGs>}rLp zmE+8EHVQNR&Ccox!*I8xcR@Yu>G&VKF_wHM>8uIsf(7HkUN7go_&hU&$fNgU#@B`k zPc65MaicHUqsFgFxVyW-4RtFCvQ!I)oPkFo$sW)sp3kvE2ahXMHZa}|ceL`nd7nf2 zO9{Gn4DjF7&^Pur;zSF7k3Mzi5wD|4d=--SBKX$d(Z#64*%r-T>qBB~ay7mv{|b0{!d%?ObPK%(TB3b@pXX_I3Q{(yFe0y6mvh(m)$@%SbL6n=idD1mtLTfi_z=TG{;g}p-)7%) zi=uFG)X_dq9s_>XZTi}eme)gm7P)wmyNTj`4rKm=SD-C1yP@%0jaBZ}kNl|Ok44};fXAlRq25#FZG^P z_G(%@tfHkvc>fS=ShM~u)Q>&kp+bqMqfCt=^<_eN_?Oux0W|vbLJz0L5$Co^yuHOWd^8T!H~^e=N_N+6?tN_P9(cde zM1VFn(CZV7u|kNDU*?isaUQ07CK%}I3vM6Dk9H-XK-HVsY~6S9V_`5bj`N%U_eo=SILK0IO_5gA z%=yl1GmLD?l;eO)pQqzYl^2_O!Gd);aY~UBcwE@q1*5!RF7;bAEv@R!u&&bOVZ*$c zMQ%xUbOsjQPQedAf*LiZV z{t?d=symvXy+MSx&eYAamQqmfFxY19&NwbCniU8oXL3aNR6JwEc^scjGcOtDCy;Rf zaj^(ZkX`ECT%<9rX+2RkQ+QpP=e5|e29HX?ZY+$-B<`iKr_&?a@rDJ@DVg zq}A3_&kYBJ4lB}thY%jZl+uuNP>@VdxpZq{`%|R}PVAqk>1y8g;z@d`td^PEj!+xE ziyUHinQ@1$l#3hd5T`XB2bxhXJWHYS3#i_N$Tc z9N)ZW_sD98*Lt3N|!oM(vw?@i9GOv z5!M{e#+(y><)$0eueKcTf8*dd+wd6q2k$vN{Ia`&EDC!W+dvWQ?&ERFv(g^HUZ;JS z$62FSHE72qzT7~?0F@-&%TR%W3#fQe^PPYyE=(2 zrmv}M{?7*)z8t3zl^Kn;dy0k)_P#OPNE*6R-d#qpu#FvlmpA&GE^H!NR?$Utw1UE` zx=8=(#@38*HP){u#a&!oF8Cw-zV{qQa2|k5&l`hELJZu5bpmJY#?^(u_s_~8-}r{*-Y|3#fPM(K z=H&G5Tu_lBcE(}7(@$kT6^_N{+7Pi=buZ&wc;Yk{>Zq?jfNcN0Ei*c_XY9G)i$ZSO zuK?H9-ziTG{M)R*CuLSlyR+Jsoc`OzM=#$VuK}}laAm`jbx-)6N&-fTfVe~O*K;xa zE14eOl=L7m_NI)A+dQ9q+6z1hey?w1i{?ZE4#V&jI-MUZ8*583J$ZXHRbA>l`_5Z2xTHgv@-0u59!T(wPpVQt1 z6P{+?M!z~is8j?vTHn<>n_M<0hi?*jFF8L+Ge3|neE^n_SzbAg)wkJJ&lFP478ix4 zLdwJ#sW;4i6pq} z0p=w0rDXRIVh4`(fg|rFO|wB$UkRy5cGVY!WI=fHAnIpd%%!6k;wb8q-IA0`Y|nXW zCyXY-pW#VrUCtlJIS>G{CpNXbB+-*90ejE;k&Z$nw|JP4zq`eWKTY5!O17VS6l%z? zam|HyP)?U7a$6`fT*Uqd=rg&;pbaB7XD0Y*y6TYyD%PvEgZO~+sssc zK;YXN#g@Z>?boiO0^9gHopvmrbHmkF7DR6s75(63Pv8vpHiFvUkW(amQ5G((xcA{= z2+P-T{2`qst_H;R2ah))_0Whdbwz0~fLir)L9XZV-i?HQsaSM(2g0iKP3`Dev)VNz>n~zamZu^YXh!~*HYybn zq?*a_SY}z_&~{thUhMr530072FFW0*a7bQ(NW6>y#j*?hx)Mw!I3g3C5VNr2%Ey&c z7l>8hS)a7Yb3|9*@wc|9fJ&((*>QxJ!EDyz);`@Z=Q z9cH`K0S2SqO@pzqD7)B(jfnuN^_p=+P-$c6!$GgBe&^5iV@Madk{33)Go^29Gkuc0 z+Bl1)`U2O)rKaQo%Xfk{XaqJ-`zXcoa2*o$y45{B{bB^;2RX;Q`(VEg-K&s~udn=I zb-_z#bGCms(K9y=absVrxJb&4x~1nTG1VZ?WBR@%r7&@KY@Dh-6{UpWxRer(g5G+z z-*65$h(gUpWhQ50K$gZYh;lVMWdGx1hjQL22f3U*nd9Fz? zSCU+UZ4lRwKZT@?z*D(WPTbF>*mBHjlb}LJLq9;eV^xF~OeSxWgAQ9K6w^^b-N*VE z?|uJZpVBU7&;aDy;X;D}bblQl@@*J&qQGYLVSmt-kD+%wsbX-P!3h`xGZ>q*_+16I zVQ=Yyb;X&vUCNz&!|04<;UhMyDKc`}T0{0$DF%9XfeASH58g|Zcn`Gy!$Gkf0rNiR z^MgDP4|v2A0_*cQx@G=T{NJ@np!1vXW!h8?I=DfaQ&rFBPT$*=xR+};j$IEBaLr~Z z&2Ru=47jAcx@q0JYz?ErFH=)x5If8p2HLutLCtC7TM=oI{!Eru_je4gN)nDx}FPW zY6ydmkhHwpK>_i@qhoZKjp*shehR^0-5g<&2Uqi=5zdrLE-@~DE${K;D@ z4^=p``cS!rkM++w${@1ry-;G@;>_Zd(ycABqN=M*ZhCal_O^%M=#0?Enp$dn20&vP zUM$|^SfMZk)6>dMk4v-C)M zAf`+&nvh+JQQ|C`eI8%x6fQgoe_@>6EB@>mxTS86CG8D3kygD&8&mN>r7M{^;ZU)= zm)=E#=hK`SrmCh#<#ayE#+>Z3cI=IzN5iIsz;a|1+vll%!LbrDaz_!W&jO)C zKROzEhUPxEsA>mXxX9aI^&XdnW!Wu7wFnKoXnHuM_=Wqav)iTq`GWaW zxn5QKsNoi)zee=+W8~EYj9>$%1(H}0e|aio1b1OTJlcTq-@=1id#0>Un4Y8-)-q(0 zCL;AG-jIn-Qnfk!jQbvWPgTc1e+o}LKG}pTVtM1Sbu(w&{f=%azV|vYdWT8RmyH}F z^Hol-jvL9}v!u|a#R{xh%|_qtR$d4Ka#omF;=kdV8E|}K(>7VG|FSYNqTq1<);~ti zJ0F>3MK>$Rr{s&>CLhJZ3Idufz80h0^>A!D15;-vsI5G4FCfq60)p^Lq5dp2-hqow zw)hc_Y%9mj-d<3}QEJB4wO2u-wXj*NHUGQiL<6z^BvYL6s=)TEaA}8+J!Q(QR`fHU z&qj7;gZo#UcH|fSS-YHWtW^!B-%T_;^PglR zGhS`qe)UNj1{!({(-lswb|;NJbJ`yV&mVZmo`FG)_l%Yk+nK}QA3eD4HCGtyj1=qiCBYtG)@E_Z@?_d#$QNwk36qm?n}dLwjyGQWtY>z0rkOpsYxp(Agz z;Z^tyjL*l(VDZsKeYUjz7wtG|=^{bV=y`R^440kkI7#NI_s^nF$lA*o`U0_=*fs_~ zgdL}FA>yWT!aD-~!;Bgs*0D?uOYTrd$A?B=<7Bk5PaZ_i<#YpiA^US*NP-@ra&#RV zzvZJ`fwjy-Dg*I(zA*Xuz^S_bl5=AQx*%~K^8JlG5(y3txH)622*%o2xnq~ z1gLtWCkpFcdm5zAFhV$o>>7Ybpd?<}7jwXuHxMT)5xZPi3`$nNkG5?SVTJN`!s_YYp>N8mRAl z>Z{Q;^`V=i_`K;rrdHYj!|KlC6W` z08-FHPl8H0PgKv%dwbpo7fqqkoiSJ_@*p&??FVbCh#fiP?aFF<;sJ3KN19#(=7;*+ zP$DI7*KL=ugvD%l>}Mx3-_xAX1tI+T*6I)4YQECzV5SXzd=lN!uGW*CqyB#S z#mTLbTG1b68uQw?uR#=9aCNUX`_S5dFxrdCir7Xa2Uk=s*2Zn%sl)`KaaPMcp>k+1 zFW&C*(37vR%{NNNI=nFR9|h4?8jbjEO_Ovx8Zy)NdW+a{%^{;x-`bcvo%{31!+x1W zr#*RpU-YU~-BG@Tvq>5HTAl_P9Hbuk!O&VSXl9zalR3sLeQ`p>W$U+iEU*2#A#J|o z7YF%;B!4e{Tz`4BYW6j^yok^yxGz$Y>ia=aiek1If-cfC)4+l;AN)hGi&B#z0G%go zH>P+TRnA+s1;(y)H#tOEOKQXw`P!YJ?4DO`2AXxG}V z3``Ajp1(S#BvE+!++Uf>KSv!Kbie#t|7x`Q`gEl^zqDmnrIUL{v_%VZ<+hM@2Bxz3i^K0mAW%p$ReBpT55qK&P& zNWOLG%qLq|M%cSl7gR?E%6s10d^BU8oBDUzXvS1{7Avi(+o95ETdINYMyLI{R+@sP zF-29r3gA_gHyS)qvdpF`NN6i>T}HJ9sw)^ID=9wlAsCFpRZv;!RI`pTV!2OF_$rDgU6ay! z!A}#&YM5*fYhD+H$j?!w^U6`G{>4*8tio`o@_&~Ve_Nr?%})dF2ZW`^nbI%g0bW z;BwMRH8HQb21h`kgKuov#U7He+FGx`Ov<`E3w)xw+Dg#*-at3fOjhtJgYLOPRXaWp zpa!0^jXV!lBd=8^^_*IH2xQpOvGP}kp$2BfL&HMZrj$>Yy_^#h(Imil=)R z4;67YF@7q`lM>q?4yBg8g@3Oh8Pc5cJ~jFriOD0gb!Q=y!F`vf15s&TUZ*?dIm%0A z8E~xDJ&cE8&?n<@rJfJMBV;QTs6>!^`IEE(+S@vkfACfVmY+KXK2?)D4w8o@M_Xar zC&Kl_UygpC1C23KcWHs=biYV+u^b{x&CyAZgKcXQa-?=#J#NG`JkyE;TL zE%-7yWKjnxnH42&c)i$AYsWKVTpvvyJ=9@YXI31UG8Hk;*bU{`whLEId_LJwx-1U##YrVr}B4 zrV)Zu$TG_}q@Fi^*Zi`(6GG?(ZlU?MOvCX+26B)bz-{Q#P)8GOu*S5&S zu+--m(xHC0MlXDSy*MF)s`uko?ee7LyK?bpWqO_5-ihfo%$L!`vmx$$p=Ja8zun3w zYOFsx-5Kpsk~1;f86pTwf&hVfkLwkzpo7$kzTD8_mlKz9q(+Q+4cPjk(rLx}zPGdX zs!^qttvG+vu{yk7BW=0vT>G_U{tD^39f@ov17zwQ4BWH8jLPD8)9=h9( zjSpGAH?S%83i7GOyMwb(tHqS=FRAVKXucddKi_AqMmKx{J2^IjNM5@ zV1?adAQ>JNWrSLqA1EN7C;f)gygRNZB}g6Z!nT%{{HgfKw`8HAU-cH-F8rz7>#FAo ztyUF(Pi{wp1fk8z*l2w}@6r(fT6^)RLX=vqFn-WcuJk5(yfgb^iZXVv;d(dbd)YtsS1M6cjGzl zk8kSv^F{HjA?_H;c}CTV+fI)zk4aOpe^%Zlg5w9B@qXpFIOf! z1=`D`T_26au>=T?xLEn?2oz@Z2kK891(Y#JC%6U7E@gfDZOJR}y;x;%H~De7=b_2a zlbI7yigxax<{%iD$$`OfomGXs92{xq76cMU;;7;qU};Q{f#9srnmoImbe?&H;XleP43FaJN3vfwH! zmd}qAgKQhGh-MoL$pK3LMwxLB1s6ykOZZgu&mnJ&Au`EcYLdrM(9c-%I9+npi3THIeBe!IxpC8Z<0-mF2ihVSk| zBelLohSj8bmld2Y9lkk!Svwaim#>{L;}nEd`tW2rZ0X^$6NcjN2?gZsHwbT?%S)b= za8#zV`Z>rV{npSoi~DAO+iJeun^$}nlYTr$P}0ueAZI_zdwD%aE{B#%*!7-*PfW8e z+7#DahkKYfATP8hrU*QHMRc0#dc1Ery>^V`8rEq;XK^X{>|o`iA@T3fb4qEU-@wEi zD&w*l{;@Un`AtiAMEVauL%Q}qbjm$w2fV%Jw?Q2u0!Y-MA2Kv=7pH48u>>{M1a%E@ zT5Ph`wLyLN?-I;q`)_#~bC z2Q8sXT57Nu)@FRV*>nhwuN$kvZKdMT@zWO3tZ}Ae^43+>Cm8h2CpedYT$qPFw3Mws zPJaVBJ+YQulLlMlvtNal(t@P7bxDIgL!%<6Lrk13y%&lK)=(5n(2NssyDN0J#a)=3 zT}8x-a#rIqb8j#*y4flkMYe~#nCu3pzC&uh5FJo))+1aPsGoCXH{Jd)^D-hzH{gqn z3T8BIUit^Eeu0Yd@x-g%TOireu@?|vd9&QB6E2G|)8GWoGPhdHqvRqEM*m*?>g@aX z_r}~s5cN;U62KVk6{GoKbJUjizxVoz$hQW*rzJ5l&U4RP5hZ~I5!9RvVUxo%AiaLS zUAc5Z0@Oood%^v$VqC-@iSX&=H@CZBrXqj+8eH%CM)Op2n_U+k;y!U(`M* z>~dMZDX~RAwK7`e82UHrcElNW?wfpguJ0n6x7_*E>2W4nens4xaZ@MC%Ug!`^ze8ix^WxeTatQiH=0SW3A%g)-1i8DVKQ=(C5%B zp_s4Ptji98)e$=iPV{82n7%B%2ZGRDI(G7>NWydsVl8KLvQ(B$po2i>I=muiB7?k; zGhH2GsAzj=kZ5RvYLF)%(&>)`xpKzWyXzAzWDQCpN3g{;UG%~ki@n3kJtoeACX$RW z4yMvXfy}avcUq54P)#kmu(X)yXqbhQe>F<6Y}sdO5q|iWH0ZDLuk&O?VIGkMpRSl{ ziwa}*!uNJ5tCyDY@I|xGc=}tF75KiuBtubVa`Cx`7G904rp50oaGw6RnS?Zq@Z=pJD-p17WN;)>A?(44uMWVX!)Ib?X{yq4n zb6A1JdUo)Z%^*Z$CWu_!G+#I{3;^wlwR8&u0|&_KLRMnQBV6EHELB-dXba-2VqXxG zWhs|?$;k!Vw=ev0&LJt%TGfr~00VAX_BP#SM`i@xyzq$M8^}Rm^swB{h&6x`KhPM46vl@ zHLbk6wh#qwA|OWB$C)rb;DMwm^=N}J3-odXMl5A+anKZ+JIlu`uwzL&(-HCQA z@cd@Dqglieh7TzF=ij7OAy5r-SZxYuKDqH365=Lhr~jj>j>J^oH69XCk9+a2d5EFf zTtB(=`$!0mby)b>^QIFTiHaYO?iU{Z-u!x_wQcU%#gvI(6*;CS0Cje~I@&X`f{2v$ zy61wbJNn1yZA;Iw$WwpcTR$3P^Z(Y3guI+h)cEz_-Z*+~mu?BDvvwBR-TQXybi5sT ztDgOR<N{??5=`Y>r2I6i_P20Bj?z71*wy?o7YAbqRwbxhW;vX;t_(hqN! z22-hF-i)Y)uG4CeUisPKfr2XVopA90Uj*~{$A5kP@zu{ies0I+*|>{|Q;Oynxt-;q zj)3Esp@qF*8-RcPSXS@C@YAY)@$k4%r(UUF6_Gp_SOl#5jQ}DXbB?E_)HAM{EZ3=* zOB*zdu6>RqNPkvH4HnHH{lL)GXU`F_2)`kjEwl;UDmLFJQQWNr$6xMjy2AkcN6%|- zTNEu$aj@f%C5y6{XjO54bo#CR`ZQ}3vF?M5bzkcXR^T-I!2XNK3HK~Q#V{*qCf^;> z>x}^(?~gNch$*bMX*+HBAozJ-f&sM{(s}S7cS~ zz;i-=^u?z%Lwo@YUGsiRQrNJv2O%Vq%(40|LKe=JUy&X{QgRWJK>wS7f0$$UbkMcI zQ8fi?Kd?@v*jnN*UXe~+v!#1pt6r)~DL?THAr|Z8d){(Ot4WiTidB|dL`d)$^Hv~7 z>S;2o9k$bC!=&>nFo=)#OYw`-qW7!QH?Bg9G}n%>eT8?`bu7{}Pd0ALd(QoO*Xdn+ zh3=d)iRCu%(l36awasO_6n;SeeCU@D)~>t9JlLh)^4|wMnCl?nXSG~ET4F`(p?vn+ zR$?9hUns-7OUL-1x7>_OYrzJf3zBX51qHZ{lt;xmM2`o>(vQpK35|g0RURpe0 zdd-6W$8J-bhQv<_yfN-YrH@O&yQyIN?QnSka~*28LnSm;P`=8E7@Pq2lo4l2{~y%@ zAP}d4j~awT;DxW8j(Va(emK5=`AS}86(qM?Nkb#N+bw?XOjMNT#m7#X`>}irPqL5N zlU!F3P2;{0ZLAGYZL6e4F?F*7l{d$EonvBk!_g-}e(UA~BaHr?5v?z4 zE`vpy3=ScPyv)yU=MdO5n@*P7;#Y3*dG1}MHvaCaY&xTzFa~s=IWuB-gTwcla&_2O zqhLc29qe=E6*PmaR>$qk!SDvKa7p+L?JUjb!Vo>vAi0Wlt%a#Tn?66Ij8ipP@RVAO zBk^51Kqj&PopJ$HCTGB*sdj5Mo9W5x$jvs+}tY$E9@xcY{t)|aCVz>>tD85;amL_ z1{0-DWBoAawk9+-3L}~^ECuv#9SPvQ^5_F1lP=3O<ip zJ05h(-S+M579HZ|*z2~OImC`Z-;U3^s`Vq=bca1Y3g~gpnvF`MZlmtNueJ_Eb$Y6a z#`E_pnM8|G{T*~z9zp#IHgeGE#B#hXgOQcUKX07<5AwmjU1?(BO-pN3n}06U;6hD- z#YS4+24cjX+u3SPdc2ao2xq7N8t*Fu=JQ;{$5WW1sH#MIGwi4F?>>X=Ev>uzs&LWD z^W33sb^A@4cECzfxw64aF0KcS$xnrJodVR;Y-4isOYG9`&lhGCIusYJy#w&XW$kIb zw8zqhVm^X}-C$KAIQ8PSkL}$tk)s}$N1Z62(fI;h*wnRB!+P${+vQCcYiDRK|BvcO zO(DZS|M8N{26vbhz_^aRNB_es7;x--ZTNRts1#c_-qmXvb6sf*6{Tak($w)u%cIy~ zZb0)OJxeUvSIB`cR>>5-)1g<+Ve;g7hH&L;_tJ?-WOiLr_C&`Y?f12~LGQR)(UTC= zw_Ua6jYp4m^3cXHRTjsdE)`V&=E~|T=mYgMS6y4g zY+(}D>RCo0X8n?LCoyTU+_d{7B~C!DBJsY^!dGZOK*>o^ z)HCqk4Rw=p7?1J~kc(aAW0hjpG7yAmuTATV4gD^OaNvsExuPKkn%;?rG(^M_(hD52 z?C=kDlK5}7xv)&8vazf+$hIKwBsxkdpQCwnzQE56hjHK%0Lc9Wt~Kty2WRxzCqYoB z{zfXXvl3}X{$Nf|)hoxicXflsUMp)EwgVkyxfZrahrL@5%2;8F=jSQHy;v*u%evTf z2fn-$ovm4DCU7b4Hn?d8ErH$&1jY0ONzN#LjLj0514L6Vm_ewH)v4QigURhRHl}OA}0Oyd~Gbcdom_No89Bp0j^4o8Y@V3^eZFlX>YXLQ4>K z!mvp4A?Yq!dtA3s`v%6;;{wXKE|zbMQsK-og}A$dd#KKxxd%lIj(j>PQ4(VlyF8=$ z*ZLFUZ)03Cent#37r~%XeExFxNh6x7(k=JZ_U^l9{u&tXO|Y86!e*6ic#-9yM5XIS zh3*o#iE^QgzPBU3pGUZK!oB(YI9QyzXBuE~ zLBtkS(P!drHkYYgjTNAk@ddz%%U@@FI2>Z;zf`6H9Gl@^sE%j97 z3JAz#Ba_`Enb1PHzj5K~%wtW{kSB-YjW3If*WjqR;-X)H>d%c!mcDk6oIFtpUDMn0 zDS5x*pE&*`)PPBkPr8b7cQUjqc=pkeT4d0D8s!hxG#P22@XyjRUp{zNzFNt!LZn+g)Dp-rfJ3jn8HCAeW<;5 zm$L*G%uGLa`0Qy@gMuSO>37vzYQoi%h?E1ZT4>_?+o{_@K&o|HyO)E!QS!+A`6rQ& zUma>FHGVe|%+`P_7DTw~eK4Q>w$+z%&7S_M(nGyL8PyATln3Y_ zb|u#TSc!>Jb<8cX0SWWIhuxWI3<;!0GnKAL8}Rc^^Y|mR9@_q`?1KXwe%51Nq3%3OiwMB4YKG1k;RoW+eJ|Carck$IZA$ zC-tET@0*xOQ@@*41kTBG1^xKl*Q#0II4P3lYWfd&>F5W*usroRGKe={x$EQL05)lI z^k`&&q0@h^o`uJ#Z)a%S;;MM)Vk|*BfeV-l+mg;yWwsJH9Z)vB8r8n5+sBlnxN4@X zcss{l{x8u>M*<<|%5#hE*q5sB+AkHXeu{-{EcZ#XPg_iSkGJKuw`yvpI;V;NV z6^nZjAAU~nsg&q8Vmx1c1#A)+inkT1VY)(}+w|KmOG;jNekwT0N5^~{ z&nD)X`({4|sR)(ca9Zo3aWav*T1NY;P8M~?#%lshmS#Pq$!TXM7knv;kP0F$IaT_p z?RqNf_RzpD1X^W;!u)_yvCT%Ni!*+nB2(PKII1dZzD^6GP}I=F4a-dz!5tbW4<<;W zy(M2=V2~(;boPzzzHhl0aNYfPfKaK_NG8g0xZNG;R}|#Er!x~iZ zPk`s#ydNLHF#g4(JEZT6F{CUtIlo&E6q{~Kn^=5a%)UA&j;dsS{Oa1Ylr3Lv+(STM z_o~Wkr`=K%PdY6vnJ~!gP}l8nxHy5{-x)Ifwq8e84bD>4+or>Xf)68IYys8suRj?$ z@2(8NTV32ZYm1_7Kd|AcGvQ)(`Xy47U+5I*i3R+=QxKXeG(OKqq3 zRL^z5Hpr>nVUSnx#9&#Y)NZ9`DO~s^H`VYb=>F8Zi}xO{TsZpq6sRMVk{h;M_@Rkk zo!f&hLgvAFKPf zm7Th**7IYBZkFyC79;9>W*5RbD*joRh1kVc$^BnWD&K8JE%C&)|#c; zLeqQ(=A_XUTnFZJbK3w88`|*X7%gjUe71c>18HL|)1+;;a$La3?beleQ{LazriNT0 zEA78Qzj@znQeIqUt#}m9W@=vJ9b_2IUXbb*Db2Dv!sNM=?r!9-6C<#$Ro({u3e#=Ol}$DYftS() z7AMn<2qR^k4wR49kTQ$k!05vA9b4e$?)kR_msdRP4vs)R7@aS9uB%qd*4sV_qCK7@ zf9eFda{T+utRg&!A&7FDs>~&A9&TO|UR{5bX2S6LJ)DoER>)iOfH zH_4%|lKsjwOBPX@{qyjuCEU$^W428tw8dt{Cf}j3aP8ebZY?ew`)#h->#>)M>|aPO z`{LC%|MaGE$OjOv`;zGx6_bzg=pf&i*T&~J?~=QapFNLC=`vgoJGa6XXNbSaow()26y z=K}>6Ge4{JrE=|*pZb3y?3b`Deq{!-;>8`p4h}nD~MLyub z53F&Ih7gu3Y2cpmGZz>02J77+mC$c)(g3yRy#1p+xbD|8+ha&-HVw3>SFp;bcbX z1SiNcGOt|plj#&lg&4?b&9qt>Ur@4vV5)9Cx|{u9v7$#q?(7V$2*deu&*_fraWlv$Qr@rW7l64G5P8Qz8YA|BdC~Vp_^F^a{vPbt1iZ)#{{5cUh@0x zs=ym4-hm2d`aVQ#za^`Fzjo4sJ48>J-8B)>tZu|__6rnct$l3(sDHZ-YF0y zj&Zg*j=f}zknBG6TFqJ7xRNj~Ks36UIsA(KIyEP`*_C84(EjRhxtdqtpD;c*q_*T- z8@OuKYg`alr)09&L!i}~!E<%`j}kZW463qK%SYUxgLiNmN{qVy=)*4*8JLdEcHfQi z)Ag$EVTwEz-9@F#gN;f_Gs3I+yhi zO?oz9hSUVE6+Lf1wMKPK(Va~;F_2tnF`~^j$vN3`vCrooeILP%T1y`_S`E{=tg-*o0*I*7z}m61 zd_b;OS3dFuvki}s8$+PfoRk_xC2HxW;C|(#QiirpuL4fsbbVPeqE@aSaJ<4)bHGB} zGVrz$|J?m?cY8gH4njuLrX%dK$Zo950Oz2W>2ZO>BYkEq%OdN&;z<}Ouz|e}MW8dg z_&H5uKJb%vm6arCdAh5u8}xWPGW@kB_BJ~wU@~E*>P?fKy{{4sEQ_c1Kx$XK_$Q2m zm?aitaHYG*Q;~b9J=v9$qXSXaf1OEq4(_@7r4~64xJw_cE&ORF`J#e8p|JhD_@||0 zLoWOM=1HsylRWJarrW&!7s}FIsOY^4iJGSCBk|#h%?5yx*|Hx7A*q+jXKqy=m|Pe{ zH1(u?9G|Ie2wbz`vbfVjmTyCZLUjaYhZF>g3<*P>IHfSwn(*kp2jMP zKKlL+6S$-*FaZWC)+Q+R#X+ZOz$CNF4Jdd?tCoHj8*ezoY|J%3Rh5ut!w1frzs|_S zS@nk@yz1`b7zfwda~KE-9^362P(@<#aRPxn^mXgVPG3a#zIm)UL${e~ylu ze)6}GFu*3CQ$t0k$q*X#(a1Gg@{%SKF{&tpQ!#On6<#y=Nuh2HjzpM{^j z8O`VLcqM4)SGcAgHu#51eTL5eF>+!mDoO@9v#Pqs87O&}>rksftCCqMXi!JyKzhA@SkR)TSj4nAIErtwQ;g zo#SRVvG+B!$`+25J`THB9VM8mExkOB$5-58jz%2B_>KCeU_O29JsqxHrXDSOTaM`m zGUt%Jy1amUJ7s)y^}ad_Tw@FrK{>L=s}FckXk%21UX*JAUjF3buIKh_kLd(r*mHye zka>qd12qsnV+rW1E+Oqz$1=}L>#SWuCcqrj+7JL`*{92mKdxg(tL*UEa-=X;R)2hG8g?Zl)#Oz+>{>?7{4}5EaI8 z>R_W01Zm|jM%1%J3xJQ62QDHt7PyH6o5!CAKtJ*5kdPMj7fu*JV=#$Gm~>ZMo-Y3+ z(>O0C%JU(sLZlGreVN)){QIp z3Y9p3wAm_s#FQ`wPpe*LXx*J>a+=+ap3D$SW}meS7XMSD`dg&Qg*O%hd)P=cz2AR3 z<|Q;;qtV$-o%jAPb$x2YJod5X@z{dWPx*p)JhyZ3Ps|J(kmV@pd^iQw8Q1Q!gwzY^ z?~k~5z$PW`6)fTx7cT_=_MZcRr9upYTSFLhwMHtBMGkE}#Km zdTTr4SR?4P)=xR?F_%l0Uf#zcfcX{5To&EGMmfFrEF%ZkG^ekq@a zy$H@S*3;Kl8eK)QHqt`&s$!V9t?TaUjrwuop$zg^VL(uA1oZe}GT2cTt z>zXarAU+@KLFzQhE_Sp_DETuVHyjl^N0$^^ii^TOgm(B&dzOYo>u(yCm5ky`Mo}tr zHtZaoJq9_$H}vl*oe5^OGCOB0g4?UiG*ptWF#IV&S-AXa@yTyLs|qU+42dc!H^w_= zt@|VRD2*JvB-IVUiI4x3-c|AZKVm(i{4{vK+Kc6K zWhP+*rg;U zSL+dVftmZ`mC@I0%Ru56+%Ep^%IOb7EES4xS;>vn?l@XLJQSmaPB-+4Z7`f4?sa#F zCNCaU?ND`IcFx>>0j_t=dk3akVXm6MvpkAZIPARhS-_=9o8n|KcaH3 z`VTCf7+QJ3)_|M;_$q>bva4Ww?ZWDG1!&D_UDcu4A>^^4i!_Uq*t|>+QHlH`W>?-| zlWAt`A)V>qh?f29ZLXR`)cygsT|E9=6NC#?9ugrzIo~2<&{;IaG%aqNS=kl~@5T0E zTW1e?c)XYq;&LUEv@~sd@gzeO@=9Wd-#%AxvTs5gcEw@n)5pNqv^p#24h3DbWeM-i zVmYJ@%L)YtGlqxA)W%-p+Bel-zXs!sn-{W7+$|1w^#i^utL+GzzDMQ zdHAdD+6&SeqdqE~{!sAswKCkYNXkfHbfs%dYgBQkTrN*xYgOFRYEAzvhMuUB{$0s? zyy2wCwEmRe(^H?IUDuVX+euk)x6}7L1h|Igl2n|KP+`etm0Puv)$b#Mr+;w5S-)|! zA@9#r9G|`Y27|#uOQBW2Q-JBB{or9Eip(06j{oPwA)c1@$O~q>VPmd2ht!#>8pl74 zB96Fmre?6j5W4KenF-x0n7o+)KdiT7wa^(jI)Ri!02HEb6?uYfQOua4d-KyCGTts> zDw{y;mXY!Iax9Z*-Hz^p>M=WG1#ySft-jS{CloCjOuMpPfK=8UT>i6V?C<$la;z=Z(laG;wQb?F=lV}^c;{G#vZaMQIG;3J$_jtpd0H<8qLf(<+Zf+ z_Fr6bk2OvQ5!apRl6wZPTHdmU2Q|>@Wc%af4H<7_egjZ>PCQ`$djOVd$U>PxJE49M zIs{RMs?4&Eo8SG)5kWqk5tgZYTl^-Jispr{5A>az(Ty=r^`2v)l3Y)UB(x*_k!Td2 z8>uWYG|sgYM_*lG{g0RT*h2Mr&+W%glht^c6m{XkhE6YnOvg4MMx2s^OQrgeg4q7Q zpj);8{ixiK!OCgejmG5I-dhNFX4dn=dchu7V_H~(%3Iy9XmGLQG3-n)*1}_zm0m6( zQpc)niX)((2kkq8GX%ZSe#B!h_q#G3s_ZQ5Mtm>;eJv z4776L#RV0RmA~wQL>0gqtaF4eH!TX92AO%WV`6B|MuW{09G$*|rs-B3J;JlK&yM=g zcVFDAg=TPCvfn(A0F9mFv*SuvWuv1PO@UrA70t^}c=VANZIC6qV9EKNo&FIK4uswS z(~i6}EaFfjHg!v+TT&i9<&3l-a=7ZEF`l+s?pP35U81dEC4#d+Ne}1{g;mEgetP7# zl3AWczVdbEW%CW7z(G%39RUdJcCtZLI))%I53CH}Amd#g8)oc&VI_X=K0L3BdV`z+ zGgY_ifg)ydl0&HhL_7`5AX&sSZR?FrSDt6c4T|p@PGaAv6$>jIs)bn%Q(_(3&xt`l zx5qAcqYTITp3I-4JPaF%NB%KqmizSI;e8`nBH{^*aCMmMg^nBJC;ns(AVhCM`)YAT z-Im#V3+nd=+)E7Bj&ep`aIvqD{a^i#EUn!v0YtaXKTOiAUjB3|b=#mxgK&$x`9h># zn;lWY7fvNgmc{${7FK;Smkh}a$J9rD&?lAO2Ojv`@9@5WO09KP^$#qwt3H=b1!os==dpySp!c=U-#HfA$t_0-l z2g&GiDI+zqH=@(mah)Y+=He2sw_`_hUCTl%#_7SB@=m!}##}}qc_YmhjEV2&Y(~y= zPFqZ)m2yA(i8&@nRT%1j`!$WBoI}A3-*gAXf~b&fdX{64f1eWcDL5j>%k><-Bi8+@ zWy8wKD2AbAiFN49wxm*h)L9hk^{dS5zTX)Np4e|fGqHbR=A{4Rj8t0z~K(jc8YkX#m?lX(8-@GAO z<37^?$Z(x(pUL#ZcE>&w?Byn%s_vOGx5^70$))fxEowxfg!4^g=938JRG*^MpA@T+ z+I*;-tgMm-Hd9zVIPkKYd3@{()WJ`WKU~#b5{vBPPMwEdbwYE_xV6&%QGH|KzRgdH zNsk&V{?8$5S|9FFEsHI5Hz&*kIk%Dy!pJ#l`M@#jGvSdik@NF+48gbGGAS%iWw9peo>tYvKS!On0F-LK(Aiz{JTJK{FhI<{A9 z`@pn2yXQf=Hxx6c_uwV1|E6VWkzW@IRqH?1j06fJG9`2WNEUiYc`n1Q0zx__^;@!B zx1MmN*0tW59I%)c@#t6bw%iL>UsL6@c=w5dX1*|_f8lEkVpYHJX7Nb{+f4!U62G3U zj&^XR#$=E27dy;C7|>AUL92fB${rLTZ0Phahmy*s)WDkQ6;tTqW`*FNmc1lEGzB=6 zIo&@`2%YTZp>baR_eM!s>WaCS0i%zMFEevmi$o$D#khocyT=IsE|<%b(SDT+2nHYS z66>HyJVp6t$S|5 z2v6Aam3_U&rS#5ioFLeXGzw-cTV

atyK_sD60U=mK;s}+C5jk3N@uK`3!!+T-9Td z!~O42!3g8oyR;L%l-TO3QaE-5(X6K@O4_n9G#ZdEuq1*rIE|!baR{V9>K4D-*vp_@ zwJlW5@I}SxWa(Nw!~So!M>qQn-|+;9Iy>{pu>{o!|0xNq(@6J_k;53T-B8n~nRyMZEZlPzaiXLjJBZ-gxa92+=lTI}nR9yid*#Jd27ft^`6$^4E>~ zy+3rOYvG)!m?i7UhWAm;mNE|3w_|~L5p|*QEriMj5+x2`iYeJ0Rynte-?Fhm>Kri_ zl2&R0qTk8d7RQV>`kKjFtCXL|6)oEBg2}SBOmsJ}-J!@G*+A3M}H5(DLwPom4*&*fLgT8K|z z;2a!EuP+{q((>^b77L!L57abj`ARoYn>Jskrn97G$#}hT-2>4Ez4?;22GM-O-v7Jz z!1v)AzZgj~-`+rSohv^ObF4>0n->cfB68~IjnN?mOAg}-hVM(By+C*?stY-ThnJ6d zM0gTBP_=i$jJa;UdyE>Ui^G&ES#iQ)ca)9jq1#Kc(gcpQ)_7Czrg<-JPp!JqsIbHD z$A8NBapPr5r!>ry*T2PW=(e$vz|XD^pGylAVOi4JrXJ23Ds#(_GYcmRpCX5`WX_)kGs%^bGgCI!mSUgaQaHmKMA?Gr;^&M#9Y5N5R;iZ5sQ2Y z#RCk>P^IlQM8kB)L#3N`I9#%;KTEi@%}ydF+TLx;>K!O4Y`82|u($TswR*%-sm~+` zQyti^COHn`hk6#eX<1H{%c-p_dM&DT>Q%XEZ^BHi}%0#Nrh4 z#Pty{SB4HYf5132So%ZKdP!I(>$!RwZKW-Bs*ir6%AI_K%mG!W!HRAmIhwU#?cJsd zpD2m2))kH9mltskMW+Un?>vZM1=5y}UV43}Kz;9a%%EpjuJ4=}(W@mRlO^}jnA&*3 zNSYv93hTT3I3Us6^7S9v))k@tq+22TW%6m#)}B~h1cGbDR)0{;!_%>^Rxzc|*tvEN zG)FPhHuaaG#G)IJ?#5gIp=W)GmZIq#Hd=zruv&d^-7DRpggBuw8uSrQv`k-6P1X+R z>in;%8q_eUG{J6hl`nm-spqctGYU>v%BK3+N>>ky@?D?=F|-JN?jL!mpFrBKPma0m zN&?|Qb#jN|Ql$cic9yUo#BWVxX90aoXF;tsi?Y%o4zX931-Dqglz0!B+0gVbbjBYV zrlv?tojD_RH;yXr5TJUAPWOWGD6^VF2*p{^=~pfcE~We}!;-~_Q<<~D`{8gIv$*O5 zlraxE>J_{MHP*OEYrQ6Rm2*;FCiCKi`-Y8;U@aaXG{xjoBv|b6z;oCPc@=^w_|+2i zFh!I@9wjGMT;itf3Wc=%c*;`#2`GGXt=5k%d}{*Nk% zY;Hb~(>BNeKd>MBs{tomQe$w}j7;zs=gR@Tx{odI2}NcCLWUne0-TpbMlDmL@pcs@ z!JpCOE;S0a4&F}a2K3q7O`weP z+JCW+MJmb?{@xC%%Im1~@tP;~jUalK{r*Q)S}S#mmG>#1!KS3e{IPsDQ3LPbj14{o z*R2UEom<#+x)oGX)vkSN?`C%R*t}2h-o3>!Og#2ARFC%IHn`77wczGX(RpQ3?5v;L zZ8=m}xcq&mCr0Ch@$#eYK)?AXj+sx$;)A8I5VTb1!rq@!s30flo~qfRm86-8#aOGW z#%9rE2mOJ$IK1wST4XuZ!j1F_B>k`S%f^i=qgCRNFfi||%bZ!KDsx3%kyF@0l^2KU4o4s1c zfunMzq9IxL6d+_-fjCne#ID-Z!mHf-?KfZ=Z3O#{1`xkr91v{|@Pd<|#HeJZFHa~( z;^I;efiG(NQIVDMEU|0 zSOhx500Vz{$v4$&CBSP6WwKt29uQaQ*soPkO&wIW+e%1lRnJb1>EZ%t8x3ZrKknb!KT3FNAWMUissmiqTkn^&T@>9ObYw-#aI1iJ6t!If0iiC z?s)O+=6D?wCE8_Mb$`P#wyogonAmX=XK~Ux0!mGk6Z zO(9~!ylMNCGlGE5dKuW%-VvM?TE=OIPZ6R=AE+JMdQ|f^L6X?NsiKVUU_%{YUESuA z%=w7wTkE858fO8+#9tF{h!1KdfZFL^3C{5Jcc6`hsD~95NSu1Kwtyk(?H*cOgnd%z$Wa;XjyQt`>p4qR`}wQYVtCBlrwVpA zE!|(T-4AU(*v>!_w;)=KuS-xjerQdI-!1AGVbEzD`$3L@--jg^zO8Z}_zaYDhQ~&F zX^xtFL70uSZ0R=5E)B|ev8Q3$uo}h8=K!B6BGr9R0GXgD#D-bm-zd%rBCQw{F7<<< z(?CV^$ovfO@5o28#dd`}ML7LBd52`l7pk}KvdJm@$8WwSO|5V=>y4$u+p;gy z=kmp^x;(ommg=^k_TYFD9w~7iiEbU+0UU{|z^CUo$_{;h49qYW0lj&H894Ux7km2!R)tAAMQ#3Nxio%-yQmVJs_lU7o8a3pv^z`RIr3EbPtYgPOL{2$fG-pO@VRwG}Mjnl__vZyg7 z&3*K7zh-DlM#PxoE}$I1vugPor)yW6Cu{pqR+3=R*-zn7XL?c5jrP=PWAE@d+R*06 z^B;@`dI}o)7*e8u3YTfd?#n!RX?`CCF&k4jTPYR_Z}YcSiA|fssR_h3%lm`v0OPjB zOZ^1^{9$T$_fv=>(&})Qa$4)WAWWjH(v^NTaGIypsZJ}Gi~qetOc zhQHWl_1KBUGA;&Sv?ZsFMX$HHURBhaS0T5R}kS1_KN#GT#`rbluPdQdv%k)Im zr#_)iv4&>>!hf+l7d(9Jnfjm7N5GsuV7#_1`}KUhO;Al z8rTgT#UcA23cTjs+muKwFp@fk7&G!N;1FP71tA9MzP-uM@J{RVwlyfR9pWm123LMs z6Qh5Mbw(w=P)g3M2->}~X(e+-Yn9*edo9cc@=DGhFHEg%N*Gn_?}9}DmzqI$IL7h& zaBTrU!e*wSfu{V2PS(PCF!1|a z%u+64RknqpdyV!Zx)TnKNi^;8o0y6jp$YwzvqM?>e_Cz|IWvp=2{&aDRcNhG^w?rJ>l>|=I5GCPnW7s@MhlU5C|q-QFRDIFAdy+H~=oUO;lsXrC-7#l>; zUY?wH=sf2=_yA8prV>}&wGmeXDWP<>?kFlQWixla`r~JJmY7G=yQ4Tp?7y?i^Vfr{ zgN)Nh-ZT7n$ywItyQPstU)OLdk;Dgo`@70(iytYT4=x&oHkt)BOes}ug2f` zX%fI;ttlYbinu<1ynD&MTnaage>%JO_8JY&*UGBJilW%Bb`REV9c8G^vS>YBNd=_% z?La8D)O1(%v|^?ex4O8J*H!QfbQ=@)gu@7s4oP;v+tIsbRTJLOnF8{m${|?eKDN*; zN&(ad)-FUD1h4Ic?*#!0RHLG!EbVP(an`-k`l-?s(Y>mzhrewCp? z^=6j)jFo3UU|U|zW)@-0oEsuMY~8W`*;xndJbXE%bSKG5U;q;p&xNEL=@$MR3ci;n z4~g4?%@ys6J;tE2T+E}A*D|D`gS+{EQ2M&Nt+>nPx0f&SBO}Y@&f0sxQg!<2(a-WV zzvSbJP<~QR3jS04BZh6PUAh)i(3t$|+W7&jG(b6Dd&`#+Mg0_#S^pGBhEm>oBS2L0 zC?D81>{>$Q1)Q!$OyQGUhlriqVrcy3KAS-Xog+M?j1wSzwBmGkqe$&6Ne8j}kz?_d zaZRlz(yLY)R}fWsgY$*OU~8bMxrMKes;nqq^w$t%+R4yGZd;4#`b0TmzQPGU z_sXglhjfSUA#B=bZq-#2C_r!dpzD>-T~^%@F=1gn`ha7vQNL%ruPPTM8ToL?2-cD$ zvix3^(|GXya1aE3ete|Sa??*IDTg!#3p&@P@Q<*h_lZb=T4qyvO+V!XC=!jD3>&<4 zU7j|_Z%!CtcRmN`On(1n@DIR1e!x-}s7e6?ehL)>etIf^E4u;iDPZ-Tq<=oM-m8db zHOV1_nT`B!2*?Hdg|yPbYn7gBT8vrT+KK*3@G{&@=jv<@v$CIb0J?5+ZVFjI=EH7x zKWS)kul(4-<=HOtLglQu}Uw1!geRj{}Pc8RX zTuz?sSmO_02X`H2gEyPr_m08GdP3t69`yCR{DMN~UikQb%=7o{z$$x&n zdAhpBTp~srPNMR6#P5(&YVmB-Wz{4)P9S}GCsw@0TxHuFB|4F&%`SvT=VWVHfJl}lVMH+PQZ@ynR~X1sf2 zEh<|OYgd}Cn-9D!Wc6A5{6y$rjfOqu_tFRTXd@L!+C_-y2FQ*XQx#EBFnDPvljBa%`{2c>qggdY1`vHO&FjhsU z{FP;xEj6^F6J7r6P6@46d-vfJrtOFFE2wrY5}Xo_eiT1rtuK3t@9j%DW|ufBm-pzD z1X93_*+NWc$ZBb%O(AG0>G22vwe8E&av0f|9@1Q z!(m#vhg|18@iUD-#>IuG?H0$59Is7=ZrDGVFjb^2If^n*4%4QFQEK7*_Wil;x_IBW zb`sRr9RXr?wD$k8_m)v@bzR$NS_-tq+TspH3dOy6fg;5z5Fog_7q}-b7qClZJdP66OX=}Mavg&;x%N?fC5@I0p`*aASN_y%%&d6v&7(hj4XVs1mzpbfh zga=()gX@wmtfL+kXc;abbgeq4uB+(m>$jPszooVV1Vrv|qdrYJ+g#hyATb`-yFE30VvQMm zU8Kj9tGS%VK&QrMaq0{=Lx*(XNli8IRek&`)T_%90wjsHHuw+*fXT_^2ZY4LsF&S4 z^UuF@x${l#%q6{#v{>7o!JKDX-5y%%N>PnfDNi|6ubmBGz!C!}BDPe7&ENNATk@k~ zg(u|{F8~pC)4>ZUctW8msWJqSlI}gd24qXa{e%9=L1<L@4quszV)9I$n6l??Zp~DbR7em`ub`!`OobceNJg!3!6nfKk0Ajzx``7q z(2tm*%7(!Fz>p(Dy>k9+b6xIC^dYc(lTdDcmrj_J(dLp;M7%(go4}3Id|RbAkLngj z<+UM`a@h>`dw2bH?)Edre@~1*tsUh=li6e7Fd2N{I?$`HRhX?9!1B!0-vFIjJkz6w z_aWZ}OCpM>2Eo1V@?$Uq+dROlUx93aDwA@7?3RIN5CBQ4%Yu~&75}{n4Ua=O$7Q`b zU4z!6>a83{!_!Jo2|YVcr15#1-oiGlFlL8J3yb4*O2A=HmzzRH&204+cNV4%{=C<) z71zqik9Ij{dY&>>Q-?eeV8xm6rwoQ7wZ`iXPar-Gntg^K zVhVY4Y^wa}vp|T(T_NN#-~s|_$XTJsnyKJEPwfl5(MhJ#b7*=6O60_@PmEm^=bdIX zGjsT^Z4hjqpYvO2x#*uF!vE@)f^HUL)0u@|`V}v=_s0`Oc29 z*)0|l(c0?*{y}yBcd4+qW{_GU1+#qy!T`XN3jFMWk1eBj(=T7Z$<1JrcKCjw$9_&V zT#tnG9n3kd;KQi_)Gt{hVgw&s>g{9M5$+UEHg1*RNk*d@J`$t?&+cnF@thi=;=DV+ z(`8PRxL=K&g+#A0rpw%i>7$ItW*RYz8x~$lDi-jM#tZ#N%;@L-CZv;Cz7sIf5&WlG zS{SGn?E#NYInxZAn%OoP0#JA4Lgzr8l5vfT*O?&!5nm%JA?rVWb*}x*FaTrgv7LpR zeL}03)b;&if_Eyxz!S4A3*;K0=SpdlkHOq134+n;)}_KkolU(drHDX|n7?it&Q;$d zi8o93jgBOwgXSJUv~18UCJiS%8B_@&7uq&Cf^dIRzpqsW05?_Yh7V7lYtu|8;fW_n zUO_~RNN_^4yEf1Hox-EP^j|$tJjr6d?W+$gi-5Ti!v<@GOYT#lgao@<#u}!{wcEPy z^7LN`IgIMrnrt-GE3>8L88RL6;C3kYzF#x{4nCbGR!u_S!m*oerLQs%`EfJ5epT36 z+atk=C(-wiHC}m^?VzonGMj2n6LRR0+j|xle~y2(gg6Fab~*`H1+$E_97xu_Z@v9f!+~o|Luk6>UQ<@K0))X+hWHQHpoU_SYTy2tnC<++bq-8I zki~_}g@o@7hN-14AkJHGoiaJC$eseo%Edf@Y&6N8QQb2z_1-u&WrH1&0Jx2rj#y81 zRrxh8+@?&alnE$=_uuIK8d7pAFPXaoVDGyUnF24UXHfY#3CRSn)a@DvyWAhw)P>} zIZi<>_MNdd)uK^22B^zFFgw)Df+{Z;WFBQ1EJnTYr8f-Pf-|`FpK484YZgoZ2BRfT zl`&WGLu{^g?x(G0x=p4WfzxHw)@<_q!5&;bYOtjX8M|GL6M% zU+%eN4`GdzXAlo3_4hZoLzB*m-vhHnOCMx5%f-m0y@ll&q0f}7fcySbSb!7z0c?l* zjJhD2=2s<^qD{9BL5x&A>~&Pjg0SRANM zl7{+@glpiuT`V9+b}x3WK=7pI;ECt4wv~AHyejI85vtG8j^Lxz3J*7Q=G0iJj{6e0 zr#GcKlkDIvvyG^w=yToo5r~-bB#bvTT>q6C%bu=`TwxjV47iqqE9qYJIZeJLu$%uXGbl@ zOk9DSKo|xY(&x6O;&!GjD~ks4Pk*M)PLiShd`}=prq0tPKuTR@G#(%YsO|HWG$b(PauMKKH#8~D{Ph>?mP`tnKU5ce|C zB&NahVs;}iKj*g1k_#k+ulPAd9>JSlk&6HDnG3LPr|Iz}V3R~&<70n3p4#v?3@-^Q zV??Nt^KYD%CpT>wz$gtpP|@T~c;NknZ3ugVmo)1_{bvNzrWa!&Y2 zgfkzUYXi?q-v_#56z`#Xc2|0)hjmY#bxr*6?6j60PCSQFm?Qp^C2BPv8tPa5S&87+ zLtCFn$^c*;rEC$QxZ5n!_f@rp>VN^@+MBpER@OQ_boY%VW<13C0raD^wA3-^BNM?t zaYVBUS7PI+%E(55X}>;SE*vc+?2GsX9J=8owxa_Lq^+*70dg%TKu$5O)5Kh?q?k7a zuN*S&a#GJxjS5{4TUJIiYbk=%QT<>$8Xb?B1@ZrK-%Kg6NF|zRA5J$FDJ+I_a2pg-9 zEh^;CJ3|lY-*$lIrh(4h-!B{M?c_tIssSd8T1EIC|>+vz=kFzhE|k*}Q)_q@tAKgxdflfnDC6dj%ST7b&rN#62-l8I)r~Lr_Jl;Mua)XrZwHk0n8! z6}1DR$~8+dl|rv7$4rZb`UIqbCTH780(-|HH6IYwj{UUnej5NS0|K{pFp>W#gRe{Y z*Whytdsau~$;!^$k2L`9Q75k9KRNF-|7A9~I`;rkX;|^owLj6ioXl_i6b>MqVc`wn zIJ56ox^$&c_ZDLY;sKuf{iG7( zOq)9093snpwHsPs2*yBqG^HMujOqSk6wlmfx6+$)2}zO#GS|I6jK{WyIoZL7G-TP2=Sl zscE}p6y0LyZnDTgHsDcqp$V1#d!Tc_Xrl}q)z_7@Y}^BXX9Vnd;O9@~DZjd3sZ3wy zgw*~pPPtKg9ldU=86Qd++~ht@92=%%a3!ZzeXTZ#f7La?tPF&=fl$#R2?jkd*Tc6< z0i7i4AoEvC4lQMBBRGQgjG3=#$qd}L)#ITyI^($#(1mhthl(V(59lOi*i~%_WyEJM zv>(}wyYAz06_K^g%Ia4MgSv=HEiFbyk zvn7WPCbN+Et<-I1>p~`5fvM_V&H&p3To6#Rj5Ph6_oPl~xl@M8Bm!8(05cKDZMTxn zTfB>45uBnTrJ^oP0gBcYp7%~{rCsvAaDW@Z$DB*(&1dm7C4^ozO9D8n+8hijZC&Mo zo~E7IjeCtxp7EDG0^%7N`_$d-CVE$@8WUXxDH`#)yiL{xFDijNG{#3){bxj~F=mxn zgeGIqjL!$x&HwV*FVVD)@Y1xbPoXqJZpA(Hn?KK4V%7`zp``>#3R}!Vb)^X0YGcC1 zNsB*78~gSU&i&Ihn0|Ou`0Coq>lVeXYcLiILN)euOc=(@?6)!pW|w@W<>lC&SlivK zGgk4<-5yX|&OH7$2Q=nGfb_4=n?R_Mr*{r!9w2P-XUE^OvumQC#avy?e3wCj7PIT9 zYqirprm{KCkmXSnf+`FxCyd>nDJ?3CJ;RV5`+Z?egI3Av{sGnQ$aOA|16EQ2NAMQ{ zC>xS~xAcvZ`hgihv7yT9D(;uP?^!kUB93>ZYPQJvu}5Qa;>8)~aIOcnB#tozKAP=I z5$~SH-i&K@->-w0|Y{smo5tQ-Muzr5t@Bea)4$g+$2~37q@CUiM7!tS^+Wd zB*3SDn1f{v0YIF%aCM2%zmVDwDR}G^?%+5vuzy&(8CX&hAe#G;dD~A6+l}bsK0~&G zv)7(yxeIKi!sus!a8^lBNS8$90Yn*YZ~e8|OiLU+W1%A7%FyBugAGQ=Y;mrOK1BBM z3sB&CsOSggQf=FhvWlO19`%PXUzfgYvohr3s!+!~VOUFx$)Z(b1u(j7^Im6QI=7aM zEA>;(KVoQMgOADnH>&`x4`$)5D4>a}M}qvDs0WNwv_du53S+Ve)Esm^`&jkOE#x&B z@U2-iDe9^NWnR(^=;+~1V!spQIH-_vl_J8U)%uHj{6@@rRl57PH0G|zby)-v*?`|& z33Gh18oCu#2^7$#P$}1&l5li?S=P`oRO9Vho$xn&EVn#5EKHkLpS=AI!&B^XEsW#? zp@FWu{yl<8q|l^YyDj6nFa!f!5A1g=){2Z{Mrxs{=U4rrbcsSI}d29^-<;hcsy;2+gN!Ig%bhJfkfy9egJ8BVS(rq`)WwpvbyZ5Vl5{HY4P!6&%@ z6vaT0+f3q&d zw+MvToiWCOKnDh}Ekk~-%Jhq*kSi{d@pI%8cM+^LW}ybQiB4;kF?kPgILGI?Rg1^^ z8m-sue~PIhhnzwVT>I{#LRA+sqK)~fUEAy@Uyn+74?{5fRHLW(YdSlXx?SDm@#6IG zFxpr4(_PmPiw10CHPW;7YyrlkbiGIEZk|Te{kSJn4chaHOE75IDYUzNIUU@!+^FPF zH7=BOUHx6}RR+^U5-)=319D11a3xQnm&o{p0Gt1r=4Sp(m6oZ55P-I;*(G8Gd71uu zd9)89BqsbNee%cB3ZX`*%MX~j8Sl*1MSv-c;;s7Kn$$li7^2MSt~RFbfggX!$xNFp zc^=TYK87;suZeH*$0B)|OOwfS(bqo0vnkK4eV5_6#uczKR<~~VO_k?XoHG!6-T8W^ z18{>u^XsKv%rk2h1T36H0FLBpj9dUj3M-5C0S8)QmReB6Gl7=( z&RCC7Zi^*$x~?U#b~_Jn)kX?zFAQK*mL8~n#0lwuQ3Bh|X&Obvb%=gZOuo6&EEDaB zvFVY}Ulpon4ltTsX{(&2dxWeN&fd*Qcd>Fh*+pVu8@#pXIXr#RiDv}OVA8xlsx;T? ztw^)xo4n1@6g1zmO!LF=i|0Oz0!ZJ7hei&*O@+CUg#~&wbmgV@ohhU(8)1f^`_Xox zw4ITxAC8@93@MiNf!jsdB1nsX(%>#0MRLf$R7M7JXi%vkgD`uSb9wGl54=#F`&MW~ zh~Zx?)hG8G0aR3I+fQ-3Hi$j&iPh-$7>ExMA=ezAA*?VK9QRshsLm)b={cq<3(+J_ z(|k2ZNrq0DX@d|f$yN;V^(!;AJKYVH23O=;xThvDu+rxcK{Q1&5L7OdP+hprd;-Sw z0=?M`EH1`>y{G%GcR9&mbDSmC`0XJV09G6$t7;CLCe1VZGb6em>qlUf0V&OOrM^m+K+Vxc6SoThPf#rG@FGrKPVf)PPm zf#JV{JTBk$1O{D1e*+e_lt2q}?HwW0AHu!bFG;Wp*#+>Z&3R3QDb2f>`M(s!=r$6z z@;}sI7ZtZMl576!M5hz;1oTgB>jFA5EmVQNt|>IUG-_A2=kP$6TftmcBJ4aM2MtvZ z$VIx=Vo7(oc3&`B5dLK#{JdweHV<^Kg_|k%opp}Xr8|s-(4Lp_&XOnB4&)zC)Xo)= zRa!3I^=}e+fWy#9ef1bu4BI+pNK zbed0fxF@fu&Tzo7W~OkRgQ)tJTPWIZ69_p~tOPLtEOneo@lBzIOxmqBjnS(r$_w!f zC0a;fx1uv$jbF4yw-fUE$ZsF^H^wi68EEFDtI|=#HzSJ@Ci)03-1^vzHi`I{-O`{X zgEkJCmtzgO(+-bb>3{XpYu#0EL=M{{iwoECh9g$aS(pHAVAppjy7pIRBP)8?;&-g% zT(bi$P%BZHGpJ>v80%IdUt@Y+%6@Sk_g$!}!Z-bA1>m$~L8g3X1W=eO%~d$v!CaNs z)5AJ@>Myhb7LYNi2axDml}u_y&THrL^bofD`UGNa)8x_^x*;_}#%Slb?|A2RX5-!4 zXwMS?SEjn{I1b&^9O5SN%2`KU-U#uu0a@@<1mHKZQ%9edq9@WfZgi3>?$fyPWyyR~=-DkNL5Ck-$UN2YOa4&MMSr8u7`HC5L_wGXpRqhWTJuu1)UvpNl zmXnq>Ka(1TQA|s6o8J|!+|u4=RNq2YtfQC_9iQeI``2{E*P=Z)#Ah;-95q5^4q?1_ zu?v$Kt`h?4-!znX%#hbgPrI$#P~hV?x=XX5Ks&{rSk zXprtgG=Mn=T`Ybl`Rb#I9ux0Cf{?n%;0Wk#KVi;oKT#l~MRIdA>0btko13Sjd1-gP z_HKMWQV>GHrMRq^`RxbL+*N*8bNoyKVVsY6*9Lxir{&I;OP+zw{a9tbWdk^z;pu2V z&PXFTQgoEKNef*IHP4U(^@8brV$elEo(9oaxoKhP3sRQkF&Ldi9zt}DT~Ofyxx+|z z5SVSfqaXQwoKFAIN4tyu=el%#)eVCT^6tT&z$b9`Z!t56(3TLRp`JsDP``Ic0M1jf z0r?VMnw7rUq^W!ttc*D%--neYdfog1)X4mfT9VlP)BP?{&FDEx+{V}gi01($eR?*@ z-Z0c=J{Ej?)-l>OEZPP{4|fQD;+Chg-Sx{?HV1UHaVaF?_nasJO|=+x=7^nH5;i_l z-*Rcpdrt3qmirU#&+glr)GNOMGNTdt_bdOK zDkA-&g*52bW-&U%e*+GC({KY0e>QZ=L=wmQ0MZVXq><(aM<`Ii&BiL6p;QHE_`jDr z!!#j$qRL%wtV$E{%Yzb0Xw?G+f`lsGqD$1SXl%)aH%XK?kS?uM3>#K z)tz9PBp_%NRgO5;QWdbj%Gp^kR*~Iz z1)B~89iXYsktA@YkV%lQ!!k(GC38wqH{SPJlFu6uFC~T9oj|+qy8m%(Q7ovD4`_D@ zxM~Z9lh+rwWOys9VDqLHa52fx3eCG4os~{bV^=a%Aj7c-amALI}yez#OoL zOU)iYn6=l^?U(lLV`^`;)s4E~`f|j=(*~`r!;*W=3DV0SFPIp!Wch_pVJE2oG`AmS#pYpdk{h)7k9issknYIN~w*kTkb#G3|OltLz98B==%?#vEKwq^ADiB z?-K+2e)1tz44&Q~7ytslY z%5ybm8eP&`8#RUv@xEg>+^FG;Ff^fs^c*^j^vLhSpzLOmaJ`xA$D@H3nC$?E_{SkG z-Qtf!rxXI9cif@g7q0*FEZVuTl*m@I6v7$h4#5k1{i_mK{jVe7p{+o4jVvxt`}zIG z>EE=?I7Ea`GH#R@|KcV~j8PKJzLw7Cr8^slavKu2Bhb0@N=GfJB~Y!8I=*DcUQ3%w zxE2C6BpT+_xZnIoZTR1l|F{galsXMik*`sQ6r~;#RJ^Y`e4F;mOXkuWe0jCo9Ogn! zqVXm_&;L5igHgCRxbxTNXtA#d^K?>*{5)K4fq7eQH1zn%C^A3$g7 z%Cms^J$mcwT6XK}=JS)u!CSG}LAMWoSnf*n`qO=8bDYu++O{fLW^>Hr4la52wYcQp zX74P|Ghj3XAmY%v3w5|8OMPJO(>Ic(u5dE=nu@nE!ihY|bKo9+YOavyg}gOZ$C;sU zP)AMv`Jzl*|LkU(qQlGk>;crRC)5HkfB#Vdn(Q>`-|_(cbf(DLC;oZ#li~rS6Bj7v zKY0UX=M>5dlnOd2lMX#8lZ~9H<89761vLjY|4D~Z zTH&a`CJO8NwCr=}ox2)!=Y11HG?2yAiR)$aDVJVPY?;n;kk#ECtD|K( z+t!M~|3`BA1`r1}QjeR@k^jAse%zBWEx*u8T-H6P<8<$Ca&XxZ-2F>lcu}62F#zn@ z-VW-DCm^`I$ZQ8D5vt@#p_@A`9QXqh_s%Ri6W-2H^LMxV7f`c2K=}*q#-w`~-A(c0 z2*V#h_1%{i_~$JTpwo)`H@UaV!0gMpWswpT{%40*T6o%`{n9(t-N&g%@!odP0Y^CH zLMs4k^KP0_?5gGe2TzV0^~Db6l{8w1HwgUGmR^9AOPk7e@{1I;rE{T!l?wJ3zsL56 zKN(fr*!g(s)AqXT*Wlvu+bd2g;z@Nmrr{Gs49hY!G^AuqZ$i=1{)?Ir&;-zbVJ?8P zrSUIQ3$Q7w{$u<8D(uBbd_lHycY<6~8`oofXnnyfX6s>F=RoW+@74cverf(dD{$$2 z*h-+oluLv5*JZ=fJv-r3 z8?)5i>8)tp0qX~l1H_48R6wx4l>F3c(atGFsCnDOHG+GWMkqe6s;M)-n~$A%SCy9; zGOZ*rz13k)PBejBGduhMl0Zup`?y8-MO<-~malA05a`7vMrEeKQ0>p3YqcY5x#-nu z;CyEc@Xd7rd~HMD<>BRa11yG&duXE*?y=@_Wpj|m_fp-^s8eigZ%W9GR(G? z`i~wes=EK;tzc3hTP}o)G2vE5c1*wGh0Uq7_gOP!(1r@@n(>@lB7vie7^j_Tggg6r zc%WK+zXz@s&T*o{cjl*ND3x#t3DH9R@Op)TH9YD2w>i|^*kwK5JA+>)xHTAIzGp$B zM)e$;90D#^Op@=6i?Kzw1EyMf9r_60D-b1%Zn_nGDT9_2$0bm+b#;;B8r!a3Q8Q!e zf2Bqg(Va{w#&gK-W+%;XcU+^#$#ueOwt_+}JaM^RDUV@?_ghnnOLZY(6FdI~%%ULx zqd+MpJ!4ef$Kkn1O~$rAQw{iYFXvGjDpmX*Kx%s_8aQ7>(L&K3!fB(*1ESs*Sjk&k zEBeW<(6*UPMnZaij@P{7t|=6uwIe~!0So$^usv3EL1U*Y+XkE|0_+A_?-F8Fu-OKN zYNn(;l_&CzI$CI}S|I`{a^IlUNLa&N)39q?dAW@DBpJ8@-%V0{U8jfV!f{^=O-eC& z1PZ~qNL1{VMfXXMoaz3O8Vw*xFF6?tp#Sh1J#$$&hPJ8sBW_k_haqtfpb7sl{TZ-W z-@&oP+ABGm1zg#RaiQvWs>J8wrFkL=M2kCZS;;@ucT0FxKa-SizYZB37wsXe^c%jy zzs#W<%;V&+buB&7G$&slty4l~yy-KSz^*VDhYLJ_Zcp;y`vj&8Pw>uaCg%G5PvV0r zhY28(d`6Q#9k`sQzfy^2um}!$moGIGE}g4tWV3MCUaTftX`2xE+<1rlqES-ojd8G1 zs8C=e5Ry!{^zzQT#W z@~3niWleNujjKaG>n|mhx~8in7oAp$DUXdD8Kd7>nyV_#pA=Ii?E!%F!{pm?Di+f$ zR8;4}Bpqh8wmj^sNjWY@=0O@(qs)E0qVM~T+cUqXWF`l`Prf?N={G_4j7O>L*1XL* ztE-BaEm&5mzTM|{im3P%C-@9z*JBlJE}r<_;)TdpMeV9~{WdOtMLeHgT!bpTo!Z#Y zM+iA9ayqhf37+n(mdiA4XpHGnXpmdhn{8HaNx< zRuED}UF*HKr1v>LzPHi^$jzj|V^wu)*@VX2th;;spd^`O$!KsDhnQ`a zFx@(1NmG@-)Ihh4s`ouVSMj^aP`G?RpiSeXX?T+hGD6UQtZNr?%MBV8lk$MzF$`LeQwA876iw9 z+P8}O<*U)Po-DIy?-#}#9`QXipPkcc)?irKXwEfn)76Bk+F(Fgxl;jkwrif~J7gqG(TJDN-GX8HbmORbtaqVAqaawIvgZW1GguTa{H zj0w_(q^?*xQgmq;w7)FMKY(stKXibYUOfRleS-A#84A)9;12`>JpsL@WD`-+w}0gu zl|{i?**1FgarIbH&+gm99OyY9@u_}P*)}D*FHPa*tFqxb$fX6la7Ekg$KMGh0pGRsxV=zcb z{uyFf_97cz?*D_~b?&$QMk|--i?ux=vi5uIN;2NBNZu#$pF?`ZNPk;?BX|o>!=rm| zzOohzhI6>{yXx6CJ$a_~sWY+diEQ9T$$UeBzjopuk)k^42I4jXZ+Sj;H2$09GHT!1Rr&&3@9bEjQ(DZMl2U?CnQerXz ze)h)7j4G3NiNBd!vRBBf-GI(cQDTh=)E~d$vh)GkPP}pYBI)`58^navYxYjeA{kQi zbC?C5Z3R%>_Ub47?#E=E&z5T6@)>T7N8X;qq)keS!66Y93T!*mkq30o-!!#+M-&

Nb*4Dxdo5wjb~QC-z(H!#L{zHGHPX?mq&o_-D6SFof$Zl*o!flPSTO$xt&A5O-6lzm`7PEN)aW=Ir2L@|}FgMogXI6x>% z{=Uu2z>nOh!To6FGq*|Dd@ksb>)?g{QfOjL_ej;9YG^J^d2o$#D@#=fw^c}BsF zTp7W*4jPNZULko{Wg`Xh@h47ExG@yJTike4Cd|*$sfVwffU9-hxYV3aRQS0?>Ljbe z?A_M>J>ncFw73y&ylYwVUOr^1FFR=@s!@@`5o^l9xFXFm58Q96+~wG*`7qb3pHq$F z88&dXI<^>Q?u~Zov>ZTAh@-0HpIMieTHmmcuZUn+Yi}_kx zg||S9AYwduKn&OR#1szT!@b(7fR0S0h=C7!h+B=svLNYUoyb5-7&94*Yyz|cg^T6!W zg|zRg1p`)FAidfL-*MILG13~pDXsmDAGj##R=QlGsq5nT6FMe2bgo6GTJKBbC!XWf zev|g|U)%l@K768l8w;S!g?kkq_g;#P&=>B;4s?L{SkdKvNPbZmXpd7U zY*}yG6W(Z;5du7_*CA1tQw6bs&3NSJ?}=xMJVe;>T0JBZIpc+4dW~kgq!-CWM|$2T znf~rs!gzG^(|_ii=)C+?W59<7#`)j9!}fjO-rudTg<9El=Z;c_ZjZk{ffC!i=VoQ_ zKz8rAhc`@m3}#r5v1~It1oP^{$Ui-*I0WfTMSZO7lTyEv1m*s%YYYvd;+5||3tw|G zcQrup4uCw6tXwH;YiZj5KmI5%579&0b$r0zgJWG{Qcc=yFTy@Br%-c`ZqiscGtTlIY= z8icRy36Zf9mH_t2z{@}=<~MJuA&!CQoGPCR8rS{+f{$vlt#N;2J{3-{j2;Bc(+wSE z98mw(-z8(?!zQDzd5cB;R?b=Eh>M-eo#QqR0=HCo02xBo|B^NZE5@TF-aqGk{>mCG zug1Ds$T$!vsoIKP&wLh`oj7-hVSLiA%+-)xW=VOX^g=mCB9@;wltsg*Gwu- zGhX#oW|{Wx!l+=>SkIrPC_QY~U}MCjTQ43xjbYe<}W)V;{Ht?8^kpz>&L# z*6EXz%%5fT?9@=JoeK2?)=xS=>Dgah&YzpU#aB#8teE)m4NhzyydEj?mQP9MbE_ti z-JdzbZxhp#zV&9Qv?uE0}`W$HE3mnjjlQFAsK@Qew&W^}WZNHv)LHAwzv#|Yi!-!-L< zoM6c}5i}&O&9udPU*r7!bL_%RUz4CA>FGAYXMWAk&`~J7mb3P25Tl#3^>~R$c&&2O zkh|z|S-e3a{UO$#?(ric@wtpJ=)yfDZt^~!F!1*ayyg@wWuUjS`dS=p@%@8#Nv=ZA z_#X)6TTkvJR<)j@FL}~&YvgEOLXib_)W1t*m4yw__ES4`i%qE?ZHVhgjio64TIFxs zB%E!vI!!cSj{B+k_w0!4lDlQAUbVV+=35|ZoQo;LpisqK=E~rc*^!|}9lmqc-(8UP zlN749{vgW0@SJ?J0v@nVZiN1d8pXipgEY^Aw}M32PiXlBS1}_P8Q1L=v|cqO{|U54 zmWdcC&(kRX`VQmA680vz^SW@iv|`llf~q7}6HS=4hm~ypLe8Y%;A`s1{qV*=7S`PM z@$#+R^eWh=e^)e4aza4O99!q&gx`c%n?s|nC-x_+N?GEY#Dsym#$SAuSG)U9;^xDi zmPX`u`!1o~UcLYNH3h@;MIBFeP_jYVWkeb!d%_BvIxGL-S-W}i(DF0T3!RZ~X@bv> z-^9|YsmIkET5g(Ah)EW}@`f6BBXfm0@DmNXCXje;N6DTekS?bG;8=t;Z<}zsV7bS`$Tx4~mUXiEoVn_BY#-!2&PC zvmEqfoAl87ln=ojRJ|cjw%-dGJ>9RV3ehfHH%D8P``E{4S#ZRxevKoCOsO9!Uiq$h zGFRw$oPNewSa^7D(aobknZ+EG?mc&o9ZR*-G-WoEt6PT}>>umws%c;WZi@5CLO^h+s+jB@&=>P8gE z%9+Nh3bA9&>p4Y|J%TN3Id<&O24W8TDOT%}$rGH|uKL#r9yUmyS)bl9b2h^j1;*`x z3fL-?=pcEnNWP%2E_!_-q*a(Vd#(KM+ww0vq+6vV{Tt_;z^42MWAQbMm7ze3#QV7G6*kz=krjZ$p@~_eex176D2&gQ6!yoz1amiux z>}V|RKDtf~88VOGTB7ozEZ^tX?J!_UCfzS1gy-)~Z+r$GK(<%kz6@gQbO#*Y1NB4G zBa?Ly+R@hZGFa)JY5{= zAF<+|o9ZZ+Z_Gwej?I2O<15|%I@g`5+ZiB=Hj;)_e!)ksd+=4LW#O-vPu(c6{yJuz z-EXjEt$H#k-@ws2Ms&G@O-?=(eR%-c1k~ira0g^o-BiC>GPuqzK?(a^8bG95-YqM) zVVUpUf)Lid$|%1bp1wETeE_lD`7G&LYVGu{7VI{x28m2>20KY_F(wda?d|h1<1`wC zP~Zf}O*=#EJ~>#&m#Dg5lD`@UP{>=#gasXw+Fi!yjOqStN|HXZ zUp)N^rxls_r|6U>A)K@*``cDN=uR$t03n2l}qWG*0JlLEDHJA&kE zTx}-z)3S(@A>hir!f;&O6EcM9v}dyR2{1jqJkXY*s?;xMF3|9kp(m58Tx`dT8(6sd zuE0&Di-p&C7R+X*9JIAlX4q$-%t?{EAI;=H@j81TsU%^0jxp>@r6GQTRnK{1r5`&% zKl-Z>;u^KxTMz?p1KnNzo3EdUM!2=$tc^%vCicNbeO|I;uy^#>RAaomih~E0Lh9^U zOp@h;o*&aE^6Sqd1oZvv<(btbei#rzJA)|T$W62TI8$o6d|@r5VR6^>Dt2dQDEsrN zH?HJF-H2-lD~(&zpF<^#L-*>^O4>cs^8(y)S*HQ%gVfO8ANaD;BRg8lzseZ4Um#VZ z8|V#ZXtj@w*0Fl&f-R{9aTQM}2F4*K`W-tR==A;^UIxZfKS5aT=jFrB7RboeaXyV=V+u-S8 zYAsOA%I!~xSxV@Q^??-QvvLmM#^5-A*2;_a+)RqS`6;qqdA>o==JXPJbLteK&S}5T zOhnEO`#xsbNnO(xOuA#69)A~=Dp#J*^(SSLht$4xLartv`--Z}%go8X!?~UB=^%du zM#nK89#nF!@EuA19)3vaeu`HXTC)lAIc&M^`?@pd^Dw6K$ZeAtvxo*T^o#&U$xqKa z4$biMFibG#$!$SX(Cs@Xb36bo$DXOy!}S3AzH$+30q8&FV4AQV6W?n<0Fub9fX3(SyP@DgRYh(Yz%h|iDqBsm(q*O8_ z*xY9}tE@5~1xPpRp9(ky$g<}>Wg|vs?|xQ`foo+WW^c6rypJruK&V01!;A=e^6W>O z1H-^nZ7dQ9OX?USMuZgkd7xt?)uE%3q6;Yc2+5-G zc^pvxd=sVQ*tjl*v)|&^kU`4IUq4?QB!JPyB{#*ZbGSY#K7dpoKp+ET`FOFzHxyMY z*x$;d6XDLR~j-M%8;YFvgSzLulP{9Eip z?Hev)V_bRlq5-<8nOR23(DN|NSR60m70l+pdI+zrseE0%s6*x;?=1J9WYnsra+C|* zbLiRQh1zG9tmkI$pv8;s4cmOo@AG*lYdQ)>zq3k|M;oMcD+{Z!&tlQRn6E4ct|e`^ zlYX;XJFvvzb#O+Pevvk#OX^s_FyAZtY0f!X$}jtoZ8c`_E)yPIBjzT{w2B#_C0i!!!cnJPps|9scYHvBDtv`xT$B=SLA_^ZCm7X{2u^{Sur4O}ds zw2%?j$OFZjj29Kh&aVM3_nnoDLz%m6pC*=zy!+FZrMLwbuQXS-NYfz$BBN#LPQgQy zTgRSG1+Cd_9=~fuOni*W;)U2TONw*vz+LyTu>GyVg3oJGnZa9W^}Doul)L4Q&(q02 zvR__{(nn?`us&t?^;F{Qnl-#whFRz#shvE7YdFv8XmO##-95UCDW~WrE=h39mI*+h zr*qH7BS8hF(j)5f=S7aZN&94zGC$GE&5mC;6O0e~a=tvOez*Ht#6tR)$=LWGrRts2 zD}}Sw`&8lSpz9+5=|R5JCXz`Uqkx%^&>&k&YqLIz<@CEjdO({hT_fIF@vzr^%=W!G zdlO9(DI=0{xPC$J=q(qmAXq&}Ea8n+b^jb#Y2#e0pSpvcl?6*GGPf@CZ&lM6OjrMO ziVF8m4J)lpZ{`h{Y1f0m*NSK1u+Leb&X3A$ZA~%fx4Gkm$-4!s-W!@rr^MKr`Et%x z>kIKu@=kAhnZo6B(H9+#@|W65W|Hg(M_37Uj#9KyWSsoDdOoeumVqMe&?Mwoxt2-D z#Y+Stqge^*@Tw3|{uhn>f6l@-lNbGKLGz?C=PnOQfk?Q~30|=HRqf)K=TCEXN!shca@ukW(fMyvG9r~py z5^*82G**d`yfi58Cow-_MA0CDRM0bFkFz`A>yoVU8$w1f-|3_kHFBv{yvRpMmwRm< zNE&DuKV^vgi7Ox>$AR7g>lqjp_SH8XjKQ&%Ke{sLzd<3cxFZ({Q}!?|5kXV_h^JtP zoylTqny(S(5L92W5|72QP;n&5ONuFD#e0SOUZv; z(;EFz+KNnYJmTQv_rpBnGO(qa32!BISF0e4(RxQIA1iLHN$;b#S_@r?F`YoYgv(&o zPNMX^|GOl@OoL>PWD&sA_(4iI6HtkeL$5OuKBYR(2}4$ok8p~ zzNzi_tD|9wh5Iq^#u|Hdo>IjKAgYNhE>ML%|B8dWBHi``@ccrxx z&9hiF(|D-7_Pmh*N+!22l1#>Cc#LjC`^IDxW@68#C-z{Y{PJgUQtcJ?C&e8k3(J$K z?UZ7o#42f(c>Kx>T?#%Z7Dajm=4jd(wrCiWUhQn8#%F4mEHsPWJlifFl5*W01Y`QR zq{u%>v)PKtBWHXlXY^xG=X1w|$i*&2rimpKuEKShQ?XF1%xJGduKO?Y6JJbADLa;& z1*eWXag6vO=eEn|Q}8D)p1q|9|1SX9KqkLFkRZ*d#@7jnxhYd8vmcb0Fa*gL4AwjF zwxZKWtRH0pA!nM<73czPgJb26T?s7Y$|KP$7$LTeZ^WvfWzsRdv0}svaafp2Pu&Fs z%#DlA0k6D>7As&VCfmA6q~!(32m!G;nfH?X1+zRUvLIz3obEmq!qSYT2V8}8@UgyF z7kSq^wO<6iD3fX9+i&C>tK!-MqW?Wed zcGv`MDLn5Q>#QC``}`TDq*YT}br8UB@{gs9ui&T>5xR%o4zQKWHKaDo(QLkD%E^ZQ zbCH$_!74edZp^=qw5a112F|sAzyobp(55aqs^_IUsPEyVajdlq^0`1LJzH{vNNxZQ zmM;A+Qy?+10F)fY4$K~p@FfD=5uZk_%Z;xJe)CtXAJmpXowD?fiP6n`^_#}Ey*g_K zpt5iiPE=@~tX*DPFL}MXaO|lT5aHiTo$F|(6=|s!w=&@P8bHJdi{I)?2k1=*4d!`r z?Jich%uo&4bT5k1F!rRH@Sy|EEO?ei5Y_`#h0<>IW5Pl&52?^$;&&v%-W#=-w^iL|M0c4c<(5;C{m>EJ_O z%XeZhg8a#^vEbQRBd|+HTACw4;(Yr%XtHVG*)<=eDop~vg7$Z#v2#Dm#o>;wJ8wTD z*m%3G0|oAGPw&bSe?pXK}+`e;T@#=Kd8`Rp85Y8m@dWW|>?mpuTNT z8@zE-RamT8py8m=JkCPtBATv_(};Q_zupy_Q(sk=7lZ1;SsX|#9^_i*8I`$btU`{g zR*?lY0Cm5!G6WL9S1l#0^kyIp;;?(Xb20NtLj!x&iVE0rqjU^9sY>^5Z1kJKkyKRY zlQ5Tq+X+AbM-X}l?;G~U;Dx+sSRr98H#2kxLjqKH-HrentWk5C8WO%uL9c)rX)-Ik za=On1*qZaaShRIrU;qpRS}s_)TM*~j4XEUB9vTHc)|ww?2z5(+XTg&LE6{*)w6+%b zx*F0vQ}{=fPvwhkO4l%gTD30N&ON?Iu=$Sh6z;OMjs|6^?Ee6$gjycA&2`piI-5Af z2d-rRT7joBt!K}^s%*kXxmxo^dK6IRD_e0Y&!gmD?@saXR5PS`(;FaVBXz^p9>7PG zR-J0$0J5sABI_{@ZgA}>D(5UKS&C_*&9>in=O0TEW@d^NblX0YHrYJeR@XNqIVTav zJ*zHr+Tbh{Z=tSPM|=ZTp?%;LVQZ?#@Df05v^2%Rp`!XlAEHz0<%^Rv*AQk`4yyjd z&>aJgc!@BX!LqfV>m!)X|Dhdyip%fP2gz_*w2%hPIfO%B8dJTe|YM`EQCzCjO zfVY_ydAh5kQnk-5;CsqvO9lDF0GtZ1B|+(I*{^V^z6#wu;i9Jl#L}9;N3JN=6nT+e z`H|Zd|D{oKd=&M_GAEZysu>bkKpIHQ+w!-|Bm#=Tq~0IW+*33j`mLqZA-yeW1i)V<|f<^{UE7%B@`g1!E!I_cx? z(qdoMS$Lz=XTTA3BPW|1rT!nmc+yjC@?~{3cINPc(6xHZvF-9bhwz=2CSW-bdY*np zl-5_4yMHMIF?!GC8Tmq3y;UlQ^g!I z4ZAK8$pp9#qVP4!x+%U0_d09hp>xry+xtSRhVH$5e#FdQ;F@@4diDh`6F1$L=lD-+ zzE}yyQh$S#9WK@x#_r!E*nb7y*Qz?{%mYWr(2-;00ja6qn}_iCJs-(FHe=2bFErnM zYAsNm0rZw>_i8xZ7cA)((e|1?h(q2D)oOH2_<<$keIXYC^oC=n9j>LzBKw*7N=?zv zA*Eo>wT&9{M8QznELG@4?8HV3<*d+~;EL*ToU*JDn267RF0o%Uh0UVoCLZU4GJfJh z2hY|pDcw|4q;@c7yHY-oA5INo1Fh9Wu5FQoK|^P01s*UEgvI5o&;^TDbd)7_ax>Gw zdJ|2|Tt;%ZWQl!h@KbA$Oak9R0#Hyk@CZ6Qs_!pfT6m1w+%M^vP;btw;IvK_h4|ac zF!fW92WAYb24=9)!S49k(UUSR#&|RL@yhjM=u9(Tn*I~o?ht#ksiYJ3eP#2?r8sNA z0j;WFrEAVD{3^Yg+;|P(Vb$3ko*u*a4)j_ktENP!^FM42jXdR;jaogrOT(R zUM~27Bm+?vo+2H zf&OuFs81}gTR`g)-J)$hVHB9C%HNTE5b)f}ZZ+Fx4X>ydg0E08g4xa#%#^&;_-Y4l z0%8--vd@4rP;SNT##%{N_5Qnf(rAm^%MC#}c-Z5Qx~Gx!H;5Lv4qQw*9ccNz{6+K3 z*d>~;Jwd?u7&g7f;2!xz0bjv;MHbrA4;Q~UdiW2VD(njOiDk81 zi^E8!3pO$Wx$MpOpplIU?(*Ai^c)Q38-|CqP<*K!{#i9FBvh~-e*u%8F-+?Dp*`hpKHoqP0}gPhXDC$wvH@qcDd~SyyB&J8FS_6X`IhI;}_AQi!+B$~$I^v+!c?1g16zX*g1-lCHTfrnq($=g@LuW0TcwUw{ zOwbE@(Yv15Y;l8AMO2|3u;tZM$tOzX+uy^5poT+Wwl%je8q7STGQ@{L2 zOQm_;2#enFTB9BEgLvh_8?kVLd z9;{Z59D9Uo;gAJE1Xy=Os68ly)xfqo+Cb=sNIjQyTd)(J8xtD!-~|N{m_w8s(W@9D z`7fI)C_PpMMnK%E8tEaRg7s?cCFwiwd=?t!)kkZXymjKyFm#6xY&-})UAm8#Fr5z2 z6B|1yAk9E{9s!$EORxocOOs|@r=-)vo4O|cZ52sKx9b> zNE<0mAk3D5;&(2A)o?|NpA>BS{0J#|pD2X_*5~v-0l3pU-XC})=Hxe%>+ogTNACMf zdrKd}c-!MjONO4eNqFzPKHNidP+hXXt1ceZ@;!&kciJXt)%diUZ4t9^017@4*3Kua zT}{DgE)UWipzF?@b%08?JtGxm+rsLXI$qE;E5$CFbRLq1$Bo(aM8U}bS@Q5d76W^< zKpM8-;&qosfcBq7@1qR+g$jM2(O7Le#)owA@#`6a;8*QW1|j$2w>bMJn^hLXN0;nr<^J1oe#M;f;v0ugpk!ORZ)A zwMSWr89n+!D_4o5=UA}KdhRlc>QJ9#7!40sf+_=RadKnOu?@TC!2tDeXTcXuu2}nN zX^3$)x+E4Wkjvnl8+N9J31|acv!Fn|BxUe6c@BXZJmU);dduJ$KX8ZJajWs9T0 zSSw8*)H;Jr& zyUal(+Hqr;9RS=25hJq)$o3y8-F->gkBLESZiu%!IQUkK&7V}&n#KvUUw?y0%QtH6 z@-!BIdJUNPgQ4fc2Sf5oP)<|6EL#b-PZIZ^TIU{&6}mYYSHG&a>nuk&1L-bDVO6=*kzkm3@xKCnJjDXQi86bqEK&KKSm`JZ!|^1p)S zQkG`n!(*qf!@{<9zv3PPY6n^V0s*uRPm%0COT8uAE6;=tu~$wf;XV)$sZNR8xLm}~D8G}p6`+ZgajcSU2dZ?MK?1)pglWEsu?gm%O;aln3|QBBgKyJV z=@(csY2cd~$s12%1)O(c)!QcA>&%{Dg+ixERJ^_AE&2_!vFu_U&wS}nxihnBHRK3MgJIevc<1HIKp z?Kn|{rdwxW4w$e9iqANdB0}KdFfsUluyN<4Vs__?}?$vw?6;{S_q5h*emh;M}!q^3F}GB{Ri(i zvUkb6za!Xumv~-hWIhlU#9cU*@^SE(>k(Tn&WL6iVgcGX{LZfnu(Y0gE3ow zQeaNl;qN_w`OEei9KA`Ki&w+%6*-6p%2NU8ke1JfDS5{MH_-D$vQ0yX8F9{R3WABI z2aC*U@A^%T(AHV%dc*w?i&IV)jHV$f8n+{$-Oh4@Lc2wq4ylL9(jK)U<+%rd>*3ci zDv5TyaqBmMe5TE6b{{Fxd>LLTq&{3$}(zVjDJuy8d?E` z+Zyp~4HI1J8<)D_oSIEwdEL5kH1JR=&`TnG8AvfhMyoo)oRArN_=ODRZ5;-UFgl0L zJAe=r8y-5sSx|T*PlQ8vR=%geagQ{h;Fa5~KE$AH@B`fZdZMY#Hv4^k76C^=96r*! zbJq9`l7_1s9o+k2{roxdknX8t`E1Q`NaJAsamX_)gX>+3yW z(eEh4I%1|cBfA%=&dPrA8akDgFD`=R4-G?(sSj(KqSt;!3kEAo=Qh%Lv1NYPSridgLcX*{`=qo=f|oLmB3MkQfq&vsZ7)OBxzEVWYL1=fZ0NM5oftnH9_{2H10z>DTE>pXkUnpO@)`zeRq$$3^%6$`o#=h z&9c}7l#5`ebLlMWMAoaME^F2s-)sXZ0rsvjvUx!)y_pS&sYC0q%Jd*KHysxs@t+0)AxN2bXRd;;=Gsps(jW>^?)cP&e=t zDXTj~3O)@dXqWBK_!go(31@f5&pa>Pyf9t7Gy8Z}5!SytfH#J~qfW(Bk57a0PNmXo z>pGJSUXu&J1;&R21mOl-A31jGLVn4TVaq4HA+7emx)8oL{-ns!`3Q4vZjM}R`dEFY zpq|4s7itxKW1~N=v!Erb!x-#p2;owQt8>1TN0wJ`dhAMsD(K#1jd;pn=QgjZminx- z*HAn(@41F);S+7Zxpdj2iV{{V70s&vOQ@#>8ICpDr45qV>A z@-h^+0&>iFmDRoDiV39e97h!PB7&D(`@uBaht(8%EihDTm5%!iA6evV%k2Sg&O#s) zboA0dXt`!#$-G0W1C+_IaXBNbfM_eXg0l_QabqiqzK5utX!OJuIN;BZ zXN7f$^p1)khbKuTNcS?_7JxWB5EjsNl*WBnZ#>d>(=w$YK{$A-ILfHYNC3Nag_Y=< z!a(k3Fabsn5NEYi(`(Yf)P~(NM;IWVVP6G1n@C>!@4_Y4{{WD`29e?0gKg^&xYV2n zS&AS5@CE}~wt&jMhAX36lCQ!m52ova6X1V8d53A}R38g>IZHs)^C7>8^<=Y3?D|6n zl>xEvlFqN}q7>AxMfv_C)rb73(pFBg#A!wKQ!LZ&3MbkH=P+6FJ%`A5^{3MD(Tn;Y z4pM5);QTrrygjtn@8FnII3@o8TGrqmFe+~TNE@vU`Xv-e3Cbl-A_LL~DqXr6S1RCi1 zhb(ZhY`-(RmauOubBV7vH3I^dp<^iA(b+dIlq1<0edW$nC{Sl*=%MBjT(lP{d54=<+++FY|zj&TjdE)$e1X^3L zO7lc(Qp0vzK;gXs;-KOe+CVCdFi)NdX4300$^hW;1kW-wWx;h|<{a6iFmJ|M_z6AE zQA*}2kh+dG4v^^P!{8DJluXUx3>~GyzFnpm05(1ht%BL7*6)sxr7i0I6R`J}Si@3= zuyl`@x%f;lV0By*HhuX0;u_8%4~L&Na`+;}vh|gE{78`|0V;RyAAgBM#(dfj)(}-- zbr$~s73)hEUHo_O%E>dh&E@$X!{$4ywpE`F7h-e)@b%VvS?P0u=17|?Q~M&ezePVJ zJ#DoY;fINLQ3NlIik8qm!V1wf3XF)h5e3$EiTsnkYX1O+6eQhIN?h5mSWpY$L=mw# zE+aj@6X!6>vtg<>|3rvad^F2F5DS(%iw9(wqGGqGwkcglhQ~z!Vfg z?1x#a-{+Vvq?^(Crl-ePG3OEmMFQMSTs@-T-ey|QdO2>-cNOEL*T(YALAU;qbMCW5 z7n!cdCZJDZ2DF0eUjG0Be!uGDO$Uf5itj{&`hKhU6byAr z2=}ElLv@fQ5HKdjz6~bd(9KNOE)zu24nbcFhU-xR*jgB~tUqHxP<$m6@Ge&bw}P+z zaVWe(pgs~FlQsx13Nma@zKkm(17qTC8(`A;_aN{U0k_17VDl_~ArD3$#CR0l*;~sM zhiK6=+N0F&pY9!xG-XAn&8d+7C!Ai7c12TSnwA0gPe zD)1ZzgW!(4BMCF2P%t#5+rPzViTYq`kBd_JZ*Scl$rsAT&^GA&16h`=qm{KItJ z(y$JSJZU}MnB5iHvaa2^Z8zIo>_xdsFLK}v0oq#tCeM6E7ISv+nQqC|iKJK8Ger*G z7jY#|L?;09!#yImsbIOG;1<*;GVvJ1db7}KEN0$VE{9~%Nu;oM_Ho+riO^2W=b9!Z zOJ`3Mf(J6#SL-n=Nu)V|3zVx{?mgfFf{z;2dZN7sqM;Gb&=D~6^IWMeAzmv4Td@_o zdX)8wDfiHt@ZG8y2)}B+27$V|!_DZAsvTv@l0dhIzy{SEu}SQjL)Hz~BU+&!RM+7~ z6`< zMbFl2n%4}aR_;tra3;X&kOI_x%9rE)U&PWV(Bbzc{kr9b`)Uugr%!blB$Hx2=USWU z^F+4Iki_+nYxRb6s#!Z-P_IHD3?`PH*{4flH?;MXYc0gTYxJ5xURW*w@CL=_F*aOB z5|3+00d2#YO!ytlTChOlu%_FIYSD%;Ty7~kgs><&!u1?wIBUlnm}@RsJ+V4P;PIpu z?$t+L(A~vv9-9R$bgEG#I;;Ue9lTnFI#E`V*w$@6jZ5@9xkDqn4}b~AqnD($Jpftf ze@J+k>L>tEQSe(hoXa1j<3;X3NI6vL)nI%I3tBm1i#vHHSS^}}c8%<-;cipaRT&nV z1cIq@Uk;2OI7h-4!nII%_O{46+wnjFVRoAl2tM=)wZIk{&>s_Cbkyb@$2jV|cKDWy za7#;N@pWgj2s#Q6iEO8`s%MU=!0>t3oIWvIbzWbeJ1stAc|5q zR=%*$AsH=bo0^1m1t^Uz4jYTMM7Vt7w^oR2=@F(qf;)(Hvc9mM_+UOzl7Yj1v8Dmz z6*H?t<6S1<7%N8lpP0Us!>?%_df3cOq#X!FEzD05A{hKdAmrrbDCL z(BE1Q>A@1ld)qPk7fBpwiYgaK`o+xlmoe!J5cK#BCM~fu=59sXNd(_+R{kJ09$Rpx z-Qo_gf%&~&e-+W9W;fPZ5Z)D?VQ3lxzAW{i&rex#9>&`~(8>WoZSg-Vj54n^9Oiaa zoABkxT%KKV$af3in`!3jQ;@i zE)hhUQ}sbH&;W)sL@hkIa}{T$YHUY>>)IuJ;J?V(L)`vlBlL!|5^qI7;5xn&3Z29z z#SZbjqhRhX3GU09C_6+Nvr0@8N^~XauI|7!(x=}&KDi5k&{uB@P6@f7I4>!+IEnQ# z%{=pf8e_m4P@5rf4nKXH7lH|6wqYRH1iHKl!5p%h+b=!Toz@|VvU_OC1+CzU4^eOp z03AGfl4UDS(zZt(&XKU?s=zw_e>6qV0iaI)_$){qvf`8KF*{AB)Hs8l4lM8dhz5Wh z9yd`-wy%~a3recG!phmu@^pZX@ydN*8UPy~0#jX8GxzBc&t8NE2Xi@q3JN{|7PNJX zyGvcUcgl@K#CdrZ1r%WTTA}64DN5njNU3YBSjv`wAZ8xD8cbH@Rl2)BNvFpBp<|!P3?IdK8Agc7GYjQF zxWoF0vVm`-xpI4a6F^-e!}@ ztQL&uO(3{laK|Ts5Tx}I_|HgtWt!yJQ#(y-tWedj3IO&Ja1>GDAh}EH5ddXZhJqd0 zs)gwsrnz*TvC=BEI!zR>ECGNZjtiNG{D_j}n2Hy%js^F2YT02%^-QNnv=Yz_;)3XP zmP2$`FC%G4-#2~}ltY}QQ?igNNE(`&a|AKv-PBF(ZQMlR&`2S#qf)phtpd;vp8*z* z=cGZT!YIQhXSWa<>y_Q6fi4HZq(br8-qH4x=y{MXxZi`MCIg}H^Id8w!Bi<71kS-Qpf+_GiaMg=+0AU7dPYu8BUT9f z*y)O%&&c*4A=$vHk)sKsds7T|5e))xQGp+@h_w;1we*{y_J|;H)Gy{7CUwZsO4wTk zP#q1W=npGB#5WH>^*+kyJS6<@nwkLhr?=9)D&sBUI+nYq(^726{WO~w-P7ILKrkza z4vkHs?=)fcL)M4U*heB#R8EGyCx)XGp0Pf-xR4_OM- zQO71dX4%^qnpp4VA1EkepnNaEH)xjn7mf*(bIm%#*iOHz;y`J;L~Ez44NeS*Dh;u( zi&uv&Mg+Ys55&K`?2uqlMxPKk?5#@I8yP)65UQ2Rdhf;)1#7#|fij@u_pV{E3VH+M zO8)=|387G)x7*}Vz0KHWEvxzFW1&p*vYHF$nWBedz%eqzEgY8y`hy%*bkyE0s@LfXtW9r-f#$N5g7iF5 zsSfh=>V{*!wG>Ukw*LS!^md_xjNX4VB7krnF8EHj`WC&mGM?*7$#J_@n(-b@z2F)E zaC{90^;CV+iZ3x;4Fv2O@)$amv2S=3QICNGElXhzv@v~c=`USsk~+J(TkG+3429M! z+12UR8-DM>k;e6(UyCASQe2_pusX%LM`j&OxM$bmP+HLzmH1`{tk>n&SnCKZRvP?j z`juWViCQbU1kF%2@rL)c-O^PN_Ex{@R#e4R@q;FVD;fknF22=4g+zb6dAsTc_C{KInoqaEx~m z#8(Nxu)C=Cnn$%Gye^9zOz|6p6I9%7p}_@PJsGl}Fg18()Nzg8Bs`3Myi3?4b-kvr zdFGg&C^c$mr21@}5bHQAh!wp z%!^v$W5w4~7EZdRj2EV!^HOJ(D-qr}UE81J;9^NZ-;W-pp*&3$DbQDxH|weX~9HZhF{w>9fD{kj-m zUa}P^&TbE@;z!VROAZlgpG1(so938&r%gXs#L(|zCew;4AAKpnhJunw%V><2U2L!!NR zNp67@ic+Yp+__GH^zl&Ds6VSE;G0&-koqa^!_iM@iPecHYxlHzFgPW-;or!-vvroi zN2+1*&#l zU1?I5ly-8SNUc@r6m{8EK-Hbhy?!Oa2JEt_b)wXhO~M{`%9SC5CmDQbU8)?S^ej`P zG_jPa0y&I6JxBzR>vk~1sSCg6>|oNH8u-?WaXQx+Bx<2-tGGZPbW}b-@wN^GA=iCZ zWrV>fs1J}C$*}~kxM{QfU{EyygXFF`0MnFU&7z?H0EG9j`3}jUs9gGFO0Nw_yym83 zCz6wo#`X&4dre$rqp4NFn=YolZ@k#gK$~ezmB~2y!6{n!xu=)|LOyN?Cj=J2b%9^3 zBB;2D7gtGHu(lbs`sEdv6@~OBo)+iPJ;+roXjd!xPLwcBBQBDA!zmhRawYLJVDL^7 zr2`VmW$uwH3$YBN;?^&C+RA6!>L|MXp_Gj$aved~d?~@4F#%@kEVs~U?>ST&q^?q0lwv3c%Rv?65B2yrG@Bu+bPl}wm zl}8Ql^GEt1&Ciy70_5T2d_ptd6+Y0Qc>hD>%UbB42 z`pBCh?=;7-h&twU<_&^<)XYe_1a+gVGIgY|Inb8vuX*m#2o$dXHp>>Cp6HL~YPB&O zU0=lB!_5xl7gyFh)Vx#FtQR)%GzxtXiG{Zm4`jL_ucX_ptv4$M&cKQ>{o_ow~t%kssP!WJNrfR zJsJ|C)q2%deVlzUA~CX2pc z@&hAU=q8M67u4%6IzK>HNKvkXLGl)hXFCn*bq|>i4&Rln_{4*4@4SRp5-Z zI{l>?;zM_z_-}g;kncqgcmqPbE@&w>FJx;;_MBwpo|q8G?Mb2+)rshNKr|{CoveXb z07pZ;Sf8OTK14K3D1JZ0GJE!Fj_nKf%bwq=h1A7)@S_K-ERIlSdsEcdjik1dZG!{v zRHOZ8Snvbb@uk)->_hC1scCPuZfjcdX7Rf;dqnOouLyBZKgk56yIu)OV=jCHY=OG4 z)B)J$H0sp`e*8XDoY|0y2j6zyuo@Gjd`%7aQrJ!tadoxQRz^BMh&!GWDMa*-iE0?! z@^Z_rhNzoKdIQ=D4BkQGP9x)7Aa2Xr*IqcMNw`27;I&A2!QA-?dV-ogl^8n0cWmG8 zo1}m<>+%+%>#@tW*Bq5?R9|_uTMP5PN$P7n5UMtlPE{+1C{xqra77(gnRq>r{H3DU zA5Y7p-TwfH_ptd6&CDk7j+xmvYuXnp^=6G{GeZv@AfC7-I@RcA3i8|*vLr@qaXjaA zOsqPmwh0Q*Bgx@}njI&$D-KUyOU=(FZ8qOm)*63!&f5LN5$eWNm(ZDt$HDF568!C9 za`Mv(XxjHrd99;exW_9PAw#QF&`!c^)!J;HKEY8gaGQQ9Z@WVc#$R|p$@LCRFs_+^?XJA6i14ptQY^4f5WizsA+YZuMz z+Q<<5Vez0^(bhcmrR^VVO0nJoRq|y@N3Ng+1sy&;Bd3E~K9Taa97)uat66V4EU+IX zmoIkQVlRBqXs0H#ksgo%K}X3dG-c=FThT*-fCPZM;$CmxB|!0Z>h^aSlM#9 zNl`^$XHH`9ODbIxNmR>EDM@}SXg_RAG102}P124qW*zRKPrR#a5$H6i_=4p}EMVR? zE9n<;iW5QGszvPWRj;h4PO@Dep|6jul&U6X*irM0R9z1bv_jkJnEJ{U%3;t^iun(f zjiK~LZ`c~6+6bh+pR5%9+y}}O8PF*C3&6EnBCaAfK62(AW(U#~7gsENpo7t9A0n0^hTIekA0!KBD?%C*uzQSaHJXF$AO;lyx5~|%LGn?{36F}| zuTJm?C}0njqS8zw%{P2&ksWT1Lzg&-at z!{j@(I;t){STs&nM#5Robfu16xe^OO^34S7JZ_~RJy~a0h~^<%b0IVyba0ysc0bfS z=-yo@xj!l~6LL;D`c7QK!kt1ewL5y;6NWhW z2AhpGH@05|Jo6v2b5-vg3vDB0J?M{VhQGM~0GV5({c{AP_#@&5!0#?9oU=IZyBWSt znW9}1um~N|`=OKrgX2)uscqeSu@-?@ zqHHzjpJ)J6c>;wvkCCK7#%+{kx2UKFiM>Z>NHe0ld;F&qMz5h0ICu}NSmn;!(f}LK zgXJ}OUulyz*E%|AP&ReHf0j>&gYrZG#mdJ<2xCw-`FYSPdaIATFjlPVoS$Fky^D38 zh9QGt=dutS)yqCGL>T>NKgZv}lXYLuOrqk;>Abb^)74Eb8UT5C-A{4DE~-}y>i{$JcPdKo(+TqR#oN3eknAUhr|smvD#bgfGC#@cz5w%bi9ygzsAH_Oz)F2O{t zL^qX3XpBQ^Y1dA=baltLkOqRH zqu0ULnp>G}U<*sk-j6`6;m-ta$)JC@do8G{RWw_dI~4+H@c^|0tgH{s8)BA_6QDij zOdmp}dV#xe`R2j$4X`;?EXUohHC3f_Ln`lCpla2(2%2aKdab3oiW$eks7FT5a*K3y z1~w;}2m^bovid@rjR(n4TG7iJ{{Vpz9zLlWC+|oB{{S{>xlo?pTc44vK*nmMp0sQxg!=h=JgW=2-yJ-uO zqQgO|^QgQ40_RcGfvUoHpa+Mr`3~9BYK7d^Q==vqp^befxHhwE`Tv-OXW-E5zDS~ut_aod?R@N|4ql!$>jRS9lZ#-)fZ!y>A&(0{3? z4)^}zhmqMydTs>+qJ(r{DUXB=tsJou1lVpkH{6kRPChJ}9u04OM$9lixzh&g57nZ?hQu6a{drpul3 z`$3yF*gr_yS?Yg$r+V|SFs5piT6}z8!Mv=W&)>CUM!vM2Sg8|^eUB^-=ciGBlx1)!ox4)L? zF<;%Y>*4J@K9|I`7i^!jK&;(hcSEhMR08T=1<`Dq`)tOuh7iE>LG zbEnJ(RR-*H%l!@F{xi1zX{|xq5$*76zt{dd-oxZOKIyutcbC8~g$Sa$de3ntezSLt zK@UBaue8>%s;xQ9rU}vkps$D7jU~P#hW?W$;L(2)#SU>osQdXn=KFe*Iwigjw5juy zslsqiY6HVlyjGTVXI2km-3hotR5;dW8?eG@Sact(-GfmVxK^bBE@b*Up)^1{d=u+< zXpf>fEU}zt=`7=0=sV;0o7w7f2Kh2f8NK+}J;5Ig9y+5)_!vAcq5$V2I*e{0`^L;} zRjz5z_%$HG?iq?`96|AD_=RHfs&q;}3UJhqC<8C+S?@S`Cl!L+5coZ(htl{p5=D3E z9I~`sUb6N6x|xA>)Y)#30wn>jmf!)Z&s`Wi^q2b31leeA}a@ z01s3joYk_~N9ztfTEFAH>^?)i74RulU9gX8)PY4G3APU^p)NOOR+?Enn`s|tjd5nz z(r!1zSFuYcaXSC8vQG?Mk$hx@kYw0(n^F?mH_E*+- ziCM*sG(lrfy)dtU--zwKW2d6h(*w#0wBF2$Yl94ZAQ^Wo{buu8mY;b=t=ANuoz8(B z4~M?+SAq+6r|nRs<%N~yavIByA5tTh+9n!dbw8sVITfTFd%(B`g%iJ?9RPS-U!qsC zE3_I@G?zDy3-pc2w`9Lts59?Uu3xG(^^~uFBB{NNcvXIUDmr`~)5GaS#Z2IV&v$hB z=6gbi=>r4KSP#58gVhmterBW&s1swP7CuX!M36@bWTzcRF8Mfs05Bgmv1oMXtwDgb z^#_*FdwmX%o#(^Rc)6=Ne^T>a{{ZqH_8%eIhOJ8C`m*ZFhA-^w-0wgu_l_`*<2iA| zPif7s-qD->=MSu=S?BX59-H>e_CQDif{vdKlP!dGVgtWNwXWl&yI`BsV(Ga~$)BVb zeaQ>&{{ZG+8C>8Af!H7df{%i`g@EuX#5%Iy%^^Zm>rW)w)JbJ|{?lTN49seOWT$$= zFe%$e6{bEfY2k1Jc?%fTi)_Th80mE0yQ(wvm*{Ts)m12<1m3U>$<`-Wj0-##L{u!+ zQABD&USAtShkvY;#q1ip&9Q#aFb41Q57D=+`pT)0@KklrhmdE&q;uk?rlt=^A&oZs zL-V<~51U;Yg%k3ydu9$dLO!E3`G~NrWHRp_)DEI26`QLZCLr_omVWQfQL)L-K^&cQ zbbY5khIxK?U&6PziF~ZO9&dm7FMAJ=?&a{ksLodvucWH=Iqg(Rp&bO#0o&owj^-JD zQJMtpZ1-nB5xajnCeY9CdDRXxQA1OM`kjat6BJ&#?2tfceWLLAH6pfjqj*Ou5eB7GZcc z9SG^8?>08g8w6UEmJXMlq z%n~al6Fn9^Hl)uKKTVvHor0fYQ|Z(oY=NJgxe4Plzn6FW2D+=;xfw5dz@IMv`#ZkX zo9tk6flP3(VQR<2NR19S3d+~xZO#e5BiP%wCdeCcD!MyLm8?ogbn{RI(mvH<877g- z!XqxhoL`{$--G0#h3@)}*yX1FNQtlYwBdE8U!t#Ps5ey})|CUsE2^y)ABUt6K96Ch zPYw^_e-dX(uI#c&wYTQ#z!=4DB%1FPvG#xgRGhrnezyK|A9oDaCWZ~nb6SMUbt-H=mS(fq<2(cF+ZrY*IU z7%j0PIbMmIC@ki_k#Sv$#Su~L*eWY@J9{H+uM)tr7^z%kq;4Q6YC<1uE)6VHB?8(~ znf@Ac+fj%2bV59RNL7M#y~CZsdZWv4^#-Pcjo7@oB!svFhvI7sZtVHN$`tVqzK=G8 zMRO9g!$XhDRSIGf@-CeSm8td54A^g?axNaas!of+J?z46Q<<_51!a3P^Jt$)JS+Y1 zZ9O&~WE>wk#!erPk+Upzwwfs#k&-2ng1Od=D(iwLV!ZuMG+DmcqV~J{7BKgBv(1Zn z2dgj9ONu;5U0R*BmAov2XPpP^3BTlf&v9iv-kPp2|sktP_Ju3VqfyjA#e zSE_8rbFWBBpH7{{OdgAfRl)YBTAq^c?Nfj3K!9|QwXkX^hU2IW&cmZ=^yqld0{}C4 zN9Lzg^Pbq1Vo;#r%_ouE!JrW`0{gy)6Uqmqoe7K|T(9m|au>nFvhMol8gyTbi&P!5 z?<76?G;dP0FJOi8mpGk*rkg0kjaK^9hJ-Q)9-^OTi68H*HDPB?2gremNCubarct0I zS(K<3?CUrMWiazl-z?vN+OxE;KhVAIe}NhkJ-__|@$UP@6E&a6+x0WNkHH9Q%Ygq- zj82pf{k-slR)x*6Nrhm*5dNVz<^z2xImAiXAZ<{T%nElBVlzYOET+xIMjKGf9aGEkrG2} znb1-KjX(AJi>pztSaPB_P4EIzr!cqco&DU+fMG>hRUa{SyT+rV@A5~&n6GuG9TvX6 z-UJR6dMj&-fCAb}AyQtOH4SFF9k&C|UHckE8&P^u7isWRcw+D#kQopa%Ec)#5BG>Z%k1e@A~NX z5P7=VMtI3P?-%c_YKLxfxaRwY@JL83Oj8Q)DT>soKQ-oLup7pryb)E;=*%Y@iV>Tz z?JspmRjhoi~x)1tv@Q~LY458$mI zw+iJ93~rUpkY~^bV+pcngmV(twzjt3DtT2@NH`cI_Y86BH^eMeRKhb#rk({ZlLX%# zc5k|sRcZ?zpGWw9s_R&)3PS5`tDbXX%oJa#EL>&jPZVP8^c%m-Q!T%-xgAcpWZNTX zHnumlQSo0s$~*ALLIY8WhnEnr#9;Bb;H9Z6N}N-<1x!C9Y5Q2Yfns_F43>1uQiVP_ z^Wc5WU09FeIn$cB>&e+S2;mOyy=uYTf}_D5tTgUAQ@{MscfXzcEZa0*EN5kiHbbv` zFZ}ZAzF>USDsD;XVux`4X@D%g7xk;N#*90HBTV`n8mMPh?8+Bt+4>Q`z`3&H2ru;$ zlWy?{@3(#|2P#r$Z#2cNZ!%>+IqU9E`@U;=j4}FHK0W5OT{EO-0^EI(;_;Suft_iB zAwa%g&9dmQ&Ih@#y?!g>jPQtWYgox1QS6P0G4IWV%28IqLOD8ag3$7-PBU7asJr|C zPwc-P`Eem`lOb9#$6V}6iswZNpCnn$m}e24`UT>v>~)e8Z+Lt9E&C!y{n%b!v_Zx0 z+5|3~s(zncrt?j0wIR+l{D*i1@^nmvHE|`GLmi>`I9<~2$AeR6vExy8VDXb#TspHI z;nEXKJ~tYZg}p$X`?Y@VflkgFTfdCl zoxM)Fvq^E@oi~0~EAup8ikV)E?X}nV_CP$(U3o5Nt1p&z#EN0-9YSX#v3Vr!L8Ges zN%bKW+LsT-vSFZ}`>VUH8Twn;h|G<@sm=a{-^83Kog8!g&-7QW8P!?mF?asPl6kIZYafQ{^1vcHFEXtUiD-g)u#S9wTkj)DuRvn=oODn=m2y-KpT; zzfY$>Uvo}6=rO$z%`Ll+iSvb6-LV-aTo$C7ALimA?g%WO#D|)!zTi5GoaUo#~L_jW`cv2MA*Ja#Ug;NZo4G zlNnm+et`^P+wH0BglRBLQj_N>s+~c;6yG=w%^03fv)v#n3@&NrZI{k4iVpoS@ux+S z-Aqn0A{c9oT*H5Ui#DSUp>-8R%D(2N&e-?srm5)L=RvH~9P%K#X@H}VB(fj5zWCVX zvs_iw|LCerBZ^Q`fxSu3a=(jf!f^afm347fEyua?$(BR(rMQY!aeFzh2G!3=N>KG& z#-rH71i5=5-B4HZ;2q6FODv#Q4*m1mTm4CZNZ#I zA4Vm*A=>v`8Q-zxx#O$D)eT_V@7>A5a`!J#hKTnp1m5R}1~o`}zf6KcJhRSi))|D( z9}vloGTAx!3$%LoCj#g3|G)plgMY*Lw^X}i*0h^$>oEHN!l7QoqQZ!B&4r!PZh@}` zIf<_fBwkF#R)t)J7D?1bVI~5TlsiX#sZ?VKm|C_h5ETs{SEQ2Q#VC=Yu9u=cEeYUQ zGwOAqf|@c;p}T7os(js7cO^B#gu8TqRI|Vv`{`#_Zs&^?^W~FD*VDSQ|VNrVb1Xj)LR2?+G;Zhx~iEEdTfO2uRt{zll+*JMi;fvL$XZn0y-=Ih$xnXg5uNaAVE>g;~X%j9Y(iBz(x-K?FI@x{7 zJW|}Tgn4kX{f)<1`d4G`NK!=-DN80E4vxF51dI&opkA;(%9=3^uL@<(b|u;@$>iSTCYk$-G(6o>Jr-@CBMO^WGYtHJ3bC2IBvV-ps@=b+#^Nd6LHc z>XA`J1!5_PN+|Y@B`oFHtdaT66*%3=sjQ)XL|-sY@*qwZnOyF;me00jbtsymkG0;A zjNpXfMalR=wO_Q^>z_V!6&-NnIq68>7essOo$%uD&<+!hl!E~tX*|842F|D#^70TOdFQLy=jJ>TSfOO(kt(D7SXbx%60c3%V5&qVujA{l9`1(Uk7jm zeI^gM*up^pJaU0h08d=_a2WXitLumA|17Gt{C|%BpVNIo^q=B~`6i+TC!+cJw_7Iu z51{@{6OLp4?_j#Y$SUke*m%{yK!bRjVE9n$e@f%O$^Cio58{FUjOI6`IQ$pr7pQ|D zaYyVDx$+O;f0F^Eu=^n6@}Kx@RWLk~G8cGE;~tF0QCAgIPCKt@k(55FnO$e7kosTc z+&;qTQoZ{d?*LInHzgcK=>o$A!|%8UDbKnW>jt*N5CIKnBeAobz;+^ zhR0uzVE6zLCT&m_4yp?bjc@#FT%Ph5=)ShH>;+Fa;^w2IPWh8L=Cyz0|BFqWw%(Rp z*HGsP!M}d`gXbTiFu-Xd;^qS_V2Jyu;P$BP_P|pM)JMtJ@-Ol~0q2$Q@LiWx8z+l^ z_P7yI z9J5u8YZ~zWZ~5Kqzu^pmEC0}(juZ~j24d2--7S^>2D43!DmZ~tvG_MJKoiwMwC#Rk z;b&42_mtJQ7~uWG2q@=wV2FF-_oKj50x3+|Kx~JaU!cr3K#&SIj52E-NBnQ`?h}bvvC0cePnJ{Ud+icTxQ#kQ?6jCvxz^1JU;1SvqP1kl}E5 zv;T%JiVx0Fr|bVz&v*LZOx~O;=j7ITFYPb+yfARK5-&`Vn|_40+Vu8a_)`=ZxDV-D z01Wf?sG&P3ki(x1ho=udQkOqnWFL3)Iv&yx4icX)uk7W!e~!NXmtMO526Ng?`2z+% zg@IlE#Gfvh3H}28M#KE{5GA{f~J)a;v}3sCqf zkQtfkKpSqh!v6iGmJCrx-q&}wWC?dL*gf9KWmnpgDbML&1q%ar{{eF<4_@yE`$tlq zD1lFVz^9>nmxp%mSDCilC_Vw6$2Xaa9$ATd1CH5uJA6AJ$0E{cO{StAQ7x4#?C`(f z^#2716COYb$D=$^2J*^&f^*vvsbtbI6Ra-r${fRl)At1{qTqmW&=uAn)p7= z8^82GSxJY$#NfC-gR1P8u)izrDNszmD>2atOmG552`bA>=h3AJunN7xiM9BUg?O&a zckMW+dNS3zm9@O=*dB4je1CZN3p9*?&)pr@gBu}#p#jsw1c8#u9Rzf89Mddzo^*N$ zjZk-Mdq|OO{nYVLNl=7e$@Dzt&@%m8ydfBNZMq^3sDp5X{B}eP{+;q~)qD^nKnKte z##H{qy{C6J?p|%;UJ#m+S@w|I8mHw32c1>uaHD_*{uc{X|ExyB1d z5c=$-wp_tm;PgK5>K1tT-+6)q;|HLeLE~5!h!Y4e1pj9pug(ndcX{NuKcCm^c_2L= zb_15K#Rs;0j^*#}bMN3uY49kZx}L#dz&DQIFisbkP&*7<08~+6R`VxM;1lUTa7K32 zw+h;-6J67#`YzT{(r<@vZ5}A<4bAviG^qaq!Bz=iB9V$M@Hh_mh#>?9UQh*}+5@0| z{w3wFI{;_(oW_q2&L9o*eb&@^&g{EgcTzwPTWFYHQiU6bT(1bUOtio`5%3XC5g7P5 z_6`nP26z+x1?RV*-&IH4+!|%aU90s1#?<~~uITy~s9^admK(XdWeWog8NLU+vw#m@ z`N3ggc)xjF(}9*d4Dd$$qh+9*=zdzM3ck2W*`)mis*hpE%1VEJlkwh?fA1lx2u$dS zA{CsI*98W?W&u<~_Z?s^{oOgQ&Hy5KHG#)}uIk@_HeCFgqXG9=0Ykzc4rR4F%sLJA zj`FzImv}DeTM#i%+89vb!@E+y!F)gg6a*)00R$ZUBjYb&fP%`Lmpl&EgbwoTQqeFB z(daX<&6b;1bv|Sn`izVE^6P@tDZ6+9u019|LGW39{u>VKtfQy2y(~$R zlJ^4kQuUEYjL%n7x2q9>#VCVZ3?hrCxX%H`(erxXJ`{2b++p++zwiHX0bp)P?c=+= zCcM_$+ym)Zg{{&po#%-zIoPX;?{csO!!25Xxz9Ad%5Scyir)fZU|_ocA0mGL@Ec9B zc;S)Y#*_8W>cNIDd<2E>pM-8(Y=0t~lL>ms57ahS5vAKdTz+-pf2168-TgZhAPk^s z+wfZGvZ*=ca&SqcZY{UjzSwH=p$@a>fw4c~HVz>$ZUH*YsQZGT~v!VQaLhMRf^*`qa zz&HHZTortF3#nU&Rqfd~Un)JjjTT&pb=qom-Aca%-p$We#Dy=#TemXbZ^d};wIlj6 zkGp|!@Sp2X7Jozgt(*TQ{r^7Zv>WH0^gqJ?mCFBNpc9AtpRopLBfS6jssEjY*nef@ zuW8c19{+1#;(t$KT#*?VEB_LB(nkCvyY7 zlRy6PoeWTk;{zkk6E!G$JK_x+X2la`KFD2%4#u$cK=*MmgZMSSyqA%~qS6{qadU+r z9DlbM7u#?=v6bddtpv;OF~%yktCAUm7zH5Yd}PYWvq_^-(dLb3Uy3X_8@`@>YHeyq zM7_3+p6)qk5$|0(aUkr55D_fnY6y6_4myb<0qm{FLB=8;C)U`Jz2pw5P(EhrlzTdP z!Vi`}8`z>ju=P5>p*iXim%D=S+vIuCHgE|P%p{6XY2Q}P(_;nQ-aA^oi{^m#=ywwz zX2hOqYopr;rl>F!vRrFauskZ{mF*dG=5&%%H^W{3JA{Q}L$m5k(0B5Wc2ELi)yZNEUrv)q>yes)@Ctj1<8 zMOPtWrtjB0?d$Bsghn?Mu^H>+kMTZ|7w6_~EkCxi=id}3zVc(y`vr=FJsDA*s*EE< zlbmguMA55Ze(m0}c-F<|Lu`B9y8fMV!LEI5ZOM{@D`G;IYRnFsMIyI%!9Ku=!nNJl z<5^pTA+Qwem_J#kb)hX%@r-?UPS=TnvifkPW{|WFZ8UF@D|Qqv;So!4k)D zxGsTwu>9Up8yaHgQdyruHh@i}6VhnnJQh&b(iOUY?3sf~!5ZccccaWP(fBd}4ZZ9o z`ns=4+v@Hgbg_Nv-kV>8!FWHa=eDW0ioR489O_m;sIyKn8)9J|f{4C2+MOh0hpDc$-tAW`O-iHv;OSVNmyn zF;u+t9rPycjC%P&q-oAD=d&^7d@~3OD>690(qT7up*DZyVc@e;U0}DaS~IO?J8aM9vT@>97?ZjH@3ka z1iCbNk8u)AVhO2XzmbXPX$m~PVimJgAjS5WbS4Rfwx2y}b+ZxxKK+~8E!l0vL8Szs zLbV==yA5`jUOv@qG}}pbt|>Z6H4X7mS=J#c$ZPD@izR7;C=Zf_uSBD8p2rv176oML z<~*Yy!R1*BdTHZfq-Y&AHEf#tFuZN|Fw=B%1!pw!rRm*EebbRhDTAs{pW@n&bpX$6 ze6Fk=uDmfJN2qxQCTA_JR0`*y7Mr8_jPnE;c~)qv!+ z{%wtdWG#a>+}uKBC09WT$5*~{kEMcIs->xtt$Kd4w))*DB3YPx*moy5dKFCYlR_`? zi>f>D8T&9apGE|!uKbQq8JQ{)4%MhEdWU6a=0|L)7WepTKB4a9)H!}#VYkpr)sBVT zDXfK$*qqmoxhwo@)lHP=rB@*r<*98<7ApD3-zg{(SW0PaQdM6nLlj*ES`_*f&LKr?DT`7WY&R(bP^X0( zst%N#wY@lR{Wao)uN_lCQtFR}Cak(i4f8oqZDI~FPT1NUzvm{s&Hhqbaq*fvdWyQ6 z#G^i3F*3m4YFk>ej4fiVxySDPO`3dz{Gt4C-&@S&@&=;E`kkr><_{P@uEbAT?kL(R z%KMA&RV}0lbhLFxPxN}rQ<@EcO|WH!I0^>i*~)het~JyoAKkKYi$dNzlw%t75f^n^ zvr*Q1qufgAM_4Aczik%SoFB*?9Y$19XeK^?&U2up`=o2|<9w|hI3B8YVSp~m4KqZ2#u(sZ?KdpXI6JNn1fve?U# z$qPQlc~YGzFC7|KVsl8$0@;g)GK!t{U~0yGAzg1}x5p`qdkNi}&5Ft))7DB4@wjWd zW4JiVi4w=H7uPSjTfX10C%!3oTSM`1od|8CJMUne37?}#+U*{f_5>}SS+m!89>0v; z!Y-?0KDdqCedgSGr~+hZwZ40yT|_bEKiYA~Ub410_PzM>f!Fxp_{4!vH%S})SI>+m zXnUCpC_~NsISrd-p4}+%&NFu-#xwPbykr&~G7_yJt^pg~i^~?pzQEo;yBGXz7`@^m zu!wtT1{c+V4@O9ugGfg4@MPzDmBT*hlD8bC$5MQQW`u=kIixB- zQ=l@~VCh-*8DAf^EYSSE+f#+2X*cNcVM+Stn_Bm1G~ArJ;K?k6;e7n*Vy|ecRyH$| zZ2UV)Lmi}Vif?BdH=u%M?6>uA!;8}|2BpuGVY_pSVhTEIlcTL}n>5uV(cc>9jpUs$ zZ{N|J(!V0~@9#MHM(Y=V&~w2>FJ5kdrjj{&>kpAXK0*FaL0hvFb_mj;)AUB_ia7Cb zO^V{My4kk_4rdMWDtyW-V)c8D(4iqN(MQvxD-0^w!P^e2EIy{_y76(Y)#+>zG~+H0 ztd;9z-Eq#cPALcRx-Hh;O{Gaq(1Yk&%acbcdeuy3zR2O^ZW~-D2+T!?kyfRR4nBMk ztUY|M2&%Wsc~Ts9!M)QRGV^jli1_mraYNXg5_?SSTeSjP3?HA8adGRn`O$l!y%kGj zZo|2xI@+#B$L+FJAJ1{4p0{!2IQs6NaYWih&SVF@xhjNI4iCl~QFJZ0MMZuHXxciW z3)$x*hkC7*u@Q(0e<5%3mqi=Xcj!AuPJGnoe>H_GntrD*6pJPfxktjcJ9Y3Q`I<%I z>A;dxu;HZ5miK)Wy4%Gg@Gu{vwjJ>+ax(JN2>)Fd$mMc z_^x(My)T7m!;sA}`=Tx9t^6~Do^4H&;UdB7i%3>BDB-rW$O9uo0Sf4~5@@M4fp2k6=<13q11U67k91^MfI@z^&vD&nB`bioa;-9Su z%@WpXFA-) zx^qab#o}?V|Fg9bEC@V63)}J~FS}5Vm6g?xU`)V1_Sbpr>y9EioG}s0+0BTk$4;gF0{k7vc0uW4DpZ%O7u&><;UEcuE`2eq<Wu%_3&v+&2xl%tx~tp3rim8#JP#dVjMxdrX6Nx-oyJrti&O z9&M{#yJEe%ylB{4sl7G>Y@Wo*##@SxF?a;5y-ZD6vs*h6IcqLXH)GSVCSDBCP&_22 zi#*n9%UxXHG_A3JvH2mhkpk7KYN%P_m>`n6xS>io!BE#!($NV^8BN^b$AV#;UFugm zm)fb($THNCRKN8=ymw7Em96@Ol#h$OBND_g5)kMnw&WbCBl(GVjy0RcY~*qqTf>;j zCOorq!{*X&u*y2cKkCj{2~3MjK?j-j`#YPR(e&Kq6E7$8KY~$BO&k684JZc!tT_dY zsP>34x%A!2dE9c*OB%~I18o9tH6yzTy4T0WyDoX=nJWx;Xr>#F-e`AA?1FmPmxxqV)yJ0>?9Fi8 z;r$j*_#$a)Hxtc@+pV%wx6yLhnnQ0R)6Ik}(XEYdhpp$ww|+qR^G zXxXq`-=ZZ1YOcLzY-r(tRhE4hbmju5Ze9|_?-8gMFTH7@USQa*4|nvTpd8?hSz1_m zmbP9m(dz!?(|W*Hm5s2&C7ttS1_iXeM>RdNWh?dJW}EiAoAu>gq&2|q?j4>_^QC(! zceLwG{w%S*VpX3O5L|=?l3bQ@jcumsfx7N1@RV%b7orl@=EH>LMZPQp3KrtAjnVV^ zg&Wm58e|N&kv*D{;?lbg_VlCPX%-{W^5Ldd^i5TiqZpM1pY*&(%Tr?XU+0!Ju<-tc@As<_n{R2V87rv7zY#0tklpOD2okGQR*wDYCI zy`@{$^o&{Uno$Q}IYn+JXsC_5=GMem9*;+-O}n+PrJ5hS=pwjxVMvXxv&RE%#zTG7 z@5E&sIOm4w>U87_h+D-CNt;F^lQ@=pIl@k{xK?%h4aTvbCo!%bDcdzwVF)fyIG+ou zADn`^)d_l`vwIflQB@l1mz2vBRQ*_PC*n8Euev9%li+wYYP1wth#VN~bcUH=O+ykD z+9DB;vd+ySZ4ZTX1?T`C+%hjqHQ?yrhNncM{Y$pZI*4`w?*%rtj>Qo36Eu9FrR14d4 ztERqsps`r;HTYXCJAOZ4^-tqK_REy*YEo_g&S`N|!~|L5HEZht;%BX!`9@mFO(Y(> zP@zg-7=M72+6CsV1pBFuvwk4BFX2qKC2(q^PIJ+YGm(NR@UjrAal`5OwuIXiCr2R6 zE$7&Gz>XT|)wGJKcBh&K^Ua%{3*$M3etIjizR?2^VeU(`Iz%#CVU&r~DdZcD*E-xA zpXdPbn$SZAg^)KKIHhp53UW+`K$)~;XW^q_#fQ_77AAd#1??hbh!@Sw6U!aE8xmPz8r=PJl1#TN( z`XbJHHXVPCudq43%fGfa*7LMxQQa2uk*sxB+z~0s7J7|Ze#roZ;<0bAjk1GF-pE-L zt^Mf9LPIuCeX5&=_bjO>`xgkMQ4de>(j(l}+R$^O;S|=Ou!0dZ^e~X)V6SQZ?g@tL z*S%QYWRkGT*=%egZuTpuC}`+jtW&8^!%d~_35baE>IA7?o_xz9f4@Peo^!uCq$xk2}y(6DwSCy-(1aVc8WX>joX# zj8jf4q_K>v(R$+*wX+rtq!+x;$oW|7Se%Zm39bUAK0Yw^9sFP=+#)v1nDyO@#x4Gc zoG6J_$fEOYv$}u^msdz*O)FY?qM90@P#m2&WvhbP+GE8|noEjheU&u-JT=zN0!yRk z4T^_t6Z<}D{&?9Z5``YvI-S7LPsg#!A#Pv176jlV#PVCqk$RBfMrdEFF6oAf^*K(# z_o05POxoyBPDyWO3I@sS*i?jLA~xpfJaC+zYvNJhkkJ!i{Bl-CCRb|Y{c!CD`j2g2 zBV?#h4IfUmK@VT?nr+-G5$(pxJ}~2mTVh+RPZtoa#=YX&^?ostd&zbl(Nfi=;l|2A zKh(=)$4Cf&o6{DWC?imNs1-_=SCpcOJB3+SG*iIpp0@a$7~>s6E)1<8c?oHl^>pg+ zxpG$RDwGrS_Vz}J`yI)xAf}|}f${A7CkxvYs8F2&wxvGak7%Q68$r8iiD^U(xN;$M zXNF(gr^I^83F+#*4R;MBjgCC+uX1h#%F5u}BaS(Go_fkgGbY$Ev`A9C_`upCR%Xx! zB%;$Dy*-JB5lhgF=!)`9X{YThhZT)3$yB9Bg1(kLDO6|Av7e3GH#TQ>L&nGpuWscd zaV(w{C>!n$wB28zu7ifMD2&vVj$=iY-bzQQnLDYufPAln4o8%OWappFLF%~@XNth_ z1F|8kebX`S)afvdSS{#cg{(6BIZ3UiyA@S<+H*^`>@|8;ibz%?4c!|?`wE0FTklCl zZb~|`S-rqo)?p*qT&!mDy%Ch;rD35J)c}x zPU@*0UYt4(9XvlRJomuX4fpk$DqmZ}6?OT+8Rc-{US@gR(LReg_c_HL<=EY6RfKBV$G(QQ2RE&9M=l3!h^)s zv>n~V@^^J~#RP7qMJfhVL&#RetJRH@ypx859FHi0UuSW(AJo>+ybMus+%a__;H%hO z-l#)KKjLU7F^|!0AsKAewCp{6;Ab}L;$8IS!pCktlKX zA;(_R!g0182U+|~;`$&LO561!7++zb9GL#Ee1u9ma2ZE#VwcR{86TiJCY3^)PqVa> zD9|BAuBKNU&gMsV90P$<{UeXb^q)el z!Zoh)TdI;nZK*ei-$h*EG~1d{K_@%$w1smUBjsd-NgErvXXfG+#A)(0)!mvk(udyg zJ`B+hpoD@E!DQO}Hl0#PJ;6uWY`v%JwON~iW-Ho8L?X#y`4~{|$Z?@kAv+KeJH0D^ zf9-0L@4N11ujHrw+8g&W#hm$q%|$QivR~jT5L)6xp^Ug>P;csuZM?Po@@I?0=^Q`{ z2niu8eUd7(U+A4CV-Km*4dl!v9~ot&n;Iu}<32CEmYm$X3d{FSklEtDu*6B8l&vMbA=hjS zS60l{Eecwa9mdeiN(8H25*G{AHJ4PWAIo!4M($Xbp)_-XuGd7SqD2@=t3vs;3+xZ1OZamUcFro?5a8`~@Fs5@yHU29Ek6+Ya^yF%wnX zA&2N}e<^Mu7h9o8_>91!fvltGHY(r|vsqPB@X?R9_VAI4x8zahy%kBs(#IHmd6-NU z?vJYyXzmi1KDFyzH4zwGsXmh6Y1>uE`Qy=)vi#)bXL0MIAY(N5;9c6FeD82Q_Fdy? z0x66~-*{VFNMME^-;E|Sj1-vG^c&zGKzFs&bZj7#jG=riXvKM60q<&dtLwC=4|dI< zl14QQaMd@0SpgHTf7go$bzB2U?W)Mx>PBa}%ze=z|NgL4Z~3BHftFe`jqbJ3Agw1U zQ0@YWOTkxE(do~kC@*)Md(cDhQe!B~-*Y69+kfueDtwYFJ!-lU^x_$oAC6jWe{w!m z7rP&8C5Yc1f={mbwcyDscS7zy$+jyu z4C0T~fi`Ww_Pz_yqDMcypQRT41&WzR0bJo4T0Q1dtouxH;fvGfL8_fs-@AIzxPXci*_ z-ywNL-P1`o12^FpSa8m;87C>?T=vLWmcqg;;q`ZaL)<-3h`*a+3R>h!meMP7FVDBmVHbL%^!{zDj=hO$ z4up(i>d+U{h$vhq7ijHJ^a^~1!w)iVuxW2_?oCAqo84|`T=3eLiq3_ZN1bJCYro{1 z7<3g%JNmOV(cw&$d>&1XEyi>6if@r0ykU|m6eT`nF{>`}(9^~w_X$y34Az$9r#?)k z+&b4jwEQL`VtW&zV+GWLY%Ph2-g^n=yl7-KvA}m1;bfeOF+ziI?0t@v(#1u zC5@Fa`w%t6>pG8=#z%$moH$YK{pN<;EeEa$kWj{2{`PIAMW)^EE2iUyunYbig5615 z-9?nD*g)g*&T|q(e1FC{(!x16W8e0WO8*5@(cP+@op=jJ5n-v7jcy{61TL0zu_Tm} z=clh6U}I)Hedy)VzW(78BanyK$B`+^z4zvwFVgLJfdMv5J-eWTUn|oay6aWfTEi&A zi;`2)WwJhSuON6IMkKSV4b0U#&`^$Bo-D!H{Hcay&N)#|Tff|*-oG-!*phO?erQ^B zS3~1iWLHCq7O)OjcP}07Bb5?HLA68x%hQb0u4b&wgrxPiWPsx)X>R!o8__p3Z+lDI zrA0f>FlC_%d%6pvZVy}lPu}lH&3$psr|!osfUd%xmgiTT zlC&U>t$|@Cc#=x-=(s6h2^Q#ACosx)AsZf)#LURl!gP!Sr z`!P4ao}t3e`-*vU)CNxA3n9x@HFAPyHc!&0dzdBkOyL#Wx>(ygwj5=4$nD4Wlgy5K zrWFp~l1EB-k@t`CFM4j0v$rmJ4827A_GQj^=g@o!N$XT=l9Ok>OTIm%4t(%tQMJ+D zQ`Wbe**Z||8{PYqs{F321kQJ@OB*dQmWEfKmUbV9n7_m8MB7&-i+fwts8E8e)yA=$ z^9i^R)-MoI;n4+fioF5u%%(+o;(*AR zp1Eso=+G$mRM2R0vTzxZ6y(^XeTH<-(jp+etJ~`7`>1y7ZVTu2k#3_Ck5d|L_Z=Yz zZ1eNWp)sL^A>}gh#jl*fP@0!%jbB>7dNd%f`Ww07h-*Y;FrGo5e?#DuX=nZ40_95P4$3gLsZXm{D=eN?b{eZEZgg%VxjuL*RvtOVW7`I=K zSWIv7{pe^`E>wTqGJd|AeglcNO1hecjNWG>1~2^mR^}mi2jT@6nFENev-ZY2*dfIG zHgg)Wb=ux|&vgjtg}5LTZ!?cg7k`1$S^SD58N%6qfuuB>V>FvXHJc+en-qQ=U6ORe z4oKRs9Esa+94R_{5EF>{1Mz@+*eb;3ie?dkaVef~58Hsa+?XyRk}k!6v^UBF!HK?S z?=_ij`MzKrkG~#Uy_a>=INr3}6|?-kT|B|`U;9%qP=O69NWcF;AS4119gn2CxpQ_v zTxBQz811pdkCXq{pMw0?{*>bvd9erwu_vD(*U8snnlxSFr+_g+*(%y@_w^)&1>q0% zxRKNMoWDTdVmt0Ma)A?ra}mjJ(aO)InL0@8oeD)>vhBtVwSO+9#rg$$)ZUW~!%fUQ ze&ZZCbCZC`=Rkf%gXOXa;LJ)i`tccco$c?Sk6Q}5{4j=l?_&DV%V zj9##=MaT9qeiBc~w$phts9-Zjia1dgcnB#MQZ^J=zI7aNmU;$twYaxhimw}^MLu$> zLPWWgnrEyyBgsPjk0fX34076UR9A^KR8I#Ju_}~W)aT~#a);-qktr~1jh|SKio_o{ z*;{o^F|NFdUrZ$(PS>6P&PDR+m0@Jb@M_DtqRtG7CteA&9VRu%w!B2N;*|FwTA^=% zvM=J}@GsDtsG*?eyfuY;F^^z%jckM)Cei?M4gb$DNr6A=G5LB+yp~6a$bxS8%mQo!ZC!*|NP4 zLYn?H>}Gr=6#5*BSw?{*>DLA;GLkr4Q@%AP2d(0cg{BXpwgi^nL={zkcVs3nHOKj; zC-<3(NcRc7Q!E$a!JYK!3k2;0b@}z_^;Y7y7}*!wS~*Eh3~xDhBb?c}f(4iInMt$O zQ2bW1#yNrl-u8K2jVCHS$ehiJWXVd6I}~RSU}QfNByysE&OUW1$r@=HG3bF)6eTrA z5VNfPKG|9tveEbgC)j;pPsZwsY7XeU-;0q4?AP92Hr?MYsXC2{(OgM~PXV=u&h#hq z7kL@T>JHy`kwD#j1oRT3RWn(CfP%oZ!ujNZbsse4*SVV_6_=dhhxeW7N=KufZF!23 z8fwzl6fX3xbr)&!?BNu2xo6W?4<}JM>eT*Xn-V<3b_zzQ>#)@LWjVyhDqT= zRWy>LU78j6`qh~F_48k#*&?&t@Z{l&!?BAO<0JXI{?TxXFkOX+Z|qbstzo@si?h1% zj{|3ww+k$_ME0|K>C7JS&kI+jrCjOUlAmWe&9)}(Ik(D8i{~dJLKMxOgH;@ohnvPV`up{S=gN-!L89Yr5U?ykfQV zlG9=ao_kdZ7w=>?7+aW@ZJ*-#I0^|+#DxaDyhDFV#r)~T8?TTfhnh%&A4`<+N&X$% z%+=1d@>n+#2Gg|DK{(y(gGnsH6HiQ_jrJ2M1|Ksw(`6PKtMXoNqIi%>4MbThh9ZzZ zuHfL}w2Z2GjHUK^Zx_lo*DHn|GV@MSdpS+#oGnq>rSu+PS-~03IJ}n>3~wJ}NJV+D z3$R1q7Ni1S&6lmew?0gyJ1tVT=6#>ep@{&|B@Ov6%CGZ_V}8l$3e&Q0-{dp1%J;W0 z#JHo%v&5*X6>Uo>Iyn1ER*zABRS)TX`Ru2`*mwUe?@Hx;;X52#5=BKS)k_^c70TB9 z_0p$8iCi`khiBQX3SnC9SgvSK> z^vY0l^0%`%0bBow-Iq$Io6gFiAs)-$Z|o)fetIpw8XF+Bc?iIUhPxDo=}qz)F4&1gU2ZFqhCwH-jh)5xSqxvlHpf1 zwf()@hB~wzR4A?Yc8$IF#o$e6gW~f0qtUeo?oQVr&@*@1>0|TqJ4U`aCNE%flrtG2#do3WCm#sDCvXk|hZjIH3y*Yffo5zjO)Fz1-+5i{Gb`}LWL!UZq>ps` z6Sb}g@P&e~Ub$2;$)$2|Ovgmev^D=kz6dyK64JWF%2Jl?_CZwFAh`7Zq3o@~+Iqsa z-ylJY6?ZG{?rz0B!3pl}ZpF1NR-`xuf)m`m#fn?8QlPlIb?5)>>)qFPuutCPV2;+x zHM3Ucd1jt_es_Yw4_`_Gea6-?4!dX|>$@MSbaOsPxcmeTJ7c=hK_jNdf|Ql#a?aDI zy@M*O9wRcx>ya^ILxMFZD2w_p`^Kgl3C;_oo|1J?=UfWd8&=H4=5IWnwe{FsH)TLh z*+l+z8tcXc5b98??Hx9ly-9o)Sk_z zxq)!#`?X!(sA8W=Fh(vXx-{9}j+;K)@#KO+QYK_5#Yl83A z%R;VtS+Yaq^`aU#0MZorN@^7*;}Hw)^C=IYYY{!i`l7|$y6}#%q9EFCMNabS967jdL zk8F=kf)blrSq!Qc7GU~5csapn_M^Lrjm&dV;KY(!48lhl46g>;{oT@SH)IU(la%5))X#HDHQ)QiQUXRl*D}*S#?X4xrOy zh9_nJVK%nze&w98aHi;VN7fd*k3DVV-sT^>PH%r%_wl?et8+v6 zSVm~DPSTZQ`&+do3ed;G6II{mXrQw%`vW+-;bPz>)DApq&>3vjYP(V*~L9a#;2qCoA*tHsNM*3kj2}30P-junT&-Ue|r;^|R)9fU}Q9NFf z-KWrp;SmX{jWZ>E5HRV|^%c33V7*-6w28A)2W=w36lvY@)d@P^ExG`W(W$MeT)-HH zQsM3xsf=5V7uD~B;O7jJy0&vakqA;o=kKL8ffORT2g?F?Z-lxC-XYO6?}LO2jULx` z)!x1V-TCz7R!w#mrQGSxHg}u9ctbHGbBP%g$TR~I5AHzKi>b&hJ4gF|rKzlNGpXU= zR04#=G!yE8w)v$ToL1S}@`TAn*jY+l)t3MM zT-L)zdAjm4lN3L9Qe&$ zLqYY1(%uOHl~kUTLCU^4IN&7F_VWXcTdXW&?FBYs=U|2uC}<96pFL%g1h2C-2*{Nx{S;Y9FsuxH!lzCM+(l?z}kKTCiU!5l7FBf!T-}jT57yrPsH&@cDW;zmBvfL>h7pDPv+Gwuv}k}4cP;JD z@30|1_NdkESsDfhj9a2Bcldeirr$STREuIxh7Ih6NP!X4_lrXspxuNzpkxL%YymCZYUb8(%>}MMykN*ay%z8uxd$>lb zQWJ1z8iC3OKV9OXTGSX#_=zeasG(beJlpT5wKftdtzD3AJ6LPdu4L`Hu(}76*eV0s zWrP%j1T|Q(*cBAxuTHbeXB;fl_4ir3F`st4BKom-mCyk3d@}0rPC5uUBH($<5+w4= z0jRnYbtCrrnwMnO4b_3_Kqa6r@V?-jXMyg9QJ8s|1*_tTFp}Ob963ew_=r178JM-A zIH`faK6A9==aoDuY~Q+C@9atacz1n9V6j~L``KRw2 z9%(Kr@hmBAv51pJ+}D17d_TZ(<|pOvNb7K!#OV7lKNTQZ6O~<#0^kKLW1&U>gY$sP zx85~?>m&zkLw`(FWAyNH;gs3%Yv_1|(#T0Oow8VL5(ODa1%RY;qWxZig6k!&M58 zb(UukE4xGI8(TdD1;ITS;G2m`VJu!rKDOFR)H!W=bGc$SMD6xU*`*+azpQS3vZ+Ax znT%>xYgMv`NYDSnBF^dO?^rl4C|!qjs0-qZ2)8#;Tit3@U-RJ%Bp>f}VSRL8y-v1HqhR0it^$Er zMV;R$)v(r#nZR49Bxk- zzT<%FjSjgLlJwx#^784S=Vdi=CW5;)HQaw#oWqV7vFMnYEpqF2`4pJ1b4tKCbKC_p7;=1MX_dA| z<#>0gk}f+^FL_^$?agpHi!*N@qVw2;A6^UBddp0TQxAP^#sz)B3rs1`O+$?EV~KN4=LeX!Nc3E$>%&M5H|2`AN8sU4w9=I^qd4 zC`d!R=5Rb#I#RD)2!<>N2^T_ajUQ`Z+1N6lYo?rlHdKm(O~!lH?VV-hmLgWu9MG&I zLCm0GT(j2jrmktjLhNW32?xmfBXl>Wf1$g=7K_CdY`R^i0)!McDn|bUEXZ;oZG^q` zGhT?e-Pet5B@Sv);;+g~qNpW6W2F28veUSJ)ww^EM%*(oTbQ!DjG{C zP;LDQBPL42)S}UOg{ORVqm9+Gze;1*c)ct51v~;b!h{*fVH73YV7)3A=ZSMm!JHv0 zgt5LrfetFF2nw?3mUp)t(5jKq=F?$IczMCQmUXfuXQ`9JMd6Ew#bQ0kF1-G5c23<4 zlbL$`&F0_Xk%dxw#e*;%|ueeH34AxNRFjzq?6}&BmjAY264T%66lN z+3<;}mXf>eUW3O2VBMFmD?b`mHGVF#+Ms(nCgc(Hyn1X0wfKAATE*Je+}y)P(spG) zr68Lv7@5H(21~iC*W0?Iu8QEA1*-IffIF1Y$@(g~f}r~)yNX-e;`0C-;03<{?cU(( zQo7@;zLSNnuewkeNdO`|Q_5H%F<8zSD?Fh@FU;xj=G^!VfM&S3aK5 z_uCE3mce+F)?YQh`&Wy4N;wTn^%O+jYc*6xa84VwV`XDcoY(%G6&Dwv;foU&rDN^i zkf)4X%}5Rvq3v68v6$!|rWIdI`7T^Mgix<@l~3dc*=vWK8)$vefQr#njFI6PlgXNZ z&`H~aBw)ZM&SQyrUMw`|XNrcC5EFX7hT$83Q-=KL7o&KZ*$8J=;s)*ZGNVv>Wvmdw z9Ewc^A@*K;Im8=?5A@5T2|OJsqYXtC;by(TS>fcsWL>q3lJ2_)LW(p@+g8LGPK>(4 ziI4ygYR(icftZpM1z8dzU|*Z00y&>2k0ZY2cYCYFe}LS>ehW%t^KGQ}yd1?7zpE&W z4JN~TEQ=Www<=oh3-CeI7Ego27Ngp}c#~Fu?;^Zc@_rA$zTPxB1GNsj$4z? z0j&xBNc3)gg_R#B9QUS%IckxLbU7}OulxjJ*XBEkycEjRGR4p;@W}8KzGwJ^sHeCD zfsA;Y#$Al8L^keR?%%E^jyq@rV+WYuPII_$1?g{a7@Na|CFURK`nubMT&s1Z>u-7o ziZOxx5a-|u-uc&! zda;3O)aUEj)kBz|2zP%Fo=gE+B{)BQkBZD36s0*stUs$XE%aRYqY|;b;jpjaq;Wib zQLMaOWCe=zi*{ZQQOfDDSEWqbv;U42+tG$p(_IJ6tV%n50U>+7y$h|f>9I89+)ns_ z#dekP@M;JxAR?Pgtl_JJ6~ zlgpNPi=&|Jh#cTR-@&Dd`$jcoXw?=!l)CK9QEzAXL5;nhjIPg*BSA#UK6RjJPr{z7e{MG4SmMve7V&2LD zF8uYx807SH?{PwJH}W$l3|5IS+!#Qgk zTLS)iYmU+8_Z~!$)ZySn7l_E4trMu!#4aXVQ*TFf=K|%Yh%p;#o@k5p_b`?M@7K~! zh{;VF7mEya|FRe1sUoPQm>%`T7*wqEW#4N{GuYm6oQ2nJJ6&j0hS}_`nTBl?on{Os zX9v%6S|+?kyZ-@*X|>I&>ATHR2kIFU&v$OCU-w@0Z|>3Sp$N9i$7a`TnTw)4KGg}F zm`ISOUHd#PSFFXcq3c8(<45-L#Ik|@hD*G7_qQ;#6=mPyUmPU^ylDIqoA(+_fZ=gP zmw60g>9!y*k7tfqAoOECns5Wvx_KWVkawHp5U%VZ;B6k0K6tlMV+HykOqBM0p2>zn z+u0+WIC5dm0d6N_Ii-PV02rI=3?}qn2b@U%T0lQt_%svYkJ9deG%gN8qqMm|6-a)R zKUb%9+FKZ#Lo1pt^}J9Q$8v36HAAQ`jkolUe1{J0z$brJ!#_qhRcRH4ELXym^E zYDj#K&9Rpe-rmx^Qix2@MX+OZ%SPPn?BzxnlM)})t4Gga^ewT%AK_cy;<`b(Zv`>y zbtqm0zfrb8Z!<*U$U?z{mmT`bkn8}ir=IRR1AEQGRfR=5-DVEla(J`)dN%AMmK{}mVvBGW*5r?_w%WnM;+&`AfYzNs4cJI>vJj01Ym0ZW z*^&^c@F)3bnpAcjkyDFQEEP*8*A{66IElJI1Tex03cV+g#VVCU0C2K-&+#Wxk-TN7 z%z0zid1})zL{7$ZSr|HM9MwSRoc)*R&f!fpL3HGf(S^Z7HR?J9|GyO7+z&K0L_k2= zaBhMT5m>&Jv#JB<5u*(XWM2ALyEoS#>!@aC>eSQ?6>_E{<3rtSn&YpkDN|j%r4581 z;M0Ds9sHHLtu}1*wmh6AVSrCyM4`AOz#~-9OFT4*5i2VzbS!w}=Zw33DABpXNuO;i zkS1*y?EK_idcYR=iXaR6#9tee>bCpD-FNqFffQL9tIF8$SDFHIk|9EXYQOz>tT&OS z2jen4maeOu3b9^sd|Z-kYL_w&s)y~t<-1CqhM^tPt+WtSACfalE@RSfVvFmpb457W zIx&3?fO2EE+}07Piq(UNH@;$wtpav4OW&K7y;d+IS#6oH0^f4yeDP%IkwTNsq5FvC zT>ZD^Q0+nko&~ek zXPFN*Y=$9N5(PhEY+rmF?_2Nt-@f5`V3+>b|VNCM?hHn^N5;5kA|_dhuaB^(XUWj8QTcvQiL=5RO9(a2iu~r#eCpO)`2K6IYPp zHg&mDlp!Zz7`DqmywKt?N;tGJRR4Y7Hm-1=d|Zt`yPDhZTdR%Y0V;7_X8ij0gD;CU zQVx#bg#p%d9Xnz2hTDuZyJGm>tHXgRC{!=4!&sA!k96Rbjk-Fypj^R3VBE}kyISf0 zbP^%E76`klBF3Cr2v8XR5!9fJLnC75<3Pypj9|d>VPz5Q^wY8Pp!GW6_8YbCqItT~ zolR_KrSnZpU@g}kUgnz5Ppe6Kou&JuUEYqa^sLA(f~?F2dV6sKfyT80ACJE(k-Iq2 zn(&f&S<-6Cp;{XT=(fi3j*Ob|R?k(VX}bZl;m)9F{jqXwUGi|$JVC$rcZHpWgEf7H zcynaF>QPP}%DzN%L||ikQ^MP6UiJ%u>dFT*dQ4q+#CFz>cS-6U)gj_xoL1glhqbBZ1eX5L=B=ePUxaI6KxeKt-lek ziIx3Q!(;ed(_j&}OzHGy-qqQvQ*&49OHKY>t%8amFydp@C89yaS?L?2kI~X?AbqxE z;Ci!%G z#)aWGj|;*-_jmki@7$3x1X@mVtlvWDT1ZZ7He5b3_lhcnBHTd`_XaQPuI|#ZjC>v}-b{H^ zq~&3nFv#Rv+`d~;iz=I0z1MyCuuevH>xPzjj(DCv{Kx(uBJM`0YWrNZp<>M z`Q&7fq{RsXuoi0^mhDQEeh`%gK$i}nnf2e?WXy!3GH%SOdkA->YYnzP$>byTo5XOK zpU)F|*V1Jz0K%`=R4xn*i_FmW8h#fB=v$T)V8}+-_j<(dW#%u4o8H%8)*BIxPJVsA zh_eqKqFAJFMrz;^@LnCbG)gKs`*Qz}qT+nAYm? zslJ>S3W{0}?7uOPfpzIhId8I<38QA#i-r4>^{4z90-^W?7-xUmE{xTc*O7Ifq!+(h|-iJTwHBsRGq4FNz z1N2sqsH=j00>shMy@=?tPoHyvE)Jc|ia0IV@Xeo-Q;yN1BB(At{JCodt)E4Y`0t;P ztFk@igv=y!QhDdBx}MVmu6x#r4fBF7l>gb-E-&;N?z~YI>>cm%XlufxCa_9gI#8tD zd}JPdge&)y+V9@G9``t4jY?n65+A1yf_e9E!1yoq#cnS~frKF-kJT>Ole?mAiz}Uj z6$ySEehDq?phW`UPjeff`+675KXFaqiJPDBR5YlL_H{lG3~os?`3F!VDsBtCB)ewq z2G`^?4BDJjyDqc9?PII$*8?%&_?Wb%WZg^H+A(IofGrdBuVXptxS3l_>Y8^5cR%G@FUia^lm^3)r$jFfT=*@aACg3R=`MUK~ zjlF6#>1W15%P8^L8W``NoVOb`RI1D z<3;GScRcjyx%ISnv_i5Y{wCPELe9^%u(L_abhU!B(nNf=EvVPK>>r>h5bl67(=F7$ zd18JR0==g0ctJhwm3gY04n4Xt?|1>9_S!tvt-$Z)cUttqweE1TsA;3U{woBew7=T; zd-}2byz2Z^AP~nKE?)`zPriZ+2Vo*0!9T$Nk+1v@UL^weV-XE^gLMDQ`NeNIUM2kh z;8o!KmH+T6kA#B&EkAgmNp9wBgpdB_-PvYu=*!dpvYFv{E9wq6U(x$tU-`edpt%97 z-)R5rkvbb*3%7vU^ixKe8dj)st*M7S)%3ujEiU1hTB4J9WF<NKo>b%p!fP-+v#i=2f0vEDUl&8E2XZ`^|0UB_K+RQI*@@tS_Ie zrP!aoNV80!T^_kTSqQ-+M>BR)HA}};N(_V?bo*T!4D%W=@T_G384poqud3;)4crv^ z6QA}L0ukt^3J^x{YNrDV-?-5@9NuH+OLyGIUoc6V@5z`%`#yI>2swHiSt$oDDHI~Ul;faI7%%N(+J4OqTx&@wcs}>RRKp1{ zwH?k7obBP{s>;fpZTA*X2k#{;o?RQG?SYKHla5@uIJWknF?2Dye{?EN_Y_sM4WGjJ zlr~2sBX~JdT&0lg<~O$|b)GFIPEtujTqUo(gD}qLAI?Di;sEHgKH%V!K!LuasQ2JS zMr-|4aK;vz487J~`&GA471hZfZA#aH>My$s zhn?W|Hzbb7J2-dz&cpS!TEW`D2HVW1AAS01xubn)3F9&DZ(8Z~r`ZR*C~oq==gEF_ z3LA-jN`vTRqHLKg0?EamU;R9C8U+yTrG4*5l&_6tqKuzSo96f)Mm%a~GhQt;H9!6X ze1Fs0akKCRV!haoc`v7C8io*?&eEKWG@`%fu}+9XHi2Q2#Rm7utNCjZ@$jPY-!B#S zCF=9ENrq0{%^n<;DleM2A?Fm{-AFw!QrUK;He|zv(Z#F#$#--ZnZT~qgr*HV=P8tf z(lYC+z6-9m98+n97MP|h(5|oe+^4FFh`y|;qldl^scrqVx=t$cltu|}3}VzA7AtGD z+Ymwvw+rZbq18o@eR8p}rVLuz5}hA9aZ6U1z!MPqlB;?kum;m_3~tK z$R`7+hy!$4NFZYZG0Ye~#8GwnYBl5g#pqmA9X5a|k|7GW@81p<(+9bs{Akq9zo0~O zxN$&wv*86zo`A% zlQ6JGuTfQ(BPm#=f07TgPd&kVDVBmMee&TwowRY=zhof)VY6+#1<9UKLU4Aet;(YW z&JaAbDofJE+e5f5y+e;(ky&`FglUB|tCq@Jbk z{kE-^>#+Ybu5zkZfLwN+k;iMVM9wU>SrHapf#iSe4rVEFM_#g}_3apoIA@`-j-}rJ zl>1b_p;JaEG5)B~yscZb8#-|H`A9eKo0x(yYklKSb=3O~8uU;{#w}S0LGaLlezzXr zKK8flK!h|SUcB-m)+}zB!5?MCR^hL6q*Y_==VKBl@6RQ+^UX^alFLD#;RvO}u&@`+ zQ#f_y1&#of>V5qNk7M)QWQdd5=Qj+z{fWh?Q#=g~IrxMWyY-H-!(PiqPxyY;eeiaSBClIhIp(*`p7Yqa_r#U zUWE_Ef8GDT+oCTy07|4e2Ojy}11e*dDgq^Ike{j_mnQ-toxxZB~A^b!^H+!l96T386CxLO|tP3(ZQ2)xcs(MT(^(lRJ)pVQ7-d5(7?3O#7 zUz{c(zv?s~@`geh+c#!K!#X{ zS=%Kkh*pWzc^d`^Cm0r3CX;~(hrq-J6_hx8bEkc+l)Jr`KFn_L__y7Txav4Nm*=l` zL~cA{`DT^nb>!)NBjN3DLab>r%z7iB;Oh-US&0SsnW+S~IuSBLlIk;0uLLT#a%VA1u2Y&2lK<@Us%|WF zJ5!mcI(2R~v!6L^K#77EwGW4KW)C6#g6KaYGqV{99>Sy3?E3~I2lq{TU&2&a*ULj> zQCxC*iV;-KDj?nfB>{@W=1s|DzE8J0dDJpzn6|S1=cK-MUlAi(3+pnjbipSa8^sG; zf=ohP827{@b_J@bi`T@O_%KANbHsc50ld z((l34k8wjrRgqXau=}Ur0Zf)ZqFTX<0}q4Q(8Z#C38Mb+it<`T#3GQAJZqLqNpUoj zCmGil6$&!e3GZh0!yADPm9LzkMxi%3*VF z8B7X(@pE18^gYQpVjcV4pBKoJS-+yOnU(}IF*5YppLPiM^(A zY$_4Lp%JkHAl{*Od)t8)T(a8!0*_Q^&U*{oN(>e`N_xkMg*N&@#?o&-#%I*H6{qEj z-#&BrvCm)Qlnrd)nhDfW**)#UO|z)aN?xwRI#Tbn&c20In{qokA^a$=c^=sltcg!Z z_*8M@!HK^>AH?S(S~f(on!|cEQ=WuH&!AG51GB^a==eSyNPAF*qp@}*0AE7#54&|@ z@>Z4IjYIOYuR>d!w12%bRO`#D=p%ahw0g&k85ot!<*!Hp_XvKQvu zNbxzS=)QbFWRN}dwq#9q-guy}5 zGkLbhcQVhFJDN-mWI;Fm$+UeI(?y$yo$wa|8AC?&=he4|x#xpd?|%T%uzUaphA7-V zdaF=>sgpzyHznM`=GrrP>0L#PQG;qwP&IfGBjqIJ;=G%63(Z$WrvM|48V!iz3#=G@SVXjrPk}M+Q z_%;;@e!IES+k7V2fka^-Ki$PFoGTzcK6jHGrDEVvO5+;MUx6-yY&EWY zR#s*pckxUguTHxbm41&#IKRmMMIGB34N%ec6*!~NwV0>gyfF?QeU*D- zSfW|!C>2cYGxLbcJ?lYk7P?S+_6x4k`$oqBc*xBz6kPHK+d4>G=;-I#t+OVK|A*6I-~ki3|)=Pm57Ja{S#?bFYnM%{s3QBKcdV-~-YiYeXlN+^AlrnrWY zcptlL%j=AS=q}5{<(^cyDtCA_kr9<8!}MDx!v56PsGEjnp!HhM9brDLQWxpo9KvfM zp1)*zpW!G2q9@+lCnc@1&!-sT1BSY`AKyv0MMOCsI64Mow>Hm{{mwmBR}!N9 zn&}=$&)hP!vTVk*4T?(D_d?G^yxn`CHl~SwS?nuaoPZwA&7ZkD{i+_F0PFrvzdlqu4Q2O4tiJO?$4!VwW zfCctj`CjCG*9Rqnj^Zl0H=AYp3#L)jm7p~22*O(i>_|R}pQhCE9T=X58jbq?s7hXx z=AC3Jd{)1U(dBR))@kT+4yjNmRjB^f1u&S0t#7cKcbc+2>B$XH3dV(`5M*Libau=d ze*()(l4Rhh&{T6GS=k@`Tuw3-SG})ivyA)!?jqky6Gp6|dU{!xgq?u_Nx%D!WsC=T zIvXHp>vGd`LT`UgF!bOOMdE1>kUk@tA~<9T!O%kKumB&u+7=O8fb*>3Xeb;@b8`D3 z#F_9HpMXC<1A$lw^7P(A66rT?`mN2wUktdl^L4gu8s-O-r^dl-RMipKhSLyG#hu2V zTqWBUPZFLT5UpprV;fjg)0l|sbt%l>cslV6?G&npqv+B7vao4t^g>=NfBVRoDSNNJ zKr+s>s2YDI;d)8in10h{dTzz5fC(Bq-90sq3jwlwV#KzGMLS%i+~ppLcH zXANlfO=sZ~#Jxk^kkTrm>98EE?1HBd)CYfF#KXR&-dwzX#pGeiGnAm(-8;~;*(*f< z;cPEyY2IzXw*4+i)L{pO;O%3-ZT}mw1d(Qjx?PLCLi}yuu|{^EbqUf<{>pr^ae^J24U zxDJwK5FL2n&H7i1R$LS9S`j>! zJc%9Om828$POmcQTFU~obXVr)_9Xux#HD9>inDn%Ojn1`3Ial>A@i47`OAOPebm{e z4{75qoCFH1PLia{oGmX=xerC)_K!}`Cz-|y9hrqjyIhLv^i9T=(gBjBxx98^We7T!mvx_J?V5$GCGbn|^K;)${Dg zwFy{|cUH@@K!q%N;)}#|?-@^fN&_KG$R`BAmD;%~D_Z!ln41vSl0-Z98$Hbxdo_7L zp>>^f;p4Fqb)CO9+;e!IjXN9dZY`N?VV)(#lOa z%S6nBWUD`zn5*b{ZYKSQ>JAz}ESzmxLefbPMRqN%H)lDORzfF{4cPSo^5>hW>?!sh z-P!krru(VFTo*z0@cC@6L9-icqg5vRNdgK9{F^L2jDR7=`~Yv1a!nl3)X)AqADq5Eu>`bOIW zd%$O$+n?-(wYh427fpO+OuRtdf`PKgClvXsUDqiBBa&-9$F(vy)+-wPnZ76bWEOL; zeS{WF5LEAWtww>Y)o(?L^h$i^LG${4P12k>{>2h>&pFo|`l%RlD~>}myc5`ne6@33 zWuMianB@TkC_^0y9JA+=-vNNaRjJKfi!fa6Eu00HE9!eS>=8E}qnf&diQb|mu zygCJmLOvHPjE_6PO|QfJOgv#j3sp-gF@)5K&i4D|8?KThvE!ppZ=l>bW*yrnsYqcZ zM$n9Or$DfP$ViTK`EbmVkM9LhXP_8%0N1upLy~C?^ZbS*H<#MT`%W-O&9`E!I0M|L zQDD38th-9g2!J5-x608z@ZttlQy{!OAGln>A@Q080Mai{RJ}VcP$*FYRIVCH$}5mq zl|Fg2l|K@5;te0@%Z(Pf*)MahyqLRXKP3Tn`3%NR4@j0Gd^a$S$xqk^#J)7+Z?i}0 zPh1-4va>fSGLsW?8nnzH-$I&n-|Hzn#dRp|8Q{Vd;=`NLIZY<$y7GiMa>wJESBx>n zXjx!o^}`xJl=WKf7Orpx#G&oT0Hfef6G8PYQA(c!^6!0@NE^rasSEkPx$F+@GY!F_ z%+Sf)q~|-8s=q}weM^jEnASpW8Jp#UaseFgq`Ox~$IFbYX$vG6)@Xa#w3++nPIuPk zrfD5$^_uZ*7L9wPlUi5dlU{wC+=ur3iK&GK3a3fYks?u!t!lcOWJbheY*k@XMB`5v zYdhub_fh7J$Dv;k0(??4Ek#J(^6xBNqln&K?K?*?Z|52?&Sea(g;(o`KQted2xhWci6}i5&JwoUh&5`qc!V zQW{p!os74I-g}r5&LVxu{k`EFPByBy$jZA3zS4@oYBroCwW$nY^)@nGbxKs>Qrn$g zq<0_DE;kilqxfrJF0=T|b3*D81R&)ztaOq!4KD8*F6?T zRMYd5s4;VR5wC*`KRUGDhv)D6Ec+TjW>(~USkE-+bq50R33#=wHI{)L3nj%ys%Tq) z;|Pk5sz3Ur^05`de@f7kGd5Z6kN4ECL3|u85l3u&s#91pFo}*lN0ShOYVP$1BQy-$ zIuTeeNSG(+qeR&QW7aB?jTL2Y#V3aBpsK$B%T8%rwm~wMa|&;EjB}gy4_?AZGek3X zN^@B2c>e)Z`#wj31I(%!ncXTR%`57#r>0GjMkoFO%)*X-z62y%`XqVY8SAWx)Huh+ zL+PG4_Lrq$NJHL93-xZjEU2>Gu$FIQN(4SWRaV3l%KCa;><%;?I?Il?q9Z&^^A0w4 zYzA>mKt{$g$Eqc=VwW?<312>K?7uT$euf6Aa`=JDN4JaryM3Gb^bdgdj-<|yDraC| z9}W2M9y#$e0Vf@ER7VGYyTrJdFx#r9Hc9ca%X7JSkjY>LKfKLGq~t?%^DTPIR-a{a zDC`24R!e7ganr3b0&dQVhXPd;jyBB@52Qy!>Hq|9d5~)y56c>|<+MMm(DBGZNWIHQ zR0`}c^3FF*;(Zjy3{;yZWYinK#uj!Xzyz+7C8JNm6wPCnqfF+?GM)B_U3JQwdxjYk3$VGaF?fJKsIjvSY*40 z!``X24K)g^FEccyy%aIs5|NfC&hf|*Fl%T*}cDcJpIN0FoF!IKz13L1!f zINlE~H41FhZl89wN>yBTN>+*A#!o}7S79d?rW@-dj|km_{#UX3|J$R{RPAcUbR{@6 z@%W}~b9~Iar}04Tbz__ZR(wB@pJVVg~`J!C;6a6zdQ+5LBAeEuQu0- zc)?K5ERkAOtKAQ1K|h7ib9htIAckYl#+9f0mgHB@jmxT4ect`3Z`~ zQum2Ow#T9{=`@W)tgp$y(r>7+Z^-}e7ib?8n5jvkh;duw&iXyuh*-dKoky6>fvQAd!+{N1 z2WzW};&^BR%FE@!2$;s*AqrU*cf@CwpMaH>5AmdydR?>QO!F}>FsJ4gjtMi@WzWr- zTvsKzgJ#8Wsr?fZmZ*j})}-665#**~BU!hjApb~o7H+PFdq z@gKl9PHR%0BKxnVeG3^H7Fs{p z4}vt@Luas>x0aIJQm3%CC4iNcHjtGFD|EcMCRS_vQTX*W^H=?aa&E{K&HW4ZL>zy^ zA!16W+M8H8GYq`B6#}l*hAZky3xDa|#9YpUC)G?dF77_0s0G7j`fIWX&38Y4A5^P& z-C~8whkOn*xS79wD*TR8wUrRF(&LfPko5M2iu8ucNDSCPr9@4g}nB z%~k;WWq|MW%+Dg4AI3!%#JCFslM*k~y(6>1>d%4T8?I|HfF((kK6@_Fzy&33dSO#3 zwr~ZqHiz`u1f9lyNEpv!4AJ_yr5VA_>k!|Y%cJXc!nB7Myk)!3hM&%xA1C3D4{X4ZE_JK5^x&*wR!xX%D zY!W7Rkl*|8rqv<$?XHQ{vX_IcShgU-5tL~hECc9R<(?H;Wn(nL3}`3|lM8m}ajqpp zJ;xs37D*UjJmlDV+T`Tr30 zmQhjs@Bi=y-QA5ycc)7?(%t2vg>%BRXpkQAP-hjXA^tdE63N&#ZK5&RD*rZo1<;M<;5Hn&j;=`J9`t zMu%yCf3iDm?8Ot{8k+HDuRxk|w3Kr(1)rKO3wQR5+N2GQ!&eo{01JiX3@=5#x0C(W z0}QgH_3?Y#*FcYsESEXf*9=*~lpUgk(f!VLh9!SW=_{ni;zU(O zO`wuYxj77&s)F^64fVW|^Y9h`LEPdsqj$tO))b$dVYzpL(=X<33~mvb)M>^S+G~;V zeNdB3De3G!rYVZXv)o=NSFFw+NprHYihOLZ{6bSbZ0sQE%>}nK#ps;$NL=9L#KUcm z08KQCmDkLmXjaY~op{}7=Nbsl3)>}2n!^;h*TXTa5o!wjq6O$S2=lFHeWAoG&xcO3 z#2NHt#?MSukdpr`#$@f1`Yb0wNmaqy$M{ozd~6DCz3-o_(q>u{b}Gf8!s0#s1~de- zsh8-;s3{6Q?;OdX;IKBa=?sp#^mMF|AD-)SoUyM8#-UbRk@cH)GV)2_swqs!hHZ&x z@*^Z&gzDBgBYXt$n-uyp|GEX^FITUzegT8% zDZ&ufrt$e+G2)ec!MEJkqQf_YwVY%TSH!R7IrF%3jj#6-#7z1^-* zeDh;xE6K4+S&_(mJX~uBmDuthfXaOr%KABHUiA`lKPIJGx6zM|qM!v!gaZ|iaS>yz zZTTCNB^AEHHoJRPcgd3aHS925z~`!nAwgyzsG`3x zb`|HT5H)j~QdT*A0Pyw%MSfydJY@4N0WB^iMojD)t-U#LIV#dwe|fZL3jjrVzV>B* z!RQoMn10Gwuxh3n|9SL-X!AjmbE@&`3}URA=KRb+1Bqdtg6BlzTo2f{V9*~aHV?( zBOvN_)pn-O&^N^S5XQe?#qi4I+kIP`Q}7?sYDBuP7-pal+y^(@IV#&O=HJ+a=ClqEDN2p7cb0 zAF^H+Q}8I9pflPik>zMwFq->OK+36X433+eCqKPtss*93lxg-B8;C0N<%G*{sT3T+ zO6GukDX)eVku%aQb_XSSNcSn5>RA+~>;)LBLHqH;lyJ>qG#~An=AQ{ z!_~gjfxakWm8+MiuP;!2Z+wS(BI1uBV|O@YzW6to=C%3n4ho*USNYlrAc`A$&7aMk zM=@`<Z#75pBQ2~U#mrf~}U@VfS(aOZ2hZ)an#a>NU6X%JdtdO8vZmc3*< zwheZgm}JCOwuTb$N@)`2AFt8mzC9Av@&>lz8)a1Ico=;R{rI$;#>}no(WYKD{lglu zw-`4m6Fy~-Uv|Va+_uCC5&gxl3gAvNhF@%KC7^Sc9_ZzT!>bKXaweI-nihX5tNN}e zP+5~yWmoaik?6smJRCM-NZAN7o{zw;h{3?2B=trh^e86!ffj*7LIO+I+{@V{0(uVA z^U>|sP@w$nmxeKbd6y>5x>ndCI$0Oo0_Pg^R?_ei2bV)w`!z>;2`>Jt+ZCZ9L%nNx zVs~MazEuV`A<9a>(;{{a>FOnukC`l=!O{Bqs^2lod36D{v_opWUgv;A(YJDI~4jFA0qL$Hb!!SY~5=6VY zt=kP#T2KGJk;iBjd~IsLL?&GosC6(x-z4DsJ$P<`=HycpP}zy*b8S=4z-5JTsCHAkRZfG5-^x|$<5@R7{=5L) zZc+#(eV)F)lXJMyLJz^D-+5lGjh3cYzT@EkQb(%2U-0)|C4%Q8xNEea?C^E9#hEY{ zuL8~-#?A9Lhx;?W;KO2I&cMQU!Jn)T6sal5#UV>UhoXT&kw{kaPepOfENMBZI$R8r zuGW7Vh~KTeSBeZK*h6ibqNFoUOD@+Oa9=Fz5&LG0s3iA$**{UF{Ib2k2{U=M zg4r6+VyV;5ACCfaX@S6g(LDMtX?X`;(GVQ>s+uX51$GmY3D(!Hh=gR$hVy^a>?Xkh zNkJ>mm?_#lU=oPviWrx7Fol@g8M~nK2ZO5J>H?|ye3l`QwGJQY&wTks0`j1H&%5$i% znE>b*qX^Cj>+#Hb(c4em|B!wz(o4-tr)+x^#eO_5w%qwni_fL{|AU}k=Pug zrjhH9d$mn#E{YkB`rS{60{mHhjkZS*&+4P_Js(L=L~IsJxfXx(Y<$w{sb-xsFD+x~ zjh6kwMYQFk@k5$+Fol0bNXocAkFwHqSw-qb@e1Xj!~4~1X-z@=9+~6W-064WLy)j% zR0hz=XTB-lmGpKmYGS(3O5 z7OT{r)wW?u$7cM|rzM6x$ec!ZQ0Cjfp@vYNHu6GojdWqDRD~UM@+@6aof&NKWQ_0- zd1v8@?J)gYIXeKZ8X=EjP}N2kwaj?Atn73U7h~`nGoOL1*3Y5i0bak{VLkhh(%ahR zPm3zV1YiWVLjWxX0S4E;bUrcWLevC$=TAvLjJ(VvpoE587f6603 zwOzRMXD8J?7XjCwR$FS5LjNVXGe`=%;Ua}cs68rJ9Dm2EDr7>#VRr%mlyA5LO@?f+)Nauq zd^KCIL?)9g&O@@L_%>bj`A#+OsM6#HN%nW7U*XXzMM3{D_4cFvrUU5(cI{Yb!~Ab$pEJRX@_K4+iyb$BWXJTaA`R`T-RQ5iZ+^ffO(x6y799=YM~C|3Fhw9>i& z1HpxK>eL@T_1*qy+tD`8w;|kb(TU5cnrtx%KQ@Y~5bn@uUoSn_>H)7H$#)2e`<#~N z`shg>s(AvY)T;<<;5kY$|1;7i6eARZT&q_ytPhT+2=I{*#W;5e}g;(3ccWhkyO zr&?gS{&k36R1b<5P}iX9E8_ zKV`>7gP_-f80wnTEiC96m0S&f&h60N%vcW-#JgX-KqZ?%j_ZW8?@e8L;$47Ck6^-< z0?H2F)Y*qJ`mYU*20w&4f6E;eRckS~L%LF75`s3u>QuV-Tm+olR>jXqUH>8Qnwm!X zY*D>PuYhK?Z-m<9J8g$0x6NF^glh22d!QZ-pti%lWT8^_)IRU{@y5^`;u^x~k(rDa z*eO1zCYt zE$;0E_EVWBL>qa#xgiFlOZ@7^?5@y4^M@Hda|cVwsfP69!4D;m40QV(EZsO3yfY@I z{D$71k~cFZ?^}2pJUXAW#$@qBr$dfb8boedY>_{iB|7i39nWBAf{5ws;}4)VDW+dM z_7s5bXmE?|$(yP#_#B|R2C5LIb=EzGOVm7WbzX|af#=uSJ0cG7Yhz+ha6}gzKt=_= zk+!c`pl`JyK=hcO@Pj@NG>+xK4)Z9HnZ7Gi%JLIB4eHc$<>Ly#RpKb6n+iR$0Mofk8yT69Gw-zhwHK zCamkW-%kqFdxIA86SO=(h^HRVkxIxx#F%4VJm_mEquL_e@QI>`f1?z&h3iv*MV&FW)+?w z(=$FgsjtU0lQBpymT_38jU}1V+9!oh+GLNGNI!ZaVVd>*ex9Ql(Xh{K8`I%J<=dGQ zzlR5Dn0rO_EP@Eq3|`ENV;jv1GO=aDYn8`xafcO7v%!b`*mKR8B^S!-O_jCXyu0$a&{->Tu{tx+%% z5vXnKoAe?nB-^~1--W9-wBrm3x@%Www9jq(dqjWlQguE|F4z+6Sa>FJOa#uUnuSXJ zJ_OTR%;YMS4&>-K`w11ECnjj|Bz4A>O@u8_nhctJwD0zne4ZP@PAguoD z5c7>I;**M59fp+LA*Q~$zcSC3HJpo3}HBg7pNN*n#kHy;*ieiP)x?fwKB*tXeZI=h5xb0O62~*euMrT6LR) z(mQcbG?cx2f=UsKSOb?zo_|>`7V}&B$McQRL2i;$ju__q5jO8;9)TH3Q7?$%FTAG_ zrI1XezjEFenCo9v$3*1bh5ms<0LYhTnZpX7A>v<}(r{F4+A@z#&7n`RL+N z*=Z^)n%-4;Runu1`tUmq5pF@naFsohAhzO4PrG8J@7t^9DQjy|1pg$o>E^wFttO+W zKjAy+4D<#NJZvIKu~p!ngydCJB2bswEZQNlwA z2N-fYQtE)Mz*#EfU!mqTJvH{}#9G$GOHtlDw18YN)k-|jAN(4b1+v=m3KsI@@|d*w z5$6ADSNT(6P^og=bZfE7E5rtFkRJC>gEie_jCJM&oi*A&Z&97dxxi1!<4e`as44(J z*H*59ky64>%ktTg^jK5H7S>X;G)|m2sK1>ZwlOA#JGdQY-F5&kwva9J1 z5%)A^UiI+riuP};V|CDv$MX+HZ}Y2U|9uyNi#q@aO5@A-dN13T>5Z{39K=m*0{3bE zs;JSy53}%`&7hr0dPh@HhdA0Vn-EeeTU~zBD~E1>2PV6eE{p917}C?k^SEsJhbpr* zXv)e<5tUzW(VmerKXXi$U{r)gMOMcu~iCk8tUybDzfVBS6*{ z{hH1HURd8Lf?) z`-EtV$`rwbLRowAhI=`f^M$oq=No~*mvBZRe7`D)aw->Lgx9D_wkTPXHSbmETJ|=p2C(5Q1%N`&R z&DK_v$d!t8-=(YUIqk)}R)CSURh)6j@LoGu`6x%~A>wrfln%4%Ekl<_wXBe|*1rI2 zD)G`JvFb2vqT0UI>FE%8%@h&j{K+vOy2v&qeW5W6-1)UIJ!W;G*Vj<%sY>x7T(xCJ zr2Z@ zIG3A539Qul7y4q=G4zw&$o}hmp^WJiY_{$WGwC$>4}kWsK3}{+%G2f9j(q)8sXSuk zMLNb2C$#mE`9k(XddC=mj4bT^TcwK3)6E4|Vd!S#B)a2jJvRB5R^nSKuc;?H31U0U zk|+}xRnOP>*)fMr6UyZVGW=;MU1C4)aU{q>u@3W;cOsy6+d9Qie`rP| zi2o9@Jk#Kts-DnWExcBdrBOIXl2>|it-sIPVk1Z^axfv~Y=C!W9zR2e=Ea=2v`gax z5A7qMr(qB37EOTDRdUUY@zgTkxk|3xjgfLJDchwQ5xH#e;*r1EL$g&WZ2OY#6$Ob{ ztsP3m7YSSU2D@@~%C2OR{;Pp=0JdV~YjoD@OxsV=yK2AYzSq=dEqYH_b%c9ThoGe; z7*z}`XsyKTG3nNF-J{i$hG6*`->8wvw>g9pXR-XXEUma8mQAU-e%)tTZ zn3bco^CIQ<<%v2v?ZBRk)f>7-hxrO=y=*f58^7nY0@Jyj-z)k&8dufjo{zj0t=A=d zF`IZSc&>$69y(4QcyyU?Yx}!*rF@HTmO@8YY@02Lo&0TI3OyGc>vADIMter?{iRVk z6;wF*X~ie3rB`%%i-tO@ovx#HqKW6W{{evLHC^>`VGE(lg6s4Mu422g(#v_%cD|Wc z-`^-pVVhh+P)T0+MEw@K&Hu84HQ_Ay3n#%z$hn^2=4Dx_fS{HtC{2VT*J=q`7m@T!L@SUM@)e2wn~#ZVy1<_o zZvZOx7k?6?m?|I!gDZ1MYUhmG?-d{V1aC67Z2aFSW$<1{=K+*m8LLE0AAS^ooq}Lr z=c+kBOxlD|$k~sr^BPT;dn;x#$=QNqFQ1CMFlrMV`tf!;yRV5_JH-3J=&98Aa7&~o zf#Ui1A7=&I50U1Q9jX%)x!YgEl_!&}hcu(cEq^;~a0af-HgBGB+h-ZCuRjCwDCkaU z$NA0;ri?;I{WCapQWl3~0L5>-H)k!?(flVGVj6i4DoCrl2Pz3hyjKch_c`*17Lf$n zEtuRCVR{o+N_3)EjQIQju}+_wO4`>rD0xI(VT1>^PjzpZ9;*9_`MxF2v%Z7o30OE@ zsH+;Q&2nj3_=4O@QtT_`yaL%H6?f-;p(XazGt2_IHCi)ARu_JEV`NDQ93Qh7IvV{|f+LN4k>H_LG^tZP$W>p}g+G(>+43s^ zAOut0T-=!ZY%Ehby9}a|T9@HuiwV?)DOc$+TNl({9uZ%cFQ$vs2Y+63ISYye-GO&F zcp>{yky^bVfkCggR$9YvKGCm>#D$jQ-T#^uD=RIpt%6xA%GqmB=$-;h;>|qbguj(d zyI30{vd_=oZ<*E+C*b0LrgJ zP@vC}cr*qoGVtb>HbsMQM$BBavi$2>Z(vxNQo|FY;GHQ4d7twup{>siRhvgXnWVyf zGu#EIz2T)xHA@E;6k;^QZsb3(`a{1$8g)I%-2d1p8RA@Obz)Yp6WWJYhLb)UJbXR= zI^;jzjpR?Z{^)n}jmJYuXgsQ8bS|F9{^Os3mq~a3blsZR>l8OOvl+~~Tf=@2rp5%V zLGPnM$8WB9mU&`*zYq9+8hYF*^!iqoAJfs%5fdVzDK$ZKOP=DFjnqLir8csyYQb}h z(9I1nX8YbJn6mcD00;ibEnir_s4MGlUKCb{jSr{mniu#sSghquaL7+XNVhf$A5x3OIbok zeU86eYmXS7v@|tS8YFJOzk`ip#jVNTCZREr;m1y4bofjr;$zRpcqIiZwaJ=xgz1##`YI?Jaiyz|+EOj6ps1CCUs9 z3>!?ZGBH0xRnn*RLoK62h1RM^Dw!mATP_2(ajJx6S+ZCuyw02pQ@a1d3vW2cq2fvE z5d}YOG;G#P{fHOC>&bxmR%+Mxag6vWZ8yf z1ztAvVVU=`Y$Nlr2pJO9$GKnIEtrNnDqI^b)3I1>(J^KA&ADNA^SKLt+Is;UCj!5fPG$%xP1w%D+e|(8QI(|kp;H9xkeC!x ziJ>A$BC)vUoQ)sf7yd6Z{Ivf;7eOP2i9d)iQQgkU1>(|YE@@^mN44niEOWW zi-dTF6OAOJ9SGezd>5nt{~r;Z)_01LOLf?!A*eY-9lZ0gttk9oeJz&Ui1x{DB9fxm zG^$y8RWyqwMeuGr`p|u1hUwgn39d34h-fLwcCGKT_t+@27=g)UFF$c0r4)#~1ENV9 zw;x_Y>X+7))dWzso)0hcaQwQR3qQ>B?#CbEJL#lse$2)7{}14;fAWXoKFDntZaD{T zklZNwnmk8iUPQf!w*WN&9sUQ90Q@fI?>8CEYAE;SSA;=6-)0+4$gpUYxbo{j1BO_DFJS=^dY*&h1$*k7boYf-Lq%K$ zgP&Plh@7H~wqQ>~O8G%mptG2pMs4V-(w*xDVHm+UnGN>MiZUgv`ENYvu*lJ3KU}pj z*t?;%nmZBcfLEhZ!p!+?w6L#Plrp)HRm5)-RT3@PHggw{q#kc2G}x&4jPB=-(qo2~ zv^9ceJbF`G(O+o?$2cN&8JOQHTzn1Dm`~oqo{MZp%46qf&`lQBWWvdea&O* zC9vpo#hm)WUjG3WItKzsIV;w2s;v2-x*p#_s=4LICroiz^;>y%5?2?=*{>H^q(^y1 zqj9DER~z?R{GTX1$<9h+R;hFd;80;sE`rvWR&R|WqF-uMx^8x&(nV_H@W!;8Xw*1! zC<02;QxuWHT{NW9w?1Q3XyuFxA;^HVp>Y7~`tU~a0wFdDiXGB&^Iq3Zq+lT}hcu{m z`Vqj&9Pl9O5aG&}?-r-Ua}#Ku``lD2%j{JRs8jBH{LEk)d{-N_QE|nYRvL&66Q#4; zyY;9M^g1#I=p7~i=FPFvP1W0mvl$%b8M0y#U9+XXcXxihZzsrW1eV$# zRcZ!N^IS1T%DVBm19Js{^2b4qGYo6Po2!6x10ys05uKvS{s|!_4tzp`0pk)N?d%~+ zuX{f^2Z5DyySsMa0nmYMNcK;bOEi9~#g1+aJx-|#RkGV~07C4MEQ69a5NxH-NFI00 z8wW27#E*eM8I`Fd_2i+)3#=5l)wY-|A$@M-Ktvc0>irw@wu^Vbeg-VTbD6*nM+#gm zcS|<7_~y(~=-92Sc);!BtVYS5}O?8n(P9OBC$hF0KAN>m}u#zwXC9QU;DepvaZJCxlK`L=Yv9 zQ*aBZP63*4LHscn*E3Z2a-s%&qwA0!%oU7(wF6e|v3gKYU3EBB$lA4~s`sb5c+a&Y zLKpA@Tgr*6(+=*Jn*3nfC;Mu^0wI(wns2K|oEKF@BV#yh4zd_{i*K_hnSQB z>Alp`g4(A#spocme*x>b_9?{>7qF&H=l;zHz$%6TH;3(b0k-(F%Diq%0e#d8 z@M=5{64l!ckE$=#&X$y@0G;4`{R(CEj>Jx59~O^+DKIFuuV~|V zWRCT8DH_-6u{chQJO1!@aqG+PaD<4gd?Q(LFmOJpoMLM=QF<*nu7-1%TTOZ)787!d zrVWIRb{P=&Yfhd~(Rj*B<-`qPlNy3 zsppM2URDG-ptjOt^JohEc}aa4?}zM&Bm4Ga zu{BiWMC)IqqyATK3HWvvi%AidyTIFPoJ1mJhfi+KHy5Urj=K8s6rx@i0K~_SoDF}e zCD*42m5l3h6iO*typu&c?c~F^^Ton|^u#qoi6dB^ie>2MW9g9N5&uACD*pj;(J6Ut z_L3=tdMyi!O{p}@Waik(m*+0^NDTKF?)09wZhxnY@ru@2Bs<25?z=G(`6>@%>df<&l5LyaPA7@;r3sGSx<-l8y(36P6EP=#U`v);%@PM4*;t$j(v%g?D< zmR4d5WPRj7_jP29@kL z;k1Rl8=VM?rLp>8)3qpnKJHYS`X4~O_6%HYl*28$@#g4(aQrX^k1TO9i%1AAR!z%? z@N8K#T?cNazwl3=4cN6O_CH41Y*GJk?jYc+%RbsW&ZJZ6?M}Hz{JQOuElht?Vm&r1ab8VJpfWr?p&_}ELh%jgZoE4iXh@pyS81Ui0a_S{B-3}e$&GWVC zMdvSQa2oe)^;8HIdW?~BGj}vOW($){V&K1(&dKC;mM2 zjNKQRKPf13A3sve-OON1_K^j@CvF7g&R20kg!E+{e@{JM*oNfrm_t8>4d${5n48182RuT@4+-p=a8QjNcb3!m z()<-ZtED? z^nDS{sbv0I9o-5g%;ki|A5vY&3{OPqD-}l4jq~uXZ=JrZ?Kh~Jn+(-tkXonT_Eu=T z-_2uP!Dp@EO8Wh(WJz``>B1_FYsKdWE`_-T`A45#hXc7mpl6}7YWd@B%FsTnt#>w7 zn!(u)dHgKVe?*)~1#^_Ck4-HgjK${zG=_3IF;D-&(g%Mpb0cNW_To4nFmmH%bFiM+^YJW07@OZL{JK1cMrj73 z4+dA9muhiIp+2X6#siD6(CXG@W8^vW8KoiCstOd%L{(Scde0hh$bTARqQeilOgH6O z`FDUpA;RCGOC1FflxqK_{{UEhU@@XpaJ?dylYj`5N>bAEohskM0UTni#lS8+&y@Un z3kUNT5}k)fn4Q(dSk<2*pZbcBBke0`jU*zo81_xE>(Tg6E=+}#UJ=R+Y1lD4e+yKG z2}=%hfEbMu;KEtn-QV>xuzvSU-1{)o447?@1N$1oFQ?HZuAO5X;;%B0rsev#!i zU1ql!11%C*AqNUc)9#1^VNoIPWl4**F^C)9(r25V_9Oxmp^fB$|dGLx(l# z%p_>NcWALp>Awr#;01D|+4h+33g@%0h<^>@C#1@w#`Awoy{P>Sw>H#?pPl#y;}XSVp*h0u(^cF(0u-#JG?e>Q zxS1vOm98F?FaZ$$a@i2yK{wi-0v*liB>UW^w`OoM1vp3l>XD+mEWqfF$?}8@;opiW z>a|SW3{;)tOW3@N! zViZfp#m#Sed@BY_zLpI~byH@TTK9Rh5G5XCcNB78P7x5(7f0vY8Ijhk(S`t!yBJvb z%?~^Da{J$8-z!E%Xrj^w%9sD5{VZ%uu;CQTGhTYDC_+P;B_;m1zf#imn)qXEbyw^M zRVMh;oK>YLllDai3;Lkh+Yb8o16fms7+<+Q%v1_n!dw3oVgnJ}V>~R%NC1-#7k=Jl z9b@jce2`5-?ctCdOKz=U+#vc#u=h3R$R(E{j!<0=mzy!NrOItwhy=Bu{U?z6BU+738CAuXJ#|#q{th$jy)|p*B*iR@R zA1zCLubR$S4E<@62Bi?X? z!HyH^wKU=;wdsav0%)0?w5omSnKN{X9Hz=`fe12&B;^KtNxQ7OJUZ}qG=~|B&ae3B zPh*Xe3X2>SI2ce!psZH6c{3U&eu0hcbaz_o$k{!*j2c09b%f*TvJ%4Gn$k6;??V0JFkkBP5 zP2uAovO83$k*l>w+w76QEw2?;GH&}yyT#r(70$VZWTkhoeQT@|IsozL6Hg|v!n;rs zN@&QA<6)7to>lykf(Z%cB0V5uM z=zE6zfS8Z5!TuH~ptx7sElr$1C|*cnVgzfxQYh_=H0M`v5oAix;B7P_&aMYfr$mPH zd|XxG7sOgQ#gw3;CdL%kBDFN)VMSKbpCMS1XTe+AH>NxVlxCwY!m`__MmzYcC ztx}uHEm;pJrF98^C!^OJ_6C|1Y?;dRDPQHAP?@?Jn?Dt#l&LX}3wXno6)#5__jz%& zYaYh0^LkJ7*z$O>_UBDs=+Y~u+u0UUso@>^N7OI^!B!EPAMM^M?VzdnO1`^N4fIsJ z;{NdLkA){v=t6-f#m7}~M_r(Qku2;X8a41stdZE47*vSHZ~tO_Z3U$qzM&&2wdyiB z@D+x|>)%VYq~NiA(Y&3}XIuv;XX~T|tn9|t=-R|P?#nbt4D_X(eWG1Kuo5aYTQdCz zupO1|bHFF+A}ZhKHPZDlR~evm>h!4e+BE9fm+Z#5#Yl0-UACjqfy*RMO}n3CyIz5c ztDU=yCZ%r8zeRA}fzRITNEsK#bgsX%7YunDr_@OQueiJHbvy^K-lR5TL^LNPk*|BC z#h6B?@Ez;IIHSEG1HB8rZlIRd?RcM#_q7Nwp7 z59AaFzh*aNU%wn5eB7cPF0@`bT@;qfsV9uPbm_Tyi~+$lfush%n8%elh-W7hgOf2D zhQ|@{D?i2 zw$BESyL0)W>?|)nkw3$$PM+qkp8CcZh=cB@$MEoqR--igwFvO*6uFdkDP^pCXb3&# z!=ygRcYihap9(*{a_P`W>e)yortOLjc=de;QZ_|~80JaF{GiYKP6>T~f0&Y{Eit`m zu^hDI8xf9hGnI4jD=H9u$dL|lV%eGcLqzhhmV^fqLvypdk%;(4#Ol!(J33(=eJJr) zJA}aJ(-(7j+`(W=99itQi&>iERG<-u317Nbq&%YyEBhMf^^C*^jBGDkf5GoQ9)v=5 z60g#7e=53Et;Oqk#dPj5&!xz}n=rH((hCtNz=&q3RQBWelcz;~d9L~+9wUw=TC*zU zA!RZf61C4nN)6$V^%G`_8mEYAhVgOw5yw$PTdQN;KG{jJy1ja9mej+4NZaKLJg`uCIR|xw6=~?YV*u zL_U(?G}8{z)*3(KxCpz1QN+=5>e@nqQ<^jAWMRCinMo3&erJ^F9=`xQ_bCH7L@1MH zeUOt;6*{igO^oo;2Ah~`@Z=b+t>W(SaG&+^VI*?VBSTV&8f@6<+C+{|je(#7AD1v| z_6W2vEq1H9pA#BV2IKAGhdCDxG{WL^2W+2So&x=N!rTb!`o}joJgc{y!Zqs;`*z4Y z17}Q1lj|+sSUIcSMzyJ&MoiQ5!kRk&H?X1-XvCkP7WVlO`ljYLB&wcjQ$*2Dqwy)0 zqgJ^vM5j)Vb8ErZxWN6{TV zrk5>oah~x;KJOjR^qHXEKBR8pSapZma-ZfKQv+B@?*J*XNyfT^8RvgP9sClOJ?vmA zNZ2F|Yhq-@SYIfrTZ%l<{_z!00H>XD0u-}lh|zaGKAUXq)MzC_gwI4m4$b-#dnffbx~sz&;Yh%sqndv^%Ale4;x)2OC!*zkYL#WB{v zDxBrFJ?-hEe66i={@2>*tAg#W0>k|F>Ppc{rAZC)9l7Evnf z^e0v4{KYb>Ntx9a+vW!E*J^%0)8(0D`X`ne+A&=d&v0Gcv2>jZU~f{<8JPzk*xF5efvzP2oE0H^o;X)C39<7R)Z%=!gr1Kpk$n zTsS2|2$1x7ocUAoMODq%cyCn_2U=%{^X4!rHNK@>B-DS5IX&mhIy~Y$DNhmzL!Yd4rzE{cq@e)?%*ehxink*QLg-@oYRVd>;lVt)=~| z+2GNPOuUE;9k$BaZgyToPi#Q)ef5f;OT{?5_R`cFqsoaj?=7#}$qfgel$vqSBKA7C;M6-(q)EDw_rG>>!M zxdodLrMNopo?1@NCwvKHtEJVnE?km^PKJ`>nv8W(ZU7LNXbj_>62T{3*b8hVPZN4H zlC|+Kd=m_H_nPBB^Nstizc}~*@uZUnFhL43AYZQ~V~L9E*~Qm+g^WZN=}EmA1LzLm z&1f&|$<)OJ%72& zZa;?|@Png*G|bR7eu18TRmgR_ZO0g4E@>Y9gC`26C>Ep37+kGx;G zAs(6g`$3QSX=vaqhu08ld?+4h!xQQxg=9x-+j^YSr||8soXEI9H%A`99wCQOW7 zbMazLhu(Jo)U+QLqO@?+IITKQwhU~{?zndrBl9l55xaV5i?vo;;9VaGj%gG_nyVkG z^FO&B9#Kdio_wtuntk<_|PZW~)0Ufnq$7(gqG-wFes+9d{G@M%<0xLlQbZC(55A~_SvQT;F59SGTB?7ZF@)syh(bopl1uomX_K{Dmga3<+W9Q|g5dwRVqq;)Op+FL&mVg*Z}lw;gftO< zE8l<4ie(b2OjszSDnI}FAQo-Zb_P zd{V-^;=spxz%_D$H?>S0Z5k6!c_33T+d>G- zbB|7wR3#bus?!Qo_rUKkIr!Va%3-C)lR-FYTIy!;J7V%q{^~GN-zIS9>HqciRzYpN z54dn}cXul^I20|G;8t9MyE_yq?(P()KycUM?oJCqi zGQ0EIqb3Zk<5Ji_2zBf_xO@Be=yGhci;EL zxeOC!Xi~SKrih$Oj)K(u7XhuOyhEi2R=QD2=Pdm@6^YY{n9p?jZzyWS<0+*j|}^WIoqv15XB6+ z9Ds{T=_J)^oW(QEQ8>%)lq^H9$<_E@Cw*d}##BstoK9SZ8O$TVN7B|Pyj6)or;nl7 z@fSjJm{Xq032n9F+;P#$=cs)dh<-76yek!5wlai3%cCx?vOjCc?pkTLMbV(6w7!yO zw&Wi(-p1yUu)MmxOzPi|x&VJHlY=$=hKp*C;)tje>B~>?aucm$JdWvKK1tyMHvRX^$9^2gU@^ znSYM0*Hh{wfI<0*r>#@{*aAcTE;y0GO(^=UWQg8q{fDOSeP5@=)-@(s7~uCM>y z#pMRW&35rGn|NCVR!O4Iu<7PvcDX_2Ogl8C*84EyadjKBibqGF(F+SgSm8PW>-VbL zD@*dzjd=N)m02j;nR_*iSzQzZ#c|m6j)$DRoS<_RxhKSU}SI0(YncKg) zsm|cErGu6>DP-qy3zx77Nz&u`ee|=adX;W*{RvMofe@D@>~&ab*gm4X%3vVKpCJB$Rw?nukB6spc=7&v&)Y|N zRZ;fUx(mtrM%)T#VI)G-MkJR$Y>*iUp776w{qWhzVpp!_muX|mavkA&aGa`nZ2`=m z_s`rXLW-m6lqHR7QHp8gK*;>EM6S3GQjve`Q9Ld7DQ&6p0){KUq4-rvEJccBr4*k; z4sE&Uwij#twy~@uE*PZ1vDSO$3Dqce+}CuX2M5YF(&!E1Pm@IAS`@g5l(v#zWba?g zn;`22wNRW-L8FQ8EM)O4s!1;@Z(`zv6_s6v`R3oqD)}ywy6|b(P>0l4lREUD9ii;CsY$B(8U?fxCbtvhzOOB?}~7cL&sc)7l~9&>riGRUbl3nE~%S_ z=;HI$oBP^91(E)O^s1zmEZH!u(Pmw7sc5~QdyOC@}pSvp5( z_UOHtD@)hfb>2Pc62+p%vKFgUwA>6e{iNUJPy(;+QJZj$0%H%mit!h=2K_46QWw0kg(Je71ltNcbCa0dk z7NvHDJS=c9X9<$=#OMf~my>XCi4j3^oXhwJ7+!~eag;$e#FOki8j_V=*SbG~lbx^S z_FTK#OC}HZuk`!~L$ZeVwFi(GBuT-#;mu4A7TSLR8PG^IMQ^Rnioe>7sTMYl1aqk^ zq6lr0L6~SATBDTCuti|KkPsB*RNk>3iWVcX+K9Jn7U%%oKio87!!^wl>M9DMq@>MN z5%%m*)^vz}urEfaT0Jqml#_EqX zWQzlrW3x(c8;d{c<%pE^^GvTu7cCNm0e|lW)D4QW-~R-JpaLGHn(CA=>cV0qP=5(J zq>IGby|vPXdLf%@bZXe@xMkzl@ZU8wsPBb7a4|5JNKP5xPIlY7Nz&W0Y+L53@YsVt zJ&qM{rF3Iu$Cu3)rk79PX|F&Camjyuso^*%rw5=S9g7>=CC{`L&HVV5m~T<*0IXl* zuPHx|o5DV;0a>K30XiWHWdhzb6Fy0Ot);fn|OFNVg*qcayk?7$GGXzw)KAt_9 zaV8mLTkVD}%iWFvIxB0fUJ-TKI_1=DHAzQa#xx6a^f!J?QXqz_fk^GknxW`AEuqaE^AVpkkHK|zP&0O`f{f$F3 zW+*<$j5J8o`RC#I$2a@~t+mwjl4yRF{+TJM8anq{M+gpR90rv7e-*mWS&cH1Y#h7w2D>jQlIXopct~#g zrcQ<96EN+1fSE0a__HpO6b|$I%pw`8=VF(5W45~|>0^oJ1jBMVJ}s8WAGtVRI{2T! z6A0DQl76SUh>1e|7lIGTjYIcw;o8Tv8C8C}zQL>3aVOCyG<(CyRXUTkTu}#L9L#HV zgV1Se`drw{PaoCDXY}3=KU@MQ>MWJRO@33mKOgUf=Ws+m#XJ`1Z3Zm?8%2U)SS2K~ ze!_p3{J`iyWEwevqNj22F69$GbX;!XR0X8gbe_ZiNOch(edo~>AU&zi5e!5|16xMU zYCB&Wn0vikB-P`fKz1w(kl&v*`}~^XK8x||1t|P_YM>Y<#ZCyRlzG%;@f=4hV3Q~c|ISSRV z-#uot=vSa;<@SnL*T}DN%@R8)ch$I)({7k#p{ANwfiiSFum%};^&f}xD9F&Rulb5& zrzTA2i7%h6{EG#yM4J?+jwDgzCr4TVZQnT9XJJ7Z>?~_>Aw`sxPtS=(45UNJb>JGw z>KWf&N%!iW5dmU;k=USl$4nJ6nT8R5eeI=NIh*9L$eNj}q49{GWGTk4z6$F}=L6vY zVmez-H^EhurORELt*K+1k68x5^glf6)ioyWx3Sd*c58^z*VdvR?^&e}E(j|ss^dOj8~<$%YpPZ=kM)h!+J9^02~DJJQAg{Q%tWfz+Qo2a zUfa%&nRg(&k0skrhcUK7GI>^hJU#xW%(sMYs%^m7Op1Qe$v;wYZt2N6l8 z+=g$`akI0!_tK>T5LDQnC6bfmf$d;s{*0heY;rya+F4$NEOm%cZ1whI=qiEKbje*j z?$^_@mr0{KJ2Bn^?0CZLqrZdd*$Nh?QcEWNT}=o{K;=!P&%7dj6<1Mk;0&!}M1WDj zlS+1Cszv|ugBBY!NTMP_tV$Aqq_sxmR@!6L@Kxj|dfB{#dtT+O+kLkBJ1AsQ@STAC z`0{vi>O1MHI?$0d7wZzgjDd)n5* z=#5M7n1k2cpm^AOA1~)1voYM8ATBIymy*+zF7_~SjVjy=bQk!}N457|)*(s36;@56 zDUnLefZb(r@z>(;+qkuOd&6Bp9$5b*%9J?KApdngE+VI+fW|xW^a1Zpxg^T^ej&P5 zrZ|tBP$Uf?#{8?tkz9>y8inXm-h~%V9yDO8Gd5%m(moeK`C$VZWq?>grqF$Y&VN9I zJxnSf88~A012BhY{*a~f13v&vp@ROZIPZeC^H_)^paYIta_N>0OcWa?)lEBsyeAJD z&XqjWn9Qjf1H1^5o_mOrO@E~{L~a?+#JX%f7x%6zY<6~+BlW(%PVIvygiFu|(c zPc<{=aOPVE3FNAb(W1GaGB8Clb+QY~lA+tA_hckxva_V%>$M>@qXP3F+|E5kz2{vl z;Z?$)=J8t&ZdiK?<*ayYoJLr0Ysj%A{#{?9kNDoM3unU)j4VH2rO{fqK9`k0_qj3# zodGQ0!w2ySHmM6-@cbSJ^@GBvWF&(?hoGx^=E7edYjd#Hr__?9aE0g81FrW$TDU9P z`uouQDS#>A2$LM-XDEVPPB7muCvnI3%m=vYU`Z6sy;{8+i!7*e`If1wx-k6KJB^3lG$;D;_me)B>ihZCZmqh?LK~-6X`MMC|tf zMivDVh)!P1M{zNFjQK7LI$AmXAwg5-A>I6$jucX5v{Mj6Tu!N4vZd~D7dbgu8O+-> zEQgm;`2xxeV%ip&@NUuL%y~OWC#`57uv;`-hZKSd5SAz%Hmqhmp9OPU{3wnjLwR7@ z(U-p&QCKpbaQ71M+Z}osRbe9+3rtF!PtKm-mNGa9lZ~M?5J^n^=Kdhc{A$kph2VE} z)XxLL;#7txibelvSwozDB@Hbqp2C-+XMq?MVc<=pb>2!u`<6zc!3O@GQUv7yVbiT~ zcD!boW{LI2juf8n2JsLxm+IVY^KaE21imLpB``gX#6n+68199wc~uaJFpJSI1?>|KKWgQYO#H(-3R?{L=EhT;=sSga+rV642GS4Ehr2?@}`{9&Gzn_$$%E59^0H?z647s+yi z{>y!C;Xh6}bwM(>K8U4$cMeZ2!Xzjy4g2M z81>17w;QAStXT8q^r*tKqD2@s;qiwtL~gVFDLe!;k^Y62iuHR+w;UqDBtzGd@wtUi z)i@%XWWe9m`>>umMIXi#g(ckZm7-EZou!*U4&Q_2EKxR-yK!a*jzthUKMhwd|2ELG zRrNxPTHa|JO7X#{9rNVcyZ2WL59{lF^RjulD=|?;F^7blzuGm{MwDPbV;MIt*n*mO zjLV}17n-#MVHSZYYq#}A-IUpjt7%l}FMLRk^1$*T zCgW?8skCuSQ9XkUW2J>PwR5f=+OPgevQoGuI@9}LhxUk**BW?>f0b&FW~VMW?d#7j zSXJ84D?LDoh%Bvu>*5fem95P_c+iT5v4kzz4uu^Fz_HJVjZZ=vAFjU}=vzZBy(8`zpK;t(e$d~LuNvjH-A)02QHAEe23$gCUn&RN>!oGQQwAp*^JNFJlt( zItP0E5TrW*Knx40sQ42IhTC|fow77s(0F;*9Ytkhd%i=_wXBK~z_~-;8knno{%3%m zBLNHPVF?;WDg!fz+}C+QDWYF4_XqjT<*^VkgRE{h>H@cn0$U53)Y)rct6A`=;0#t zGhX0IhcJgXO*I9kE(t?E8{-13pi$g0tGARxelieo==5jfP|a)o@jLYP3)nR=y9cp& zjdHzSD}ajJ;l0y#v|xYwpY!1*2}a9kDbYU}nGu^o>4BEe-2k`a;|p}C4s=VT=CUsa z6H4_J5=Yuwzh6n*^T4>jpD5p(;(S}yF*;Bw624lMgI~vbIV+Nj{@7>vaC2xd@sQa^ zE>&nr!7pGqIzmme@Qd7AW`=XUy%n1Ynqd{ERJL=X%wkKvU=H;n>78s9Ahkp&VHE{C#pnTyW!2TX2!XR#&c1;C+aKt#&^V(pYa@@ z6pcL?T>_#0(wU?ccd@t|z1hkJjS>JY_h_7T-qW#ViwKNcF&!6rYLxAT(&- zD|F3X=vWUMHnB<^Y#YmFwA8!Wicv9Dj4AyYep61*E!jFd*QQ#LXt1 z{ioX~6}L;7;zbu_FO_6?TWKo1oi_?`6}L{cBVZ4IT#KL29qv;>_r#W%d!39?Z55NY zdU)R;q!TJ0(HC2(*eEd@G@l)q@v4BYYyYHygx-Qer-H(uV#C9?gdQ{BHOqM(;Qf#V zU3uoC7W9h<_}r-waz#&k${D06xpHLM!lvD1yXR)TQ?Ixa1)xHqlST$AF?btRXa|mp!ZfzUSQ8ZAgvjx!PEn0)7yf!1qGwqVSr$T zhsp?+6xU~9TY0RwGLO#mXlV|SnWM$Jn_O-JN*DG%<;qr!;pP)RTp)wi4eU=7 z@-xhEx8pzc11IW_K#{ArWhMPZeh0kUi_-{Gi&b5fcWy*-t?%+uVZ9l;Kdsod+Zd5pQ z?1p2zGA*%xOl&tRVGCx0GEoKRUWxB1&|ucL2eD(6o7*|o(aVJ))~%CmradOUFco1H z%!-S=Kl&gh(!9(L)fB@(B(f3gKry2ivVGvn-q@S3hZ( zDV75K^v8kMu)o)+10M3`zttmtETx`!#44A4T}`g}Vz%vJPrYa(fq$ej?{5V03Qavj zc!2$NT`SsgFE^e29&dK7`^i)fMIC4L8Nn<6Q>Q|P)Sn4$E9c*ZF;+C%m13FCO0~0X z(F9s(0pp`mZ_7}&^Fok02O89}Ujs?wZ4G}YS!X(+Jj25diC-;-*t|hGl8&lhnV2#B zjc0dM>vjwRM%(-dJp|+(4@lxy0S|HPuCaQ2Qn!6#Z3DzShNp;Z6B%nKkre#^b6dNE z4IkRF$Q6}elK#!@>PUD7usG4&8 zO^Q60=DTB%EUZ}K+#T+s^??Vtc+dHLmM1)4i6#lx@m{QR{r&-(d81u5>E7Ly|CZ1| zZl0I^{oP^*&-2UGisc&R2Dr{Q6#AvGDZK;N_9P?A=C`eKFJyjfC27w7s~**KgYOcM z?Jf4HsQBI%cl?E=36=5{ty331^PlLTy$BDVi!gtlP=^gW>eOlfAd9_uYrKhY@=Z`g z2*6AtA=gzr(8+$4DFn58(A%J5Kj-94FeyFmK+C<2WSx5N2-(L(q~DPQLb4}s()QZE z=Y^l|(lsL@=W3|Anx;H|k9fq1X1_S}VT<>{He9%4(a1R^urNGodKgAibX4#~pZF~o zZC5~)In?L!MokG}zrxiq_1EEbz@LSEP~nGy?>Ni zM4+6*CF`0EhUTrxNfCCTMF-kw6;pnWcY$fHD_Q<}S&-sUPkIX)B+X0e0NJz91Uij7LAnom)Q`*P_yD!C&5ry5hi`zh9l$}a|RuNgwS2lvn1+N{Zrl$ zO&~rzH%uMxx$Zx-XH0LPyrcH3$CmPO{4G-!YTE$GGW!MuT5+w5NG{rXsdhBv2CtWM z3j_*C&$!@J$9lytOAE5J{|HY0**>#*Rq4-1SOgQ<^#~4a_{8$#8%h9k4=2P9#VfA; zgxE@CCva6Rr?}=!Yd7{;TyAr*mH~Xzp>0f}Q&GIAV((*}#IC=9uso%_I!lN}8Qopu z&r#t3H2RqBZ{KKjCPLW%hQy1l{e}{ilVTI2sVgk3q=iwa>+TWhU}1^3#O0luM(hE@wWxU}OvHoeV#ff@*<))pCvn z(M3Q8J)jQnx%a<8^B+ij&xE)2hSo87n!}xP<*#xC3tR%7k2~}9IVJ{6wZy*UR%N~g zbXQ2F+%UuMcf^3&zSbDHn?Ce zqK&hADDj5u4BLQ~=z_zcd^?JH>0Y4VcTapt)SjyfsM(p!UEngx;zc?9Q{kdnt70|A zxlD+J$3r9vZiefOK8vuzOw-`tVVcUH73HrB`}osHDNHg~2K_xpar)14{D{Fi@kB4G z_S$pZbQB(bu5=TFd`{6OGe(R$q|5lsgRgT z;YY{oU*V}RcYry*DPJbv(Th$9k!%*Pd||5PUk=7Db{h>?IS&^i({p2ZQ0~qnO0`5q z(IxNsiH^+;ww$KsnwFFeG;;oQLKv|)!|21sLq?9v7DICAZ?EGH2k9O%NqKo4Iz|&U zTW$*4uuWm}xCgcOiDR|h*vtIonx$7M($VP{CSEV71}TUI6!*#aybF<`%UUi$I)Y1h zZ7B_MFM67NX7}pCNY^Z2((;f*PdZY5eWToWFDA<3q2FzY$BX7@`mg?j$p;8|coB>k z%UwJQ4Vq4noXbmNH;|0wrx8BBf?7(eztA+^_8BmH+@>oKIKVLERyluKok$btu@ zS8nX8wdFX>yDCk`rty}yACJ85GEwU04Y&4_Ttovt&>L!@@T z6B1_587U&6E`(Deg=LNO)Pi~s;RAcRB)_f!tIK7`F9oC<8@+^8?kE0PR|`!(X`?w8 z<{UTFMMR?vr$nanzZw7%a;r0bv8pxJHs8#p#&|!h{^nX4fA#*4r81UjMa#}C+Be*_ zbWM(IWZ2AFQY(7dOTr3`oW8|Q#?e;zrEk}tBzk`2xW(_>HdJ7A?Dz6HM`Zo^^F}`V zXNYlyJGcu}+Z;CnDAA zhyJd#ESeXyT)7fiI-yT>5(GphmPG`$ip zo!1pH6u}2-)f+qd)1do_zl@M!w1%@PbwxfP^WTAYNNOGg8N#zXLF08`G`(BekO42n zFT8(zui!b|ua+eqmb$?P1D4YEMYw-8*_$z@&^@-mKrBUUA zUV;b2cTXo^5X8XqKA!9XL)$ zxWx|sb83o>grkX;izwQX_G*&WB4UWKSZReFfzsn;k!@3-j zN)CIq@UNN!*@v7XyTx_)B^6Fc(?i$i9Rie)?=ky(3;@Rgl@WBAOl@7+I4zGpAsf47_9J?FSM{@F z;v57cI(v-;e&s-8*ZhU>VjUOL+7eiAg(~DaWNGPtZEjI09kzx|%< zNrPnw%4T_JupHBf?E3Kh=1^W=?IRlI3CN5dF)RtMFED&ENXsC~`pKwzqZBvUhXSrYduE70)7bK! zqj4)%dNC#GWII ze{K!$G=@ER3Nb_5TvI~loxh(}?jN+u(j&;>_51`ZS#N<%H_0(q9B4(dO`Cc54nipJ zm?Qf&1NNm8O-ke&?bx>}@*D~=(;w}a%C-ooG8%hjY!%+WWO_#M_ARKvn0VoWk@9a1 z7Nktp_UawoDi!w^OQ3YV^RQ4az=pEB1In>;@wXH#n#029lWmht7MjDmI3fgUI83~XUo38XMWq=p+rRV*V zaqI)Dq<5Xej15k_qQ{SI&&0^p<%f^))R}+ppC~Pm+SEa(_AnkPb!5y z@BY0n<;$@fq$|%DQ#39hE8V{W#IhXG+MTNe)qj$M;k5+Ro1(e7x~>nvvI^5heOY-U z5CbAx^fmda=FeeDBNQ!-CXhXp;^jobtAE$tIBNRn#Kr^?73^E#)6B)*|GCWPRZ2Cc zT#Q{^E%YzXX}FSU&!(0Vs6_)T!eOZE33*5gyb(B~?)~ZD8$4O*rLtSm^3l*R`XuR4 zpyOpSNV@X#dBLx$ja@GXGJve}p%HAY;h;&)ufNZlIqvm|_=Wuzy>#U8iI)ahZj|Wd z@HPD(fSL7`|J}&Z+i8ppKJLOe32dKQa9*Wv(z~z z60FG6=xqOV$jZCDiZPFgA9&;;Bj`T@%4K@C6De?;o}q1>{=n5ymdLYev5vCB6p>Be z^SN7awr@o+suC@rNRPNRccF=#g`GnLnkal?~5)h$gle0+7=Kh%=tC)9r2)8f;Fk=BzaK@MGWVHRrU%CLPmP_4Fi&Iv<<8tUOFq-V-~ZW9bMNP16KR( zO6~4iHQXi~j(*g;!dy4#Bf(4^{sF0{=H2&P%E?b8YZ7 z9Nh9?d82}KBSic0KYTVxqVV0M-C~f0=Vg{Q3;^&AC=t0?MJ=P|09^Uv5|;k}-^t-! z+pnL7U9Cp>T`HA$a&rX{mG$EH%X^^!m^(mR<-+aY9;#@iGGvY%J91 zvu`^_jJoZn#Rz=Us{hm>SKwYIk6KBh>$x2gHH8gK&7HV3{`NGGvn)d60KyRkGUn&t z`&JC|?ypMDB^0PWMj@(Cxw1Yxr?w6|&J+ca+I2rB+T9kYq+H$3hv3Nw>?WOK$wx>` zm8eM8S)Y{?X@c>AQ2#}&+sTG;Fi+Mr!)$ePr_D|Bya~$BxlL!d@sSf8iq4SEIh~GL zm&JZCQj`0-Rm`Gl?TQE22%p7HAy7WiJrf`k^e|&);{ieYJALhNH2w1~VARKfi4H;AoIg-BhCI;gqSG0iGVcG6k!(Il=+D5P^& z=#sy9ycf7n_14vn*oSR8v{dM`Le>YN7}8W8v-S>qn79icba6ms*<58nFS5F&l>%TG z8qoP7%#t;GFv2Xac$VfUiw;!T%U^+6FOmVef6Xq;SGwOJS`9R zE(4Tp5hLWv)jDjeH`OQpT{7p{<#)+uiu49^Y3a1Lc7N6Xs%~zbPER0#z&%KSLac%n{rgg*ngbg~d~ z7`D@i#Ls4iCU}+Lg#x3SBVTKW0Zk)~sN3k8!nN66ZP47@ff&qQ8>!j%F<_e$S<;0cfIdo2JRwpYYnpaVKC|eX0K2d3=72lD8{psPCq)P zj}KYqE;6J-o}4p0#hxE)OXqBQ@rpd!k{Ya$X-J{U&`LSBqq7Gf6Y13(XO&8L6uU)^ zTO{(5QO2UK+!3F@k;9w=EtUO0z<>=k?gXW;Wq)5l7`ZOQmY{IznHSx35Tks)u77q4PbM{{Xv*eClVRHIr!hDa;cJ=knTumaGb` zyph^XM_S5tb@L_U35^DSIqBO2Y`Bf`P@n@44j_z^GOn%fqnZ()lsZqHUMtfiU06E{}$c?gkd7$IuKVI!d-hF%5@P@DpUB{-w!4E~a7eS|wD lccgyW^DCw-XW3~&NYk{5ITe-YyRFH*3jYC${sI2|_&F%lW83JcW81c!bZm4wwr$&X(y?u$V>=bw&X>FUex7sA`~9dV>+#?_$ zejo{90c989Q%?vVEj7n`wvMyOd!72W|{ucZsgd&kv^jQ#5qX6ASXIcsV*VoJ- z)>fb>T`lsyrVeWQRCg)HNYm*&(+)gjHa6l?cAWuj5G5JxsU{J$c7UgN8T<*Px~T33SO zDP8bnHgYz-z?Eo?>}mDi)&7y48AOTLhPvQc{dlw_DvX(c=7?v-&E(M}AI{#Lv*OS8 z{(FBL10;465O-MQz5KJ65Q&@Vm<1G-gMJ;W_4NOJW8lbcOClTQfct=X!raVQ*&En! znJw1%YbFX}xXJo8;RDXU$M!$#&Jf#>1UP>)HdP#i9gR)MjxBf+7e8?@E*{OMobvD! z!{*gF!#`qoaQ#20T8s@MCdh5-c0SwSTJf}RE4`RR$wgzH1+|?~CSBAnZEnH8BbRJV zTiz?CBsM?a`TxjxN<^>3n6`1Ed{+xFmP&3c9#!(3DuEp_QN4?xs8uG=fU+uTO`R#o zwLF>y$iNCh{8!$;;%_8CN{gv80Zb+1;Kt84PlHd`|K{$PadU?qnV;HHBVE@;_HO$(s2@r8|e^} zBE}?n z_=-|AHl-|9Z2rH^{x>A9H|V$#OGndfyR>a$kw^Kw1vBxhT<=bKyXE$+mnkw^6Da`h z4pw4(f97V~)?WQMeV3BjiyL3;DXK-rK3cix0pvq82 z(y@Ouo35jZOW)vkYYn5)O;o#b(7h&=k(DMHzh?bvq|s|9&7PI8Jt%Tk0ZPOeo5|Dd zCc1w0tkzBuD86@ zcRYrbrv&WY+?G^-`I54!Oi) z^x{T9Txos$Xs)`jwL~}-&A8ilKejcV7HO_F!@JIS@58q<&pavl<8dhwKnEq=!(Rns z$*Fhg&aP|5(|7$>j|`UQa80%s>-oJQo{60jFwnY9RCGtau@xKsyE}RwkX|;CY*!l_ z8$Ws~_%zv!Vhfq$l53^J`1U^Wx?_2j1eYS_5w#-|KAsBkIt#fc2}+wGT~Dz4U;>}E zwm$VYa)e~zUD<$_Rz#SH9;Y7gIQ=Szf-;4WaX0O$z6CF;wX4kP7A1r%;fgQY3!=h? z#MR0jVCb=&(TnlO`U~!P7b36iFYu>g#H+?-q@fr)Re3eVaD6pEl~ir0UJFVG^wXvA ze=>}(G~^{mszkpKW`PX7|HDlL|XcTdntVHBt532D){HL!4b8} ze6O2fX7wEpu(JaOu51#(Up(upDn#G33CVLdKan$`<-nM6g6O9iH-NWYEp6E^C3Ew$ zz*7-8RL!2*nZ1<((vxXwo%QjxhyM6{9Xqw^TVri9=elsUoNy|=bWRJ0)g zBzR|0d}Y77_*f%qNs|X?6p#=0umO3U*9T~p+)sZCEDYKvuMyhMsuJmBo zXb<~sMt0tQB3lm+I^BMQwRYduh7>|l zbCYO#^;*sqD}Qr>YMN8`CqQ?nqXBQHggF=4%O{c@6g3|zcaAb8v+Xh#7H}vTCw}XM zE*~_m9xdetl;1+>Qkz|-YWyZqC!XkqvfFXH>E?^{TwN1rOb) zAqObQFHdN2V=fOmiHLgRipCsmc}?J{RJL}X8TARb&7V~ktfhHImqS*_F{bX%5LF?C-5*D#Bz zK}I8Lc8bINi-r>Q)C$A1M^a}e^eKLfj~<1_>C493xM_`sFddXuFRS@WC*prd4gR2@ z67v{g4CrX_un|s)C91CFr?+$_ZnUAGo@a?K!!lY`3w3#S3s4Yz*7+=P=OFy| zP@q6U)j^Sj0OKO&wphLS5~yex`BnKaJcbIg`mZ|!9F-pLil}G`&}Sp0%aM8c#W0bV2^I20v$Ot~d?KQK9$S0QNIW-| zORQRuKd?yq%&kGtag50-6|K^C_#Nsorne9Y1U<}1sF0{yu^Y>pdXG^gZj27PLYT0k~VW*Ah&niLMywEH7J`fT5=YBmK)=cz_AD4%$;Z>US}v1 zQ1|o78lhu1;7@;YLTY-UO@42K<#3_}DKH;v<+v-MzA~Ju)IKD6(r$gGo>NQA%^@KW z`oiN=41`~1u!pLeUseX`HJL~%qVsrtgM@uJ=v##TC?%<&u<8?dQ(o$3GSuFJMlyj! zndVxprvpc%kNx_vms^{sKJ|mCHK;Y_vue0B5#k@mLV$&pUXfh=O(f2RWS<6YdaaVC zJcZgAGaR>TV!A^qhomDCQYfH*zoVSCtmaxfk&Ae6eSJN`1zbc)L=hKk#Y) z3nf!i4mp2rw|668G9R79oa3ncX;tjVpu%y~Fi8X8btP^C`KZ{#uaYYC5HbTJ&_)Vl z)iVKFqgomL@pJA@HVO`y(VJI4aAB8YcOmm@JYm(eUf?UVE%VX_SqE>7D;uZtf=db|77YezwI6Lr4hqIB{;A}BL^C_()d}&{lyn6e zkPf0VmdE0w;Y{G|0;V_N1HyS#m7s4>Dx z>7a0;{?}SrBOyDpP@asHW z+*ZmZe=FVy*D+yj3hU$jZOSze#j~Ymc1^CiIL5&_IG#!u+w*efVo1XG)W4nHvd-)> zCJ31a53`^&eEmA+I5YB3YAN;i8TYn>JWwW4dlHOL>3j?$FV61R3st(Cxl!>zHdjp~ znZ=bQ>!Pe2mC)c&ZVi*_RGf|R4I9?t`aGenVWgPN*w-M&*b-uEnML+eq~AAt5}vNN zTYuCC0qn}ki3{)V4L!tsBPWtKZ?E*k=mu5pfrg=bTyD3Mp1CIzV-iYU+`E(B7ndy? zK}u_!P}Qrb>X52}W2fjeeKh+|Z^Ui@3gYW?^)lU+OB^Jt_IIGl#CtJ=<2-v7z({{sp!gVg!*b3P z`%J>(Z@6zu{-p0uo!tuY+fF#>l+Dp8yRovs1*VK9HBLZ-dO{}bf%LMg+*;c40 z+=YV6x&8}Vbd!Jvm_l!&R+iqT75W6A>7wbvGf}9zQvUp=HD8IH#tb5=Y#3(+U*23s z)(N|mQ7PN4HAo%xk8aHB2S}1CaQTTu5l`^w(LW*eU)+Tp6m=1Ghgvq)IFZ&I z8NHH1>uV;iRF))DunNzJ(osCgL|n>9sl%1Zl5|306n4d9Sw=#vA!QyBAdp^)W8SHV z;gI9Z=q!dRkUXDf2|BeVme6AcdtenNi$_?Yk_> zACmOS9#C!k5((*69GMefX5vF zjN50`%MAhKN$PT1b0cU}OX))?!_u+>2H|y^78i1F_ZVj*=*f74Lh^+AUk6Ts14T_% zMmE__prk%a-&ItF7s)`w5SL4Kqm=6%#X82KjGIu446VxGpf)dyfKDG9@u}mm@31s> zXcg6KL=^9`HZ5N-UD=jHr4eF)4L~Uwam?v9+<5jZORGya6;E*n(uBVpL-Zac(r3l2 zB^tiI!AD5<3x*Tq^=Xhtc!r9*XsY-!S{%zxW+-%si!HJo7H%_51v>gD$0MAQ5=by>tlw&;}N-Bk}gt6)uRVl#U?n+KK zCYDnLHUR3o+4HP=3X?C$SuPz!gHnB>r1d(kGndpB+9``@!NXgCY=1%$H=s}G z{wgEWALwghW7P&GLNeU;V&HPf4@p-qX0uC#P?>rVxLKHD5fnB=C{bZ>Y6YURETRGv z6i7c-0+1(`;0wQPc2yVrmM)%=y`Drg$=k%i#*+~+4-mDLfJ3L#fOmfQc4|i8znbR% zGkQJj;_Z>u(h8~%V6#*EV>45y0#69%uC=fXEj}WWo|CFqDNwG5EsTd(SZ|MGk$^bbxr$C)Hf>jM_)Z5t z5HKp68;A|x;lj9|H)N zzi%{=YF*@6t-n~gek2}l+<*0Ye=*g&T4I=E3m`Ss^_BZL8>+S&5v2CGHb32bqvZVZ zL;j4*E!3GajI;6&R8~uf81W4;&J^jRtweF(wv(YZ)pD0>!b*lKxcXR<>0K zjHM{!K}of8(3+c&Jz+`I8Ic!5)^NxGB+{Pfi>1_$Ep{>lMYM$-1sJ!BE@$7bMEI-K z7PrmZS2hU-h$E34ARZ0OUb)%}!*Fpd?0*IYJXJ_Sh!R~71ahhdWAIc`T>m)VHRwtv zH-=Z_`J()V_ZDVqqai^d;JakuB2=8m1&=sdB=T0ofFeJ$9@^1xj2U^}R=WBQ%0HzOt5^*PQSU%)Q4ngBv_tm*1@!teT(vZ5p}#hno0;hsJ0~euGk=cP|_e zh-!8RA$Xs+scZiZ%zk^8<^I?~sUGqpSFc9DcD9W$HzdM&q18{q2JDeSft#d6@A4yhAi;P4#bIN0O@QnF~YomX0G zE^nlms~@0Vjzfmc7rDV(Idd-Cy&US?P3Y(A4JDCHy^PqtvnReo={{4tg-2YZzl@`C zqtjL8*WhuUQEbP{s|H%pq^jAw2&gQ6{6>YzZEM4ed~wd9&&dUB83}T9OrgwD#uUsb$vU!d-;~yGYg0!v%FIDrNRS1@#2^SJdB7z;0e)~-d^|i6 z85y|iYet-|#1kysBcUHsFr4ne+v6i*GH196=+BY!Po-ltGw-&y>iTwR{eyIzFcLLF zY(S)0(?BaLZAcAn3MvZF$?J0qrEUTy+wFT^_>e<2wd0XtxNS91AuSpLZklqaGJO=i zYbYUxW*8`|s@@dV0OP;ZS-QjO zIHUfA;QGBW9ekffdrM|$vlr(8{?rDgR6D*Y>-QxoSw#J`I*uMA{-;{Jo~y8J|56R~@Yj>mWP+yz{U&XuiF zyAyGBMORqnPEgBz_C>tC@9>~uv<9r^0RXDP>_JXFGW1@_%G|)3?@)oG{R?lHraX9- z(+|%br{;p6DePJEx3u57{9Yg_bHo-uLENO>thEgAFVDX%AQ$BDjfXv1%YYe(wdP@8 z2KQ=zomt{)5Mwr2*9Lxn(mCX0iOnqZ@6%g%#>TivLzA7t<BCCSrKqJ9PSj zsB&C4{mB2-QjX1q;evLi$=dl6yaZjCBMp?{7|tPf90fz=@2>tUwIw8)XA9~%~42;;T2CFlHi_qJY%C1@g>Lxo6D@f3W;X7UhH zvEH5U{OwK1C5dzV%h}IrjH!WOzIXhD==1(xO&|E#kKLZ);PL8C(~iw%Z;}pO)dYQA zU3gtx9iVq27Xv%rD_QF~71XsNhvmsu(moKBB!tR4#7!QUQdK%Wr1p=FL;E7ON8HpIO^8o4KWffOv$rueU&+*4 z{HonXu$s5i?mVz^3Q{@8!xyxJ&gecMKkicHh$P>PG$$oJuCH&VTMnX^ZpQN;r`A^y z4=;q5XXtT~c411N7Lb(_8sluWz4tal~Kf5BRpYnry>J+v@U&)o6S7V4? zQ5Z;C9hIR=V!xXL!@;EOO@`RYRC;dvPHxyCB^sOrM)7rEc34CSMt`$>;j~)81m*dh z--g>?UV)zd7?kA&wGZ>t_1h|r6ywk8MGa_R`f2*!o3mt+7t*IW4BDx6G&JchD_h-JzghO9GqC4yF z-^+;EXP2mL2aTv+^Nf}U^GjNz?E)$(>|}(~^tsSRc?gge0{2bgiwF=;Yir5R5icke za7eMonCO!s_~5GODEcB0HMBn>BD1bF{)-}**cBaHwQ|TxGJO3>MZMkyi89QUKUbDUSgD;36+ZE+?K;_l*(%Es~?s2>qu+(76C%EA8{X$(D z?`N5n@|0Nrk=RceihZwA+_oiE-I++&8w`CfTgFh!<@^ml&i-_C-L?W>$LW5)p)xb?t@8hD=*y6v8WT^-wurJzR!ell#b!H*?pn!zqT|`dC~dCD=^E@GI&NR-%V*E*P9B9FsK}HZftphbdFR z;2@ezTLq()zeGGL46u{Z4!1XLH;aA3>rA~C7h)2eypbX+^sFfk?Oet2PJ)}HEOXp_<>)R@s>QgP;+BkKQkN4htL~^lePMrvV@f)PsCt9p zU#a0Yt_5w}aR4;#ZyD>6$F5@$hD4FRK_a{12w3+E6G9XSfH+^(v|r@{8AGk%v;s1+ z&~xfJpOv6U^EKV%Kv3PjOuJEu+rJ!+x=f(e0HD%OkEr~lBCc2MF zJ6d+Kab&iDsax*CHKf@6+n{3NQO2*fqjFW=t^LL|+0-DXXJ>(jOk+_FU%q?^(y2RS zl3LNngmN;X1Ou?BC#w1J>tW)Zoh{Xw64;Tuz(iHQZ((Yw($>}_MFOF^=i`_JzUnDAJZG5|b}^*;b+>IUqG06pHGpY+Dg>s(D0%g4|Ql zc>R$+84(ReLVn!UgFj4NigEMgvTgNAQFO**7p{90&|QQ#7h1F;`fVP@Fs6cCf2&uV zecNpq_ol}8*+M9m-LWDcvoNOAl_ZLZ z!z=wtp3^C3Ud^3Uf%Qhg{)AHdDUM8EWlD=xY3b|4{%9HS=ZKY~;yTfKGTw&u$i8S& zE||GTIK);%BQ>{=8l4H0IF>e?whr-Va8Pt^aS`V!_>u(4)-?jm2m=08P;3@|As@7^ z0NS3?9HQ|T)UOw({6OhJT`k|Zc3gJwmr}nP9M(qbVuzKYnNgSLSA&XZ(nYmwWXEC4Mqv^V-Kqz-2^#Z~Z%JBI$^L7G!? zsjFZH9%!g67s#jCE!3D+?CN+w(at}DaTV+GGPjNMVVk2 zCl1vayW>qI*%SK`Wm9Y!Twi>xtlk`D!(iO@g!mvSkvOg;ey!{wDN$!ZG7DA04KSr7 zHV+dv9pl5B0V8gqC$dcv(*M?}IFD#vM-9up=MrmarEJ@y%!>gW)LfW8zs)$>T(APbTP7SqnaGXWh0r?IVgGC3;<%{aLeZ zP4Ei{`&IeR{EW7E7q)pK57c8^gybnY>kIXrz!HklPOzX*d8gY-Qr3tod#6e%^lr_s zfQ`C@q^@Rqg~EwJmjEdWox}|tH#fI6EuCvr4%XuIKBYJldjT}r3;7*NnLsoJ*5dN# z6%4yr1Dc)WgoK>tlm$vmRhLwTT7{Om?4drorVjR4a3#BOG)-BgD+#4Ifz9;z@DWLT zMX$Ay34S!VXu9NJ$&6F33{(jVwlAR6Bt!Ms0_(;aW@O05HT%Q?9Pu!HuOC>igLB-XTUWQ@znIOVx`oa;M4_NL2A!D2-G~y!w8C*)B2g%-iQG;XrH5L}u zp`g7SA9PrYs|1D2kg@%mZ&jCd%7Q`rT-%go8@HxP=3l5vRtz*Cr;^fMpTaSCfHROi zmX*sTLqnnmHyfBJ6dmudi4J+E!sKyTL~4%u^;%+fFJP{?nB%eA0dodNsYe4E`hht; zT`5D-R>_`?aw>y0{*q+IgCsI#EUm9hNhCu>>m?Nlv9?0oVIO(7RFSsBM^fI)tuu-M z1Ju$e%Ru}cZKVtLl4ixaUBvV#%znu!EyrG|`Q9yC7PL6nWpS7^tRNP{fAnat{@yqg(vZO`YngWlF?t4~ zX^rS#FR3afO;IP%S#1wQ*9oifn1EJJM9n~@q}{il*8kH?+AtD*CDDMdLs|Gt^5>P3 zy&glA3_^GYgcG@0ux0-Hwv(#r&1RegEJG99skan>*{5ZanT%gZrDo^Un(FgvPe+}j zi=Cod6A?bIA*^%FSJG0L>70hlzl6||E69W!u#R1{wN3;=ov$@r7AWg72u=3TX@T|^ zzGybdLo>ki(N89OLx=eS#uST=YvEQ$lOwtzokzpDDDJj)7I?TAShd0s*(k7bzlV%5 zhIF(!<<&3AagEh~+a1rpU12d-kZNpnZ2i;R{Dzbc^Z`?o4>WX4{o<+|&5j0DXCpJf z1_ouXPG~3_Urp3UsXd_bWcmIHWKo(@Sx%~L-$p67J{}J*WJxBoD)S^3c}8|$34VRc z;z{#@7Wbjs6?Pz$X~0V0p!!in2_aFIZ2fG#XC&h2tYriIdqH{~kYz|V`Rc0Fg|FTQ zt}Z#loft0{^u8lGm!OZFaFEY7_pVr7j_bgE2<5Sb9X`y{`acxq3*){5t{ z#Y=0nx6?&TI%z=DCvxNPqq$}Hj%|ifY-27!qZ<9yIC=C8Vzqkgh~LVV1%0Y~&}P-^ z)bmcmvTms*RF3DkOlsQOh72C}bhmMdc5aqr$M$=jUF!zObNgS$5=Aiox2)3Uf@?s* z#CATI{`VP=`b_bT)CIJoY+Sdit!W)mZEH@@faZyy0M}M8*?SLc^`1bCSx~%RHQcvi zFr1u<=}FQLw1h?)PH@yYSK;=h`?e!!$}N|2aw%ZZ8SQoz+PXgU7+(hCRli%0_zs;3 zYw%8dTUfC7$HTE}{yx2%M=>z)Ltz8acl~G~Za@CbWM=m-frktUM68!3m{ z3vA6I_3;Wv{aw3_Z5TbBCL!lE6Y=;=Tq0{X&z#{^!7^w4w5p4jzC>BwRTsR~A#t$N zf5*h^=%A=%DxEFAq|NtVEw{}{CzdIX9X^e6Ya-l7Jn7`vT=(gX-)2__ykjl>wczS3oiXOx-U}a+=E$=#i&&|A2V5J?X^L3YAM&`J=DT0dWbY zyQOOT$En>MI(y`V&kRlsl#mSekt@=QF}zrRtBl)ZZu6y0(+@3g*LJN4KDAP=%-!D+ zkrG8a*NnclhTWkI;u_@)Ce!({GpO2|@i-9brdlO0oxHw@zrTU@@ovXD%)5*5JnNQBOB^FK zf6Z?5NfNNaqoii?M{_bj zj1?iIK;w8+y2oxUokRp(1}-$TbvlP)n;rb8ECy>un&s*fyAxaKQP44wM=56{+G@s9 z^>ZZFR6);C$X|SI0+GIxXq0RamDcPGIij>(tTYF-2G zDE~tLaK47MSq$qX!VZ<0pLe|+B+^5|3nOQ7+C`5u#U~jKs@f67AG`F%P+`(p20gVP zvHj+%!!af3SyO*qw&9{V0yFacs z8HxK}s_7)Awh38vl{MYbqy-iAR4`G1MY^98K}v>vX)PqOo?K&(%_fG_t#JQ(y2~!2 zOMHxp#(Y-5?7836wm{SvLiQL$@;Y4bym^cn;W&er5WH9)W;i%KHr9YTyN_PNXp+Za z+Q?l^_fN>Fd7%us=L!S|LLv%E`gb+Rc7}w|bGW?j6bo5K5*bkcUW)OfII`4IR;AQn z08SwMo5IiKA}W`XQDO08e36FGXHFZtPuhi+1kw&T3#G4ci41R-T*$HOZxAvL?xE$f#p>lz$dCIxy~0ik7qnpyT=Hf7&lUOk(>U< zN>7FzG|dleO@>l8Uq6Sx)O;6|TK4afw;i{ek1`jjk8>=hF~3(~L1=~7Php9qZ9t&` zBs3@t_~+aq=Q((Hj`Kl@aQWP0-&8!I>PUOOwPYiRV@m)(Z4^w}cNs3t|3H9N2CAE%=XC5)l<4&7EP|ja_6jDoa zmcX*L!4&qnTkuXz+E2u2%^`O2fdKLIodYf|L_xS9m&7ROi6N^NF^7CMpx}xT-%D4%7W>Wfkx-GS83S6}@v#Id zb0}`_7rHBg=*HbL=Zt9nRn^xagbxV5=QC~f?X^Gz&r>BEmF&PejOVK*xeqjF-ew5r z%{RDyrw$)$?>klM)_u3RXJkw$&}}+uf6de@aZsC0gs!Mn9mw6H^eWMId#cX&NNb|1 zeK9l46X8|y5Y9CQ*t>guJu^&nHTW>8u&3IeeTVV{&k#W^jWgkXc|?yLuN`pr_C0%u z)aTp@NvOROn^Daqt1jD>**oVQydL*;9|!>4G5R6oeT%cTJVo|+q}vCCuh>mBzmWG{ z)*++ebHjgXWTG=SH06;XvaVp^dl)B!9b!48e4U6*Q1Tn_455TvjzX(?C2|=hK;3Y9U2_Q&3Y0>C(y!=gP9`>qIh7&aMos zwt+Q}QhknLxXN#sR?kR9Ph&>3G*C=R^-C|HE2HjlN7uM&;B$^!v(3dwjYjlTu&%+< z3-i=6PcJr&;WKm~rx_iXmwH`#wBx#00C8oGvuJda3d!qY9Y1u)y{VcT+sHWqMm-*rygud9){6WUo`|KL%@-}0=J^J8H z5Z(2Tq-EC7_rtdD>|XESZF4})>x%h()7d*X932(s!LZpI*Mwq_&qz^ff*D1N+u9~|?I2l0(pst~Z- zu5nlRtT@EJXsI?`F-OPCgn&k@k%*8CB(ozBuO}4G?A_EP?zc*=1Jq->Qb?YM$Au`Y z2#xCLr~C(0IEHIT8HOROev0tYlA2%N%=ewyrQObojHpk_V`;R_?y~*mMqOET(J zmvZ!k?Po_7_O{+g++sW!Nsp3A(omFRiS-33)r8$5rm^ar;ZVPJT&<=e5ICJ2@@zT< zscV;?k&C$_Z+M2%w{gk8DTBfvP{E2z{mZcPD)o8y0K|fth~B2yLyJPwdz4k zqsWXLw$@AoXxedq7etHa?0XJ2xa{dfkJzIf>gA{mO>VuZ?1aLgrP*1F@fkB77}R|S z()6Udn)rBT=zN?WyWE=c?rD73>A_gFSr6t7mD{|Qb@3Bue;=N@?A~B}T|elIz*#`s zjp91rmtvKRngnF~<5;)jV(8xBdpuW73FdWS<-Qs}UbB6mVEiia?b8U+U)x6D3=5-ezJl%XE z(aw3;GxYwReYNeZ?dw8r;89A?x#0aDf@^`0&+O9(Ksf5S?jUEBE4U2I%O7^iDBx&L zcDPa!MR&)l9Am!IE$G)FoTCyEY3VFip$Xy4GALHR)A{|SB1CD`RHQ+f(%wO{?RGC? zhQb+O=}Du_5h(f`Zy+r2NnKv5S5r@n{I}oGsvo3$-Vb~Cn#SH>-zXFLL=;n+%H9U4 zo2|XPBKFkksY$u}dg^_1l^WMQqhEu$do!(X=Wh*>>CO#gR+%&B3v%60wbc7}Q*CP4pGdthPm_q8a za1rc(K--9xW+IP6I-lhSEX(3_hKEgWOh-peE6k~)>l2<;A%Gcgvf;ipCm!X%o`QxS1Rge@7O&G>nglRd}*a}OcrWsaG@5cZ;H^@d$@Vr=!$cQ7FL+u5lNfQ%8$6poTJsG1MPw*EriRTwaW$c#maHAwUg>0LbS zDQ8YfnRPF{kwcrrV$j*@YG^FHe;mCje^f=RmWqpQ*wFSHt%~e$(CXPBxR+qZ(;8(3 zt%Kf&JSqSnqOJ?*jrzj#ID4<3kcgq@8Tkc5f|u)MusU=#jqL#WwQ3gP?Hk!nF3Z#G0S9hH!P3+&YC5h>Q?_D_=Eh9t^2R1|3Y z=Z1-i1L}Lbee21=O>|>IrVof&OKbtPqjp*>Q+;v=Bvhw#ix&fTKdAJME{3WojDG;e z`XZ^u$n5BqaH|Hii08_4jug$tl(#^$0`PmYzcY7U9h-?$%4#(tvAhkrUN5t>m}_fQdi zl8)a@T0<#Ui;u|Wc$<&7M3K+{v^)-af1qZ&jX}dUbp}novoB~lkoMB8`xCs~I=eg{ zL0tKNJmP;}3$L+K;L&|oEE9F=OlbVM5whz2M0I+O&oUQYFqAQ4tbRt&_v?W>GU;B{ zf%RvfbEgf=dvi->7r|2|2YvBmv}Zb7ZeczOm#!y1&!hpn0rBOeM(C8syROYLGLBW( z8NzU88>Zvf)ZqU|)Hg6z7B0(9FyX|R*tTu%*qqq5ZQHgnv6CI!wr!j9a?ZW)z2C6D zUR_;XRjup$hJRZp2Si!cX}}xg3?=!ZqZ7a*puEB^sz$EqR3SrdpS}A2yIBT&+$xW8A$J|7Ys`kA^-g3Mz{w!6crFvn)j6uo~>J31yCf+3a+H zme+=kde`bz%O)6R83%cc17g3kD23V66_HSl25Y}BSK;C;oKQV6mig#tUw{bxHlNVS z2B&9AA5zBou;A}ZV01Pk_z)4vG)SWzqfQVQ@@`G2NRO7r+{M2BJoG{lQSfz=Xc;09p=GmF~IGnqM>Rwyst$U zO%m!b;6rT+Ml7(g_RD=souKdM(U}(Si7wD$w>sopnZ*-9hh8+Mu3$hA-01ej;8k8` z+-MbXW+T**WreXI-tp%L*~Bg19{$DOEd5K^1}vU|UK<|xXuT7;D59G31HJi<`kT*4 zwwfZ^_*j+O8+HB1edX%``}NfxyBpc^u&c@ckDces6BaOZw3qvJiC^aMdJK#{qvIU%Hn&gSMaliA`O&}Z6EWTSLQWFev0KBBkFTK`vx0`d zQAnrt@}*h(uE{@e%>Qp1suA37lbkPhEQ`d)`?|)SVMjPi@vx<9Un9#|;rG<$Tg^+a zYj^aGIC}M4&oSBN8C>HNqr*GT+`mttEaR_)OQ*2d%fb?@O3dTD)e znT2|wYnLOY&rLM_4&N8v<*V1{apY4le(O)}?pwo;W!HZw{!jYIm(v7j7Q^_yU;JUH z$_5PW$k*)#lq#GiZ0N=yn&kf_R=(Lzdy$M5GJvor%cdg>FhUzyu0EgQ73L0 z>1yJN9$KVPmXtFh3nEf>cDg$bVsJyqh9vxgXspI~RBV(3^i4_NqLjQ?eYwveGx)AM zo8#JcG3qbS%Tvj(E|LY8BfEk0eWG4Sk_RG+NBJ>zf_$P5Vq{naER<;b!b_&2QPi~; zG`Yu3^E|2(@`-l#6lW-JU!9$9jl@u?!@)pxC|)hE7mD{TeL>Q@I~upJgfX3@eyrG? z;LYrnxm1GHCjx;2JbEYWW1QloIMlSS_fHNEg?V*Xwi?Ro zQ2u#E0C{-B$%mZp@qHI=Hwci=#;qQr^$hE5J$2G34jy=4x!v2Oz#Z{&*w_B1{c^dF zjad3KJj;up@9yw=ryjv80{Mxe%22tWx@1PZY|XA$=A|EHIo~hRmKOhu=&J4ZT3%7Z zLY+jRQDTMFU0yMqi!I2}k}Bq*GXK|4(_L=pVc#!j=5@w=Mq^ zV*_cmYda(r#IkKNooTziwks*1@dH}QAe@tb4#E1Lsr|oT3I{?v#W)kXU}68{qYN5S zuRMH@AUUyxx%>Hjm~Z7?I4te119#FI)eK;BB%KpBq72{t^2{QufgpU32VDc3ySCgL z(YTWO7X7#t@u+k|Nb}coOOyeDkysWiBE7D*RQ3@WkLYLt6tdWm0Y>Ct`BJaSV`gYw zc{jN2pJl&b$NsV+aL(n?=?7xNl?|Q%sblorz6}No3h%A%Qry^!7(+QFdq zjsz$nIM-1!3{jl{a=ZPh0ABi;`^ zrteZ@E%eM%n*sgP5<~S2)->*8ih5 zt3OonA^*R7kp7Q9@X%97oy2Zk)W5Dze)x>h#u+wJdJa26+Q)e$8z*j^RqJpRaH zbx)!Rz4-<@zsR%QymcYju3&pEr+cvy9*%|pjR`B^LA=~diS}IH1}q?E;k)6uT6PjD za7ggvZMAY;7Cg<#<`0LF5&!78FLSy=eDL(shuFK3;C zYV0RH{QoqqEFvi2Zqc7IVfa>a37%yFr1X)hzO^h&}Yhu;Z6ifgVN$Cs+}=M5~D%`yX%&xlg^vDg{ep6G?s z!ZNY3rxQu~hseA$7+DZjwLFM;*d`(+MV-#(uw~cKKIu7q@7)f)j6oBok(ff1cNg3x zUKoIiq$Qu3aKrsncOG!JRg=S+vepKU>X_O8>)PZ1-?E#dL9=p9KsB`wgm(wXPAvW! zXQoJG4l72Lb-epv;IE%}EVW3*cAWo6g+$5-$xvL&qE*8}PZ@wOmMg5eg)io<0Lt<) z@_7MTnO6|Lc#M=0%bOIklHFk_hrtW08%G^kr!0Xolr(6diph%BQrH!~g>aPPEQPk_ z@!l}`=y%zBh^llD=R++ywR<}&`J3liG|Q{JezrVJac634vvk~`CF2YV6iRC%WMD^L zjw@P6;34IVeVccR!_XIOo@$KMsuy`D+ydWY@gfmS*@j>iQ@dH1F(ut`wI||lm@^4% ztDh7fV9-6obM}8o?kuDK)aR;X z(+^uAvblx7)8+1c5*(j@2nNqMp63oO2O8wS2~g7l&fUraxzjl+hw7WCPi1HLswYEh zjiWw`pUX5x7%-Gnmn>E@4BO-M3w4^A=6srg4-fxIQ!_UvQl4BHGffrmO=gg-z*H}< zlL&D6kNwrJ`CAtYX z8sVPoC|a7?>4gG1l^^Egs z+RY7z(>*O+2p^Na_Du&odh^;e)6EhFi7ct6uR9$QKAdeIca{Pu61g9I)nDLtvj$gM z4)}J%faZCG1>GNBFJRy1`|18?hEE%k)GmaSZePs^J$x5u<(BmAqy@(bbfU19;X~-J-J_|){1tNoHf0PQ|{{U-LU4ryNM>yN2E3f*#F_-Z zd#{?V5PGov0{9GIQRnOe)uQqBq1!y!-wWPzod0eZ?Sv3VnC5fhx+JE|w%JaDlT#o# zHcgx${IREz^g~bn=s1JS2!KR-<)ik-$C+^FLefrp0qn1iw}|F_7Fk=UmDrwb_fsYL zL}ZJ&b*8~{vh02KFR{1WGW1w?F%(inNo3B%n|CwEC64S}Dy~o4)6LsX*V5_uHThTP zN-gD%QVe36MxmJ0gp$H`9A)eRDlH zx9>3${@wG?Jdy+!pYmN-{}M#p1XpYI?Pe7g&R5I z(d51`yniw!bORw%2Juq7p<3fVAHI-Y2Cc<&ynoY-zk==UnM_a?p>p=zCvcNTV#5;u zsm&m{7w)NS0Y%t^_1dlMDJDs3@s2!tNa=I9H{Ot@L|=J@NluE z@3yj;cidomps}-*j|mNK)kn2?J5Ne-ea`i}<`p9Hro5Q|kXX_mw#5hezBh>5Y>YeZ zz=|du)^wyFWXxH`PsklYxWQe$(!Pf)7)+j1gxuNg!}gAKgDyEtj)ZoJgLdbjJZh7S zp(n`m0}_BKPp}e6iAIM+ll-xUuXEFsRY#Fvu0o6d`!A4&TG>A7nou9~$na;PF+I?01Oj%51adm(%-!sg(%ezca}f5f+EC9syj z;@Dn6d30eUWsfeWAsNi6D*UWD1}nawT$Gg+Ye$A;BH9`zJ7#R5)MP!C_@-4b%!MLT zVSwok>HFK@r?69OuyD9fo8qUwQvH=&hoQ*M;(J%jw9g^@uX|Uux`W@q&xtl*7f?Vb zXq1tbm2M~8nnV$Z@p%a(kXkMLc$21^s8^;HB7|OprB6gelb>518+(_yv7kgMIcQ7` zvPwfX&|n9VS+d6#l(f_JJIrz{)7jmhWP$qlQYAr;KOws|-3MZZW{f;@;49yq@i(fe z!j=vpS_KvBgRFZei)23QX6>IDuTl)3ofbT_hYFL-t`9PfRMRiVKRoe{TVouea|aXK zw>Z^JhF*i^c>W$0iV;;N6OjorfkrZTZzHK8lCjx@8&_=R%{>;n58g4@x`_FF`X&iY z@jZ6)jPQL7ZYidniFNTTBt|x~0+;WAKL38%(NC8nD>IGmxCDn8W^-96xu;s81omCq!~FBMIpc=z_5zq zb|2`W@xIf))dyQ!V~v=}C9KZY68AY{()dFnX5VPF3ew|NQ5OgH7!{NI)ZN_ZGMSgx zt%NmfR~)o$!BAVnsJD4ZVXeKSbHA8WUGbu*;us^Rq$yUWk*~gJ2tkZMe1iysY_x_r zM}bwF87i)-FMsVhW z{tZCfD4`YBqP%+Zu;p`sc~I^Bx9qH5;xYyiq@w}rt3Yw3uQ8Fv12{GcmrS7F{-`FN z<48*b40XRR&9V~3*g@zKOgKh68cNmD$f(P zO11t>V6W10Z||!5G5NMzc)mi1z4ZYuLFa3!maHhMANJNolI|>xe)l!LPHEi*pYmZ} zZ)2ES!w)d%!Wiv#i9d`s8v)mfpG-seK5{KCx61{Itd;?}O`Y%uIUAeHNt!_>QIDPo z3IQ|;q5SouQ|I>9<>uAD=cy~n(nNS8D}UKHK66{suGb^#>4_Fd*+b*vv^~H6< z^48P``%9#5+>I9jjo#i+<5v{j-951{>=Eg&YoAv#xkLRxLjw=%`LW{L9sSE1T^d>) za_>itQ#_ih-c3CycyReS;_jVv(|9l9-W(H^5%qbB%-F{_e_j3C9gbN_&jw!^QvlkFDy5?K(SE$V5^mC(&eMao+bRDSeu2+{!(@iNYO7|YUV}l!sz`Wwf6_PcJ=M$ZmT6y655w6E@_i0GonFsa{ByIU|CG@(O1`gP}1EOmhG?x zUjO=^Ep`X)!je}dOZ8~k9M-N7yast>y&Tya;&VqG7yxd?=jE}=P_VDc=38oI77V}z{(PKdqeD12_TM( zPZDA)w7^99bA(wHe1QHw|6Q6qij?z_-`1Dg1TwMFa7<;FBS(e*s z!wrY+&nJ*?7o5JZfpc&Cp^BgWxh#>I6Zn|HCdORti6(Yt4tj1g;*^g|pNazt4-(pQ z8!XTVxrrN#vF@CDUgG+vg`dVV{ID#Zxn7d92B_Ua={SaCzD7yo!2jr|z1@&)$sU5z zq3aHTRnsT_M{XXms^z-!pZH+XGgbua#Ub9^@WGGXYMaYy^I6Lc4CXfd(i@FMggTaR zfVEPnPN3bI=IFp4cojjywxzo8z8MW49@{j{`z$7S(u#UbQ#4_AdU8og2%6X1QdtT# z1Ab=C_`rja^XzN}1WjRP=0ZUHh!HaHfeCVkJHR|1kQ{8Mgf^qJp1CkpxpWUOEstCo zk!or2csAwVyTX|$%*M?QdaPcq*Op<4=IPBXxF0@Xq{OAxuGOq;W-~)@%cvVkIQC4n zKR}UTz^rtW@F^inGOgqY_xAKGs}!Vua<-gpME>1_A&i)*Cslam-YMJRMsb5YoDU1T zbY?AMae!;1VovoK`bhS9BGSMCE@l5%?ZMlWj%@`9nL2!Y35j0x2)LP*;o(_yKI*lv ztuS<_A+ONUv`!pKF|OR$uG;R@DP$+8Ps?_OAgt^lka5Y9VY4R?oyL_z4p{VBJ<_OJ zSuD>MvEz=kHzi@KObJaBDQMPaW7IrnSF0puRx?c@IURIfl%%DS$D1WYarc9g_jOlj z2I#5W@8<6=Y80?J5O+uUxH3*{a$7!*w3975t&N_{tE$yZ%F={M)0N$#F}xw6r}J znna1Xryu7keh0LV>nn{%y~wu?GsXPmKKEi-P%pWD!(5t&c_mAB5cJm*nrSj&$75M} z@*FuIG;PS<*8OLM!kVGV81g=IXOjq!iB8)8c>4JIv*DoG%wp5*wXY*c{b3nqg=~mL zsog!((Y12aKQ|x&!+)RWwTM4d-xDZ=^4a#UB~Hz5?H=v&rdELweDcX0hW}r!m&{?* z&jQb(WS(Ds&zG?Rdh>Mvfojw2;N8|M?Y7HmnNgiEU3l4wHuZ zeYZF}AYt5%pz8>63u1quYcOA=(Rzyd-QNl$w~v$wI4xUthil1w9X4~tf&GP0!ntSO z(I$*ObE7q>I>`d|9AZq9zN6Sl!}(e7H}CXh*6eVq{V^uiOvDY&AZedn86nxwI3oEP zR{bHIcen9b=Nhc1CN}F#3HA%!uqk8W@HFYYd1NiHfOoMwHcwJk!+Sm#+q{EVpl7H6fq(0 zjrrxbRChka{NF*-@6f&{OV?nQNapgEX6VO!`1VntH)_x)=5ygK1x5O~t`y&0pN z^u$}aUo=6OifJ6!w3jShl=wvd+Pf2LkeQa3|22GuamYC>%6za90Gq96C*wvB)g3rF z$!qy-c@HY%x^7s*Zal&s{@lqN7Z^@W(s!lE#OZjtWeMQN5h0N&R$$5TlihU$i6aNB zJ9@g3a2$Cn^WC@iYEqEi^e~vvAS7!c(q!KvHO`yIf3169sorCEM@R=fy;z$=Sp(-q zdYHtYl*tgPS1IxFIFeLoXw?yWElUuEinKNO?? zxx(d=Z6}4tIi7?2^1aI;8u<7P&iZIHYeUK0IUG*n(BYZARGB}C^;95dJ7MQA^^gJ5 zsy=R%I86{8DxPKX@dfeBmh_AHt{Y}u{>DjtwXV!HAm|bTeX9g7@dsIv?1gmDEjP1NT`g`svEtXiCIkq`}o39gq8` zPY-W30cOSdra%EoTY`8qF{0Tp*b0}2_}~$exmkNMfh*=j!Yn&KzynaO3R({-AWSj0 z$+$jFCDXDokqd@NiN(<^Wy=2|G^|-YB8kvZxuBhKx>;CczZ}+hf^anRVaRN67J|Ls zgq_KM4E@GZ+`S%e8-ofw9Ni(NGH z46MRmSVRC+0xZO+Llo$^9@N9oRo{y=?<&!(yBZgR)u&Xd#DW=sdIT=vkAq4?p58D0 zMc8FQwlS=)%`PHVtJS!+@?M` ztFs);3Vv?HiS`Nyr@iIHG@saFXbqqDXuhB91eG%rjSf`ckigmd%b%~b z2+p$SfR|xb0lP9=&c&x7kqm3?3L|fUkQN*W0iC5ve zmpe{Hs&XD4c24Zd>(kIqxNHo)!86<;B2(0!HT7Nlx=ocddC}CbenPYiG#~5@j(ZWAbuT*z2 z6S@T7IN^+I5zOK$yf&w(G{bDl%vmK7X}PEDZ=`SGC?NF=5&uQXxP`aa2vH{;ny#5x z-KJs=E@wV(DbD<1E|KAf$vwDS`T0opUM7Z~wKMrBAuFXS0Z4k8tA9Kf*d{#7k-?df zp5Kh@sw<~fFdmTv9H1}(8C3p^ZeGlqlDohgL~lmYZ}4KIgrwfGVX7HWqHV&?;Z&{z zJy-+;&bjtx@k%Wn4NEPAM%&E*<$LE?D!91?bYw!#Oq}LY+NnpfQws}yE}gZNqT}As zW+Q=SU4avFL>~HoD()=S79)(^MgG)#MPKqj3BYr4Kdh{qB|?+q5uD|mcJtOtl27aa zopnQxrYTi4G&{cH-m}Ljew0@1+ub&$&iJl zR!EA|;@E)M8fm(39w%Q8`@b(+Cc9PD=Tqn0tXvaJp9CL}Fe-gKl#h@P2`<%1R&Qw( z^74-Sh@-?nYA`yC0^a?LGb&|8V1_?TNM2J|Hsq&A1Gq;3~1FtVaE{(ss1q>SCDr5jdNx+GN4MD;^s$VU40&sUjCP~OxZK#v<2QM8MoR=Q zU2{tdTdu89sC2`{w~C!tVUBQ`{E&VeAf-St`3iNmu7G{B*F_&v#2au_VbAY<&k~Lb zFZ^MCEe}gBZ75U9rs`@b?OyHmc*#L~@~E{<+8=XXSMRF)`!p-OO(P_RS0MXo$K+tC z&vgB+gVeD@Eg`<>kGM^J+&oKuxedGAPYflg`UF->w`MlOHr=KMQQSCBZbF1F&Z0$cdP@Vt4GuX!Wz7eVu6isAYjol;9i zJmTM#`KrSz_oPtN2dP^QWx6$1)$mTBNpwPuk0|`#xxeJ;9S@RQPtkmFm$YfAr7eUL z(|pq;Z)9Qdj9vW1i)b}!o*H2FNBypiHlziP6Me}VRMlvI@!7VPrgTO!B!0)+2)KdP zHW&e=(#WG_X)?3pXov-#7%4+W8Mtpq^aogyA4v|7tU;itN$YRq^|!8s80wYDDAN~u z71dlB&P+M~x~#$0+G2f*hwNy2N-W{WC^4@`#bOmo4JQ1_Ivkb)@LV#wBc(c2wN^<< z-R|>(;j$#8aRFUs8jqWLXi9xNouznrFYx6yk4jg>{&7*F0;crws2m&=Ss`8Bn$+8< zsqa{L>V&(!ngE)1y(ZJx5>cZ-DH5#81K+UF0D!nb>FZjd0Gsu5#$u|ECtp?_2f7i- zel!^9)J|n4sKxWE^JbPo)n=f6YOyR=b}f?s&BxxAAGlk|2~v&V1ile(P%V^47yEf zI36g4_`1-wPv8$dQCI8yzEXdlAIVowj z_j`{}x%d27-**D;{2HGK{>Qx>JSm!+n-KC3$U7dYG@gC4g!oPy-_rIEvgi}d{oYR6 zWuSUm_m+n$P!Ta5>JJfxm{mC3W?Y}mLtyiFL|>QcR1t)m!C21+ydDNWrh;6ElM(i^b*@=5NS4X>NW9}`sX)~%N zF^UgUUB(?{srRFfmZrNM zRXS6~HMEuVvqit+BM&X+}^d_1FcsmubQNig}+8@>JYA0-_a-}~ES6$Z{tuqkx*&(d6=s~myIt#9O zB1gNePJtCuSni!8emVI;h&@GGaW@88b1M2ox!_yX$(8q4a;Q&K+U|gEp^2}9Ae(yv zLXVnOO_#sC!>l3>@o3BvLSmX_dFu`6HB16}1fb3co=sT!QCXjOTh|G_WBxv@VrhP;buoN3WAYP;55?}xhC zJr#EkP05I;IsHU5NYouDLYMcM4jgmi5t6PMS(uxIGdlp2&aqlp-G@476`F{PcxSe{fqfl>wN2|jKuSE)I= zpr_5}sA};1m$?zD=h3V$K{;aCh_8qA5{IQTHIpKET2h>2C1|s{Lvj0C$PdB`Rb2o5 z#|sAc{2)tyw*3WA?ogU_ENM*R>S6v!6CCCNm9I`{^!-cRAp%iY!L{082)YZ!TdLClOU1#}C}b+R9(n$uNW(c@uV1tGW# z+PA(wn;F&rjPuX2t$)dmI_<{$h`c%E$qmJp-PCf$_GHGWceArqzTa^D%Eeex{ikn@ zQ`bjq`s9$X3O>mK{JxFixv8JBZ?>&en~`zeBSnT{_tUp)Ia@=*Xhiw%mf&I|xUzp1 zF((xWm%}AFz=mQaKZBQBB|2qnz6~D(O?6#&yZUCXQ(ArJ?jcimkIM!hO-qM(7YX1V zw+$zS+N_5GsEAE+p~;W&BwXg?SO^)|@+PlwaZ7x-*o5J4rZFgP?jd$5VE?Q1GBvz2 z#S^aSM3RUMm-J+-y3!}eEWPu{=( z$=t1%{*{3*{r{?IhDeVr5rJbD^cR20W!s{p#kjl{dky1zd|p>eoqvioVsq8?Qx_OF zP42JEm*tdh-A&3c+^ya*tLLRXpc7C&&(l~)Gi19I%O+v`W@4v})6g>s#U0j!vx;JD z?r(m1&U>rMJLpNOJI4-mk20dpA8`$?99k1qvaJIrq^0!ch!^loX8jK>5e831xuJ-&Qqcg0GDv5_#vRB&m`73j0*gBiZ z@-am`rTY|X%LT4JbMjkBcg6oCT=81svFfa@hHy9QY!U2(B3S-`c-yIZv9un9e0;f+ z{=g}%C@j{dh7#?>)NOn{rF6?2Q89NAx3w{KAo9r%m&fI`ddQY=M~O&W^nY?^pGDxU z5OBF@9fGfcV@@#*+C+)0oh>5t8FE}|75W{*wU887feO^8*8S=PVvAZHo+P5!FW{@G4sOG|ujwtkC+H9A{&BjWApgVnevaH#& z2y|I78qY`E8!t^Y2Uj40WlsEdca|!W;QXQd__Z6#ZRg5f%oAu9Y;R#$gCPq^#pjRq zRP&nM_0aTBL2Am#`x$b3iG!whOOwb`W&w_}DlOEqaaTPoH`vV~z_i}W`zOioNXKz_ zq+Hwp*cs^rUK(p}l!4pWkKKEMhSu`;rR#97(BzlX<3ntBBsATymSBe=1{zQ}N>~SN z`5zMWgObv(;$rZXOWc3%Px@OaFZvm!MnnCsd-trocYdsaFsFx%iruId!u%bF<&VlW zC&T1#)8zYpQeSreGyHLYLy&_GgP61JA(vJ+nlv3MQycnXq7m2#TqoDz9NO2mv}>P& zlaE&Vs5ux6aQxOjDC6PBT9&;O9$trdXKm59GqYvNcF(OX^GCY2ke*|H1x@KEvq?IF zGZbo;8*gDT(}w*D!_-19cyxb3%7JcUfj!2Q4v~tgoQ3TG%c-NrulIaf0fp-))#9MT zEf^zdXFpGhD!%-fP1nKghw9G`V7(uWw$9+8l7kH~DMOd~Lzy_9u=i}~tdupCAzwyy z#!zZ`XuQT9X+d9p+|#1BA0n4^XkEvt4P7=yNp2_)7m*K_v(Se|OoD+LC~W z7tg$TG{OEGw{?c60OEvqLNIdo7iGHe3&Iq|Kx!G}hz&zg&9LYWw$^4yc={-R%c18F zkZ5zVirHG1--daRzht^vM2gO)>J7MinB_~!& zI^UX&us9@EKu6KK=i;rd6nNZl$`32#sKpydKMa&Mx}%)-zKP$HpPfRgxJ^1fo8QhQ-|VusKiK}0EsS(b zi_`-Pamr!R=<2xYm%IoixgQy9yTmnWINNeBeHe3iMf;)!XYqKgUzSKIKw(qXqSXDU z1Rq-odjH)eGdCMj-1Zap2G0m3ec`#ovq-+f=dKT_QxRn`=J>Q{5w#fim!5$DEosN1RG5kGNZC3!$(8?+7(0Al zYESekP|K*O$b7emD4Le20!Nkf-eV_FM1!HWN%Op^8dGu7!(1l>kaT2bf~kXHR$^>j z_#|8edpD;XpPt9gu5w=N&C6x!;7x9A=f$2M!z@e?~OH2IZv^P7FDu z1UX)4de3|yQ8){U9(dF;$r$?ms{Av5v%N|~&IZoxDa`=LW<*KM`&WxhgYO+s)%;ju zfkfCMhp7P761E(J>m1~7v$V21m=SvDs9_%m3GxB4&)){gSchNf$oFaaAAwTbqoP*5 z%y{Jd$(Yt}WW4JC|Uc?kabhsXB_)h#D$wRc~ET)HJ+N6 zPd1EyE+dUna=QAZ8Y>k&_s@RA$fadwM&uOpAeNX?v@GkL$pJHWBiYo+`3joJc}jQb zwV;6vZ>AJ4z(&8wr)P%*wTtUnqxoTmc3e?Zc57CD&+AsgbRrH$=>6_!V5z{hRkMUI zEeN0V0gElE*uW}i8OE|?5N&@p@9Ri>QTufj{qV7l;n8*(Gx|eWm1_bS!XUrv zMEg3dyY^d3pPCP#>r^yu%2yOS%B0ne&BzrB>60M0Ay?%k74i6fddL)QE$)CEU&wdkNAKiLIV|YaaG%$a*x)GnWx-UHxWG^rk+zTa02dDo|2Drb z8mRYfyo6(|*e+8aukXP=P%@7%b7%Rj8TSBc#HDXFun||rb2BcI)MHD_h;)$$^Um$a zIug~{7xzEw#cwoxLumIAXL5j)@20hoY#aI-l(djFiN4w3Iit(E8_DteRA^O}Xq~N- zTR*1Rx>?~|lB%eQ2)la_*hN1nM?|F8vD7qZc@&3wDdgBz#r$*9B17ma&*3gXMRUI5 z5`2SVAfOd9Ksi>5xnuaq@K5s&4O@D3j+_#KEb+~3*0^^oG`~wQ@ar--dwI{Y6`mdW zQro_QWW(ClKz$wB!r&FtZ3826yvaf5F|^K$T`#!-D&X;imin#?)?mqBwHd9_g150d zKhuxoU>EsM+X$pIR-9N*^xlS>*YD6qC(A-HI)H#=Aj9P5<7P%6{^2z3fqKs;o7HAb z2t59kmSVD8M{!)B!I^}4e--R_h&)yG>{!EIV<$uv2OfI}p;CIrYx4T~{xauF*;WMqQHuw`d|||IFU!~`78lAw^0MT`%P|VdIYeIp>>dXr z`+y}PlyZ<^=n=I!E^g=;Ar`T}?UXOGT?Q`YWG3Vmi%VRrL$8anEIlgk!s5L3P1e4J zScF_`?u;pW_6_y+ro=dq({tde7!o(}-tX>U`qfVq-$tgoJF$CfhE2@y+Sp)j$P4pUzb+;MY+bIaZEVdYN; zBAnt+3-hxW+r5HDdwPW6E=G7uy5d*xP&CKE4(TN79oUNF;Jre43)|{7b{3=v-QL_z zd8P|eKBm};vD#)CulX7a0yGCyegd`|0)jp>ZkkBll!Q`>N8&WIuoV6D`;gD0?E2t| zzuXbs=*d-bZCXYDyOG;Le9=SPxpKfEC7v2p_mgr(;%if*lsUnfUG%VeteKS&H)B_L z){r8c__IN=J=wQWGl{WS@7c8ZoRQhl*d0PZ2tH3C3vK@X#l}izxH&mYPCzm`k|>H~ zCAKvq`M~W+YKja&;m1x=IUQCaBh)ZzTXI=TpJ=Jhbd5x|$tv^LG_o;7A|5!>^6DIe zaXLY(5wwySzIab*X57h3IT%#NUUx7>G&8!5s2udX5@D#Q`eR(ZClN(H#euSQ zgvwtUlH%Xom#i_-{s=&|)>pVNWrzM%Iv_er!ZX}$uB>U+&D#*SN(A|BlTErr9)+hs zKysG1hD>CqSuer{BRLN_%t!`y7TUFmlC(y^+~1EaV3TTNESl=cQx@u}PM4amNm-VG!Tfyr~_Od zO!JaP1?|ZS%YV|(K`H0^hrj9<}?DG`+OYbSa9!ZB3b@}Ix4i}LEv^eq9d~oV5 z-tW1HVPPE*mW+s-w)FQVJZ}X@>=XO^tzK=~C5fcltAYL!{s@MhIz%ct5@{E^Bdj{9 zc*8`K&SW?v3D;+q#6>N>-B!HOKRC*@*zr0-n>kz)B#XM=lbi@sF3C^UE5Y9}Jxja{ z3FovAVNfpTh+kZ zzvPm-`BV6bF~J!-6w+hw{n*hjHZjXQK!Kpz6qm;}Gs07bH#!mo^AY6QWKVyExp)(d zTnPyxWCHLNJj9FM9w@_rKPq)TT!g=J^?P*mZ=Ez1Wi06 z;LvP3mks|n3!tnXic;WbuLhp>av84(clfs{3hD~2aj1xj5T{IYA3CmBl?Vs#%3>OT zFM(ro4Pt~l;ASBjb5{Fx4dNQPgaKNL7-(6lgPj>2uWycZbxISls{;y58(Kf(;6x!G zlP0^7;RHkMB_yVW=?6VsZw);6BYCeQbMyfqqaE0OEa%I=uK0ZN%cHR;kWJ=E(yL4) zbaDKwWLdh1dJx}Ad!6Gan(ZlVpikyO89_*r9GVT)BtFg^o}x4(f4Y#WwV)S)>3c`p z0=k)RIp0l2stg4FGyO_;%pmr;!&||?xLW@&@%>8!+`u(w-|~l9sAt`;X0&UQ0KLe$ zvxuab9Eje%BSTk#SZf^7R6=rnabt!yCI2i4=Yu7}%#)*%XL$A@*RrK6*Z4Aqc*OZ; zo1hi-TJOxO)vtB$Wtv1Z=P5Q#>DVluk30C#ooZ`&TBE#Xv~lX11k*hfmk;qUWw}xv zW2Vg)Mo}!u-0dD}03^ySS~k~&Te3zN+!qZBw9>tU(^eYJr!}-H{oh;5mgTca3mwhU z(zAy#gH1(!@AKJnY>mctK`u+hx+&gb#p9WA0nLt_wat>o0ISKev6)}u8U|h~unnF0 zy?e-jA^lGTy~9m^y7z|)P;Uo8vN_0Zk8nJt?b*sdg)pon;iEApJa>Db9%Uuk(P=Z* z|6Kl-v4Ve51Z$O&iI9Q72ik21g$8VF7##oCwcTH6hZ2 ze_hFTs^b!X>IB7Nv@hXPV-HwtpBjQ5e&0$PZaQ7K4YpX(sv>W0Du<2>Flj5!0#O|f z6zSwOv;TqCT_;6vMP26IuVgkV@Eel^qieUrKp)5Inln;>1wnAjo9No!&fJH2MHBcqNe zQ|=c9)9}2kKJoaeGzO2eNKS;ELW`4QCpx2@AC+A7g^28r_qJ!55<m^(`hE~2bDN&43+(iI|UC7DP}xB&GXE!YYls|6ivtfA;kTA1A8Fx0Pnc@5oy@J zW0P6xIHg8u|42d=QTgwHb|JFUjnpVCsxzTG07Uy*YhlR|uP`E=)igTV>#=$EF4F0v z#Xr+22l!4WO8yUb@BAIv_I-hN)Uj>bw$m{>?AW$#vtz5{4lA~kj&0jc#eV5~@8>(- zcz?pHpK9z=<5Z1P3v;bC=iZZMI~f{3$9t{C7Q!9Bl&kWt-qeh^rV*eUro_m!egU7b zOBoSL<(W23P3OjzsntMPt(M1ql2?$M`5mZ|tZ%ZhJk?{|YFp$~FzK%}zXi!nP8uTR zTDlhHRS#|03TETTM^+Ij6|B1K+klksts_@*VsDiqAAUA%UOHMMa5+fQk@xk74VFYy zylOPP6grJnpmn`vc}t}X&uo5ed zV|iwUuYIt}^Isb=pzYm!N+>&y1rfY>w|mW3myiGD>i(UHAdp=7hD{4}O5K2HN9P#| z|CzxXBxCWT>~Eok$er*7Mho@1PSJ+#UPs4asUWtR*D*X?K|c}urEGv{{5)HF08FQc zf1B?lWsB+j2CY$OOvkCjAwoWEG{0(S(t%G>B&Vi71FlSsh0SSIxu_Zou1V+YsNVPm z$uzzG5VTHTtFxtD)B?FQ_IZ~s*KXs&i^sxhU?pZy<#b+{|Mek(5*vgAKzD~{oORay zp>8C8fm%BEY+4(6bfKFKCBfGuYH&5x!F^N3Ooav1^GdU)@L$vBToW zK2<7ofN12bF15(%46K?6UoSba)ThWRC#_#Pr?AIJhR6G2@4-`|q*!^{ z#_O|u5*WnRaVAEyDN7gV+p+N`jFA~8?&-SFx3$QIK17ovwO)v*ukCY9bS_v1_apbi z53bd{(|lH^{!I}A`<|taP(EJh$>ZPUt6An?@h%d;gi<-0%Zm+cs`JmY&1AA*hxPfZI#q#}L|+0D#K zBn5E`k!qK!K^tKImE8m_>6#yA^u>OvUYS84Aj}K^p7`D4B`zx^A1TMCH=?~Ke1TgC zLLzlsSrf7|Cvk?g9dfo{bvW7azj%fU9Uv@{Q4se)+(W?3ehL(l*yTS-fh@cDee<%h z4XJsdrLoD9vgy`w3FJLTU3D!4N{pOt%3FVs?tiqrNSxL}2e?LOOml@9NQNcscfyOO zLRyyiO?U1;GdwVIPI+$neAWDUj=Q??!Q1#|-c|78mFE6(%Wbvn>MulK;}cYd=a|w! z+u-6?$2o;mTLwfnQdLzX671Rx?Z|AnO&$9_Dc7RAh^^n0qE*bAT5^fJB8`zUMp<7aF>P^aLlpHZPn6Iy>S`Hau_m}_3C>C zy*+cFeiRa9im$e_0b>knwYpn6f3UcR&s~T+3wU9*hqsOT&lurc!M|c!5S*Fk%D!>3 znD5?ozQpw>LI=GY#Td*8egx|xolJ}XbHI0 zsa?YDFybLH@}9EYS1=wgrEm4Lygl2EwK_`9?Pv5PxyaD^GY4+` zade)Zso$ElMQ|BdCH-ygf6 zBKmXkI&Dr3Iqnw_Ay&2QBo9uJt8MC&m3CqTiIwG5dWD>+bPJiMZh7Fe3lKkas;>7V zJfGqrZ)Z|0LE#%MF?~BKTYviC-i1LD&ZiV3psUC)?x*Hd_OHaWUN*&A2_^>64#+N$ zV`i6BsHa`_I9F@y{jZNB1ipbeW1S&KR+)q0)}7M~|Cc?EjC@sp{u+*8UKXo<@;H8L zPzxUh4X&b7pzCcx`8b;tZI*-&5G)|R5iE+$6%>U~w*tHaF`Yj;FwM`4V{j5pw#Rqd zx42IUeWjb;cCbz4+f+nvM)($oI95_B0T$_u5_(T7f+y3t?@G85j;Qz10~AL&oh%?iCyO7aX*$O&o?opYOSKCWNfZc?_Ik~vV3Zs)#|QKPp}|L z&?80!mvF9A=(84Yy*--jEZV%lm*@&RiO>B=MX3w>t$)gR_p0^Oy0=5DS$Bed&Ufw{ zGiWrDmvSxIx~ujNU;eY96-R78Cgr3g(k@>*Hsg_8;6Kn&OlW)jVt_98F-kX`pJ_M!WEPvW;%dIoFMSTERMaws^f8}#Ufi5nVr86S z$D~)0R!COse0FKSV>@cVY^3ttp_*6WWX0qby^E2^@ zbDFW{Y!+J+Cg-XfB@&R+UKBJ3KXWtA?MdqeualCKQ{F$Cz5V0YE<*S;b9s9tqvgEZ zYjwB1>7()Ccys4zZyl<}SiZ1E3s^j!9i6}8t2%a6N)j<1-tT&H^Ev+EHz&Au7tN5V zwxf}W+r%($#UAnuxZI2u;8x9RIr(a{3#IDUt+u?&!7|WI8}wHF*k)K$9K-g|Jd($eD7r+^5Z4&q_??MmTHRo zhy1Kv3Q{&Jw~SQ^UrFt?vdUY@>h0M*&3H~iNfyepBvWsXXFI~qe#(F5|BDgy`UA-L z0R+@UcM1g3Pc9hccAhrk_4Zg>ukzAAp!krIlYjesHD>-L@OFC%{c3B|g68wqWXkGE z-bvC?&zGt*vz%LENk+`fm+A{Gb9YZ3kjpBby>?VoR4YyZI`Dh|GjE?1{KjCZbCR-3 zesYL_MrjhRbR z+^38%sbVo!NCS`0Bl1mS_>5B);+S}T?!ofC7t!J( z`WlDZF{4NJ|S>mpGvVw$EXXei8jY>$8XQ z*8c>8$RLoJ%O_yGOjQQ|ci~^xOvL<5PYN4Gm*N+nqoOQ0D5NZ4I=6n z#H8~JG8ky^vEp5mt-%Rx6Da1o`6y1rKg`OpRQ`~rr^wR^VzQO#WjTf7o5(GenYMOZG)4=I3`q+y5VWZeSCCu}$Y2HWNKz&m~ z)6lv3p3eHkMQ%3|#jmpo7APod6OU6sW+z2yO1o~cbaqCiO(qKpLe|83+^-xQ&nCMK zaXa|Y#0xo_E8DX&xhqFuE@P*-yk=8P=90j^jCv#gR7{?4x%tS*U~{Wd{3)-u^Urq4 zF3evk!bx85IXwIV&t6IdrDi8mBB@XRPjzTt_{7CaRK<%BkO&d5K>%f@-Od?*8OO|N zopkXmSQ;NOggTjW*#}GH2KL#VZ_;gjKrStn%6KL|Zl5I%m}@+XonZxh7ktriEC_Cb zYdA}}XwzHKDzV=nd#dUL_e&F>AaZD1ly=)B*Kv+w`)c@j1t`LQA?%Y&N%5e@7VAlrW3FP)#yMiL#(zDhHbG_ z<@onjrUTc3+!4^JVU$oeIq)PZ`Mn*x`!8!I$bs)CqBA#pTJAvNy8Si>WU=;)MztV( zcUe4mhtPJd-qi>W`uLNc3!_6k?_bv(t%CpZ`-3F5pnhB-b}Cmv{*4zAF+vKzq^y2) zma(5Uma{@_^TZLEKWxyjW6G4k#b3VbN!ndB3{gT#?8xoCo$;$p6v z6k)M7y75(nSWcpaD+1t#<5fU zd7gi!dbTi%(cfK!sUYh{QuVa*khrFxdc)AALZ~-oUX0Q6 zmwjFl$xImO#y2wvA00dy(YWdSpuh|htt z-#M7g%QMMeBzQ8I?m!r66*MX6wTE3*ig!_n&VjLn7#%C+xFokW(X{STR)GaM#~ox} zHb+>XM`aGa=CbgTy28u$@)`hH`DQI_Er0sG%@;Cn87quk*@GE7qtPbc=iy9PT?l0E zP(~^}l%+znZ!7#J8H=2oVMKSK!|_UK%zi-F8M#zN{ODx@zUXm<3`+FruiLXB0NcwL zVNe(FWsKn9%kJKfmo_xt97s4B(t%pc!RpZ6-T#X!TO9t$9v|9c zU}t$XXyi-n>rmoMAnf#Ztpcxc3v84r&wg=iYEGw|W(WLCBOdGx!qbExJ`j2sUjKSe zMtFq5PuKI_1d8>g(1^YfE!e#3HSzX{=ZS#_7MPATY?oiPuSI;t`a7(adKsy|Bgsq1 z)22L0d5G(4^BPV%!qjh#K3w7~kP#PFHJSTreH=T+_(e5H=#XEYIOpc~{(8}EV0^M| zV;^9p$&heS2Jb+cr^p+k$7$b6ABglJ*0Z2Ho>&gUmowiONM!qTo?{{O6Oc_JpBVDl z@Dl$JDO&`ot%h6bLK?gRvXe20)MG-!&2#nL<3UE>=!O=pPQ^YFuk;D-c;BX}od0A& zT%!8^<&O>vGK=`1X>co8M1LeiZw`J@2Y@T^l$*g@{HjlcM5-nXwH$RyT&vCzgftwi zGJp_d6~@F6)NdOajWhgH)ChM4ntHY9sy9}ok?IFHwCrf|<+&NUd}dG)TKYE@G5SoK zVRfYv5oa^{{sb*hsgUosOa3O9v}dzjeUY#&KF8Pc?+3`IJCOhC5C)Hlc5OsTxg_GWlVx-}Yd@StDty)P~PfGE*#=%RDP>)q>WpjV2K z!-FdpUO7)|`?n7M$q3%e+WhbP4LxvRndz~YV1|o(L4>*srIAZLv^P5~QTVVVxwx>9 zM<-yDYW-O!q-2wBa}!nlQiPiVE1q%UAU$Y0(8Impb~q<0S6aq3HP-cg;I!tA=pFTHiE_c z5Phzaf#@P?o5rir%t)gY6!;9ub{B`$mZB>3CWaapzOar@91GUZ*|7}i!`>3I2bBfB zVEta$iD*A+8BJ#M?~u-sm=H95BorhF-%u&mgf9lCKX3*fiS6;sNw$8|gj{OV>)mi8 zsY=?B?)g!t>>4sHftb?tg097Ud$YJS5PesB{xY2n-?-p9YwA37R4?dCYC|U8rtq}+ z${EZH%`z4w0$)BN*XzlqXdyK# zoACCy`U^z7Zj;h>p#gB8nd3jTwnL~tY=KV}KGLVvcgG`PwHwsOvSql91pSV|3p>C< zX>pD1g=I*zlTTuM5L#GyjLsq6gc?fi7&&a{1bE_`G)bH3fhm@YVDxa*MKWkb{z2`v z(q(`l&L{nChMF=lKgWM+Nbw{r#JZ#_^prpxXL2G=Ss^lXj*GO^5Vz3*50fiQC~vaK z5$=kbGN&TB|I-GBvv@>;$`<0@sA1r}dMi4puq7JUk)-$a+K#mFG+Diu)`RFu(!T4u z;QTKAo6n_gA+ov#pc!W-mymjOBXqt&yK^LA$dku-ZJ;yb-KG$cJKB%Uhj0j9+cv_R zmbm1-CT^AQ>X$%bc7?{I=-1t5C>FkeR!$+uAJu0oc_r?d?Mt?#|uot*jXo@jg6|q7aQxN6;3-BY``ZY-WaA| z(G0t4+4A(5^x`oD-$%UIe+!kVt8S54rSaKro1sQp{&;Zm#@|hC&O>;x zU6hKFk3yPX>a4jOh?$xEAl$=fa;ikSnytz|^X_H;NO~q3a)?{qiqLEH~K^6&F_Y^wJUBxpi4YPS@&-?`{!xf7kYdr;8?T>>)sOo<+F*iem&|n zpTqf$)B6<#-B>K5jxg%XC3mnFE@tBm@9QdMp;RBKfVbciIo+6gpkBMoHt##trnjUy zv2V;v3A+MKd&&1sLT~Jk?pS0b6J`+QtCyGCug6Y;Bveo-M_nHU%nl=EuU zXwG&qu8TCEe5gKNAMR(eMA)x@sSmz1?JxY-GTdqw?T$a#y@+c@;0BlX23He!=6@S? zBUEyt(m+=>J&If=W$G2-ko&M&FoxZ151yLqJ3#SaM2}*b3619t_7cCZ_T}t%TkB?t zbPu;6QHorkfTB;?Gc6=I_b&MkkfPhF+sKtq@uE`xrel>NEZ7q8#yBnccG z{sDWf3x}+LNSno}(U~sEX6Sj)86tPOCbEt}W}+44-A6nnH8HY=Nz!!ok5_%g1^vl} z*hqTte2T;cf;2>xRk*fOBn=b&-BlCkwVqJ4Qm5ofdJiR6JTM-sL)Ha}{qlVk-M9Ql zCm<_|ua~Wt_*2Y&9kUnc*=|XZt$^09N0j`Cy&C6sD`d@D{ap{r!3Sh1~eyP5~Ofz>p5Ye2cGUJfAwAEX4(!Q?$;%rWb4O&7^)5=thHtjdwG|L*3PCCE>??f%e?D$UeF`ygg0-NKV1om zhbPviy$*_X!NH_+v@TX`smEy@OR&BLtJCIqN7?=WmFjxG#1akdl_5yALwTbSnc4o| z01-|a9M6IvDQw(jg_x-sIb1<{u`r!AmM`2kk-Z|>4y2!x+%tkjqdy2~@*1}AkkD9U z9qPD%F!YwBL;a421&(ewwSY2omS#;jeQ-Evk3ow*+sYV+<@K3%=A=XE4I?CsA<1s< z`$?E=F26!2J)`ubt&%$s9ax0eosmd%!ZCam@DZdg8l70p=6CCDaj#B>+m);aj6eE#5Ysa5#@ z*8YF}B>_hMLXTK;*p)ymHJ>w{bIZT^`!5oC01F6yC_r8e{RGY+ zU%UlEL=1`wOK7>TuSON@6u$S^S4mNXdjK1cNXBb!Cp}{;o893Sf!h*G&7SFpHuwkwsHKs4dv`Y zjf+9}J=m?r%Aavd8al$?YQAo59)gNE;%)yy;zk2eUEo_n;`05ElOx60A#~?|es&OH zK;NZike{qA_}~7JL!gkts2ach;QC5EK$#zjTaV9}3k`4s=n!@qL$FJMzY(h z{-L$VI6Jb!M$coDuk@M(A2x>{jP<;;A0`QmP*%k|nrlWp?rkS%^)6)e zU9GP3U3tCTJgef#_T*Mr02LL1{5qh~lmBT-$Mf~Ls0^CSd@gDE`DhR@lavQR;BW!& zD$!E0xm#keQvFGW#TD#t#LUMXOzT^y;d;4Hq&h2z)lTTk_B>BMoKNGChAmN@i3JIF zGRv&Ds;h6($zahL359cGlLT}Tv)4;X9q&yhMLHHepU(i3txKw|OC(99iB`boyt1-* zt8oAAvAK(G`4Ep2*LnU^wzX4&pFN;*$eioT#;xC0z>%MCl(zZ0HihOsIApV%q$SK( zX5*?^-u%QNKTNUcZDt+`2h7^}-pjph#XIx2JPJCmJUY&wua&`Txs`G<%aBq;&&S?T^wKZ!(XJ%g1Qu zdfoOGG#%S@ejZ%kb!+GR+#l!rlxSxuXOwmEcpx`&Vl&;!ZdKn^DPht#XsKe-64=1o zP|&EOx;r?ssZq|pddgnm*u3&!{as5tf92%9w1!qOu~=T&)4s0rI$7)1_0q%h!cA{> zsMmS^!cIk8H*ZGSFx7e@EHyWd!2SD5x9Fc(+vH-A;ME6ds|6Viq=Q9*N6PSxxc+>d6CR{Oq-gWUm5!$!s&fZ8A;2ac(cLJb5~4w!s>*3H;B1F zy+pSD&<%#;vCHT46zZo0#uD}OO)HuGLi6pltBSd=g7PLjFI9(2+V5kZvHW>l>CINP zr0Q*Kkf9u}T9L4RBEDH-~BW6pDi~I^XG+JD~j-`~05Dpe9d~Iwn-0tK;3}omjQt z_x>VirCz0n8kOksqFJ$I6XUrhp*J7l>F@>4t&gle zRaCv~h(2E{d}y6^y-BS%mTDD2q^~r%c1(QCRBEa;ep_{&FKuwD$Ooqg9ntN4>c1(o^Uho7LqwjDptUjQhDihLa=}s_dlR=`MZJPZx8H zYUb2RZ)W2SL}AbyOVao@--bg@%r3SeV1&3+l^%cmd1|n$?9%&CX&>X@cx@Cgk162Q z0fWM6F+9?Acfjg5H-dit+%M=OvHALF4Y6B&5*FxroTJs?o~aO3^$|Uhqgn{*cN5Mr z?@W~UE27GD{(Q9v8kO=l6MK%qpX0GuA5}N+)4CCB-4eac`!xBESB&-NS-VocdZng~ zI+gWK3idvO^AKSuxRs0ybVLdr{bR=L&}i;yUyg;EPi&ThX& z3JYB0md1Et|9KGf=Dhv?LB4ie(95Wzr*)p~#=lc2FjD=(Wu-H(t%Si#46jviz34*w zFxPxy2CB5Gky;jrlI;AAO}ltTav6xHw;T1aI>YCzvoI)Q!C26b|y>#QKNx z!&^(a9E!t>Mcw_QNC#yVfp#GibaKz~B2}saZeK~kR#2F)t<&5ROLle&>PB%9vbjvE zS|xId<9#vGV+`NLDNxC0VUn{G^3PtEpBLPUsUUz}LSgv0oZ3sfqs;0Rv8CKIf#1y2 zqxQWHe|KjBS*@NlxjXTJOdX?(qSFd zJHuVo$P-vnd?EnD%nG@JlO7>Jug;z4+qDE*MYD8jE~ohFVrOVKvEpZxOG*cZUwn)g zltpS4pWltEXP_o=8j@wPMhyEb(o$A`v&sxD%O~RBaVFu`Vm-_$((~RW_W3yc##1>4 zfb@HJu_1WVSbs+(oJkt-e}8=0r0wEY#%-Lp9LrK7e@UEQsg-tnJ9rpbd4Y6s-Wu+G z9uyT~L`AL4<)EISWEghey<)wt_IQ)gY`gJ}882P@ct3H-_8oTNF|j@t-idD1@_LYx zTUTP=tQ?~Ej5lNx$-4fM4@dE}kzF*yU_sO@&*osZS-#8)U4e1~TnvDE5&J*riT+s! z5~9mi#eFVZr4k_>DyV)6v1mr61F&{$Y15t3?5(CsfSD@1o~sx}GQc!{_0?g*A)Cf z1zuD!$0%#3zx*?I3})Kr@kY1US8hhxTF&eMfl(xp-h~NIL<;-DoR#*OFo3{ zmu4_4$NZVugt_IJ3Rk$I(j!;VPVgd;j35wimO*lAY&Rz#<=;(O_&WS?({sO9oBTo7 z`H{|@!zt8s)><1r%lLIb?p<>~8kB3GNDl&glZ!IKC05ckIXNw~n{~(fd6|{D8lGdc zHD=*-z2>`Q?`qRz8^n2-FvID zrfgJkz>P}X_FYPGw%YTi2K>?4nqS!{74w7ffi5j-P?jo{{M7O3(z!4%17J%-vQdk7 z2W=e-R&tAyTN7{ud^KAVGGsWK!s^FJdDOq|uJk zeFhEos$YsKKc6oBq(CFVOQT7mz`y~*pA@6b|7y1$C)?(noPH=c*BNgVLLN@%|Ee=C z@xOdjZjrp$X?^X6|JsFri?)BO3{;SSo~Nn@F`5)MT=B5oggSm`^c;uV#DjcAY% zLKD|!Mthg7-Ni^{tewh-CWOx;RdcsBnxJZ2mOy2(JxC1_ zfGO;`K>_sye;Td+QEl$H@EmO7F$XJQV!kxap&dBQFe<>OoYcCV@gHrSBp8&?!goJk zf0&+IeGB$zu2SnOV=673W2md(D5ncdhRQ3M$jGoMqyRA?>)EbxF=;q;;1?K4TpX%y zssPL91tY#|JxWB3za$4m0g31)5A`P-PhEm!D4|8>>Vo2`2G&`)`%EL>;!S~E!0n08 zSx)g0e_hX9McOYFw~)t(ifrx~Ls`%?`ht3BO^2I3YU2q$1JWnC(R8(2aOKt(pll?OH^bNhXnhWJYp~;wgs7Dur|eA zi`Q2orX|809LR!E52`|)%(j9rVGopfLP-aWnTpzRv1@@W2h&f$B_WCfe1)sJYISJ? z1A$2CaV7|Gq9MUAI9?ErSfvZ*Ye!dVAe@MT!Qf9KCA9^Gh5okoA!`RQzx;*NnihUh zF(P6J2!(gjRBDO^hxP3BD6N=~^KoQ`d|&hMMQZPeVPbV)j#Xc35H~g@=zi~m8Iv8T zqM_9X&1)N-d`cT&uwcE~XY^NuG&3Miim!NG>_V^hA{aHV+iO&u1RpK_eqmZ|n);lMS$xPO$k>))Iq=#a8o`Pd#6YxU`A{`{DPc1Q|Kc<8~`6 zi%y@>czuu24n_1Dqd6&!YpQ>#Z1=jkH(t^QdephVlWI zSh+uUHQ(H?N>d%`wZS%#5RKK(zOzQPvh8#7-K}{IRVn2)iieC|dLle%&!>ZEDJh3x zTbU^goDx=ssVaP!wP7H30-0M|>3bh31w%A!@<&;unpn{$71E;#HlEz?hg0K})cYU< zEpM@Jm&kORId}iv@Yx$$%w>A68q6ea4*TUgh zb$d<)tHPM4a3R7FdM`EuUgOcxDPsmZyW`z=E^xPJw@6T?LEP2FLOb%jAHlP37Dc3@ zzxbM(bo)MtMCgO0mk5F^+(CY-aq>N*Z5RE>EtC^#;nO;z_maWy%qgI02(ikzi#2Wn z+Pysm!@iK({Q`sJ&+mFfZhjU>uzbhXctOotaHo1R^ymOABgUF2Vy*EtGAQ|!G5R>* zygcTBr%wQXamJRQN)j`fVZ>#@nhD8JE$WS%^Q?;) zCqhYrlYt&02P9r*9A3`=mHs++k-xum<*#4*w8cJL=yn}<0vG^E2~|acMw{gr8qI2S zQyX#5PZiAu3rMPqi~elc{*?_YIHxXixSG@Q)x`#65NakVTO+qE?j|e+m=sQCuVi60 zh7xNptw@sPVk=HcB-t5HHRAb;Zfx6tn}o0f^jbmNZqCa^a2^ltCY?s?w^=j8q_B>dC-QIuTcW&;Si;k(F9(QJ7AGly?b`?a=bhP+y>)<;SVM4vEHI2IUWbh zM^1=_Bea{ufI`sK#!CswLE+|?i%bJ44S#?s;{aowV1J3z4b$1cbvWk=)A>H9LvKK-c2yK4L3nUdaoMGLFqGfPFO}+od6CWyY8zT2Rc>(2-qJL{girs(J)4sYy7@+ zB66Nfc+sDSgZwxsKc*RES%YgQu<8U6qACYTN!Tl9(Ohl8FK8xfVn+*bp9hse{GPM@ z?ZXs9B#&*+ZCuvynfomVOsW$V#T|~cCm$IXoD+-zcchcE`w4<{@q(;h12=%(3+s&@ z*AJlXSb7U`S^6jULj0WIyJ)}Q~azt|QZ>Gy}lLq0tfUHN}=80v{nx(lucwryJ@Vf4LqG#=X*LdF*i6JddH^9_vRE6IBE zfrOb-S7&R&ec*T7=$t7K*lR~HMCaPc@FNN=x5)-`{H=iTblcxlT)~V2s{sWb=upL% z4yGp5eBP+`>r?Mg)bm@(vATgeCgK_cigRcu*Xx*i!bbO^Rzdj5`T0bQ>dHESwIZ~k z{5()PZEiikC*mK;jCdnQdHb-Y!TSK6&_Sz~HFo!z>RGBnhTcf>Itac`wiK3Bec^C- z;&PRVCcQD{8;9a>1Wah`1)95NnSt%3FUO)f0e3Q8eFBKIz4;{Hl{LknBeDD5Sz*8! zP9w`Uiw$SgXjPGNns3$>v{?{h#W$uHzKpO2lwN}QX6I@5{H_Z{*267Am^DqX-{||z z7*QrYi`BV%+5lmHP(!lo!tbvZy~Ch2VutMN*+RDcB4x>{Zc-gCuuu$R%lYL;bYLQ~ z0Xf^x+E}&kJW3-yh{_R^dGF_tTkNVdQU`lUp zM3W7IMlIf0zpFS2sG|iKO~uMRbtApN1sKy&aTvLaf%hTAZK^*b5UkZSb^IGyOkXS@@8fbU zknqd{J4_jUwOX(@JY>Fq*8)SYX9E99L_vYm6x{*6GD2spRe(H96S`w*83ZlvhuT=B z8DWz*ukOM`7#P!n<%eB($Wu0maeZ%_|mEZifRIyTZ4J*+$Ip zJ=P>EYBtk2>7%@!EL;u3mzZyX_e+q(x==(!U|$%_RErIo?;X404RC)6$FyfAMgtBf z%pphZOt%$^TfpIHvw$BCo+YZ5Z>#c$Bu}QOM|iv@XM)^4B1S}j1kw~~)?WIcGt^!K z?(X1l#KFL%_~ev}0|@EJ97Pn$TA&_48^+9rF3IYJrOZ3LUxR3Dp%F95E(c6t1$zG^ zjCy^m(Nfo}*DsuxeI6hec+=$ieZbWT-*-pw%;=a-GXT!4umu(HC)|a)$s*kuDgjub zeea^qS9--C0NhyuDmBw)?7cyeu6rlTeR^gbRF)YynT5cl|LCX`I0HSnI@{`ORPa>4 zB+cg$sZOlTr2ZMI^bNh@D--E_w`9f9Jd1lqPgx2zJuBc2KS{JX_AL7$ar9#>V2y^Q zWpj84*H`}#OCt4zcCry%8GTre2#L*N%hlK35O3~!i8rrO!UbE*;&}6xf4Oi~l$r__ z_z@!lkXyZ4K#e748OSVyAUsOjTCiJVeSE)&cm(bqS0|a0dS0RG%^qHZ-kE`rMYd z^H{K}%;`V|Pv zS!^A(uN)cb&V4x*onvy)( z^NiVkWs9;B7Li?qOv#);d!yEOXmEjjoMtb}PPicj~cE$MIpDba#B!OW10EXVqx8dIu5kDjBVuUDoX;NVD zq@_bbj2>87xZ@Rp;X_bpMVAr{JpN7yJQvH_!hss*o;riIxJcHwG=+G>(O1Cl?kvuD(x411+=Z6kAEIc-Sh7Hz_Py^BrvIiUXjfe-TP?4WbL-J7GvY z8YGz09s<{4wu;=i5l~L(?la>DjS2$N;%4t?n*N5|R3y;DHj-bTmC;nHjQm~ix2<0C zaZH8-Gag|3x}YKOq0N3CKQ>Pbb#j)cBHO|bv)@4;1ROD{)`EBGj#jdn0t$22%4Wk;<1ScX~~pr!<_ed zlO^kh4M7~>6%Fdi%yKpfb-ipIMAczFOeUNREj>mkRyKLSZld=u2!Q)r(ZUHOg#Cqo zeuN-x<0JiiA-33u53ak$0(1xO%yEE}PKk|tQtDn`>2+Yl z{OM1pR-Rg~mD%QkvUaQ@49?(3j~24_ou!O?H0vc_ELDsIK_#>4+`c@CMc)W04s4t| z#-Hy&YRMs`sCuChDQ2kR^&t!UYS)KG&p`x}Jq);N6Q1}okt=;qP z#2r61Jdkm~Nc-RCkd+m@%hci(bR3ZJS<=9dfIR~9!jAv``sVt7+5yk)`_|;3hu9#z zz8-cgzI=G0xhajIJx3R2F_3ElGv;Z(7+!nl(zyB=Rrd=?WC zv-!gr(QHu>nKb!}@Xoq>1*}?rbL+O&4N-i$lqC2B2WSFYu8o+9-(ORNhB6;FO^K%f zLFrk3PH8a1*7ZEZMN9DKIRbT9RaWS(-%}pbXap`r`iMk(U=DrVQ9D+q9+Cf_@75ku9G5T>ZFV`Py{Ttw7iw?x9ty(_U>dsdWuSOYr`< z6SO2-rLL?QnGXYDM3@^+%k#W*N{w1$>_(%CO4A>M6>C;?8Iygp`Gc&)hDMrmH`ix4j!M;wiPxU2AsnOhAsp0_E@&OMLEtiqGK{opu zjZsCEM{hRAgE%2is$L%)e@ZXYa(3XVMv3RZk1ueP%Ey?L-A-ocD2@ipeNt>j=OE@V@3C1IEa^_{o& zu4d>W1x<+U$MBUZrbfezQTB7ku>ld;-*ei(Y9J88b$Js-roM+mL@gR_9_U@4d41Br z>8Jl6&fY32uBD9@B{&3kcXxMp*Wm8%?(XjH1W0fR5F8r!27@ zJWLoM4ZI*+pJXprQ!f(L664^MKc^8LB7hlQX@n(YT#b5*b?s(-LJf5z2t^Q9dvrcNS?utG*Ux?Z z;*LRcIR67ZI$BF9l&ex}$#J15PqWd9c2sPxr3hwYcfOwNrsFIlIqfB!DcEQGV8)RO zZPdhCD{LwbTqgtefRk*>Tw-A(0Bq7#!gD7$aB3YwMauTbA};I%3weJDy{R(i1-E2G z*JJwNsq%>IH`&C*VYe;$Q;m#8)~Q2K-M!mxXXrse&7W&u)>`z9bayXWtf%}jOBVWr z^<0T!MPu8|=h#{+KMM+K1NZ|ud|%iKAYsFmp<7 z%4=7yf(ed-u|=?>3Mr|VTJHeOr<(z``S6Jpuax*4aQWupMhkoerRmg})4tP)j-;BG$y^&&#+pY&@>fsA z3&rL=E=u9@C9A-sen=i}Tx^q$<7E#*vMRlZD+1aE z!89$(Sq)gDY|1yev_F|siKsaJ+%Me)G?L3YA(A8ChqAcQla5-^{g_Z$P>ci1PV`Kf zo(;`LwiGCBh0$Ho6}p*<_N+v52tc{ZIX9Ux^kZv3rb0SZv}1>HsBlglN|l|sU< zsp5X6d?;-vX2Cp(YyW#lX9iZ;oU%0^1dF zFWIVXaREqUl*cLZd{GSut7b}Z`X9{s2-EvCe%zi5T5 z;Gl!bT%i!o1p|pLM8^v%g;G`FJs9DCjRr-YEFudpHy^^F8iTE#xB_VSX_4EAJE>5Q zj;y#clVK1=RVDmS>>H2hkdm{l`+7T^M~irSISV1*=S&|gHFg;}faq8Ays<<%f~O+!N(7V9?WLf5%ytw_!=$6_-FeohN|agV)7Ydhvs z4laq-nf>#4l&4@bm^}_F^vy#+yn#kT=IBixiiQr5wpDwx;@u8c((gavJU)j&+5YO3pMnr%jlf>`a}N z&KFoa3b8|Z%ift4FcXaJ@01q}DWmK=`3use?QzoI?9--UGsqbRg7+$s02Qsjtw5g1e5;Ry;r9jo+hD8>JSLWgDG6| z!t^s{_OnNXe*g8s3lgEn99o&l&be>X(Q>>ScT)wGRPRqW-a=1^z6caJ59ahoM>PDU zIQaHtE{x@Its6sV`=Dm|*CZcL-g$=&#GFA?7(CT)g+#CLgdNyEoYlSh6tFlUB*0n^ z#0qB)!*AVV$o%j0v60NZ$1C7(-oGrQ?KpSKF964Jyhh<)@Z0Z1Zm3WZY&BtH9Aul~ zV*0x9aTC9XZZ^2fD`nRsR&tz6(64cR9+eTX=op|W%UW;#$NZa@7zvM*a8zE>N%I7i zl~#f%H_t8hDsn!2x)Ytoq$8Wz*I1I6fsoQBeD`iwdTCH|#;cb@e zR}!2kMgZy>#!WWX><%~55@2j1REl9tf4924^gQAh+=n`2qM-(7)XR+@Lapg);4MyD zcduS@Eef%YgfB`s0(eg}f9bI)bnZoTU^1J5sR0ppUJPo`*9TU#iq^ha#Mb9LY={CURy_e|bdK zUr@;+Ay4>QPQD|9B5WB8+g0XBt@7Z;IU}}9aa20cXiwyw5$@tQYk&K*yk&ps_y=g& zI&a{Gj$_R~m2pN#yeUcFtoc8S2K(=F#=3d!j1nY^{H44a_J_2jlGOrnIQkQx-uH?R zY!tOyC;?wGM0XZ5aphi0BCMbX~u+OgDWujQK8=3?<)$VPhn!HxH`w9qJ z(ko0*)sw5bdwJ;UiX6GgRp&Y%`1S#}vYrsnB4>tHHkAku{H(|ei02ZRO*mZC5!16_ zebzPM13!X)9=@cEZ4kIE7oaW;vxs*HYEP@WO2N*lAwke7Ad!Kp8}_KufJl<@5^ zrN4Ec>p|%wQ0g)DFNV_Ri}W3j+@LQPLc#i4h;&v9bIdKl|ldJA0&n=fjP9 zoIY-J$3w9j*G_1WH%|29@yM#brazXpqKM3Vb37~xaY_Y$|fEcDlWnLa%QYZ ze?}c37Glk|kQ|=T1H>iaWH9>g%lLzA7%879urYQuAF;l&tY=1Kn1K>@eMx5nhjiV; zN^e9OaX3w>!L}!GMCits;k{=ZB>I<^Bto}-DiK=#$Yh`>K`ya@sO_8#1M#4&)`#)T z0q^R!&~O7o1yv#iTO~m1^xUhMsiG#lD6#I({=2gxsnitSSd~UDtlQZ-LnM9Q3}D-U zQk9jkN!puX9-kCJo*jzmDXaYQI_5+C1!1Gu6u3}fH;)+N-P|{McT@Mh7+azRxOG5? zQJf6k(!>VBxyA#MwSsB5cmGWr!H39bTFUi$YGBo?+gSA4e*6SGJV-vLaKn*ZkCRU+ zxea$?GCO9j&)GAqspro^oN#G!oYn?i;F@GyQz=_3)X>g+{m!PI7f83QocPpx2-4Ao z@{|Vl0DlwyVxEWw3`PAOP`Vf*>PJ?8^NmpQ1{IZoLhwyYL1Bk0CTR~@n-$a{0r7nc z@s%0)A{rt8N5(IM5*C74( zWsplmU`o2#jellBKeNYH7cHR9Wlt9)y>jsSXlmz#95*SY3BFeFfgqVB`RXpm!)WHk zMp^djZ$*mE(2NaV%q27PjMlcG{#){ahW-xAauQw0-%hwpFkDdUV=; zNa{JLI_l$nbCOrfQzXLlTEw^txPcqgjMpuHU zYx%;oa2t}^Eim3+PERzu%L982n5UoTn+Y%HN`%--fbart_|IS=j3stBZf!xSYMx3g zTh?S_V_0eV_t=1o^fvjTCRH1S;yhW0U77TW`~53e6T-YKw1deCS8WW^fmj3|7E?wn zXB6q@71XU&85Vjn{9>lzHo{H0EpClOB2|YoYio6cZ+jqP0bwqvBd+kZ3zmy+P$L2V z22j4o{SBbpA23=lB6A%;rzWRjLT1??(7RHIw|{0$GSZiiUrAFdEqZMHP)5 zALR6ftCbWVK-dPw*zwAfkChckbZo6h18;2?Hg$7>F}7{1J(D zNVZF9!HJIRyjLk~pt z_^i>PnaT&@>WMP7yLYidDU06T^GOcMy%$6<*MsB-y~*a z?fpEY+ozbs_+AGQz}NKb!pG@^!vbrqWK$a!2jjRZ=d2(x)`yLbJ|`ISP-SvvsKy7+ z-q*+2URMZ0zQ^u}39cC%pJ&ac&MevM}9rZqNn^ zRnw~>tpBd1wnA1Pe)SRrrYZox7ro@FSf6{kmcg5jhiO4crMC|skso0-7384d*Q&&| z?!fmAER1JQ7-*Aj9o3VXA`>zmx>RvVtWJ!ivtT?5{Oe^|swqvG?M^4;+?Q*R0%#c? zsX86yedZ?a)X~q*wI=@+s8ZnZ#625_|3h}jFXiNf8`Z0T;~@`|!6A=cS-qf^0;~|L zn}w|;tA*l918NL{hMn{CysbUdXnhU+umypFsQMg>5|SMxI&b73>^nITgFBKwF1~y_ zcKvlf3y4~jP7LXEA~Mc9k?qaIxvb~6u6+o^PnZPBB*RROJ!TyPcxZRdcmgRjfbsE^ z)HFPXbBbzHWMZAlAsu}pi;bg(ragaze~V~3GTC1a7YJ)jQN{lLj;;Ur zY{!>*i|T&9nqWiFirYn?j$%sc@>N)DH;_(8yj90}p`bP3MNqqi^?3XIJif57FnnxS<1d9|Br&YlwEVn2w)MjM7&!S@1LUdg zQPr@NT>LVb?oilRqV8?Wrpn+y21J=Q#$FEn$g7u$C;o zjmHGhGB9n_3H<3t>1v4q%~_3r-~J#TT0r>^VEH8E&?jhByL|^(Ga_Hz`-wDJt!c2wjf*-U?rO6ZWLb$x_dl4K$x@ zY{OBgb)-7ex%5n4WJ%tBYp32B4rFa2O+BI7yWGD=OFl~#B^yt3yl(@QaX>Lm;yN^K z(!HH+V@)MGmSQ7Y@5k#R`y+CVJn6&+Lt$iY`ULVaK@;3K07A8(>;8N_UlTPP@E2{P zdOirg^a{?-%;*LXu3>E<1&|I@_EAM?I&?(Z!LQ)z54@+eaCSI&qBq)VW|%d7JQV!K z3liH@L~bP;wOp~eoTWr%_B*$KR1ryXBZ8Qy^@Vc~a6*?O`wzqa#$yVZK^tP-P1ttZ z{$}Hjh&K1*lUf9VTitn+vQZAhpb`b zPph5KKyQ@`==X4)6qPxE4e@;1)1rcWOTvQyS0nbwvm}iet1J%qw)}7 zWcRmrBf2&POj}+Y!pCvw?$tla)aL*u|5{DC!nC~UpWj3EQ`}0JU{s7;0JB}K7clm9 zU3ao0xfxZ>Y|pXvNYhR+0PkBn#FN)kMg9DqhGkZu#qCwNG#UhV$bQ`)BC7)qGK-92 zy={=tmw8I8{NCT?yrm6~&=?FW5eQL!K(_Ih0@SNz>oyamcKRMvZ&P(iG58i3UM_Wx zT|u{2>!=uBR1o^rh0XeLG09woc@;n(4Thv%ckyX+zF#{|3y`>f?}$!?LcHLY>kiY* zpaLIwT^osDWZT)D#p~18p`&BfR;M~dGD=J1TRABn{ui29$lr5@3 zPDVuuD@{Z>EJ1JzQU2b_$w8>v)rddR>R5-pwv9H4O!r2w2S4*2HBKvPko_u(JQh+4 zY3p<*h`Pg5Mn-5@?t=kiOy~O_qH!L)9XD!

?1|1GeVRh0%XX?_4_y4;uf7)C^j3 zkPa(|1MGqolS^+p^X##`QvWOdKZ+OQeY{M~C)udl+7cd`@+GrginTZi_BM`9=n z#(Jp;z+A$!m)vF|D%2Va7SaO)(T0uih0D##|DtJh#Gj;D3slAZui2f86ZFd8 z?nI`Q&|)qSSd8CzIUQ4AG`z=I_@kVWqgpq}PZD+$e;(!Sg_2}^3ckI$X5b$GMH!xd z+2i~x;~!yP0T)PE(pjXGMMw=-ThKTnAF>iC4+lxF-Rjnl?B>k7<&w-IJIzoLL;_MVv7NZ})@vN8 zeSwb(oxn;H)?GV;ID9R&IyFp@gEfNspfoNiA0A2hFV~ev*+2mVZXXY#1>^oeVgM4b z^y^?8ZUO?ZOKPJ9#b@U5LpNI9E7$;@IkM*Njt38oycE?iVC&e$60SshqI-EU6vKg_OD3gTd*PcAqx1>gIr=(miN4rkqg2_`Q zLWBM&^A6HdfI>be+B{y(X5P4MY~gHi;j|LI!HEn8JON6B-A5nFsH<={q3t-bwx)xx zTff9?22UWIWA}BnkCRRx2n|}^b_0$Sg0Yf490+Wtz{@5f!&CZwrfF)$b=pE5KM+vf z$nrCrqB#6?9s`vj4z`GuY7{GClx~Qn{>>v_D*nwQ7m!d*WkBbefa6r*PhFW!YCZ5& zV`$R{WLtnQX&DJy8nP7!chtQPXB84nVsQeI`8Ls$#DTZJ_1G(1r5RoR(Z?pf6$m_(|m-UMJ z8~9ylkuBl+i!Uiy8l%ZXOYIi$s2ypehIbI|Yp=e9jhzfVLb@7+cY z?{Iv6afkxBuMizP_aY;&8CRE*lGBYE8hA^c2m~>(%rBimo?A6en#y?b@$`Y{vJ1pl zH5ouYQ*es7c9?8jeLy>6e7LXevPwOYj4B4P6)-4i0Y4Vn<*mUe zl|9ig;F5y$*8*sU!G1FLW&Ji=+)y-2IDOmoBM8cbS}HeFw(JV`LA(f+qCmf2Zw&~; zyl*eGiw_m|*dZc+Y6G;#y`CTk?o`oAlqxIeXd|hj3%$MjbtB8QH>|V+dpFdVRHev6 z$0=dA1KXSdZt`G7obq!(+`;{3Za->|w?vhW^D2Jd2@MW{jR=X}SVYW{k_7zL<0!WP zV^0Bx%5Nykk|9ASak7Wb=j7FbHlN>W|Hj{_{~GhuD%V_}zz_XpOO*O`l5vg#WgGTu z%g#=1ml$$XnHHfGwKWBY{~LdU-}82tKbTUk6G)}MX3#3~o6!tSoo7UeL#_YxGq~~Y zwEH5bA4l}^YH*AFDU_mc=1ESppTnCGi$1ME@ShmOLc2b$Ob>>hYtmn>``7xG$~*D$ zg!Ti?e!t0bc^+M{;}XicqtodElN*lh#+NQx&%L(ED^$7%1_9a~zxKi4ZkMC&MG5;9 zyYc(kA$du%mAja4RLPp)dQT&|L>yl&$1|^Jy1H+y#~G|*MhM?93D>4!p~h6GDesF+ zLtd2>v4ce;HbZeQfs~3yTVFklM_)U+>eldA8eJTxR$Q2k0ueqIe&LR)Ayn{-GMz7t z#CEwCRkZ-^veaOc^5tfvSu)i@EkH|p;8mdg87tU-i1rDuISIqOWny2UD?bX_u3l6H zU!SlU>cLN$5H2K#>UiRJQ^d|ByK=Z1HTGc#V5f4Bom6Jb>|~W@i8`~Rm$cxcKAnny zzDQ13frRUdvW$g<`kIcZZ!R5Rt(i-5OrDe*Y@P~WME}@wA?I#(lIt^QyVAAI-I$`? zI(=e7Gxn{**OGT0-*Gq>i;^am=6*34h3?GPL@R6T+dAC2;dW2{hg*K?f%rk;D;oxa zs`pVwWK8BCZoG8mp7+JmO8%>1?hCBM2M>Dm9s+PY?_H` z{(!8qksz!SpH=Aig8-LF#y_6<<+5KnW=5;3HhPDN?;6uL`1!c9a1WwpoGIJl{gG-r zp9w>KXfF##xjYsm@Fs5t51C1kP)p127lQ*YHO0andwNWT(~K>E+qL`4!#$awd>WcTVkTR7N0a5`2?_ z38MnjPako+Cw9ajZx=VY!gM8+p=TXAlqm|TNMM^KLDG8$)ys)4c`v*I4S@m!!$!u; z-kJNRU*eAt;r|^@Of*Wm@9M@z+PCFen;TDZXYe?Dl8=z;O&fpX*znf7myaul6$QV~ zDGd{ukmcRxg5>GI2mRL}R5JA*w|}(mzln(e%uj{-gTeWlFRI5&q>#8_@zOh=!DpUf zhm4w@ZaGk~tUF<0BP*D!9ctT{WI8&@G-Mf54|O?tZnt=l6Lhj_O|!$7|CQ%fbOT}T z{u)8eZk~v481<#)Ezm*ZoC1@qsM3K1u^NzuXuN!@$h7lE5*R-chK%n`JiUOZTyF>b znT0;^3$~-P^J#Mnl3m&Xa73b1NxIWC4NpQ-gTBl$G0;9G-@l{SDKAnfKGiHQ6?!64 z(+v4e4Joz&iwT8Vl|3=&;Vu!7w{3|GQNufs7C7X|dc<2k6vX0@sFFM%8D$qerSA=n zxDi0;zhE;ouPHt1$5yxB|jXbQU-dWs!q>8%Ze%5}kl7w9)_`!g13 zn`{-fbLG%S6EYFl`O-gd98O8n%}Uw(baiY3;NO!W80xksO&nkNPAlYxP&;uFbNI+e ztJ#o)2tgbtG9=y}lu@w5y{!!&H&l~3)i0535|I27uHIA-mBc+~=yF~RF$nhImFyfl zpW9CPYf7}J)E4ZR+j~v;+Tb76{xkounJ4?-CZQT6CDIxj5)^uGvWpc&Mn-&l-)%j) z)^q2~D!n9CZsf{zh6oEom(|r4_SbBh!zF5rjauzcKF)yGgR7Tmb{<$^#npsV>_%l> z-E(JZmpFAhp@V+m3OgrbR}Km%Jbl0Ups9U2NrW|W6vhnBpgGgAVyC5O22mHb+k2BcmhYP5B$-HEwH_HVK#UJXQq= zTr6cV*Jv=Bckh&JR`o*HX&1q)!QP5-rG8|Z-f*O{otm++XU&A{_t0ywve3CA*djZ@ z{fQvRU~f(6trX=u@K8dlBDE7j##gr}`x$VtYcVS5ZqC$c)vMUNqi?n}Fi=&Di1(H0$o=zE&SK!{JG;{fA+0UXT)vS$_DP?q;C`-Guj$D@(j| z6g_pfCsxxzAPT&avCvRIF=^G?Kl#Vzu)%v6EG>z5O^ptVJ+7fcScH0>DUt!DcRNV7 z<$lWQ@LFhkP?bO>wZ03Fi-}bPSg1W9N>cJ==@IyR)De39CDcU76_P^fz;pwMe~qCe zlEVAwK0v}N)XBV^i2iT<5PMIoH`L_<{ysX#%8a;HqFB??wt&*bLvPs&E~@-gvsW#v z>qhy^fj{I%H7{_x-G4bCL`)B1C=SCS_ejv+6fR0o05zy}f79u<_T(FgmrsK2$lj`f zw7&2^RfR4rG_-imzl7uOfc_=$e?Nwh=#zaA29kRGBk2DX1OE3u_-7$9b`kV{uKy7u zJ|!W=zwBHQB)R<`P5S@8pX{Ie|L^O6pZ5R5{Q?sKpoy#W$gPXWNCO)*HbE>5=CXQ! z%))*#xmLN3|D3$M-)!qp)jc%;;}-mH$d9Pty9IrMU!=~{p>?aE=d^=for&7PsWxr2 z*p(*iFw_mylh{tYJebC}zGuQIk=rSq4tNVR?-hc>?A#B4j+q@{ra=N; zhMPBX8y;}vwE+ji^DLf6{UuZYf!9S=|OIxa4Z{oT|@#(vW^XC61f;TS6^UIv|;(+GRReN3X zhqawez=~V;b8~7Oh!?SnOOX5s*|?Vw{2tbBcFn%`mcP>mO1~DJI|IV{$u;#F zm7y3<$N`!L$XUrr@pKL9)MmlzE6ezbdaN!hOs%T0ZYn2a^JRE9@5sf)q(^aKDyNYp z^3it*69U_spM6v9G}FwJh|K;tm_IPT1-xgoRsCey{{7|s((S|N2hM&rEFUi~A()Ru zp<*gR8|zVc08+Xgbui5wlpm=pW$c~DY%=b@nifQSAV57PHx>;~W;E2oQm9V=*`8JVed@jEQL*>5Qsp5SC0D40gxU{MleWP-kz%HW3JlEqF=2%Z9| zCVOz4N8%2B|k`z2j$bH^4AjFsyl*K{g39T$@{Y{TQ4+8gGBWMaTbj%}JXlcdux=B$yA zWtE!;VreRIU*)U`j)a86&?Ul0(NWe*ATTWXZ&M2{`AZ1M&7pom$A-vSeaBuhQ_qw; zhra#>u=-V`0xoOIfu_NYv{@Mug@fVn9mAn&RN@Ax3{7eD@g^E5r1&35ZILSJsD(E) z&@7@uvVC*^rs2Fc5lBIAQuJ+)xV;HQJj`pjmLXj41QAh<{+l+bNJub}kY53C7Mebw zuSqR~h?f>DYz`va`{p91)yn`xK6(G^rf9NgDs(P9cmr$@d;1AYBQ$Z!8Vm`Ye%jze zrJ$i7$)on9)J@r&qqv$+^Zt~Jg(D{N!4uBi;q2U3;%MA-GPU+dM;NWg)V-(S1EQbw z3bE3$LW(Bj>So$btv|6gm{G6Z^#v|X7ZsqsM+U^k5Hw{g$Zw zEpAv7wQ65`Tr)LpVvj?U-rdy(H}?LX#@AywYN!S#M8gX?U*90kW}Cm>UJ;pvY6uB< zKh_p%>?)@>mvanPN-zm%KZ#cE?UAh3i)&|y5yCRrsI8?Fp%2jb3?6K5+4QM~hyrWA zbEckgFQ!mL;;9^|?LONBN%D@8RggCiX?@1mgVPi|w^sO}E2ziN zjjiFGqp_gT64~HKTjnKlCO+hC&2W7-06-#AX$i(sR4OUCW~%jo8%$AaWf;QA1BS>> zb(or3R3|ydDW9ge1`>LFX{xcs51fZv1VDu$ls_tXHqaz&87<-$q@W#5xKAA;MO^G) zBoVGTA(g1|cXS9d*qRf9@F-DGC1u_4JxTi5GN$4QDmR6+w3!rEVNrs=!q#WBL=GH{ z;e}k`PJ7_{dVc`tt{Jx*V%D9_+0)dJ}9$%DIw=13lX$f(#jr=Z-zHc_`tp3A^As zO~nb_v^YV`Yw67hdaqQk#jRchzL1W@O&#fEgUFx8w973IVIkOwW^zJV%?wlQkc2Xy z0e|6f;7EA{Ng9qkfNyK>D4s)}uYMckg?+`SG)>AE$&5^b-8hzfnA%J%hXqw}pjRqJ zN-6c<-Eq30(5?3Fu}CZQ!y`PgRomL&Y*yiQM1+er(MieaU&JpE8TkXqzqvAbLG@_q zMP{q8(y5Og)2UHW(FOf#hU7>?@HSw?;_vg?`0z0Cm?W{}F#7J3+78?LMN5s*VsuL{}%jemXb!jjwv?@uHYvPOp-t4bP=2d5c0N8SHu{KRR zXqDSIBCm#<_`qSQ+E#^MYGF*tXMVr;q$X7pv$bL3gb_)@=1CgcnpWjx`FJ~CPgd#B ze&KlEt7=cwc?RO>N2Bwy(jbkZALQ8>NJ4NcN?SHXJOhM3&k{f+Jt<;FrQncM~%gXvz z!qWCcBmmIO3qtVnK@c^0>i9BB(xb~S61=(4a5llyby5VVV7%n`*klW7=Ip_5Z^GZ# zy1S%jUTZ%hcR@an%zz8xv(9Qf@+Kz33_gMdrPT&hYHHkMG@AqG%3;(;O|YOrzFQ*k z91e=rU2$)BXcM#)gm$>iwh8n%tFj}iODg{yy97q#NKjz?GBe$3M)-2Ltre4?zQ2{X zhhAZZ_L?5_xmnZ-NY@idNEMaYsS)Ahb7>)v2g-?16DAS1MvacO-jB%XaoNwSb91TR z25T;W6bfDmtcrpGxUgK#rt2r;gB?kL%}%S4ZM+ZFOM@ zWyB?)>ZY=s@TN=0_7(pRrJP6$g*82!Xkm0-uMdFi$P-Vr1<<3GSeciXOhUA~xW!&x zOd~5&B6?tH5}!A+EW;6V$W!`!`{xKBs%D;ZR@m%K-^HUAy(`yR%w#?oU-5kjx) zl$6X!U%Kmj2s%c&&lR@m+kT=D@_N^?SZWzCVl9l*B5g(?Ee#!VDJV%-k!vKW7`8K2hCsrxxG~cAjjBGhks$p?bT#d6gdljJEtSZ@Lzl1)`V?PG{3@ z{uQWDf>8hRHE^z7+^C*`u)xRPazk?PTfL`b#P*kF0{g)6v(!*0r~qx!s|h(5<0l$J zTyh*pR3v%3Mk!cV{HfH4!2|ZHm|f1ThX0t$KUi)KUdp#){O-E9%HRbKZgo%bq$cGl z9`3y3PRBLw7yK<|KK6S(yBsWB99b5H7Q^s?ME*kEhMQ7JfLIb*NfP+VwB?R;W$~(; znie(9p&un*$=cj_p}BLN@Pp3;ixu}cyV+v19pz~@QI^fAMjvOhjO5$G=pZegl-MyT_W(l3e0YO#muR~3`;5?#&I&%feapEwJSb6q+o z6o|d;pM~*?Vx))e?T)zpC^TGKFIU?6Va~TEfGo!aGLo6Mc0YvOC3r?NosMv^!jYc! z3y!fzdap1{3Hpn4ad})VFMe)&Qjn$({WZT)g1$losjj^+q}Go){msQ$1PW21J`d+< zg=fb^ax5@zR%(XMYqgr&E-ToyfhGyVp{Y!Dc*4}h%xBl&^++7B=Q8SBDUX8^1tLb> zA8EU3F$;Lz3*4TwLTe;+L5e~l3UvcpOs>8>Z?dzYobaeHO1TiLPrYN^q+j&+!W$UA z`aKvQ3BWcgE`i^96t~?{l?evGKO^9T150ZAdV)fj@9!LhTw*2ShUT5D^Nk}W^gk9i znMhxHRN%T_c{-l1Pd1h_(3z!Fhw||_u+?V#`lMsg+tMr;{*nES08HhYs%#^Si&aY zdqg}weRS`uz3UKDIYnLpaAArsbY)PX%8f=X@(_aULj3?RAPtWz_jrhx#U{K`pq z(sC7@*A>8v;UCHSe#62NijglrUfrAaDOKl%p%FLRb994eju;>zO^g zB!~hz0M82J9ASFK+MX9n0g4=n-432HG7R+u3jd z==t#RY;vxXuyf+DJ}~nIX8J-emY=RrMNKuP46If>CK^Ap(Ubn?T-T=;iU#>hb*r6G zP*U5Uec1w?y~3n&rO}}+nyrCPT;H^2DHXAL>w!-<#RR%nhjUUir zk;)ZbweGTfWBU2{1V|wkRuG^Ik|Rk=HkiKka)C+OE)nLw7QT~RiK66WiQT2s6w$v1 zTg?p8vgX4n*DQ0Yr54a>$G*6D^`*jfy<(Uz5RpPKL3{an;I0#e6VW>1xWwmbqKeRw zsTwb-x^46Mq(@VuS>v~cf;z(y z^~E(XpvhsHrSBEIKiRIJr?+*$!w-u-HM zz?o?Kyqc0Jd);a&{G12NYXSr^w|FR;QJg(=wLHoGby9wH>MrftSr(TAAvx4IOATuL*9Qz z%m*sCY3Wk{XnRM6*ybxS5*CP+TDv+n9wa~vvUeAqv zgoQM>W(RlD#FlPd6RYCc!DB@dg8dseAV znp>qqjmC2gN-_i(Z(Lx4)n=J_MxIb_Iq2!L>fFZ^M|FQltX}w5oli2<5dU`;z(pgM z(oYGyD|b|e*w3_WEU-jQE0GgQ!U%QpyDY{P-uLQ?!^d=04;HilRZm$=$&xG4ngDelV%dT~(=dTw9&T)7*AonkgAhKKC`1*Dpf*5BH5qbS_?ReZuSZwzq*Ka@2+O-iu zE@K3g8@vnpEPNqSbNki(gSWBMV^OcUKK1(p0*RDz@1iUUdlUu5V5Yg;bVYX*bCiUM zv0VM2zLJn|BErl|9|PT3AH#}IPBZUKz`+Cp+*oiS@|TN0M(gbtT`m`I=PR5W{12-x z7)*GcxW}z%TZKmLmfzisB}-U|Sn37O1Fo%1PvRTSx&dv$NOJf~wX=MH5DDu?!hl=< z*Kh5%Z9|2-|= zj!uvheZMas{#zzuj;k^czR?~Twj%z$JbX`6nDv$8?$1!NdTP*A;Ao%+NI_GNKX@bP zmDQ*}FAZQpW21t%03FQ}I4~4TX%{tw#D5yFryuAA4&&`WZW$RLZbZ)`d%>iJRVS;K zNLF^rFy`<;zLtnYbxKL;GLGj7Km=8rQQ6c)GIf^-Exvl^Y-E9;>(TVk%kIu+Sp+6_d~5<9*Xr zXd=5JuhZC`ms^rovlqHinlCuE8}(n?J?w>TeDD!Y1q1t<)$vvv;o)rpeoDqP$Q*C5ab(Nus{(ZV#jkNrP8NWqiSmvXkRmEQ!_BqaF-A~_y)EKC(M_9 z)TR-#e#+bzn=D@T2?dmW@MRgrT)8ZZCgC_d*+!jTC2f|UTlzf&vc10-8Sb>VUOZfg z60U(=F|)-}AiJdK$s(Z>k85wYtfK{vZ)K2FJAvgt?vTPotnUjrT&iAH#U?=(!oL6S!;o5wk}4^~%PH&G4id9g8(OEn}evB`N5IpbfdR9=$oc zWUJ(DRV>+* z&(?=9g%h^Y<%Swn8{hm3axP;wN6ZGd)4E9&U@hUjp?fR+2<6XY!Jx8@W3(V?eBV!( z?Ti&Dr@tT!r(gtZ>50pPbDE$k#^+lfsU;?ZfnG8f~GEOvHdZ?vX><0v#WH_c$I z5E@#!5EW}}(9Ay^_jk|Qmgc=V%trb^O;9XrV?pvUfT2Rc1M-m6juWP)_Zs+u^ zBAmjwAc&lgOWHa}ho^x-t!()8fMz@1%cDt_{<}J$>@Wu3Rn7*OUwDu8$)?1leq7C6DwEbgOT<*O}>Qncm)O;t#^1K zJ}+;@!JqS{5F%Rp@(3oM1f1^g6Kp&k#hXbsrG+OrIenA;QHrDl?aKh3t3U~pmedR= z2yl{O>ob-iCSZ(D_3WOK$mMcC8nBhYaXgC?a-XzVVUe}CYqCK2njmps*o&3Ki?&lX z-h=4>>M~BSfOsX%pOlXsky&B-+4#K{6Sd5^zwS$9Oq8g>vj6SGX;5jC0G2?hAjz%! z>0&d4%>!QVS+^Hb$*gqUFR?SGUnNCJS4~+2!MXZTO!X#nP$>&X$QbwuUqqpZLW~eZ zi@l9^D|#bg5k<^Dwpwp$2u+TA2v)f;Lm!*uV!uHhjv(93OzA_APEQWg%}`8KR(MMCFl6=--yk zN`}RVIGgHCWO8GoQOFJ*5kt(jtAW2coy4WG;5C0-dI2)AV+%JrX1`YOt48lq}-PE$W@fP1cw%drB6@fPB zdvpM#W&}_f3K-uyYqL9mhtzJw#c=Q0JOV;>U|a6)d1e{daIRqa%=1y;MSidjO|{{0 ztaApUA+f;c)s9FjNTmL1qWS(0k9_mDMMi$t(AKVaOX4rh7 zTWUXeneyILs=QD?bZi~*?gSy`j1&4Y(;{5Q8@AFDwrRdF{+tx#97nkUW7xV=zg!wWtg`rGW9H16wa#R3(`Mh=l}ARVt7c!fQ= zqnW)Zrb>1pJo_E4l|vDN-ju4!wOO^8=Jfq;_EIR(ggL%Y3F6!;bWwa~aSFk!Zd9r2 zWqfu*UOh!{J70&jnYm8yl(Dg2sXHok1U13C;>sgjd7;I#Q4Rg9W^#1rofV@wSdU0(N`bEbhXSSCMO^EylB7yf2yJ$LjI+>il@hSHU zNuo%|?slgwLiN<}0ME#=-hZ&fR;Zu-)Grz}%kJ1~Thu<0+uYPgG+nsU^wO0u5t@%RU^eHY1PH`(Nz1dmMu%^3ER9ZQWj+z&ma?s39&p%i!{tB;6*?1}2n8o@< zQZSf=M*gtjC{Tlv6E2nGj|b*_RwxVoL_3#v;#=B@rW!^+rHFHpAJd76L|e?hAxo_8 zIqX_gIRL1*gda?!_=n$YGzHPFf|7eqIwez$j5_x;8YcBvpH!Qs#A9MK=TmjxP}05| zZu=*9&E0J5aP>{a{Asambu}39R2fI|r|MU5CkvW@T9`Ey?tbH&X-d`Vc zjK&A1H-EU4{DXN0Bt+>ZVyWF7wTY9#8*=o+DS9UT8%*7Edb$h4WH0Qac;hYVh%MAg z+E_Ry^>KPR1Ef(cY|2u^_IX5GPHXmg4G%z@k~!9#YSIEw5ps(p@@2EZq6L{~2kDG# zU@^`Pnnnj1a6KU;@Bl@`+F-yqnD9#0Iav@!!={oaqJMufOio(d7(jZYD6b`gke&$z z%9mh#fGrIFXmu} z@?c6u(Pz)Z+AWhK12$dcJaCaxx~`-vDk*|BGogo%)-u))pWVjHDXT$y>jQ+34yIp# zksR3hFW5%QE^B-PwSfRQ*u$RD85eoi$L8Ky=H~~0_sCF`&s{DD8Z5^m6G$U>b0{?9 z6-zRiL&)>rPd5Yu(+Z9=*0j5!#~+jc#nH$7vh}jrg_ihtT1`DG97UZBvcAQ-epP2C zz9T%P#ns`YH)C)g_boMRtL;3hT%h_}o1&L796{R&7t44-El+jZ%e98oB%u?% zd$Od9^ghs`!Y`!rO}Gm0X2hc&kaaJy{gcf;Xx`XIgX<=1ZOR4m7@u1W1IoW?C~E)9 zJh_d1%jWUpIa3kpVkmGXC>=)OwP%aZHdi=ET;=uMA$2(PZ&->v*hd>k<==(#Pfj)A(E&GD6 zn{QJtBU0Z)+B*@T-E2jRSTGWY?n7euWp$b~tu>l(-~>G3f)A|!OkhWDyor;tfgDj| zvJs&zPxw-=CUXqMbH2Pj#SdQ^8dIEi912a)PW5+Zd*Y39kQLRuc6bOF5(zl91 z(x+=bs{Ri7(pE>!$m)N>_omaKbJ+!%Nv2QcM;in{Pptp1qdP2(fPOxX#6b8iBD z=?cvSSSfAIw)9^}=u$s6ySD`25K^<%LpMK6*v{YjcO}^fd-}V50f`sx;GK!2l6v&^ zi-MD_4mjdbNBt-Ctn3{5UE#!| zla-~Fm7$aJJbphJwX#aUpdW1dirZXS7t{=-O_`PIJ&AdH{0kqQjZKimKKLz4X*Z(j zco4VJSF`&cyDZ?1vd1tWnDicxIe?lIn1zGnB&4L!7!b_0-bPuNny)?A%~!+S9pOpg z1v8{EUHQ8OwXOCDy4p2g->T(T=^Ysp**Ms zTA+DclZc3@A*g%v$oeKr_eEzm4utt?Z7lQ9oU72C2UD!*KsokWBKWR$s}bKLV<^}< zd^aRvwcZD(@oT6!H%#oqqF~7jNmMy?TfiW)pKc_4_@m2@csd#O$9?W z2h58~vOHYE1goX|6|IXEJEU=%YT?rx=Drxm*bw(6G-QyJjs_b6em=|E)$k!1qUD@2 zg_wR%g?70GA|=#TERhEb?S)eB@&H>n+nFpyCAj_Su_LEfu_H(3hK0#NkfgCtVv9-3 zVesK%SVs*pcz!V``hq_&@ERD&6NbN`e;lL z19bbWvc|OQEUh?&oieGLvp8DPsP>mE4~gNRna31G1y^!Oyx=+p?pDK8jmRV-;F>*15HPpLlZosUxk0O0MjZ;j_0fh?g5 zepNaf1c3yap+s0Q7dCqa8g~e44(FMp`|lj1LNQ7?72!B0W7>n`RmFnU7+l+quNK75 zKjrcE?9@qwq#?y1XOe;rIPiIwP7_~=NL@k-%5WlcJ{XcaE&b*J0eWwL^}?8?$)vgi zK7rdO+9>(=qvW?$e^EnD5rAzLBWymb$Z&R_$N2y(s#mvvi#+8isRPao7qk?1HY4dC zw~>l_Fma=I*g>431G{Y+Mpuz3ITAVnUlR|2a1Q?Pv}fE7D=)9d>t@a3P3|6HLdLPO z+7AVK#FP}s+wlh4Q-@Gp6t*=iX(+g8#VZV64G4y}MfjM~UUu@?6~yp?lPW-i;WhLF z)XJ(!8B;ze(*!aZb^E|6(WtV{q=O$^)HE1p@(KDhjp+&G(hBjjY)sSWaZy+E)xFUL zyYU3dfs}^~gK|zG<_}L;8~|B+p5q^>df@9@>HrJSEO7=zd*X@jZVHh0P7Az`(fE4d zal4TP1qCM{3>l6?LbXpW+V+P!asYT#&~zCW5f%eL2$O|l3nfDkt?_;%Ze!_cQ->FWCWIq}YvOw`X55FN!-*F2uR^!Hu{?RqXcx_OXI~aw|uI0_O5K4)iiKpS^ z=Ydl}k#*BvH0t%5INoGlZR*wfi~ZXGdI;eME_>yAL|NcGa;{N(mtj=l@`O+{BhBU2HWZO)(~F3QOz&29Reep9CpDh^?C9tchAY=9M>#xz z__bG4So6Uq7eaj>Bo>?W7gD|u)i%mn2Bb-pv6|CBw^;`wNSX}>@z;TIeR4%BLmk8+ zJ+pbqHW!$fu;E9k#LJ%gSkls6T$p`tA+ALGl~UT6c{PiCp=h3ySQuCdW)V%C?R;0& z%e|n1zq$9i0jHPLBw&b$(ZcG^tc`OFB6{<1TI?4R=YOPNn2%=$GMe|ZR^|ezktXfGnSicjymVq@JDy}E_x%l{DO+V`So#Y= zv#Cs+@?QM(A>9xQ(7`>yA^#`Ag0t@Oq&DSENWJSzE92C4Gmk%Ry*^fBx(OR z`d;Q%RskZhI%O`%C{t*P^q&N}xz&Tt}HFm`3w?qF^Xd z4$Lhs`sKd@lTsxnHvRwYS}4=@LG;qlV|y^IOYOiL^UNS@p>SEFMtRCaR) zGo?%o zl0j*!wFXuM+`@^H=~9JvI+85I>v9)(Y(*F&0?I+8>;wmyqJLWV0znf*Ya`5paf^0E!^1_gBf~c5S$1wy0S_~if7Zo%ESx5r0G|$03c`Lr2cHMCw z3xA_wYEnz@8eB9imxzhtj2C<5^s?T|_f2eAH3Y5TY<$xa)t{^WT961att}tU!8DCD zYK9je;dPbcyV@F((xh1VcToIU7F)wl@~z0M=I!NhLdyzISCt}Yd#qo@&nnC|GA?q1 zU%cR6WCzJ=J!#hwIA|R1upO#dtkQOyyb7@<%;4VIat^pYs-xn`ks;(hBfs%7_g2fOG#uw12KtxQO!GCu22*3bt*N@6L3? zg^}-GOZMcg(V9@s|LT1!#(NtKpYr}KJ~>w5FPZlP<P_?;6aE^yy%>|gGBWvUbgHuS2vRsYLHM6l-IW~$KrBZM`Gh^+L4 zm_}Q6j!#Nm&N+)X0_u^=BECb)$NKR)>$z|ff-Y5za*x$uGKc@c1OlWO%XHzblPEv) zrGXBa|6weCz+@u9X!C~99YUo4M)+*_Urg%1+y^>17CCqXt6oz7Z|Hl>#l|zJlb${8)D8WF2MFpVvFjR1e4dQO6>BluNNO z>6#K2Ic_MFzo3nJn|`U~n(Has%$X=#horjKOex=@!eP?*26^Q1jzotfRI`ncdVpI~ zFCB`+6r&Vl9R1{3Q>R&^$HR2k{ZZyj-On+TvUAWv=UU9Vf`B} z159q4&Og{C0{IBBv27`0V}#bv;Z+tmqwM{Kl&g`)z^FkWD+toqnXOZ38pfYuS!)$Q zHz~p5T%!(=cVgdj5GRN{k@A6UGGoZhs6Zm)oV?E#NgighI&*;B>L?XPIL^%4oMqze zNIs2&4%ys^%`Qi~DyU>NgR76o*t206Sdw_;Ramj`jj^z9a$&nt4wQ=|=>B|J z^QNiO84cA|WXyuC7L{wx)gobvpPtk-Gl3JCc-B*Md@;oj{Ka^kS&g>AuD)$}D-|l3 z+vu-&slG@xdk1)u@0Sg#w;}FTT~61+j?$EDY({{}N&`6C(}qQ>Vh=BkNCrd+AJjGG z=~&U)VFBf$+=sVlUArAp!o0yuNryXwd!>RNrA2N}MUYQ+m%Y4JS0Zh@mg5k`87wR; zj5i@glD7p=@1|v=@SF{faNRnyXWszCwf(c-pp=c7>~zG%&sfr(?Hibta_j;q6(O)k zr~A$QUB!7%PjU89ogkPFZ;mns%v0;F+xv%(!n)WQk6UIZM#7y7jw~@^=*#xBwRe42*I(xK^?d? z%Kxn1zUkkzY6>g6{%0YFf`Va@p?S48SJ$|3vvamvtwHdvq$DmxJ77#VNu>2R&h3p^fZW>vjU;ySu;GlJ-?xQe~B613i7V(<(Jf*iXX|af3_4hmKYWX`f9*=Cv*(sg`Ikj77 zv~(Q&XfBh7$*oOt_SYE&;62KjWb=yrba#eRi*;rzsSM>fhp=2@N_h*9CbK=GAY41lkV2V1C&CuJjfZLQrRZ%7O7;@dC}Vi zGy=XsOA#m*8bcNSjc&;OX3gJI`)3Fo20m6o)f4rbo2_tSA{rQR$pF)LYJ6fk@@2hs z$qAD+1ej=9kq;k-nL7hoom>9l;VE`| z;3F+$$RH8*gkPV?AP6w#WCR-__R-r~y4jBTw7^aW$?iThIIJ{?tQorS>=9|37V0qQKvKn(s14QKK|JjmYZ2Y|WGtKh!JG>N^zOT-E- zZ)U`OK$FLQ%RZdlAkG2io0RBzE?$zgp1(*mFs1=90|g@`V!#~B)tYXrFu^#>F_6@V z(Ol3D57b5=kq%##JoI;|akd97HY%Bi} zC2XCdnCIP33(pxB9$X2_T+7>sVW#3$tS<(P9ec{HQeL2JD?nMsMiW@3mrC5O2@$WXzI)l);J9)$kiLgfQ_nH9OO_s}HSitF0Jb^kYwH z$T3-g?W6`c%%b2>zfD|i^G)6@l$KEZWFXT#2_i7q-nM65>RS|wb~p3AgQ5ul*XRn* zYvdU}=30d@(eV07cbaAL6+UY;LeQHh9UE1Xf=qX!^cNyR{+ek`D^5#5#+0$pm&Rcm zL^H~R^e;hTko_A!WaR281pC)Uq#`wjk=I;Hu$TS!b%{1)1iZ(qHWFuqJ zb+tm-r4qWZqw;6j`sFJdG8W9w^5F}D6N#GxaSW0dt~HX+z9BLf6y*`%<6+tOJ>moK zLa3VLcv2$de}!rBP?96uxty+EIdJ^Au|%DuV;el}q!71bBxR(E+ubOPE^=F2p$`o9 zwaRU3J6mo!Df4cE`r2<}wG$C$x2B=WdjE&D85Bm!^BLhXn?N<+uit>D91e|MwEy#S z-g;+1QI%^?1Jhb`9uE(7l|g7JTMw zTDV8K4&%d)mHwu1K|YG%mAtUR*Oal3XCk8M;#Z&4)??C(xJdHRi}A5vG>w}QXllRV zfgGdFw~Ei7MQ7rTMiXN*U%K|y-wF12eNV?fmEpj{PcT8N5OX{%7TqHzl-`Jvfb}y4 zi(T)a5Tg%4UNWEQ#PZ`5#if>Z)s6auJdDybTSenLjA*%MzP2J>4ZDG}k5og_dh!`^ zWFD-?%G{8g5QKwb!+}G;X0C zWufixx>ft>OCGmswh0g91BW!;h^OCRP01mu;5hhy0kf2_Phh4;(jwjPzb*{=|3hrUoWFJdS6jgF&$eF4 z1Ik5d{(p4h$o23)k5H0s-Sv+5NWte4$M10ujM65?yvJc*`}0qerAn1d8$B;9SiRf6 zz1>R3q1qYqOg=DNU=bFUNGs(`E7Sn3$lHIdc0x{;b-J(Zf9xn~=GRNj^=Vt3#rY|$ zZ+f+9{|y{Td}l7)v@ns>NslzJ2)ylGg{`hWlsD_y=~qiYLTeSu2S_`4Kgggsq@2>6 zbozJIr%~I2fN#jyI})JEr|e>^vni)a6k4*6Q8rjN`T`26p)JT(&GRc?G9qcSpRvC% ze8+GQyCTGF%|7TsD5SJk!S-N>yR;uCLg0R=xz z>c#+F8TV5rhz7Za?}`U%hhDu#?3=tCd6S3{x*(YK8}}#max*m2Wy<$5Hk=6Sh0!l$ zldwWV#N5kULG~AVd+7;sZ2|1tZ6@E1&{T6nz;rWjc88!Rk@kVhr`yP5)GgL^# zPWS_oGlb@Z!ELQXg?#uet{= zLd(Q2+5u4eu#X>=N!qZr@JIFcQMvWwyp6kom{<1dFz7&@#?Y zL$UZ_abLvMoE!9)?&YX>K~V&qp$6(7#z&wm&Ng2bc58vCxJTtCH>v=5`Xs!yz{=H! z{<8Y~hd7j@B^oepgBfdQnRP|$F@cEFE{Itc62J4LtWsL;FyPoE!3gr9DxdU&?cjO7 zW4^-NB$cb;6)a$;8R>U~oBqSQ^tBH@IQEpsW3F~ky43gf+x`uu$N>Cy#ku8xFP|dD zyYs+LPWN=B4X%+rh1~$P4gf$6?+Oeh=`$Ar>$klyFgM5E?S=nJfHn69a0^~KGkuXk zG<><|Grv)vEw`hGGA`+{pZnsmwdG__tRm4`Y| zcxMvkOd5lE)|;#oY`*k0U+w-=T9Pth7An@xn!j!#;-N@zkD<~h#VC6qNJdda#NOwZ zZW`h~EgMpqhl=4c85R*Pg0WA)3xfIVXwK_FqB_Xl#=0#`rQ{#xREzo_P0);r7U^ME zY|@p$y(L7)l7btp{Bfl!&CbpaaG})hpN-L?*mGwM_xWY-SL%5`9DM6| zTl&zK5aaueXgeq)<7PN>sd&WM0M-38nd7j$-RO1Sdz{e15VjI(1{gykznvc=ecrhG|S3SktF0y%QxF1QtU$btuc_Rbq z=XBl(qD{}w?k75e_gyD-b?YHHwJ3w5!>WeRy*Si#8n{mpJf9yM^zn}tAWD-mq`~Wv z^QqQ5@45F;^W(6up2Sp()eHky#ZNR33VW{~zRpMl-5(z{wV)7@6W!qrO%~X5eN>(g zKfuO_=Uq|zrEe^6n2rhJ>C>eS`Z5}mtt5Dp&nq~A=Q9Q;_T0hNH49g8}57jua9okxrdK5 zj18d}R#+VYIqy5fuP4!WM=U1nj%X0}`ZwPz)VmSHBoGbuGkgcrMTk@Riv$h?Tm*DA z6sIr(#aTHaW++QI0B^5But71)s_3@8^|7yrJeJm_%kB6y0xMsgoLQyaY^Pn4Id0^NRv z#6M?vh6;b_bQZvC~kQ*D^mW#Za3WzlbkLrV-e#ePyA`n}T=X-R&;}*Tc32JEhamMi=n@y!6 zPHayBGx9r5;i@)n<=s2U=p(kta%2-~G|_0YvFfSIX|s82i$}Ha zN!b0nGH&-hj5d6s2Gp#61BurWVB)jJkNDk1Fx|x~RvQ87Z2AZPJtx#oZ##wIpv=dB zsTzf;99~e(h$#HXsekeqdsLbOHZ4Bga@Q%_sU=>70Ixl+^Q2d8TMc%-nQTyaB5*Rw zZ+s`c)xREvX2T241vX_jhRgeEvHK%%C!K7f6G%kse|3~$Ddq;$vkVDv{ngj#?i* zCJJ3<0vbIZoM$k%-56Naj;B5v#zNzM(2p=W7e0rJ*y`c?Sl!;)&RQrxQds8!P&m66BY@UgX6n)B-uddEBuaTKBjg(%(){;s} z+gG-24>bxqAv__PCyxJ6URIJrf4pQIC6x(nNrxY?KNLw!IQ5ClaNIVxwiq7z+qnHG zhi}-dh5=`4H^R1e$@cwc%+BYpgZ-c89oz@rDBNNM39A*`j>{<}_||*=wTHKi<%2xQ zsBe3*ojH|(yt3=${1Be}Mtt?pyLD+SObO*MY^!QJf>ubf_=!{Dzm|dqy~+_OEIV1A zV|&RHhe$}-0WQB>qcVKRJ;hpU-F8tsy6w^CP!r z*+4L)>DJl=RaJ1nWOYey?sLWT`q%}oOaO{ya@z=rIW|0{E?YR&YfPSp;+#1i+D^Ar z-_58!I%||X7Vrb@;tC5}lB?z3Of6s(P$ad9Z2 zuP!n_Fz28He_GD&ToW-lQ3>zH=+Bo^OrmeYAFVkEN}Ynzt@I<{hmVoo0|U>z<16g? z<3X6lp}{qds5SC~&C7`Gr}SiIVn#O-aI;CQ050pFEH_cw*RcwY+tEv6q`Y9y&z5+$ z41x5tnV@G5MN3IBPf->ed_hB5NKjVga`U~BGUrc+)kgv_eet{-72}knw%9T*_9Ub_V&Sf7#19NvA0 zJ{R=NOm)vJZ9P;5yW1~v|A;;`u;q&KzEjoC!d5^%G1TDDMr8o4%kx-3Yo1u9VTE?B zP&bU)hHe-O=Sr!v#frEelfDz-M8V`oe~K?LLhonE3hB{P))Kz*om?quRSFCTxzB)p z1oEx2=aUm*i_{tY&}{*eQMHo*O`M=4cOzE0=J$`Lu5S3<5q3u|ZK?-&dgB#8$oSWG z;s-qjk)eIX9k<<*mau1Qob+eq`MZRrTtrw8_OQ`&UuQ4#6Ct_tf#iPp?+^54h~|3* zE}?Ahe-8T;5~T50rT;Rs_VUQcbeXFw0dg1JL5I4N!*UOAcyFk*=58xBo>wQF=6Jzg zT>=Qis+!YE)KoHd2@OR|-+lSWRG>NRgic4S940fznM|w+@wS=WHmWM!U!vVE-(PM% z>ns2}KBj=yJc6ilFKbU9j~u@PhIUIhqJnD^-7S>%)g@5-*&A0iYaE`S-A`=dulP4I zJ`gdz8&F)xL_Mqych4q^9v2^Mg=>QMCh2e=MFY&6r~@d#x+d-C7N@0^RAIIj*)9w; z6yeXG2Ahc}8tRIsM9!Kji?Ao@D$PsiD&lxjhW!yD^#*qWU4bu6Rt9s4l5=}`jkN$f zfe!^y*)~KLiTi*oH(Y9y!INskr*{DFYkh{ri|mWl`(bf-&y}HA4iCmpoH8`jBTKFb6B>S@gVb`#A-Wl@ z0>+d3(5pE(hOO3;1L8Ih>||@uB(7{a{D*JphnP8#nZEm8^>wSg*+?k{){5QQFrx_b zBG(-E%)RvF&oz&fa*^&zsauoOX{TGYdeG9x3oMkL3ymqF4Dl6)mpe49C7QLooq$-? z@UF1RU$BZ*XMjZ_y;y2FR1Upf_%AUi(UbI=H$kRrIHcT&dZc=Ccnyf*o!_`seLb=u z z8`)G}Z#q zV!U6Z;3@m9UGnN00w^F{|0Jp97E_}O6(izp)Pt=)fb3T%3k(X`*M`F72ji7M$r8P) zI?;%U7?_d4w10SJ|9h=U`(erfX*yQnB6_wdDFHBR%VL`9<0arvsAbeXs+0(#o3hlOm&TRXubTZ;`M{~r;Tzn65Y3Xr?)Mr#}_8vN%$RcfINbv);`36w`rsv7ZKLZ;Skceew{>OeIoy3y#>H zX_?guw=iAuTMiJ4C;iks-_>ri%{y!g5ReB^r_V6jpMWv0xVY0x*Ss0B7l#OZoa?^q zdl`3V$p;UWnwf)&5!XyY#Kr0i3NCNfe-HwV>b}^(PfA`}ik0DDsOi@w*={j=t2@Ks z#Mqa4FHoS9es%Nt@Ej<>CVW41=x8yA`~7|FhivuNvQDR27Wz~pXP{G$;a1%@r10Z0 zV#eCMF?%*W*T1y3U;t4UGyI&v1MLO29Q?XXh&-{}4tMyQy?#_!8{P*O>N=W8Jd^6G zl^{enb;5~OOA7i<=kF9T`0x@QOhfC}Xi?Mfy`L-T~&*Ox2|^+s1*S?qfI7Y+*2)NDR^qByd;rjT#e26`zK z8j+K$uKfv&FiYC}i3HTL=4I7Fe|u4H0lYZj2AH4>Xj1+GKb!H5-gX&P12-WQLzWPd zEL<$_b6ADo+Lr42W@V$hz9;n#t{it&gWYF*b9M?NL2<~K8Pef|L0ptIJM^J>NZ~M` zoETGbPa^+~sJ#jPRAw6-ail^HWKSSA*NZ27*zV2a(7gM7(UW?+Ef0w9 z&Hkpb9%Y)z@ud}RR+YZ=Fa2606dYFdP>GeA!FtK6KF5IJOxy+)s4EHgi`rJtECqa) z<5#FV_<`-dg0L|_NbgD}Fz>4s2NqG~2P0Xq$E2_Pg5IxcWKLYjBsrE}#6Ut=#4fwd z+6(Qjb|1)HBM}bIjH$qQBN!6yX6yjKWQ8?R=?Q$`P<{Sa7&v~k|!&1 zcWZD%(m;!EedN#2nhlP@y6h>fs|YYf-VXYHj}Yg~>_GpvG8KN$Ju2P?8M)8A=?!Nt zT%`t8Wf10eG+mw$0V2qlc%ozLDuP%ZStZxtw#^)Q4Ta#(xm0phx;KfZ6o0M!v&KOB zW31u+YtiBBzZc$=0&;0Kn1P*CZ|>HurTqEsv%v3%`k-wV8`^ zQGRFm9V*lbcvKSyXjoBCVS`!xQeEGI$e&Z`%0oK=hrBuN6P?1(eF~JH1p5QyrEg^` zC}+c>orFDz;WxraGyGv}Q_}+L+|ZaC+!8s?l{TOeD2#$*8QhtNiJ{F03lmYORz}6I zK)>9`P0MJN%Cq8X=S2%#RPdyD1C=0RfSX15y;hE2gsxaGO75m@xOJr@Zl?VA$Oc7M ztVuxu5Dq)}eXJF@r+&Tn$?4c}-J=+O80lpvCGPPoQO^Op1W#?6hRL(t3xEB3An;`c;s|}b6x~{wz=WSz z*Z6X6jw`q4)a^_Bt(c?UdwxFEOZh?}q}#EKM_DxdOzeG9vFX;`un(DWImIv8h2+GQ zOy;N#l9_o0kuR|t%B_$rxmDp`<`~`GEUQhqRh~Tj9kZ;{2o(sROez5%{?Jrg(V2c@ zinAJC6{ka`RGL%#xVpZVyKa-d=1=JmYIn_F+nt^tbz?eY<@7vBG<9dX(Cp^bqKxuw z#we8y?nxshnr{E(t20^k+%3yle8rWe{;^oRqCFDmZR}^9wOl?$?Yym?-?4`;czp_g z-L=XcP4H*D1k*ji{Boip6aO5+r-)BV{qSU@)b)RfATl}r(eoSd!IokF zi-rGuq|Sn0PeQy=hS)xyWNQ_2|0H`ldtP8Oh0+B0)Wy5mdg1f*8Lyk>ah*WW$kJLK z;MI`P#ZG`0H}LwOTma+nrrFl}o~P$i&phL>sJOr+#QKzm7dW zED*x!`M7x~-Hbl%XiBwjRc=DZ_^ue8J)hQfJr`A-C-}ZO>WsNf&Yz!D{A{N{NOHxu zl5k6CpT0@)ZfklP!^br!iyck5NUR#Y#D%8&a6UWMOy&qPw@Zmxn#D&fv8t}-JJ$Lq zlz8|z+H)o~w(pQ#i!d#JQDxU4J$T{28QizMpa=q=xW$uLW(PCHDF(u|KGXCEh-2CB ze5;?IOhLwVKywBSyzCR;ulvL|#WmK@x79n%uBVAsvt&Ht@a7@u!aB4U=b15lh$0N2)uvng#6lSfXn`CQg!&pui$kSMzT1W1^tQ}hH zR*9V#w5v2}82>$j*5(Dz{vQ4ShYh#lHiBXt=6K_4}P65Yb|T=qUJNg%apdV4j+jAhaMEW$#=I@JJCG*> z+fsUE@)=$3wT&r56MLRr1lxte9B5o^G-Haa9_B+`a7=ZX5dOmYoB!quP)G)b1@0q4 z?E(VF8E}Ury%3hWEQwXE=_Kq8zne~^`u{PfN&gL3n*=5haU3-Q;0Bu)zmqT;CFr={ zP~y}ubC*Rmo$3nM?hu(-Qy-*?%we{dfh(TM3d}TT`Tog_dl6O&iA3TWFlnT+pP=4G z7~vj&uQ3|;L{Dlo7*gS2CPG+UErJU$a=JsQ(5?Z9MTr6cco7xvBJuI&LqkP0Kj!{G zmGr|J`C%|Qgaf=rl|>uW49|Y=FaKN%Ix1w9?Qo{3TRx1b>U$G9=pawbKYzV8`sm1Nc#iyJWI@Fnb2!Hjai>Q!|6vH!T`Hln@%-RTYqpi$g+& z5R=w3B=lu*DqHm?g}0UpB*eb%SQZ+a?)DVN))bVc@sAh8@!4~RJWLl(pUyKn>Y4m^ zf|tOQQlD`JYpE0?KR%NX8TW~St0$=q>EltuaB%GW6?j9zg*&k~zX^4}CWO}4~hi+9DDF) z6=?p0j%lYGVfKDY_WlBDUrtgRBf@}cU{BuF55(p3x)WZF|MfVomm?$+nugQ0`{2b^ zo9p_^Uj^`otiX+16K-=DO`$7T+|ul3sUsvb=NT>gZHZ-s!RGe)ov$DdLwZRFwe70{ z?-jGA2%yOJEz^_tgNhLUEC%n6>r!=saO8pjz_rCdI=GeTy?P+@Z z>FX+rtwp~~EML%ICO{dzqJQ(_>iZgT-tzmXjQiu}xW18dn~L2lwjF~&&<`sz#I{bT~U`CQ^-!fE&WZM}MX zHJIX6?WYc-<|?tM12f?0ND``*x*wWBAlNPCOd>W4K8W$vMI|d*-I-yvoBQhQv~J1C z@Sb4f%^VIE$fE0hPTkSE^F^vwZT^IYVQr&d1(GA&UEA?HDv6&;6l<&VJV6rSSbsh_ z&lefU{66feHPoJ52888KAMC|*b3vdrD@mVlXRkUC+zj8Gl&I7 z!EYnH6PLd)U@(JI@vq=QZpBk75lJ* zz>w~Bg@n>M+3Ue{VC~VJnt>CKlsZC~G_JpwFB@iVlbe$}0%AuaAtEXzM>Wse8b+c& zDH$=G(;>yS`4uteHEHdIIULA^!Z@}KN8vUjMvc%T?f{O}fpb2O{NjdlRWad%TC?C~ zw~boE-BS_Gb?H<@u@WKBkSw1d<%Z{p7y9jKp#AR!=q% zKK=gZZGFX1xsA!lYAcw(flAHZ#E^;6%eEUKU+f$FuC}9*Ud5_XbU4TRx9!jkp|?En ziVtHXTi~tqTJjoFk=PSw2OE@1C2OmG(2)Obbx%&|8*rzVax3p^Br6566>F80Mc;Pd zS1KJlF_h8LVbp-x`mWeahx*I4{%L#QYwrm-4C#>uX+c~VQ4BRG-JRnz;>U+!%yggF{PX^KxF(ODsxA=2olAU_B%r2f(Szk#$@w z>lwcp$#RNRw5RJgFxC;O36hQJyBouk5=kQ`*n)bu$S|#_<`U*-c*;3oy}pDu6GVU2 z7iCY|QVjAZp2NWWEj@F!>=yWiGmeyo3>E;SDs-RI) zwpDe?N!~>UALs>(0w!`wcnP>BgP%fa#aZ((DA5jyUP!&U6P)U$U@kuckI{&pwHln# zfFGv359g|T+Rrwk;TrwVjZ`*{7C2_q)=lC@Bc(z|oOK8vkKJZ9^+${$KqpLLY!D|= z0FDxAKWii(%-|EqwV3qKk*$B5Y3G;Prs%-d$_VvZ>~Bw8WL#;3EcW^=_CCV|WWZX$ zS3+RafP)iqhc6P_y)9ruoKx$X(bo#iRL>jx6n;g!s9eZ9FzbLaG@jR$KFZ^lxw*9< zHe1djzD&H@S8d-nW#z^tOZIMz;Ec;#Ck3My0-ma8E;;bhlmxdkOrV}nBGuB(5kC@B&Heq zL*!Vl65C7i{tvF+GAgbnSQmyskl@Y)he3k7yAvD|+}+*XCCFgGEx5b8ySuvvcOCf1 zd(K_!e)s>Z-o5v9S9RATHMtdT)}XR=IU%d|y6}Z{-I2j0`nX@=gt?WIiHp0bM5zf1 zF96WM!}#eKaNLv7o_wwQ#50ACl4^l$FwD}~AG!@`7CF-0Yv_411Ca0M{#?xP?}9Qa z-t}PmV!W8U_m4A^g#o$DOOAI)9|HW5U;LMt|8a4%`VT3*>-+O+7y@xHNio>zmX(=k^PHtgQqNu4H zSTgdRwSGC6#2jzDNP@ziLYDOA+KE%4&Xq3H>%KGWs{`Bl^MD&VJ1o2s7IjG}RW8i@ z{O_VNn;mWLpnSysNwb}+E^k#b+x}5bHOl4$sg786I`{Py+Ion<__nNDch8EeWN?KdySaI~;aP3BS0MYZZ4-iK}5xk^0qc zCq6~CnILGW2vv(UT<97jcexk;|rvOdDd%K zj(RLpZ1UZheH^}vuL`C>-!Hc!v!AgFU1qS(%?(L|G#4q{?Jmq5K(PS?VBP47d$-13 zhgaT8NLnAlG@oZGBMk-_jWxfqzaG6b?$-EYmn~PRr*U518@?*UFlm&ty4q(KR4c|` zY~M^;)ETS~pQoB4S9Pr$47Tj^pt9I;59B3teqW;+j>#QAQ4=Z+VpOSNQ% z+&z@ixQ&9s|LfS>oUqOwhtvbPr@qk(cC&L=MN5DKCqGp6&jwbckfG^zG|ze8*=8%? zV;38$LQ*m&&}m=FU#EZ8BxAgvX>K}OShbsVc4RDbC4KoOkDO&_FfTXWECUnVrzz6Q zcwUo|%%TtM$w02F=cKAA{qg>F?!Q|8PvU>I{Lg<&mC>`T?ICooGQu!bX3i0HH^9!M#_E>Orwj(<$RH{UCG9&d$|1q65{B&hU+ zLs8NGg;!pMeBC+hl#uQ-<7KXaMc&6DAtyspu6)BTfDbe|Ys=1}$mGxh=r3TMV@J^b189sWm^D`}IeF^eo^}X@ z*4{KtmwZOolLit93jXb1=jemI>!BVqRm;KgCYGz)$-ne=m1g%L1T%1hQ>uB%HnoJv zxpPkL`~wHB===)K5oV-fBMf|zFkDal%xH`(XyNsVo**`Qr43pX_xshwx-2d`CG2lc z4n5a;G%0;SDKRb%?)$)4-tuM@87!AwGasm52=_UR^KcUCgdNE# zr$fxE5ZkKqQx}cX?>**en)!6IqrPJ3O%3^KBK$%=6$-7u-TzrHDVdALbb?_~lrlRM zG@vIWgI*oxtiP;jOS0kg4Y9#QZTBTxx6+OwML_5}oRTmlI!ILCE{mV4s7Tbxh)0FC zSafol1$E12pkW^2YS8q+aLM&xJ83aGBsU)!c4?5^5a={6p}0C1*2S8gW5)2kZxOeN6cx^)B3@_$y7rm(|OkAk|5FU*CDjC?GJHyN5S4<;EI>8(n zCnrRAovqD^PS6MaYU{Eiu;>U*NC85Yo%U+2iW4&FVbLCAm#Av%P7K#+lLKvHcsg7% zb4k~^;Dc+Jt_h8Jb1?2XrE0uLwEu2;!2_{JHBq)1lu^Z#<~UKB9vl=CBPtFJU&Zv{ zR2QG_9qo#I*n0Dn{$A_tde5~LBzqYv1P4SN5r#J|S6NASOizD*xI2|>EyXH>KnSKb zlMVg*w`8$>zOvKf&O7y|ICsr9hoY0_^(c;q)63aH|96x8ed73x;P3xxE-7TF-jETc zj3iTsnsIL=e=Wx3;RhC+HK(xVa(`pbd&8#1D)SRL9?I-jc|$FLtI9YS{wc*9N`N_N$hR*I_NJ68@bOENWIqSFmoS9tNh_QNVGT-Nw+DekEn`&paa=bX$WNVZcd9#iqdD$>^u)6wJ$Sv+W2(VcJFXi(o;aS^H+8^% zPO9k1;Gh+z-51F%?Y|23gI!1Zz0_`qa(}uUC?dg$#zaqqfHUD=YfPl0)(+|IueRoO zLXaG1EN+{HuuRUPsF)t@RtjL7^Eylz7nLI|-@BOFv$t2^!TPjIOniZ2KmF|@0fAh~ zcYSIq8<3c1O}0vPLWXm;n{J)tT|>yn?&9mj^Gu^Aw&?8BnYG4ot@@3DU5aej?v@<* zve!ggL*Ls!ZTRnp4jIOlx~s#(!^vEZ1u6~C691}gj?nmX?i#d-`U2giK5ww@A-C{4 zm53Y?N^_MCb(?4@erj<2%0Z7v3xIJ(z|v+69b7l>cC0@*`j%@(GknsRj}k*1d?Y?f zc)_tEG*UfOFDi%XZk}G$Yu8+uKIsLsFey#c&+Su}y!1`tS?bvCTU}TpnUQxB5nq0! zy=#@0&wCgf^H_MWiSmzM0bv5V>`Gp*;=dBw>a;wH-3tVaG2|0mEPV=(!9reqg9;me zX0Ii#zijMWtgvcT>|u?zg5t^PfW~wFmo3cUQ7=>vjB#cIQx3HJ@yk>ivrh&DHXyGU zvl`stQP7v%VZ^VaWJ}UIi5UJu=aW)oqe~`PZ5J!#b)FO!eiF5s!nJ7I4iOy{%-U$@OZr%IpGbZuO zO8tokvKttx{p~US4VL)(YL^xdvCQeA7uJB_tzPQO=DS+$tvnTsyDJBV&$ieF+c@3_ zxgA4fhb6vPpDvIfk57#HX8I0Qp8F8UhhL0$2clM96cMFol0=bYW^|Sl#TIgb>J7$x z9<<-XjCooJ!BGZS>nhig2>Zi|XCI-J5>8a|aA_ot(M}bS$7u#D?)Q*ANptV!b*|i0 zfI;Mn<1-S;N-hTc>=8lX=1zn-YpX`P5m?li6#D9xvSSu4vEKp!pbD~*B9ohD0s0Lc z0o|W^$f~ih3BU@dh(T>BN8<8MooLU1VZ3g88J&h+t3Ufn6Y-)}VvCmS2>Q7~z%Lk; zBpzqe;ra1l7K>lq<$(;78Is)L>&4B3gc0a>xadyijtMeZ9tY<4Rnfw?hDW~?JVSD< z4Oa&jREB~wHy9r~jC2S=5bg%L&^J^j==dSd246J~tJ%2LS*WeiNGD4(nv#TLa*v%R zy>6SH|9PhJb3$X9zim~is{WHZ4iKZ(zeX>E{cYdc|2k9bM`x+hwifoV_~?)X7n%^- zZWfb^-=|A$%>K5gc-cI;qwsli!^+KG?lZ|uTx#OOV`|0!VP<|2K6>OOM?wOM} z@^9e(i!k3*^iPzS(*Z8I`p6;wxAYN1=!26To}G97|NH$1!F~>oO7=#~`{n2E2be zlcHWo?M-Gn(i!U!PKokuH*ez3>5k9+fTv7~Cpq|vo|E$8%p&`z@F)#=wAD$IvtDyR z*+L!KzW6E$Y|i5BJ$yyLWOADWEi;LH#n)~v#<>a52ErM)1`Tj_)xwz#gsV{@$=~g; z*Rd@^S;19fbKwNaITZKZ*E)MWLe{Ya+Dxxc>$(q+-f`XO44pp3b0TJSU+)8Wo#V{w zMBk7P!_UL{d9IJdPIB~VIfq#2IikMBT@Yl3N~+vd{sZ^!P;cZB$NG6L8U97_ZhOY zQ}*gPdW)<}> z-|Ci{N$1Oyt%Ii~LP)v~?)cTv__+Qae@HFd^VIBgNCyeSP8@Sg+Hc;|5vw+(fn^B> zr05R9I;Y#NAw2+>I~Sq@38aS->ltuY(LRc-tc8B

Y)#Xfu9v*wR@0IRi|uW(7~13t^IhlEK^QIW_DZPBkQroUaUs#k5{9@D3hl9I$Kx?QvjX1;nm*Zyi%m-!=04iKX_^w>93=Emkd_=3 z!uA%9w*Z+#YbSW{9UWVlCiD{bg3@A=(}l;J`>GJ=L#Crq#jjktwW zU-Y761!3$#90=6}%D3OrfP+?+^0piyr|4*LK?xP>2D{3*UKCeBcKRzp>&1Gd>_FQGx#(?}@oNkJ5Sl^`VS3>3M zGXJXnOGO<2dD=Filwg18t_dNleV=&)mmW_I-F78uybobwxxKmgy(i17kQa7_CEg;z z>vBLfnWmH3_aFclg}9w37DKr^z(4ZoanaTrtJ(*Lnu8Zqi2HF&deHZ_Mxs)4g6#m$ z%agL7Vg>HmkskjM(DGCz(9s0*>a2U2pVzmy5*kc6_*`Be7_9W#4pVm}NZl z*UD+(VqjPP6{#HYJTCKhD2VqG0K7f#y&!b|v<|d4dFE)}-_7dX`Q|?6iOIkZW?fL_ zz6!qZ!BPFaJf&WX(CO<%$IH+7>PV`>+;_`%!>r(Hhv9Rx6me^ACF*sM|G1VU@CzT$ zTRp1{?JLQ-WqD8T`;+t>bDyMA+2_w#Nn11SpO@+n$ni~ry3Z#FfkrqvWo`HzK1aQb z&Zj*$la(U1vtf0oUTbbI;HENCz#E3om7Lhmd$=bTo|bBB@Xb<9YvN=Y6d#y>O#4d8 z>+Y>IyO5?m)|G0{b@x&e?{d*}FP&avIdg&Kp{|aNZY%;tRDn zNxgHX@0Sqm!3$m@QaF_APwKdrABx7p38$3CI%#&T!o$ma!^+&~M`ig2|4Jp*JEndU zV|?CqJ|OCK{|D=cvjxhw?Po-DMAFW9p^v{clK!4i9?Q_ER$0rj2WI{JwB|?8=i}b%hpk}YW1HHkfv=b zwn(UM??u)oG!I?h??D>r z=&tzIR%iSM8g$U}QCB=+SVQl*Bamion+USNYo=9J9=tjp&rFEWecSItX{5j~pKuIj zdoUeQFZf|ILi(ELYNz&O>d=KL{SJe3E-M^Gn*-%e^Zggj#!>_-UC~9l=ryX~Nwu|` zk1}6H=Po-CgOeoW{nT#8n&7oi4f8H&@V~BA+~seiHoH z*G?ql|3;jt+m2x)FTmdM2$Ttyw(HFgUb@;LTXM*b72*+xEM^;&j%Z)U=*qEF-MLmL z=WKDOi$+r)F@{njNV_*Ux&Wq zen)~>A!ZDPKvDmok>nfv)`yE={NpbeDROKfs9MUwWLlgYk_g3!09$&C1&6!NK7XF znpP6~gBNhb72>0D8BJ6f%FuwC^*AT6J?mNqO~!61^i5PEHBmM4n=&OC`&5*F(3i3{ zd-^u;`bVWu0yZCSWo9E`scopo0XcIWQsb~voUcFNgfb&i>#pM{cilN6g@DZ zf~X!fG%P1#@BVpYG~(Nd%B{bV*vdI~I{+|;j0lqg)9 zn(OaO#lWmM00%lk;daR9csLW_hfQYg!a0TnJT4;zAAMpj3KFMX-B-CJ0N|Sx?onB9 z0iCA*+GjQXTuhH>;YV-!AFAg)_aoS}t=lR;fP-1+ROB8HY~{=n)VzEIW*=c&S`#Ov zEy4VOGoEO+*k|3egKY?zBwtQp@QqY_S+s*xsN&3f;ULjq`%(qNr6<1#0 zNbw6c>MC(G*~`17U$mQUcC&=M%urVwt;N@~aVX;4Q;Lkk$NRB3w2*rRQA$t<2T=2I z*h^vGAVkzYT15F8@fn~yR$Ujc{y>rGLoWH5^)4y)>4@0f%O{sWb>od>1fIQjKZg<5 zq)8hY`QhZlGueXMyxS@}US+cJy4qv-Vvs2Xgf(ARCUTd~+eOSW>wx!DIqQefyB{e@098omY}BiPL{TR5u2p23)z zMm>i1OBNler6FB0Ng)!gOD&5sj1J|&q+G+iQ7XcTki@zu?8H%vW2cV4(Y8n8*~#Q>r)g#jIgyX82DoD$tameBQpB&XFyB zjOBYVt0wBfT9gx6|6tA@??|Xe5NhhIMkeaeFU-|yB2&bYI1zN9$@IWU^W5afqKsNll^IMV)CMG=vXuKf@P*V=KIZ~%bMM&wi-Wq;KDR6rc7oefTf@|(ma{To0@tALGO|a ztqyBzSNG-XdFzeCh~z@>~Wcy*v1neY7DSykM=$&Pb z9gN0GzkXEUTXV8vVuYKHG{pnb9f}hFR$-qWs$8RPFZ~}qLc+I$#p=H$N>-bf-1z!5 zP8)Lg)=s;BElsjP3qAz!Ib(nD(u7~bV&TJT59oS+s?zqSe}bY0&G*R0(XKdFcUUgj zJD6!HF&m82|Cs)@$w4&}th3SC6%Hz>HU>2kTyiP6ItggB2&*-iv0orbaY;EU&lA3G z>Jkn%x-k;aLZ+jzxvkjM6qt?Ku^Ds%f^QD9iQ_TY0N87Xcs*SZ zlp4s*TShOzyn|P}Ck-H9Lfw~JZ7K<#`*XelQ|daO@4 zXrBzUF=UTuKyQ#dczjK}77Xl_t6|FsI0h@PmUhS3uq}Td;(1l|uj;e>$G3X@`FN_$ zVaXxa`Ia7^Pnr(?TEU){Gtt2wP?J|51UJCxJXDh`|0~ zqMkdLiblBPZ_T4*4zTK}^m#yxbdUIRQQWtJ1f|yD1^+=7T09SNTOGp5!l}@5 zSz3&UdDm$^P4%oh;^2dggmDHT$7T~m{7$L;et48hn&k;pz zw>jQaFgG~ANyRCNU|`gqrbVIH$ETTYmlS`En(XQIo=k3;dlDFY)2GUEC7{npe~!k0 zthc;gXbydzQcw+EiZf4EiOF+E*~WoiDY@TW?&OVlf>*ZRFH`csl@ioY`RYQj9?7HBb?C$VxiKZJw3eRq^e3xf%H>H0pt6< z3GNY!u@W?NqaYoT3JmQaOK2_dlJ61K82quMd`*v$l#sp)uK{f})sS4*R!rdesnzX@ zLI@|ic&7lP7yl>_@X-2V|Efl=!4sk1+}vbb~c*n}Xu^y^ac9!%CiQ_Qq2qfu)3r zL0*^z&5G*kp_z~z%-<{b;JUx$^=rmv=>F;wq**D>>NpMIsrJ$#k@!Rv4EQa&M?!yn zovC)(vk3HXdg~4{`Y3A;EN}G~BH*dp=W{F18C4mU45DO~nE*vE*fDuGv2NXgrjJ z(a(Zou>Wf`HWd@JH}}J${i9EATtH1Ri&?+Cu$3^W((Fu=0K7SXD_O?6eV&RxTS8UT z5_>aCC^Y{3OGd(NP;9EYjQ70^aM0~3y^)mM^^1U0g@ld;ZPp+VvkI&Z7?%7#8WKAb zJC-io;O@l{lv`QQfcXhSM9?cbqC)3+=L135-26Zg+KKZ27eO#@q^)6i#?uDbNZdre zVwK5Ry?}ijHivZE+N^+4z+NR1b)(JjB(z1S)HhlOnT6az@rqh2 zzLQWJ9UROI6W(#@QrFC@Mn<03FrG*RZ4_?Z0E{{;9m6WN2-q1Id+F$_c0Cgjnkn*j zTDQyq8p0F7U-IEPFxUwRn<$?R`OIIao$Q^CS|$H&A1Ubu?Qv;0tidM>E_ZINS_xnwv#g9AuabVrI$m z`%&J;0UgvkyqOa2dn1g=kN~_$io)ZtxR7{SMc*j{M^#7t0~W|FoZ=clz{MIu@;h_2l!sMQ>6{nP? zW}9a*t&+at)7;^j{zA;Pzob?}7l*e@Y3;zkBfDsor)EqSrd@S(U}@0=Q$ufHe7bg}kNAFp zPCbXW-*p?L$NWK^HcGEfrv|XxUg>0lHH#J5si=*pF%OsGa)3YBR=Zymm8=t(HiR9 z8+ajqt^NZoE_F~9f}Yy5akx}t4<4fbPz!yGnoRtuWHysnfjaGfr3UGa|DX#lSLq+Y z{}%-EUzZEE58#D=@jR^LpK$O0xIg|w)wx3dDH+!26tVsvt0B(Ce5A_jWv_u5mCprJ;F>)c@Sl(%dL8vAL2%cyaMw2E-&VgBIdl`st-XeKMX2}4nt(or*-esSYWp;QQfx>z9uNbeCk`(9Hp{Q{^DUsq( z@Xc~fPfg!&zmU;FDh{CDGSXaJ!|4BFA{8Io?lljiaByAnV@(q(wRL(2fI|X~@W1q6 zn+jVY3LsOs)x_U1Rq-XF-KA_5|L>o}dGRqH5j$k0uH#t}IzDy06vSFbSH#;cEJk8- zIxHqi?gvIM8(3~0Pt++uYuIHxq~k7_2cDq%NPpkUi55bn}>m zz}F`jHp2xRUp0AXeEG6_H+7ThuZQbVG1Q`PzhJ}j+=DGM%-HRfbMyHW?85806iz-n zqmYgcmGMIfXW*n&vuaqs*R?`Uq~s@vcoZYP$^TG%6|FwTvJkP+anHG(kkaQ>iLydn z%a7*Q7;$0Ye8%05sz}|JufSKzK4CA3tM@MI&HrGyH`OL~W2bFKN?V?qcf69L>54TDATC8=up zn>d9yv{7{Zf!F+b^~5ULwaK@uKGK&pR)LFBn9Ow;Clr}B@pZ6VnI!@-%W8;JiukLw zJdes^%l2(Zo!qU$E@}z)8d_=NWW;=H+SLugdeOyIT()>6#y|nz`qB!TU&;zv=On1b zg)-W;sc0DadJPrc0r$gK?&r7&=gdJ5L^-a@hc@uJ4Xz4Ik~)|6yoP%XJuR>|unnWr z7{AJ|(O}M3eX}^u{0Y;;^_I&%+Z>MV^mM(id-v+k@BKQZQKjWm(N;dc?2WJ=xhSBE zQ_-}DMUcwRkC6B;0U)6IGfM1v*)SNy@30{-xBi4Sld34^jfDKFswD^eTF1>3+M=*N z^_hrAr=0|OKfcgFNgw8U_3M-bjkEO!d2iV`nY;N@^*Qp7MFmPXLZGoZKZ;FUNsjOq zSa1uKZcGz-+CTf0QnAj`qW;}Q5*l-T@HWz1K(@p z0VlI@VGI{A>Z~tB;%EpD)g9Xv4g1&ah;oK#J)VWgIXPM}jz#m)F?^OT^Wn9h3uV|) z@(b#)c^&+iDt(QCbUwke8`P-_1N8`t{o;>6agQjtovuHGydF!zeWFk;uuy8%R4DP0 zF0W?F0&#J)Vfn}H99BI7TJfM_3qQ0t3ikDs~P7CvhsKKoyY_$ zH?ZV)89%y(puIj$#HA+#Zj`th)+|^Ksh^4@mpW)>g~m_gy=L2&6$!j)rA1yAMZ0kj<=Rt-CD-N6eu;{JR*Nl zJbqU(LVgYXo&pP4uLU=)Ov{0`8q76jG9j`thp3p@P{^5;bXFE8A)E!Ghn=|L4tx2o2aU}#Ik6^*l^y3p!8R(O_O z;iaEELvjM5Vz9&RU7-5VGszLDfuxGDerGQ3HXXhoe%Bam3n-O|@xG)aQ(_}mea%u7 zp8N(G`rAjzmHes|M~BTZzY1*a`p_CO-j+9~-GwEl~o~E2omX8vVFa6pGhiF&QbbKad9*LOH^(t`F^VbnQsT zZ+8RxvvgJ?4H320Vt^*iOZLM_OgdU9X9U_)3ie|u_%h^ft|NSvhwr2Rg6IsAWoCWL zXrD0Qm{c_}KR?VL2$o?xKIe-+#>M|rDx>Y!M)rFF9yxB%a28} z)d_JDN#b^%(xO!5e*)&`Uw-!n*?CdmmI=vl4HtS^624QYbvw_tOSg91zh{C!Em3a{ z6LK1_aFOv(L`j@St6{$4+@3bdVmqTvg+D|Ie7Q-P9l@6)#KQEm$-afl^%cMl3k!% z5Eoeg=bTSp}*ol-w{)w z7S?)#eGs!@j3s;=zA`-;tG;b3tj>fCX6MZ3I$B+&xG+{^*n8O9p{f)+cQM+9_L*S9 zQY786hV0^`ML8w|b9d67)4g77gg816*@7I7WCg)t38eA`Pb>aYTRSoyHnFs+ov#jh zl;tz);tVUix4^g&`nf|99{4veIfHN+BQHGEIg#YDl0)G|HDro8a>XpE=FRDQp2=e-y zalClT`|oT1pFNFgW)6zcl=WaryhFyZD(HCpB_{Z#{cUP!1fDQKH^%IuRKJiTgcDST z@EehH1CBNmN^Szr-RRcC#DX^wYv~wgC74?dC^aJ8OBa94*Lx~)KJD|vL6~Ta-OI)2 zOAWdKV0p$NoZz={**q+@(6zQ7UT$)z=XubW3NqTPloxW)gC@^EA>xqs6m*s|OHnm48P(OLBI_%JF_bFxqq>+gtS;9>XJ)sDl%4se? zM27^K>kR02eULz&%%&>{#AK!6(>V{dJGw2HZCVm zz#w6y>!HDWu7h6aUKIl$X2N{M`5ZT^xMBy!$AB?kz}_5%U!z=DNgoqki#N(8jLvZU z!2wBzEhjC!j#i0XU&YD+AR-4mR14D36d9(d9JCvK=N>saCl$TVGGaq+^=H}Eh+AW% zTf#P%iNsPJW7enlraft5Sf+l<=vC`(>6?OdisXmkm;~i>M4^tQcO?r7Ypezfz znINeyENv6D$SrSyXj%8r8XorJvJX+LNt(_56a61i$^QlqlHf&h6&J^zyB)saF@6C9 zFmD8lplR@n{X1t%KLwB?YmDc4skt>@@}}9-_Q&`~Gz36i#Z8N#<7X!bE=Uv{6&%eN zF;#Pz6+!8&LajNm)Q>a+8^gzkuT?`g)wM=P(Y7u>SWh&}4^-3&w7? zTglhR`A83_kR&568vbGdjESrgkOtUaGM&?>U=mdQbjbXAc;)cRaC|3L6vtNwie9IY z-aI6S+Rqh#c=Sy_xc)R{%%M67JZ^}?*s0>xm+^FhSqL@p_@@H;M+Dmnkk|@zf6=ri5)KpHYG`elc(_#2=~R))$)m+lfAPK!h-_kp}m1o!y64Y*E& znH*Y~L0biXH2%lS=|5xEDY!@h9SAcQZpOk*F zf+Kp-thFCF1F7|04ejX*LaJC|PprZbyoTvFIs6&N4j6_BC^#rxCHen*=dzn7&ZZs5 z8825sYR5%Gok>p>ouOGUA{iu8n!Bd;FtLs!25CD06quu)_h`S<7b#a14x})N$l(*& zw={fj9MV&uoJb_}+}m@ZIuco*?ln6GWPcPT5XEU~en~4PCPi-?P|rEzF*z!IjV9vq z(+N3Fg!kV$<(DMYRmcQQgka0PB24e5#~!|LRrdWVb~NdoR|^(T1ne;pP~Pw3B;^L_ zFt?;e@3@IH9%$(f4*OjV=ND12?o4f401p!lHKdTi;VwI<-uwB9V?BAV%Gv}{87Cj$ zJXx>FAntL{W@_~DEKS|8FQukSOFiRjWa8aggst96&F|zy?~rSxoin|oixCwZvecWv z(3P|FTQy9If$Vx`IA=rvVefFH!Ij>g9p*kcap=xX2?fvJXE39N0VDZl2G+VCa&(*aXcc|XcfI=GF zR>kq>)c@lJP;g%y5gR!6N;dR8B67>jo$F;`k|&|NEuC@X55D-r@c}ecqP0Fjjan$9 zv#fVp^u6nF8^xCwjK@K09q1!2@w8yP1O8$td(%0%>T;#JwGm{V=cqCN35j&<~IfCu#Q zx`*Dp_j#0_DA7)GgK78>&}Ga2K>W4~O2gIo35qIx_al0}b|2Gh!1UC8X5xRDHGCWz z$6C6+HdO0;LwbD<73gkw)^R&s58K^qrL(o_nHXo-85DRYuxQ0i=;Qb6Nr}H2TjYpUG(Ufl^?^w-na=5(5HyAsB~o zZ)u+mX1vTomo655K5SRQrX#Ky5c}5d%17oTi#@oDx8j9*cq3BJM#%12M4kmCxRGtF ztVT+JUKCUHg#-v3-_1o(m>tn^7c@aC22X=z{9`g-b|fZQ|HShbBQ(DY$}$l09rXm{ zs9Et%MSkxDI0E!2L~#_@==dJ*Y$&m@26vVhPpUS>DoLLq8VFDOG}}G22Ef-^K~GlC z`y!IB=+~u@6OybCe)>LZVKl|4A_;|i2ib;$D3_hB&T8E4sDm!;L7V)WL3kBu-|0_6 z`yCu$ykk&~qm$mk#$~xbAgvVRqu7k!ay^6ws9z~z!i#@D@y($j#dc{(^@y~gocjR) z=K8PK2wA-f)1>I0;;YjI=Qkv0)}IK$uE<&VpOI}Pmh=NFXysdNV|mcPN$#+z{<>h4 z9!(TMS3QCh{u7lx*Tn-$e@dM#BJ^Ghn-a)_acHeqvZvj5C>H|Q4L2x#Rcfbx;ELrN zteX|v`qU-pTo2E5PL3Rs-|rIS1EuXsS>=tYkr}UPK7*3(L%dJZa)SIyo3hmO zFgxb(*ty9AWPOsH1yl^>+LoDk)oFvrax8Jar(zQ6rDRgn-*TAFtZB!Yp%jJn3Jb;t z3;Dss)T%F4cSO4Wqe2E^rE3tm70KpEZ%e(K1M~AdSMwaYLdPYbT}>i&Xcn?-x{P*+ z_WDZ0(e1457SGNAPaNZ#K%++s4Z~R$hQY?p=bv?Fx>26C@1z@LWg-;ia}{%3kFPkL z0hr=Mto`KXicmz)r>0jU-U8i?9Z%$Q*2Uc!o3CUC`~A5VyF{1ZcW*)z@S{ELH{fyX zhnX-sZPRswW))R8q~(ph6~L(Itg&Hh`p(&Hi{^pYyy(m3%d&LS$p*RGZ4%z)ROc&( zLmBvS)AkM8o~WH@sS2Cr8sX|%-Z!9~59sbqsD^&Hl_miG>cAU%8R7fRj_gF<^_7v> z_|5roAnD>0S&-#7nJ9PtXr#qS$pFuxuy!QAD&?e(DptUU3Oy$O!~jDW@q|F!xihql0qys=NL3+l=Ols|S1_clk`;Lsd0FZW zAQp=u_VgYOB(VwR9d$&{-aHvC$TQ%g>nvAcc(Z0y%gCLeDp)Xqo^- z9q(CoFTa?d?x)aY+1ls@Hlhi7Yw=QJFh8v9Hs2)hg{TqrX$oZMghZ6FwCV^qpP#aJ z7R>p}e1rA9Jny%!qE;1;UUZGT&n%mqbg0KqUzMyVpUMmaiJEZS&SATDDNDy%5C{dg`HI7vL%Z)B@*#wX?%XY z51lHfgw(l|8c3{^aO;Pf;*x41atEkMk|Ls?3Ub@MWB}VO{gR{-F81D52~D_=(VtDx z&QB;~8z4e;^2Jt=yk3zfM)q%#6T8_@b9bFcZ}%sszx&{~ygwmKH6=Kc{ZakbUyIWc zu4rMtEzT1Qp!l?P_|HA$_i(Gf-;|#taWBFsE3acbb^iJ;#x)m1MddD|AM5 zpiT8n*G+*i%Z+u}4sfkV@&BW2X^}a=@%FU(IMC<=dy*hkM756~@Ju3@m30+9EGZiu zic2VabtNe$CtgO5EhQx-Ss={;MFux8z!D~<1k{sC)`eJK|LtaSr_t7A0;ORVJ32W! z_4mAZ)6domjPMxB;EN^T86X;Z9d&K&y8Xsuq7Rm~)cdCH7kZ&Hgl0pq2FBkxgbiNr zg+ZabLgCvx>&qf~`s{Nvd$C*CsRi0esntVt&8e;xib%voY_WT^V{y82r!|Dt;hp)S zdO}!c*mVBXTy4f{2LWXxj$FnKvpR>e?YRl&W`=dyDK2;BynN28TKE3;wkz@Of5I@? z*U|Efk@s^bcLR6eyhM(-rUG8Ts|0Br62l}9_pv&=klD|55VhE>ObRwMEzvsgbNo9yXLHjdtHSFt%7p~0U)#b!#>BY8f{~F8Ubk$7q9XFJ3vJazt@r$)ss50sOKW~g z-+5M$&p-H?)aPxUpXuB0AlF=1?-*g$hiEc_ZC^)S?KgkAyE)&`;O^D3Qdi4u2*N(M z;1&Nk5>jr2D&+^$d0mlVxp`|6oWDGGZe^0gyP2%(Jif}|^K*uo-{9ykM-Dfg#S`_? zKHn_F!QBIb5xs1tT#|#5v2|!Fi3C zG|ofKj(#gWD$EkIImC`JW2vN%(2Z_W;TX0LQ>oyL+u+lhB!@pY-AFfx+s*CDdXpsW zKx%QPB}+u%&N99v%qfXpDKbuQzXVkVpPcbN%$u5#ywgdsRt|&UpxybJSzfaPazk~y zYsx=5O_12JqfI^c*SZTi)BQVgt&&6Q$&cFLb5Iq<1YJBjYb) zM=MB(hTIR8d!2666Pr;R*Pqd})Dwd@!S~h2hKa!zfV!m1cupLBy3$dvj<LJ;q^A-SAoh} z1XT!e7~}Ep^$+T=I1}~TVpj%41@Wai0dFV_KONAkMU#iq8<7WB-_OrCohI^H)BrLR zm`6~v)eOs1ns!&X9-nSjnJ8cA5+-%SYedP`u77r){5^*7oLEa4V))*l$7R1HKgo%?MmhOo-+eK6Zam)gQp2lRU7!2K} z=KY)9p`w$Ft-}e+eV=J&C{5Vtx9Gp!*gO3JS0IU z@+HkAw=FOy=p;p^>=`6ebRYs@AaU1GLVD@X=1$RZ-Cv zbXN`{v{?$T(xOaHZ@hctyF^@B)>7Cx{=4dJMSs5P_RIT)gqot6ki05}RK3}P+X21us8>Y0-uoUiAp%pg=c0K+`cK*W8RZMPV@h9b&0u44Ooj@| z5De6jfl(@9T3Y=>luxn0Vh*8rGht7DhqCFd-h)&5+&l0uu)9x*!be2BLn7j)dCC03 zvL!wh-4{Ze{iRwB9LkFV7EH5b%9S1Zd%%*jhK#&;Eoq+#h4r&e0F74sux=bKx9*@H z*S>S7r2gdTSwvzs8Ef|Y$&YUxaotWT>H>7q&;#rx3Y%SzOVcB&W4N#Y{a9qPyb3T_iIdd1{@NrXtSWmv*#a=qzzp9rVI6XDW&B0PUv5>D|V3e&#*xz=TXg3?hP(zhWI zRb{r%R3=6eiGILKX?|-imMIDxmgq<%-(P6!FT!QuSM4R}|a8eyT8 zQK_kIz}oc8B{oj*aC&fK@;1SyES|d7PW$s@CI7hM$N-FA;#@_3?4PG@x6bez;sAm) zOev$L!Ri=1Fvi=jgX@n3qk^a-n7hi63vb_*lv%>Rs8NMD4iokS-8vw>J{3RO|0a<| z#0WzqD52PxKbji4BxY@*@82{d^HED3BR|Fmxe7!Cj9=_S^%gsuhK(A%{E7aw#2gp(@lZu|=FQu>zFA8C%Az{ycGzUpLDGKu=IMEiKU#03a3<-2m|H^5pGSiaZh(;T-?H$x~~GnRHn6@aiiW` zTo~$Iw2^_z6kk7KK0QgIW7xk9KV@3~y#2bsSG@;{T z6V39(Q~S@4l#xO(n5^k;`(r3CMVl3! zcgxlMV)j>HA#ASpZ_X|+VX3w7-$>awY)L@hmm==Yh66DCHh2sI^W28&ax07-={=s` zw950-F;H4{%ETNU5-Tk~-XS$-xDm9&7=&OKwNG3N0a8MkC*yBfe;8rxkl2&cv2GgM&ia6@#ZRkh}#*^b^ca}rTdU{ z8auEyQUj=>N`O_lTj%1gJtyWh`fd|?=-+z7Q@=v6SHErye{#Mo_?S4!jbo6b8W;@ zx}!M;+x2t6BLM> zfE7+s4-ss#hTen$o2}BJi;E7oDHp~yP@wcES-;>MX$<&KG3bmw z<-Nov6`ok4IAO8q&YLcgjDh$-FFV;oa%`D*JKnqjb|hE1I7|Db%vI2Z5h%vxx6SgC z_=ChBqqonXE+T0IDUdVJW9RAfd6DC5#Rv$Dh;9ihW3ot$e27lS3fQ`9kx#$n5AC&z zjP0jsQPcH(6+HwsZz}Hkf66UANH0WgeSa4Q3ixXk zEp2gs+<9J(`#vG7sIYEZS$9)Bcq2TlH(P1A%QGfaqSMrDNt>%mv zf_2C$mjnMV|8J`2F~jzIy0WPb-sZt;Wk(Uxw8^;P`l2g6qUEsAL=G2FvLk}+iU`vS znhFJxQtsf#pgRwDVT(fKkG}ZVYcIRQYD=Lrzcm8se3op=Dn302VLX^vS##-Y(B5?m z!S{1FckxE%Q?1LrMViSNHh|XF-%+fJI;1zM;5MDXH@_-(;T$2Uo)>3lvG)^+qk*$w zGB|dG>=P8qS-r~66xn=l|qUQI*s$=rVjvwszx zcE*2xtDmNj2+D97U2tI(?ik@UbTpYipd>~xCe91z=l^dl2?<@vB#aeG&RG5bb02v8 z$I$Jx9sVBp{}?*ce_?&5a4CN({}*2SzxAFgvK*?HZ_qTZ4zhQ2g}MM4y}fNCN}XFy*@< zGXiTv?*D!nhXmsIyW7>e)d;9pzaH@}xLIzX*F`=zE9;j~0eu!;6Oc+5?8S zq%Mm=oid%$FzhD{9Cb&sQ^|r(u&ZGLUBV{e3UK3_aBGEvjqf5^$wxHLJ^!FgIzbM? zgI4wh1l@|s{0ef;l2S7{5S`{{1Dm+ua}Rg?-KyjMq{s zg%_N29~Xl~8+6~diviFMtD_TUAk)ff`?`KqB;J~Oc)RSsJb+HihN}40*$H`Ppy_#P zn_?UJL`gX%sH_-BHvT(8*GG)|rxR-P&ro>i@ImFDr5y@XS%UgQYrnoF6bvG{j*+$3 z7lNl}aCja5W-HLxT>pk#wHoc~(}<|fI5AW!!b!zeN+ALP< z-v#E+aSa8rt%UZr9<|hUNZMYrtl={tqVd#t+3Hc_II4y;SYWukAAxQQHwA|tuxw>O z!U)xZ*~VcrptSQ%r>Pm@e!lm@djb90a}nQW5Ya`ZXo4BboMt95a($13_KX?UEoKE~QJw_rHr zC=Xs>Jdh9Q^V_ts?1}%*z~%jOiAa~wl32vMibVJ9S%9yp1AKFHEx5=>updMqZ)ezsrZqot71!z7{u6T(+w)tDcA_-r2V;0}4@=#cx^Rx~b-><0HU5M7!F z)u*i#`K*?xqU1l%kMmhAfm6%BhH9cWMsAq(n-9wrO|kE`vP8 zv?BI<9cSy*fK0Pg&V>djW`F~hqv28c?6fbw#>DIW+Ngi_l;EDby09>D$=BV1* zmDR_23feIeWxRvoO9F+R+f?9Y$4bZ#QDAE)DfYxz?`aOEA~E9bNz);~z657;<-GSS zez)dKi7}0O2+i_^J$Y}dj{?1na~`jx>b&3A`Xn$jo{V3ZT)Wg@+{mkrs@TH#Z0Z`5gV5dN2m1ha(gYtBxJ#G_C;iyBZV6iG?%{O}%`EvPVm*}`B zhd(UAAkHe|MOrg$^3Airq6E!t5)O%5+R2=((lQ@B707I%yih(mmlMwZ)DKv-tqXHr z`3vSUWkA0#55!xVt#?J)Qm>lZmQ}3JDH04bH}5V>;6T#cy;}b+?v_Ib!$^eZ$4CbM z$)jp;@MAtO{z5Zpk(iQ>V76`%PsPTCNH9Nj*6*2%Cp5OySW#N%wpqo$BVtoZ z?*F||yWM|*y-0~h4j=Tl=85|89Q^K^hbv4VCGY8;7Gc&1(vOrEQS$)wo(s>ZsgaX4 z5mK8Fd9j&5nKevL^PoH3Z6EpbhmBD86|6npK*Yi?g-t}bl*s%eLW_lk_K?3v=^Rrt zZ%KAuw#E(a#hpw8`XPw^V&r=Q4l~yvp8wkK%4=2L%yM7t5*3({VW*^G#Aq|hijOvZ%dy+KzV6-q_~_rS4ep@MjL|IVa2TZIv@<%gfz`2sWUAm?Z=A^)MfOkt57Wy zbqG*S9|J4}J>5-}apomtx*4=%vHser+5S3O=+ZBt4v1&17_PxnHaL-d0n;Y?FKwmw*NrZ(q_rh6<}i|(tHNfTE1_GVzr8|SZZ#>;ip6-9&5;H zl#OYx=US`#3hiN$MHGTtpmI$XH6PDZ+#jSgZz5@}1=p6I5mtPc-cHPrc8c1_wHOpBkp^q8=48>T)F&w7|-L(6drcOaoDD9A!pv5*-tx#N}SXutc2G6%p0`N zJd24$`FkdxRRV$JT8#}-9H+=k1fuTqh z+HE9)@Heq9=P3zcFXpqatFEJA$*7y{2`iVa4s0O(uKJwVMu7bAm|QRPZLP9eJ>vdP z0VFxQ4)^*Hzjan9ByiE32-0qE>S=pCE>o4?`NNlwUU0b?axx@gh#0jQQ7aqOz*~MZ z`|W&|-q4_|KqmG=_&7FmL?Qw7MP&SwSo4aI5Xs9g6$)sPY_0(kwE%nwYhnFS7YiwA znjbT3T3Xw;eO~)cAbp$bRq;g!Z;x3bqH)G74=f!UOYYNxh6l4h5$zDfk-g0$;S~dB zEsIUA2tK~DdAs9)ZsTSR&{2X}nwVSOEGlcDpzj_P^Y;~ zk+{#O@-fIgN!EpPzxsEx{!+6sUy<^Y5ngSBw>Oufc&qL(b7kRQ<%{Yoe)elO+7Sz! zTYGNUuh;rMA5=~-2{GbKHIF0fuOO1(%Of)S zVbBj@6(718LW$L#+dkw3)B+3w0u=9nz~tW-S@7R{Nw8s%75dh<8>Em7xy|~fS)Yc9 z76F@(GlKME-;bL6uEV0?hxC0=pcz31snwW#QU#i!0ybeWF+Ka_khp2Y;ola!*xuMs z3DY9&scY!>)}pL7L;OA)jSKMu1)O~RN2PRyNJuvslvog|bpf_>iYXHB2xGOe$i?8o^fs(7 zlOLsw$~t!lNrtN?h_3O!9|{pF7VB2S}Ozt(2}Atcfi zF}OU((THwF=(8%eL|hY|J9z4W=|+46S$eCn+i ziiD^HIOm|d4s~BrLffr0nEp+rZZsySRu%}8tm#DU#p8uDbTo4X(xlsmC(GipO^p-0|N5*`REs3BEGzrtqJRn z&2K-VEsF5AycuWZq-2WpOtUW4N7yWl`YX_c%*2rLL?HQ`f8)p1jD*FYI7U0uud5>{ z8bPS+n8(b>nAzBLp9grI9@H$C^1lEAT}CBCz#49xM4MXC-a?G@O%>2*>R6O&EHP8C<>Ok zd!96#Z5eZu$8|p<@Nl%ZHlmX`jTVp0`Vb@QDxJG>@R}5v-H=m>BS)3&C^ucz^&xgT1FWs=Q<&-P9V;Hs~$Q6w@QA!UXj@KWKjA}-Bk5l7vmQVU4gZh70LeMUr z4^}e=qU9r}_4pfCR4fG&t={KMUJ4IuHX_-J4w9Eio}A@3TIXq(MOiKY1S|>Pf=LV? zI;`T|*jjffDY+)qsxIoqwZ2eh#0oFo6za@*vAsA9r43a)SK$IB5QQS?MCg`MOx86jv()hP+)TV z5Yw~GC2ccqKJpQ3HseX(N-689MjdUlUiN>y5z=P8p!;y};yDs$gbThua9&OidVIUi zu+anlXs77kYP4PVPbP8C+Vkd&xu_KXW(81T@z^ze*bFTce8PY4)Mv#28K#;bgK8PO zcYUAWlLTH2y)Jw=MGP$Xp5Rf1@NW2D&qGTeUy+;0**vkc+%J8rkN)ayWo>qbWD8zm z*rf7YhULASeA)nbUi1s9Ju_KIQ66+O!RS3*kzCD}4H3fp1iMRIy^8;M$L#!gvEAy- zeub`ke}O%!@g%=?w|X3jJvjcQ^?Lr-dYDQxdMnbnYBTiW+Nf9ek0cWLZMThV^B45%D44 zHpVVNjXj5q2V1bUmeVU(+>I3WaURQ?lt)cnJ1V=Y6P8nFjwB1&{qgS~7^8ug>d?KX z({27MTh(}BSl|#++Un`JW(@rX`mh3R54j`fZqkX3nA9@m>lUoy&h}>>!$gAs7F)|o z!;|Riv1#2^ohfmD;~i~WT#`4^j9lk}QdGvBSf#Z)5^!H%{ID&U*;H2u9Mm))4aVt6>D z7PPFYACkQ*e@N;qDk*XT?%-ZG9WTB9@&d<0x6Ed-sN#lSa362tzaY?^BzK`8orwb` zpXm)!(_tWSvQ|OC3d$>D{ur!qN1OTb%fwd>;>?$qw*KX*-3>S-S4(kbSiH+94>f1u z)^11;$s6F4!G_BWC+;8MjhIwee8DJsulTE=AYZc2Y58L!)y_#rdt}7MSrCeiSnKoc z?9UJ>)I}GKxcv=nx(|91nY%j&^}9xa=AV2{-C)s_OsaOxN&BkkU0`A)N{zr{h^`&v z&>z8Pbkp^OV|5FQiK1HGsXE~R-Y{-tQ( zae7^^&#&!#pi7Sz{z@9Y*x4UMlkWxi+P&Vxc9JxnloaxiWZOL%^_>VG=L_u8 z?YZGU;3#nUhEY=I2QrOFvHm^l$8Lte5wffpD#n2A1`~Rn&lQ9ZZ=+z++Zh>B%aN{* zK(1iRD85&{5hSIK9lys+17Rtyd$-HjTMquF*OswPPgE$aDPKg)AKB7o3$b@e+wG^P z2=uUte&?q3NVzAMgCR*<;y9Zrg^a=&EOblFsV{D!R(<)Yt3W@oO9PIz3QGmab|5TjSXamVjX+8g=`?)vM^aG7 zz{}-DovYzpn46p&N^VjURK($g=t|gRuB(;6h8HX9tN@IBY+ zS>2!jhlGSHr6A_RyizADj!UKVzpPaQk<6(_oCf$J4>;S~Ls`AYgLK3;@xio=e>Xu! z6hDUe)+-(OwPU7eZvUa6c)@QzunI$b!e-jH+$3Q+|M4Lnb*tt^><*8+ zf$H|ot$#^3#mvQr8hblV1fj$8hyJy=iy&olfql%7ZIhxUF{#6V1RfY}w6Dv-M&N}J z-kgMozbw?k`v|Kwe&KZvcEFqe4EB`n2AIJn-sS#+E2vY4semtFV%XWuEyTyi8COj? z)Uae@t*`6T$?8|2>3hnDk4NG?>@l_t{0U?-=40}}fUl3iga$2n`8=a{xgiu{B0GOE zxOk5TX6{F1x+OGe!rYS>$rqBy@6{+5bv=LpuQ8jsnbry>2Q;9CItlTUep9EE=GELVsB*w=_l6&Vc!PjdV}JC zVAmFklo#+3A|?y^NF+R9rSQ51w|_%>0$~93+|X|@w?o6insY2efQ}3O+mlI4B5EnY zF|tlLo|%}uWg)^|bIvW~_;7Ge#^;->1)}WA zCVv~mcX8RPY{6H^*O5Rp%P{G5)44bH(d=-2=$TDF^gF@Kq^sIXXe7|@| zsPDvdJ>oTf|Fs7N{ccsV61t4~j$L;1uoKnwiW}3O5ID9+`TBI?x*Mp2%0lk4<@^53 zgdih(c|2=-neqtJUr0&TDv6+SB)AHHyfhCqLw<$iy0Fsk!LYzZ^e1u`P9w7Wl*Zy@v{e$Mm0!c+&NPX;u2>fgoI4?705S3JaUiN3ko=nMK2YLVm|UE zq!}N?ja~;8I}DE2CaU)FnXohrzDWs5KR#6JNFvT$A6NwyGv0irlV{F9yZU9Xk9t^x z9|)qpo>*^CO|gac==qLJ2F@wq_$y5LC)z^zKJil_&>i0Rifq{ret!}hbVoyL#^^y= zF`xmg-CuRY=t6STz5~WFu)hQI=6@-Y^Ij+_c6}>ne^LiaD z@Dx~m(wyv{g7^397Kx#Cy_QR5l=586;ke*}a8W)_WK#|3Q!+Qpm|QTn`;UV)X?=_6 zKxsI_J$VJFjjhL$QSx5@bQVNH-teII3v}k_W-{LWKWQ@M6C;wTj2O-opDDqHoM&;| zDMU#1$`rYllyNpQ`h;Iz>tW~PD#i_tVFMTo;0YG^V`wM5vA9=V^CuHl?f-HcR@=?) znw8Ua2SDGrm%G3{&!Y}FAI2C#_qki>6VAdFzBvj4|H4@jTd~+LxL!@{@lJ)lC#|Cq zCr0~ztPZ+=M3yg80Vu7zb5%BS?tf|ud%R5vl z%F2oDfn_yh$QoVw_=op$Tyqkz*9c6ym}lnss7p+|P28zrJ4O;*2N&g%GBnm5>9T}v z6HK7&wO*B1b6q@_X)(k8t#4~fyd`I-1+jeV>Gs z84E)S7lbX2PndB-aEgbqAf0v>R3}JDE|r;BIh~WNl^T9`VsbEShD(R#hOa-FLDXR- z_%Y_=6=vp*x5j&YvAVPw1=2uMpA!yYq8GzM0+94Z>9X{7&1+WC{^pV9lx{iPzR4S! zbq9N=!nWhr^&gKQNr@X^tpJzrHExhvR1Y_%bV^2dG)YkEOo1W0-S};wEulTxcX9sR)M15E{ETYEPAhQ1gWkG<38f+a0TMNM#V_`APFhgyF<(Y=44ZrRoSO#9qv>>f^28 z!Rfk#n;%ao_|gb{vP9auUEL2bg!__ zWGFNnF1b@~gpc=*wy8jvPCyBL?tZ#xU4BNsPm-Z+yz04x+PvzSFz#l|7K@{%HPLa4 zaIK9gCVMli?x8y7%|#+w1mpN=VAajd8MCrReG7zDZQxzZwUyuIA8gG9w?`^Nl;(V=Z!?IVZyWgM# zzh%-*X`udHE5>__G|TO=RP$Vv_f~eOXy-)zcIcrU-dugZ_AoE}V8O zh9R6*PKea2k&ad0r>qB9K$&QtoMq!bwN=;(8r9Uz36G5a<$CT4^m~7QcLwI^S|gv%jf+d`Ml29%bIDQ{m2U;{zZsc z&J*+G_8qZ(D13zKg7k6(`#^DxgQ38CdDRNjA_7Y+rK#5Y!ZYE3jIF7~96{QPef($A z)N6po`yN%jFiAl7s4eL7=1n->Pcyr?bI=J!~;P6K>e%va*E{@9D$0v1a4jmluv_H?0TwS-`H5+^3 zANVw5yixjqlTZ8-{|I`JqKq#U|Kb01bHANYnjFjdNH}hlFBrrZW&)Vg;7N!H>Lm~8 zB~`@{GzEB1w1P3myrX>*;9Y~=-4Pg1Nd}aW{0815y%A%K3%H^1`5*~qYHWsMf{%t_oPjqpD^&Zt;b|K9>=SpAi>P7~a(-VJM3fWM?fhMW_ zUO*s20P;|g^vbCr%M>+-8!G%TuSqJ_qhSedbhR(E3X-uxEN6(z&EwL?0lT4QS2|&- z2d6_Sr*8{y$kniDb}gX5YL;c?7Br{pgle6Y`d$#80W}}%)+LiCuljj!A=in?+N^Ya zjOwShFof$sLXu$04quU|L7L*sf+fdtmyzZsCY9_|&eqMH3?+RW7d-Bm5~t zUCQK%=#N{e4h_v1nLRm2mJEh3#?EBsz%vm=P)#usZb)}KeWVXVIL#)e1&egbxU|SJ zi~&=cJBuzoZ(;ozAPnJpjF}EG3+F&7LSbsLLll|uq#IB_w&JGfqKk06SR1E1(pqWy zBsQi_apPS?pGVZwOe+BNHx(e~s)3`bMAWE~)Dzj#e&}-)?n-hB_*mSdi0%z`M5fa5 zg0=o_E=!y^2imv@MGJK(8g)2>3l4oF^l98@DSFXCA6Fl&)t&pAd{@h zpA$<%xfNGaQ$&zY3ye*PDnxAD7?Cbnse^ewqSzkAEZr^~f0WBRt`i&S!B~H$}5oip}zNAUiJ%?V}>?lXqSV`Ms z(!b+aA9^3B6Ehr1W?tUou(G~zX*8fsdDp);#uQX*7�lYGaSQpn|zh%L=ZSp7zV_ zch~*`LlQw_$gLT5x@vdehrPpAxX@&Qoso_g^WdGlq)eBS5poL>x%u%|dYE+7- zNMnK#Pa}!7wJmuZu=I<$t}Fy^%7vI6dq5FgRyfd`+9=Xejr#!7Kb5mPew-xtPzXsPZJi&c74w1Kp`;p{ng7T+YgNGaM9^H6G#p<_p2^c{(y2bcCi;ILE~%M z8w9ikPGoiDogCX-PcqWg5B}iST}pE~AhpgtOG6Rx{li}Ip?AB$QdR+YlSJiJzQ*_#(v&Z=-itw8(V1IKC5Ta zgqaThJBNMdz=kCVyC>B)SFuXbn<|CZ5NWw+CE+baC=p2z8u?i55}F&{h%7*#s(rKp zXQpaxkHTIw*RFuS(NV({w<3$?5R6Lp#V;Wkn}-sjIGxQ=s{UW^t|Nn^Zwq+0V2plW zFrPKAtO+!<5Kz}Pm(Uzdl!u3@$h+TjtXL~@XwbfBj)rL^&^^=AvtYxRtz*i}R>f;Y z`k7=3wM?}NwYAq8k?cF(5pwP5Q02PXWQ32QkZan51}D-j<%ov+DPTBJK^4#^WgEwB z_ZywZ=Rzkwh8u!}p$oHQSH}`7@2BZTdHq~m!re(MAG*r6bNMTF)A}f6`O=oTB{c3s zZ--8yI4e(O&5Au^HAbSpNgn-?n01aI`}B1MaSPiMCF{*lGXF5uCZeV5v&EB==__Ev zy2X;=mG&N0FXw~c+Q7zr5c(qi@@?vxD;YVHbS95QD(WJoyM+-x0&-u;tMnObt+wHn z=lP>o)JP@DseUWzp*f|Fj>RQ;yYSEp?MBwt39~#RCR`LW0q> zcQ^+HD)Bv$tK8U1qa$kR=DAw2@2u zz~JPx79$Prw7l3lD8TLP@`gv%&*>;i7WlzfXj!TpxvJE+eyunY&5Gmi&zE|(LR-QN zVAg)7RK(Qo_sU04nUL3tmt?9+tk?Ex2(AylEhwudcxsO=9_E+KlM?x=f-@qo(6lkW z49)q4f7Hn}$uaN?{M!6y865}b+aqK$pj$Qp=S;WWUuFZNk_+e2Z(pY^$~_JV10&RK zqB*PhkMXqkf>kT~Pt9+QH#P>#A%v+mmqc;P(2Ui(<$c`wf!T!3CNK@2f0eg>CN-h> z*D-_~&QMGCWH{5U3^#x5m~gM0$&u;eFjr=`6(8`4qayHXqAuJVGfuJNBoqE>c{+1% z6vx{|G1^$1*r|m&adK0_+uVPlb**<|Xoc(V#}g94>k#h0tZys@_FJN9=%Gy3E7Myi ziMFlU9k4r7W=5b=D4e3MhVN?YKs*hgHk+3v3OKrT-miW_2}u5`K|-{k^Dp~GBlr*@ zLJwj>D-+AUlutVh&$;23d{637NpWH%=u)>9!@sk={fxq0$ZGcCUY~89bcCXg2VJEK zHuG1DnJ>Lw0T7DqvY*iHd6|zSfIX$40xj!B8jUt(vFIU{L@>XF_nn+rfH}RO%)%3q znvgFOUzQm1b!>Nn!98IQGWe=#@Jaw<^daS)if&Up zd5ZKE^FY{OO)J*LX&1plDOKTK3geTN^-S0w!S8l5SJg@zstx6J29FsXgjqP5UZ%qt z<%kxu_jN`rwV}5@Bpy|%O{FSg_RS<+)QVlWzk89|2){W3=WBhkEtk}_Wab_r4flQ$ z#&3;BQ=~$1Jg1_};Iex1b2j2Hj_VG=aRgi;g|l6GM_Z?oNcc9Xa-(kt89ogz%qFUN zkBGFM8n@q=SH8gE5;TdMrVu^4Jzu) zF}5mgODet`5rgnhb(X{6>cB%Xg9vH-p0yk$dpG;#-bAqrTPRc=YC~V7b_HLODy_6b zlQSy>rVSOf@tO)o#RKjlw-_KJZQ9?7eZFm=U*z4~Shu^6f6Fw>e(vk&kMAtu1V0qE zK@lZb@&M@~HSM%-*Hc1bFq>3QXW}ZG{jSm2Mt@=#;(>A5owrnoAoHWeo-1q#ncTLanLtR?$L) z46vhiV3l_jW*DQcrvr2XDbVYj3^vlufe-pT{VW0Tn9w9$&hmO=VStt_dv0&E%|9cE zH^ZAlaKySd>;7~@9t%AI|A9lMX#S9(6Qrc>T8__|UL~^KuBfum zOp-NkOJ2V1*@`0fXCF`!R{yXKeh6eD{`OIFX^RMuZt1W)Pt_8Oh>IT2sb#yq_dzy!LEU8=8xR?E%*0}?rg*( zPZv}OPG8}K%DWuc)e`wAqgFl_ZYFX(KAkFx;f3V}*Mm~;C3(8qcx}bQ%eG!i9QL+| zMu>-Mh6-#yJoZJU^dI1pgYI;e-11);f{PRG)MKYx>$`u)TRHX`nz;yjskM)80vp`9 z={W7rxC3D0(c%T)lG3&SPz;aQ^lD+tQqSx@x6le20>BMmx2C2-ZN_~b#N74Ht3 zep_^eX{T_o3nPC^@!diHKk$ipZR{Sj;o8Eo;i!(iK3 zE`=Nk_PDN6WAF5#dxd9+*+=FxPpa;^%LYTv@KXod7R2U=`+i!z!#%REt+F}ao7{>f z^NWch>suz*%W=QFZ53Lz9d`H&UEGvPo4#oy_fL=VSW+EYl8`c;;aiMpk|?qnvhH3R ztl(9F-8-=4qfNY4y|e7r(F^E`TP<^>0w-w&5B3c3FK@CTU1xQ@uuhbOS>Agvpllw= zF7*WS7czo8saH?T$k4*OpBBNJ*5TKDFsX>;)YEaq*XBbe;uh(ud@NbFD6$QC<;b>8 zmuWQAsj7v4mxagwvF-V9Enp1=D@^`(Dm@y8Yo@|>-eV!|umXzx1lEXp%*y4WjyRC9 zSFNQKsjAffv{}~fj`<+)A84-%7Z%~M_Nt{L67r?oo%)jVAq2AAAd%R498rS`$?Wqo&qE_6TOaVc2eDm?aSthQ zhh9|Z$d2F9tdJdEX6o)a0Uh^SfPrkbDZ)Fl#1{$lfCRBhzx)0|IYY(~6DSfwdM58w zF{>YnFZw=if7Z$Ls=N+lbV+7YxISG2jaOi*4Dpfx)gbqHQDSsLde+%zBUnE5Y~Qa@ z|*r%cYe9SJ2bSQ~xV0oLBo@cRGzCko{48md^lb6d!RC&VX7G z(BR3ny(DS<$WSD$LR*Eu4XmMfn)`zpID??@E!scuPka26l|jp5jgM#8j|KgmVIG?i zLvPNR*MI{41oHr>_6H@4cX#~N9B(+jKIw=tVH&9e*He?#TR|n=7nXm(;MN$a{*YNV z-Wt8S=vENg6!XeCwItQ8fn*q6ZDOLhr=f+Xx^*s9DD8HO6<_#Qcx8C_>B={q#%vfx zTTn3*Mllk`=D@{FW9k9);#!Zq$7@Btce>PuTY_CTFIpDkSY9@tg_!46X3QZdWuDaT zi5{pHnJ%_-Mm3_{EpHaShU7N7&u73H)gQaE_Hd}oo2ht#x7g>1(lBPZS!@ckv05G8 z*uGpJQ&glH#N&(1zDRwk!yYJx?zK4DKC`ua3mz_zJ+JKPGhQK=ymuSc?Qv%K&cjXV zN^r6a+Z=ay)N?J9Pg4hK-k8%bX88e{nxosG`ae{CV_+p+)^*3}?%39iZQDji9ox2T zcan~6+qP}nwv*0x`#V&B?oPs(Exx%Zp8e4T7Y;sZ{Cpjej?X7!_a9&9X$BQn%4x%;Sk&JPg$f*2F0C|q~7*4H62-;OrN;^ z#DmUX>Q5y?rasc~al1yuCu1w3%J+u7ocDm}5^h%a0WRjFqW8zv3@&%J24g-m73HL0!3t8BwDbl4I z@dvCW_93!&xg@^!mp3b59@D}FNugV1Sk>xCsPGQ%{z^2xNd}s!`*?xgF)24;NCkd& zjgvlMWa!EnN6$XLgj=$c;AM-v7aV~I8RU*cV!wUe?j@Q5jbpR5%`dzE&H*DJhu^s!rs2d5~^g+cE0>Yuyuop9^;r(4Wu# zm8oKYW(CogjdJSdHB6^U1cpB7t*^SVv?nkd79#d1Ww_2hjwZg=1uF_9Cp1Ct*9ej~ zT(eC#KL4~+_jpu;uG@_Fa zUA|kHvK*5_JDTfEY~)9kr~dnBTIha@jZ=;rtLhUlezKALEby-EXm-(&`Bhrq|8qig z(d;m?)KSZGv00X7iO7(uhssO zIsQtG|8?^8zp+WbFfITVH~)14_9*D&v6eaLMk6C=M`rN-%A(TJ(U;**2We@T+`POA zbVo4@a;liW&thnBVy3197X6V*NEB7poEl8_^%J#FDDmi9T|-F>(Sl$iqW$?(GG$U+ zZ)Fvg;PCKo&mZ1HE;WX3KYk$T8yLuy(HXEPSy|D9hlk6nseR30n65LCBPJ$x3}v;H zB8+5ig5dF3LOAY+J{B&>;BwKw+U$ttPuRAyv=mlX$9jFdu;9(%vRhN10Y)hM1F@J) z$lJ_;hV?fyAr}j{!JH7{>BBE(s5xW!R0^rFGryOYho4GHF=vosgn-NW>59};EEE(J z_I;(IUdbCmpFf=Ud6kuLEiEn3u&^Rk+8sEc8-I#%QgTrkG#V}KZc!CfIDqc}JO53{ z#B?`|&N-8zI2WD){e`NOa2gl!7by`L*$`sA*XzB2zJ4xP05d9=^SNzcWuS(JoVt3A zJ_I8pMImacKhW(1p%EM z#_jIujy{HM}W-nV6U$r>1rs{3Q?o$}`j(9%*tL@i{#2FOg`yW4k;3!5l&TqdQWets|FyRVdqMu|`xB{@(d|I($B*4eu8b7sw9_M{2OKHj{oxB8N-t$AnMXYl$UIJ~r()+s zgH++lxbVx@O3E~bc{Q%9C^$^)+&rk zKvTD16?mgKviN94RN+)3ETRe*TX{_3%(FAIrrEH7*hATMR~1L0ZD9li#6+*PKOb+g z2?+AWAt1uzqYyfE7ap%pv zARA}`Y{`^t_kF z;g`SO*dO}8>KMK;SXOmtZkeZ?CZFBhos-1|yRZSvpt4}Nf^ zRY*h?ngg>G*+wHbH@9e-bcLW(5>Iw7XP2w0@Nq2Um#z9-a-t zYb;mc%L5{bUad6I#VVbveryi={Jam97ZkCFUzif~B}vZeSSU1Vl3;tQ@?|ik@wW+y zTG)Dl5^ioSb!i$8s;Mt8FO#}NZ~O@Y0nyAXSeo@I3NI~g!_o3>Gs?0uyuj5O(%j5W zrb8oO&Y11tF4lEmEkT57dg`(Tzm5@h!{| zFUue*=xp`2Y=z_Kb$IFHHd)N8t*h{(>Mf;WZuL#=$F!s6JeJvCY9+nJqb2_|3)lEf)e{d*fNR!Ms-k;Y*(r`<^oTcX}h z7|RAx2~o~Wnwj7@liMc}kLPv!OHRG;50QS6bQayv_fwt)A*Z?SoZ51&wnf68Q-oE7 z5GCPW5l7H>%uLy$Y?b zi@F8PV308EKfRAg-hk!>;GEo`vsgNu^q@i>f%5U|(igSpA4q~Yu#V2^Iolx@uYNX?bVKC1Lo zux61W@(&6b3BmqU#af?pl8S*UPN=*{kKtAITCnw`m_(bDRhMQGif+g12MFG!@sNO6 z_XCRBDha>P+R^g+0tJ<6uY;U-=AULnfQnxl-ehqfCcwfo0~7?;^vBPcNSe1)d2xrx zLNH_W{kdh;xk#MNt(vYdQml!KLmB`EMw07k=6?EqL??*_iQ;{1I$}*C&YZ)De2^lc z!|$X)gb@}BGN3W9hV8F&B}`?~5lITl^$yye8_w4Tvs5nk?`khFeyjupImEf7z2tfn zilDJay1<>B^7=X=SE6_%Q}?Nb>qbHkGP$qv1tjqhUu9_EPn!Y@Z|5CgL6M$&3^#Bw zhy$G}2;xegLSQg=5v-p&$4P&8G9S%4<5Y?$v=Vh(HqAQDIYD6-(8%0f@APGO3cDBJ zCP8}470Uby3$5p9B^y_=R`s})WoD))Czm-mNbIYKfNqnz){=?cZJpY%DcD(*S^CSU zZI1&8D2RO~rmTtuC>lmSkc4X7U3i>r&1SJ@_O7D3S`1HBsngLueX!%5W-p8grhSnp$+C zDzQ9QbpZuxja3Z64tXwZ9rRN-N&(G92zYT|(`fN6`7vOQ`rB)x^!tW%TdNv`&wULqKiiZ3#D)W1t`NSg%MtRMIe|)snIL%Sk7_a_4NR zX>HCH&Wn5LoUsl*Y*yx!CIJ0(gon z65r?JA`mL4H$oyR9ncVFWMo_*h#?BZ_B%{+0E;uCiCG+}P|4?TP0G$u$~K(@MZ0{8 zncFN#&?U`Ynk-28jU-zY?t~4JbV)@lD_E@7qSx?#%IH~DuZL*x7V%o~Do7MP0#7BV zCeKl!0RP!(f}GmcfZk911yPR2sJQWX*>I>@nZJD%t;po1wrdfU+=_39pSL>A zyb-ncfT~Uzr4q`7r$xV2JeTIZaXqW=D?dGSR?oayua2r7$jjp#6oHI*ZOO|cU|3eJ-}o2DKKxF zca~X{=Vf!g3%!%jZW3TUrLB zkQ)V!&SM6Iq@wDt)KZs*Qye$(&XzXge=It9InrFjzDX}#)@@suSuxis^ht~v{sqXg za3Hbji0nz08RC>kxwC@e4wM3g*Z*a zFK3y&uNU5DtIC?-I&DuVay=hbsfUUS4%Avg*#QLE1>9eMt{jJtOWaWqQSFv{yqfB# z4ID73)~#xbBn{TLL3~w({WhE5?BCd?r~AMH-iieZ)_igW3UsOV|C2(8jpBEAl9uAL zJ=sHoIL}BW9OPy;hZA5p~7ehO3gP{`Hj3@$PX8*4-%TrWvp5X#*4 zBd}RmPEsmXb;Q-o-Dog}y@)Zi-6F3=Wv0HDA}G>Zs_lZ(n;4zTf_`>gO=EJ@3ffZ? zoY}+q!jXi#Ami7xeFNfZI+-7-o=MIJf-O|#qtwg~AT0J{3i5qt%G!!4^HSoToG9ya z$I~T>k647x=ALG`e(s7}{lci<(OmCCtS?@?#$~EcyGuP$$@+exk-tI{Utc}@KKffI zEs`b0C9Ae+Q=An{N+8>a#lN8EwTjGU^Fa9SH0L&2Jkm=6^5h-((I@&6t7! zoWB(HvC;n>w-p7Bdy48dN&oMGIFbJ{CI3`o|3zc~pRdrs@hRx@#FGDEMEFevyEP}B z3Kwf@u|Ubf_#GVT_O7+J&|v&@!S{1VAMUU``JUCuB2K!&zXx~kKo@@wS1ixG1K+?< z{Mpt(rZ4x!n0@a&}giE#6k z_jxd=Fej(ygs&%rw0Y{h%>%Ig3Z58geqpGvq53(~B&f0f^GUSzY*E`w{8p^Ci0Rpp z^)9k<4ZXrx3)(N9WD+YslHMgJ+V)#%E^qFi0`J{8Y__!RH&ItjrSmlrey`pH2Jfz| z4zKGqtsL7uTu)u};BFA`c{XAb&WImF;&o1cmn~qMRwu)ff#er)GBpP1kRX5l%e8lI zXI>11(6%cy&ZY~>*6ZYreBBW(XvGNrlbYBo>+E{0df%u%biKcC8~CYr6VgF)lREZz z^ou{wc_MWEqMTz@!`lWDLD@Y6qG=RClk-(_*s{>ybLr-{1hD-={aca$e`*O>Z|Ua0 zj5qlu{D*`3=j;Fdds!M#GK)uhb19<|7j1CvXtB6S4R5Lz-~{5p0{sHSga$8xJ@|h+ z`p@nN1M%hLPc+^U;>W=^EL!eoyQsGxToDD6#I&zs8KPYWtQ{r5x$*FO1-;0Wo4$Cc z2n!SE=JxIw;M=#w$_&eXn~eTMIuIe(VwS87493whZoVzrQyrIdiCbK>qPN})!37du@mzd@-?QE%rSqCp#LTw)=7y8Ih_i^G-uG%fAPg-0X$#SW=B zSm3WExG{1fhQx&T!1PcI=2a$gf89+uakyfUnhNJywd1fcH>epeVNxlUT=+7sXT$L~ z-dq3zqB^4koc2y5cq|^m;Ut-k)X4>R@(s|0+ z1Wy}x*q(U92P|$CmWo}d9U;Km3AKM)Y6S(|q3IoMW+b;2Ll?9~R5$frnPxW3E-2}Q zz2m-+SKI>n*8-g^*!1@vCCv+j6gr!rUqpV(ZI_BV0NylkDMuf(e?6tT$uUazTgvUo zqFhzzr*c}wQklp+kgul1c%3_R{)*Q?Jo)yVb;1dZm2{KufekQ`T?G2hyEP2!v_JxA zUvMZLwg&jJyiuJL%!Ws(`nM?n)h=sqt|tX93abeI$-H^CZ!GA!#%n6XMj27iIxwk} zx7Wl_ePfj^C=^i|*Y{Fuwl5V-Gfbqt%t06PlDKr3=GVRKX6n@#D+Z+cN4-h(CMzoD zseLth;{n*6p76m-biE&RG{PcINR?Smh|WIutcmB8K}+ai7FW&R+xR@;$9Pww>n1G% zyA07&OZTcB9Pl=M?r5I(c4Jv|(cA|l?@V70qnG}WsSwenR@1n!r3yk30nDHl%iSTK zwz|HM7-_HHPFxXj<-o+OKe0Q{&jUC~A?mDK{Kz(9#i3HGclL^I_Q%GiF|Z=M4GMP$ zF#+^%e>@Zr#QBwtSVVO^nJP0l!6S6q@Ev$*gUKk-=BipT(W51M&g8fw3)UV1%&d@E zZ3lPkWjd`61T$d?oBBn_<d98?;Yfu_28%t59LB%6eLQDe}15;T&;(|-s$wR z(zb}&mat?$D1FyL25GeU>4msCc@nmxma^Ps#i4@Ay0dFxJ?JI`>Uv(&B(GTw_dSVi zcU7ZOaCuWCxuYDb?(!X}uwZc~$Ye|CwYw&n(GJE~8awK6W34)m03R~|vZhL=_kJWk z7H5FO3ox#g`w0#-zv;3dnH4nw#FZh%+L<*UF4F1#ki^t>-*Hm%$c|iYoKA<_Q-!UR?_vNfNd{qiA?J2W6!O8A0l3dSkA~G^yvvG{{ z*6ZZ=bW}q9tM(`tv_W+G^pkj8I1El5Eu%X_h47i8e-?2(uNQG#-w}_Qn$exUG@9Fy zwicMBPJy2$76G45^-Q4Z`cDQILutP<5T$alhk8u;HZ_}3G}tPE-JRtE6A*T0LgwuE*n0j42REMX2`x;XH@(lnmY+#!5%8+i zBAVu)yCo8p4@}WaZ|0UO+{0)5S`4=?=((~UP&AJ2;M|%Mr7KN{7Urb-@{$P7jiv)rQ{_J6W(}=BUY;X}?Mc1@Dy7dE&CkVn%c_ zjHgkVh^EQ;Br(d|oeZtpGXpE|M^Rs?^Ul|eWSzM1y2(>9IjEiH_eZ+C?J2qKm@<7t zK3f{bgFgUC5H9J7PI98-UfC0js3$Z(i z*xa4pN^YZzgH!rFH?&##U}~;pU_pAoPwU-VHusq}G1SbDy$*6XnIWF7%&`0qSPVC3 z&JHu?yw5->ySbKu=Fg|6Szz;*&g9Ii81J(d-d1Z%OYY}?!PvqH+#y(TX=!0+)`F0Kw;Plm4-gU&`D07U zT4t7YEpT2=UY^p1*BfkfS%Cd{9o0|w7-_)RAVbr>l&-3*%z66G&YOG?!qG1jgHbO8 z1Kuo^(YD>!+v`Vq3e_UKH9s8~O0Ij!hu0f_Mb{&<)rO89IZj}fs8JJd-YEMJ_bO>< zn7;?oP19`X>gCn32RGKSRM08G-TqME`g+p^aC%*`T~Xs$iJ1|X!&;~hNHP<1bR@oF$?Mtb_Z~c&aO4^MliRg0_{zyFpN4^m zInY_e)@J={te#vS702A%9H$e(v+7RU0}*c_bm;Phch~v@tuk|Z=%)CGh8V3dpFIfO zo1`O3RClU^q6Vxp)T5t#Ycq^d!on@{tS(>a!k@2?*e~EoCXrtbdAYneE~q1ImpQP{#LBrqE`RZB5ZV=CF)2q6l2kbv+LN2A=m-RvuC ziBx49q|8~bHpGNvKk`cE-$#}LMvjc`1tc6HzoF4Xu0!+KmuPsMf>H#pJNouYZ6Jp1 zyd7sgm6j@HQV&|&FoV(}5YkejF;ZPxK6!5t4WJZ46vXUUQJ})`i>{h)#ctrW znf~;+x?goO9g+Gpzu=7LMQA!g&U)&bm(kABtHA!ni{{YNeP*U8KYs95oN(IP&QPuv zom~#jxF{+@LmI~HZHvf&YvthKJhzp<(lP(lBTL_Tty}P>I3So7T z`8M7O3q+afK(on>@7_a|upsClBbj{%F>mDnFYN!rlB<{?UkL%c7cl1NGL;S9xgJmx z%B`jy73voO{1zYk!W|#=auU=D!1r@l+!g^A`+%88zwBoUjGwb`H*T|L)31x$Z>uu43ZE@!fD@6IvX)(sZ3$h;Fl;iZvJ;sK5_d(ZOM{htZ2fDtJQMC_pq1K1bo-9cKjwy6Cg|MVrlM;Z>IH z>WfzKtC%XCV~(V{^8b(FKhV&xkpPGwX^vOg{y!)1k1QV)glXy+ZNf6){1SQ?PZ=9z zg0`XUDG+dWW!=w_IMK#xU-Aj$S>eRGH=tdakzqz~TmaiUSF1Kc+jYZP(SdQm%bLeh zK?7bE!M6KEwcisZr5l#bHCh;%?Og(GBFRf>%2gF#c2jryn z<^|YJo4_HFfle*(c`{ARy{EU0HK?u))NgE?Jy6bi&^Y}3UPAVwA7V?mJU}(ZF^%B? z6Hr^?3;x3dLDZ_euwzpc=!^${MRv;dmq3mV>E7`YyQ#RsfoJi!B$`$mqyS7zX<+^x zU$Ph=4--E$xe0s{{}+P-6wLrHiWJsr;eW^3-GFH*i`)6J5Loy)B7nDdA_P8rn%en3 zoCvm9*YJ8hbqJo0W+R4kaMd-||D>)g&3BtBI~ z9$>E>an(jvpp9E++5_P_o#3)8*&0#jFM44S(e5GdrJE1_bbIzTLrCBTrD2_V0gT^e zfNw~Dp68Gx^-407kKz~YkkC+mk8N1yoYqpb37j5g1ow9BI24`lDM)U&OL?uWJ%mg> zhLq2*y6$9Lan4zV#~3U|ZNcv5)vh~~DCd>Vq$1X7WnVv>Sc+{-dHHpDWPPq*+G#!@ zCeGm-O!qlxiPl5JFLTZUyvUQk+^w$7`<9-z2E^0Hs0Jo`$xOA&{Cr*p{8Up4LfBMY z<+b^Oois7fhJV1-r6JQ~K(I<9nbYOwBl9(JQpcB4u2{%Z0t8 zt<1&x9&hl6FmaX#Fx z0&Z#I;?Kr)hq2|=z?*K3#|22>tTcged;BfwCz(NUn0woQavhOauGXMCKKRyZK@8Ka zM}2{6J_I(a&Ru2oi%Hqb&q-g%39;O|?B~_jX6{9yd)OcDsCK#GaX=IRVL4ff`TjJT zZ&wZ)tTz%BoaTYYx0*a%HmuYRc4p7a6L*b2VnW?}H-yzbP=6k7JW-b7)0tUyO)j0! z%ZkN{oAB`4Z<+Dlj>w9OZHRZSb>$RzzlBS!=y>vP$iCk!#pqv>X((tZ#;kI_L>o1k zmx{@FfstwPGA~?;Z9Z~$ULo1UlF^Oc@%793yrMK-tVbBzK))-$ReQb3DlNr-aRVj) z`NA76^^})w3q_e??HY>y(xOpgZ$QD_Y0JEnXy9RucZwy_tC{LtufF>6D9Ufu>~e(l z{-9%laBAV7RivAEz8cosp?u9*hL-we!Qkjo^St04nvml04O?DSG06X2tGJ48Al|19 zKZ@bx7!e)P83kqEY6$PK?OGl`m7}DzqH=I^S8y~7>u5S@w6=qsoD%U#L_&}E@wdZl z!3Y|eGOEY7Y6BC*R00ToCZl-#go3u}JX_T!D=O;Yf2R3j{CQzJXL2SutQxBLyIqBXwh$_96TPO0htx%A;s@~^j ztLN*FCIw8iph&!jpRbRW`&UJ*GfDnyHXT?k5omh>O>Fwt{Q?*`HHdCAb9Q}!Fy-ou z@GOXc88UrrZ7&B3Nr~U?u$iH1s#C(A!o@9mXWd?+tL?k(R^q}uv$0l zp+7Zw9T*S0ody5ft zepQYnnciCg8dc&w5udKG5G)C%FkyX1UC$!-65%4yr`5ahBFS%;-|q4G5LD-n^dIl) z6#T@W;~;l4>9oZTHKSerJkTe4oL2MI({;W+tl>`NCG8jKUx{+RJjau79)-{upmcr= zsELPfOGK#?)!4ctTnFk-HNJ%6xucr1wFhr?RpZqtX#Yk%ADTbA>QRp0y&3&@G7z}& z4t(3S>r3kLtG1fj?Qv-YaMtowcC_MeYS+R6JamR1*ZrQhyznes7JJb{``lbM!_0Ud z_d0cWB0oZV>n$qz9~PTxTFHG^!?B1-1_=U@Uw2QTJ#Q>@L=AR+aKF6{K_3b8n7yMm zTd6~4`ka16S+HePQ4h~lwCUeSf*sPR!yjr#g~(sf!*E`qX{O}@pW=BWzLiWx$czhK zyL{$;B^I?Un0Qt2a3TV7ALwz3(ZFPz?hM{OKdOXkCE#Wb!#D@H>A{}T_Yi_1mC5w& zE^%3XZ<=?C@tvji8A{J*RxMCw=>~|e^tI2MVJ?y{TX^kZ#1qu|J!PCk!Qg) zSBZLEJ2c&@;Ag|O+5Rgs{Y$zT#wuAMm zX2`f5vfblKUu;}v7z48CUb!8*zEYN9Kj}mF^2m{ap+3!~s|RiGTg*jjy!js@diP#iL)et)yQvXT zOGDyUSf1~5lXwCCL3FxKf4r4uaSY_f5!M}-GQ>+Z!r8WT0z}PFg;Hy_6Zp0-a>X@W zM>v-M%zRJ&5a1Q;D|u5ru8+;`FJXi2o`{ti;E z)Aj|vGcMo+eOy@h0qgysWM#SDn}zjKdMLw1D+o~t(qKHBqS^GYfL!Mfc;*<$W%C`a zG#*MfGJHp9c%kHeIy<}94M2wYC zwNso>CNYvjX>=1}-U^7k>h-7{gX8^HS=a&a+8V(V7F&IJ_%{7+Z0kX1pFSQspjr!d zg)e64NJs>PVP0F~`t-}L1I}DwxI28zztldO2X8__pV0L-zmFQc0k(L*a#ks z!|(Pn1@E;C`S=Y`vF@-J-3d=AypUsa!jPaQ6qn!y=W(&_vUT}DMkMiuG(F)Cj3~0= z=GoGtjkh5xl1=o&Iuz#YEh~7OSf^N>4*FeiRr?%~QvX@7QDAmu4~8?Z^?EbR_CAYz zbS7%jgU{q)KwpZ+H{TLX7c_d%98Y5eaX6yuU#YLs9Rx6$qI^6&hU1iS>sfX0;g8zE&H|aEb zR7GQ1v66@3c#kuAC2p)cf^sf2iEbl}UhtU{IouKfPybF-^%zlSa+iYhF6|3c7+~*IxH(3*nwh86S{^OH)Th9;s_C~$$+G;i6GD1iCKNZp% zoGN)D*Hn=gRXN*MXJQSO2zVV+CP|sZ;Lk2(LT+YnS(2U3LawiQ?iAKnaXxU|Jb>Fe z!^`6dUdF@4Ux&G>AuP0gaLa1orFy=6Vds|+7EtoJlkxn}7^AN@XQwnC?;GNMF_4VDDv^8I;P-Hvs zkgX=~t}@?vWS8IfM_Bb;?$+JNIa<-ibkw6Hn%bfAqPzTcXx-q3Ea#i-PbZq+f@_wW z9T~jq>}e<~TVSucj#Pf5lkmb)R>RkvF7ZV)yvgnyYS?r)qI-udG)yFOc0kr0(FiJU z+NE3bMjd6jp-l)`50LS`(qi+Miw$ivG{51pZOsJ$Sj^4n8*6) zK@;8&g-B|aw@t-lk1q>b zRlm7HQ5KH(;5v|P>sp{6PPCyZi^wehSc>G@tbyK>jM+iw^8mFuKkb!q>4J)|&J0S- zpxYfW(HRUAFt98gz9`1{oyD?5^_aD(ZQ&ZV+F+sT=o6_@R&$2F*KZDZgzTAJEsob_ zu(Ioz4PUV%Ir>@D_-3{{zCRG_XJ%a<$y#W3B$~pezd}BSHp;E%T8GbSiBZ|{xLR!a zc?8W4+98&mlLIm^FfbF8!`9+f$;w71O6eX?hwb^2@Nxo{cC%N-6V3H``#o7((Of6m zVITm_BLxnIl2#4KBh4J*N z@YbYh#iD<&;9F?X+-M!i8MU0AU06b_C(W^CcUc=}ZI!KW02>a3cAo|e(284*S=fI-W8b*A`Z0;VP`; z)WaeX4+U_Qwz?mbQ;!oQC#x41q81)$GGG*zv35gMoAc?4`sFcrcYd|}R8`({i%{zk7(RxhCCIgtw`>w+K(bh&})@KY< z1>Ki|f`Udk?j$P9S}6Q9O2mpG%OUf!()tyX?17+{FjU&q@+{OEx#X)93&b}qhPRA| z$j3#vv@&hOyl7S;D=%uct;9wq9(;8!QxKABtGds36=_ z8ef{~@7#_5ID50QvaUc>5S*^PH14UiZPs)-@C^(sUey)@ZeEdj>$8?peh?3HB++P~ zZQx6^x_!*{@j;~~9~KIbB~I1rj}rCgP-mRx)yE&A!0wof5yT_d&Zf7stZHB@bbT^Y z4#74Z^qkpLNm{{R5J+?m$VoT77k(2vAPH~tMJV+dh%i=LUs~zJywGmVUhj{l zC4Y+K%md(g0AJQ9v_M`Wv%(RcJ1Yr8Mb%#-)EGBeA~~L^i7T@+dy0&HUr;PFrG;)9 zv${H}q@$Y`)GE)~w_-89!Yim1*NH^;R?g?rjqn@QiTO->RkfG!?$ZR!?2+V%u?yb~ z78gW}00Q)-@wetgV4Q>D{&*C?5#9zThdN6heKeLw!fo`(4w6W>3Ot<3f__+BtY5S| zJdBkE(vCG9z(I_4g8-z@6Ls?2=rs)CKIq6f)pWi&>SSqjtdx}|Jixx6 z-I=3zU_Zjk;;Z7G5f#{jx)@3gw!#?Y43|$*QgI7UU{GEudH1J;6lq?Dg1-0mH^e`m zkFW*D@yi4tXlr_8o}J#nh+S1?<~;&rg7}#;&+-XTxht;_L65pnj}_2dS}B3^K9GC?+ragH@^>ELaGN+lF@3-A32%qMqI2z*!(%dT%eQ~S zjQsK@!44mtqiO3hqd# zfRo&cnlcn7*F@hEK=`XXOMZxMl;`h>ww*}T_*#-l`e!eOtW;@%hvG|dt^c?s%oZI+ zF(E?GCKKD16!De1pf^<;rg~uiXr_e&LCh4?P@(ywJIEXc4_5T`h2?#hzVnAejgRdf zi|7Y2w+fQkX$v*RTl7;F0&Y^+mpAyguT^|22(LlRY%D9Q!9{XnY!{Ya2F{M54Y5Fp zsnM~8J>#OlwJcLIuSNxZ^SDbu-(Z`hP6A3soj(%%_*?%7&7Lo1QsrgPEh*%m@g-Sg zp=(%o{W8vEzIni=YwXROLTtuD+!i`2h@Lf^3pz2k+>_=rzT!*S6im_JyW&b&W7V$P zB%G$pPf^}AO@b2ZMr?Kkf5&NcJ?Ad^C*;7$+m=AzW1Dr(hr9nN`vUSd;)GKS28Z`0 z6-%#+=UXJk*GL?rE{4=(RX5b^?T};OH2r(FFhhOg4`D2Sm(@yQvwfMO1=;ijV>j$? zX@MZG&FVR*VxDM$YrX=}9YtA2CsQTiPI8}4j~Rsr@zOpmAX;I2d1+0>mn(4h=x*E{ zQp95f1DPCIfNB^q-@izJN${JishFL}J8T(Xz59g;?YN|!qA(^*lkR8u!$LF4X(#Ax zz^Hr58`^RW)V6-5vsN_PYJzpu^OpVk5l;7vbbE#mq*BKeFg>6O_<`;<-p+Ac=v{J` zOV^5&c|#3KJZc_KaY_d7sasx87c1CRpMq2H)jJ6)8Vhi z;rmv_n{MZz%O^afRLc^FC$WUhq&RMB@;yaJZEoBg!dddJMmIpI z*WCT&!cymCC?-HR2-(-ub^N8p>u(+#wk?Wcj;VPWjfkYrqW{tUR3E!* zy!~5dg<#KkxbD*ZSXbXEiWRU5cCA40hM`m9aEbc1kjdjA4{Wv6>kH1v2uZcfG~fO5 z-t~WLY#Z8;3RF!ZRYLvr>2)Z_&4298WvS!!4R%?wkrw;EunH?Ta`p45bh<=+*Hu#Y6c#$)CVkU+@XFm0wG5=on(cK9F z{EM;1NE07#PhxyyGJ?{*;|hm)Vln=Wv$H>xl$3n?_| zIZ+leoc2sf=E&qSBcYkaM(b$m>s@wFE+(EA&tP5aI$>!BEjUn9zUv9cGh5QP4CKeN zF6d7^qK#KHDerUIGf*>_1dOkyTYTPGODMj7P`6bC*GJ;N+;l8k11pp?; z=j8lsSe_$1(;@rWFwfq zC1LuN6HO4OLIMBpP|_lXUq8*w0}2g)b)Ro`aACO}QaR8f3j6nY5*Ljm!nBTKkkMp< zXBlP@XFcqfdB=Ila_o@1*U*%S_9oEyDKaHpm@)KvV%gxB3#An*2?xHs=Vg~nKtC1= z+nYAE?MAFAgV@q7tH|x9V^E&v&9(luf ze(`v_Xquu`WY+5ihr#E4KWBz*BwLH}_I%#CkHTa!HdMxb1f(BbP?E91yDhfkE_Qlx z46vmp8bOp&Wi~0wX)=OZHhFry%`I9V8Brs=DLF_QRL(S16?NXUXI{dBatzZAH>KEG z&Fl*6O&)9ROb2IC6Hxiy% zM_a~-6NssZQ;9QhQVvY-o^SR9hv_qv zxptnK{`K&MN5pJcg*6dCY8-OGCX&BbPiP3NIY2p z_L>Sj&uEs6)k>DrM2$oC!!T`oyEhY-C7!2nwoKAkNIv;;RNw465@p(9AJ;IZhF~G> zQsHO%4O1G46`EXt7oR!4dFIVy@fE^wO~`s;-&*zi15L2qv}mw%G|zi_V}jl41q*V` z)Xqe@m3!faE8fAa0EM~Y+vd)^nFqk>@ou}bEI$c_K$m9539G8(8WgS(%WbhmaJ@nc z%SzH9qGG?q;ryHfJ;CXOITeD(Ze#aRV%Bd6UUD357_b2O$V zxplBbY3i2y<){BfH34EDx%D)yi^;k4IRiQ=+sIwNa7egkI%Gm3)M7KmbYQgUQuGzFEW$hn8s&%%{2tEN;eRr5(+STxsb^pZ0TAcu zh1#HZ58rsS2RZuh=-co*hyjg4z%oboYJPt25DYwdct3)%KB}2Mv(1t5VSCl>`q}e< zK_1EXnNoH1J8wh?!ZW={2qocJ)nyMA9DFHu+6!pxOnfLKKXRE{42i4J+XIr%UqrUq zzN&kR&@TNo{hjy-5c%rnfvHxA(Q7qWi$)1e)E zI?&TYQ@{I&2^m~&#(B=c#w)7`ZpmO>!E0t=N5g3C9g|>S24*<@q9FaltNs1(3eUw? zal=)`j;Vq0@@or;k^=^3DbL$7XMcnYYAxtB;?xn(AcYJ4F^-Fs{6j?oW!+z}fX-Dw za#-NyMtP3&lvdn+YU~*CUcRPmm<7A_bsF{i*+L82b@J`LiA7hZ%poPC_B~s_E4c62 ztGDpGfO9Gt5i!EKn>ka&iqVBbK_$Lt;HGR=)@Z@+sqdaMs6)QND!*H^y!@(z1>OTn z^mE9s$5Kaw77h=|cuLdeV=g06iGeqa=^3UtVi% zb`{sX_S|$bWCvWo!e$(*;Z^2!*PenG1@+qR?o2rhfsb?B{AFK8@pNu{%jiMw?=>7H zG(UA8dw~*boyMf!lk$9Ceb={8S2mI!`ESGV$*^9R3N~b?~ zgPDGk4z12)DyPZvI|8i5eAt9rw|$BArb22*V4e1p(THrZ3A0rGyG`>^%}(7d%0-XF z?un_jO$TVMr_<@TFwA*-aBW|$v7hFWx8SD+?isf@&p&uqzu8ghDIEcp@gnzVId*+A zH~ksUmw4$9kk!mlpSG56@y1ayqct1JXJlw3_Mlr09s zpMh;?irA46)M^FFavM+wzhU`EpX}t_x4Vnc*}2I`{q@z_*}+TKc($({c06_dNiRVFtEo1w4n6s zxyWQKL|&KGq~cjK*I*kl^|Pc)IK zr2#2_?p)HfUGyBBj&0m+*n#B2f(P2zdH4`TDcRxyFkutlfP0Tn1d#5w*a;6&%unbG zpKEtINlcryxXZ2h35}n+#{^>nTQV-waMJvo%^|tx=-W@ErD&ojB0BPLg%lzt+d~5{ z#@fw$27UL+d(hF2YFrnZT7xjAEt3Ul0P}zSwQPCUsJ1N>*#=cOQxo;5wuEY z^N^Vu9IdRUq$S=osE_3UUEca~2aaa+3L07J(7)h#?f0|XJzsf~r+S-Xn2~kl@agv) z3JAffg(-fh>qv36^nqE;kOQ{>PuBrwZl$~YGZ%#CJYbtKf-;s^TMyQ)X{2a*3-k5V z72T~rMUWkNUJ-d1^6`vrZR3ypozEzl)N-R#Uw&Q`@6)mps&*guyLE%N5@atp3WnNm zcdJP5t?-qb#eE-Xy_TbW7?J!rDpAY~o$8pAWk=(I;Wp6MX!LhW@H$F@}@O&)>q@y(CDp{cqPiV)4JBW$i6 zqW9;3D<_ymObmeVELT8I$*Vmw4%*_&=*+vj)29H@{G)0bK-GMG$h_pf+1phAyGwm1 zIrtA(K(MvR8#va!90HFtx!(Is^<=vJp6uDtH?Nyf{wX8ES?S8IHqPBiJS1hi2>%YP zec`di&=}9@c#-+q)MQXyUr!#GjH-_QHZ#zIx1=wkk5KklSrVZ)zl3{myaK3>c9Gm- zL?JTL!z-=2U&FkioEi^@z>TdPwZ&NG^IJistV55MiNRsYDUFyrtI;vsv2wyI8!p4< zD>hkp?i@nNq8J2q@Ij5C>~Q0XkQ3+@eyH$9DW$ax-EQ`K9T0F!yx{fH;;^%#DB(AJ zaKrr%?=`6p-U?i#Q_ljeD9F;Q2tTa|G*7Q!qBEBuU!-@J;xV%aA2Q)Su1FfhN&Huu|* z4*ZAp4G1~g+zq%yJI<6y+-4#-!tp-XXZR% z^iT65btIj($%itbMQuG>zzs&Ev}?bz;b?;Q$=Q&Kn(yK5*W|OvJIo9M9{d{XZ z6YjoGtMVS{(kOCgh|@b>qfhoy)B~F@*a1Dk``KOyLWkaVsd~C^w3npU1BuKXVr?fQ zr|KpQGMCfkW|&cEC7HALL4K<-WTn&*Wwtuv2V4f;-21pm*eSDDe4*{N-_Y7wTtoiu z!0CL=>he_b#`2`})#WtpPA|)a%~RLkmDZlr%XEg-zP<~g+QD(Q4jlI|Nd@T)Q1!^H z^eR|bF*_y4*rL(sUE)oe3(h+Nv|KUrRK*>WbC>j=nBfL|Y$1wez^)aZLH$$c)o7_Obw@iQ>LgBRP1{2F zfpAYX?!M9zsY{FuFUCT){+Eqr_rbS=-N5aV4pR2WJEXfTl7Oz$ZcZ3HS=jpJz@w@i zCw?a?>llkVFr62fQ!VLr&vAqE;#2dk;G)0~nbvrTyTRl^&Mv==C8YEXN51TOCKd*6 zigxBK5(VromqSpQegtCb#qt3faDUfP7@xV@6z$wy(eGW%KK6GSL-{PHUPpZBqIxim zuM3u(6t7nvy4;`Z_xO4x<=Y%tLBa4=4F9`mdj>fuo#}WlYiOCf#!lfnxme-K6(&7H zJQLzh2H2j~fC*xN!c+IA1L)?izXkTRw4XD~ir4)xA+FCs>giW3YiL>G(|@Y-a((;F ze9l2t@Zp|h%;TCFT9q@=1>|x?)9H|WVkF$~EhY%Fv24W;0eftdaoI_*+87V9M(;KdLFqxOnDGw~KyGdul@$LTHqTb1rY6sT0 z!mhoa6Pld=A+M~p0Wo`@YFwBV37!8cmp8eH(5W=OZ!e&enl7eZr}8r6YF@vR$vVoI zZ)-eMA8coNs8E2ZuFMII6^ z5}03VZS3N!C57@_MCl>s@ks$zQ-qDEd#hPo8=5``oyV5VRsy#l^_xKa+3O#i;N&;l z07FWC)QJ9(VsF03p6~|?-EgPso_X}g%vq)YINV{kPCe`-iJ+303TCj z^I4O$?bc`3Hx9l`uz#mj;UTtGwzOqmNOIp^bSIAcrWt)91Ips>A{&R5m)bL`DX4G5*Ac`y4^(Ir=Yy z9#Zi$a{&kp0-41=X2gkq&>&gs{wbFPXWznSI{oI5yLt4^-SES%3)#>9@t*RKXD&S> z!^heQZ3RF81a;tC+DNUrZPQ1_VD3gDL+!%^4jd)#dHWM zLr&6G$Jc=4+S8A$02^F>I>xWqP-?E1LKpdP4x*(8Iq&?2yd60NS~_O#%6jw`YDxuz zkYS5$8Er4_xDabgJNGntsJiHoRgy!u8=DU~Cx)vL6l+!5m*zB~_0{RtZ8Z_~w>7I= zt#voks-Z&#YMHQnOzNtuMrEs4RXN4ecMF?l~%55{d!{;MRknr8H7h4x29xGce4?V{K#}MQ9;gF8Sr4}nH0gWM?mcsgVU{svlkC@;{ z1Oh^O??{dr02X@mMzYdIjq-YwXm=<#{HeK(ObSNQ-uRbZi-X-qhZ0Ay{l}}O-0xpa z%@ol26Zo1lgagyx0{~fF^zswSBd{?e4d=Ky5^D4Q8M=_I@GKnSZ%1a1+j)rR`bn1+aFo$xvS`#h7Z4$UO{Be9Y}u4- zPU>Q6-G?3JY6lZ2rks;!?@g=HhK2fa(L#d!r}TZRNRQOh$I2}dF>aApg_67BL+V~v zX2d4aqZX?4G!{v<&IJ)7)Lh55{+kFw(Ykxit2Hbdy)4;QvWq^$Fk%dSbvlr>f>|H^ ztx2t;h0EQnJ?w_CewdjHk&PkvjCT66^-a!IxLb=AH+6F8#r6#>{pVlob3<9wI2m5i&CHIy~^F*0f{ zsjCWTXw@n~n2Huw%94mqCcmGyQ95>D3CJN`5~5*H6^M#_z>xSdl%GkgOo|)=YZFdP zsXlqoUrYQZHBv*6&MfKR)3wK;8ts2${BDi3kzcs(klNNsGu=n()yi#1nfXd?UPdec0A2tRZ1z+EWZTC>)+(!Yq8iSkpq4;IHo_L zAVzOID=^vS`>MTd|BukFmr}oY6*BtTD@`hZssVO?sdJX9R>KW;9X7xD=WY^kFkzI{ zu#VNL>_vQWkUgbL(e8J`8MfFN1xGfo+}+gPj*}-D=uN>*)!Bt0a$~>TPP>${a%X_% z2>|b%UIZQC)Z4==q8G#c)JQ#I`K9WGt27UB&iR9|7!ra@>a#JSRZDPQI z`G*Ag3aM_{rjE<11zhHzY7W|pl2Gdc56!W|f-_rP$q3*4yBfXRkuMWQb`GbV(x&(4 zyBHLK9}-CEaA&$d&eArgY}hGXyV6L~gU0fkw|#s`)l&0j?3#|pyGt@nu@sx8Z<~&r zl2UflNEDhj?AIo#8};*&B`0EQEc0uHb@k$3zwwC1i6G3(DAqQJ(H)GvKdbc}%QxTI zISS$$`FT`{eVS`uV`0O)BJ$;ooH6!sMnQ2kRlPwbxJ#9rCf{nne_iZ9I)f-V-aD_H zhqW>_$n1-cS+Bfz&V3K^7`Hj%MjrZEJ$epI^^@|r!+0BaU5=AC{HGCkczaK+FGg1{ zi3~ce(~Csq?0Dx%=&oo1=4=frEYG-y8Wm&$xnGPy|N1Y+pm4?aIsy&v9xXbJ>dy)f zD|TP?KV0`=E94x)-dDFqM=f^?hBu0&{p)J}f3|QY=U-8>7_N|a2tA&Au&So-0LpX| zji(>7EnxR7$%-gZDoWI-^rLrWpII$(Nm|jo?>JLO3h9do4Zu2N=`j*Te?{aqZJgpp z0i?Notec^~CzDoubmSxPpUZnAZ=#-^=lH$3r{DT7p9?t{nS2Sl)R`5Iz?Eunp@X}u zfu-@??@M1(XmJ?Q)4)F|)AmUb33&DbSR{N7Z7j!~K$Eo`l%wYfG4#SuCkouEl0XxJ zutBYa^h!c73j^h-1ZCR?RNvw*=>N1GM)X8^3xVTY;l-+Td%p`J*R)pl-$!7}^YIj< z8`%Tf*`^<@%Nw|By>Tug|JUc0emsP{Z?cRSN7%s>0cROKEnq>b=nU-;dcZa;NJe*0OKOnea+*XhaS+0N)tmtM*lkADuBi+oI6LHJ*FB3^c6g(T1 z;syFK_7D?KrYry@=v#8ZCHH%>lJ)J98_o4orbaeFu#I-T%kL1YtfZP^nB|ezl^4Ea z5Kt1l8J3($h=^?aq4RhQI>_Qb+gEc*M_uuBs6|DG?BjaAcunKx|5sz+R=ATh(U(k~ zV3(p?ndz6?;8qJIc%1hF?Axdz2VJ zdp!F-Ld|55Y%SwpGV5!hKyhh>JV}lp+3hH;R?G0h{FABwPU62EV}6{1=qzLDX+f#T z$%7b+=>^bNC`MuYiQ!q7i(%dPVXT_8f>iqG@#T$oUh=1WwxlYdl|Izx%dn{skY13_ zI-Qc+vL?%e1@9$==5?wBh1PX>-!;wD4=@$3aNPi4>c-#weVA45I-a+`ASu8(IW4;B zG$D!wop`pSV7RQL@epy6EyXMN(^*Ng-Ag=L1g%u~Ma_ll(Vs}`gm&J?%^nZd4ao** zIujy&2q+;R*pjdJi8o%Pk3Vv^dIL(J)zn%2-}?Ob(6SsUSp;>6pw3dx64P;yS7X&o zQ4zI5{8fYd%c_Rni$3Sv{jxuBWh{xevKh}`wY?XClMT{QY18z4q&&+{x9zJrV9A3R z#?|sLZGRNKo4`Y%6!Hx8-G@6Nn>7|{MztVy#Ji`w@2t1sZ7P(hASAS2ahy8qXLOHzTc z%nv)|#)b@0Y=dk;EzkEPAEY@W&z5PDV6t>svss>@B*mB|NML}gO_({jzY_LCC{XJA zee`3tQk1O^!9kJQc`-AJfoge_8MO~(906edF}5`cyd~Q`x!-*(Z|QW%(1-CFXI*2E zX?j&M@K*)@mNX~XD$Z~EMX|U}!~aN)s})-IyzzAhYx%Lx2W!}f2cG5UP#Okb+UU4^ee z7V5C%9s@1ZVDM(ghhVysWYh){D^Vm+tO42bYlbZ(-qj*&O2%_&nvx5Gt;GB#>hR|< z7E$A-B0TX~Xuup~_$3;4NIFm;&HOrHz3>VhT_w-0x?1`Fl_fa+g4`S{#r}(U-gNQQ z2Bxi;ZiAkr>mPjl07_`@q0Tfa*TB1>T#GAkon~!-Vnx|1ce~}QmBc3@V%D>zD0z9l zbft0&N=uBl*`8eMexJRXYap+xZQm1psIb*&=eI#6?-hVMN!UoyNci~luD0@I>c6@8k^y}hyp$|Dtsvo|cA)7!-q3*j8`xzL5`C%8 zC(bqodNVCMLFiQmh7B0C7+;Bg#nqP=h%+=X0&IV5Gyhz;%y6z1r6ip0E6=_7i5IZw z1TicvC2~5?zT+KU+A;NpI2;G!z;Lz0mq6`0P-!BtkwTkZ#%0mXweJ z^o|FkVfce5{~^u3DyL(8a&m=ATZD%q|; znH>(HTFm*_T>sIsb1gE3HR^TazXRMF!=3&=3qc@kw|Mek%TmOGcNkU`7u=}1#+Hiz zK*EI-aN&*(G4xEUQ0|v^o*_Xm>H(D|)dwSKu)+1ill}|NzeG7teCY!LDZ<%Pg)Dt0t+Bmm=?7XB0-S?I#x1g+IpZL+8VWr zl3P~o&VYP+J6T*+LmlO=VsXQe(P5SFtXC6<#4D3Z)fVxJ4E~B~$)?cJJ~;;bqc~`E zob|PGp+TM3n{%&Oopvu7&A;NC(=#@fjEPtXxxWtb3ApYHhfX4l+R&=B!|yO&{p@ue zY~T8gDE7fX0xE#nHbF;;+wlFkv!P^mx+tY(zW1$1&IOJJ8V!P9E#nb>~PI&spf{X^^`7SZW6xmI=_Ue0`GX=4_B)*>``CA;jxV-DeRr9amcbP6CH2e#-|C6MhFr6y)T2Z+l#h;x1b7spBDYRQiTMK zw8Y}VN6`d!G9Eyr0Ky(c-FK+nW{Eq!#!`o9WkT+pJXs_`TBo_89=DD^F#tu9>3}6N z5=@+J^y#aF$lSBl-b#&TJjom9I&e?_D>_EW;pQh4DrF6PTNKMk1zFD{hn)96Zgra# zbVs7T3i``8?{kDJFhN6z3#>VQgx<`CDitm{A*#kweQ-JR#0^uf54B5y$#9@pNCjg7 z7O))-+JXZWW-~@JF6Pp&(h61o4v$z03TxF>g|}<8F0G@FGaHO$*qXv(&yWRDSii0- zWTvy{9~}0B5J*m3%vN=%!dE`dR=ckQI*jFykMao_70i}s&%hMIiWf4@^HDqhVDLM5 zY?;{I&lmp!Ny8}1uVw|Y*d>OOSjVL3w(QLP$Kk~gEs4ti`&ddEtl}ggbV;=F_p)n+r%<#AluIOJ@ng6i^6rq)7F=_;Ynj z%}af#0vqGXJ#cdGaeU^oiw-^R{YpePDA;9bRS2HW$0fJ4TKJ4YXT*n$R_`C z|7(B6@e)JiG?CV`f*Dr)d8)L9RXl;79!%daU*+dVqn^D&QmhZhE{f*U3>ez>VJsnx zn0(FUM3nv+Zp-)AK(^oH&Vpq;SU&wQEMM~>9E2iC8Oyp1vsL9p*sH;xpSeV20fXxf zmKylg;6UL|wkXUn-)@BItXPJrcgz`bB33H9eBd7)BfR}}z%n0U->N(y#tgp*jfG;C z0Exaa1<@Z+p-RH>tP=4$bX?V{g$CC+iV&Ud7T;#TaC=%uU+^)l#18+%&j;Y|iL5+* zKVKY)e&6?1NLaiuqer^&N;=4@rr0WhMPjGrjkys6_wD1lMf_>x{kaFq#M9MfkPG2M ziKR~_|8Gb8ZxN?cE!!y5v>IHsq{Z1^>N`*|E}rb_74neO9=WelD84-h885|)p&V%t zFb36Vlc6!-k_NH1J)i)mW6P7jLtHI81y_7)Z*Jk)GHrCNEH;^J+49&8H2nZE0Kx_Q z#av6kWKE;VnHt_Chr6U_ha6Em?Ofioh`+J;`~u}?>Y;F+P$&mqp`@{=$4}1WOc>4` zjpS0Wv(#P5gIs0dg}9`_08|xU0e`x2z&}>4B^RzBK-eOfauR_)Y=2P|g_+a;BmA@k;30_|y&n9U_Fb5X1vVX=-2naVHh)+A{(cMhQ&!Y? KQ6_Kp;r{_>^a&pT literal 0 HcmV?d00001 diff --git a/en/2023.4/assets/freqtrade_poweredby.svg b/en/2023.4/assets/freqtrade_poweredby.svg new file mode 100644 index 000000000..957ec6401 --- /dev/null +++ b/en/2023.4/assets/freqtrade_poweredby.svg @@ -0,0 +1,44 @@ + + + + + + + + + image/svg+xml + + + + + + + + + + + + + + + + + + + + + + + + + + + + Freqtrade + + + + + poweredby + + diff --git a/en/2023.4/assets/frequi_url.png b/en/2023.4/assets/frequi_url.png new file mode 100644 index 0000000000000000000000000000000000000000..bd6ef52b62d3e0cd8013922e22bea795828fe802 GIT binary patch literal 11391 zcma)Cg;!Kxv>rkbke0>)r9nViLP5H_JES`tIuwxZ&LM~HZV(uRA*8#z8M^23d+YrH z@2*?xuJ!G+@4ow-`|Su*R+PfVB*z2*0N66p;;H}u(#FfZ=4<4ab7m1T&C3PDNm|Dh z0Kn<{??6go!Jz;Er~xwK-_$(}kC%P*)n?a)&y-xmT*M}5QGWsm2!Mg6-|B;HSVD>8 zrimAI=R1!NAdB_h#mg3l3yxZCzBgGJ%bkr66~~RB$ws{&4TQFqtc{qBjP+jwIJZ?q z@m^EM_602kEp6QlVB>9|DCPEhtyFd1W9}|xW-QTyy-!+p4L#+ws*s%3EaObpaIQh(fO;pNw;{hHk%SdEUl=hm@&SvMN442|H?ONm8REwi z`T|4bY#faE!@kge)5j&dY>x8)gk06a_SZa>?i3OSHYu3hcXpg{(e&PVZyFUwZxx>| zEIK}4I$_-K;R1mk4j#=azGJk)Z+|tTy-1Hof}XetJjxA(p?g$@dOARVp|7A)ZT@^D z8d>U=^o0p!Ba$~gg@C~JyRWxhD!T7O;WQk+y~bIJmR%!V+)hGkp4~q%Jl6s<9-MuR zN~FoRkG+-vF(LRSqhEsMZ>S-pVf!Tw*1ZBT0~deFsvfi=69zVG7R#Rt3IA+Io-PRj z9U1fQmXMcf9<}4|EqXScQ7Aey`Y7hb46JG!LQ6PGQj<~95huG!NZMU@{D@Y7PCGSW zcM&*?(&^Q?PYUhl+vja}_I6ep!+f-V1$^q@GGDW?uX_H%0SXi|_TJk(h~3Cl2=?K5 zEs0AEL)sKdr0>VwH`Z}`*Pdj9Very3Cjhjvvh_D)OflP(G9NqX#Rpt>6JB09W$IC7 z0LcL;bS5N(Shqxs@_dtz3LCpz;}kH=2MWNsCAMnyPTGA9V~th4-x_(&yLVoN*stnq z{59te@QAv9{?*>}T+Pb+ghbPg?7zbe`oN*5Lugj!)N0h$c>#>b5VQ=~R!_H!Nj& zRUST@aKKr>z$TkDoE7{C!W?|M#_i-$+A)&H3#3o=6(d`onz&EcZMxS&AAPY|`^B}V zzAAcwQdWxZE(F^aDKRk#aVR3A7J($~dAH|8n zHW3Q?;A2y=t zOCG$i-Od>&nInpVU2keXif_xrJE6cNL}fpr3V^9rEi^8BUcJUYG$ zag5s*a~;I4`@{Vs&**kbRo1{J+@nwIP)(AN{`}#6s580yfo~ky$a=g0weJBs=6ON< zT72z!e67+`DX+b@`(gR_u0I%OFssT~xdV)wlI8r4)nQLEQ{BsjQQv2xz!j#Io8Hq3 zSZ0a}=&0F03)fLAuB}ixws8%9TCNvM$>avA&t7;Mg0+p!0jQ<*s@DZ$wrFjG_BgV= zj_-_!nnb-0Q4k#^Q!oxolaB$^m0hUqjCgIP3UF~5HlA#uC+SwT;8WWXX32!NWrawcLYYN6vdTR0BV!s_e@QURi6E~CpZI#*>~NP zgbbRN*jmZUMt1qhKPZ2^-%n;xvuSj^Q;`63|JZJ2Xc7t>0Iw{UYS-`IG5jDg^k3f( zU&;APDkH}N_*`L7w2FWoTC!bA?+9IALZf2T)!Uu6bAmqcIC{cZ7P~G)QkaTdf=m(S_j>?C}&!B-DwHsr+ZIU8rz<*t@N@V#g4G6C&Y%?v$3hC|j2NwvthM?lMG7 zlKOUrh^5wC;bCQAawE2%eds~Axx1|mnF!o;ZaW0IdG-@IjW67;@URNr9BUg(?`Y^# zu<_w|YTM-<7=QvoW1^Uz52>EY(0TT;2{RsQJibBnK@tA@()y}Y{F?hl9B4}sROE-G zwZ}=b^ui5hjE9?MW|gfdbyqj8?8faf$7HM!4SbU;nt`kqeB9OQlcNJgMR`)gjeAi} zDqTEv_#armr5zxWy(81(5UWfJa^lP7aP=!E+TO;_UR= z;=k&XUO^-NBi|)|2L#Y-J8;$4u5*yi*hX&U^V9}_8h@NE#NMhjqIbvXl^pHVZHnd3 z*m|AZKq4&y=@H9kud}1Gt?Lzt+LGp;BmcaWOoV#4u^Z|@QGbrZ>A-HQP7Y!?H+(+q z6?aoR;~nVTb6r_^C5Z^aa{DSxqAuRtQkY@|LBN*}@4wW3W%2eLjj&1u{IcF9$0cy> z9uU;mJQvP#4c+x}kUyXbn#k+ZvJubXyc+FC=sQ(Lr!j@4k%dJN^==!M-r)*0dRf~< ziobqc_&$$5clCv}SxOxS_e|Q4!W2V;zKH@joM{#>E3quE+x5?f6A%x&*LUH|Z8t|> z#z7J{OSnYoUY3vvRo}}@m9)$fz2<@r_f`LUAE~#J;%Uw>XhpHD6s|qw> zX>H$oGUt@z0C=Iazd)FPEfdVk8SLws^}R#xw)R!DLYveA3OPkm9qq+$DLAN=wP!V6~8p~8rqlPWEj{FG|}Z{m#(Y(;wl?v(^{Zn z{`j@E7cyXIW$kOGRO7#FH6wiG3J3u`jdX2b3g1Zt>_SNe+i3puKFUdjzsBVPfgVz# z^z0-DkNTDHIKcKKb+P4Ym0=FA@&-I;|WK2R3 zEzUljGH6}@3t`$_0kf2O+s>{#5P#M)NO(BT`!yWhMODX7`g3@Z0*yL86cwi~FxcgJ zE`^U@8*rfG9N(D#wd``&)K}Dy_a+y;PJi1Q(K=vdhF^1=v5Iq}b6$#z)ZEJAek{yi zpR3PC-!06!0p;rCnifkS-rzY{eR^b)ajHfTsP(evGw^=GUYpx30xTJ=bxi(C0(xIk zU^&7*~xax6#rLelw=AOS6fc=H zYPTb4YAYVk1nkwRRm#4>-yVnH~kdMyrLpy{DTX+V|i#tCr8G$(2bjfn#QN#MR z$hJ)EiOWRDC=`)mZBO(Ih&+3nyY|1A8G}HeT2Ac`SQ$wfM5pi4(C!aorUl$uU?j9b zgXXpW{@zB`Yu_2uIZu+W=Ac_zO(3UCsd3zM&<2oUch9UM&PV<27OFv$zZ;q$w5LyN zczpM9ymIb0dBIC*9rji2EAzgCh%_VSuXN4NnL^NYMt*&-)APquq5CsI?^Q0Hk75kC z|B=I_o60$r8S%NB9U9Kpx%jcA7P&O7H0s zi=wdE?(_X37x=SM`_p2)0!33hoM*0FLCVOl!_+L=-*wi!%m?05R?foq{p8@(PB|oW z?Z{f7;U>6?@*UulGJIz=f9%VB#%#)GVT9Z`7sJmU$C~Y4a3OlK#5qbK-dcIsbAR75I1|=k!N38M7iG`2N7ty&;-zGKDhjj1Cf6oIuNx7W92_jD%LnHp zMiac>r@bJBX`bB!@^sE!b6Psuy?6q^%L*EkOjSh*t4IkAqekF;fJ>x$$fH!h{HZ4? zAfrEa{N#VXFz=`@{c0R&zDIrguP%1)Q|=j=SZ>(Kv2oj}(}ASD%PFKG`A2hb4x?UC zygqDWRSpIH{T33#b9PMB+|P5{#?Y^D&p*FgL-KD{97VrJ;2Hf^f#BgyqQ(pm%1E_* zb&YTKUh!T-654Dg*s<8~jv)AT`{%!~{PnLOQTSU$|E=>7^zPn~SM=YGdpF54Nx#%a zQbU0MMbWSOdw@;e-95Kj@;L5NLNVy1?O%I0*PnTq;>CKU`Om)K0wjEO-( z;k2q>c|2@GIVdOtcRGkutf&x^7ywYXmjT6CVf_u5V%OyM1c-{SOxT4PmrB3nz$duo zR>#&9VXg!Ehyd^l0nfTG0Kk_pK6_vUrTYR$HO7+kx@-ze32VFd;2U2X6vvHe^CA=i zf+=PZ%=Y-E>C_<|d_rHrV3XWg5imrO#lfl$+8Z*Uu*8I+R8sPJTSZrA=z zmo4!2Z4AK~Bj)QRVQLvXnz+bCS&5(t*MAMaCt$q^84vc1{;$@kedos{I1NsM5;?=9i5qU;$v zjl%~Z0O;a;d+WTbi{3U)TN=hh>!CuvX(YKMBm{Ia5i|8__YitB_DCJdB#!pQ!9fTp z1e{39eqB2r%8nVfp8?!|L(q3ASt&~U`d8|Vv^*i|H+frH2*xjJ2oJ#|>NmVe@-XXP z)S=cYyDDPUj6i;Y_07pE zI4&$PbS*vX_2M!xI(&0%@@M__0@50~+qpu@5;zHmd$w&F9xC@!{GjX8e=^eppxkCs z`(2iXHXXvk?DnI~y0oU0(XIEtZ)9<%r?alMDg!u&5lsJ$^8^8oKvicsX8157#b)=U zFj5o}`x%P^zS%lZBi$y_#k?L)7cd*{6K>K`@4>XY-gG8;o!9*qSX4^+XS75 ztMC1HG^I}!IIEbrND74ZdESUS|I6nJk`+byv_Ogo zmw7EN?@sE#e_1)aoeBmWMk;TEVbPFb#fs9(=i9$=G=+t(6BAb>BMyAtvrCt9r(I)h zQwt_~p^0t?NlCKO@3DCMG7% z&dvv^sh2u+1+v>fu}M_Z^nIHJpJv~B_V2No?8_s4FV2;t_Tr*sWs9oV+9*g=fI!NHGoK7!Cu7! z7Lw3Zw3m%EBrvadZ?AZ_$&TOs#7AF$=}3o3C9d?~fQ!pgl7Wsk@iilK9q%*cCON$M ztEXE4KsjBnC24z_B5u0DcXnQ6#{n!!rDNto8P$!l(Fi zjr(TCReG?sgiVXA@k*!JDq&;j014!6KO+x8P5AFYJ2v~`d3{LLMVMn8r-iM9WSnj( za@i@T-IKPkJ|E);dh^3S1&>bq4diL`lp{qhE9FQ4WgRRMM_j(jRU@@+WWe087#VBc zS}HO7ilOt2azFCh!kog^ztw-#TPjvJnU@%YNX8j0|$PtfD-3 zW0Sw8^J883CLk{_&!VUxRQyBCfa}gM*>Gt0JI`m<$XBd0!4L917Uk0`%U_e~teiC+ zFJ|J?bv28PGjlA2uC5C8^5?=YqVyE!NM)o=!a)zEpOyvDdFx%2;(Q&qYqd6u7k;;n zpiyk$+ny)yd=6SkerKP7xVqbEU&Nf>xuQCZ0%%*>$21WrI~aOqbk-nQU0dyNjsZl3p#l!AX_o7Sv>Aj@Op6Eo}nTlK*&|(O{+lfN|M0 z-LWh8TCTv>tWLVvpc^}_ERTy!H8R*V*QmBNJA>$S6uyr&lm zK5^W7MMLosNJTYe70vUeJ3~w|AXB$i{bGQ8WJBJ(^1gi(m@~Gm<>7>prbO()r)LM`iV_s>NBsW)pjP zIY*=Ie*{{U%O5&+8pGvvsm$4r^Co)xF_HQ|?Sw-cht|c2W$}SYg!$^@#^p2Nl7UJZ z#^ON(TLXVJxL>6#&rSPQw>X~JN*2+H?*@ev8At7$V1{rSSuCq8uV9?yYwgpD>UMKR zPR=>Z&Er>dgtKl>lc&*2RZRRv;llDH;5a4+fRVv5XVTs)f!9jHln3;)Q&W%6H(gym zcbp27F+%5mH~J&K0C5kmQvw#Kn1pO4ugBXe)3KGGdjc1ftXE1okN}0% zJQf|UtzOlq$&HM*a&Pq>SAHfntWf*FzaK8fuo)&~wLev)3y5ULHqt2;3qPpLpdVWR zJ|k={ewAv7J&5UBr<2x!Qwxe8As6@|ciGYL*YZ~?}Og@{4Rz5H5;$6fdae+fm> zJ@^WF&60hXF@7-u*QdD_iUyee(dL3ByZg{rFI~r@Rd&FiaWS(mTbY%x3vy#E`>VJZ z|L^t+3h*x~iWO zHu$`(Z7GKlGN8S5#NCn;8?tTN5Pa0KY~Y~${w_?`jSMukK&Fh@qRufDAd(8Csj}iZ zauOjT%c|0#7!k|wO=OnsCtUz>3tG1@8)=d_y^cZgN+f^EY$`k+|9Hw*$`94M&@4a|0%3*sBmM-%c;5nTLzK=xGh zaI_Z7k0XNEPV-|hKmDV4gBqu*mdMCU_C~PvC^l%Dp)`W=dRj69a2I1?H{EvIAjqi9AY*jaTT=q3cNK?X~mvl6x zEV(kGnK-|wi#PF0PWF!*7VJL6?PU|O+#g>PGALR#EFipYO>{q=oSvR?=?QUt$WcI! zi9upO{pM6xTnPy9biZ%%f0eMl1WoT;@3LQ(44`;lQCdPZWJO)Ts3Mbb*vPjiF}22% z9O+r^`QaQaPHy?g}&wM5kNM&U&W~%{` z)YMNccHTOG@a*=Shi4g}0ETgLt{w?1R>E6X&PNVHK>K;yJPH=lGX+wUB+N7FjUeQ` z3T-N&CI?tgSv4_D1T3toJ(v|z{b<+4edt|cJ5@B4UX{UQW?7e!qaGRrK!FwFqATg9 zx9Q56pNkgiPfhRU1Y=f1fwo1x*l$wXNLRUx?={7AswhM(U7Ogrb!m3BaI$#0DLQ2*HqNckEV=uFX3@CRPUHN>}uMLuEPXPhyhbzi4Y_v*Iebi=Z`1)ytP+lX2ZvA zHjhg0Cje{Q3qJR66ZIL;Q3nfS7;s6v=7|Wp-%tGwHXd&1UB~R!HE}kT=dh=`69>3m zw@0$>kMFzG{IPWP3IwP|p<11mogcs9+|i|QK|@Rg$r0taB!B7Fuc4FA>&Z=}_yyBl zJmf;A4)b8?MPf##Pp~^ zr|a}b!=W3TDyal#HvP2IpKC?dP>5P^*V+9JCH08Im5UGqDiixwedVl6h=}fu_bN~c3s55~nzj(a7cp3&7S8~HUN=L><#Nu8a%Fk)p}RYLlNsm&N`^uW=GPgkP& zn7K~3J~{LyBU2X}qfNia^wizdY@jpZi&^Kb(%0EOGcM)!X^e zY^m{eDd{DMhTfr%85*`eYJ{DKbU8)zZiYfC8Krzh*n}Yl++wxg^ReYwbUArQ&MPCq zx#qRfas%yxf6pPFa=?Z0{CLH&7UkTcSJnNobF=o&ft210AGiDHDj{bSyKb*2-0A}- zZZUaUIo|+9q;0e7J&VB4-r4Jb>5x`&RRR{mpO~b4qcf4^K4=E;2xJ0;SyeJBUolzC zD3xb$cj%8h|6;$@M~*rQ1L>5|ZBj~s-QTB7h-=>&zbyM#Z-BEnt>?D zjW@DxvMrKqV=^k+sn~)x56D!wNSx29U!OVH$a3WZJCW80_x6J(F*LX{-*~Z?(EQO_ zR*QANnfUilxoL4^7hb`;ShG2R==641Sf#tJeWgk^1-EiGgAh*>g(`?Y=@gjXSDF9O zwyhw$BX~c#;@L1*0m$^KBUEw8qI_uY%>@+b)MEUQFRtccl5O3!8_!-p#GpyE{<^AL zwu1IoY=Y@cOyz`27AonZ>Rwy)?o}uxUEyZ1*{K~RCK0@N%il0uo zzC4#E4OZQVHGwn_=ai?gVVI<7+AxUgC7n-HP28l5HpUx4gq+$Qbyq`S|4+J4pM)qt zpH|a}!hQ+OrmdXcd5F^d=8UONCA6xmLL`ERpxy}`DN$Rj4#HOC_>}kpDp&kK)Fb%# zAtj86%`yqaX+BE1^!vAJPop|XPE*9X(_E>b-Sj)W-~q=eW*&9$*xY_9sr_(tI`q)3 zxFyUhG9pV}i=41U%ut-$Vno93gIx;J+ME=la8Nee^RATRkIOGzcZ1WoFF|$X^Rc7< zLml|($lvcuSSp%wbg2C)*a&e-kn~I77AM*-;F?%l{@(!}Va#vxOYE_{*Fkb^=bb+ zo)eB)9U4U|S5iN#L&q{x61VI8x%v3seUJUhzNxbPH%5#$j3|nc)WVjx6z5i0FRD!c?yIlRB@bW}vVIWScG57< z|LQhVnN)HkF%+M~RZFP)?700nJ-dLDpa1rQfT*a{U)sTS{F_3$UD*7>PNNNRz~olI z2X){#q%|pVNy|ZWqulRM&g01MWUM5w2F+jmFS3^(LY1bkLCO?{_p1N*z*dMFwDN=) zkbJ_-l2jQblQ8ZQMypmdgKyUYB9t^Pv?zf_lg6D5F_ zc2zK42!AKKQI%_gFhR@Kk`_jbPavR47XOZ}@TYD2h^-?_v|MUslx#BUAYJXz+a#!Y zMQS`#%v;uYap}EvMDfp35DQ@`b9kIY=&uinD#12PlamJ90Bu(d+_ErLOBU;vP{n8! zOKf)pT^fdpwX94KqcmGw0ZVW_%z@le(lwG7(%iFVB)}+Dfvmh_oGLGA_R>dTKn~Cz zF*dO32smR*^>V5eTKN8W3Kq@}f9?O$8Z{J6xvqx%E!Mc3|LpYXVbve;WSQ-^QzU$E za!4rrb=4gs&PE(6Li^|FO=zFhnPm>a%if~o88uidRV7twAUfQVzeTIwzq_s?UlyE zj@@ac#bG-CU|V>x*4Pb)o3vjk?>o&&C%u!}bG0@CVj`lqJ~@@bbAzQ*@WTdC8k7-b z8r~;peCj3gade2*+i7x~>reIK5G;)ytHMI>)u)+hrL1|wFZu@Ri7IZVm`DQ99QU|s5Luf{DO zoas0$c<-Cb=Qcc&#P4mtNJ4*n>-8I1F@HqYym*zTC*K>eQO-Z81JM7t=|Voz(H|w2 z()321x4wix!*-{PlmwRA#v;(qzr;S@KM`O3=tRVtv*tqX9od%7s`d0JWg2xXnIPm9 zrTgY{l)-P)Q*~cRIfN{1oS$hB6Unw#S8j}gxh3ZBgr=SHF2ktTP(LS|rM*;@p=K9yApgP$vzy12AIg)sypwb$e9@-|K`Iw@3=7$R{+gPgn#NChW zUjdBucP4<$QYqzmCbj9fGmQ%4iT+UnCLLpBf}6pv{s;d~Z9)fQRH{_eocLcf>cNE> z%+Bd&;$cdqx5kAknZj$*)%TPFizk z$We?sGTn4A{$834Om*S=OSJyTy4M$}U95uIkNLw&zX{I+c&r zvAfM=Eq>938qnjC8exeefk8FC08X6@#dP2iUYUl z)`mkmS3p3@{?H`yQWJtLsURp$ zpHLuVi-pV6DUVo11bjDv-fKpD7(YX;bN@m=;S|w)_#jG+*R0Q0*MC0d-|2eLw>zQt z{ngDyz5%D)LKF!5KRtHFg(vEV&(*_JaX@KaQKS%P#1;V`pwIVQE<; znw!e!X!sHe`TF*<`?Vc&3zN3ohzd#EjK6;|116@jJPN7P#zlV5^nlh z2=3;mkm4oR(*SymKVhrsKdE1D?7R(41Bx*sO~HXb+Fv!_MGc)JR9QMau)O=X#!i9;T z37kk-upj^(fsR36MHs_+1RCI)NNu9}lD0S{B^g8PN?Ww(5|~L#Ng*g{WsqleV}|#l zz8@ri&cTzw_h33bHI+12+kK6WN$h#n5cD8OQt`5kw6p~9H3()bUQ8OS4Q4HTQ=1Ol z_JAocz`fLbT2^{`8n~UAo=#AUOf=SOq4pYkt;XbC&f#7lb$*7=$na!mWCQ`dBQsO0 zLFBSPj*N?#u5&pf2t4XjEGH|=pPQ8xh7tpx;US5Cx_Ju;!O`ya-yF`)b%TEt5>eP1ZX~}sjjA%FJF?h7cX8=b!DZl<6%Cv z*G0uvvU+vmnpLZ2paivG-(cd*y3$hCIcsZcYOGh{$&)A6*XX&kXZd3G8m)G$Zz-LV z^GF3VAW^Mdv!)4OM8EgqRiz~*Cji;uzl2uC9^=8I84vNp;ltJ|q-*uQwGp2ma6cY7 z;`%`!9UXO@fr&Ebapfs34OmS9^u6$)bJxrucutf>`dKPKT%%*d3XlFVKunp9 zasduxjrjs>f8V=D|J=XNZp;_Zy^WgQ$9WDjgY=z@stwiEBm9u5*|34&1Na8BMjjgf3+SHcr`5~>oz1Y?SW^=K z^bTyO6>Gar#P_W2gEMwq)ot3; zREHn~U&Dp0l6YT0&k-wLwYjb?5zGK`W6S2v+K>AM(95m2C20L|3m~rN8dprPr@t)5lsk9Hu*W z?pS990s;Ez=+Rj{x7p``4>+c0G5^pYnB1^!TL=(?HLHZ+HicG{~4F1d^5Awl_2!1jICM-!9eoLhbbT^;yHcefyTAaqRcY zmuctDopPT!%k+}x%lZRKnzykr2}}XfG_ne?nRQO~?%hkzo;@RN{P6o`&mMUWBYMTe z6i8ChtjX&gXl`nvrU>jah)2iNM%JdjqoaeaU%yVn!^70x-flljp6Q5tK}5}&X8&&G zX3fpb3E(!rH=zVI_9Gjl45w@{(ITqngWFe7@9{mX;tO25Z_8 zQHEpI+FkTU#4xu>RkN>b3Tnc3UpWzPXWm#o55GKF09j^Mh~)K7{QqbO_~(@CVq! zS<8954|P8mXN2MRs86xZ&Q4EfM@JB94b=(YGuk)s&^jiSF=t3*oNK3`rD{H`yQ?d; ztE=laAUoZx5?RC8*WKOj`%LXEkgDd>&^Q4M^z`%u0rg-It=hLCVsq!Z%^6eB-OvOT zFZ28TN&cRmgU}Elrnk43)!>Z1FCPL2K$7}gwzIc48NX}#!A1BpJP?#v5wkNprhV** z?Cpalt1oH&{r!o3eSKc&ap)iz2BTn_VV`4>9M^b3;(YY}4>#ML6{~(4mH+?%07*qo IM6N<$f(jP3KmY&$ literal 0 HcmV?d00001 diff --git a/en/2023.4/assets/javascripts/bundle.51198bba.min.js b/en/2023.4/assets/javascripts/bundle.51198bba.min.js new file mode 100644 index 000000000..31bd0414c --- /dev/null +++ b/en/2023.4/assets/javascripts/bundle.51198bba.min.js @@ -0,0 +1,29 @@ +"use strict";(()=>{var Ri=Object.create;var gr=Object.defineProperty;var ki=Object.getOwnPropertyDescriptor;var Hi=Object.getOwnPropertyNames,Ht=Object.getOwnPropertySymbols,Pi=Object.getPrototypeOf,yr=Object.prototype.hasOwnProperty,on=Object.prototype.propertyIsEnumerable;var nn=(e,t,r)=>t in e?gr(e,t,{enumerable:!0,configurable:!0,writable:!0,value:r}):e[t]=r,P=(e,t)=>{for(var r in t||(t={}))yr.call(t,r)&&nn(e,r,t[r]);if(Ht)for(var r of Ht(t))on.call(t,r)&&nn(e,r,t[r]);return e};var an=(e,t)=>{var r={};for(var n in e)yr.call(e,n)&&t.indexOf(n)<0&&(r[n]=e[n]);if(e!=null&&Ht)for(var n of Ht(e))t.indexOf(n)<0&&on.call(e,n)&&(r[n]=e[n]);return r};var Pt=(e,t)=>()=>(t||e((t={exports:{}}).exports,t),t.exports);var $i=(e,t,r,n)=>{if(t&&typeof t=="object"||typeof t=="function")for(let o of Hi(t))!yr.call(e,o)&&o!==r&&gr(e,o,{get:()=>t[o],enumerable:!(n=ki(t,o))||n.enumerable});return e};var yt=(e,t,r)=>(r=e!=null?Ri(Pi(e)):{},$i(t||!e||!e.__esModule?gr(r,"default",{value:e,enumerable:!0}):r,e));var cn=Pt((xr,sn)=>{(function(e,t){typeof xr=="object"&&typeof sn!="undefined"?t():typeof define=="function"&&define.amd?define(t):t()})(xr,function(){"use strict";function e(r){var n=!0,o=!1,i=null,s={text:!0,search:!0,url:!0,tel:!0,email:!0,password:!0,number:!0,date:!0,month:!0,week:!0,time:!0,datetime:!0,"datetime-local":!0};function a(T){return!!(T&&T!==document&&T.nodeName!=="HTML"&&T.nodeName!=="BODY"&&"classList"in T&&"contains"in T.classList)}function c(T){var Qe=T.type,De=T.tagName;return!!(De==="INPUT"&&s[Qe]&&!T.readOnly||De==="TEXTAREA"&&!T.readOnly||T.isContentEditable)}function f(T){T.classList.contains("focus-visible")||(T.classList.add("focus-visible"),T.setAttribute("data-focus-visible-added",""))}function u(T){T.hasAttribute("data-focus-visible-added")&&(T.classList.remove("focus-visible"),T.removeAttribute("data-focus-visible-added"))}function p(T){T.metaKey||T.altKey||T.ctrlKey||(a(r.activeElement)&&f(r.activeElement),n=!0)}function m(T){n=!1}function d(T){a(T.target)&&(n||c(T.target))&&f(T.target)}function h(T){a(T.target)&&(T.target.classList.contains("focus-visible")||T.target.hasAttribute("data-focus-visible-added"))&&(o=!0,window.clearTimeout(i),i=window.setTimeout(function(){o=!1},100),u(T.target))}function v(T){document.visibilityState==="hidden"&&(o&&(n=!0),G())}function G(){document.addEventListener("mousemove",N),document.addEventListener("mousedown",N),document.addEventListener("mouseup",N),document.addEventListener("pointermove",N),document.addEventListener("pointerdown",N),document.addEventListener("pointerup",N),document.addEventListener("touchmove",N),document.addEventListener("touchstart",N),document.addEventListener("touchend",N)}function oe(){document.removeEventListener("mousemove",N),document.removeEventListener("mousedown",N),document.removeEventListener("mouseup",N),document.removeEventListener("pointermove",N),document.removeEventListener("pointerdown",N),document.removeEventListener("pointerup",N),document.removeEventListener("touchmove",N),document.removeEventListener("touchstart",N),document.removeEventListener("touchend",N)}function N(T){T.target.nodeName&&T.target.nodeName.toLowerCase()==="html"||(n=!1,oe())}document.addEventListener("keydown",p,!0),document.addEventListener("mousedown",m,!0),document.addEventListener("pointerdown",m,!0),document.addEventListener("touchstart",m,!0),document.addEventListener("visibilitychange",v,!0),G(),r.addEventListener("focus",d,!0),r.addEventListener("blur",h,!0),r.nodeType===Node.DOCUMENT_FRAGMENT_NODE&&r.host?r.host.setAttribute("data-js-focus-visible",""):r.nodeType===Node.DOCUMENT_NODE&&(document.documentElement.classList.add("js-focus-visible"),document.documentElement.setAttribute("data-js-focus-visible",""))}if(typeof window!="undefined"&&typeof document!="undefined"){window.applyFocusVisiblePolyfill=e;var t;try{t=new CustomEvent("focus-visible-polyfill-ready")}catch(r){t=document.createEvent("CustomEvent"),t.initCustomEvent("focus-visible-polyfill-ready",!1,!1,{})}window.dispatchEvent(t)}typeof document!="undefined"&&e(document)})});var fn=Pt(Er=>{(function(e){var t=function(){try{return!!Symbol.iterator}catch(f){return!1}},r=t(),n=function(f){var u={next:function(){var p=f.shift();return{done:p===void 0,value:p}}};return r&&(u[Symbol.iterator]=function(){return u}),u},o=function(f){return encodeURIComponent(f).replace(/%20/g,"+")},i=function(f){return decodeURIComponent(String(f).replace(/\+/g," "))},s=function(){var f=function(p){Object.defineProperty(this,"_entries",{writable:!0,value:{}});var m=typeof p;if(m!=="undefined")if(m==="string")p!==""&&this._fromString(p);else if(p instanceof f){var d=this;p.forEach(function(oe,N){d.append(N,oe)})}else if(p!==null&&m==="object")if(Object.prototype.toString.call(p)==="[object Array]")for(var h=0;hd[0]?1:0}),f._entries&&(f._entries={});for(var p=0;p1?i(d[1]):"")}})})(typeof global!="undefined"?global:typeof window!="undefined"?window:typeof self!="undefined"?self:Er);(function(e){var t=function(){try{var o=new e.URL("b","http://a");return o.pathname="c d",o.href==="http://a/c%20d"&&o.searchParams}catch(i){return!1}},r=function(){var o=e.URL,i=function(c,f){typeof c!="string"&&(c=String(c)),f&&typeof f!="string"&&(f=String(f));var u=document,p;if(f&&(e.location===void 0||f!==e.location.href)){f=f.toLowerCase(),u=document.implementation.createHTMLDocument(""),p=u.createElement("base"),p.href=f,u.head.appendChild(p);try{if(p.href.indexOf(f)!==0)throw new Error(p.href)}catch(T){throw new Error("URL unable to set base "+f+" due to "+T)}}var m=u.createElement("a");m.href=c,p&&(u.body.appendChild(m),m.href=m.href);var d=u.createElement("input");if(d.type="url",d.value=c,m.protocol===":"||!/:/.test(m.href)||!d.checkValidity()&&!f)throw new TypeError("Invalid URL");Object.defineProperty(this,"_anchorElement",{value:m});var h=new e.URLSearchParams(this.search),v=!0,G=!0,oe=this;["append","delete","set"].forEach(function(T){var Qe=h[T];h[T]=function(){Qe.apply(h,arguments),v&&(G=!1,oe.search=h.toString(),G=!0)}}),Object.defineProperty(this,"searchParams",{value:h,enumerable:!0});var N=void 0;Object.defineProperty(this,"_updateSearchParams",{enumerable:!1,configurable:!1,writable:!1,value:function(){this.search!==N&&(N=this.search,G&&(v=!1,this.searchParams._fromString(this.search),v=!0))}})},s=i.prototype,a=function(c){Object.defineProperty(s,c,{get:function(){return this._anchorElement[c]},set:function(f){this._anchorElement[c]=f},enumerable:!0})};["hash","host","hostname","port","protocol"].forEach(function(c){a(c)}),Object.defineProperty(s,"search",{get:function(){return this._anchorElement.search},set:function(c){this._anchorElement.search=c,this._updateSearchParams()},enumerable:!0}),Object.defineProperties(s,{toString:{get:function(){var c=this;return function(){return c.href}}},href:{get:function(){return this._anchorElement.href.replace(/\?$/,"")},set:function(c){this._anchorElement.href=c,this._updateSearchParams()},enumerable:!0},pathname:{get:function(){return this._anchorElement.pathname.replace(/(^\/?)/,"/")},set:function(c){this._anchorElement.pathname=c},enumerable:!0},origin:{get:function(){var c={"http:":80,"https:":443,"ftp:":21}[this._anchorElement.protocol],f=this._anchorElement.port!=c&&this._anchorElement.port!=="";return this._anchorElement.protocol+"//"+this._anchorElement.hostname+(f?":"+this._anchorElement.port:"")},enumerable:!0},password:{get:function(){return""},set:function(c){},enumerable:!0},username:{get:function(){return""},set:function(c){},enumerable:!0}}),i.createObjectURL=function(c){return o.createObjectURL.apply(o,arguments)},i.revokeObjectURL=function(c){return o.revokeObjectURL.apply(o,arguments)},e.URL=i};if(t()||r(),e.location!==void 0&&!("origin"in e.location)){var n=function(){return e.location.protocol+"//"+e.location.hostname+(e.location.port?":"+e.location.port:"")};try{Object.defineProperty(e.location,"origin",{get:n,enumerable:!0})}catch(o){setInterval(function(){e.location.origin=n()},100)}}})(typeof global!="undefined"?global:typeof window!="undefined"?window:typeof self!="undefined"?self:Er)});var Kr=Pt((Mt,qr)=>{/*! + * clipboard.js v2.0.11 + * https://clipboardjs.com/ + * + * Licensed MIT © Zeno Rocha + */(function(t,r){typeof Mt=="object"&&typeof qr=="object"?qr.exports=r():typeof define=="function"&&define.amd?define([],r):typeof Mt=="object"?Mt.ClipboardJS=r():t.ClipboardJS=r()})(Mt,function(){return function(){var e={686:function(n,o,i){"use strict";i.d(o,{default:function(){return Ci}});var s=i(279),a=i.n(s),c=i(370),f=i.n(c),u=i(817),p=i.n(u);function m(j){try{return document.execCommand(j)}catch(O){return!1}}var d=function(O){var E=p()(O);return m("cut"),E},h=d;function v(j){var O=document.documentElement.getAttribute("dir")==="rtl",E=document.createElement("textarea");E.style.fontSize="12pt",E.style.border="0",E.style.padding="0",E.style.margin="0",E.style.position="absolute",E.style[O?"right":"left"]="-9999px";var H=window.pageYOffset||document.documentElement.scrollTop;return E.style.top="".concat(H,"px"),E.setAttribute("readonly",""),E.value=j,E}var G=function(O,E){var H=v(O);E.container.appendChild(H);var I=p()(H);return m("copy"),H.remove(),I},oe=function(O){var E=arguments.length>1&&arguments[1]!==void 0?arguments[1]:{container:document.body},H="";return typeof O=="string"?H=G(O,E):O instanceof HTMLInputElement&&!["text","search","url","tel","password"].includes(O==null?void 0:O.type)?H=G(O.value,E):(H=p()(O),m("copy")),H},N=oe;function T(j){return typeof Symbol=="function"&&typeof Symbol.iterator=="symbol"?T=function(E){return typeof E}:T=function(E){return E&&typeof Symbol=="function"&&E.constructor===Symbol&&E!==Symbol.prototype?"symbol":typeof E},T(j)}var Qe=function(){var O=arguments.length>0&&arguments[0]!==void 0?arguments[0]:{},E=O.action,H=E===void 0?"copy":E,I=O.container,q=O.target,Me=O.text;if(H!=="copy"&&H!=="cut")throw new Error('Invalid "action" value, use either "copy" or "cut"');if(q!==void 0)if(q&&T(q)==="object"&&q.nodeType===1){if(H==="copy"&&q.hasAttribute("disabled"))throw new Error('Invalid "target" attribute. Please use "readonly" instead of "disabled" attribute');if(H==="cut"&&(q.hasAttribute("readonly")||q.hasAttribute("disabled")))throw new Error(`Invalid "target" attribute. You can't cut text from elements with "readonly" or "disabled" attributes`)}else throw new Error('Invalid "target" value, use a valid Element');if(Me)return N(Me,{container:I});if(q)return H==="cut"?h(q):N(q,{container:I})},De=Qe;function $e(j){return typeof Symbol=="function"&&typeof Symbol.iterator=="symbol"?$e=function(E){return typeof E}:$e=function(E){return E&&typeof Symbol=="function"&&E.constructor===Symbol&&E!==Symbol.prototype?"symbol":typeof E},$e(j)}function wi(j,O){if(!(j instanceof O))throw new TypeError("Cannot call a class as a function")}function rn(j,O){for(var E=0;E0&&arguments[0]!==void 0?arguments[0]:{};this.action=typeof I.action=="function"?I.action:this.defaultAction,this.target=typeof I.target=="function"?I.target:this.defaultTarget,this.text=typeof I.text=="function"?I.text:this.defaultText,this.container=$e(I.container)==="object"?I.container:document.body}},{key:"listenClick",value:function(I){var q=this;this.listener=f()(I,"click",function(Me){return q.onClick(Me)})}},{key:"onClick",value:function(I){var q=I.delegateTarget||I.currentTarget,Me=this.action(q)||"copy",kt=De({action:Me,container:this.container,target:this.target(q),text:this.text(q)});this.emit(kt?"success":"error",{action:Me,text:kt,trigger:q,clearSelection:function(){q&&q.focus(),window.getSelection().removeAllRanges()}})}},{key:"defaultAction",value:function(I){return vr("action",I)}},{key:"defaultTarget",value:function(I){var q=vr("target",I);if(q)return document.querySelector(q)}},{key:"defaultText",value:function(I){return vr("text",I)}},{key:"destroy",value:function(){this.listener.destroy()}}],[{key:"copy",value:function(I){var q=arguments.length>1&&arguments[1]!==void 0?arguments[1]:{container:document.body};return N(I,q)}},{key:"cut",value:function(I){return h(I)}},{key:"isSupported",value:function(){var I=arguments.length>0&&arguments[0]!==void 0?arguments[0]:["copy","cut"],q=typeof I=="string"?[I]:I,Me=!!document.queryCommandSupported;return q.forEach(function(kt){Me=Me&&!!document.queryCommandSupported(kt)}),Me}}]),E}(a()),Ci=Ai},828:function(n){var o=9;if(typeof Element!="undefined"&&!Element.prototype.matches){var i=Element.prototype;i.matches=i.matchesSelector||i.mozMatchesSelector||i.msMatchesSelector||i.oMatchesSelector||i.webkitMatchesSelector}function s(a,c){for(;a&&a.nodeType!==o;){if(typeof a.matches=="function"&&a.matches(c))return a;a=a.parentNode}}n.exports=s},438:function(n,o,i){var s=i(828);function a(u,p,m,d,h){var v=f.apply(this,arguments);return u.addEventListener(m,v,h),{destroy:function(){u.removeEventListener(m,v,h)}}}function c(u,p,m,d,h){return typeof u.addEventListener=="function"?a.apply(null,arguments):typeof m=="function"?a.bind(null,document).apply(null,arguments):(typeof u=="string"&&(u=document.querySelectorAll(u)),Array.prototype.map.call(u,function(v){return a(v,p,m,d,h)}))}function f(u,p,m,d){return function(h){h.delegateTarget=s(h.target,p),h.delegateTarget&&d.call(u,h)}}n.exports=c},879:function(n,o){o.node=function(i){return i!==void 0&&i instanceof HTMLElement&&i.nodeType===1},o.nodeList=function(i){var s=Object.prototype.toString.call(i);return i!==void 0&&(s==="[object NodeList]"||s==="[object HTMLCollection]")&&"length"in i&&(i.length===0||o.node(i[0]))},o.string=function(i){return typeof i=="string"||i instanceof String},o.fn=function(i){var s=Object.prototype.toString.call(i);return s==="[object Function]"}},370:function(n,o,i){var s=i(879),a=i(438);function c(m,d,h){if(!m&&!d&&!h)throw new Error("Missing required arguments");if(!s.string(d))throw new TypeError("Second argument must be a String");if(!s.fn(h))throw new TypeError("Third argument must be a Function");if(s.node(m))return f(m,d,h);if(s.nodeList(m))return u(m,d,h);if(s.string(m))return p(m,d,h);throw new TypeError("First argument must be a String, HTMLElement, HTMLCollection, or NodeList")}function f(m,d,h){return m.addEventListener(d,h),{destroy:function(){m.removeEventListener(d,h)}}}function u(m,d,h){return Array.prototype.forEach.call(m,function(v){v.addEventListener(d,h)}),{destroy:function(){Array.prototype.forEach.call(m,function(v){v.removeEventListener(d,h)})}}}function p(m,d,h){return a(document.body,m,d,h)}n.exports=c},817:function(n){function o(i){var s;if(i.nodeName==="SELECT")i.focus(),s=i.value;else if(i.nodeName==="INPUT"||i.nodeName==="TEXTAREA"){var a=i.hasAttribute("readonly");a||i.setAttribute("readonly",""),i.select(),i.setSelectionRange(0,i.value.length),a||i.removeAttribute("readonly"),s=i.value}else{i.hasAttribute("contenteditable")&&i.focus();var c=window.getSelection(),f=document.createRange();f.selectNodeContents(i),c.removeAllRanges(),c.addRange(f),s=c.toString()}return s}n.exports=o},279:function(n){function o(){}o.prototype={on:function(i,s,a){var c=this.e||(this.e={});return(c[i]||(c[i]=[])).push({fn:s,ctx:a}),this},once:function(i,s,a){var c=this;function f(){c.off(i,f),s.apply(a,arguments)}return f._=s,this.on(i,f,a)},emit:function(i){var s=[].slice.call(arguments,1),a=((this.e||(this.e={}))[i]||[]).slice(),c=0,f=a.length;for(c;c{"use strict";/*! + * escape-html + * Copyright(c) 2012-2013 TJ Holowaychuk + * Copyright(c) 2015 Andreas Lubbe + * Copyright(c) 2015 Tiancheng "Timothy" Gu + * MIT Licensed + */var ns=/["'&<>]/;Go.exports=os;function os(e){var t=""+e,r=ns.exec(t);if(!r)return t;var n,o="",i=0,s=0;for(i=r.index;i0&&i[i.length-1])&&(f[0]===6||f[0]===2)){r=0;continue}if(f[0]===3&&(!i||f[1]>i[0]&&f[1]=e.length&&(e=void 0),{value:e&&e[n++],done:!e}}};throw new TypeError(t?"Object is not iterable.":"Symbol.iterator is not defined.")}function W(e,t){var r=typeof Symbol=="function"&&e[Symbol.iterator];if(!r)return e;var n=r.call(e),o,i=[],s;try{for(;(t===void 0||t-- >0)&&!(o=n.next()).done;)i.push(o.value)}catch(a){s={error:a}}finally{try{o&&!o.done&&(r=n.return)&&r.call(n)}finally{if(s)throw s.error}}return i}function D(e,t,r){if(r||arguments.length===2)for(var n=0,o=t.length,i;n1||a(m,d)})})}function a(m,d){try{c(n[m](d))}catch(h){p(i[0][3],h)}}function c(m){m.value instanceof et?Promise.resolve(m.value.v).then(f,u):p(i[0][2],m)}function f(m){a("next",m)}function u(m){a("throw",m)}function p(m,d){m(d),i.shift(),i.length&&a(i[0][0],i[0][1])}}function ln(e){if(!Symbol.asyncIterator)throw new TypeError("Symbol.asyncIterator is not defined.");var t=e[Symbol.asyncIterator],r;return t?t.call(e):(e=typeof Ee=="function"?Ee(e):e[Symbol.iterator](),r={},n("next"),n("throw"),n("return"),r[Symbol.asyncIterator]=function(){return this},r);function n(i){r[i]=e[i]&&function(s){return new Promise(function(a,c){s=e[i](s),o(a,c,s.done,s.value)})}}function o(i,s,a,c){Promise.resolve(c).then(function(f){i({value:f,done:a})},s)}}function C(e){return typeof e=="function"}function at(e){var t=function(n){Error.call(n),n.stack=new Error().stack},r=e(t);return r.prototype=Object.create(Error.prototype),r.prototype.constructor=r,r}var It=at(function(e){return function(r){e(this),this.message=r?r.length+` errors occurred during unsubscription: +`+r.map(function(n,o){return o+1+") "+n.toString()}).join(` + `):"",this.name="UnsubscriptionError",this.errors=r}});function Ve(e,t){if(e){var r=e.indexOf(t);0<=r&&e.splice(r,1)}}var Ie=function(){function e(t){this.initialTeardown=t,this.closed=!1,this._parentage=null,this._finalizers=null}return e.prototype.unsubscribe=function(){var t,r,n,o,i;if(!this.closed){this.closed=!0;var s=this._parentage;if(s)if(this._parentage=null,Array.isArray(s))try{for(var a=Ee(s),c=a.next();!c.done;c=a.next()){var f=c.value;f.remove(this)}}catch(v){t={error:v}}finally{try{c&&!c.done&&(r=a.return)&&r.call(a)}finally{if(t)throw t.error}}else s.remove(this);var u=this.initialTeardown;if(C(u))try{u()}catch(v){i=v instanceof It?v.errors:[v]}var p=this._finalizers;if(p){this._finalizers=null;try{for(var m=Ee(p),d=m.next();!d.done;d=m.next()){var h=d.value;try{mn(h)}catch(v){i=i!=null?i:[],v instanceof It?i=D(D([],W(i)),W(v.errors)):i.push(v)}}}catch(v){n={error:v}}finally{try{d&&!d.done&&(o=m.return)&&o.call(m)}finally{if(n)throw n.error}}}if(i)throw new It(i)}},e.prototype.add=function(t){var r;if(t&&t!==this)if(this.closed)mn(t);else{if(t instanceof e){if(t.closed||t._hasParent(this))return;t._addParent(this)}(this._finalizers=(r=this._finalizers)!==null&&r!==void 0?r:[]).push(t)}},e.prototype._hasParent=function(t){var r=this._parentage;return r===t||Array.isArray(r)&&r.includes(t)},e.prototype._addParent=function(t){var r=this._parentage;this._parentage=Array.isArray(r)?(r.push(t),r):r?[r,t]:t},e.prototype._removeParent=function(t){var r=this._parentage;r===t?this._parentage=null:Array.isArray(r)&&Ve(r,t)},e.prototype.remove=function(t){var r=this._finalizers;r&&Ve(r,t),t instanceof e&&t._removeParent(this)},e.EMPTY=function(){var t=new e;return t.closed=!0,t}(),e}();var Sr=Ie.EMPTY;function jt(e){return e instanceof Ie||e&&"closed"in e&&C(e.remove)&&C(e.add)&&C(e.unsubscribe)}function mn(e){C(e)?e():e.unsubscribe()}var Le={onUnhandledError:null,onStoppedNotification:null,Promise:void 0,useDeprecatedSynchronousErrorHandling:!1,useDeprecatedNextContext:!1};var st={setTimeout:function(e,t){for(var r=[],n=2;n0},enumerable:!1,configurable:!0}),t.prototype._trySubscribe=function(r){return this._throwIfClosed(),e.prototype._trySubscribe.call(this,r)},t.prototype._subscribe=function(r){return this._throwIfClosed(),this._checkFinalizedStatuses(r),this._innerSubscribe(r)},t.prototype._innerSubscribe=function(r){var n=this,o=this,i=o.hasError,s=o.isStopped,a=o.observers;return i||s?Sr:(this.currentObservers=null,a.push(r),new Ie(function(){n.currentObservers=null,Ve(a,r)}))},t.prototype._checkFinalizedStatuses=function(r){var n=this,o=n.hasError,i=n.thrownError,s=n.isStopped;o?r.error(i):s&&r.complete()},t.prototype.asObservable=function(){var r=new F;return r.source=this,r},t.create=function(r,n){return new En(r,n)},t}(F);var En=function(e){ie(t,e);function t(r,n){var o=e.call(this)||this;return o.destination=r,o.source=n,o}return t.prototype.next=function(r){var n,o;(o=(n=this.destination)===null||n===void 0?void 0:n.next)===null||o===void 0||o.call(n,r)},t.prototype.error=function(r){var n,o;(o=(n=this.destination)===null||n===void 0?void 0:n.error)===null||o===void 0||o.call(n,r)},t.prototype.complete=function(){var r,n;(n=(r=this.destination)===null||r===void 0?void 0:r.complete)===null||n===void 0||n.call(r)},t.prototype._subscribe=function(r){var n,o;return(o=(n=this.source)===null||n===void 0?void 0:n.subscribe(r))!==null&&o!==void 0?o:Sr},t}(x);var Et={now:function(){return(Et.delegate||Date).now()},delegate:void 0};var wt=function(e){ie(t,e);function t(r,n,o){r===void 0&&(r=1/0),n===void 0&&(n=1/0),o===void 0&&(o=Et);var i=e.call(this)||this;return i._bufferSize=r,i._windowTime=n,i._timestampProvider=o,i._buffer=[],i._infiniteTimeWindow=!0,i._infiniteTimeWindow=n===1/0,i._bufferSize=Math.max(1,r),i._windowTime=Math.max(1,n),i}return t.prototype.next=function(r){var n=this,o=n.isStopped,i=n._buffer,s=n._infiniteTimeWindow,a=n._timestampProvider,c=n._windowTime;o||(i.push(r),!s&&i.push(a.now()+c)),this._trimBuffer(),e.prototype.next.call(this,r)},t.prototype._subscribe=function(r){this._throwIfClosed(),this._trimBuffer();for(var n=this._innerSubscribe(r),o=this,i=o._infiniteTimeWindow,s=o._buffer,a=s.slice(),c=0;c0?e.prototype.requestAsyncId.call(this,r,n,o):(r.actions.push(this),r._scheduled||(r._scheduled=ut.requestAnimationFrame(function(){return r.flush(void 0)})))},t.prototype.recycleAsyncId=function(r,n,o){var i;if(o===void 0&&(o=0),o!=null?o>0:this.delay>0)return e.prototype.recycleAsyncId.call(this,r,n,o);var s=r.actions;n!=null&&((i=s[s.length-1])===null||i===void 0?void 0:i.id)!==n&&(ut.cancelAnimationFrame(n),r._scheduled=void 0)},t}(Wt);var Tn=function(e){ie(t,e);function t(){return e!==null&&e.apply(this,arguments)||this}return t.prototype.flush=function(r){this._active=!0;var n=this._scheduled;this._scheduled=void 0;var o=this.actions,i;r=r||o.shift();do if(i=r.execute(r.state,r.delay))break;while((r=o[0])&&r.id===n&&o.shift());if(this._active=!1,i){for(;(r=o[0])&&r.id===n&&o.shift();)r.unsubscribe();throw i}},t}(Dt);var Te=new Tn(Sn);var _=new F(function(e){return e.complete()});function Vt(e){return e&&C(e.schedule)}function Cr(e){return e[e.length-1]}function Ye(e){return C(Cr(e))?e.pop():void 0}function Oe(e){return Vt(Cr(e))?e.pop():void 0}function zt(e,t){return typeof Cr(e)=="number"?e.pop():t}var pt=function(e){return e&&typeof e.length=="number"&&typeof e!="function"};function Nt(e){return C(e==null?void 0:e.then)}function qt(e){return C(e[ft])}function Kt(e){return Symbol.asyncIterator&&C(e==null?void 0:e[Symbol.asyncIterator])}function Qt(e){return new TypeError("You provided "+(e!==null&&typeof e=="object"?"an invalid object":"'"+e+"'")+" where a stream was expected. You can provide an Observable, Promise, ReadableStream, Array, AsyncIterable, or Iterable.")}function Ni(){return typeof Symbol!="function"||!Symbol.iterator?"@@iterator":Symbol.iterator}var Yt=Ni();function Gt(e){return C(e==null?void 0:e[Yt])}function Bt(e){return pn(this,arguments,function(){var r,n,o,i;return $t(this,function(s){switch(s.label){case 0:r=e.getReader(),s.label=1;case 1:s.trys.push([1,,9,10]),s.label=2;case 2:return[4,et(r.read())];case 3:return n=s.sent(),o=n.value,i=n.done,i?[4,et(void 0)]:[3,5];case 4:return[2,s.sent()];case 5:return[4,et(o)];case 6:return[4,s.sent()];case 7:return s.sent(),[3,2];case 8:return[3,10];case 9:return r.releaseLock(),[7];case 10:return[2]}})})}function Jt(e){return C(e==null?void 0:e.getReader)}function U(e){if(e instanceof F)return e;if(e!=null){if(qt(e))return qi(e);if(pt(e))return Ki(e);if(Nt(e))return Qi(e);if(Kt(e))return On(e);if(Gt(e))return Yi(e);if(Jt(e))return Gi(e)}throw Qt(e)}function qi(e){return new F(function(t){var r=e[ft]();if(C(r.subscribe))return r.subscribe(t);throw new TypeError("Provided object does not correctly implement Symbol.observable")})}function Ki(e){return new F(function(t){for(var r=0;r=2;return function(n){return n.pipe(e?A(function(o,i){return e(o,i,n)}):de,ge(1),r?He(t):Vn(function(){return new Zt}))}}function zn(){for(var e=[],t=0;t=2,!0))}function pe(e){e===void 0&&(e={});var t=e.connector,r=t===void 0?function(){return new x}:t,n=e.resetOnError,o=n===void 0?!0:n,i=e.resetOnComplete,s=i===void 0?!0:i,a=e.resetOnRefCountZero,c=a===void 0?!0:a;return function(f){var u,p,m,d=0,h=!1,v=!1,G=function(){p==null||p.unsubscribe(),p=void 0},oe=function(){G(),u=m=void 0,h=v=!1},N=function(){var T=u;oe(),T==null||T.unsubscribe()};return y(function(T,Qe){d++,!v&&!h&&G();var De=m=m!=null?m:r();Qe.add(function(){d--,d===0&&!v&&!h&&(p=$r(N,c))}),De.subscribe(Qe),!u&&d>0&&(u=new rt({next:function($e){return De.next($e)},error:function($e){v=!0,G(),p=$r(oe,o,$e),De.error($e)},complete:function(){h=!0,G(),p=$r(oe,s),De.complete()}}),U(T).subscribe(u))})(f)}}function $r(e,t){for(var r=[],n=2;ne.next(document)),e}function K(e,t=document){return Array.from(t.querySelectorAll(e))}function z(e,t=document){let r=ce(e,t);if(typeof r=="undefined")throw new ReferenceError(`Missing element: expected "${e}" to be present`);return r}function ce(e,t=document){return t.querySelector(e)||void 0}function _e(){return document.activeElement instanceof HTMLElement&&document.activeElement||void 0}function tr(e){return L(b(document.body,"focusin"),b(document.body,"focusout")).pipe(ke(1),l(()=>{let t=_e();return typeof t!="undefined"?e.contains(t):!1}),V(e===_e()),B())}function Xe(e){return{x:e.offsetLeft,y:e.offsetTop}}function Qn(e){return L(b(window,"load"),b(window,"resize")).pipe(Ce(0,Te),l(()=>Xe(e)),V(Xe(e)))}function rr(e){return{x:e.scrollLeft,y:e.scrollTop}}function dt(e){return L(b(e,"scroll"),b(window,"resize")).pipe(Ce(0,Te),l(()=>rr(e)),V(rr(e)))}var Gn=function(){if(typeof Map!="undefined")return Map;function e(t,r){var n=-1;return t.some(function(o,i){return o[0]===r?(n=i,!0):!1}),n}return function(){function t(){this.__entries__=[]}return Object.defineProperty(t.prototype,"size",{get:function(){return this.__entries__.length},enumerable:!0,configurable:!0}),t.prototype.get=function(r){var n=e(this.__entries__,r),o=this.__entries__[n];return o&&o[1]},t.prototype.set=function(r,n){var o=e(this.__entries__,r);~o?this.__entries__[o][1]=n:this.__entries__.push([r,n])},t.prototype.delete=function(r){var n=this.__entries__,o=e(n,r);~o&&n.splice(o,1)},t.prototype.has=function(r){return!!~e(this.__entries__,r)},t.prototype.clear=function(){this.__entries__.splice(0)},t.prototype.forEach=function(r,n){n===void 0&&(n=null);for(var o=0,i=this.__entries__;o0},e.prototype.connect_=function(){!Dr||this.connected_||(document.addEventListener("transitionend",this.onTransitionEnd_),window.addEventListener("resize",this.refresh),ga?(this.mutationsObserver_=new MutationObserver(this.refresh),this.mutationsObserver_.observe(document,{attributes:!0,childList:!0,characterData:!0,subtree:!0})):(document.addEventListener("DOMSubtreeModified",this.refresh),this.mutationEventsAdded_=!0),this.connected_=!0)},e.prototype.disconnect_=function(){!Dr||!this.connected_||(document.removeEventListener("transitionend",this.onTransitionEnd_),window.removeEventListener("resize",this.refresh),this.mutationsObserver_&&this.mutationsObserver_.disconnect(),this.mutationEventsAdded_&&document.removeEventListener("DOMSubtreeModified",this.refresh),this.mutationsObserver_=null,this.mutationEventsAdded_=!1,this.connected_=!1)},e.prototype.onTransitionEnd_=function(t){var r=t.propertyName,n=r===void 0?"":r,o=va.some(function(i){return!!~n.indexOf(i)});o&&this.refresh()},e.getInstance=function(){return this.instance_||(this.instance_=new e),this.instance_},e.instance_=null,e}(),Bn=function(e,t){for(var r=0,n=Object.keys(t);r0},e}(),Xn=typeof WeakMap!="undefined"?new WeakMap:new Gn,Zn=function(){function e(t){if(!(this instanceof e))throw new TypeError("Cannot call a class as a function.");if(!arguments.length)throw new TypeError("1 argument required, but only 0 present.");var r=ya.getInstance(),n=new Aa(t,r,this);Xn.set(this,n)}return e}();["observe","unobserve","disconnect"].forEach(function(e){Zn.prototype[e]=function(){var t;return(t=Xn.get(this))[e].apply(t,arguments)}});var Ca=function(){return typeof nr.ResizeObserver!="undefined"?nr.ResizeObserver:Zn}(),eo=Ca;var to=new x,Ra=$(()=>k(new eo(e=>{for(let t of e)to.next(t)}))).pipe(g(e=>L(ze,k(e)).pipe(R(()=>e.disconnect()))),J(1));function he(e){return{width:e.offsetWidth,height:e.offsetHeight}}function ye(e){return Ra.pipe(S(t=>t.observe(e)),g(t=>to.pipe(A(({target:r})=>r===e),R(()=>t.unobserve(e)),l(()=>he(e)))),V(he(e)))}function bt(e){return{width:e.scrollWidth,height:e.scrollHeight}}function ar(e){let t=e.parentElement;for(;t&&(e.scrollWidth<=t.scrollWidth&&e.scrollHeight<=t.scrollHeight);)t=(e=t).parentElement;return t?e:void 0}var ro=new x,ka=$(()=>k(new IntersectionObserver(e=>{for(let t of e)ro.next(t)},{threshold:0}))).pipe(g(e=>L(ze,k(e)).pipe(R(()=>e.disconnect()))),J(1));function sr(e){return ka.pipe(S(t=>t.observe(e)),g(t=>ro.pipe(A(({target:r})=>r===e),R(()=>t.unobserve(e)),l(({isIntersecting:r})=>r))))}function no(e,t=16){return dt(e).pipe(l(({y:r})=>{let n=he(e),o=bt(e);return r>=o.height-n.height-t}),B())}var cr={drawer:z("[data-md-toggle=drawer]"),search:z("[data-md-toggle=search]")};function oo(e){return cr[e].checked}function Ke(e,t){cr[e].checked!==t&&cr[e].click()}function Ue(e){let t=cr[e];return b(t,"change").pipe(l(()=>t.checked),V(t.checked))}function Ha(e,t){switch(e.constructor){case HTMLInputElement:return e.type==="radio"?/^Arrow/.test(t):!0;case HTMLSelectElement:case HTMLTextAreaElement:return!0;default:return e.isContentEditable}}function Pa(){return L(b(window,"compositionstart").pipe(l(()=>!0)),b(window,"compositionend").pipe(l(()=>!1))).pipe(V(!1))}function io(){let e=b(window,"keydown").pipe(A(t=>!(t.metaKey||t.ctrlKey)),l(t=>({mode:oo("search")?"search":"global",type:t.key,claim(){t.preventDefault(),t.stopPropagation()}})),A(({mode:t,type:r})=>{if(t==="global"){let n=_e();if(typeof n!="undefined")return!Ha(n,r)}return!0}),pe());return Pa().pipe(g(t=>t?_:e))}function le(){return new URL(location.href)}function ot(e){location.href=e.href}function ao(){return new x}function so(e,t){if(typeof t=="string"||typeof t=="number")e.innerHTML+=t.toString();else if(t instanceof Node)e.appendChild(t);else if(Array.isArray(t))for(let r of t)so(e,r)}function M(e,t,...r){let n=document.createElement(e);if(t)for(let o of Object.keys(t))typeof t[o]!="undefined"&&(typeof t[o]!="boolean"?n.setAttribute(o,t[o]):n.setAttribute(o,""));for(let o of r)so(n,o);return n}function fr(e){if(e>999){let t=+((e-950)%1e3>99);return`${((e+1e-6)/1e3).toFixed(t)}k`}else return e.toString()}function co(){return location.hash.substring(1)}function Vr(e){let t=M("a",{href:e});t.addEventListener("click",r=>r.stopPropagation()),t.click()}function $a(e){return L(b(window,"hashchange"),e).pipe(l(co),V(co()),A(t=>t.length>0),J(1))}function fo(e){return $a(e).pipe(l(t=>ce(`[id="${t}"]`)),A(t=>typeof t!="undefined"))}function zr(e){let t=matchMedia(e);return er(r=>t.addListener(()=>r(t.matches))).pipe(V(t.matches))}function uo(){let e=matchMedia("print");return L(b(window,"beforeprint").pipe(l(()=>!0)),b(window,"afterprint").pipe(l(()=>!1))).pipe(V(e.matches))}function Nr(e,t){return e.pipe(g(r=>r?t():_))}function ur(e,t={credentials:"same-origin"}){return ue(fetch(`${e}`,t)).pipe(fe(()=>_),g(r=>r.status!==200?Tt(()=>new Error(r.statusText)):k(r)))}function We(e,t){return ur(e,t).pipe(g(r=>r.json()),J(1))}function po(e,t){let r=new DOMParser;return ur(e,t).pipe(g(n=>n.text()),l(n=>r.parseFromString(n,"text/xml")),J(1))}function pr(e){let t=M("script",{src:e});return $(()=>(document.head.appendChild(t),L(b(t,"load"),b(t,"error").pipe(g(()=>Tt(()=>new ReferenceError(`Invalid script: ${e}`))))).pipe(l(()=>{}),R(()=>document.head.removeChild(t)),ge(1))))}function lo(){return{x:Math.max(0,scrollX),y:Math.max(0,scrollY)}}function mo(){return L(b(window,"scroll",{passive:!0}),b(window,"resize",{passive:!0})).pipe(l(lo),V(lo()))}function ho(){return{width:innerWidth,height:innerHeight}}function bo(){return b(window,"resize",{passive:!0}).pipe(l(ho),V(ho()))}function vo(){return Q([mo(),bo()]).pipe(l(([e,t])=>({offset:e,size:t})),J(1))}function lr(e,{viewport$:t,header$:r}){let n=t.pipe(Z("size")),o=Q([n,r]).pipe(l(()=>Xe(e)));return Q([r,t,o]).pipe(l(([{height:i},{offset:s,size:a},{x:c,y:f}])=>({offset:{x:s.x-c,y:s.y-f+i},size:a})))}(()=>{function e(n,o){parent.postMessage(n,o||"*")}function t(...n){return n.reduce((o,i)=>o.then(()=>new Promise(s=>{let a=document.createElement("script");a.src=i,a.onload=s,document.body.appendChild(a)})),Promise.resolve())}var r=class extends EventTarget{constructor(n){super(),this.url=n,this.m=i=>{i.source===this.w&&(this.dispatchEvent(new MessageEvent("message",{data:i.data})),this.onmessage&&this.onmessage(i))},this.e=(i,s,a,c,f)=>{if(s===`${this.url}`){let u=new ErrorEvent("error",{message:i,filename:s,lineno:a,colno:c,error:f});this.dispatchEvent(u),this.onerror&&this.onerror(u)}};let o=document.createElement("iframe");o.hidden=!0,document.body.appendChild(this.iframe=o),this.w.document.open(),this.w.document.write(` + + + + + + + + + + + + + + + + + + + + + + + + + +

+
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Backtesting

+

This page explains how to validate your strategy performance by using Backtesting.

+

Backtesting requires historic data to be available. +To learn how to get data for the pairs and exchange you're interested in, head over to the Data Downloading section of the documentation.

+

Backtesting command reference

+
usage: freqtrade backtesting [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                             [-d PATH] [--userdir PATH] [-s NAME]
+                             [--strategy-path PATH] [-i TIMEFRAME]
+                             [--timerange TIMERANGE]
+                             [--data-format-ohlcv {json,jsongz,hdf5}]
+                             [--max-open-trades INT]
+                             [--stake-amount STAKE_AMOUNT] [--fee FLOAT]
+                             [-p PAIRS [PAIRS ...]] [--eps] [--dmmp]
+                             [--enable-protections]
+                             [--dry-run-wallet DRY_RUN_WALLET]
+                             [--timeframe-detail TIMEFRAME_DETAIL]
+                             [--strategy-list STRATEGY_LIST [STRATEGY_LIST ...]]
+                             [--export {none,trades,signals}]
+                             [--export-filename PATH]
+                             [--breakdown {day,week,month} [{day,week,month} ...]]
+                             [--cache {none,day,week,month}]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  -i TIMEFRAME, --timeframe TIMEFRAME
+                        Specify timeframe (`1m`, `5m`, `30m`, `1h`, `1d`).
+  --timerange TIMERANGE
+                        Specify what timerange of data to use.
+  --data-format-ohlcv {json,jsongz,hdf5}
+                        Storage format for downloaded candle (OHLCV) data.
+                        (default: `json`).
+  --max-open-trades INT
+                        Override the value of the `max_open_trades`
+                        configuration setting.
+  --stake-amount STAKE_AMOUNT
+                        Override the value of the `stake_amount` configuration
+                        setting.
+  --fee FLOAT           Specify fee ratio. Will be applied twice (on trade
+                        entry and exit).
+  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]
+                        Limit command to these pairs. Pairs are space-
+                        separated.
+  --eps, --enable-position-stacking
+                        Allow buying the same pair multiple times (position
+                        stacking).
+  --dmmp, --disable-max-market-positions
+                        Disable applying `max_open_trades` during backtest
+                        (same as setting `max_open_trades` to a very high
+                        number).
+  --enable-protections, --enableprotections
+                        Enable protections for backtesting.Will slow
+                        backtesting down by a considerable amount, but will
+                        include configured protections
+  --dry-run-wallet DRY_RUN_WALLET, --starting-balance DRY_RUN_WALLET
+                        Starting balance, used for backtesting / hyperopt and
+                        dry-runs.
+  --timeframe-detail TIMEFRAME_DETAIL
+                        Specify detail timeframe for backtesting (`1m`, `5m`,
+                        `30m`, `1h`, `1d`).
+  --strategy-list STRATEGY_LIST [STRATEGY_LIST ...]
+                        Provide a space-separated list of strategies to
+                        backtest. Please note that timeframe needs to be set
+                        either in config or via command line. When using this
+                        together with `--export trades`, the strategy-name is
+                        injected into the filename (so `backtest-data.json`
+                        becomes `backtest-data-SampleStrategy.json`
+  --export {none,trades,signals}
+                        Export backtest results (default: trades).
+  --export-filename PATH, --backtest-filename PATH
+                        Use this filename for backtest results.Requires
+                        `--export` to be set as well. Example: `--export-filen
+                        ame=user_data/backtest_results/backtest_today.json`
+  --breakdown {day,week,month} [{day,week,month} ...]
+                        Show backtesting breakdown per [day, week, month].
+  --cache {none,day,week,month}
+                        Load a cached backtest result no older than specified
+                        age (default: day).
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+Strategy arguments:
+  -s NAME, --strategy NAME
+                        Specify strategy class name which will be used by the
+                        bot.
+  --strategy-path PATH  Specify additional strategy lookup path.
+
+

Test your strategy with Backtesting

+

Now you have good Entry and exit strategies and some historic data, you want to test it against +real data. This is what we call backtesting.

+

Backtesting will use the crypto-currencies (pairs) from your config file and load historical candle (OHLCV) data from user_data/data/<exchange> by default. +If no data is available for the exchange / pair / timeframe combination, backtesting will ask you to download them first using freqtrade download-data. +For details on downloading, please refer to the Data Downloading section in the documentation.

+

The result of backtesting will confirm if your bot has better odds of making a profit than a loss.

+

All profit calculations include fees, and freqtrade will use the exchange's default fees for the calculation.

+
+

Using dynamic pairlists for backtesting

+

Using dynamic pairlists is possible (not all of the handlers are allowed to be used in backtest mode), however it relies on the current market conditions - which will not reflect the historic status of the pairlist. +Also, when using pairlists other than StaticPairlist, reproducibility of backtesting-results cannot be guaranteed. +Please read the pairlists documentation for more information.

+

To achieve reproducible results, best generate a pairlist via the test-pairlist command and use that as static pairlist.

+
+
+

Note

+

By default, Freqtrade will export backtesting results to user_data/backtest_results. +The exported trades can be used for further analysis or can be used by the plotting sub-command (freqtrade plot-dataframe) in the scripts directory.

+
+

Starting balance

+

Backtesting will require a starting balance, which can be provided as --dry-run-wallet <balance> or --starting-balance <balance> command line argument, or via dry_run_wallet configuration setting. +This amount must be higher than stake_amount, otherwise the bot will not be able to simulate any trade.

+

Dynamic stake amount

+

Backtesting supports dynamic stake amount by configuring stake_amount as "unlimited", which will split the starting balance into max_open_trades pieces. +Profits from early trades will result in subsequent higher stake amounts, resulting in compounding of profits over the backtesting period.

+

Example backtesting commands

+

With 5 min candle (OHLCV) data (per default)

+
freqtrade backtesting --strategy AwesomeStrategy
+
+

Where --strategy AwesomeStrategy / -s AwesomeStrategy refers to the class name of the strategy, which is within a python file in the user_data/strategies directory.

+
+

With 1 min candle (OHLCV) data

+
freqtrade backtesting --strategy AwesomeStrategy --timeframe 1m
+
+
+

Providing a custom starting balance of 1000 (in stake currency)

+
freqtrade backtesting --strategy AwesomeStrategy --dry-run-wallet 1000
+
+
+

Using a different on-disk historical candle (OHLCV) data source

+

Assume you downloaded the history data from the Bittrex exchange and kept it in the user_data/data/bittrex-20180101 directory. +You can then use this data for backtesting as follows:

+
freqtrade backtesting --strategy AwesomeStrategy --datadir user_data/data/bittrex-20180101 
+
+
+

Comparing multiple Strategies

+
freqtrade backtesting --strategy-list SampleStrategy1 AwesomeStrategy --timeframe 5m
+
+

Where SampleStrategy1 and AwesomeStrategy refer to class names of strategies.

+
+

Prevent exporting trades to file

+
freqtrade backtesting --strategy backtesting --export none --config config.json 
+
+

Only use this if you're sure you'll not want to plot or analyze your results further.

+
+

Exporting trades to file specifying a custom filename

+
freqtrade backtesting --strategy backtesting --export trades --export-filename=backtest_samplestrategy.json
+
+

Please also read about the strategy startup period.

+
+

Supplying custom fee value

+

Sometimes your account has certain fee rebates (fee reductions starting with a certain account size or monthly volume), which are not visible to ccxt. +To account for this in backtesting, you can use the --fee command line option to supply this value to backtesting. +This fee must be a ratio, and will be applied twice (once for trade entry, and once for trade exit).

+

For example, if the commission fee per order is 0.1% (i.e., 0.001 written as ratio), then you would run backtesting as the following:

+
freqtrade backtesting --fee 0.001
+
+
+

Note

+

Only supply this option (or the corresponding configuration parameter) if you want to experiment with different fee values. By default, Backtesting fetches the default fee from the exchange pair/market info.

+
+
+

Running backtest with smaller test-set by using timerange

+

Use the --timerange argument to change how much of the test-set you want to use.

+

For example, running backtesting with the --timerange=20190501- option will use all available data starting with May 1st, 2019 from your input data.

+
freqtrade backtesting --timerange=20190501-
+
+

You can also specify particular date ranges.

+

The full timerange specification:

+
    +
  • Use data until 2018/01/31: --timerange=-20180131
  • +
  • Use data since 2018/01/31: --timerange=20180131-
  • +
  • Use data since 2018/01/31 till 2018/03/01 : --timerange=20180131-20180301
  • +
  • Use data between POSIX / epoch timestamps 1527595200 1527618600: --timerange=1527595200-1527618600
  • +
+

Understand the backtesting result

+

The most important in the backtesting is to understand the result.

+

A backtesting result will look like that:

+
========================================================= BACKTESTING REPORT =========================================================
+| Pair     | Entries |   Avg Profit % |   Cum Profit % |   Tot Profit BTC |   Tot Profit % | Avg Duration |  Wins Draws Loss   Win%  |
+|:---------|--------:|---------------:|---------------:|-----------------:|---------------:|:-------------|-------------------------:|
+| ADA/BTC  |      35 |          -0.11 |          -3.88 |      -0.00019428 |          -1.94 | 4:35:00      |    14     0    21   40.0 |
+| ARK/BTC  |      11 |          -0.41 |          -4.52 |      -0.00022647 |          -2.26 | 2:03:00      |     3     0     8   27.3 |
+| BTS/BTC  |      32 |           0.31 |           9.78 |       0.00048938 |           4.89 | 5:05:00      |    18     0    14   56.2 |
+| DASH/BTC |      13 |          -0.08 |          -1.07 |      -0.00005343 |          -0.53 | 4:39:00      |     6     0     7   46.2 |
+| ENG/BTC  |      18 |           1.36 |          24.54 |       0.00122807 |          12.27 | 2:50:00      |     8     0    10   44.4 |
+| EOS/BTC  |      36 |           0.08 |           3.06 |       0.00015304 |           1.53 | 3:34:00      |    16     0    20   44.4 |
+| ETC/BTC  |      26 |           0.37 |           9.51 |       0.00047576 |           4.75 | 6:14:00      |    11     0    15   42.3 |
+| ETH/BTC  |      33 |           0.30 |           9.96 |       0.00049856 |           4.98 | 7:31:00      |    16     0    17   48.5 |
+| IOTA/BTC |      32 |           0.03 |           1.09 |       0.00005444 |           0.54 | 3:12:00      |    14     0    18   43.8 |
+| LSK/BTC  |      15 |           1.75 |          26.26 |       0.00131413 |          13.13 | 2:58:00      |     6     0     9   40.0 |
+| LTC/BTC  |      32 |          -0.04 |          -1.38 |      -0.00006886 |          -0.69 | 4:49:00      |    11     0    21   34.4 |
+| NANO/BTC |      17 |           1.26 |          21.39 |       0.00107058 |          10.70 | 1:55:00      |    10     0     7   58.5 |
+| NEO/BTC  |      23 |           0.82 |          18.97 |       0.00094936 |           9.48 | 2:59:00      |    10     0    13   43.5 |
+| REQ/BTC  |       9 |           1.17 |          10.54 |       0.00052734 |           5.27 | 3:47:00      |     4     0     5   44.4 |
+| XLM/BTC  |      16 |           1.22 |          19.54 |       0.00097800 |           9.77 | 3:15:00      |     7     0     9   43.8 |
+| XMR/BTC  |      23 |          -0.18 |          -4.13 |      -0.00020696 |          -2.07 | 5:30:00      |    12     0    11   52.2 |
+| XRP/BTC  |      35 |           0.66 |          22.96 |       0.00114897 |          11.48 | 3:49:00      |    12     0    23   34.3 |
+| ZEC/BTC  |      22 |          -0.46 |         -10.18 |      -0.00050971 |          -5.09 | 2:22:00      |     7     0    15   31.8 |
+| TOTAL    |     429 |           0.36 |         152.41 |       0.00762792 |          76.20 | 4:12:00      |   186     0   243   43.4 |
+====================================================== LEFT OPEN TRADES REPORT ======================================================
+| Pair     |  Entries |   Avg Profit % |   Cum Profit % |   Tot Profit BTC |   Tot Profit % | Avg Duration   |  Win Draw Loss Win% |
+|:---------|---------:|---------------:|---------------:|-----------------:|---------------:|:---------------|--------------------:|
+| ADA/BTC  |        1 |           0.89 |           0.89 |       0.00004434 |           0.44 | 6:00:00        |    1    0    0  100 |
+| LTC/BTC  |        1 |           0.68 |           0.68 |       0.00003421 |           0.34 | 2:00:00        |    1    0    0  100 |
+| TOTAL    |        2 |           0.78 |           1.57 |       0.00007855 |           0.78 | 4:00:00        |    2    0    0  100 |
+==================== EXIT REASON STATS ====================
+| Exit Reason        |   Exits |  Wins |  Draws |  Losses |
+|:-------------------|--------:|------:|-------:|--------:|
+| trailing_stop_loss |     205 |   150 |      0 |      55 |
+| stop_loss          |     166 |     0 |      0 |     166 |
+| exit_signal        |      56 |    36 |      0 |      20 |
+| force_exit         |       2 |     0 |      0 |       2 |
+
+================== SUMMARY METRICS ==================
+| Metric                      | Value               |
+|-----------------------------+---------------------|
+| Backtesting from            | 2019-01-01 00:00:00 |
+| Backtesting to              | 2019-05-01 00:00:00 |
+| Max open trades             | 3                   |
+|                             |                     |
+| Total/Daily Avg Trades      | 429 / 3.575         |
+| Starting balance            | 0.01000000 BTC      |
+| Final balance               | 0.01762792 BTC      |
+| Absolute profit             | 0.00762792 BTC      |
+| Total profit %              | 76.2%               |
+| CAGR %                      | 460.87%             |
+| Sortino                     | 1.88                |
+| Sharpe                      | 2.97                |
+| Calmar                      | 6.29                |
+| Profit factor               | 1.11                |
+| Expectancy                  | -0.15               |
+| Avg. stake amount           | 0.001      BTC      |
+| Total trade volume          | 0.429      BTC      |
+|                             |                     |
+| Long / Short                | 352 / 77            |
+| Total profit Long %         | 1250.58%            |
+| Total profit Short %        | -15.02%             |
+| Absolute profit Long        | 0.00838792 BTC      |
+| Absolute profit Short       | -0.00076 BTC        |
+|                             |                     |
+| Best Pair                   | LSK/BTC 26.26%      |
+| Worst Pair                  | ZEC/BTC -10.18%     |
+| Best Trade                  | LSK/BTC 4.25%       |
+| Worst Trade                 | ZEC/BTC -10.25%     |
+| Best day                    | 0.00076 BTC         |
+| Worst day                   | -0.00036 BTC        |
+| Days win/draw/lose          | 12 / 82 / 25        |
+| Avg. Duration Winners       | 4:23:00             |
+| Avg. Duration Loser         | 6:55:00             |
+| Rejected Entry signals      | 3089                |
+| Entry/Exit Timeouts         | 0 / 0               |
+| Canceled Trade Entries      | 34                  |
+| Canceled Entry Orders       | 123                 |
+| Replaced Entry Orders       | 89                  |
+|                             |                     |
+| Min balance                 | 0.00945123 BTC      |
+| Max balance                 | 0.01846651 BTC      |
+| Max % of account underwater | 25.19%              |
+| Absolute Drawdown (Account) | 13.33%              |
+| Drawdown                    | 0.0015 BTC          |
+| Drawdown high               | 0.0013 BTC          |
+| Drawdown low                | -0.0002 BTC         |
+| Drawdown Start              | 2019-02-15 14:10:00 |
+| Drawdown End                | 2019-04-11 18:15:00 |
+| Market change               | -5.88%              |
+=====================================================
+
+

Backtesting report table

+

The 1st table contains all trades the bot made, including "left open trades".

+

The last line will give you the overall performance of your strategy, +here:

+
| TOTAL    |    429 |           0.36 |         152.41 |       0.00762792 |          76.20 | 4:12:00      |   186     0   243   43.4 |
+
+

The bot has made 429 trades for an average duration of 4:12:00, with a performance of 76.20% (profit), that means it has +earned a total of 0.00762792 BTC starting with a capital of 0.01 BTC.

+

The column Avg Profit % shows the average profit for all trades made while the column Cum Profit % sums up all the profits/losses. +The column Tot Profit % shows instead the total profit % in relation to the starting balance. +In the above results, we have a starting balance of 0.01 BTC and an absolute profit of 0.00762792 BTC - so the Tot Profit % will be (0.00762792 / 0.01) * 100 ~= 76.2%.

+

Your strategy performance is influenced by your entry strategy, your exit strategy, and also by the minimal_roi and stop_loss you have set.

+

For example, if your minimal_roi is only "0": 0.01 you cannot expect the bot to make more profit than 1% (because it will exit every time a trade reaches 1%).

+
"minimal_roi": {
+    "0":  0.01
+},
+
+

On the other hand, if you set a too high minimal_roi like "0": 0.55 +(55%), there is almost no chance that the bot will ever reach this profit. +Hence, keep in mind that your performance is an integral mix of all different elements of the strategy, your configuration, and the crypto-currency pairs you have set up.

+

Exit reasons table

+

The 2nd table contains a recap of exit reasons. +This table can tell you which area needs some additional work (e.g. all or many of the exit_signal trades are losses, so you should work on improving the exit signal, or consider disabling it).

+

Left open trades table

+

The 3rd table contains all trades the bot had to force_exit at the end of the backtesting period to present you the full picture. +This is necessary to simulate realistic behavior, since the backtest period has to end at some point, while realistically, you could leave the bot running forever. +These trades are also included in the first table, but are also shown separately in this table for clarity.

+

Summary metrics

+

The last element of the backtest report is the summary metrics table. +It contains some useful key metrics about performance of your strategy on backtesting data.

+
================== SUMMARY METRICS ==================
+| Metric                      | Value               |
+|-----------------------------+---------------------|
+| Backtesting from            | 2019-01-01 00:00:00 |
+| Backtesting to              | 2019-05-01 00:00:00 |
+| Max open trades             | 3                   |
+|                             |                     |
+| Total/Daily Avg Trades      | 429 / 3.575         |
+| Starting balance            | 0.01000000 BTC      |
+| Final balance               | 0.01762792 BTC      |
+| Absolute profit             | 0.00762792 BTC      |
+| Total profit %              | 76.2%               |
+| CAGR %                      | 460.87%             |
+| Sortino                     | 1.88                |
+| Sharpe                      | 2.97                |
+| Calmar                      | 6.29                |
+| Profit factor               | 1.11                |
+| Expectancy                  | -0.15               |
+| Avg. stake amount           | 0.001      BTC      |
+| Total trade volume          | 0.429      BTC      |
+|                             |                     |
+| Long / Short                | 352 / 77            |
+| Total profit Long %         | 1250.58%            |
+| Total profit Short %        | -15.02%             |
+| Absolute profit Long        | 0.00838792 BTC      |
+| Absolute profit Short       | -0.00076 BTC        |
+|                             |                     |
+| Best Pair                   | LSK/BTC 26.26%      |
+| Worst Pair                  | ZEC/BTC -10.18%     |
+| Best Trade                  | LSK/BTC 4.25%       |
+| Worst Trade                 | ZEC/BTC -10.25%     |
+| Best day                    | 0.00076 BTC         |
+| Worst day                   | -0.00036 BTC        |
+| Days win/draw/lose          | 12 / 82 / 25        |
+| Avg. Duration Winners       | 4:23:00             |
+| Avg. Duration Loser         | 6:55:00             |
+| Rejected Entry signals      | 3089                |
+| Entry/Exit Timeouts         | 0 / 0               |
+| Canceled Trade Entries      | 34                  |
+| Canceled Entry Orders       | 123                 |
+| Replaced Entry Orders       | 89                  |
+|                             |                     |
+| Min balance                 | 0.00945123 BTC      |
+| Max balance                 | 0.01846651 BTC      |
+| Max % of account underwater | 25.19%              |
+| Absolute Drawdown (Account) | 13.33%              |
+| Drawdown                    | 0.0015 BTC          |
+| Drawdown high               | 0.0013 BTC          |
+| Drawdown low                | -0.0002 BTC         |
+| Drawdown Start              | 2019-02-15 14:10:00 |
+| Drawdown End                | 2019-04-11 18:15:00 |
+| Market change               | -5.88%              |
+=====================================================
+
+
    +
  • Backtesting from / Backtesting to: Backtesting range (usually defined with the --timerange option).
  • +
  • Max open trades: Setting of max_open_trades (or --max-open-trades) - or number of pairs in the pairlist (whatever is lower).
  • +
  • Total/Daily Avg Trades: Identical to the total trades of the backtest output table / Total trades divided by the backtesting duration in days (this will give you information about how many trades to expect from the strategy).
  • +
  • Starting balance: Start balance - as given by dry-run-wallet (config or command line).
  • +
  • Final balance: Final balance - starting balance + absolute profit.
  • +
  • Absolute profit: Profit made in stake currency.
  • +
  • Total profit %: Total profit. Aligned to the TOTAL row's Tot Profit % from the first table. Calculated as (End capital − Starting capital) / Starting capital.
  • +
  • CAGR %: Compound annual growth rate.
  • +
  • Sortino: Annualized Sortino ratio.
  • +
  • Sharpe: Annualized Sharpe ratio.
  • +
  • Calmar: Annualized Calmar ratio.
  • +
  • Profit factor: profit / loss.
  • +
  • Avg. stake amount: Average stake amount, either stake_amount or the average when using dynamic stake amount.
  • +
  • Total trade volume: Volume generated on the exchange to reach the above profit.
  • +
  • Best Pair / Worst Pair: Best and worst performing pair, and it's corresponding Cum Profit %.
  • +
  • Best Trade / Worst Trade: Biggest single winning trade and biggest single losing trade.
  • +
  • Best day / Worst day: Best and worst day based on daily profit.
  • +
  • Days win/draw/lose: Winning / Losing days (draws are usually days without closed trade).
  • +
  • Avg. Duration Winners / Avg. Duration Loser: Average durations for winning and losing trades.
  • +
  • Rejected Entry signals: Trade entry signals that could not be acted upon due to max_open_trades being reached.
  • +
  • Entry/Exit Timeouts: Entry/exit orders which did not fill (only applicable if custom pricing is used).
  • +
  • Canceled Trade Entries: Number of trades that have been canceled by user request via adjust_entry_price.
  • +
  • Canceled Entry Orders: Number of entry orders that have been canceled by user request via adjust_entry_price.
  • +
  • Replaced Entry Orders: Number of entry orders that have been replaced by user request via adjust_entry_price.
  • +
  • Min balance / Max balance: Lowest and Highest Wallet balance during the backtest period.
  • +
  • Max % of account underwater: Maximum percentage your account has decreased from the top since the simulation started. +Calculated as the maximum of (Max Balance - Current Balance) / (Max Balance).
  • +
  • Absolute Drawdown (Account): Maximum Account Drawdown experienced. Calculated as (Absolute Drawdown) / (DrawdownHigh + startingBalance).
  • +
  • Drawdown: Maximum, absolute drawdown experienced. Difference between Drawdown High and Subsequent Low point.
  • +
  • Drawdown high / Drawdown low: Profit at the beginning and end of the largest drawdown period. A negative low value means initial capital lost.
  • +
  • Drawdown Start / Drawdown End: Start and end datetime for this largest drawdown (can also be visualized via the plot-dataframe sub-command).
  • +
  • Market change: Change of the market during the backtest period. Calculated as average of all pairs changes from the first to the last candle using the "close" column.
  • +
  • Long / Short: Split long/short values (Only shown when short trades were made).
  • +
  • Total profit Long % / Absolute profit Long: Profit long trades only (Only shown when short trades were made).
  • +
  • Total profit Short % / Absolute profit Short: Profit short trades only (Only shown when short trades were made).
  • +
+

Daily / Weekly / Monthly breakdown

+

You can get an overview over daily / weekly or monthly results by using the --breakdown <> switch.

+

To visualize daily and weekly breakdowns, you can use the following:

+
freqtrade backtesting --strategy MyAwesomeStrategy --breakdown day week
+
+
======================== DAY BREAKDOWN =========================
+|        Day |   Tot Profit USDT |   Wins |   Draws |   Losses |
+|------------+-------------------+--------+---------+----------|
+| 03/07/2021 |           200.0   |      2 |       0 |        0 |
+| 04/07/2021 |           -50.31  |      0 |       0 |        2 |
+| 05/07/2021 |           220.611 |      3 |       2 |        0 |
+| 06/07/2021 |           150.974 |      3 |       0 |        2 |
+| 07/07/2021 |           -70.193 |      1 |       0 |        2 |
+| 08/07/2021 |           212.413 |      2 |       0 |        3 |
+
+

The output will show a table containing the realized absolute Profit (in stake currency) for the given timeperiod, as well as wins, draws and losses that materialized (closed) on this day. Below that there will be a second table for the summarized values of weeks indicated by the date of the closing Sunday. The same would apply to a monthly breakdown indicated by the last day of the month.

+

Backtest result caching

+

To save time, by default backtest will reuse a cached result from within the last day when the backtested strategy and config match that of a previous backtest. To force a new backtest despite existing result for an identical run specify --cache none parameter.

+
+

Warning

+

Caching is automatically disabled for open-ended timeranges (--timerange 20210101-), as freqtrade cannot ensure reliably that the underlying data didn't change. It can also use cached results where it shouldn't if the original backtest had missing data at the end, which was fixed by downloading more data. +In this instance, please use --cache none once to force a fresh backtest.

+
+

Further backtest-result analysis

+

To further analyze your backtest results, you can export the trades. +You can then load the trades to perform further analysis as shown in the data analysis backtesting section.

+

Assumptions made by backtesting

+

Since backtesting lacks some detailed information about what happens within a candle, it needs to take a few assumptions:

+
    +
  • Exchange trading limits are respected
  • +
  • Entries happen at open-price
  • +
  • All orders are filled at the requested price (no slippage, no unfilled orders)
  • +
  • Exit-signal exits happen at open-price of the consecutive candle
  • +
  • Exit-signal is favored over Stoploss, because exit-signals are assumed to trigger on candle's open
  • +
  • ROI
      +
    • exits are compared to high - but the ROI value is used (e.g. ROI = 2%, high=5% - so the exit will be at 2%)
    • +
    • exits are never "below the candle", so a ROI of 2% may result in a exit at 2.4% if low was at 2.4% profit
    • +
    • Force-exits caused by <N>=-1 ROI entries use low as exit value, unless N falls on the candle open (e.g. 120: -1 for 1h candles)
    • +
    +
  • +
  • Stoploss exits happen exactly at stoploss price, even if low was lower, but the loss will be 2 * fees higher than the stoploss price
  • +
  • Stoploss is evaluated before ROI within one candle. So you can often see more trades with the stoploss exit reason comparing to the results obtained with the same strategy in the Dry Run/Live Trade modes
  • +
  • Low happens before high for stoploss, protecting capital first
  • +
  • Trailing stoploss
      +
    • Trailing Stoploss is only adjusted if it's below the candle's low (otherwise it would be triggered)
    • +
    • On trade entry candles that trigger trailing stoploss, the "minimum offset" (stop_positive_offset) is assumed (instead of high) - and the stop is calculated from this point. This rule is NOT applicable to custom-stoploss scenarios, since there's no information about the stoploss logic available.
    • +
    • High happens first - adjusting stoploss
    • +
    • Low uses the adjusted stoploss (so exits with large high-low difference are backtested correctly)
    • +
    • ROI applies before trailing-stop, ensuring profits are "top-capped" at ROI if both ROI and trailing stop applies
    • +
    +
  • +
  • Exit-reason does not explain if a trade was positive or negative, just what triggered the exit (this can look odd if negative ROI values are used)
  • +
  • Evaluation sequence (if multiple signals happen on the same candle)
      +
    • Exit-signal
    • +
    • Stoploss
    • +
    • ROI
    • +
    • Trailing stoploss
    • +
    +
  • +
+

Taking these assumptions, backtesting tries to mirror real trading as closely as possible. However, backtesting will never replace running a strategy in dry-run mode. +Also, keep in mind that past results don't guarantee future success.

+

In addition to the above assumptions, strategy authors should carefully read the Common Mistakes section, to avoid using data in backtesting which is not available in real market conditions.

+

Trading limits in backtesting

+

Exchanges have certain trading limits, like minimum (and maximum) base currency, or minimum/maximum stake (quote) currency. +These limits are usually listed in the exchange documentation as "trading rules" or similar and can be quite different between different pairs.

+

Backtesting (as well as live and dry-run) does honor these limits, and will ensure that a stoploss can be placed below this value - so the value will be slightly higher than what the exchange specifies. +Freqtrade has however no information about historic limits.

+

This can lead to situations where trading-limits are inflated by using a historic price, resulting in minimum amounts > 50$.

+

For example:

+

BTC minimum tradable amount is 0.001. +BTC trades at 22.000$ today (0.001 BTC is related to this) - but the backtesting period includes prices as high as 50.000$. +Today's minimum would be 0.001 * 22_000 - or 22$.
+However the limit could also be 50$ - based on 0.001 * 50_000 in some historic setting.

+

Trading precision limits

+

Most exchanges pose precision limits on both price and amounts, so you cannot buy 1.0020401 of a pair, or at a price of 1.24567123123.
+Instead, these prices and amounts will be rounded or truncated (based on the exchange definition) to the defined trading precision. +The above values may for example be rounded to an amount of 1.002, and a price of 1.24567.

+

These precision values are based on current exchange limits (as described in the above section), as historic precision limits are not available.

+

Improved backtest accuracy

+

One big limitation of backtesting is it's inability to know how prices moved intra-candle (was high before close, or viceversa?). +So assuming you run backtesting with a 1h timeframe, there will be 4 prices for that candle (Open, High, Low, Close).

+

While backtesting does take some assumptions (read above) about this - this can never be perfect, and will always be biased in one way or the other. +To mitigate this, freqtrade can use a lower (faster) timeframe to simulate intra-candle movements.

+

To utilize this, you can append --timeframe-detail 5m to your regular backtesting command.

+
freqtrade backtesting --strategy AwesomeStrategy --timeframe 1h --timeframe-detail 5m
+
+

This will load 1h data as well as 5m data for the timeframe. The strategy will be analyzed with the 1h timeframe, and Entry orders will only be placed at the main timeframe, however Order fills and exit signals will be evaluated at the 5m candle, simulating intra-candle movements.

+

All callback functions (custom_exit(), custom_stoploss(), ... ) will be running for each 5m candle once the trade is opened (so 12 times in the above example of 1h timeframe, and 5m detailed timeframe).

+

--timeframe-detail must be smaller than the original timeframe, otherwise backtesting will fail to start.

+

Obviously this will require more memory (5m data is bigger than 1h data), and will also impact runtime (depending on the amount of trades and trade durations). +Also, data must be available / downloaded already.

+
+

Tip

+

You can use this function as the last part of strategy development, to ensure your strategy is not exploiting one of the backtesting assumptions. Strategies that perform similarly well with this mode have a good chance to perform well in dry/live modes too (although only forward-testing (dry-mode) can really confirm a strategy).

+
+

Backtesting multiple strategies

+

To compare multiple strategies, a list of Strategies can be provided to backtesting.

+

This is limited to 1 timeframe value per run. However, data is only loaded once from disk so if you have multiple +strategies you'd like to compare, this will give a nice runtime boost.

+

All listed Strategies need to be in the same directory.

+
freqtrade backtesting --timerange 20180401-20180410 --timeframe 5m --strategy-list Strategy001 Strategy002 --export trades
+
+

This will save the results to user_data/backtest_results/backtest-result-<strategy>.json, injecting the strategy-name into the target filename. +There will be an additional table comparing win/losses of the different strategies (identical to the "Total" row in the first table). +Detailed output for all strategies one after the other will be available, so make sure to scroll up to see the details per strategy.

+
=========================================================== STRATEGY SUMMARY ===========================================================================
+| Strategy    |  Entries |   Avg Profit % |   Cum Profit % |   Tot Profit BTC |   Tot Profit % | Avg Duration   |  Wins |  Draws | Losses | Drawdown % |
+|:------------|---------:|---------------:|---------------:|-----------------:|---------------:|:---------------|------:|-------:|-------:|-----------:|
+| Strategy1   |      429 |           0.36 |         152.41 |       0.00762792 |          76.20 | 4:12:00        |   186 |      0 |    243 |       45.2 |
+| Strategy2   |     1487 |          -0.13 |        -197.58 |      -0.00988917 |         -98.79 | 4:43:00        |   662 |      0 |    825 |     241.68 |
+
+

Next step

+

Great, your strategy is profitable. What if the bot can give your the optimal parameters to use for your strategy? +Your next step is to learn how to find optimal parameters with Hyperopt

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/bot-basics/index.html b/en/2023.4/bot-basics/index.html new file mode 100644 index 000000000..68aae8eab --- /dev/null +++ b/en/2023.4/bot-basics/index.html @@ -0,0 +1,1386 @@ + + + + + + + + + + + + + + + + + + + + + + + + Freqtrade Basics - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Freqtrade basics

+

This page provides you some basic concepts on how Freqtrade works and operates.

+

Freqtrade terminology

+
    +
  • Strategy: Your trading strategy, telling the bot what to do.
  • +
  • Trade: Open position.
  • +
  • Open Order: Order which is currently placed on the exchange, and is not yet complete.
  • +
  • Pair: Tradable pair, usually in the format of Base/Quote (e.g. XRP/USDT).
  • +
  • Timeframe: Candle length to use (e.g. "5m", "1h", ...).
  • +
  • Indicators: Technical indicators (SMA, EMA, RSI, ...).
  • +
  • Limit order: Limit orders which execute at the defined limit price or better.
  • +
  • Market order: Guaranteed to fill, may move price depending on the order size.
  • +
  • Current Profit: Currently pending (unrealized) profit for this trade. This is mainly used throughout the bot and UI.
  • +
  • Realized Profit: Already realized profit. Only relevant in combination with partial exits - which also explains the calculation logic for this.
  • +
  • Total Profit: Combined realized and unrealized profit. The relative number (%) is calculated against the total investment in this trade.
  • +
+

Fee handling

+

All profit calculations of Freqtrade include fees. For Backtesting / Hyperopt / Dry-run modes, the exchange default fee is used (lowest tier on the exchange). For live operations, fees are used as applied by the exchange (this includes BNB rebates etc.).

+

Bot execution logic

+

Starting freqtrade in dry-run or live mode (using freqtrade trade) will start the bot and start the bot iteration loop. +This will also run the bot_start() callback.

+

By default, the bot loop runs every few seconds (internals.process_throttle_secs) and performs the following actions:

+
    +
  • Fetch open trades from persistence.
  • +
  • Calculate current list of tradable pairs.
  • +
  • Download OHLCV data for the pairlist including all informative pairs
    + This step is only executed once per Candle to avoid unnecessary network traffic.
  • +
  • Call bot_loop_start() strategy callback.
  • +
  • Analyze strategy per pair.
      +
    • Call populate_indicators()
    • +
    • Call populate_entry_trend()
    • +
    • Call populate_exit_trend()
    • +
    +
  • +
  • Check timeouts for open orders.
      +
    • Calls check_entry_timeout() strategy callback for open entry orders.
    • +
    • Calls check_exit_timeout() strategy callback for open exit orders.
    • +
    • Calls adjust_entry_price() strategy callback for open entry orders.
    • +
    +
  • +
  • Verifies existing positions and eventually places exit orders.
      +
    • Considers stoploss, ROI and exit-signal, custom_exit() and custom_stoploss().
    • +
    • Determine exit-price based on exit_pricing configuration setting or by using the custom_exit_price() callback.
    • +
    • Before a exit order is placed, confirm_trade_exit() strategy callback is called.
    • +
    +
  • +
  • Check position adjustments for open trades if enabled by calling adjust_trade_position() and place additional order if required.
  • +
  • Check if trade-slots are still available (if max_open_trades is reached).
  • +
  • Verifies entry signal trying to enter new positions.
      +
    • Determine entry-price based on entry_pricing configuration setting, or by using the custom_entry_price() callback.
    • +
    • In Margin and Futures mode, leverage() strategy callback is called to determine the desired leverage.
    • +
    • Determine stake size by calling the custom_stake_amount() callback.
    • +
    • Before an entry order is placed, confirm_trade_entry() strategy callback is called.
    • +
    +
  • +
+

This loop will be repeated again and again until the bot is stopped.

+

Backtesting / Hyperopt execution logic

+

backtesting or hyperopt do only part of the above logic, since most of the trading operations are fully simulated.

+
    +
  • Load historic data for configured pairlist.
  • +
  • Calls bot_start() once.
  • +
  • Calculate indicators (calls populate_indicators() once per pair).
  • +
  • Calculate entry / exit signals (calls populate_entry_trend() and populate_exit_trend() once per pair).
  • +
  • Loops per candle simulating entry and exit points.
      +
    • Calls bot_loop_start() strategy callback.
    • +
    • Check for Order timeouts, either via the unfilledtimeout configuration, or via check_entry_timeout() / check_exit_timeout() strategy callbacks.
    • +
    • Calls adjust_entry_price() strategy callback for open entry orders.
    • +
    • Check for trade entry signals (enter_long / enter_short columns).
    • +
    • Confirm trade entry / exits (calls confirm_trade_entry() and confirm_trade_exit() if implemented in the strategy).
    • +
    • Call custom_entry_price() (if implemented in the strategy) to determine entry price (Prices are moved to be within the opening candle).
    • +
    • In Margin and Futures mode, leverage() strategy callback is called to determine the desired leverage.
    • +
    • Determine stake size by calling the custom_stake_amount() callback.
    • +
    • Check position adjustments for open trades if enabled and call adjust_trade_position() to determine if an additional order is requested.
    • +
    • Call custom_stoploss() and custom_exit() to find custom exit points.
    • +
    • For exits based on exit-signal, custom-exit and partial exits: Call custom_exit_price() to determine exit price (Prices are moved to be within the closing candle).
    • +
    +
  • +
  • Generate backtest report output
  • +
+
+

Note

+

Both Backtesting and Hyperopt include exchange default Fees in the calculation. Custom fees can be passed to backtesting / hyperopt by specifying the --fee argument.

+
+
+

Callback call frequency

+

Backtesting will call each callback at max. once per candle (--timeframe-detail modifies this behavior to once per detailed candle). +Most callbacks will be called once per iteration in live (usually every ~5s) - which can cause backtesting mismatches.

+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/bot-usage/index.html b/en/2023.4/bot-usage/index.html new file mode 100644 index 000000000..86af55342 --- /dev/null +++ b/en/2023.4/bot-usage/index.html @@ -0,0 +1,1572 @@ + + + + + + + + + + + + + + + + + + + + + + + + Start the bot - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Start the bot

+

This page explains the different parameters of the bot and how to run it.

+
+

Note

+

If you've used setup.sh, don't forget to activate your virtual environment (source .env/bin/activate) before running freqtrade commands.

+
+
+

Up-to-date clock

+

The clock on the system running the bot must be accurate, synchronized to a NTP server frequently enough to avoid problems with communication to the exchanges.

+
+

Bot commands

+
usage: freqtrade [-h] [-V]
+                 {trade,create-userdir,new-config,new-strategy,download-data,convert-data,convert-trade-data,list-data,backtesting,edge,hyperopt,hyperopt-list,hyperopt-show,list-exchanges,list-hyperopts,list-markets,list-pairs,list-strategies,list-timeframes,show-trades,test-pairlist,install-ui,plot-dataframe,plot-profit,webserver}
+                 ...
+
+Free, open source crypto trading bot
+
+positional arguments:
+  {trade,create-userdir,new-config,new-strategy,download-data,convert-data,convert-trade-data,list-data,backtesting,edge,hyperopt,hyperopt-list,hyperopt-show,list-exchanges,list-hyperopts,list-markets,list-pairs,list-strategies,list-timeframes,show-trades,test-pairlist,install-ui,plot-dataframe,plot-profit,webserver}
+    trade               Trade module.
+    create-userdir      Create user-data directory.
+    new-config          Create new config
+    new-strategy        Create new strategy
+    download-data       Download backtesting data.
+    convert-data        Convert candle (OHLCV) data from one format to
+                        another.
+    convert-trade-data  Convert trade data from one format to another.
+    list-data           List downloaded data.
+    backtesting         Backtesting module.
+    edge                Edge module.
+    hyperopt            Hyperopt module.
+    hyperopt-list       List Hyperopt results
+    hyperopt-show       Show details of Hyperopt results
+    list-exchanges      Print available exchanges.
+    list-hyperopts      Print available hyperopt classes.
+    list-markets        Print markets on exchange.
+    list-pairs          Print pairs on exchange.
+    list-strategies     Print available strategies.
+    list-timeframes     Print available timeframes for the exchange.
+    show-trades         Show trades.
+    test-pairlist       Test your pairlist configuration.
+    install-ui          Install FreqUI
+    plot-dataframe      Plot candles with indicators.
+    plot-profit         Generate plot showing profits.
+    webserver           Webserver module.
+
+optional arguments:
+  -h, --help            show this help message and exit
+  -V, --version         show program's version number and exit
+
+

Bot trading commands

+
usage: freqtrade trade [-h] [-v] [--logfile FILE] [-V] [-c PATH] [-d PATH]
+                       [--userdir PATH] [-s NAME] [--strategy-path PATH]
+                       [--db-url PATH] [--sd-notify] [--dry-run]
+                       [--dry-run-wallet DRY_RUN_WALLET]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --db-url PATH         Override trades database URL, this is useful in custom
+                        deployments (default: `sqlite:///tradesv3.sqlite` for
+                        Live Run mode, `sqlite:///tradesv3.dryrun.sqlite` for
+                        Dry Run).
+  --sd-notify           Notify systemd service manager.
+  --dry-run             Enforce dry-run for trading (removes Exchange secrets
+                        and simulates trades).
+  --dry-run-wallet DRY_RUN_WALLET, --starting-balance DRY_RUN_WALLET
+                        Starting balance, used for backtesting / hyperopt and
+                        dry-runs.
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+Strategy arguments:
+  -s NAME, --strategy NAME
+                        Specify strategy class name which will be used by the
+                        bot.
+  --strategy-path PATH  Specify additional strategy lookup path.
+
+

How to specify which configuration file be used?

+

The bot allows you to select which configuration file you want to use by means of +the -c/--config command line option:

+
freqtrade trade -c path/far/far/away/config.json
+
+

Per default, the bot loads the config.json configuration file from the current +working directory.

+

How to use multiple configuration files?

+

The bot allows you to use multiple configuration files by specifying multiple +-c/--config options in the command line. Configuration parameters +defined in the latter configuration files override parameters with the same name +defined in the previous configuration files specified in the command line earlier.

+

For example, you can make a separate configuration file with your key and secret +for the Exchange you use for trading, specify default configuration file with +empty key and secret values while running in the Dry Mode (which does not actually +require them):

+
freqtrade trade -c ./config.json
+
+

and specify both configuration files when running in the normal Live Trade Mode:

+
freqtrade trade -c ./config.json -c path/to/secrets/keys.config.json
+
+

This could help you hide your private Exchange key and Exchange secret on you local machine +by setting appropriate file permissions for the file which contains actual secrets and, additionally, +prevent unintended disclosure of sensitive private data when you publish examples +of your configuration in the project issues or in the Internet.

+

See more details on this technique with examples in the documentation page on +configuration.

+

Where to store custom data

+

Freqtrade allows the creation of a user-data directory using freqtrade create-userdir --userdir someDirectory. +This directory will look as follows:

+
user_data/
+├── backtest_results
+├── data
+├── hyperopts
+├── hyperopt_results
+├── plot
+└── strategies
+
+

You can add the entry "user_data_dir" setting to your configuration, to always point your bot to this directory. +Alternatively, pass in --userdir to every command. +The bot will fail to start if the directory does not exist, but will create necessary subdirectories.

+

This directory should contain your custom strategies, custom hyperopts and hyperopt loss functions, backtesting historical data (downloaded using either backtesting command or the download script) and plot outputs.

+

It is recommended to use version control to keep track of changes to your strategies.

+

How to use --strategy?

+

This parameter will allow you to load your custom strategy class. +To test the bot installation, you can use the SampleStrategy installed by the create-userdir subcommand (usually user_data/strategy/sample_strategy.py).

+

The bot will search your strategy file within user_data/strategies. +To use other directories, please read the next section about --strategy-path.

+

To load a strategy, simply pass the class name (e.g.: CustomStrategy) in this parameter.

+

Example: +In user_data/strategies you have a file my_awesome_strategy.py which has +a strategy class called AwesomeStrategy to load it:

+
freqtrade trade --strategy AwesomeStrategy
+
+

If the bot does not find your strategy file, it will display in an error +message the reason (File not found, or errors in your code).

+

Learn more about strategy file in +Strategy Customization.

+

How to use --strategy-path?

+

This parameter allows you to add an additional strategy lookup path, which gets +checked before the default locations (The passed path must be a directory!):

+
freqtrade trade --strategy AwesomeStrategy --strategy-path /some/directory
+
+

How to install a strategy?

+

This is very simple. Copy paste your strategy file into the directory +user_data/strategies or use --strategy-path. And voila, the bot is ready to use it.

+

How to use --db-url?

+

When you run the bot in Dry-run mode, per default no transactions are +stored in a database. If you want to store your bot actions in a DB +using --db-url. This can also be used to specify a custom database +in production mode. Example command:

+
freqtrade trade -c config.json --db-url sqlite:///tradesv3.dry_run.sqlite
+
+

Next step

+

The optimal strategy of the bot will change with time depending of the market trends. The next step is to +Strategy Customization.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/configuration/index.html b/en/2023.4/configuration/index.html new file mode 100644 index 000000000..ac9c2cdfd --- /dev/null +++ b/en/2023.4/configuration/index.html @@ -0,0 +1,3082 @@ + + + + + + + + + + + + + + + + + + + + + + + + Configuration - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Configure the bot

+

Freqtrade has many configurable features and possibilities. +By default, these settings are configured via the configuration file (see below).

+

The Freqtrade configuration file

+

The bot uses a set of configuration parameters during its operation that all together conform to the bot configuration. It normally reads its configuration from a file (Freqtrade configuration file).

+

Per default, the bot loads the configuration from the config.json file, located in the current working directory.

+

You can specify a different configuration file used by the bot with the -c/--config command-line option.

+

If you used the Quick start method for installing +the bot, the installation script should have already created the default configuration file (config.json) for you.

+

If the default configuration file is not created we recommend to use freqtrade new-config --config config.json to generate a basic configuration file.

+

The Freqtrade configuration file is to be written in JSON format.

+

Additionally to the standard JSON syntax, you may use one-line // ... and multi-line /* ... */ comments in your configuration files and trailing commas in the lists of parameters.

+

Do not worry if you are not familiar with JSON format -- simply open the configuration file with an editor of your choice, make some changes to the parameters you need, save your changes and, finally, restart the bot or, if it was previously stopped, run it again with the changes you made to the configuration. The bot validates the syntax of the configuration file at startup and will warn you if you made any errors editing it, pointing out problematic lines.

+

Environment variables

+

Set options in the Freqtrade configuration via environment variables. +This takes priority over the corresponding value in configuration or strategy.

+

Environment variables must be prefixed with FREQTRADE__ to be loaded to the freqtrade configuration.

+

__ serves as level separator, so the format used should correspond to FREQTRADE__{section}__{key}. +As such - an environment variable defined as export FREQTRADE__STAKE_AMOUNT=200 would result in {stake_amount: 200}.

+

A more complex example might be export FREQTRADE__EXCHANGE__KEY=<yourExchangeKey> to keep your exchange key secret. This will move the value to the exchange.key section of the configuration. +Using this scheme, all configuration settings will also be available as environment variables.

+

Please note that Environment variables will overwrite corresponding settings in your configuration, but command line Arguments will always win.

+

Common example:

+
FREQTRADE__TELEGRAM__CHAT_ID=<telegramchatid>
+FREQTRADE__TELEGRAM__TOKEN=<telegramToken>
+FREQTRADE__EXCHANGE__KEY=<yourExchangeKey>
+FREQTRADE__EXCHANGE__SECRET=<yourExchangeSecret>
+
+
+

Note

+

Environment variables detected are logged at startup - so if you can't find why a value is not what you think it should be based on the configuration, make sure it's not loaded from an environment variable.

+
+

Multiple configuration files

+

Multiple configuration files can be specified and used by the bot or the bot can read its configuration parameters from the process standard input stream.

+

You can specify additional configuration files in add_config_files. Files specified in this parameter will be loaded and merged with the initial config file. The files are resolved relative to the initial configuration file. +This is similar to using multiple --config parameters, but simpler in usage as you don't have to specify all files for all commands.

+
+

Use multiple configuration files to keep secrets secret

+

You can use a 2nd configuration file containing your secrets. That way you can share your "primary" configuration file, while still keeping your API keys for yourself. +The 2nd file should only specify what you intend to override. +If a key is in more than one of the configurations, then the "last specified configuration" wins (in the above example, config-private.json).

+

For one-off commands, you can also use the below syntax by specifying multiple "--config" parameters.

+
freqtrade trade --config user_data/config1.json --config user_data/config-private.json <...>
+
+

The below is equivalent to the example above - but having 2 configuration files in the configuration, for easier reuse.

+
user_data/config.json
"add_config_files": [
+    "config1.json",
+    "config-private.json"
+]
+
+
freqtrade trade --config user_data/config.json <...>
+
+
+
+config collision handling +

If the same configuration setting takes place in both config.json and config-import.json, then the parent configuration wins. +In the below case, max_open_trades would be 3 after the merging - as the reusable "import" configuration has this key overwritten.

+
user_data/config.json
{
+    "max_open_trades": 3,
+    "stake_currency": "USDT",
+    "add_config_files": [
+        "config-import.json"
+    ]
+}
+
+
user_data/config-import.json
{
+    "max_open_trades": 10,
+    "stake_amount": "unlimited",
+}
+
+

Resulting combined configuration:

+
Result
{
+    "max_open_trades": 3,
+    "stake_currency": "USDT",
+    "stake_amount": "unlimited"
+}
+
+

If multiple files are in the add_config_files section, then they will be assumed to be at identical levels, having the last occurrence override the earlier config (unless a parent already defined such a key).

+
+

Configuration parameters

+

The table below will list all configuration parameters available.

+

Freqtrade can also load many options via command line (CLI) arguments (check out the commands --help output for details).

+

Configuration option prevalence

+

The prevalence for all Options is as follows:

+
    +
  • CLI arguments override any other option
  • +
  • Environment Variables
  • +
  • Configuration files are used in sequence (the last file wins) and override Strategy configurations.
  • +
  • Strategy configurations are only used if they are not set via configuration or command-line arguments. These options are marked with Strategy Override in the below table.
  • +
+

Parameters table

+

Mandatory parameters are marked as Required, which means that they are required to be set in one of the possible ways.

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
ParameterDescription
max_open_tradesRequired. Number of open trades your bot is allowed to have. Only one open trade per pair is possible, so the length of your pairlist is another limitation that can apply. If -1 then it is ignored (i.e. potentially unlimited open trades, limited by the pairlist). More information below. Strategy Override.
Datatype: Positive integer or -1.
stake_currencyRequired. Crypto-currency used for trading.
Datatype: String
stake_amountRequired. Amount of crypto-currency your bot will use for each trade. Set it to "unlimited" to allow the bot to use all available balance. More information below.
Datatype: Positive float or "unlimited".
tradable_balance_ratioRatio of the total account balance the bot is allowed to trade. More information below.
Defaults to 0.99 99%).
Datatype: Positive float between 0.1 and 1.0.
available_capitalAvailable starting capital for the bot. Useful when running multiple bots on the same exchange account. More information below.
Datatype: Positive float.
amend_last_stake_amountUse reduced last stake amount if necessary. More information below.
Defaults to false.
Datatype: Boolean
last_stake_amount_min_ratioDefines minimum stake amount that has to be left and executed. Applies only to the last stake amount when it's amended to a reduced value (i.e. if amend_last_stake_amount is set to true). More information below.
Defaults to 0.5.
Datatype: Float (as ratio)
amount_reserve_percentReserve some amount in min pair stake amount. The bot will reserve amount_reserve_percent + stoploss value when calculating min pair stake amount in order to avoid possible trade refusals.
Defaults to 0.05 (5%).
Datatype: Positive Float as ratio.
timeframeThe timeframe to use (e.g 1m, 5m, 15m, 30m, 1h ...). Usually missing in configuration, and specified in the strategy. Strategy Override.
Datatype: String
fiat_display_currencyFiat currency used to show your profits. More information below.
Datatype: String
dry_runRequired. Define if the bot must be in Dry Run or production mode.
Defaults to true.
Datatype: Boolean
dry_run_walletDefine the starting amount in stake currency for the simulated wallet used by the bot running in Dry Run mode.
Defaults to 1000.
Datatype: Float
cancel_open_orders_on_exitCancel open orders when the /stop RPC command is issued, Ctrl+C is pressed or the bot dies unexpectedly. When set to true, this allows you to use /stop to cancel unfilled and partially filled orders in the event of a market crash. It does not impact open positions.
Defaults to false.
Datatype: Boolean
process_only_new_candlesEnable processing of indicators only when new candles arrive. If false each loop populates the indicators, this will mean the same candle is processed many times creating system load but can be useful of your strategy depends on tick data not only candle. Strategy Override.
Defaults to true.
Datatype: Boolean
minimal_roiRequired. Set the threshold as ratio the bot will use to exit a trade. More information below. Strategy Override.
Datatype: Dict
stoplossRequired. Value as ratio of the stoploss used by the bot. More details in the stoploss documentation. Strategy Override.
Datatype: Float (as ratio)
trailing_stopEnables trailing stoploss (based on stoploss in either configuration or strategy file). More details in the stoploss documentation. Strategy Override.
Datatype: Boolean
trailing_stop_positiveChanges stoploss once profit has been reached. More details in the stoploss documentation. Strategy Override.
Datatype: Float
trailing_stop_positive_offsetOffset on when to apply trailing_stop_positive. Percentage value which should be positive. More details in the stoploss documentation. Strategy Override.
Defaults to 0.0 (no offset).
Datatype: Float
trailing_only_offset_is_reachedOnly apply trailing stoploss when the offset is reached. stoploss documentation. Strategy Override.
Defaults to false.
Datatype: Boolean
feeFee used during backtesting / dry-runs. Should normally not be configured, which has freqtrade fall back to the exchange default fee. Set as ratio (e.g. 0.001 = 0.1%). Fee is applied twice for each trade, once when buying, once when selling.
Datatype: Float (as ratio)
futures_funding_rateUser-specified funding rate to be used when historical funding rates are not available from the exchange. This does not overwrite real historical rates. It is recommended that this be set to 0 unless you are testing a specific coin and you understand how the funding rate will affect freqtrade's profit calculations. More information here
Defaults to None.
Datatype: Float
trading_modeSpecifies if you want to trade regularly, trade with leverage, or trade contracts whose prices are derived from matching cryptocurrency prices. leverage documentation.
Defaults to "spot".
Datatype: String
margin_modeWhen trading with leverage, this determines if the collateral owned by the trader will be shared or isolated to each trading pair leverage documentation.
Datatype: String
liquidation_bufferA ratio specifying how large of a safety net to place between the liquidation price and the stoploss to prevent a position from reaching the liquidation price leverage documentation.
Defaults to 0.05.
Datatype: Float
Unfilled timeout
unfilledtimeout.entryRequired. How long (in minutes or seconds) the bot will wait for an unfilled entry order to complete, after which the order will be cancelled and repeated at current (new) price, as long as there is a signal. Strategy Override.
Datatype: Integer
unfilledtimeout.exitRequired. How long (in minutes or seconds) the bot will wait for an unfilled exit order to complete, after which the order will be cancelled and repeated at current (new) price, as long as there is a signal. Strategy Override.
Datatype: Integer
unfilledtimeout.unitUnit to use in unfilledtimeout setting. Note: If you set unfilledtimeout.unit to "seconds", "internals.process_throttle_secs" must be inferior or equal to timeout Strategy Override.
Defaults to "minutes".
Datatype: String
unfilledtimeout.exit_timeout_countHow many times can exit orders time out. Once this number of timeouts is reached, an emergency exit is triggered. 0 to disable and allow unlimited order cancels. Strategy Override.
Defaults to 0.
Datatype: Integer
Pricing
entry_pricing.price_sideSelect the side of the spread the bot should look at to get the entry rate. More information below.
Defaults to "same".
Datatype: String (either ask, bid, same or other).
entry_pricing.price_last_balanceRequired. Interpolate the bidding price. More information below.
entry_pricing.use_order_bookEnable entering using the rates in Order Book Entry.
Defaults to true.
Datatype: Boolean
entry_pricing.order_book_topBot will use the top N rate in Order Book "price_side" to enter a trade. I.e. a value of 2 will allow the bot to pick the 2nd entry in Order Book Entry.
Defaults to 1.
Datatype: Positive Integer
entry_pricing. check_depth_of_market.enabledDo not enter if the difference of buy orders and sell orders is met in Order Book. Check market depth.
Defaults to false.
Datatype: Boolean
entry_pricing. check_depth_of_market.bids_to_ask_deltaThe difference ratio of buy orders and sell orders found in Order Book. A value below 1 means sell order size is greater, while value greater than 1 means buy order size is higher. Check market depth
Defaults to 0.
Datatype: Float (as ratio)
exit_pricing.price_sideSelect the side of the spread the bot should look at to get the exit rate. More information below.
Defaults to "same".
Datatype: String (either ask, bid, same or other).
exit_pricing.price_last_balanceInterpolate the exiting price. More information below.
exit_pricing.use_order_bookEnable exiting of open trades using Order Book Exit.
Defaults to true.
Datatype: Boolean
exit_pricing.order_book_topBot will use the top N rate in Order Book "price_side" to exit. I.e. a value of 2 will allow the bot to pick the 2nd ask rate in Order Book Exit
Defaults to 1.
Datatype: Positive Integer
custom_price_max_distance_ratioConfigure maximum distance ratio between current and custom entry or exit price.
Defaults to 0.02 2%).
Datatype: Positive float
TODO
use_exit_signalUse exit signals produced by the strategy in addition to the minimal_roi. Strategy Override.
Defaults to true.
Datatype: Boolean
exit_profit_onlyWait until the bot reaches exit_profit_offset before taking an exit decision. Strategy Override.
Defaults to false.
Datatype: Boolean
exit_profit_offsetExit-signal is only active above this value. Only active in combination with exit_profit_only=True. Strategy Override.
Defaults to 0.0.
Datatype: Float (as ratio)
ignore_roi_if_entry_signalDo not exit if the entry signal is still active. This setting takes preference over minimal_roi and use_exit_signal. Strategy Override.
Defaults to false.
Datatype: Boolean
ignore_buying_expired_candle_afterSpecifies the number of seconds until a buy signal is no longer used.
Datatype: Integer
order_typesConfigure order-types depending on the action ("entry", "exit", "stoploss", "stoploss_on_exchange"). More information below. Strategy Override.
Datatype: Dict
order_time_in_forceConfigure time in force for entry and exit orders. More information below. Strategy Override.
Datatype: Dict
position_adjustment_enableEnables the strategy to use position adjustments (additional buys or sells). More information here.
Strategy Override.
Defaults to false.
Datatype: Boolean
max_entry_position_adjustmentMaximum additional order(s) for each open trade on top of the first entry Order. Set it to -1 for unlimited additional orders. More information here.
Strategy Override.
Defaults to -1.
Datatype: Positive Integer or -1
Exchange
exchange.nameRequired. Name of the exchange class to use. List below.
Datatype: String
exchange.sandboxUse the 'sandbox' version of the exchange, where the exchange provides a sandbox for risk-free integration. See here in more details.
Datatype: Boolean
exchange.keyAPI key to use for the exchange. Only required when you are in production mode.
Keep it in secret, do not disclose publicly.
Datatype: String
exchange.secretAPI secret to use for the exchange. Only required when you are in production mode.
Keep it in secret, do not disclose publicly.
Datatype: String
exchange.passwordAPI password to use for the exchange. Only required when you are in production mode and for exchanges that use password for API requests.
Keep it in secret, do not disclose publicly.
Datatype: String
exchange.uidAPI uid to use for the exchange. Only required when you are in production mode and for exchanges that use uid for API requests.
Keep it in secret, do not disclose publicly.
Datatype: String
exchange.pair_whitelistList of pairs to use by the bot for trading and to check for potential trades during backtesting. Supports regex pairs as .*/BTC. Not used by VolumePairList. More information.
Datatype: List
exchange.pair_blacklistList of pairs the bot must absolutely avoid for trading and backtesting. More information.
Datatype: List
exchange.ccxt_configAdditional CCXT parameters passed to both ccxt instances (sync and async). This is usually the correct place for additional ccxt configurations. Parameters may differ from exchange to exchange and are documented in the ccxt documentation. Please avoid adding exchange secrets here (use the dedicated fields instead), as they may be contained in logs.
Datatype: Dict
exchange.ccxt_sync_configAdditional CCXT parameters passed to the regular (sync) ccxt instance. Parameters may differ from exchange to exchange and are documented in the ccxt documentation
Datatype: Dict
exchange.ccxt_async_configAdditional CCXT parameters passed to the async ccxt instance. Parameters may differ from exchange to exchange and are documented in the ccxt documentation
Datatype: Dict
exchange.markets_refresh_intervalThe interval in minutes in which markets are reloaded.
Defaults to 60 minutes.
Datatype: Positive Integer
exchange.skip_pair_validationSkip pairlist validation on startup.
Defaults to false
Datatype: Boolean
exchange.skip_open_order_updateSkips open order updates on startup should the exchange cause problems. Only relevant in live conditions.
Defaults to false
Datatype: Boolean
exchange.unknown_fee_rateFallback value to use when calculating trading fees. This can be useful for exchanges which have fees in non-tradable currencies. The value provided here will be multiplied with the "fee cost".
Defaults to None
Datatype:* float
exchange.log_responsesLog relevant exchange responses. For debug mode only - use with care.
Defaults to false
Datatype: Boolean
experimental.block_bad_exchangesBlock exchanges known to not work with freqtrade. Leave on default unless you want to test if that exchange works now.
Defaults to true.
Datatype: Boolean
Plugins
edge.*Please refer to edge configuration document for detailed explanation of all possible configuration options.
pairlistsDefine one or more pairlists to be used. More information.
Defaults to StaticPairList.
Datatype: List of Dicts
protectionsDefine one or more protections to be used. More information.
Datatype: List of Dicts
Telegram
telegram.enabledEnable the usage of Telegram.
Datatype: Boolean
telegram.tokenYour Telegram bot token. Only required if telegram.enabled is true.
Keep it in secret, do not disclose publicly.
Datatype: String
telegram.chat_idYour personal Telegram account id. Only required if telegram.enabled is true.
Keep it in secret, do not disclose publicly.
Datatype: String
telegram.balance_dust_levelDust-level (in stake currency) - currencies with a balance below this will not be shown by /balance.
Datatype: float
telegram.reloadAllow "reload" buttons on telegram messages.
Defaults to true.
Datatype:* boolean
telegram.notification_settings.*Detailed notification settings. Refer to the telegram documentation for details.
Datatype: dictionary
telegram.allow_custom_messagesEnable the sending of Telegram messages from strategies via the dataprovider.send_msg() function.
Datatype: Boolean
Webhook
webhook.enabledEnable usage of Webhook notifications
Datatype: Boolean
webhook.urlURL for the webhook. Only required if webhook.enabled is true. See the webhook documentation for more details.
Datatype: String
webhook.entryPayload to send on entry. Only required if webhook.enabled is true. See the webhook documentation for more details.
Datatype: String
webhook.entry_cancelPayload to send on entry order cancel. Only required if webhook.enabled is true. See the webhook documentation for more details.
Datatype: String
webhook.entry_fillPayload to send on entry order filled. Only required if webhook.enabled is true. See the webhook documentation for more details.
Datatype: String
webhook.exitPayload to send on exit. Only required if webhook.enabled is true. See the webhook documentation for more details.
Datatype: String
webhook.exit_cancelPayload to send on exit order cancel. Only required if webhook.enabled is true. See the webhook documentation for more details.
Datatype: String
webhook.exit_fillPayload to send on exit order filled. Only required if webhook.enabled is true. See the webhook documentation for more details.
Datatype: String
webhook.statusPayload to send on status calls. Only required if webhook.enabled is true. See the webhook documentation for more details.
Datatype: String
webhook.allow_custom_messagesEnable the sending of Webhook messages from strategies via the dataprovider.send_msg() function.
Datatype: Boolean
Rest API / FreqUI / Producer-Consumer
api_server.enabledEnable usage of API Server. See the API Server documentation for more details.
Datatype: Boolean
api_server.listen_ip_addressBind IP address. See the API Server documentation for more details.
Datatype: IPv4
api_server.listen_portBind Port. See the API Server documentation for more details.
Datatype: Integer between 1024 and 65535
api_server.verbosityLogging verbosity. info will print all RPC Calls, while "error" will only display errors.
Datatype: Enum, either info or error. Defaults to info.
api_server.usernameUsername for API server. See the API Server documentation for more details.
Keep it in secret, do not disclose publicly.
Datatype: String
api_server.passwordPassword for API server. See the API Server documentation for more details.
Keep it in secret, do not disclose publicly.
Datatype: String
api_server.ws_tokenAPI token for the Message WebSocket. See the API Server documentation for more details.
Keep it in secret, do not disclose publicly.
Datatype: String
bot_nameName of the bot. Passed via API to a client - can be shown to distinguish / name bots.
Defaults to freqtrade
Datatype: String
external_message_consumerEnable Producer/Consumer mode for more details.
Datatype: Dict
Other
initial_stateDefines the initial application state. If set to stopped, then the bot has to be explicitly started via /start RPC command.
Defaults to stopped.
Datatype: Enum, either stopped or running
force_entry_enableEnables the RPC Commands to force a Trade entry. More information below.
Datatype: Boolean
disable_dataframe_checksDisable checking the OHLCV dataframe returned from the strategy methods for correctness. Only use when intentionally changing the dataframe and understand what you are doing. Strategy Override.
Defaults to False.
Datatype: Boolean
internals.process_throttle_secsSet the process throttle, or minimum loop duration for one bot iteration loop. Value in second.
Defaults to 5 seconds.
Datatype: Positive Integer
internals.heartbeat_intervalPrint heartbeat message every N seconds. Set to 0 to disable heartbeat messages.
Defaults to 60 seconds.
Datatype: Positive Integer or 0
internals.sd_notifyEnables use of the sd_notify protocol to tell systemd service manager about changes in the bot state and issue keep-alive pings. See here for more details.
Datatype: Boolean
strategyRequired Defines Strategy class to use. Recommended to be set via --strategy NAME.
Datatype: ClassName
strategy_pathAdds an additional strategy lookup path (must be a directory).
Datatype: String
recursive_strategy_searchSet to true to recursively search sub-directories inside user_data/strategies for a strategy.
Datatype: Boolean
user_data_dirDirectory containing user data.
Defaults to ./user_data/.
Datatype: String
db_urlDeclares database URL to use. NOTE: This defaults to sqlite:///tradesv3.dryrun.sqlite if dry_run is true, and to sqlite:///tradesv3.sqlite for production instances.
Datatype: String, SQLAlchemy connect string
logfileSpecifies logfile name. Uses a rolling strategy for log file rotation for 10 files with the 1MB limit per file.
Datatype: String
add_config_filesAdditional config files. These files will be loaded and merged with the current config file. The files are resolved relative to the initial file.
Defaults to [].
Datatype: List of strings
dataformat_ohlcvData format to use to store historical candle (OHLCV) data.
Defaults to json.
Datatype: String
dataformat_tradesData format to use to store historical trades data.
Defaults to jsongz.
Datatype: String
reduce_df_footprintRecast all numeric columns to float32/int32, with the objective of reducing ram/disk usage (and decreasing train/inference timing in FreqAI). (Currently only affects FreqAI use-cases)
Datatype: Boolean.
Default: False.
+

Parameters in the strategy

+

The following parameters can be set in the configuration file or strategy. +Values set in the configuration file always overwrite values set in the strategy.

+
    +
  • minimal_roi
  • +
  • timeframe
  • +
  • stoploss
  • +
  • max_open_trades
  • +
  • trailing_stop
  • +
  • trailing_stop_positive
  • +
  • trailing_stop_positive_offset
  • +
  • trailing_only_offset_is_reached
  • +
  • use_custom_stoploss
  • +
  • process_only_new_candles
  • +
  • order_types
  • +
  • order_time_in_force
  • +
  • unfilledtimeout
  • +
  • disable_dataframe_checks
  • +
  • use_exit_signal
  • +
  • exit_profit_only
  • +
  • exit_profit_offset
  • +
  • ignore_roi_if_entry_signal
  • +
  • ignore_buying_expired_candle_after
  • +
  • position_adjustment_enable
  • +
  • max_entry_position_adjustment
  • +
+

Configuring amount per trade

+

There are several methods to configure how much of the stake currency the bot will use to enter a trade. All methods respect the available balance configuration as explained below.

+

Minimum trade stake

+

The minimum stake amount will depend on exchange and pair and is usually listed in the exchange support pages.

+

Assuming the minimum tradable amount for XRP/USD is 20 XRP (given by the exchange), and the price is 0.6\(, the minimum stake amount to buy this pair is 20 * 0.6 ~= 12. +This exchange has also a limit on USD - where all orders must be > 10\) - which however does not apply in this case.

+

To guarantee safe execution, freqtrade will not allow buying with a stake-amount of 10.1$, instead, it'll make sure that there's enough space to place a stoploss below the pair (+ an offset, defined by amount_reserve_percent, which defaults to 5%).

+

With a reserve of 5%, the minimum stake amount would be ~12.6$ (12 * (1 + 0.05)). If we take into account a stoploss of 10% on top of that - we'd end up with a value of ~14$ (12.6 / (1 - 0.1)).

+

To limit this calculation in case of large stoploss values, the calculated minimum stake-limit will never be more than 50% above the real limit.

+
+

Warning

+

Since the limits on exchanges are usually stable and are not updated often, some pairs can show pretty high minimum limits, simply because the price increased a lot since the last limit adjustment by the exchange. Freqtrade adjusts the stake-amount to this value, unless it's > 30% more than the calculated/desired stake-amount - in which case the trade is rejected.

+
+

Tradable balance

+

By default, the bot assumes that the complete amount - 1% is at it's disposal, and when using dynamic stake amount, it will split the complete balance into max_open_trades buckets per trade. +Freqtrade will reserve 1% for eventual fees when entering a trade and will therefore not touch that by default.

+

You can configure the "untouched" amount by using the tradable_balance_ratio setting.

+

For example, if you have 10 ETH available in your wallet on the exchange and tradable_balance_ratio=0.5 (which is 50%), then the bot will use a maximum amount of 5 ETH for trading and considers this as an available balance. The rest of the wallet is untouched by the trades.

+
+

Danger

+

This setting should not be used when running multiple bots on the same account. Please look at Available Capital to the bot instead.

+
+
+

Warning

+

The tradable_balance_ratio setting applies to the current balance (free balance + tied up in trades). Therefore, assuming the starting balance of 1000, a configuration with tradable_balance_ratio=0.99 will not guarantee that 10 currency units will always remain available on the exchange. For example, the free amount may reduce to 5 units if the total balance is reduced to 500 (either by a losing streak or by withdrawing balance).

+
+

Assign available Capital

+

To fully utilize compounding profits when using multiple bots on the same exchange account, you'll want to limit each bot to a certain starting balance. +This can be accomplished by setting available_capital to the desired starting balance.

+

Assuming your account has 10.000 USDT and you want to run 2 different strategies on this exchange. +You'd set available_capital=5000 - granting each bot an initial capital of 5000 USDT. +The bot will then split this starting balance equally into max_open_trades buckets. +Profitable trades will result in increased stake-sizes for this bot - without affecting the stake-sizes of the other bot.

+
+

Incompatible with tradable_balance_ratio

+

Setting this option will replace any configuration of tradable_balance_ratio.

+
+

Amend last stake amount

+

Assuming we have the tradable balance of 1000 USDT, stake_amount=400, and max_open_trades=3. +The bot would open 2 trades and will be unable to fill the last trading slot, since the requested 400 USDT are no longer available since 800 USDT are already tied in other trades.

+

To overcome this, the option amend_last_stake_amount can be set to True, which will enable the bot to reduce stake_amount to the available balance to fill the last trade slot.

+

In the example above this would mean:

+
    +
  • Trade1: 400 USDT
  • +
  • Trade2: 400 USDT
  • +
  • Trade3: 200 USDT
  • +
+
+

Note

+

This option only applies with Static stake amount - since Dynamic stake amount divides the balances evenly.

+
+
+

Note

+

The minimum last stake amount can be configured using last_stake_amount_min_ratio - which defaults to 0.5 (50%). This means that the minimum stake amount that's ever used is stake_amount * 0.5. This avoids very low stake amounts, that are close to the minimum tradable amount for the pair and can be refused by the exchange.

+
+

Static stake amount

+

The stake_amount configuration statically configures the amount of stake-currency your bot will use for each trade.

+

The minimal configuration value is 0.0001, however, please check your exchange's trading minimums for the stake currency you're using to avoid problems.

+

This setting works in combination with max_open_trades. The maximum capital engaged in trades is stake_amount * max_open_trades. +For example, the bot will at most use (0.05 BTC x 3) = 0.15 BTC, assuming a configuration of max_open_trades=3 and stake_amount=0.05.

+
+

Note

+

This setting respects the available balance configuration.

+
+

Dynamic stake amount

+

Alternatively, you can use a dynamic stake amount, which will use the available balance on the exchange, and divide that equally by the number of allowed trades (max_open_trades).

+

To configure this, set stake_amount="unlimited". We also recommend to set tradable_balance_ratio=0.99 (99%) - to keep a minimum balance for eventual fees.

+

In this case a trade amount is calculated as:

+
currency_balance / (max_open_trades - current_open_trades)
+
+

To allow the bot to trade all the available stake_currency in your account (minus tradable_balance_ratio) set

+
"stake_amount" : "unlimited",
+"tradable_balance_ratio": 0.99,
+
+
+

Compounding profits

+

This configuration will allow increasing/decreasing stakes depending on the performance of the bot (lower stake if the bot is losing, higher stakes if the bot has a winning record since higher balances are available), and will result in profit compounding.

+
+
+

When using Dry-Run Mode

+

When using "stake_amount" : "unlimited", in combination with Dry-Run, Backtesting or Hyperopt, the balance will be simulated starting with a stake of dry_run_wallet which will evolve. +It is therefore important to set dry_run_wallet to a sensible value (like 0.05 or 0.01 for BTC and 1000 or 100 for USDT, for example), otherwise, it may simulate trades with 100 BTC (or more) or 0.05 USDT (or less) at once - which may not correspond to your real available balance or is less than the exchange minimal limit for the order amount for the stake currency.

+
+

Dynamic stake amount with position adjustment

+

When you want to use position adjustment with unlimited stakes, you must also implement custom_stake_amount to a return a value depending on your strategy. +Typical value would be in the range of 25% - 50% of the proposed stakes, but depends highly on your strategy and how much you wish to leave into the wallet as position adjustment buffer.

+

For example if your position adjustment assumes it can do 2 additional buys with the same stake amounts then your buffer should be 66.6667% of the initially proposed unlimited stake amount.

+

Or another example if your position adjustment assumes it can do 1 additional buy with 3x the original stake amount then custom_stake_amount should return 25% of proposed stake amount and leave 75% for possible later position adjustments.

+

Prices used for orders

+

Prices for regular orders can be controlled via the parameter structures entry_pricing for trade entries and exit_pricing for trade exits. +Prices are always retrieved right before an order is placed, either by querying the exchange tickers or by using the orderbook data.

+
+

Note

+

Orderbook data used by Freqtrade are the data retrieved from exchange by the ccxt's function fetch_order_book(), i.e. are usually data from the L2-aggregated orderbook, while the ticker data are the structures returned by the ccxt's fetch_ticker()/fetch_tickers() functions. Refer to the ccxt library documentation for more details.

+
+
+

Using market orders

+

Please read the section Market order pricing section when using market orders.

+
+

Entry price

+

Enter price side

+

The configuration setting entry_pricing.price_side defines the side of the orderbook the bot looks for when buying.

+

The following displays an orderbook.

+
...
+103
+102
+101  # ask
+-------------Current spread
+99   # bid
+98
+97
+...
+
+

If entry_pricing.price_side is set to "bid", then the bot will use 99 as entry price.
+In line with that, if entry_pricing.price_side is set to "ask", then the bot will use 101 as entry price.

+

Depending on the order direction (long/short), this will lead to different results. Therefore we recommend to use "same" or "other" for this configuration instead. +This would result in the following pricing matrix:

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
directionOrdersettingpricecrosses spread
longbuyask101yes
longbuybid99no
longbuysame99no
longbuyother101yes
shortsellask101no
shortsellbid99yes
shortsellsame101no
shortsellother99yes
+

Using the other side of the orderbook often guarantees quicker filled orders, but the bot can also end up paying more than what would have been necessary. +Taker fees instead of maker fees will most likely apply even when using limit buy orders. +Also, prices at the "other" side of the spread are higher than prices at the "bid" side in the orderbook, so the order behaves similar to a market order (however with a maximum price).

+

Entry price with Orderbook enabled

+

When entering a trade with the orderbook enabled (entry_pricing.use_order_book=True), Freqtrade fetches the entry_pricing.order_book_top entries from the orderbook and uses the entry specified as entry_pricing.order_book_top on the configured side (entry_pricing.price_side) of the orderbook. 1 specifies the topmost entry in the orderbook, while 2 would use the 2nd entry in the orderbook, and so on.

+

Entry price without Orderbook enabled

+

The following section uses side as the configured entry_pricing.price_side (defaults to "same").

+

When not using orderbook (entry_pricing.use_order_book=False), Freqtrade uses the best side price from the ticker if it's below the last traded price from the ticker. Otherwise (when the side price is above the last price), it calculates a rate between side and last price based on entry_pricing.price_last_balance.

+

The entry_pricing.price_last_balance configuration parameter controls this. A value of 0.0 will use side price, while 1.0 will use the last price and values between those interpolate between ask and last price.

+

Check depth of market

+

When check depth of market is enabled (entry_pricing.check_depth_of_market.enabled=True), the entry signals are filtered based on the orderbook depth (sum of all amounts) for each orderbook side.

+

Orderbook bid (buy) side depth is then divided by the orderbook ask (sell) side depth and the resulting delta is compared to the value of the entry_pricing.check_depth_of_market.bids_to_ask_delta parameter. The entry order is only executed if the orderbook delta is greater than or equal to the configured delta value.

+
+

Note

+

A delta value below 1 means that ask (sell) orderbook side depth is greater than the depth of the bid (buy) orderbook side, while a value greater than 1 means opposite (depth of the buy side is higher than the depth of the sell side).

+
+

Exit price

+

Exit price side

+

The configuration setting exit_pricing.price_side defines the side of the spread the bot looks for when exiting a trade.

+

The following displays an orderbook:

+
...
+103
+102
+101  # ask
+-------------Current spread
+99   # bid
+98
+97
+...
+
+

If exit_pricing.price_side is set to "ask", then the bot will use 101 as exiting price.
+In line with that, if exit_pricing.price_side is set to "bid", then the bot will use 99 as exiting price.

+

Depending on the order direction (long/short), this will lead to different results. Therefore we recommend to use "same" or "other" for this configuration instead. +This would result in the following pricing matrix:

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
DirectionOrdersettingpricecrosses spread
longsellask101no
longsellbid99yes
longsellsame101no
longsellother99yes
shortbuyask101yes
shortbuybid99no
shortbuysame99no
shortbuyother101yes
+

Exit price with Orderbook enabled

+

When exiting with the orderbook enabled (exit_pricing.use_order_book=True), Freqtrade fetches the exit_pricing.order_book_top entries in the orderbook and uses the entry specified as exit_pricing.order_book_top from the configured side (exit_pricing.price_side) as trade exit price.

+

1 specifies the topmost entry in the orderbook, while 2 would use the 2nd entry in the orderbook, and so on.

+

Exit price without Orderbook enabled

+

The following section uses side as the configured exit_pricing.price_side (defaults to "ask").

+

When not using orderbook (exit_pricing.use_order_book=False), Freqtrade uses the best side price from the ticker if it's above the last traded price from the ticker. Otherwise (when the side price is below the last price), it calculates a rate between side and last price based on exit_pricing.price_last_balance.

+

The exit_pricing.price_last_balance configuration parameter controls this. A value of 0.0 will use side price, while 1.0 will use the last price and values between those interpolate between side and last price.

+

Market order pricing

+

When using market orders, prices should be configured to use the "correct" side of the orderbook to allow realistic pricing detection. +Assuming both entry and exits are using market orders, a configuration similar to the following must be used

+
  "order_types": {
+    "entry": "market",
+    "exit": "market"
+    // ...
+  },
+  "entry_pricing": {
+    "price_side": "other",
+    // ...
+  },
+  "exit_pricing":{
+    "price_side": "other",
+    // ...
+  },
+
+

Obviously, if only one side is using limit orders, different pricing combinations can be used.

+

Understand minimal_roi

+

The minimal_roi configuration parameter is a JSON object where the key is a duration +in minutes and the value is the minimum ROI as a ratio. +See the example below:

+
"minimal_roi": {
+    "40": 0.0,    # Exit after 40 minutes if the profit is not negative
+    "30": 0.01,   # Exit after 30 minutes if there is at least 1% profit
+    "20": 0.02,   # Exit after 20 minutes if there is at least 2% profit
+    "0":  0.04    # Exit immediately if there is at least 4% profit
+},
+
+

Most of the strategy files already include the optimal minimal_roi value. +This parameter can be set in either Strategy or Configuration file. If you use it in the configuration file, it will override the +minimal_roi value from the strategy file. +If it is not set in either Strategy or Configuration, a default of 1000% {"0": 10} is used, and minimal ROI is disabled unless your trade generates 1000% profit.

+
+

Special case to forceexit after a specific time

+

A special case presents using "<N>": -1 as ROI. This forces the bot to exit a trade after N Minutes, no matter if it's positive or negative, so represents a time-limited force-exit.

+
+

Understand force_entry_enable

+

The force_entry_enable configuration parameter enables the usage of force-enter (/forcelong, /forceshort) commands via Telegram and REST API. +For security reasons, it's disabled by default, and freqtrade will show a warning message on startup if enabled. +For example, you can send /forceenter ETH/BTC to the bot, which will result in freqtrade buying the pair and holds it until a regular exit-signal (ROI, stoploss, /forceexit) appears.

+

This can be dangerous with some strategies, so use with care.

+

See the telegram documentation for details on usage.

+

Ignoring expired candles

+

When working with larger timeframes (for example 1h or more) and using a low max_open_trades value, the last candle can be processed as soon as a trade slot becomes available. When processing the last candle, this can lead to a situation where it may not be desirable to use the buy signal on that candle. For example, when using a condition in your strategy where you use a cross-over, that point may have passed too long ago for you to start a trade on it.

+

In these situations, you can enable the functionality to ignore candles that are beyond a specified period by setting ignore_buying_expired_candle_after to a positive number, indicating the number of seconds after which the buy signal becomes expired.

+

For example, if your strategy is using a 1h timeframe, and you only want to buy within the first 5 minutes when a new candle comes in, you can add the following configuration to your strategy:

+
  {
+    //...
+    "ignore_buying_expired_candle_after": 300,
+    // ...
+  }
+
+
+

Note

+

This setting resets with each new candle, so it will not prevent sticking-signals from executing on the 2nd or 3rd candle they're active. Best use a "trigger" selector for buy signals, which are only active for one candle.

+
+

Understand order_types

+

The order_types configuration parameter maps actions (entry, exit, stoploss, emergency_exit, force_exit, force_entry) to order-types (market, limit, ...) as well as configures stoploss to be on the exchange and defines stoploss on exchange update interval in seconds.

+

This allows to enter using limit orders, exit using limit-orders, and create stoplosses using market orders. +It also allows to set the +stoploss "on exchange" which means stoploss order would be placed immediately once the buy order is fulfilled.

+

order_types set in the configuration file overwrites values set in the strategy as a whole, so you need to configure the whole order_types dictionary in one place.

+

If this is configured, the following 4 values (entry, exit, stoploss and stoploss_on_exchange) need to be present, otherwise, the bot will fail to start.

+

For information on (emergency_exit,force_exit, force_entry, stoploss_on_exchange,stoploss_on_exchange_interval,stoploss_on_exchange_limit_ratio) please see stop loss documentation stop loss on exchange

+

Syntax for Strategy:

+
order_types = {
+    "entry": "limit",
+    "exit": "limit",
+    "emergency_exit": "market",
+    "force_entry": "market",
+    "force_exit": "market",
+    "stoploss": "market",
+    "stoploss_on_exchange": False,
+    "stoploss_on_exchange_interval": 60,
+    "stoploss_on_exchange_limit_ratio": 0.99,
+}
+
+

Configuration:

+
"order_types": {
+    "entry": "limit",
+    "exit": "limit",
+    "emergency_exit": "market",
+    "force_entry": "market",
+    "force_exit": "market",
+    "stoploss": "market",
+    "stoploss_on_exchange": false,
+    "stoploss_on_exchange_interval": 60
+}
+
+
+

Market order support

+

Not all exchanges support "market" orders. +The following message will be shown if your exchange does not support market orders: +"Exchange <yourexchange> does not support market orders." and the bot will refuse to start.

+
+
+

Using market orders

+

Please carefully read the section Market order pricing section when using market orders.

+
+
+

Stoploss on exchange

+

stoploss_on_exchange_interval is not mandatory. Do not change its value if you are +unsure of what you are doing. For more information about how stoploss works please +refer to the stoploss documentation.

+

If stoploss_on_exchange is enabled and the stoploss is cancelled manually on the exchange, then the bot will create a new stoploss order.

+
+
+

Warning: stoploss_on_exchange failures

+

If stoploss on exchange creation fails for some reason, then an "emergency exit" is initiated. By default, this will exit the trade using a market order. The order-type for the emergency-exit can be changed by setting the emergency_exit value in the order_types dictionary - however, this is not advised.

+
+

Understand order_time_in_force

+

The order_time_in_force configuration parameter defines the policy by which the order +is executed on the exchange. Three commonly used time in force are:

+

GTC (Good Till Canceled):

+

This is most of the time the default time in force. It means the order will remain +on exchange till it is cancelled by the user. It can be fully or partially fulfilled. +If partially fulfilled, the remaining will stay on the exchange till cancelled.

+

FOK (Fill Or Kill):

+

It means if the order is not executed immediately AND fully then it is cancelled by the exchange.

+

IOC (Immediate Or Canceled):

+

It is the same as FOK (above) except it can be partially fulfilled. The remaining part +is automatically cancelled by the exchange.

+

PO (Post only):

+

Post only order. The order is either placed as a maker order, or it is canceled. +This means the order must be placed on orderbook for at at least time in an unfilled state.

+

time_in_force config

+

The order_time_in_force parameter contains a dict with entry and exit time in force policy values. +This can be set in the configuration file or in the strategy. +Values set in the configuration file overwrites values set in the strategy.

+

The possible values are: GTC (default), FOK or IOC.

+
"order_time_in_force": {
+    "entry": "GTC",
+    "exit": "GTC"
+},
+
+
+

Warning

+

This is ongoing work. For now, it is supported only for binance, gate and kucoin. +Please don't change the default value unless you know what you are doing and have researched the impact of using different values for your particular exchange.

+
+

What values can be used for fiat_display_currency?

+

The fiat_display_currency configuration parameter sets the base currency to use for the +conversion from coin to fiat in the bot Telegram reports.

+

The valid values are:

+
"AUD", "BRL", "CAD", "CHF", "CLP", "CNY", "CZK", "DKK", "EUR", "GBP", "HKD", "HUF", "IDR", "ILS", "INR", "JPY", "KRW", "MXN", "MYR", "NOK", "NZD", "PHP", "PKR", "PLN", "RUB", "SEK", "SGD", "THB", "TRY", "TWD", "ZAR", "USD"
+
+

In addition to fiat currencies, a range of crypto currencies is supported.

+

The valid values are:

+
"BTC", "ETH", "XRP", "LTC", "BCH", "USDT"
+
+

Using Dry-run mode

+

We recommend starting the bot in the Dry-run mode to see how your bot will +behave and what is the performance of your strategy. In the Dry-run mode, the +bot does not engage your money. It only runs a live simulation without +creating trades on the exchange.

+
    +
  1. Edit your config.json configuration file.
  2. +
  3. Switch dry-run to true and specify db_url for a persistence database.
  4. +
+
"dry_run": true,
+"db_url": "sqlite:///tradesv3.dryrun.sqlite",
+
+
    +
  1. Remove your Exchange API key and secret (change them by empty values or fake credentials):
  2. +
+
"exchange": {
+    "name": "bittrex",
+    "key": "key",
+    "secret": "secret",
+    ...
+}
+
+

Once you will be happy with your bot performance running in the Dry-run mode, you can switch it to production mode.

+
+

Note

+

A simulated wallet is available during dry-run mode and will assume a starting capital of dry_run_wallet (defaults to 1000).

+
+

Considerations for dry-run

+
    +
  • API-keys may or may not be provided. Only Read-Only operations (i.e. operations that do not alter account state) on the exchange are performed in dry-run mode.
  • +
  • Wallets (/balance) are simulated based on dry_run_wallet.
  • +
  • Orders are simulated, and will not be posted to the exchange.
  • +
  • Market orders fill based on orderbook volume the moment the order is placed.
  • +
  • Limit orders fill once the price reaches the defined level - or time out based on unfilledtimeout settings.
  • +
  • In combination with stoploss_on_exchange, the stop_loss price is assumed to be filled.
  • +
  • Open orders (not trades, which are stored in the database) are kept open after bot restarts, with the assumption that they were not filled while being offline.
  • +
+

Switch to production mode

+

In production mode, the bot will engage your money. Be careful, since a wrong strategy can lose all your money. +Be aware of what you are doing when you run it in production mode.

+

When switching to Production mode, please make sure to use a different / fresh database to avoid dry-run trades messing with your exchange money and eventually tainting your statistics.

+

Setup your exchange account

+

You will need to create API Keys (usually you get key and secret, some exchanges require an additional password) from the Exchange website and you'll need to insert this into the appropriate fields in the configuration or when asked by the freqtrade new-config command. +API Keys are usually only required for live trading (trading for real money, bot running in "production mode", executing real orders on the exchange) and are not required for the bot running in dry-run (trade simulation) mode. When you set up the bot in dry-run mode, you may fill these fields with empty values.

+

To switch your bot in production mode

+

Edit your config.json file.

+

Switch dry-run to false and don't forget to adapt your database URL if set:

+
"dry_run": false,
+
+

Insert your Exchange API key (change them by fake API keys):

+
{
+    "exchange": {
+        "name": "bittrex",
+        "key": "af8ddd35195e9dc500b9a6f799f6f5c93d89193b",
+        "secret": "08a9dc6db3d7b53e1acebd9275677f4b0a04f1a5",
+        //"password": "", // Optional, not needed by all exchanges)
+        // ...
+    }
+    //...
+}
+
+

You should also make sure to read the Exchanges section of the documentation to be aware of potential configuration details specific to your exchange.

+
+

Keep your secrets secret

+

To keep your secrets secret, we recommend using a 2nd configuration for your API keys. +Simply use the above snippet in a new configuration file (e.g. config-private.json) and keep your settings in this file. +You can then start the bot with freqtrade trade --config user_data/config.json --config user_data/config-private.json <...> to have your keys loaded.

+

NEVER share your private configuration file or your exchange keys with anyone!

+
+

Using proxy with Freqtrade

+

To use a proxy with freqtrade, export your proxy settings using the variables "HTTP_PROXY" and "HTTPS_PROXY" set to the appropriate values. +This will have the proxy settings applied to everything (telegram, coingecko, ...) except for exchange requests.

+
export HTTP_PROXY="http://addr:port"
+export HTTPS_PROXY="http://addr:port"
+freqtrade
+
+

Proxy exchange requests

+

To use a proxy for exchange connections - you will have to define the proxies as part of the ccxt configuration.

+
{ 
+  "exchange": {
+    "ccxt_config": {
+      "aiohttp_proxy": "http://addr:port",
+      "proxies": {
+        "http": "http://addr:port",
+        "https": "http://addr:port"
+      },
+    }
+  }
+}
+
+

Next step

+

Now you have configured your config.json, the next step is to start your bot.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/data-analysis/index.html b/en/2023.4/data-analysis/index.html new file mode 100644 index 000000000..f22d1bfd6 --- /dev/null +++ b/en/2023.4/data-analysis/index.html @@ -0,0 +1,1502 @@ + + + + + + + + + + + + + + + + + + + + + + + + Jupyter Notebooks - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Analyzing bot data with Jupyter notebooks

+

You can analyze the results of backtests and trading history easily using Jupyter notebooks. Sample notebooks are located at user_data/notebooks/ after initializing the user directory with freqtrade create-userdir --userdir user_data.

+

Quick start with docker

+

Freqtrade provides a docker-compose file which starts up a jupyter lab server. +You can run this server using the following command: docker compose -f docker/docker-compose-jupyter.yml up

+

This will create a dockercontainer running jupyter lab, which will be accessible using https://127.0.0.1:8888/lab. +Please use the link that's printed in the console after startup for simplified login.

+

For more information, Please visit the Data analysis with Docker section.

+

Pro tips

+
    +
  • See jupyter.org for usage instructions.
  • +
  • Don't forget to start a Jupyter notebook server from within your conda or venv environment or use nb_conda_kernels*
  • +
  • Copy the example notebook before use so your changes don't get overwritten with the next freqtrade update.
  • +
+

Using virtual environment with system-wide Jupyter installation

+

Sometimes it can be desired to use a system-wide installation of Jupyter notebook, and use a jupyter kernel from the virtual environment. +This prevents you from installing the full jupyter suite multiple times per system, and provides an easy way to switch between tasks (freqtrade / other analytics tasks).

+

For this to work, first activate your virtual environment and run the following commands:

+
# Activate virtual environment
+source .env/bin/activate
+
+pip install ipykernel
+ipython kernel install --user --name=freqtrade
+# Restart jupyter (lab / notebook)
+# select kernel "freqtrade" in the notebook
+
+
+

Note

+

This section is provided for completeness, the Freqtrade Team won't provide full support for problems with this setup and will recommend to install Jupyter in the virtual environment directly, as that is the easiest way to get jupyter notebooks up and running. For help with this setup please refer to the Project Jupyter documentation or help channels.

+
+
+

Warning

+

Some tasks don't work especially well in notebooks. For example, anything using asynchronous execution is a problem for Jupyter. Also, freqtrade's primary entry point is the shell cli, so using pure python in a notebook bypasses arguments that provide required objects and parameters to helper functions. You may need to set those values or create expected objects manually.

+
+ + + + + + + + + + + + + + + + + + + + + + +
TaskTool
Bot operationsCLI
Repetitive tasksShell scripts
Data analysis & visualizationNotebook
+
    +
  1. +

    Use the CLI to

    +

    * download historical data + * run a backtest + * run with real-time data + * export results

    +
  2. +
  3. +

    Collect these actions in shell scripts

    +

    * save complicated commands with arguments + * execute multi-step operations + * automate testing strategies and preparing data for analysis

    +
  4. +
  5. +

    Use a notebook to

    +

    * visualize data + * mangle and plot to generate insights

    +
  6. +
+

Example utility snippets

+

Change directory to root

+

Jupyter notebooks execute from the notebook directory. The following snippet searches for the project root, so relative paths remain consistent.

+
import os
+from pathlib import Path
+
+# Change directory
+# Modify this cell to insure that the output shows the correct path.
+# Define all paths relative to the project root shown in the cell output
+project_root = "somedir/freqtrade"
+i=0
+try:
+    os.chdir(project_root)
+    assert Path('LICENSE').is_file()
+except:
+    while i<4 and (not Path('LICENSE').is_file()):
+        os.chdir(Path(Path.cwd(), '../'))
+        i+=1
+    project_root = Path.cwd()
+print(Path.cwd())
+
+

Load multiple configuration files

+

This option can be useful to inspect the results of passing in multiple configs. +This will also run through the whole Configuration initialization, so the configuration is completely initialized to be passed to other methods.

+
import json
+from freqtrade.configuration import Configuration
+
+# Load config from multiple files
+config = Configuration.from_files(["config1.json", "config2.json"])
+
+# Show the config in memory
+print(json.dumps(config['original_config'], indent=2))
+
+

For Interactive environments, have an additional configuration specifying user_data_dir and pass this in last, so you don't have to change directories while running the bot. +Best avoid relative paths, since this starts at the storage location of the jupyter notebook, unless the directory is changed.

+
{
+    "user_data_dir": "~/.freqtrade/"
+}
+
+

Further Data analysis documentation

+ +

Feel free to submit an issue or Pull Request enhancing this document if you would like to share ideas on how to best analyze the data.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/data-download/index.html b/en/2023.4/data-download/index.html new file mode 100644 index 000000000..1bade4974 --- /dev/null +++ b/en/2023.4/data-download/index.html @@ -0,0 +1,2021 @@ + + + + + + + + + + + + + + + + + + + + + + + + Data Downloading - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Data Downloading

+

Getting data for backtesting and hyperopt

+

To download data (candles / OHLCV) needed for backtesting and hyperoptimization use the freqtrade download-data command.

+

If no additional parameter is specified, freqtrade will download data for "1m" and "5m" timeframes for the last 30 days. +Exchange and pairs will come from config.json (if specified using -c/--config). +Otherwise --exchange becomes mandatory.

+

You can use a relative timerange (--days 20) or an absolute starting point (--timerange 20200101-). For incremental downloads, the relative approach should be used.

+
+

Tip: Updating existing data

+

If you already have backtesting data available in your data-directory and would like to refresh this data up to today, freqtrade will automatically calculate the data missing for the existing pairs and the download will occur from the latest available point until "now", neither --days or --timerange parameters are required. Freqtrade will keep the available data and only download the missing data. +If you are updating existing data after inserting new pairs that you have no data for, use --new-pairs-days xx parameter. Specified number of days will be downloaded for new pairs while old pairs will be updated with missing data only. +If you use --days xx parameter alone - data for specified number of days will be downloaded for all pairs. Be careful, if specified number of days is smaller than gap between now and last downloaded candle - freqtrade will delete all existing data to avoid gaps in candle data.

+
+

Usage

+
usage: freqtrade download-data [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                               [-d PATH] [--userdir PATH]
+                               [-p PAIRS [PAIRS ...]] [--pairs-file FILE]
+                               [--days INT] [--new-pairs-days INT]
+                               [--include-inactive-pairs]
+                               [--timerange TIMERANGE] [--dl-trades]
+                               [--exchange EXCHANGE]
+                               [-t TIMEFRAMES [TIMEFRAMES ...]] [--erase]
+                               [--data-format-ohlcv {json,jsongz,hdf5,feather,parquet}]
+                               [--data-format-trades {json,jsongz,hdf5}]
+                               [--trading-mode {spot,margin,futures}]
+                               [--prepend]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]
+                        Limit command to these pairs. Pairs are space-
+                        separated.
+  --pairs-file FILE     File containing a list of pairs. Takes precedence over
+                        --pairs or pairs configured in the configuration.
+  --days INT            Download data for given number of days.
+  --new-pairs-days INT  Download data of new pairs for given number of days.
+                        Default: `None`.
+  --include-inactive-pairs
+                        Also download data from inactive pairs.
+  --timerange TIMERANGE
+                        Specify what timerange of data to use.
+  --dl-trades           Download trades instead of OHLCV data. The bot will
+                        resample trades to the desired timeframe as specified
+                        as --timeframes/-t.
+  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no
+                        config is provided.
+  -t TIMEFRAMES [TIMEFRAMES ...], --timeframes TIMEFRAMES [TIMEFRAMES ...]
+                        Specify which tickers to download. Space-separated
+                        list. Default: `1m 5m`.
+  --erase               Clean all existing data for the selected
+                        exchange/pairs/timeframes.
+  --data-format-ohlcv {json,jsongz,hdf5,feather,parquet}
+                        Storage format for downloaded candle (OHLCV) data.
+                        (default: `json`).
+  --data-format-trades {json,jsongz,hdf5}
+                        Storage format for downloaded trades data. (default:
+                        `jsongz`).
+  --trading-mode {spot,margin,futures}, --tradingmode {spot,margin,futures}
+                        Select Trading mode
+  --prepend             Allow data prepending. (Data-appending is disabled)
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH, --data-dir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+
+

Startup period

+

download-data is a strategy-independent command. The idea is to download a big chunk of data once, and then iteratively increase the amount of data stored.

+

For that reason, download-data does not care about the "startup-period" defined in a strategy. It's up to the user to download additional days if the backtest should start at a specific point in time (while respecting startup period).

+
+

Pairs file

+

In alternative to the whitelist from config.json, a pairs.json file can be used. +If you are using Binance for example:

+
    +
  • create a directory user_data/data/binance and copy or create the pairs.json file in that directory.
  • +
  • update the pairs.json file to contain the currency pairs you are interested in.
  • +
+
mkdir -p user_data/data/binance
+touch user_data/data/binance/pairs.json
+
+

The format of the pairs.json file is a simple json list. +Mixing different stake-currencies is allowed for this file, since it's only used for downloading.

+
[
+    "ETH/BTC",
+    "ETH/USDT",
+    "BTC/USDT",
+    "XRP/ETH"
+]
+
+
+

Downloading all data for one quote currency

+

Often, you'll want to download data for all pairs of a specific quote-currency. In such cases, you can use the following shorthand: +freqtrade download-data --exchange binance --pairs .*/USDT <...>. The provided "pairs" string will be expanded to contain all active pairs on the exchange. +To also download data for inactive (delisted) pairs, add --include-inactive-pairs to the command.

+
+
+Permission denied errors +

If your configuration directory user_data was made by docker, you may get the following error:

+
cp: cannot create regular file 'user_data/data/binance/pairs.json': Permission denied
+
+

You can fix the permissions of your user-data directory as follows:

+
sudo chown -R $UID:$GID user_data
+
+
+

Start download

+

Then run:

+
freqtrade download-data --exchange binance
+
+

This will download historical candle (OHLCV) data for all the currency pairs you defined in pairs.json.

+

Alternatively, specify the pairs directly

+
freqtrade download-data --exchange binance --pairs ETH/USDT XRP/USDT BTC/USDT
+
+

or as regex (to download all active USDT pairs)

+
freqtrade download-data --exchange binance --pairs .*/USDT
+
+

Other Notes

+
    +
  • To use a different directory than the exchange specific default, use --datadir user_data/data/some_directory.
  • +
  • To change the exchange used to download the historical data from, please use a different configuration file (you'll probably need to adjust rate limits etc.)
  • +
  • To use pairs.json from some other directory, use --pairs-file some_other_dir/pairs.json.
  • +
  • To download historical candle (OHLCV) data for only 10 days, use --days 10 (defaults to 30 days).
  • +
  • To download historical candle (OHLCV) data from a fixed starting point, use --timerange 20200101- - which will download all data from January 1st, 2020.
  • +
  • Use --timeframes to specify what timeframe download the historical candle (OHLCV) data for. Default is --timeframes 1m 5m which will download 1-minute and 5-minute data.
  • +
  • To use exchange, timeframe and list of pairs as defined in your configuration file, use the -c/--config option. With this, the script uses the whitelist defined in the config as the list of currency pairs to download data for and does not require the pairs.json file. You can combine -c/--config with most other options.
  • +
+

Download additional data before the current timerange

+

Assuming you downloaded all data from 2022 (--timerange 20220101-) - but you'd now like to also backtest with earlier data. +You can do so by using the --prepend flag, combined with --timerange - specifying an end-date.

+
freqtrade download-data --exchange binance --pairs ETH/USDT XRP/USDT BTC/USDT --prepend --timerange 20210101-20220101
+
+
+

Note

+

Freqtrade will ignore the end-date in this mode if data is available, updating the end-date to the existing data start point.

+
+

Data format

+

Freqtrade currently supports the following data-formats:

+
    +
  • json - plain "text" json files
  • +
  • jsongz - a gzip-zipped version of json files
  • +
  • hdf5 - a high performance datastore
  • +
  • feather - a dataformat based on Apache Arrow (OHLCV only)
  • +
  • parquet - columnar datastore (OHLCV only)
  • +
+

By default, OHLCV data is stored as json data, while trades data is stored as jsongz data.

+

This can be changed via the --data-format-ohlcv and --data-format-trades command line arguments respectively. +To persist this change, you should also add the following snippet to your configuration, so you don't have to insert the above arguments each time:

+
    // ...
+    "dataformat_ohlcv": "hdf5",
+    "dataformat_trades": "hdf5",
+    // ...
+
+

If the default data-format has been changed during download, then the keys dataformat_ohlcv and dataformat_trades in the configuration file need to be adjusted to the selected dataformat as well.

+
+

Note

+

You can convert between data-formats using the convert-data and convert-trade-data methods.

+
+

Dataformat comparison

+

The following comparisons have been made with the following data, and by using the linux time command.

+
Found 6 pair / timeframe combinations.
++----------+-------------+--------+---------------------+---------------------+
+|     Pair |   Timeframe |   Type |                From |                  To |
+|----------+-------------+--------+---------------------+---------------------|
+| BTC/USDT |          5m |   spot | 2017-08-17 04:00:00 | 2022-09-13 19:25:00 |
+| ETH/USDT |          1m |   spot | 2017-08-17 04:00:00 | 2022-09-13 19:26:00 |
+| BTC/USDT |          1m |   spot | 2017-08-17 04:00:00 | 2022-09-13 19:30:00 |
+| XRP/USDT |          5m |   spot | 2018-05-04 08:10:00 | 2022-09-13 19:15:00 |
+| XRP/USDT |          1m |   spot | 2018-05-04 08:11:00 | 2022-09-13 19:22:00 |
+| ETH/USDT |          5m |   spot | 2017-08-17 04:00:00 | 2022-09-13 19:20:00 |
++----------+-------------+--------+---------------------+---------------------+
+
+

Timings have been taken in a not very scientific way with the following command, which forces reading the data into memory.

+
time freqtrade list-data --show-timerange --data-format-ohlcv <dataformat>
+
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
FormatSizetiming
json149Mb25.6s
jsongz39Mb27s
hdf5145Mb3.9s
feather72Mb3.5s
parquet83Mb3.8s
+

Size has been taken from the BTC/USDT 1m spot combination for the timerange specified above.

+

To have a best performance/size mix, we recommend the use of either feather or parquet.

+

Sub-command convert data

+
usage: freqtrade convert-data [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                              [-d PATH] [--userdir PATH]
+                              [-p PAIRS [PAIRS ...]] --format-from
+                              {json,jsongz,hdf5,feather,parquet} --format-to
+                              {json,jsongz,hdf5,feather,parquet} [--erase]
+                              [--exchange EXCHANGE]
+                              [-t TIMEFRAMES [TIMEFRAMES ...]]
+                              [--trading-mode {spot,margin,futures}]
+                              [--candle-types {spot,futures,mark,index,premiumIndex,funding_rate} [{spot,futures,mark,index,premiumIndex,funding_rate} ...]]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]
+                        Limit command to these pairs. Pairs are space-
+                        separated.
+  --format-from {json,jsongz,hdf5,feather,parquet}
+                        Source format for data conversion.
+  --format-to {json,jsongz,hdf5,feather,parquet}
+                        Destination format for data conversion.
+  --erase               Clean all existing data for the selected
+                        exchange/pairs/timeframes.
+  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no
+                        config is provided.
+  -t TIMEFRAMES [TIMEFRAMES ...], --timeframes TIMEFRAMES [TIMEFRAMES ...]
+                        Specify which tickers to download. Space-separated
+                        list. Default: `1m 5m`.
+  --trading-mode {spot,margin,futures}, --tradingmode {spot,margin,futures}
+                        Select Trading mode
+  --candle-types {spot,futures,mark,index,premiumIndex,funding_rate} [{spot,futures,mark,index,premiumIndex,funding_rate} ...]
+                        Select candle type to use
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH, --data-dir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+
Example converting data
+

The following command will convert all candle (OHLCV) data available in ~/.freqtrade/data/binance from json to jsongz, saving diskspace in the process. +It'll also remove original json data files (--erase parameter).

+
freqtrade convert-data --format-from json --format-to jsongz --datadir ~/.freqtrade/data/binance -t 5m 15m --erase
+
+

Sub-command convert trade data

+
usage: freqtrade convert-trade-data [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                                    [-d PATH] [--userdir PATH]
+                                    [-p PAIRS [PAIRS ...]] --format-from
+                                    {json,jsongz,hdf5,feather,parquet}
+                                    --format-to
+                                    {json,jsongz,hdf5,feather,parquet}
+                                    [--erase] [--exchange EXCHANGE]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]
+                        Limit command to these pairs. Pairs are space-
+                        separated.
+  --format-from {json,jsongz,hdf5,feather,parquet}
+                        Source format for data conversion.
+  --format-to {json,jsongz,hdf5,feather,parquet}
+                        Destination format for data conversion.
+  --erase               Clean all existing data for the selected
+                        exchange/pairs/timeframes.
+  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no
+                        config is provided.
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH, --data-dir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+
Example converting trades
+

The following command will convert all available trade-data in ~/.freqtrade/data/kraken from jsongz to json. +It'll also remove original jsongz data files (--erase parameter).

+
freqtrade convert-trade-data --format-from jsongz --format-to json --datadir ~/.freqtrade/data/kraken --erase
+
+

Sub-command trades to ohlcv

+

When you need to use --dl-trades (kraken only) to download data, conversion of trades data to ohlcv data is the last step. +This command will allow you to repeat this last step for additional timeframes without re-downloading the data.

+
usage: freqtrade trades-to-ohlcv [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                                 [-d PATH] [--userdir PATH]
+                                 [-p PAIRS [PAIRS ...]]
+                                 [-t TIMEFRAMES [TIMEFRAMES ...]]
+                                 [--exchange EXCHANGE]
+                                 [--data-format-ohlcv {json,jsongz,hdf5,feather,parquet}]
+                                 [--data-format-trades {json,jsongz,hdf5}]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]
+                        Limit command to these pairs. Pairs are space-
+                        separated.
+  -t TIMEFRAMES [TIMEFRAMES ...], --timeframes TIMEFRAMES [TIMEFRAMES ...]
+                        Specify which tickers to download. Space-separated
+                        list. Default: `1m 5m`.
+  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no
+                        config is provided.
+  --data-format-ohlcv {json,jsongz,hdf5,feather,parquet}
+                        Storage format for downloaded candle (OHLCV) data.
+                        (default: `json`).
+  --data-format-trades {json,jsongz,hdf5}
+                        Storage format for downloaded trades data. (default:
+                        `jsongz`).
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH, --data-dir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+

Example trade-to-ohlcv conversion

+
freqtrade trades-to-ohlcv --exchange kraken -t 5m 1h 1d --pairs BTC/EUR ETH/EUR
+
+

Sub-command list-data

+

You can get a list of downloaded data using the list-data sub-command.

+
usage: freqtrade list-data [-h] [-v] [--logfile FILE] [-V] [-c PATH] [-d PATH]
+                           [--userdir PATH] [--exchange EXCHANGE]
+                           [--data-format-ohlcv {json,jsongz,hdf5,feather,parquet}]
+                           [-p PAIRS [PAIRS ...]]
+                           [--trading-mode {spot,margin,futures}]
+                           [--show-timerange]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no
+                        config is provided.
+  --data-format-ohlcv {json,jsongz,hdf5,feather,parquet}
+                        Storage format for downloaded candle (OHLCV) data.
+                        (default: `json`).
+  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]
+                        Limit command to these pairs. Pairs are space-
+                        separated.
+  --trading-mode {spot,margin,futures}, --tradingmode {spot,margin,futures}
+                        Select Trading mode
+  --show-timerange      Show timerange available for available data. (May take
+                        a while to calculate).
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH, --data-dir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+

Example list-data

+
> freqtrade list-data --userdir ~/.freqtrade/user_data/
+
+Found 33 pair / timeframe combinations.
+pairs       timeframe
+----------  -----------------------------------------
+ADA/BTC     5m, 15m, 30m, 1h, 2h, 4h, 6h, 12h, 1d
+ADA/ETH     5m, 15m, 30m, 1h, 2h, 4h, 6h, 12h, 1d
+ETH/BTC     5m, 15m, 30m, 1h, 2h, 4h, 6h, 12h, 1d
+ETH/USDT    5m, 15m, 30m, 1h, 2h, 4h
+
+

Trades (tick) data

+

By default, download-data sub-command downloads Candles (OHLCV) data. Some exchanges also provide historic trade-data via their API. +This data can be useful if you need many different timeframes, since it is only downloaded once, and then resampled locally to the desired timeframes.

+

Since this data is large by default, the files use gzip by default. They are stored in your data-directory with the naming convention of <pair>-trades.json.gz (ETH_BTC-trades.json.gz). Incremental mode is also supported, as for historic OHLCV data, so downloading the data once per week with --days 8 will create an incremental data-repository.

+

To use this mode, simply add --dl-trades to your call. This will swap the download method to download trades, and resamples the data locally.

+
+

do not use

+

You should not use this unless you're a kraken user. Most other exchanges provide OHLCV data with sufficient history.

+
+

Example call:

+
freqtrade download-data --exchange kraken --pairs XRP/EUR ETH/EUR --days 20 --dl-trades
+
+
+

Note

+

While this method uses async calls, it will be slow, since it requires the result of the previous call to generate the next request to the exchange.

+
+
+

Warning

+

The historic trades are not available during Freqtrade dry-run and live trade modes because all exchanges tested provide this data with a delay of few 100 candles, so it's not suitable for real-time trading.

+
+
+

Kraken user

+

Kraken users should read this before starting to download data.

+
+

Next step

+

Great, you now have backtest data downloaded, so you can now start backtesting your strategy.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/deprecated/index.html b/en/2023.4/deprecated/index.html new file mode 100644 index 000000000..c98c74910 --- /dev/null +++ b/en/2023.4/deprecated/index.html @@ -0,0 +1,1523 @@ + + + + + + + + + + + + + + + + + + + + + + + + Deprecated Features - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Deprecated features

+

This page contains description of the command line arguments, configuration parameters +and the bot features that were declared as DEPRECATED by the bot development team +and are no longer supported. Please avoid their usage in your configuration.

+

Removed features

+

the --refresh-pairs-cached command line option

+

--refresh-pairs-cached in the context of backtesting, hyperopt and edge allows to refresh candle data for backtesting. +Since this leads to much confusion, and slows down backtesting (while not being part of backtesting) this has been singled out as a separate freqtrade sub-command freqtrade download-data.

+

This command line option was deprecated in 2019.7-dev (develop branch) and removed in 2019.9.

+

The --dynamic-whitelist command line option

+

This command line option was deprecated in 2018 and removed freqtrade 2019.6-dev (develop branch) and in freqtrade 2019.7. +Please refer to pairlists instead.

+

the --live command line option

+

--live in the context of backtesting allowed to download the latest tick data for backtesting. +Did only download the latest 500 candles, so was ineffective in getting good backtest data. +Removed in 2019-7-dev (develop branch) and in freqtrade 2019.8.

+

ticker_interval (now timeframe)

+

Support for ticker_interval terminology was deprecated in 2020.6 in favor of timeframe - and compatibility code was removed in 2022.3.

+

Allow running multiple pairlists in sequence

+

The former "pairlist" section in the configuration has been removed, and is replaced by "pairlists" - being a list to specify a sequence of pairlists.

+

The old section of configuration parameters ("pairlist") has been deprecated in 2019.11 and has been removed in 2020.4.

+

deprecation of bidVolume and askVolume from volume-pairlist

+

Since only quoteVolume can be compared between assets, the other options (bidVolume, askVolume) have been deprecated in 2020.4, and have been removed in 2020.9.

+

Using order book steps for exit price

+

Using order_book_min and order_book_max used to allow stepping the orderbook and trying to find the next ROI slot - trying to place sell-orders early. +As this does however increase risk and provides no benefit, it's been removed for maintainability purposes in 2021.7.

+

Legacy Hyperopt mode

+

Using separate hyperopt files was deprecated in 2021.4 and was removed in 2021.9. +Please switch to the new Parametrized Strategies to benefit from the new hyperopt interface.

+

Strategy changes between V2 and V3

+

Isolated Futures / short trading was introduced in 2022.4. This required major changes to configuration settings, strategy interfaces, ...

+

We have put a great effort into keeping compatibility with existing strategies, so if you just want to continue using freqtrade in spot markets, there are no changes necessary. +While we may drop support for the current interface sometime in the future, we will announce this separately and have an appropriate transition period.

+

Please follow the Strategy migration guide to migrate your strategy to the new format to start using the new functionalities.

+

webhooks - changes with 2022.4

+

buy_tag has been renamed to enter_tag

+

This should apply only to your strategy and potentially to webhooks. +We will keep a compatibility layer for 1-2 versions (so both buy_tag and enter_tag will still work), but support for this in webhooks will disappear after that.

+

Naming changes

+

Webhook terminology changed from "sell" to "exit", and from "buy" to "entry", removing "webhook" in the process.

+
    +
  • webhookbuy, webhookentry -> entry
  • +
  • webhookbuyfill, webhookentryfill -> entry_fill
  • +
  • webhookbuycancel, webhookentrycancel -> entry_cancel
  • +
  • webhooksell, webhookexit -> exit
  • +
  • webhooksellfill, webhookexitfill -> exit_fill
  • +
  • webhooksellcancel, webhookexitcancel -> exit_cancel
  • +
+

Removal of populate_any_indicators

+

version 2023.3 saw the removal of populate_any_indicators in favor of split methods for feature engineering and targets. Please read the migration document for full details.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/developer/index.html b/en/2023.4/developer/index.html new file mode 100644 index 000000000..3f17480aa --- /dev/null +++ b/en/2023.4/developer/index.html @@ -0,0 +1,2196 @@ + + + + + + + + + + + + + + + + + + + + + + Contributors Guide - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Development Help

+

This page is intended for developers of Freqtrade, people who want to contribute to the Freqtrade codebase or documentation, or people who want to understand the source code of the application they're running.

+

All contributions, bug reports, bug fixes, documentation improvements, enhancements and ideas are welcome. We track issues on GitHub and also have a dev channel on discord where you can ask questions.

+

Documentation

+

Documentation is available at https://freqtrade.io and needs to be provided with every new feature PR.

+

Special fields for the documentation (like Note boxes, ...) can be found here.

+

To test the documentation locally use the following commands.

+
pip install -r docs/requirements-docs.txt
+mkdocs serve
+
+

This will spin up a local server (usually on port 8000) so you can see if everything looks as you'd like it to.

+

Developer setup

+

To configure a development environment, you can either use the provided DevContainer, or use the setup.sh script and answer "y" when asked "Do you want to install dependencies for dev [y/N]? ". +Alternatively (e.g. if your system is not supported by the setup.sh script), follow the manual installation process and run pip3 install -e .[all].

+

This will install all required tools for development, including pytest, ruff, mypy, and coveralls.

+

Then install the git hook scripts by running pre-commit install, so your changes will be verified locally before committing. +This avoids a lot of waiting for CI already, as some basic formatting checks are done locally on your machine.

+

Before opening a pull request, please familiarize yourself with our Contributing Guidelines.

+

Devcontainer setup

+

The fastest and easiest way to get started is to use VSCode with the Remote container extension. +This gives developers the ability to start the bot with all required dependencies without needing to install any freqtrade specific dependencies on your local machine.

+

Devcontainer dependencies

+ +

For more information about the Remote container extension, best consult the documentation.

+

Tests

+

New code should be covered by basic unittests. Depending on the complexity of the feature, Reviewers may request more in-depth unittests. +If necessary, the Freqtrade team can assist and give guidance with writing good tests (however please don't expect anyone to write the tests for you).

+

How to run tests

+

Use pytest in root folder to run all available testcases and confirm your local environment is setup correctly

+
+

feature branches

+

Tests are expected to pass on the develop and stable branches. Other branches may be work in progress with tests not working yet.

+
+

Checking log content in tests

+

Freqtrade uses 2 main methods to check log content in tests, log_has() and log_has_re() (to check using regex, in case of dynamic log-messages). +These are available from conftest.py and can be imported in any test module.

+

A sample check looks as follows:

+
from tests.conftest import log_has, log_has_re
+
+def test_method_to_test(caplog):
+    method_to_test()
+
+    assert log_has("This event happened", caplog)
+    # Check regex with trailing number ...
+    assert log_has_re(r"This dynamic event happened and produced \d+", caplog)
+
+

Debug configuration

+

To debug freqtrade, we recommend VSCode with the following launch configuration (located in .vscode/launch.json). +Details will obviously vary between setups - but this should work to get you started.

+
{
+    "name": "freqtrade trade",
+    "type": "python",
+    "request": "launch",
+    "module": "freqtrade",
+    "console": "integratedTerminal",
+    "args": [
+        "trade",
+        // Optional:
+        // "--userdir", "user_data",
+        "--strategy", 
+        "MyAwesomeStrategy",
+    ]
+},
+
+

Command line arguments can be added in the "args" array. +This method can also be used to debug a strategy, by setting the breakpoints within the strategy.

+

A similar setup can also be taken for Pycharm - using freqtrade as module name, and setting the command line arguments as "parameters".

+
+

Startup directory

+

This assumes that you have the repository checked out, and the editor is started at the repository root level (so setup.py is at the top level of your repository).

+
+

ErrorHandling

+

Freqtrade Exceptions all inherit from FreqtradeException. +This general class of error should however not be used directly. Instead, multiple specialized sub-Exceptions exist.

+

Below is an outline of exception inheritance hierarchy:

+
+ FreqtradeException
+|
++---+ OperationalException
+|
++---+ DependencyException
+|   |
+|   +---+ PricingError
+|   |
+|   +---+ ExchangeError
+|       |
+|       +---+ TemporaryError
+|       |
+|       +---+ DDosProtection
+|       |
+|       +---+ InvalidOrderException
+|           |
+|           +---+ RetryableOrderError
+|           |
+|           +---+ InsufficientFundsError
+|
++---+ StrategyError
+
+
+

Plugins

+

Pairlists

+

You have a great idea for a new pair selection algorithm you would like to try out? Great. +Hopefully you also want to contribute this back upstream.

+

Whatever your motivations are - This should get you off the ground in trying to develop a new Pairlist Handler.

+

First of all, have a look at the VolumePairList Handler, and best copy this file with a name of your new Pairlist Handler.

+

This is a simple Handler, which however serves as a good example on how to start developing.

+

Next, modify the class-name of the Handler (ideally align this with the module filename).

+

The base-class provides an instance of the exchange (self._exchange) the pairlist manager (self._pairlistmanager), as well as the main configuration (self._config), the pairlist dedicated configuration (self._pairlistconfig) and the absolute position within the list of pairlists.

+
        self._exchange = exchange
+        self._pairlistmanager = pairlistmanager
+        self._config = config
+        self._pairlistconfig = pairlistconfig
+        self._pairlist_pos = pairlist_pos
+
+
+

Tip

+

Don't forget to register your pairlist in constants.py under the variable AVAILABLE_PAIRLISTS - otherwise it will not be selectable.

+
+

Now, let's step through the methods which require actions:

+

Pairlist configuration

+

Configuration for the chain of Pairlist Handlers is done in the bot configuration file in the element "pairlists", an array of configuration parameters for each Pairlist Handlers in the chain.

+

By convention, "number_assets" is used to specify the maximum number of pairs to keep in the pairlist. Please follow this to ensure a consistent user experience.

+

Additional parameters can be configured as needed. For instance, VolumePairList uses "sort_key" to specify the sorting value - however feel free to specify whatever is necessary for your great algorithm to be successful and dynamic.

+

short_desc

+

Returns a description used for Telegram messages.

+

This should contain the name of the Pairlist Handler, as well as a short description containing the number of assets. Please follow the format "PairlistName - top/bottom X pairs".

+

gen_pairlist

+

Override this method if the Pairlist Handler can be used as the leading Pairlist Handler in the chain, defining the initial pairlist which is then handled by all Pairlist Handlers in the chain. Examples are StaticPairList and VolumePairList.

+

This is called with each iteration of the bot (only if the Pairlist Handler is at the first location) - so consider implementing caching for compute/network heavy calculations.

+

It must return the resulting pairlist (which may then be passed into the chain of Pairlist Handlers).

+

Validations are optional, the parent class exposes a _verify_blacklist(pairlist) and _whitelist_for_active_markets(pairlist) to do default filtering. Use this if you limit your result to a certain number of pairs - so the end-result is not shorter than expected.

+

filter_pairlist

+

This method is called for each Pairlist Handler in the chain by the pairlist manager.

+

This is called with each iteration of the bot - so consider implementing caching for compute/network heavy calculations.

+

It gets passed a pairlist (which can be the result of previous pairlists) as well as tickers, a pre-fetched version of get_tickers().

+

The default implementation in the base class simply calls the _validate_pair() method for each pair in the pairlist, but you may override it. So you should either implement the _validate_pair() in your Pairlist Handler or override filter_pairlist() to do something else.

+

If overridden, it must return the resulting pairlist (which may then be passed into the next Pairlist Handler in the chain).

+

Validations are optional, the parent class exposes a _verify_blacklist(pairlist) and _whitelist_for_active_markets(pairlist) to do default filters. Use this if you limit your result to a certain number of pairs - so the end result is not shorter than expected.

+

In VolumePairList, this implements different methods of sorting, does early validation so only the expected number of pairs is returned.

+
sample
+
    def filter_pairlist(self, pairlist: List[str], tickers: Dict) -> List[str]:
+        # Generate dynamic whitelist
+        pairs = self._calculate_pairlist(pairlist, tickers)
+        return pairs
+
+

Protections

+

Best read the Protection documentation to understand protections. +This Guide is directed towards Developers who want to develop a new protection.

+

No protection should use datetime directly, but use the provided date_now variable for date calculations. This preserves the ability to backtest protections.

+
+

Writing a new Protection

+

Best copy one of the existing Protections to have a good example. +Don't forget to register your protection in constants.py under the variable AVAILABLE_PROTECTIONS - otherwise it will not be selectable.

+
+

Implementation of a new protection

+

All Protection implementations must have IProtection as parent class. +For that reason, they must implement the following methods:

+
    +
  • short_desc()
  • +
  • global_stop()
  • +
  • stop_per_pair().
  • +
+

global_stop() and stop_per_pair() must return a ProtectionReturn object, which consists of:

+
    +
  • lock pair - boolean
  • +
  • lock until - datetime - until when should the pair be locked (will be rounded up to the next new candle)
  • +
  • reason - string, used for logging and storage in the database
  • +
  • lock_side - long, short or '*'.
  • +
+

The until portion should be calculated using the provided calculate_lock_end() method.

+

All Protections should use "stop_duration" / "stop_duration_candles" to define how long a a pair (or all pairs) should be locked. +The content of this is made available as self._stop_duration to the each Protection.

+

If your protection requires a look-back period, please use "lookback_period" / "lockback_period_candles" to keep all protections aligned.

+

Global vs. local stops

+

Protections can have 2 different ways to stop trading for a limited :

+
    +
  • Per pair (local)
  • +
  • For all Pairs (globally)
  • +
+
Protections - per pair
+

Protections that implement the per pair approach must set has_local_stop=True. +The method stop_per_pair() will be called whenever a trade closed (exit order completed).

+
Protections - global protection
+

These Protections should do their evaluation across all pairs, and consequently will also lock all pairs from trading (called a global PairLock). +Global protection must set has_global_stop=True to be evaluated for global stops. +The method global_stop() will be called whenever a trade closed (exit order completed).

+
Protections - calculating lock end time
+

Protections should calculate the lock end time based on the last trade it considers. +This avoids re-locking should the lookback-period be longer than the actual lock period.

+

The IProtection parent class provides a helper method for this in calculate_lock_end().

+
+

Implement a new Exchange (WIP)

+
+

Note

+

This section is a Work in Progress and is not a complete guide on how to test a new exchange with Freqtrade.

+
+
+

Note

+

Make sure to use an up-to-date version of CCXT before running any of the below tests. +You can get the latest version of ccxt by running pip install -U ccxt with activated virtual environment. +Native docker is not supported for these tests, however the available dev-container will support all required actions and eventually necessary changes.

+
+

Most exchanges supported by CCXT should work out of the box.

+

To quickly test the public endpoints of an exchange, add a configuration for your exchange to test_ccxt_compat.py and run these tests with pytest --longrun tests/exchange/test_ccxt_compat.py. +Completing these tests successfully a good basis point (it's a requirement, actually), however these won't guarantee correct exchange functioning, as this only tests public endpoints, but no private endpoint (like generate order or similar).

+

Also try to use freqtrade download-data for an extended timerange (multiple months) and verify that the data downloaded correctly (no holes, the specified timerange was actually downloaded).

+

These are prerequisites to have an exchange listed as either Supported or Community tested (listed on the homepage). +The below are "extras", which will make an exchange better (feature-complete) - but are not absolutely necessary for either of the 2 categories.

+

Additional tests / steps to complete:

+
    +
  • Verify data provided by fetch_ohlcv() - and eventually adjust ohlcv_candle_limit for this exchange
  • +
  • Check L2 orderbook limit range (API documentation) - and eventually set as necessary
  • +
  • Check if balance shows correctly (*)
  • +
  • Create market order (*)
  • +
  • Create limit order (*)
  • +
  • Complete trade (enter + exit) (*)
      +
    • Compare result calculation between exchange and bot
    • +
    • Ensure fees are applied correctly (check the database against the exchange)
    • +
    +
  • +
+

(*) Requires API keys and Balance on the exchange.

+

Stoploss On Exchange

+

Check if the new exchange supports Stoploss on Exchange orders through their API.

+

Since CCXT does not provide unification for Stoploss On Exchange yet, we'll need to implement the exchange-specific parameters ourselves. Best look at binance.py for an example implementation of this. You'll need to dig through the documentation of the Exchange's API on how exactly this can be done. CCXT Issues may also provide great help, since others may have implemented something similar for their projects.

+

Incomplete candles

+

While fetching candle (OHLCV) data, we may end up getting incomplete candles (depending on the exchange). +To demonstrate this, we'll use daily candles ("1d") to keep things simple. +We query the api (ct.fetch_ohlcv()) for the timeframe and look at the date of the last entry. If this entry changes or shows the date of a "incomplete" candle, then we should drop this since having incomplete candles is problematic because indicators assume that only complete candles are passed to them, and will generate a lot of false buy signals. By default, we're therefore removing the last candle assuming it's incomplete.

+

To check how the new exchange behaves, you can use the following snippet:

+
import ccxt
+from datetime import datetime
+from freqtrade.data.converter import ohlcv_to_dataframe
+ct = ccxt.binance()
+timeframe = "1d"
+pair = "XLM/BTC"  # Make sure to use a pair that exists on that exchange!
+raw = ct.fetch_ohlcv(pair, timeframe=timeframe)
+
+# convert to dataframe
+df1 = ohlcv_to_dataframe(raw, timeframe, pair=pair, drop_incomplete=False)
+
+print(df1.tail(1))
+print(datetime.utcnow())
+
+
                         date      open      high       low     close  volume  
+499 2019-06-08 00:00:00+00:00  0.000007  0.000007  0.000007  0.000007   26264344.0  
+2019-06-09 12:30:27.873327
+
+

The output will show the last entry from the Exchange as well as the current UTC date. +If the day shows the same day, then the last candle can be assumed as incomplete and should be dropped (leave the setting "ohlcv_partial_candle" from the exchange-class untouched / True). Otherwise, set "ohlcv_partial_candle" to False to not drop Candles (shown in the example above). +Another way is to run this command multiple times in a row and observe if the volume is changing (while the date remains the same).

+

Update binance cached leverage tiers

+

Updating leveraged tiers should be done regularly - and requires an authenticated account with futures enabled.

+
import ccxt
+import json
+from pathlib import Path
+
+exchange = ccxt.binance({
+    'apiKey': '<apikey>',
+    'secret': '<secret>'
+    'options': {'defaultType': 'swap'}
+    })
+_ = exchange.load_markets()
+
+lev_tiers = exchange.fetch_leverage_tiers()
+
+# Assumes this is running in the root of the repository.
+file = Path('freqtrade/exchange/binance_leverage_tiers.json')
+json.dump(dict(sorted(lev_tiers.items())), file.open('w'), indent=2)
+
+

This file should then be contributed upstream, so others can benefit from this, too.

+

Updating example notebooks

+

To keep the jupyter notebooks aligned with the documentation, the following should be ran after updating a example notebook.

+
jupyter nbconvert --ClearOutputPreprocessor.enabled=True --inplace freqtrade/templates/strategy_analysis_example.ipynb
+jupyter nbconvert --ClearOutputPreprocessor.enabled=True --to markdown freqtrade/templates/strategy_analysis_example.ipynb --stdout > docs/strategy_analysis_example.md
+
+

Continuous integration

+

This documents some decisions taken for the CI Pipeline.

+
    +
  • CI runs on all OS variants, Linux (ubuntu), macOS and Windows.
  • +
  • Docker images are build for the branches stable and develop, and are built as multiarch builds, supporting multiple platforms via the same tag.
  • +
  • Docker images containing Plot dependencies are also available as stable_plot and develop_plot.
  • +
  • Docker images contain a file, /freqtrade/freqtrade_commit containing the commit this image is based of.
  • +
  • Full docker image rebuilds are run once a week via schedule.
  • +
  • Deployments run on ubuntu.
  • +
  • ta-lib binaries are contained in the build_helpers directory to avoid fails related to external unavailability.
  • +
  • All tests must pass for a PR to be merged to stable or develop.
  • +
+

Creating a release

+

This part of the documentation is aimed at maintainers, and shows how to create a release.

+

Create release branch

+

First, pick a commit that's about one week old (to not include latest additions to releases).

+
# create new branch
+git checkout -b new_release <commitid>
+
+

Determine if crucial bugfixes have been made between this commit and the current state, and eventually cherry-pick these.

+
    +
  • Merge the release branch (stable) into this branch.
  • +
  • Edit freqtrade/__init__.py and add the version matching the current date (for example 2019.7 for July 2019). Minor versions can be 2019.7.1 should we need to do a second release that month. Version numbers must follow allowed versions from PEP0440 to avoid failures pushing to pypi.
  • +
  • Commit this part.
  • +
  • push that branch to the remote and create a PR against the stable branch.
  • +
  • Update develop version to next version following the pattern 2019.8-dev.
  • +
+

Create changelog from git commits

+
+

Note

+

Make sure that the stable branch is up-to-date!

+
+
# Needs to be done before merging / pulling that branch.
+git log --oneline --no-decorate --no-merges stable..new_release
+
+

To keep the release-log short, best wrap the full git changelog into a collapsible details section.

+
<details>
+<summary>Expand full changelog</summary>
+
+... Full git changelog
+
+</details>
+
+

FreqUI release

+

If FreqUI has been updated substantially, make sure to create a release before merging the release branch. +Make sure that freqUI CI on the release is finished and passed before merging the release.

+

Create github release / tag

+

Once the PR against stable is merged (best right after merging):

+
    +
  • Use the button "Draft a new release" in the Github UI (subsection releases).
  • +
  • Use the version-number specified as tag.
  • +
  • Use "stable" as reference (this step comes after the above PR is merged).
  • +
  • Use the above changelog as release comment (as codeblock)
  • +
+

Releases

+

pypi

+
+

Note

+

This process is now automated as part of Github Actions.

+
+

To create a pypi release, please run the following commands:

+

Additional requirement: wheel, twine (for uploading), account on pypi with proper permissions.

+
python setup.py sdist bdist_wheel
+
+# For pypi test (to check if some change to the installation did work)
+twine upload --repository-url https://test.pypi.org/legacy/ dist/*
+
+# For production:
+twine upload dist/*
+
+

Please don't push non-releases to the productive / real pypi instance.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/docker_quickstart/index.html b/en/2023.4/docker_quickstart/index.html new file mode 100644 index 000000000..5969f0446 --- /dev/null +++ b/en/2023.4/docker_quickstart/index.html @@ -0,0 +1,1700 @@ + + + + + + + + + + + + + + + + + + + + + + + + Quickstart with Docker - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Using Freqtrade with Docker

+

This page explains how to run the bot with Docker. It is not meant to work out of the box. You'll still need to read through the documentation and understand how to properly configure it.

+

Install Docker

+

Start by downloading and installing Docker / Docker Desktop for your platform:

+ +
+

Docker compose install

+

Freqtrade documentation assumes the use of Docker desktop (or the docker compose plugin).
+While the docker-compose standalone installation still works, it will require changing all docker compose commands from docker compose to docker-compose to work (e.g. docker compose up -d will become docker-compose up -d).

+
+

Freqtrade with docker

+

Freqtrade provides an official Docker image on Dockerhub, as well as a docker compose file ready for usage.

+
+

Note

+
    +
  • The following section assumes that docker is installed and available to the logged in user.
  • +
  • All below commands use relative directories and will have to be executed from the directory containing the docker-compose.yml file.
  • +
+
+

Docker quick start

+

Create a new directory and place the docker-compose file in this directory.

+
mkdir ft_userdata
+cd ft_userdata/
+# Download the docker-compose file from the repository
+curl https://raw.githubusercontent.com/freqtrade/freqtrade/stable/docker-compose.yml -o docker-compose.yml
+
+# Pull the freqtrade image
+docker compose pull
+
+# Create user directory structure
+docker compose run --rm freqtrade create-userdir --userdir user_data
+
+# Create configuration - Requires answering interactive questions
+docker compose run --rm freqtrade new-config --config user_data/config.json
+
+

The above snippet creates a new directory called ft_userdata, downloads the latest compose file and pulls the freqtrade image. +The last 2 steps in the snippet create the directory with user_data, as well as (interactively) the default configuration based on your selections.

+
+

How to edit the bot configuration?

+

You can edit the configuration at any time, which is available as user_data/config.json (within the directory ft_userdata) when using the above configuration.

+

You can also change the both Strategy and commands by editing the command section of your docker-compose.yml file.

+
+

Adding a custom strategy

+
    +
  1. The configuration is now available as user_data/config.json
  2. +
  3. Copy a custom strategy to the directory user_data/strategies/
  4. +
  5. Add the Strategy' class name to the docker-compose.yml file
  6. +
+

The SampleStrategy is run by default.

+
+

SampleStrategy is just a demo!

+

The SampleStrategy is there for your reference and give you ideas for your own strategy. +Please always backtest your strategy and use dry-run for some time before risking real money! +You will find more information about Strategy development in the Strategy documentation.

+
+

Once this is done, you're ready to launch the bot in trading mode (Dry-run or Live-trading, depending on your answer to the corresponding question you made above).

+
docker compose up -d
+
+
+

Default configuration

+

While the configuration generated will be mostly functional, you will still need to verify that all options correspond to what you want (like Pricing, pairlist, ...) before starting the bot.

+
+

Accessing the UI

+

If you've selected to enable FreqUI in the new-config step, you will have freqUI available at port localhost:8080.

+

You can now access the UI by typing localhost:8080 in your browser.

+
+UI Access on a remote servers +

If you're running on a VPS, you should consider using either a ssh tunnel, or setup a VPN (openVPN, wireguard) to connect to your bot. +This will ensure that freqUI is not directly exposed to the internet, which is not recommended for security reasons (freqUI does not support https out of the box). +Setup of these tools is not part of this tutorial, however many good tutorials can be found on the internet. +Please also read the API configuration with docker section to learn more about this configuration.

+
+

Monitoring the bot

+

You can check for running instances with docker compose ps. +This should list the service freqtrade as running. If that's not the case, best check the logs (see next point).

+

Docker compose logs

+

Logs will be written to: user_data/logs/freqtrade.log.
+You can also check the latest log with the command docker compose logs -f.

+

Database

+

The database will be located at: user_data/tradesv3.sqlite

+

Updating freqtrade with docker

+

Updating freqtrade when using docker is as simple as running the following 2 commands:

+
# Download the latest image
+docker compose pull
+# Restart the image
+docker compose up -d
+
+

This will first pull the latest image, and will then restart the container with the just pulled version.

+
+

Check the Changelog

+

You should always check the changelog for breaking changes / manual interventions required and make sure the bot starts correctly after the update.

+
+

Editing the docker-compose file

+

Advanced users may edit the docker-compose file further to include all possible options or arguments.

+

All freqtrade arguments will be available by running docker compose run --rm freqtrade <command> <optional arguments>.

+
+

docker compose for trade commands

+

Trade commands (freqtrade trade <...>) should not be ran via docker compose run - but should use docker compose up -d instead. +This makes sure that the container is properly started (including port forwardings) and will make sure that the container will restart after a system reboot. +If you intend to use freqUI, please also ensure to adjust the configuration accordingly, otherwise the UI will not be available.

+
+
+

docker compose run --rm

+

Including --rm will remove the container after completion, and is highly recommended for all modes except trading mode (running with freqtrade trade command).

+
+
+Using docker without docker +

"docker compose run --rm" will require a compose file to be provided. +Some freqtrade commands that don't require authentication such as list-pairs can be run with "docker run --rm" instead.
+For example docker run --rm freqtradeorg/freqtrade:stable list-pairs --exchange binance --quote BTC --print-json.
+This can be useful for fetching exchange information to add to your config.json without affecting your running containers.

+
+

Example: Download data with docker

+

Download backtesting data for 5 days for the pair ETH/BTC and 1h timeframe from Binance. The data will be stored in the directory user_data/data/ on the host.

+
docker compose run --rm freqtrade download-data --pairs ETH/BTC --exchange binance --days 5 -t 1h
+
+

Head over to the Data Downloading Documentation for more details on downloading data.

+

Example: Backtest with docker

+

Run backtesting in docker-containers for SampleStrategy and specified timerange of historical data, on 5m timeframe:

+
docker compose run --rm freqtrade backtesting --config user_data/config.json --strategy SampleStrategy --timerange 20190801-20191001 -i 5m
+
+

Head over to the Backtesting Documentation to learn more.

+

Additional dependencies with docker

+

If your strategy requires dependencies not included in the default image - it will be necessary to build the image on your host. +For this, please create a Dockerfile containing installation steps for the additional dependencies (have a look at docker/Dockerfile.custom for an example).

+

You'll then also need to modify the docker-compose.yml file and uncomment the build step, as well as rename the image to avoid naming collisions.

+
    image: freqtrade_custom
+    build:
+      context: .
+      dockerfile: "./Dockerfile.<yourextension>"
+
+

You can then run docker compose build --pull to build the docker image, and run it using the commands described above.

+

Plotting with docker

+

Commands freqtrade plot-profit and freqtrade plot-dataframe (Documentation) are available by changing the image to *_plot in your docker-compose.yml file. +You can then use these commands as follows:

+
docker compose run --rm freqtrade plot-dataframe --strategy AwesomeStrategy -p BTC/ETH --timerange=20180801-20180805
+
+

The output will be stored in the user_data/plot directory, and can be opened with any modern browser.

+

Data analysis using docker compose

+

Freqtrade provides a docker-compose file which starts up a jupyter lab server. +You can run this server using the following command:

+
docker compose -f docker/docker-compose-jupyter.yml up
+
+

This will create a docker-container running jupyter lab, which will be accessible using https://127.0.0.1:8888/lab. +Please use the link that's printed in the console after startup for simplified login.

+

Since part of this image is built on your machine, it is recommended to rebuild the image from time to time to keep freqtrade (and dependencies) up-to-date.

+
docker compose -f docker/docker-compose-jupyter.yml build --no-cache
+
+

Troubleshooting

+

Docker on Windows

+
    +
  • Error: "Timestamp for this request is outside of the recvWindow."
      +
    • +

      The market api requests require a synchronized clock but the time in the docker container shifts a bit over time into the past. + To fix this issue temporarily you need to run wsl --shutdown and restart docker again (a popup on windows 10 will ask you to do so). + A permanent solution is either to host the docker container on a linux host or restart the wsl from time to time with the scheduler.

      +
      taskkill /IM "Docker Desktop.exe" /F
      +wsl --shutdown
      +start "" "C:\Program Files\Docker\Docker\Docker Desktop.exe"
      +
      +
    • +
    +
  • +
+
+

Warning

+

Due to the above, we do not recommend the usage of docker on windows for production setups, but only for experimentation, datadownload and backtesting. +Best use a linux-VPS for running freqtrade reliably.

+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/edge/index.html b/en/2023.4/edge/index.html new file mode 100644 index 000000000..a3c859aed --- /dev/null +++ b/en/2023.4/edge/index.html @@ -0,0 +1,1979 @@ + + + + + + + + + + + + + + + + + + + + + + + + Edge Positioning - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Edge positioning

+

The Edge Positioning module uses probability to calculate your win rate and risk reward ratio. It will use these statistics to control your strategy trade entry points, position size and, stoploss.

+
+

Warning

+

When using Edge positioning with a dynamic whitelist (VolumePairList), make sure to also use AgeFilter and set it to at least calculate_since_number_of_days to avoid problems with missing data.

+
+
+

Note

+

Edge Positioning only considers its own buy/sell/stoploss signals. It ignores the stoploss, trailing stoploss, and ROI settings in the strategy configuration file. +Edge Positioning improves the performance of some trading strategies and decreases the performance of others.

+
+

Introduction

+

Trading strategies are not perfect. They are frameworks that are susceptible to the market and its indicators. Because the market is not at all predictable, sometimes a strategy will win and sometimes the same strategy will lose.

+

To obtain an edge in the market, a strategy has to make more money than it loses. Making money in trading is not only about how often the strategy makes or loses money.

+
+

It doesn't matter how often, but how much!

+

A bad strategy might make 1 penny in ten transactions but lose 1 dollar in one transaction. If one only checks the number of winning trades, it would be misleading to think that the strategy is actually making a profit.

+
+

The Edge Positioning module seeks to improve a strategy's winning probability and the money that the strategy will make on the long run.

+

We raise the following question1:

+
+

Which trade is a better option?

+

a) A trade with 80% of chance of losing 100$ and 20% chance of winning 200$
+b) A trade with 100% of chance of losing 30$

+
+
+Answer +

The expected value of a) is smaller than the expected value of b).
+Hence, b) represents a smaller loss in the long run.
+However, the answer is: it depends

+
+

Another way to look at it is to ask a similar question:

+
+

Which trade is a better option?

+

a) A trade with 80% of chance of winning 100$ and 20% chance of losing 200$
+b) A trade with 100% of chance of winning 30$

+
+

Edge positioning tries to answer the hard questions about risk/reward and position size automatically, seeking to minimizes the chances of losing of a given strategy.

+

Trading, winning and losing

+

Let's call \(o\) the return of a single transaction \(o\) where \(o \in \mathbb{R}\). The collection \(O = \{o_1, o_2, ..., o_N\}\) is the set of all returns of transactions made during a trading session. We say that \(N\) is the cardinality of \(O\), or, in lay terms, it is the number of transactions made in a trading session.

+
+

Example

+

In a session where a strategy made three transactions we can say that \(O = \{3.5, -1, 15\}\). That means that \(N = 3\) and \(o_1 = 3.5\), \(o_2 = -1\), \(o_3 = 15\).

+
+

A winning trade is a trade where a strategy made money. Making money means that the strategy closed the position in a value that returned a profit, after all deducted fees. Formally, a winning trade will have a return \(o_i > 0\). Similarly, a losing trade will have a return \(o_j \leq 0\). With that, we can discover the set of all winning trades, \(T_{win}\), as follows:

+
\[ T_{win} = \{ o \in O | o > 0 \} \]
+

Similarly, we can discover the set of losing trades \(T_{lose}\) as follows:

+
\[ T_{lose} = \{o \in O | o \leq 0\} \]
+
+

Example

+

In a section where a strategy made four transactions \(O = \{3.5, -1, 15, 0\}\):
+\(T_{win} = \{3.5, 15\}\)
+\(T_{lose} = \{-1, 0\}\)

+
+

Win Rate and Lose Rate

+

The win rate \(W\) is the proportion of winning trades with respect to all the trades made by a strategy. We use the following function to compute the win rate:

+
\[W = \frac{|T_{win}|}{N}\]
+

Where \(W\) is the win rate, \(N\) is the number of trades and, \(T_{win}\) is the set of all trades where the strategy made money.

+

Similarly, we can compute the rate of losing trades:

+
\[ + L = \frac{|T_{lose}|}{N} +\]
+

Where \(L\) is the lose rate, \(N\) is the amount of trades made and, \(T_{lose}\) is the set of all trades where the strategy lost money. Note that the above formula is the same as calculating \(L = 1 – W\) or \(W = 1 – L\)

+

Risk Reward Ratio

+

Risk Reward Ratio (\(R\)) is a formula used to measure the expected gains of a given investment against the risk of loss. It is basically what you potentially win divided by what you potentially lose. Formally:

+
\[ R = \frac{\text{potential_profit}}{\text{potential_loss}} \]
+
+Worked example of \(R\) calculation +

Let's say that you think that the price of stonecoin today is 10.0$. You believe that, because they will start mining stonecoin, it will go up to 15.0$ tomorrow. There is the risk that the stone is too hard, and the GPUs can't mine it, so the price might go to 0$ tomorrow. You are planning to invest 100$, which will give you 10 shares (100 / 10).

+

Your potential profit is calculated as:

+

\(\begin{aligned} + \text{potential_profit} &= (\text{potential_price} - \text{entry_price}) * \frac{\text{investment}}{\text{entry_price}} \\ + &= (15 - 10) * (100 / 10) \\ + &= 50 +\end{aligned}\)

+

Since the price might go to 0$, the 100$ dollars invested could turn into 0.

+

We do however use a stoploss of 15% - so in the worst case, we'll sell 15% below entry price (or at 8.5$).

+

\(\begin{aligned} + \text{potential_loss} &= (\text{entry_price} - \text{stoploss}) * \frac{\text{investment}}{\text{entry_price}} \\ + &= (10 - 8.5) * (100 / 10)\\ + &= 15 +\end{aligned}\)

+

We can compute the Risk Reward Ratio as follows:

+

\(\begin{aligned} + R &= \frac{\text{potential_profit}}{\text{potential_loss}}\\ + &= \frac{50}{15}\\ + &= 3.33 +\end{aligned}\)
+What it effectively means is that the strategy have the potential to make 3.33$ for each 1$ invested.

+
+

On a long horizon, that is, on many trades, we can calculate the risk reward by dividing the strategy' average profit on winning trades by the strategy' average loss on losing trades. We can calculate the average profit, \(\mu_{win}\), as follows:

+
\[ \text{average_profit} = \mu_{win} = \frac{\text{sum_of_profits}}{\text{count_winning_trades}} = \frac{\sum^{o \in T_{win}} o}{|T_{win}|} \]
+

Similarly, we can calculate the average loss, \(\mu_{lose}\), as follows:

+
\[ \text{average_loss} = \mu_{lose} = \frac{\text{sum_of_losses}}{\text{count_losing_trades}} = \frac{\sum^{o \in T_{lose}} o}{|T_{lose}|} \]
+

Finally, we can calculate the Risk Reward ratio, \(R\), as follows:

+
\[ R = \frac{\text{average_profit}}{\text{average_loss}} = \frac{\mu_{win}}{\mu_{lose}}\\ \]
+
+Worked example of \(R\) calculation using mean profit/loss +

Let's say the strategy that we are using makes an average win \(\mu_{win} = 2.06\) and an average loss \(\mu_{loss} = 4.11\).
+We calculate the risk reward ratio as follows:
+\(R = \frac{\mu_{win}}{\mu_{loss}} = \frac{2.06}{4.11} = 0.5012...\)

+
+

Expectancy

+

By combining the Win Rate \(W\) and and the Risk Reward ratio \(R\) to create an expectancy ratio \(E\). A expectance ratio is the expected return of the investment made in a trade. We can compute the value of \(E\) as follows:

+
\[E = R * W - L\]
+
+

Calculating \(E\)

+

Let's say that a strategy has a win rate \(W = 0.28\) and a risk reward ratio \(R = 5\). What this means is that the strategy is expected to make 5 times the investment around on 28% of the trades it makes. Working out the example:
+\(E = R * W - L = 5 * 0.28 - 0.72 = 0.68\) +

+
+

The expectancy worked out in the example above means that, on average, this strategy' trades will return 1.68 times the size of its losses. Said another way, the strategy makes 1.68$ for every 1$ it loses, on average.

+

This is important for two reasons: First, it may seem obvious, but you know right away that you have a positive return. Second, you now have a number you can compare to other candidate systems to make decisions about which ones you employ.

+

It is important to remember that any system with an expectancy greater than 0 is profitable using past data. The key is finding one that will be profitable in the future.

+

You can also use this value to evaluate the effectiveness of modifications to this system.

+
+

Note

+

It's important to keep in mind that Edge is testing your expectancy using historical data, there's no guarantee that you will have a similar edge in the future. It's still vital to do this testing in order to build confidence in your methodology but be wary of "curve-fitting" your approach to the historical data as things are unlikely to play out the exact same way for future trades.

+
+

How does it work?

+

Edge combines dynamic stoploss, dynamic positions, and whitelist generation into one isolated module which is then applied to the trading strategy. If enabled in config, Edge will go through historical data with a range of stoplosses in order to find buy and sell/stoploss signals. It then calculates win rate and expectancy over N trades for each stoploss. Here is an example:

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
PairStoplossWin RateRisk Reward RatioExpectancy
XZC/ETH-0.010.501.1763840.088
XZC/ETH-0.020.511.1159410.079
XZC/ETH-0.030.521.3596700.228
XZC/ETH-0.040.511.2345390.117
+

The goal here is to find the best stoploss for the strategy in order to have the maximum expectancy. In the above example stoploss at \(3%\) leads to the maximum expectancy according to historical data.

+

Edge module then forces stoploss value it evaluated to your strategy dynamically.

+

Position size

+

Edge dictates the amount at stake for each trade to the bot according to the following factors:

+
    +
  • Allowed capital at risk
  • +
  • Stoploss
  • +
+

Allowed capital at risk is calculated as follows:

+
Allowed capital at risk = (Capital available_percentage) X (Allowed risk per trade)
+
+

Stoploss is calculated as described above with respect to historical data.

+

The position size is calculated as follows:

+
Position size = (Allowed capital at risk) / Stoploss
+
+

Example:

+

Let's say the stake currency is ETH and there is \(10\) ETH on the wallet. The capital available percentage is \(50%\) and the allowed risk per trade is \(1\%\). Thus, the available capital for trading is \(10 * 0.5 = 5\) ETH and the allowed capital at risk would be \(5 * 0.01 = 0.05\) ETH.

+
    +
  • Trade 1: The strategy detects a new buy signal in the XLM/ETH market. Edge Positioning calculates a stoploss of \(2\%\) and a position of \(0.05 / 0.02 = 2.5\) ETH. The bot takes a position of \(2.5\) ETH in the XLM/ETH market.
  • +
+
    +
  • Trade 2: The strategy detects a buy signal on the BTC/ETH market while Trade 1 is still open. Edge Positioning calculates the stoploss of \(4\%\) on this market. Thus, Trade 2 position size is \(0.05 / 0.04 = 1.25\) ETH.
  • +
+
+

Available Capital \(\neq\) Available in wallet

+

The available capital for trading didn't change in Trade 2 even with Trade 1 still open. The available capital is not the free amount in the wallet.

+
+
    +
  • Trade 3: The strategy detects a buy signal in the ADA/ETH market. Edge Positioning calculates a stoploss of \(1\%\) and a position of \(0.05 / 0.01 = 5\) ETH. Since Trade 1 has \(2.5\) ETH blocked and Trade 2 has \(1.25\) ETH blocked, there is only \(5 - 1.25 - 2.5 = 1.25\) ETH available. Hence, the position size of Trade 3 is \(1.25\) ETH.
  • +
+
+

Available Capital Updates

+

The available capital does not change before a position is sold. After a trade is closed the Available Capital goes up if the trade was profitable or goes down if the trade was a loss.

+
+
    +
  • The strategy detects a sell signal in the XLM/ETH market. The bot exits Trade 1 for a profit of \(1\) ETH. The total capital in the wallet becomes \(11\) ETH and the available capital for trading becomes \(5.5\) ETH.
  • +
+
    +
  • Trade 4 The strategy detects a new buy signal int the XLM/ETH market. Edge Positioning calculates the stoploss of \(2\%\), and the position size of \(0.055 / 0.02 = 2.75\) ETH.
  • +
+

Edge command reference

+
usage: freqtrade edge [-h] [-v] [--logfile FILE] [-V] [-c PATH] [-d PATH]
+                      [--userdir PATH] [-s NAME] [--strategy-path PATH]
+                      [-i TIMEFRAME] [--timerange TIMERANGE]
+                      [--data-format-ohlcv {json,jsongz,hdf5}]
+                      [--max-open-trades INT] [--stake-amount STAKE_AMOUNT]
+                      [--fee FLOAT] [-p PAIRS [PAIRS ...]]
+                      [--stoplosses STOPLOSS_RANGE]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  -i TIMEFRAME, --timeframe TIMEFRAME
+                        Specify timeframe (`1m`, `5m`, `30m`, `1h`, `1d`).
+  --timerange TIMERANGE
+                        Specify what timerange of data to use.
+  --data-format-ohlcv {json,jsongz,hdf5}
+                        Storage format for downloaded candle (OHLCV) data.
+                        (default: `None`).
+  --max-open-trades INT
+                        Override the value of the `max_open_trades`
+                        configuration setting.
+  --stake-amount STAKE_AMOUNT
+                        Override the value of the `stake_amount` configuration
+                        setting.
+  --fee FLOAT           Specify fee ratio. Will be applied twice (on trade
+                        entry and exit).
+  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]
+                        Limit command to these pairs. Pairs are space-
+                        separated.
+  --stoplosses STOPLOSS_RANGE
+                        Defines a range of stoploss values against which edge
+                        will assess the strategy. The format is "min,max,step"
+                        (without any space). Example:
+                        `--stoplosses=-0.01,-0.1,-0.001`
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+Strategy arguments:
+  -s NAME, --strategy NAME
+                        Specify strategy class name which will be used by the
+                        bot.
+  --strategy-path PATH  Specify additional strategy lookup path.
+
+

Configurations

+

Edge module has following configuration options:

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
ParameterDescription
enabledIf true, then Edge will run periodically.
Defaults to false.
Datatype: Boolean
process_throttle_secsHow often should Edge run in seconds.
Defaults to 3600 (once per hour).
Datatype: Integer
calculate_since_number_of_daysNumber of days of data against which Edge calculates Win Rate, Risk Reward and Expectancy.
Note that it downloads historical data so increasing this number would lead to slowing down the bot.
Defaults to 7.
Datatype: Integer
allowed_riskRatio of allowed risk per trade.
Defaults to 0.01 (1%)).
Datatype: Float
stoploss_range_minMinimum stoploss.
Defaults to -0.01.
Datatype: Float
stoploss_range_maxMaximum stoploss.
Defaults to -0.10.
Datatype: Float
stoploss_range_stepAs an example if this is set to -0.01 then Edge will test the strategy for [-0.01, -0,02, -0,03 ..., -0.09, -0.10] ranges.
Note than having a smaller step means having a bigger range which could lead to slow calculation.
If you set this parameter to -0.001, you then slow down the Edge calculation by a factor of 10.
Defaults to -0.001.
Datatype: Float
minimum_winrateIt filters out pairs which don't have at least minimum_winrate.
This comes handy if you want to be conservative and don't comprise win rate in favour of risk reward ratio.
Defaults to 0.60.
Datatype: Float
minimum_expectancyIt filters out pairs which have the expectancy lower than this number.
Having an expectancy of 0.20 means if you put 10$ on a trade you expect a 12$ return.
Defaults to 0.20.
Datatype: Float
min_trade_numberWhen calculating W, R and E (expectancy) against historical data, you always want to have a minimum number of trades. The more this number is the more Edge is reliable.
Having a win rate of 100% on a single trade doesn't mean anything at all. But having a win rate of 70% over past 100 trades means clearly something.
Defaults to 10 (it is highly recommended not to decrease this number).
Datatype: Integer
max_trade_duration_minuteEdge will filter out trades with long duration. If a trade is profitable after 1 month, it is hard to evaluate the strategy based on it. But if most of trades are profitable and they have maximum duration of 30 minutes, then it is clearly a good sign.
NOTICE: While configuring this value, you should take into consideration your timeframe. As an example filtering out trades having duration less than one day for a strategy which has 4h interval does not make sense. Default value is set assuming your strategy interval is relatively small (1m or 5m, etc.).
Defaults to 1440 (one day).
Datatype: Integer
remove_pumpsEdge will remove sudden pumps in a given market while going through historical data. However, given that pumps happen very often in crypto markets, we recommend you keep this off.
Defaults to false.
Datatype: Boolean
+

Running Edge independently

+

You can run Edge independently in order to see in details the result. Here is an example:

+
freqtrade edge
+
+

An example of its output:

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
pairstoplosswin raterisk reward ratiorequired risk rewardexpectancytotal number of tradesaverage duration (min)
AGI/BTC-0.020.645.860.563.411454
NXS/BTC-0.030.642.990.571.541126
LEND/BTC-0.020.822.050.221.501136
VIA/BTC-0.010.553.010.831.191148
MTH/BTC-0.090.562.820.801.121852
ARDR/BTC-0.040.423.141.400.731242
BCPT/BTC-0.010.711.340.400.671430
WINGS/BTC-0.020.561.970.800.652742
VIBE/BTC-0.020.830.910.200.591235
MCO/BTC-0.020.790.970.270.551431
GNT/BTC-0.020.502.061.000.531824
HOT/BTC-0.010.177.724.810.502097
SNM/BTC-0.030.711.060.420.451738
APPC/BTC-0.020.442.281.270.442543
NEBL/BTC-0.030.631.290.580.441959
+

Edge produced the above table by comparing calculate_since_number_of_days to minimum_expectancy to find min_trade_number historical information based on the config file. The timerange Edge uses for its comparisons can be further limited by using the --timerange switch.

+

In live and dry-run modes, after the process_throttle_secs has passed, Edge will again process calculate_since_number_of_days against minimum_expectancy to find min_trade_number. If no min_trade_number is found, the bot will return "whitelist empty". Depending on the trade strategy being deployed, "whitelist empty" may be return much of the time - or all of the time. The use of Edge may also cause trading to occur in bursts, though this is rare.

+

If you encounter "whitelist empty" a lot, condsider tuning calculate_since_number_of_days, minimum_expectancy and min_trade_number to align to the trading frequency of your strategy.

+

Update cached pairs with the latest data

+

Edge requires historic data the same way as backtesting does. +Please refer to the Data Downloading section of the documentation for details.

+

Precising stoploss range

+
freqtrade edge --stoplosses=-0.01,-0.1,-0.001 #min,max,step
+
+

Advanced use of timerange

+
freqtrade edge --timerange=20181110-20181113
+
+

Doing --timerange=-20190901 will get all available data until September 1st (excluding September 1st 2019).

+

The full timerange specification:

+
    +
  • Use tickframes till 2018/01/31: --timerange=-20180131
  • +
  • Use tickframes since 2018/01/31: --timerange=20180131-
  • +
  • Use tickframes since 2018/01/31 till 2018/03/01 : --timerange=20180131-20180301
  • +
  • Use tickframes between POSIX timestamps 1527595200 1527618600: --timerange=1527595200-1527618600
  • +
+
+
+
    +
  1. +

    Question extracted from MIT Opencourseware S096 - Mathematics with applications in Finance: https://ocw.mit.edu/courses/mathematics/18-s096-topics-in-mathematics-with-applications-in-finance-fall-2013/ 

    +
  2. +
+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/exchanges/index.html b/en/2023.4/exchanges/index.html new file mode 100644 index 000000000..30edd6384 --- /dev/null +++ b/en/2023.4/exchanges/index.html @@ -0,0 +1,1910 @@ + + + + + + + + + + + + + + + + + + + + + + + + Exchange-specific Notes - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Exchange-specific Notes

+

This page combines common gotchas and Information which are exchange-specific and most likely don't apply to other exchanges.

+

Exchange configuration

+

Freqtrade is based on CCXT library that supports over 100 cryptocurrency +exchange markets and trading APIs. The complete up-to-date list can be found in the +CCXT repo homepage. +However, the bot was tested by the development team with only a few exchanges. +A current list of these can be found in the "Home" section of this documentation.

+

Feel free to test other exchanges and submit your feedback or PR to improve the bot or confirm exchanges that work flawlessly..

+

Some exchanges require special configuration, which can be found below.

+

Sample exchange configuration

+

A exchange configuration for "binance" would look as follows:

+
"exchange": {
+    "name": "binance",
+    "key": "your_exchange_key",
+    "secret": "your_exchange_secret",
+    "ccxt_config": {},
+    "ccxt_async_config": {},
+    // ... 
+
+

Setting rate limits

+

Usually, rate limits set by CCXT are reliable and work well. +In case of problems related to rate-limits (usually DDOS Exceptions in your logs), it's easy to change rateLimit settings to other values.

+
"exchange": {
+    "name": "kraken",
+    "key": "your_exchange_key",
+    "secret": "your_exchange_secret",
+    "ccxt_config": {"enableRateLimit": true},
+    "ccxt_async_config": {
+        "enableRateLimit": true,
+        "rateLimit": 3100
+    },
+
+

This configuration enables kraken, as well as rate-limiting to avoid bans from the exchange. +"rateLimit": 3100 defines a wait-event of 3.1s between each call. This can also be completely disabled by setting "enableRateLimit" to false.

+
+

Note

+

Optimal settings for rate-limiting depend on the exchange and the size of the whitelist, so an ideal parameter will vary on many other settings. +We try to provide sensible defaults per exchange where possible, if you encounter bans please make sure that "enableRateLimit" is enabled and increase the "rateLimit" parameter step by step.

+
+

Binance

+
+

Server location and geo-ip restrictions

+

Please be aware that binance restrict api access regarding the server country. The currents and non exhaustive countries blocked are United States, Malaysia (Singapour), Ontario (Canada). Please go to binance terms > b. Eligibility to find up to date list.

+
+

Binance supports time_in_force.

+
+

Stoploss on Exchange

+

Binance supports stoploss_on_exchange and uses stop-loss-limit orders. It provides great advantages, so we recommend to benefit from it by enabling stoploss on exchange. +On futures, Binance supports both stop-limit as well as stop-market orders. You can use either "limit" or "market" in the order_types.stoploss configuration setting to decide which type to use.

+
+

Binance Blacklist recommendation

+

For Binance, it is suggested to add "BNB/<STAKE>" to your blacklist to avoid issues, unless you are willing to maintain enough extra BNB on the account or unless you're willing to disable using BNB for fees. +Binance accounts may use BNB for fees, and if a trade happens to be on BNB, further trades may consume this position and make the initial BNB trade unsellable as the expected amount is not there anymore.

+

Binance sites

+

Binance has been split into 2, and users must use the correct ccxt exchange ID for their exchange, otherwise API keys are not recognized.

+
    +
  • binance.com - International users. Use exchange id: binance.
  • +
  • binance.us - US based users. Use exchange id: binanceus.
  • +
+

Binance RSA keys

+

Freqtrade supports binance RSA API keys.

+

We recommend to use them as environment variable.

+
export FREQTRADE__EXCHANGE__SECRET="$(cat ./rsa_binance.private)"
+
+

They can however also be configured via configuration file. Since json doesn't support multi-line strings, you'll have to replace all newlines with \n to have a valid json file.

+
// ...
+ "key": "<someapikey>",
+ "secret": "-----BEGIN PRIVATE KEY-----\nMIIEvQIBABACAFQA<...>s8KX8=\n-----END PRIVATE KEY-----"
+// ...
+
+

Binance Futures

+

Binance has specific (unfortunately complex) Futures Trading Quantitative Rules which need to be followed, and which prohibit a too low stake-amount (among others) for too many orders. +Violating these rules will result in a trading restriction.

+

When trading on Binance Futures market, orderbook must be used because there is no price ticker data for futures.

+
  "entry_pricing": {
+      "use_order_book": true,
+      "order_book_top": 1,
+      "check_depth_of_market": {
+          "enabled": false,
+          "bids_to_ask_delta": 1
+      }
+  },
+  "exit_pricing": {
+      "use_order_book": true,
+      "order_book_top": 1
+  },
+
+

Binance futures settings

+

Users will also have to have the futures-setting "Position Mode" set to "One-way Mode", and "Asset Mode" set to "Single-Asset Mode". +These settings will be checked on startup, and freqtrade will show an error if this setting is wrong.

+

Binance futures settings

+

Freqtrade will not attempt to change these settings.

+

Kraken

+
+

Stoploss on Exchange

+

Kraken supports stoploss_on_exchange and can use both stop-loss-market and stop-loss-limit orders. It provides great advantages, so we recommend to benefit from it. +You can use either "limit" or "market" in the order_types.stoploss configuration setting to decide which type to use.

+
+

Historic Kraken data

+

The Kraken API does only provide 720 historic candles, which is sufficient for Freqtrade dry-run and live trade modes, but is a problem for backtesting. +To download data for the Kraken exchange, using --dl-trades is mandatory, otherwise the bot will download the same 720 candles over and over, and you'll not have enough backtest data.

+

Due to the heavy rate-limiting applied by Kraken, the following configuration section should be used to download data:

+
    "ccxt_async_config": {
+        "enableRateLimit": true,
+        "rateLimit": 3100
+    },
+
+
+

Downloading data from kraken

+

Downloading kraken data will require significantly more memory (RAM) than any other exchange, as the trades-data needs to be converted into candles on your machine. +It will also take a long time, as freqtrade will need to download every single trade that happened on the exchange for the pair / timerange combination, therefore please be patient.

+
+
+

rateLimit tuning

+

Please pay attention that rateLimit configuration entry holds delay in milliseconds between requests, NOT requests\sec rate. +So, in order to mitigate Kraken API "Rate limit exceeded" exception, this configuration should be increased, NOT decreased.

+
+

Bittrex

+

Order types

+

Bittrex does not support market orders. If you have a message at the bot startup about this, you should change order type values set in your configuration and/or in the strategy from "market" to "limit". See some more details on this here in the FAQ.

+

Bittrex also does not support VolumePairlist due to limited / split API constellation at the moment. +Please use StaticPairlist. Other pairlists (other than VolumePairlist) should not be affected.

+

Volume pairlist

+

Bittrex does not support the direct usage of VolumePairList. This can however be worked around by using the advanced mode with lookback_days: 1 (or more), which will emulate 24h volume.

+

Read more in the pairlist documentation.

+

Restricted markets

+

Bittrex split its exchange into US and International versions. +The International version has more pairs available, however the API always returns all pairs, so there is currently no automated way to detect if you're affected by the restriction.

+

If you have restricted pairs in your whitelist, you'll get a warning message in the log on Freqtrade startup for each restricted pair.

+

The warning message will look similar to the following:

+
[...] Message: bittrex {"success":false,"message":"RESTRICTED_MARKET","result":null,"explanation":null}"
+
+

If you're an "International" customer on the Bittrex exchange, then this warning will probably not impact you. +If you're a US customer, the bot will fail to create orders for these pairs, and you should remove them from your whitelist.

+

You can get a list of restricted markets by using the following snippet:

+
import ccxt
+ct = ccxt.bittrex()
+lm = ct.load_markets()
+
+res = [p for p, x in lm.items() if 'US' in x['info']['prohibitedIn']]
+print(res)
+
+

Kucoin

+

Kucoin requires a passphrase for each api key, you will therefore need to add this key into the configuration so your exchange section looks as follows:

+
"exchange": {
+    "name": "kucoin",
+    "key": "your_exchange_key",
+    "secret": "your_exchange_secret",
+    "password": "your_exchange_api_key_password",
+    // ...
+}
+
+

Kucoin supports time_in_force.

+
+

Stoploss on Exchange

+

Kucoin supports stoploss_on_exchange and can use both stop-loss-market and stop-loss-limit orders. It provides great advantages, so we recommend to benefit from it. +You can use either "limit" or "market" in the order_types.stoploss configuration setting to decide which type of stoploss shall be used.

+
+

Kucoin Blacklists

+

For Kucoin, it is suggested to add "KCS/<STAKE>" to your blacklist to avoid issues, unless you are willing to maintain enough extra KCS on the account or unless you're willing to disable using KCS for fees. +Kucoin accounts may use KCS for fees, and if a trade happens to be on KCS, further trades may consume this position and make the initial KCS trade unsellable as the expected amount is not there anymore.

+

Huobi

+
+

Stoploss on Exchange

+

Huobi supports stoploss_on_exchange and uses stop-limit orders. It provides great advantages, so we recommend to benefit from it by enabling stoploss on exchange.

+
+

OKX (former OKEX)

+

OKX requires a passphrase for each api key, you will therefore need to add this key into the configuration so your exchange section looks as follows:

+
"exchange": {
+    "name": "okx",
+    "key": "your_exchange_key",
+    "secret": "your_exchange_secret",
+    "password": "your_exchange_api_key_password",
+    // ...
+}
+
+
+

Warning

+

OKX only provides 100 candles per api call. Therefore, the strategy will only have a pretty low amount of data available in backtesting mode.

+
+
+

Futures

+

OKX Futures has the concept of "position mode" - which can be "Buy/Sell" or long/short (hedge mode). +Freqtrade supports both modes (we recommend to use Buy/Sell mode) - but changing the mode mid-trading is not supported and will lead to exceptions and failures to place trades. +OKX also only provides MARK candles for the past ~3 months. Backtesting futures prior to that date will therefore lead to slight deviations, as funding-fees cannot be calculated correctly without this data.

+
+

Gate.io

+
+

Stoploss on Exchange

+

Gate.io supports stoploss_on_exchange and uses stop-loss-limit orders. It provides great advantages, so we recommend to benefit from it by enabling stoploss on exchange..

+
+

Gate.io allows the use of POINT to pay for fees. As this is not a tradable currency (no regular market available), automatic fee calculations will fail (and default to a fee of 0). +The configuration parameter exchange.unknown_fee_rate can be used to specify the exchange rate between Point and the stake currency. Obviously, changing the stake-currency will also require changes to this value.

+

Bybit

+

Futures trading on bybit is currently supported for USDT markets, and will use isolated futures mode. +Users with unified accounts (there's no way back) can create a Sub-account which will start as "non-unified", and can therefore use isolated futures. +On startup, freqtrade will set the position mode to "One-way Mode" for the whole (sub)account. This avoids making this call over and over again (slowing down bot operations), but means that changes to this setting may result in exceptions and errors.

+

As bybit doesn't provide funding rate history, the dry-run calculation is used for live trades as well.

+
+

Stoploss on Exchange

+

Bybit (futures only) supports stoploss_on_exchange and uses stop-loss-limit orders. It provides great advantages, so we recommend to benefit from it by enabling stoploss on exchange. +On futures, Bybit supports both stop-limit as well as stop-market orders. You can use either "limit" or "market" in the order_types.stoploss configuration setting to decide which type to use.

+
+

All exchanges

+

Should you experience constant errors with Nonce (like InvalidNonce), it is best to regenerate the API keys. Resetting Nonce is difficult and it's usually easier to regenerate the API keys.

+

Random notes for other exchanges

+
    +
  • The Ocean (exchange id: theocean) exchange uses Web3 functionality and requires web3 python package to be installed:
  • +
+
$ pip3 install web3
+
+

Getting latest price / Incomplete candles

+

Most exchanges return current incomplete candle via their OHLCV/klines API interface. +By default, Freqtrade assumes that incomplete candle is fetched from the exchange and removes the last candle assuming it's the incomplete candle.

+

Whether your exchange returns incomplete candles or not can be checked using the helper script from the Contributor documentation.

+

Due to the danger of repainting, Freqtrade does not allow you to use this incomplete candle.

+

However, if it is based on the need for the latest price for your strategy - then this requirement can be acquired using the data provider from within the strategy.

+

Advanced Freqtrade Exchange configuration

+

Advanced options can be configured using the _ft_has_params setting, which will override Defaults and exchange-specific behavior.

+

Available options are listed in the exchange-class as _ft_has_default.

+

For example, to test the order type FOK with Kraken, and modify candle limit to 200 (so you only get 200 candles per API call):

+
"exchange": {
+    "name": "kraken",
+    "_ft_has_params": {
+        "order_time_in_force": ["GTC", "FOK"],
+        "ohlcv_candle_limit": 200
+        }
+    //...
+}
+
+
+

Warning

+

Please make sure to fully understand the impacts of these settings before modifying them.

+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/faq/index.html b/en/2023.4/faq/index.html new file mode 100644 index 000000000..149d1b9e4 --- /dev/null +++ b/en/2023.4/faq/index.html @@ -0,0 +1,1902 @@ + + + + + + + + + + + + + + + + + + + + + + + + FAQ - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Freqtrade FAQ

+

Supported Markets

+

Freqtrade supports spot trading, as well as (isolated) futures trading for some selected exchanges. Please refer to the documentation start page for an uptodate list of supported exchanges.

+

Can my bot open short positions?

+

Freqtrade can open short positions in futures markets. +This requires the strategy to be made for this - and "trading_mode": "futures" in the configuration. +Please make sure to read the relevant documentation page first.

+

In spot markets, you can in some cases use leveraged spot tokens, which reflect an inverted pair (eg. BTCUP/USD, BTCDOWN/USD, ETHBULL/USD, ETHBEAR/USD,...) which can be traded with Freqtrade.

+

Can my bot trade options or futures?

+

Futures trading is supported for selected exchanges. Please refer to the documentation start page for an uptodate list of supported exchanges.

+

Beginner Tips & Tricks

+
    +
  • When you work with your strategy & hyperopt file you should use a proper code editor like VSCode or PyCharm. A good code editor will provide syntax highlighting as well as line numbers, making it easy to find syntax errors (most likely pointed out by Freqtrade during startup).
  • +
+

Freqtrade common issues

+

Can freqtrade open multiple positions on the same pair in parallel?

+

No. Freqtrade will only open one position per pair at a time. +You can however use the adjust_trade_position() callback to adjust an open position.

+

Backtesting provides an option for this in --eps - however this is only there to highlight "hidden" signals, and will not work in live.

+

The bot does not start

+

Running the bot with freqtrade trade --config config.json shows the output freqtrade: command not found.

+

This could be caused by the following reasons:

+
    +
  • The virtual environment is not active.
      +
    • Run source .env/bin/activate to activate the virtual environment.
    • +
    +
  • +
  • The installation did not complete successfully. +
  • +
+

I have waited 5 minutes, why hasn't the bot made any trades yet?

+
    +
  • Depending on the buy strategy, the amount of whitelisted coins, the +situation of the market etc, it can take up to hours to find a good entry +position for a trade. Be patient!
  • +
+
    +
  • It may be because of a configuration error. It's best to check the logs, they usually tell you if the bot is simply not getting buy signals (only heartbeat messages), or if there is something wrong (errors / exceptions in the log).
  • +
+

I have made 12 trades already, why is my total profit negative?

+

I understand your disappointment but unfortunately 12 trades is just +not enough to say anything. If you run backtesting, you can see that the +current algorithm does leave you on the plus side, but that is after +thousands of trades and even there, you will be left with losses on +specific coins that you have traded tens if not hundreds of times. We +of course constantly aim to improve the bot but it will always be a +gamble, which should leave you with modest wins on monthly basis but +you can't say much from few trades.

+

I’d like to make changes to the config. Can I do that without having to kill the bot?

+

Yes. You can edit your config and use the /reload_config command to reload the configuration. The bot will stop, reload the configuration and strategy and will restart with the new configuration and strategy.

+

Why does my bot not sell everything it bought?

+

This is called "coin dust" and can happen on all exchanges. +It happens because many exchanges subtract fees from the "receiving currency" - so you buy 100 COIN - but you only get 99.9 COIN. +As COIN is trading in full lot sizes (1COIN steps), you cannot sell 0.9 COIN (or 99.9 COIN) - but you need to round down to 99 COIN.

+

This is not a bot-problem, but will also happen while manual trading.

+

While freqtrade can handle this (it'll sell 99 COIN), fees are often below the minimum tradable lot-size (you can only trade full COIN, not 0.9 COIN). +Leaving the dust (0.9 COIN) on the exchange makes usually sense, as the next time freqtrade buys COIN, it'll eat into the remaining small balance, this time selling everything it bought, and therefore slowly declining the dust balance (although it most likely will never reach exactly 0).

+

Where possible (e.g. on binance), the use of the exchange's dedicated fee currency will fix this. +On binance, it's sufficient to have BNB in your account, and have "Pay fees in BNB" enabled in your profile. Your BNB balance will slowly decline (as it's used to pay fees) - but you'll no longer encounter dust (Freqtrade will include the fees in the profit calculations). +Other exchanges don't offer such possibilities, where it's simply something you'll have to accept or move to a different exchange.

+

I want to use incomplete candles

+

Freqtrade will not provide incomplete candles to strategies. Using incomplete candles will lead to repainting and consequently to strategies with "ghost" buys, which are impossible to both backtest, and verify after they happened.

+

You can use "current" market data by using the dataprovider's orderbook or ticker methods - which however cannot be used during backtesting.

+

Is there a setting to only Exit the trades being held and not perform any new Entries?

+

You can use the /stopentry command in Telegram to prevent future trade entry, followed by /forceexit all (sell all open trades).

+

I want to run multiple bots on the same machine

+

Please look at the advanced setup documentation Page.

+

I'm getting "Missing data fillup" messages in the log

+

This message is just a warning that the latest candles had missing candles in them. +Depending on the exchange, this can indicate that the pair didn't have a trade for the timeframe you are using - and the exchange does only return candles with volume. +On low volume pairs, this is a rather common occurrence.

+

If this happens for all pairs in the pairlist, this might indicate a recent exchange downtime. Please check your exchange's public channels for details.

+

Irrespectively of the reason, Freqtrade will fill up these candles with "empty" candles, where open, high, low and close are set to the previous candle close - and volume is empty. In a chart, this will look like a _ - and is aligned with how exchanges usually represent 0 volume candles.

+

I'm getting "Price jump between 2 candles detected"

+

This message is a warning that the candles had a price jump of > 30%. +This might be a sign that the pair stopped trading, and some token exchange took place (e.g. COCOS in 2021 - where price jumped from 0.0000154 to 0.01621). +This message is often accompanied by "Missing data fillup" - as trading on such pairs is often stopped for some time.

+

I'm getting "Outdated history for pair xxx" in the log

+

The bot is trying to tell you that it got an outdated last candle (not the last complete candle). +As a consequence, Freqtrade will not enter a trade for this pair - as trading on old information is usually not what is desired.

+

This warning can point to one of the below problems:

+
    +
  • Exchange downtime -> Check your exchange status page / blog / twitter feed for details.
  • +
  • Wrong system time -> Ensure your system-time is correct.
  • +
  • Barely traded pair -> Check the pair on the exchange webpage, look at the timeframe your strategy uses. If the pair does not have any volume in some candles (usually visualized with a "volume 0" bar, and a "_" as candle), this pair did not have any trades in this timeframe. These pairs should ideally be avoided, as they can cause problems with order-filling.
  • +
  • API problem -> API returns wrong data (this only here for completeness, and should not happen with supported exchanges).
  • +
+

I'm getting the "RESTRICTED_MARKET" message in the log

+

Currently known to happen for US Bittrex users.

+

Read the Bittrex section about restricted markets for more information.

+

I'm getting the "Exchange XXX does not support market orders." message and cannot run my strategy

+

As the message says, your exchange does not support market orders and you have one of the order types set to "market". Your strategy was probably written with other exchanges in mind and sets "market" orders for "stoploss" orders, which is correct and preferable for most of the exchanges supporting market orders (but not for Bittrex and Gate.io).

+

To fix this, redefine order types in the strategy to use "limit" instead of "market":

+
    order_types = {
+        ...
+        "stoploss": "limit",
+        ...
+    }
+
+

The same fix should be applied in the configuration file, if order types are defined in your custom config rather than in the strategy.

+

I'm trying to start the bot live, but get an API permission error

+

Errors like Invalid API-key, IP, or permissions for action mean exactly what they actually say. +Your API key is either invalid (copy/paste error? check for leading/trailing spaces in the config), expired, or the IP you're running the bot from is not enabled in the Exchange's API console. +Usually, the permission "Spot Trading" (or the equivalent in the exchange you use) will be necessary. +Futures will usually have to be enabled specifically.

+

How do I search the bot logs for something?

+

By default, the bot writes its log into stderr stream. This is implemented this way so that you can easily separate the bot's diagnostics messages from Backtesting, Edge and Hyperopt results, output from other various Freqtrade utility sub-commands, as well as from the output of your custom print()'s you may have inserted into your strategy. So if you need to search the log messages with the grep utility, you need to redirect stderr to stdout and disregard stdout.

+
    +
  • In unix shells, this normally can be done as simple as: +
    $ freqtrade --some-options 2>&1 >/dev/null | grep 'something'
    +
    +(note, 2>&1 and >/dev/null should be written in this order)
  • +
+
    +
  • Bash interpreter also supports so called process substitution syntax, you can grep the log for a string with it as: +
    $ freqtrade --some-options 2> >(grep 'something') >/dev/null
    +
    +or +
    $ freqtrade --some-options 2> >(grep -v 'something' 1>&2)
    +
  • +
+
    +
  • You can also write the copy of Freqtrade log messages to a file with the --logfile option: +
    $ freqtrade --logfile /path/to/mylogfile.log --some-options
    +
    +and then grep it as: +
    $ cat /path/to/mylogfile.log | grep 'something'
    +
    +or even on the fly, as the bot works and the log file grows: +
    $ tail -f /path/to/mylogfile.log | grep 'something'
    +
    +from a separate terminal window.
  • +
+

On Windows, the --logfile option is also supported by Freqtrade and you can use the findstr command to search the log for the string of interest: +

> type \path\to\mylogfile.log | findstr "something"
+

+

Hyperopt module

+

Why does freqtrade not have GPU support?

+

First of all, most indicator libraries don't have GPU support - as such, there would be little benefit for indicator calculations. +The GPU improvements would only apply to pandas-native calculations - or ones written by yourself.

+

For hyperopt, freqtrade is using scikit-optimize, which is built on top of scikit-learn. +Their statement about GPU support is pretty clear.

+

GPU's also are only good at crunching numbers (floating point operations). +For hyperopt, we need both number-crunching (find next parameters) and running python code (running backtesting). +As such, GPU's are not too well suited for most parts of hyperopt.

+

The benefit of using GPU would therefore be pretty slim - and will not justify the complexity introduced by trying to add GPU support.

+

There is however nothing preventing you from using GPU-enabled indicators within your strategy if you think you must have this - you will however probably be disappointed by the slim gain that will give you (compared to the complexity).

+

How many epochs do I need to get a good Hyperopt result?

+

Per default Hyperopt called without the -e/--epochs command line option will only +run 100 epochs, means 100 evaluations of your triggers, guards, ... Too few +to find a great result (unless if you are very lucky), so you probably +have to run it for 10000 or more. But it will take an eternity to +compute.

+

Since hyperopt uses Bayesian search, running for too many epochs may not produce greater results.

+

It's therefore recommended to run between 500-1000 epochs over and over until you hit at least 10000 epochs in total (or are satisfied with the result). You can best judge by looking at the results - if the bot keeps discovering better strategies, it's best to keep on going.

+
freqtrade hyperopt --hyperopt-loss SharpeHyperOptLossDaily --strategy SampleStrategy -e 1000
+
+

Why does it take a long time to run hyperopt?

+
    +
  • Discovering a great strategy with Hyperopt takes time. Study www.freqtrade.io, the Freqtrade Documentation page, join the Freqtrade discord community. While you patiently wait for the most advanced, free crypto bot in the world, to hand you a possible golden strategy specially designed just for you.
  • +
+
    +
  • If you wonder why it can take from 20 minutes to days to do 1000 epochs here are some answers:
  • +
+

This answer was written during the release 0.15.1, when we had:

+
    +
  • 8 triggers
  • +
  • 9 guards: let's say we evaluate even 10 values from each
  • +
  • 1 stoploss calculation: let's say we want 10 values from that too to be evaluated
  • +
+

The following calculation is still very rough and not very precise +but it will give the idea. With only these triggers and guards there is +already 8*10^9*10 evaluations. A roughly total of 80 billion evaluations. +Did you run 100 000 evaluations? Congrats, you've done roughly 1 / 100 000 th +of the search space, assuming that the bot never tests the same parameters more than once.

+
    +
  • The time it takes to run 1000 hyperopt epochs depends on things like: The available cpu, hard-disk, ram, timeframe, timerange, indicator settings, indicator count, amount of coins that hyperopt test strategies on and the resulting trade count - which can be 650 trades in a year or 100000 trades depending if the strategy aims for big profits by trading rarely or for many low profit trades.
  • +
+

Example: 4% profit 650 times vs 0,3% profit a trade 10000 times in a year. If we assume you set the --timerange to 365 days.

+

Example: +freqtrade --config config.json --strategy SampleStrategy --hyperopt SampleHyperopt -e 1000 --timerange 20190601-20200601

+

Edge module

+

Edge implements interesting approach for controlling position size, is there any theory behind it?

+

The Edge module is mostly a result of brainstorming of @mishaker and @creslinux freqtrade team members.

+

You can find further info on expectancy, win rate, risk management and position size in the following sources:

+ +

Official channels

+

Freqtrade is using exclusively the following official channels:

+ +

Nobody affiliated with the freqtrade project will ask you about your exchange keys or anything else exposing your funds to exploitation. +Should you be asked to expose your exchange keys or send funds to some random wallet, then please don't follow these instructions.

+

Failing to follow these guidelines will not be responsibility of freqtrade.

+

"Freqtrade token"

+

Freqtrade does not have a Crypto token offering.

+

Token offerings you find on the internet referring Freqtrade, FreqAI or freqUI must be considered to be a scam, trying to exploit freqtrade's popularity for their own, nefarious gains.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/freqai-configuration/index.html b/en/2023.4/freqai-configuration/index.html new file mode 100644 index 000000000..c4edf0d99 --- /dev/null +++ b/en/2023.4/freqai-configuration/index.html @@ -0,0 +1,1872 @@ + + + + + + + + + + + + + + + + + + + + + + + + Configuration - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Configuration

+

FreqAI is configured through the typical Freqtrade config file and the standard Freqtrade strategy. Examples of FreqAI config and strategy files can be found in config_examples/config_freqai.example.json and freqtrade/templates/FreqaiExampleStrategy.py, respectively.

+

Setting up the configuration file

+

Although there are plenty of additional parameters to choose from, as highlighted in the parameter table, a FreqAI config must at minimum include the following parameters (the parameter values are only examples):

+
    "freqai": {
+        "enabled": true,
+        "purge_old_models": 2,
+        "train_period_days": 30,
+        "backtest_period_days": 7,
+        "identifier" : "unique-id",
+        "feature_parameters" : {
+            "include_timeframes": ["5m","15m","4h"],
+            "include_corr_pairlist": [
+                "ETH/USD",
+                "LINK/USD",
+                "BNB/USD"
+            ],
+            "label_period_candles": 24,
+            "include_shifted_candles": 2,
+            "indicator_periods_candles": [10, 20]
+        },
+        "data_split_parameters" : {
+            "test_size": 0.25
+        }
+    }
+
+

A full example config is available in config_examples/config_freqai.example.json.

+

Building a FreqAI strategy

+

The FreqAI strategy requires including the following lines of code in the standard Freqtrade strategy:

+
    # user should define the maximum startup candle count (the largest number of candles
+    # passed to any single indicator)
+    startup_candle_count: int = 20
+
+    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+
+        # the model will return all labels created by user in `set_freqai_labels()`
+        # (& appended targets), an indication of whether or not the prediction should be accepted,
+        # the target mean/std values for each of the labels created by user in
+        # `feature_engineering_*` for each training period.
+
+        dataframe = self.freqai.start(dataframe, metadata, self)
+
+        return dataframe
+
+    def feature_engineering_expand_all(self, dataframe: DataFrame, period, **kwargs) -> DataFrame:
+        """
+        *Only functional with FreqAI enabled strategies*
+        This function will automatically expand the defined features on the config defined
+        `indicator_periods_candles`, `include_timeframes`, `include_shifted_candles`, and
+        `include_corr_pairs`. In other words, a single feature defined in this function
+        will automatically expand to a total of
+        `indicator_periods_candles` * `include_timeframes` * `include_shifted_candles` *
+        `include_corr_pairs` numbers of features added to the model.
+
+        All features must be prepended with `%` to be recognized by FreqAI internals.
+
+        :param df: strategy dataframe which will receive the features
+        :param period: period of the indicator - usage example:
+        dataframe["%-ema-period"] = ta.EMA(dataframe, timeperiod=period)
+        """
+
+        dataframe["%-rsi-period"] = ta.RSI(dataframe, timeperiod=period)
+        dataframe["%-mfi-period"] = ta.MFI(dataframe, timeperiod=period)
+        dataframe["%-adx-period"] = ta.ADX(dataframe, timeperiod=period)
+        dataframe["%-sma-period"] = ta.SMA(dataframe, timeperiod=period)
+        dataframe["%-ema-period"] = ta.EMA(dataframe, timeperiod=period)
+
+        return dataframe
+
+    def feature_engineering_expand_basic(self, dataframe: DataFrame, **kwargs) -> DataFrame:
+        """
+        *Only functional with FreqAI enabled strategies*
+        This function will automatically expand the defined features on the config defined
+        `include_timeframes`, `include_shifted_candles`, and `include_corr_pairs`.
+        In other words, a single feature defined in this function
+        will automatically expand to a total of
+        `include_timeframes` * `include_shifted_candles` * `include_corr_pairs`
+        numbers of features added to the model.
+
+        Features defined here will *not* be automatically duplicated on user defined
+        `indicator_periods_candles`
+
+        All features must be prepended with `%` to be recognized by FreqAI internals.
+
+        :param df: strategy dataframe which will receive the features
+        dataframe["%-pct-change"] = dataframe["close"].pct_change()
+        dataframe["%-ema-200"] = ta.EMA(dataframe, timeperiod=200)
+        """
+        dataframe["%-pct-change"] = dataframe["close"].pct_change()
+        dataframe["%-raw_volume"] = dataframe["volume"]
+        dataframe["%-raw_price"] = dataframe["close"]
+        return dataframe
+
+    def feature_engineering_standard(self, dataframe: DataFrame, **kwargs) -> DataFrame:
+        """
+        *Only functional with FreqAI enabled strategies*
+        This optional function will be called once with the dataframe of the base timeframe.
+        This is the final function to be called, which means that the dataframe entering this
+        function will contain all the features and columns created by all other
+        freqai_feature_engineering_* functions.
+
+        This function is a good place to do custom exotic feature extractions (e.g. tsfresh).
+        This function is a good place for any feature that should not be auto-expanded upon
+        (e.g. day of the week).
+
+        All features must be prepended with `%` to be recognized by FreqAI internals.
+
+        :param df: strategy dataframe which will receive the features
+        usage example: dataframe["%-day_of_week"] = (dataframe["date"].dt.dayofweek + 1) / 7
+        """
+        dataframe["%-day_of_week"] = (dataframe["date"].dt.dayofweek + 1) / 7
+        dataframe["%-hour_of_day"] = (dataframe["date"].dt.hour + 1) / 25
+        return dataframe
+
+    def set_freqai_targets(self, dataframe: DataFrame, **kwargs) -> DataFrame:
+        """
+        *Only functional with FreqAI enabled strategies*
+        Required function to set the targets for the model.
+        All targets must be prepended with `&` to be recognized by the FreqAI internals.
+
+        :param df: strategy dataframe which will receive the targets
+        usage example: dataframe["&-target"] = dataframe["close"].shift(-1) / dataframe["close"]
+        """
+        dataframe["&-s_close"] = (
+            dataframe["close"]
+            .shift(-self.freqai_info["feature_parameters"]["label_period_candles"])
+            .rolling(self.freqai_info["feature_parameters"]["label_period_candles"])
+            .mean()
+            / dataframe["close"]
+            - 1
+            )
+        return dataframe
+
+

Notice how the feature_engineering_*() is where features are added. Meanwhile set_freqai_targets() adds the labels/targets. A full example strategy is available in templates/FreqaiExampleStrategy.py.

+
+

Note

+

The self.freqai.start() function cannot be called outside the populate_indicators().

+
+
+

Note

+

Features must be defined in feature_engineering_*(). Defining FreqAI features in populate_indicators() +will cause the algorithm to fail in live/dry mode. In order to add generalized features that are not associated with a specific pair or timeframe, you should use feature_engineering_standard() +(as exemplified in freqtrade/templates/FreqaiExampleStrategy.py).

+
+

Important dataframe key patterns

+

Below are the values you can expect to include/use inside a typical strategy dataframe (df[]):

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
DataFrame KeyDescription
df['&*']Any dataframe column prepended with & in set_freqai_targets() is treated as a training target (label) inside FreqAI (typically following the naming convention &-s*). For example, to predict the close price 40 candles into the future, you would set df['&-s_close'] = df['close'].shift(-self.freqai_info["feature_parameters"]["label_period_candles"]) with "label_period_candles": 40 in the config. FreqAI makes the predictions and gives them back under the same key (df['&-s_close']) to be used in populate_entry/exit_trend().
Datatype: Depends on the output of the model.
df['&*_std/mean']Standard deviation and mean values of the defined labels during training (or live tracking with fit_live_predictions_candles). Commonly used to understand the rarity of a prediction (use the z-score as shown in templates/FreqaiExampleStrategy.py and explained here to evaluate how often a particular prediction was observed during training or historically with fit_live_predictions_candles).
Datatype: Float.
df['do_predict']Indication of an outlier data point. The return value is integer between -2 and 2, which lets you know if the prediction is trustworthy or not. do_predict==1 means that the prediction is trustworthy. If the Dissimilarity Index (DI, see details here) of the input data point is above the threshold defined in the config, FreqAI will subtract 1 from do_predict, resulting in do_predict==0. If use_SVM_to_remove_outliers() is active, the Support Vector Machine (SVM, see details here) may also detect outliers in training and prediction data. In this case, the SVM will also subtract 1 from do_predict. If the input data point was considered an outlier by the SVM but not by the DI, or vice versa, the result will be do_predict==0. If both the DI and the SVM considers the input data point to be an outlier, the result will be do_predict==-1. As with the SVM, if use_DBSCAN_to_remove_outliers is active, DBSCAN (see details here) may also detect outliers and subtract 1 from do_predict. Hence, if both the SVM and DBSCAN are active and identify a datapoint that was above the DI threshold as an outlier, the result will be do_predict==-2. A particular case is when do_predict == 2, which means that the model has expired due to exceeding expired_hours.
Datatype: Integer between -2 and 2.
df['DI_values']Dissimilarity Index (DI) values are proxies for the level of confidence FreqAI has in the prediction. A lower DI means the prediction is close to the training data, i.e., higher prediction confidence. See details about the DI here.
Datatype: Float.
df['%*']Any dataframe column prepended with % in feature_engineering_*() is treated as a training feature. For example, you can include the RSI in the training feature set (similar to in templates/FreqaiExampleStrategy.py) by setting df['%-rsi']. See more details on how this is done here.
Note: Since the number of features prepended with % can multiply very quickly (10s of thousands of features are easily engineered using the multiplictative functionality of, e.g., include_shifted_candles and include_timeframes as described in the parameter table), these features are removed from the dataframe that is returned from FreqAI to the strategy. To keep a particular type of feature for plotting purposes, you would prepend it with %%.
Datatype: Depends on the output of the model.
+

Setting the startup_candle_count

+

The startup_candle_count in the FreqAI strategy needs to be set up in the same way as in the standard Freqtrade strategy (see details here). This value is used by Freqtrade to ensure that a sufficient amount of data is provided when calling the dataprovider, to avoid any NaNs at the beginning of the first training. You can easily set this value by identifying the longest period (in candle units) which is passed to the indicator creation functions (e.g., TA-Lib functions). In the presented example, startup_candle_count is 20 since this is the maximum value in indicators_periods_candles.

+
+

Note

+

There are instances where the TA-Lib functions actually require more data than just the passed period or else the feature dataset gets populated with NaNs. Anecdotally, multiplying the startup_candle_count by 2 always leads to a fully NaN free training dataset. Hence, it is typically safest to multiply the expected startup_candle_count by 2. Look out for this log message to confirm that the data is clean:

+
2022-08-31 15:14:04 - freqtrade.freqai.data_kitchen - INFO - dropped 0 training points due to NaNs in populated dataset 4319.
+
+
+

Creating a dynamic target threshold

+

Deciding when to enter or exit a trade can be done in a dynamic way to reflect current market conditions. FreqAI allows you to return additional information from the training of a model (more info here). For example, the &*_std/mean return values describe the statistical distribution of the target/label during the most recent training. Comparing a given prediction to these values allows you to know the rarity of the prediction. In templates/FreqaiExampleStrategy.py, the target_roi and sell_roi are defined to be 1.25 z-scores away from the mean which causes predictions that are closer to the mean to be filtered out.

+
dataframe["target_roi"] = dataframe["&-s_close_mean"] + dataframe["&-s_close_std"] * 1.25
+dataframe["sell_roi"] = dataframe["&-s_close_mean"] - dataframe["&-s_close_std"] * 1.25
+
+

To consider the population of historical predictions for creating the dynamic target instead of information from the training as discussed above, you would set fit_live_predictions_candles in the config to the number of historical prediction candles you wish to use to generate target statistics.

+
    "freqai": {
+        "fit_live_predictions_candles": 300,
+    }
+
+

If this value is set, FreqAI will initially use the predictions from the training data and subsequently begin introducing real prediction data as it is generated. FreqAI will save this historical data to be reloaded if you stop and restart a model with the same identifier.

+

Using different prediction models

+

FreqAI has multiple example prediction model libraries that are ready to be used as is via the flag --freqaimodel. These libraries include CatBoost, LightGBM, and XGBoost regression, classification, and multi-target models, and can be found in freqai/prediction_models/.

+

Regression and classification models differ in what targets they predict - a regression model will predict a target of continuous values, for example what price BTC will be at tomorrow, whilst a classifier will predict a target of discrete values, for example if the price of BTC will go up tomorrow or not. This means that you have to specify your targets differently depending on which model type you are using (see details below).

+

All of the aforementioned model libraries implement gradient boosted decision tree algorithms. They all work on the principle of ensemble learning, where predictions from multiple simple learners are combined to get a final prediction that is more stable and generalized. The simple learners in this case are decision trees. Gradient boosting refers to the method of learning, where each simple learner is built in sequence - the subsequent learner is used to improve on the error from the previous learner. If you want to learn more about the different model libraries you can find the information in their respective docs:

+ +

There are also numerous online articles describing and comparing the algorithms. Some relatively lightweight examples would be CatBoost vs. LightGBM vs. XGBoost — Which is the best algorithm? and XGBoost, LightGBM or CatBoost — which boosting algorithm should I use?. Keep in mind that the performance of each model is highly dependent on the application and so any reported metrics might not be true for your particular use of the model.

+

Apart from the models already available in FreqAI, it is also possible to customize and create your own prediction models using the IFreqaiModel class. You are encouraged to inherit fit(), train(), and predict() to customize various aspects of the training procedures. You can place custom FreqAI models in user_data/freqaimodels - and freqtrade will pick them up from there based on the provided --freqaimodel name - which has to correspond to the class name of your custom model. +Make sure to use unique names to avoid overriding built-in models.

+

Setting model targets

+

Regressors

+

If you are using a regressor, you need to specify a target that has continuous values. FreqAI includes a variety of regressors, such as the CatboostRegressorvia the flag --freqaimodel CatboostRegressor. An example of how you could set a regression target for predicting the price 100 candles into the future would be

+
df['&s-close_price'] = df['close'].shift(-100)
+
+

If you want to predict multiple targets, you need to define multiple labels using the same syntax as shown above.

+

Classifiers

+

If you are using a classifier, you need to specify a target that has discrete values. FreqAI includes a variety of classifiers, such as the CatboostClassifier via the flag --freqaimodel CatboostClassifier. If you elects to use a classifier, the classes need to be set using strings. For example, if you want to predict if the price 100 candles into the future goes up or down you would set

+
df['&s-up_or_down'] = np.where( df["close"].shift(-100) > df["close"], 'up', 'down')
+
+

If you want to predict multiple targets you must specify all labels in the same label column. You could, for example, add the label same to define where the price was unchanged by setting

+
df['&s-up_or_down'] = np.where( df["close"].shift(-100) > df["close"], 'up', 'down')
+df['&s-up_or_down'] = np.where( df["close"].shift(-100) == df["close"], 'same', df['&s-up_or_down'])
+
+

PyTorch Module

+

Quick start

+

The easiest way to quickly run a pytorch model is with the following command (for regression task):

+
freqtrade trade --config config_examples/config_freqai.example.json --strategy FreqaiExampleStrategy --freqaimodel PyTorchMLPRegressor --strategy-path freqtrade/templates 
+
+
+

Installation/docker

+

The PyTorch module requires large packages such as torch, which should be explicitly requested during ./setup.sh -i by answering "y" to the question "Do you also want dependencies for freqai-rl or PyTorch (~700mb additional space required) [y/N]?". +Users who prefer docker should ensure they use the docker image appended with _freqaitorch.

+
+

Structure

+

Model

+

You can construct your own Neural Network architecture in PyTorch by simply defining your nn.Module class inside your custom IFreqaiModel file and then using that class in your def train() function. Here is an example of logistic regression model implementation using PyTorch (should be used with nn.BCELoss criterion) for classification tasks.

+
class LogisticRegression(nn.Module):
+    def __init__(self, input_size: int):
+        super().__init__()
+        # Define your layers
+        self.linear = nn.Linear(input_size, 1)
+        self.activation = nn.Sigmoid()
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        # Define the forward pass
+        out = self.linear(x)
+        out = self.activation(out)
+        return out
+
+class MyCoolPyTorchClassifier(BasePyTorchClassifier):
+    """
+    This is a custom IFreqaiModel showing how a user might setup their own 
+    custom Neural Network architecture for their training.
+    """
+
+    @property
+    def data_convertor(self) -> PyTorchDataConvertor:
+        return DefaultPyTorchDataConvertor(target_tensor_type=torch.float)
+
+    def __init__(self, **kwargs) -> None:
+        super().__init__(**kwargs)
+        config = self.freqai_info.get("model_training_parameters", {})
+        self.learning_rate: float = config.get("learning_rate",  3e-4)
+        self.model_kwargs: Dict[str, Any] = config.get("model_kwargs",  {})
+        self.trainer_kwargs: Dict[str, Any] = config.get("trainer_kwargs",  {})
+
+    def fit(self, data_dictionary: Dict, dk: FreqaiDataKitchen, **kwargs) -> Any:
+        """
+        User sets up the training and test data to fit their desired model here
+        :param data_dictionary: the dictionary holding all data for train, test,
+            labels, weights
+        :param dk: The datakitchen object for the current coin/model
+        """
+
+        class_names = self.get_class_names()
+        self.convert_label_column_to_int(data_dictionary, dk, class_names)
+        n_features = data_dictionary["train_features"].shape[-1]
+        model = LogisticRegression(
+            input_dim=n_features
+        )
+        model.to(self.device)
+        optimizer = torch.optim.AdamW(model.parameters(), lr=self.learning_rate)
+        criterion = torch.nn.CrossEntropyLoss()
+        init_model = self.get_init_model(dk.pair)
+        trainer = PyTorchModelTrainer(
+            model=model,
+            optimizer=optimizer,
+            criterion=criterion,
+            model_meta_data={"class_names": class_names},
+            device=self.device,
+            init_model=init_model,
+            data_convertor=self.data_convertor,
+            **self.trainer_kwargs,
+        )
+        trainer.fit(data_dictionary, self.splits)
+        return trainer
+
+

Trainer

+

The PyTorchModelTrainer performs the idiomatic PyTorch train loop: +Define our model, loss function, and optimizer, and then move them to the appropriate device (GPU or CPU). Inside the loop, we iterate through the batches in the dataloader, move the data to the device, compute the prediction and loss, backpropagate, and update the model parameters using the optimizer.

+

In addition, the trainer is responsible for the following: + - saving and loading the model + - converting the data from pandas.DataFrame to torch.Tensor.

+

Integration with Freqai module

+

Like all freqai models, PyTorch models inherit IFreqaiModel. IFreqaiModel declares three abstract methods: train, fit, and predict. we implement these methods in three levels of hierarchy. +From top to bottom:

+
    +
  1. BasePyTorchModel - Implements the train method. all BasePyTorch* inherit it. responsible for general data preparation (e.g., data normalization) and calling the fit method. Sets device attribute used by children classes. Sets model_type attribute used by the parent class.
  2. +
  3. BasePyTorch* - Implements the predict method. Here, the * represents a group of algorithms, such as classifiers or regressors. responsible for data preprocessing, predicting, and postprocessing if needed.
  4. +
  5. PyTorch*Classifier / PyTorch*Regressor - implements the fit method. responsible for the main train flaw, where we initialize the trainer and model objects.
  6. +
+

image

+

Full example

+

Building a PyTorch regressor using MLP (multilayer perceptron) model, MSELoss criterion, and AdamW optimizer.

+
class PyTorchMLPRegressor(BasePyTorchRegressor):
+    def __init__(self, **kwargs) -> None:
+        super().__init__(**kwargs)
+        config = self.freqai_info.get("model_training_parameters", {})
+        self.learning_rate: float = config.get("learning_rate",  3e-4)
+        self.model_kwargs: Dict[str, Any] = config.get("model_kwargs",  {})
+        self.trainer_kwargs: Dict[str, Any] = config.get("trainer_kwargs",  {})
+
+    def fit(self, data_dictionary: Dict, dk: FreqaiDataKitchen, **kwargs) -> Any:
+        n_features = data_dictionary["train_features"].shape[-1]
+        model = PyTorchMLPModel(
+            input_dim=n_features,
+            output_dim=1,
+            **self.model_kwargs
+        )
+        model.to(self.device)
+        optimizer = torch.optim.AdamW(model.parameters(), lr=self.learning_rate)
+        criterion = torch.nn.MSELoss()
+        init_model = self.get_init_model(dk.pair)
+        trainer = PyTorchModelTrainer(
+            model=model,
+            optimizer=optimizer,
+            criterion=criterion,
+            device=self.device,
+            init_model=init_model,
+            target_tensor_type=torch.float,
+            **self.trainer_kwargs,
+        )
+        trainer.fit(data_dictionary)
+        return trainer
+
+

Here we create a PyTorchMLPRegressor class that implements the fit method. The fit method specifies the training building blocks: model, optimizer, criterion, and trainer. We inherit both BasePyTorchRegressor and BasePyTorchModel, where the former implements the predict method that is suitable for our regression task, and the latter implements the train method.

+
+Setting Class Names for Classifiers +

When using classifiers, the user must declare the class names (or targets) by overriding the IFreqaiModel.class_names attribute. This is achieved by setting self.freqai.class_names in the FreqAI strategy inside the set_freqai_targets method.

+

For example, if you are using a binary classifier to predict price movements as up or down, you can set the class names as follows: +

def set_freqai_targets(self, dataframe: DataFrame, metadata: Dict, **kwargs) -> DataFrame:
+    self.freqai.class_names = ["down", "up"]
+    dataframe['&s-up_or_down'] = np.where(dataframe["close"].shift(-100) >
+                                              dataframe["close"], 'up', 'down')
+
+    return dataframe
+
+To see a full example, you can refer to the classifier test strategy class.

+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/freqai-developers/index.html b/en/2023.4/freqai-developers/index.html new file mode 100644 index 000000000..dd4f79902 --- /dev/null +++ b/en/2023.4/freqai-developers/index.html @@ -0,0 +1,1399 @@ + + + + + + + + + + + + + + + + + + + + + + + + Developer guide - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Development

+

Project architecture

+

The architecture and functions of FreqAI are generalized to encourages development of unique features, functions, models, etc.

+

The class structure and a detailed algorithmic overview is depicted in the following diagram:

+

image

+

As shown, there are three distinct objects comprising FreqAI:

+
    +
  • IFreqaiModel - A singular persistent object containing all the necessary logic to collect, store, and process data, engineer features, run training, and inference models.
  • +
  • FreqaiDataKitchen - A non-persistent object which is created uniquely for each unique asset/model. Beyond metadata, it also contains a variety of data processing tools.
  • +
  • FreqaiDataDrawer - A singular persistent object containing all the historical predictions, models, and save/load methods.
  • +
+

There are a variety of built-in prediction models which inherit directly from IFreqaiModel. Each of these models have full access to all methods in IFreqaiModel and can therefore override any of those functions at will. However, advanced users will likely stick to overriding fit(), train(), predict(), and data_cleaning_train/predict().

+

Data handling

+

FreqAI aims to organize model files, prediction data, and meta data in a way that simplifies post-processing and enhances crash resilience by automatic data reloading. The data is saved in a file structure,user_data_dir/models/, which contains all the data associated with the trainings and backtests. The FreqaiDataKitchen() relies heavily on the file structure for proper training and inferencing and should therefore not be manually modified.

+

File structure

+

The file structure is automatically generated based on the model identifier set in the config. The following structure shows where the data is stored for post processing:

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
StructureDescription
config_*.jsonA copy of the model specific configuration file.
historic_predictions.pklA file containing all historic predictions generated during the lifetime of the identifier model during live deployment. historic_predictions.pkl is used to reload the model after a crash or a config change. A backup file is always held in case of corruption on the main file. FreqAI automatically detects corruption and replaces the corrupted file with the backup.
pair_dictionary.jsonA file containing the training queue as well as the on disk location of the most recently trained model.
sub-train-*_TIMESTAMPA folder containing all the files associated with a single model, such as:
*_metadata.json - Metadata for the model, such as normalization max/min, expected training feature list, etc.
*_model.* - The model file saved to disk for reloading from a crash. Can be joblib (typical boosting libs), zip (stable_baselines), hd5 (keras type), etc.
*_pca_object.pkl - The Principal component analysis (PCA) transform (if principal_component_analysis: True is set in the config) which will be used to transform unseen prediction features.
*_svm_model.pkl - The Support Vector Machine (SVM) model (if use_SVM_to_remove_outliers: True is set in the config) which is used to detect outliers in unseen prediction features.
*_trained_df.pkl - The dataframe containing all the training features used to train the identifier model. This is used for computing the Dissimilarity Index (DI) and can also be used for post-processing.
*_trained_dates.df.pkl - The dates associated with the trained_df.pkl, which is useful for post-processing.
+

The example file structure would look like this:

+
├── models
+│   └── unique-id
+│       ├── config_freqai.example.json
+│       ├── historic_predictions.backup.pkl
+│       ├── historic_predictions.pkl
+│       ├── pair_dictionary.json
+│       ├── sub-train-1INCH_1662821319
+│       │   ├── cb_1inch_1662821319_metadata.json
+│       │   ├── cb_1inch_1662821319_model.joblib
+│       │   ├── cb_1inch_1662821319_pca_object.pkl
+│       │   ├── cb_1inch_1662821319_svm_model.joblib
+│       │   ├── cb_1inch_1662821319_trained_dates_df.pkl
+│       │   └── cb_1inch_1662821319_trained_df.pkl
+│       ├── sub-train-1INCH_1662821371
+│       │   ├── cb_1inch_1662821371_metadata.json
+│       │   ├── cb_1inch_1662821371_model.joblib
+│       │   ├── cb_1inch_1662821371_pca_object.pkl
+│       │   ├── cb_1inch_1662821371_svm_model.joblib
+│       │   ├── cb_1inch_1662821371_trained_dates_df.pkl
+│       │   └── cb_1inch_1662821371_trained_df.pkl
+│       ├── sub-train-ADA_1662821344
+│       │   ├── cb_ada_1662821344_metadata.json
+│       │   ├── cb_ada_1662821344_model.joblib
+│       │   ├── cb_ada_1662821344_pca_object.pkl
+│       │   ├── cb_ada_1662821344_svm_model.joblib
+│       │   ├── cb_ada_1662821344_trained_dates_df.pkl
+│       │   └── cb_ada_1662821344_trained_df.pkl
+│       └── sub-train-ADA_1662821399
+│           ├── cb_ada_1662821399_metadata.json
+│           ├── cb_ada_1662821399_model.joblib
+│           ├── cb_ada_1662821399_pca_object.pkl
+│           ├── cb_ada_1662821399_svm_model.joblib
+│           ├── cb_ada_1662821399_trained_dates_df.pkl
+│           └── cb_ada_1662821399_trained_df.pkl
+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/freqai-feature-engineering/index.html b/en/2023.4/freqai-feature-engineering/index.html new file mode 100644 index 000000000..6611b2357 --- /dev/null +++ b/en/2023.4/freqai-feature-engineering/index.html @@ -0,0 +1,1694 @@ + + + + + + + + + + + + + + + + + + + + + + + + Feature engineering - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Feature engineering

+

Defining the features

+

Low level feature engineering is performed in the user strategy within a set of functions called feature_engineering_*. These function set the base features such as, RSI, MFI, EMA, SMA, time of day, volume, etc. The base features can be custom indicators or they can be imported from any technical-analysis library that you can find. FreqAI is equipped with a set of functions to simplify rapid large-scale feature engineering:

+ + + + + + + + + + + + + + + + + + + + + + + + + +
FunctionDescription
feature_engineering_expand_all()This optional function will automatically expand the defined features on the config defined indicator_periods_candles, include_timeframes, include_shifted_candles, and include_corr_pairs.
feature_engineering_expand_basic()This optional function will automatically expand the defined features on the config defined include_timeframes, include_shifted_candles, and include_corr_pairs. Note: this function does not expand across include_periods_candles.
feature_engineering_standard()This optional function will be called once with the dataframe of the base timeframe. This is the final function to be called, which means that the dataframe entering this function will contain all the features and columns from the base asset created by the other feature_engineering_expand functions. This function is a good place to do custom exotic feature extractions (e.g. tsfresh). This function is also a good place for any feature that should not be auto-expanded upon (e.g., day of the week).
set_freqai_targets()Required function to set the targets for the model. All targets must be prepended with & to be recognized by the FreqAI internals.
+

Meanwhile, high level feature engineering is handled within "feature_parameters":{} in the FreqAI config. Within this file, it is possible to decide large scale feature expansions on top of the base_features such as "including correlated pairs" or "including informative timeframes" or even "including recent candles."

+

It is advisable to start from the template feature_engineering_* functions in the source provided example strategy (found in templates/FreqaiExampleStrategy.py) to ensure that the feature definitions are following the correct conventions. Here is an example of how to set the indicators and labels in the strategy:

+
    def feature_engineering_expand_all(self, dataframe: DataFrame, period, metadata, **kwargs) -> DataFrame:
+        """
+        *Only functional with FreqAI enabled strategies*
+        This function will automatically expand the defined features on the config defined
+        `indicator_periods_candles`, `include_timeframes`, `include_shifted_candles`, and
+        `include_corr_pairs`. In other words, a single feature defined in this function
+        will automatically expand to a total of
+        `indicator_periods_candles` * `include_timeframes` * `include_shifted_candles` *
+        `include_corr_pairs` numbers of features added to the model.
+
+        All features must be prepended with `%` to be recognized by FreqAI internals.
+
+        Access metadata such as the current pair/timeframe/period with:
+
+        `metadata["pair"]` `metadata["tf"]`  `metadata["period"]`
+
+        :param df: strategy dataframe which will receive the features
+        :param period: period of the indicator - usage example:
+        :param metadata: metadata of current pair
+        dataframe["%-ema-period"] = ta.EMA(dataframe, timeperiod=period)
+        """
+
+        dataframe["%-rsi-period"] = ta.RSI(dataframe, timeperiod=period)
+        dataframe["%-mfi-period"] = ta.MFI(dataframe, timeperiod=period)
+        dataframe["%-adx-period"] = ta.ADX(dataframe, timeperiod=period)
+        dataframe["%-sma-period"] = ta.SMA(dataframe, timeperiod=period)
+        dataframe["%-ema-period"] = ta.EMA(dataframe, timeperiod=period)
+
+        bollinger = qtpylib.bollinger_bands(
+            qtpylib.typical_price(dataframe), window=period, stds=2.2
+        )
+        dataframe["bb_lowerband-period"] = bollinger["lower"]
+        dataframe["bb_middleband-period"] = bollinger["mid"]
+        dataframe["bb_upperband-period"] = bollinger["upper"]
+
+        dataframe["%-bb_width-period"] = (
+            dataframe["bb_upperband-period"]
+            - dataframe["bb_lowerband-period"]
+        ) / dataframe["bb_middleband-period"]
+        dataframe["%-close-bb_lower-period"] = (
+            dataframe["close"] / dataframe["bb_lowerband-period"]
+        )
+
+        dataframe["%-roc-period"] = ta.ROC(dataframe, timeperiod=period)
+
+        dataframe["%-relative_volume-period"] = (
+            dataframe["volume"] / dataframe["volume"].rolling(period).mean()
+        )
+
+        return dataframe
+
+    def feature_engineering_expand_basic(self, dataframe: DataFrame, metadata, **kwargs) -> DataFrame:
+        """
+        *Only functional with FreqAI enabled strategies*
+        This function will automatically expand the defined features on the config defined
+        `include_timeframes`, `include_shifted_candles`, and `include_corr_pairs`.
+        In other words, a single feature defined in this function
+        will automatically expand to a total of
+        `include_timeframes` * `include_shifted_candles` * `include_corr_pairs`
+        numbers of features added to the model.
+
+        Features defined here will *not* be automatically duplicated on user defined
+        `indicator_periods_candles`
+
+        Access metadata such as the current pair/timeframe with:
+
+        `metadata["pair"]` `metadata["tf"]`
+
+        All features must be prepended with `%` to be recognized by FreqAI internals.
+
+        :param df: strategy dataframe which will receive the features
+        :param metadata: metadata of current pair
+        dataframe["%-pct-change"] = dataframe["close"].pct_change()
+        dataframe["%-ema-200"] = ta.EMA(dataframe, timeperiod=200)
+        """
+        dataframe["%-pct-change"] = dataframe["close"].pct_change()
+        dataframe["%-raw_volume"] = dataframe["volume"]
+        dataframe["%-raw_price"] = dataframe["close"]
+        return dataframe
+
+    def feature_engineering_standard(self, dataframe: DataFrame, metadata, **kwargs) -> DataFrame:
+        """
+        *Only functional with FreqAI enabled strategies*
+        This optional function will be called once with the dataframe of the base timeframe.
+        This is the final function to be called, which means that the dataframe entering this
+        function will contain all the features and columns created by all other
+        freqai_feature_engineering_* functions.
+
+        This function is a good place to do custom exotic feature extractions (e.g. tsfresh).
+        This function is a good place for any feature that should not be auto-expanded upon
+        (e.g. day of the week).
+
+        Access metadata such as the current pair with:
+
+        `metadata["pair"]`
+
+        All features must be prepended with `%` to be recognized by FreqAI internals.
+
+        :param df: strategy dataframe which will receive the features
+        :param metadata: metadata of current pair
+        usage example: dataframe["%-day_of_week"] = (dataframe["date"].dt.dayofweek + 1) / 7
+        """
+        dataframe["%-day_of_week"] = (dataframe["date"].dt.dayofweek + 1) / 7
+        dataframe["%-hour_of_day"] = (dataframe["date"].dt.hour + 1) / 25
+        return dataframe
+
+    def set_freqai_targets(self, dataframe: DataFrame, metadata, **kwargs) -> DataFrame:
+        """
+        *Only functional with FreqAI enabled strategies*
+        Required function to set the targets for the model.
+        All targets must be prepended with `&` to be recognized by the FreqAI internals.
+
+        Access metadata such as the current pair with:
+
+        `metadata["pair"]`
+
+        :param df: strategy dataframe which will receive the targets
+        :param metadata: metadata of current pair
+        usage example: dataframe["&-target"] = dataframe["close"].shift(-1) / dataframe["close"]
+        """
+        dataframe["&-s_close"] = (
+            dataframe["close"]
+            .shift(-self.freqai_info["feature_parameters"]["label_period_candles"])
+            .rolling(self.freqai_info["feature_parameters"]["label_period_candles"])
+            .mean()
+            / dataframe["close"]
+            - 1
+            )
+
+        return dataframe
+
+

In the presented example, the user does not wish to pass the bb_lowerband as a feature to the model, +and has therefore not prepended it with %. The user does, however, wish to pass bb_width to the +model for training/prediction and has therefore prepended it with %.

+

After having defined the base features, the next step is to expand upon them using the powerful feature_parameters in the configuration file:

+
    "freqai": {
+        //...
+        "feature_parameters" : {
+            "include_timeframes": ["5m","15m","4h"],
+            "include_corr_pairlist": [
+                "ETH/USD",
+                "LINK/USD",
+                "BNB/USD"
+            ],
+            "label_period_candles": 24,
+            "include_shifted_candles": 2,
+            "indicator_periods_candles": [10, 20]
+        },
+        //...
+    }
+
+

The include_timeframes in the config above are the timeframes (tf) of each call to feature_engineering_expand_*() in the strategy. In the presented case, the user is asking for the 5m, 15m, and 4h timeframes of the rsi, mfi, roc, and bb_width to be included in the feature set.

+

You can ask for each of the defined features to be included also for informative pairs using the include_corr_pairlist. This means that the feature set will include all the features from feature_engineering_expand_*() on all the include_timeframes for each of the correlated pairs defined in the config (ETH/USD, LINK/USD, and BNB/USD in the presented example).

+

include_shifted_candles indicates the number of previous candles to include in the feature set. For example, include_shifted_candles: 2 tells FreqAI to include the past 2 candles for each of the features in the feature set.

+

In total, the number of features the user of the presented example strat has created is: length of include_timeframes * no. features in feature_engineering_expand_*() * length of include_corr_pairlist * no. include_shifted_candles * length of indicator_periods_candles + \(= 3 * 3 * 3 * 2 * 2 = 108\).

+

Gain finer control over feature_engineering_* functions with metadata

+

All feature_engineering_* and set_freqai_targets() functions are passed a metadata dictionary which contains information about the pair, tf (timeframe), and period that FreqAI is automating for feature building. As such, a user can use metadata inside feature_engineering_* functions as criteria for blocking/reserving features for certain timeframes, periods, pairs etc.

+
def feature_engineering_expand_all(self, dataframe: DataFrame, period, metadata, **kwargs) -> DataFrame:
+    if metadata["tf"] == "1h":
+        dataframe["%-roc-period"] = ta.ROC(dataframe, timeperiod=period)
+
+

This will block ta.ROC() from being added to any timeframes other than "1h".

+

Returning additional info from training

+

Important metrics can be returned to the strategy at the end of each model training by assigning them to dk.data['extra_returns_per_train']['my_new_value'] = XYZ inside the custom prediction model class.

+

FreqAI takes the my_new_value assigned in this dictionary and expands it to fit the dataframe that is returned to the strategy. You can then use the returned metrics in your strategy through dataframe['my_new_value']. An example of how return values can be used in FreqAI are the &*_mean and &*_std values that are used to created a dynamic target threshold.

+

Another example, where the user wants to use live metrics from the trade database, is shown below:

+
    "freqai": {
+        "extra_returns_per_train": {"total_profit": 4}
+    }
+
+

You need to set the standard dictionary in the config so that FreqAI can return proper dataframe shapes. These values will likely be overridden by the prediction model, but in the case where the model has yet to set them, or needs a default initial value, the pre-set values are what will be returned.

+

Feature normalization

+

FreqAI is strict when it comes to data normalization. The train features, \(X^{train}\), are always normalized to [-1, 1] using a shifted min-max normalization:

+
\[X^{train}_{norm} = 2 * \frac{X^{train} - X^{train}.min()}{X^{train}.max() - X^{train}.min()} - 1\]
+

All other data (test data and unseen prediction data in dry/live/backtest) is always automatically normalized to the training feature space according to industry standards. FreqAI stores all the metadata required to ensure that test and prediction features will be properly normalized and that predictions are properly denormalized. For this reason, it is not recommended to eschew industry standards and modify FreqAI internals - however - advanced users can do so by inheriting train() in their custom IFreqaiModel and using their own normalization functions.

+

Data dimensionality reduction with Principal Component Analysis

+

You can reduce the dimensionality of your features by activating the principal_component_analysis in the config:

+
    "freqai": {
+        "feature_parameters" : {
+            "principal_component_analysis": true
+        }
+    }
+
+

This will perform PCA on the features and reduce their dimensionality so that the explained variance of the data set is >= 0.999. Reducing data dimensionality makes training the model faster and hence allows for more up-to-date models.

+

Inlier metric

+

The inlier_metric is a metric aimed at quantifying how similar the features of a data point are to the most recent historical data points.

+

You define the lookback window by setting inlier_metric_window and FreqAI computes the distance between the present time point and each of the previous inlier_metric_window lookback points. A Weibull function is fit to each of the lookback distributions and its cumulative distribution function (CDF) is used to produce a quantile for each lookback point. The inlier_metric is then computed for each time point as the average of the corresponding lookback quantiles. The figure below explains the concept for an inlier_metric_window of 5.

+

inlier-metric

+

FreqAI adds the inlier_metric to the training features and hence gives the model access to a novel type of temporal information.

+

This function does not remove outliers from the data set.

+

Weighting features for temporal importance

+

FreqAI allows you to set a weight_factor to weight recent data more strongly than past data via an exponential function:

+
\[ W_i = \exp(\frac{-i}{\alpha*n}) \]
+

where \(W_i\) is the weight of data point \(i\) in a total set of \(n\) data points. Below is a figure showing the effect of different weight factors on the data points in a feature set.

+

weight-factor

+

Outlier detection

+

Equity and crypto markets suffer from a high level of non-patterned noise in the form of outlier data points. FreqAI implements a variety of methods to identify such outliers and hence mitigate risk.

+

Identifying outliers with the Dissimilarity Index (DI)

+

The Dissimilarity Index (DI) aims to quantify the uncertainty associated with each prediction made by the model.

+

You can tell FreqAI to remove outlier data points from the training/test data sets using the DI by including the following statement in the config:

+
    "freqai": {
+        "feature_parameters" : {
+            "DI_threshold": 1
+        }
+    }
+
+

The DI allows predictions which are outliers (not existent in the model feature space) to be thrown out due to low levels of certainty. To do so, FreqAI measures the distance between each training data point (feature vector), \(X_{a}\), and all other training data points:

+
\[ d_{ab} = \sqrt{\sum_{j=1}^p(X_{a,j}-X_{b,j})^2} \]
+

where \(d_{ab}\) is the distance between the normalized points \(a\) and \(b\), and \(p\) is the number of features, i.e., the length of the vector \(X\). The characteristic distance, \(\overline{d}\), for a set of training data points is simply the mean of the average distances:

+
\[ \overline{d} = \sum_{a=1}^n(\sum_{b=1}^n(d_{ab}/n)/n) \]
+

\(\overline{d}\) quantifies the spread of the training data, which is compared to the distance between a new prediction feature vectors, \(X_k\) and all the training data:

+
\[ d_k = \arg \min d_{k,i} \]
+

This enables the estimation of the Dissimilarity Index as:

+
\[ DI_k = d_k/\overline{d} \]
+

You can tweak the DI through the DI_threshold to increase or decrease the extrapolation of the trained model. A higher DI_threshold means that the DI is more lenient and allows predictions further away from the training data to be used whilst a lower DI_threshold has the opposite effect and hence discards more predictions.

+

Below is a figure that describes the DI for a 3D data set.

+

DI

+

Identifying outliers using a Support Vector Machine (SVM)

+

You can tell FreqAI to remove outlier data points from the training/test data sets using a Support Vector Machine (SVM) by including the following statement in the config:

+
    "freqai": {
+        "feature_parameters" : {
+            "use_SVM_to_remove_outliers": true
+        }
+    }
+
+

The SVM will be trained on the training data and any data point that the SVM deems to be beyond the feature space will be removed.

+

FreqAI uses sklearn.linear_model.SGDOneClassSVM (details are available on scikit-learn's webpage here (external website)) and you can elect to provide additional parameters for the SVM, such as shuffle, and nu.

+

The parameter shuffle is by default set to False to ensure consistent results. If it is set to True, running the SVM multiple times on the same data set might result in different outcomes due to max_iter being to low for the algorithm to reach the demanded tol. Increasing max_iter solves this issue but causes the procedure to take longer time.

+

The parameter nu, very broadly, is the amount of data points that should be considered outliers and should be between 0 and 1.

+

Identifying outliers with DBSCAN

+

You can configure FreqAI to use DBSCAN to cluster and remove outliers from the training/test data set or incoming outliers from predictions, by activating use_DBSCAN_to_remove_outliers in the config:

+
    "freqai": {
+        "feature_parameters" : {
+            "use_DBSCAN_to_remove_outliers": true
+        }
+    }
+
+

DBSCAN is an unsupervised machine learning algorithm that clusters data without needing to know how many clusters there should be.

+

Given a number of data points \(N\), and a distance \(\varepsilon\), DBSCAN clusters the data set by setting all data points that have \(N-1\) other data points within a distance of \(\varepsilon\) as core points. A data point that is within a distance of \(\varepsilon\) from a core point but that does not have \(N-1\) other data points within a distance of \(\varepsilon\) from itself is considered an edge point. A cluster is then the collection of core points and edge points. Data points that have no other data points at a distance \(<\varepsilon\) are considered outliers. The figure below shows a cluster with \(N = 3\).

+

dbscan

+

FreqAI uses sklearn.cluster.DBSCAN (details are available on scikit-learn's webpage here (external website)) with min_samples (\(N\)) taken as ¼ of the no. of time points (candles) in the feature set. eps (\(\varepsilon\)) is computed automatically as the elbow point in the k-distance graph computed from the nearest neighbors in the pairwise distances of all data points in the feature set.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/freqai-parameter-table/index.html b/en/2023.4/freqai-parameter-table/index.html new file mode 100644 index 000000000..7ebde8a1e --- /dev/null +++ b/en/2023.4/freqai-parameter-table/index.html @@ -0,0 +1,1753 @@ + + + + + + + + + + + + + + + + + + + + + + + + Parameter table - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Parameter table

+

The table below will list all configuration parameters available for FreqAI. Some of the parameters are exemplified in config_examples/config_freqai.example.json.

+

Mandatory parameters are marked as Required and have to be set in one of the suggested ways.

+

General configuration parameters

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
ParameterDescription
General configuration parameters within the config.freqai tree
freqaiRequired.
The parent dictionary containing all the parameters for controlling FreqAI.
Datatype: Dictionary.
train_period_daysRequired.
Number of days to use for the training data (width of the sliding window).
Datatype: Positive integer.
backtest_period_daysRequired.
Number of days to inference from the trained model before sliding the train_period_days window defined above, and retraining the model during backtesting (more info here). This can be fractional days, but beware that the provided timerange will be divided by this number to yield the number of trainings necessary to complete the backtest.
Datatype: Float.
identifierRequired.
A unique ID for the current model. If models are saved to disk, the identifier allows for reloading specific pre-trained models/data.
Datatype: String.
live_retrain_hoursFrequency of retraining during dry/live runs.
Datatype: Float > 0.
Default: 0 (models retrain as often as possible).
expiration_hoursAvoid making predictions if a model is more than expiration_hours old.
Datatype: Positive integer.
Default: 0 (models never expire).
purge_old_modelsNumber of models to keep on disk (not relevant to backtesting). Default is 2, which means that dry/live runs will keep the latest 2 models on disk. Setting to 0 keeps all models. This parameter also accepts a boolean to maintain backwards compatibility.
Datatype: Integer.
Default: 2.
save_backtest_modelsSave models to disk when running backtesting. Backtesting operates most efficiently by saving the prediction data and reusing them directly for subsequent runs (when you wish to tune entry/exit parameters). Saving backtesting models to disk also allows to use the same model files for starting a dry/live instance with the same model identifier.
Datatype: Boolean.
Default: False (no models are saved).
fit_live_predictions_candlesNumber of historical candles to use for computing target (label) statistics from prediction data, instead of from the training dataset (more information can be found here).
Datatype: Positive integer.
continual_learningUse the final state of the most recently trained model as starting point for the new model, allowing for incremental learning (more information can be found here).
Datatype: Boolean.
Default: False.
write_metrics_to_diskCollect train timings, inference timings and cpu usage in json file.
Datatype: Boolean.
Default: False
data_kitchen_thread_count
Designate the number of threads you want to use for data processing (outlier methods, normalization, etc.). This has no impact on the number of threads used for training. If user does not set it (default), FreqAI will use max number of threads - 2 (leaving 1 physical core available for Freqtrade bot and FreqUI)
Datatype: Positive integer.
+

Feature parameters

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
ParameterDescription
Feature parameters within the freqai.feature_parameters sub dictionary
feature_parametersA dictionary containing the parameters used to engineer the feature set. Details and examples are shown here.
Datatype: Dictionary.
include_timeframesA list of timeframes that all indicators in feature_engineering_expand_*() will be created for. The list is added as features to the base indicators dataset.
Datatype: List of timeframes (strings).
include_corr_pairlistA list of correlated coins that FreqAI will add as additional features to all pair_whitelist coins. All indicators set in feature_engineering_expand_*() during feature engineering (see details here) will be created for each correlated coin. The correlated coins features are added to the base indicators dataset.
Datatype: List of assets (strings).
label_period_candlesNumber of candles into the future that the labels are created for. This is used in feature_engineering_expand_all() (see templates/FreqaiExampleStrategy.py for detailed usage). You can create custom labels and choose whether to make use of this parameter or not.
Datatype: Positive integer.
include_shifted_candlesAdd features from previous candles to subsequent candles with the intent of adding historical information. If used, FreqAI will duplicate and shift all features from the include_shifted_candles previous candles so that the information is available for the subsequent candle.
Datatype: Positive integer.
weight_factorWeight training data points according to their recency (see details here).
Datatype: Positive float (typically < 1).
indicator_max_period_candlesNo longer used (#7325). Replaced by startup_candle_count which is set in the strategy. startup_candle_count is timeframe independent and defines the maximum period used in feature_engineering_*() for indicator creation. FreqAI uses this parameter together with the maximum timeframe in include_time_frames to calculate how many data points to download such that the first data point does not include a NaN.
Datatype: Positive integer.
indicator_periods_candlesTime periods to calculate indicators for. The indicators are added to the base indicator dataset.
Datatype: List of positive integers.
principal_component_analysisAutomatically reduce the dimensionality of the data set using Principal Component Analysis. See details about how it works here
Datatype: Boolean.
Default: False.
plot_feature_importancesCreate a feature importance plot for each model for the top/bottom plot_feature_importances number of features. Plot is stored in user_data/models/<identifier>/sub-train-<COIN>_<timestamp>.html.
Datatype: Integer.
Default: 0.
DI_thresholdActivates the use of the Dissimilarity Index for outlier detection when set to > 0. See details about how it works here.
Datatype: Positive float (typically < 1).
use_SVM_to_remove_outliersTrain a support vector machine to detect and remove outliers from the training dataset, as well as from incoming data points. See details about how it works here.
Datatype: Boolean.
svm_paramsAll parameters available in Sklearn's SGDOneClassSVM(). See details about some select parameters here.
Datatype: Dictionary.
use_DBSCAN_to_remove_outliersCluster data using the DBSCAN algorithm to identify and remove outliers from training and prediction data. See details about how it works here.
Datatype: Boolean.
inlier_metric_windowIf set, FreqAI adds an inlier_metric to the training feature set and set the lookback to be the inlier_metric_window, i.e., the number of previous time points to compare the current candle to. Details of how the inlier_metric is computed can be found here.
Datatype: Integer.
Default: 0.
noise_standard_deviationIf set, FreqAI adds noise to the training features with the aim of preventing overfitting. FreqAI generates random deviates from a gaussian distribution with a standard deviation of noise_standard_deviation and adds them to all data points. noise_standard_deviation should be kept relative to the normalized space, i.e., between -1 and 1. In other words, since data in FreqAI is always normalized to be between -1 and 1, noise_standard_deviation: 0.05 would result in 32% of the data being randomly increased/decreased by more than 2.5% (i.e., the percent of data falling within the first standard deviation).
Datatype: Integer.
Default: 0.
outlier_protection_percentageEnable to prevent outlier detection methods from discarding too much data. If more than outlier_protection_percentage % of points are detected as outliers by the SVM or DBSCAN, FreqAI will log a warning message and ignore outlier detection, i.e., the original dataset will be kept intact. If the outlier protection is triggered, no predictions will be made based on the training dataset.
Datatype: Float.
Default: 30.
reverse_train_test_orderSplit the feature dataset (see below) and use the latest data split for training and test on historical split of the data. This allows the model to be trained up to the most recent data point, while avoiding overfitting. However, you should be careful to understand the unorthodox nature of this parameter before employing it.
Datatype: Boolean.
Default: False (no reversal).
shuffle_after_splitSplit the data into train and test sets, and then shuffle both sets individually.
Datatype: Boolean.
Default: False.
buffer_train_data_candlesCut buffer_train_data_candles off the beginning and end of the training data after the indicators were populated. The main example use is when predicting maxima and minima, the argrelextrema function cannot know the maxima/minima at the edges of the timerange. To improve model accuracy, it is best to compute argrelextrema on the full timerange and then use this function to cut off the edges (buffer) by the kernel. In another case, if the targets are set to a shifted price movement, this buffer is unnecessary because the shifted candles at the end of the timerange will be NaN and FreqAI will automatically cut those off of the training dataset.
Datatype: Integer.
Default: 0.
+

Data split parameters

+ + + + + + + + + + + + + + + + + + + + + + + + + +
ParameterDescription
Data split parameters within the freqai.data_split_parameters sub dictionary
data_split_parametersInclude any additional parameters available from scikit-learn test_train_split(), which are shown here (external website).
Datatype: Dictionary.
test_sizeThe fraction of data that should be used for testing instead of training.
Datatype: Positive float < 1.
shuffleShuffle the training data points during training. Typically, to not remove the chronological order of data in time-series forecasting, this is set to False.
Datatype: Boolean.
Defaut: False.
+

Model training parameters

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
ParameterDescription
Model training parameters within the freqai.model_training_parameters sub dictionary
model_training_parametersA flexible dictionary that includes all parameters available by the selected model library. For example, if you use LightGBMRegressor, this dictionary can contain any parameter available by the LightGBMRegressor here (external website). If you select a different model, this dictionary can contain any parameter from that model. A list of the currently available models can be found here.
Datatype: Dictionary.
n_estimatorsThe number of boosted trees to fit in the training of the model.
Datatype: Integer.
learning_rateBoosting learning rate during training of the model.
Datatype: Float.
n_jobs, thread_count, task_typeSet the number of threads for parallel processing and the task_type (gpu or cpu). Different model libraries use different parameter names.
Datatype: Float.
+

Reinforcement Learning parameters

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
ParameterDescription
Reinforcement Learning Parameters within the freqai.rl_config sub dictionary
rl_configA dictionary containing the control parameters for a Reinforcement Learning model.
Datatype: Dictionary.
train_cyclesTraining time steps will be set based on the `train_cycles * number of training data points.
Datatype: Integer.
cpu_countNumber of processors to dedicate to the Reinforcement Learning training process.
Datatype: int.
max_trade_duration_candlesGuides the agent training to keep trades below desired length. Example usage shown in prediction_models/ReinforcementLearner.py within the customizable calculate_reward() function.
Datatype: int.
model_typeModel string from stable_baselines3 or SBcontrib. Available strings include: 'TRPO', 'ARS', 'RecurrentPPO', 'MaskablePPO', 'PPO', 'A2C', 'DQN'. User should ensure that model_training_parameters match those available to the corresponding stable_baselines3 model by visiting their documentaiton. PPO doc (external website)
Datatype: string.
policy_typeOne of the available policy types from stable_baselines3
Datatype: string.
max_training_drawdown_pctThe maximum drawdown that the agent is allowed to experience during training.
Datatype: float.
Default: 0.8
cpu_countNumber of threads/cpus to dedicate to the Reinforcement Learning training process (depending on if ReinforcementLearning_multiproc is selected or not). Recommended to leave this untouched, by default, this value is set to the total number of physical cores minus 1.
Datatype: int.
model_reward_parametersParameters used inside the customizable calculate_reward() function in ReinforcementLearner.py
Datatype: int.
add_state_infoTell FreqAI to include state information in the feature set for training and inferencing. The current state variables include trade duration, current profit, trade position. This is only available in dry/live runs, and is automatically switched to false for backtesting.
Datatype: bool.
Default: False.
net_archNetwork architecture which is well described in stable_baselines3 doc. In summary: [<shared layers>, dict(vf=[<non-shared value network layers>], pi=[<non-shared policy network layers>])]. By default this is set to [128, 128], which defines 2 shared hidden layers with 128 units each.
randomize_starting_positionRandomize the starting point of each episode to avoid overfitting.
Datatype: bool.
Default: False.
drop_ohlc_from_featuresDo not include the normalized ohlc data in the feature set passed to the agent during training (ohlc will still be used for driving the environment in all cases)
Datatype: Boolean.
Default: False
progress_barDisplay a progress bar with the current progress, elapsed time and estimated remaining time.
Datatype: Boolean.
Default: False.
+

PyTorch parameters

+

general

+ + + + + + + + + + + + + + + + + + + + + + + + + +
ParameterDescription
Model training parameters within the freqai.model_training_parameters sub dictionary
learning_rateLearning rate to be passed to the optimizer.
Datatype: float.
Default: 3e-4.
model_kwargsParameters to be passed to the model class.
Datatype: dict.
Default: {}.
trainer_kwargsParameters to be passed to the trainer class.
Datatype: dict.
Default: {}.
+

trainer_kwargs

+ + + + + + + + + + + + + + + + + + + + + + + + + +
ParameterDescription
Model training parameters within the freqai.model_training_parameters.model_kwargs sub dictionary
max_itersThe number of training iterations to run. iteration here refers to the number of times we call self.optimizer.step(). used to calculate n_epochs.
Datatype: int.
Default: 100.
batch_sizeThe size of the batches to use during training..
Datatype: int.
Default: 64.
max_n_eval_batchesThe maximum number batches to use for evaluation..
Datatype: int, optional.
Default: None.
+

Additional parameters

+ + + + + + + + + + + + + + + + + + + + + + + + + +
ParameterDescription
Extraneous parameters
freqai.kerasIf the selected model makes use of Keras (typical for TensorFlow-based prediction models), this flag needs to be activated so that the model save/loading follows Keras standards.
Datatype: Boolean.
Default: False.
freqai.conv_widthThe width of a convolutional neural network input tensor. This replaces the need for shifting candles (include_shifted_candles) by feeding in historical data points as the second dimension of the tensor. Technically, this parameter can also be used for regressors, but it only adds computational overhead and does not change the model training/prediction.
Datatype: Integer.
Default: 2.
freqai.reduce_df_footprintRecast all numeric columns to float32/int32, with the objective of reducing ram/disk usage and decreasing train/inference timing. This parameter is set in the main level of the Freqtrade configuration file (not inside FreqAI).
Datatype: Boolean.
Default: False.
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/freqai-reinforcement-learning/index.html b/en/2023.4/freqai-reinforcement-learning/index.html new file mode 100644 index 000000000..24e935cf5 --- /dev/null +++ b/en/2023.4/freqai-reinforcement-learning/index.html @@ -0,0 +1,1632 @@ + + + + + + + + + + + + + + + + + + + + + + + + Reinforcement Learning - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Reinforcement Learning

+
+

Installation size

+

Reinforcement learning dependencies include large packages such as torch, which should be explicitly requested during ./setup.sh -i by answering "y" to the question "Do you also want dependencies for freqai-rl (~700mb additional space required) [y/N]?". +Users who prefer docker should ensure they use the docker image appended with _freqairl.

+
+

Background and terminology

+

What is RL and why does FreqAI need it?

+

Reinforcement learning involves two important components, the agent and the training environment. During agent training, the agent moves through historical data candle by candle, always making 1 of a set of actions: Long entry, long exit, short entry, short exit, neutral). During this training process, the environment tracks the performance of these actions and rewards the agent according to a custom user made calculate_reward() (here we offer a default reward for users to build on if they wish details here). The reward is used to train weights in a neural network.

+

A second important component of the FreqAI RL implementation is the use of state information. State information is fed into the network at each step, including current profit, current position, and current trade duration. These are used to train the agent in the training environment, and to reinforce the agent in dry/live (this functionality is not available in backtesting). FreqAI + Freqtrade is a perfect match for this reinforcing mechanism since this information is readily available in live deployments.

+

Reinforcement learning is a natural progression for FreqAI, since it adds a new layer of adaptivity and market reactivity that Classifiers and Regressors cannot match. However, Classifiers and Regressors have strengths that RL does not have such as robust predictions. Improperly trained RL agents may find "cheats" and "tricks" to maximize reward without actually winning any trades. For this reason, RL is more complex and demands a higher level of understanding than typical Classifiers and Regressors.

+

The RL interface

+

With the current framework, we aim to expose the training environment via the common "prediction model" file, which is a user inherited BaseReinforcementLearner object (e.g. freqai/prediction_models/ReinforcementLearner). Inside this user class, the RL environment is available and customized via MyRLEnv as shown below.

+

We envision the majority of users focusing their effort on creative design of the calculate_reward() function details here, while leaving the rest of the environment untouched. Other users may not touch the environment at all, and they will only play with the configuration settings and the powerful feature engineering that already exists in FreqAI. Meanwhile, we enable advanced users to create their own model classes entirely.

+

The framework is built on stable_baselines3 (torch) and OpenAI gym for the base environment class. But generally speaking, the model class is well isolated. Thus, the addition of competing libraries can be easily integrated into the existing framework. For the environment, it is inheriting from gym.env which means that it is necessary to write an entirely new environment in order to switch to a different library.

+

Important considerations

+

As explained above, the agent is "trained" in an artificial trading "environment". In our case, that environment may seem quite similar to a real Freqtrade backtesting environment, but it is NOT. In fact, the RL training environment is much more simplified. It does not incorporate any of the complicated strategy logic, such as callbacks like custom_exit, custom_stoploss, leverage controls, etc. The RL environment is instead a very "raw" representation of the true market, where the agent has free will to learn the policy (read: stoploss, take profit, etc.) which is enforced by the calculate_reward(). Thus, it is important to consider that the agent training environment is not identical to the real world.

+

Running Reinforcement Learning

+

Setting up and running a Reinforcement Learning model is the same as running a Regressor or Classifier. The same two flags, --freqaimodel and --strategy, must be defined on the command line:

+
freqtrade trade --freqaimodel ReinforcementLearner --strategy MyRLStrategy --config config.json
+
+

where ReinforcementLearner will use the templated ReinforcementLearner from freqai/prediction_models/ReinforcementLearner (or a custom user defined one located in user_data/freqaimodels). The strategy, on the other hand, follows the same base feature engineering with feature_engineering_* as a typical Regressor. The difference lies in the creation of the targets, Reinforcement Learning doesn't require them. However, FreqAI requires a default (neutral) value to be set in the action column:

+
    def set_freqai_targets(self, dataframe, **kwargs) -> DataFrame:
+        """
+        *Only functional with FreqAI enabled strategies*
+        Required function to set the targets for the model.
+        All targets must be prepended with `&` to be recognized by the FreqAI internals.
+
+        More details about feature engineering available:
+
+        https://www.freqtrade.io/en/latest/freqai-feature-engineering
+
+        :param df: strategy dataframe which will receive the targets
+        usage example: dataframe["&-target"] = dataframe["close"].shift(-1) / dataframe["close"]
+        """
+        # For RL, there are no direct targets to set. This is filler (neutral)
+        # until the agent sends an action.
+        dataframe["&-action"] = 0
+        return dataframe
+
+

Most of the function remains the same as for typical Regressors, however, the function below shows how the strategy must pass the raw price data to the agent so that it has access to raw OHLCV in the training environment:

+
    def feature_engineering_standard(self, dataframe: DataFrame, **kwargs) -> DataFrame:
+        # The following features are necessary for RL models
+        dataframe[f"%-raw_close"] = dataframe["close"]
+        dataframe[f"%-raw_open"] = dataframe["open"]
+        dataframe[f"%-raw_high"] = dataframe["high"]
+        dataframe[f"%-raw_low"] = dataframe["low"]
+    return dataframe
+
+

Finally, there is no explicit "label" to make - instead it is necessary to assign the &-action column which will contain the agent's actions when accessed in populate_entry/exit_trends(). In the present example, the neutral action to 0. This value should align with the environment used. FreqAI provides two environments, both use 0 as the neutral action.

+

After users realize there are no labels to set, they will soon understand that the agent is making its "own" entry and exit decisions. This makes strategy construction rather simple. The entry and exit signals come from the agent in the form of an integer - which are used directly to decide entries and exits in the strategy:

+
    def populate_entry_trend(self, df: DataFrame, metadata: dict) -> DataFrame:
+
+        enter_long_conditions = [df["do_predict"] == 1, df["&-action"] == 1]
+
+        if enter_long_conditions:
+            df.loc[
+                reduce(lambda x, y: x & y, enter_long_conditions), ["enter_long", "enter_tag"]
+            ] = (1, "long")
+
+        enter_short_conditions = [df["do_predict"] == 1, df["&-action"] == 3]
+
+        if enter_short_conditions:
+            df.loc[
+                reduce(lambda x, y: x & y, enter_short_conditions), ["enter_short", "enter_tag"]
+            ] = (1, "short")
+
+        return df
+
+    def populate_exit_trend(self, df: DataFrame, metadata: dict) -> DataFrame:
+        exit_long_conditions = [df["do_predict"] == 1, df["&-action"] == 2]
+        if exit_long_conditions:
+            df.loc[reduce(lambda x, y: x & y, exit_long_conditions), "exit_long"] = 1
+
+        exit_short_conditions = [df["do_predict"] == 1, df["&-action"] == 4]
+        if exit_short_conditions:
+            df.loc[reduce(lambda x, y: x & y, exit_short_conditions), "exit_short"] = 1
+
+        return df
+
+

It is important to consider that &-action depends on which environment they choose to use. The example above shows 5 actions, where 0 is neutral, 1 is enter long, 2 is exit long, 3 is enter short and 4 is exit short.

+

Configuring the Reinforcement Learner

+

In order to configure the Reinforcement Learner the following dictionary must exist in the freqai config:

+
        "rl_config": {
+            "train_cycles": 25,
+            "add_state_info": true,
+            "max_trade_duration_candles": 300,
+            "max_training_drawdown_pct": 0.02,
+            "cpu_count": 8,
+            "model_type": "PPO",
+            "policy_type": "MlpPolicy",
+            "model_reward_parameters": {
+                "rr": 1,
+                "profit_aim": 0.025
+            }
+        }
+
+

Parameter details can be found here, but in general the train_cycles decides how many times the agent should cycle through the candle data in its artificial environment to train weights in the model. model_type is a string which selects one of the available models in stable_baselines(external link).

+
+

Note

+

If you would like to experiment with continual_learning, then you should set that value to true in the main freqai configuration dictionary. This will tell the Reinforcement Learning library to continue training new models from the final state of previous models, instead of retraining new models from scratch each time a retrain is initiated.

+
+
+

Note

+

Remember that the general model_training_parameters dictionary should contain all the model hyperparameter customizations for the particular model_type. For example, PPO parameters can be found here.

+
+

Creating a custom reward function

+

As you begin to modify the strategy and the prediction model, you will quickly realize some important differences between the Reinforcement Learner and the Regressors/Classifiers. Firstly, the strategy does not set a target value (no labels!). Instead, you set the calculate_reward() function inside the MyRLEnv class (see below). A default calculate_reward() is provided inside prediction_models/ReinforcementLearner.py to demonstrate the necessary building blocks for creating rewards, but users are encouraged to create their own custom reinforcement learning model class (see below) and save it to user_data/freqaimodels. It is inside the calculate_reward() where creative theories about the market can be expressed. For example, you can reward your agent when it makes a winning trade, and penalize the agent when it makes a losing trade. Or perhaps, you wish to reward the agent for entering trades, and penalize the agent for sitting in trades too long. Below we show examples of how these rewards are all calculated:

+
    from freqtrade.freqai.prediction_models.ReinforcementLearner import ReinforcementLearner
+    from freqtrade.freqai.RL.Base5ActionRLEnv import Actions, Base5ActionRLEnv, Positions
+
+
+    class MyCoolRLModel(ReinforcementLearner):
+        """
+        User created RL prediction model.
+
+        Save this file to `freqtrade/user_data/freqaimodels`
+
+        then use it with:
+
+        freqtrade trade --freqaimodel MyCoolRLModel --config config.json --strategy SomeCoolStrat
+
+        Here the users can override any of the functions
+        available in the `IFreqaiModel` inheritance tree. Most importantly for RL, this
+        is where the user overrides `MyRLEnv` (see below), to define custom
+        `calculate_reward()` function, or to override any other parts of the environment.
+
+        This class also allows users to override any other part of the IFreqaiModel tree.
+        For example, the user can override `def fit()` or `def train()` or `def predict()`
+        to take fine-tuned control over these processes.
+
+        Another common override may be `def data_cleaning_predict()` where the user can
+        take fine-tuned control over the data handling pipeline.
+        """
+        class MyRLEnv(Base5ActionRLEnv):
+            """
+            User made custom environment. This class inherits from BaseEnvironment and gym.env.
+            Users can override any functions from those parent classes. Here is an example
+            of a user customized `calculate_reward()` function.
+            """
+            def calculate_reward(self, action: int) -> float:
+                # first, penalize if the action is not valid
+                if not self._is_valid(action):
+                    return -2
+                pnl = self.get_unrealized_profit()
+
+                factor = 100
+
+                pair = self.pair.replace(':', '')
+
+                # you can use feature values from dataframe
+                # Assumes the shifted RSI indicator has been generated in the strategy.
+                rsi_now = self.raw_features[f"%-rsi-period_10_shift-1_{pair}_"
+                                f"{self.config['timeframe']}"].iloc[self._current_tick]
+
+                # reward agent for entering trades
+                if (action in (Actions.Long_enter.value, Actions.Short_enter.value)
+                        and self._position == Positions.Neutral):
+                    if rsi_now < 40:
+                        factor = 40 / rsi_now
+                    else:
+                        factor = 1
+                    return 25 * factor
+
+                # discourage agent from not entering trades
+                if action == Actions.Neutral.value and self._position == Positions.Neutral:
+                    return -1
+                max_trade_duration = self.rl_config.get('max_trade_duration_candles', 300)
+                trade_duration = self._current_tick - self._last_trade_tick
+                if trade_duration <= max_trade_duration:
+                    factor *= 1.5
+                elif trade_duration > max_trade_duration:
+                    factor *= 0.5
+                # discourage sitting in position
+                if self._position in (Positions.Short, Positions.Long) and \
+                action == Actions.Neutral.value:
+                    return -1 * trade_duration / max_trade_duration
+                # close long
+                if action == Actions.Long_exit.value and self._position == Positions.Long:
+                    if pnl > self.profit_aim * self.rr:
+                        factor *= self.rl_config['model_reward_parameters'].get('win_reward_factor', 2)
+                    return float(pnl * factor)
+                # close short
+                if action == Actions.Short_exit.value and self._position == Positions.Short:
+                    if pnl > self.profit_aim * self.rr:
+                        factor *= self.rl_config['model_reward_parameters'].get('win_reward_factor', 2)
+                    return float(pnl * factor)
+                return 0.
+
+

Using Tensorboard

+

Reinforcement Learning models benefit from tracking training metrics. FreqAI has integrated Tensorboard to allow users to track training and evaluation performance across all coins and across all retrainings. Tensorboard is activated via the following command:

+
cd freqtrade
+tensorboard --logdir user_data/models/unique-id
+
+

where unique-id is the identifier set in the freqai configuration file. This command must be run in a separate shell to view the output in their browser at 127.0.0.1:6006 (6006 is the default port used by Tensorboard).

+

tensorboard

+

Custom logging

+

FreqAI also provides a built in episodic summary logger called self.tensorboard_log for adding custom information to the Tensorboard log. By default, this function is already called once per step inside the environment to record the agent actions. All values accumulated for all steps in a single episode are reported at the conclusion of each episode, followed by a full reset of all metrics to 0 in preparation for the subsequent episode.

+

self.tensorboard_log can also be used anywhere inside the environment, for example, it can be added to the calculate_reward function to collect more detailed information about how often various parts of the reward were called:

+
        class MyRLEnv(Base5ActionRLEnv):
+            """
+            User made custom environment. This class inherits from BaseEnvironment and gym.env.
+            Users can override any functions from those parent classes. Here is an example
+            of a user customized `calculate_reward()` function.
+            """
+            def calculate_reward(self, action: int) -> float:
+                if not self._is_valid(action):
+                    self.tensorboard_log("invalid")
+                    return -2
+
+
+

Note

+

The self.tensorboard_log() function is designed for tracking incremented objects only i.e. events, actions inside the training environment. If the event of interest is a float, the float can be passed as the second argument e.g. self.tensorboard_log("float_metric1", 0.23). In this case the metric values are not incremented.

+
+

Choosing a base environment

+

FreqAI provides three base environments, Base3ActionRLEnvironment, Base4ActionEnvironment and Base5ActionEnvironment. As the names imply, the environments are customized for agents that can select from 3, 4 or 5 actions. The Base3ActionEnvironment is the simplest, the agent can select from hold, long, or short. This environment can also be used for long-only bots (it automatically follows the can_short flag from the strategy), where long is the enter condition and short is the exit condition. Meanwhile, in the Base4ActionEnvironment, the agent can enter long, enter short, hold neutral, or exit position. Finally, in the Base5ActionEnvironment, the agent has the same actions as Base4, but instead of a single exit action, it separates exit long and exit short. The main changes stemming from the environment selection include:

+
    +
  • the actions available in the calculate_reward
  • +
  • the actions consumed by the user strategy
  • +
+

All of the FreqAI provided environments inherit from an action/position agnostic environment object called the BaseEnvironment, which contains all shared logic. The architecture is designed to be easily customized. The simplest customization is the calculate_reward() (see details here). However, the customizations can be further extended into any of the functions inside the environment. You can do this by simply overriding those functions inside your MyRLEnv in the prediction model file. Or for more advanced customizations, it is encouraged to create an entirely new environment inherited from BaseEnvironment.

+
+

Note

+

Only the Base3ActionRLEnv can do long-only training/trading (set the user strategy attribute can_short = False).

+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/freqai-running/index.html b/en/2023.4/freqai-running/index.html new file mode 100644 index 000000000..54fb738e5 --- /dev/null +++ b/en/2023.4/freqai-running/index.html @@ -0,0 +1,1569 @@ + + + + + + + + + + + + + + + + + + + + + + + + Running FreqAI - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Running FreqAI

+

There are two ways to train and deploy an adaptive machine learning model - live deployment and historical backtesting. In both cases, FreqAI runs/simulates periodic retraining of models as shown in the following figure:

+

freqai-window

+

Live deployments

+

FreqAI can be run dry/live using the following command:

+
freqtrade trade --strategy FreqaiExampleStrategy --config config_freqai.example.json --freqaimodel LightGBMRegressor
+
+

When launched, FreqAI will start training a new model, with a new identifier, based on the config settings. Following training, the model will be used to make predictions on incoming candles until a new model is available. New models are typically generated as often as possible, with FreqAI managing an internal queue of the coin pairs to try to keep all models equally up to date. FreqAI will always use the most recently trained model to make predictions on incoming live data. If you do not want FreqAI to retrain new models as often as possible, you can set live_retrain_hours to tell FreqAI to wait at least that number of hours before training a new model. Additionally, you can set expired_hours to tell FreqAI to avoid making predictions on models that are older than that number of hours.

+

Trained models are by default saved to disk to allow for reuse during backtesting or after a crash. You can opt to purge old models to save disk space by setting "purge_old_models": true in the config.

+

To start a dry/live run from a saved backtest model (or from a previously crashed dry/live session), you only need to specify the identifier of the specific model:

+
    "freqai": {
+        "identifier": "example",
+        "live_retrain_hours": 0.5
+    }
+
+

In this case, although FreqAI will initiate with a pre-trained model, it will still check to see how much time has elapsed since the model was trained. If a full live_retrain_hours has elapsed since the end of the loaded model, FreqAI will start training a new model.

+

Automatic data download

+

FreqAI automatically downloads the proper amount of data needed to ensure training of a model through the defined train_period_days and startup_candle_count (see the parameter table for detailed descriptions of these parameters).

+

Saving prediction data

+

All predictions made during the lifetime of a specific identifier model are stored in historic_predictions.pkl to allow for reloading after a crash or changes made to the config.

+

Purging old model data

+

FreqAI stores new model files after each successful training. These files become obsolete as new models are generated to adapt to new market conditions. If you are planning to leave FreqAI running for extended periods of time with high frequency retraining, you should enable purge_old_models in the config:

+
    "freqai": {
+        "purge_old_models": true,
+    }
+
+

This will automatically purge all models older than the two most recently trained ones to save disk space.

+

Backtesting

+

The FreqAI backtesting module can be executed with the following command:

+
freqtrade backtesting --strategy FreqaiExampleStrategy --strategy-path freqtrade/templates --config config_examples/config_freqai.example.json --freqaimodel LightGBMRegressor --timerange 20210501-20210701
+
+

If this command has never been executed with the existing config file, FreqAI will train a new model +for each pair, for each backtesting window within the expanded --timerange.

+

Backtesting mode requires downloading the necessary data before deployment (unlike in dry/live mode where FreqAI handles the data downloading automatically). You should be careful to consider that the time range of the downloaded data is more than the backtesting time range. This is because FreqAI needs data prior to the desired backtesting time range in order to train a model to be ready to make predictions on the first candle of the set backtesting time range. More details on how to calculate the data to download can be found here.

+
+

Model reuse

+

Once the training is completed, you can execute the backtesting again with the same config file and +FreqAI will find the trained models and load them instead of spending time training. This is useful +if you want to tweak (or even hyperopt) buy and sell criteria inside the strategy. If you +want to retrain a new model with the same config file, you should simply change the identifier. +This way, you can return to using any model you wish by simply specifying the identifier.

+
+
+

Note

+

Backtesting calls set_freqai_targets() one time for each backtest window (where the number of windows is the full backtest timerange divided by the backtest_period_days parameter). Doing this means that the targets simulate dry/live behavior without look ahead bias. However, the definition of the features in feature_engineering_*() is performed once on the entire backtest timerange. This means that you should be sure that features do look-ahead into the future. +More details about look-ahead bias can be found in Common Mistakes.

+
+
+

Saving prediction data

+

To allow for tweaking your strategy (not the features!), FreqAI will automatically save the predictions during backtesting so that they can be reused for future backtests and live runs using the same identifier model. This provides a performance enhancement geared towards enabling high-level hyperopting of entry/exit criteria.

+

An additional directory called backtesting_predictions, which contains all the predictions stored in hdf format, will be created in the unique-id folder.

+

To change your features, you must set a new identifier in the config to signal to FreqAI to train new models.

+

To save the models generated during a particular backtest so that you can start a live deployment from one of them instead of training a new model, you must set save_backtest_models to True in the config.

+

Backtest live collected predictions

+

FreqAI allow you to reuse live historic predictions through the backtest parameter --freqai-backtest-live-models. This can be useful when you want to reuse predictions generated in dry/run for comparison or other study.

+

The --timerange parameter must not be informed, as it will be automatically calculated through the data in the historic predictions file.

+

Downloading data to cover the full backtest period

+

For live/dry deployments, FreqAI will download the necessary data automatically. However, to use backtesting functionality, you need to download the necessary data using download-data (details here). You need to pay careful attention to understanding how much additional data needs to be downloaded to ensure that there is a sufficient amount of training data before the start of the backtesting time range. The amount of additional data can be roughly estimated by moving the start date of the time range backwards by train_period_days and the startup_candle_count (see the parameter table for detailed descriptions of these parameters) from the beginning of the desired backtesting time range.

+

As an example, to backtest the --timerange 20210501-20210701 using the example config which sets train_period_days to 30, together with startup_candle_count: 40 on a maximum include_timeframes of 1h, the start date for the downloaded data needs to be 20210501 - 30 days - 40 * 1h / 24 hours = 20210330 (31.7 days earlier than the start of the desired training time range).

+

Deciding the size of the sliding training window and backtesting duration

+

The backtesting time range is defined with the typical --timerange parameter in the configuration file. The duration of the sliding training window is set by train_period_days, whilst backtest_period_days is the sliding backtesting window, both in number of days (backtest_period_days can be +a float to indicate sub-daily retraining in live/dry mode). In the presented example config (found in config_examples/config_freqai.example.json), the user is asking FreqAI to use a training period of 30 days and backtest on the subsequent 7 days. After the training of the model, FreqAI will backtest the subsequent 7 days. The "sliding window" then moves one week forward (emulating FreqAI retraining once per week in live mode) and the new model uses the previous 30 days (including the 7 days used for backtesting by the previous model) to train. This is repeated until the end of --timerange. This means that if you set --timerange 20210501-20210701, FreqAI will have trained 8 separate models at the end of --timerange (because the full range comprises 8 weeks).

+
+

Note

+

Although fractional backtest_period_days is allowed, you should be aware that the --timerange is divided by this value to determine the number of models that FreqAI will need to train in order to backtest the full range. For example, by setting a --timerange of 10 days, and a backtest_period_days of 0.1, FreqAI will need to train 100 models per pair to complete the full backtest. Because of this, a true backtest of FreqAI adaptive training would take a very long time. The best way to fully test a model is to run it dry and let it train constantly. In this case, backtesting would take the exact same amount of time as a dry run.

+
+

Defining model expirations

+

During dry/live mode, FreqAI trains each coin pair sequentially (on separate threads/GPU from the main Freqtrade bot). This means that there is always an age discrepancy between models. If you are training on 50 pairs, and each pair requires 5 minutes to train, the oldest model will be over 4 hours old. This may be undesirable if the characteristic time scale (the trade duration target) for a strategy is less than 4 hours. You can decide to only make trade entries if the model is less than a certain number of hours old by setting the expiration_hours in the config file:

+
    "freqai": {
+        "expiration_hours": 0.5,
+    }
+
+

In the presented example config, the user will only allow predictions on models that are less than ½ hours old.

+

Controlling the model learning process

+

Model training parameters are unique to the selected machine learning library. FreqAI allows you to set any parameter for any library using the model_training_parameters dictionary in the config. The example config (found in config_examples/config_freqai.example.json) shows some of the example parameters associated with Catboost and LightGBM, but you can add any parameters available in those libraries or any other machine learning library you choose to implement.

+

Data split parameters are defined in data_split_parameters which can be any parameters associated with scikit-learn's train_test_split() function. train_test_split() has a parameters called shuffle which allows to shuffle the data or keep it unshuffled. This is particularly useful to avoid biasing training with temporally auto-correlated data. More details about these parameters can be found the scikit-learn website (external website).

+

The FreqAI specific parameter label_period_candles defines the offset (number of candles into the future) used for the labels. In the presented example config, the user is asking for labels that are 24 candles in the future.

+

Continual learning

+

You can choose to adopt a continual learning scheme by setting "continual_learning": true in the config. By enabling continual_learning, after training an initial model from scratch, subsequent trainings will start from the final model state of the preceding training. This gives the new model a "memory" of the previous state. By default, this is set to False which means that all new models are trained from scratch, without input from previous models.

+
+Continual learning enforces a constant parameter space +

Since continual_learning means that the model parameter space cannot change between trainings, principal_component_analysis is automatically disabled when continual_learning is enabled. Hint: PCA changes the parameter space and the number of features, learn more about PCA here.

+
+

Hyperopt

+

You can hyperopt using the same command as for typical Freqtrade hyperopt:

+
freqtrade hyperopt --hyperopt-loss SharpeHyperOptLoss --strategy FreqaiExampleStrategy --freqaimodel LightGBMRegressor --strategy-path freqtrade/templates --config config_examples/config_freqai.example.json --timerange 20220428-20220507
+
+

hyperopt requires you to have the data pre-downloaded in the same fashion as if you were doing backtesting. In addition, you must consider some restrictions when trying to hyperopt FreqAI strategies:

+
    +
  • The --analyze-per-epoch hyperopt parameter is not compatible with FreqAI.
  • +
  • It's not possible to hyperopt indicators in the feature_engineering_*() and set_freqai_targets() functions. This means that you cannot optimize model parameters using hyperopt. Apart from this exception, it is possible to optimize all other spaces.
  • +
  • The backtesting instructions also apply to hyperopt.
  • +
+

The best method for combining hyperopt and FreqAI is to focus on hyperopting entry/exit thresholds/criteria. You need to focus on hyperopting parameters that are not used in your features. For example, you should not try to hyperopt rolling window lengths in the feature creation, or any part of the FreqAI config which changes predictions. In order to efficiently hyperopt the FreqAI strategy, FreqAI stores predictions as dataframes and reuses them. Hence the requirement to hyperopt entry/exit thresholds/criteria only.

+

A good example of a hyperoptable parameter in FreqAI is a threshold for the Dissimilarity Index (DI) DI_values beyond which we consider data points as outliers:

+
di_max = IntParameter(low=1, high=20, default=10, space='buy', optimize=True, load=True)
+dataframe['outlier'] = np.where(dataframe['DI_values'] > self.di_max.value/10, 1, 0)
+
+

This specific hyperopt would help you understand the appropriate DI_values for your particular parameter space.

+

Using Tensorboard

+

CatBoost models benefit from tracking training metrics via Tensorboard. You can take advantage of the FreqAI integration to track training and evaluation performance across all coins and across all retrainings. Tensorboard is activated via the following command:

+
cd freqtrade
+tensorboard --logdir user_data/models/unique-id
+
+

where unique-id is the identifier set in the freqai configuration file. This command must be run in a separate shell if you wish to view the output in your browser at 127.0.0.1:6060 (6060 is the default port used by Tensorboard).

+

tensorboard

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/freqai/index.html b/en/2023.4/freqai/index.html new file mode 100644 index 000000000..366d474ca --- /dev/null +++ b/en/2023.4/freqai/index.html @@ -0,0 +1,1496 @@ + + + + + + + + + + + + + + + + + + + + + + + + Introduction - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+ +
+ + + +
+
+ + + + + + + +

freqai-logo

+

FreqAI

+

Introduction

+

FreqAI is a software designed to automate a variety of tasks associated with training a predictive machine learning model to generate market forecasts given a set of input signals. In general, FreqAI aims to be a sandbox for easily deploying robust machine learning libraries on real-time data (details).

+
+

Note

+

FreqAI is, and always will be, a not-for-profit, open-source project. FreqAI does not have a crypto token, FreqAI does not sell signals, and FreqAI does not have a domain besides the present freqtrade documentation.

+
+

Features include:

+
    +
  • Self-adaptive retraining - Retrain models during live deployments to self-adapt to the market in a supervised manner
  • +
  • Rapid feature engineering - Create large rich feature sets (10k+ features) based on simple user-created strategies
  • +
  • High performance - Threading allows for adaptive model retraining on a separate thread (or on GPU if available) from model inferencing (prediction) and bot trade operations. Newest models and data are kept in RAM for rapid inferencing
  • +
  • Realistic backtesting - Emulate self-adaptive training on historic data with a backtesting module that automates retraining
  • +
  • Extensibility - The generalized and robust architecture allows for incorporating any machine learning library/method available in Python. Eight examples are currently available, including classifiers, regressors, and a convolutional neural network
  • +
  • Smart outlier removal - Remove outliers from training and prediction data sets using a variety of outlier detection techniques
  • +
  • Crash resilience - Store trained models to disk to make reloading from a crash fast and easy, and purge obsolete files for sustained dry/live runs
  • +
  • Automatic data normalization - Normalize the data in a smart and statistically safe way
  • +
  • Automatic data download - Compute timeranges for data downloads and update historic data (in live deployments)
  • +
  • Cleaning of incoming data - Handle NaNs safely before training and model inferencing
  • +
  • Dimensionality reduction - Reduce the size of the training data via Principal Component Analysis
  • +
  • Deploying bot fleets - Set one bot to train models while a fleet of consumers use signals.
  • +
+

Quick start

+

The easiest way to quickly test FreqAI is to run it in dry mode with the following command:

+
freqtrade trade --config config_examples/config_freqai.example.json --strategy FreqaiExampleStrategy --freqaimodel LightGBMRegressor --strategy-path freqtrade/templates
+
+

You will see the boot-up process of automatic data downloading, followed by simultaneous training and trading.

+

An example strategy, prediction model, and config to use as a starting points can be found in +freqtrade/templates/FreqaiExampleStrategy.py, freqtrade/freqai/prediction_models/LightGBMRegressor.py, and +config_examples/config_freqai.example.json, respectively.

+

General approach

+

You provide FreqAI with a set of custom base indicators (the same way as in a typical Freqtrade strategy) as well as target values (labels). For each pair in the whitelist, FreqAI trains a model to predict the target values based on the input of custom indicators. The models are then consistently retrained, with a predetermined frequency, to adapt to market conditions. FreqAI offers the ability to both backtest strategies (emulating reality with periodic retraining on historic data) and deploy dry/live runs. In dry/live conditions, FreqAI can be set to constant retraining in a background thread to keep models as up to date as possible.

+

An overview of the algorithm, explaining the data processing pipeline and model usage, is shown below.

+

freqai-algo

+

Important machine learning vocabulary

+

Features - the parameters, based on historic data, on which a model is trained. All features for a single candle are stored as a vector. In FreqAI, you build a feature data set from anything you can construct in the strategy.

+

Labels - the target values that the model is trained toward. Each feature vector is associated with a single label that is defined by you within the strategy. These labels intentionally look into the future and are what you are training the model to be able to predict.

+

Training - the process of "teaching" the model to match the feature sets to the associated labels. Different types of models "learn" in different ways which means that one might be better than another for a specific application. More information about the different models that are already implemented in FreqAI can be found here.

+

Train data - a subset of the feature data set that is fed to the model during training to "teach" the model how to predict the targets. This data directly influences weight connections in the model.

+

Test data - a subset of the feature data set that is used to evaluate the performance of the model after training. This data does not influence nodal weights within the model.

+

Inferencing - the process of feeding a trained model new unseen data on which it will make a prediction.

+

Install prerequisites

+

The normal Freqtrade install process will ask if you wish to install FreqAI dependencies. You should reply "yes" to this question if you wish to use FreqAI. If you did not reply yes, you can manually install these dependencies after the install with:

+
pip install -r requirements-freqai.txt
+
+
+

Note

+

Catboost will not be installed on arm devices (raspberry, Mac M1, ARM based VPS, ...), since it does not provide wheels for this platform.

+
+
+

python 3.11

+

Some dependencies (Catboost, Torch) currently don't support python 3.11. Freqtrade therefore only supports python 3.10 for these models/dependencies. +Tests involving these dependencies are skipped on 3.11.

+
+

Usage with docker

+

If you are using docker, a dedicated tag with FreqAI dependencies is available as :freqai. As such - you can replace the image line in your docker compose file with image: freqtradeorg/freqtrade:develop_freqai. This image contains the regular FreqAI dependencies. Similar to native installs, Catboost will not be available on ARM based devices.

+

FreqAI position in open-source machine learning landscape

+

Forecasting chaotic time-series based systems, such as equity/cryptocurrency markets, requires a broad set of tools geared toward testing a wide range of hypotheses. Fortunately, a recent maturation of robust machine learning libraries (e.g. scikit-learn) has opened up a wide range of research possibilities. Scientists from a diverse range of fields can now easily prototype their studies on an abundance of established machine learning algorithms. Similarly, these user-friendly libraries enable "citzen scientists" to use their basic Python skills for data exploration. However, leveraging these machine learning libraries on historical and live chaotic data sources can be logistically difficult and expensive. Additionally, robust data collection, storage, and handling presents a disparate challenge. FreqAI aims to provide a generalized and extensible open-sourced framework geared toward live deployments of adaptive modeling for market forecasting. The FreqAI framework is effectively a sandbox for the rich world of open-source machine learning libraries. Inside the FreqAI sandbox, users find they can combine a wide variety of third-party libraries to test creative hypotheses on a free live 24/7 chaotic data source - cryptocurrency exchange data.

+

Citing FreqAI

+

FreqAI is published in the Journal of Open Source Software. If you find FreqAI useful in your research, please use the following citation:

+
@article{Caulk2022, 
+    doi = {10.21105/joss.04864},
+    url = {https://doi.org/10.21105/joss.04864},
+    year = {2022}, publisher = {The Open Journal},
+    volume = {7}, number = {80}, pages = {4864},
+    author = {Robert A. Caulk and Elin Törnquist and Matthias Voppichler and Andrew R. Lawless and Ryan McMullan and Wagner Costa Santos and Timothy C. Pogue and Johan van der Vlugt and Stefan P. Gehring and Pascal Schmidt},
+    title = {FreqAI: generalizing adaptive modeling for chaotic time-series market forecasts},
+    journal = {Journal of Open Source Software} } 
+
+

Common pitfalls

+

FreqAI cannot be combined with dynamic VolumePairlists (or any pairlist filter that adds and removes pairs dynamically). +This is for performance reasons - FreqAI relies on making quick predictions/retrains. To do this effectively, +it needs to download all the training data at the beginning of a dry/live instance. FreqAI stores and appends +new candles automatically for future retrains. This means that if new pairs arrive later in the dry run due to a volume pairlist, it will not have the data ready. However, FreqAI does work with the ShufflePairlist or a VolumePairlist which keeps the total pairlist constant (but reorders the pairs according to volume).

+

Credits

+

FreqAI is developed by a group of individuals who all contribute specific skillsets to the project.

+

Conception and software development: +Robert Caulk @robcaulk

+

Theoretical brainstorming and data analysis: +Elin Törnquist @th0rntwig

+

Code review and software architecture brainstorming: +@xmatthias

+

Software development: +Wagner Costa @wagnercosta +Emre Suzen @aemr3 +Timothy Pogue @wizrds

+

Beta testing and bug reporting: +Stefan Gehring @bloodhunter4rc, @longyu, Andrew Lawless @paranoidandy, Pascal Schmidt @smidelis, Ryan McMullan @smarmau, Juha Nykänen @suikula, Johan van der Vlugt @jooopiert, Richárd Józsa @richardjosza

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/hyperopt/index.html b/en/2023.4/hyperopt/index.html new file mode 100644 index 000000000..06334c797 --- /dev/null +++ b/en/2023.4/hyperopt/index.html @@ -0,0 +1,2694 @@ + + + + + + + + + + + + + + + + + + + + + + + + Hyperopt - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Hyperopt

+

This page explains how to tune your strategy by finding the optimal +parameters, a process called hyperparameter optimization. The bot uses algorithms included in the scikit-optimize package to accomplish this. +The search will burn all your CPU cores, make your laptop sound like a fighter jet and still take a long time.

+

In general, the search for best parameters starts with a few random combinations (see below for more details) and then uses Bayesian search with a ML regressor algorithm (currently ExtraTreesRegressor) to quickly find a combination of parameters in the search hyperspace that minimizes the value of the loss function.

+

Hyperopt requires historic data to be available, just as backtesting does (hyperopt runs backtesting many times with different parameters). +To learn how to get data for the pairs and exchange you're interested in, head over to the Data Downloading section of the documentation.

+
+

Bug

+

Hyperopt can crash when used with only 1 CPU Core as found out in Issue #1133

+
+
+

Note

+

Since 2021.4 release you no longer have to write a separate hyperopt class, but can configure the parameters directly in the strategy. +The legacy method is still supported, but it is no longer the recommended way of setting up hyperopt. +The legacy documentation is available at Legacy Hyperopt.

+
+

Install hyperopt dependencies

+

Since Hyperopt dependencies are not needed to run the bot itself, are heavy, can not be easily built on some platforms (like Raspberry PI), they are not installed by default. Before you run Hyperopt, you need to install the corresponding dependencies, as described in this section below.

+
+

Note

+

Since Hyperopt is a resource intensive process, running it on a Raspberry Pi is not recommended nor supported.

+
+

Docker

+

The docker-image includes hyperopt dependencies, no further action needed.

+

Easy installation script (setup.sh) / Manual installation

+
source .env/bin/activate
+pip install -r requirements-hyperopt.txt
+
+

Hyperopt command reference

+
usage: freqtrade hyperopt [-h] [-v] [--logfile FILE] [-V] [-c PATH] [-d PATH]
+                          [--userdir PATH] [-s NAME] [--strategy-path PATH]
+                          [--recursive-strategy-search] [--freqaimodel NAME]
+                          [--freqaimodel-path PATH] [-i TIMEFRAME]
+                          [--timerange TIMERANGE]
+                          [--data-format-ohlcv {json,jsongz,hdf5}]
+                          [--max-open-trades INT]
+                          [--stake-amount STAKE_AMOUNT] [--fee FLOAT]
+                          [-p PAIRS [PAIRS ...]] [--hyperopt-path PATH]
+                          [--eps] [--dmmp] [--enable-protections]
+                          [--dry-run-wallet DRY_RUN_WALLET]
+                          [--timeframe-detail TIMEFRAME_DETAIL] [-e INT]
+                          [--spaces {all,buy,sell,roi,stoploss,trailing,protection,trades,default} [{all,buy,sell,roi,stoploss,trailing,protection,trades,default} ...]]
+                          [--print-all] [--no-color] [--print-json] [-j JOBS]
+                          [--random-state INT] [--min-trades INT]
+                          [--hyperopt-loss NAME] [--disable-param-export]
+                          [--ignore-missing-spaces] [--analyze-per-epoch]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  -i TIMEFRAME, --timeframe TIMEFRAME
+                        Specify timeframe (`1m`, `5m`, `30m`, `1h`, `1d`).
+  --timerange TIMERANGE
+                        Specify what timerange of data to use.
+  --data-format-ohlcv {json,jsongz,hdf5}
+                        Storage format for downloaded candle (OHLCV) data.
+                        (default: `json`).
+  --max-open-trades INT
+                        Override the value of the `max_open_trades`
+                        configuration setting.
+  --stake-amount STAKE_AMOUNT
+                        Override the value of the `stake_amount` configuration
+                        setting.
+  --fee FLOAT           Specify fee ratio. Will be applied twice (on trade
+                        entry and exit).
+  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]
+                        Limit command to these pairs. Pairs are space-
+                        separated.
+  --hyperopt-path PATH  Specify additional lookup path for Hyperopt Loss
+                        functions.
+  --eps, --enable-position-stacking
+                        Allow buying the same pair multiple times (position
+                        stacking).
+  --dmmp, --disable-max-market-positions
+                        Disable applying `max_open_trades` during backtest
+                        (same as setting `max_open_trades` to a very high
+                        number).
+  --enable-protections, --enableprotections
+                        Enable protections for backtesting.Will slow
+                        backtesting down by a considerable amount, but will
+                        include configured protections
+  --dry-run-wallet DRY_RUN_WALLET, --starting-balance DRY_RUN_WALLET
+                        Starting balance, used for backtesting / hyperopt and
+                        dry-runs.
+  --timeframe-detail TIMEFRAME_DETAIL
+                        Specify detail timeframe for backtesting (`1m`, `5m`,
+                        `30m`, `1h`, `1d`).
+  -e INT, --epochs INT  Specify number of epochs (default: 100).
+  --spaces {all,buy,sell,roi,stoploss,trailing,protection,trades,default} [{all,buy,sell,roi,stoploss,trailing,protection,trades,default} ...]
+                        Specify which parameters to hyperopt. Space-separated
+                        list.
+  --print-all           Print all results, not only the best ones.
+  --no-color            Disable colorization of hyperopt results. May be
+                        useful if you are redirecting output to a file.
+  --print-json          Print output in JSON format.
+  -j JOBS, --job-workers JOBS
+                        The number of concurrently running jobs for
+                        hyperoptimization (hyperopt worker processes). If -1
+                        (default), all CPUs are used, for -2, all CPUs but one
+                        are used, etc. If 1 is given, no parallel computing
+                        code is used at all.
+  --random-state INT    Set random state to some positive integer for
+                        reproducible hyperopt results.
+  --min-trades INT      Set minimal desired number of trades for evaluations
+                        in the hyperopt optimization path (default: 1).
+  --hyperopt-loss NAME, --hyperoptloss NAME
+                        Specify the class name of the hyperopt loss function
+                        class (IHyperOptLoss). Different functions can
+                        generate completely different results, since the
+                        target for optimization is different. Built-in
+                        Hyperopt-loss-functions are:
+                        ShortTradeDurHyperOptLoss, OnlyProfitHyperOptLoss,
+                        SharpeHyperOptLoss, SharpeHyperOptLossDaily,
+                        SortinoHyperOptLoss, SortinoHyperOptLossDaily,
+                        CalmarHyperOptLoss, MaxDrawDownHyperOptLoss,
+                        MaxDrawDownRelativeHyperOptLoss,
+                        ProfitDrawDownHyperOptLoss
+  --disable-param-export
+                        Disable automatic hyperopt parameter export.
+  --ignore-missing-spaces, --ignore-unparameterized-spaces
+                        Suppress errors for any requested Hyperopt spaces that
+                        do not contain any parameters.
+  --analyze-per-epoch   Run populate_indicators once per epoch.
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+Strategy arguments:
+  -s NAME, --strategy NAME
+                        Specify strategy class name which will be used by the
+                        bot.
+  --strategy-path PATH  Specify additional strategy lookup path.
+  --recursive-strategy-search
+                        Recursively search for a strategy in the strategies
+                        folder.
+  --freqaimodel NAME    Specify a custom freqaimodels.
+  --freqaimodel-path PATH
+                        Specify additional lookup path for freqaimodels.
+
+

Hyperopt checklist

+

Checklist on all tasks / possibilities in hyperopt

+

Depending on the space you want to optimize, only some of the below are required:

+
    +
  • define parameters with space='buy' - for entry signal optimization
  • +
  • define parameters with space='sell' - for exit signal optimization
  • +
+
+

Note

+

populate_indicators needs to create all indicators any of the spaces may use, otherwise hyperopt will not work.

+
+

Rarely you may also need to create a nested class named HyperOpt and implement

+
    +
  • roi_space - for custom ROI optimization (if you need the ranges for the ROI parameters in the optimization hyperspace that differ from default)
  • +
  • generate_roi_table - for custom ROI optimization (if you need the ranges for the values in the ROI table that differ from default or the number of entries (steps) in the ROI table which differs from the default 4 steps)
  • +
  • stoploss_space - for custom stoploss optimization (if you need the range for the stoploss parameter in the optimization hyperspace that differs from default)
  • +
  • trailing_space - for custom trailing stop optimization (if you need the ranges for the trailing stop parameters in the optimization hyperspace that differ from default)
  • +
  • max_open_trades_space - for custom max_open_trades optimization (if you need the ranges for the max_open_trades parameter in the optimization hyperspace that differ from default)
  • +
+
+

Quickly optimize ROI, stoploss and trailing stoploss

+

You can quickly optimize the spaces roi, stoploss and trailing without changing anything in your strategy.

+
# Have a working strategy at hand.
+freqtrade hyperopt --hyperopt-loss SharpeHyperOptLossDaily --spaces roi stoploss trailing --strategy MyWorkingStrategy --config config.json -e 100
+
+
+

Hyperopt execution logic

+

Hyperopt will first load your data into memory and will then run populate_indicators() once per Pair to generate all indicators, unless --analyze-per-epoch is specified.

+

Hyperopt will then spawn into different processes (number of processors, or -j <n>), and run backtesting over and over again, changing the parameters that are part of the --spaces defined.

+

For every new set of parameters, freqtrade will run first populate_entry_trend() followed by populate_exit_trend(), and then run the regular backtesting process to simulate trades.

+

After backtesting, the results are passed into the loss function, which will evaluate if this result was better or worse than previous results.
+Based on the loss function result, hyperopt will determine the next set of parameters to try in the next round of backtesting.

+

Configure your Guards and Triggers

+

There are two places you need to change in your strategy file to add a new buy hyperopt for testing:

+
    +
  • Define the parameters at the class level hyperopt shall be optimizing.
  • +
  • Within populate_entry_trend() - use defined parameter values instead of raw constants.
  • +
+

There you have two different types of indicators: 1. guards and 2. triggers.

+
    +
  1. Guards are conditions like "never buy if ADX < 10", or never buy if current price is over EMA10.
  2. +
  3. Triggers are ones that actually trigger buy in specific moment, like "buy when EMA5 crosses over EMA10" or "buy when close price touches lower Bollinger band".
  4. +
+
+

Guards and Triggers

+

Technically, there is no difference between Guards and Triggers.
+However, this guide will make this distinction to make it clear that signals should not be "sticking". +Sticking signals are signals that are active for multiple candles. This can lead into entering a signal late (right before the signal disappears - which means that the chance of success is a lot lower than right at the beginning).

+
+

Hyper-optimization will, for each epoch round, pick one trigger and possibly multiple guards.

+

Exit signal optimization

+

Similar to the entry-signal above, exit-signals can also be optimized. +Place the corresponding settings into the following methods

+
    +
  • Define the parameters at the class level hyperopt shall be optimizing, either naming them sell_*, or by explicitly defining space='sell'.
  • +
  • Within populate_exit_trend() - use defined parameter values instead of raw constants.
  • +
+

The configuration and rules are the same than for buy signals.

+

Solving a Mystery

+

Let's say you are curious: should you use MACD crossings or lower Bollinger Bands to trigger your long entries. +And you also wonder should you use RSI or ADX to help with those decisions. +If you decide to use RSI or ADX, which values should I use for them?

+

So let's use hyperparameter optimization to solve this mystery.

+

Defining indicators to be used

+

We start by calculating the indicators our strategy is going to use.

+
class MyAwesomeStrategy(IStrategy):
+
+    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        Generate all indicators used by the strategy
+        """
+        dataframe['adx'] = ta.ADX(dataframe)
+        dataframe['rsi'] = ta.RSI(dataframe)
+        macd = ta.MACD(dataframe)
+        dataframe['macd'] = macd['macd']
+        dataframe['macdsignal'] = macd['macdsignal']
+        dataframe['macdhist'] = macd['macdhist']
+
+        bollinger = ta.BBANDS(dataframe, timeperiod=20, nbdevup=2.0, nbdevdn=2.0)
+        dataframe['bb_lowerband'] = bollinger['lowerband']
+        dataframe['bb_middleband'] = bollinger['middleband']
+        dataframe['bb_upperband'] = bollinger['upperband']
+        return dataframe
+
+

Hyperoptable parameters

+

We continue to define hyperoptable parameters:

+
class MyAwesomeStrategy(IStrategy):
+    buy_adx = DecimalParameter(20, 40, decimals=1, default=30.1, space="buy")
+    buy_rsi = IntParameter(20, 40, default=30, space="buy")
+    buy_adx_enabled = BooleanParameter(default=True, space="buy")
+    buy_rsi_enabled = CategoricalParameter([True, False], default=False, space="buy")
+    buy_trigger = CategoricalParameter(["bb_lower", "macd_cross_signal"], default="bb_lower", space="buy")
+
+

The above definition says: I have five parameters I want to randomly combine to find the best combination.
+buy_rsi is an integer parameter, which will be tested between 20 and 40. This space has a size of 20.
+buy_adx is a decimal parameter, which will be evaluated between 20 and 40 with 1 decimal place (so values are 20.1, 20.2, ...). This space has a size of 200.
+Then we have three category variables. First two are either True or False. +We use these to either enable or disable the ADX and RSI guards. +The last one we call trigger and use it to decide which buy trigger we want to use.

+
+

Parameter space assignment

+

Parameters must either be assigned to a variable named buy_* or sell_* - or contain space='buy' | space='sell' to be assigned to a space correctly. +If no parameter is available for a space, you'll receive the error that no space was found when running hyperopt.
+Parameters with unclear space (e.g. adx_period = IntParameter(4, 24, default=14) - no explicit nor implicit space) will not be detected and will therefore be ignored.

+
+

So let's write the buy strategy using these values:

+
    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        conditions = []
+        # GUARDS AND TRENDS
+        if self.buy_adx_enabled.value:
+            conditions.append(dataframe['adx'] > self.buy_adx.value)
+        if self.buy_rsi_enabled.value:
+            conditions.append(dataframe['rsi'] < self.buy_rsi.value)
+
+        # TRIGGERS
+        if self.buy_trigger.value == 'bb_lower':
+            conditions.append(dataframe['close'] < dataframe['bb_lowerband'])
+        if self.buy_trigger.value == 'macd_cross_signal':
+            conditions.append(qtpylib.crossed_above(
+                dataframe['macd'], dataframe['macdsignal']
+            ))
+
+        # Check that volume is not 0
+        conditions.append(dataframe['volume'] > 0)
+
+        if conditions:
+            dataframe.loc[
+                reduce(lambda x, y: x & y, conditions),
+                'enter_long'] = 1
+
+        return dataframe
+
+

Hyperopt will now call populate_entry_trend() many times (epochs) with different value combinations.
+It will use the given historical data and simulate buys based on the buy signals generated with the above function.
+Based on the results, hyperopt will tell you which parameter combination produced the best results (based on the configured loss function).

+
+

Note

+

The above setup expects to find ADX, RSI and Bollinger Bands in the populated indicators. +When you want to test an indicator that isn't used by the bot currently, remember to +add it to the populate_indicators() method in your strategy or hyperopt file.

+
+

Parameter types

+

There are four parameter types each suited for different purposes.

+
    +
  • IntParameter - defines an integral parameter with upper and lower boundaries of search space.
  • +
  • DecimalParameter - defines a floating point parameter with a limited number of decimals (default 3). Should be preferred instead of RealParameter in most cases.
  • +
  • RealParameter - defines a floating point parameter with upper and lower boundaries and no precision limit. Rarely used as it creates a space with a near infinite number of possibilities.
  • +
  • CategoricalParameter - defines a parameter with a predetermined number of choices.
  • +
  • BooleanParameter - Shorthand for CategoricalParameter([True, False]) - great for "enable" parameters.
  • +
+
+

Disabling parameter optimization

+

Each parameter takes two boolean parameters: +* load - when set to False it will not load values configured in buy_params and sell_params. +* optimize - when set to False parameter will not be included in optimization process. +Use these parameters to quickly prototype various ideas.

+
+
+

Warning

+

Hyperoptable parameters cannot be used in populate_indicators - as hyperopt does not recalculate indicators for each epoch, so the starting value would be used in this case.

+
+

Optimizing an indicator parameter

+

Assuming you have a simple strategy in mind - a EMA cross strategy (2 Moving averages crossing) - and you'd like to find the ideal parameters for this strategy. +By default, we assume a stoploss of 5% - and a take-profit (minimal_roi) of 10% - which means freqtrade will sell the trade once 10% profit has been reached.

+
from pandas import DataFrame
+from functools import reduce
+
+import talib.abstract as ta
+
+from freqtrade.strategy import (BooleanParameter, CategoricalParameter, DecimalParameter, 
+                                IStrategy, IntParameter)
+import freqtrade.vendor.qtpylib.indicators as qtpylib
+
+class MyAwesomeStrategy(IStrategy):
+    stoploss = -0.05
+    timeframe = '15m'
+    minimal_roi = {
+        "0":  0.10
+    }
+    # Define the parameter spaces
+    buy_ema_short = IntParameter(3, 50, default=5)
+    buy_ema_long = IntParameter(15, 200, default=50)
+
+
+    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """Generate all indicators used by the strategy"""
+
+        # Calculate all ema_short values
+        for val in self.buy_ema_short.range:
+            dataframe[f'ema_short_{val}'] = ta.EMA(dataframe, timeperiod=val)
+
+        # Calculate all ema_long values
+        for val in self.buy_ema_long.range:
+            dataframe[f'ema_long_{val}'] = ta.EMA(dataframe, timeperiod=val)
+
+        return dataframe
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        conditions = []
+        conditions.append(qtpylib.crossed_above(
+                dataframe[f'ema_short_{self.buy_ema_short.value}'], dataframe[f'ema_long_{self.buy_ema_long.value}']
+            ))
+
+        # Check that volume is not 0
+        conditions.append(dataframe['volume'] > 0)
+
+        if conditions:
+            dataframe.loc[
+                reduce(lambda x, y: x & y, conditions),
+                'enter_long'] = 1
+        return dataframe
+
+    def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        conditions = []
+        conditions.append(qtpylib.crossed_above(
+                dataframe[f'ema_long_{self.buy_ema_long.value}'], dataframe[f'ema_short_{self.buy_ema_short.value}']
+            ))
+
+        # Check that volume is not 0
+        conditions.append(dataframe['volume'] > 0)
+
+        if conditions:
+            dataframe.loc[
+                reduce(lambda x, y: x & y, conditions),
+                'exit_long'] = 1
+        return dataframe
+
+

Breaking it down:

+

Using self.buy_ema_short.range will return a range object containing all entries between the Parameters low and high value. +In this case (IntParameter(3, 50, default=5)), the loop would run for all numbers between 3 and 50 ([3, 4, 5, ... 49, 50]). +By using this in a loop, hyperopt will generate 48 new columns (['buy_ema_3', 'buy_ema_4', ... , 'buy_ema_50']).

+

Hyperopt itself will then use the selected value to create the buy and sell signals.

+

While this strategy is most likely too simple to provide consistent profit, it should serve as an example how optimize indicator parameters.

+
+

Note

+

self.buy_ema_short.range will act differently between hyperopt and other modes. For hyperopt, the above example may generate 48 new columns, however for all other modes (backtesting, dry/live), it will only generate the column for the selected value. You should therefore avoid using the resulting column with explicit values (values other than self.buy_ema_short.value).

+
+
+

Note

+

range property may also be used with DecimalParameter and CategoricalParameter. RealParameter does not provide this property due to infinite search space.

+
+
+Performance tip +

During normal hyperopting, indicators are calculated once and supplied to each epoch, linearly increasing RAM usage as a factor of increasing cores. As this also has performance implications, hyperopt provides --analyze-per-epoch which will move the execution of populate_indicators() to the epoch process, calculating a single value per parameter per epoch instead of using the .range functionality. In this case, .range functionality will only return the actually used value. This will reduce RAM usage, but increase CPU usage. However, your hyperopting run will be less likely to fail due to Out Of Memory (OOM) issues.

+

In either case, you should try to use space ranges as small as possible this will improve CPU/RAM usage in both scenarios.

+
+

Optimizing protections

+

Freqtrade can also optimize protections. How you optimize protections is up to you, and the following should be considered as example only.

+

The strategy will simply need to define the "protections" entry as property returning a list of protection configurations.

+
from pandas import DataFrame
+from functools import reduce
+
+import talib.abstract as ta
+
+from freqtrade.strategy import (BooleanParameter, CategoricalParameter, DecimalParameter, 
+                                IStrategy, IntParameter)
+import freqtrade.vendor.qtpylib.indicators as qtpylib
+
+class MyAwesomeStrategy(IStrategy):
+    stoploss = -0.05
+    timeframe = '15m'
+    # Define the parameter spaces
+    cooldown_lookback = IntParameter(2, 48, default=5, space="protection", optimize=True)
+    stop_duration = IntParameter(12, 200, default=5, space="protection", optimize=True)
+    use_stop_protection = BooleanParameter(default=True, space="protection", optimize=True)
+
+
+    @property
+    def protections(self):
+        prot = []
+
+        prot.append({
+            "method": "CooldownPeriod",
+            "stop_duration_candles": self.cooldown_lookback.value
+        })
+        if self.use_stop_protection.value:
+            prot.append({
+                "method": "StoplossGuard",
+                "lookback_period_candles": 24 * 3,
+                "trade_limit": 4,
+                "stop_duration_candles": self.stop_duration.value,
+                "only_per_pair": False
+            })
+
+        return prot
+
+    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        # ...
+
+

You can then run hyperopt as follows: +freqtrade hyperopt --hyperopt-loss SharpeHyperOptLossDaily --strategy MyAwesomeStrategy --spaces protection

+
+

Note

+

The protection space is not part of the default space, and is only available with the Parameters Hyperopt interface, not with the legacy hyperopt interface (which required separate hyperopt files). +Freqtrade will also automatically change the "--enable-protections" flag if the protection space is selected.

+
+
+

Warning

+

If protections are defined as property, entries from the configuration will be ignored. +It is therefore recommended to not define protections in the configuration.

+
+

Migrating from previous property setups

+

A migration from a previous setup is pretty simple, and can be accomplished by converting the protections entry to a property. +In simple terms, the following configuration will be converted to the below.

+
class MyAwesomeStrategy(IStrategy):
+    protections = [
+        {
+            "method": "CooldownPeriod",
+            "stop_duration_candles": 4
+        }
+    ]
+
+

Result

+
class MyAwesomeStrategy(IStrategy):
+
+    @property
+    def protections(self):
+        return [
+            {
+                "method": "CooldownPeriod",
+                "stop_duration_candles": 4
+            }
+        ]
+
+

You will then obviously also change potential interesting entries to parameters to allow hyper-optimization.

+

Optimizing max_entry_position_adjustment

+

While max_entry_position_adjustment is not a separate space, it can still be used in hyperopt by using the property approach shown above.

+
from pandas import DataFrame
+from functools import reduce
+
+import talib.abstract as ta
+
+from freqtrade.strategy import (BooleanParameter, CategoricalParameter, DecimalParameter, 
+                                IStrategy, IntParameter)
+import freqtrade.vendor.qtpylib.indicators as qtpylib
+
+class MyAwesomeStrategy(IStrategy):
+    stoploss = -0.05
+    timeframe = '15m'
+
+    # Define the parameter spaces
+    max_epa = CategoricalParameter([-1, 0, 1, 3, 5, 10], default=1, space="buy", optimize=True)
+
+    @property
+    def max_entry_position_adjustment(self):
+        return self.max_epa.value
+
+
+    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        # ...
+
+
+Using IntParameter +

You can also use the IntParameter for this optimization, but you must explicitly return an integer: +

max_epa = IntParameter(-1, 10, default=1, space="buy", optimize=True)
+
+@property
+def max_entry_position_adjustment(self):
+    return int(self.max_epa.value)
+

+
+

Loss-functions

+

Each hyperparameter tuning requires a target. This is usually defined as a loss function (sometimes also called objective function), which should decrease for more desirable results, and increase for bad results.

+

A loss function must be specified via the --hyperopt-loss <Class-name> argument (or optionally via the configuration under the "hyperopt_loss" key). +This class should be in its own file within the user_data/hyperopts/ directory.

+

Currently, the following loss functions are builtin:

+
    +
  • ShortTradeDurHyperOptLoss - (default legacy Freqtrade hyperoptimization loss function) - Mostly for short trade duration and avoiding losses.
  • +
  • OnlyProfitHyperOptLoss - takes only amount of profit into consideration.
  • +
  • SharpeHyperOptLoss - optimizes Sharpe Ratio calculated on trade returns relative to standard deviation.
  • +
  • SharpeHyperOptLossDaily - optimizes Sharpe Ratio calculated on daily trade returns relative to standard deviation.
  • +
  • SortinoHyperOptLoss - optimizes Sortino Ratio calculated on trade returns relative to downside standard deviation.
  • +
  • SortinoHyperOptLossDaily - optimizes Sortino Ratio calculated on daily trade returns relative to downside standard deviation.
  • +
  • MaxDrawDownHyperOptLoss - Optimizes Maximum absolute drawdown.
  • +
  • MaxDrawDownRelativeHyperOptLoss - Optimizes both maximum absolute drawdown while also adjusting for maximum relative drawdown.
  • +
  • CalmarHyperOptLoss - Optimizes Calmar Ratio calculated on trade returns relative to max drawdown.
  • +
  • ProfitDrawDownHyperOptLoss - Optimizes by max Profit & min Drawdown objective. DRAWDOWN_MULT variable within the hyperoptloss file can be adjusted to be stricter or more flexible on drawdown purposes.
  • +
+

Creation of a custom loss function is covered in the Advanced Hyperopt part of the documentation.

+

Execute Hyperopt

+

Once you have updated your hyperopt configuration you can run it. +Because hyperopt tries a lot of combinations to find the best parameters it will take time to get a good result.

+

We strongly recommend to use screen or tmux to prevent any connection loss.

+
freqtrade hyperopt --config config.json --hyperopt-loss <hyperoptlossname> --strategy <strategyname> -e 500 --spaces all
+
+

The -e option will set how many evaluations hyperopt will do. Since hyperopt uses Bayesian search, running too many epochs at once may not produce greater results. Experience has shown that best results are usually not improving much after 500-1000 epochs.
+Doing multiple runs (executions) with a few 1000 epochs and different random state will most likely produce different results.

+

The --spaces all option determines that all possible parameters should be optimized. Possibilities are listed below.

+
+

Note

+

Hyperopt will store hyperopt results with the timestamp of the hyperopt start time. +Reading commands (hyperopt-list, hyperopt-show) can use --hyperopt-filename <filename> to read and display older hyperopt results. +You can find a list of filenames with ls -l user_data/hyperopt_results/.

+
+

Execute Hyperopt with different historical data source

+

If you would like to hyperopt parameters using an alternate historical data set that +you have on-disk, use the --datadir PATH option. By default, hyperopt uses data from directory user_data/data.

+

Running Hyperopt with a smaller test-set

+

Use the --timerange argument to change how much of the test-set you want to use. +For example, to use one month of data, pass --timerange 20210101-20210201 (from january 2021 - february 2021) to the hyperopt call.

+

Full command:

+
freqtrade hyperopt --strategy <strategyname> --timerange 20210101-20210201
+
+

Running Hyperopt with Smaller Search Space

+

Use the --spaces option to limit the search space used by hyperopt. +Letting Hyperopt optimize everything is a huuuuge search space. +Often it might make more sense to start by just searching for initial buy algorithm. +Or maybe you just want to optimize your stoploss or roi table for that awesome new buy strategy you have.

+

Legal values are:

+
    +
  • all: optimize everything
  • +
  • buy: just search for a new buy strategy
  • +
  • sell: just search for a new sell strategy
  • +
  • roi: just optimize the minimal profit table for your strategy
  • +
  • stoploss: search for the best stoploss value
  • +
  • trailing: search for the best trailing stop values
  • +
  • trades: search for the best max open trades values
  • +
  • protection: search for the best protection parameters (read the protections section on how to properly define these)
  • +
  • default: all except trailing and protection
  • +
  • space-separated list of any of the above values for example --spaces roi stoploss
  • +
+

The default Hyperopt Search Space, used when no --space command line option is specified, does not include the trailing hyperspace. We recommend you to run optimization for the trailing hyperspace separately, when the best parameters for other hyperspaces were found, validated and pasted into your custom strategy.

+

Understand the Hyperopt Result

+

Once Hyperopt is completed you can use the result to update your strategy. +Given the following result from hyperopt:

+
Best result:
+
+    44/100:    135 trades. Avg profit  0.57%. Total profit  0.03871918 BTC (0.7722%). Avg duration 180.4 mins. Objective: 1.94367
+
+    # Buy hyperspace params:
+    buy_params = {
+        'buy_adx': 44,
+        'buy_rsi': 29,
+        'buy_adx_enabled': False,
+        'buy_rsi_enabled': True,
+        'buy_trigger': 'bb_lower'
+    }
+
+

You should understand this result like:

+
    +
  • The buy trigger that worked best was bb_lower.
  • +
  • You should not use ADX because 'buy_adx_enabled': False.
  • +
  • You should consider using the RSI indicator ('buy_rsi_enabled': True) and the best value is 29.0 ('buy_rsi': 29.0)
  • +
+

Automatic parameter application to the strategy

+

When using Hyperoptable parameters, the result of your hyperopt-run will be written to a json file next to your strategy (so for MyAwesomeStrategy.py, the file would be MyAwesomeStrategy.json).
+This file is also updated when using the hyperopt-show sub-command, unless --disable-param-export is provided to either of the 2 commands.

+

Your strategy class can also contain these results explicitly. Simply copy hyperopt results block and paste them at class level, replacing old parameters (if any). New parameters will automatically be loaded next time strategy is executed.

+

Transferring your whole hyperopt result to your strategy would then look like:

+
class MyAwesomeStrategy(IStrategy):
+    # Buy hyperspace params:
+    buy_params = {
+        'buy_adx': 44,
+        'buy_rsi': 29,
+        'buy_adx_enabled': False,
+        'buy_rsi_enabled': True,
+        'buy_trigger': 'bb_lower'
+    }
+
+
+

Note

+

Values in the configuration file will overwrite Parameter-file level parameters - and both will overwrite parameters within the strategy. +The prevalence is therefore: config > parameter file > strategy *_params > parameter default

+
+

Understand Hyperopt ROI results

+

If you are optimizing ROI (i.e. if optimization search-space contains 'all', 'default' or 'roi'), your result will look as follows and include a ROI table:

+
Best result:
+
+    44/100:    135 trades. Avg profit  0.57%. Total profit  0.03871918 BTC (0.7722%). Avg duration 180.4 mins. Objective: 1.94367
+
+    # ROI table:
+    minimal_roi = {
+        0: 0.10674,
+        21: 0.09158,
+        78: 0.03634,
+        118: 0
+    }
+
+

In order to use this best ROI table found by Hyperopt in backtesting and for live trades/dry-run, copy-paste it as the value of the minimal_roi attribute of your custom strategy:

+
    # Minimal ROI designed for the strategy.
+    # This attribute will be overridden if the config file contains "minimal_roi"
+    minimal_roi = {
+        0: 0.10674,
+        21: 0.09158,
+        78: 0.03634,
+        118: 0
+    }
+
+

As stated in the comment, you can also use it as the value of the minimal_roi setting in the configuration file.

+

Default ROI Search Space

+

If you are optimizing ROI, Freqtrade creates the 'roi' optimization hyperspace for you -- it's the hyperspace of components for the ROI tables. By default, each ROI table generated by the Freqtrade consists of 4 rows (steps). Hyperopt implements adaptive ranges for ROI tables with ranges for values in the ROI steps that depend on the timeframe used. By default the values vary in the following ranges (for some of the most used timeframes, values are rounded to 3 digits after the decimal point):

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
# step1m5m1h1d
100.011...0.11900.03...0.3100.068...0.71100.121...1.258
22...80.007...0.04210...400.02...0.11120...4800.045...0.2522880...115200.081...0.446
34...200.003...0.01520...1000.01...0.04240...12000.022...0.0915760...288000.040...0.162
46...440.030...2200.0360...26400.08640...633600.0
+

These ranges should be sufficient in most cases. The minutes in the steps (ROI dict keys) are scaled linearly depending on the timeframe used. The ROI values in the steps (ROI dict values) are scaled logarithmically depending on the timeframe used.

+

If you have the generate_roi_table() and roi_space() methods in your custom hyperopt, remove them in order to utilize these adaptive ROI tables and the ROI hyperoptimization space generated by Freqtrade by default.

+

Override the roi_space() method if you need components of the ROI tables to vary in other ranges. Override the generate_roi_table() and roi_space() methods and implement your own custom approach for generation of the ROI tables during hyperoptimization if you need a different structure of the ROI tables or other amount of rows (steps).

+

A sample for these methods can be found in the overriding pre-defined spaces section.

+
+

Reduced search space

+

To limit the search space further, Decimals are limited to 3 decimal places (a precision of 0.001). This is usually sufficient, every value more precise than this will usually result in overfitted results. You can however overriding pre-defined spaces to change this to your needs.

+
+

Understand Hyperopt Stoploss results

+

If you are optimizing stoploss values (i.e. if optimization search-space contains 'all', 'default' or 'stoploss'), your result will look as follows and include stoploss:

+
Best result:
+
+    44/100:    135 trades. Avg profit  0.57%. Total profit  0.03871918 BTC (0.7722%). Avg duration 180.4 mins. Objective: 1.94367
+
+    # Buy hyperspace params:
+    buy_params = {
+        'buy_adx': 44,
+        'buy_rsi': 29,
+        'buy_adx_enabled': False,
+        'buy_rsi_enabled': True,
+        'buy_trigger': 'bb_lower'
+    }
+
+    stoploss: -0.27996
+
+

In order to use this best stoploss value found by Hyperopt in backtesting and for live trades/dry-run, copy-paste it as the value of the stoploss attribute of your custom strategy:

+
    # Optimal stoploss designed for the strategy
+    # This attribute will be overridden if the config file contains "stoploss"
+    stoploss = -0.27996
+
+

As stated in the comment, you can also use it as the value of the stoploss setting in the configuration file.

+

Default Stoploss Search Space

+

If you are optimizing stoploss values, Freqtrade creates the 'stoploss' optimization hyperspace for you. By default, the stoploss values in that hyperspace vary in the range -0.35...-0.02, which is sufficient in most cases.

+

If you have the stoploss_space() method in your custom hyperopt file, remove it in order to utilize Stoploss hyperoptimization space generated by Freqtrade by default.

+

Override the stoploss_space() method and define the desired range in it if you need stoploss values to vary in other range during hyperoptimization. A sample for this method can be found in the overriding pre-defined spaces section.

+
+

Reduced search space

+

To limit the search space further, Decimals are limited to 3 decimal places (a precision of 0.001). This is usually sufficient, every value more precise than this will usually result in overfitted results. You can however overriding pre-defined spaces to change this to your needs.

+
+

Understand Hyperopt Trailing Stop results

+

If you are optimizing trailing stop values (i.e. if optimization search-space contains 'all' or 'trailing'), your result will look as follows and include trailing stop parameters:

+
Best result:
+
+    45/100:    606 trades. Avg profit  1.04%. Total profit  0.31555614 BTC ( 630.48%). Avg duration 150.3 mins. Objective: -1.10161
+
+    # Trailing stop:
+    trailing_stop = True
+    trailing_stop_positive = 0.02001
+    trailing_stop_positive_offset = 0.06038
+    trailing_only_offset_is_reached = True
+
+

In order to use these best trailing stop parameters found by Hyperopt in backtesting and for live trades/dry-run, copy-paste them as the values of the corresponding attributes of your custom strategy:

+
    # Trailing stop
+    # These attributes will be overridden if the config file contains corresponding values.
+    trailing_stop = True
+    trailing_stop_positive = 0.02001
+    trailing_stop_positive_offset = 0.06038
+    trailing_only_offset_is_reached = True
+
+

As stated in the comment, you can also use it as the values of the corresponding settings in the configuration file.

+

Default Trailing Stop Search Space

+

If you are optimizing trailing stop values, Freqtrade creates the 'trailing' optimization hyperspace for you. By default, the trailing_stop parameter is always set to True in that hyperspace, the value of the trailing_only_offset_is_reached vary between True and False, the values of the trailing_stop_positive and trailing_stop_positive_offset parameters vary in the ranges 0.02...0.35 and 0.01...0.1 correspondingly, which is sufficient in most cases.

+

Override the trailing_space() method and define the desired range in it if you need values of the trailing stop parameters to vary in other ranges during hyperoptimization. A sample for this method can be found in the overriding pre-defined spaces section.

+
+

Reduced search space

+

To limit the search space further, Decimals are limited to 3 decimal places (a precision of 0.001). This is usually sufficient, every value more precise than this will usually result in overfitted results. You can however overriding pre-defined spaces to change this to your needs.

+
+

Reproducible results

+

The search for optimal parameters starts with a few (currently 30) random combinations in the hyperspace of parameters, random Hyperopt epochs. These random epochs are marked with an asterisk character (*) in the first column in the Hyperopt output.

+

The initial state for generation of these random values (random state) is controlled by the value of the --random-state command line option. You can set it to some arbitrary value of your choice to obtain reproducible results.

+

If you have not set this value explicitly in the command line options, Hyperopt seeds the random state with some random value for you. The random state value for each Hyperopt run is shown in the log, so you can copy and paste it into the --random-state command line option to repeat the set of the initial random epochs used.

+

If you have not changed anything in the command line options, configuration, timerange, Strategy and Hyperopt classes, historical data and the Loss Function -- you should obtain same hyper-optimization results with same random state value used.

+

Output formatting

+

By default, hyperopt prints colorized results -- epochs with positive profit are printed in the green color. This highlighting helps you find epochs that can be interesting for later analysis. Epochs with zero total profit or with negative profits (losses) are printed in the normal color. If you do not need colorization of results (for instance, when you are redirecting hyperopt output to a file) you can switch colorization off by specifying the --no-color option in the command line.

+

You can use the --print-all command line option if you would like to see all results in the hyperopt output, not only the best ones. When --print-all is used, current best results are also colorized by default -- they are printed in bold (bright) style. This can also be switched off with the --no-color command line option.

+
+

Windows and color output

+

Windows does not support color-output natively, therefore it is automatically disabled. To have color-output for hyperopt running under windows, please consider using WSL.

+
+

Position stacking and disabling max market positions

+

In some situations, you may need to run Hyperopt (and Backtesting) with the +--eps/--enable-position-staking and --dmmp/--disable-max-market-positions arguments.

+

By default, hyperopt emulates the behavior of the Freqtrade Live Run/Dry Run, where only one +open trade is allowed for every traded pair. The total number of trades open for all pairs +is also limited by the max_open_trades setting. During Hyperopt/Backtesting this may lead to +some potential trades to be hidden (or masked) by previously open trades.

+

The --eps/--enable-position-stacking argument allows emulation of buying the same pair multiple times, +while --dmmp/--disable-max-market-positions disables applying max_open_trades +during Hyperopt/Backtesting (which is equal to setting max_open_trades to a very high +number).

+
+

Note

+

Dry/live runs will NOT use position stacking - therefore it does make sense to also validate the strategy without this as it's closer to reality.

+
+

You can also enable position stacking in the configuration file by explicitly setting +"position_stacking"=true.

+

Out of Memory errors

+

As hyperopt consumes a lot of memory (the complete data needs to be in memory once per parallel backtesting process), it's likely that you run into "out of memory" errors. +To combat these, you have multiple options:

+
    +
  • Reduce the amount of pairs.
  • +
  • Reduce the timerange used (--timerange <timerange>).
  • +
  • Avoid using --timeframe-detail (this loads a lot of additional data into memory).
  • +
  • Reduce the number of parallel processes (-j <n>).
  • +
  • Increase the memory of your machine.
  • +
  • Use --analyze-per-epoch if you're using a lot of parameters with .range functionality.
  • +
+

The objective has been evaluated at this point before.

+

If you see The objective has been evaluated at this point before. - then this is a sign that your space has been exhausted, or is close to that. +Basically all points in your space have been hit (or a local minima has been hit) - and hyperopt does no longer find points in the multi-dimensional space it did not try yet. +Freqtrade tries to counter the "local minima" problem by using new, randomized points in this case.

+

Example:

+
buy_ema_short = IntParameter(5, 20, default=10, space="buy", optimize=True)
+# This is the only parameter in the buy space
+
+

The buy_ema_short space has 15 possible values (5, 6, ... 19, 20). If you now run hyperopt for the buy space, hyperopt will only have 15 values to try before running out of options. +Your epochs should therefore be aligned to the possible values - or you should be ready to interrupt a run if you norice a lot of The objective has been evaluated at this point before. warnings.

+

Show details of Hyperopt results

+

After you run Hyperopt for the desired amount of epochs, you can later list all results for analysis, select only best or profitable once, and show the details for any of the epochs previously evaluated. This can be done with the hyperopt-list and hyperopt-show sub-commands. The usage of these sub-commands is described in the Utils chapter.

+

Validate backtesting results

+

Once the optimized strategy has been implemented into your strategy, you should backtest this strategy to make sure everything is working as expected.

+

To achieve same the results (number of trades, their durations, profit, etc.) as during Hyperopt, please use the same configuration and parameters (timerange, timeframe, ...) used for hyperopt --dmmp/--disable-max-market-positions and --eps/--enable-position-stacking for Backtesting.

+

Should results not match, please double-check to make sure you transferred all conditions correctly. +Pay special care to the stoploss, max_open_trades and trailing stoploss parameters, as these are often set in configuration files, which override changes to the strategy. +You should also carefully review the log of your backtest to ensure that there were no parameters inadvertently set by the configuration (like stoploss, max_open_trades or trailing_stop).

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/images/logo.png b/en/2023.4/images/logo.png new file mode 100644 index 0000000000000000000000000000000000000000..8a7ffdd7091e40e0713c71967fa2b4bc32cad64c GIT binary patch literal 11025 zcmZWvbySpn6Mg9J25A(eQ9wkx8>AcQ4rM8+r8}gN?(PO9MY_9}?gnZ2p7-x>&)GP8 zcF%8jo|!v$?wv3dC21^la&!Oyuw-Quixw)}hJJ>p#8atV@JASiFI}soU02)A6LR8&7{W!zJ zi%|0ZMOZL?V3;9r8Ks|WIG0ug7Y`1BX|MFKmi1)a52d#92*)6C?K`x5gYNLamPnff zXZiYz0WP6r^tMoU!8$5!ph$OOY#;+eyy=_x)PZ4$_sgC_dMd=_X86kNUH)2?)^GP8 zl_jMo-&`u%V#6&1Xl;h!rCd-vfW4EuU zOTi|PQt8kvHfF*U7&J~yM&_YfHJ1f^AsX;>OKvjEkv>Oe>Q0W0RdHJH4zie3($LUw z*4NjkWfO)fi+@8w7YsIlk^Iulo3NraRM*nd1L{?LoEaG0aHWdHR-l}WL8mc2!C0RN z2ng~npFMSTf3gb-K4BKH{w`;0L;f7d!`wBubO3vL;_p#Y@CAvneX?`n~y;MSQOQTgJe|rmkXT$&EPunK`l4Yq|NW$HKybG-)sw>NXOmAmXvsp7D_gEDgRW$7`Fg z6^e(eb&K(u5&G6()h38)=q2E#`(6rlv248OjsPHTM>|-{E-oh&`=bt!(6Z+zst~KA zuK;XNHxZKQWH^bCau;-T!$)|Cu_C5%TQ|9+K`t&YQ+MhAKV*t-(UXhVd6Lt`I%4t} z*kJ&s4;|?EDz3_2j6|JgKo8t9Dhazq(((5#Xvk6VNFDJhj1|3A@u}fDYtpf-VvzuCnqBzz2If^1Xx0BYJRu|0Vkq#L9?F0}wE=REjG!AC|$%MZ=Xh$iq$ zsIZ+%?*#O&Exb^S(yu4WEiLx;_R%V6B_EKGkdpIqbCtsC(*RCrgnc&djso15fHTI9 z{Nmzb=IQQ%kQE2FRo|B@%G$2aI! zXeLb&kBp4WvgtKWJI+_bhA)_X#zKYHCdRdmpARvNUG zqTo<@;V0|`;^EW<6kD+o4`rEMpzwnd@18l^Kf zKVw~gQmnen8H60Uwr%YPF4(!Qem|`6z#n4d&p5fr5f6#1zbGv&jRX#m2HoGx_?maCyn%#GGpNxZHW8 z+rshdtlP(LEm`!VGl}XIEImgM4n_nF0p*C**qvx9{3pF6MgkItXTN^^LcX76zW{OZh!<8xn7pggc)+GGT6Oqey`27S-3g$n-} zct$@_$*7_=ir|ER>Fo;=W z1scxet^r~ZGH-KpCg|l^U0#uLmJ+3gR+S-}OEhx|coCJ)631#;C-xx)-==H6ZvM4t z=Ldsp4~uqx&~s&n){A_S^;>#Z;rBwh&ieLkiyXnMaZm8)-5;1e2~XEmzAJ7!!>RI~ zHz#c5+*acbIjy5v!fX3#8Oh0v&6|+Rkq^|%jrL31-QC@T!Y_AY51o_~J0qXIU+tOJ zKag^n{8L=wH2Id*z`A z>YYGhg2ib19YRIG(&A!qn+T53Elj9=t-i65Ty#P3`_>QX@ZXopsqeP=3ZgY5db@4s zz1O^NCZx`Ii{g#W$7g3PkOZa5Ira@(_?5A{2j*sm=;qLRI9-?%!@m<-g8}fio>O=e zAK-eNq~{Q}{m^Dub&3RIUTY{UEL3vvV{mX?ao76*B4IzTGIWMHn*bAF7NxRbs`x60sJXJr`>LA-hFx>080f0Oc!ci zM#et5HcfT)1R2yYM_XH4gI35s_V|;%V(_=(Y0G=Ri;|xo99Oxj#&*csy)b*ya~0wA zi`7f{=$(P%Y~xlRv1ssBX1(3$;_L+a1@ldavsr)ryo>HZorJCi1qk$8>B1 zPt3v2&S>qjy}eDgd^!4)CxTayc27!bdHr3j>cH^u6)B#co}P!Wy>7)UlcuJtt1Hj8 zs=ky|@KZsu;h-Q0Zb8?Fb6E&;y!MOreyhG|Lf-r{jx!({oG8l6FEu%= zw8nw`T5~kx(+9WV6t2(N<%5h0DNeF`hLclMCG@#7HrFW^3h8{GnMOs^RRqy9a-z$# zeC|s^l$?*|Yl`kJdV71f!=uXenj8uw%m2LvQHwHX=0`#T;a(vEoXEe^HDAd+!$!M> z?uS=Eto>4>$DS2fH{P<@l@!|te?-t8erOb2{6bpnFP#dbpj`33Ss9o!a9bGw(j;cf zb*8MsPrp(~`<$yR`HL+G7qXnlvG^KUAAaIuG{gVIf9%1 zAI93cx)>j@J_-I@9M9L7PJ;(xc(G*}rvxLQU>BockxNW5L$YUVO1f=ow6$+uA|lII z{)z}61rep=K(Xz3DNY4T)R_mEQ$C$E)3XT+ua(RlB9U-8XlHrTugN{UYBG7TR3Si^LFer_ozCwH~k7ZZ~lXH+($DlPA5eErPtF=4N+Wz13T+P1RU z)zr{%E~IZ1NCZelkrT}=FAG?mZ}jfw=H{}WZ5xvo?wqup3hChJ64V z!|?s}p`*~(jllE5G>3b*$5y4d*jOiv-f$w4!^14mG`!< zd~zhEG-am}Drt_2;IKn_9^(IcDnCb-uVeObPJ0yhtXm@+bCMRj;;95`2vr*XKH!?8 z4@CUc*-DK26WKkcJTWox@m>g9QhR&jxM?jK!k2O_<{y6x*HuZ0!R^I zuYt7y=ME~!8wFgR;<%{u%|4VXPwGOr3g90!qEnD`5bubc7B9Vt*pQL*!3!vooL{?xvVsi~=mz2MgkiBu8Fh|b?fzx}64L}~Smn}?Nc z*i%KyIQ68TsXULk{`4ZbUdft`+5ko3*LrRE=#{UHT_%V@Le8oKU!ghuZ^6;T#N_Ot ztO{8_x`}780ZLac^bWG8^6PFVO_b&3>1IuOtzF&dRUBSwU2p--4XJUWvl^^rH?v1g zS@5;iS4vjxYT&))XhH!io;}ncNzPv!9=ZmQ5+4LZkba$4ZfSpM*@W;9ydM*wAMy`3 zCcRj+tr}Pmh)YXj*-Ae?uiQfwzD4r6O5c+pX>b;R&UJZcKVv}kT5~5|U2`!4XL@~~ zFR9}a61JMM{~a$Sd)=(K)i&9VWeSRQ-+$3^36}VdaErs=R7aPRSQvoO(|-ekr2jZG*Ns$FNHm@4w5)|O43 z{|>WL=d9`S8z)Oef%<`*E#Bi)T-ogb2&ReHA>!P8k~H=8^|4j$_u<2tf@#Z;DMhO8 zwAfgYzma6(V2>8(by${RCAN9Q(s93@(HcU$_nV7+hxmvNnM>CM-ur*Heg%}A^!XA; zA~$noY2I9H4+ZeHPNHh?&8MWKhz&Yd1E`rNiu|S`F`>?QqL`a4k*aTv7Rk${>fc9e zHpvz6+w!Ibe9Cu0Na8T^5rN$~;#^!Nqq9jZNC9zP^tV0NFkkU@L{7Nnn$SwGIxz)M*mi8Fk&(RZZ3hQrXw%Ndru60Ua_kb!&wNCbb4CUWYNsT7 zukDYW8L+b%>RH^(TFtO0rg2)g_xv0wLI|8g$1g>lL4}adA(iKaUMNt5Sl1YP!0KfD z)0f4svU1s2S0ssY1tu*|)2IL!zHcLuzR}&J16#AzVvpp-6HX^Wtb=Je3dvNbH@`OA zs00_2bVpy_<&rTH=QvwT>yzV&2R33J7op*T?(iE@4uvzVoAMhJwu#F;$h|LG%x{nQO^(sL~ZD zTP6LGqZmcj0(Z|0Zl(-p=zBP~x=MYx^Cq*OHcaRNN7Vg_sycTvtaeGWXg6_vz@ zlV_+-C>Osl3VPKuhU0HmM969AI|6+D+((`sZiLT9Kb>xAX=w1sUCutEUVZoOI;`sO`D4Z1ILMl<4mQz6Znu+Vappw1Ql2!=d(4XkBeTl{uz}GO zWY-~gB{urCJDGI8BWf#@sJr@cXIlkl|rb zhjOxZa&m*u$4gCFhs0p>cp!B`YZeSiMEI?p*qTqZ8j=Oy(j;xNqFQbZDoaA+pF zO5U>+xKlh^9EPbq&ibpZuAw-TB{;fs5EPT{Jr>r7ZQ;Dpflaf=@wd92u6lrx?;w?< z-S&I>x8$fSEh&kHg!E5|=xq7mUvkz(f&s>Tks#KUpE1VXw@x>oQPYEYiMQ&kXc?Pf z5|#wjYVdafve<5+<21NJbEEi+=AQ>)cO4@cWla%)haO za<$2v);SRRA>RF>vHKXK%hN)Z!5*)r<%dAWT};2D{)byPCyW+iC-jXNn6)DeLm0uW zF83xOGt(!#u`$U)W9Ar?Ksd2p@tqYXc-s0Vx!v7i8k|}C%h7nHcW~U2Y$K-2(qwgB zEMT&_7$^W|IH*gatwrK6Esd5-rm@rKj($fAsIQ?yr{Z48`sCYXIHldUuEGs0n&o_i zHtUzkO&~w<^xNGL6jS_nPS#2Q=Y#ajA?f2rxhI{;3tb|0+O>dadNjiCZ8``QA$^ec z#_bf`D<5v~uQTu>SY|W~b?fvfaJEF}>oE{IfxEHBRkisLpqH8yHjU+!^_vp+Gpn?O zL^|`^w_%FP$~f+QMHLkzK0Cu?YCb+b^0KmF4Xv#|c}OwwqN1Wek`@{hV+rzt1tIVI z;$R%=N=hLwcLhyN%}A*{w0S3bP@y^j2gaI@8A^^%MIS1YpHs3l1fxd(%6;k_u@2fp zagdG}(|*tcI)aKeqoM$xlf;th&WCR1^F`IUT@ZH@LT;kyB4Y}w+n<|myMWRZR{DFnWt z-=C;2w%YXajr8?V?q9i53Sc=7{;HglDeqytzvK>dZQs!n@X&F}tblV++Voy|uy^Z% zuAOz$8~8q5FFk(bOi4@A`}Cy~nztvAg)U_?G&(Azsjq)@$({<*_Nl2U-&mDC9hvt# z#$DdV25R*o9v1>i`vgMwP@!xujL}aRo%>JTWW|pgy8=9XHL&VrPW z!u0n?*;rU?Tb}O^3o2W;mHvVJk-DO!q~>nY0vz^!owOdp*LuQaJ>izo z9YznlN3<|r)W_XgnY%`GejDXS2+h)l1qoom^tjkJ2DDHd@!@enn|1DJolg`$Pepa`3l;3~nwt zBbQxve}7+^f}ETorWQG%t6a5&U7l$PS*h!E5_=?&+l3U)l4_45EStiqjS<6t>!RPY zdo^&BQQ=i$U2=@2w)ZzF(G>5q% z3L!?@6Gv_Iz!}3M#w@RI5;lDf^VRG4d=)lu9^d zy4(^c&Y*f}4I?2ab#zcHtB=Dwh+b3rTRn1PW;h;TG#%=mosvwwdiAR_-tE25Rrn*` z*mZP5YHEw3O-Q9~!vWo1ba$(1?bA)iOQIRx-&)qRpFicC)z{_3zUrdDr>taJa`C-x zx7*DZL&{R)z6l8Q+PzhTT{(K0`BgIGW9AQWDnx-gjqLugW5<(bW!@WXr!vh?Yc7d$ z8falsJoxedX{&Kk^A{HvQmEAHT>ccYo|t*#ceKs^GuQ|({S7Ml3hhW~R-kySf#Iy16B%VU5I|dT6vesy} z4Z$xN*I)&7gkAOWNIF7*Uj{f>N3=gG&O1#ieZSMW_;-YSvZbcIP z^haaTzbhz2WWqj=*f)oPvDhSY{~cb=uM#Dt=Dz`q7U76QoH)fkdEKm%Z7#chpfvGFMCgyI!F(=v)ReYdSoLI(T^ zA5z0|e}8|Eg#Yu}cUyVYMr4v%jqq`Ma5t`)-ur4oI@fuow)Wk7GtJK)7>D3|wYJ?o z$f=-aZ(mXDH~M&Y*~=@?%AQ+`(|>|kma*HZGxrPqL?AlC8CHmxhQRayw}=#MJo<9t z)C@`MRbL7#Kgft!c#)Q53v#rYxD~9e4xt;vFyV?0+V~3ib0@2dWsc#Dkv!_ZG#QTIy$&0n*!ag85aYN zO!fHT!QdDY2!$ zxwtg+UcZ2<)jj@GKEdH^1;OPh^mCpRn!Pa&R=!UQ8l!Q+$+~i;K*r{xV)YEF)be|oyoveqjj)9epl4T%kDz^?T3+hzeO!VfKKcnBS<*NXPSam9n~~CU!)4_<=hCs4zqRn!Q6R=WqRf zUHXL@jI&ht)Z3q*Y;K7+-W3N9VJF3kq|k2oPk|pgF8Bo=_-kr^D}w|qND3R5-NN^n zo8M0&4Xa|n;-+Hc!$Jp$OHF5@F--yOuM)Qv)VDDz0cI@f> z2mt7D|Lp=Ou}DM)q0yG>W1VLrW0G^rFrkNyX;)~*rKWlWo9pO6(%-~eW!rEWVFdLZNJupQ0$HY|0%gBs3$4TW)aAI!C+obWk9KeB`1Hz)#=KbKBOdSAK32Zpz z3aAs`rVq!ZYhL*o7dM$vv6=@CDH-ueNv<_HIXR{pLM{i>k)Z4GD1xD^MC>Ye4BXl0OoL#O|gECbf{h zladK1X<(*-AYa|EgJ?4<3!tB4z~}j|92WQC{rmS{w!dqssHkubT5-F2iJpK6AmSD~ z4<_!U=3OO!P^N)B^m<5+5AD!nF$ff^rE;i2ujYU(xFo5BvP(d%ik*Ywv^s$t#6S*2 zPDhOrb@Tt+(9)vQju%5KL8q>@)&NmHG?wcw`-NIbh#{Ey8m}Fl{lyg&6#6XNT3Qx@ z0mJAT)4p(r6%dKaWMSV*I$Tf!obWXd0zP@&R_|Fp=Y&ul@Gb>&)6vn{pkR`YbKw({ zkf78%Up=h77*4xC^#7n2JMiBSIXE3S3z|RT<1nkyMqH^jYWddd->DRd4{U+^9Q->j z+?+mQLQY|7_y@Y_qPq|*EUXG#q2ohi?K~PvP@rHr4KV2N@vc$n z!xfVOwVTe62N0y5ym13(gdA4|Vv@5P8fN}-My^p|gVY$~Pi9}KMp>Z00e4=Sw?~gT z3Ih5oBYuGt$5yqL@B5f zKp^!`!4wWE!cSLI-j2V{rEoCH`al8HQ>>rCcd5l}0RQt`;>+W%aL}SRup(nm6#a;n z@B?a`5&?GWASrw>&&){wuEky>0;%9ErBs8*qSAi55h14X%Titz)c&n(wn! zMjXVa%uG36-^tZGO6jGe39;Ni8yPuTl1kY`GPdJ+q0s&*3iVax^Lcb%h@};Gy<1|} zQY}$WB#w|g!w^TM1v6F51qpwN*(z`rR`c;Zx`aSo`C$rap{NMnCg#8b`*d%Qa-0%PRfmDVPFWBZ*{zwy!wfaq1qOB$G04X;1!NseIXb>1sb1OG;~pixj9go0RW&s=9KR?c zAZX9Hi=#EX1$VC#jo^(P(XpVIjJUXXno?1`DfT{iMfMHFPzT1wszA0Y9`X^i8&_C| zbgK-1>*meaY&*rv0Q~T&)1Pao4V?CIYMW2Li^!ElXkZt>yHLTLWzB(lStwQHFf5hyPrRQzBy-q z{sDS*_sz|GI_i3U`QSbR0SUW4=iTJLjg1XCfQQbU+7Ew=jzIw$s$&2y<9rqu4=8?~ zp)8t`O=36D&DSg&3L3y&KT*uGs;;h13L;6W%-tKNBn{eaq)=3-VU_Ug z{Jhma4Zw~!?UKXyPX4o*+1AGzet1Q2!($j6iy}b3-wV4<$I#G_1P|}4rkJ2IVg(5y zLlI-gnRAU;ieW}k;O`lR>em-G$Nc9A_Pu+c0QzCGx|-W!1c~2(-psw{@ZB!|F5!SKVUR6+de0^v?eFtt>btkf)}nv!->u@LsQ~WbZy84 z+|Sm{hmwfE$noHd!B56Cntt;9ceMRlK56Uvc`t=P7-L3agK7c4 zB7w+{q`@UroL@w)|AY~6fi>p+{rh))aBuAtgn-aLeC(p68u5${bSepSgKQI60XR>w9EMMi)UOK)XH=$D1*0E; ze%IJ4j7eg^)c=Bkup6ETN{0K9nP8)y2x_;~gFc7+5okwz%D2%f{0`l~j_b6#L@;KZuHj5C8xG literal 0 HcmV?d00001 diff --git a/en/2023.4/includes/pairlists/index.html b/en/2023.4/includes/pairlists/index.html new file mode 100644 index 000000000..1d4ea3786 --- /dev/null +++ b/en/2023.4/includes/pairlists/index.html @@ -0,0 +1,1653 @@ + + + + + + + + + + + + + + + + + + + + Pairlists - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Pairlists

+ +

Pairlists and Pairlist Handlers

+

Pairlist Handlers define the list of pairs (pairlist) that the bot should trade. They are configured in the pairlists section of the configuration settings.

+

In your configuration, you can use Static Pairlist (defined by the StaticPairList Pairlist Handler) and Dynamic Pairlist (defined by the VolumePairList Pairlist Handler).

+

Additionally, AgeFilter, PrecisionFilter, PriceFilter, ShuffleFilter, SpreadFilter and VolatilityFilter act as Pairlist Filters, removing certain pairs and/or moving their positions in the pairlist.

+

If multiple Pairlist Handlers are used, they are chained and a combination of all Pairlist Handlers forms the resulting pairlist the bot uses for trading and backtesting. Pairlist Handlers are executed in the sequence they are configured. You should always configure either StaticPairList or VolumePairList as the starting Pairlist Handler.

+

Inactive markets are always removed from the resulting pairlist. Explicitly blacklisted pairs (those in the pair_blacklist configuration setting) are also always removed from the resulting pairlist.

+

Pair blacklist

+

The pair blacklist (configured via exchange.pair_blacklist in the configuration) disallows certain pairs from trading. +This can be as simple as excluding DOGE/BTC - which will remove exactly this pair.

+

The pair-blacklist does also support wildcards (in regex-style) - so BNB/.* will exclude ALL pairs that start with BNB. +You may also use something like .*DOWN/BTC or .*UP/BTC to exclude leveraged tokens (check Pair naming conventions for your exchange!)

+

Available Pairlist Handlers

+ +
+

Testing pairlists

+

Pairlist configurations can be quite tricky to get right. Best use the test-pairlist utility sub-command to test your configuration quickly.

+
+

Static Pair List

+

By default, the StaticPairList method is used, which uses a statically defined pair whitelist from the configuration. The pairlist also supports wildcards (in regex-style) - so .*/BTC will include all pairs with BTC as a stake.

+

It uses configuration from exchange.pair_whitelist and exchange.pair_blacklist.

+
"pairlists": [
+    {"method": "StaticPairList"}
+],
+
+

By default, only currently enabled pairs are allowed. +To skip pair validation against active markets, set "allow_inactive": true within the StaticPairList configuration. +This can be useful for backtesting expired pairs (like quarterly spot-markets). +This option must be configured along with exchange.skip_pair_validation in the exchange configuration.

+

When used in a "follow-up" position (e.g. after VolumePairlist), all pairs in 'pair_whitelist' will be added to the end of the pairlist.

+

Volume Pair List

+

VolumePairList employs sorting/filtering of pairs by their trading volume. It selects number_assets top pairs with sorting based on the sort_key (which can only be quoteVolume).

+

When used in the chain of Pairlist Handlers in a non-leading position (after StaticPairList and other Pairlist Filters), VolumePairList considers outputs of previous Pairlist Handlers, adding its sorting/selection of the pairs by the trading volume.

+

When used in the leading position of the chain of Pairlist Handlers, the pair_whitelist configuration setting is ignored. Instead, VolumePairList selects the top assets from all available markets with matching stake-currency on the exchange.

+

The refresh_period setting allows to define the period (in seconds), at which the pairlist will be refreshed. Defaults to 1800s (30 minutes). +The pairlist cache (refresh_period) on VolumePairList is only applicable to generating pairlists. +Filtering instances (not the first position in the list) will not apply any cache and will always use up-to-date data.

+

VolumePairList is per default based on the ticker data from exchange, as reported by the ccxt library:

+
    +
  • The quoteVolume is the amount of quote (stake) currency traded (bought or sold) in last 24 hours.
  • +
+
"pairlists": [
+    {
+        "method": "VolumePairList",
+        "number_assets": 20,
+        "sort_key": "quoteVolume",
+        "min_value": 0,
+        "refresh_period": 1800
+    }
+],
+
+

You can define a minimum volume with min_value - which will filter out pairs with a volume lower than the specified value in the specified timerange.

+
VolumePairList Advanced mode
+

VolumePairList can also operate in an advanced mode to build volume over a given timerange of specified candle size. It utilizes exchange historical candle data, builds a typical price (calculated by (open+high+low)/3) and multiplies the typical price with every candle's volume. The sum is the quoteVolume over the given range. This allows different scenarios, for a more smoothened volume, when using longer ranges with larger candle sizes, or the opposite when using a short range with small candles.

+

For convenience lookback_days can be specified, which will imply that 1d candles will be used for the lookback. In the example below the pairlist would be created based on the last 7 days:

+
"pairlists": [
+    {
+        "method": "VolumePairList",
+        "number_assets": 20,
+        "sort_key": "quoteVolume",
+        "min_value": 0,
+        "refresh_period": 86400,
+        "lookback_days": 7
+    }
+],
+
+
+

Range look back and refresh period

+

When used in conjunction with lookback_days and lookback_timeframe the refresh_period can not be smaller than the candle size in seconds. As this will result in unnecessary requests to the exchanges API.

+
+
+

Performance implications when using lookback range

+

If used in first position in combination with lookback, the computation of the range based volume can be time and resource consuming, as it downloads candles for all tradable pairs. Hence it's highly advised to use the standard approach with VolumeFilter to narrow the pairlist down for further range volume calculation.

+
+
+Unsupported exchanges (Bittrex, Gemini) +

On some exchanges (like Bittrex and Gemini), regular VolumePairList does not work as the api does not natively provide 24h volume. This can be worked around by using candle data to build the volume. +To roughly simulate 24h volume, you can use the following configuration. +Please note that These pairlists will only refresh once per day.

+
"pairlists": [
+    {
+        "method": "VolumePairList",
+        "number_assets": 20,
+        "sort_key": "quoteVolume",
+        "min_value": 0,
+        "refresh_period": 86400,
+        "lookback_days": 1
+    }
+],
+
+
+

More sophisticated approach can be used, by using lookback_timeframe for candle size and lookback_period which specifies the amount of candles. This example will build the volume pairs based on a rolling period of 3 days of 1h candles:

+
"pairlists": [
+    {
+        "method": "VolumePairList",
+        "number_assets": 20,
+        "sort_key": "quoteVolume",
+        "min_value": 0,
+        "refresh_period": 3600,
+        "lookback_timeframe": "1h",
+        "lookback_period": 72
+    }
+],
+
+
+

Note

+

VolumePairList does not support backtesting mode.

+
+

ProducerPairList

+

With ProducerPairList, you can reuse the pairlist from a Producer without explicitly defining the pairlist on each consumer.

+

Consumer mode is required for this pairlist to work.

+

The pairlist will perform a check on active pairs against the current exchange configuration to avoid attempting to trade on invalid markets.

+

You can limit the length of the pairlist with the optional parameter number_assets. Using "number_assets"=0 or omitting this key will result in the reuse of all producer pairs valid for the current setup.

+
"pairlists": [
+    {
+        "method": "ProducerPairList",
+        "number_assets": 5,
+        "producer_name": "default",
+    }
+],
+
+
+

Combining pairlists

+

This pairlist can be combined with all other pairlists and filters for further pairlist reduction, and can also act as an "additional" pairlist, on top of already defined pairs. +ProducerPairList can also be used multiple times in sequence, combining the pairs from multiple producers. +Obviously in complex such configurations, the Producer may not provide data for all pairs, so the strategy must be fit for this.

+
+

RemotePairList

+

It allows the user to fetch a pairlist from a remote server or a locally stored json file within the freqtrade directory, enabling dynamic updates and customization of the trading pairlist.

+

The RemotePairList is defined in the pairlists section of the configuration settings. It uses the following configuration options:

+
"pairlists": [
+    {
+        "method": "RemotePairList",
+        "pairlist_url": "https://example.com/pairlist",
+        "number_assets": 10,
+        "refresh_period": 1800,
+        "keep_pairlist_on_failure": true,
+        "read_timeout": 60,
+        "bearer_token": "my-bearer-token"
+    }
+]
+
+

The pairlist_url option specifies the URL of the remote server where the pairlist is located, or the path to a local file (if file:/// is prepended). This allows the user to use either a remote server or a local file as the source for the pairlist.

+

The user is responsible for providing a server or local file that returns a JSON object with the following structure:

+
{
+    "pairs": ["XRP/USDT", "ETH/USDT", "LTC/USDT"],
+    "refresh_period": 1800,
+}
+
+

The pairs property should contain a list of strings with the trading pairs to be used by the bot. The refresh_period property is optional and specifies the number of seconds that the pairlist should be cached before being refreshed.

+

The optional keep_pairlist_on_failure specifies whether the previous received pairlist should be used if the remote server is not reachable or returns an error. The default value is true.

+

The optional read_timeout specifies the maximum amount of time (in seconds) to wait for a response from the remote source, The default value is 60.

+

The optional bearer_token will be included in the requests Authorization Header.

+
+

Note

+

In case of a server error the last received pairlist will be kept if keep_pairlist_on_failure is set to true, when set to false a empty pairlist is returned.

+
+

AgeFilter

+

Removes pairs that have been listed on the exchange for less than min_days_listed days (defaults to 10) or more than max_days_listed days (defaults None mean infinity).

+

When pairs are first listed on an exchange they can suffer huge price drops and volatility +in the first few days while the pair goes through its price-discovery period. Bots can often +be caught out buying before the pair has finished dropping in price.

+

This filter allows freqtrade to ignore pairs until they have been listed for at least min_days_listed days and listed before max_days_listed.

+

OffsetFilter

+

Offsets an incoming pairlist by a given offset value.

+

As an example it can be used in conjunction with VolumeFilter to remove the top X volume pairs. Or to split a larger pairlist on two bot instances.

+

Example to remove the first 10 pairs from the pairlist, and takes the next 20 (taking items 10-30 of the initial list):

+
"pairlists": [
+    // ...
+    {
+        "method": "OffsetFilter",
+        "offset": 10,
+        "number_assets": 20
+    }
+],
+
+
+

Warning

+

When OffsetFilter is used to split a larger pairlist among multiple bots in combination with VolumeFilter +it can not be guaranteed that pairs won't overlap due to slightly different refresh intervals for the +VolumeFilter.

+
+
+

Note

+

An offset larger than the total length of the incoming pairlist will result in an empty pairlist.

+
+

PerformanceFilter

+

Sorts pairs by past trade performance, as follows:

+
    +
  1. Positive performance.
  2. +
  3. No closed trades yet.
  4. +
  5. Negative performance.
  6. +
+

Trade count is used as a tie breaker.

+

You can use the minutes parameter to only consider performance of the past X minutes (rolling window). +Not defining this parameter (or setting it to 0) will use all-time performance.

+

The optional min_profit (as ratio -> a setting of 0.01 corresponds to 1%) parameter defines the minimum profit a pair must have to be considered. +Pairs below this level will be filtered out. +Using this parameter without minutes is highly discouraged, as it can lead to an empty pairlist without a way to recover.

+
"pairlists": [
+    // ...
+    {
+        "method": "PerformanceFilter",
+        "minutes": 1440,  // rolling 24h
+        "min_profit": 0.01  // minimal profit 1%
+    }
+],
+
+

As this Filter uses past performance of the bot, it'll have some startup-period - and should only be used after the bot has a few 100 trades in the database.

+
+

Backtesting

+

PerformanceFilter does not support backtesting mode.

+
+

PrecisionFilter

+

Filters low-value coins which would not allow setting stoplosses.

+
+

Backtesting

+

PrecisionFilter does not support backtesting mode using multiple strategies.

+
+

PriceFilter

+

The PriceFilter allows filtering of pairs by price. Currently the following price filters are supported:

+
    +
  • min_price
  • +
  • max_price
  • +
  • max_value
  • +
  • low_price_ratio
  • +
+

The min_price setting removes pairs where the price is below the specified price. This is useful if you wish to avoid trading very low-priced pairs. +This option is disabled by default, and will only apply if set to > 0.

+

The max_price setting removes pairs where the price is above the specified price. This is useful if you wish to trade only low-priced pairs. +This option is disabled by default, and will only apply if set to > 0.

+

The max_value setting removes pairs where the minimum value change is above a specified value. +This is useful when an exchange has unbalanced limits. For example, if step-size = 1 (so you can only buy 1, or 2, or 3, but not 1.1 Coins) - and the price is pretty high (like 20$) as the coin has risen sharply since the last limit adaption. +As a result of the above, you can only buy for 20$, or 40$ - but not for 25$. +On exchanges that deduct fees from the receiving currency (e.g. binance) - this can result in high value coins / amounts that are unsellable as the amount is slightly below the limit.

+

The low_price_ratio setting removes pairs where a raise of 1 price unit (pip) is above the low_price_ratio ratio. +This option is disabled by default, and will only apply if set to > 0.

+

For PriceFilter at least one of its min_price, max_price or low_price_ratio settings must be applied.

+

Calculation example:

+

Min price precision for SHITCOIN/BTC is 8 decimals. If its price is 0.00000011 - one price step above would be 0.00000012, which is ~9% higher than the previous price value. You may filter out this pair by using PriceFilter with low_price_ratio set to 0.09 (9%) or with min_price set to 0.00000011, correspondingly.

+
+

Low priced pairs

+

Low priced pairs with high "1 pip movements" are dangerous since they are often illiquid and it may also be impossible to place the desired stoploss, which can often result in high losses since price needs to be rounded to the next tradable price - so instead of having a stoploss of -5%, you could end up with a stoploss of -9% simply due to price rounding.

+
+

ShuffleFilter

+

Shuffles (randomizes) pairs in the pairlist. It can be used for preventing the bot from trading some of the pairs more frequently then others when you want all pairs be treated with the same priority.

+

By default, ShuffleFilter will shuffle pairs once per candle. +To shuffle on every iteration, set "shuffle_frequency" to "iteration" instead of the default of "candle".

+
    {
+        "method": "ShuffleFilter", 
+        "shuffle_frequency": "candle",
+        "seed": 42
+    }
+
+
+

Tip

+

You may set the seed value for this Pairlist to obtain reproducible results, which can be useful for repeated backtesting sessions. If seed is not set, the pairs are shuffled in the non-repeatable random order. ShuffleFilter will automatically detect runmodes and apply the seed only for backtesting modes - if a seed value is set.

+
+

SpreadFilter

+

Removes pairs that have a difference between asks and bids above the specified ratio, max_spread_ratio (defaults to 0.005).

+

Example:

+

If DOGE/BTC maximum bid is 0.00000026 and minimum ask is 0.00000027, the ratio is calculated as: 1 - bid/ask ~= 0.037 which is > 0.005 and this pair will be filtered out.

+

RangeStabilityFilter

+

Removes pairs where the difference between lowest low and highest high over lookback_days days is below min_rate_of_change or above max_rate_of_change. Since this is a filter that requires additional data, the results are cached for refresh_period.

+

In the below example: +If the trading range over the last 10 days is <1% or >99%, remove the pair from the whitelist.

+
"pairlists": [
+    {
+        "method": "RangeStabilityFilter",
+        "lookback_days": 10,
+        "min_rate_of_change": 0.01,
+        "max_rate_of_change": 0.99,
+        "refresh_period": 1440
+    }
+]
+
+
+

Tip

+

This Filter can be used to automatically remove stable coin pairs, which have a very low trading range, and are therefore extremely difficult to trade with profit. +Additionally, it can also be used to automatically remove pairs with extreme high/low variance over a given amount of time.

+
+

VolatilityFilter

+

Volatility is the degree of historical variation of a pairs over time, it is measured by the standard deviation of logarithmic daily returns. Returns are assumed to be normally distributed, although actual distribution might be different. In a normal distribution, 68% of observations fall within one standard deviation and 95% of observations fall within two standard deviations. Assuming a volatility of 0.05 means that the expected returns for 20 out of 30 days is expected to be less than 5% (one standard deviation). Volatility is a positive ratio of the expected deviation of return and can be greater than 1.00. Please refer to the wikipedia definition of volatility.

+

This filter removes pairs if the average volatility over a lookback_days days is below min_volatility or above max_volatility. Since this is a filter that requires additional data, the results are cached for refresh_period.

+

This filter can be used to narrow down your pairs to a certain volatility or avoid very volatile pairs.

+

In the below example: +If the volatility over the last 10 days is not in the range of 0.05-0.50, remove the pair from the whitelist. The filter is applied every 24h.

+
"pairlists": [
+    {
+        "method": "VolatilityFilter",
+        "lookback_days": 10,
+        "min_volatility": 0.05,
+        "max_volatility": 0.50,
+        "refresh_period": 86400
+    }
+]
+
+

Full example of Pairlist Handlers

+

The below example blacklists BNB/BTC, uses VolumePairList with 20 assets, sorting pairs by quoteVolume and applies PrecisionFilter and PriceFilter, filtering all assets where 1 price unit is > 1%. Then the SpreadFilter and VolatilityFilter is applied and pairs are finally shuffled with the random seed set to some predefined value.

+
"exchange": {
+    "pair_whitelist": [],
+    "pair_blacklist": ["BNB/BTC"]
+},
+"pairlists": [
+    {
+        "method": "VolumePairList",
+        "number_assets": 20,
+        "sort_key": "quoteVolume"
+    },
+    {"method": "AgeFilter", "min_days_listed": 10},
+    {"method": "PrecisionFilter"},
+    {"method": "PriceFilter", "low_price_ratio": 0.01},
+    {"method": "SpreadFilter", "max_spread_ratio": 0.005},
+    {
+        "method": "RangeStabilityFilter",
+        "lookback_days": 10,
+        "min_rate_of_change": 0.01,
+        "refresh_period": 1440
+    },
+    {
+        "method": "VolatilityFilter",
+        "lookback_days": 10,
+        "min_volatility": 0.05,
+        "max_volatility": 0.50,
+        "refresh_period": 86400
+    },
+    {"method": "ShuffleFilter", "seed": 42}
+],
+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/includes/pricing/index.html b/en/2023.4/includes/pricing/index.html new file mode 100644 index 000000000..158039c4d --- /dev/null +++ b/en/2023.4/includes/pricing/index.html @@ -0,0 +1,1488 @@ + + + + + + + + + + + + + + + + + + + + Pricing - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Pricing

+ +

Prices used for orders

+

Prices for regular orders can be controlled via the parameter structures entry_pricing for trade entries and exit_pricing for trade exits. +Prices are always retrieved right before an order is placed, either by querying the exchange tickers or by using the orderbook data.

+
+

Note

+

Orderbook data used by Freqtrade are the data retrieved from exchange by the ccxt's function fetch_order_book(), i.e. are usually data from the L2-aggregated orderbook, while the ticker data are the structures returned by the ccxt's fetch_ticker()/fetch_tickers() functions. Refer to the ccxt library documentation for more details.

+
+
+

Using market orders

+

Please read the section Market order pricing section when using market orders.

+
+

Entry price

+

Enter price side

+

The configuration setting entry_pricing.price_side defines the side of the orderbook the bot looks for when buying.

+

The following displays an orderbook.

+
...
+103
+102
+101  # ask
+-------------Current spread
+99   # bid
+98
+97
+...
+
+

If entry_pricing.price_side is set to "bid", then the bot will use 99 as entry price.
+In line with that, if entry_pricing.price_side is set to "ask", then the bot will use 101 as entry price.

+

Depending on the order direction (long/short), this will lead to different results. Therefore we recommend to use "same" or "other" for this configuration instead. +This would result in the following pricing matrix:

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
directionOrdersettingpricecrosses spread
longbuyask101yes
longbuybid99no
longbuysame99no
longbuyother101yes
shortsellask101no
shortsellbid99yes
shortsellsame101no
shortsellother99yes
+

Using the other side of the orderbook often guarantees quicker filled orders, but the bot can also end up paying more than what would have been necessary. +Taker fees instead of maker fees will most likely apply even when using limit buy orders. +Also, prices at the "other" side of the spread are higher than prices at the "bid" side in the orderbook, so the order behaves similar to a market order (however with a maximum price).

+

Entry price with Orderbook enabled

+

When entering a trade with the orderbook enabled (entry_pricing.use_order_book=True), Freqtrade fetches the entry_pricing.order_book_top entries from the orderbook and uses the entry specified as entry_pricing.order_book_top on the configured side (entry_pricing.price_side) of the orderbook. 1 specifies the topmost entry in the orderbook, while 2 would use the 2nd entry in the orderbook, and so on.

+

Entry price without Orderbook enabled

+

The following section uses side as the configured entry_pricing.price_side (defaults to "same").

+

When not using orderbook (entry_pricing.use_order_book=False), Freqtrade uses the best side price from the ticker if it's below the last traded price from the ticker. Otherwise (when the side price is above the last price), it calculates a rate between side and last price based on entry_pricing.price_last_balance.

+

The entry_pricing.price_last_balance configuration parameter controls this. A value of 0.0 will use side price, while 1.0 will use the last price and values between those interpolate between ask and last price.

+

Check depth of market

+

When check depth of market is enabled (entry_pricing.check_depth_of_market.enabled=True), the entry signals are filtered based on the orderbook depth (sum of all amounts) for each orderbook side.

+

Orderbook bid (buy) side depth is then divided by the orderbook ask (sell) side depth and the resulting delta is compared to the value of the entry_pricing.check_depth_of_market.bids_to_ask_delta parameter. The entry order is only executed if the orderbook delta is greater than or equal to the configured delta value.

+
+

Note

+

A delta value below 1 means that ask (sell) orderbook side depth is greater than the depth of the bid (buy) orderbook side, while a value greater than 1 means opposite (depth of the buy side is higher than the depth of the sell side).

+
+

Exit price

+

Exit price side

+

The configuration setting exit_pricing.price_side defines the side of the spread the bot looks for when exiting a trade.

+

The following displays an orderbook:

+
...
+103
+102
+101  # ask
+-------------Current spread
+99   # bid
+98
+97
+...
+
+

If exit_pricing.price_side is set to "ask", then the bot will use 101 as exiting price.
+In line with that, if exit_pricing.price_side is set to "bid", then the bot will use 99 as exiting price.

+

Depending on the order direction (long/short), this will lead to different results. Therefore we recommend to use "same" or "other" for this configuration instead. +This would result in the following pricing matrix:

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
DirectionOrdersettingpricecrosses spread
longsellask101no
longsellbid99yes
longsellsame101no
longsellother99yes
shortbuyask101yes
shortbuybid99no
shortbuysame99no
shortbuyother101yes
+

Exit price with Orderbook enabled

+

When exiting with the orderbook enabled (exit_pricing.use_order_book=True), Freqtrade fetches the exit_pricing.order_book_top entries in the orderbook and uses the entry specified as exit_pricing.order_book_top from the configured side (exit_pricing.price_side) as trade exit price.

+

1 specifies the topmost entry in the orderbook, while 2 would use the 2nd entry in the orderbook, and so on.

+

Exit price without Orderbook enabled

+

The following section uses side as the configured exit_pricing.price_side (defaults to "ask").

+

When not using orderbook (exit_pricing.use_order_book=False), Freqtrade uses the best side price from the ticker if it's above the last traded price from the ticker. Otherwise (when the side price is below the last price), it calculates a rate between side and last price based on exit_pricing.price_last_balance.

+

The exit_pricing.price_last_balance configuration parameter controls this. A value of 0.0 will use side price, while 1.0 will use the last price and values between those interpolate between side and last price.

+

Market order pricing

+

When using market orders, prices should be configured to use the "correct" side of the orderbook to allow realistic pricing detection. +Assuming both entry and exits are using market orders, a configuration similar to the following must be used

+
  "order_types": {
+    "entry": "market",
+    "exit": "market"
+    // ...
+  },
+  "entry_pricing": {
+    "price_side": "other",
+    // ...
+  },
+  "exit_pricing":{
+    "price_side": "other",
+    // ...
+  },
+
+

Obviously, if only one side is using limit orders, different pricing combinations can be used.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/includes/protections/index.html b/en/2023.4/includes/protections/index.html new file mode 100644 index 000000000..86ec5e956 --- /dev/null +++ b/en/2023.4/includes/protections/index.html @@ -0,0 +1,1436 @@ + + + + + + + + + + + + + + + + + + + + Protections - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Protections

+ +

Protections

+
+

Beta feature

+

This feature is still in it's testing phase. Should you notice something you think is wrong please let us know via Discord or via Github Issue.

+
+

Protections will protect your strategy from unexpected events and market conditions by temporarily stop trading for either one pair, or for all pairs. +All protection end times are rounded up to the next candle to avoid sudden, unexpected intra-candle buys.

+
+

Note

+

Not all Protections will work for all strategies, and parameters will need to be tuned for your strategy to improve performance.

+
+
+

Tip

+

Each Protection can be configured multiple times with different parameters, to allow different levels of protection (short-term / long-term).

+
+
+

Backtesting

+

Protections are supported by backtesting and hyperopt, but must be explicitly enabled by using the --enable-protections flag.

+
+
+

Setting protections from the configuration

+

Setting protections from the configuration via "protections": [], key should be considered deprecated and will be removed in a future version. +It is also no longer guaranteed that your protections apply to the strategy in cases where the strategy defines protections as property.

+
+

Available Protections

+
    +
  • StoplossGuard Stop trading if a certain amount of stoploss occurred within a certain time window.
  • +
  • MaxDrawdown Stop trading if max-drawdown is reached.
  • +
  • LowProfitPairs Lock pairs with low profits
  • +
  • CooldownPeriod Don't enter a trade right after selling a trade.
  • +
+

Common settings to all Protections

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
ParameterDescription
methodProtection name to use.
Datatype: String, selected from available Protections
stop_duration_candlesFor how many candles should the lock be set?
Datatype: Positive integer (in candles)
stop_durationhow many minutes should protections be locked.
Cannot be used together with stop_duration_candles.
Datatype: Float (in minutes)
lookback_period_candlesOnly trades that completed within the last lookback_period_candles candles will be considered. This setting may be ignored by some Protections.
Datatype: Positive integer (in candles).
lookback_periodOnly trades that completed after current_time - lookback_period will be considered.
Cannot be used together with lookback_period_candles.
This setting may be ignored by some Protections.
Datatype: Float (in minutes)
trade_limitNumber of trades required at minimum (not used by all Protections).
Datatype: Positive integer
+
+

Durations

+

Durations (stop_duration* and lookback_period* can be defined in either minutes or candles). +For more flexibility when testing different timeframes, all below examples will use the "candle" definition.

+
+

Stoploss Guard

+

StoplossGuard selects all trades within lookback_period in minutes (or in candles when using lookback_period_candles). +If trade_limit or more trades resulted in stoploss, trading will stop for stop_duration in minutes (or in candles when using stop_duration_candles).

+

This applies across all pairs, unless only_per_pair is set to true, which will then only look at one pair at a time.

+

Similarly, this protection will by default look at all trades (long and short). For futures bots, setting only_per_side will make the bot only consider one side, and will then only lock this one side, allowing for example shorts to continue after a series of long stoplosses.

+

required_profit will determine the required relative profit (or loss) for stoplosses to consider. This should normally not be set and defaults to 0.0 - which means all losing stoplosses will be triggering a block.

+

The below example stops trading for all pairs for 4 candles after the last trade if the bot hit stoploss 4 times within the last 24 candles.

+
@property
+def protections(self):
+    return [
+        {
+            "method": "StoplossGuard",
+            "lookback_period_candles": 24,
+            "trade_limit": 4,
+            "stop_duration_candles": 4,
+            "required_profit": 0.0,
+            "only_per_pair": False,
+            "only_per_side": False
+        }
+    ]
+
+
+

Note

+

StoplossGuard considers all trades with the results "stop_loss", "stoploss_on_exchange" and "trailing_stop_loss" if the resulting profit was negative. +trade_limit and lookback_period will need to be tuned for your strategy.

+
+

MaxDrawdown

+

MaxDrawdown uses all trades within lookback_period in minutes (or in candles when using lookback_period_candles) to determine the maximum drawdown. If the drawdown is below max_allowed_drawdown, trading will stop for stop_duration in minutes (or in candles when using stop_duration_candles) after the last trade - assuming that the bot needs some time to let markets recover.

+

The below sample stops trading for 12 candles if max-drawdown is > 20% considering all pairs - with a minimum of trade_limit trades - within the last 48 candles. If desired, lookback_period and/or stop_duration can be used.

+
@property
+def protections(self):
+    return  [
+        {
+            "method": "MaxDrawdown",
+            "lookback_period_candles": 48,
+            "trade_limit": 20,
+            "stop_duration_candles": 12,
+            "max_allowed_drawdown": 0.2
+        },
+    ]
+
+

Low Profit Pairs

+

LowProfitPairs uses all trades for a pair within lookback_period in minutes (or in candles when using lookback_period_candles) to determine the overall profit ratio. +If that ratio is below required_profit, that pair will be locked for stop_duration in minutes (or in candles when using stop_duration_candles).

+

For futures bots, setting only_per_side will make the bot only consider one side, and will then only lock this one side, allowing for example shorts to continue after a series of long losses.

+

The below example will stop trading a pair for 60 minutes if the pair does not have a required profit of 2% (and a minimum of 2 trades) within the last 6 candles.

+
@property
+def protections(self):
+    return [
+        {
+            "method": "LowProfitPairs",
+            "lookback_period_candles": 6,
+            "trade_limit": 2,
+            "stop_duration": 60,
+            "required_profit": 0.02,
+            "only_per_pair": False,
+        }
+    ]
+
+

Cooldown Period

+

CooldownPeriod locks a pair for stop_duration in minutes (or in candles when using stop_duration_candles) after selling, avoiding a re-entry for this pair for stop_duration minutes.

+

The below example will stop trading a pair for 2 candles after closing a trade, allowing this pair to "cool down".

+
@property
+def protections(self):
+    return  [
+        {
+            "method": "CooldownPeriod",
+            "stop_duration_candles": 2
+        }
+    ]
+
+
+

Note

+

This Protection applies only at pair-level, and will never lock all pairs globally. +This Protection does not consider lookback_period as it only looks at the latest trade.

+
+

Full example of Protections

+

All protections can be combined at will, also with different parameters, creating a increasing wall for under-performing pairs. +All protections are evaluated in the sequence they are defined.

+

The below example assumes a timeframe of 1 hour:

+
    +
  • Locks each pair after selling for an additional 5 candles (CooldownPeriod), giving other pairs a chance to get filled.
  • +
  • Stops trading for 4 hours (4 * 1h candles) if the last 2 days (48 * 1h candles) had 20 trades, which caused a max-drawdown of more than 20%. (MaxDrawdown).
  • +
  • Stops trading if more than 4 stoploss occur for all pairs within a 1 day (24 * 1h candles) limit (StoplossGuard).
  • +
  • Locks all pairs that had 2 Trades within the last 6 hours (6 * 1h candles) with a combined profit ratio of below 0.02 (<2%) (LowProfitPairs).
  • +
  • Locks all pairs for 2 candles that had a profit of below 0.01 (<1%) within the last 24h (24 * 1h candles), a minimum of 4 trades.
  • +
+
from freqtrade.strategy import IStrategy
+
+class AwesomeStrategy(IStrategy)
+    timeframe = '1h'
+
+    @property
+    def protections(self):
+        return [
+            {
+                "method": "CooldownPeriod",
+                "stop_duration_candles": 5
+            },
+            {
+                "method": "MaxDrawdown",
+                "lookback_period_candles": 48,
+                "trade_limit": 20,
+                "stop_duration_candles": 4,
+                "max_allowed_drawdown": 0.2
+            },
+            {
+                "method": "StoplossGuard",
+                "lookback_period_candles": 24,
+                "trade_limit": 4,
+                "stop_duration_candles": 2,
+                "only_per_pair": False
+            },
+            {
+                "method": "LowProfitPairs",
+                "lookback_period_candles": 6,
+                "trade_limit": 2,
+                "stop_duration_candles": 60,
+                "required_profit": 0.02
+            },
+            {
+                "method": "LowProfitPairs",
+                "lookback_period_candles": 24,
+                "trade_limit": 4,
+                "stop_duration_candles": 2,
+                "required_profit": 0.01
+            }
+        ]
+    # ...
+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/index.html b/en/2023.4/index.html new file mode 100644 index 000000000..68a06358b --- /dev/null +++ b/en/2023.4/index.html @@ -0,0 +1,1490 @@ + + + + + + + + + + + + + + + + + + + + + + Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Home

+ +

freqtrade

+

Freqtrade CI +DOI +Coverage Status +Maintainability

+ +

Star +Fork +Download

+

Introduction

+

Freqtrade is a free and open source crypto trading bot written in Python. It is designed to support all major exchanges and be controlled via Telegram or webUI. It contains backtesting, plotting and money management tools as well as strategy optimization by machine learning.

+
+

DISCLAIMER

+

This software is for educational purposes only. Do not risk money which you are afraid to lose. USE THE SOFTWARE AT YOUR OWN RISK. THE AUTHORS AND ALL AFFILIATES ASSUME NO RESPONSIBILITY FOR YOUR TRADING RESULTS.

+

Always start by running a trading bot in Dry-run and do not engage money before you understand how it works and what profit/loss you should expect.

+

We strongly recommend you to have basic coding skills and Python knowledge. Do not hesitate to read the source code and understand the mechanisms of this bot, algorithms and techniques implemented in it.

+
+

freqtrade screenshot

+

Features

+
    +
  • Develop your Strategy: Write your strategy in python, using pandas. Example strategies to inspire you are available in the strategy repository.
  • +
  • Download market data: Download historical data of the exchange and the markets your may want to trade with.
  • +
  • Backtest: Test your strategy on downloaded historical data.
  • +
  • Optimize: Find the best parameters for your strategy using hyperoptimization which employs machining learning methods. You can optimize buy, sell, take profit (ROI), stop-loss and trailing stop-loss parameters for your strategy.
  • +
  • Select markets: Create your static list or use an automatic one based on top traded volumes and/or prices (not available during backtesting). You can also explicitly blacklist markets you don't want to trade.
  • +
  • Run: Test your strategy with simulated money (Dry-Run mode) or deploy it with real money (Live-Trade mode).
  • +
  • Run using Edge (optional module): The concept is to find the best historical trade expectancy by markets based on variation of the stop-loss and then allow/reject markets to trade. The sizing of the trade is based on a risk of a percentage of your capital.
  • +
  • Control/Monitor: Use Telegram or a WebUI (start/stop the bot, show profit/loss, daily summary, current open trades results, etc.).
  • +
  • Analyze: Further analysis can be performed on either Backtesting data or Freqtrade trading history (SQL database), including automated standard plots, and methods to load the data into interactive environments.
  • +
+

Supported exchange marketplaces

+

Please read the exchange specific notes to learn about eventual, special configurations needed for each exchange.

+ +

Supported Futures Exchanges (experimental)

+ +

Please make sure to read the exchange specific notes, as well as the trading with leverage documentation before diving in.

+

Community tested

+

Exchanges confirmed working by the community:

+ +

Requirements

+

Hardware requirements

+

To run this bot we recommend you a linux cloud instance with a minimum of:

+
    +
  • 2GB RAM
  • +
  • 1GB disk space
  • +
  • 2vCPU
  • +
+

Software requirements

+
    +
  • Docker (Recommended)
  • +
+

Alternatively

+
    +
  • Python 3.8+
  • +
  • pip (pip3)
  • +
  • git
  • +
  • TA-Lib
  • +
  • virtualenv (Recommended)
  • +
+

Support

+

Help / Discord

+

For any questions not covered by the documentation or for further information about the bot, or to simply engage with like-minded individuals, we encourage you to join the Freqtrade discord server.

+

Ready to try?

+

Begin by reading the installation guide for docker (recommended), or for installation without docker.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/installation/index.html b/en/2023.4/installation/index.html new file mode 100644 index 000000000..afbe93ecd --- /dev/null +++ b/en/2023.4/installation/index.html @@ -0,0 +1,2187 @@ + + + + + + + + + + + + + + + + + + + + + + + + Linux/MacOS/Raspberry - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+ +
+
+ + + +
+
+ + + + + + + +

Installation

+

This page explains how to prepare your environment for running the bot.

+

The freqtrade documentation describes various ways to install freqtrade

+ +

Please consider using the prebuilt docker images to get started quickly while evaluating how freqtrade works.

+
+

Information

+

For Windows installation, please use the windows installation guide.

+

The easiest way to install and run Freqtrade is to clone the bot Github repository and then run the ./setup.sh script, if it's available for your platform.

+
+

Version considerations

+

When cloning the repository the default working branch has the name develop. This branch contains all last features (can be considered as relatively stable, thanks to automated tests). +The stable branch contains the code of the last release (done usually once per month on an approximately one week old snapshot of the develop branch to prevent packaging bugs, so potentially it's more stable).

+
+
+

Note

+

Python3.8 or higher and the corresponding pip are assumed to be available. The install-script will warn you and stop if that's not the case. git is also needed to clone the Freqtrade repository.
+Also, python headers (python<yourversion>-dev / python<yourversion>-devel) must be available for the installation to complete successfully.

+
+
+

Up-to-date clock

+

The clock on the system running the bot must be accurate, synchronized to a NTP server frequently enough to avoid problems with communication to the exchanges.

+
+
+

Running setup.py install for gym did not run successfully.

+

If you get an error related with gym we suggest you to downgrade setuptools it to version 65.5.0 you can do it with the following command: +

pip install setuptools==65.5.0
+

+
+
+

Requirements

+

These requirements apply to both Script Installation and Manual Installation.

+
+

ARM64 systems

+

If you are running an ARM64 system (like a MacOS M1 or an Oracle VM), please use docker to run freqtrade. +While native installation is possible with some manual effort, this is not supported at the moment.

+
+

Install guide

+ +

Install code

+

We've included/collected install instructions for Ubuntu, MacOS, and Windows. These are guidelines and your success may vary with other distros. +OS Specific steps are listed first, the Common section below is necessary for all systems.

+
+

Note

+

Python3.8 or higher and the corresponding pip are assumed to be available.

+
+
+
+
+

Install necessary dependencies

+
# update repository
+sudo apt-get update
+
+# install packages
+sudo apt install -y python3-pip python3-venv python3-dev python3-pandas git curl
+
+
+
+

The following assumes the latest Raspbian Buster lite image. +This image comes with python3.9 preinstalled, making it easy to get freqtrade up and running.

+

Tested using a Raspberry Pi 3 with the Raspbian Buster lite image, all updates applied.

+
sudo apt-get install python3-venv libatlas-base-dev cmake curl
+# Use pywheels.org to speed up installation
+sudo echo "[global]\nextra-index-url=https://www.piwheels.org/simple" > tee /etc/pip.conf
+
+git clone https://github.com/freqtrade/freqtrade.git
+cd freqtrade
+
+bash setup.sh -i
+
+
+

Installation duration

+

Depending on your internet speed and the Raspberry Pi version, installation can take multiple hours to complete. +Due to this, we recommend to use the pre-build docker-image for Raspberry, by following the Docker quickstart documentation

+
+
+

Note

+

The above does not install hyperopt dependencies. To install these, please use python3 -m pip install -e .[hyperopt]. +We do not advise to run hyperopt on a Raspberry Pi, since this is a very resource-heavy operation, which should be done on powerful machine.

+
+
+
+
+
+

Freqtrade repository

+

Freqtrade is an open source crypto-currency trading bot, whose code is hosted on github.com

+
# Download `develop` branch of freqtrade repository
+git clone https://github.com/freqtrade/freqtrade.git
+
+# Enter downloaded directory
+cd freqtrade
+
+# your choice (1): novice user
+git checkout stable
+
+# your choice (2): advanced user
+git checkout develop
+
+

(1) This command switches the cloned repository to the use of the stable branch. It's not needed, if you wish to stay on the (2) develop branch.

+

You may later switch between branches at any time with the git checkout stable/git checkout develop commands.

+
+Install from pypi +

An alternative way to install Freqtrade is from pypi. The downside is that this method requires ta-lib to be correctly installed beforehand, and is therefore currently not the recommended way to install Freqtrade.

+
pip install freqtrade
+
+
+
+

Script Installation

+

First of the ways to install Freqtrade, is to use provided the Linux/MacOS ./setup.sh script, which install all dependencies and help you configure the bot.

+

Make sure you fulfill the Requirements and have downloaded the Freqtrade repository.

+

Use /setup.sh -install (Linux/MacOS)

+

If you are on Debian, Ubuntu or MacOS, freqtrade provides the script to install freqtrade.

+
# --install, Install freqtrade from scratch
+./setup.sh -i
+
+

Activate your virtual environment

+

Each time you open a new terminal, you must run source .env/bin/activate to activate your virtual environment.

+
# then activate your .env
+source ./.env/bin/activate
+
+

Congratulations

+

You are ready, and run the bot

+

Other options of /setup.sh script

+

You can as well update, configure and reset the codebase of your bot with ./script.sh

+
# --update, Command git pull to update.
+./setup.sh -u
+# --reset, Hard reset your develop/stable branch.
+./setup.sh -r
+
+
** --install **
+
+With this option, the script will install the bot and most dependencies:
+You will need to have git and python3.8+ installed beforehand for this to work.
+
+* Mandatory software as: `ta-lib`
+* Setup your virtualenv under `.env/`
+
+This option is a combination of installation tasks and `--reset`
+
+** --update **
+
+This option will pull the last version of your current branch and update your virtualenv. Run the script with this option periodically to update your bot.
+
+** --reset **
+
+This option will hard reset your branch (only if you are on either `stable` or `develop`) and recreate your virtualenv.
+
+
+

Manual Installation

+

Make sure you fulfill the Requirements and have downloaded the Freqtrade repository.

+

Install TA-Lib

+

TA-Lib script installation

+
sudo ./build_helpers/install_ta-lib.sh
+
+
+

Note

+

This will use the ta-lib tar.gz included in this repository.

+
+
TA-Lib manual installation
+

Official installation guide

+
wget http://prdownloads.sourceforge.net/ta-lib/ta-lib-0.4.0-src.tar.gz
+tar xvzf ta-lib-0.4.0-src.tar.gz
+cd ta-lib
+sed -i.bak "s|0.00000001|0.000000000000000001 |g" src/ta_func/ta_utility.h
+./configure --prefix=/usr/local
+make
+sudo make install
+# On debian based systems (debian, ubuntu, ...) - updating ldconfig might be necessary.
+sudo ldconfig  
+cd ..
+rm -rf ./ta-lib*
+
+

Setup Python virtual environment (virtualenv)

+

You will run freqtrade in separated virtual environment

+
# create virtualenv in directory /freqtrade/.env
+python3 -m venv .env
+
+# run virtualenv
+source .env/bin/activate
+
+

Install python dependencies

+
python3 -m pip install --upgrade pip
+python3 -m pip install -e .
+
+

Congratulations

+

You are ready, and run the bot

+

(Optional) Post-installation Tasks

+
+

Note

+

If you run the bot on a server, you should consider using Docker or a terminal multiplexer like screen or tmux to avoid that the bot is stopped on logout.

+
+

On Linux with software suite systemd, as an optional post-installation task, you may wish to setup the bot to run as a systemd service or configure it to send the log messages to the syslog/rsyslog or journald daemons. See Advanced Logging for details.

+
+

Installation with Conda

+

Freqtrade can also be installed with Miniconda or Anaconda. We recommend using Miniconda as it's installation footprint is smaller. Conda will automatically prepare and manage the extensive library-dependencies of the Freqtrade program.

+

What is Conda?

+

Conda is a package, dependency and environment manager for multiple programming languages: conda docs

+

Installation with conda

+

Install Conda

+

Installing on linux

+

Installing on windows

+

Answer all questions. After installation, it is mandatory to turn your terminal OFF and ON again.

+

Freqtrade download

+

Download and install freqtrade.

+
# download freqtrade
+git clone https://github.com/freqtrade/freqtrade.git
+
+# enter downloaded directory 'freqtrade'
+cd freqtrade      
+
+

Freqtrade install: Conda Environment

+
conda create --name freqtrade python=3.10
+
+
+

Creating Conda Environment

+

The conda command create -n automatically installs all nested dependencies for the selected libraries, general structure of installation command is:

+
# choose your own packages
+conda env create -n [name of the environment] [python version] [packages]
+
+
+

Enter/exit freqtrade environment

+

To check available environments, type

+
conda env list
+
+

Enter installed environment

+
# enter conda environment
+conda activate freqtrade
+
+# exit conda environment - don't do it now
+conda deactivate
+
+

Install last python dependencies with pip

+
python3 -m pip install --upgrade pip
+python3 -m pip install -r requirements.txt
+python3 -m pip install -e .
+
+

Patch conda libta-lib (Linux only)

+
# Ensure that the environment is active!
+conda activate freqtrade
+
+cd build_helpers
+bash install_ta-lib.sh ${CONDA_PREFIX} nosudo
+
+

Congratulations

+

You are ready, and run the bot

+

Important shortcuts

+
# list installed conda environments
+conda env list
+
+# activate base environment
+conda activate
+
+# activate freqtrade environment
+conda activate freqtrade
+
+#deactivate any conda environments
+conda deactivate                              
+
+

Further info on anaconda

+
+

New heavy packages

+

It may happen that creating a new Conda environment, populated with selected packages at the moment of creation takes less time than installing a large, heavy library or application, into previously set environment.

+
+
+

pip install within conda

+

The documentation of conda says that pip should NOT be used within conda, because internal problems can occur. +However, they are rare. Anaconda Blogpost

+

Nevertheless, that is why, the conda-forge channel is preferred:

+
    +
  • more libraries are available (less need for pip)
  • +
  • conda-forge works better with pip
  • +
  • the libraries are newer
  • +
+
+

Happy trading!

+
+

You are ready

+

You've made it this far, so you have successfully installed freqtrade.

+

Initialize the configuration

+
# Step 1 - Initialize user folder
+freqtrade create-userdir --userdir user_data
+
+# Step 2 - Create a new configuration file
+freqtrade new-config --config config.json
+
+

You are ready to run, read Bot Configuration, remember to start with dry_run: True and verify that everything is working.

+

To learn how to setup your configuration, please refer to the Bot Configuration documentation page.

+

Start the Bot

+
freqtrade trade --config config.json --strategy SampleStrategy
+
+
+

Warning

+

You should read through the rest of the documentation, backtest the strategy you're going to use, and use dry-run before enabling trading with real money.

+
+
+

Troubleshooting

+

Common problem: "command not found"

+

If you used (1)Script or (2)Manual installation, you need to run the bot in virtual environment. If you get error as below, make sure venv is active.

+
# if:
+bash: freqtrade: command not found
+
+# then activate your .env
+source ./.env/bin/activate
+
+

MacOS installation error

+

Newer versions of MacOS may have installation failed with errors like error: command 'g++' failed with exit status 1.

+

This error will require explicit installation of the SDK Headers, which are not installed by default in this version of MacOS. +For MacOS 10.14, this can be accomplished with the below command.

+
open /Library/Developer/CommandLineTools/Packages/macOS_SDK_headers_for_macOS_10.14.pkg
+
+

If this file is inexistent, then you're probably on a different version of MacOS, so you may need to consult the internet for specific resolution details.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/javascripts/config.js b/en/2023.4/javascripts/config.js new file mode 100644 index 000000000..80e81ba59 --- /dev/null +++ b/en/2023.4/javascripts/config.js @@ -0,0 +1,12 @@ +window.MathJax = { + tex: { + inlineMath: [["\\(", "\\)"]], + displayMath: [["\\[", "\\]"]], + processEscapes: true, + processEnvironments: true + }, + options: { + ignoreHtmlClass: ".*|", + processHtmlClass: "arithmatex" + } +}; diff --git a/en/2023.4/leverage/index.html b/en/2023.4/leverage/index.html new file mode 100644 index 000000000..472799c25 --- /dev/null +++ b/en/2023.4/leverage/index.html @@ -0,0 +1,1634 @@ + + + + + + + + + + + + + + + + + + + + + + + + Short / Leverage - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Trading with Leverage

+
+

Beta feature

+

This feature is still in it's testing phase. Should you notice something you think is wrong please let us know via Discord or via Github Issue.

+
+
+

Multiple bots on one account

+

You can't run 2 bots on the same account with leverage. For leveraged / margin trading, freqtrade assumes it's the only user of the account, and all liquidation levels are calculated based on this assumption.

+
+
+

Trading with leverage is very risky

+

Do not trade with a leverage > 1 using a strategy that hasn't shown positive results in a live run using the spot market. Check the stoploss of your strategy. With a leverage of 2, a stoploss of 0.5 (50%) would be too low, and these trades would be liquidated before reaching that stoploss. +We do not assume any responsibility for eventual losses that occur from using this software or this mode.

+

Please only use advanced trading modes when you know how freqtrade (and your strategy) works. +Also, never risk more than what you can afford to lose.

+
+

If you already have an existing strategy, please read the strategy migration guide to migrate your strategy from a freqtrade v2 strategy, to strategy of version 3 which can short and trade futures.

+

Shorting

+

Shorting is not possible when trading with trading_mode set to spot. To short trade, trading_mode must be set to margin(currently unavailable) or futures, with margin_mode set to cross(currently unavailable) or isolated

+

For a strategy to short, the strategy class must set the class variable can_short = True

+

Please read strategy customization for instructions on how to set signals to enter and exit short trades.

+

Understand trading_mode

+

The possible values are: spot (default), margin(Currently unavailable) or futures.

+

Spot

+

Regular trading mode (low risk)

+
    +
  • Long trades only (No short trades).
  • +
  • No leverage.
  • +
  • No Liquidation.
  • +
  • Profits gained/lost are equal to the change in value of the assets (minus trading fees).
  • +
+

Leverage trading modes

+

With leverage, a trader borrows capital from the exchange. The capital must be re-payed fully to the exchange (potentially with interest), and the trader keeps any profits, or pays any losses, from any trades made using the borrowed capital.

+

Because the capital must always be re-payed, exchanges will liquidate (forcefully sell the traders assets) a trade made using borrowed capital when the total value of assets in the leverage account drops to a certain point (a point where the total value of losses is less than the value of the collateral that the trader actually owns in the leverage account), in order to ensure that the trader has enough capital to pay the borrowed assets back to the exchange. The exchange will also charge a liquidation fee, adding to the traders losses.

+

For this reason, DO NOT TRADE WITH LEVERAGE IF YOU DON'T KNOW EXACTLY WHAT YOUR DOING. LEVERAGE TRADING IS HIGH RISK, AND CAN RESULT IN THE VALUE OF YOUR ASSETS DROPPING TO 0 VERY QUICKLY, WITH NO CHANCE OF INCREASING IN VALUE AGAIN.

+

Margin (currently unavailable)

+

Trading occurs on the spot market, but the exchange lends currency to you in an amount equal to the chosen leverage. You pay the amount lent to you back to the exchange with interest, and your profits/losses are multiplied by the leverage specified.

+

Futures

+

Perpetual swaps (also known as Perpetual Futures) are contracts traded at a price that is closely tied to the underlying asset they are based off of (ex.). You are not trading the actual asset but instead are trading a derivative contract. Perpetual swap contracts can last indefinitely, in contrast to futures or option contracts.

+

In addition to the gains/losses from the change in price of the futures contract, traders also exchange funding fees, which are gains/losses worth an amount that is derived from the difference in price between the futures contract and the underlying asset. The difference in price between a futures contract and the underlying asset varies between exchanges.

+

To trade in futures markets, you'll have to set trading_mode to "futures". +You will also have to pick a "margin mode" (explanation below) - with freqtrade currently only supporting isolated margin.

+
"trading_mode": "futures",
+"margin_mode": "isolated"
+
+
Pair namings
+

Freqtrade follows the ccxt naming conventions for futures. +A futures pair will therefore have the naming of base/quote:settle (e.g. ETH/USDT:USDT).

+

Margin mode

+

On top of trading_mode - you will also have to configure your margin_mode. +While freqtrade currently only supports one margin mode, this will change, and by configuring it now you're all set for future updates.

+

The possible values are: isolated, or cross(currently unavailable).

+

Isolated margin mode

+

Each market(trading pair), keeps collateral in a separate account

+
"margin_mode": "isolated"
+
+

Cross margin mode (currently unavailable)

+

One account is used to share collateral between markets (trading pairs). Margin is taken from total account balance to avoid liquidation when needed.

+
"margin_mode": "cross"
+
+

Please read the exchange specific notes for exchanges that support this mode and how they differ.

+

Set leverage to use

+

Different strategies and risk profiles will require different levels of leverage. +While you could configure one static leverage value - freqtrade offers you the flexibility to adjust this via strategy leverage callback - which allows you to use different leverages by pair, or based on some other factor benefitting your strategy result.

+

If not implemented, leverage defaults to 1x (no leverage).

+
+

Warning

+

Higher leverage also equals higher risk - be sure you fully understand the implications of using leverage!

+
+

Understand liquidation_buffer

+

Defaults to 0.05

+

A ratio specifying how large of a safety net to place between the liquidation price and the stoploss to prevent a position from reaching the liquidation price. +This artificial liquidation price is calculated as:

+

freqtrade_liquidation_price = liquidation_price ± (abs(open_rate - liquidation_price) * liquidation_buffer)

+
    +
  • ± = + for long trades
  • +
  • ± = - for short trades
  • +
+

Possible values are any floats between 0.0 and 0.99

+

ex: If a trade is entered at a price of 10 coin/USDT, and the liquidation price of this trade is 8 coin/USDT, then with liquidation_buffer set to 0.05 the minimum stoploss for this trade would be \(8 + ((10 - 8) * 0.05) = 8 + 0.1 = 8.1\)

+
+

A liquidation_buffer of 0.0, or a low liquidation_buffer is likely to result in liquidations, and liquidation fees

+

Currently Freqtrade is able to calculate liquidation prices, but does not calculate liquidation fees. Setting your liquidation_buffer to 0.0, or using a low liquidation_buffer could result in your positions being liquidated. Freqtrade does not track liquidation fees, so liquidations will result in inaccurate profit/loss results for your bot. If you use a low liquidation_buffer, it is recommended to use stoploss_on_exchange if your exchange supports this.

+
+

Unavailable funding rates

+

For futures data, exchanges commonly provide the futures candles, the marks, and the funding rates. However, it is common that whilst candles and marks might be available, the funding rates are not. This can affect backtesting timeranges, i.e. you may only be able to test recent timeranges and not earlier, experiencing the No data found. Terminating. error. To get around this, add the futures_funding_rate config option as listed in configuration.md, and it is recommended that you set this to 0, unless you know a given specific funding rate for your pair, exchange and timerange. Setting this to anything other than 0 can have drastic effects on your profit calculations within strategy, e.g. within the custom_exit, custom_stoploss, etc functions.

+
+

This will mean your backtests are inaccurate.

+

This will not overwrite funding rates that are available from the exchange, but bear in mind that setting a false funding rate will mean backtesting results will be inaccurate for historical timeranges where funding rates are not available.

+
+

Developer

+

Margin mode

+

For shorts, the currency which pays the interest fee for the borrowed currency is purchased at the same time of the closing trade (This means that the amount purchased in short closing trades is greater than the amount sold in short opening trades).

+

For longs, the currency which pays the interest fee for the borrowed will already be owned by the user and does not need to be purchased. The interest is subtracted from the close_value of the trade.

+

All Fees are included in current_profit calculations during the trade.

+

Futures mode

+

Funding fees are either added or subtracted from the total amount of a trade

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/overrides/main.html b/en/2023.4/overrides/main.html new file mode 100644 index 000000000..cba627ead --- /dev/null +++ b/en/2023.4/overrides/main.html @@ -0,0 +1,44 @@ +{% extends "base.html" %} + + + +{% block site_nav %} + + + {% if nav %} + {% if page and page.meta and page.meta.hide %} + {% set hidden = "hidden" if "navigation" in page.meta.hide %} + {% endif %} + + {% endif %} + + + {% if page.toc and not "toc.integrate" in features %} + {% if page and page.meta and page.meta.hide %} + {% set hidden = "hidden" if "toc" in page.meta.hide %} + {% endif %} + + {% endif %} +{% endblock %} + +{% block footer %} + {{ super() }} + + + + + +{% endblock %} diff --git a/en/2023.4/plotting/index.html b/en/2023.4/plotting/index.html new file mode 100644 index 000000000..f3cb4e1ef --- /dev/null +++ b/en/2023.4/plotting/index.html @@ -0,0 +1,1666 @@ + + + + + + + + + + + + + + + + + + + + + + + + Plotting - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Plotting

+

This page explains how to plot prices, indicators and profits.

+

Installation / Setup

+

Plotting modules use the Plotly library. You can install / upgrade this by running the following command:

+
pip install -U -r requirements-plot.txt
+
+

Plot price and indicators

+

The freqtrade plot-dataframe subcommand shows an interactive graph with three subplots:

+
    +
  • Main plot with candlesticks and indicators following price (sma/ema)
  • +
  • Volume bars
  • +
  • Additional indicators as specified by --indicators2
  • +
+

plot-dataframe

+

Possible arguments:

+
usage: freqtrade plot-dataframe [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                                [-d PATH] [--userdir PATH] [-s NAME]
+                                [--strategy-path PATH] [-p PAIRS [PAIRS ...]]
+                                [--indicators1 INDICATORS1 [INDICATORS1 ...]]
+                                [--indicators2 INDICATORS2 [INDICATORS2 ...]]
+                                [--plot-limit INT] [--db-url PATH]
+                                [--trade-source {DB,file}] [--export EXPORT]
+                                [--export-filename PATH]
+                                [--timerange TIMERANGE] [-i TIMEFRAME]
+                                [--no-trades]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]
+                        Limit command to these pairs. Pairs are space-
+                        separated.
+  --indicators1 INDICATORS1 [INDICATORS1 ...]
+                        Set indicators from your strategy you want in the
+                        first row of the graph. Space-separated list. Example:
+                        `ema3 ema5`. Default: `['sma', 'ema3', 'ema5']`.
+  --indicators2 INDICATORS2 [INDICATORS2 ...]
+                        Set indicators from your strategy you want in the
+                        third row of the graph. Space-separated list. Example:
+                        `fastd fastk`. Default: `['macd', 'macdsignal']`.
+  --plot-limit INT      Specify tick limit for plotting. Notice: too high
+                        values cause huge files. Default: 750.
+  --db-url PATH         Override trades database URL, this is useful in custom
+                        deployments (default: `sqlite:///tradesv3.sqlite` for
+                        Live Run mode, `sqlite:///tradesv3.dryrun.sqlite` for
+                        Dry Run).
+  --trade-source {DB,file}
+                        Specify the source for trades (Can be DB or file
+                        (backtest file)) Default: file
+  --export EXPORT       Export backtest results, argument are: trades.
+                        Example: `--export=trades`
+  --export-filename PATH
+                        Save backtest results to the file with this filename.
+                        Requires `--export` to be set as well. Example:
+                        `--export-filename=user_data/backtest_results/backtest
+                        _today.json`
+  --timerange TIMERANGE
+                        Specify what timerange of data to use.
+  -i TIMEFRAME, --timeframe TIMEFRAME
+                        Specify timeframe (`1m`, `5m`, `30m`, `1h`, `1d`).
+  --no-trades           Skip using trades from backtesting file and DB.
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+Strategy arguments:
+  -s NAME, --strategy NAME
+                        Specify strategy class name which will be used by the
+                        bot.
+  --strategy-path PATH  Specify additional strategy lookup path.
+
+

Example:

+
freqtrade plot-dataframe -p BTC/ETH --strategy AwesomeStrategy
+
+

The -p/--pairs argument can be used to specify pairs you would like to plot.

+
+

Note

+

The freqtrade plot-dataframe subcommand generates one plot-file per pair.

+
+

Specify custom indicators. +Use --indicators1 for the main plot and --indicators2 for the subplot below (if values are in a different range than prices).

+
freqtrade plot-dataframe --strategy AwesomeStrategy -p BTC/ETH --indicators1 sma ema --indicators2 macd
+
+

Further usage examples

+

To plot multiple pairs, separate them with a space:

+
freqtrade plot-dataframe --strategy AwesomeStrategy -p BTC/ETH XRP/ETH
+
+

To plot a timerange (to zoom in)

+
freqtrade plot-dataframe --strategy AwesomeStrategy -p BTC/ETH --timerange=20180801-20180805
+
+

To plot trades stored in a database use --db-url in combination with --trade-source DB:

+
freqtrade plot-dataframe --strategy AwesomeStrategy --db-url sqlite:///tradesv3.dry_run.sqlite -p BTC/ETH --trade-source DB
+
+

To plot trades from a backtesting result, use --export-filename <filename>

+
freqtrade plot-dataframe --strategy AwesomeStrategy --export-filename user_data/backtest_results/backtest-result.json -p BTC/ETH
+
+

Plot dataframe basics

+

plot-dataframe2

+

The plot-dataframe subcommand requires backtesting data, a strategy and either a backtesting-results file or a database, containing trades corresponding to the strategy.

+

The resulting plot will have the following elements:

+
    +
  • Green triangles: Buy signals from the strategy. (Note: not every buy signal generates a trade, compare to cyan circles.)
  • +
  • Red triangles: Sell signals from the strategy. (Also, not every sell signal terminates a trade, compare to red and green squares.)
  • +
  • Cyan circles: Trade entry points.
  • +
  • Red squares: Trade exit points for trades with loss or 0% profit.
  • +
  • Green squares: Trade exit points for profitable trades.
  • +
  • Indicators with values corresponding to the candle scale (e.g. SMA/EMA), as specified with --indicators1.
  • +
  • Volume (bar chart at the bottom of the main chart).
  • +
  • Indicators with values in different scales (e.g. MACD, RSI) below the volume bars, as specified with --indicators2.
  • +
+
+

Bollinger Bands

+

Bollinger bands are automatically added to the plot if the columns bb_lowerband and bb_upperband exist, and are painted as a light blue area spanning from the lower band to the upper band.

+
+

Advanced plot configuration

+

An advanced plot configuration can be specified in the strategy in the plot_config parameter.

+

Additional features when using plot_config include:

+
    +
  • Specify colors per indicator
  • +
  • Specify additional subplots
  • +
  • Specify indicator pairs to fill area in between
  • +
+

The sample plot configuration below specifies fixed colors for the indicators. Otherwise, consecutive plots may produce different color schemes each time, making comparisons difficult. +It also allows multiple subplots to display both MACD and RSI at the same time.

+

Plot type can be configured using type key. Possible types are:

+
    +
  • scatter corresponding to plotly.graph_objects.Scatter class (default).
  • +
  • bar corresponding to plotly.graph_objects.Bar class.
  • +
+

Extra parameters to plotly.graph_objects.* constructor can be specified in plotly dict.

+

Sample configuration with inline comments explaining the process:

+
@property
+def plot_config(self):
+    """
+        There are a lot of solutions how to build the return dictionary.
+        The only important point is the return value.
+        Example:
+            plot_config = {'main_plot': {}, 'subplots': {}}
+
+    """
+    plot_config = {}
+    plot_config['main_plot'] = {
+        # Configuration for main plot indicators.
+        # Assumes 2 parameters, emashort and emalong to be specified.
+        f'ema_{self.emashort.value}': {'color': 'red'},
+        f'ema_{self.emalong.value}': {'color': '#CCCCCC'},
+        # By omitting color, a random color is selected.
+        'sar': {},
+        # fill area between senkou_a and senkou_b
+        'senkou_a': {
+            'color': 'green', #optional
+            'fill_to': 'senkou_b',
+            'fill_label': 'Ichimoku Cloud', #optional
+            'fill_color': 'rgba(255,76,46,0.2)', #optional
+        },
+        # plot senkou_b, too. Not only the area to it.
+        'senkou_b': {}
+    }
+    plot_config['subplots'] = {
+         # Create subplot MACD
+        "MACD": {
+            'macd': {'color': 'blue', 'fill_to': 'macdhist'},
+            'macdsignal': {'color': 'orange'},
+            'macdhist': {'type': 'bar', 'plotly': {'opacity': 0.9}}
+        },
+        # Additional subplot RSI
+        "RSI": {
+            'rsi': {'color': 'red'}
+        }
+    }
+
+    return plot_config
+
+
+As attribute (former method) +

Assigning plot_config is also possible as Attribute (this used to be the default way). +This has the disadvantage that strategy parameters are not available, preventing certain configurations from working.

+
    plot_config = {
+        'main_plot': {
+            # Configuration for main plot indicators.
+            # Specifies `ema10` to be red, and `ema50` to be a shade of gray
+            'ema10': {'color': 'red'},
+            'ema50': {'color': '#CCCCCC'},
+            # By omitting color, a random color is selected.
+            'sar': {},
+        # fill area between senkou_a and senkou_b
+        'senkou_a': {
+            'color': 'green', #optional
+            'fill_to': 'senkou_b',
+            'fill_label': 'Ichimoku Cloud', #optional
+            'fill_color': 'rgba(255,76,46,0.2)', #optional
+        },
+        # plot senkou_b, too. Not only the area to it.
+        'senkou_b': {}
+        },
+        'subplots': {
+            # Create subplot MACD
+            "MACD": {
+                'macd': {'color': 'blue', 'fill_to': 'macdhist'},
+                'macdsignal': {'color': 'orange'},
+                'macdhist': {'type': 'bar', 'plotly': {'opacity': 0.9}}
+            },
+            # Additional subplot RSI
+            "RSI": {
+                'rsi': {'color': 'red'}
+            }
+        }
+    }
+
+
+
+

Note

+

The above configuration assumes that ema10, ema50, senkou_a, senkou_b, +macd, macdsignal, macdhist and rsi are columns in the DataFrame created by the strategy.

+
+
+

Warning

+

plotly arguments are only supported with plotly library and will not work with freq-ui.

+
+
+

Trade position adjustments

+

If position_adjustment_enable / adjust_trade_position() is used, the trade initial buy price is averaged over multiple orders and the trade start price will most likely appear outside the candle range.

+
+

Plot profit

+

plot-profit

+

The plot-profit subcommand shows an interactive graph with three plots:

+
    +
  • Average closing price for all pairs.
  • +
  • The summarized profit made by backtesting. +Note that this is not the real-world profit, but more of an estimate.
  • +
  • Profit for each individual pair.
  • +
  • Parallelism of trades.
  • +
  • Underwater (Periods of drawdown).
  • +
+

The first graph is good to get a grip of how the overall market progresses.

+

The second graph will show if your algorithm works or doesn't. +Perhaps you want an algorithm that steadily makes small profits, or one that acts less often, but makes big swings. +This graph will also highlight the start (and end) of the Max drawdown period.

+

The third graph can be useful to spot outliers, events in pairs that cause profit spikes.

+

The forth graph can help you analyze trade parallelism, showing how often max_open_trades have been maxed out.

+

Possible options for the freqtrade plot-profit subcommand:

+
usage: freqtrade plot-profit [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                             [-d PATH] [--userdir PATH] [-s NAME]
+                             [--strategy-path PATH] [-p PAIRS [PAIRS ...]]
+                             [--timerange TIMERANGE] [--export EXPORT]
+                             [--export-filename PATH] [--db-url PATH]
+                             [--trade-source {DB,file}] [-i TIMEFRAME]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]
+                        Limit command to these pairs. Pairs are space-
+                        separated.
+  --timerange TIMERANGE
+                        Specify what timerange of data to use.
+  --export EXPORT       Export backtest results, argument are: trades.
+                        Example: `--export=trades`
+  --export-filename PATH, --backtest-filename PATH
+                        Use backtest results from this filename.
+                        Requires `--export` to be set as well. Example:
+                        `--export-filename=user_data/backtest_results/backtest
+                        _today.json`
+  --db-url PATH         Override trades database URL, this is useful in custom
+                        deployments (default: `sqlite:///tradesv3.sqlite` for
+                        Live Run mode, `sqlite:///tradesv3.dryrun.sqlite` for
+                        Dry Run).
+  --trade-source {DB,file}
+                        Specify the source for trades (Can be DB or file
+                        (backtest file)) Default: file
+  -i TIMEFRAME, --timeframe TIMEFRAME
+                        Specify timeframe (`1m`, `5m`, `30m`, `1h`, `1d`).
+  --auto-open           Automatically open generated plot.
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+Strategy arguments:
+  -s NAME, --strategy NAME
+                        Specify strategy class name which will be used by the
+                        bot.
+  --strategy-path PATH  Specify additional strategy lookup path.
+
+

The -p/--pairs argument, can be used to limit the pairs that are considered for this calculation.

+

Examples:

+

Use custom backtest-export file

+
freqtrade plot-profit  -p LTC/BTC --export-filename user_data/backtest_results/backtest-result.json
+
+

Use custom database

+
freqtrade plot-profit  -p LTC/BTC --db-url sqlite:///tradesv3.sqlite --trade-source DB
+
+
freqtrade --datadir user_data/data/binance_save/ plot-profit -p LTC/BTC
+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/plugins/index.html b/en/2023.4/plugins/index.html new file mode 100644 index 000000000..0167b572f --- /dev/null +++ b/en/2023.4/plugins/index.html @@ -0,0 +1,2209 @@ + + + + + + + + + + + + + + + + + + + + + + + + Plugins - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Plugins

+

Pairlists and Pairlist Handlers

+

Pairlist Handlers define the list of pairs (pairlist) that the bot should trade. They are configured in the pairlists section of the configuration settings.

+

In your configuration, you can use Static Pairlist (defined by the StaticPairList Pairlist Handler) and Dynamic Pairlist (defined by the VolumePairList Pairlist Handler).

+

Additionally, AgeFilter, PrecisionFilter, PriceFilter, ShuffleFilter, SpreadFilter and VolatilityFilter act as Pairlist Filters, removing certain pairs and/or moving their positions in the pairlist.

+

If multiple Pairlist Handlers are used, they are chained and a combination of all Pairlist Handlers forms the resulting pairlist the bot uses for trading and backtesting. Pairlist Handlers are executed in the sequence they are configured. You should always configure either StaticPairList or VolumePairList as the starting Pairlist Handler.

+

Inactive markets are always removed from the resulting pairlist. Explicitly blacklisted pairs (those in the pair_blacklist configuration setting) are also always removed from the resulting pairlist.

+

Pair blacklist

+

The pair blacklist (configured via exchange.pair_blacklist in the configuration) disallows certain pairs from trading. +This can be as simple as excluding DOGE/BTC - which will remove exactly this pair.

+

The pair-blacklist does also support wildcards (in regex-style) - so BNB/.* will exclude ALL pairs that start with BNB. +You may also use something like .*DOWN/BTC or .*UP/BTC to exclude leveraged tokens (check Pair naming conventions for your exchange!)

+

Available Pairlist Handlers

+ +
+

Testing pairlists

+

Pairlist configurations can be quite tricky to get right. Best use the test-pairlist utility sub-command to test your configuration quickly.

+
+

Static Pair List

+

By default, the StaticPairList method is used, which uses a statically defined pair whitelist from the configuration. The pairlist also supports wildcards (in regex-style) - so .*/BTC will include all pairs with BTC as a stake.

+

It uses configuration from exchange.pair_whitelist and exchange.pair_blacklist.

+
"pairlists": [
+    {"method": "StaticPairList"}
+],
+
+

By default, only currently enabled pairs are allowed. +To skip pair validation against active markets, set "allow_inactive": true within the StaticPairList configuration. +This can be useful for backtesting expired pairs (like quarterly spot-markets). +This option must be configured along with exchange.skip_pair_validation in the exchange configuration.

+

When used in a "follow-up" position (e.g. after VolumePairlist), all pairs in 'pair_whitelist' will be added to the end of the pairlist.

+

Volume Pair List

+

VolumePairList employs sorting/filtering of pairs by their trading volume. It selects number_assets top pairs with sorting based on the sort_key (which can only be quoteVolume).

+

When used in the chain of Pairlist Handlers in a non-leading position (after StaticPairList and other Pairlist Filters), VolumePairList considers outputs of previous Pairlist Handlers, adding its sorting/selection of the pairs by the trading volume.

+

When used in the leading position of the chain of Pairlist Handlers, the pair_whitelist configuration setting is ignored. Instead, VolumePairList selects the top assets from all available markets with matching stake-currency on the exchange.

+

The refresh_period setting allows to define the period (in seconds), at which the pairlist will be refreshed. Defaults to 1800s (30 minutes). +The pairlist cache (refresh_period) on VolumePairList is only applicable to generating pairlists. +Filtering instances (not the first position in the list) will not apply any cache and will always use up-to-date data.

+

VolumePairList is per default based on the ticker data from exchange, as reported by the ccxt library:

+
    +
  • The quoteVolume is the amount of quote (stake) currency traded (bought or sold) in last 24 hours.
  • +
+
"pairlists": [
+    {
+        "method": "VolumePairList",
+        "number_assets": 20,
+        "sort_key": "quoteVolume",
+        "min_value": 0,
+        "refresh_period": 1800
+    }
+],
+
+

You can define a minimum volume with min_value - which will filter out pairs with a volume lower than the specified value in the specified timerange.

+
VolumePairList Advanced mode
+

VolumePairList can also operate in an advanced mode to build volume over a given timerange of specified candle size. It utilizes exchange historical candle data, builds a typical price (calculated by (open+high+low)/3) and multiplies the typical price with every candle's volume. The sum is the quoteVolume over the given range. This allows different scenarios, for a more smoothened volume, when using longer ranges with larger candle sizes, or the opposite when using a short range with small candles.

+

For convenience lookback_days can be specified, which will imply that 1d candles will be used for the lookback. In the example below the pairlist would be created based on the last 7 days:

+
"pairlists": [
+    {
+        "method": "VolumePairList",
+        "number_assets": 20,
+        "sort_key": "quoteVolume",
+        "min_value": 0,
+        "refresh_period": 86400,
+        "lookback_days": 7
+    }
+],
+
+
+

Range look back and refresh period

+

When used in conjunction with lookback_days and lookback_timeframe the refresh_period can not be smaller than the candle size in seconds. As this will result in unnecessary requests to the exchanges API.

+
+
+

Performance implications when using lookback range

+

If used in first position in combination with lookback, the computation of the range based volume can be time and resource consuming, as it downloads candles for all tradable pairs. Hence it's highly advised to use the standard approach with VolumeFilter to narrow the pairlist down for further range volume calculation.

+
+
+Unsupported exchanges (Bittrex, Gemini) +

On some exchanges (like Bittrex and Gemini), regular VolumePairList does not work as the api does not natively provide 24h volume. This can be worked around by using candle data to build the volume. +To roughly simulate 24h volume, you can use the following configuration. +Please note that These pairlists will only refresh once per day.

+
"pairlists": [
+    {
+        "method": "VolumePairList",
+        "number_assets": 20,
+        "sort_key": "quoteVolume",
+        "min_value": 0,
+        "refresh_period": 86400,
+        "lookback_days": 1
+    }
+],
+
+
+

More sophisticated approach can be used, by using lookback_timeframe for candle size and lookback_period which specifies the amount of candles. This example will build the volume pairs based on a rolling period of 3 days of 1h candles:

+
"pairlists": [
+    {
+        "method": "VolumePairList",
+        "number_assets": 20,
+        "sort_key": "quoteVolume",
+        "min_value": 0,
+        "refresh_period": 3600,
+        "lookback_timeframe": "1h",
+        "lookback_period": 72
+    }
+],
+
+
+

Note

+

VolumePairList does not support backtesting mode.

+
+

ProducerPairList

+

With ProducerPairList, you can reuse the pairlist from a Producer without explicitly defining the pairlist on each consumer.

+

Consumer mode is required for this pairlist to work.

+

The pairlist will perform a check on active pairs against the current exchange configuration to avoid attempting to trade on invalid markets.

+

You can limit the length of the pairlist with the optional parameter number_assets. Using "number_assets"=0 or omitting this key will result in the reuse of all producer pairs valid for the current setup.

+
"pairlists": [
+    {
+        "method": "ProducerPairList",
+        "number_assets": 5,
+        "producer_name": "default",
+    }
+],
+
+
+

Combining pairlists

+

This pairlist can be combined with all other pairlists and filters for further pairlist reduction, and can also act as an "additional" pairlist, on top of already defined pairs. +ProducerPairList can also be used multiple times in sequence, combining the pairs from multiple producers. +Obviously in complex such configurations, the Producer may not provide data for all pairs, so the strategy must be fit for this.

+
+

RemotePairList

+

It allows the user to fetch a pairlist from a remote server or a locally stored json file within the freqtrade directory, enabling dynamic updates and customization of the trading pairlist.

+

The RemotePairList is defined in the pairlists section of the configuration settings. It uses the following configuration options:

+
"pairlists": [
+    {
+        "method": "RemotePairList",
+        "pairlist_url": "https://example.com/pairlist",
+        "number_assets": 10,
+        "refresh_period": 1800,
+        "keep_pairlist_on_failure": true,
+        "read_timeout": 60,
+        "bearer_token": "my-bearer-token"
+    }
+]
+
+

The pairlist_url option specifies the URL of the remote server where the pairlist is located, or the path to a local file (if file:/// is prepended). This allows the user to use either a remote server or a local file as the source for the pairlist.

+

The user is responsible for providing a server or local file that returns a JSON object with the following structure:

+
{
+    "pairs": ["XRP/USDT", "ETH/USDT", "LTC/USDT"],
+    "refresh_period": 1800,
+}
+
+

The pairs property should contain a list of strings with the trading pairs to be used by the bot. The refresh_period property is optional and specifies the number of seconds that the pairlist should be cached before being refreshed.

+

The optional keep_pairlist_on_failure specifies whether the previous received pairlist should be used if the remote server is not reachable or returns an error. The default value is true.

+

The optional read_timeout specifies the maximum amount of time (in seconds) to wait for a response from the remote source, The default value is 60.

+

The optional bearer_token will be included in the requests Authorization Header.

+
+

Note

+

In case of a server error the last received pairlist will be kept if keep_pairlist_on_failure is set to true, when set to false a empty pairlist is returned.

+
+

AgeFilter

+

Removes pairs that have been listed on the exchange for less than min_days_listed days (defaults to 10) or more than max_days_listed days (defaults None mean infinity).

+

When pairs are first listed on an exchange they can suffer huge price drops and volatility +in the first few days while the pair goes through its price-discovery period. Bots can often +be caught out buying before the pair has finished dropping in price.

+

This filter allows freqtrade to ignore pairs until they have been listed for at least min_days_listed days and listed before max_days_listed.

+

OffsetFilter

+

Offsets an incoming pairlist by a given offset value.

+

As an example it can be used in conjunction with VolumeFilter to remove the top X volume pairs. Or to split a larger pairlist on two bot instances.

+

Example to remove the first 10 pairs from the pairlist, and takes the next 20 (taking items 10-30 of the initial list):

+
"pairlists": [
+    // ...
+    {
+        "method": "OffsetFilter",
+        "offset": 10,
+        "number_assets": 20
+    }
+],
+
+
+

Warning

+

When OffsetFilter is used to split a larger pairlist among multiple bots in combination with VolumeFilter +it can not be guaranteed that pairs won't overlap due to slightly different refresh intervals for the +VolumeFilter.

+
+
+

Note

+

An offset larger than the total length of the incoming pairlist will result in an empty pairlist.

+
+

PerformanceFilter

+

Sorts pairs by past trade performance, as follows:

+
    +
  1. Positive performance.
  2. +
  3. No closed trades yet.
  4. +
  5. Negative performance.
  6. +
+

Trade count is used as a tie breaker.

+

You can use the minutes parameter to only consider performance of the past X minutes (rolling window). +Not defining this parameter (or setting it to 0) will use all-time performance.

+

The optional min_profit (as ratio -> a setting of 0.01 corresponds to 1%) parameter defines the minimum profit a pair must have to be considered. +Pairs below this level will be filtered out. +Using this parameter without minutes is highly discouraged, as it can lead to an empty pairlist without a way to recover.

+
"pairlists": [
+    // ...
+    {
+        "method": "PerformanceFilter",
+        "minutes": 1440,  // rolling 24h
+        "min_profit": 0.01  // minimal profit 1%
+    }
+],
+
+

As this Filter uses past performance of the bot, it'll have some startup-period - and should only be used after the bot has a few 100 trades in the database.

+
+

Backtesting

+

PerformanceFilter does not support backtesting mode.

+
+

PrecisionFilter

+

Filters low-value coins which would not allow setting stoplosses.

+
+

Backtesting

+

PrecisionFilter does not support backtesting mode using multiple strategies.

+
+

PriceFilter

+

The PriceFilter allows filtering of pairs by price. Currently the following price filters are supported:

+
    +
  • min_price
  • +
  • max_price
  • +
  • max_value
  • +
  • low_price_ratio
  • +
+

The min_price setting removes pairs where the price is below the specified price. This is useful if you wish to avoid trading very low-priced pairs. +This option is disabled by default, and will only apply if set to > 0.

+

The max_price setting removes pairs where the price is above the specified price. This is useful if you wish to trade only low-priced pairs. +This option is disabled by default, and will only apply if set to > 0.

+

The max_value setting removes pairs where the minimum value change is above a specified value. +This is useful when an exchange has unbalanced limits. For example, if step-size = 1 (so you can only buy 1, or 2, or 3, but not 1.1 Coins) - and the price is pretty high (like 20$) as the coin has risen sharply since the last limit adaption. +As a result of the above, you can only buy for 20$, or 40$ - but not for 25$. +On exchanges that deduct fees from the receiving currency (e.g. binance) - this can result in high value coins / amounts that are unsellable as the amount is slightly below the limit.

+

The low_price_ratio setting removes pairs where a raise of 1 price unit (pip) is above the low_price_ratio ratio. +This option is disabled by default, and will only apply if set to > 0.

+

For PriceFilter at least one of its min_price, max_price or low_price_ratio settings must be applied.

+

Calculation example:

+

Min price precision for SHITCOIN/BTC is 8 decimals. If its price is 0.00000011 - one price step above would be 0.00000012, which is ~9% higher than the previous price value. You may filter out this pair by using PriceFilter with low_price_ratio set to 0.09 (9%) or with min_price set to 0.00000011, correspondingly.

+
+

Low priced pairs

+

Low priced pairs with high "1 pip movements" are dangerous since they are often illiquid and it may also be impossible to place the desired stoploss, which can often result in high losses since price needs to be rounded to the next tradable price - so instead of having a stoploss of -5%, you could end up with a stoploss of -9% simply due to price rounding.

+
+

ShuffleFilter

+

Shuffles (randomizes) pairs in the pairlist. It can be used for preventing the bot from trading some of the pairs more frequently then others when you want all pairs be treated with the same priority.

+

By default, ShuffleFilter will shuffle pairs once per candle. +To shuffle on every iteration, set "shuffle_frequency" to "iteration" instead of the default of "candle".

+
    {
+        "method": "ShuffleFilter", 
+        "shuffle_frequency": "candle",
+        "seed": 42
+    }
+
+
+

Tip

+

You may set the seed value for this Pairlist to obtain reproducible results, which can be useful for repeated backtesting sessions. If seed is not set, the pairs are shuffled in the non-repeatable random order. ShuffleFilter will automatically detect runmodes and apply the seed only for backtesting modes - if a seed value is set.

+
+

SpreadFilter

+

Removes pairs that have a difference between asks and bids above the specified ratio, max_spread_ratio (defaults to 0.005).

+

Example:

+

If DOGE/BTC maximum bid is 0.00000026 and minimum ask is 0.00000027, the ratio is calculated as: 1 - bid/ask ~= 0.037 which is > 0.005 and this pair will be filtered out.

+

RangeStabilityFilter

+

Removes pairs where the difference between lowest low and highest high over lookback_days days is below min_rate_of_change or above max_rate_of_change. Since this is a filter that requires additional data, the results are cached for refresh_period.

+

In the below example: +If the trading range over the last 10 days is <1% or >99%, remove the pair from the whitelist.

+
"pairlists": [
+    {
+        "method": "RangeStabilityFilter",
+        "lookback_days": 10,
+        "min_rate_of_change": 0.01,
+        "max_rate_of_change": 0.99,
+        "refresh_period": 1440
+    }
+]
+
+
+

Tip

+

This Filter can be used to automatically remove stable coin pairs, which have a very low trading range, and are therefore extremely difficult to trade with profit. +Additionally, it can also be used to automatically remove pairs with extreme high/low variance over a given amount of time.

+
+

VolatilityFilter

+

Volatility is the degree of historical variation of a pairs over time, it is measured by the standard deviation of logarithmic daily returns. Returns are assumed to be normally distributed, although actual distribution might be different. In a normal distribution, 68% of observations fall within one standard deviation and 95% of observations fall within two standard deviations. Assuming a volatility of 0.05 means that the expected returns for 20 out of 30 days is expected to be less than 5% (one standard deviation). Volatility is a positive ratio of the expected deviation of return and can be greater than 1.00. Please refer to the wikipedia definition of volatility.

+

This filter removes pairs if the average volatility over a lookback_days days is below min_volatility or above max_volatility. Since this is a filter that requires additional data, the results are cached for refresh_period.

+

This filter can be used to narrow down your pairs to a certain volatility or avoid very volatile pairs.

+

In the below example: +If the volatility over the last 10 days is not in the range of 0.05-0.50, remove the pair from the whitelist. The filter is applied every 24h.

+
"pairlists": [
+    {
+        "method": "VolatilityFilter",
+        "lookback_days": 10,
+        "min_volatility": 0.05,
+        "max_volatility": 0.50,
+        "refresh_period": 86400
+    }
+]
+
+

Full example of Pairlist Handlers

+

The below example blacklists BNB/BTC, uses VolumePairList with 20 assets, sorting pairs by quoteVolume and applies PrecisionFilter and PriceFilter, filtering all assets where 1 price unit is > 1%. Then the SpreadFilter and VolatilityFilter is applied and pairs are finally shuffled with the random seed set to some predefined value.

+
"exchange": {
+    "pair_whitelist": [],
+    "pair_blacklist": ["BNB/BTC"]
+},
+"pairlists": [
+    {
+        "method": "VolumePairList",
+        "number_assets": 20,
+        "sort_key": "quoteVolume"
+    },
+    {"method": "AgeFilter", "min_days_listed": 10},
+    {"method": "PrecisionFilter"},
+    {"method": "PriceFilter", "low_price_ratio": 0.01},
+    {"method": "SpreadFilter", "max_spread_ratio": 0.005},
+    {
+        "method": "RangeStabilityFilter",
+        "lookback_days": 10,
+        "min_rate_of_change": 0.01,
+        "refresh_period": 1440
+    },
+    {
+        "method": "VolatilityFilter",
+        "lookback_days": 10,
+        "min_volatility": 0.05,
+        "max_volatility": 0.50,
+        "refresh_period": 86400
+    },
+    {"method": "ShuffleFilter", "seed": 42}
+],
+
+

Protections

+
+

Beta feature

+

This feature is still in it's testing phase. Should you notice something you think is wrong please let us know via Discord or via Github Issue.

+
+

Protections will protect your strategy from unexpected events and market conditions by temporarily stop trading for either one pair, or for all pairs. +All protection end times are rounded up to the next candle to avoid sudden, unexpected intra-candle buys.

+
+

Note

+

Not all Protections will work for all strategies, and parameters will need to be tuned for your strategy to improve performance.

+
+
+

Tip

+

Each Protection can be configured multiple times with different parameters, to allow different levels of protection (short-term / long-term).

+
+
+

Backtesting

+

Protections are supported by backtesting and hyperopt, but must be explicitly enabled by using the --enable-protections flag.

+
+
+

Setting protections from the configuration

+

Setting protections from the configuration via "protections": [], key should be considered deprecated and will be removed in a future version. +It is also no longer guaranteed that your protections apply to the strategy in cases where the strategy defines protections as property.

+
+

Available Protections

+
    +
  • StoplossGuard Stop trading if a certain amount of stoploss occurred within a certain time window.
  • +
  • MaxDrawdown Stop trading if max-drawdown is reached.
  • +
  • LowProfitPairs Lock pairs with low profits
  • +
  • CooldownPeriod Don't enter a trade right after selling a trade.
  • +
+

Common settings to all Protections

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
ParameterDescription
methodProtection name to use.
Datatype: String, selected from available Protections
stop_duration_candlesFor how many candles should the lock be set?
Datatype: Positive integer (in candles)
stop_durationhow many minutes should protections be locked.
Cannot be used together with stop_duration_candles.
Datatype: Float (in minutes)
lookback_period_candlesOnly trades that completed within the last lookback_period_candles candles will be considered. This setting may be ignored by some Protections.
Datatype: Positive integer (in candles).
lookback_periodOnly trades that completed after current_time - lookback_period will be considered.
Cannot be used together with lookback_period_candles.
This setting may be ignored by some Protections.
Datatype: Float (in minutes)
trade_limitNumber of trades required at minimum (not used by all Protections).
Datatype: Positive integer
+
+

Durations

+

Durations (stop_duration* and lookback_period* can be defined in either minutes or candles). +For more flexibility when testing different timeframes, all below examples will use the "candle" definition.

+
+

Stoploss Guard

+

StoplossGuard selects all trades within lookback_period in minutes (or in candles when using lookback_period_candles). +If trade_limit or more trades resulted in stoploss, trading will stop for stop_duration in minutes (or in candles when using stop_duration_candles).

+

This applies across all pairs, unless only_per_pair is set to true, which will then only look at one pair at a time.

+

Similarly, this protection will by default look at all trades (long and short). For futures bots, setting only_per_side will make the bot only consider one side, and will then only lock this one side, allowing for example shorts to continue after a series of long stoplosses.

+

required_profit will determine the required relative profit (or loss) for stoplosses to consider. This should normally not be set and defaults to 0.0 - which means all losing stoplosses will be triggering a block.

+

The below example stops trading for all pairs for 4 candles after the last trade if the bot hit stoploss 4 times within the last 24 candles.

+
@property
+def protections(self):
+    return [
+        {
+            "method": "StoplossGuard",
+            "lookback_period_candles": 24,
+            "trade_limit": 4,
+            "stop_duration_candles": 4,
+            "required_profit": 0.0,
+            "only_per_pair": False,
+            "only_per_side": False
+        }
+    ]
+
+
+

Note

+

StoplossGuard considers all trades with the results "stop_loss", "stoploss_on_exchange" and "trailing_stop_loss" if the resulting profit was negative. +trade_limit and lookback_period will need to be tuned for your strategy.

+
+

MaxDrawdown

+

MaxDrawdown uses all trades within lookback_period in minutes (or in candles when using lookback_period_candles) to determine the maximum drawdown. If the drawdown is below max_allowed_drawdown, trading will stop for stop_duration in minutes (or in candles when using stop_duration_candles) after the last trade - assuming that the bot needs some time to let markets recover.

+

The below sample stops trading for 12 candles if max-drawdown is > 20% considering all pairs - with a minimum of trade_limit trades - within the last 48 candles. If desired, lookback_period and/or stop_duration can be used.

+
@property
+def protections(self):
+    return  [
+        {
+            "method": "MaxDrawdown",
+            "lookback_period_candles": 48,
+            "trade_limit": 20,
+            "stop_duration_candles": 12,
+            "max_allowed_drawdown": 0.2
+        },
+    ]
+
+

Low Profit Pairs

+

LowProfitPairs uses all trades for a pair within lookback_period in minutes (or in candles when using lookback_period_candles) to determine the overall profit ratio. +If that ratio is below required_profit, that pair will be locked for stop_duration in minutes (or in candles when using stop_duration_candles).

+

For futures bots, setting only_per_side will make the bot only consider one side, and will then only lock this one side, allowing for example shorts to continue after a series of long losses.

+

The below example will stop trading a pair for 60 minutes if the pair does not have a required profit of 2% (and a minimum of 2 trades) within the last 6 candles.

+
@property
+def protections(self):
+    return [
+        {
+            "method": "LowProfitPairs",
+            "lookback_period_candles": 6,
+            "trade_limit": 2,
+            "stop_duration": 60,
+            "required_profit": 0.02,
+            "only_per_pair": False,
+        }
+    ]
+
+

Cooldown Period

+

CooldownPeriod locks a pair for stop_duration in minutes (or in candles when using stop_duration_candles) after selling, avoiding a re-entry for this pair for stop_duration minutes.

+

The below example will stop trading a pair for 2 candles after closing a trade, allowing this pair to "cool down".

+
@property
+def protections(self):
+    return  [
+        {
+            "method": "CooldownPeriod",
+            "stop_duration_candles": 2
+        }
+    ]
+
+
+

Note

+

This Protection applies only at pair-level, and will never lock all pairs globally. +This Protection does not consider lookback_period as it only looks at the latest trade.

+
+

Full example of Protections

+

All protections can be combined at will, also with different parameters, creating a increasing wall for under-performing pairs. +All protections are evaluated in the sequence they are defined.

+

The below example assumes a timeframe of 1 hour:

+
    +
  • Locks each pair after selling for an additional 5 candles (CooldownPeriod), giving other pairs a chance to get filled.
  • +
  • Stops trading for 4 hours (4 * 1h candles) if the last 2 days (48 * 1h candles) had 20 trades, which caused a max-drawdown of more than 20%. (MaxDrawdown).
  • +
  • Stops trading if more than 4 stoploss occur for all pairs within a 1 day (24 * 1h candles) limit (StoplossGuard).
  • +
  • Locks all pairs that had 2 Trades within the last 6 hours (6 * 1h candles) with a combined profit ratio of below 0.02 (<2%) (LowProfitPairs).
  • +
  • Locks all pairs for 2 candles that had a profit of below 0.01 (<1%) within the last 24h (24 * 1h candles), a minimum of 4 trades.
  • +
+
from freqtrade.strategy import IStrategy
+
+class AwesomeStrategy(IStrategy)
+    timeframe = '1h'
+
+    @property
+    def protections(self):
+        return [
+            {
+                "method": "CooldownPeriod",
+                "stop_duration_candles": 5
+            },
+            {
+                "method": "MaxDrawdown",
+                "lookback_period_candles": 48,
+                "trade_limit": 20,
+                "stop_duration_candles": 4,
+                "max_allowed_drawdown": 0.2
+            },
+            {
+                "method": "StoplossGuard",
+                "lookback_period_candles": 24,
+                "trade_limit": 4,
+                "stop_duration_candles": 2,
+                "only_per_pair": False
+            },
+            {
+                "method": "LowProfitPairs",
+                "lookback_period_candles": 6,
+                "trade_limit": 2,
+                "stop_duration_candles": 60,
+                "required_profit": 0.02
+            },
+            {
+                "method": "LowProfitPairs",
+                "lookback_period_candles": 24,
+                "trade_limit": 4,
+                "stop_duration_candles": 2,
+                "required_profit": 0.01
+            }
+        ]
+    # ...
+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/producer-consumer/index.html b/en/2023.4/producer-consumer/index.html new file mode 100644 index 000000000..98bfc3726 --- /dev/null +++ b/en/2023.4/producer-consumer/index.html @@ -0,0 +1,1507 @@ + + + + + + + + + + + + + + + + + + + + + + + + Producer/Consumer mode - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Producer / Consumer mode

+

freqtrade provides a mechanism whereby an instance (also called consumer) may listen to messages from an upstream freqtrade instance (also called producer) using the message websocket. Mainly, analyzed_df and whitelist messages. This allows the reuse of computed indicators (and signals) for pairs in multiple bots without needing to compute them multiple times.

+

See Message Websocket in the Rest API docs for setting up the api_server configuration for your message websocket (this will be your producer).

+
+

Note

+

We strongly recommend to set ws_token to something random and known only to yourself to avoid unauthorized access to your bot.

+
+

Configuration

+

Enable subscribing to an instance by adding the external_message_consumer section to the consumer's config file.

+
{
+    //...
+   "external_message_consumer": {
+        "enabled": true,
+        "producers": [
+            {
+                "name": "default", // This can be any name you'd like, default is "default"
+                "host": "127.0.0.1", // The host from your producer's api_server config
+                "port": 8080, // The port from your producer's api_server config
+                "secure": false, // Use a secure websockets connection, default false
+                "ws_token": "sercet_Ws_t0ken" // The ws_token from your producer's api_server config
+            }
+        ],
+        // The following configurations are optional, and usually not required
+        // "wait_timeout": 300,
+        // "ping_timeout": 10,
+        // "sleep_time": 10,
+        // "remove_entry_exit_signals": false,
+        // "message_size_limit": 8
+    }
+    //...
+}
+
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
ParameterDescription
enabledRequired. Enable consumer mode. If set to false, all other settings in this section are ignored.
Defaults to false.
Datatype: boolean .
producersRequired. List of producers
Datatype: Array.
producers.nameRequired. Name of this producer. This name must be used in calls to get_producer_pairs() and get_producer_df() if more than one producer is used.
Datatype: string
producers.hostRequired. The hostname or IP address from your producer.
Datatype: string
producers.portRequired. The port matching the above host.
Defaults to 8080.
Datatype: Integer
producers.secureOptional. Use ssl in websockets connection. Default False.
Datatype: string
producers.ws_tokenRequired. ws_token as configured on the producer.
Datatype: string
Optional settings
wait_timeoutTimeout until we ping again if no message is received.
Defaults to 300.
Datatype: Integer - in seconds.
ping_timeoutPing timeout
Defaults to 10.
Datatype: Integer - in seconds.
sleep_timeSleep time before retrying to connect.
Defaults to 10.
Datatype: Integer - in seconds.
remove_entry_exit_signalsRemove signal columns from the dataframe (set them to 0) on dataframe receipt.
Defaults to false.
Datatype: Boolean.
message_size_limitSize limit per message
Defaults to 8.
Datatype: Integer - Megabytes.
+

Instead of (or as well as) calculating indicators in populate_indicators() the follower instance listens on the connection to a producer instance's messages (or multiple producer instances in advanced configurations) and requests the producer's most recently analyzed dataframes for each pair in the active whitelist.

+

A consumer instance will then have a full copy of the analyzed dataframes without the need to calculate them itself.

+

Examples

+

Example - Producer Strategy

+

A simple strategy with multiple indicators. No special considerations are required in the strategy itself.

+
class ProducerStrategy(IStrategy):
+    #...
+    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        Calculate indicators in the standard freqtrade way which can then be broadcast to other instances
+        """
+        dataframe['rsi'] = ta.RSI(dataframe)
+        bollinger = qtpylib.bollinger_bands(qtpylib.typical_price(dataframe), window=20, stds=2)
+        dataframe['bb_lowerband'] = bollinger['lower']
+        dataframe['bb_middleband'] = bollinger['mid']
+        dataframe['bb_upperband'] = bollinger['upper']
+        dataframe['tema'] = ta.TEMA(dataframe, timeperiod=9)
+
+        return dataframe
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        Populates the entry signal for the given dataframe
+        """
+        dataframe.loc[
+            (
+                (qtpylib.crossed_above(dataframe['rsi'], self.buy_rsi.value)) &
+                (dataframe['tema'] <= dataframe['bb_middleband']) &
+                (dataframe['tema'] > dataframe['tema'].shift(1)) &
+                (dataframe['volume'] > 0)
+            ),
+            'enter_long'] = 1
+
+        return dataframe
+
+
+

FreqAI

+

You can use this to setup FreqAI on a powerful machine, while you run consumers on simple machines like raspberries, which can interpret the signals generated from the producer in different ways.

+
+

Example - Consumer Strategy

+

A logically equivalent strategy which calculates no indicators itself, but will have the same analyzed dataframes available to make trading decisions based on the indicators calculated in the producer. In this example the consumer has the same entry criteria, however this is not necessary. The consumer may use different logic to enter/exit trades, and only use the indicators as specified.

+
class ConsumerStrategy(IStrategy):
+    #...
+    process_only_new_candles = False # required for consumers
+
+    _columns_to_expect = ['rsi_default', 'tema_default', 'bb_middleband_default']
+
+    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        Use the websocket api to get pre-populated indicators from another freqtrade instance.
+        Use `self.dp.get_producer_df(pair)` to get the dataframe
+        """
+        pair = metadata['pair']
+        timeframe = self.timeframe
+
+        producer_pairs = self.dp.get_producer_pairs()
+        # You can specify which producer to get pairs from via:
+        # self.dp.get_producer_pairs("my_other_producer")
+
+        # This func returns the analyzed dataframe, and when it was analyzed
+        producer_dataframe, _ = self.dp.get_producer_df(pair)
+        # You can get other data if the producer makes it available:
+        # self.dp.get_producer_df(
+        #   pair,
+        #   timeframe="1h",
+        #   candle_type=CandleType.SPOT,
+        #   producer_name="my_other_producer"
+        # )
+
+        if not producer_dataframe.empty:
+            # If you plan on passing the producer's entry/exit signal directly,
+            # specify ffill=False or it will have unintended results
+            merged_dataframe = merge_informative_pair(dataframe, producer_dataframe,
+                                                      timeframe, timeframe,
+                                                      append_timeframe=False,
+                                                      suffix="default")
+            return merged_dataframe
+        else:
+            dataframe[self._columns_to_expect] = 0
+
+        return dataframe
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        Populates the entry signal for the given dataframe
+        """
+        # Use the dataframe columns as if we calculated them ourselves
+        dataframe.loc[
+            (
+                (qtpylib.crossed_above(dataframe['rsi_default'], self.buy_rsi.value)) &
+                (dataframe['tema_default'] <= dataframe['bb_middleband_default']) &
+                (dataframe['tema_default'] > dataframe['tema_default'].shift(1)) &
+                (dataframe['volume'] > 0)
+            ),
+            'enter_long'] = 1
+
+        return dataframe
+
+
+

Using upstream signals

+

By setting remove_entry_exit_signals=false, you can also use the producer's signals directly. They should be available as enter_long_default (assuming suffix="default" was used) - and can be used as either signal directly, or as additional indicator.

+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/requirements-docs.txt b/en/2023.4/requirements-docs.txt new file mode 100644 index 000000000..91b0e993b --- /dev/null +++ b/en/2023.4/requirements-docs.txt @@ -0,0 +1,6 @@ +markdown==3.3.7 +mkdocs==1.4.2 +mkdocs-material==9.1.7 +mdx_truly_sane_lists==1.3 +pymdown-extensions==9.11 +jinja2==3.1.2 diff --git a/en/2023.4/rest-api/index.html b/en/2023.4/rest-api/index.html new file mode 100644 index 000000000..2ed200e23 --- /dev/null +++ b/en/2023.4/rest-api/index.html @@ -0,0 +1,2023 @@ + + + + + + + + + + + + + + + + + + + + + + + + REST API & FreqUI - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

REST API & FreqUI

+

FreqUI

+

Freqtrade provides a builtin webserver, which can serve FreqUI, the freqtrade UI.

+

By default, the UI is not included in the installation (except for docker images), and must be installed explicitly with freqtrade install-ui. +This same command can also be used to update freqUI, should there be a new release.

+

Once the bot is started in trade / dry-run mode (with freqtrade trade) - the UI will be available under the configured port below (usually http://127.0.0.1:8080).

+
+

developers

+

Developers should not use this method, but instead use the method described in the freqUI repository to get the source-code of freqUI.

+
+

Configuration

+

Enable the rest API by adding the api_server section to your configuration and setting api_server.enabled to true.

+

Sample configuration:

+
    "api_server": {
+        "enabled": true,
+        "listen_ip_address": "127.0.0.1",
+        "listen_port": 8080,
+        "verbosity": "error",
+        "enable_openapi": false,
+        "jwt_secret_key": "somethingrandom",
+        "CORS_origins": [],
+        "username": "Freqtrader",
+        "password": "SuperSecret1!",
+        "ws_token": "sercet_Ws_t0ken"
+    },
+
+
+

Security warning

+

By default, the configuration listens on localhost only (so it's not reachable from other systems). We strongly recommend to not expose this API to the internet and choose a strong, unique password, since others will potentially be able to control your bot.

+
+
+API/UI Access on a remote servers +

If you're running on a VPS, you should consider using either a ssh tunnel, or setup a VPN (openVPN, wireguard) to connect to your bot. +This will ensure that freqUI is not directly exposed to the internet, which is not recommended for security reasons (freqUI does not support https out of the box). +Setup of these tools is not part of this tutorial, however many good tutorials can be found on the internet.

+
+

You can then access the API by going to http://127.0.0.1:8080/api/v1/ping in a browser to check if the API is running correctly. +This should return the response:

+
{"status":"pong"}
+
+

All other endpoints return sensitive info and require authentication and are therefore not available through a web browser.

+

Security

+

To generate a secure password, best use a password manager, or use the below code.

+
import secrets
+secrets.token_hex()
+
+
+

JWT token

+

Use the same method to also generate a JWT secret key (jwt_secret_key).

+
+
+

Password selection

+

Please make sure to select a very strong, unique password to protect your bot from unauthorized access. +Also change jwt_secret_key to something random (no need to remember this, but it'll be used to encrypt your session, so it better be something unique!).

+
+

Configuration with docker

+

If you run your bot using docker, you'll need to have the bot listen to incoming connections. The security is then handled by docker.

+
    "api_server": {
+        "enabled": true,
+        "listen_ip_address": "0.0.0.0",
+        "listen_port": 8080,
+        "username": "Freqtrader",
+        "password": "SuperSecret1!",
+        //...
+    },
+
+

Make sure that the following 2 lines are available in your docker-compose file:

+
    ports:
+      - "127.0.0.1:8080:8080"
+
+
+

Security warning

+

By using 8080:8080 in the docker port mapping, the API will be available to everyone connecting to the server under the correct port, so others may be able to control your bot.

+
+

Rest API

+

Consuming the API

+

You can consume the API by using the script scripts/rest_client.py. +The client script only requires the requests module, so Freqtrade does not need to be installed on the system.

+
python3 scripts/rest_client.py <command> [optional parameters]
+
+

By default, the script assumes 127.0.0.1 (localhost) and port 8080 to be used, however you can specify a configuration file to override this behaviour.

+

Minimalistic client config

+
{
+    "api_server": {
+        "enabled": true,
+        "listen_ip_address": "0.0.0.0",
+        "listen_port": 8080,
+        "username": "Freqtrader",
+        "password": "SuperSecret1!",
+        //...
+    }
+}
+
+
python3 scripts/rest_client.py --config rest_config.json <command> [optional parameters]
+
+

Available endpoints

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
CommandDescription
pingSimple command testing the API Readiness - requires no authentication.
startStarts the trader.
stopStops the trader.
stopbuyStops the trader from opening new trades. Gracefully closes open trades according to their rules.
reload_configReloads the configuration file.
tradesList last trades. Limited to 500 trades per call.
trade/<tradeid>Get specific trade.
delete_trade <trade_id>Remove trade from the database. Tries to close open orders. Requires manual handling of this trade on the exchange.
show_configShows part of the current configuration with relevant settings to operation.
logsShows last log messages.
statusLists all open trades.
countDisplays number of trades used and available.
locksDisplays currently locked pairs.
delete_lock <lock_id>Deletes (disables) the lock by id.
profitDisplay a summary of your profit/loss from close trades and some stats about your performance.
forceexit <trade_id>Instantly exits the given trade (Ignoring minimum_roi).
forceexit allInstantly exits all open trades (Ignoring minimum_roi).
forceenter <pair> [rate]Instantly enters the given pair. Rate is optional. (force_entry_enable must be set to True)
forceenter <pair> <side> [rate]Instantly longs or shorts the given pair. Rate is optional. (force_entry_enable must be set to True)
performanceShow performance of each finished trade grouped by pair.
balanceShow account balance per currency.
daily <n>Shows profit or loss per day, over the last n days (n defaults to 7).
statsDisplay a summary of profit / loss reasons as well as average holding times.
whitelistShow the current whitelist.
blacklist [pair]Show the current blacklist, or adds a pair to the blacklist.
edgeShow validated pairs by Edge if it is enabled.
pair_candlesReturns dataframe for a pair / timeframe combination while the bot is running. Alpha
pair_historyReturns an analyzed dataframe for a given timerange, analyzed by a given strategy. Alpha
plot_configGet plot config from the strategy (or nothing if not configured). Alpha
strategiesList strategies in strategy directory. Alpha
strategy <strategy>Get specific Strategy content. Alpha
available_pairsList available backtest data. Alpha
versionShow version.
sysinfoShow information about the system load.
healthShow bot health (last bot loop).
+
+

Alpha status

+

Endpoints labeled with Alpha status above may change at any time without notice.

+
+

Possible commands can be listed from the rest-client script using the help command.

+
python3 scripts/rest_client.py help
+
+
Possible commands:
+
+available_pairs
+    Return available pair (backtest data) based on timeframe / stake_currency selection
+
+        :param timeframe: Only pairs with this timeframe available.
+        :param stake_currency: Only pairs that include this timeframe
+
+balance
+    Get the account balance.
+
+blacklist
+    Show the current blacklist.
+
+        :param add: List of coins to add (example: "BNB/BTC")
+
+cancel_open_order
+    Cancel open order for trade.
+
+        :param trade_id: Cancels open orders for this trade.
+
+count
+    Return the amount of open trades.
+
+daily
+    Return the profits for each day, and amount of trades.
+
+delete_lock
+    Delete (disable) lock from the database.
+
+        :param lock_id: ID for the lock to delete
+
+delete_trade
+    Delete trade from the database.
+        Tries to close open orders. Requires manual handling of this asset on the exchange.
+
+        :param trade_id: Deletes the trade with this ID from the database.
+
+edge
+    Return information about edge.
+
+forcebuy
+    Buy an asset.
+
+        :param pair: Pair to buy (ETH/BTC)
+        :param price: Optional - price to buy
+
+forceenter
+    Force entering a trade
+
+        :param pair: Pair to buy (ETH/BTC)
+        :param side: 'long' or 'short'
+        :param price: Optional - price to buy
+
+forceexit
+    Force-exit a trade.
+
+        :param tradeid: Id of the trade (can be received via status command)
+        :param ordertype: Order type to use (must be market or limit)
+        :param amount: Amount to sell. Full sell if not given
+
+health
+    Provides a quick health check of the running bot.
+
+locks
+    Return current locks
+
+logs
+    Show latest logs.
+
+        :param limit: Limits log messages to the last <limit> logs. No limit to get the entire log.
+
+pair_candles
+    Return live dataframe for <pair><timeframe>.
+
+        :param pair: Pair to get data for
+        :param timeframe: Only pairs with this timeframe available.
+        :param limit: Limit result to the last n candles.
+
+pair_history
+    Return historic, analyzed dataframe
+
+        :param pair: Pair to get data for
+        :param timeframe: Only pairs with this timeframe available.
+        :param strategy: Strategy to analyze and get values for
+        :param timerange: Timerange to get data for (same format than --timerange endpoints)
+
+performance
+    Return the performance of the different coins.
+
+ping
+    simple ping
+
+plot_config
+    Return plot configuration if the strategy defines one.
+
+profit
+    Return the profit summary.
+
+reload_config
+    Reload configuration.
+
+show_config
+        Returns part of the configuration, relevant for trading operations.
+
+start
+    Start the bot if it's in the stopped state.
+
+stats
+    Return the stats report (durations, sell-reasons).
+
+status
+    Get the status of open trades.
+
+stop
+    Stop the bot. Use `start` to restart.
+
+stopbuy
+    Stop buying (but handle sells gracefully). Use `reload_config` to reset.
+
+strategies
+    Lists available strategies
+
+strategy
+    Get strategy details
+
+        :param strategy: Strategy class name
+
+sysinfo
+    Provides system information (CPU, RAM usage)
+
+trade
+    Return specific trade
+
+        :param trade_id: Specify which trade to get.
+
+trades
+    Return trades history, sorted by id
+
+        :param limit: Limits trades to the X last trades. Max 500 trades.
+        :param offset: Offset by this amount of trades.
+
+version
+    Return the version of the bot.
+
+whitelist
+    Show the current whitelist.
+
+

Message WebSocket

+

The API Server includes a websocket endpoint for subscribing to RPC messages from the freqtrade Bot. +This can be used to consume real-time data from your bot, such as entry/exit fill messages, whitelist changes, populated indicators for pairs, and more.

+

This is also used to setup Producer/Consumer mode in Freqtrade.

+

Assuming your rest API is set to 127.0.0.1 on port 8080, the endpoint is available at http://localhost:8080/api/v1/message/ws.

+

To access the websocket endpoint, the ws_token is required as a query parameter in the endpoint URL.

+

To generate a safe ws_token you can run the following code:

+
>>> import secrets
+>>> secrets.token_urlsafe(25)
+'hZ-y58LXyX_HZ8O1cJzVyN6ePWrLpNQv4Q'
+
+

You would then add that token under ws_token in your api_server config. Like so:

+
"api_server": {
+    "enabled": true,
+    "listen_ip_address": "127.0.0.1",
+    "listen_port": 8080,
+    "verbosity": "error",
+    "enable_openapi": false,
+    "jwt_secret_key": "somethingrandom",
+    "CORS_origins": [],
+    "username": "Freqtrader",
+    "password": "SuperSecret1!",
+    "ws_token": "hZ-y58LXyX_HZ8O1cJzVyN6ePWrLpNQv4Q" // <-----
+},
+
+

You can now connect to the endpoint at http://localhost:8080/api/v1/message/ws?token=hZ-y58LXyX_HZ8O1cJzVyN6ePWrLpNQv4Q.

+
+

Reuse of example tokens

+

Please do not use the above example token. To make sure you are secure, generate a completely new token.

+
+

Using the WebSocket

+

Once connected to the WebSocket, the bot will broadcast RPC messages to anyone who is subscribed to them. To subscribe to a list of messages, you must send a JSON request through the WebSocket like the one below. The data key must be a list of message type strings.

+
{
+  "type": "subscribe",
+  "data": ["whitelist", "analyzed_df"] // A list of string message types
+}
+
+

For a list of message types, please refer to the RPCMessageType enum in freqtrade/enums/rpcmessagetype.py

+

Now anytime those types of RPC messages are sent in the bot, you will receive them through the WebSocket as long as the connection is active. They typically take the same form as the request:

+
{
+  "type": "analyzed_df",
+  "data": {
+      "key": ["NEO/BTC", "5m", "spot"],
+      "df": {}, // The dataframe
+      "la": "2022-09-08 22:14:41.457786+00:00"
+  }
+}
+
+

Reverse Proxy setup

+

When using Nginx, the following configuration is required for WebSockets to work (Note this configuration is incomplete, it's missing some information and can not be used as is):

+

Please make sure to replace <freqtrade_listen_ip> (and the subsequent port) with the IP and Port matching your configuration/setup.

+
http {
+    map $http_upgrade $connection_upgrade {
+        default upgrade;
+        '' close;
+    }
+
+    #...
+
+    server {
+        #...
+
+        location / {
+            proxy_http_version 1.1;
+            proxy_pass http://<freqtrade_listen_ip>:8080;
+            proxy_set_header Upgrade $http_upgrade;
+            proxy_set_header Connection $connection_upgrade;
+            proxy_set_header Host $host;
+        }
+    }
+}
+
+

To properly configure your reverse proxy (securely), please consult it's documentation for proxying websockets.

+
    +
  • Traefik: Traefik supports websockets out of the box, see the documentation
  • +
  • Caddy: Caddy v2 supports websockets out of the box, see the documentation
  • +
+
+

SSL certificates

+

You can use tools like certbot to setup ssl certificates to access your bot's UI through encrypted connection by using any fo the above reverse proxies. +While this will protect your data in transit, we do not recommend to run the freqtrade API outside of your private network (VPN, SSH tunnel).

+
+

OpenAPI interface

+

To enable the builtin openAPI interface (Swagger UI), specify "enable_openapi": true in the api_server configuration. +This will enable the Swagger UI at the /docs endpoint. By default, that's running at http://localhost:8080/docs - but it'll depend on your settings.

+

Advanced API usage using JWT tokens

+
+

Note

+

The below should be done in an application (a Freqtrade REST API client, which fetches info via API), and is not intended to be used on a regular basis.

+
+

Freqtrade's REST API also offers JWT (JSON Web Tokens). +You can login using the following command, and subsequently use the resulting access_token.

+
> curl -X POST --user Freqtrader http://localhost:8080/api/v1/token/login
+{"access_token":"eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpYXQiOjE1ODkxMTk2ODEsIm5iZiI6MTU4OTExOTY4MSwianRpIjoiMmEwYmY0NWUtMjhmOS00YTUzLTlmNzItMmM5ZWVlYThkNzc2IiwiZXhwIjoxNTg5MTIwNTgxLCJpZGVudGl0eSI6eyJ1IjoiRnJlcXRyYWRlciJ9LCJmcmVzaCI6ZmFsc2UsInR5cGUiOiJhY2Nlc3MifQ.qt6MAXYIa-l556OM7arBvYJ0SDI9J8bIk3_glDujF5g","refresh_token":"eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpYXQiOjE1ODkxMTk2ODEsIm5iZiI6MTU4OTExOTY4MSwianRpIjoiZWQ1ZWI3YjAtYjMwMy00YzAyLTg2N2MtNWViMjIxNWQ2YTMxIiwiZXhwIjoxNTkxNzExNjgxLCJpZGVudGl0eSI6eyJ1IjoiRnJlcXRyYWRlciJ9LCJ0eXBlIjoicmVmcmVzaCJ9.d1AT_jYICyTAjD0fiQAr52rkRqtxCjUGEMwlNuuzgNQ"}
+
+> access_token="eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpYXQiOjE1ODkxMTk2ODEsIm5iZiI6MTU4OTExOTY4MSwianRpIjoiMmEwYmY0NWUtMjhmOS00YTUzLTlmNzItMmM5ZWVlYThkNzc2IiwiZXhwIjoxNTg5MTIwNTgxLCJpZGVudGl0eSI6eyJ1IjoiRnJlcXRyYWRlciJ9LCJmcmVzaCI6ZmFsc2UsInR5cGUiOiJhY2Nlc3MifQ.qt6MAXYIa-l556OM7arBvYJ0SDI9J8bIk3_glDujF5g"
+# Use access_token for authentication
+> curl -X GET --header "Authorization: Bearer ${access_token}" http://localhost:8080/api/v1/count
+
+

Since the access token has a short timeout (15 min) - the token/refresh request should be used periodically to get a fresh access token:

+
> curl -X POST --header "Authorization: Bearer ${refresh_token}"http://localhost:8080/api/v1/token/refresh
+{"access_token":"eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpYXQiOjE1ODkxMTk5NzQsIm5iZiI6MTU4OTExOTk3NCwianRpIjoiMDBjNTlhMWUtMjBmYS00ZTk0LTliZjAtNWQwNTg2MTdiZDIyIiwiZXhwIjoxNTg5MTIwODc0LCJpZGVudGl0eSI6eyJ1IjoiRnJlcXRyYWRlciJ9LCJmcmVzaCI6ZmFsc2UsInR5cGUiOiJhY2Nlc3MifQ.1seHlII3WprjjclY6DpRhen0rqdF4j6jbvxIhUFaSbs"}
+
+

CORS

+

This whole section is only necessary in cross-origin cases (where you multiple bot API's running on localhost:8081, localhost:8082, ...), and want to combine them into one FreqUI instance.

+
+Technical explanation +

All web-based front-ends are subject to CORS - Cross-Origin Resource Sharing. +Since most of the requests to the Freqtrade API must be authenticated, a proper CORS policy is key to avoid security problems. +Also, the standard disallows * CORS policies for requests with credentials, so this setting must be set appropriately.

+
+

Users can allow access from different origin URL's to the bot API via the CORS_origins configuration setting. +It consists of a list of allowed URL's that are allowed to consume resources from the bot's API.

+

Assuming your application is deployed as https://frequi.freqtrade.io/home/ - this would mean that the following configuration becomes necessary:

+
{
+    //...
+    "jwt_secret_key": "somethingrandom",
+    "CORS_origins": ["https://frequi.freqtrade.io"],
+    //...
+}
+
+

In the following (pretty common) case, FreqUI is accessible on http://localhost:8080/trade (this is what you see in your navbar when navigating to freqUI). +freqUI url

+

The correct configuration for this case is http://localhost:8080 - the main part of the URL including the port.

+
{
+    //...
+    "jwt_secret_key": "somethingrandom",
+    "CORS_origins": ["http://localhost:8080"],
+    //...
+}
+
+
+

Note

+

We strongly recommend to also set jwt_secret_key to something random and known only to yourself to avoid unauthorized access to your bot.

+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/sandbox-testing/index.html b/en/2023.4/sandbox-testing/index.html new file mode 100644 index 000000000..80d3bae4d --- /dev/null +++ b/en/2023.4/sandbox-testing/index.html @@ -0,0 +1,1505 @@ + + + + + + + + + + + + + + + + + + + + + + + + Sandbox Testing - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Sandbox API testing

+

Some exchanges provide sandboxes or testbeds for risk-free testing, while running the bot against a real exchange. +With some configuration, freqtrade (in combination with ccxt) provides access to these.

+

This document is an overview to configure Freqtrade to be used with sandboxes. +This can be useful to developers and trader alike.

+
+

Warning

+

Sandboxes usually have very low volume, and either a very wide spread, or no orders available at all. +Therefore, sandboxes will usually not do a good job of showing you how a strategy would work in real trading.

+
+

Exchanges known to have a sandbox / testnet

+ +
+

Note

+

We did not test correct functioning of all of the above testnets. Please report your experiences with each sandbox.

+
+
+

Configure a Sandbox account

+

When testing your API connectivity, make sure to use the appropriate sandbox / testnet URL.

+

In general, you should follow these steps to enable an exchange's sandbox:

+
    +
  • Figure out if an exchange has a sandbox (most likely by using google or the exchange's support documents)
  • +
  • Create a sandbox account (often the sandbox-account requires separate registration)
  • +
  • Add some test assets to account
  • +
  • Create API keys
  • +
+

Add test funds

+

Usually, sandbox exchanges allow depositing funds directly via web-interface. +You should make sure to have a realistic amount of funds available to your test-account, so results are representable of your real account funds.

+
+

Warning

+

Test exchanges will NEVER require your real credit card or banking details!

+
+

Configure freqtrade to use a exchange's sandbox

+

Sandbox URLs

+

Freqtrade makes use of CCXT which in turn provides a list of URLs to Freqtrade. +These include ['test'] and ['api'].

+
    +
  • [Test] if available will point to an Exchanges sandbox.
  • +
  • [Api] normally used, and resolves to live API target on the exchange.
  • +
+

To make use of sandbox / test add "sandbox": true, to your config.json

+
  "exchange": {
+        "name": "coinbasepro",
+        "sandbox": true,
+        "key": "5wowfxemogxeowo;heiohgmd",
+        "secret": "/ZMH1P62rCVmwefewrgcewX8nh4gob+lywxfwfxwwfxwfNsH1ySgvWCUR/w==",
+        "password": "1bkjfkhfhfu6sr",
+        "outdated_offset": 5
+        "pair_whitelist": [
+            "BTC/USD"
+        ]
+  },
+  "datadir": "user_data/data/coinbasepro_sandbox"
+
+

Also the following information:

+
    +
  • api-key (created for the sandbox webpage)
  • +
  • api-secret (noted earlier)
  • +
  • password (the passphrase - noted earlier)
  • +
+
+

Different data directory

+

We also recommend to set datadir to something identifying downloaded data as sandbox data, to avoid having sandbox data mixed with data from the real exchange. +This can be done by adding the "datadir" key to the configuration. +Now, whenever you use this configuration, your data directory will be set to this directory.

+
+
+

You should now be ready to test your sandbox

+

Ensure Freqtrade logs show the sandbox URL, and trades made are shown in sandbox. Also make sure to select a pair which shows at least some decent value (which very often is BTC/).

+

Common problems with sandbox exchanges

+

Sandbox exchange instances often have very low volume, which can cause some problems which usually are not seen on a real exchange instance.

+

Old Candles problem

+

Since Sandboxes often have low volume, candles can be quite old and show no volume. +To disable the error "Outdated history for pair ...", best increase the parameter "outdated_offset" to a number that seems realistic for the sandbox you're using.

+

Unfilled orders

+

Sandboxes often have very low volumes - which means that many trades can go unfilled, or can go unfilled for a very long time.

+

To mitigate this, you can try to match the first order on the opposite orderbook side using the following configuration:

+

jsonc + "order_types": { + "entry": "limit", + "exit": "limit" + // ... + }, + "entry_pricing": { + "price_side": "other", + // ... + }, + "exit_pricing":{ + "price_side": "other", + // ... + },

+

The configuration is similar to the suggested configuration for market orders - however by using limit-orders you can avoid moving the price too much, and you can set the worst price you might get.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/search/search_index.json b/en/2023.4/search/search_index.json new file mode 100644 index 000000000..58511f9a2 --- /dev/null +++ b/en/2023.4/search/search_index.json @@ -0,0 +1 @@ +{"config":{"lang":["en"],"separator":"[\\s\\-]+","pipeline":["stopWordFilter"]},"docs":[{"location":"","title":"Home","text":"

Star Fork Download

"},{"location":"#introduction","title":"Introduction","text":"

Freqtrade is a free and open source crypto trading bot written in Python. It is designed to support all major exchanges and be controlled via Telegram or webUI. It contains backtesting, plotting and money management tools as well as strategy optimization by machine learning.

DISCLAIMER

This software is for educational purposes only. Do not risk money which you are afraid to lose. USE THE SOFTWARE AT YOUR OWN RISK. THE AUTHORS AND ALL AFFILIATES ASSUME NO RESPONSIBILITY FOR YOUR TRADING RESULTS.

Always start by running a trading bot in Dry-run and do not engage money before you understand how it works and what profit/loss you should expect.

We strongly recommend you to have basic coding skills and Python knowledge. Do not hesitate to read the source code and understand the mechanisms of this bot, algorithms and techniques implemented in it.

"},{"location":"#features","title":"Features","text":"
  • Develop your Strategy: Write your strategy in python, using pandas. Example strategies to inspire you are available in the strategy repository.
  • Download market data: Download historical data of the exchange and the markets your may want to trade with.
  • Backtest: Test your strategy on downloaded historical data.
  • Optimize: Find the best parameters for your strategy using hyperoptimization which employs machining learning methods. You can optimize buy, sell, take profit (ROI), stop-loss and trailing stop-loss parameters for your strategy.
  • Select markets: Create your static list or use an automatic one based on top traded volumes and/or prices (not available during backtesting). You can also explicitly blacklist markets you don't want to trade.
  • Run: Test your strategy with simulated money (Dry-Run mode) or deploy it with real money (Live-Trade mode).
  • Run using Edge (optional module): The concept is to find the best historical trade expectancy by markets based on variation of the stop-loss and then allow/reject markets to trade. The sizing of the trade is based on a risk of a percentage of your capital.
  • Control/Monitor: Use Telegram or a WebUI (start/stop the bot, show profit/loss, daily summary, current open trades results, etc.).
  • Analyze: Further analysis can be performed on either Backtesting data or Freqtrade trading history (SQL database), including automated standard plots, and methods to load the data into interactive environments.
"},{"location":"#supported-exchange-marketplaces","title":"Supported exchange marketplaces","text":"

Please read the exchange specific notes to learn about eventual, special configurations needed for each exchange.

  • Binance
  • Bittrex
  • Gate.io
  • Huobi
  • Kraken
  • OKX (Former OKEX)
  • potentially many others through . (We cannot guarantee they will work)
"},{"location":"#supported-futures-exchanges-experimental","title":"Supported Futures Exchanges (experimental)","text":"
  • Binance
  • Gate.io
  • OKX
  • Bybit

Please make sure to read the exchange specific notes, as well as the trading with leverage documentation before diving in.

"},{"location":"#community-tested","title":"Community tested","text":"

Exchanges confirmed working by the community:

  • Bitvavo
  • Kucoin
"},{"location":"#requirements","title":"Requirements","text":""},{"location":"#hardware-requirements","title":"Hardware requirements","text":"

To run this bot we recommend you a linux cloud instance with a minimum of:

  • 2GB RAM
  • 1GB disk space
  • 2vCPU
"},{"location":"#software-requirements","title":"Software requirements","text":"
  • Docker (Recommended)

Alternatively

  • Python 3.8+
  • pip (pip3)
  • git
  • TA-Lib
  • virtualenv (Recommended)
"},{"location":"#support","title":"Support","text":""},{"location":"#help-discord","title":"Help / Discord","text":"

For any questions not covered by the documentation or for further information about the bot, or to simply engage with like-minded individuals, we encourage you to join the Freqtrade discord server.

"},{"location":"#ready-to-try","title":"Ready to try?","text":"

Begin by reading the installation guide for docker (recommended), or for installation without docker.

"},{"location":"advanced-backtesting/","title":"Advanced Backtesting Analysis","text":""},{"location":"advanced-backtesting/#analyze-the-buyentry-and-sellexit-tags","title":"Analyze the buy/entry and sell/exit tags","text":"

It can be helpful to understand how a strategy behaves according to the buy/entry tags used to mark up different buy conditions. You might want to see more complex statistics about each buy and sell condition above those provided by the default backtesting output. You may also want to determine indicator values on the signal candle that resulted in a trade opening.

Note

The following buy reason analysis is only available for backtesting, not hyperopt.

We need to run backtesting with the --export option set to signals to enable the exporting of signals and trades:

freqtrade backtesting -c <config.json> --timeframe <tf> --strategy <strategy_name> --timerange=<timerange> --export=signals\n

This will tell freqtrade to output a pickled dictionary of strategy, pairs and corresponding DataFrame of the candles that resulted in buy signals. Depending on how many buys your strategy makes, this file may get quite large, so periodically check your user_data/backtest_results folder to delete old exports.

Before running your next backtest, make sure you either delete your old backtest results or run backtesting with the --cache none option to make sure no cached results are used.

If all goes well, you should now see a backtest-result-{timestamp}_signals.pkl file in the user_data/backtest_results folder.

To analyze the entry/exit tags, we now need to use the freqtrade backtesting-analysis command with --analysis-groups option provided with space-separated arguments (default 0 1 2):

freqtrade backtesting-analysis -c <config.json> --analysis-groups 0 1 2 3 4 5\n

This command will read from the last backtesting results. The --analysis-groups option is used to specify the various tabular outputs showing the profit fo each group or trade, ranging from the simplest (0) to the most detailed per pair, per buy and per sell tag (4):

  • 1: profit summaries grouped by enter_tag
  • 2: profit summaries grouped by enter_tag and exit_tag
  • 3: profit summaries grouped by pair and enter_tag
  • 4: profit summaries grouped by pair, enter_ and exit_tag (this can get quite large)
  • 5: profit summaries grouped by exit_tag

More options are available by running with the -h option.

"},{"location":"advanced-backtesting/#using-export-filename","title":"Using export-filename","text":"

Normally, backtesting-analysis uses the latest backtest results, but if you wanted to go back to a previous backtest output, you need to supply the --export-filename option. You can supply the same parameter to backtest-analysis with the name of the final backtest output file. This allows you to keep historical versions of backtest results and re-analyse them at a later date:

freqtrade backtesting -c <config.json> --timeframe <tf> --strategy <strategy_name> --timerange=<timerange> --export=signals --export-filename=/tmp/mystrat_backtest.json\n

You should see some output similar to below in the logs with the name of the timestamped filename that was exported:

2022-06-14 16:28:32,698 - freqtrade.misc - INFO - dumping json to \"/tmp/mystrat_backtest-2022-06-14_16-28-32.json\"\n

You can then use that filename in backtesting-analysis:

freqtrade backtesting-analysis -c <config.json> --export-filename=/tmp/mystrat_backtest-2022-06-14_16-28-32.json\n
"},{"location":"advanced-backtesting/#tuning-the-buy-tags-and-sell-tags-to-display","title":"Tuning the buy tags and sell tags to display","text":"

To show only certain buy and sell tags in the displayed output, use the following two options:

--enter-reason-list : Space-separated list of enter signals to analyse. Default: \"all\"\n--exit-reason-list : Space-separated list of exit signals to analyse. Default: \"all\"\n

For example:

freqtrade backtesting-analysis -c <config.json> --analysis-groups 0 2 --enter-reason-list enter_tag_a enter_tag_b --exit-reason-list roi custom_exit_tag_a stop_loss\n
"},{"location":"advanced-backtesting/#outputting-signal-candle-indicators","title":"Outputting signal candle indicators","text":"

The real power of freqtrade backtesting-analysis comes from the ability to print out the indicator values present on signal candles to allow fine-grained investigation and tuning of buy signal indicators. To print out a column for a given set of indicators, use the --indicator-list option:

freqtrade backtesting-analysis -c <config.json> --analysis-groups 0 2 --enter-reason-list enter_tag_a enter_tag_b --exit-reason-list roi custom_exit_tag_a stop_loss --indicator-list rsi rsi_1h bb_lowerband ema_9 macd macdsignal\n

The indicators have to be present in your strategy's main DataFrame (either for your main timeframe or for informative timeframes) otherwise they will simply be ignored in the script output.

"},{"location":"advanced-backtesting/#filtering-the-trade-output-by-date","title":"Filtering the trade output by date","text":"

To show only trades between dates within your backtested timerange, supply the usual timerange option in YYYYMMDD-[YYYYMMDD] format:

--timerange : Timerange to filter output trades, start date inclusive, end date exclusive. e.g. 20220101-20221231\n

For example, if your backtest timerange was 20220101-20221231 but you only want to output trades in January:

freqtrade backtesting-analysis -c <config.json> --timerange 20220101-20220201\n
"},{"location":"advanced-hyperopt/","title":"Advanced Hyperopt","text":"

This page explains some advanced Hyperopt topics that may require higher coding skills and Python knowledge than creation of an ordinal hyperoptimization class.

"},{"location":"advanced-hyperopt/#creating-and-using-a-custom-loss-function","title":"Creating and using a custom loss function","text":"

To use a custom loss function class, make sure that the function hyperopt_loss_function is defined in your custom hyperopt loss class. For the sample below, you then need to add the command line parameter --hyperopt-loss SuperDuperHyperOptLoss to your hyperopt call so this function is being used.

A sample of this can be found below, which is identical to the Default Hyperopt loss implementation. A full sample can be found in userdata/hyperopts.

from datetime import datetime\nfrom typing import Any, Dict\n\nfrom pandas import DataFrame\n\nfrom freqtrade.constants import Config\nfrom freqtrade.optimize.hyperopt import IHyperOptLoss\n\nTARGET_TRADES = 600\nEXPECTED_MAX_PROFIT = 3.0\nMAX_ACCEPTED_TRADE_DURATION = 300\n\nclass SuperDuperHyperOptLoss(IHyperOptLoss):\n\"\"\"\n    Defines the default loss function for hyperopt\n    \"\"\"\n\n    @staticmethod\n    def hyperopt_loss_function(results: DataFrame, trade_count: int,\n                               min_date: datetime, max_date: datetime,\n                               config: Config, processed: Dict[str, DataFrame],\n                               backtest_stats: Dict[str, Any],\n                               *args, **kwargs) -> float:\n\"\"\"\n        Objective function, returns smaller number for better results\n        This is the legacy algorithm (used until now in freqtrade).\n        Weights are distributed as follows:\n        * 0.4 to trade duration\n        * 0.25: Avoiding trade loss\n        * 1.0 to total profit, compared to the expected value (`EXPECTED_MAX_PROFIT`) defined above\n        \"\"\"\n        total_profit = results['profit_ratio'].sum()\n        trade_duration = results['trade_duration'].mean()\n\n        trade_loss = 1 - 0.25 * exp(-(trade_count - TARGET_TRADES) ** 2 / 10 ** 5.8)\n        profit_loss = max(0, 1 - total_profit / EXPECTED_MAX_PROFIT)\n        duration_loss = 0.4 * min(trade_duration / MAX_ACCEPTED_TRADE_DURATION, 1)\n        result = trade_loss + profit_loss + duration_loss\n        return result\n

Currently, the arguments are:

  • results: DataFrame containing the resulting trades. The following columns are available in results (corresponds to the output-file of backtesting when used with --export trades): pair, profit_ratio, profit_abs, open_date, open_rate, fee_open, close_date, close_rate, fee_close, amount, trade_duration, is_open, exit_reason, stake_amount, min_rate, max_rate, stop_loss_ratio, stop_loss_abs
  • trade_count: Amount of trades (identical to len(results))
  • min_date: Start date of the timerange used
  • min_date: End date of the timerange used
  • config: Config object used (Note: Not all strategy-related parameters will be updated here if they are part of a hyperopt space).
  • processed: Dict of Dataframes with the pair as keys containing the data used for backtesting.
  • backtest_stats: Backtesting statistics using the same format as the backtesting file \"strategy\" substructure. Available fields can be seen in generate_strategy_stats() in optimize_reports.py.

This function needs to return a floating point number (float). Smaller numbers will be interpreted as better results. The parameters and balancing for this is up to you.

Note

This function is called once per epoch - so please make sure to have this as optimized as possible to not slow hyperopt down unnecessarily.

*args and **kwargs

Please keep the arguments *args and **kwargs in the interface to allow us to extend this interface in the future.

"},{"location":"advanced-hyperopt/#overriding-pre-defined-spaces","title":"Overriding pre-defined spaces","text":"

To override a pre-defined space (roi_space, generate_roi_table, stoploss_space, trailing_space, max_open_trades_space), define a nested class called Hyperopt and define the required spaces as follows:

from freqtrade.optimize.space import Categorical, Dimension, Integer, SKDecimal\n\nclass MyAwesomeStrategy(IStrategy):\n    class HyperOpt:\n        # Define a custom stoploss space.\n        def stoploss_space():\n            return [SKDecimal(-0.05, -0.01, decimals=3, name='stoploss')]\n\n        # Define custom ROI space\n        def roi_space() -> List[Dimension]:\n            return [\n                Integer(10, 120, name='roi_t1'),\n                Integer(10, 60, name='roi_t2'),\n                Integer(10, 40, name='roi_t3'),\n                SKDecimal(0.01, 0.04, decimals=3, name='roi_p1'),\n                SKDecimal(0.01, 0.07, decimals=3, name='roi_p2'),\n                SKDecimal(0.01, 0.20, decimals=3, name='roi_p3'),\n            ]\n\n        def generate_roi_table(params: Dict) -> Dict[int, float]:\n\n            roi_table = {}\n            roi_table[0] = params['roi_p1'] + params['roi_p2'] + params['roi_p3']\n            roi_table[params['roi_t3']] = params['roi_p1'] + params['roi_p2']\n            roi_table[params['roi_t3'] + params['roi_t2']] = params['roi_p1']\n            roi_table[params['roi_t3'] + params['roi_t2'] + params['roi_t1']] = 0\n\n            return roi_table\n\n        def trailing_space() -> List[Dimension]:\n            # All parameters here are mandatory, you can only modify their type or the range.\n            return [\n                # Fixed to true, if optimizing trailing_stop we assume to use trailing stop at all times.\n                Categorical([True], name='trailing_stop'),\n\n                SKDecimal(0.01, 0.35, decimals=3, name='trailing_stop_positive'),\n                # 'trailing_stop_positive_offset' should be greater than 'trailing_stop_positive',\n                # so this intermediate parameter is used as the value of the difference between\n                # them. The value of the 'trailing_stop_positive_offset' is constructed in the\n                # generate_trailing_params() method.\n                # This is similar to the hyperspace dimensions used for constructing the ROI tables.\n                SKDecimal(0.001, 0.1, decimals=3, name='trailing_stop_positive_offset_p1'),\n\n                Categorical([True, False], name='trailing_only_offset_is_reached'),\n        ]\n\n        # Define a custom max_open_trades space\n        def max_open_trades_space(self) -> List[Dimension]:\n            return [\n                Integer(-1, 10, name='max_open_trades'),\n            ]\n

Note

All overrides are optional and can be mixed/matched as necessary.

"},{"location":"advanced-hyperopt/#dynamic-parameters","title":"Dynamic parameters","text":"

Parameters can also be defined dynamically, but must be available to the instance once the * bot_start() callback has been called.

class MyAwesomeStrategy(IStrategy):\n\n    def bot_start(self, **kwargs) -> None:\n        self.buy_adx = IntParameter(20, 30, default=30, optimize=True)\n\n    # ...\n

Warning

Parameters created this way will not show up in the list-strategies parameter count.

"},{"location":"advanced-hyperopt/#overriding-base-estimator","title":"Overriding Base estimator","text":"

You can define your own estimator for Hyperopt by implementing generate_estimator() in the Hyperopt subclass.

class MyAwesomeStrategy(IStrategy):\n    class HyperOpt:\n        def generate_estimator(dimensions: List['Dimension'], **kwargs):\n            return \"RF\"\n

Possible values are either one of \"GP\", \"RF\", \"ET\", \"GBRT\" (Details can be found in the scikit-optimize documentation), or \"an instance of a class that inherits from RegressorMixin (from sklearn) and where the predict method has an optional return_std argument, which returns std(Y | x) along with E[Y | x]\".

Some research will be necessary to find additional Regressors.

Example for ExtraTreesRegressor (\"ET\") with additional parameters:

class MyAwesomeStrategy(IStrategy):\n    class HyperOpt:\n        def generate_estimator(dimensions: List['Dimension'], **kwargs):\n            from skopt.learning import ExtraTreesRegressor\n            # Corresponds to \"ET\" - but allows additional parameters.\n            return ExtraTreesRegressor(n_estimators=100)\n

The dimensions parameter is the list of skopt.space.Dimension objects corresponding to the parameters to be optimized. It can be used to create isotropic kernels for the skopt.learning.GaussianProcessRegressor estimator. Here's an example:

class MyAwesomeStrategy(IStrategy):\n    class HyperOpt:\n        def generate_estimator(dimensions: List['Dimension'], **kwargs):\n            from skopt.utils import cook_estimator\n            from skopt.learning.gaussian_process.kernels import (Matern, ConstantKernel)\n            kernel_bounds = (0.0001, 10000)\n            kernel = (\n                ConstantKernel(1.0, kernel_bounds) * \n                Matern(length_scale=np.ones(len(dimensions)), length_scale_bounds=[kernel_bounds for d in dimensions], nu=2.5)\n            )\n            kernel += (\n                ConstantKernel(1.0, kernel_bounds) * \n                Matern(length_scale=np.ones(len(dimensions)), length_scale_bounds=[kernel_bounds for d in dimensions], nu=1.5)\n            )\n\n            return cook_estimator(\"GP\", space=dimensions, kernel=kernel, n_restarts_optimizer=2)\n

Note

While custom estimators can be provided, it's up to you as User to do research on possible parameters and analyze / understand which ones should be used. If you're unsure about this, best use one of the Defaults (\"ET\" has proven to be the most versatile) without further parameters.

"},{"location":"advanced-hyperopt/#space-options","title":"Space options","text":"

For the additional spaces, scikit-optimize (in combination with Freqtrade) provides the following space types:

  • Categorical - Pick from a list of categories (e.g. Categorical(['a', 'b', 'c'], name=\"cat\"))
  • Integer - Pick from a range of whole numbers (e.g. Integer(1, 10, name='rsi'))
  • SKDecimal - Pick from a range of decimal numbers with limited precision (e.g. SKDecimal(0.1, 0.5, decimals=3, name='adx')). Available only with freqtrade.
  • Real - Pick from a range of decimal numbers with full precision (e.g. Real(0.1, 0.5, name='adx')

You can import all of these from freqtrade.optimize.space, although Categorical, Integer and Real are only aliases for their corresponding scikit-optimize Spaces. SKDecimal is provided by freqtrade for faster optimizations.

from freqtrade.optimize.space import Categorical, Dimension, Integer, SKDecimal, Real  # noqa\n

SKDecimal vs. Real

We recommend to use SKDecimal instead of the Real space in almost all cases. While the Real space provides full accuracy (up to ~16 decimal places) - this precision is rarely needed, and leads to unnecessary long hyperopt times.

Assuming the definition of a rather small space (SKDecimal(0.10, 0.15, decimals=2, name='xxx')) - SKDecimal will have 5 possibilities ([0.10, 0.11, 0.12, 0.13, 0.14, 0.15]).

A corresponding real space Real(0.10, 0.15 name='xxx') on the other hand has an almost unlimited number of possibilities ([0.10, 0.010000000001, 0.010000000002, ... 0.014999999999, 0.01500000000]).

"},{"location":"advanced-setup/","title":"Advanced Post-installation Tasks","text":"

This page explains some advanced tasks and configuration options that can be performed after the bot installation and may be uselful in some environments.

If you do not know what things mentioned here mean, you probably do not need it.

"},{"location":"advanced-setup/#running-multiple-instances-of-freqtrade","title":"Running multiple instances of Freqtrade","text":"

This section will show you how to run multiple bots at the same time, on the same machine.

"},{"location":"advanced-setup/#things-to-consider","title":"Things to consider","text":"
  • Use different database files.
  • Use different Telegram bots (requires multiple different configuration files; applies only when Telegram is enabled).
  • Use different ports (applies only when Freqtrade REST API webserver is enabled).
"},{"location":"advanced-setup/#different-database-files","title":"Different database files","text":"

In order to keep track of your trades, profits, etc., freqtrade is using a SQLite database where it stores various types of information such as the trades you performed in the past and the current position(s) you are holding at any time. This allows you to keep track of your profits, but most importantly, keep track of ongoing activity if the bot process would be restarted or would be terminated unexpectedly.

Freqtrade will, by default, use separate database files for dry-run and live bots (this assumes no database-url is given in either configuration nor via command line argument). For live trading mode, the default database will be tradesv3.sqlite and for dry-run it will be tradesv3.dryrun.sqlite.

The optional argument to the trade command used to specify the path of these files is --db-url, which requires a valid SQLAlchemy url. So when you are starting a bot with only the config and strategy arguments in dry-run mode, the following 2 commands would have the same outcome.

freqtrade trade -c MyConfig.json -s MyStrategy\n# is equivalent to\nfreqtrade trade -c MyConfig.json -s MyStrategy --db-url sqlite:///tradesv3.dryrun.sqlite\n

It means that if you are running the trade command in two different terminals, for example to test your strategy both for trades in USDT and in another instance for trades in BTC, you will have to run them with different databases.

If you specify the URL of a database which does not exist, freqtrade will create one with the name you specified. So to test your custom strategy with BTC and USDT stake currencies, you could use the following commands (in 2 separate terminals):

# Terminal 1:\nfreqtrade trade -c MyConfigBTC.json -s MyCustomStrategy --db-url sqlite:///user_data/tradesBTC.dryrun.sqlite\n# Terminal 2:\nfreqtrade trade -c MyConfigUSDT.json -s MyCustomStrategy --db-url sqlite:///user_data/tradesUSDT.dryrun.sqlite\n

Conversely, if you wish to do the same thing in production mode, you will also have to create at least one new database (in addition to the default one) and specify the path to the \"live\" databases, for example:

# Terminal 1:\nfreqtrade trade -c MyConfigBTC.json -s MyCustomStrategy --db-url sqlite:///user_data/tradesBTC.live.sqlite\n# Terminal 2:\nfreqtrade trade -c MyConfigUSDT.json -s MyCustomStrategy --db-url sqlite:///user_data/tradesUSDT.live.sqlite\n

For more information regarding usage of the sqlite databases, for example to manually enter or remove trades, please refer to the SQL Cheatsheet.

"},{"location":"advanced-setup/#multiple-instances-using-docker","title":"Multiple instances using docker","text":"

To run multiple instances of freqtrade using docker you will need to edit the docker-compose.yml file and add all the instances you want as separate services. Remember, you can separate your configuration into multiple files, so it's a good idea to think about making them modular, then if you need to edit something common to all bots, you can do that in a single config file.

---\nversion: '3'\nservices:\n  freqtrade1:\n    image: freqtradeorg/freqtrade:stable\n    # image: freqtradeorg/freqtrade:develop\n    # Use plotting image\n    # image: freqtradeorg/freqtrade:develop_plot\n    # Build step - only needed when additional dependencies are needed\n    # build:\n    #   context: .\n    #   dockerfile: \"./docker/Dockerfile.custom\"\n    restart: always\n    container_name: freqtrade1\n    volumes:\n      - \"./user_data:/freqtrade/user_data\"\n    # Expose api on port 8080 (localhost only)\n    # Please read the https://www.freqtrade.io/en/latest/rest-api/ documentation\n    # before enabling this.\n     ports:\n     - \"127.0.0.1:8080:8080\"\n    # Default command used when running `docker compose up`\n    command: >\n      trade\n      --logfile /freqtrade/user_data/logs/freqtrade1.log\n      --db-url sqlite:////freqtrade/user_data/tradesv3_freqtrade1.sqlite\n      --config /freqtrade/user_data/config.json\n      --config /freqtrade/user_data/config.freqtrade1.json\n      --strategy SampleStrategy\n\n  freqtrade2:\n    image: freqtradeorg/freqtrade:stable\n    # image: freqtradeorg/freqtrade:develop\n    # Use plotting image\n    # image: freqtradeorg/freqtrade:develop_plot\n    # Build step - only needed when additional dependencies are needed\n    # build:\n    #   context: .\n    #   dockerfile: \"./docker/Dockerfile.custom\"\n    restart: always\n    container_name: freqtrade2\n    volumes:\n      - \"./user_data:/freqtrade/user_data\"\n    # Expose api on port 8080 (localhost only)\n    # Please read the https://www.freqtrade.io/en/latest/rest-api/ documentation\n    # before enabling this.\n    ports:\n      - \"127.0.0.1:8081:8080\"\n    # Default command used when running `docker compose up`\n    command: >\n      trade\n      --logfile /freqtrade/user_data/logs/freqtrade2.log\n      --db-url sqlite:////freqtrade/user_data/tradesv3_freqtrade2.sqlite\n      --config /freqtrade/user_data/config.json\n      --config /freqtrade/user_data/config.freqtrade2.json\n      --strategy SampleStrategy\n
You can use whatever naming convention you want, freqtrade1 and 2 are arbitrary. Note, that you will need to use different database files, port mappings and telegram configurations for each instance, as mentioned above.

"},{"location":"advanced-setup/#configure-the-bot-running-as-a-systemd-service","title":"Configure the bot running as a systemd service","text":"

Copy the freqtrade.service file to your systemd user directory (usually ~/.config/systemd/user) and update WorkingDirectory and ExecStart to match your setup.

Note

Certain systems (like Raspbian) don't load service unit files from the user directory. In this case, copy freqtrade.service into /etc/systemd/user/ (requires superuser permissions).

After that you can start the daemon with:

systemctl --user start freqtrade\n

For this to be persistent (run when user is logged out) you'll need to enable linger for your freqtrade user.

sudo loginctl enable-linger \"$USER\"\n

If you run the bot as a service, you can use systemd service manager as a software watchdog monitoring freqtrade bot state and restarting it in the case of failures. If the internals.sd_notify parameter is set to true in the configuration or the --sd-notify command line option is used, the bot will send keep-alive ping messages to systemd using the sd_notify (systemd notifications) protocol and will also tell systemd its current state (Running or Stopped) when it changes.

The freqtrade.service.watchdog file contains an example of the service unit configuration file which uses systemd as the watchdog.

Note

The sd_notify communication between the bot and the systemd service manager will not work if the bot runs in a Docker container.

"},{"location":"advanced-setup/#advanced-logging","title":"Advanced Logging","text":"

On many Linux systems the bot can be configured to send its log messages to syslog or journald system services. Logging to a remote syslog server is also available on Windows. The special values for the --logfile command line option can be used for this.

"},{"location":"advanced-setup/#logging-to-syslog","title":"Logging to syslog","text":"

To send Freqtrade log messages to a local or remote syslog service use the --logfile command line option with the value in the following format:

  • --logfile syslog:<syslog_address> -- send log messages to syslog service using the <syslog_address> as the syslog address.

The syslog address can be either a Unix domain socket (socket filename) or a UDP socket specification, consisting of IP address and UDP port, separated by the : character.

So, the following are the examples of possible usages:

  • --logfile syslog:/dev/log -- log to syslog (rsyslog) using the /dev/log socket, suitable for most systems.
  • --logfile syslog -- same as above, the shortcut for /dev/log.
  • --logfile syslog:/var/run/syslog -- log to syslog (rsyslog) using the /var/run/syslog socket. Use this on MacOS.
  • --logfile syslog:localhost:514 -- log to local syslog using UDP socket, if it listens on port 514.
  • --logfile syslog:<ip>:514 -- log to remote syslog at IP address and port 514. This may be used on Windows for remote logging to an external syslog server.

Log messages are send to syslog with the user facility. So you can see them with the following commands:

  • tail -f /var/log/user, or
  • install a comprehensive graphical viewer (for instance, 'Log File Viewer' for Ubuntu).

On many systems syslog (rsyslog) fetches data from journald (and vice versa), so both --logfile syslog or --logfile journald can be used and the messages be viewed with both journalctl and a syslog viewer utility. You can combine this in any way which suites you better.

For rsyslog the messages from the bot can be redirected into a separate dedicated log file. To achieve this, add

if $programname startswith \"freqtrade\" then -/var/log/freqtrade.log\n

to one of the rsyslog configuration files, for example at the end of the /etc/rsyslog.d/50-default.conf.

For syslog (rsyslog), the reduction mode can be switched on. This will reduce the number of repeating messages. For instance, multiple bot Heartbeat messages will be reduced to a single message when nothing else happens with the bot. To achieve this, set in /etc/rsyslog.conf:

# Filter duplicated messages\n$RepeatedMsgReduction on\n
"},{"location":"advanced-setup/#logging-to-journald","title":"Logging to journald","text":"

This needs the cysystemd python package installed as dependency (pip install cysystemd), which is not available on Windows. Hence, the whole journald logging functionality is not available for a bot running on Windows.

To send Freqtrade log messages to journald system service use the --logfile command line option with the value in the following format:

  • --logfile journald -- send log messages to journald.

Log messages are send to journald with the user facility. So you can see them with the following commands:

  • journalctl -f -- shows Freqtrade log messages sent to journald along with other log messages fetched by journald.
  • journalctl -f -u freqtrade.service -- this command can be used when the bot is run as a systemd service.

There are many other options in the journalctl utility to filter the messages, see manual pages for this utility.

On many systems syslog (rsyslog) fetches data from journald (and vice versa), so both --logfile syslog or --logfile journald can be used and the messages be viewed with both journalctl and a syslog viewer utility. You can combine this in any way which suites you better.

"},{"location":"backtesting/","title":"Backtesting","text":"

This page explains how to validate your strategy performance by using Backtesting.

Backtesting requires historic data to be available. To learn how to get data for the pairs and exchange you're interested in, head over to the Data Downloading section of the documentation.

"},{"location":"backtesting/#backtesting-command-reference","title":"Backtesting command reference","text":"
usage: freqtrade backtesting [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                             [-d PATH] [--userdir PATH] [-s NAME]\n                             [--strategy-path PATH] [-i TIMEFRAME]\n                             [--timerange TIMERANGE]\n                             [--data-format-ohlcv {json,jsongz,hdf5}]\n                             [--max-open-trades INT]\n                             [--stake-amount STAKE_AMOUNT] [--fee FLOAT]\n                             [-p PAIRS [PAIRS ...]] [--eps] [--dmmp]\n                             [--enable-protections]\n                             [--dry-run-wallet DRY_RUN_WALLET]\n                             [--timeframe-detail TIMEFRAME_DETAIL]\n                             [--strategy-list STRATEGY_LIST [STRATEGY_LIST ...]]\n                             [--export {none,trades,signals}]\n                             [--export-filename PATH]\n                             [--breakdown {day,week,month} [{day,week,month} ...]]\n                             [--cache {none,day,week,month}]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  -i TIMEFRAME, --timeframe TIMEFRAME\n                        Specify timeframe (`1m`, `5m`, `30m`, `1h`, `1d`).\n  --timerange TIMERANGE\n                        Specify what timerange of data to use.\n  --data-format-ohlcv {json,jsongz,hdf5}\n                        Storage format for downloaded candle (OHLCV) data.\n                        (default: `json`).\n  --max-open-trades INT\n                        Override the value of the `max_open_trades`\n                        configuration setting.\n  --stake-amount STAKE_AMOUNT\n                        Override the value of the `stake_amount` configuration\n                        setting.\n  --fee FLOAT           Specify fee ratio. Will be applied twice (on trade\n                        entry and exit).\n  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]\n                        Limit command to these pairs. Pairs are space-\n                        separated.\n  --eps, --enable-position-stacking\n                        Allow buying the same pair multiple times (position\n                        stacking).\n  --dmmp, --disable-max-market-positions\n                        Disable applying `max_open_trades` during backtest\n                        (same as setting `max_open_trades` to a very high\n                        number).\n  --enable-protections, --enableprotections\n                        Enable protections for backtesting.Will slow\n                        backtesting down by a considerable amount, but will\n                        include configured protections\n  --dry-run-wallet DRY_RUN_WALLET, --starting-balance DRY_RUN_WALLET\n                        Starting balance, used for backtesting / hyperopt and\n                        dry-runs.\n  --timeframe-detail TIMEFRAME_DETAIL\n                        Specify detail timeframe for backtesting (`1m`, `5m`,\n                        `30m`, `1h`, `1d`).\n  --strategy-list STRATEGY_LIST [STRATEGY_LIST ...]\n                        Provide a space-separated list of strategies to\n                        backtest. Please note that timeframe needs to be set\n                        either in config or via command line. When using this\n                        together with `--export trades`, the strategy-name is\n                        injected into the filename (so `backtest-data.json`\n                        becomes `backtest-data-SampleStrategy.json`\n  --export {none,trades,signals}\n                        Export backtest results (default: trades).\n  --export-filename PATH, --backtest-filename PATH\n                        Use this filename for backtest results.Requires\n                        `--export` to be set as well. Example: `--export-filen\n                        ame=user_data/backtest_results/backtest_today.json`\n  --breakdown {day,week,month} [{day,week,month} ...]\n                        Show backtesting breakdown per [day, week, month].\n  --cache {none,day,week,month}\n                        Load a cached backtest result no older than specified\n                        age (default: day).\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n\nStrategy arguments:\n  -s NAME, --strategy NAME\n                        Specify strategy class name which will be used by the\n                        bot.\n  --strategy-path PATH  Specify additional strategy lookup path.\n
"},{"location":"backtesting/#test-your-strategy-with-backtesting","title":"Test your strategy with Backtesting","text":"

Now you have good Entry and exit strategies and some historic data, you want to test it against real data. This is what we call backtesting.

Backtesting will use the crypto-currencies (pairs) from your config file and load historical candle (OHLCV) data from user_data/data/<exchange> by default. If no data is available for the exchange / pair / timeframe combination, backtesting will ask you to download them first using freqtrade download-data. For details on downloading, please refer to the Data Downloading section in the documentation.

The result of backtesting will confirm if your bot has better odds of making a profit than a loss.

All profit calculations include fees, and freqtrade will use the exchange's default fees for the calculation.

Using dynamic pairlists for backtesting

Using dynamic pairlists is possible (not all of the handlers are allowed to be used in backtest mode), however it relies on the current market conditions - which will not reflect the historic status of the pairlist. Also, when using pairlists other than StaticPairlist, reproducibility of backtesting-results cannot be guaranteed. Please read the pairlists documentation for more information.

To achieve reproducible results, best generate a pairlist via the test-pairlist command and use that as static pairlist.

Note

By default, Freqtrade will export backtesting results to user_data/backtest_results. The exported trades can be used for further analysis or can be used by the plotting sub-command (freqtrade plot-dataframe) in the scripts directory.

"},{"location":"backtesting/#starting-balance","title":"Starting balance","text":"

Backtesting will require a starting balance, which can be provided as --dry-run-wallet <balance> or --starting-balance <balance> command line argument, or via dry_run_wallet configuration setting. This amount must be higher than stake_amount, otherwise the bot will not be able to simulate any trade.

"},{"location":"backtesting/#dynamic-stake-amount","title":"Dynamic stake amount","text":"

Backtesting supports dynamic stake amount by configuring stake_amount as \"unlimited\", which will split the starting balance into max_open_trades pieces. Profits from early trades will result in subsequent higher stake amounts, resulting in compounding of profits over the backtesting period.

"},{"location":"backtesting/#example-backtesting-commands","title":"Example backtesting commands","text":"

With 5 min candle (OHLCV) data (per default)

freqtrade backtesting --strategy AwesomeStrategy\n

Where --strategy AwesomeStrategy / -s AwesomeStrategy refers to the class name of the strategy, which is within a python file in the user_data/strategies directory.

With 1 min candle (OHLCV) data

freqtrade backtesting --strategy AwesomeStrategy --timeframe 1m\n

Providing a custom starting balance of 1000 (in stake currency)

freqtrade backtesting --strategy AwesomeStrategy --dry-run-wallet 1000\n

Using a different on-disk historical candle (OHLCV) data source

Assume you downloaded the history data from the Bittrex exchange and kept it in the user_data/data/bittrex-20180101 directory. You can then use this data for backtesting as follows:

freqtrade backtesting --strategy AwesomeStrategy --datadir user_data/data/bittrex-20180101 

Comparing multiple Strategies

freqtrade backtesting --strategy-list SampleStrategy1 AwesomeStrategy --timeframe 5m\n

Where SampleStrategy1 and AwesomeStrategy refer to class names of strategies.

Prevent exporting trades to file

freqtrade backtesting --strategy backtesting --export none --config config.json 

Only use this if you're sure you'll not want to plot or analyze your results further.

Exporting trades to file specifying a custom filename

freqtrade backtesting --strategy backtesting --export trades --export-filename=backtest_samplestrategy.json\n

Please also read about the strategy startup period.

Supplying custom fee value

Sometimes your account has certain fee rebates (fee reductions starting with a certain account size or monthly volume), which are not visible to ccxt. To account for this in backtesting, you can use the --fee command line option to supply this value to backtesting. This fee must be a ratio, and will be applied twice (once for trade entry, and once for trade exit).

For example, if the commission fee per order is 0.1% (i.e., 0.001 written as ratio), then you would run backtesting as the following:

freqtrade backtesting --fee 0.001\n

Note

Only supply this option (or the corresponding configuration parameter) if you want to experiment with different fee values. By default, Backtesting fetches the default fee from the exchange pair/market info.

Running backtest with smaller test-set by using timerange

Use the --timerange argument to change how much of the test-set you want to use.

For example, running backtesting with the --timerange=20190501- option will use all available data starting with May 1st, 2019 from your input data.

freqtrade backtesting --timerange=20190501-\n

You can also specify particular date ranges.

The full timerange specification:

  • Use data until 2018/01/31: --timerange=-20180131
  • Use data since 2018/01/31: --timerange=20180131-
  • Use data since 2018/01/31 till 2018/03/01 : --timerange=20180131-20180301
  • Use data between POSIX / epoch timestamps 1527595200 1527618600: --timerange=1527595200-1527618600
"},{"location":"backtesting/#understand-the-backtesting-result","title":"Understand the backtesting result","text":"

The most important in the backtesting is to understand the result.

A backtesting result will look like that:

========================================================= BACKTESTING REPORT =========================================================\n| Pair     | Entries |   Avg Profit % |   Cum Profit % |   Tot Profit BTC |   Tot Profit % | Avg Duration |  Wins Draws Loss   Win%  |\n|:---------|--------:|---------------:|---------------:|-----------------:|---------------:|:-------------|-------------------------:|\n| ADA/BTC  |      35 |          -0.11 |          -3.88 |      -0.00019428 |          -1.94 | 4:35:00      |    14     0    21   40.0 |\n| ARK/BTC  |      11 |          -0.41 |          -4.52 |      -0.00022647 |          -2.26 | 2:03:00      |     3     0     8   27.3 |\n| BTS/BTC  |      32 |           0.31 |           9.78 |       0.00048938 |           4.89 | 5:05:00      |    18     0    14   56.2 |\n| DASH/BTC |      13 |          -0.08 |          -1.07 |      -0.00005343 |          -0.53 | 4:39:00      |     6     0     7   46.2 |\n| ENG/BTC  |      18 |           1.36 |          24.54 |       0.00122807 |          12.27 | 2:50:00      |     8     0    10   44.4 |\n| EOS/BTC  |      36 |           0.08 |           3.06 |       0.00015304 |           1.53 | 3:34:00      |    16     0    20   44.4 |\n| ETC/BTC  |      26 |           0.37 |           9.51 |       0.00047576 |           4.75 | 6:14:00      |    11     0    15   42.3 |\n| ETH/BTC  |      33 |           0.30 |           9.96 |       0.00049856 |           4.98 | 7:31:00      |    16     0    17   48.5 |\n| IOTA/BTC |      32 |           0.03 |           1.09 |       0.00005444 |           0.54 | 3:12:00      |    14     0    18   43.8 |\n| LSK/BTC  |      15 |           1.75 |          26.26 |       0.00131413 |          13.13 | 2:58:00      |     6     0     9   40.0 |\n| LTC/BTC  |      32 |          -0.04 |          -1.38 |      -0.00006886 |          -0.69 | 4:49:00      |    11     0    21   34.4 |\n| NANO/BTC |      17 |           1.26 |          21.39 |       0.00107058 |          10.70 | 1:55:00      |    10     0     7   58.5 |\n| NEO/BTC  |      23 |           0.82 |          18.97 |       0.00094936 |           9.48 | 2:59:00      |    10     0    13   43.5 |\n| REQ/BTC  |       9 |           1.17 |          10.54 |       0.00052734 |           5.27 | 3:47:00      |     4     0     5   44.4 |\n| XLM/BTC  |      16 |           1.22 |          19.54 |       0.00097800 |           9.77 | 3:15:00      |     7     0     9   43.8 |\n| XMR/BTC  |      23 |          -0.18 |          -4.13 |      -0.00020696 |          -2.07 | 5:30:00      |    12     0    11   52.2 |\n| XRP/BTC  |      35 |           0.66 |          22.96 |       0.00114897 |          11.48 | 3:49:00      |    12     0    23   34.3 |\n| ZEC/BTC  |      22 |          -0.46 |         -10.18 |      -0.00050971 |          -5.09 | 2:22:00      |     7     0    15   31.8 |\n| TOTAL    |     429 |           0.36 |         152.41 |       0.00762792 |          76.20 | 4:12:00      |   186     0   243   43.4 |\n====================================================== LEFT OPEN TRADES REPORT ======================================================\n| Pair     |  Entries |   Avg Profit % |   Cum Profit % |   Tot Profit BTC |   Tot Profit % | Avg Duration   |  Win Draw Loss Win% |\n|:---------|---------:|---------------:|---------------:|-----------------:|---------------:|:---------------|--------------------:|\n| ADA/BTC  |        1 |           0.89 |           0.89 |       0.00004434 |           0.44 | 6:00:00        |    1    0    0  100 |\n| LTC/BTC  |        1 |           0.68 |           0.68 |       0.00003421 |           0.34 | 2:00:00        |    1    0    0  100 |\n| TOTAL    |        2 |           0.78 |           1.57 |       0.00007855 |           0.78 | 4:00:00        |    2    0    0  100 |\n==================== EXIT REASON STATS ====================\n| Exit Reason        |   Exits |  Wins |  Draws |  Losses |\n|:-------------------|--------:|------:|-------:|--------:|\n| trailing_stop_loss |     205 |   150 |      0 |      55 |\n| stop_loss          |     166 |     0 |      0 |     166 |\n| exit_signal        |      56 |    36 |      0 |      20 |\n| force_exit         |       2 |     0 |      0 |       2 |\n\n================== SUMMARY METRICS ==================\n| Metric                      | Value               |\n|-----------------------------+---------------------|\n| Backtesting from            | 2019-01-01 00:00:00 |\n| Backtesting to              | 2019-05-01 00:00:00 |\n| Max open trades             | 3                   |\n|                             |                     |\n| Total/Daily Avg Trades      | 429 / 3.575         |\n| Starting balance            | 0.01000000 BTC      |\n| Final balance               | 0.01762792 BTC      |\n| Absolute profit             | 0.00762792 BTC      |\n| Total profit %              | 76.2%               |\n| CAGR %                      | 460.87%             |\n| Sortino                     | 1.88                |\n| Sharpe                      | 2.97                |\n| Calmar                      | 6.29                |\n| Profit factor               | 1.11                |\n| Expectancy                  | -0.15               |\n| Avg. stake amount           | 0.001      BTC      |\n| Total trade volume          | 0.429      BTC      |\n|                             |                     |\n| Long / Short                | 352 / 77            |\n| Total profit Long %         | 1250.58%            |\n| Total profit Short %        | -15.02%             |\n| Absolute profit Long        | 0.00838792 BTC      |\n| Absolute profit Short       | -0.00076 BTC        |\n|                             |                     |\n| Best Pair                   | LSK/BTC 26.26%      |\n| Worst Pair                  | ZEC/BTC -10.18%     |\n| Best Trade                  | LSK/BTC 4.25%       |\n| Worst Trade                 | ZEC/BTC -10.25%     |\n| Best day                    | 0.00076 BTC         |\n| Worst day                   | -0.00036 BTC        |\n| Days win/draw/lose          | 12 / 82 / 25        |\n| Avg. Duration Winners       | 4:23:00             |\n| Avg. Duration Loser         | 6:55:00             |\n| Rejected Entry signals      | 3089                |\n| Entry/Exit Timeouts         | 0 / 0               |\n| Canceled Trade Entries      | 34                  |\n| Canceled Entry Orders       | 123                 |\n| Replaced Entry Orders       | 89                  |\n|                             |                     |\n| Min balance                 | 0.00945123 BTC      |\n| Max balance                 | 0.01846651 BTC      |\n| Max % of account underwater | 25.19%              |\n| Absolute Drawdown (Account) | 13.33%              |\n| Drawdown                    | 0.0015 BTC          |\n| Drawdown high               | 0.0013 BTC          |\n| Drawdown low                | -0.0002 BTC         |\n| Drawdown Start              | 2019-02-15 14:10:00 |\n| Drawdown End                | 2019-04-11 18:15:00 |\n| Market change               | -5.88%              |\n=====================================================\n
"},{"location":"backtesting/#backtesting-report-table","title":"Backtesting report table","text":"

The 1st table contains all trades the bot made, including \"left open trades\".

The last line will give you the overall performance of your strategy, here:

| TOTAL    |    429 |           0.36 |         152.41 |       0.00762792 |          76.20 | 4:12:00      |   186     0   243   43.4 |\n

The bot has made 429 trades for an average duration of 4:12:00, with a performance of 76.20% (profit), that means it has earned a total of 0.00762792 BTC starting with a capital of 0.01 BTC.

The column Avg Profit % shows the average profit for all trades made while the column Cum Profit % sums up all the profits/losses. The column Tot Profit % shows instead the total profit % in relation to the starting balance. In the above results, we have a starting balance of 0.01 BTC and an absolute profit of 0.00762792 BTC - so the Tot Profit % will be (0.00762792 / 0.01) * 100 ~= 76.2%.

Your strategy performance is influenced by your entry strategy, your exit strategy, and also by the minimal_roi and stop_loss you have set.

For example, if your minimal_roi is only \"0\": 0.01 you cannot expect the bot to make more profit than 1% (because it will exit every time a trade reaches 1%).

\"minimal_roi\": {\n\"0\":  0.01\n},\n

On the other hand, if you set a too high minimal_roi like \"0\": 0.55 (55%), there is almost no chance that the bot will ever reach this profit. Hence, keep in mind that your performance is an integral mix of all different elements of the strategy, your configuration, and the crypto-currency pairs you have set up.

"},{"location":"backtesting/#exit-reasons-table","title":"Exit reasons table","text":"

The 2nd table contains a recap of exit reasons. This table can tell you which area needs some additional work (e.g. all or many of the exit_signal trades are losses, so you should work on improving the exit signal, or consider disabling it).

"},{"location":"backtesting/#left-open-trades-table","title":"Left open trades table","text":"

The 3rd table contains all trades the bot had to force_exit at the end of the backtesting period to present you the full picture. This is necessary to simulate realistic behavior, since the backtest period has to end at some point, while realistically, you could leave the bot running forever. These trades are also included in the first table, but are also shown separately in this table for clarity.

"},{"location":"backtesting/#summary-metrics","title":"Summary metrics","text":"

The last element of the backtest report is the summary metrics table. It contains some useful key metrics about performance of your strategy on backtesting data.

================== SUMMARY METRICS ==================\n| Metric                      | Value               |\n|-----------------------------+---------------------|\n| Backtesting from            | 2019-01-01 00:00:00 |\n| Backtesting to              | 2019-05-01 00:00:00 |\n| Max open trades             | 3                   |\n|                             |                     |\n| Total/Daily Avg Trades      | 429 / 3.575         |\n| Starting balance            | 0.01000000 BTC      |\n| Final balance               | 0.01762792 BTC      |\n| Absolute profit             | 0.00762792 BTC      |\n| Total profit %              | 76.2%               |\n| CAGR %                      | 460.87%             |\n| Sortino                     | 1.88                |\n| Sharpe                      | 2.97                |\n| Calmar                      | 6.29                |\n| Profit factor               | 1.11                |\n| Expectancy                  | -0.15               |\n| Avg. stake amount           | 0.001      BTC      |\n| Total trade volume          | 0.429      BTC      |\n|                             |                     |\n| Long / Short                | 352 / 77            |\n| Total profit Long %         | 1250.58%            |\n| Total profit Short %        | -15.02%             |\n| Absolute profit Long        | 0.00838792 BTC      |\n| Absolute profit Short       | -0.00076 BTC        |\n|                             |                     |\n| Best Pair                   | LSK/BTC 26.26%      |\n| Worst Pair                  | ZEC/BTC -10.18%     |\n| Best Trade                  | LSK/BTC 4.25%       |\n| Worst Trade                 | ZEC/BTC -10.25%     |\n| Best day                    | 0.00076 BTC         |\n| Worst day                   | -0.00036 BTC        |\n| Days win/draw/lose          | 12 / 82 / 25        |\n| Avg. Duration Winners       | 4:23:00             |\n| Avg. Duration Loser         | 6:55:00             |\n| Rejected Entry signals      | 3089                |\n| Entry/Exit Timeouts         | 0 / 0               |\n| Canceled Trade Entries      | 34                  |\n| Canceled Entry Orders       | 123                 |\n| Replaced Entry Orders       | 89                  |\n|                             |                     |\n| Min balance                 | 0.00945123 BTC      |\n| Max balance                 | 0.01846651 BTC      |\n| Max % of account underwater | 25.19%              |\n| Absolute Drawdown (Account) | 13.33%              |\n| Drawdown                    | 0.0015 BTC          |\n| Drawdown high               | 0.0013 BTC          |\n| Drawdown low                | -0.0002 BTC         |\n| Drawdown Start              | 2019-02-15 14:10:00 |\n| Drawdown End                | 2019-04-11 18:15:00 |\n| Market change               | -5.88%              |\n=====================================================\n
  • Backtesting from / Backtesting to: Backtesting range (usually defined with the --timerange option).
  • Max open trades: Setting of max_open_trades (or --max-open-trades) - or number of pairs in the pairlist (whatever is lower).
  • Total/Daily Avg Trades: Identical to the total trades of the backtest output table / Total trades divided by the backtesting duration in days (this will give you information about how many trades to expect from the strategy).
  • Starting balance: Start balance - as given by dry-run-wallet (config or command line).
  • Final balance: Final balance - starting balance + absolute profit.
  • Absolute profit: Profit made in stake currency.
  • Total profit %: Total profit. Aligned to the TOTAL row's Tot Profit % from the first table. Calculated as (End capital \u2212 Starting capital) / Starting capital.
  • CAGR %: Compound annual growth rate.
  • Sortino: Annualized Sortino ratio.
  • Sharpe: Annualized Sharpe ratio.
  • Calmar: Annualized Calmar ratio.
  • Profit factor: profit / loss.
  • Avg. stake amount: Average stake amount, either stake_amount or the average when using dynamic stake amount.
  • Total trade volume: Volume generated on the exchange to reach the above profit.
  • Best Pair / Worst Pair: Best and worst performing pair, and it's corresponding Cum Profit %.
  • Best Trade / Worst Trade: Biggest single winning trade and biggest single losing trade.
  • Best day / Worst day: Best and worst day based on daily profit.
  • Days win/draw/lose: Winning / Losing days (draws are usually days without closed trade).
  • Avg. Duration Winners / Avg. Duration Loser: Average durations for winning and losing trades.
  • Rejected Entry signals: Trade entry signals that could not be acted upon due to max_open_trades being reached.
  • Entry/Exit Timeouts: Entry/exit orders which did not fill (only applicable if custom pricing is used).
  • Canceled Trade Entries: Number of trades that have been canceled by user request via adjust_entry_price.
  • Canceled Entry Orders: Number of entry orders that have been canceled by user request via adjust_entry_price.
  • Replaced Entry Orders: Number of entry orders that have been replaced by user request via adjust_entry_price.
  • Min balance / Max balance: Lowest and Highest Wallet balance during the backtest period.
  • Max % of account underwater: Maximum percentage your account has decreased from the top since the simulation started. Calculated as the maximum of (Max Balance - Current Balance) / (Max Balance).
  • Absolute Drawdown (Account): Maximum Account Drawdown experienced. Calculated as (Absolute Drawdown) / (DrawdownHigh + startingBalance).
  • Drawdown: Maximum, absolute drawdown experienced. Difference between Drawdown High and Subsequent Low point.
  • Drawdown high / Drawdown low: Profit at the beginning and end of the largest drawdown period. A negative low value means initial capital lost.
  • Drawdown Start / Drawdown End: Start and end datetime for this largest drawdown (can also be visualized via the plot-dataframe sub-command).
  • Market change: Change of the market during the backtest period. Calculated as average of all pairs changes from the first to the last candle using the \"close\" column.
  • Long / Short: Split long/short values (Only shown when short trades were made).
  • Total profit Long % / Absolute profit Long: Profit long trades only (Only shown when short trades were made).
  • Total profit Short % / Absolute profit Short: Profit short trades only (Only shown when short trades were made).
"},{"location":"backtesting/#daily-weekly-monthly-breakdown","title":"Daily / Weekly / Monthly breakdown","text":"

You can get an overview over daily / weekly or monthly results by using the --breakdown <> switch.

To visualize daily and weekly breakdowns, you can use the following:

freqtrade backtesting --strategy MyAwesomeStrategy --breakdown day week\n
======================== DAY BREAKDOWN =========================\n|        Day |   Tot Profit USDT |   Wins |   Draws |   Losses |\n|------------+-------------------+--------+---------+----------|\n| 03/07/2021 |           200.0   |      2 |       0 |        0 |\n| 04/07/2021 |           -50.31  |      0 |       0 |        2 |\n| 05/07/2021 |           220.611 |      3 |       2 |        0 |\n| 06/07/2021 |           150.974 |      3 |       0 |        2 |\n| 07/07/2021 |           -70.193 |      1 |       0 |        2 |\n| 08/07/2021 |           212.413 |      2 |       0 |        3 |\n

The output will show a table containing the realized absolute Profit (in stake currency) for the given timeperiod, as well as wins, draws and losses that materialized (closed) on this day. Below that there will be a second table for the summarized values of weeks indicated by the date of the closing Sunday. The same would apply to a monthly breakdown indicated by the last day of the month.

"},{"location":"backtesting/#backtest-result-caching","title":"Backtest result caching","text":"

To save time, by default backtest will reuse a cached result from within the last day when the backtested strategy and config match that of a previous backtest. To force a new backtest despite existing result for an identical run specify --cache none parameter.

Warning

Caching is automatically disabled for open-ended timeranges (--timerange 20210101-), as freqtrade cannot ensure reliably that the underlying data didn't change. It can also use cached results where it shouldn't if the original backtest had missing data at the end, which was fixed by downloading more data. In this instance, please use --cache none once to force a fresh backtest.

"},{"location":"backtesting/#further-backtest-result-analysis","title":"Further backtest-result analysis","text":"

To further analyze your backtest results, you can export the trades. You can then load the trades to perform further analysis as shown in the data analysis backtesting section.

"},{"location":"backtesting/#assumptions-made-by-backtesting","title":"Assumptions made by backtesting","text":"

Since backtesting lacks some detailed information about what happens within a candle, it needs to take a few assumptions:

  • Exchange trading limits are respected
  • Entries happen at open-price
  • All orders are filled at the requested price (no slippage, no unfilled orders)
  • Exit-signal exits happen at open-price of the consecutive candle
  • Exit-signal is favored over Stoploss, because exit-signals are assumed to trigger on candle's open
  • ROI
    • exits are compared to high - but the ROI value is used (e.g. ROI = 2%, high=5% - so the exit will be at 2%)
    • exits are never \"below the candle\", so a ROI of 2% may result in a exit at 2.4% if low was at 2.4% profit
    • Force-exits caused by <N>=-1 ROI entries use low as exit value, unless N falls on the candle open (e.g. 120: -1 for 1h candles)
  • Stoploss exits happen exactly at stoploss price, even if low was lower, but the loss will be 2 * fees higher than the stoploss price
  • Stoploss is evaluated before ROI within one candle. So you can often see more trades with the stoploss exit reason comparing to the results obtained with the same strategy in the Dry Run/Live Trade modes
  • Low happens before high for stoploss, protecting capital first
  • Trailing stoploss
    • Trailing Stoploss is only adjusted if it's below the candle's low (otherwise it would be triggered)
    • On trade entry candles that trigger trailing stoploss, the \"minimum offset\" (stop_positive_offset) is assumed (instead of high) - and the stop is calculated from this point. This rule is NOT applicable to custom-stoploss scenarios, since there's no information about the stoploss logic available.
    • High happens first - adjusting stoploss
    • Low uses the adjusted stoploss (so exits with large high-low difference are backtested correctly)
    • ROI applies before trailing-stop, ensuring profits are \"top-capped\" at ROI if both ROI and trailing stop applies
  • Exit-reason does not explain if a trade was positive or negative, just what triggered the exit (this can look odd if negative ROI values are used)
  • Evaluation sequence (if multiple signals happen on the same candle)
    • Exit-signal
    • Stoploss
    • ROI
    • Trailing stoploss

Taking these assumptions, backtesting tries to mirror real trading as closely as possible. However, backtesting will never replace running a strategy in dry-run mode. Also, keep in mind that past results don't guarantee future success.

In addition to the above assumptions, strategy authors should carefully read the Common Mistakes section, to avoid using data in backtesting which is not available in real market conditions.

"},{"location":"backtesting/#trading-limits-in-backtesting","title":"Trading limits in backtesting","text":"

Exchanges have certain trading limits, like minimum (and maximum) base currency, or minimum/maximum stake (quote) currency. These limits are usually listed in the exchange documentation as \"trading rules\" or similar and can be quite different between different pairs.

Backtesting (as well as live and dry-run) does honor these limits, and will ensure that a stoploss can be placed below this value - so the value will be slightly higher than what the exchange specifies. Freqtrade has however no information about historic limits.

This can lead to situations where trading-limits are inflated by using a historic price, resulting in minimum amounts > 50$.

For example:

BTC minimum tradable amount is 0.001. BTC trades at 22.000$ today (0.001 BTC is related to this) - but the backtesting period includes prices as high as 50.000$. Today's minimum would be 0.001 * 22_000 - or 22$. However the limit could also be 50$ - based on 0.001 * 50_000 in some historic setting.

"},{"location":"backtesting/#trading-precision-limits","title":"Trading precision limits","text":"

Most exchanges pose precision limits on both price and amounts, so you cannot buy 1.0020401 of a pair, or at a price of 1.24567123123. Instead, these prices and amounts will be rounded or truncated (based on the exchange definition) to the defined trading precision. The above values may for example be rounded to an amount of 1.002, and a price of 1.24567.

These precision values are based on current exchange limits (as described in the above section), as historic precision limits are not available.

"},{"location":"backtesting/#improved-backtest-accuracy","title":"Improved backtest accuracy","text":"

One big limitation of backtesting is it's inability to know how prices moved intra-candle (was high before close, or viceversa?). So assuming you run backtesting with a 1h timeframe, there will be 4 prices for that candle (Open, High, Low, Close).

While backtesting does take some assumptions (read above) about this - this can never be perfect, and will always be biased in one way or the other. To mitigate this, freqtrade can use a lower (faster) timeframe to simulate intra-candle movements.

To utilize this, you can append --timeframe-detail 5m to your regular backtesting command.

freqtrade backtesting --strategy AwesomeStrategy --timeframe 1h --timeframe-detail 5m\n

This will load 1h data as well as 5m data for the timeframe. The strategy will be analyzed with the 1h timeframe, and Entry orders will only be placed at the main timeframe, however Order fills and exit signals will be evaluated at the 5m candle, simulating intra-candle movements.

All callback functions (custom_exit(), custom_stoploss(), ... ) will be running for each 5m candle once the trade is opened (so 12 times in the above example of 1h timeframe, and 5m detailed timeframe).

--timeframe-detail must be smaller than the original timeframe, otherwise backtesting will fail to start.

Obviously this will require more memory (5m data is bigger than 1h data), and will also impact runtime (depending on the amount of trades and trade durations). Also, data must be available / downloaded already.

Tip

You can use this function as the last part of strategy development, to ensure your strategy is not exploiting one of the backtesting assumptions. Strategies that perform similarly well with this mode have a good chance to perform well in dry/live modes too (although only forward-testing (dry-mode) can really confirm a strategy).

"},{"location":"backtesting/#backtesting-multiple-strategies","title":"Backtesting multiple strategies","text":"

To compare multiple strategies, a list of Strategies can be provided to backtesting.

This is limited to 1 timeframe value per run. However, data is only loaded once from disk so if you have multiple strategies you'd like to compare, this will give a nice runtime boost.

All listed Strategies need to be in the same directory.

freqtrade backtesting --timerange 20180401-20180410 --timeframe 5m --strategy-list Strategy001 Strategy002 --export trades\n

This will save the results to user_data/backtest_results/backtest-result-<strategy>.json, injecting the strategy-name into the target filename. There will be an additional table comparing win/losses of the different strategies (identical to the \"Total\" row in the first table). Detailed output for all strategies one after the other will be available, so make sure to scroll up to see the details per strategy.

=========================================================== STRATEGY SUMMARY ===========================================================================\n| Strategy    |  Entries |   Avg Profit % |   Cum Profit % |   Tot Profit BTC |   Tot Profit % | Avg Duration   |  Wins |  Draws | Losses | Drawdown % |\n|:------------|---------:|---------------:|---------------:|-----------------:|---------------:|:---------------|------:|-------:|-------:|-----------:|\n| Strategy1   |      429 |           0.36 |         152.41 |       0.00762792 |          76.20 | 4:12:00        |   186 |      0 |    243 |       45.2 |\n| Strategy2   |     1487 |          -0.13 |        -197.58 |      -0.00988917 |         -98.79 | 4:43:00        |   662 |      0 |    825 |     241.68 |\n
"},{"location":"backtesting/#next-step","title":"Next step","text":"

Great, your strategy is profitable. What if the bot can give your the optimal parameters to use for your strategy? Your next step is to learn how to find optimal parameters with Hyperopt

"},{"location":"bot-basics/","title":"Freqtrade basics","text":"

This page provides you some basic concepts on how Freqtrade works and operates.

"},{"location":"bot-basics/#freqtrade-terminology","title":"Freqtrade terminology","text":"
  • Strategy: Your trading strategy, telling the bot what to do.
  • Trade: Open position.
  • Open Order: Order which is currently placed on the exchange, and is not yet complete.
  • Pair: Tradable pair, usually in the format of Base/Quote (e.g. XRP/USDT).
  • Timeframe: Candle length to use (e.g. \"5m\", \"1h\", ...).
  • Indicators: Technical indicators (SMA, EMA, RSI, ...).
  • Limit order: Limit orders which execute at the defined limit price or better.
  • Market order: Guaranteed to fill, may move price depending on the order size.
  • Current Profit: Currently pending (unrealized) profit for this trade. This is mainly used throughout the bot and UI.
  • Realized Profit: Already realized profit. Only relevant in combination with partial exits - which also explains the calculation logic for this.
  • Total Profit: Combined realized and unrealized profit. The relative number (%) is calculated against the total investment in this trade.
"},{"location":"bot-basics/#fee-handling","title":"Fee handling","text":"

All profit calculations of Freqtrade include fees. For Backtesting / Hyperopt / Dry-run modes, the exchange default fee is used (lowest tier on the exchange). For live operations, fees are used as applied by the exchange (this includes BNB rebates etc.).

"},{"location":"bot-basics/#bot-execution-logic","title":"Bot execution logic","text":"

Starting freqtrade in dry-run or live mode (using freqtrade trade) will start the bot and start the bot iteration loop. This will also run the bot_start() callback.

By default, the bot loop runs every few seconds (internals.process_throttle_secs) and performs the following actions:

  • Fetch open trades from persistence.
  • Calculate current list of tradable pairs.
  • Download OHLCV data for the pairlist including all informative pairs This step is only executed once per Candle to avoid unnecessary network traffic.
  • Call bot_loop_start() strategy callback.
  • Analyze strategy per pair.
    • Call populate_indicators()
    • Call populate_entry_trend()
    • Call populate_exit_trend()
  • Check timeouts for open orders.
    • Calls check_entry_timeout() strategy callback for open entry orders.
    • Calls check_exit_timeout() strategy callback for open exit orders.
    • Calls adjust_entry_price() strategy callback for open entry orders.
  • Verifies existing positions and eventually places exit orders.
    • Considers stoploss, ROI and exit-signal, custom_exit() and custom_stoploss().
    • Determine exit-price based on exit_pricing configuration setting or by using the custom_exit_price() callback.
    • Before a exit order is placed, confirm_trade_exit() strategy callback is called.
  • Check position adjustments for open trades if enabled by calling adjust_trade_position() and place additional order if required.
  • Check if trade-slots are still available (if max_open_trades is reached).
  • Verifies entry signal trying to enter new positions.
    • Determine entry-price based on entry_pricing configuration setting, or by using the custom_entry_price() callback.
    • In Margin and Futures mode, leverage() strategy callback is called to determine the desired leverage.
    • Determine stake size by calling the custom_stake_amount() callback.
    • Before an entry order is placed, confirm_trade_entry() strategy callback is called.

This loop will be repeated again and again until the bot is stopped.

"},{"location":"bot-basics/#backtesting-hyperopt-execution-logic","title":"Backtesting / Hyperopt execution logic","text":"

backtesting or hyperopt do only part of the above logic, since most of the trading operations are fully simulated.

  • Load historic data for configured pairlist.
  • Calls bot_start() once.
  • Calculate indicators (calls populate_indicators() once per pair).
  • Calculate entry / exit signals (calls populate_entry_trend() and populate_exit_trend() once per pair).
  • Loops per candle simulating entry and exit points.
    • Calls bot_loop_start() strategy callback.
    • Check for Order timeouts, either via the unfilledtimeout configuration, or via check_entry_timeout() / check_exit_timeout() strategy callbacks.
    • Calls adjust_entry_price() strategy callback for open entry orders.
    • Check for trade entry signals (enter_long / enter_short columns).
    • Confirm trade entry / exits (calls confirm_trade_entry() and confirm_trade_exit() if implemented in the strategy).
    • Call custom_entry_price() (if implemented in the strategy) to determine entry price (Prices are moved to be within the opening candle).
    • In Margin and Futures mode, leverage() strategy callback is called to determine the desired leverage.
    • Determine stake size by calling the custom_stake_amount() callback.
    • Check position adjustments for open trades if enabled and call adjust_trade_position() to determine if an additional order is requested.
    • Call custom_stoploss() and custom_exit() to find custom exit points.
    • For exits based on exit-signal, custom-exit and partial exits: Call custom_exit_price() to determine exit price (Prices are moved to be within the closing candle).
  • Generate backtest report output

Note

Both Backtesting and Hyperopt include exchange default Fees in the calculation. Custom fees can be passed to backtesting / hyperopt by specifying the --fee argument.

Callback call frequency

Backtesting will call each callback at max. once per candle (--timeframe-detail modifies this behavior to once per detailed candle). Most callbacks will be called once per iteration in live (usually every ~5s) - which can cause backtesting mismatches.

"},{"location":"bot-usage/","title":"Start the bot","text":"

This page explains the different parameters of the bot and how to run it.

Note

If you've used setup.sh, don't forget to activate your virtual environment (source .env/bin/activate) before running freqtrade commands.

Up-to-date clock

The clock on the system running the bot must be accurate, synchronized to a NTP server frequently enough to avoid problems with communication to the exchanges.

"},{"location":"bot-usage/#bot-commands","title":"Bot commands","text":"
usage: freqtrade [-h] [-V]\n                 {trade,create-userdir,new-config,new-strategy,download-data,convert-data,convert-trade-data,list-data,backtesting,edge,hyperopt,hyperopt-list,hyperopt-show,list-exchanges,list-hyperopts,list-markets,list-pairs,list-strategies,list-timeframes,show-trades,test-pairlist,install-ui,plot-dataframe,plot-profit,webserver}\n                 ...\n\nFree, open source crypto trading bot\n\npositional arguments:\n  {trade,create-userdir,new-config,new-strategy,download-data,convert-data,convert-trade-data,list-data,backtesting,edge,hyperopt,hyperopt-list,hyperopt-show,list-exchanges,list-hyperopts,list-markets,list-pairs,list-strategies,list-timeframes,show-trades,test-pairlist,install-ui,plot-dataframe,plot-profit,webserver}\n    trade               Trade module.\n    create-userdir      Create user-data directory.\n    new-config          Create new config\n    new-strategy        Create new strategy\n    download-data       Download backtesting data.\n    convert-data        Convert candle (OHLCV) data from one format to\n                        another.\n    convert-trade-data  Convert trade data from one format to another.\n    list-data           List downloaded data.\n    backtesting         Backtesting module.\n    edge                Edge module.\n    hyperopt            Hyperopt module.\n    hyperopt-list       List Hyperopt results\n    hyperopt-show       Show details of Hyperopt results\n    list-exchanges      Print available exchanges.\n    list-hyperopts      Print available hyperopt classes.\n    list-markets        Print markets on exchange.\n    list-pairs          Print pairs on exchange.\n    list-strategies     Print available strategies.\n    list-timeframes     Print available timeframes for the exchange.\n    show-trades         Show trades.\n    test-pairlist       Test your pairlist configuration.\n    install-ui          Install FreqUI\n    plot-dataframe      Plot candles with indicators.\n    plot-profit         Generate plot showing profits.\n    webserver           Webserver module.\n\noptional arguments:\n  -h, --help            show this help message and exit\n  -V, --version         show program's version number and exit\n
"},{"location":"bot-usage/#bot-trading-commands","title":"Bot trading commands","text":"
usage: freqtrade trade [-h] [-v] [--logfile FILE] [-V] [-c PATH] [-d PATH]\n                       [--userdir PATH] [-s NAME] [--strategy-path PATH]\n                       [--db-url PATH] [--sd-notify] [--dry-run]\n                       [--dry-run-wallet DRY_RUN_WALLET]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --db-url PATH         Override trades database URL, this is useful in custom\n                        deployments (default: `sqlite:///tradesv3.sqlite` for\n                        Live Run mode, `sqlite:///tradesv3.dryrun.sqlite` for\n                        Dry Run).\n  --sd-notify           Notify systemd service manager.\n  --dry-run             Enforce dry-run for trading (removes Exchange secrets\n                        and simulates trades).\n  --dry-run-wallet DRY_RUN_WALLET, --starting-balance DRY_RUN_WALLET\n                        Starting balance, used for backtesting / hyperopt and\n                        dry-runs.\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n\nStrategy arguments:\n  -s NAME, --strategy NAME\n                        Specify strategy class name which will be used by the\n                        bot.\n  --strategy-path PATH  Specify additional strategy lookup path.\n
"},{"location":"bot-usage/#how-to-specify-which-configuration-file-be-used","title":"How to specify which configuration file be used?","text":"

The bot allows you to select which configuration file you want to use by means of the -c/--config command line option:

freqtrade trade -c path/far/far/away/config.json\n

Per default, the bot loads the config.json configuration file from the current working directory.

"},{"location":"bot-usage/#how-to-use-multiple-configuration-files","title":"How to use multiple configuration files?","text":"

The bot allows you to use multiple configuration files by specifying multiple -c/--config options in the command line. Configuration parameters defined in the latter configuration files override parameters with the same name defined in the previous configuration files specified in the command line earlier.

For example, you can make a separate configuration file with your key and secret for the Exchange you use for trading, specify default configuration file with empty key and secret values while running in the Dry Mode (which does not actually require them):

freqtrade trade -c ./config.json\n

and specify both configuration files when running in the normal Live Trade Mode:

freqtrade trade -c ./config.json -c path/to/secrets/keys.config.json\n

This could help you hide your private Exchange key and Exchange secret on you local machine by setting appropriate file permissions for the file which contains actual secrets and, additionally, prevent unintended disclosure of sensitive private data when you publish examples of your configuration in the project issues or in the Internet.

See more details on this technique with examples in the documentation page on configuration.

"},{"location":"bot-usage/#where-to-store-custom-data","title":"Where to store custom data","text":"

Freqtrade allows the creation of a user-data directory using freqtrade create-userdir --userdir someDirectory. This directory will look as follows:

user_data/\n\u251c\u2500\u2500 backtest_results\n\u251c\u2500\u2500 data\n\u251c\u2500\u2500 hyperopts\n\u251c\u2500\u2500 hyperopt_results\n\u251c\u2500\u2500 plot\n\u2514\u2500\u2500 strategies\n

You can add the entry \"user_data_dir\" setting to your configuration, to always point your bot to this directory. Alternatively, pass in --userdir to every command. The bot will fail to start if the directory does not exist, but will create necessary subdirectories.

This directory should contain your custom strategies, custom hyperopts and hyperopt loss functions, backtesting historical data (downloaded using either backtesting command or the download script) and plot outputs.

It is recommended to use version control to keep track of changes to your strategies.

"},{"location":"bot-usage/#how-to-use-strategy","title":"How to use --strategy?","text":"

This parameter will allow you to load your custom strategy class. To test the bot installation, you can use the SampleStrategy installed by the create-userdir subcommand (usually user_data/strategy/sample_strategy.py).

The bot will search your strategy file within user_data/strategies. To use other directories, please read the next section about --strategy-path.

To load a strategy, simply pass the class name (e.g.: CustomStrategy) in this parameter.

Example: In user_data/strategies you have a file my_awesome_strategy.py which has a strategy class called AwesomeStrategy to load it:

freqtrade trade --strategy AwesomeStrategy\n

If the bot does not find your strategy file, it will display in an error message the reason (File not found, or errors in your code).

Learn more about strategy file in Strategy Customization.

"},{"location":"bot-usage/#how-to-use-strategy-path","title":"How to use --strategy-path?","text":"

This parameter allows you to add an additional strategy lookup path, which gets checked before the default locations (The passed path must be a directory!):

freqtrade trade --strategy AwesomeStrategy --strategy-path /some/directory\n
"},{"location":"bot-usage/#how-to-install-a-strategy","title":"How to install a strategy?","text":"

This is very simple. Copy paste your strategy file into the directory user_data/strategies or use --strategy-path. And voila, the bot is ready to use it.

"},{"location":"bot-usage/#how-to-use-db-url","title":"How to use --db-url?","text":"

When you run the bot in Dry-run mode, per default no transactions are stored in a database. If you want to store your bot actions in a DB using --db-url. This can also be used to specify a custom database in production mode. Example command:

freqtrade trade -c config.json --db-url sqlite:///tradesv3.dry_run.sqlite\n
"},{"location":"bot-usage/#next-step","title":"Next step","text":"

The optimal strategy of the bot will change with time depending of the market trends. The next step is to Strategy Customization.

"},{"location":"configuration/","title":"Configure the bot","text":"

Freqtrade has many configurable features and possibilities. By default, these settings are configured via the configuration file (see below).

"},{"location":"configuration/#the-freqtrade-configuration-file","title":"The Freqtrade configuration file","text":"

The bot uses a set of configuration parameters during its operation that all together conform to the bot configuration. It normally reads its configuration from a file (Freqtrade configuration file).

Per default, the bot loads the configuration from the config.json file, located in the current working directory.

You can specify a different configuration file used by the bot with the -c/--config command-line option.

If you used the Quick start method for installing the bot, the installation script should have already created the default configuration file (config.json) for you.

If the default configuration file is not created we recommend to use freqtrade new-config --config config.json to generate a basic configuration file.

The Freqtrade configuration file is to be written in JSON format.

Additionally to the standard JSON syntax, you may use one-line // ... and multi-line /* ... */ comments in your configuration files and trailing commas in the lists of parameters.

Do not worry if you are not familiar with JSON format -- simply open the configuration file with an editor of your choice, make some changes to the parameters you need, save your changes and, finally, restart the bot or, if it was previously stopped, run it again with the changes you made to the configuration. The bot validates the syntax of the configuration file at startup and will warn you if you made any errors editing it, pointing out problematic lines.

"},{"location":"configuration/#environment-variables","title":"Environment variables","text":"

Set options in the Freqtrade configuration via environment variables. This takes priority over the corresponding value in configuration or strategy.

Environment variables must be prefixed with FREQTRADE__ to be loaded to the freqtrade configuration.

__ serves as level separator, so the format used should correspond to FREQTRADE__{section}__{key}. As such - an environment variable defined as export FREQTRADE__STAKE_AMOUNT=200 would result in {stake_amount: 200}.

A more complex example might be export FREQTRADE__EXCHANGE__KEY=<yourExchangeKey> to keep your exchange key secret. This will move the value to the exchange.key section of the configuration. Using this scheme, all configuration settings will also be available as environment variables.

Please note that Environment variables will overwrite corresponding settings in your configuration, but command line Arguments will always win.

Common example:

FREQTRADE__TELEGRAM__CHAT_ID=<telegramchatid>\nFREQTRADE__TELEGRAM__TOKEN=<telegramToken>\nFREQTRADE__EXCHANGE__KEY=<yourExchangeKey>\nFREQTRADE__EXCHANGE__SECRET=<yourExchangeSecret>\n

Note

Environment variables detected are logged at startup - so if you can't find why a value is not what you think it should be based on the configuration, make sure it's not loaded from an environment variable.

"},{"location":"configuration/#multiple-configuration-files","title":"Multiple configuration files","text":"

Multiple configuration files can be specified and used by the bot or the bot can read its configuration parameters from the process standard input stream.

You can specify additional configuration files in add_config_files. Files specified in this parameter will be loaded and merged with the initial config file. The files are resolved relative to the initial configuration file. This is similar to using multiple --config parameters, but simpler in usage as you don't have to specify all files for all commands.

Use multiple configuration files to keep secrets secret

You can use a 2nd configuration file containing your secrets. That way you can share your \"primary\" configuration file, while still keeping your API keys for yourself. The 2nd file should only specify what you intend to override. If a key is in more than one of the configurations, then the \"last specified configuration\" wins (in the above example, config-private.json).

For one-off commands, you can also use the below syntax by specifying multiple \"--config\" parameters.

freqtrade trade --config user_data/config1.json --config user_data/config-private.json <...>\n

The below is equivalent to the example above - but having 2 configuration files in the configuration, for easier reuse.

user_data/config.json
\"add_config_files\": [\n\"config1.json\",\n\"config-private.json\"\n]\n
freqtrade trade --config user_data/config.json <...>\n
config collision handling

If the same configuration setting takes place in both config.json and config-import.json, then the parent configuration wins. In the below case, max_open_trades would be 3 after the merging - as the reusable \"import\" configuration has this key overwritten.

user_data/config.json
{\n\"max_open_trades\": 3,\n\"stake_currency\": \"USDT\",\n\"add_config_files\": [\n\"config-import.json\"\n]\n}\n
user_data/config-import.json
{\n\"max_open_trades\": 10,\n\"stake_amount\": \"unlimited\",\n}\n

Resulting combined configuration:

Result
{\n\"max_open_trades\": 3,\n\"stake_currency\": \"USDT\",\n\"stake_amount\": \"unlimited\"\n}\n

If multiple files are in the add_config_files section, then they will be assumed to be at identical levels, having the last occurrence override the earlier config (unless a parent already defined such a key).

"},{"location":"configuration/#configuration-parameters","title":"Configuration parameters","text":"

The table below will list all configuration parameters available.

Freqtrade can also load many options via command line (CLI) arguments (check out the commands --help output for details).

"},{"location":"configuration/#configuration-option-prevalence","title":"Configuration option prevalence","text":"

The prevalence for all Options is as follows:

  • CLI arguments override any other option
  • Environment Variables
  • Configuration files are used in sequence (the last file wins) and override Strategy configurations.
  • Strategy configurations are only used if they are not set via configuration or command-line arguments. These options are marked with Strategy Override in the below table.
"},{"location":"configuration/#parameters-table","title":"Parameters table","text":"

Mandatory parameters are marked as Required, which means that they are required to be set in one of the possible ways.

Parameter Description max_open_trades Required. Number of open trades your bot is allowed to have. Only one open trade per pair is possible, so the length of your pairlist is another limitation that can apply. If -1 then it is ignored (i.e. potentially unlimited open trades, limited by the pairlist). More information below. Strategy Override. Datatype: Positive integer or -1. stake_currency Required. Crypto-currency used for trading. Datatype: String stake_amount Required. Amount of crypto-currency your bot will use for each trade. Set it to \"unlimited\" to allow the bot to use all available balance. More information below. Datatype: Positive float or \"unlimited\". tradable_balance_ratio Ratio of the total account balance the bot is allowed to trade. More information below. Defaults to 0.99 99%). Datatype: Positive float between 0.1 and 1.0. available_capital Available starting capital for the bot. Useful when running multiple bots on the same exchange account. More information below. Datatype: Positive float. amend_last_stake_amount Use reduced last stake amount if necessary. More information below. Defaults to false. Datatype: Boolean last_stake_amount_min_ratio Defines minimum stake amount that has to be left and executed. Applies only to the last stake amount when it's amended to a reduced value (i.e. if amend_last_stake_amount is set to true). More information below. Defaults to 0.5. Datatype: Float (as ratio) amount_reserve_percent Reserve some amount in min pair stake amount. The bot will reserve amount_reserve_percent + stoploss value when calculating min pair stake amount in order to avoid possible trade refusals. Defaults to 0.05 (5%). Datatype: Positive Float as ratio. timeframe The timeframe to use (e.g 1m, 5m, 15m, 30m, 1h ...). Usually missing in configuration, and specified in the strategy. Strategy Override. Datatype: String fiat_display_currency Fiat currency used to show your profits. More information below. Datatype: String dry_run Required. Define if the bot must be in Dry Run or production mode. Defaults to true. Datatype: Boolean dry_run_wallet Define the starting amount in stake currency for the simulated wallet used by the bot running in Dry Run mode.Defaults to 1000. Datatype: Float cancel_open_orders_on_exit Cancel open orders when the /stop RPC command is issued, Ctrl+C is pressed or the bot dies unexpectedly. When set to true, this allows you to use /stop to cancel unfilled and partially filled orders in the event of a market crash. It does not impact open positions. Defaults to false. Datatype: Boolean process_only_new_candles Enable processing of indicators only when new candles arrive. If false each loop populates the indicators, this will mean the same candle is processed many times creating system load but can be useful of your strategy depends on tick data not only candle. Strategy Override. Defaults to true. Datatype: Boolean minimal_roi Required. Set the threshold as ratio the bot will use to exit a trade. More information below. Strategy Override. Datatype: Dict stoploss Required. Value as ratio of the stoploss used by the bot. More details in the stoploss documentation. Strategy Override. Datatype: Float (as ratio) trailing_stop Enables trailing stoploss (based on stoploss in either configuration or strategy file). More details in the stoploss documentation. Strategy Override. Datatype: Boolean trailing_stop_positive Changes stoploss once profit has been reached. More details in the stoploss documentation. Strategy Override. Datatype: Float trailing_stop_positive_offset Offset on when to apply trailing_stop_positive. Percentage value which should be positive. More details in the stoploss documentation. Strategy Override. Defaults to 0.0 (no offset). Datatype: Float trailing_only_offset_is_reached Only apply trailing stoploss when the offset is reached. stoploss documentation. Strategy Override. Defaults to false. Datatype: Boolean fee Fee used during backtesting / dry-runs. Should normally not be configured, which has freqtrade fall back to the exchange default fee. Set as ratio (e.g. 0.001 = 0.1%). Fee is applied twice for each trade, once when buying, once when selling. Datatype: Float (as ratio) futures_funding_rate User-specified funding rate to be used when historical funding rates are not available from the exchange. This does not overwrite real historical rates. It is recommended that this be set to 0 unless you are testing a specific coin and you understand how the funding rate will affect freqtrade's profit calculations. More information here Defaults to None. Datatype: Float trading_mode Specifies if you want to trade regularly, trade with leverage, or trade contracts whose prices are derived from matching cryptocurrency prices. leverage documentation. Defaults to \"spot\". Datatype: String margin_mode When trading with leverage, this determines if the collateral owned by the trader will be shared or isolated to each trading pair leverage documentation. Datatype: String liquidation_buffer A ratio specifying how large of a safety net to place between the liquidation price and the stoploss to prevent a position from reaching the liquidation price leverage documentation. Defaults to 0.05. Datatype: Float Unfilled timeout unfilledtimeout.entry Required. How long (in minutes or seconds) the bot will wait for an unfilled entry order to complete, after which the order will be cancelled and repeated at current (new) price, as long as there is a signal. Strategy Override. Datatype: Integer unfilledtimeout.exit Required. How long (in minutes or seconds) the bot will wait for an unfilled exit order to complete, after which the order will be cancelled and repeated at current (new) price, as long as there is a signal. Strategy Override. Datatype: Integer unfilledtimeout.unit Unit to use in unfilledtimeout setting. Note: If you set unfilledtimeout.unit to \"seconds\", \"internals.process_throttle_secs\" must be inferior or equal to timeout Strategy Override. Defaults to \"minutes\". Datatype: String unfilledtimeout.exit_timeout_count How many times can exit orders time out. Once this number of timeouts is reached, an emergency exit is triggered. 0 to disable and allow unlimited order cancels. Strategy Override.Defaults to 0. Datatype: Integer Pricing entry_pricing.price_side Select the side of the spread the bot should look at to get the entry rate. More information below. Defaults to \"same\". Datatype: String (either ask, bid, same or other). entry_pricing.price_last_balance Required. Interpolate the bidding price. More information below. entry_pricing.use_order_book Enable entering using the rates in Order Book Entry. Defaults to true. Datatype: Boolean entry_pricing.order_book_top Bot will use the top N rate in Order Book \"price_side\" to enter a trade. I.e. a value of 2 will allow the bot to pick the 2nd entry in Order Book Entry. Defaults to 1. Datatype: Positive Integer entry_pricing. check_depth_of_market.enabled Do not enter if the difference of buy orders and sell orders is met in Order Book. Check market depth. Defaults to false. Datatype: Boolean entry_pricing. check_depth_of_market.bids_to_ask_delta The difference ratio of buy orders and sell orders found in Order Book. A value below 1 means sell order size is greater, while value greater than 1 means buy order size is higher. Check market depth Defaults to 0. Datatype: Float (as ratio) exit_pricing.price_side Select the side of the spread the bot should look at to get the exit rate. More information below. Defaults to \"same\". Datatype: String (either ask, bid, same or other). exit_pricing.price_last_balance Interpolate the exiting price. More information below. exit_pricing.use_order_book Enable exiting of open trades using Order Book Exit. Defaults to true. Datatype: Boolean exit_pricing.order_book_top Bot will use the top N rate in Order Book \"price_side\" to exit. I.e. a value of 2 will allow the bot to pick the 2nd ask rate in Order Book ExitDefaults to 1. Datatype: Positive Integer custom_price_max_distance_ratio Configure maximum distance ratio between current and custom entry or exit price. Defaults to 0.02 2%). Datatype: Positive float TODO use_exit_signal Use exit signals produced by the strategy in addition to the minimal_roi. Strategy Override. Defaults to true. Datatype: Boolean exit_profit_only Wait until the bot reaches exit_profit_offset before taking an exit decision. Strategy Override. Defaults to false. Datatype: Boolean exit_profit_offset Exit-signal is only active above this value. Only active in combination with exit_profit_only=True. Strategy Override. Defaults to 0.0. Datatype: Float (as ratio) ignore_roi_if_entry_signal Do not exit if the entry signal is still active. This setting takes preference over minimal_roi and use_exit_signal. Strategy Override. Defaults to false. Datatype: Boolean ignore_buying_expired_candle_after Specifies the number of seconds until a buy signal is no longer used. Datatype: Integer order_types Configure order-types depending on the action (\"entry\", \"exit\", \"stoploss\", \"stoploss_on_exchange\"). More information below. Strategy Override. Datatype: Dict order_time_in_force Configure time in force for entry and exit orders. More information below. Strategy Override. Datatype: Dict position_adjustment_enable Enables the strategy to use position adjustments (additional buys or sells). More information here. Strategy Override. Defaults to false. Datatype: Boolean max_entry_position_adjustment Maximum additional order(s) for each open trade on top of the first entry Order. Set it to -1 for unlimited additional orders. More information here. Strategy Override. Defaults to -1. Datatype: Positive Integer or -1 Exchange exchange.name Required. Name of the exchange class to use. List below. Datatype: String exchange.sandbox Use the 'sandbox' version of the exchange, where the exchange provides a sandbox for risk-free integration. See here in more details. Datatype: Boolean exchange.key API key to use for the exchange. Only required when you are in production mode.Keep it in secret, do not disclose publicly. Datatype: String exchange.secret API secret to use for the exchange. Only required when you are in production mode.Keep it in secret, do not disclose publicly. Datatype: String exchange.password API password to use for the exchange. Only required when you are in production mode and for exchanges that use password for API requests.Keep it in secret, do not disclose publicly. Datatype: String exchange.uid API uid to use for the exchange. Only required when you are in production mode and for exchanges that use uid for API requests.Keep it in secret, do not disclose publicly. Datatype: String exchange.pair_whitelist List of pairs to use by the bot for trading and to check for potential trades during backtesting. Supports regex pairs as .*/BTC. Not used by VolumePairList. More information. Datatype: List exchange.pair_blacklist List of pairs the bot must absolutely avoid for trading and backtesting. More information. Datatype: List exchange.ccxt_config Additional CCXT parameters passed to both ccxt instances (sync and async). This is usually the correct place for additional ccxt configurations. Parameters may differ from exchange to exchange and are documented in the ccxt documentation. Please avoid adding exchange secrets here (use the dedicated fields instead), as they may be contained in logs. Datatype: Dict exchange.ccxt_sync_config Additional CCXT parameters passed to the regular (sync) ccxt instance. Parameters may differ from exchange to exchange and are documented in the ccxt documentation Datatype: Dict exchange.ccxt_async_config Additional CCXT parameters passed to the async ccxt instance. Parameters may differ from exchange to exchange and are documented in the ccxt documentation Datatype: Dict exchange.markets_refresh_interval The interval in minutes in which markets are reloaded. Defaults to 60 minutes. Datatype: Positive Integer exchange.skip_pair_validation Skip pairlist validation on startup.Defaults to false Datatype: Boolean exchange.skip_open_order_update Skips open order updates on startup should the exchange cause problems. Only relevant in live conditions.Defaults to false Datatype: Boolean exchange.unknown_fee_rate Fallback value to use when calculating trading fees. This can be useful for exchanges which have fees in non-tradable currencies. The value provided here will be multiplied with the \"fee cost\".Defaults to None Datatype:* float exchange.log_responses Log relevant exchange responses. For debug mode only - use with care.Defaults to false Datatype: Boolean experimental.block_bad_exchanges Block exchanges known to not work with freqtrade. Leave on default unless you want to test if that exchange works now. Defaults to true. Datatype: Boolean Plugins edge.* Please refer to edge configuration document for detailed explanation of all possible configuration options. pairlists Define one or more pairlists to be used. More information. Defaults to StaticPairList. Datatype: List of Dicts protections Define one or more protections to be used. More information. Datatype: List of Dicts Telegram telegram.enabled Enable the usage of Telegram. Datatype: Boolean telegram.token Your Telegram bot token. Only required if telegram.enabled is true. Keep it in secret, do not disclose publicly. Datatype: String telegram.chat_id Your personal Telegram account id. Only required if telegram.enabled is true. Keep it in secret, do not disclose publicly. Datatype: String telegram.balance_dust_level Dust-level (in stake currency) - currencies with a balance below this will not be shown by /balance. Datatype: float telegram.reload Allow \"reload\" buttons on telegram messages. Defaults to true. Datatype:* boolean telegram.notification_settings.* Detailed notification settings. Refer to the telegram documentation for details. Datatype: dictionary telegram.allow_custom_messages Enable the sending of Telegram messages from strategies via the dataprovider.send_msg() function. Datatype: Boolean Webhook webhook.enabled Enable usage of Webhook notifications Datatype: Boolean webhook.url URL for the webhook. Only required if webhook.enabled is true. See the webhook documentation for more details. Datatype: String webhook.entry Payload to send on entry. Only required if webhook.enabled is true. See the webhook documentation for more details. Datatype: String webhook.entry_cancel Payload to send on entry order cancel. Only required if webhook.enabled is true. See the webhook documentation for more details. Datatype: String webhook.entry_fill Payload to send on entry order filled. Only required if webhook.enabled is true. See the webhook documentation for more details. Datatype: String webhook.exit Payload to send on exit. Only required if webhook.enabled is true. See the webhook documentation for more details. Datatype: String webhook.exit_cancel Payload to send on exit order cancel. Only required if webhook.enabled is true. See the webhook documentation for more details. Datatype: String webhook.exit_fill Payload to send on exit order filled. Only required if webhook.enabled is true. See the webhook documentation for more details. Datatype: String webhook.status Payload to send on status calls. Only required if webhook.enabled is true. See the webhook documentation for more details. Datatype: String webhook.allow_custom_messages Enable the sending of Webhook messages from strategies via the dataprovider.send_msg() function. Datatype: Boolean Rest API / FreqUI / Producer-Consumer api_server.enabled Enable usage of API Server. See the API Server documentation for more details. Datatype: Boolean api_server.listen_ip_address Bind IP address. See the API Server documentation for more details. Datatype: IPv4 api_server.listen_port Bind Port. See the API Server documentation for more details. Datatype: Integer between 1024 and 65535 api_server.verbosity Logging verbosity. info will print all RPC Calls, while \"error\" will only display errors. Datatype: Enum, either info or error. Defaults to info. api_server.username Username for API server. See the API Server documentation for more details. Keep it in secret, do not disclose publicly. Datatype: String api_server.password Password for API server. See the API Server documentation for more details. Keep it in secret, do not disclose publicly. Datatype: String api_server.ws_token API token for the Message WebSocket. See the API Server documentation for more details. Keep it in secret, do not disclose publicly. Datatype: String bot_name Name of the bot. Passed via API to a client - can be shown to distinguish / name bots. Defaults to freqtrade Datatype: String external_message_consumer Enable Producer/Consumer mode for more details. Datatype: Dict Other initial_state Defines the initial application state. If set to stopped, then the bot has to be explicitly started via /start RPC command. Defaults to stopped. Datatype: Enum, either stopped or running force_entry_enable Enables the RPC Commands to force a Trade entry. More information below. Datatype: Boolean disable_dataframe_checks Disable checking the OHLCV dataframe returned from the strategy methods for correctness. Only use when intentionally changing the dataframe and understand what you are doing. Strategy Override. Defaults to False. Datatype: Boolean internals.process_throttle_secs Set the process throttle, or minimum loop duration for one bot iteration loop. Value in second. Defaults to 5 seconds. Datatype: Positive Integer internals.heartbeat_interval Print heartbeat message every N seconds. Set to 0 to disable heartbeat messages. Defaults to 60 seconds. Datatype: Positive Integer or 0 internals.sd_notify Enables use of the sd_notify protocol to tell systemd service manager about changes in the bot state and issue keep-alive pings. See here for more details. Datatype: Boolean strategy Required Defines Strategy class to use. Recommended to be set via --strategy NAME. Datatype: ClassName strategy_path Adds an additional strategy lookup path (must be a directory). Datatype: String recursive_strategy_search Set to true to recursively search sub-directories inside user_data/strategies for a strategy. Datatype: Boolean user_data_dir Directory containing user data. Defaults to ./user_data/. Datatype: String db_url Declares database URL to use. NOTE: This defaults to sqlite:///tradesv3.dryrun.sqlite if dry_run is true, and to sqlite:///tradesv3.sqlite for production instances. Datatype: String, SQLAlchemy connect string logfile Specifies logfile name. Uses a rolling strategy for log file rotation for 10 files with the 1MB limit per file. Datatype: String add_config_files Additional config files. These files will be loaded and merged with the current config file. The files are resolved relative to the initial file. Defaults to []. Datatype: List of strings dataformat_ohlcv Data format to use to store historical candle (OHLCV) data. Defaults to json. Datatype: String dataformat_trades Data format to use to store historical trades data. Defaults to jsongz. Datatype: String reduce_df_footprint Recast all numeric columns to float32/int32, with the objective of reducing ram/disk usage (and decreasing train/inference timing in FreqAI). (Currently only affects FreqAI use-cases) Datatype: Boolean. Default: False."},{"location":"configuration/#parameters-in-the-strategy","title":"Parameters in the strategy","text":"

The following parameters can be set in the configuration file or strategy. Values set in the configuration file always overwrite values set in the strategy.

  • minimal_roi
  • timeframe
  • stoploss
  • max_open_trades
  • trailing_stop
  • trailing_stop_positive
  • trailing_stop_positive_offset
  • trailing_only_offset_is_reached
  • use_custom_stoploss
  • process_only_new_candles
  • order_types
  • order_time_in_force
  • unfilledtimeout
  • disable_dataframe_checks
  • use_exit_signal
  • exit_profit_only
  • exit_profit_offset
  • ignore_roi_if_entry_signal
  • ignore_buying_expired_candle_after
  • position_adjustment_enable
  • max_entry_position_adjustment
"},{"location":"configuration/#configuring-amount-per-trade","title":"Configuring amount per trade","text":"

There are several methods to configure how much of the stake currency the bot will use to enter a trade. All methods respect the available balance configuration as explained below.

"},{"location":"configuration/#minimum-trade-stake","title":"Minimum trade stake","text":"

The minimum stake amount will depend on exchange and pair and is usually listed in the exchange support pages.

Assuming the minimum tradable amount for XRP/USD is 20 XRP (given by the exchange), and the price is 0.6\\(, the minimum stake amount to buy this pair is 20 * 0.6 ~= 12. This exchange has also a limit on USD - where all orders must be > 10\\) - which however does not apply in this case.

To guarantee safe execution, freqtrade will not allow buying with a stake-amount of 10.1$, instead, it'll make sure that there's enough space to place a stoploss below the pair (+ an offset, defined by amount_reserve_percent, which defaults to 5%).

With a reserve of 5%, the minimum stake amount would be ~12.6$ (12 * (1 + 0.05)). If we take into account a stoploss of 10% on top of that - we'd end up with a value of ~14$ (12.6 / (1 - 0.1)).

To limit this calculation in case of large stoploss values, the calculated minimum stake-limit will never be more than 50% above the real limit.

Warning

Since the limits on exchanges are usually stable and are not updated often, some pairs can show pretty high minimum limits, simply because the price increased a lot since the last limit adjustment by the exchange. Freqtrade adjusts the stake-amount to this value, unless it's > 30% more than the calculated/desired stake-amount - in which case the trade is rejected.

"},{"location":"configuration/#tradable-balance","title":"Tradable balance","text":"

By default, the bot assumes that the complete amount - 1% is at it's disposal, and when using dynamic stake amount, it will split the complete balance into max_open_trades buckets per trade. Freqtrade will reserve 1% for eventual fees when entering a trade and will therefore not touch that by default.

You can configure the \"untouched\" amount by using the tradable_balance_ratio setting.

For example, if you have 10 ETH available in your wallet on the exchange and tradable_balance_ratio=0.5 (which is 50%), then the bot will use a maximum amount of 5 ETH for trading and considers this as an available balance. The rest of the wallet is untouched by the trades.

Danger

This setting should not be used when running multiple bots on the same account. Please look at Available Capital to the bot instead.

Warning

The tradable_balance_ratio setting applies to the current balance (free balance + tied up in trades). Therefore, assuming the starting balance of 1000, a configuration with tradable_balance_ratio=0.99 will not guarantee that 10 currency units will always remain available on the exchange. For example, the free amount may reduce to 5 units if the total balance is reduced to 500 (either by a losing streak or by withdrawing balance).

"},{"location":"configuration/#assign-available-capital","title":"Assign available Capital","text":"

To fully utilize compounding profits when using multiple bots on the same exchange account, you'll want to limit each bot to a certain starting balance. This can be accomplished by setting available_capital to the desired starting balance.

Assuming your account has 10.000 USDT and you want to run 2 different strategies on this exchange. You'd set available_capital=5000 - granting each bot an initial capital of 5000 USDT. The bot will then split this starting balance equally into max_open_trades buckets. Profitable trades will result in increased stake-sizes for this bot - without affecting the stake-sizes of the other bot.

Incompatible with tradable_balance_ratio

Setting this option will replace any configuration of tradable_balance_ratio.

"},{"location":"configuration/#amend-last-stake-amount","title":"Amend last stake amount","text":"

Assuming we have the tradable balance of 1000 USDT, stake_amount=400, and max_open_trades=3. The bot would open 2 trades and will be unable to fill the last trading slot, since the requested 400 USDT are no longer available since 800 USDT are already tied in other trades.

To overcome this, the option amend_last_stake_amount can be set to True, which will enable the bot to reduce stake_amount to the available balance to fill the last trade slot.

In the example above this would mean:

  • Trade1: 400 USDT
  • Trade2: 400 USDT
  • Trade3: 200 USDT

Note

This option only applies with Static stake amount - since Dynamic stake amount divides the balances evenly.

Note

The minimum last stake amount can be configured using last_stake_amount_min_ratio - which defaults to 0.5 (50%). This means that the minimum stake amount that's ever used is stake_amount * 0.5. This avoids very low stake amounts, that are close to the minimum tradable amount for the pair and can be refused by the exchange.

"},{"location":"configuration/#static-stake-amount","title":"Static stake amount","text":"

The stake_amount configuration statically configures the amount of stake-currency your bot will use for each trade.

The minimal configuration value is 0.0001, however, please check your exchange's trading minimums for the stake currency you're using to avoid problems.

This setting works in combination with max_open_trades. The maximum capital engaged in trades is stake_amount * max_open_trades. For example, the bot will at most use (0.05 BTC x 3) = 0.15 BTC, assuming a configuration of max_open_trades=3 and stake_amount=0.05.

Note

This setting respects the available balance configuration.

"},{"location":"configuration/#dynamic-stake-amount","title":"Dynamic stake amount","text":"

Alternatively, you can use a dynamic stake amount, which will use the available balance on the exchange, and divide that equally by the number of allowed trades (max_open_trades).

To configure this, set stake_amount=\"unlimited\". We also recommend to set tradable_balance_ratio=0.99 (99%) - to keep a minimum balance for eventual fees.

In this case a trade amount is calculated as:

currency_balance / (max_open_trades - current_open_trades)\n

To allow the bot to trade all the available stake_currency in your account (minus tradable_balance_ratio) set

\"stake_amount\" : \"unlimited\",\n\"tradable_balance_ratio\": 0.99,\n

Compounding profits

This configuration will allow increasing/decreasing stakes depending on the performance of the bot (lower stake if the bot is losing, higher stakes if the bot has a winning record since higher balances are available), and will result in profit compounding.

When using Dry-Run Mode

When using \"stake_amount\" : \"unlimited\", in combination with Dry-Run, Backtesting or Hyperopt, the balance will be simulated starting with a stake of dry_run_wallet which will evolve. It is therefore important to set dry_run_wallet to a sensible value (like 0.05 or 0.01 for BTC and 1000 or 100 for USDT, for example), otherwise, it may simulate trades with 100 BTC (or more) or 0.05 USDT (or less) at once - which may not correspond to your real available balance or is less than the exchange minimal limit for the order amount for the stake currency.

"},{"location":"configuration/#dynamic-stake-amount-with-position-adjustment","title":"Dynamic stake amount with position adjustment","text":"

When you want to use position adjustment with unlimited stakes, you must also implement custom_stake_amount to a return a value depending on your strategy. Typical value would be in the range of 25% - 50% of the proposed stakes, but depends highly on your strategy and how much you wish to leave into the wallet as position adjustment buffer.

For example if your position adjustment assumes it can do 2 additional buys with the same stake amounts then your buffer should be 66.6667% of the initially proposed unlimited stake amount.

Or another example if your position adjustment assumes it can do 1 additional buy with 3x the original stake amount then custom_stake_amount should return 25% of proposed stake amount and leave 75% for possible later position adjustments.

"},{"location":"configuration/#prices-used-for-orders","title":"Prices used for orders","text":"

Prices for regular orders can be controlled via the parameter structures entry_pricing for trade entries and exit_pricing for trade exits. Prices are always retrieved right before an order is placed, either by querying the exchange tickers or by using the orderbook data.

Note

Orderbook data used by Freqtrade are the data retrieved from exchange by the ccxt's function fetch_order_book(), i.e. are usually data from the L2-aggregated orderbook, while the ticker data are the structures returned by the ccxt's fetch_ticker()/fetch_tickers() functions. Refer to the ccxt library documentation for more details.

Using market orders

Please read the section Market order pricing section when using market orders.

"},{"location":"configuration/#entry-price","title":"Entry price","text":""},{"location":"configuration/#enter-price-side","title":"Enter price side","text":"

The configuration setting entry_pricing.price_side defines the side of the orderbook the bot looks for when buying.

The following displays an orderbook.

...\n103\n102\n101  # ask\n-------------Current spread\n99   # bid\n98\n97\n...\n

If entry_pricing.price_side is set to \"bid\", then the bot will use 99 as entry price. In line with that, if entry_pricing.price_side is set to \"ask\", then the bot will use 101 as entry price.

Depending on the order direction (long/short), this will lead to different results. Therefore we recommend to use \"same\" or \"other\" for this configuration instead. This would result in the following pricing matrix:

direction Order setting price crosses spread long buy ask 101 yes long buy bid 99 no long buy same 99 no long buy other 101 yes short sell ask 101 no short sell bid 99 yes short sell same 101 no short sell other 99 yes

Using the other side of the orderbook often guarantees quicker filled orders, but the bot can also end up paying more than what would have been necessary. Taker fees instead of maker fees will most likely apply even when using limit buy orders. Also, prices at the \"other\" side of the spread are higher than prices at the \"bid\" side in the orderbook, so the order behaves similar to a market order (however with a maximum price).

"},{"location":"configuration/#entry-price-with-orderbook-enabled","title":"Entry price with Orderbook enabled","text":"

When entering a trade with the orderbook enabled (entry_pricing.use_order_book=True), Freqtrade fetches the entry_pricing.order_book_top entries from the orderbook and uses the entry specified as entry_pricing.order_book_top on the configured side (entry_pricing.price_side) of the orderbook. 1 specifies the topmost entry in the orderbook, while 2 would use the 2nd entry in the orderbook, and so on.

"},{"location":"configuration/#entry-price-without-orderbook-enabled","title":"Entry price without Orderbook enabled","text":"

The following section uses side as the configured entry_pricing.price_side (defaults to \"same\").

When not using orderbook (entry_pricing.use_order_book=False), Freqtrade uses the best side price from the ticker if it's below the last traded price from the ticker. Otherwise (when the side price is above the last price), it calculates a rate between side and last price based on entry_pricing.price_last_balance.

The entry_pricing.price_last_balance configuration parameter controls this. A value of 0.0 will use side price, while 1.0 will use the last price and values between those interpolate between ask and last price.

"},{"location":"configuration/#check-depth-of-market","title":"Check depth of market","text":"

When check depth of market is enabled (entry_pricing.check_depth_of_market.enabled=True), the entry signals are filtered based on the orderbook depth (sum of all amounts) for each orderbook side.

Orderbook bid (buy) side depth is then divided by the orderbook ask (sell) side depth and the resulting delta is compared to the value of the entry_pricing.check_depth_of_market.bids_to_ask_delta parameter. The entry order is only executed if the orderbook delta is greater than or equal to the configured delta value.

Note

A delta value below 1 means that ask (sell) orderbook side depth is greater than the depth of the bid (buy) orderbook side, while a value greater than 1 means opposite (depth of the buy side is higher than the depth of the sell side).

"},{"location":"configuration/#exit-price","title":"Exit price","text":""},{"location":"configuration/#exit-price-side","title":"Exit price side","text":"

The configuration setting exit_pricing.price_side defines the side of the spread the bot looks for when exiting a trade.

The following displays an orderbook:

...\n103\n102\n101  # ask\n-------------Current spread\n99   # bid\n98\n97\n...\n

If exit_pricing.price_side is set to \"ask\", then the bot will use 101 as exiting price. In line with that, if exit_pricing.price_side is set to \"bid\", then the bot will use 99 as exiting price.

Depending on the order direction (long/short), this will lead to different results. Therefore we recommend to use \"same\" or \"other\" for this configuration instead. This would result in the following pricing matrix:

Direction Order setting price crosses spread long sell ask 101 no long sell bid 99 yes long sell same 101 no long sell other 99 yes short buy ask 101 yes short buy bid 99 no short buy same 99 no short buy other 101 yes"},{"location":"configuration/#exit-price-with-orderbook-enabled","title":"Exit price with Orderbook enabled","text":"

When exiting with the orderbook enabled (exit_pricing.use_order_book=True), Freqtrade fetches the exit_pricing.order_book_top entries in the orderbook and uses the entry specified as exit_pricing.order_book_top from the configured side (exit_pricing.price_side) as trade exit price.

1 specifies the topmost entry in the orderbook, while 2 would use the 2nd entry in the orderbook, and so on.

"},{"location":"configuration/#exit-price-without-orderbook-enabled","title":"Exit price without Orderbook enabled","text":"

The following section uses side as the configured exit_pricing.price_side (defaults to \"ask\").

When not using orderbook (exit_pricing.use_order_book=False), Freqtrade uses the best side price from the ticker if it's above the last traded price from the ticker. Otherwise (when the side price is below the last price), it calculates a rate between side and last price based on exit_pricing.price_last_balance.

The exit_pricing.price_last_balance configuration parameter controls this. A value of 0.0 will use side price, while 1.0 will use the last price and values between those interpolate between side and last price.

"},{"location":"configuration/#market-order-pricing","title":"Market order pricing","text":"

When using market orders, prices should be configured to use the \"correct\" side of the orderbook to allow realistic pricing detection. Assuming both entry and exits are using market orders, a configuration similar to the following must be used

  \"order_types\": {\n    \"entry\": \"market\",\n    \"exit\": \"market\"\n    // ...\n  },\n  \"entry_pricing\": {\n    \"price_side\": \"other\",\n    // ...\n  },\n  \"exit_pricing\":{\n    \"price_side\": \"other\",\n    // ...\n  },\n

Obviously, if only one side is using limit orders, different pricing combinations can be used.

"},{"location":"configuration/#understand-minimal_roi","title":"Understand minimal_roi","text":"

The minimal_roi configuration parameter is a JSON object where the key is a duration in minutes and the value is the minimum ROI as a ratio. See the example below:

\"minimal_roi\": {\n\"40\": 0.0,    # Exit after 40 minutes if the profit is not negative\n\"30\": 0.01,   # Exit after 30 minutes if there is at least 1% profit\n\"20\": 0.02,   # Exit after 20 minutes if there is at least 2% profit\n\"0\":  0.04    # Exit immediately if there is at least 4% profit\n},\n

Most of the strategy files already include the optimal minimal_roi value. This parameter can be set in either Strategy or Configuration file. If you use it in the configuration file, it will override the minimal_roi value from the strategy file. If it is not set in either Strategy or Configuration, a default of 1000% {\"0\": 10} is used, and minimal ROI is disabled unless your trade generates 1000% profit.

Special case to forceexit after a specific time

A special case presents using \"<N>\": -1 as ROI. This forces the bot to exit a trade after N Minutes, no matter if it's positive or negative, so represents a time-limited force-exit.

"},{"location":"configuration/#understand-force_entry_enable","title":"Understand force_entry_enable","text":"

The force_entry_enable configuration parameter enables the usage of force-enter (/forcelong, /forceshort) commands via Telegram and REST API. For security reasons, it's disabled by default, and freqtrade will show a warning message on startup if enabled. For example, you can send /forceenter ETH/BTC to the bot, which will result in freqtrade buying the pair and holds it until a regular exit-signal (ROI, stoploss, /forceexit) appears.

This can be dangerous with some strategies, so use with care.

See the telegram documentation for details on usage.

"},{"location":"configuration/#ignoring-expired-candles","title":"Ignoring expired candles","text":"

When working with larger timeframes (for example 1h or more) and using a low max_open_trades value, the last candle can be processed as soon as a trade slot becomes available. When processing the last candle, this can lead to a situation where it may not be desirable to use the buy signal on that candle. For example, when using a condition in your strategy where you use a cross-over, that point may have passed too long ago for you to start a trade on it.

In these situations, you can enable the functionality to ignore candles that are beyond a specified period by setting ignore_buying_expired_candle_after to a positive number, indicating the number of seconds after which the buy signal becomes expired.

For example, if your strategy is using a 1h timeframe, and you only want to buy within the first 5 minutes when a new candle comes in, you can add the following configuration to your strategy:

  {\n//...\n\"ignore_buying_expired_candle_after\": 300,\n// ...\n}\n

Note

This setting resets with each new candle, so it will not prevent sticking-signals from executing on the 2nd or 3rd candle they're active. Best use a \"trigger\" selector for buy signals, which are only active for one candle.

"},{"location":"configuration/#understand-order_types","title":"Understand order_types","text":"

The order_types configuration parameter maps actions (entry, exit, stoploss, emergency_exit, force_exit, force_entry) to order-types (market, limit, ...) as well as configures stoploss to be on the exchange and defines stoploss on exchange update interval in seconds.

This allows to enter using limit orders, exit using limit-orders, and create stoplosses using market orders. It also allows to set the stoploss \"on exchange\" which means stoploss order would be placed immediately once the buy order is fulfilled.

order_types set in the configuration file overwrites values set in the strategy as a whole, so you need to configure the whole order_types dictionary in one place.

If this is configured, the following 4 values (entry, exit, stoploss and stoploss_on_exchange) need to be present, otherwise, the bot will fail to start.

For information on (emergency_exit,force_exit, force_entry, stoploss_on_exchange,stoploss_on_exchange_interval,stoploss_on_exchange_limit_ratio) please see stop loss documentation stop loss on exchange

Syntax for Strategy:

order_types = {\n    \"entry\": \"limit\",\n    \"exit\": \"limit\",\n    \"emergency_exit\": \"market\",\n    \"force_entry\": \"market\",\n    \"force_exit\": \"market\",\n    \"stoploss\": \"market\",\n    \"stoploss_on_exchange\": False,\n    \"stoploss_on_exchange_interval\": 60,\n    \"stoploss_on_exchange_limit_ratio\": 0.99,\n}\n

Configuration:

\"order_types\": {\n\"entry\": \"limit\",\n\"exit\": \"limit\",\n\"emergency_exit\": \"market\",\n\"force_entry\": \"market\",\n\"force_exit\": \"market\",\n\"stoploss\": \"market\",\n\"stoploss_on_exchange\": false,\n\"stoploss_on_exchange_interval\": 60\n}\n

Market order support

Not all exchanges support \"market\" orders. The following message will be shown if your exchange does not support market orders: \"Exchange <yourexchange> does not support market orders.\" and the bot will refuse to start.

Using market orders

Please carefully read the section Market order pricing section when using market orders.

Stoploss on exchange

stoploss_on_exchange_interval is not mandatory. Do not change its value if you are unsure of what you are doing. For more information about how stoploss works please refer to the stoploss documentation.

If stoploss_on_exchange is enabled and the stoploss is cancelled manually on the exchange, then the bot will create a new stoploss order.

Warning: stoploss_on_exchange failures

If stoploss on exchange creation fails for some reason, then an \"emergency exit\" is initiated. By default, this will exit the trade using a market order. The order-type for the emergency-exit can be changed by setting the emergency_exit value in the order_types dictionary - however, this is not advised.

"},{"location":"configuration/#understand-order_time_in_force","title":"Understand order_time_in_force","text":"

The order_time_in_force configuration parameter defines the policy by which the order is executed on the exchange. Three commonly used time in force are:

GTC (Good Till Canceled):

This is most of the time the default time in force. It means the order will remain on exchange till it is cancelled by the user. It can be fully or partially fulfilled. If partially fulfilled, the remaining will stay on the exchange till cancelled.

FOK (Fill Or Kill):

It means if the order is not executed immediately AND fully then it is cancelled by the exchange.

IOC (Immediate Or Canceled):

It is the same as FOK (above) except it can be partially fulfilled. The remaining part is automatically cancelled by the exchange.

PO (Post only):

Post only order. The order is either placed as a maker order, or it is canceled. This means the order must be placed on orderbook for at at least time in an unfilled state.

"},{"location":"configuration/#time_in_force-config","title":"time_in_force config","text":"

The order_time_in_force parameter contains a dict with entry and exit time in force policy values. This can be set in the configuration file or in the strategy. Values set in the configuration file overwrites values set in the strategy.

The possible values are: GTC (default), FOK or IOC.

\"order_time_in_force\": {\n    \"entry\": \"GTC\",\n    \"exit\": \"GTC\"\n},\n

Warning

This is ongoing work. For now, it is supported only for binance, gate and kucoin. Please don't change the default value unless you know what you are doing and have researched the impact of using different values for your particular exchange.

"},{"location":"configuration/#what-values-can-be-used-for-fiat_display_currency","title":"What values can be used for fiat_display_currency?","text":"

The fiat_display_currency configuration parameter sets the base currency to use for the conversion from coin to fiat in the bot Telegram reports.

The valid values are:

\"AUD\", \"BRL\", \"CAD\", \"CHF\", \"CLP\", \"CNY\", \"CZK\", \"DKK\", \"EUR\", \"GBP\", \"HKD\", \"HUF\", \"IDR\", \"ILS\", \"INR\", \"JPY\", \"KRW\", \"MXN\", \"MYR\", \"NOK\", \"NZD\", \"PHP\", \"PKR\", \"PLN\", \"RUB\", \"SEK\", \"SGD\", \"THB\", \"TRY\", \"TWD\", \"ZAR\", \"USD\"\n

In addition to fiat currencies, a range of crypto currencies is supported.

The valid values are:

\"BTC\", \"ETH\", \"XRP\", \"LTC\", \"BCH\", \"USDT\"\n
"},{"location":"configuration/#using-dry-run-mode","title":"Using Dry-run mode","text":"

We recommend starting the bot in the Dry-run mode to see how your bot will behave and what is the performance of your strategy. In the Dry-run mode, the bot does not engage your money. It only runs a live simulation without creating trades on the exchange.

  1. Edit your config.json configuration file.
  2. Switch dry-run to true and specify db_url for a persistence database.
\"dry_run\": true,\n\"db_url\": \"sqlite:///tradesv3.dryrun.sqlite\",\n
  1. Remove your Exchange API key and secret (change them by empty values or fake credentials):
\"exchange\": {\n\"name\": \"bittrex\",\n\"key\": \"key\",\n\"secret\": \"secret\",\n...\n}\n

Once you will be happy with your bot performance running in the Dry-run mode, you can switch it to production mode.

Note

A simulated wallet is available during dry-run mode and will assume a starting capital of dry_run_wallet (defaults to 1000).

"},{"location":"configuration/#considerations-for-dry-run","title":"Considerations for dry-run","text":"
  • API-keys may or may not be provided. Only Read-Only operations (i.e. operations that do not alter account state) on the exchange are performed in dry-run mode.
  • Wallets (/balance) are simulated based on dry_run_wallet.
  • Orders are simulated, and will not be posted to the exchange.
  • Market orders fill based on orderbook volume the moment the order is placed.
  • Limit orders fill once the price reaches the defined level - or time out based on unfilledtimeout settings.
  • In combination with stoploss_on_exchange, the stop_loss price is assumed to be filled.
  • Open orders (not trades, which are stored in the database) are kept open after bot restarts, with the assumption that they were not filled while being offline.
"},{"location":"configuration/#switch-to-production-mode","title":"Switch to production mode","text":"

In production mode, the bot will engage your money. Be careful, since a wrong strategy can lose all your money. Be aware of what you are doing when you run it in production mode.

When switching to Production mode, please make sure to use a different / fresh database to avoid dry-run trades messing with your exchange money and eventually tainting your statistics.

"},{"location":"configuration/#setup-your-exchange-account","title":"Setup your exchange account","text":"

You will need to create API Keys (usually you get key and secret, some exchanges require an additional password) from the Exchange website and you'll need to insert this into the appropriate fields in the configuration or when asked by the freqtrade new-config command. API Keys are usually only required for live trading (trading for real money, bot running in \"production mode\", executing real orders on the exchange) and are not required for the bot running in dry-run (trade simulation) mode. When you set up the bot in dry-run mode, you may fill these fields with empty values.

"},{"location":"configuration/#to-switch-your-bot-in-production-mode","title":"To switch your bot in production mode","text":"

Edit your config.json file.

Switch dry-run to false and don't forget to adapt your database URL if set:

\"dry_run\": false,\n

Insert your Exchange API key (change them by fake API keys):

{\n\"exchange\": {\n\"name\": \"bittrex\",\n\"key\": \"af8ddd35195e9dc500b9a6f799f6f5c93d89193b\",\n\"secret\": \"08a9dc6db3d7b53e1acebd9275677f4b0a04f1a5\",\n//\"password\": \"\", // Optional, not needed by all exchanges)\n// ...\n}\n//...\n}\n

You should also make sure to read the Exchanges section of the documentation to be aware of potential configuration details specific to your exchange.

Keep your secrets secret

To keep your secrets secret, we recommend using a 2nd configuration for your API keys. Simply use the above snippet in a new configuration file (e.g. config-private.json) and keep your settings in this file. You can then start the bot with freqtrade trade --config user_data/config.json --config user_data/config-private.json <...> to have your keys loaded.

NEVER share your private configuration file or your exchange keys with anyone!

"},{"location":"configuration/#using-proxy-with-freqtrade","title":"Using proxy with Freqtrade","text":"

To use a proxy with freqtrade, export your proxy settings using the variables \"HTTP_PROXY\" and \"HTTPS_PROXY\" set to the appropriate values. This will have the proxy settings applied to everything (telegram, coingecko, ...) except for exchange requests.

export HTTP_PROXY=\"http://addr:port\"\nexport HTTPS_PROXY=\"http://addr:port\"\nfreqtrade\n
"},{"location":"configuration/#proxy-exchange-requests","title":"Proxy exchange requests","text":"

To use a proxy for exchange connections - you will have to define the proxies as part of the ccxt configuration.

{ \"exchange\": {\n\"ccxt_config\": {\n\"aiohttp_proxy\": \"http://addr:port\",\n\"proxies\": {\n\"http\": \"http://addr:port\",\n\"https\": \"http://addr:port\"\n},\n}\n}\n}\n
"},{"location":"configuration/#next-step","title":"Next step","text":"

Now you have configured your config.json, the next step is to start your bot.

"},{"location":"data-analysis/","title":"Analyzing bot data with Jupyter notebooks","text":"

You can analyze the results of backtests and trading history easily using Jupyter notebooks. Sample notebooks are located at user_data/notebooks/ after initializing the user directory with freqtrade create-userdir --userdir user_data.

"},{"location":"data-analysis/#quick-start-with-docker","title":"Quick start with docker","text":"

Freqtrade provides a docker-compose file which starts up a jupyter lab server. You can run this server using the following command: docker compose -f docker/docker-compose-jupyter.yml up

This will create a dockercontainer running jupyter lab, which will be accessible using https://127.0.0.1:8888/lab. Please use the link that's printed in the console after startup for simplified login.

For more information, Please visit the Data analysis with Docker section.

"},{"location":"data-analysis/#pro-tips","title":"Pro tips","text":"
  • See jupyter.org for usage instructions.
  • Don't forget to start a Jupyter notebook server from within your conda or venv environment or use nb_conda_kernels*
  • Copy the example notebook before use so your changes don't get overwritten with the next freqtrade update.
"},{"location":"data-analysis/#using-virtual-environment-with-system-wide-jupyter-installation","title":"Using virtual environment with system-wide Jupyter installation","text":"

Sometimes it can be desired to use a system-wide installation of Jupyter notebook, and use a jupyter kernel from the virtual environment. This prevents you from installing the full jupyter suite multiple times per system, and provides an easy way to switch between tasks (freqtrade / other analytics tasks).

For this to work, first activate your virtual environment and run the following commands:

# Activate virtual environment\nsource .env/bin/activate\n\npip install ipykernel\nipython kernel install --user --name=freqtrade\n# Restart jupyter (lab / notebook)\n# select kernel \"freqtrade\" in the notebook\n

Note

This section is provided for completeness, the Freqtrade Team won't provide full support for problems with this setup and will recommend to install Jupyter in the virtual environment directly, as that is the easiest way to get jupyter notebooks up and running. For help with this setup please refer to the Project Jupyter documentation or help channels.

Warning

Some tasks don't work especially well in notebooks. For example, anything using asynchronous execution is a problem for Jupyter. Also, freqtrade's primary entry point is the shell cli, so using pure python in a notebook bypasses arguments that provide required objects and parameters to helper functions. You may need to set those values or create expected objects manually.

"},{"location":"data-analysis/#recommended-workflow","title":"Recommended workflow","text":"Task Tool Bot operations CLI Repetitive tasks Shell scripts Data analysis & visualization Notebook
  1. Use the CLI to

    * download historical data * run a backtest * run with real-time data * export results

  2. Collect these actions in shell scripts

    * save complicated commands with arguments * execute multi-step operations * automate testing strategies and preparing data for analysis

  3. Use a notebook to

    * visualize data * mangle and plot to generate insights

"},{"location":"data-analysis/#example-utility-snippets","title":"Example utility snippets","text":""},{"location":"data-analysis/#change-directory-to-root","title":"Change directory to root","text":"

Jupyter notebooks execute from the notebook directory. The following snippet searches for the project root, so relative paths remain consistent.

import os\nfrom pathlib import Path\n\n# Change directory\n# Modify this cell to insure that the output shows the correct path.\n# Define all paths relative to the project root shown in the cell output\nproject_root = \"somedir/freqtrade\"\ni=0\ntry:\n    os.chdir(project_root)\n    assert Path('LICENSE').is_file()\nexcept:\n    while i<4 and (not Path('LICENSE').is_file()):\n        os.chdir(Path(Path.cwd(), '../'))\n        i+=1\n    project_root = Path.cwd()\nprint(Path.cwd())\n
"},{"location":"data-analysis/#load-multiple-configuration-files","title":"Load multiple configuration files","text":"

This option can be useful to inspect the results of passing in multiple configs. This will also run through the whole Configuration initialization, so the configuration is completely initialized to be passed to other methods.

import json\nfrom freqtrade.configuration import Configuration\n\n# Load config from multiple files\nconfig = Configuration.from_files([\"config1.json\", \"config2.json\"])\n\n# Show the config in memory\nprint(json.dumps(config['original_config'], indent=2))\n

For Interactive environments, have an additional configuration specifying user_data_dir and pass this in last, so you don't have to change directories while running the bot. Best avoid relative paths, since this starts at the storage location of the jupyter notebook, unless the directory is changed.

{\n\"user_data_dir\": \"~/.freqtrade/\"\n}\n
"},{"location":"data-analysis/#further-data-analysis-documentation","title":"Further Data analysis documentation","text":"
  • Strategy debugging - also available as Jupyter notebook (user_data/notebooks/strategy_analysis_example.ipynb)
  • Plotting
  • Tag Analysis

Feel free to submit an issue or Pull Request enhancing this document if you would like to share ideas on how to best analyze the data.

"},{"location":"data-download/","title":"Data Downloading","text":""},{"location":"data-download/#getting-data-for-backtesting-and-hyperopt","title":"Getting data for backtesting and hyperopt","text":"

To download data (candles / OHLCV) needed for backtesting and hyperoptimization use the freqtrade download-data command.

If no additional parameter is specified, freqtrade will download data for \"1m\" and \"5m\" timeframes for the last 30 days. Exchange and pairs will come from config.json (if specified using -c/--config). Otherwise --exchange becomes mandatory.

You can use a relative timerange (--days 20) or an absolute starting point (--timerange 20200101-). For incremental downloads, the relative approach should be used.

Tip: Updating existing data

If you already have backtesting data available in your data-directory and would like to refresh this data up to today, freqtrade will automatically calculate the data missing for the existing pairs and the download will occur from the latest available point until \"now\", neither --days or --timerange parameters are required. Freqtrade will keep the available data and only download the missing data. If you are updating existing data after inserting new pairs that you have no data for, use --new-pairs-days xx parameter. Specified number of days will be downloaded for new pairs while old pairs will be updated with missing data only. If you use --days xx parameter alone - data for specified number of days will be downloaded for all pairs. Be careful, if specified number of days is smaller than gap between now and last downloaded candle - freqtrade will delete all existing data to avoid gaps in candle data.

"},{"location":"data-download/#usage","title":"Usage","text":"
usage: freqtrade download-data [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                               [-d PATH] [--userdir PATH]\n                               [-p PAIRS [PAIRS ...]] [--pairs-file FILE]\n                               [--days INT] [--new-pairs-days INT]\n                               [--include-inactive-pairs]\n                               [--timerange TIMERANGE] [--dl-trades]\n                               [--exchange EXCHANGE]\n                               [-t TIMEFRAMES [TIMEFRAMES ...]] [--erase]\n                               [--data-format-ohlcv {json,jsongz,hdf5,feather,parquet}]\n                               [--data-format-trades {json,jsongz,hdf5}]\n                               [--trading-mode {spot,margin,futures}]\n                               [--prepend]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]\n                        Limit command to these pairs. Pairs are space-\n                        separated.\n  --pairs-file FILE     File containing a list of pairs. Takes precedence over\n                        --pairs or pairs configured in the configuration.\n  --days INT            Download data for given number of days.\n  --new-pairs-days INT  Download data of new pairs for given number of days.\n                        Default: `None`.\n  --include-inactive-pairs\n                        Also download data from inactive pairs.\n  --timerange TIMERANGE\n                        Specify what timerange of data to use.\n  --dl-trades           Download trades instead of OHLCV data. The bot will\n                        resample trades to the desired timeframe as specified\n                        as --timeframes/-t.\n  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no\n                        config is provided.\n  -t TIMEFRAMES [TIMEFRAMES ...], --timeframes TIMEFRAMES [TIMEFRAMES ...]\n                        Specify which tickers to download. Space-separated\n                        list. Default: `1m 5m`.\n  --erase               Clean all existing data for the selected\n                        exchange/pairs/timeframes.\n  --data-format-ohlcv {json,jsongz,hdf5,feather,parquet}\n                        Storage format for downloaded candle (OHLCV) data.\n                        (default: `json`).\n  --data-format-trades {json,jsongz,hdf5}\n                        Storage format for downloaded trades data. (default:\n                        `jsongz`).\n  --trading-mode {spot,margin,futures}, --tradingmode {spot,margin,futures}\n                        Select Trading mode\n  --prepend             Allow data prepending. (Data-appending is disabled)\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH, --data-dir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n

Startup period

download-data is a strategy-independent command. The idea is to download a big chunk of data once, and then iteratively increase the amount of data stored.

For that reason, download-data does not care about the \"startup-period\" defined in a strategy. It's up to the user to download additional days if the backtest should start at a specific point in time (while respecting startup period).

"},{"location":"data-download/#pairs-file","title":"Pairs file","text":"

In alternative to the whitelist from config.json, a pairs.json file can be used. If you are using Binance for example:

  • create a directory user_data/data/binance and copy or create the pairs.json file in that directory.
  • update the pairs.json file to contain the currency pairs you are interested in.
mkdir -p user_data/data/binance\ntouch user_data/data/binance/pairs.json\n

The format of the pairs.json file is a simple json list. Mixing different stake-currencies is allowed for this file, since it's only used for downloading.

[\n\"ETH/BTC\",\n\"ETH/USDT\",\n\"BTC/USDT\",\n\"XRP/ETH\"\n]\n

Downloading all data for one quote currency

Often, you'll want to download data for all pairs of a specific quote-currency. In such cases, you can use the following shorthand: freqtrade download-data --exchange binance --pairs .*/USDT <...>. The provided \"pairs\" string will be expanded to contain all active pairs on the exchange. To also download data for inactive (delisted) pairs, add --include-inactive-pairs to the command.

Permission denied errors

If your configuration directory user_data was made by docker, you may get the following error:

cp: cannot create regular file 'user_data/data/binance/pairs.json': Permission denied\n

You can fix the permissions of your user-data directory as follows:

sudo chown -R $UID:$GID user_data\n
"},{"location":"data-download/#start-download","title":"Start download","text":"

Then run:

freqtrade download-data --exchange binance\n

This will download historical candle (OHLCV) data for all the currency pairs you defined in pairs.json.

Alternatively, specify the pairs directly

freqtrade download-data --exchange binance --pairs ETH/USDT XRP/USDT BTC/USDT\n

or as regex (to download all active USDT pairs)

freqtrade download-data --exchange binance --pairs .*/USDT\n
"},{"location":"data-download/#other-notes","title":"Other Notes","text":"
  • To use a different directory than the exchange specific default, use --datadir user_data/data/some_directory.
  • To change the exchange used to download the historical data from, please use a different configuration file (you'll probably need to adjust rate limits etc.)
  • To use pairs.json from some other directory, use --pairs-file some_other_dir/pairs.json.
  • To download historical candle (OHLCV) data for only 10 days, use --days 10 (defaults to 30 days).
  • To download historical candle (OHLCV) data from a fixed starting point, use --timerange 20200101- - which will download all data from January 1st, 2020.
  • Use --timeframes to specify what timeframe download the historical candle (OHLCV) data for. Default is --timeframes 1m 5m which will download 1-minute and 5-minute data.
  • To use exchange, timeframe and list of pairs as defined in your configuration file, use the -c/--config option. With this, the script uses the whitelist defined in the config as the list of currency pairs to download data for and does not require the pairs.json file. You can combine -c/--config with most other options.
"},{"location":"data-download/#download-additional-data-before-the-current-timerange","title":"Download additional data before the current timerange","text":"

Assuming you downloaded all data from 2022 (--timerange 20220101-) - but you'd now like to also backtest with earlier data. You can do so by using the --prepend flag, combined with --timerange - specifying an end-date.

freqtrade download-data --exchange binance --pairs ETH/USDT XRP/USDT BTC/USDT --prepend --timerange 20210101-20220101\n

Note

Freqtrade will ignore the end-date in this mode if data is available, updating the end-date to the existing data start point.

"},{"location":"data-download/#data-format","title":"Data format","text":"

Freqtrade currently supports the following data-formats:

  • json - plain \"text\" json files
  • jsongz - a gzip-zipped version of json files
  • hdf5 - a high performance datastore
  • feather - a dataformat based on Apache Arrow (OHLCV only)
  • parquet - columnar datastore (OHLCV only)

By default, OHLCV data is stored as json data, while trades data is stored as jsongz data.

This can be changed via the --data-format-ohlcv and --data-format-trades command line arguments respectively. To persist this change, you should also add the following snippet to your configuration, so you don't have to insert the above arguments each time:

    // ...\n    \"dataformat_ohlcv\": \"hdf5\",\n    \"dataformat_trades\": \"hdf5\",\n    // ...\n

If the default data-format has been changed during download, then the keys dataformat_ohlcv and dataformat_trades in the configuration file need to be adjusted to the selected dataformat as well.

Note

You can convert between data-formats using the convert-data and convert-trade-data methods.

"},{"location":"data-download/#dataformat-comparison","title":"Dataformat comparison","text":"

The following comparisons have been made with the following data, and by using the linux time command.

Found 6 pair / timeframe combinations.\n+----------+-------------+--------+---------------------+---------------------+\n|     Pair |   Timeframe |   Type |                From |                  To |\n|----------+-------------+--------+---------------------+---------------------|\n| BTC/USDT |          5m |   spot | 2017-08-17 04:00:00 | 2022-09-13 19:25:00 |\n| ETH/USDT |          1m |   spot | 2017-08-17 04:00:00 | 2022-09-13 19:26:00 |\n| BTC/USDT |          1m |   spot | 2017-08-17 04:00:00 | 2022-09-13 19:30:00 |\n| XRP/USDT |          5m |   spot | 2018-05-04 08:10:00 | 2022-09-13 19:15:00 |\n| XRP/USDT |          1m |   spot | 2018-05-04 08:11:00 | 2022-09-13 19:22:00 |\n| ETH/USDT |          5m |   spot | 2017-08-17 04:00:00 | 2022-09-13 19:20:00 |\n+----------+-------------+--------+---------------------+---------------------+\n

Timings have been taken in a not very scientific way with the following command, which forces reading the data into memory.

time freqtrade list-data --show-timerange --data-format-ohlcv <dataformat>\n
Format Size timing json 149Mb 25.6s jsongz 39Mb 27s hdf5 145Mb 3.9s feather 72Mb 3.5s parquet 83Mb 3.8s

Size has been taken from the BTC/USDT 1m spot combination for the timerange specified above.

To have a best performance/size mix, we recommend the use of either feather or parquet.

"},{"location":"data-download/#sub-command-convert-data","title":"Sub-command convert data","text":"
usage: freqtrade convert-data [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                              [-d PATH] [--userdir PATH]\n                              [-p PAIRS [PAIRS ...]] --format-from\n                              {json,jsongz,hdf5,feather,parquet} --format-to\n                              {json,jsongz,hdf5,feather,parquet} [--erase]\n                              [--exchange EXCHANGE]\n                              [-t TIMEFRAMES [TIMEFRAMES ...]]\n                              [--trading-mode {spot,margin,futures}]\n                              [--candle-types {spot,futures,mark,index,premiumIndex,funding_rate} [{spot,futures,mark,index,premiumIndex,funding_rate} ...]]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]\n                        Limit command to these pairs. Pairs are space-\n                        separated.\n  --format-from {json,jsongz,hdf5,feather,parquet}\n                        Source format for data conversion.\n  --format-to {json,jsongz,hdf5,feather,parquet}\n                        Destination format for data conversion.\n  --erase               Clean all existing data for the selected\n                        exchange/pairs/timeframes.\n  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no\n                        config is provided.\n  -t TIMEFRAMES [TIMEFRAMES ...], --timeframes TIMEFRAMES [TIMEFRAMES ...]\n                        Specify which tickers to download. Space-separated\n                        list. Default: `1m 5m`.\n  --trading-mode {spot,margin,futures}, --tradingmode {spot,margin,futures}\n                        Select Trading mode\n  --candle-types {spot,futures,mark,index,premiumIndex,funding_rate} [{spot,futures,mark,index,premiumIndex,funding_rate} ...]\n                        Select candle type to use\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH, --data-dir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n
"},{"location":"data-download/#example-converting-data","title":"Example converting data","text":"

The following command will convert all candle (OHLCV) data available in ~/.freqtrade/data/binance from json to jsongz, saving diskspace in the process. It'll also remove original json data files (--erase parameter).

freqtrade convert-data --format-from json --format-to jsongz --datadir ~/.freqtrade/data/binance -t 5m 15m --erase\n
"},{"location":"data-download/#sub-command-convert-trade-data","title":"Sub-command convert trade data","text":"
usage: freqtrade convert-trade-data [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                                    [-d PATH] [--userdir PATH]\n                                    [-p PAIRS [PAIRS ...]] --format-from\n                                    {json,jsongz,hdf5,feather,parquet}\n                                    --format-to\n                                    {json,jsongz,hdf5,feather,parquet}\n                                    [--erase] [--exchange EXCHANGE]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]\n                        Limit command to these pairs. Pairs are space-\n                        separated.\n  --format-from {json,jsongz,hdf5,feather,parquet}\n                        Source format for data conversion.\n  --format-to {json,jsongz,hdf5,feather,parquet}\n                        Destination format for data conversion.\n  --erase               Clean all existing data for the selected\n                        exchange/pairs/timeframes.\n  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no\n                        config is provided.\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH, --data-dir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n
"},{"location":"data-download/#example-converting-trades","title":"Example converting trades","text":"

The following command will convert all available trade-data in ~/.freqtrade/data/kraken from jsongz to json. It'll also remove original jsongz data files (--erase parameter).

freqtrade convert-trade-data --format-from jsongz --format-to json --datadir ~/.freqtrade/data/kraken --erase\n
"},{"location":"data-download/#sub-command-trades-to-ohlcv","title":"Sub-command trades to ohlcv","text":"

When you need to use --dl-trades (kraken only) to download data, conversion of trades data to ohlcv data is the last step. This command will allow you to repeat this last step for additional timeframes without re-downloading the data.

usage: freqtrade trades-to-ohlcv [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                                 [-d PATH] [--userdir PATH]\n                                 [-p PAIRS [PAIRS ...]]\n                                 [-t TIMEFRAMES [TIMEFRAMES ...]]\n                                 [--exchange EXCHANGE]\n                                 [--data-format-ohlcv {json,jsongz,hdf5,feather,parquet}]\n                                 [--data-format-trades {json,jsongz,hdf5}]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]\n                        Limit command to these pairs. Pairs are space-\n                        separated.\n  -t TIMEFRAMES [TIMEFRAMES ...], --timeframes TIMEFRAMES [TIMEFRAMES ...]\n                        Specify which tickers to download. Space-separated\n                        list. Default: `1m 5m`.\n  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no\n                        config is provided.\n  --data-format-ohlcv {json,jsongz,hdf5,feather,parquet}\n                        Storage format for downloaded candle (OHLCV) data.\n                        (default: `json`).\n  --data-format-trades {json,jsongz,hdf5}\n                        Storage format for downloaded trades data. (default:\n                        `jsongz`).\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH, --data-dir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n
"},{"location":"data-download/#example-trade-to-ohlcv-conversion","title":"Example trade-to-ohlcv conversion","text":"
freqtrade trades-to-ohlcv --exchange kraken -t 5m 1h 1d --pairs BTC/EUR ETH/EUR\n
"},{"location":"data-download/#sub-command-list-data","title":"Sub-command list-data","text":"

You can get a list of downloaded data using the list-data sub-command.

usage: freqtrade list-data [-h] [-v] [--logfile FILE] [-V] [-c PATH] [-d PATH]\n                           [--userdir PATH] [--exchange EXCHANGE]\n                           [--data-format-ohlcv {json,jsongz,hdf5,feather,parquet}]\n                           [-p PAIRS [PAIRS ...]]\n                           [--trading-mode {spot,margin,futures}]\n                           [--show-timerange]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no\n                        config is provided.\n  --data-format-ohlcv {json,jsongz,hdf5,feather,parquet}\n                        Storage format for downloaded candle (OHLCV) data.\n                        (default: `json`).\n  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]\n                        Limit command to these pairs. Pairs are space-\n                        separated.\n  --trading-mode {spot,margin,futures}, --tradingmode {spot,margin,futures}\n                        Select Trading mode\n  --show-timerange      Show timerange available for available data. (May take\n                        a while to calculate).\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH, --data-dir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n
"},{"location":"data-download/#example-list-data","title":"Example list-data","text":"
> freqtrade list-data --userdir ~/.freqtrade/user_data/\n\nFound 33 pair / timeframe combinations.\npairs       timeframe\n----------  -----------------------------------------\nADA/BTC     5m, 15m, 30m, 1h, 2h, 4h, 6h, 12h, 1d\nADA/ETH     5m, 15m, 30m, 1h, 2h, 4h, 6h, 12h, 1d\nETH/BTC     5m, 15m, 30m, 1h, 2h, 4h, 6h, 12h, 1d\nETH/USDT    5m, 15m, 30m, 1h, 2h, 4h\n
"},{"location":"data-download/#trades-tick-data","title":"Trades (tick) data","text":"

By default, download-data sub-command downloads Candles (OHLCV) data. Some exchanges also provide historic trade-data via their API. This data can be useful if you need many different timeframes, since it is only downloaded once, and then resampled locally to the desired timeframes.

Since this data is large by default, the files use gzip by default. They are stored in your data-directory with the naming convention of <pair>-trades.json.gz (ETH_BTC-trades.json.gz). Incremental mode is also supported, as for historic OHLCV data, so downloading the data once per week with --days 8 will create an incremental data-repository.

To use this mode, simply add --dl-trades to your call. This will swap the download method to download trades, and resamples the data locally.

do not use

You should not use this unless you're a kraken user. Most other exchanges provide OHLCV data with sufficient history.

Example call:

freqtrade download-data --exchange kraken --pairs XRP/EUR ETH/EUR --days 20 --dl-trades\n

Note

While this method uses async calls, it will be slow, since it requires the result of the previous call to generate the next request to the exchange.

Warning

The historic trades are not available during Freqtrade dry-run and live trade modes because all exchanges tested provide this data with a delay of few 100 candles, so it's not suitable for real-time trading.

Kraken user

Kraken users should read this before starting to download data.

"},{"location":"data-download/#next-step","title":"Next step","text":"

Great, you now have backtest data downloaded, so you can now start backtesting your strategy.

"},{"location":"deprecated/","title":"Deprecated features","text":"

This page contains description of the command line arguments, configuration parameters and the bot features that were declared as DEPRECATED by the bot development team and are no longer supported. Please avoid their usage in your configuration.

"},{"location":"deprecated/#removed-features","title":"Removed features","text":""},{"location":"deprecated/#the-refresh-pairs-cached-command-line-option","title":"the --refresh-pairs-cached command line option","text":"

--refresh-pairs-cached in the context of backtesting, hyperopt and edge allows to refresh candle data for backtesting. Since this leads to much confusion, and slows down backtesting (while not being part of backtesting) this has been singled out as a separate freqtrade sub-command freqtrade download-data.

This command line option was deprecated in 2019.7-dev (develop branch) and removed in 2019.9.

"},{"location":"deprecated/#the-dynamic-whitelist-command-line-option","title":"The --dynamic-whitelist command line option","text":"

This command line option was deprecated in 2018 and removed freqtrade 2019.6-dev (develop branch) and in freqtrade 2019.7. Please refer to pairlists instead.

"},{"location":"deprecated/#the-live-command-line-option","title":"the --live command line option","text":"

--live in the context of backtesting allowed to download the latest tick data for backtesting. Did only download the latest 500 candles, so was ineffective in getting good backtest data. Removed in 2019-7-dev (develop branch) and in freqtrade 2019.8.

"},{"location":"deprecated/#ticker_interval-now-timeframe","title":"ticker_interval (now timeframe)","text":"

Support for ticker_interval terminology was deprecated in 2020.6 in favor of timeframe - and compatibility code was removed in 2022.3.

"},{"location":"deprecated/#allow-running-multiple-pairlists-in-sequence","title":"Allow running multiple pairlists in sequence","text":"

The former \"pairlist\" section in the configuration has been removed, and is replaced by \"pairlists\" - being a list to specify a sequence of pairlists.

The old section of configuration parameters (\"pairlist\") has been deprecated in 2019.11 and has been removed in 2020.4.

"},{"location":"deprecated/#deprecation-of-bidvolume-and-askvolume-from-volume-pairlist","title":"deprecation of bidVolume and askVolume from volume-pairlist","text":"

Since only quoteVolume can be compared between assets, the other options (bidVolume, askVolume) have been deprecated in 2020.4, and have been removed in 2020.9.

"},{"location":"deprecated/#using-order-book-steps-for-exit-price","title":"Using order book steps for exit price","text":"

Using order_book_min and order_book_max used to allow stepping the orderbook and trying to find the next ROI slot - trying to place sell-orders early. As this does however increase risk and provides no benefit, it's been removed for maintainability purposes in 2021.7.

"},{"location":"deprecated/#legacy-hyperopt-mode","title":"Legacy Hyperopt mode","text":"

Using separate hyperopt files was deprecated in 2021.4 and was removed in 2021.9. Please switch to the new Parametrized Strategies to benefit from the new hyperopt interface.

"},{"location":"deprecated/#strategy-changes-between-v2-and-v3","title":"Strategy changes between V2 and V3","text":"

Isolated Futures / short trading was introduced in 2022.4. This required major changes to configuration settings, strategy interfaces, ...

We have put a great effort into keeping compatibility with existing strategies, so if you just want to continue using freqtrade in spot markets, there are no changes necessary. While we may drop support for the current interface sometime in the future, we will announce this separately and have an appropriate transition period.

Please follow the Strategy migration guide to migrate your strategy to the new format to start using the new functionalities.

"},{"location":"deprecated/#webhooks-changes-with-20224","title":"webhooks - changes with 2022.4","text":""},{"location":"deprecated/#buy_tag-has-been-renamed-to-enter_tag","title":"buy_tag has been renamed to enter_tag","text":"

This should apply only to your strategy and potentially to webhooks. We will keep a compatibility layer for 1-2 versions (so both buy_tag and enter_tag will still work), but support for this in webhooks will disappear after that.

"},{"location":"deprecated/#naming-changes","title":"Naming changes","text":"

Webhook terminology changed from \"sell\" to \"exit\", and from \"buy\" to \"entry\", removing \"webhook\" in the process.

  • webhookbuy, webhookentry -> entry
  • webhookbuyfill, webhookentryfill -> entry_fill
  • webhookbuycancel, webhookentrycancel -> entry_cancel
  • webhooksell, webhookexit -> exit
  • webhooksellfill, webhookexitfill -> exit_fill
  • webhooksellcancel, webhookexitcancel -> exit_cancel
"},{"location":"deprecated/#removal-of-populate_any_indicators","title":"Removal of populate_any_indicators","text":"

version 2023.3 saw the removal of populate_any_indicators in favor of split methods for feature engineering and targets. Please read the migration document for full details.

"},{"location":"developer/","title":"Development Help","text":"

This page is intended for developers of Freqtrade, people who want to contribute to the Freqtrade codebase or documentation, or people who want to understand the source code of the application they're running.

All contributions, bug reports, bug fixes, documentation improvements, enhancements and ideas are welcome. We track issues on GitHub and also have a dev channel on discord where you can ask questions.

"},{"location":"developer/#documentation","title":"Documentation","text":"

Documentation is available at https://freqtrade.io and needs to be provided with every new feature PR.

Special fields for the documentation (like Note boxes, ...) can be found here.

To test the documentation locally use the following commands.

pip install -r docs/requirements-docs.txt\nmkdocs serve\n

This will spin up a local server (usually on port 8000) so you can see if everything looks as you'd like it to.

"},{"location":"developer/#developer-setup","title":"Developer setup","text":"

To configure a development environment, you can either use the provided DevContainer, or use the setup.sh script and answer \"y\" when asked \"Do you want to install dependencies for dev [y/N]? \". Alternatively (e.g. if your system is not supported by the setup.sh script), follow the manual installation process and run pip3 install -e .[all].

This will install all required tools for development, including pytest, ruff, mypy, and coveralls.

Then install the git hook scripts by running pre-commit install, so your changes will be verified locally before committing. This avoids a lot of waiting for CI already, as some basic formatting checks are done locally on your machine.

Before opening a pull request, please familiarize yourself with our Contributing Guidelines.

"},{"location":"developer/#devcontainer-setup","title":"Devcontainer setup","text":"

The fastest and easiest way to get started is to use VSCode with the Remote container extension. This gives developers the ability to start the bot with all required dependencies without needing to install any freqtrade specific dependencies on your local machine.

"},{"location":"developer/#devcontainer-dependencies","title":"Devcontainer dependencies","text":"
  • VSCode
  • docker
  • Remote container extension documentation

For more information about the Remote container extension, best consult the documentation.

"},{"location":"developer/#tests","title":"Tests","text":"

New code should be covered by basic unittests. Depending on the complexity of the feature, Reviewers may request more in-depth unittests. If necessary, the Freqtrade team can assist and give guidance with writing good tests (however please don't expect anyone to write the tests for you).

"},{"location":"developer/#how-to-run-tests","title":"How to run tests","text":"

Use pytest in root folder to run all available testcases and confirm your local environment is setup correctly

feature branches

Tests are expected to pass on the develop and stable branches. Other branches may be work in progress with tests not working yet.

"},{"location":"developer/#checking-log-content-in-tests","title":"Checking log content in tests","text":"

Freqtrade uses 2 main methods to check log content in tests, log_has() and log_has_re() (to check using regex, in case of dynamic log-messages). These are available from conftest.py and can be imported in any test module.

A sample check looks as follows:

from tests.conftest import log_has, log_has_re\n\ndef test_method_to_test(caplog):\n    method_to_test()\n\n    assert log_has(\"This event happened\", caplog)\n    # Check regex with trailing number ...\n    assert log_has_re(r\"This dynamic event happened and produced \\d+\", caplog)\n
"},{"location":"developer/#debug-configuration","title":"Debug configuration","text":"

To debug freqtrade, we recommend VSCode with the following launch configuration (located in .vscode/launch.json). Details will obviously vary between setups - but this should work to get you started.

{\n\"name\": \"freqtrade trade\",\n\"type\": \"python\",\n\"request\": \"launch\",\n\"module\": \"freqtrade\",\n\"console\": \"integratedTerminal\",\n\"args\": [\n\"trade\",\n// Optional:\n// \"--userdir\", \"user_data\",\n\"--strategy\", \"MyAwesomeStrategy\",\n]\n},\n

Command line arguments can be added in the \"args\" array. This method can also be used to debug a strategy, by setting the breakpoints within the strategy.

A similar setup can also be taken for Pycharm - using freqtrade as module name, and setting the command line arguments as \"parameters\".

Startup directory

This assumes that you have the repository checked out, and the editor is started at the repository root level (so setup.py is at the top level of your repository).

"},{"location":"developer/#errorhandling","title":"ErrorHandling","text":"

Freqtrade Exceptions all inherit from FreqtradeException. This general class of error should however not be used directly. Instead, multiple specialized sub-Exceptions exist.

Below is an outline of exception inheritance hierarchy:

+ FreqtradeException\n|\n+---+ OperationalException\n|\n+---+ DependencyException\n|   |\n|   +---+ PricingError\n|   |\n|   +---+ ExchangeError\n|       |\n|       +---+ TemporaryError\n|       |\n|       +---+ DDosProtection\n|       |\n|       +---+ InvalidOrderException\n|           |\n|           +---+ RetryableOrderError\n|           |\n|           +---+ InsufficientFundsError\n|\n+---+ StrategyError\n
"},{"location":"developer/#plugins","title":"Plugins","text":""},{"location":"developer/#pairlists","title":"Pairlists","text":"

You have a great idea for a new pair selection algorithm you would like to try out? Great. Hopefully you also want to contribute this back upstream.

Whatever your motivations are - This should get you off the ground in trying to develop a new Pairlist Handler.

First of all, have a look at the VolumePairList Handler, and best copy this file with a name of your new Pairlist Handler.

This is a simple Handler, which however serves as a good example on how to start developing.

Next, modify the class-name of the Handler (ideally align this with the module filename).

The base-class provides an instance of the exchange (self._exchange) the pairlist manager (self._pairlistmanager), as well as the main configuration (self._config), the pairlist dedicated configuration (self._pairlistconfig) and the absolute position within the list of pairlists.

        self._exchange = exchange\n        self._pairlistmanager = pairlistmanager\n        self._config = config\n        self._pairlistconfig = pairlistconfig\n        self._pairlist_pos = pairlist_pos\n

Tip

Don't forget to register your pairlist in constants.py under the variable AVAILABLE_PAIRLISTS - otherwise it will not be selectable.

Now, let's step through the methods which require actions:

"},{"location":"developer/#pairlist-configuration","title":"Pairlist configuration","text":"

Configuration for the chain of Pairlist Handlers is done in the bot configuration file in the element \"pairlists\", an array of configuration parameters for each Pairlist Handlers in the chain.

By convention, \"number_assets\" is used to specify the maximum number of pairs to keep in the pairlist. Please follow this to ensure a consistent user experience.

Additional parameters can be configured as needed. For instance, VolumePairList uses \"sort_key\" to specify the sorting value - however feel free to specify whatever is necessary for your great algorithm to be successful and dynamic.

"},{"location":"developer/#short_desc","title":"short_desc","text":"

Returns a description used for Telegram messages.

This should contain the name of the Pairlist Handler, as well as a short description containing the number of assets. Please follow the format \"PairlistName - top/bottom X pairs\".

"},{"location":"developer/#gen_pairlist","title":"gen_pairlist","text":"

Override this method if the Pairlist Handler can be used as the leading Pairlist Handler in the chain, defining the initial pairlist which is then handled by all Pairlist Handlers in the chain. Examples are StaticPairList and VolumePairList.

This is called with each iteration of the bot (only if the Pairlist Handler is at the first location) - so consider implementing caching for compute/network heavy calculations.

It must return the resulting pairlist (which may then be passed into the chain of Pairlist Handlers).

Validations are optional, the parent class exposes a _verify_blacklist(pairlist) and _whitelist_for_active_markets(pairlist) to do default filtering. Use this if you limit your result to a certain number of pairs - so the end-result is not shorter than expected.

"},{"location":"developer/#filter_pairlist","title":"filter_pairlist","text":"

This method is called for each Pairlist Handler in the chain by the pairlist manager.

This is called with each iteration of the bot - so consider implementing caching for compute/network heavy calculations.

It gets passed a pairlist (which can be the result of previous pairlists) as well as tickers, a pre-fetched version of get_tickers().

The default implementation in the base class simply calls the _validate_pair() method for each pair in the pairlist, but you may override it. So you should either implement the _validate_pair() in your Pairlist Handler or override filter_pairlist() to do something else.

If overridden, it must return the resulting pairlist (which may then be passed into the next Pairlist Handler in the chain).

Validations are optional, the parent class exposes a _verify_blacklist(pairlist) and _whitelist_for_active_markets(pairlist) to do default filters. Use this if you limit your result to a certain number of pairs - so the end result is not shorter than expected.

In VolumePairList, this implements different methods of sorting, does early validation so only the expected number of pairs is returned.

"},{"location":"developer/#sample","title":"sample","text":"
    def filter_pairlist(self, pairlist: List[str], tickers: Dict) -> List[str]:\n        # Generate dynamic whitelist\n        pairs = self._calculate_pairlist(pairlist, tickers)\n        return pairs\n
"},{"location":"developer/#protections","title":"Protections","text":"

Best read the Protection documentation to understand protections. This Guide is directed towards Developers who want to develop a new protection.

No protection should use datetime directly, but use the provided date_now variable for date calculations. This preserves the ability to backtest protections.

Writing a new Protection

Best copy one of the existing Protections to have a good example. Don't forget to register your protection in constants.py under the variable AVAILABLE_PROTECTIONS - otherwise it will not be selectable.

"},{"location":"developer/#implementation-of-a-new-protection","title":"Implementation of a new protection","text":"

All Protection implementations must have IProtection as parent class. For that reason, they must implement the following methods:

  • short_desc()
  • global_stop()
  • stop_per_pair().

global_stop() and stop_per_pair() must return a ProtectionReturn object, which consists of:

  • lock pair - boolean
  • lock until - datetime - until when should the pair be locked (will be rounded up to the next new candle)
  • reason - string, used for logging and storage in the database
  • lock_side - long, short or '*'.

The until portion should be calculated using the provided calculate_lock_end() method.

All Protections should use \"stop_duration\" / \"stop_duration_candles\" to define how long a a pair (or all pairs) should be locked. The content of this is made available as self._stop_duration to the each Protection.

If your protection requires a look-back period, please use \"lookback_period\" / \"lockback_period_candles\" to keep all protections aligned.

"},{"location":"developer/#global-vs-local-stops","title":"Global vs. local stops","text":"

Protections can have 2 different ways to stop trading for a limited :

  • Per pair (local)
  • For all Pairs (globally)
"},{"location":"developer/#protections-per-pair","title":"Protections - per pair","text":"

Protections that implement the per pair approach must set has_local_stop=True. The method stop_per_pair() will be called whenever a trade closed (exit order completed).

"},{"location":"developer/#protections-global-protection","title":"Protections - global protection","text":"

These Protections should do their evaluation across all pairs, and consequently will also lock all pairs from trading (called a global PairLock). Global protection must set has_global_stop=True to be evaluated for global stops. The method global_stop() will be called whenever a trade closed (exit order completed).

"},{"location":"developer/#protections-calculating-lock-end-time","title":"Protections - calculating lock end time","text":"

Protections should calculate the lock end time based on the last trade it considers. This avoids re-locking should the lookback-period be longer than the actual lock period.

The IProtection parent class provides a helper method for this in calculate_lock_end().

"},{"location":"developer/#implement-a-new-exchange-wip","title":"Implement a new Exchange (WIP)","text":"

Note

This section is a Work in Progress and is not a complete guide on how to test a new exchange with Freqtrade.

Note

Make sure to use an up-to-date version of CCXT before running any of the below tests. You can get the latest version of ccxt by running pip install -U ccxt with activated virtual environment. Native docker is not supported for these tests, however the available dev-container will support all required actions and eventually necessary changes.

Most exchanges supported by CCXT should work out of the box.

To quickly test the public endpoints of an exchange, add a configuration for your exchange to test_ccxt_compat.py and run these tests with pytest --longrun tests/exchange/test_ccxt_compat.py. Completing these tests successfully a good basis point (it's a requirement, actually), however these won't guarantee correct exchange functioning, as this only tests public endpoints, but no private endpoint (like generate order or similar).

Also try to use freqtrade download-data for an extended timerange (multiple months) and verify that the data downloaded correctly (no holes, the specified timerange was actually downloaded).

These are prerequisites to have an exchange listed as either Supported or Community tested (listed on the homepage). The below are \"extras\", which will make an exchange better (feature-complete) - but are not absolutely necessary for either of the 2 categories.

Additional tests / steps to complete:

  • Verify data provided by fetch_ohlcv() - and eventually adjust ohlcv_candle_limit for this exchange
  • Check L2 orderbook limit range (API documentation) - and eventually set as necessary
  • Check if balance shows correctly (*)
  • Create market order (*)
  • Create limit order (*)
  • Complete trade (enter + exit) (*)
    • Compare result calculation between exchange and bot
    • Ensure fees are applied correctly (check the database against the exchange)

(*) Requires API keys and Balance on the exchange.

"},{"location":"developer/#stoploss-on-exchange","title":"Stoploss On Exchange","text":"

Check if the new exchange supports Stoploss on Exchange orders through their API.

Since CCXT does not provide unification for Stoploss On Exchange yet, we'll need to implement the exchange-specific parameters ourselves. Best look at binance.py for an example implementation of this. You'll need to dig through the documentation of the Exchange's API on how exactly this can be done. CCXT Issues may also provide great help, since others may have implemented something similar for their projects.

"},{"location":"developer/#incomplete-candles","title":"Incomplete candles","text":"

While fetching candle (OHLCV) data, we may end up getting incomplete candles (depending on the exchange). To demonstrate this, we'll use daily candles (\"1d\") to keep things simple. We query the api (ct.fetch_ohlcv()) for the timeframe and look at the date of the last entry. If this entry changes or shows the date of a \"incomplete\" candle, then we should drop this since having incomplete candles is problematic because indicators assume that only complete candles are passed to them, and will generate a lot of false buy signals. By default, we're therefore removing the last candle assuming it's incomplete.

To check how the new exchange behaves, you can use the following snippet:

import ccxt\nfrom datetime import datetime\nfrom freqtrade.data.converter import ohlcv_to_dataframe\nct = ccxt.binance()\ntimeframe = \"1d\"\npair = \"XLM/BTC\"  # Make sure to use a pair that exists on that exchange!\nraw = ct.fetch_ohlcv(pair, timeframe=timeframe)\n\n# convert to dataframe\ndf1 = ohlcv_to_dataframe(raw, timeframe, pair=pair, drop_incomplete=False)\n\nprint(df1.tail(1))\nprint(datetime.utcnow())\n
                         date      open      high       low     close  volume  \n499 2019-06-08 00:00:00+00:00  0.000007  0.000007  0.000007  0.000007   26264344.0  \n2019-06-09 12:30:27.873327\n

The output will show the last entry from the Exchange as well as the current UTC date. If the day shows the same day, then the last candle can be assumed as incomplete and should be dropped (leave the setting \"ohlcv_partial_candle\" from the exchange-class untouched / True). Otherwise, set \"ohlcv_partial_candle\" to False to not drop Candles (shown in the example above). Another way is to run this command multiple times in a row and observe if the volume is changing (while the date remains the same).

"},{"location":"developer/#update-binance-cached-leverage-tiers","title":"Update binance cached leverage tiers","text":"

Updating leveraged tiers should be done regularly - and requires an authenticated account with futures enabled.

import ccxt\nimport json\nfrom pathlib import Path\n\nexchange = ccxt.binance({\n    'apiKey': '<apikey>',\n    'secret': '<secret>'\n    'options': {'defaultType': 'swap'}\n    })\n_ = exchange.load_markets()\n\nlev_tiers = exchange.fetch_leverage_tiers()\n\n# Assumes this is running in the root of the repository.\nfile = Path('freqtrade/exchange/binance_leverage_tiers.json')\njson.dump(dict(sorted(lev_tiers.items())), file.open('w'), indent=2)\n

This file should then be contributed upstream, so others can benefit from this, too.

"},{"location":"developer/#updating-example-notebooks","title":"Updating example notebooks","text":"

To keep the jupyter notebooks aligned with the documentation, the following should be ran after updating a example notebook.

jupyter nbconvert --ClearOutputPreprocessor.enabled=True --inplace freqtrade/templates/strategy_analysis_example.ipynb\njupyter nbconvert --ClearOutputPreprocessor.enabled=True --to markdown freqtrade/templates/strategy_analysis_example.ipynb --stdout > docs/strategy_analysis_example.md\n
"},{"location":"developer/#continuous-integration","title":"Continuous integration","text":"

This documents some decisions taken for the CI Pipeline.

  • CI runs on all OS variants, Linux (ubuntu), macOS and Windows.
  • Docker images are build for the branches stable and develop, and are built as multiarch builds, supporting multiple platforms via the same tag.
  • Docker images containing Plot dependencies are also available as stable_plot and develop_plot.
  • Docker images contain a file, /freqtrade/freqtrade_commit containing the commit this image is based of.
  • Full docker image rebuilds are run once a week via schedule.
  • Deployments run on ubuntu.
  • ta-lib binaries are contained in the build_helpers directory to avoid fails related to external unavailability.
  • All tests must pass for a PR to be merged to stable or develop.
"},{"location":"developer/#creating-a-release","title":"Creating a release","text":"

This part of the documentation is aimed at maintainers, and shows how to create a release.

"},{"location":"developer/#create-release-branch","title":"Create release branch","text":"

First, pick a commit that's about one week old (to not include latest additions to releases).

# create new branch\ngit checkout -b new_release <commitid>\n

Determine if crucial bugfixes have been made between this commit and the current state, and eventually cherry-pick these.

  • Merge the release branch (stable) into this branch.
  • Edit freqtrade/__init__.py and add the version matching the current date (for example 2019.7 for July 2019). Minor versions can be 2019.7.1 should we need to do a second release that month. Version numbers must follow allowed versions from PEP0440 to avoid failures pushing to pypi.
  • Commit this part.
  • push that branch to the remote and create a PR against the stable branch.
  • Update develop version to next version following the pattern 2019.8-dev.
"},{"location":"developer/#create-changelog-from-git-commits","title":"Create changelog from git commits","text":"

Note

Make sure that the stable branch is up-to-date!

# Needs to be done before merging / pulling that branch.\ngit log --oneline --no-decorate --no-merges stable..new_release\n

To keep the release-log short, best wrap the full git changelog into a collapsible details section.

<details>\n<summary>Expand full changelog</summary>\n\n... Full git changelog\n\n</details>\n
"},{"location":"developer/#frequi-release","title":"FreqUI release","text":"

If FreqUI has been updated substantially, make sure to create a release before merging the release branch. Make sure that freqUI CI on the release is finished and passed before merging the release.

"},{"location":"developer/#create-github-release-tag","title":"Create github release / tag","text":"

Once the PR against stable is merged (best right after merging):

  • Use the button \"Draft a new release\" in the Github UI (subsection releases).
  • Use the version-number specified as tag.
  • Use \"stable\" as reference (this step comes after the above PR is merged).
  • Use the above changelog as release comment (as codeblock)
"},{"location":"developer/#releases","title":"Releases","text":""},{"location":"developer/#pypi","title":"pypi","text":"

Note

This process is now automated as part of Github Actions.

To create a pypi release, please run the following commands:

Additional requirement: wheel, twine (for uploading), account on pypi with proper permissions.

python setup.py sdist bdist_wheel\n\n# For pypi test (to check if some change to the installation did work)\ntwine upload --repository-url https://test.pypi.org/legacy/ dist/*\n\n# For production:\ntwine upload dist/*\n

Please don't push non-releases to the productive / real pypi instance.

"},{"location":"docker_quickstart/","title":"Using Freqtrade with Docker","text":"

This page explains how to run the bot with Docker. It is not meant to work out of the box. You'll still need to read through the documentation and understand how to properly configure it.

"},{"location":"docker_quickstart/#install-docker","title":"Install Docker","text":"

Start by downloading and installing Docker / Docker Desktop for your platform:

  • Mac
  • Windows
  • Linux

Docker compose install

Freqtrade documentation assumes the use of Docker desktop (or the docker compose plugin). While the docker-compose standalone installation still works, it will require changing all docker compose commands from docker compose to docker-compose to work (e.g. docker compose up -d will become docker-compose up -d).

"},{"location":"docker_quickstart/#freqtrade-with-docker","title":"Freqtrade with docker","text":"

Freqtrade provides an official Docker image on Dockerhub, as well as a docker compose file ready for usage.

Note

  • The following section assumes that docker is installed and available to the logged in user.
  • All below commands use relative directories and will have to be executed from the directory containing the docker-compose.yml file.
"},{"location":"docker_quickstart/#docker-quick-start","title":"Docker quick start","text":"

Create a new directory and place the docker-compose file in this directory.

mkdir ft_userdata\ncd ft_userdata/\n# Download the docker-compose file from the repository\ncurl https://raw.githubusercontent.com/freqtrade/freqtrade/stable/docker-compose.yml -o docker-compose.yml\n\n# Pull the freqtrade image\ndocker compose pull\n\n# Create user directory structure\ndocker compose run --rm freqtrade create-userdir --userdir user_data\n\n# Create configuration - Requires answering interactive questions\ndocker compose run --rm freqtrade new-config --config user_data/config.json\n

The above snippet creates a new directory called ft_userdata, downloads the latest compose file and pulls the freqtrade image. The last 2 steps in the snippet create the directory with user_data, as well as (interactively) the default configuration based on your selections.

How to edit the bot configuration?

You can edit the configuration at any time, which is available as user_data/config.json (within the directory ft_userdata) when using the above configuration.

You can also change the both Strategy and commands by editing the command section of your docker-compose.yml file.

"},{"location":"docker_quickstart/#adding-a-custom-strategy","title":"Adding a custom strategy","text":"
  1. The configuration is now available as user_data/config.json
  2. Copy a custom strategy to the directory user_data/strategies/
  3. Add the Strategy' class name to the docker-compose.yml file

The SampleStrategy is run by default.

SampleStrategy is just a demo!

The SampleStrategy is there for your reference and give you ideas for your own strategy. Please always backtest your strategy and use dry-run for some time before risking real money! You will find more information about Strategy development in the Strategy documentation.

Once this is done, you're ready to launch the bot in trading mode (Dry-run or Live-trading, depending on your answer to the corresponding question you made above).

docker compose up -d\n

Default configuration

While the configuration generated will be mostly functional, you will still need to verify that all options correspond to what you want (like Pricing, pairlist, ...) before starting the bot.

"},{"location":"docker_quickstart/#accessing-the-ui","title":"Accessing the UI","text":"

If you've selected to enable FreqUI in the new-config step, you will have freqUI available at port localhost:8080.

You can now access the UI by typing localhost:8080 in your browser.

UI Access on a remote servers

If you're running on a VPS, you should consider using either a ssh tunnel, or setup a VPN (openVPN, wireguard) to connect to your bot. This will ensure that freqUI is not directly exposed to the internet, which is not recommended for security reasons (freqUI does not support https out of the box). Setup of these tools is not part of this tutorial, however many good tutorials can be found on the internet. Please also read the API configuration with docker section to learn more about this configuration.

"},{"location":"docker_quickstart/#monitoring-the-bot","title":"Monitoring the bot","text":"

You can check for running instances with docker compose ps. This should list the service freqtrade as running. If that's not the case, best check the logs (see next point).

"},{"location":"docker_quickstart/#docker-compose-logs","title":"Docker compose logs","text":"

Logs will be written to: user_data/logs/freqtrade.log. You can also check the latest log with the command docker compose logs -f.

"},{"location":"docker_quickstart/#database","title":"Database","text":"

The database will be located at: user_data/tradesv3.sqlite

"},{"location":"docker_quickstart/#updating-freqtrade-with-docker","title":"Updating freqtrade with docker","text":"

Updating freqtrade when using docker is as simple as running the following 2 commands:

# Download the latest image\ndocker compose pull\n# Restart the image\ndocker compose up -d\n

This will first pull the latest image, and will then restart the container with the just pulled version.

Check the Changelog

You should always check the changelog for breaking changes / manual interventions required and make sure the bot starts correctly after the update.

"},{"location":"docker_quickstart/#editing-the-docker-compose-file","title":"Editing the docker-compose file","text":"

Advanced users may edit the docker-compose file further to include all possible options or arguments.

All freqtrade arguments will be available by running docker compose run --rm freqtrade <command> <optional arguments>.

docker compose for trade commands

Trade commands (freqtrade trade <...>) should not be ran via docker compose run - but should use docker compose up -d instead. This makes sure that the container is properly started (including port forwardings) and will make sure that the container will restart after a system reboot. If you intend to use freqUI, please also ensure to adjust the configuration accordingly, otherwise the UI will not be available.

docker compose run --rm

Including --rm will remove the container after completion, and is highly recommended for all modes except trading mode (running with freqtrade trade command).

Using docker without docker

\"docker compose run --rm\" will require a compose file to be provided. Some freqtrade commands that don't require authentication such as list-pairs can be run with \"docker run --rm\" instead. For example docker run --rm freqtradeorg/freqtrade:stable list-pairs --exchange binance --quote BTC --print-json. This can be useful for fetching exchange information to add to your config.json without affecting your running containers.

"},{"location":"docker_quickstart/#example-download-data-with-docker","title":"Example: Download data with docker","text":"

Download backtesting data for 5 days for the pair ETH/BTC and 1h timeframe from Binance. The data will be stored in the directory user_data/data/ on the host.

docker compose run --rm freqtrade download-data --pairs ETH/BTC --exchange binance --days 5 -t 1h\n

Head over to the Data Downloading Documentation for more details on downloading data.

"},{"location":"docker_quickstart/#example-backtest-with-docker","title":"Example: Backtest with docker","text":"

Run backtesting in docker-containers for SampleStrategy and specified timerange of historical data, on 5m timeframe:

docker compose run --rm freqtrade backtesting --config user_data/config.json --strategy SampleStrategy --timerange 20190801-20191001 -i 5m\n

Head over to the Backtesting Documentation to learn more.

"},{"location":"docker_quickstart/#additional-dependencies-with-docker","title":"Additional dependencies with docker","text":"

If your strategy requires dependencies not included in the default image - it will be necessary to build the image on your host. For this, please create a Dockerfile containing installation steps for the additional dependencies (have a look at docker/Dockerfile.custom for an example).

You'll then also need to modify the docker-compose.yml file and uncomment the build step, as well as rename the image to avoid naming collisions.

    image: freqtrade_custom\nbuild:\ncontext: .\ndockerfile: \"./Dockerfile.<yourextension>\"\n

You can then run docker compose build --pull to build the docker image, and run it using the commands described above.

"},{"location":"docker_quickstart/#plotting-with-docker","title":"Plotting with docker","text":"

Commands freqtrade plot-profit and freqtrade plot-dataframe (Documentation) are available by changing the image to *_plot in your docker-compose.yml file. You can then use these commands as follows:

docker compose run --rm freqtrade plot-dataframe --strategy AwesomeStrategy -p BTC/ETH --timerange=20180801-20180805\n

The output will be stored in the user_data/plot directory, and can be opened with any modern browser.

"},{"location":"docker_quickstart/#data-analysis-using-docker-compose","title":"Data analysis using docker compose","text":"

Freqtrade provides a docker-compose file which starts up a jupyter lab server. You can run this server using the following command:

docker compose -f docker/docker-compose-jupyter.yml up\n

This will create a docker-container running jupyter lab, which will be accessible using https://127.0.0.1:8888/lab. Please use the link that's printed in the console after startup for simplified login.

Since part of this image is built on your machine, it is recommended to rebuild the image from time to time to keep freqtrade (and dependencies) up-to-date.

docker compose -f docker/docker-compose-jupyter.yml build --no-cache\n
"},{"location":"docker_quickstart/#troubleshooting","title":"Troubleshooting","text":""},{"location":"docker_quickstart/#docker-on-windows","title":"Docker on Windows","text":"
  • Error: \"Timestamp for this request is outside of the recvWindow.\"
    • The market api requests require a synchronized clock but the time in the docker container shifts a bit over time into the past. To fix this issue temporarily you need to run wsl --shutdown and restart docker again (a popup on windows 10 will ask you to do so). A permanent solution is either to host the docker container on a linux host or restart the wsl from time to time with the scheduler.

      taskkill /IM \"Docker Desktop.exe\" /F\nwsl --shutdown\nstart \"\" \"C:\\Program Files\\Docker\\Docker\\Docker Desktop.exe\"\n

Warning

Due to the above, we do not recommend the usage of docker on windows for production setups, but only for experimentation, datadownload and backtesting. Best use a linux-VPS for running freqtrade reliably.

"},{"location":"edge/","title":"Edge positioning","text":"

The Edge Positioning module uses probability to calculate your win rate and risk reward ratio. It will use these statistics to control your strategy trade entry points, position size and, stoploss.

Warning

When using Edge positioning with a dynamic whitelist (VolumePairList), make sure to also use AgeFilter and set it to at least calculate_since_number_of_days to avoid problems with missing data.

Note

Edge Positioning only considers its own buy/sell/stoploss signals. It ignores the stoploss, trailing stoploss, and ROI settings in the strategy configuration file. Edge Positioning improves the performance of some trading strategies and decreases the performance of others.

"},{"location":"edge/#introduction","title":"Introduction","text":"

Trading strategies are not perfect. They are frameworks that are susceptible to the market and its indicators. Because the market is not at all predictable, sometimes a strategy will win and sometimes the same strategy will lose.

To obtain an edge in the market, a strategy has to make more money than it loses. Making money in trading is not only about how often the strategy makes or loses money.

It doesn't matter how often, but how much!

A bad strategy might make 1 penny in ten transactions but lose 1 dollar in one transaction. If one only checks the number of winning trades, it would be misleading to think that the strategy is actually making a profit.

The Edge Positioning module seeks to improve a strategy's winning probability and the money that the strategy will make on the long run.

We raise the following question1:

Which trade is a better option?

a) A trade with 80% of chance of losing 100$ and 20% chance of winning 200$ b) A trade with 100% of chance of losing 30$

Answer

The expected value of a) is smaller than the expected value of b). Hence, b) represents a smaller loss in the long run. However, the answer is: it depends

Another way to look at it is to ask a similar question:

Which trade is a better option?

a) A trade with 80% of chance of winning 100$ and 20% chance of losing 200$ b) A trade with 100% of chance of winning 30$

Edge positioning tries to answer the hard questions about risk/reward and position size automatically, seeking to minimizes the chances of losing of a given strategy.

"},{"location":"edge/#trading-winning-and-losing","title":"Trading, winning and losing","text":"

Let's call \\(o\\) the return of a single transaction \\(o\\) where \\(o \\in \\mathbb{R}\\). The collection \\(O = \\{o_1, o_2, ..., o_N\\}\\) is the set of all returns of transactions made during a trading session. We say that \\(N\\) is the cardinality of \\(O\\), or, in lay terms, it is the number of transactions made in a trading session.

Example

In a session where a strategy made three transactions we can say that \\(O = \\{3.5, -1, 15\\}\\). That means that \\(N = 3\\) and \\(o_1 = 3.5\\), \\(o_2 = -1\\), \\(o_3 = 15\\).

A winning trade is a trade where a strategy made money. Making money means that the strategy closed the position in a value that returned a profit, after all deducted fees. Formally, a winning trade will have a return \\(o_i > 0\\). Similarly, a losing trade will have a return \\(o_j \\leq 0\\). With that, we can discover the set of all winning trades, \\(T_{win}\\), as follows:

\\[ T_{win} = \\{ o \\in O | o > 0 \\} \\]

Similarly, we can discover the set of losing trades \\(T_{lose}\\) as follows:

\\[ T_{lose} = \\{o \\in O | o \\leq 0\\} \\]

Example

In a section where a strategy made four transactions \\(O = \\{3.5, -1, 15, 0\\}\\): \\(T_{win} = \\{3.5, 15\\}\\) \\(T_{lose} = \\{-1, 0\\}\\)

"},{"location":"edge/#win-rate-and-lose-rate","title":"Win Rate and Lose Rate","text":"

The win rate \\(W\\) is the proportion of winning trades with respect to all the trades made by a strategy. We use the following function to compute the win rate:

\\[W = \\frac{|T_{win}|}{N}\\]

Where \\(W\\) is the win rate, \\(N\\) is the number of trades and, \\(T_{win}\\) is the set of all trades where the strategy made money.

Similarly, we can compute the rate of losing trades:

\\[ L = \\frac{|T_{lose}|}{N} \\]

Where \\(L\\) is the lose rate, \\(N\\) is the amount of trades made and, \\(T_{lose}\\) is the set of all trades where the strategy lost money. Note that the above formula is the same as calculating \\(L = 1 \u2013 W\\) or \\(W = 1 \u2013 L\\)

"},{"location":"edge/#risk-reward-ratio","title":"Risk Reward Ratio","text":"

Risk Reward Ratio (\\(R\\)) is a formula used to measure the expected gains of a given investment against the risk of loss. It is basically what you potentially win divided by what you potentially lose. Formally:

\\[ R = \\frac{\\text{potential_profit}}{\\text{potential_loss}} \\] Worked example of \\(R\\) calculation

Let's say that you think that the price of stonecoin today is 10.0$. You believe that, because they will start mining stonecoin, it will go up to 15.0$ tomorrow. There is the risk that the stone is too hard, and the GPUs can't mine it, so the price might go to 0$ tomorrow. You are planning to invest 100$, which will give you 10 shares (100 / 10).

Your potential profit is calculated as:

\\(\\begin{aligned} \\text{potential_profit} &= (\\text{potential_price} - \\text{entry_price}) * \\frac{\\text{investment}}{\\text{entry_price}} \\\\ &= (15 - 10) * (100 / 10) \\\\ &= 50 \\end{aligned}\\)

Since the price might go to 0$, the 100$ dollars invested could turn into 0.

We do however use a stoploss of 15% - so in the worst case, we'll sell 15% below entry price (or at 8.5$).

\\(\\begin{aligned} \\text{potential_loss} &= (\\text{entry_price} - \\text{stoploss}) * \\frac{\\text{investment}}{\\text{entry_price}} \\\\ &= (10 - 8.5) * (100 / 10)\\\\ &= 15 \\end{aligned}\\)

We can compute the Risk Reward Ratio as follows:

\\(\\begin{aligned} R &= \\frac{\\text{potential_profit}}{\\text{potential_loss}}\\\\ &= \\frac{50}{15}\\\\ &= 3.33 \\end{aligned}\\) What it effectively means is that the strategy have the potential to make 3.33$ for each 1$ invested.

On a long horizon, that is, on many trades, we can calculate the risk reward by dividing the strategy' average profit on winning trades by the strategy' average loss on losing trades. We can calculate the average profit, \\(\\mu_{win}\\), as follows:

\\[ \\text{average_profit} = \\mu_{win} = \\frac{\\text{sum_of_profits}}{\\text{count_winning_trades}} = \\frac{\\sum^{o \\in T_{win}} o}{|T_{win}|} \\]

Similarly, we can calculate the average loss, \\(\\mu_{lose}\\), as follows:

\\[ \\text{average_loss} = \\mu_{lose} = \\frac{\\text{sum_of_losses}}{\\text{count_losing_trades}} = \\frac{\\sum^{o \\in T_{lose}} o}{|T_{lose}|} \\]

Finally, we can calculate the Risk Reward ratio, \\(R\\), as follows:

\\[ R = \\frac{\\text{average_profit}}{\\text{average_loss}} = \\frac{\\mu_{win}}{\\mu_{lose}}\\\\ \\] Worked example of \\(R\\) calculation using mean profit/loss

Let's say the strategy that we are using makes an average win \\(\\mu_{win} = 2.06\\) and an average loss \\(\\mu_{loss} = 4.11\\). We calculate the risk reward ratio as follows: \\(R = \\frac{\\mu_{win}}{\\mu_{loss}} = \\frac{2.06}{4.11} = 0.5012...\\)

"},{"location":"edge/#expectancy","title":"Expectancy","text":"

By combining the Win Rate \\(W\\) and and the Risk Reward ratio \\(R\\) to create an expectancy ratio \\(E\\). A expectance ratio is the expected return of the investment made in a trade. We can compute the value of \\(E\\) as follows:

\\[E = R * W - L\\]

Calculating \\(E\\)

Let's say that a strategy has a win rate \\(W = 0.28\\) and a risk reward ratio \\(R = 5\\). What this means is that the strategy is expected to make 5 times the investment around on 28% of the trades it makes. Working out the example: \\(E = R * W - L = 5 * 0.28 - 0.72 = 0.68\\)

The expectancy worked out in the example above means that, on average, this strategy' trades will return 1.68 times the size of its losses. Said another way, the strategy makes 1.68$ for every 1$ it loses, on average.

This is important for two reasons: First, it may seem obvious, but you know right away that you have a positive return. Second, you now have a number you can compare to other candidate systems to make decisions about which ones you employ.

It is important to remember that any system with an expectancy greater than 0 is profitable using past data. The key is finding one that will be profitable in the future.

You can also use this value to evaluate the effectiveness of modifications to this system.

Note

It's important to keep in mind that Edge is testing your expectancy using historical data, there's no guarantee that you will have a similar edge in the future. It's still vital to do this testing in order to build confidence in your methodology but be wary of \"curve-fitting\" your approach to the historical data as things are unlikely to play out the exact same way for future trades.

"},{"location":"edge/#how-does-it-work","title":"How does it work?","text":"

Edge combines dynamic stoploss, dynamic positions, and whitelist generation into one isolated module which is then applied to the trading strategy. If enabled in config, Edge will go through historical data with a range of stoplosses in order to find buy and sell/stoploss signals. It then calculates win rate and expectancy over N trades for each stoploss. Here is an example:

Pair Stoploss Win Rate Risk Reward Ratio Expectancy XZC/ETH -0.01 0.50 1.176384 0.088 XZC/ETH -0.02 0.51 1.115941 0.079 XZC/ETH -0.03 0.52 1.359670 0.228 XZC/ETH -0.04 0.51 1.234539 0.117

The goal here is to find the best stoploss for the strategy in order to have the maximum expectancy. In the above example stoploss at \\(3%\\) leads to the maximum expectancy according to historical data.

Edge module then forces stoploss value it evaluated to your strategy dynamically.

"},{"location":"edge/#position-size","title":"Position size","text":"

Edge dictates the amount at stake for each trade to the bot according to the following factors:

  • Allowed capital at risk
  • Stoploss

Allowed capital at risk is calculated as follows:

Allowed capital at risk = (Capital available_percentage) X (Allowed risk per trade)\n

Stoploss is calculated as described above with respect to historical data.

The position size is calculated as follows:

Position size = (Allowed capital at risk) / Stoploss\n

Example:

Let's say the stake currency is ETH and there is \\(10\\) ETH on the wallet. The capital available percentage is \\(50%\\) and the allowed risk per trade is \\(1\\%\\). Thus, the available capital for trading is \\(10 * 0.5 = 5\\) ETH and the allowed capital at risk would be \\(5 * 0.01 = 0.05\\) ETH.

  • Trade 1: The strategy detects a new buy signal in the XLM/ETH market. Edge Positioning calculates a stoploss of \\(2\\%\\) and a position of \\(0.05 / 0.02 = 2.5\\) ETH. The bot takes a position of \\(2.5\\) ETH in the XLM/ETH market.
  • Trade 2: The strategy detects a buy signal on the BTC/ETH market while Trade 1 is still open. Edge Positioning calculates the stoploss of \\(4\\%\\) on this market. Thus, Trade 2 position size is \\(0.05 / 0.04 = 1.25\\) ETH.

Available Capital \\(\\neq\\) Available in wallet

The available capital for trading didn't change in Trade 2 even with Trade 1 still open. The available capital is not the free amount in the wallet.

  • Trade 3: The strategy detects a buy signal in the ADA/ETH market. Edge Positioning calculates a stoploss of \\(1\\%\\) and a position of \\(0.05 / 0.01 = 5\\) ETH. Since Trade 1 has \\(2.5\\) ETH blocked and Trade 2 has \\(1.25\\) ETH blocked, there is only \\(5 - 1.25 - 2.5 = 1.25\\) ETH available. Hence, the position size of Trade 3 is \\(1.25\\) ETH.

Available Capital Updates

The available capital does not change before a position is sold. After a trade is closed the Available Capital goes up if the trade was profitable or goes down if the trade was a loss.

  • The strategy detects a sell signal in the XLM/ETH market. The bot exits Trade 1 for a profit of \\(1\\) ETH. The total capital in the wallet becomes \\(11\\) ETH and the available capital for trading becomes \\(5.5\\) ETH.
  • Trade 4 The strategy detects a new buy signal int the XLM/ETH market. Edge Positioning calculates the stoploss of \\(2\\%\\), and the position size of \\(0.055 / 0.02 = 2.75\\) ETH.
"},{"location":"edge/#edge-command-reference","title":"Edge command reference","text":"
usage: freqtrade edge [-h] [-v] [--logfile FILE] [-V] [-c PATH] [-d PATH]\n                      [--userdir PATH] [-s NAME] [--strategy-path PATH]\n                      [-i TIMEFRAME] [--timerange TIMERANGE]\n                      [--data-format-ohlcv {json,jsongz,hdf5}]\n                      [--max-open-trades INT] [--stake-amount STAKE_AMOUNT]\n                      [--fee FLOAT] [-p PAIRS [PAIRS ...]]\n                      [--stoplosses STOPLOSS_RANGE]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  -i TIMEFRAME, --timeframe TIMEFRAME\n                        Specify timeframe (`1m`, `5m`, `30m`, `1h`, `1d`).\n  --timerange TIMERANGE\n                        Specify what timerange of data to use.\n  --data-format-ohlcv {json,jsongz,hdf5}\n                        Storage format for downloaded candle (OHLCV) data.\n                        (default: `None`).\n  --max-open-trades INT\n                        Override the value of the `max_open_trades`\n                        configuration setting.\n  --stake-amount STAKE_AMOUNT\n                        Override the value of the `stake_amount` configuration\n                        setting.\n  --fee FLOAT           Specify fee ratio. Will be applied twice (on trade\n                        entry and exit).\n  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]\n                        Limit command to these pairs. Pairs are space-\n                        separated.\n  --stoplosses STOPLOSS_RANGE\n                        Defines a range of stoploss values against which edge\n                        will assess the strategy. The format is \"min,max,step\"\n                        (without any space). Example:\n                        `--stoplosses=-0.01,-0.1,-0.001`\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n\nStrategy arguments:\n  -s NAME, --strategy NAME\n                        Specify strategy class name which will be used by the\n                        bot.\n  --strategy-path PATH  Specify additional strategy lookup path.\n
"},{"location":"edge/#configurations","title":"Configurations","text":"

Edge module has following configuration options:

Parameter Description enabled If true, then Edge will run periodically. Defaults to false. Datatype: Boolean process_throttle_secs How often should Edge run in seconds. Defaults to 3600 (once per hour). Datatype: Integer calculate_since_number_of_days Number of days of data against which Edge calculates Win Rate, Risk Reward and Expectancy. Note that it downloads historical data so increasing this number would lead to slowing down the bot. Defaults to 7. Datatype: Integer allowed_risk Ratio of allowed risk per trade. Defaults to 0.01 (1%)). Datatype: Float stoploss_range_min Minimum stoploss. Defaults to -0.01. Datatype: Float stoploss_range_max Maximum stoploss. Defaults to -0.10. Datatype: Float stoploss_range_step As an example if this is set to -0.01 then Edge will test the strategy for [-0.01, -0,02, -0,03 ..., -0.09, -0.10] ranges. Note than having a smaller step means having a bigger range which could lead to slow calculation. If you set this parameter to -0.001, you then slow down the Edge calculation by a factor of 10. Defaults to -0.001. Datatype: Float minimum_winrate It filters out pairs which don't have at least minimum_winrate. This comes handy if you want to be conservative and don't comprise win rate in favour of risk reward ratio. Defaults to 0.60. Datatype: Float minimum_expectancy It filters out pairs which have the expectancy lower than this number. Having an expectancy of 0.20 means if you put 10$ on a trade you expect a 12$ return. Defaults to 0.20. Datatype: Float min_trade_number When calculating W, R and E (expectancy) against historical data, you always want to have a minimum number of trades. The more this number is the more Edge is reliable. Having a win rate of 100% on a single trade doesn't mean anything at all. But having a win rate of 70% over past 100 trades means clearly something. Defaults to 10 (it is highly recommended not to decrease this number). Datatype: Integer max_trade_duration_minute Edge will filter out trades with long duration. If a trade is profitable after 1 month, it is hard to evaluate the strategy based on it. But if most of trades are profitable and they have maximum duration of 30 minutes, then it is clearly a good sign.NOTICE: While configuring this value, you should take into consideration your timeframe. As an example filtering out trades having duration less than one day for a strategy which has 4h interval does not make sense. Default value is set assuming your strategy interval is relatively small (1m or 5m, etc.).Defaults to 1440 (one day). Datatype: Integer remove_pumps Edge will remove sudden pumps in a given market while going through historical data. However, given that pumps happen very often in crypto markets, we recommend you keep this off.Defaults to false. Datatype: Boolean"},{"location":"edge/#running-edge-independently","title":"Running Edge independently","text":"

You can run Edge independently in order to see in details the result. Here is an example:

freqtrade edge\n

An example of its output:

pair stoploss win rate risk reward ratio required risk reward expectancy total number of trades average duration (min) AGI/BTC -0.02 0.64 5.86 0.56 3.41 14 54 NXS/BTC -0.03 0.64 2.99 0.57 1.54 11 26 LEND/BTC -0.02 0.82 2.05 0.22 1.50 11 36 VIA/BTC -0.01 0.55 3.01 0.83 1.19 11 48 MTH/BTC -0.09 0.56 2.82 0.80 1.12 18 52 ARDR/BTC -0.04 0.42 3.14 1.40 0.73 12 42 BCPT/BTC -0.01 0.71 1.34 0.40 0.67 14 30 WINGS/BTC -0.02 0.56 1.97 0.80 0.65 27 42 VIBE/BTC -0.02 0.83 0.91 0.20 0.59 12 35 MCO/BTC -0.02 0.79 0.97 0.27 0.55 14 31 GNT/BTC -0.02 0.50 2.06 1.00 0.53 18 24 HOT/BTC -0.01 0.17 7.72 4.81 0.50 209 7 SNM/BTC -0.03 0.71 1.06 0.42 0.45 17 38 APPC/BTC -0.02 0.44 2.28 1.27 0.44 25 43 NEBL/BTC -0.03 0.63 1.29 0.58 0.44 19 59

Edge produced the above table by comparing calculate_since_number_of_days to minimum_expectancy to find min_trade_number historical information based on the config file. The timerange Edge uses for its comparisons can be further limited by using the --timerange switch.

In live and dry-run modes, after the process_throttle_secs has passed, Edge will again process calculate_since_number_of_days against minimum_expectancy to find min_trade_number. If no min_trade_number is found, the bot will return \"whitelist empty\". Depending on the trade strategy being deployed, \"whitelist empty\" may be return much of the time - or all of the time. The use of Edge may also cause trading to occur in bursts, though this is rare.

If you encounter \"whitelist empty\" a lot, condsider tuning calculate_since_number_of_days, minimum_expectancy and min_trade_number to align to the trading frequency of your strategy.

"},{"location":"edge/#update-cached-pairs-with-the-latest-data","title":"Update cached pairs with the latest data","text":"

Edge requires historic data the same way as backtesting does. Please refer to the Data Downloading section of the documentation for details.

"},{"location":"edge/#precising-stoploss-range","title":"Precising stoploss range","text":"
freqtrade edge --stoplosses=-0.01,-0.1,-0.001 #min,max,step\n
"},{"location":"edge/#advanced-use-of-timerange","title":"Advanced use of timerange","text":"
freqtrade edge --timerange=20181110-20181113\n

Doing --timerange=-20190901 will get all available data until September 1st (excluding September 1st 2019).

The full timerange specification:

  • Use tickframes till 2018/01/31: --timerange=-20180131
  • Use tickframes since 2018/01/31: --timerange=20180131-
  • Use tickframes since 2018/01/31 till 2018/03/01 : --timerange=20180131-20180301
  • Use tickframes between POSIX timestamps 1527595200 1527618600: --timerange=1527595200-1527618600
  1. Question extracted from MIT Opencourseware S096 - Mathematics with applications in Finance: https://ocw.mit.edu/courses/mathematics/18-s096-topics-in-mathematics-with-applications-in-finance-fall-2013/ \u21a9

"},{"location":"exchanges/","title":"Exchange-specific Notes","text":"

This page combines common gotchas and Information which are exchange-specific and most likely don't apply to other exchanges.

"},{"location":"exchanges/#exchange-configuration","title":"Exchange configuration","text":"

Freqtrade is based on CCXT library that supports over 100 cryptocurrency exchange markets and trading APIs. The complete up-to-date list can be found in the CCXT repo homepage. However, the bot was tested by the development team with only a few exchanges. A current list of these can be found in the \"Home\" section of this documentation.

Feel free to test other exchanges and submit your feedback or PR to improve the bot or confirm exchanges that work flawlessly..

Some exchanges require special configuration, which can be found below.

"},{"location":"exchanges/#sample-exchange-configuration","title":"Sample exchange configuration","text":"

A exchange configuration for \"binance\" would look as follows:

\"exchange\": {\n\"name\": \"binance\",\n\"key\": \"your_exchange_key\",\n\"secret\": \"your_exchange_secret\",\n\"ccxt_config\": {},\n\"ccxt_async_config\": {},\n// ... \n
"},{"location":"exchanges/#setting-rate-limits","title":"Setting rate limits","text":"

Usually, rate limits set by CCXT are reliable and work well. In case of problems related to rate-limits (usually DDOS Exceptions in your logs), it's easy to change rateLimit settings to other values.

\"exchange\": {\n\"name\": \"kraken\",\n\"key\": \"your_exchange_key\",\n\"secret\": \"your_exchange_secret\",\n\"ccxt_config\": {\"enableRateLimit\": true},\n\"ccxt_async_config\": {\n\"enableRateLimit\": true,\n\"rateLimit\": 3100\n},\n

This configuration enables kraken, as well as rate-limiting to avoid bans from the exchange. \"rateLimit\": 3100 defines a wait-event of 3.1s between each call. This can also be completely disabled by setting \"enableRateLimit\" to false.

Note

Optimal settings for rate-limiting depend on the exchange and the size of the whitelist, so an ideal parameter will vary on many other settings. We try to provide sensible defaults per exchange where possible, if you encounter bans please make sure that \"enableRateLimit\" is enabled and increase the \"rateLimit\" parameter step by step.

"},{"location":"exchanges/#binance","title":"Binance","text":"

Server location and geo-ip restrictions

Please be aware that binance restrict api access regarding the server country. The currents and non exhaustive countries blocked are United States, Malaysia (Singapour), Ontario (Canada). Please go to binance terms > b. Eligibility to find up to date list.

Binance supports time_in_force.

Stoploss on Exchange

Binance supports stoploss_on_exchange and uses stop-loss-limit orders. It provides great advantages, so we recommend to benefit from it by enabling stoploss on exchange. On futures, Binance supports both stop-limit as well as stop-market orders. You can use either \"limit\" or \"market\" in the order_types.stoploss configuration setting to decide which type to use.

"},{"location":"exchanges/#binance-blacklist-recommendation","title":"Binance Blacklist recommendation","text":"

For Binance, it is suggested to add \"BNB/<STAKE>\" to your blacklist to avoid issues, unless you are willing to maintain enough extra BNB on the account or unless you're willing to disable using BNB for fees. Binance accounts may use BNB for fees, and if a trade happens to be on BNB, further trades may consume this position and make the initial BNB trade unsellable as the expected amount is not there anymore.

"},{"location":"exchanges/#binance-sites","title":"Binance sites","text":"

Binance has been split into 2, and users must use the correct ccxt exchange ID for their exchange, otherwise API keys are not recognized.

  • binance.com - International users. Use exchange id: binance.
  • binance.us - US based users. Use exchange id: binanceus.
"},{"location":"exchanges/#binance-rsa-keys","title":"Binance RSA keys","text":"

Freqtrade supports binance RSA API keys.

We recommend to use them as environment variable.

export FREQTRADE__EXCHANGE__SECRET=\"$(cat ./rsa_binance.private)\"\n

They can however also be configured via configuration file. Since json doesn't support multi-line strings, you'll have to replace all newlines with \\n to have a valid json file.

// ...\n\"key\": \"<someapikey>\",\n\"secret\": \"-----BEGIN PRIVATE KEY-----\\nMIIEvQIBABACAFQA<...>s8KX8=\\n-----END PRIVATE KEY-----\"\n// ...\n
"},{"location":"exchanges/#binance-futures","title":"Binance Futures","text":"

Binance has specific (unfortunately complex) Futures Trading Quantitative Rules which need to be followed, and which prohibit a too low stake-amount (among others) for too many orders. Violating these rules will result in a trading restriction.

When trading on Binance Futures market, orderbook must be used because there is no price ticker data for futures.

  \"entry_pricing\": {\n      \"use_order_book\": true,\n      \"order_book_top\": 1,\n      \"check_depth_of_market\": {\n          \"enabled\": false,\n          \"bids_to_ask_delta\": 1\n      }\n  },\n  \"exit_pricing\": {\n      \"use_order_book\": true,\n      \"order_book_top\": 1\n  },\n
"},{"location":"exchanges/#binance-futures-settings","title":"Binance futures settings","text":"

Users will also have to have the futures-setting \"Position Mode\" set to \"One-way Mode\", and \"Asset Mode\" set to \"Single-Asset Mode\". These settings will be checked on startup, and freqtrade will show an error if this setting is wrong.

Freqtrade will not attempt to change these settings.

"},{"location":"exchanges/#kraken","title":"Kraken","text":"

Stoploss on Exchange

Kraken supports stoploss_on_exchange and can use both stop-loss-market and stop-loss-limit orders. It provides great advantages, so we recommend to benefit from it. You can use either \"limit\" or \"market\" in the order_types.stoploss configuration setting to decide which type to use.

"},{"location":"exchanges/#historic-kraken-data","title":"Historic Kraken data","text":"

The Kraken API does only provide 720 historic candles, which is sufficient for Freqtrade dry-run and live trade modes, but is a problem for backtesting. To download data for the Kraken exchange, using --dl-trades is mandatory, otherwise the bot will download the same 720 candles over and over, and you'll not have enough backtest data.

Due to the heavy rate-limiting applied by Kraken, the following configuration section should be used to download data:

    \"ccxt_async_config\": {\n\"enableRateLimit\": true,\n\"rateLimit\": 3100\n},\n

Downloading data from kraken

Downloading kraken data will require significantly more memory (RAM) than any other exchange, as the trades-data needs to be converted into candles on your machine. It will also take a long time, as freqtrade will need to download every single trade that happened on the exchange for the pair / timerange combination, therefore please be patient.

rateLimit tuning

Please pay attention that rateLimit configuration entry holds delay in milliseconds between requests, NOT requests\\sec rate. So, in order to mitigate Kraken API \"Rate limit exceeded\" exception, this configuration should be increased, NOT decreased.

"},{"location":"exchanges/#bittrex","title":"Bittrex","text":""},{"location":"exchanges/#order-types","title":"Order types","text":"

Bittrex does not support market orders. If you have a message at the bot startup about this, you should change order type values set in your configuration and/or in the strategy from \"market\" to \"limit\". See some more details on this here in the FAQ.

Bittrex also does not support VolumePairlist due to limited / split API constellation at the moment. Please use StaticPairlist. Other pairlists (other than VolumePairlist) should not be affected.

"},{"location":"exchanges/#volume-pairlist","title":"Volume pairlist","text":"

Bittrex does not support the direct usage of VolumePairList. This can however be worked around by using the advanced mode with lookback_days: 1 (or more), which will emulate 24h volume.

Read more in the pairlist documentation.

"},{"location":"exchanges/#restricted-markets","title":"Restricted markets","text":"

Bittrex split its exchange into US and International versions. The International version has more pairs available, however the API always returns all pairs, so there is currently no automated way to detect if you're affected by the restriction.

If you have restricted pairs in your whitelist, you'll get a warning message in the log on Freqtrade startup for each restricted pair.

The warning message will look similar to the following:

[...] Message: bittrex {\"success\":false,\"message\":\"RESTRICTED_MARKET\",\"result\":null,\"explanation\":null}\"\n

If you're an \"International\" customer on the Bittrex exchange, then this warning will probably not impact you. If you're a US customer, the bot will fail to create orders for these pairs, and you should remove them from your whitelist.

You can get a list of restricted markets by using the following snippet:

import ccxt\nct = ccxt.bittrex()\nlm = ct.load_markets()\n\nres = [p for p, x in lm.items() if 'US' in x['info']['prohibitedIn']]\nprint(res)\n
"},{"location":"exchanges/#kucoin","title":"Kucoin","text":"

Kucoin requires a passphrase for each api key, you will therefore need to add this key into the configuration so your exchange section looks as follows:

\"exchange\": {\n\"name\": \"kucoin\",\n\"key\": \"your_exchange_key\",\n\"secret\": \"your_exchange_secret\",\n\"password\": \"your_exchange_api_key_password\",\n// ...\n}\n

Kucoin supports time_in_force.

Stoploss on Exchange

Kucoin supports stoploss_on_exchange and can use both stop-loss-market and stop-loss-limit orders. It provides great advantages, so we recommend to benefit from it. You can use either \"limit\" or \"market\" in the order_types.stoploss configuration setting to decide which type of stoploss shall be used.

"},{"location":"exchanges/#kucoin-blacklists","title":"Kucoin Blacklists","text":"

For Kucoin, it is suggested to add \"KCS/<STAKE>\" to your blacklist to avoid issues, unless you are willing to maintain enough extra KCS on the account or unless you're willing to disable using KCS for fees. Kucoin accounts may use KCS for fees, and if a trade happens to be on KCS, further trades may consume this position and make the initial KCS trade unsellable as the expected amount is not there anymore.

"},{"location":"exchanges/#huobi","title":"Huobi","text":"

Stoploss on Exchange

Huobi supports stoploss_on_exchange and uses stop-limit orders. It provides great advantages, so we recommend to benefit from it by enabling stoploss on exchange.

"},{"location":"exchanges/#okx-former-okex","title":"OKX (former OKEX)","text":"

OKX requires a passphrase for each api key, you will therefore need to add this key into the configuration so your exchange section looks as follows:

\"exchange\": {\n\"name\": \"okx\",\n\"key\": \"your_exchange_key\",\n\"secret\": \"your_exchange_secret\",\n\"password\": \"your_exchange_api_key_password\",\n// ...\n}\n

Warning

OKX only provides 100 candles per api call. Therefore, the strategy will only have a pretty low amount of data available in backtesting mode.

Futures

OKX Futures has the concept of \"position mode\" - which can be \"Buy/Sell\" or long/short (hedge mode). Freqtrade supports both modes (we recommend to use Buy/Sell mode) - but changing the mode mid-trading is not supported and will lead to exceptions and failures to place trades. OKX also only provides MARK candles for the past ~3 months. Backtesting futures prior to that date will therefore lead to slight deviations, as funding-fees cannot be calculated correctly without this data.

"},{"location":"exchanges/#gateio","title":"Gate.io","text":"

Stoploss on Exchange

Gate.io supports stoploss_on_exchange and uses stop-loss-limit orders. It provides great advantages, so we recommend to benefit from it by enabling stoploss on exchange..

Gate.io allows the use of POINT to pay for fees. As this is not a tradable currency (no regular market available), automatic fee calculations will fail (and default to a fee of 0). The configuration parameter exchange.unknown_fee_rate can be used to specify the exchange rate between Point and the stake currency. Obviously, changing the stake-currency will also require changes to this value.

"},{"location":"exchanges/#bybit","title":"Bybit","text":"

Futures trading on bybit is currently supported for USDT markets, and will use isolated futures mode. Users with unified accounts (there's no way back) can create a Sub-account which will start as \"non-unified\", and can therefore use isolated futures. On startup, freqtrade will set the position mode to \"One-way Mode\" for the whole (sub)account. This avoids making this call over and over again (slowing down bot operations), but means that changes to this setting may result in exceptions and errors.

As bybit doesn't provide funding rate history, the dry-run calculation is used for live trades as well.

Stoploss on Exchange

Bybit (futures only) supports stoploss_on_exchange and uses stop-loss-limit orders. It provides great advantages, so we recommend to benefit from it by enabling stoploss on exchange. On futures, Bybit supports both stop-limit as well as stop-market orders. You can use either \"limit\" or \"market\" in the order_types.stoploss configuration setting to decide which type to use.

"},{"location":"exchanges/#all-exchanges","title":"All exchanges","text":"

Should you experience constant errors with Nonce (like InvalidNonce), it is best to regenerate the API keys. Resetting Nonce is difficult and it's usually easier to regenerate the API keys.

"},{"location":"exchanges/#random-notes-for-other-exchanges","title":"Random notes for other exchanges","text":"
  • The Ocean (exchange id: theocean) exchange uses Web3 functionality and requires web3 python package to be installed:
$ pip3 install web3\n
"},{"location":"exchanges/#getting-latest-price-incomplete-candles","title":"Getting latest price / Incomplete candles","text":"

Most exchanges return current incomplete candle via their OHLCV/klines API interface. By default, Freqtrade assumes that incomplete candle is fetched from the exchange and removes the last candle assuming it's the incomplete candle.

Whether your exchange returns incomplete candles or not can be checked using the helper script from the Contributor documentation.

Due to the danger of repainting, Freqtrade does not allow you to use this incomplete candle.

However, if it is based on the need for the latest price for your strategy - then this requirement can be acquired using the data provider from within the strategy.

"},{"location":"exchanges/#advanced-freqtrade-exchange-configuration","title":"Advanced Freqtrade Exchange configuration","text":"

Advanced options can be configured using the _ft_has_params setting, which will override Defaults and exchange-specific behavior.

Available options are listed in the exchange-class as _ft_has_default.

For example, to test the order type FOK with Kraken, and modify candle limit to 200 (so you only get 200 candles per API call):

\"exchange\": {\n\"name\": \"kraken\",\n\"_ft_has_params\": {\n\"order_time_in_force\": [\"GTC\", \"FOK\"],\n\"ohlcv_candle_limit\": 200\n}\n//...\n}\n

Warning

Please make sure to fully understand the impacts of these settings before modifying them.

"},{"location":"faq/","title":"Freqtrade FAQ","text":""},{"location":"faq/#supported-markets","title":"Supported Markets","text":"

Freqtrade supports spot trading, as well as (isolated) futures trading for some selected exchanges. Please refer to the documentation start page for an uptodate list of supported exchanges.

"},{"location":"faq/#can-my-bot-open-short-positions","title":"Can my bot open short positions?","text":"

Freqtrade can open short positions in futures markets. This requires the strategy to be made for this - and \"trading_mode\": \"futures\" in the configuration. Please make sure to read the relevant documentation page first.

In spot markets, you can in some cases use leveraged spot tokens, which reflect an inverted pair (eg. BTCUP/USD, BTCDOWN/USD, ETHBULL/USD, ETHBEAR/USD,...) which can be traded with Freqtrade.

"},{"location":"faq/#can-my-bot-trade-options-or-futures","title":"Can my bot trade options or futures?","text":"

Futures trading is supported for selected exchanges. Please refer to the documentation start page for an uptodate list of supported exchanges.

"},{"location":"faq/#beginner-tips-tricks","title":"Beginner Tips & Tricks","text":"
  • When you work with your strategy & hyperopt file you should use a proper code editor like VSCode or PyCharm. A good code editor will provide syntax highlighting as well as line numbers, making it easy to find syntax errors (most likely pointed out by Freqtrade during startup).
"},{"location":"faq/#freqtrade-common-issues","title":"Freqtrade common issues","text":""},{"location":"faq/#can-freqtrade-open-multiple-positions-on-the-same-pair-in-parallel","title":"Can freqtrade open multiple positions on the same pair in parallel?","text":"

No. Freqtrade will only open one position per pair at a time. You can however use the adjust_trade_position() callback to adjust an open position.

Backtesting provides an option for this in --eps - however this is only there to highlight \"hidden\" signals, and will not work in live.

"},{"location":"faq/#the-bot-does-not-start","title":"The bot does not start","text":"

Running the bot with freqtrade trade --config config.json shows the output freqtrade: command not found.

This could be caused by the following reasons:

  • The virtual environment is not active.
    • Run source .env/bin/activate to activate the virtual environment.
  • The installation did not complete successfully.
    • Please check the Installation documentation.
"},{"location":"faq/#i-have-waited-5-minutes-why-hasnt-the-bot-made-any-trades-yet","title":"I have waited 5 minutes, why hasn't the bot made any trades yet?","text":"
  • Depending on the buy strategy, the amount of whitelisted coins, the situation of the market etc, it can take up to hours to find a good entry position for a trade. Be patient!
  • It may be because of a configuration error. It's best to check the logs, they usually tell you if the bot is simply not getting buy signals (only heartbeat messages), or if there is something wrong (errors / exceptions in the log).
"},{"location":"faq/#i-have-made-12-trades-already-why-is-my-total-profit-negative","title":"I have made 12 trades already, why is my total profit negative?","text":"

I understand your disappointment but unfortunately 12 trades is just not enough to say anything. If you run backtesting, you can see that the current algorithm does leave you on the plus side, but that is after thousands of trades and even there, you will be left with losses on specific coins that you have traded tens if not hundreds of times. We of course constantly aim to improve the bot but it will always be a gamble, which should leave you with modest wins on monthly basis but you can't say much from few trades.

"},{"location":"faq/#id-like-to-make-changes-to-the-config-can-i-do-that-without-having-to-kill-the-bot","title":"I\u2019d like to make changes to the config. Can I do that without having to kill the bot?","text":"

Yes. You can edit your config and use the /reload_config command to reload the configuration. The bot will stop, reload the configuration and strategy and will restart with the new configuration and strategy.

"},{"location":"faq/#why-does-my-bot-not-sell-everything-it-bought","title":"Why does my bot not sell everything it bought?","text":"

This is called \"coin dust\" and can happen on all exchanges. It happens because many exchanges subtract fees from the \"receiving currency\" - so you buy 100 COIN - but you only get 99.9 COIN. As COIN is trading in full lot sizes (1COIN steps), you cannot sell 0.9 COIN (or 99.9 COIN) - but you need to round down to 99 COIN.

This is not a bot-problem, but will also happen while manual trading.

While freqtrade can handle this (it'll sell 99 COIN), fees are often below the minimum tradable lot-size (you can only trade full COIN, not 0.9 COIN). Leaving the dust (0.9 COIN) on the exchange makes usually sense, as the next time freqtrade buys COIN, it'll eat into the remaining small balance, this time selling everything it bought, and therefore slowly declining the dust balance (although it most likely will never reach exactly 0).

Where possible (e.g. on binance), the use of the exchange's dedicated fee currency will fix this. On binance, it's sufficient to have BNB in your account, and have \"Pay fees in BNB\" enabled in your profile. Your BNB balance will slowly decline (as it's used to pay fees) - but you'll no longer encounter dust (Freqtrade will include the fees in the profit calculations). Other exchanges don't offer such possibilities, where it's simply something you'll have to accept or move to a different exchange.

"},{"location":"faq/#i-want-to-use-incomplete-candles","title":"I want to use incomplete candles","text":"

Freqtrade will not provide incomplete candles to strategies. Using incomplete candles will lead to repainting and consequently to strategies with \"ghost\" buys, which are impossible to both backtest, and verify after they happened.

You can use \"current\" market data by using the dataprovider's orderbook or ticker methods - which however cannot be used during backtesting.

"},{"location":"faq/#is-there-a-setting-to-only-exit-the-trades-being-held-and-not-perform-any-new-entries","title":"Is there a setting to only Exit the trades being held and not perform any new Entries?","text":"

You can use the /stopentry command in Telegram to prevent future trade entry, followed by /forceexit all (sell all open trades).

"},{"location":"faq/#i-want-to-run-multiple-bots-on-the-same-machine","title":"I want to run multiple bots on the same machine","text":"

Please look at the advanced setup documentation Page.

"},{"location":"faq/#im-getting-missing-data-fillup-messages-in-the-log","title":"I'm getting \"Missing data fillup\" messages in the log","text":"

This message is just a warning that the latest candles had missing candles in them. Depending on the exchange, this can indicate that the pair didn't have a trade for the timeframe you are using - and the exchange does only return candles with volume. On low volume pairs, this is a rather common occurrence.

If this happens for all pairs in the pairlist, this might indicate a recent exchange downtime. Please check your exchange's public channels for details.

Irrespectively of the reason, Freqtrade will fill up these candles with \"empty\" candles, where open, high, low and close are set to the previous candle close - and volume is empty. In a chart, this will look like a _ - and is aligned with how exchanges usually represent 0 volume candles.

"},{"location":"faq/#im-getting-price-jump-between-2-candles-detected","title":"I'm getting \"Price jump between 2 candles detected\"","text":"

This message is a warning that the candles had a price jump of > 30%. This might be a sign that the pair stopped trading, and some token exchange took place (e.g. COCOS in 2021 - where price jumped from 0.0000154 to 0.01621). This message is often accompanied by \"Missing data fillup\" - as trading on such pairs is often stopped for some time.

"},{"location":"faq/#im-getting-outdated-history-for-pair-xxx-in-the-log","title":"I'm getting \"Outdated history for pair xxx\" in the log","text":"

The bot is trying to tell you that it got an outdated last candle (not the last complete candle). As a consequence, Freqtrade will not enter a trade for this pair - as trading on old information is usually not what is desired.

This warning can point to one of the below problems:

  • Exchange downtime -> Check your exchange status page / blog / twitter feed for details.
  • Wrong system time -> Ensure your system-time is correct.
  • Barely traded pair -> Check the pair on the exchange webpage, look at the timeframe your strategy uses. If the pair does not have any volume in some candles (usually visualized with a \"volume 0\" bar, and a \"_\" as candle), this pair did not have any trades in this timeframe. These pairs should ideally be avoided, as they can cause problems with order-filling.
  • API problem -> API returns wrong data (this only here for completeness, and should not happen with supported exchanges).
"},{"location":"faq/#im-getting-the-restricted_market-message-in-the-log","title":"I'm getting the \"RESTRICTED_MARKET\" message in the log","text":"

Currently known to happen for US Bittrex users.

Read the Bittrex section about restricted markets for more information.

"},{"location":"faq/#im-getting-the-exchange-xxx-does-not-support-market-orders-message-and-cannot-run-my-strategy","title":"I'm getting the \"Exchange XXX does not support market orders.\" message and cannot run my strategy","text":"

As the message says, your exchange does not support market orders and you have one of the order types set to \"market\". Your strategy was probably written with other exchanges in mind and sets \"market\" orders for \"stoploss\" orders, which is correct and preferable for most of the exchanges supporting market orders (but not for Bittrex and Gate.io).

To fix this, redefine order types in the strategy to use \"limit\" instead of \"market\":

    order_types = {\n        ...\n        \"stoploss\": \"limit\",\n        ...\n    }\n

The same fix should be applied in the configuration file, if order types are defined in your custom config rather than in the strategy.

"},{"location":"faq/#im-trying-to-start-the-bot-live-but-get-an-api-permission-error","title":"I'm trying to start the bot live, but get an API permission error","text":"

Errors like Invalid API-key, IP, or permissions for action mean exactly what they actually say. Your API key is either invalid (copy/paste error? check for leading/trailing spaces in the config), expired, or the IP you're running the bot from is not enabled in the Exchange's API console. Usually, the permission \"Spot Trading\" (or the equivalent in the exchange you use) will be necessary. Futures will usually have to be enabled specifically.

"},{"location":"faq/#how-do-i-search-the-bot-logs-for-something","title":"How do I search the bot logs for something?","text":"

By default, the bot writes its log into stderr stream. This is implemented this way so that you can easily separate the bot's diagnostics messages from Backtesting, Edge and Hyperopt results, output from other various Freqtrade utility sub-commands, as well as from the output of your custom print()'s you may have inserted into your strategy. So if you need to search the log messages with the grep utility, you need to redirect stderr to stdout and disregard stdout.

  • In unix shells, this normally can be done as simple as:
    $ freqtrade --some-options 2>&1 >/dev/null | grep 'something'\n
    (note, 2>&1 and >/dev/null should be written in this order)
  • Bash interpreter also supports so called process substitution syntax, you can grep the log for a string with it as:
    $ freqtrade --some-options 2> >(grep 'something') >/dev/null\n
    or
    $ freqtrade --some-options 2> >(grep -v 'something' 1>&2)\n
  • You can also write the copy of Freqtrade log messages to a file with the --logfile option:
    $ freqtrade --logfile /path/to/mylogfile.log --some-options\n
    and then grep it as:
    $ cat /path/to/mylogfile.log | grep 'something'\n
    or even on the fly, as the bot works and the log file grows:
    $ tail -f /path/to/mylogfile.log | grep 'something'\n
    from a separate terminal window.

On Windows, the --logfile option is also supported by Freqtrade and you can use the findstr command to search the log for the string of interest:

> type \\path\\to\\mylogfile.log | findstr \"something\"\n

"},{"location":"faq/#hyperopt-module","title":"Hyperopt module","text":""},{"location":"faq/#why-does-freqtrade-not-have-gpu-support","title":"Why does freqtrade not have GPU support?","text":"

First of all, most indicator libraries don't have GPU support - as such, there would be little benefit for indicator calculations. The GPU improvements would only apply to pandas-native calculations - or ones written by yourself.

For hyperopt, freqtrade is using scikit-optimize, which is built on top of scikit-learn. Their statement about GPU support is pretty clear.

GPU's also are only good at crunching numbers (floating point operations). For hyperopt, we need both number-crunching (find next parameters) and running python code (running backtesting). As such, GPU's are not too well suited for most parts of hyperopt.

The benefit of using GPU would therefore be pretty slim - and will not justify the complexity introduced by trying to add GPU support.

There is however nothing preventing you from using GPU-enabled indicators within your strategy if you think you must have this - you will however probably be disappointed by the slim gain that will give you (compared to the complexity).

"},{"location":"faq/#how-many-epochs-do-i-need-to-get-a-good-hyperopt-result","title":"How many epochs do I need to get a good Hyperopt result?","text":"

Per default Hyperopt called without the -e/--epochs command line option will only run 100 epochs, means 100 evaluations of your triggers, guards, ... Too few to find a great result (unless if you are very lucky), so you probably have to run it for 10000 or more. But it will take an eternity to compute.

Since hyperopt uses Bayesian search, running for too many epochs may not produce greater results.

It's therefore recommended to run between 500-1000 epochs over and over until you hit at least 10000 epochs in total (or are satisfied with the result). You can best judge by looking at the results - if the bot keeps discovering better strategies, it's best to keep on going.

freqtrade hyperopt --hyperopt-loss SharpeHyperOptLossDaily --strategy SampleStrategy -e 1000\n
"},{"location":"faq/#why-does-it-take-a-long-time-to-run-hyperopt","title":"Why does it take a long time to run hyperopt?","text":"
  • Discovering a great strategy with Hyperopt takes time. Study www.freqtrade.io, the Freqtrade Documentation page, join the Freqtrade discord community. While you patiently wait for the most advanced, free crypto bot in the world, to hand you a possible golden strategy specially designed just for you.
  • If you wonder why it can take from 20 minutes to days to do 1000 epochs here are some answers:

This answer was written during the release 0.15.1, when we had:

  • 8 triggers
  • 9 guards: let's say we evaluate even 10 values from each
  • 1 stoploss calculation: let's say we want 10 values from that too to be evaluated

The following calculation is still very rough and not very precise but it will give the idea. With only these triggers and guards there is already 8*10^9*10 evaluations. A roughly total of 80 billion evaluations. Did you run 100 000 evaluations? Congrats, you've done roughly 1 / 100 000 th of the search space, assuming that the bot never tests the same parameters more than once.

  • The time it takes to run 1000 hyperopt epochs depends on things like: The available cpu, hard-disk, ram, timeframe, timerange, indicator settings, indicator count, amount of coins that hyperopt test strategies on and the resulting trade count - which can be 650 trades in a year or 100000 trades depending if the strategy aims for big profits by trading rarely or for many low profit trades.

Example: 4% profit 650 times vs 0,3% profit a trade 10000 times in a year. If we assume you set the --timerange to 365 days.

Example: freqtrade --config config.json --strategy SampleStrategy --hyperopt SampleHyperopt -e 1000 --timerange 20190601-20200601

"},{"location":"faq/#edge-module","title":"Edge module","text":""},{"location":"faq/#edge-implements-interesting-approach-for-controlling-position-size-is-there-any-theory-behind-it","title":"Edge implements interesting approach for controlling position size, is there any theory behind it?","text":"

The Edge module is mostly a result of brainstorming of @mishaker and @creslinux freqtrade team members.

You can find further info on expectancy, win rate, risk management and position size in the following sources:

  • https://www.tradeciety.com/ultimate-math-guide-for-traders/
  • https://samuraitradingacademy.com/trading-expectancy/
  • https://www.learningmarkets.com/determining-expectancy-in-your-trading/
  • https://www.lonestocktrader.com/make-money-trading-positive-expectancy/
  • https://www.babypips.com/trading/trade-expectancy-matter
"},{"location":"faq/#official-channels","title":"Official channels","text":"

Freqtrade is using exclusively the following official channels:

  • Freqtrade discord server
  • Freqtrade documentation (https://freqtrade.io)
  • Freqtrade github organization

Nobody affiliated with the freqtrade project will ask you about your exchange keys or anything else exposing your funds to exploitation. Should you be asked to expose your exchange keys or send funds to some random wallet, then please don't follow these instructions.

Failing to follow these guidelines will not be responsibility of freqtrade.

"},{"location":"faq/#freqtrade-token","title":"\"Freqtrade token\"","text":"

Freqtrade does not have a Crypto token offering.

Token offerings you find on the internet referring Freqtrade, FreqAI or freqUI must be considered to be a scam, trying to exploit freqtrade's popularity for their own, nefarious gains.

"},{"location":"freqai-configuration/","title":"Configuration","text":"

FreqAI is configured through the typical Freqtrade config file and the standard Freqtrade strategy. Examples of FreqAI config and strategy files can be found in config_examples/config_freqai.example.json and freqtrade/templates/FreqaiExampleStrategy.py, respectively.

"},{"location":"freqai-configuration/#setting-up-the-configuration-file","title":"Setting up the configuration file","text":"

Although there are plenty of additional parameters to choose from, as highlighted in the parameter table, a FreqAI config must at minimum include the following parameters (the parameter values are only examples):

    \"freqai\": {\n\"enabled\": true,\n\"purge_old_models\": 2,\n\"train_period_days\": 30,\n\"backtest_period_days\": 7,\n\"identifier\" : \"unique-id\",\n\"feature_parameters\" : {\n\"include_timeframes\": [\"5m\",\"15m\",\"4h\"],\n\"include_corr_pairlist\": [\n\"ETH/USD\",\n\"LINK/USD\",\n\"BNB/USD\"\n],\n\"label_period_candles\": 24,\n\"include_shifted_candles\": 2,\n\"indicator_periods_candles\": [10, 20]\n},\n\"data_split_parameters\" : {\n\"test_size\": 0.25\n}\n}\n

A full example config is available in config_examples/config_freqai.example.json.

"},{"location":"freqai-configuration/#building-a-freqai-strategy","title":"Building a FreqAI strategy","text":"

The FreqAI strategy requires including the following lines of code in the standard Freqtrade strategy:

    # user should define the maximum startup candle count (the largest number of candles\n    # passed to any single indicator)\n    startup_candle_count: int = 20\n\n    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n\n        # the model will return all labels created by user in `set_freqai_labels()`\n        # (& appended targets), an indication of whether or not the prediction should be accepted,\n        # the target mean/std values for each of the labels created by user in\n        # `feature_engineering_*` for each training period.\n\n        dataframe = self.freqai.start(dataframe, metadata, self)\n\n        return dataframe\n\n    def feature_engineering_expand_all(self, dataframe: DataFrame, period, **kwargs) -> DataFrame:\n\"\"\"\n        *Only functional with FreqAI enabled strategies*\n        This function will automatically expand the defined features on the config defined\n        `indicator_periods_candles`, `include_timeframes`, `include_shifted_candles`, and\n        `include_corr_pairs`. In other words, a single feature defined in this function\n        will automatically expand to a total of\n        `indicator_periods_candles` * `include_timeframes` * `include_shifted_candles` *\n        `include_corr_pairs` numbers of features added to the model.\n\n        All features must be prepended with `%` to be recognized by FreqAI internals.\n\n        :param df: strategy dataframe which will receive the features\n        :param period: period of the indicator - usage example:\n        dataframe[\"%-ema-period\"] = ta.EMA(dataframe, timeperiod=period)\n        \"\"\"\n\n        dataframe[\"%-rsi-period\"] = ta.RSI(dataframe, timeperiod=period)\n        dataframe[\"%-mfi-period\"] = ta.MFI(dataframe, timeperiod=period)\n        dataframe[\"%-adx-period\"] = ta.ADX(dataframe, timeperiod=period)\n        dataframe[\"%-sma-period\"] = ta.SMA(dataframe, timeperiod=period)\n        dataframe[\"%-ema-period\"] = ta.EMA(dataframe, timeperiod=period)\n\n        return dataframe\n\n    def feature_engineering_expand_basic(self, dataframe: DataFrame, **kwargs) -> DataFrame:\n\"\"\"\n        *Only functional with FreqAI enabled strategies*\n        This function will automatically expand the defined features on the config defined\n        `include_timeframes`, `include_shifted_candles`, and `include_corr_pairs`.\n        In other words, a single feature defined in this function\n        will automatically expand to a total of\n        `include_timeframes` * `include_shifted_candles` * `include_corr_pairs`\n        numbers of features added to the model.\n\n        Features defined here will *not* be automatically duplicated on user defined\n        `indicator_periods_candles`\n\n        All features must be prepended with `%` to be recognized by FreqAI internals.\n\n        :param df: strategy dataframe which will receive the features\n        dataframe[\"%-pct-change\"] = dataframe[\"close\"].pct_change()\n        dataframe[\"%-ema-200\"] = ta.EMA(dataframe, timeperiod=200)\n        \"\"\"\n        dataframe[\"%-pct-change\"] = dataframe[\"close\"].pct_change()\n        dataframe[\"%-raw_volume\"] = dataframe[\"volume\"]\n        dataframe[\"%-raw_price\"] = dataframe[\"close\"]\n        return dataframe\n\n    def feature_engineering_standard(self, dataframe: DataFrame, **kwargs) -> DataFrame:\n\"\"\"\n        *Only functional with FreqAI enabled strategies*\n        This optional function will be called once with the dataframe of the base timeframe.\n        This is the final function to be called, which means that the dataframe entering this\n        function will contain all the features and columns created by all other\n        freqai_feature_engineering_* functions.\n\n        This function is a good place to do custom exotic feature extractions (e.g. tsfresh).\n        This function is a good place for any feature that should not be auto-expanded upon\n        (e.g. day of the week).\n\n        All features must be prepended with `%` to be recognized by FreqAI internals.\n\n        :param df: strategy dataframe which will receive the features\n        usage example: dataframe[\"%-day_of_week\"] = (dataframe[\"date\"].dt.dayofweek + 1) / 7\n        \"\"\"\n        dataframe[\"%-day_of_week\"] = (dataframe[\"date\"].dt.dayofweek + 1) / 7\n        dataframe[\"%-hour_of_day\"] = (dataframe[\"date\"].dt.hour + 1) / 25\n        return dataframe\n\n    def set_freqai_targets(self, dataframe: DataFrame, **kwargs) -> DataFrame:\n\"\"\"\n        *Only functional with FreqAI enabled strategies*\n        Required function to set the targets for the model.\n        All targets must be prepended with `&` to be recognized by the FreqAI internals.\n\n        :param df: strategy dataframe which will receive the targets\n        usage example: dataframe[\"&-target\"] = dataframe[\"close\"].shift(-1) / dataframe[\"close\"]\n        \"\"\"\n        dataframe[\"&-s_close\"] = (\n            dataframe[\"close\"]\n            .shift(-self.freqai_info[\"feature_parameters\"][\"label_period_candles\"])\n            .rolling(self.freqai_info[\"feature_parameters\"][\"label_period_candles\"])\n            .mean()\n            / dataframe[\"close\"]\n            - 1\n            )\n        return dataframe\n

Notice how the feature_engineering_*() is where features are added. Meanwhile set_freqai_targets() adds the labels/targets. A full example strategy is available in templates/FreqaiExampleStrategy.py.

Note

The self.freqai.start() function cannot be called outside the populate_indicators().

Note

Features must be defined in feature_engineering_*(). Defining FreqAI features in populate_indicators() will cause the algorithm to fail in live/dry mode. In order to add generalized features that are not associated with a specific pair or timeframe, you should use feature_engineering_standard() (as exemplified in freqtrade/templates/FreqaiExampleStrategy.py).

"},{"location":"freqai-configuration/#important-dataframe-key-patterns","title":"Important dataframe key patterns","text":"

Below are the values you can expect to include/use inside a typical strategy dataframe (df[]):

DataFrame Key Description df['&*'] Any dataframe column prepended with & in set_freqai_targets() is treated as a training target (label) inside FreqAI (typically following the naming convention &-s*). For example, to predict the close price 40 candles into the future, you would set df['&-s_close'] = df['close'].shift(-self.freqai_info[\"feature_parameters\"][\"label_period_candles\"]) with \"label_period_candles\": 40 in the config. FreqAI makes the predictions and gives them back under the same key (df['&-s_close']) to be used in populate_entry/exit_trend(). Datatype: Depends on the output of the model. df['&*_std/mean'] Standard deviation and mean values of the defined labels during training (or live tracking with fit_live_predictions_candles). Commonly used to understand the rarity of a prediction (use the z-score as shown in templates/FreqaiExampleStrategy.py and explained here to evaluate how often a particular prediction was observed during training or historically with fit_live_predictions_candles). Datatype: Float. df['do_predict'] Indication of an outlier data point. The return value is integer between -2 and 2, which lets you know if the prediction is trustworthy or not. do_predict==1 means that the prediction is trustworthy. If the Dissimilarity Index (DI, see details here) of the input data point is above the threshold defined in the config, FreqAI will subtract 1 from do_predict, resulting in do_predict==0. If use_SVM_to_remove_outliers() is active, the Support Vector Machine (SVM, see details here) may also detect outliers in training and prediction data. In this case, the SVM will also subtract 1 from do_predict. If the input data point was considered an outlier by the SVM but not by the DI, or vice versa, the result will be do_predict==0. If both the DI and the SVM considers the input data point to be an outlier, the result will be do_predict==-1. As with the SVM, if use_DBSCAN_to_remove_outliers is active, DBSCAN (see details here) may also detect outliers and subtract 1 from do_predict. Hence, if both the SVM and DBSCAN are active and identify a datapoint that was above the DI threshold as an outlier, the result will be do_predict==-2. A particular case is when do_predict == 2, which means that the model has expired due to exceeding expired_hours. Datatype: Integer between -2 and 2. df['DI_values'] Dissimilarity Index (DI) values are proxies for the level of confidence FreqAI has in the prediction. A lower DI means the prediction is close to the training data, i.e., higher prediction confidence. See details about the DI here. Datatype: Float. df['%*'] Any dataframe column prepended with % in feature_engineering_*() is treated as a training feature. For example, you can include the RSI in the training feature set (similar to in templates/FreqaiExampleStrategy.py) by setting df['%-rsi']. See more details on how this is done here. Note: Since the number of features prepended with % can multiply very quickly (10s of thousands of features are easily engineered using the multiplictative functionality of, e.g., include_shifted_candles and include_timeframes as described in the parameter table), these features are removed from the dataframe that is returned from FreqAI to the strategy. To keep a particular type of feature for plotting purposes, you would prepend it with %%. Datatype: Depends on the output of the model."},{"location":"freqai-configuration/#setting-the-startup_candle_count","title":"Setting the startup_candle_count","text":"

The startup_candle_count in the FreqAI strategy needs to be set up in the same way as in the standard Freqtrade strategy (see details here). This value is used by Freqtrade to ensure that a sufficient amount of data is provided when calling the dataprovider, to avoid any NaNs at the beginning of the first training. You can easily set this value by identifying the longest period (in candle units) which is passed to the indicator creation functions (e.g., TA-Lib functions). In the presented example, startup_candle_count is 20 since this is the maximum value in indicators_periods_candles.

Note

There are instances where the TA-Lib functions actually require more data than just the passed period or else the feature dataset gets populated with NaNs. Anecdotally, multiplying the startup_candle_count by 2 always leads to a fully NaN free training dataset. Hence, it is typically safest to multiply the expected startup_candle_count by 2. Look out for this log message to confirm that the data is clean:

2022-08-31 15:14:04 - freqtrade.freqai.data_kitchen - INFO - dropped 0 training points due to NaNs in populated dataset 4319.\n
"},{"location":"freqai-configuration/#creating-a-dynamic-target-threshold","title":"Creating a dynamic target threshold","text":"

Deciding when to enter or exit a trade can be done in a dynamic way to reflect current market conditions. FreqAI allows you to return additional information from the training of a model (more info here). For example, the &*_std/mean return values describe the statistical distribution of the target/label during the most recent training. Comparing a given prediction to these values allows you to know the rarity of the prediction. In templates/FreqaiExampleStrategy.py, the target_roi and sell_roi are defined to be 1.25 z-scores away from the mean which causes predictions that are closer to the mean to be filtered out.

dataframe[\"target_roi\"] = dataframe[\"&-s_close_mean\"] + dataframe[\"&-s_close_std\"] * 1.25\ndataframe[\"sell_roi\"] = dataframe[\"&-s_close_mean\"] - dataframe[\"&-s_close_std\"] * 1.25\n

To consider the population of historical predictions for creating the dynamic target instead of information from the training as discussed above, you would set fit_live_predictions_candles in the config to the number of historical prediction candles you wish to use to generate target statistics.

    \"freqai\": {\n\"fit_live_predictions_candles\": 300,\n}\n

If this value is set, FreqAI will initially use the predictions from the training data and subsequently begin introducing real prediction data as it is generated. FreqAI will save this historical data to be reloaded if you stop and restart a model with the same identifier.

"},{"location":"freqai-configuration/#using-different-prediction-models","title":"Using different prediction models","text":"

FreqAI has multiple example prediction model libraries that are ready to be used as is via the flag --freqaimodel. These libraries include CatBoost, LightGBM, and XGBoost regression, classification, and multi-target models, and can be found in freqai/prediction_models/.

Regression and classification models differ in what targets they predict - a regression model will predict a target of continuous values, for example what price BTC will be at tomorrow, whilst a classifier will predict a target of discrete values, for example if the price of BTC will go up tomorrow or not. This means that you have to specify your targets differently depending on which model type you are using (see details below).

All of the aforementioned model libraries implement gradient boosted decision tree algorithms. They all work on the principle of ensemble learning, where predictions from multiple simple learners are combined to get a final prediction that is more stable and generalized. The simple learners in this case are decision trees. Gradient boosting refers to the method of learning, where each simple learner is built in sequence - the subsequent learner is used to improve on the error from the previous learner. If you want to learn more about the different model libraries you can find the information in their respective docs:

  • CatBoost: https://catboost.ai/en/docs/
  • LightGBM: https://lightgbm.readthedocs.io/en/v3.3.2/#
  • XGBoost: https://xgboost.readthedocs.io/en/stable/#

There are also numerous online articles describing and comparing the algorithms. Some relatively lightweight examples would be CatBoost vs. LightGBM vs. XGBoost \u2014 Which is the best algorithm? and XGBoost, LightGBM or CatBoost \u2014 which boosting algorithm should I use?. Keep in mind that the performance of each model is highly dependent on the application and so any reported metrics might not be true for your particular use of the model.

Apart from the models already available in FreqAI, it is also possible to customize and create your own prediction models using the IFreqaiModel class. You are encouraged to inherit fit(), train(), and predict() to customize various aspects of the training procedures. You can place custom FreqAI models in user_data/freqaimodels - and freqtrade will pick them up from there based on the provided --freqaimodel name - which has to correspond to the class name of your custom model. Make sure to use unique names to avoid overriding built-in models.

"},{"location":"freqai-configuration/#setting-model-targets","title":"Setting model targets","text":""},{"location":"freqai-configuration/#regressors","title":"Regressors","text":"

If you are using a regressor, you need to specify a target that has continuous values. FreqAI includes a variety of regressors, such as the CatboostRegressorvia the flag --freqaimodel CatboostRegressor. An example of how you could set a regression target for predicting the price 100 candles into the future would be

df['&s-close_price'] = df['close'].shift(-100)\n

If you want to predict multiple targets, you need to define multiple labels using the same syntax as shown above.

"},{"location":"freqai-configuration/#classifiers","title":"Classifiers","text":"

If you are using a classifier, you need to specify a target that has discrete values. FreqAI includes a variety of classifiers, such as the CatboostClassifier via the flag --freqaimodel CatboostClassifier. If you elects to use a classifier, the classes need to be set using strings. For example, if you want to predict if the price 100 candles into the future goes up or down you would set

df['&s-up_or_down'] = np.where( df[\"close\"].shift(-100) > df[\"close\"], 'up', 'down')\n

If you want to predict multiple targets you must specify all labels in the same label column. You could, for example, add the label same to define where the price was unchanged by setting

df['&s-up_or_down'] = np.where( df[\"close\"].shift(-100) > df[\"close\"], 'up', 'down')\ndf['&s-up_or_down'] = np.where( df[\"close\"].shift(-100) == df[\"close\"], 'same', df['&s-up_or_down'])\n
"},{"location":"freqai-configuration/#pytorch-module","title":"PyTorch Module","text":""},{"location":"freqai-configuration/#quick-start","title":"Quick start","text":"

The easiest way to quickly run a pytorch model is with the following command (for regression task):

freqtrade trade --config config_examples/config_freqai.example.json --strategy FreqaiExampleStrategy --freqaimodel PyTorchMLPRegressor --strategy-path freqtrade/templates 

Installation/docker

The PyTorch module requires large packages such as torch, which should be explicitly requested during ./setup.sh -i by answering \"y\" to the question \"Do you also want dependencies for freqai-rl or PyTorch (~700mb additional space required) [y/N]?\". Users who prefer docker should ensure they use the docker image appended with _freqaitorch.

"},{"location":"freqai-configuration/#structure","title":"Structure","text":""},{"location":"freqai-configuration/#model","title":"Model","text":"

You can construct your own Neural Network architecture in PyTorch by simply defining your nn.Module class inside your custom IFreqaiModel file and then using that class in your def train() function. Here is an example of logistic regression model implementation using PyTorch (should be used with nn.BCELoss criterion) for classification tasks.

class LogisticRegression(nn.Module):\n    def __init__(self, input_size: int):\n        super().__init__()\n        # Define your layers\n        self.linear = nn.Linear(input_size, 1)\n        self.activation = nn.Sigmoid()\n\n    def forward(self, x: torch.Tensor) -> torch.Tensor:\n        # Define the forward pass\n        out = self.linear(x)\n        out = self.activation(out)\n        return out\n\nclass MyCoolPyTorchClassifier(BasePyTorchClassifier):\n\"\"\"\n    This is a custom IFreqaiModel showing how a user might setup their own \n    custom Neural Network architecture for their training.\n    \"\"\"\n\n    @property\n    def data_convertor(self) -> PyTorchDataConvertor:\n        return DefaultPyTorchDataConvertor(target_tensor_type=torch.float)\n\n    def __init__(self, **kwargs) -> None:\n        super().__init__(**kwargs)\n        config = self.freqai_info.get(\"model_training_parameters\", {})\n        self.learning_rate: float = config.get(\"learning_rate\",  3e-4)\n        self.model_kwargs: Dict[str, Any] = config.get(\"model_kwargs\",  {})\n        self.trainer_kwargs: Dict[str, Any] = config.get(\"trainer_kwargs\",  {})\n\n    def fit(self, data_dictionary: Dict, dk: FreqaiDataKitchen, **kwargs) -> Any:\n\"\"\"\n        User sets up the training and test data to fit their desired model here\n        :param data_dictionary: the dictionary holding all data for train, test,\n            labels, weights\n        :param dk: The datakitchen object for the current coin/model\n        \"\"\"\n\n        class_names = self.get_class_names()\n        self.convert_label_column_to_int(data_dictionary, dk, class_names)\n        n_features = data_dictionary[\"train_features\"].shape[-1]\n        model = LogisticRegression(\n            input_dim=n_features\n        )\n        model.to(self.device)\n        optimizer = torch.optim.AdamW(model.parameters(), lr=self.learning_rate)\n        criterion = torch.nn.CrossEntropyLoss()\n        init_model = self.get_init_model(dk.pair)\n        trainer = PyTorchModelTrainer(\n            model=model,\n            optimizer=optimizer,\n            criterion=criterion,\n            model_meta_data={\"class_names\": class_names},\n            device=self.device,\n            init_model=init_model,\n            data_convertor=self.data_convertor,\n            **self.trainer_kwargs,\n        )\n        trainer.fit(data_dictionary, self.splits)\n        return trainer\n
"},{"location":"freqai-configuration/#trainer","title":"Trainer","text":"

The PyTorchModelTrainer performs the idiomatic PyTorch train loop: Define our model, loss function, and optimizer, and then move them to the appropriate device (GPU or CPU). Inside the loop, we iterate through the batches in the dataloader, move the data to the device, compute the prediction and loss, backpropagate, and update the model parameters using the optimizer.

In addition, the trainer is responsible for the following: - saving and loading the model - converting the data from pandas.DataFrame to torch.Tensor.

"},{"location":"freqai-configuration/#integration-with-freqai-module","title":"Integration with Freqai module","text":"

Like all freqai models, PyTorch models inherit IFreqaiModel. IFreqaiModel declares three abstract methods: train, fit, and predict. we implement these methods in three levels of hierarchy. From top to bottom:

  1. BasePyTorchModel - Implements the train method. all BasePyTorch* inherit it. responsible for general data preparation (e.g., data normalization) and calling the fit method. Sets device attribute used by children classes. Sets model_type attribute used by the parent class.
  2. BasePyTorch* - Implements the predict method. Here, the * represents a group of algorithms, such as classifiers or regressors. responsible for data preprocessing, predicting, and postprocessing if needed.
  3. PyTorch*Classifier / PyTorch*Regressor - implements the fit method. responsible for the main train flaw, where we initialize the trainer and model objects.

"},{"location":"freqai-configuration/#full-example","title":"Full example","text":"

Building a PyTorch regressor using MLP (multilayer perceptron) model, MSELoss criterion, and AdamW optimizer.

class PyTorchMLPRegressor(BasePyTorchRegressor):\n    def __init__(self, **kwargs) -> None:\n        super().__init__(**kwargs)\n        config = self.freqai_info.get(\"model_training_parameters\", {})\n        self.learning_rate: float = config.get(\"learning_rate\",  3e-4)\n        self.model_kwargs: Dict[str, Any] = config.get(\"model_kwargs\",  {})\n        self.trainer_kwargs: Dict[str, Any] = config.get(\"trainer_kwargs\",  {})\n\n    def fit(self, data_dictionary: Dict, dk: FreqaiDataKitchen, **kwargs) -> Any:\n        n_features = data_dictionary[\"train_features\"].shape[-1]\n        model = PyTorchMLPModel(\n            input_dim=n_features,\n            output_dim=1,\n            **self.model_kwargs\n        )\n        model.to(self.device)\n        optimizer = torch.optim.AdamW(model.parameters(), lr=self.learning_rate)\n        criterion = torch.nn.MSELoss()\n        init_model = self.get_init_model(dk.pair)\n        trainer = PyTorchModelTrainer(\n            model=model,\n            optimizer=optimizer,\n            criterion=criterion,\n            device=self.device,\n            init_model=init_model,\n            target_tensor_type=torch.float,\n            **self.trainer_kwargs,\n        )\n        trainer.fit(data_dictionary)\n        return trainer\n

Here we create a PyTorchMLPRegressor class that implements the fit method. The fit method specifies the training building blocks: model, optimizer, criterion, and trainer. We inherit both BasePyTorchRegressor and BasePyTorchModel, where the former implements the predict method that is suitable for our regression task, and the latter implements the train method.

Setting Class Names for Classifiers

When using classifiers, the user must declare the class names (or targets) by overriding the IFreqaiModel.class_names attribute. This is achieved by setting self.freqai.class_names in the FreqAI strategy inside the set_freqai_targets method.

For example, if you are using a binary classifier to predict price movements as up or down, you can set the class names as follows:

def set_freqai_targets(self, dataframe: DataFrame, metadata: Dict, **kwargs) -> DataFrame:\n    self.freqai.class_names = [\"down\", \"up\"]\n    dataframe['&s-up_or_down'] = np.where(dataframe[\"close\"].shift(-100) >\n                                              dataframe[\"close\"], 'up', 'down')\n\n    return dataframe\n
To see a full example, you can refer to the classifier test strategy class.

"},{"location":"freqai-developers/","title":"Development","text":""},{"location":"freqai-developers/#project-architecture","title":"Project architecture","text":"

The architecture and functions of FreqAI are generalized to encourages development of unique features, functions, models, etc.

The class structure and a detailed algorithmic overview is depicted in the following diagram:

As shown, there are three distinct objects comprising FreqAI:

  • IFreqaiModel - A singular persistent object containing all the necessary logic to collect, store, and process data, engineer features, run training, and inference models.
  • FreqaiDataKitchen - A non-persistent object which is created uniquely for each unique asset/model. Beyond metadata, it also contains a variety of data processing tools.
  • FreqaiDataDrawer - A singular persistent object containing all the historical predictions, models, and save/load methods.

There are a variety of built-in prediction models which inherit directly from IFreqaiModel. Each of these models have full access to all methods in IFreqaiModel and can therefore override any of those functions at will. However, advanced users will likely stick to overriding fit(), train(), predict(), and data_cleaning_train/predict().

"},{"location":"freqai-developers/#data-handling","title":"Data handling","text":"

FreqAI aims to organize model files, prediction data, and meta data in a way that simplifies post-processing and enhances crash resilience by automatic data reloading. The data is saved in a file structure,user_data_dir/models/, which contains all the data associated with the trainings and backtests. The FreqaiDataKitchen() relies heavily on the file structure for proper training and inferencing and should therefore not be manually modified.

"},{"location":"freqai-developers/#file-structure","title":"File structure","text":"

The file structure is automatically generated based on the model identifier set in the config. The following structure shows where the data is stored for post processing:

Structure Description config_*.json A copy of the model specific configuration file. historic_predictions.pkl A file containing all historic predictions generated during the lifetime of the identifier model during live deployment. historic_predictions.pkl is used to reload the model after a crash or a config change. A backup file is always held in case of corruption on the main file. FreqAI automatically detects corruption and replaces the corrupted file with the backup. pair_dictionary.json A file containing the training queue as well as the on disk location of the most recently trained model. sub-train-*_TIMESTAMP A folder containing all the files associated with a single model, such as: *_metadata.json - Metadata for the model, such as normalization max/min, expected training feature list, etc. *_model.* - The model file saved to disk for reloading from a crash. Can be joblib (typical boosting libs), zip (stable_baselines), hd5 (keras type), etc. *_pca_object.pkl - The Principal component analysis (PCA) transform (if principal_component_analysis: True is set in the config) which will be used to transform unseen prediction features. *_svm_model.pkl - The Support Vector Machine (SVM) model (if use_SVM_to_remove_outliers: True is set in the config) which is used to detect outliers in unseen prediction features. *_trained_df.pkl - The dataframe containing all the training features used to train the identifier model. This is used for computing the Dissimilarity Index (DI) and can also be used for post-processing. *_trained_dates.df.pkl - The dates associated with the trained_df.pkl, which is useful for post-processing.

The example file structure would look like this:

\u251c\u2500\u2500 models\n\u2502\u00a0\u00a0 \u2514\u2500\u2500 unique-id\n\u2502\u00a0\u00a0     \u251c\u2500\u2500 config_freqai.example.json\n\u2502\u00a0\u00a0     \u251c\u2500\u2500 historic_predictions.backup.pkl\n\u2502\u00a0\u00a0     \u251c\u2500\u2500 historic_predictions.pkl\n\u2502\u00a0\u00a0     \u251c\u2500\u2500 pair_dictionary.json\n\u2502\u00a0\u00a0     \u251c\u2500\u2500 sub-train-1INCH_1662821319\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_1inch_1662821319_metadata.json\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_1inch_1662821319_model.joblib\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_1inch_1662821319_pca_object.pkl\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_1inch_1662821319_svm_model.joblib\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_1inch_1662821319_trained_dates_df.pkl\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u2514\u2500\u2500 cb_1inch_1662821319_trained_df.pkl\n\u2502\u00a0\u00a0     \u251c\u2500\u2500 sub-train-1INCH_1662821371\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_1inch_1662821371_metadata.json\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_1inch_1662821371_model.joblib\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_1inch_1662821371_pca_object.pkl\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_1inch_1662821371_svm_model.joblib\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_1inch_1662821371_trained_dates_df.pkl\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u2514\u2500\u2500 cb_1inch_1662821371_trained_df.pkl\n\u2502\u00a0\u00a0     \u251c\u2500\u2500 sub-train-ADA_1662821344\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_ada_1662821344_metadata.json\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_ada_1662821344_model.joblib\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_ada_1662821344_pca_object.pkl\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_ada_1662821344_svm_model.joblib\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u251c\u2500\u2500 cb_ada_1662821344_trained_dates_df.pkl\n\u2502\u00a0\u00a0     \u2502\u00a0\u00a0 \u2514\u2500\u2500 cb_ada_1662821344_trained_df.pkl\n\u2502\u00a0\u00a0     \u2514\u2500\u2500 sub-train-ADA_1662821399\n\u2502\u00a0\u00a0         \u251c\u2500\u2500 cb_ada_1662821399_metadata.json\n\u2502\u00a0\u00a0         \u251c\u2500\u2500 cb_ada_1662821399_model.joblib\n\u2502\u00a0\u00a0         \u251c\u2500\u2500 cb_ada_1662821399_pca_object.pkl\n\u2502\u00a0\u00a0         \u251c\u2500\u2500 cb_ada_1662821399_svm_model.joblib\n\u2502\u00a0\u00a0         \u251c\u2500\u2500 cb_ada_1662821399_trained_dates_df.pkl\n\u2502\u00a0\u00a0         \u2514\u2500\u2500 cb_ada_1662821399_trained_df.pkl\n
"},{"location":"freqai-feature-engineering/","title":"Feature engineering","text":""},{"location":"freqai-feature-engineering/#defining-the-features","title":"Defining the features","text":"

Low level feature engineering is performed in the user strategy within a set of functions called feature_engineering_*. These function set the base features such as, RSI, MFI, EMA, SMA, time of day, volume, etc. The base features can be custom indicators or they can be imported from any technical-analysis library that you can find. FreqAI is equipped with a set of functions to simplify rapid large-scale feature engineering:

Function Description feature_engineering_expand_all() This optional function will automatically expand the defined features on the config defined indicator_periods_candles, include_timeframes, include_shifted_candles, and include_corr_pairs. feature_engineering_expand_basic() This optional function will automatically expand the defined features on the config defined include_timeframes, include_shifted_candles, and include_corr_pairs. Note: this function does not expand across include_periods_candles. feature_engineering_standard() This optional function will be called once with the dataframe of the base timeframe. This is the final function to be called, which means that the dataframe entering this function will contain all the features and columns from the base asset created by the other feature_engineering_expand functions. This function is a good place to do custom exotic feature extractions (e.g. tsfresh). This function is also a good place for any feature that should not be auto-expanded upon (e.g., day of the week). set_freqai_targets() Required function to set the targets for the model. All targets must be prepended with & to be recognized by the FreqAI internals.

Meanwhile, high level feature engineering is handled within \"feature_parameters\":{} in the FreqAI config. Within this file, it is possible to decide large scale feature expansions on top of the base_features such as \"including correlated pairs\" or \"including informative timeframes\" or even \"including recent candles.\"

It is advisable to start from the template feature_engineering_* functions in the source provided example strategy (found in templates/FreqaiExampleStrategy.py) to ensure that the feature definitions are following the correct conventions. Here is an example of how to set the indicators and labels in the strategy:

    def feature_engineering_expand_all(self, dataframe: DataFrame, period, metadata, **kwargs) -> DataFrame:\n\"\"\"\n        *Only functional with FreqAI enabled strategies*\n        This function will automatically expand the defined features on the config defined\n        `indicator_periods_candles`, `include_timeframes`, `include_shifted_candles`, and\n        `include_corr_pairs`. In other words, a single feature defined in this function\n        will automatically expand to a total of\n        `indicator_periods_candles` * `include_timeframes` * `include_shifted_candles` *\n        `include_corr_pairs` numbers of features added to the model.\n\n        All features must be prepended with `%` to be recognized by FreqAI internals.\n\n        Access metadata such as the current pair/timeframe/period with:\n\n        `metadata[\"pair\"]` `metadata[\"tf\"]`  `metadata[\"period\"]`\n\n        :param df: strategy dataframe which will receive the features\n        :param period: period of the indicator - usage example:\n        :param metadata: metadata of current pair\n        dataframe[\"%-ema-period\"] = ta.EMA(dataframe, timeperiod=period)\n        \"\"\"\n\n        dataframe[\"%-rsi-period\"] = ta.RSI(dataframe, timeperiod=period)\n        dataframe[\"%-mfi-period\"] = ta.MFI(dataframe, timeperiod=period)\n        dataframe[\"%-adx-period\"] = ta.ADX(dataframe, timeperiod=period)\n        dataframe[\"%-sma-period\"] = ta.SMA(dataframe, timeperiod=period)\n        dataframe[\"%-ema-period\"] = ta.EMA(dataframe, timeperiod=period)\n\n        bollinger = qtpylib.bollinger_bands(\n            qtpylib.typical_price(dataframe), window=period, stds=2.2\n        )\n        dataframe[\"bb_lowerband-period\"] = bollinger[\"lower\"]\n        dataframe[\"bb_middleband-period\"] = bollinger[\"mid\"]\n        dataframe[\"bb_upperband-period\"] = bollinger[\"upper\"]\n\n        dataframe[\"%-bb_width-period\"] = (\n            dataframe[\"bb_upperband-period\"]\n            - dataframe[\"bb_lowerband-period\"]\n        ) / dataframe[\"bb_middleband-period\"]\n        dataframe[\"%-close-bb_lower-period\"] = (\n            dataframe[\"close\"] / dataframe[\"bb_lowerband-period\"]\n        )\n\n        dataframe[\"%-roc-period\"] = ta.ROC(dataframe, timeperiod=period)\n\n        dataframe[\"%-relative_volume-period\"] = (\n            dataframe[\"volume\"] / dataframe[\"volume\"].rolling(period).mean()\n        )\n\n        return dataframe\n\n    def feature_engineering_expand_basic(self, dataframe: DataFrame, metadata, **kwargs) -> DataFrame:\n\"\"\"\n        *Only functional with FreqAI enabled strategies*\n        This function will automatically expand the defined features on the config defined\n        `include_timeframes`, `include_shifted_candles`, and `include_corr_pairs`.\n        In other words, a single feature defined in this function\n        will automatically expand to a total of\n        `include_timeframes` * `include_shifted_candles` * `include_corr_pairs`\n        numbers of features added to the model.\n\n        Features defined here will *not* be automatically duplicated on user defined\n        `indicator_periods_candles`\n\n        Access metadata such as the current pair/timeframe with:\n\n        `metadata[\"pair\"]` `metadata[\"tf\"]`\n\n        All features must be prepended with `%` to be recognized by FreqAI internals.\n\n        :param df: strategy dataframe which will receive the features\n        :param metadata: metadata of current pair\n        dataframe[\"%-pct-change\"] = dataframe[\"close\"].pct_change()\n        dataframe[\"%-ema-200\"] = ta.EMA(dataframe, timeperiod=200)\n        \"\"\"\n        dataframe[\"%-pct-change\"] = dataframe[\"close\"].pct_change()\n        dataframe[\"%-raw_volume\"] = dataframe[\"volume\"]\n        dataframe[\"%-raw_price\"] = dataframe[\"close\"]\n        return dataframe\n\n    def feature_engineering_standard(self, dataframe: DataFrame, metadata, **kwargs) -> DataFrame:\n\"\"\"\n        *Only functional with FreqAI enabled strategies*\n        This optional function will be called once with the dataframe of the base timeframe.\n        This is the final function to be called, which means that the dataframe entering this\n        function will contain all the features and columns created by all other\n        freqai_feature_engineering_* functions.\n\n        This function is a good place to do custom exotic feature extractions (e.g. tsfresh).\n        This function is a good place for any feature that should not be auto-expanded upon\n        (e.g. day of the week).\n\n        Access metadata such as the current pair with:\n\n        `metadata[\"pair\"]`\n\n        All features must be prepended with `%` to be recognized by FreqAI internals.\n\n        :param df: strategy dataframe which will receive the features\n        :param metadata: metadata of current pair\n        usage example: dataframe[\"%-day_of_week\"] = (dataframe[\"date\"].dt.dayofweek + 1) / 7\n        \"\"\"\n        dataframe[\"%-day_of_week\"] = (dataframe[\"date\"].dt.dayofweek + 1) / 7\n        dataframe[\"%-hour_of_day\"] = (dataframe[\"date\"].dt.hour + 1) / 25\n        return dataframe\n\n    def set_freqai_targets(self, dataframe: DataFrame, metadata, **kwargs) -> DataFrame:\n\"\"\"\n        *Only functional with FreqAI enabled strategies*\n        Required function to set the targets for the model.\n        All targets must be prepended with `&` to be recognized by the FreqAI internals.\n\n        Access metadata such as the current pair with:\n\n        `metadata[\"pair\"]`\n\n        :param df: strategy dataframe which will receive the targets\n        :param metadata: metadata of current pair\n        usage example: dataframe[\"&-target\"] = dataframe[\"close\"].shift(-1) / dataframe[\"close\"]\n        \"\"\"\n        dataframe[\"&-s_close\"] = (\n            dataframe[\"close\"]\n            .shift(-self.freqai_info[\"feature_parameters\"][\"label_period_candles\"])\n            .rolling(self.freqai_info[\"feature_parameters\"][\"label_period_candles\"])\n            .mean()\n            / dataframe[\"close\"]\n            - 1\n            )\n\n        return dataframe\n

In the presented example, the user does not wish to pass the bb_lowerband as a feature to the model, and has therefore not prepended it with %. The user does, however, wish to pass bb_width to the model for training/prediction and has therefore prepended it with %.

After having defined the base features, the next step is to expand upon them using the powerful feature_parameters in the configuration file:

    \"freqai\": {\n//...\n\"feature_parameters\" : {\n\"include_timeframes\": [\"5m\",\"15m\",\"4h\"],\n\"include_corr_pairlist\": [\n\"ETH/USD\",\n\"LINK/USD\",\n\"BNB/USD\"\n],\n\"label_period_candles\": 24,\n\"include_shifted_candles\": 2,\n\"indicator_periods_candles\": [10, 20]\n},\n//...\n}\n

The include_timeframes in the config above are the timeframes (tf) of each call to feature_engineering_expand_*() in the strategy. In the presented case, the user is asking for the 5m, 15m, and 4h timeframes of the rsi, mfi, roc, and bb_width to be included in the feature set.

You can ask for each of the defined features to be included also for informative pairs using the include_corr_pairlist. This means that the feature set will include all the features from feature_engineering_expand_*() on all the include_timeframes for each of the correlated pairs defined in the config (ETH/USD, LINK/USD, and BNB/USD in the presented example).

include_shifted_candles indicates the number of previous candles to include in the feature set. For example, include_shifted_candles: 2 tells FreqAI to include the past 2 candles for each of the features in the feature set.

In total, the number of features the user of the presented example strat has created is: length of include_timeframes * no. features in feature_engineering_expand_*() * length of include_corr_pairlist * no. include_shifted_candles * length of indicator_periods_candles \\(= 3 * 3 * 3 * 2 * 2 = 108\\).

"},{"location":"freqai-feature-engineering/#gain-finer-control-over-feature_engineering_-functions-with-metadata","title":"Gain finer control over feature_engineering_* functions with metadata","text":"

All feature_engineering_* and set_freqai_targets() functions are passed a metadata dictionary which contains information about the pair, tf (timeframe), and period that FreqAI is automating for feature building. As such, a user can use metadata inside feature_engineering_* functions as criteria for blocking/reserving features for certain timeframes, periods, pairs etc.

def feature_engineering_expand_all(self, dataframe: DataFrame, period, metadata, **kwargs) -> DataFrame:\n    if metadata[\"tf\"] == \"1h\":\n        dataframe[\"%-roc-period\"] = ta.ROC(dataframe, timeperiod=period)\n

This will block ta.ROC() from being added to any timeframes other than \"1h\".

"},{"location":"freqai-feature-engineering/#returning-additional-info-from-training","title":"Returning additional info from training","text":"

Important metrics can be returned to the strategy at the end of each model training by assigning them to dk.data['extra_returns_per_train']['my_new_value'] = XYZ inside the custom prediction model class.

FreqAI takes the my_new_value assigned in this dictionary and expands it to fit the dataframe that is returned to the strategy. You can then use the returned metrics in your strategy through dataframe['my_new_value']. An example of how return values can be used in FreqAI are the &*_mean and &*_std values that are used to created a dynamic target threshold.

Another example, where the user wants to use live metrics from the trade database, is shown below:

    \"freqai\": {\n\"extra_returns_per_train\": {\"total_profit\": 4}\n}\n

You need to set the standard dictionary in the config so that FreqAI can return proper dataframe shapes. These values will likely be overridden by the prediction model, but in the case where the model has yet to set them, or needs a default initial value, the pre-set values are what will be returned.

"},{"location":"freqai-feature-engineering/#feature-normalization","title":"Feature normalization","text":"

FreqAI is strict when it comes to data normalization. The train features, \\(X^{train}\\), are always normalized to [-1, 1] using a shifted min-max normalization:

\\[X^{train}_{norm} = 2 * \\frac{X^{train} - X^{train}.min()}{X^{train}.max() - X^{train}.min()} - 1\\]

All other data (test data and unseen prediction data in dry/live/backtest) is always automatically normalized to the training feature space according to industry standards. FreqAI stores all the metadata required to ensure that test and prediction features will be properly normalized and that predictions are properly denormalized. For this reason, it is not recommended to eschew industry standards and modify FreqAI internals - however - advanced users can do so by inheriting train() in their custom IFreqaiModel and using their own normalization functions.

"},{"location":"freqai-feature-engineering/#data-dimensionality-reduction-with-principal-component-analysis","title":"Data dimensionality reduction with Principal Component Analysis","text":"

You can reduce the dimensionality of your features by activating the principal_component_analysis in the config:

    \"freqai\": {\n\"feature_parameters\" : {\n\"principal_component_analysis\": true\n}\n}\n

This will perform PCA on the features and reduce their dimensionality so that the explained variance of the data set is >= 0.999. Reducing data dimensionality makes training the model faster and hence allows for more up-to-date models.

"},{"location":"freqai-feature-engineering/#inlier-metric","title":"Inlier metric","text":"

The inlier_metric is a metric aimed at quantifying how similar the features of a data point are to the most recent historical data points.

You define the lookback window by setting inlier_metric_window and FreqAI computes the distance between the present time point and each of the previous inlier_metric_window lookback points. A Weibull function is fit to each of the lookback distributions and its cumulative distribution function (CDF) is used to produce a quantile for each lookback point. The inlier_metric is then computed for each time point as the average of the corresponding lookback quantiles. The figure below explains the concept for an inlier_metric_window of 5.

FreqAI adds the inlier_metric to the training features and hence gives the model access to a novel type of temporal information.

This function does not remove outliers from the data set.

"},{"location":"freqai-feature-engineering/#weighting-features-for-temporal-importance","title":"Weighting features for temporal importance","text":"

FreqAI allows you to set a weight_factor to weight recent data more strongly than past data via an exponential function:

\\[ W_i = \\exp(\\frac{-i}{\\alpha*n}) \\]

where \\(W_i\\) is the weight of data point \\(i\\) in a total set of \\(n\\) data points. Below is a figure showing the effect of different weight factors on the data points in a feature set.

"},{"location":"freqai-feature-engineering/#outlier-detection","title":"Outlier detection","text":"

Equity and crypto markets suffer from a high level of non-patterned noise in the form of outlier data points. FreqAI implements a variety of methods to identify such outliers and hence mitigate risk.

"},{"location":"freqai-feature-engineering/#identifying-outliers-with-the-dissimilarity-index-di","title":"Identifying outliers with the Dissimilarity Index (DI)","text":"

The Dissimilarity Index (DI) aims to quantify the uncertainty associated with each prediction made by the model.

You can tell FreqAI to remove outlier data points from the training/test data sets using the DI by including the following statement in the config:

    \"freqai\": {\n\"feature_parameters\" : {\n\"DI_threshold\": 1\n}\n}\n

The DI allows predictions which are outliers (not existent in the model feature space) to be thrown out due to low levels of certainty. To do so, FreqAI measures the distance between each training data point (feature vector), \\(X_{a}\\), and all other training data points:

\\[ d_{ab} = \\sqrt{\\sum_{j=1}^p(X_{a,j}-X_{b,j})^2} \\]

where \\(d_{ab}\\) is the distance between the normalized points \\(a\\) and \\(b\\), and \\(p\\) is the number of features, i.e., the length of the vector \\(X\\). The characteristic distance, \\(\\overline{d}\\), for a set of training data points is simply the mean of the average distances:

\\[ \\overline{d} = \\sum_{a=1}^n(\\sum_{b=1}^n(d_{ab}/n)/n) \\]

\\(\\overline{d}\\) quantifies the spread of the training data, which is compared to the distance between a new prediction feature vectors, \\(X_k\\) and all the training data:

\\[ d_k = \\arg \\min d_{k,i} \\]

This enables the estimation of the Dissimilarity Index as:

\\[ DI_k = d_k/\\overline{d} \\]

You can tweak the DI through the DI_threshold to increase or decrease the extrapolation of the trained model. A higher DI_threshold means that the DI is more lenient and allows predictions further away from the training data to be used whilst a lower DI_threshold has the opposite effect and hence discards more predictions.

Below is a figure that describes the DI for a 3D data set.

"},{"location":"freqai-feature-engineering/#identifying-outliers-using-a-support-vector-machine-svm","title":"Identifying outliers using a Support Vector Machine (SVM)","text":"

You can tell FreqAI to remove outlier data points from the training/test data sets using a Support Vector Machine (SVM) by including the following statement in the config:

    \"freqai\": {\n\"feature_parameters\" : {\n\"use_SVM_to_remove_outliers\": true\n}\n}\n

The SVM will be trained on the training data and any data point that the SVM deems to be beyond the feature space will be removed.

FreqAI uses sklearn.linear_model.SGDOneClassSVM (details are available on scikit-learn's webpage here (external website)) and you can elect to provide additional parameters for the SVM, such as shuffle, and nu.

The parameter shuffle is by default set to False to ensure consistent results. If it is set to True, running the SVM multiple times on the same data set might result in different outcomes due to max_iter being to low for the algorithm to reach the demanded tol. Increasing max_iter solves this issue but causes the procedure to take longer time.

The parameter nu, very broadly, is the amount of data points that should be considered outliers and should be between 0 and 1.

"},{"location":"freqai-feature-engineering/#identifying-outliers-with-dbscan","title":"Identifying outliers with DBSCAN","text":"

You can configure FreqAI to use DBSCAN to cluster and remove outliers from the training/test data set or incoming outliers from predictions, by activating use_DBSCAN_to_remove_outliers in the config:

    \"freqai\": {\n\"feature_parameters\" : {\n\"use_DBSCAN_to_remove_outliers\": true\n}\n}\n

DBSCAN is an unsupervised machine learning algorithm that clusters data without needing to know how many clusters there should be.

Given a number of data points \\(N\\), and a distance \\(\\varepsilon\\), DBSCAN clusters the data set by setting all data points that have \\(N-1\\) other data points within a distance of \\(\\varepsilon\\) as core points. A data point that is within a distance of \\(\\varepsilon\\) from a core point but that does not have \\(N-1\\) other data points within a distance of \\(\\varepsilon\\) from itself is considered an edge point. A cluster is then the collection of core points and edge points. Data points that have no other data points at a distance \\(<\\varepsilon\\) are considered outliers. The figure below shows a cluster with \\(N = 3\\).

FreqAI uses sklearn.cluster.DBSCAN (details are available on scikit-learn's webpage here (external website)) with min_samples (\\(N\\)) taken as \u00bc of the no. of time points (candles) in the feature set. eps (\\(\\varepsilon\\)) is computed automatically as the elbow point in the k-distance graph computed from the nearest neighbors in the pairwise distances of all data points in the feature set.

"},{"location":"freqai-parameter-table/","title":"Parameter table","text":"

The table below will list all configuration parameters available for FreqAI. Some of the parameters are exemplified in config_examples/config_freqai.example.json.

Mandatory parameters are marked as Required and have to be set in one of the suggested ways.

"},{"location":"freqai-parameter-table/#general-configuration-parameters","title":"General configuration parameters","text":"Parameter Description General configuration parameters within the config.freqai tree freqai Required. The parent dictionary containing all the parameters for controlling FreqAI. Datatype: Dictionary. train_period_days Required. Number of days to use for the training data (width of the sliding window). Datatype: Positive integer. backtest_period_days Required. Number of days to inference from the trained model before sliding the train_period_days window defined above, and retraining the model during backtesting (more info here). This can be fractional days, but beware that the provided timerange will be divided by this number to yield the number of trainings necessary to complete the backtest. Datatype: Float. identifier Required. A unique ID for the current model. If models are saved to disk, the identifier allows for reloading specific pre-trained models/data. Datatype: String. live_retrain_hours Frequency of retraining during dry/live runs. Datatype: Float > 0. Default: 0 (models retrain as often as possible). expiration_hours Avoid making predictions if a model is more than expiration_hours old. Datatype: Positive integer. Default: 0 (models never expire). purge_old_models Number of models to keep on disk (not relevant to backtesting). Default is 2, which means that dry/live runs will keep the latest 2 models on disk. Setting to 0 keeps all models. This parameter also accepts a boolean to maintain backwards compatibility. Datatype: Integer. Default: 2. save_backtest_models Save models to disk when running backtesting. Backtesting operates most efficiently by saving the prediction data and reusing them directly for subsequent runs (when you wish to tune entry/exit parameters). Saving backtesting models to disk also allows to use the same model files for starting a dry/live instance with the same model identifier. Datatype: Boolean. Default: False (no models are saved). fit_live_predictions_candles Number of historical candles to use for computing target (label) statistics from prediction data, instead of from the training dataset (more information can be found here). Datatype: Positive integer. continual_learning Use the final state of the most recently trained model as starting point for the new model, allowing for incremental learning (more information can be found here). Datatype: Boolean. Default: False. write_metrics_to_disk Collect train timings, inference timings and cpu usage in json file. Datatype: Boolean. Default: False data_kitchen_thread_count Designate the number of threads you want to use for data processing (outlier methods, normalization, etc.). This has no impact on the number of threads used for training. If user does not set it (default), FreqAI will use max number of threads - 2 (leaving 1 physical core available for Freqtrade bot and FreqUI) Datatype: Positive integer."},{"location":"freqai-parameter-table/#feature-parameters","title":"Feature parameters","text":"Parameter Description Feature parameters within the freqai.feature_parameters sub dictionary feature_parameters A dictionary containing the parameters used to engineer the feature set. Details and examples are shown here. Datatype: Dictionary. include_timeframes A list of timeframes that all indicators in feature_engineering_expand_*() will be created for. The list is added as features to the base indicators dataset. Datatype: List of timeframes (strings). include_corr_pairlist A list of correlated coins that FreqAI will add as additional features to all pair_whitelist coins. All indicators set in feature_engineering_expand_*() during feature engineering (see details here) will be created for each correlated coin. The correlated coins features are added to the base indicators dataset. Datatype: List of assets (strings). label_period_candles Number of candles into the future that the labels are created for. This is used in feature_engineering_expand_all() (see templates/FreqaiExampleStrategy.py for detailed usage). You can create custom labels and choose whether to make use of this parameter or not. Datatype: Positive integer. include_shifted_candles Add features from previous candles to subsequent candles with the intent of adding historical information. If used, FreqAI will duplicate and shift all features from the include_shifted_candles previous candles so that the information is available for the subsequent candle. Datatype: Positive integer. weight_factor Weight training data points according to their recency (see details here). Datatype: Positive float (typically < 1). indicator_max_period_candles No longer used (#7325). Replaced by startup_candle_count which is set in the strategy. startup_candle_count is timeframe independent and defines the maximum period used in feature_engineering_*() for indicator creation. FreqAI uses this parameter together with the maximum timeframe in include_time_frames to calculate how many data points to download such that the first data point does not include a NaN. Datatype: Positive integer. indicator_periods_candles Time periods to calculate indicators for. The indicators are added to the base indicator dataset. Datatype: List of positive integers. principal_component_analysis Automatically reduce the dimensionality of the data set using Principal Component Analysis. See details about how it works here Datatype: Boolean. Default: False. plot_feature_importances Create a feature importance plot for each model for the top/bottom plot_feature_importances number of features. Plot is stored in user_data/models/<identifier>/sub-train-<COIN>_<timestamp>.html. Datatype: Integer. Default: 0. DI_threshold Activates the use of the Dissimilarity Index for outlier detection when set to > 0. See details about how it works here. Datatype: Positive float (typically < 1). use_SVM_to_remove_outliers Train a support vector machine to detect and remove outliers from the training dataset, as well as from incoming data points. See details about how it works here. Datatype: Boolean. svm_params All parameters available in Sklearn's SGDOneClassSVM(). See details about some select parameters here. Datatype: Dictionary. use_DBSCAN_to_remove_outliers Cluster data using the DBSCAN algorithm to identify and remove outliers from training and prediction data. See details about how it works here. Datatype: Boolean. inlier_metric_window If set, FreqAI adds an inlier_metric to the training feature set and set the lookback to be the inlier_metric_window, i.e., the number of previous time points to compare the current candle to. Details of how the inlier_metric is computed can be found here. Datatype: Integer. Default: 0. noise_standard_deviation If set, FreqAI adds noise to the training features with the aim of preventing overfitting. FreqAI generates random deviates from a gaussian distribution with a standard deviation of noise_standard_deviation and adds them to all data points. noise_standard_deviation should be kept relative to the normalized space, i.e., between -1 and 1. In other words, since data in FreqAI is always normalized to be between -1 and 1, noise_standard_deviation: 0.05 would result in 32% of the data being randomly increased/decreased by more than 2.5% (i.e., the percent of data falling within the first standard deviation). Datatype: Integer. Default: 0. outlier_protection_percentage Enable to prevent outlier detection methods from discarding too much data. If more than outlier_protection_percentage % of points are detected as outliers by the SVM or DBSCAN, FreqAI will log a warning message and ignore outlier detection, i.e., the original dataset will be kept intact. If the outlier protection is triggered, no predictions will be made based on the training dataset. Datatype: Float. Default: 30. reverse_train_test_order Split the feature dataset (see below) and use the latest data split for training and test on historical split of the data. This allows the model to be trained up to the most recent data point, while avoiding overfitting. However, you should be careful to understand the unorthodox nature of this parameter before employing it. Datatype: Boolean. Default: False (no reversal). shuffle_after_split Split the data into train and test sets, and then shuffle both sets individually. Datatype: Boolean. Default: False. buffer_train_data_candles Cut buffer_train_data_candles off the beginning and end of the training data after the indicators were populated. The main example use is when predicting maxima and minima, the argrelextrema function cannot know the maxima/minima at the edges of the timerange. To improve model accuracy, it is best to compute argrelextrema on the full timerange and then use this function to cut off the edges (buffer) by the kernel. In another case, if the targets are set to a shifted price movement, this buffer is unnecessary because the shifted candles at the end of the timerange will be NaN and FreqAI will automatically cut those off of the training dataset. Datatype: Integer. Default: 0."},{"location":"freqai-parameter-table/#data-split-parameters","title":"Data split parameters","text":"Parameter Description Data split parameters within the freqai.data_split_parameters sub dictionary data_split_parameters Include any additional parameters available from scikit-learn test_train_split(), which are shown here (external website). Datatype: Dictionary. test_size The fraction of data that should be used for testing instead of training. Datatype: Positive float < 1. shuffle Shuffle the training data points during training. Typically, to not remove the chronological order of data in time-series forecasting, this is set to False. Datatype: Boolean. Defaut: False."},{"location":"freqai-parameter-table/#model-training-parameters","title":"Model training parameters","text":"Parameter Description Model training parameters within the freqai.model_training_parameters sub dictionary model_training_parameters A flexible dictionary that includes all parameters available by the selected model library. For example, if you use LightGBMRegressor, this dictionary can contain any parameter available by the LightGBMRegressor here (external website). If you select a different model, this dictionary can contain any parameter from that model. A list of the currently available models can be found here. Datatype: Dictionary. n_estimators The number of boosted trees to fit in the training of the model. Datatype: Integer. learning_rate Boosting learning rate during training of the model. Datatype: Float. n_jobs, thread_count, task_type Set the number of threads for parallel processing and the task_type (gpu or cpu). Different model libraries use different parameter names. Datatype: Float."},{"location":"freqai-parameter-table/#reinforcement-learning-parameters","title":"Reinforcement Learning parameters","text":"Parameter Description Reinforcement Learning Parameters within the freqai.rl_config sub dictionary rl_config A dictionary containing the control parameters for a Reinforcement Learning model. Datatype: Dictionary. train_cycles Training time steps will be set based on the `train_cycles * number of training data points. Datatype: Integer. cpu_count Number of processors to dedicate to the Reinforcement Learning training process. Datatype: int. max_trade_duration_candles Guides the agent training to keep trades below desired length. Example usage shown in prediction_models/ReinforcementLearner.py within the customizable calculate_reward() function. Datatype: int. model_type Model string from stable_baselines3 or SBcontrib. Available strings include: 'TRPO', 'ARS', 'RecurrentPPO', 'MaskablePPO', 'PPO', 'A2C', 'DQN'. User should ensure that model_training_parameters match those available to the corresponding stable_baselines3 model by visiting their documentaiton. PPO doc (external website) Datatype: string. policy_type One of the available policy types from stable_baselines3 Datatype: string. max_training_drawdown_pct The maximum drawdown that the agent is allowed to experience during training. Datatype: float. Default: 0.8 cpu_count Number of threads/cpus to dedicate to the Reinforcement Learning training process (depending on if ReinforcementLearning_multiproc is selected or not). Recommended to leave this untouched, by default, this value is set to the total number of physical cores minus 1. Datatype: int. model_reward_parameters Parameters used inside the customizable calculate_reward() function in ReinforcementLearner.py Datatype: int. add_state_info Tell FreqAI to include state information in the feature set for training and inferencing. The current state variables include trade duration, current profit, trade position. This is only available in dry/live runs, and is automatically switched to false for backtesting. Datatype: bool. Default: False. net_arch Network architecture which is well described in stable_baselines3 doc. In summary: [<shared layers>, dict(vf=[<non-shared value network layers>], pi=[<non-shared policy network layers>])]. By default this is set to [128, 128], which defines 2 shared hidden layers with 128 units each. randomize_starting_position Randomize the starting point of each episode to avoid overfitting. Datatype: bool. Default: False. drop_ohlc_from_features Do not include the normalized ohlc data in the feature set passed to the agent during training (ohlc will still be used for driving the environment in all cases) Datatype: Boolean. Default: False progress_bar Display a progress bar with the current progress, elapsed time and estimated remaining time. Datatype: Boolean. Default: False."},{"location":"freqai-parameter-table/#pytorch-parameters","title":"PyTorch parameters","text":""},{"location":"freqai-parameter-table/#general","title":"general","text":"Parameter Description Model training parameters within the freqai.model_training_parameters sub dictionary learning_rate Learning rate to be passed to the optimizer. Datatype: float. Default: 3e-4. model_kwargs Parameters to be passed to the model class. Datatype: dict. Default: {}. trainer_kwargs Parameters to be passed to the trainer class. Datatype: dict. Default: {}."},{"location":"freqai-parameter-table/#trainer_kwargs","title":"trainer_kwargs","text":"Parameter Description Model training parameters within the freqai.model_training_parameters.model_kwargs sub dictionary max_iters The number of training iterations to run. iteration here refers to the number of times we call self.optimizer.step(). used to calculate n_epochs. Datatype: int. Default: 100. batch_size The size of the batches to use during training.. Datatype: int. Default: 64. max_n_eval_batches The maximum number batches to use for evaluation.. Datatype: int, optional. Default: None."},{"location":"freqai-parameter-table/#additional-parameters","title":"Additional parameters","text":"Parameter Description Extraneous parameters freqai.keras If the selected model makes use of Keras (typical for TensorFlow-based prediction models), this flag needs to be activated so that the model save/loading follows Keras standards. Datatype: Boolean. Default: False. freqai.conv_width The width of a convolutional neural network input tensor. This replaces the need for shifting candles (include_shifted_candles) by feeding in historical data points as the second dimension of the tensor. Technically, this parameter can also be used for regressors, but it only adds computational overhead and does not change the model training/prediction. Datatype: Integer. Default: 2. freqai.reduce_df_footprint Recast all numeric columns to float32/int32, with the objective of reducing ram/disk usage and decreasing train/inference timing. This parameter is set in the main level of the Freqtrade configuration file (not inside FreqAI). Datatype: Boolean. Default: False."},{"location":"freqai-reinforcement-learning/","title":"Reinforcement Learning","text":"

Installation size

Reinforcement learning dependencies include large packages such as torch, which should be explicitly requested during ./setup.sh -i by answering \"y\" to the question \"Do you also want dependencies for freqai-rl (~700mb additional space required) [y/N]?\". Users who prefer docker should ensure they use the docker image appended with _freqairl.

"},{"location":"freqai-reinforcement-learning/#background-and-terminology","title":"Background and terminology","text":""},{"location":"freqai-reinforcement-learning/#what-is-rl-and-why-does-freqai-need-it","title":"What is RL and why does FreqAI need it?","text":"

Reinforcement learning involves two important components, the agent and the training environment. During agent training, the agent moves through historical data candle by candle, always making 1 of a set of actions: Long entry, long exit, short entry, short exit, neutral). During this training process, the environment tracks the performance of these actions and rewards the agent according to a custom user made calculate_reward() (here we offer a default reward for users to build on if they wish details here). The reward is used to train weights in a neural network.

A second important component of the FreqAI RL implementation is the use of state information. State information is fed into the network at each step, including current profit, current position, and current trade duration. These are used to train the agent in the training environment, and to reinforce the agent in dry/live (this functionality is not available in backtesting). FreqAI + Freqtrade is a perfect match for this reinforcing mechanism since this information is readily available in live deployments.

Reinforcement learning is a natural progression for FreqAI, since it adds a new layer of adaptivity and market reactivity that Classifiers and Regressors cannot match. However, Classifiers and Regressors have strengths that RL does not have such as robust predictions. Improperly trained RL agents may find \"cheats\" and \"tricks\" to maximize reward without actually winning any trades. For this reason, RL is more complex and demands a higher level of understanding than typical Classifiers and Regressors.

"},{"location":"freqai-reinforcement-learning/#the-rl-interface","title":"The RL interface","text":"

With the current framework, we aim to expose the training environment via the common \"prediction model\" file, which is a user inherited BaseReinforcementLearner object (e.g. freqai/prediction_models/ReinforcementLearner). Inside this user class, the RL environment is available and customized via MyRLEnv as shown below.

We envision the majority of users focusing their effort on creative design of the calculate_reward() function details here, while leaving the rest of the environment untouched. Other users may not touch the environment at all, and they will only play with the configuration settings and the powerful feature engineering that already exists in FreqAI. Meanwhile, we enable advanced users to create their own model classes entirely.

The framework is built on stable_baselines3 (torch) and OpenAI gym for the base environment class. But generally speaking, the model class is well isolated. Thus, the addition of competing libraries can be easily integrated into the existing framework. For the environment, it is inheriting from gym.env which means that it is necessary to write an entirely new environment in order to switch to a different library.

"},{"location":"freqai-reinforcement-learning/#important-considerations","title":"Important considerations","text":"

As explained above, the agent is \"trained\" in an artificial trading \"environment\". In our case, that environment may seem quite similar to a real Freqtrade backtesting environment, but it is NOT. In fact, the RL training environment is much more simplified. It does not incorporate any of the complicated strategy logic, such as callbacks like custom_exit, custom_stoploss, leverage controls, etc. The RL environment is instead a very \"raw\" representation of the true market, where the agent has free will to learn the policy (read: stoploss, take profit, etc.) which is enforced by the calculate_reward(). Thus, it is important to consider that the agent training environment is not identical to the real world.

"},{"location":"freqai-reinforcement-learning/#running-reinforcement-learning","title":"Running Reinforcement Learning","text":"

Setting up and running a Reinforcement Learning model is the same as running a Regressor or Classifier. The same two flags, --freqaimodel and --strategy, must be defined on the command line:

freqtrade trade --freqaimodel ReinforcementLearner --strategy MyRLStrategy --config config.json\n

where ReinforcementLearner will use the templated ReinforcementLearner from freqai/prediction_models/ReinforcementLearner (or a custom user defined one located in user_data/freqaimodels). The strategy, on the other hand, follows the same base feature engineering with feature_engineering_* as a typical Regressor. The difference lies in the creation of the targets, Reinforcement Learning doesn't require them. However, FreqAI requires a default (neutral) value to be set in the action column:

    def set_freqai_targets(self, dataframe, **kwargs) -> DataFrame:\n\"\"\"\n        *Only functional with FreqAI enabled strategies*\n        Required function to set the targets for the model.\n        All targets must be prepended with `&` to be recognized by the FreqAI internals.\n\n        More details about feature engineering available:\n\n        https://www.freqtrade.io/en/latest/freqai-feature-engineering\n\n        :param df: strategy dataframe which will receive the targets\n        usage example: dataframe[\"&-target\"] = dataframe[\"close\"].shift(-1) / dataframe[\"close\"]\n        \"\"\"\n        # For RL, there are no direct targets to set. This is filler (neutral)\n        # until the agent sends an action.\n        dataframe[\"&-action\"] = 0\n        return dataframe\n

Most of the function remains the same as for typical Regressors, however, the function below shows how the strategy must pass the raw price data to the agent so that it has access to raw OHLCV in the training environment:

    def feature_engineering_standard(self, dataframe: DataFrame, **kwargs) -> DataFrame:\n        # The following features are necessary for RL models\n        dataframe[f\"%-raw_close\"] = dataframe[\"close\"]\n        dataframe[f\"%-raw_open\"] = dataframe[\"open\"]\n        dataframe[f\"%-raw_high\"] = dataframe[\"high\"]\n        dataframe[f\"%-raw_low\"] = dataframe[\"low\"]\n    return dataframe\n

Finally, there is no explicit \"label\" to make - instead it is necessary to assign the &-action column which will contain the agent's actions when accessed in populate_entry/exit_trends(). In the present example, the neutral action to 0. This value should align with the environment used. FreqAI provides two environments, both use 0 as the neutral action.

After users realize there are no labels to set, they will soon understand that the agent is making its \"own\" entry and exit decisions. This makes strategy construction rather simple. The entry and exit signals come from the agent in the form of an integer - which are used directly to decide entries and exits in the strategy:

    def populate_entry_trend(self, df: DataFrame, metadata: dict) -> DataFrame:\n\n        enter_long_conditions = [df[\"do_predict\"] == 1, df[\"&-action\"] == 1]\n\n        if enter_long_conditions:\n            df.loc[\n                reduce(lambda x, y: x & y, enter_long_conditions), [\"enter_long\", \"enter_tag\"]\n            ] = (1, \"long\")\n\n        enter_short_conditions = [df[\"do_predict\"] == 1, df[\"&-action\"] == 3]\n\n        if enter_short_conditions:\n            df.loc[\n                reduce(lambda x, y: x & y, enter_short_conditions), [\"enter_short\", \"enter_tag\"]\n            ] = (1, \"short\")\n\n        return df\n\n    def populate_exit_trend(self, df: DataFrame, metadata: dict) -> DataFrame:\n        exit_long_conditions = [df[\"do_predict\"] == 1, df[\"&-action\"] == 2]\n        if exit_long_conditions:\n            df.loc[reduce(lambda x, y: x & y, exit_long_conditions), \"exit_long\"] = 1\n\n        exit_short_conditions = [df[\"do_predict\"] == 1, df[\"&-action\"] == 4]\n        if exit_short_conditions:\n            df.loc[reduce(lambda x, y: x & y, exit_short_conditions), \"exit_short\"] = 1\n\n        return df\n

It is important to consider that &-action depends on which environment they choose to use. The example above shows 5 actions, where 0 is neutral, 1 is enter long, 2 is exit long, 3 is enter short and 4 is exit short.

"},{"location":"freqai-reinforcement-learning/#configuring-the-reinforcement-learner","title":"Configuring the Reinforcement Learner","text":"

In order to configure the Reinforcement Learner the following dictionary must exist in the freqai config:

        \"rl_config\": {\n\"train_cycles\": 25,\n\"add_state_info\": true,\n\"max_trade_duration_candles\": 300,\n\"max_training_drawdown_pct\": 0.02,\n\"cpu_count\": 8,\n\"model_type\": \"PPO\",\n\"policy_type\": \"MlpPolicy\",\n\"model_reward_parameters\": {\n\"rr\": 1,\n\"profit_aim\": 0.025\n}\n}\n

Parameter details can be found here, but in general the train_cycles decides how many times the agent should cycle through the candle data in its artificial environment to train weights in the model. model_type is a string which selects one of the available models in stable_baselines(external link).

Note

If you would like to experiment with continual_learning, then you should set that value to true in the main freqai configuration dictionary. This will tell the Reinforcement Learning library to continue training new models from the final state of previous models, instead of retraining new models from scratch each time a retrain is initiated.

Note

Remember that the general model_training_parameters dictionary should contain all the model hyperparameter customizations for the particular model_type. For example, PPO parameters can be found here.

"},{"location":"freqai-reinforcement-learning/#creating-a-custom-reward-function","title":"Creating a custom reward function","text":"

As you begin to modify the strategy and the prediction model, you will quickly realize some important differences between the Reinforcement Learner and the Regressors/Classifiers. Firstly, the strategy does not set a target value (no labels!). Instead, you set the calculate_reward() function inside the MyRLEnv class (see below). A default calculate_reward() is provided inside prediction_models/ReinforcementLearner.py to demonstrate the necessary building blocks for creating rewards, but users are encouraged to create their own custom reinforcement learning model class (see below) and save it to user_data/freqaimodels. It is inside the calculate_reward() where creative theories about the market can be expressed. For example, you can reward your agent when it makes a winning trade, and penalize the agent when it makes a losing trade. Or perhaps, you wish to reward the agent for entering trades, and penalize the agent for sitting in trades too long. Below we show examples of how these rewards are all calculated:

    from freqtrade.freqai.prediction_models.ReinforcementLearner import ReinforcementLearner\n    from freqtrade.freqai.RL.Base5ActionRLEnv import Actions, Base5ActionRLEnv, Positions\n\n\n    class MyCoolRLModel(ReinforcementLearner):\n\"\"\"\n        User created RL prediction model.\n\n        Save this file to `freqtrade/user_data/freqaimodels`\n\n        then use it with:\n\n        freqtrade trade --freqaimodel MyCoolRLModel --config config.json --strategy SomeCoolStrat\n\n        Here the users can override any of the functions\n        available in the `IFreqaiModel` inheritance tree. Most importantly for RL, this\n        is where the user overrides `MyRLEnv` (see below), to define custom\n        `calculate_reward()` function, or to override any other parts of the environment.\n\n        This class also allows users to override any other part of the IFreqaiModel tree.\n        For example, the user can override `def fit()` or `def train()` or `def predict()`\n        to take fine-tuned control over these processes.\n\n        Another common override may be `def data_cleaning_predict()` where the user can\n        take fine-tuned control over the data handling pipeline.\n        \"\"\"\n        class MyRLEnv(Base5ActionRLEnv):\n\"\"\"\n            User made custom environment. This class inherits from BaseEnvironment and gym.env.\n            Users can override any functions from those parent classes. Here is an example\n            of a user customized `calculate_reward()` function.\n            \"\"\"\n            def calculate_reward(self, action: int) -> float:\n                # first, penalize if the action is not valid\n                if not self._is_valid(action):\n                    return -2\n                pnl = self.get_unrealized_profit()\n\n                factor = 100\n\n                pair = self.pair.replace(':', '')\n\n                # you can use feature values from dataframe\n                # Assumes the shifted RSI indicator has been generated in the strategy.\n                rsi_now = self.raw_features[f\"%-rsi-period_10_shift-1_{pair}_\"\n                                f\"{self.config['timeframe']}\"].iloc[self._current_tick]\n\n                # reward agent for entering trades\n                if (action in (Actions.Long_enter.value, Actions.Short_enter.value)\n                        and self._position == Positions.Neutral):\n                    if rsi_now < 40:\n                        factor = 40 / rsi_now\n                    else:\n                        factor = 1\n                    return 25 * factor\n\n                # discourage agent from not entering trades\n                if action == Actions.Neutral.value and self._position == Positions.Neutral:\n                    return -1\n                max_trade_duration = self.rl_config.get('max_trade_duration_candles', 300)\n                trade_duration = self._current_tick - self._last_trade_tick\n                if trade_duration <= max_trade_duration:\n                    factor *= 1.5\n                elif trade_duration > max_trade_duration:\n                    factor *= 0.5\n                # discourage sitting in position\n                if self._position in (Positions.Short, Positions.Long) and \\\n                action == Actions.Neutral.value:\n                    return -1 * trade_duration / max_trade_duration\n                # close long\n                if action == Actions.Long_exit.value and self._position == Positions.Long:\n                    if pnl > self.profit_aim * self.rr:\n                        factor *= self.rl_config['model_reward_parameters'].get('win_reward_factor', 2)\n                    return float(pnl * factor)\n                # close short\n                if action == Actions.Short_exit.value and self._position == Positions.Short:\n                    if pnl > self.profit_aim * self.rr:\n                        factor *= self.rl_config['model_reward_parameters'].get('win_reward_factor', 2)\n                    return float(pnl * factor)\n                return 0.\n
"},{"location":"freqai-reinforcement-learning/#using-tensorboard","title":"Using Tensorboard","text":"

Reinforcement Learning models benefit from tracking training metrics. FreqAI has integrated Tensorboard to allow users to track training and evaluation performance across all coins and across all retrainings. Tensorboard is activated via the following command:

cd freqtrade\ntensorboard --logdir user_data/models/unique-id\n

where unique-id is the identifier set in the freqai configuration file. This command must be run in a separate shell to view the output in their browser at 127.0.0.1:6006 (6006 is the default port used by Tensorboard).

"},{"location":"freqai-reinforcement-learning/#custom-logging","title":"Custom logging","text":"

FreqAI also provides a built in episodic summary logger called self.tensorboard_log for adding custom information to the Tensorboard log. By default, this function is already called once per step inside the environment to record the agent actions. All values accumulated for all steps in a single episode are reported at the conclusion of each episode, followed by a full reset of all metrics to 0 in preparation for the subsequent episode.

self.tensorboard_log can also be used anywhere inside the environment, for example, it can be added to the calculate_reward function to collect more detailed information about how often various parts of the reward were called:

        class MyRLEnv(Base5ActionRLEnv):\n\"\"\"\n            User made custom environment. This class inherits from BaseEnvironment and gym.env.\n            Users can override any functions from those parent classes. Here is an example\n            of a user customized `calculate_reward()` function.\n            \"\"\"\n            def calculate_reward(self, action: int) -> float:\n                if not self._is_valid(action):\n                    self.tensorboard_log(\"invalid\")\n                    return -2\n

Note

The self.tensorboard_log() function is designed for tracking incremented objects only i.e. events, actions inside the training environment. If the event of interest is a float, the float can be passed as the second argument e.g. self.tensorboard_log(\"float_metric1\", 0.23). In this case the metric values are not incremented.

"},{"location":"freqai-reinforcement-learning/#choosing-a-base-environment","title":"Choosing a base environment","text":"

FreqAI provides three base environments, Base3ActionRLEnvironment, Base4ActionEnvironment and Base5ActionEnvironment. As the names imply, the environments are customized for agents that can select from 3, 4 or 5 actions. The Base3ActionEnvironment is the simplest, the agent can select from hold, long, or short. This environment can also be used for long-only bots (it automatically follows the can_short flag from the strategy), where long is the enter condition and short is the exit condition. Meanwhile, in the Base4ActionEnvironment, the agent can enter long, enter short, hold neutral, or exit position. Finally, in the Base5ActionEnvironment, the agent has the same actions as Base4, but instead of a single exit action, it separates exit long and exit short. The main changes stemming from the environment selection include:

  • the actions available in the calculate_reward
  • the actions consumed by the user strategy

All of the FreqAI provided environments inherit from an action/position agnostic environment object called the BaseEnvironment, which contains all shared logic. The architecture is designed to be easily customized. The simplest customization is the calculate_reward() (see details here). However, the customizations can be further extended into any of the functions inside the environment. You can do this by simply overriding those functions inside your MyRLEnv in the prediction model file. Or for more advanced customizations, it is encouraged to create an entirely new environment inherited from BaseEnvironment.

Note

Only the Base3ActionRLEnv can do long-only training/trading (set the user strategy attribute can_short = False).

"},{"location":"freqai-running/","title":"Running FreqAI","text":"

There are two ways to train and deploy an adaptive machine learning model - live deployment and historical backtesting. In both cases, FreqAI runs/simulates periodic retraining of models as shown in the following figure:

"},{"location":"freqai-running/#live-deployments","title":"Live deployments","text":"

FreqAI can be run dry/live using the following command:

freqtrade trade --strategy FreqaiExampleStrategy --config config_freqai.example.json --freqaimodel LightGBMRegressor\n

When launched, FreqAI will start training a new model, with a new identifier, based on the config settings. Following training, the model will be used to make predictions on incoming candles until a new model is available. New models are typically generated as often as possible, with FreqAI managing an internal queue of the coin pairs to try to keep all models equally up to date. FreqAI will always use the most recently trained model to make predictions on incoming live data. If you do not want FreqAI to retrain new models as often as possible, you can set live_retrain_hours to tell FreqAI to wait at least that number of hours before training a new model. Additionally, you can set expired_hours to tell FreqAI to avoid making predictions on models that are older than that number of hours.

Trained models are by default saved to disk to allow for reuse during backtesting or after a crash. You can opt to purge old models to save disk space by setting \"purge_old_models\": true in the config.

To start a dry/live run from a saved backtest model (or from a previously crashed dry/live session), you only need to specify the identifier of the specific model:

    \"freqai\": {\n\"identifier\": \"example\",\n\"live_retrain_hours\": 0.5\n}\n

In this case, although FreqAI will initiate with a pre-trained model, it will still check to see how much time has elapsed since the model was trained. If a full live_retrain_hours has elapsed since the end of the loaded model, FreqAI will start training a new model.

"},{"location":"freqai-running/#automatic-data-download","title":"Automatic data download","text":"

FreqAI automatically downloads the proper amount of data needed to ensure training of a model through the defined train_period_days and startup_candle_count (see the parameter table for detailed descriptions of these parameters).

"},{"location":"freqai-running/#saving-prediction-data","title":"Saving prediction data","text":"

All predictions made during the lifetime of a specific identifier model are stored in historic_predictions.pkl to allow for reloading after a crash or changes made to the config.

"},{"location":"freqai-running/#purging-old-model-data","title":"Purging old model data","text":"

FreqAI stores new model files after each successful training. These files become obsolete as new models are generated to adapt to new market conditions. If you are planning to leave FreqAI running for extended periods of time with high frequency retraining, you should enable purge_old_models in the config:

    \"freqai\": {\n\"purge_old_models\": true,\n}\n

This will automatically purge all models older than the two most recently trained ones to save disk space.

"},{"location":"freqai-running/#backtesting","title":"Backtesting","text":"

The FreqAI backtesting module can be executed with the following command:

freqtrade backtesting --strategy FreqaiExampleStrategy --strategy-path freqtrade/templates --config config_examples/config_freqai.example.json --freqaimodel LightGBMRegressor --timerange 20210501-20210701\n

If this command has never been executed with the existing config file, FreqAI will train a new model for each pair, for each backtesting window within the expanded --timerange.

Backtesting mode requires downloading the necessary data before deployment (unlike in dry/live mode where FreqAI handles the data downloading automatically). You should be careful to consider that the time range of the downloaded data is more than the backtesting time range. This is because FreqAI needs data prior to the desired backtesting time range in order to train a model to be ready to make predictions on the first candle of the set backtesting time range. More details on how to calculate the data to download can be found here.

Model reuse

Once the training is completed, you can execute the backtesting again with the same config file and FreqAI will find the trained models and load them instead of spending time training. This is useful if you want to tweak (or even hyperopt) buy and sell criteria inside the strategy. If you want to retrain a new model with the same config file, you should simply change the identifier. This way, you can return to using any model you wish by simply specifying the identifier.

Note

Backtesting calls set_freqai_targets() one time for each backtest window (where the number of windows is the full backtest timerange divided by the backtest_period_days parameter). Doing this means that the targets simulate dry/live behavior without look ahead bias. However, the definition of the features in feature_engineering_*() is performed once on the entire backtest timerange. This means that you should be sure that features do look-ahead into the future. More details about look-ahead bias can be found in Common Mistakes.

"},{"location":"freqai-running/#saving-prediction-data_1","title":"Saving prediction data","text":"

To allow for tweaking your strategy (not the features!), FreqAI will automatically save the predictions during backtesting so that they can be reused for future backtests and live runs using the same identifier model. This provides a performance enhancement geared towards enabling high-level hyperopting of entry/exit criteria.

An additional directory called backtesting_predictions, which contains all the predictions stored in hdf format, will be created in the unique-id folder.

To change your features, you must set a new identifier in the config to signal to FreqAI to train new models.

To save the models generated during a particular backtest so that you can start a live deployment from one of them instead of training a new model, you must set save_backtest_models to True in the config.

"},{"location":"freqai-running/#backtest-live-collected-predictions","title":"Backtest live collected predictions","text":"

FreqAI allow you to reuse live historic predictions through the backtest parameter --freqai-backtest-live-models. This can be useful when you want to reuse predictions generated in dry/run for comparison or other study.

The --timerange parameter must not be informed, as it will be automatically calculated through the data in the historic predictions file.

"},{"location":"freqai-running/#downloading-data-to-cover-the-full-backtest-period","title":"Downloading data to cover the full backtest period","text":"

For live/dry deployments, FreqAI will download the necessary data automatically. However, to use backtesting functionality, you need to download the necessary data using download-data (details here). You need to pay careful attention to understanding how much additional data needs to be downloaded to ensure that there is a sufficient amount of training data before the start of the backtesting time range. The amount of additional data can be roughly estimated by moving the start date of the time range backwards by train_period_days and the startup_candle_count (see the parameter table for detailed descriptions of these parameters) from the beginning of the desired backtesting time range.

As an example, to backtest the --timerange 20210501-20210701 using the example config which sets train_period_days to 30, together with startup_candle_count: 40 on a maximum include_timeframes of 1h, the start date for the downloaded data needs to be 20210501 - 30 days - 40 * 1h / 24 hours = 20210330 (31.7 days earlier than the start of the desired training time range).

"},{"location":"freqai-running/#deciding-the-size-of-the-sliding-training-window-and-backtesting-duration","title":"Deciding the size of the sliding training window and backtesting duration","text":"

The backtesting time range is defined with the typical --timerange parameter in the configuration file. The duration of the sliding training window is set by train_period_days, whilst backtest_period_days is the sliding backtesting window, both in number of days (backtest_period_days can be a float to indicate sub-daily retraining in live/dry mode). In the presented example config (found in config_examples/config_freqai.example.json), the user is asking FreqAI to use a training period of 30 days and backtest on the subsequent 7 days. After the training of the model, FreqAI will backtest the subsequent 7 days. The \"sliding window\" then moves one week forward (emulating FreqAI retraining once per week in live mode) and the new model uses the previous 30 days (including the 7 days used for backtesting by the previous model) to train. This is repeated until the end of --timerange. This means that if you set --timerange 20210501-20210701, FreqAI will have trained 8 separate models at the end of --timerange (because the full range comprises 8 weeks).

Note

Although fractional backtest_period_days is allowed, you should be aware that the --timerange is divided by this value to determine the number of models that FreqAI will need to train in order to backtest the full range. For example, by setting a --timerange of 10 days, and a backtest_period_days of 0.1, FreqAI will need to train 100 models per pair to complete the full backtest. Because of this, a true backtest of FreqAI adaptive training would take a very long time. The best way to fully test a model is to run it dry and let it train constantly. In this case, backtesting would take the exact same amount of time as a dry run.

"},{"location":"freqai-running/#defining-model-expirations","title":"Defining model expirations","text":"

During dry/live mode, FreqAI trains each coin pair sequentially (on separate threads/GPU from the main Freqtrade bot). This means that there is always an age discrepancy between models. If you are training on 50 pairs, and each pair requires 5 minutes to train, the oldest model will be over 4 hours old. This may be undesirable if the characteristic time scale (the trade duration target) for a strategy is less than 4 hours. You can decide to only make trade entries if the model is less than a certain number of hours old by setting the expiration_hours in the config file:

    \"freqai\": {\n\"expiration_hours\": 0.5,\n}\n

In the presented example config, the user will only allow predictions on models that are less than \u00bd hours old.

"},{"location":"freqai-running/#controlling-the-model-learning-process","title":"Controlling the model learning process","text":"

Model training parameters are unique to the selected machine learning library. FreqAI allows you to set any parameter for any library using the model_training_parameters dictionary in the config. The example config (found in config_examples/config_freqai.example.json) shows some of the example parameters associated with Catboost and LightGBM, but you can add any parameters available in those libraries or any other machine learning library you choose to implement.

Data split parameters are defined in data_split_parameters which can be any parameters associated with scikit-learn's train_test_split() function. train_test_split() has a parameters called shuffle which allows to shuffle the data or keep it unshuffled. This is particularly useful to avoid biasing training with temporally auto-correlated data. More details about these parameters can be found the scikit-learn website (external website).

The FreqAI specific parameter label_period_candles defines the offset (number of candles into the future) used for the labels. In the presented example config, the user is asking for labels that are 24 candles in the future.

"},{"location":"freqai-running/#continual-learning","title":"Continual learning","text":"

You can choose to adopt a continual learning scheme by setting \"continual_learning\": true in the config. By enabling continual_learning, after training an initial model from scratch, subsequent trainings will start from the final model state of the preceding training. This gives the new model a \"memory\" of the previous state. By default, this is set to False which means that all new models are trained from scratch, without input from previous models.

Continual learning enforces a constant parameter space

Since continual_learning means that the model parameter space cannot change between trainings, principal_component_analysis is automatically disabled when continual_learning is enabled. Hint: PCA changes the parameter space and the number of features, learn more about PCA here.

"},{"location":"freqai-running/#hyperopt","title":"Hyperopt","text":"

You can hyperopt using the same command as for typical Freqtrade hyperopt:

freqtrade hyperopt --hyperopt-loss SharpeHyperOptLoss --strategy FreqaiExampleStrategy --freqaimodel LightGBMRegressor --strategy-path freqtrade/templates --config config_examples/config_freqai.example.json --timerange 20220428-20220507\n

hyperopt requires you to have the data pre-downloaded in the same fashion as if you were doing backtesting. In addition, you must consider some restrictions when trying to hyperopt FreqAI strategies:

  • The --analyze-per-epoch hyperopt parameter is not compatible with FreqAI.
  • It's not possible to hyperopt indicators in the feature_engineering_*() and set_freqai_targets() functions. This means that you cannot optimize model parameters using hyperopt. Apart from this exception, it is possible to optimize all other spaces.
  • The backtesting instructions also apply to hyperopt.

The best method for combining hyperopt and FreqAI is to focus on hyperopting entry/exit thresholds/criteria. You need to focus on hyperopting parameters that are not used in your features. For example, you should not try to hyperopt rolling window lengths in the feature creation, or any part of the FreqAI config which changes predictions. In order to efficiently hyperopt the FreqAI strategy, FreqAI stores predictions as dataframes and reuses them. Hence the requirement to hyperopt entry/exit thresholds/criteria only.

A good example of a hyperoptable parameter in FreqAI is a threshold for the Dissimilarity Index (DI) DI_values beyond which we consider data points as outliers:

di_max = IntParameter(low=1, high=20, default=10, space='buy', optimize=True, load=True)\ndataframe['outlier'] = np.where(dataframe['DI_values'] > self.di_max.value/10, 1, 0)\n

This specific hyperopt would help you understand the appropriate DI_values for your particular parameter space.

"},{"location":"freqai-running/#using-tensorboard","title":"Using Tensorboard","text":"

CatBoost models benefit from tracking training metrics via Tensorboard. You can take advantage of the FreqAI integration to track training and evaluation performance across all coins and across all retrainings. Tensorboard is activated via the following command:

cd freqtrade\ntensorboard --logdir user_data/models/unique-id\n

where unique-id is the identifier set in the freqai configuration file. This command must be run in a separate shell if you wish to view the output in your browser at 127.0.0.1:6060 (6060 is the default port used by Tensorboard).

"},{"location":"freqai/","title":"Introduction","text":""},{"location":"freqai/#freqai","title":"FreqAI","text":""},{"location":"freqai/#introduction","title":"Introduction","text":"

FreqAI is a software designed to automate a variety of tasks associated with training a predictive machine learning model to generate market forecasts given a set of input signals. In general, FreqAI aims to be a sandbox for easily deploying robust machine learning libraries on real-time data (details).

Note

FreqAI is, and always will be, a not-for-profit, open-source project. FreqAI does not have a crypto token, FreqAI does not sell signals, and FreqAI does not have a domain besides the present freqtrade documentation.

Features include:

  • Self-adaptive retraining - Retrain models during live deployments to self-adapt to the market in a supervised manner
  • Rapid feature engineering - Create large rich feature sets (10k+ features) based on simple user-created strategies
  • High performance - Threading allows for adaptive model retraining on a separate thread (or on GPU if available) from model inferencing (prediction) and bot trade operations. Newest models and data are kept in RAM for rapid inferencing
  • Realistic backtesting - Emulate self-adaptive training on historic data with a backtesting module that automates retraining
  • Extensibility - The generalized and robust architecture allows for incorporating any machine learning library/method available in Python. Eight examples are currently available, including classifiers, regressors, and a convolutional neural network
  • Smart outlier removal - Remove outliers from training and prediction data sets using a variety of outlier detection techniques
  • Crash resilience - Store trained models to disk to make reloading from a crash fast and easy, and purge obsolete files for sustained dry/live runs
  • Automatic data normalization - Normalize the data in a smart and statistically safe way
  • Automatic data download - Compute timeranges for data downloads and update historic data (in live deployments)
  • Cleaning of incoming data - Handle NaNs safely before training and model inferencing
  • Dimensionality reduction - Reduce the size of the training data via Principal Component Analysis
  • Deploying bot fleets - Set one bot to train models while a fleet of consumers use signals.
"},{"location":"freqai/#quick-start","title":"Quick start","text":"

The easiest way to quickly test FreqAI is to run it in dry mode with the following command:

freqtrade trade --config config_examples/config_freqai.example.json --strategy FreqaiExampleStrategy --freqaimodel LightGBMRegressor --strategy-path freqtrade/templates\n

You will see the boot-up process of automatic data downloading, followed by simultaneous training and trading.

An example strategy, prediction model, and config to use as a starting points can be found in freqtrade/templates/FreqaiExampleStrategy.py, freqtrade/freqai/prediction_models/LightGBMRegressor.py, and config_examples/config_freqai.example.json, respectively.

"},{"location":"freqai/#general-approach","title":"General approach","text":"

You provide FreqAI with a set of custom base indicators (the same way as in a typical Freqtrade strategy) as well as target values (labels). For each pair in the whitelist, FreqAI trains a model to predict the target values based on the input of custom indicators. The models are then consistently retrained, with a predetermined frequency, to adapt to market conditions. FreqAI offers the ability to both backtest strategies (emulating reality with periodic retraining on historic data) and deploy dry/live runs. In dry/live conditions, FreqAI can be set to constant retraining in a background thread to keep models as up to date as possible.

An overview of the algorithm, explaining the data processing pipeline and model usage, is shown below.

"},{"location":"freqai/#important-machine-learning-vocabulary","title":"Important machine learning vocabulary","text":"

Features - the parameters, based on historic data, on which a model is trained. All features for a single candle are stored as a vector. In FreqAI, you build a feature data set from anything you can construct in the strategy.

Labels - the target values that the model is trained toward. Each feature vector is associated with a single label that is defined by you within the strategy. These labels intentionally look into the future and are what you are training the model to be able to predict.

Training - the process of \"teaching\" the model to match the feature sets to the associated labels. Different types of models \"learn\" in different ways which means that one might be better than another for a specific application. More information about the different models that are already implemented in FreqAI can be found here.

Train data - a subset of the feature data set that is fed to the model during training to \"teach\" the model how to predict the targets. This data directly influences weight connections in the model.

Test data - a subset of the feature data set that is used to evaluate the performance of the model after training. This data does not influence nodal weights within the model.

Inferencing - the process of feeding a trained model new unseen data on which it will make a prediction.

"},{"location":"freqai/#install-prerequisites","title":"Install prerequisites","text":"

The normal Freqtrade install process will ask if you wish to install FreqAI dependencies. You should reply \"yes\" to this question if you wish to use FreqAI. If you did not reply yes, you can manually install these dependencies after the install with:

pip install -r requirements-freqai.txt\n

Note

Catboost will not be installed on arm devices (raspberry, Mac M1, ARM based VPS, ...), since it does not provide wheels for this platform.

python 3.11

Some dependencies (Catboost, Torch) currently don't support python 3.11. Freqtrade therefore only supports python 3.10 for these models/dependencies. Tests involving these dependencies are skipped on 3.11.

"},{"location":"freqai/#usage-with-docker","title":"Usage with docker","text":"

If you are using docker, a dedicated tag with FreqAI dependencies is available as :freqai. As such - you can replace the image line in your docker compose file with image: freqtradeorg/freqtrade:develop_freqai. This image contains the regular FreqAI dependencies. Similar to native installs, Catboost will not be available on ARM based devices.

"},{"location":"freqai/#freqai-position-in-open-source-machine-learning-landscape","title":"FreqAI position in open-source machine learning landscape","text":"

Forecasting chaotic time-series based systems, such as equity/cryptocurrency markets, requires a broad set of tools geared toward testing a wide range of hypotheses. Fortunately, a recent maturation of robust machine learning libraries (e.g. scikit-learn) has opened up a wide range of research possibilities. Scientists from a diverse range of fields can now easily prototype their studies on an abundance of established machine learning algorithms. Similarly, these user-friendly libraries enable \"citzen scientists\" to use their basic Python skills for data exploration. However, leveraging these machine learning libraries on historical and live chaotic data sources can be logistically difficult and expensive. Additionally, robust data collection, storage, and handling presents a disparate challenge. FreqAI aims to provide a generalized and extensible open-sourced framework geared toward live deployments of adaptive modeling for market forecasting. The FreqAI framework is effectively a sandbox for the rich world of open-source machine learning libraries. Inside the FreqAI sandbox, users find they can combine a wide variety of third-party libraries to test creative hypotheses on a free live 24/7 chaotic data source - cryptocurrency exchange data.

"},{"location":"freqai/#citing-freqai","title":"Citing FreqAI","text":"

FreqAI is published in the Journal of Open Source Software. If you find FreqAI useful in your research, please use the following citation:

@article{Caulk2022, doi = {10.21105/joss.04864},\nurl = {https://doi.org/10.21105/joss.04864},\nyear = {2022}, publisher = {The Open Journal},\nvolume = {7}, number = {80}, pages = {4864},\nauthor = {Robert A. Caulk and Elin T\u00f6rnquist and Matthias Voppichler and Andrew R. Lawless and Ryan McMullan and Wagner Costa Santos and Timothy C. Pogue and Johan van der Vlugt and Stefan P. Gehring and Pascal Schmidt},\ntitle = {FreqAI: generalizing adaptive modeling for chaotic time-series market forecasts},\njournal = {Journal of Open Source Software} } 
"},{"location":"freqai/#common-pitfalls","title":"Common pitfalls","text":"

FreqAI cannot be combined with dynamic VolumePairlists (or any pairlist filter that adds and removes pairs dynamically). This is for performance reasons - FreqAI relies on making quick predictions/retrains. To do this effectively, it needs to download all the training data at the beginning of a dry/live instance. FreqAI stores and appends new candles automatically for future retrains. This means that if new pairs arrive later in the dry run due to a volume pairlist, it will not have the data ready. However, FreqAI does work with the ShufflePairlist or a VolumePairlist which keeps the total pairlist constant (but reorders the pairs according to volume).

"},{"location":"freqai/#credits","title":"Credits","text":"

FreqAI is developed by a group of individuals who all contribute specific skillsets to the project.

Conception and software development: Robert Caulk @robcaulk

Theoretical brainstorming and data analysis: Elin T\u00f6rnquist @th0rntwig

Code review and software architecture brainstorming: @xmatthias

Software development: Wagner Costa @wagnercosta Emre Suzen @aemr3 Timothy Pogue @wizrds

Beta testing and bug reporting: Stefan Gehring @bloodhunter4rc, @longyu, Andrew Lawless @paranoidandy, Pascal Schmidt @smidelis, Ryan McMullan @smarmau, Juha Nyk\u00e4nen @suikula, Johan van der Vlugt @jooopiert, Rich\u00e1rd J\u00f3zsa @richardjosza

"},{"location":"hyperopt/","title":"Hyperopt","text":"

This page explains how to tune your strategy by finding the optimal parameters, a process called hyperparameter optimization. The bot uses algorithms included in the scikit-optimize package to accomplish this. The search will burn all your CPU cores, make your laptop sound like a fighter jet and still take a long time.

In general, the search for best parameters starts with a few random combinations (see below for more details) and then uses Bayesian search with a ML regressor algorithm (currently ExtraTreesRegressor) to quickly find a combination of parameters in the search hyperspace that minimizes the value of the loss function.

Hyperopt requires historic data to be available, just as backtesting does (hyperopt runs backtesting many times with different parameters). To learn how to get data for the pairs and exchange you're interested in, head over to the Data Downloading section of the documentation.

Bug

Hyperopt can crash when used with only 1 CPU Core as found out in Issue #1133

Note

Since 2021.4 release you no longer have to write a separate hyperopt class, but can configure the parameters directly in the strategy. The legacy method is still supported, but it is no longer the recommended way of setting up hyperopt. The legacy documentation is available at Legacy Hyperopt.

"},{"location":"hyperopt/#install-hyperopt-dependencies","title":"Install hyperopt dependencies","text":"

Since Hyperopt dependencies are not needed to run the bot itself, are heavy, can not be easily built on some platforms (like Raspberry PI), they are not installed by default. Before you run Hyperopt, you need to install the corresponding dependencies, as described in this section below.

Note

Since Hyperopt is a resource intensive process, running it on a Raspberry Pi is not recommended nor supported.

"},{"location":"hyperopt/#docker","title":"Docker","text":"

The docker-image includes hyperopt dependencies, no further action needed.

"},{"location":"hyperopt/#easy-installation-script-setupsh-manual-installation","title":"Easy installation script (setup.sh) / Manual installation","text":"
source .env/bin/activate\npip install -r requirements-hyperopt.txt\n
"},{"location":"hyperopt/#hyperopt-command-reference","title":"Hyperopt command reference","text":"
usage: freqtrade hyperopt [-h] [-v] [--logfile FILE] [-V] [-c PATH] [-d PATH]\n                          [--userdir PATH] [-s NAME] [--strategy-path PATH]\n                          [--recursive-strategy-search] [--freqaimodel NAME]\n                          [--freqaimodel-path PATH] [-i TIMEFRAME]\n                          [--timerange TIMERANGE]\n                          [--data-format-ohlcv {json,jsongz,hdf5}]\n                          [--max-open-trades INT]\n                          [--stake-amount STAKE_AMOUNT] [--fee FLOAT]\n                          [-p PAIRS [PAIRS ...]] [--hyperopt-path PATH]\n                          [--eps] [--dmmp] [--enable-protections]\n                          [--dry-run-wallet DRY_RUN_WALLET]\n                          [--timeframe-detail TIMEFRAME_DETAIL] [-e INT]\n                          [--spaces {all,buy,sell,roi,stoploss,trailing,protection,trades,default} [{all,buy,sell,roi,stoploss,trailing,protection,trades,default} ...]]\n                          [--print-all] [--no-color] [--print-json] [-j JOBS]\n                          [--random-state INT] [--min-trades INT]\n                          [--hyperopt-loss NAME] [--disable-param-export]\n                          [--ignore-missing-spaces] [--analyze-per-epoch]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  -i TIMEFRAME, --timeframe TIMEFRAME\n                        Specify timeframe (`1m`, `5m`, `30m`, `1h`, `1d`).\n  --timerange TIMERANGE\n                        Specify what timerange of data to use.\n  --data-format-ohlcv {json,jsongz,hdf5}\n                        Storage format for downloaded candle (OHLCV) data.\n                        (default: `json`).\n  --max-open-trades INT\n                        Override the value of the `max_open_trades`\n                        configuration setting.\n  --stake-amount STAKE_AMOUNT\n                        Override the value of the `stake_amount` configuration\n                        setting.\n  --fee FLOAT           Specify fee ratio. Will be applied twice (on trade\n                        entry and exit).\n  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]\n                        Limit command to these pairs. Pairs are space-\n                        separated.\n  --hyperopt-path PATH  Specify additional lookup path for Hyperopt Loss\n                        functions.\n  --eps, --enable-position-stacking\n                        Allow buying the same pair multiple times (position\n                        stacking).\n  --dmmp, --disable-max-market-positions\n                        Disable applying `max_open_trades` during backtest\n                        (same as setting `max_open_trades` to a very high\n                        number).\n  --enable-protections, --enableprotections\n                        Enable protections for backtesting.Will slow\n                        backtesting down by a considerable amount, but will\n                        include configured protections\n  --dry-run-wallet DRY_RUN_WALLET, --starting-balance DRY_RUN_WALLET\n                        Starting balance, used for backtesting / hyperopt and\n                        dry-runs.\n  --timeframe-detail TIMEFRAME_DETAIL\n                        Specify detail timeframe for backtesting (`1m`, `5m`,\n                        `30m`, `1h`, `1d`).\n  -e INT, --epochs INT  Specify number of epochs (default: 100).\n  --spaces {all,buy,sell,roi,stoploss,trailing,protection,trades,default} [{all,buy,sell,roi,stoploss,trailing,protection,trades,default} ...]\n                        Specify which parameters to hyperopt. Space-separated\n                        list.\n  --print-all           Print all results, not only the best ones.\n  --no-color            Disable colorization of hyperopt results. May be\n                        useful if you are redirecting output to a file.\n  --print-json          Print output in JSON format.\n  -j JOBS, --job-workers JOBS\n                        The number of concurrently running jobs for\n                        hyperoptimization (hyperopt worker processes). If -1\n                        (default), all CPUs are used, for -2, all CPUs but one\n                        are used, etc. If 1 is given, no parallel computing\n                        code is used at all.\n  --random-state INT    Set random state to some positive integer for\n                        reproducible hyperopt results.\n  --min-trades INT      Set minimal desired number of trades for evaluations\n                        in the hyperopt optimization path (default: 1).\n  --hyperopt-loss NAME, --hyperoptloss NAME\n                        Specify the class name of the hyperopt loss function\n                        class (IHyperOptLoss). Different functions can\n                        generate completely different results, since the\n                        target for optimization is different. Built-in\n                        Hyperopt-loss-functions are:\n                        ShortTradeDurHyperOptLoss, OnlyProfitHyperOptLoss,\n                        SharpeHyperOptLoss, SharpeHyperOptLossDaily,\n                        SortinoHyperOptLoss, SortinoHyperOptLossDaily,\n                        CalmarHyperOptLoss, MaxDrawDownHyperOptLoss,\n                        MaxDrawDownRelativeHyperOptLoss,\n                        ProfitDrawDownHyperOptLoss\n  --disable-param-export\n                        Disable automatic hyperopt parameter export.\n  --ignore-missing-spaces, --ignore-unparameterized-spaces\n                        Suppress errors for any requested Hyperopt spaces that\n                        do not contain any parameters.\n  --analyze-per-epoch   Run populate_indicators once per epoch.\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n\nStrategy arguments:\n  -s NAME, --strategy NAME\n                        Specify strategy class name which will be used by the\n                        bot.\n  --strategy-path PATH  Specify additional strategy lookup path.\n  --recursive-strategy-search\n                        Recursively search for a strategy in the strategies\n                        folder.\n  --freqaimodel NAME    Specify a custom freqaimodels.\n  --freqaimodel-path PATH\n                        Specify additional lookup path for freqaimodels.\n
"},{"location":"hyperopt/#hyperopt-checklist","title":"Hyperopt checklist","text":"

Checklist on all tasks / possibilities in hyperopt

Depending on the space you want to optimize, only some of the below are required:

  • define parameters with space='buy' - for entry signal optimization
  • define parameters with space='sell' - for exit signal optimization

Note

populate_indicators needs to create all indicators any of the spaces may use, otherwise hyperopt will not work.

Rarely you may also need to create a nested class named HyperOpt and implement

  • roi_space - for custom ROI optimization (if you need the ranges for the ROI parameters in the optimization hyperspace that differ from default)
  • generate_roi_table - for custom ROI optimization (if you need the ranges for the values in the ROI table that differ from default or the number of entries (steps) in the ROI table which differs from the default 4 steps)
  • stoploss_space - for custom stoploss optimization (if you need the range for the stoploss parameter in the optimization hyperspace that differs from default)
  • trailing_space - for custom trailing stop optimization (if you need the ranges for the trailing stop parameters in the optimization hyperspace that differ from default)
  • max_open_trades_space - for custom max_open_trades optimization (if you need the ranges for the max_open_trades parameter in the optimization hyperspace that differ from default)

Quickly optimize ROI, stoploss and trailing stoploss

You can quickly optimize the spaces roi, stoploss and trailing without changing anything in your strategy.

# Have a working strategy at hand.\nfreqtrade hyperopt --hyperopt-loss SharpeHyperOptLossDaily --spaces roi stoploss trailing --strategy MyWorkingStrategy --config config.json -e 100\n
"},{"location":"hyperopt/#hyperopt-execution-logic","title":"Hyperopt execution logic","text":"

Hyperopt will first load your data into memory and will then run populate_indicators() once per Pair to generate all indicators, unless --analyze-per-epoch is specified.

Hyperopt will then spawn into different processes (number of processors, or -j <n>), and run backtesting over and over again, changing the parameters that are part of the --spaces defined.

For every new set of parameters, freqtrade will run first populate_entry_trend() followed by populate_exit_trend(), and then run the regular backtesting process to simulate trades.

After backtesting, the results are passed into the loss function, which will evaluate if this result was better or worse than previous results. Based on the loss function result, hyperopt will determine the next set of parameters to try in the next round of backtesting.

"},{"location":"hyperopt/#configure-your-guards-and-triggers","title":"Configure your Guards and Triggers","text":"

There are two places you need to change in your strategy file to add a new buy hyperopt for testing:

  • Define the parameters at the class level hyperopt shall be optimizing.
  • Within populate_entry_trend() - use defined parameter values instead of raw constants.

There you have two different types of indicators: 1. guards and 2. triggers.

  1. Guards are conditions like \"never buy if ADX < 10\", or never buy if current price is over EMA10.
  2. Triggers are ones that actually trigger buy in specific moment, like \"buy when EMA5 crosses over EMA10\" or \"buy when close price touches lower Bollinger band\".

Guards and Triggers

Technically, there is no difference between Guards and Triggers. However, this guide will make this distinction to make it clear that signals should not be \"sticking\". Sticking signals are signals that are active for multiple candles. This can lead into entering a signal late (right before the signal disappears - which means that the chance of success is a lot lower than right at the beginning).

Hyper-optimization will, for each epoch round, pick one trigger and possibly multiple guards.

"},{"location":"hyperopt/#exit-signal-optimization","title":"Exit signal optimization","text":"

Similar to the entry-signal above, exit-signals can also be optimized. Place the corresponding settings into the following methods

  • Define the parameters at the class level hyperopt shall be optimizing, either naming them sell_*, or by explicitly defining space='sell'.
  • Within populate_exit_trend() - use defined parameter values instead of raw constants.

The configuration and rules are the same than for buy signals.

"},{"location":"hyperopt/#solving-a-mystery","title":"Solving a Mystery","text":"

Let's say you are curious: should you use MACD crossings or lower Bollinger Bands to trigger your long entries. And you also wonder should you use RSI or ADX to help with those decisions. If you decide to use RSI or ADX, which values should I use for them?

So let's use hyperparameter optimization to solve this mystery.

"},{"location":"hyperopt/#defining-indicators-to-be-used","title":"Defining indicators to be used","text":"

We start by calculating the indicators our strategy is going to use.

class MyAwesomeStrategy(IStrategy):\n\n    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n\"\"\"\n        Generate all indicators used by the strategy\n        \"\"\"\n        dataframe['adx'] = ta.ADX(dataframe)\n        dataframe['rsi'] = ta.RSI(dataframe)\n        macd = ta.MACD(dataframe)\n        dataframe['macd'] = macd['macd']\n        dataframe['macdsignal'] = macd['macdsignal']\n        dataframe['macdhist'] = macd['macdhist']\n\n        bollinger = ta.BBANDS(dataframe, timeperiod=20, nbdevup=2.0, nbdevdn=2.0)\n        dataframe['bb_lowerband'] = bollinger['lowerband']\n        dataframe['bb_middleband'] = bollinger['middleband']\n        dataframe['bb_upperband'] = bollinger['upperband']\n        return dataframe\n
"},{"location":"hyperopt/#hyperoptable-parameters","title":"Hyperoptable parameters","text":"

We continue to define hyperoptable parameters:

class MyAwesomeStrategy(IStrategy):\n    buy_adx = DecimalParameter(20, 40, decimals=1, default=30.1, space=\"buy\")\n    buy_rsi = IntParameter(20, 40, default=30, space=\"buy\")\n    buy_adx_enabled = BooleanParameter(default=True, space=\"buy\")\n    buy_rsi_enabled = CategoricalParameter([True, False], default=False, space=\"buy\")\n    buy_trigger = CategoricalParameter([\"bb_lower\", \"macd_cross_signal\"], default=\"bb_lower\", space=\"buy\")\n

The above definition says: I have five parameters I want to randomly combine to find the best combination. buy_rsi is an integer parameter, which will be tested between 20 and 40. This space has a size of 20. buy_adx is a decimal parameter, which will be evaluated between 20 and 40 with 1 decimal place (so values are 20.1, 20.2, ...). This space has a size of 200. Then we have three category variables. First two are either True or False. We use these to either enable or disable the ADX and RSI guards. The last one we call trigger and use it to decide which buy trigger we want to use.

Parameter space assignment

Parameters must either be assigned to a variable named buy_* or sell_* - or contain space='buy' | space='sell' to be assigned to a space correctly. If no parameter is available for a space, you'll receive the error that no space was found when running hyperopt. Parameters with unclear space (e.g. adx_period = IntParameter(4, 24, default=14) - no explicit nor implicit space) will not be detected and will therefore be ignored.

So let's write the buy strategy using these values:

    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n        conditions = []\n        # GUARDS AND TRENDS\n        if self.buy_adx_enabled.value:\n            conditions.append(dataframe['adx'] > self.buy_adx.value)\n        if self.buy_rsi_enabled.value:\n            conditions.append(dataframe['rsi'] < self.buy_rsi.value)\n\n        # TRIGGERS\n        if self.buy_trigger.value == 'bb_lower':\n            conditions.append(dataframe['close'] < dataframe['bb_lowerband'])\n        if self.buy_trigger.value == 'macd_cross_signal':\n            conditions.append(qtpylib.crossed_above(\n                dataframe['macd'], dataframe['macdsignal']\n            ))\n\n        # Check that volume is not 0\n        conditions.append(dataframe['volume'] > 0)\n\n        if conditions:\n            dataframe.loc[\n                reduce(lambda x, y: x & y, conditions),\n                'enter_long'] = 1\n\n        return dataframe\n

Hyperopt will now call populate_entry_trend() many times (epochs) with different value combinations. It will use the given historical data and simulate buys based on the buy signals generated with the above function. Based on the results, hyperopt will tell you which parameter combination produced the best results (based on the configured loss function).

Note

The above setup expects to find ADX, RSI and Bollinger Bands in the populated indicators. When you want to test an indicator that isn't used by the bot currently, remember to add it to the populate_indicators() method in your strategy or hyperopt file.

"},{"location":"hyperopt/#parameter-types","title":"Parameter types","text":"

There are four parameter types each suited for different purposes.

  • IntParameter - defines an integral parameter with upper and lower boundaries of search space.
  • DecimalParameter - defines a floating point parameter with a limited number of decimals (default 3). Should be preferred instead of RealParameter in most cases.
  • RealParameter - defines a floating point parameter with upper and lower boundaries and no precision limit. Rarely used as it creates a space with a near infinite number of possibilities.
  • CategoricalParameter - defines a parameter with a predetermined number of choices.
  • BooleanParameter - Shorthand for CategoricalParameter([True, False]) - great for \"enable\" parameters.

Disabling parameter optimization

Each parameter takes two boolean parameters: * load - when set to False it will not load values configured in buy_params and sell_params. * optimize - when set to False parameter will not be included in optimization process. Use these parameters to quickly prototype various ideas.

Warning

Hyperoptable parameters cannot be used in populate_indicators - as hyperopt does not recalculate indicators for each epoch, so the starting value would be used in this case.

"},{"location":"hyperopt/#optimizing-an-indicator-parameter","title":"Optimizing an indicator parameter","text":"

Assuming you have a simple strategy in mind - a EMA cross strategy (2 Moving averages crossing) - and you'd like to find the ideal parameters for this strategy. By default, we assume a stoploss of 5% - and a take-profit (minimal_roi) of 10% - which means freqtrade will sell the trade once 10% profit has been reached.

from pandas import DataFrame\nfrom functools import reduce\n\nimport talib.abstract as ta\n\nfrom freqtrade.strategy import (BooleanParameter, CategoricalParameter, DecimalParameter, \n                                IStrategy, IntParameter)\nimport freqtrade.vendor.qtpylib.indicators as qtpylib\n\nclass MyAwesomeStrategy(IStrategy):\n    stoploss = -0.05\n    timeframe = '15m'\n    minimal_roi = {\n        \"0\":  0.10\n    }\n    # Define the parameter spaces\n    buy_ema_short = IntParameter(3, 50, default=5)\n    buy_ema_long = IntParameter(15, 200, default=50)\n\n\n    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n\"\"\"Generate all indicators used by the strategy\"\"\"\n\n        # Calculate all ema_short values\n        for val in self.buy_ema_short.range:\n            dataframe[f'ema_short_{val}'] = ta.EMA(dataframe, timeperiod=val)\n\n        # Calculate all ema_long values\n        for val in self.buy_ema_long.range:\n            dataframe[f'ema_long_{val}'] = ta.EMA(dataframe, timeperiod=val)\n\n        return dataframe\n\n    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n        conditions = []\n        conditions.append(qtpylib.crossed_above(\n                dataframe[f'ema_short_{self.buy_ema_short.value}'], dataframe[f'ema_long_{self.buy_ema_long.value}']\n            ))\n\n        # Check that volume is not 0\n        conditions.append(dataframe['volume'] > 0)\n\n        if conditions:\n            dataframe.loc[\n                reduce(lambda x, y: x & y, conditions),\n                'enter_long'] = 1\n        return dataframe\n\n    def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n        conditions = []\n        conditions.append(qtpylib.crossed_above(\n                dataframe[f'ema_long_{self.buy_ema_long.value}'], dataframe[f'ema_short_{self.buy_ema_short.value}']\n            ))\n\n        # Check that volume is not 0\n        conditions.append(dataframe['volume'] > 0)\n\n        if conditions:\n            dataframe.loc[\n                reduce(lambda x, y: x & y, conditions),\n                'exit_long'] = 1\n        return dataframe\n

Breaking it down:

Using self.buy_ema_short.range will return a range object containing all entries between the Parameters low and high value. In this case (IntParameter(3, 50, default=5)), the loop would run for all numbers between 3 and 50 ([3, 4, 5, ... 49, 50]). By using this in a loop, hyperopt will generate 48 new columns (['buy_ema_3', 'buy_ema_4', ... , 'buy_ema_50']).

Hyperopt itself will then use the selected value to create the buy and sell signals.

While this strategy is most likely too simple to provide consistent profit, it should serve as an example how optimize indicator parameters.

Note

self.buy_ema_short.range will act differently between hyperopt and other modes. For hyperopt, the above example may generate 48 new columns, however for all other modes (backtesting, dry/live), it will only generate the column for the selected value. You should therefore avoid using the resulting column with explicit values (values other than self.buy_ema_short.value).

Note

range property may also be used with DecimalParameter and CategoricalParameter. RealParameter does not provide this property due to infinite search space.

Performance tip

During normal hyperopting, indicators are calculated once and supplied to each epoch, linearly increasing RAM usage as a factor of increasing cores. As this also has performance implications, hyperopt provides --analyze-per-epoch which will move the execution of populate_indicators() to the epoch process, calculating a single value per parameter per epoch instead of using the .range functionality. In this case, .range functionality will only return the actually used value. This will reduce RAM usage, but increase CPU usage. However, your hyperopting run will be less likely to fail due to Out Of Memory (OOM) issues.

In either case, you should try to use space ranges as small as possible this will improve CPU/RAM usage in both scenarios.

"},{"location":"hyperopt/#optimizing-protections","title":"Optimizing protections","text":"

Freqtrade can also optimize protections. How you optimize protections is up to you, and the following should be considered as example only.

The strategy will simply need to define the \"protections\" entry as property returning a list of protection configurations.

from pandas import DataFrame\nfrom functools import reduce\n\nimport talib.abstract as ta\n\nfrom freqtrade.strategy import (BooleanParameter, CategoricalParameter, DecimalParameter, \n                                IStrategy, IntParameter)\nimport freqtrade.vendor.qtpylib.indicators as qtpylib\n\nclass MyAwesomeStrategy(IStrategy):\n    stoploss = -0.05\n    timeframe = '15m'\n    # Define the parameter spaces\n    cooldown_lookback = IntParameter(2, 48, default=5, space=\"protection\", optimize=True)\n    stop_duration = IntParameter(12, 200, default=5, space=\"protection\", optimize=True)\n    use_stop_protection = BooleanParameter(default=True, space=\"protection\", optimize=True)\n\n\n    @property\n    def protections(self):\n        prot = []\n\n        prot.append({\n            \"method\": \"CooldownPeriod\",\n            \"stop_duration_candles\": self.cooldown_lookback.value\n        })\n        if self.use_stop_protection.value:\n            prot.append({\n                \"method\": \"StoplossGuard\",\n                \"lookback_period_candles\": 24 * 3,\n                \"trade_limit\": 4,\n                \"stop_duration_candles\": self.stop_duration.value,\n                \"only_per_pair\": False\n            })\n\n        return prot\n\n    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n        # ...\n

You can then run hyperopt as follows: freqtrade hyperopt --hyperopt-loss SharpeHyperOptLossDaily --strategy MyAwesomeStrategy --spaces protection

Note

The protection space is not part of the default space, and is only available with the Parameters Hyperopt interface, not with the legacy hyperopt interface (which required separate hyperopt files). Freqtrade will also automatically change the \"--enable-protections\" flag if the protection space is selected.

Warning

If protections are defined as property, entries from the configuration will be ignored. It is therefore recommended to not define protections in the configuration.

"},{"location":"hyperopt/#migrating-from-previous-property-setups","title":"Migrating from previous property setups","text":"

A migration from a previous setup is pretty simple, and can be accomplished by converting the protections entry to a property. In simple terms, the following configuration will be converted to the below.

class MyAwesomeStrategy(IStrategy):\n    protections = [\n        {\n            \"method\": \"CooldownPeriod\",\n            \"stop_duration_candles\": 4\n        }\n    ]\n

Result

class MyAwesomeStrategy(IStrategy):\n\n    @property\n    def protections(self):\n        return [\n            {\n                \"method\": \"CooldownPeriod\",\n                \"stop_duration_candles\": 4\n            }\n        ]\n

You will then obviously also change potential interesting entries to parameters to allow hyper-optimization.

"},{"location":"hyperopt/#optimizing-max_entry_position_adjustment","title":"Optimizing max_entry_position_adjustment","text":"

While max_entry_position_adjustment is not a separate space, it can still be used in hyperopt by using the property approach shown above.

from pandas import DataFrame\nfrom functools import reduce\n\nimport talib.abstract as ta\n\nfrom freqtrade.strategy import (BooleanParameter, CategoricalParameter, DecimalParameter, \n                                IStrategy, IntParameter)\nimport freqtrade.vendor.qtpylib.indicators as qtpylib\n\nclass MyAwesomeStrategy(IStrategy):\n    stoploss = -0.05\n    timeframe = '15m'\n\n    # Define the parameter spaces\n    max_epa = CategoricalParameter([-1, 0, 1, 3, 5, 10], default=1, space=\"buy\", optimize=True)\n\n    @property\n    def max_entry_position_adjustment(self):\n        return self.max_epa.value\n\n\n    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n        # ...\n
Using IntParameter

You can also use the IntParameter for this optimization, but you must explicitly return an integer:

max_epa = IntParameter(-1, 10, default=1, space=\"buy\", optimize=True)\n\n@property\ndef max_entry_position_adjustment(self):\n    return int(self.max_epa.value)\n

"},{"location":"hyperopt/#loss-functions","title":"Loss-functions","text":"

Each hyperparameter tuning requires a target. This is usually defined as a loss function (sometimes also called objective function), which should decrease for more desirable results, and increase for bad results.

A loss function must be specified via the --hyperopt-loss <Class-name> argument (or optionally via the configuration under the \"hyperopt_loss\" key). This class should be in its own file within the user_data/hyperopts/ directory.

Currently, the following loss functions are builtin:

  • ShortTradeDurHyperOptLoss - (default legacy Freqtrade hyperoptimization loss function) - Mostly for short trade duration and avoiding losses.
  • OnlyProfitHyperOptLoss - takes only amount of profit into consideration.
  • SharpeHyperOptLoss - optimizes Sharpe Ratio calculated on trade returns relative to standard deviation.
  • SharpeHyperOptLossDaily - optimizes Sharpe Ratio calculated on daily trade returns relative to standard deviation.
  • SortinoHyperOptLoss - optimizes Sortino Ratio calculated on trade returns relative to downside standard deviation.
  • SortinoHyperOptLossDaily - optimizes Sortino Ratio calculated on daily trade returns relative to downside standard deviation.
  • MaxDrawDownHyperOptLoss - Optimizes Maximum absolute drawdown.
  • MaxDrawDownRelativeHyperOptLoss - Optimizes both maximum absolute drawdown while also adjusting for maximum relative drawdown.
  • CalmarHyperOptLoss - Optimizes Calmar Ratio calculated on trade returns relative to max drawdown.
  • ProfitDrawDownHyperOptLoss - Optimizes by max Profit & min Drawdown objective. DRAWDOWN_MULT variable within the hyperoptloss file can be adjusted to be stricter or more flexible on drawdown purposes.

Creation of a custom loss function is covered in the Advanced Hyperopt part of the documentation.

"},{"location":"hyperopt/#execute-hyperopt","title":"Execute Hyperopt","text":"

Once you have updated your hyperopt configuration you can run it. Because hyperopt tries a lot of combinations to find the best parameters it will take time to get a good result.

We strongly recommend to use screen or tmux to prevent any connection loss.

freqtrade hyperopt --config config.json --hyperopt-loss <hyperoptlossname> --strategy <strategyname> -e 500 --spaces all\n

The -e option will set how many evaluations hyperopt will do. Since hyperopt uses Bayesian search, running too many epochs at once may not produce greater results. Experience has shown that best results are usually not improving much after 500-1000 epochs. Doing multiple runs (executions) with a few 1000 epochs and different random state will most likely produce different results.

The --spaces all option determines that all possible parameters should be optimized. Possibilities are listed below.

Note

Hyperopt will store hyperopt results with the timestamp of the hyperopt start time. Reading commands (hyperopt-list, hyperopt-show) can use --hyperopt-filename <filename> to read and display older hyperopt results. You can find a list of filenames with ls -l user_data/hyperopt_results/.

"},{"location":"hyperopt/#execute-hyperopt-with-different-historical-data-source","title":"Execute Hyperopt with different historical data source","text":"

If you would like to hyperopt parameters using an alternate historical data set that you have on-disk, use the --datadir PATH option. By default, hyperopt uses data from directory user_data/data.

"},{"location":"hyperopt/#running-hyperopt-with-a-smaller-test-set","title":"Running Hyperopt with a smaller test-set","text":"

Use the --timerange argument to change how much of the test-set you want to use. For example, to use one month of data, pass --timerange 20210101-20210201 (from january 2021 - february 2021) to the hyperopt call.

Full command:

freqtrade hyperopt --strategy <strategyname> --timerange 20210101-20210201\n
"},{"location":"hyperopt/#running-hyperopt-with-smaller-search-space","title":"Running Hyperopt with Smaller Search Space","text":"

Use the --spaces option to limit the search space used by hyperopt. Letting Hyperopt optimize everything is a huuuuge search space. Often it might make more sense to start by just searching for initial buy algorithm. Or maybe you just want to optimize your stoploss or roi table for that awesome new buy strategy you have.

Legal values are:

  • all: optimize everything
  • buy: just search for a new buy strategy
  • sell: just search for a new sell strategy
  • roi: just optimize the minimal profit table for your strategy
  • stoploss: search for the best stoploss value
  • trailing: search for the best trailing stop values
  • trades: search for the best max open trades values
  • protection: search for the best protection parameters (read the protections section on how to properly define these)
  • default: all except trailing and protection
  • space-separated list of any of the above values for example --spaces roi stoploss

The default Hyperopt Search Space, used when no --space command line option is specified, does not include the trailing hyperspace. We recommend you to run optimization for the trailing hyperspace separately, when the best parameters for other hyperspaces were found, validated and pasted into your custom strategy.

"},{"location":"hyperopt/#understand-the-hyperopt-result","title":"Understand the Hyperopt Result","text":"

Once Hyperopt is completed you can use the result to update your strategy. Given the following result from hyperopt:

Best result:\n\n    44/100:    135 trades. Avg profit  0.57%. Total profit  0.03871918 BTC (0.7722%). Avg duration 180.4 mins. Objective: 1.94367\n\n    # Buy hyperspace params:\n    buy_params = {\n        'buy_adx': 44,\n        'buy_rsi': 29,\n        'buy_adx_enabled': False,\n        'buy_rsi_enabled': True,\n        'buy_trigger': 'bb_lower'\n    }\n

You should understand this result like:

  • The buy trigger that worked best was bb_lower.
  • You should not use ADX because 'buy_adx_enabled': False.
  • You should consider using the RSI indicator ('buy_rsi_enabled': True) and the best value is 29.0 ('buy_rsi': 29.0)
"},{"location":"hyperopt/#automatic-parameter-application-to-the-strategy","title":"Automatic parameter application to the strategy","text":"

When using Hyperoptable parameters, the result of your hyperopt-run will be written to a json file next to your strategy (so for MyAwesomeStrategy.py, the file would be MyAwesomeStrategy.json). This file is also updated when using the hyperopt-show sub-command, unless --disable-param-export is provided to either of the 2 commands.

Your strategy class can also contain these results explicitly. Simply copy hyperopt results block and paste them at class level, replacing old parameters (if any). New parameters will automatically be loaded next time strategy is executed.

Transferring your whole hyperopt result to your strategy would then look like:

class MyAwesomeStrategy(IStrategy):\n    # Buy hyperspace params:\n    buy_params = {\n        'buy_adx': 44,\n        'buy_rsi': 29,\n        'buy_adx_enabled': False,\n        'buy_rsi_enabled': True,\n        'buy_trigger': 'bb_lower'\n    }\n

Note

Values in the configuration file will overwrite Parameter-file level parameters - and both will overwrite parameters within the strategy. The prevalence is therefore: config > parameter file > strategy *_params > parameter default

"},{"location":"hyperopt/#understand-hyperopt-roi-results","title":"Understand Hyperopt ROI results","text":"

If you are optimizing ROI (i.e. if optimization search-space contains 'all', 'default' or 'roi'), your result will look as follows and include a ROI table:

Best result:\n\n    44/100:    135 trades. Avg profit  0.57%. Total profit  0.03871918 BTC (0.7722%). Avg duration 180.4 mins. Objective: 1.94367\n\n    # ROI table:\n    minimal_roi = {\n        0: 0.10674,\n        21: 0.09158,\n        78: 0.03634,\n        118: 0\n    }\n

In order to use this best ROI table found by Hyperopt in backtesting and for live trades/dry-run, copy-paste it as the value of the minimal_roi attribute of your custom strategy:

    # Minimal ROI designed for the strategy.\n    # This attribute will be overridden if the config file contains \"minimal_roi\"\n    minimal_roi = {\n        0: 0.10674,\n        21: 0.09158,\n        78: 0.03634,\n        118: 0\n    }\n

As stated in the comment, you can also use it as the value of the minimal_roi setting in the configuration file.

"},{"location":"hyperopt/#default-roi-search-space","title":"Default ROI Search Space","text":"

If you are optimizing ROI, Freqtrade creates the 'roi' optimization hyperspace for you -- it's the hyperspace of components for the ROI tables. By default, each ROI table generated by the Freqtrade consists of 4 rows (steps). Hyperopt implements adaptive ranges for ROI tables with ranges for values in the ROI steps that depend on the timeframe used. By default the values vary in the following ranges (for some of the most used timeframes, values are rounded to 3 digits after the decimal point):

# step 1m 5m 1h 1d 1 0 0.011...0.119 0 0.03...0.31 0 0.068...0.711 0 0.121...1.258 2 2...8 0.007...0.042 10...40 0.02...0.11 120...480 0.045...0.252 2880...11520 0.081...0.446 3 4...20 0.003...0.015 20...100 0.01...0.04 240...1200 0.022...0.091 5760...28800 0.040...0.162 4 6...44 0.0 30...220 0.0 360...2640 0.0 8640...63360 0.0

These ranges should be sufficient in most cases. The minutes in the steps (ROI dict keys) are scaled linearly depending on the timeframe used. The ROI values in the steps (ROI dict values) are scaled logarithmically depending on the timeframe used.

If you have the generate_roi_table() and roi_space() methods in your custom hyperopt, remove them in order to utilize these adaptive ROI tables and the ROI hyperoptimization space generated by Freqtrade by default.

Override the roi_space() method if you need components of the ROI tables to vary in other ranges. Override the generate_roi_table() and roi_space() methods and implement your own custom approach for generation of the ROI tables during hyperoptimization if you need a different structure of the ROI tables or other amount of rows (steps).

A sample for these methods can be found in the overriding pre-defined spaces section.

Reduced search space

To limit the search space further, Decimals are limited to 3 decimal places (a precision of 0.001). This is usually sufficient, every value more precise than this will usually result in overfitted results. You can however overriding pre-defined spaces to change this to your needs.

"},{"location":"hyperopt/#understand-hyperopt-stoploss-results","title":"Understand Hyperopt Stoploss results","text":"

If you are optimizing stoploss values (i.e. if optimization search-space contains 'all', 'default' or 'stoploss'), your result will look as follows and include stoploss:

Best result:\n\n    44/100:    135 trades. Avg profit  0.57%. Total profit  0.03871918 BTC (0.7722%). Avg duration 180.4 mins. Objective: 1.94367\n\n    # Buy hyperspace params:\n    buy_params = {\n        'buy_adx': 44,\n        'buy_rsi': 29,\n        'buy_adx_enabled': False,\n        'buy_rsi_enabled': True,\n        'buy_trigger': 'bb_lower'\n    }\n\n    stoploss: -0.27996\n

In order to use this best stoploss value found by Hyperopt in backtesting and for live trades/dry-run, copy-paste it as the value of the stoploss attribute of your custom strategy:

    # Optimal stoploss designed for the strategy\n    # This attribute will be overridden if the config file contains \"stoploss\"\n    stoploss = -0.27996\n

As stated in the comment, you can also use it as the value of the stoploss setting in the configuration file.

"},{"location":"hyperopt/#default-stoploss-search-space","title":"Default Stoploss Search Space","text":"

If you are optimizing stoploss values, Freqtrade creates the 'stoploss' optimization hyperspace for you. By default, the stoploss values in that hyperspace vary in the range -0.35...-0.02, which is sufficient in most cases.

If you have the stoploss_space() method in your custom hyperopt file, remove it in order to utilize Stoploss hyperoptimization space generated by Freqtrade by default.

Override the stoploss_space() method and define the desired range in it if you need stoploss values to vary in other range during hyperoptimization. A sample for this method can be found in the overriding pre-defined spaces section.

Reduced search space

To limit the search space further, Decimals are limited to 3 decimal places (a precision of 0.001). This is usually sufficient, every value more precise than this will usually result in overfitted results. You can however overriding pre-defined spaces to change this to your needs.

"},{"location":"hyperopt/#understand-hyperopt-trailing-stop-results","title":"Understand Hyperopt Trailing Stop results","text":"

If you are optimizing trailing stop values (i.e. if optimization search-space contains 'all' or 'trailing'), your result will look as follows and include trailing stop parameters:

Best result:\n\n    45/100:    606 trades. Avg profit  1.04%. Total profit  0.31555614 BTC ( 630.48%). Avg duration 150.3 mins. Objective: -1.10161\n\n    # Trailing stop:\n    trailing_stop = True\n    trailing_stop_positive = 0.02001\n    trailing_stop_positive_offset = 0.06038\n    trailing_only_offset_is_reached = True\n

In order to use these best trailing stop parameters found by Hyperopt in backtesting and for live trades/dry-run, copy-paste them as the values of the corresponding attributes of your custom strategy:

    # Trailing stop\n    # These attributes will be overridden if the config file contains corresponding values.\n    trailing_stop = True\n    trailing_stop_positive = 0.02001\n    trailing_stop_positive_offset = 0.06038\n    trailing_only_offset_is_reached = True\n

As stated in the comment, you can also use it as the values of the corresponding settings in the configuration file.

"},{"location":"hyperopt/#default-trailing-stop-search-space","title":"Default Trailing Stop Search Space","text":"

If you are optimizing trailing stop values, Freqtrade creates the 'trailing' optimization hyperspace for you. By default, the trailing_stop parameter is always set to True in that hyperspace, the value of the trailing_only_offset_is_reached vary between True and False, the values of the trailing_stop_positive and trailing_stop_positive_offset parameters vary in the ranges 0.02...0.35 and 0.01...0.1 correspondingly, which is sufficient in most cases.

Override the trailing_space() method and define the desired range in it if you need values of the trailing stop parameters to vary in other ranges during hyperoptimization. A sample for this method can be found in the overriding pre-defined spaces section.

Reduced search space

To limit the search space further, Decimals are limited to 3 decimal places (a precision of 0.001). This is usually sufficient, every value more precise than this will usually result in overfitted results. You can however overriding pre-defined spaces to change this to your needs.

"},{"location":"hyperopt/#reproducible-results","title":"Reproducible results","text":"

The search for optimal parameters starts with a few (currently 30) random combinations in the hyperspace of parameters, random Hyperopt epochs. These random epochs are marked with an asterisk character (*) in the first column in the Hyperopt output.

The initial state for generation of these random values (random state) is controlled by the value of the --random-state command line option. You can set it to some arbitrary value of your choice to obtain reproducible results.

If you have not set this value explicitly in the command line options, Hyperopt seeds the random state with some random value for you. The random state value for each Hyperopt run is shown in the log, so you can copy and paste it into the --random-state command line option to repeat the set of the initial random epochs used.

If you have not changed anything in the command line options, configuration, timerange, Strategy and Hyperopt classes, historical data and the Loss Function -- you should obtain same hyper-optimization results with same random state value used.

"},{"location":"hyperopt/#output-formatting","title":"Output formatting","text":"

By default, hyperopt prints colorized results -- epochs with positive profit are printed in the green color. This highlighting helps you find epochs that can be interesting for later analysis. Epochs with zero total profit or with negative profits (losses) are printed in the normal color. If you do not need colorization of results (for instance, when you are redirecting hyperopt output to a file) you can switch colorization off by specifying the --no-color option in the command line.

You can use the --print-all command line option if you would like to see all results in the hyperopt output, not only the best ones. When --print-all is used, current best results are also colorized by default -- they are printed in bold (bright) style. This can also be switched off with the --no-color command line option.

Windows and color output

Windows does not support color-output natively, therefore it is automatically disabled. To have color-output for hyperopt running under windows, please consider using WSL.

"},{"location":"hyperopt/#position-stacking-and-disabling-max-market-positions","title":"Position stacking and disabling max market positions","text":"

In some situations, you may need to run Hyperopt (and Backtesting) with the --eps/--enable-position-staking and --dmmp/--disable-max-market-positions arguments.

By default, hyperopt emulates the behavior of the Freqtrade Live Run/Dry Run, where only one open trade is allowed for every traded pair. The total number of trades open for all pairs is also limited by the max_open_trades setting. During Hyperopt/Backtesting this may lead to some potential trades to be hidden (or masked) by previously open trades.

The --eps/--enable-position-stacking argument allows emulation of buying the same pair multiple times, while --dmmp/--disable-max-market-positions disables applying max_open_trades during Hyperopt/Backtesting (which is equal to setting max_open_trades to a very high number).

Note

Dry/live runs will NOT use position stacking - therefore it does make sense to also validate the strategy without this as it's closer to reality.

You can also enable position stacking in the configuration file by explicitly setting \"position_stacking\"=true.

"},{"location":"hyperopt/#out-of-memory-errors","title":"Out of Memory errors","text":"

As hyperopt consumes a lot of memory (the complete data needs to be in memory once per parallel backtesting process), it's likely that you run into \"out of memory\" errors. To combat these, you have multiple options:

  • Reduce the amount of pairs.
  • Reduce the timerange used (--timerange <timerange>).
  • Avoid using --timeframe-detail (this loads a lot of additional data into memory).
  • Reduce the number of parallel processes (-j <n>).
  • Increase the memory of your machine.
  • Use --analyze-per-epoch if you're using a lot of parameters with .range functionality.
"},{"location":"hyperopt/#the-objective-has-been-evaluated-at-this-point-before","title":"The objective has been evaluated at this point before.","text":"

If you see The objective has been evaluated at this point before. - then this is a sign that your space has been exhausted, or is close to that. Basically all points in your space have been hit (or a local minima has been hit) - and hyperopt does no longer find points in the multi-dimensional space it did not try yet. Freqtrade tries to counter the \"local minima\" problem by using new, randomized points in this case.

Example:

buy_ema_short = IntParameter(5, 20, default=10, space=\"buy\", optimize=True)\n# This is the only parameter in the buy space\n

The buy_ema_short space has 15 possible values (5, 6, ... 19, 20). If you now run hyperopt for the buy space, hyperopt will only have 15 values to try before running out of options. Your epochs should therefore be aligned to the possible values - or you should be ready to interrupt a run if you norice a lot of The objective has been evaluated at this point before. warnings.

"},{"location":"hyperopt/#show-details-of-hyperopt-results","title":"Show details of Hyperopt results","text":"

After you run Hyperopt for the desired amount of epochs, you can later list all results for analysis, select only best or profitable once, and show the details for any of the epochs previously evaluated. This can be done with the hyperopt-list and hyperopt-show sub-commands. The usage of these sub-commands is described in the Utils chapter.

"},{"location":"hyperopt/#validate-backtesting-results","title":"Validate backtesting results","text":"

Once the optimized strategy has been implemented into your strategy, you should backtest this strategy to make sure everything is working as expected.

To achieve same the results (number of trades, their durations, profit, etc.) as during Hyperopt, please use the same configuration and parameters (timerange, timeframe, ...) used for hyperopt --dmmp/--disable-max-market-positions and --eps/--enable-position-stacking for Backtesting.

Should results not match, please double-check to make sure you transferred all conditions correctly. Pay special care to the stoploss, max_open_trades and trailing stoploss parameters, as these are often set in configuration files, which override changes to the strategy. You should also carefully review the log of your backtest to ensure that there were no parameters inadvertently set by the configuration (like stoploss, max_open_trades or trailing_stop).

"},{"location":"installation/","title":"Installation","text":"

This page explains how to prepare your environment for running the bot.

The freqtrade documentation describes various ways to install freqtrade

  • Docker images (separate page)
  • Script Installation
  • Manual Installation
  • Installation with Conda

Please consider using the prebuilt docker images to get started quickly while evaluating how freqtrade works.

"},{"location":"installation/#information","title":"Information","text":"

For Windows installation, please use the windows installation guide.

The easiest way to install and run Freqtrade is to clone the bot Github repository and then run the ./setup.sh script, if it's available for your platform.

Version considerations

When cloning the repository the default working branch has the name develop. This branch contains all last features (can be considered as relatively stable, thanks to automated tests). The stable branch contains the code of the last release (done usually once per month on an approximately one week old snapshot of the develop branch to prevent packaging bugs, so potentially it's more stable).

Note

Python3.8 or higher and the corresponding pip are assumed to be available. The install-script will warn you and stop if that's not the case. git is also needed to clone the Freqtrade repository. Also, python headers (python<yourversion>-dev / python<yourversion>-devel) must be available for the installation to complete successfully.

Up-to-date clock

The clock on the system running the bot must be accurate, synchronized to a NTP server frequently enough to avoid problems with communication to the exchanges.

Running setup.py install for gym did not run successfully.

If you get an error related with gym we suggest you to downgrade setuptools it to version 65.5.0 you can do it with the following command:

pip install setuptools==65.5.0\n

"},{"location":"installation/#requirements","title":"Requirements","text":"

These requirements apply to both Script Installation and Manual Installation.

ARM64 systems

If you are running an ARM64 system (like a MacOS M1 or an Oracle VM), please use docker to run freqtrade. While native installation is possible with some manual effort, this is not supported at the moment.

"},{"location":"installation/#install-guide","title":"Install guide","text":"
  • Python >= 3.8.x
  • pip
  • git
  • virtualenv (Recommended)
  • TA-Lib (install instructions below)
"},{"location":"installation/#install-code","title":"Install code","text":"

We've included/collected install instructions for Ubuntu, MacOS, and Windows. These are guidelines and your success may vary with other distros. OS Specific steps are listed first, the Common section below is necessary for all systems.

Note

Python3.8 or higher and the corresponding pip are assumed to be available.

Debian/UbuntuRaspberryPi/Raspbian

The following assumes the latest Raspbian Buster lite image. This image comes with python3.9 preinstalled, making it easy to get freqtrade up and running.

Tested using a Raspberry Pi 3 with the Raspbian Buster lite image, all updates applied.

sudo apt-get install python3-venv libatlas-base-dev cmake curl\n# Use pywheels.org to speed up installation\nsudo echo \"[global]\\nextra-index-url=https://www.piwheels.org/simple\" > tee /etc/pip.conf\n\ngit clone https://github.com/freqtrade/freqtrade.git\ncd freqtrade\n\nbash setup.sh -i\n

Installation duration

Depending on your internet speed and the Raspberry Pi version, installation can take multiple hours to complete. Due to this, we recommend to use the pre-build docker-image for Raspberry, by following the Docker quickstart documentation

Note

The above does not install hyperopt dependencies. To install these, please use python3 -m pip install -e .[hyperopt]. We do not advise to run hyperopt on a Raspberry Pi, since this is a very resource-heavy operation, which should be done on powerful machine.

"},{"location":"installation/#install-necessary-dependencies","title":"Install necessary dependencies","text":"
# update repository\nsudo apt-get update\n\n# install packages\nsudo apt install -y python3-pip python3-venv python3-dev python3-pandas git curl\n
"},{"location":"installation/#freqtrade-repository","title":"Freqtrade repository","text":"

Freqtrade is an open source crypto-currency trading bot, whose code is hosted on github.com

# Download `develop` branch of freqtrade repository\ngit clone https://github.com/freqtrade/freqtrade.git\n\n# Enter downloaded directory\ncd freqtrade\n\n# your choice (1): novice user\ngit checkout stable\n\n# your choice (2): advanced user\ngit checkout develop\n

(1) This command switches the cloned repository to the use of the stable branch. It's not needed, if you wish to stay on the (2) develop branch.

You may later switch between branches at any time with the git checkout stable/git checkout develop commands.

Install from pypi

An alternative way to install Freqtrade is from pypi. The downside is that this method requires ta-lib to be correctly installed beforehand, and is therefore currently not the recommended way to install Freqtrade.

pip install freqtrade\n
"},{"location":"installation/#script-installation","title":"Script Installation","text":"

First of the ways to install Freqtrade, is to use provided the Linux/MacOS ./setup.sh script, which install all dependencies and help you configure the bot.

Make sure you fulfill the Requirements and have downloaded the Freqtrade repository.

"},{"location":"installation/#use-setupsh-install-linuxmacos","title":"Use /setup.sh -install (Linux/MacOS)","text":"

If you are on Debian, Ubuntu or MacOS, freqtrade provides the script to install freqtrade.

# --install, Install freqtrade from scratch\n./setup.sh -i\n
"},{"location":"installation/#activate-your-virtual-environment","title":"Activate your virtual environment","text":"

Each time you open a new terminal, you must run source .env/bin/activate to activate your virtual environment.

# then activate your .env\nsource ./.env/bin/activate\n
"},{"location":"installation/#congratulations","title":"Congratulations","text":"

You are ready, and run the bot

"},{"location":"installation/#other-options-of-setupsh-script","title":"Other options of /setup.sh script","text":"

You can as well update, configure and reset the codebase of your bot with ./script.sh

# --update, Command git pull to update.\n./setup.sh -u\n# --reset, Hard reset your develop/stable branch.\n./setup.sh -r\n
** --install **\n\nWith this option, the script will install the bot and most dependencies:\nYou will need to have git and python3.8+ installed beforehand for this to work.\n\n* Mandatory software as: `ta-lib`\n* Setup your virtualenv under `.env/`\n\nThis option is a combination of installation tasks and `--reset`\n\n** --update **\n\nThis option will pull the last version of your current branch and update your virtualenv. Run the script with this option periodically to update your bot.\n\n** --reset **\n\nThis option will hard reset your branch (only if you are on either `stable` or `develop`) and recreate your virtualenv.\n
"},{"location":"installation/#manual-installation","title":"Manual Installation","text":"

Make sure you fulfill the Requirements and have downloaded the Freqtrade repository.

"},{"location":"installation/#install-ta-lib","title":"Install TA-Lib","text":""},{"location":"installation/#ta-lib-script-installation","title":"TA-Lib script installation","text":"
sudo ./build_helpers/install_ta-lib.sh\n

Note

This will use the ta-lib tar.gz included in this repository.

"},{"location":"installation/#ta-lib-manual-installation","title":"TA-Lib manual installation","text":"

Official installation guide

wget http://prdownloads.sourceforge.net/ta-lib/ta-lib-0.4.0-src.tar.gz\ntar xvzf ta-lib-0.4.0-src.tar.gz\ncd ta-lib\nsed -i.bak \"s|0.00000001|0.000000000000000001 |g\" src/ta_func/ta_utility.h\n./configure --prefix=/usr/local\nmake\nsudo make install\n# On debian based systems (debian, ubuntu, ...) - updating ldconfig might be necessary.\nsudo ldconfig  cd ..\nrm -rf ./ta-lib*\n
"},{"location":"installation/#setup-python-virtual-environment-virtualenv","title":"Setup Python virtual environment (virtualenv)","text":"

You will run freqtrade in separated virtual environment

# create virtualenv in directory /freqtrade/.env\npython3 -m venv .env\n\n# run virtualenv\nsource .env/bin/activate\n
"},{"location":"installation/#install-python-dependencies","title":"Install python dependencies","text":"
python3 -m pip install --upgrade pip\npython3 -m pip install -e .\n
"},{"location":"installation/#congratulations_1","title":"Congratulations","text":"

You are ready, and run the bot

"},{"location":"installation/#optional-post-installation-tasks","title":"(Optional) Post-installation Tasks","text":"

Note

If you run the bot on a server, you should consider using Docker or a terminal multiplexer like screen or tmux to avoid that the bot is stopped on logout.

On Linux with software suite systemd, as an optional post-installation task, you may wish to setup the bot to run as a systemd service or configure it to send the log messages to the syslog/rsyslog or journald daemons. See Advanced Logging for details.

"},{"location":"installation/#installation-with-conda","title":"Installation with Conda","text":"

Freqtrade can also be installed with Miniconda or Anaconda. We recommend using Miniconda as it's installation footprint is smaller. Conda will automatically prepare and manage the extensive library-dependencies of the Freqtrade program.

"},{"location":"installation/#what-is-conda","title":"What is Conda?","text":"

Conda is a package, dependency and environment manager for multiple programming languages: conda docs

"},{"location":"installation/#installation-with-conda_1","title":"Installation with conda","text":""},{"location":"installation/#install-conda","title":"Install Conda","text":"

Installing on linux

Installing on windows

Answer all questions. After installation, it is mandatory to turn your terminal OFF and ON again.

"},{"location":"installation/#freqtrade-download","title":"Freqtrade download","text":"

Download and install freqtrade.

# download freqtrade\ngit clone https://github.com/freqtrade/freqtrade.git\n\n# enter downloaded directory 'freqtrade'\ncd freqtrade      
"},{"location":"installation/#freqtrade-install-conda-environment","title":"Freqtrade install: Conda Environment","text":"
conda create --name freqtrade python=3.10\n

Creating Conda Environment

The conda command create -n automatically installs all nested dependencies for the selected libraries, general structure of installation command is:

# choose your own packages\nconda env create -n [name of the environment] [python version] [packages]\n
"},{"location":"installation/#enterexit-freqtrade-environment","title":"Enter/exit freqtrade environment","text":"

To check available environments, type

conda env list\n

Enter installed environment

# enter conda environment\nconda activate freqtrade\n\n# exit conda environment - don't do it now\nconda deactivate\n

Install last python dependencies with pip

python3 -m pip install --upgrade pip\npython3 -m pip install -r requirements.txt\npython3 -m pip install -e .\n

Patch conda libta-lib (Linux only)

# Ensure that the environment is active!\nconda activate freqtrade\n\ncd build_helpers\nbash install_ta-lib.sh ${CONDA_PREFIX} nosudo\n
"},{"location":"installation/#congratulations_2","title":"Congratulations","text":"

You are ready, and run the bot

"},{"location":"installation/#important-shortcuts","title":"Important shortcuts","text":"
# list installed conda environments\nconda env list\n\n# activate base environment\nconda activate\n\n# activate freqtrade environment\nconda activate freqtrade\n\n#deactivate any conda environments\nconda deactivate                              
"},{"location":"installation/#further-info-on-anaconda","title":"Further info on anaconda","text":"

New heavy packages

It may happen that creating a new Conda environment, populated with selected packages at the moment of creation takes less time than installing a large, heavy library or application, into previously set environment.

pip install within conda

The documentation of conda says that pip should NOT be used within conda, because internal problems can occur. However, they are rare. Anaconda Blogpost

Nevertheless, that is why, the conda-forge channel is preferred:

  • more libraries are available (less need for pip)
  • conda-forge works better with pip
  • the libraries are newer

Happy trading!

"},{"location":"installation/#you-are-ready","title":"You are ready","text":"

You've made it this far, so you have successfully installed freqtrade.

"},{"location":"installation/#initialize-the-configuration","title":"Initialize the configuration","text":"
# Step 1 - Initialize user folder\nfreqtrade create-userdir --userdir user_data\n\n# Step 2 - Create a new configuration file\nfreqtrade new-config --config config.json\n

You are ready to run, read Bot Configuration, remember to start with dry_run: True and verify that everything is working.

To learn how to setup your configuration, please refer to the Bot Configuration documentation page.

"},{"location":"installation/#start-the-bot","title":"Start the Bot","text":"
freqtrade trade --config config.json --strategy SampleStrategy\n

Warning

You should read through the rest of the documentation, backtest the strategy you're going to use, and use dry-run before enabling trading with real money.

"},{"location":"installation/#troubleshooting","title":"Troubleshooting","text":""},{"location":"installation/#common-problem-command-not-found","title":"Common problem: \"command not found\"","text":"

If you used (1)Script or (2)Manual installation, you need to run the bot in virtual environment. If you get error as below, make sure venv is active.

# if:\nbash: freqtrade: command not found\n\n# then activate your .env\nsource ./.env/bin/activate\n
"},{"location":"installation/#macos-installation-error","title":"MacOS installation error","text":"

Newer versions of MacOS may have installation failed with errors like error: command 'g++' failed with exit status 1.

This error will require explicit installation of the SDK Headers, which are not installed by default in this version of MacOS. For MacOS 10.14, this can be accomplished with the below command.

open /Library/Developer/CommandLineTools/Packages/macOS_SDK_headers_for_macOS_10.14.pkg\n

If this file is inexistent, then you're probably on a different version of MacOS, so you may need to consult the internet for specific resolution details.

"},{"location":"leverage/","title":"Trading with Leverage","text":"

Beta feature

This feature is still in it's testing phase. Should you notice something you think is wrong please let us know via Discord or via Github Issue.

Multiple bots on one account

You can't run 2 bots on the same account with leverage. For leveraged / margin trading, freqtrade assumes it's the only user of the account, and all liquidation levels are calculated based on this assumption.

Trading with leverage is very risky

Do not trade with a leverage > 1 using a strategy that hasn't shown positive results in a live run using the spot market. Check the stoploss of your strategy. With a leverage of 2, a stoploss of 0.5 (50%) would be too low, and these trades would be liquidated before reaching that stoploss. We do not assume any responsibility for eventual losses that occur from using this software or this mode.

Please only use advanced trading modes when you know how freqtrade (and your strategy) works. Also, never risk more than what you can afford to lose.

If you already have an existing strategy, please read the strategy migration guide to migrate your strategy from a freqtrade v2 strategy, to strategy of version 3 which can short and trade futures.

"},{"location":"leverage/#shorting","title":"Shorting","text":"

Shorting is not possible when trading with trading_mode set to spot. To short trade, trading_mode must be set to margin(currently unavailable) or futures, with margin_mode set to cross(currently unavailable) or isolated

For a strategy to short, the strategy class must set the class variable can_short = True

Please read strategy customization for instructions on how to set signals to enter and exit short trades.

"},{"location":"leverage/#understand-trading_mode","title":"Understand trading_mode","text":"

The possible values are: spot (default), margin(Currently unavailable) or futures.

"},{"location":"leverage/#spot","title":"Spot","text":"

Regular trading mode (low risk)

  • Long trades only (No short trades).
  • No leverage.
  • No Liquidation.
  • Profits gained/lost are equal to the change in value of the assets (minus trading fees).
"},{"location":"leverage/#leverage-trading-modes","title":"Leverage trading modes","text":"

With leverage, a trader borrows capital from the exchange. The capital must be re-payed fully to the exchange (potentially with interest), and the trader keeps any profits, or pays any losses, from any trades made using the borrowed capital.

Because the capital must always be re-payed, exchanges will liquidate (forcefully sell the traders assets) a trade made using borrowed capital when the total value of assets in the leverage account drops to a certain point (a point where the total value of losses is less than the value of the collateral that the trader actually owns in the leverage account), in order to ensure that the trader has enough capital to pay the borrowed assets back to the exchange. The exchange will also charge a liquidation fee, adding to the traders losses.

For this reason, DO NOT TRADE WITH LEVERAGE IF YOU DON'T KNOW EXACTLY WHAT YOUR DOING. LEVERAGE TRADING IS HIGH RISK, AND CAN RESULT IN THE VALUE OF YOUR ASSETS DROPPING TO 0 VERY QUICKLY, WITH NO CHANCE OF INCREASING IN VALUE AGAIN.

"},{"location":"leverage/#margin-currently-unavailable","title":"Margin (currently unavailable)","text":"

Trading occurs on the spot market, but the exchange lends currency to you in an amount equal to the chosen leverage. You pay the amount lent to you back to the exchange with interest, and your profits/losses are multiplied by the leverage specified.

"},{"location":"leverage/#futures","title":"Futures","text":"

Perpetual swaps (also known as Perpetual Futures) are contracts traded at a price that is closely tied to the underlying asset they are based off of (ex.). You are not trading the actual asset but instead are trading a derivative contract. Perpetual swap contracts can last indefinitely, in contrast to futures or option contracts.

In addition to the gains/losses from the change in price of the futures contract, traders also exchange funding fees, which are gains/losses worth an amount that is derived from the difference in price between the futures contract and the underlying asset. The difference in price between a futures contract and the underlying asset varies between exchanges.

To trade in futures markets, you'll have to set trading_mode to \"futures\". You will also have to pick a \"margin mode\" (explanation below) - with freqtrade currently only supporting isolated margin.

\"trading_mode\": \"futures\",\n\"margin_mode\": \"isolated\"\n
"},{"location":"leverage/#pair-namings","title":"Pair namings","text":"

Freqtrade follows the ccxt naming conventions for futures. A futures pair will therefore have the naming of base/quote:settle (e.g. ETH/USDT:USDT).

"},{"location":"leverage/#margin-mode","title":"Margin mode","text":"

On top of trading_mode - you will also have to configure your margin_mode. While freqtrade currently only supports one margin mode, this will change, and by configuring it now you're all set for future updates.

The possible values are: isolated, or cross(currently unavailable).

"},{"location":"leverage/#isolated-margin-mode","title":"Isolated margin mode","text":"

Each market(trading pair), keeps collateral in a separate account

\"margin_mode\": \"isolated\"\n
"},{"location":"leverage/#cross-margin-mode-currently-unavailable","title":"Cross margin mode (currently unavailable)","text":"

One account is used to share collateral between markets (trading pairs). Margin is taken from total account balance to avoid liquidation when needed.

\"margin_mode\": \"cross\"\n

Please read the exchange specific notes for exchanges that support this mode and how they differ.

"},{"location":"leverage/#set-leverage-to-use","title":"Set leverage to use","text":"

Different strategies and risk profiles will require different levels of leverage. While you could configure one static leverage value - freqtrade offers you the flexibility to adjust this via strategy leverage callback - which allows you to use different leverages by pair, or based on some other factor benefitting your strategy result.

If not implemented, leverage defaults to 1x (no leverage).

Warning

Higher leverage also equals higher risk - be sure you fully understand the implications of using leverage!

"},{"location":"leverage/#understand-liquidation_buffer","title":"Understand liquidation_buffer","text":"

Defaults to 0.05

A ratio specifying how large of a safety net to place between the liquidation price and the stoploss to prevent a position from reaching the liquidation price. This artificial liquidation price is calculated as:

freqtrade_liquidation_price = liquidation_price \u00b1 (abs(open_rate - liquidation_price) * liquidation_buffer)

  • \u00b1 = + for long trades
  • \u00b1 = - for short trades

Possible values are any floats between 0.0 and 0.99

ex: If a trade is entered at a price of 10 coin/USDT, and the liquidation price of this trade is 8 coin/USDT, then with liquidation_buffer set to 0.05 the minimum stoploss for this trade would be \\(8 + ((10 - 8) * 0.05) = 8 + 0.1 = 8.1\\)

A liquidation_buffer of 0.0, or a low liquidation_buffer is likely to result in liquidations, and liquidation fees

Currently Freqtrade is able to calculate liquidation prices, but does not calculate liquidation fees. Setting your liquidation_buffer to 0.0, or using a low liquidation_buffer could result in your positions being liquidated. Freqtrade does not track liquidation fees, so liquidations will result in inaccurate profit/loss results for your bot. If you use a low liquidation_buffer, it is recommended to use stoploss_on_exchange if your exchange supports this.

"},{"location":"leverage/#unavailable-funding-rates","title":"Unavailable funding rates","text":"

For futures data, exchanges commonly provide the futures candles, the marks, and the funding rates. However, it is common that whilst candles and marks might be available, the funding rates are not. This can affect backtesting timeranges, i.e. you may only be able to test recent timeranges and not earlier, experiencing the No data found. Terminating. error. To get around this, add the futures_funding_rate config option as listed in configuration.md, and it is recommended that you set this to 0, unless you know a given specific funding rate for your pair, exchange and timerange. Setting this to anything other than 0 can have drastic effects on your profit calculations within strategy, e.g. within the custom_exit, custom_stoploss, etc functions.

This will mean your backtests are inaccurate.

This will not overwrite funding rates that are available from the exchange, but bear in mind that setting a false funding rate will mean backtesting results will be inaccurate for historical timeranges where funding rates are not available.

"},{"location":"leverage/#developer","title":"Developer","text":""},{"location":"leverage/#margin-mode_1","title":"Margin mode","text":"

For shorts, the currency which pays the interest fee for the borrowed currency is purchased at the same time of the closing trade (This means that the amount purchased in short closing trades is greater than the amount sold in short opening trades).

For longs, the currency which pays the interest fee for the borrowed will already be owned by the user and does not need to be purchased. The interest is subtracted from the close_value of the trade.

All Fees are included in current_profit calculations during the trade.

"},{"location":"leverage/#futures-mode","title":"Futures mode","text":"

Funding fees are either added or subtracted from the total amount of a trade

"},{"location":"plotting/","title":"Plotting","text":"

This page explains how to plot prices, indicators and profits.

"},{"location":"plotting/#installation-setup","title":"Installation / Setup","text":"

Plotting modules use the Plotly library. You can install / upgrade this by running the following command:

pip install -U -r requirements-plot.txt\n
"},{"location":"plotting/#plot-price-and-indicators","title":"Plot price and indicators","text":"

The freqtrade plot-dataframe subcommand shows an interactive graph with three subplots:

  • Main plot with candlesticks and indicators following price (sma/ema)
  • Volume bars
  • Additional indicators as specified by --indicators2

Possible arguments:

usage: freqtrade plot-dataframe [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                                [-d PATH] [--userdir PATH] [-s NAME]\n                                [--strategy-path PATH] [-p PAIRS [PAIRS ...]]\n                                [--indicators1 INDICATORS1 [INDICATORS1 ...]]\n                                [--indicators2 INDICATORS2 [INDICATORS2 ...]]\n                                [--plot-limit INT] [--db-url PATH]\n                                [--trade-source {DB,file}] [--export EXPORT]\n                                [--export-filename PATH]\n                                [--timerange TIMERANGE] [-i TIMEFRAME]\n                                [--no-trades]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]\n                        Limit command to these pairs. Pairs are space-\n                        separated.\n  --indicators1 INDICATORS1 [INDICATORS1 ...]\n                        Set indicators from your strategy you want in the\n                        first row of the graph. Space-separated list. Example:\n                        `ema3 ema5`. Default: `['sma', 'ema3', 'ema5']`.\n  --indicators2 INDICATORS2 [INDICATORS2 ...]\n                        Set indicators from your strategy you want in the\n                        third row of the graph. Space-separated list. Example:\n                        `fastd fastk`. Default: `['macd', 'macdsignal']`.\n  --plot-limit INT      Specify tick limit for plotting. Notice: too high\n                        values cause huge files. Default: 750.\n  --db-url PATH         Override trades database URL, this is useful in custom\n                        deployments (default: `sqlite:///tradesv3.sqlite` for\n                        Live Run mode, `sqlite:///tradesv3.dryrun.sqlite` for\n                        Dry Run).\n  --trade-source {DB,file}\n                        Specify the source for trades (Can be DB or file\n                        (backtest file)) Default: file\n  --export EXPORT       Export backtest results, argument are: trades.\n                        Example: `--export=trades`\n  --export-filename PATH\n                        Save backtest results to the file with this filename.\n                        Requires `--export` to be set as well. Example:\n                        `--export-filename=user_data/backtest_results/backtest\n                        _today.json`\n  --timerange TIMERANGE\n                        Specify what timerange of data to use.\n  -i TIMEFRAME, --timeframe TIMEFRAME\n                        Specify timeframe (`1m`, `5m`, `30m`, `1h`, `1d`).\n  --no-trades           Skip using trades from backtesting file and DB.\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n\nStrategy arguments:\n  -s NAME, --strategy NAME\n                        Specify strategy class name which will be used by the\n                        bot.\n  --strategy-path PATH  Specify additional strategy lookup path.\n

Example:

freqtrade plot-dataframe -p BTC/ETH --strategy AwesomeStrategy\n

The -p/--pairs argument can be used to specify pairs you would like to plot.

Note

The freqtrade plot-dataframe subcommand generates one plot-file per pair.

Specify custom indicators. Use --indicators1 for the main plot and --indicators2 for the subplot below (if values are in a different range than prices).

freqtrade plot-dataframe --strategy AwesomeStrategy -p BTC/ETH --indicators1 sma ema --indicators2 macd\n
"},{"location":"plotting/#further-usage-examples","title":"Further usage examples","text":"

To plot multiple pairs, separate them with a space:

freqtrade plot-dataframe --strategy AwesomeStrategy -p BTC/ETH XRP/ETH\n

To plot a timerange (to zoom in)

freqtrade plot-dataframe --strategy AwesomeStrategy -p BTC/ETH --timerange=20180801-20180805\n

To plot trades stored in a database use --db-url in combination with --trade-source DB:

freqtrade plot-dataframe --strategy AwesomeStrategy --db-url sqlite:///tradesv3.dry_run.sqlite -p BTC/ETH --trade-source DB\n

To plot trades from a backtesting result, use --export-filename <filename>

freqtrade plot-dataframe --strategy AwesomeStrategy --export-filename user_data/backtest_results/backtest-result.json -p BTC/ETH\n
"},{"location":"plotting/#plot-dataframe-basics","title":"Plot dataframe basics","text":"

The plot-dataframe subcommand requires backtesting data, a strategy and either a backtesting-results file or a database, containing trades corresponding to the strategy.

The resulting plot will have the following elements:

  • Green triangles: Buy signals from the strategy. (Note: not every buy signal generates a trade, compare to cyan circles.)
  • Red triangles: Sell signals from the strategy. (Also, not every sell signal terminates a trade, compare to red and green squares.)
  • Cyan circles: Trade entry points.
  • Red squares: Trade exit points for trades with loss or 0% profit.
  • Green squares: Trade exit points for profitable trades.
  • Indicators with values corresponding to the candle scale (e.g. SMA/EMA), as specified with --indicators1.
  • Volume (bar chart at the bottom of the main chart).
  • Indicators with values in different scales (e.g. MACD, RSI) below the volume bars, as specified with --indicators2.

Bollinger Bands

Bollinger bands are automatically added to the plot if the columns bb_lowerband and bb_upperband exist, and are painted as a light blue area spanning from the lower band to the upper band.

"},{"location":"plotting/#advanced-plot-configuration","title":"Advanced plot configuration","text":"

An advanced plot configuration can be specified in the strategy in the plot_config parameter.

Additional features when using plot_config include:

  • Specify colors per indicator
  • Specify additional subplots
  • Specify indicator pairs to fill area in between

The sample plot configuration below specifies fixed colors for the indicators. Otherwise, consecutive plots may produce different color schemes each time, making comparisons difficult. It also allows multiple subplots to display both MACD and RSI at the same time.

Plot type can be configured using type key. Possible types are:

  • scatter corresponding to plotly.graph_objects.Scatter class (default).
  • bar corresponding to plotly.graph_objects.Bar class.

Extra parameters to plotly.graph_objects.* constructor can be specified in plotly dict.

Sample configuration with inline comments explaining the process:

@property\ndef plot_config(self):\n\"\"\"\n        There are a lot of solutions how to build the return dictionary.\n        The only important point is the return value.\n        Example:\n            plot_config = {'main_plot': {}, 'subplots': {}}\n\n    \"\"\"\n    plot_config = {}\n    plot_config['main_plot'] = {\n        # Configuration for main plot indicators.\n        # Assumes 2 parameters, emashort and emalong to be specified.\n        f'ema_{self.emashort.value}': {'color': 'red'},\n        f'ema_{self.emalong.value}': {'color': '#CCCCCC'},\n        # By omitting color, a random color is selected.\n        'sar': {},\n        # fill area between senkou_a and senkou_b\n        'senkou_a': {\n            'color': 'green', #optional\n            'fill_to': 'senkou_b',\n            'fill_label': 'Ichimoku Cloud', #optional\n            'fill_color': 'rgba(255,76,46,0.2)', #optional\n        },\n        # plot senkou_b, too. Not only the area to it.\n        'senkou_b': {}\n    }\n    plot_config['subplots'] = {\n         # Create subplot MACD\n        \"MACD\": {\n            'macd': {'color': 'blue', 'fill_to': 'macdhist'},\n            'macdsignal': {'color': 'orange'},\n            'macdhist': {'type': 'bar', 'plotly': {'opacity': 0.9}}\n        },\n        # Additional subplot RSI\n        \"RSI\": {\n            'rsi': {'color': 'red'}\n        }\n    }\n\n    return plot_config\n
As attribute (former method)

Assigning plot_config is also possible as Attribute (this used to be the default way). This has the disadvantage that strategy parameters are not available, preventing certain configurations from working.

    plot_config = {\n        'main_plot': {\n            # Configuration for main plot indicators.\n            # Specifies `ema10` to be red, and `ema50` to be a shade of gray\n            'ema10': {'color': 'red'},\n            'ema50': {'color': '#CCCCCC'},\n            # By omitting color, a random color is selected.\n            'sar': {},\n        # fill area between senkou_a and senkou_b\n        'senkou_a': {\n            'color': 'green', #optional\n            'fill_to': 'senkou_b',\n            'fill_label': 'Ichimoku Cloud', #optional\n            'fill_color': 'rgba(255,76,46,0.2)', #optional\n        },\n        # plot senkou_b, too. Not only the area to it.\n        'senkou_b': {}\n        },\n        'subplots': {\n            # Create subplot MACD\n            \"MACD\": {\n                'macd': {'color': 'blue', 'fill_to': 'macdhist'},\n                'macdsignal': {'color': 'orange'},\n                'macdhist': {'type': 'bar', 'plotly': {'opacity': 0.9}}\n            },\n            # Additional subplot RSI\n            \"RSI\": {\n                'rsi': {'color': 'red'}\n            }\n        }\n    }\n

Note

The above configuration assumes that ema10, ema50, senkou_a, senkou_b, macd, macdsignal, macdhist and rsi are columns in the DataFrame created by the strategy.

Warning

plotly arguments are only supported with plotly library and will not work with freq-ui.

Trade position adjustments

If position_adjustment_enable / adjust_trade_position() is used, the trade initial buy price is averaged over multiple orders and the trade start price will most likely appear outside the candle range.

"},{"location":"plotting/#plot-profit","title":"Plot profit","text":"

The plot-profit subcommand shows an interactive graph with three plots:

  • Average closing price for all pairs.
  • The summarized profit made by backtesting. Note that this is not the real-world profit, but more of an estimate.
  • Profit for each individual pair.
  • Parallelism of trades.
  • Underwater (Periods of drawdown).

The first graph is good to get a grip of how the overall market progresses.

The second graph will show if your algorithm works or doesn't. Perhaps you want an algorithm that steadily makes small profits, or one that acts less often, but makes big swings. This graph will also highlight the start (and end) of the Max drawdown period.

The third graph can be useful to spot outliers, events in pairs that cause profit spikes.

The forth graph can help you analyze trade parallelism, showing how often max_open_trades have been maxed out.

Possible options for the freqtrade plot-profit subcommand:

usage: freqtrade plot-profit [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                             [-d PATH] [--userdir PATH] [-s NAME]\n                             [--strategy-path PATH] [-p PAIRS [PAIRS ...]]\n                             [--timerange TIMERANGE] [--export EXPORT]\n                             [--export-filename PATH] [--db-url PATH]\n                             [--trade-source {DB,file}] [-i TIMEFRAME]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]\n                        Limit command to these pairs. Pairs are space-\n                        separated.\n  --timerange TIMERANGE\n                        Specify what timerange of data to use.\n  --export EXPORT       Export backtest results, argument are: trades.\n                        Example: `--export=trades`\n  --export-filename PATH, --backtest-filename PATH\n                        Use backtest results from this filename.\n                        Requires `--export` to be set as well. Example:\n                        `--export-filename=user_data/backtest_results/backtest\n                        _today.json`\n  --db-url PATH         Override trades database URL, this is useful in custom\n                        deployments (default: `sqlite:///tradesv3.sqlite` for\n                        Live Run mode, `sqlite:///tradesv3.dryrun.sqlite` for\n                        Dry Run).\n  --trade-source {DB,file}\n                        Specify the source for trades (Can be DB or file\n                        (backtest file)) Default: file\n  -i TIMEFRAME, --timeframe TIMEFRAME\n                        Specify timeframe (`1m`, `5m`, `30m`, `1h`, `1d`).\n  --auto-open           Automatically open generated plot.\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n\nStrategy arguments:\n  -s NAME, --strategy NAME\n                        Specify strategy class name which will be used by the\n                        bot.\n  --strategy-path PATH  Specify additional strategy lookup path.\n

The -p/--pairs argument, can be used to limit the pairs that are considered for this calculation.

Examples:

Use custom backtest-export file

freqtrade plot-profit  -p LTC/BTC --export-filename user_data/backtest_results/backtest-result.json\n

Use custom database

freqtrade plot-profit  -p LTC/BTC --db-url sqlite:///tradesv3.sqlite --trade-source DB\n
freqtrade --datadir user_data/data/binance_save/ plot-profit -p LTC/BTC\n
"},{"location":"plugins/","title":"Plugins","text":""},{"location":"plugins/#pairlists-and-pairlist-handlers","title":"Pairlists and Pairlist Handlers","text":"

Pairlist Handlers define the list of pairs (pairlist) that the bot should trade. They are configured in the pairlists section of the configuration settings.

In your configuration, you can use Static Pairlist (defined by the StaticPairList Pairlist Handler) and Dynamic Pairlist (defined by the VolumePairList Pairlist Handler).

Additionally, AgeFilter, PrecisionFilter, PriceFilter, ShuffleFilter, SpreadFilter and VolatilityFilter act as Pairlist Filters, removing certain pairs and/or moving their positions in the pairlist.

If multiple Pairlist Handlers are used, they are chained and a combination of all Pairlist Handlers forms the resulting pairlist the bot uses for trading and backtesting. Pairlist Handlers are executed in the sequence they are configured. You should always configure either StaticPairList or VolumePairList as the starting Pairlist Handler.

Inactive markets are always removed from the resulting pairlist. Explicitly blacklisted pairs (those in the pair_blacklist configuration setting) are also always removed from the resulting pairlist.

"},{"location":"plugins/#pair-blacklist","title":"Pair blacklist","text":"

The pair blacklist (configured via exchange.pair_blacklist in the configuration) disallows certain pairs from trading. This can be as simple as excluding DOGE/BTC - which will remove exactly this pair.

The pair-blacklist does also support wildcards (in regex-style) - so BNB/.* will exclude ALL pairs that start with BNB. You may also use something like .*DOWN/BTC or .*UP/BTC to exclude leveraged tokens (check Pair naming conventions for your exchange!)

"},{"location":"plugins/#available-pairlist-handlers","title":"Available Pairlist Handlers","text":"
  • StaticPairList (default, if not configured differently)
  • VolumePairList
  • ProducerPairList
  • RemotePairList
  • AgeFilter
  • OffsetFilter
  • PerformanceFilter
  • PrecisionFilter
  • PriceFilter
  • ShuffleFilter
  • SpreadFilter
  • RangeStabilityFilter
  • VolatilityFilter

Testing pairlists

Pairlist configurations can be quite tricky to get right. Best use the test-pairlist utility sub-command to test your configuration quickly.

"},{"location":"plugins/#static-pair-list","title":"Static Pair List","text":"

By default, the StaticPairList method is used, which uses a statically defined pair whitelist from the configuration. The pairlist also supports wildcards (in regex-style) - so .*/BTC will include all pairs with BTC as a stake.

It uses configuration from exchange.pair_whitelist and exchange.pair_blacklist.

\"pairlists\": [\n{\"method\": \"StaticPairList\"}\n],\n

By default, only currently enabled pairs are allowed. To skip pair validation against active markets, set \"allow_inactive\": true within the StaticPairList configuration. This can be useful for backtesting expired pairs (like quarterly spot-markets). This option must be configured along with exchange.skip_pair_validation in the exchange configuration.

When used in a \"follow-up\" position (e.g. after VolumePairlist), all pairs in 'pair_whitelist' will be added to the end of the pairlist.

"},{"location":"plugins/#volume-pair-list","title":"Volume Pair List","text":"

VolumePairList employs sorting/filtering of pairs by their trading volume. It selects number_assets top pairs with sorting based on the sort_key (which can only be quoteVolume).

When used in the chain of Pairlist Handlers in a non-leading position (after StaticPairList and other Pairlist Filters), VolumePairList considers outputs of previous Pairlist Handlers, adding its sorting/selection of the pairs by the trading volume.

When used in the leading position of the chain of Pairlist Handlers, the pair_whitelist configuration setting is ignored. Instead, VolumePairList selects the top assets from all available markets with matching stake-currency on the exchange.

The refresh_period setting allows to define the period (in seconds), at which the pairlist will be refreshed. Defaults to 1800s (30 minutes). The pairlist cache (refresh_period) on VolumePairList is only applicable to generating pairlists. Filtering instances (not the first position in the list) will not apply any cache and will always use up-to-date data.

VolumePairList is per default based on the ticker data from exchange, as reported by the ccxt library:

  • The quoteVolume is the amount of quote (stake) currency traded (bought or sold) in last 24 hours.
\"pairlists\": [\n{\n\"method\": \"VolumePairList\",\n\"number_assets\": 20,\n\"sort_key\": \"quoteVolume\",\n\"min_value\": 0,\n\"refresh_period\": 1800\n}\n],\n

You can define a minimum volume with min_value - which will filter out pairs with a volume lower than the specified value in the specified timerange.

"},{"location":"plugins/#volumepairlist-advanced-mode","title":"VolumePairList Advanced mode","text":"

VolumePairList can also operate in an advanced mode to build volume over a given timerange of specified candle size. It utilizes exchange historical candle data, builds a typical price (calculated by (open+high+low)/3) and multiplies the typical price with every candle's volume. The sum is the quoteVolume over the given range. This allows different scenarios, for a more smoothened volume, when using longer ranges with larger candle sizes, or the opposite when using a short range with small candles.

For convenience lookback_days can be specified, which will imply that 1d candles will be used for the lookback. In the example below the pairlist would be created based on the last 7 days:

\"pairlists\": [\n{\n\"method\": \"VolumePairList\",\n\"number_assets\": 20,\n\"sort_key\": \"quoteVolume\",\n\"min_value\": 0,\n\"refresh_period\": 86400,\n\"lookback_days\": 7\n}\n],\n

Range look back and refresh period

When used in conjunction with lookback_days and lookback_timeframe the refresh_period can not be smaller than the candle size in seconds. As this will result in unnecessary requests to the exchanges API.

Performance implications when using lookback range

If used in first position in combination with lookback, the computation of the range based volume can be time and resource consuming, as it downloads candles for all tradable pairs. Hence it's highly advised to use the standard approach with VolumeFilter to narrow the pairlist down for further range volume calculation.

Unsupported exchanges (Bittrex, Gemini)

On some exchanges (like Bittrex and Gemini), regular VolumePairList does not work as the api does not natively provide 24h volume. This can be worked around by using candle data to build the volume. To roughly simulate 24h volume, you can use the following configuration. Please note that These pairlists will only refresh once per day.

\"pairlists\": [\n{\n\"method\": \"VolumePairList\",\n\"number_assets\": 20,\n\"sort_key\": \"quoteVolume\",\n\"min_value\": 0,\n\"refresh_period\": 86400,\n\"lookback_days\": 1\n}\n],\n

More sophisticated approach can be used, by using lookback_timeframe for candle size and lookback_period which specifies the amount of candles. This example will build the volume pairs based on a rolling period of 3 days of 1h candles:

\"pairlists\": [\n{\n\"method\": \"VolumePairList\",\n\"number_assets\": 20,\n\"sort_key\": \"quoteVolume\",\n\"min_value\": 0,\n\"refresh_period\": 3600,\n\"lookback_timeframe\": \"1h\",\n\"lookback_period\": 72\n}\n],\n

Note

VolumePairList does not support backtesting mode.

"},{"location":"plugins/#producerpairlist","title":"ProducerPairList","text":"

With ProducerPairList, you can reuse the pairlist from a Producer without explicitly defining the pairlist on each consumer.

Consumer mode is required for this pairlist to work.

The pairlist will perform a check on active pairs against the current exchange configuration to avoid attempting to trade on invalid markets.

You can limit the length of the pairlist with the optional parameter number_assets. Using \"number_assets\"=0 or omitting this key will result in the reuse of all producer pairs valid for the current setup.

\"pairlists\": [\n{\n\"method\": \"ProducerPairList\",\n\"number_assets\": 5,\n\"producer_name\": \"default\",\n}\n],\n

Combining pairlists

This pairlist can be combined with all other pairlists and filters for further pairlist reduction, and can also act as an \"additional\" pairlist, on top of already defined pairs. ProducerPairList can also be used multiple times in sequence, combining the pairs from multiple producers. Obviously in complex such configurations, the Producer may not provide data for all pairs, so the strategy must be fit for this.

"},{"location":"plugins/#remotepairlist","title":"RemotePairList","text":"

It allows the user to fetch a pairlist from a remote server or a locally stored json file within the freqtrade directory, enabling dynamic updates and customization of the trading pairlist.

The RemotePairList is defined in the pairlists section of the configuration settings. It uses the following configuration options:

\"pairlists\": [\n{\n\"method\": \"RemotePairList\",\n\"pairlist_url\": \"https://example.com/pairlist\",\n\"number_assets\": 10,\n\"refresh_period\": 1800,\n\"keep_pairlist_on_failure\": true,\n\"read_timeout\": 60,\n\"bearer_token\": \"my-bearer-token\"\n}\n]\n

The pairlist_url option specifies the URL of the remote server where the pairlist is located, or the path to a local file (if file:/// is prepended). This allows the user to use either a remote server or a local file as the source for the pairlist.

The user is responsible for providing a server or local file that returns a JSON object with the following structure:

{\n\"pairs\": [\"XRP/USDT\", \"ETH/USDT\", \"LTC/USDT\"],\n\"refresh_period\": 1800,\n}\n

The pairs property should contain a list of strings with the trading pairs to be used by the bot. The refresh_period property is optional and specifies the number of seconds that the pairlist should be cached before being refreshed.

The optional keep_pairlist_on_failure specifies whether the previous received pairlist should be used if the remote server is not reachable or returns an error. The default value is true.

The optional read_timeout specifies the maximum amount of time (in seconds) to wait for a response from the remote source, The default value is 60.

The optional bearer_token will be included in the requests Authorization Header.

Note

In case of a server error the last received pairlist will be kept if keep_pairlist_on_failure is set to true, when set to false a empty pairlist is returned.

"},{"location":"plugins/#agefilter","title":"AgeFilter","text":"

Removes pairs that have been listed on the exchange for less than min_days_listed days (defaults to 10) or more than max_days_listed days (defaults None mean infinity).

When pairs are first listed on an exchange they can suffer huge price drops and volatility in the first few days while the pair goes through its price-discovery period. Bots can often be caught out buying before the pair has finished dropping in price.

This filter allows freqtrade to ignore pairs until they have been listed for at least min_days_listed days and listed before max_days_listed.

"},{"location":"plugins/#offsetfilter","title":"OffsetFilter","text":"

Offsets an incoming pairlist by a given offset value.

As an example it can be used in conjunction with VolumeFilter to remove the top X volume pairs. Or to split a larger pairlist on two bot instances.

Example to remove the first 10 pairs from the pairlist, and takes the next 20 (taking items 10-30 of the initial list):

\"pairlists\": [\n// ...\n{\n\"method\": \"OffsetFilter\",\n\"offset\": 10,\n\"number_assets\": 20\n}\n],\n

Warning

When OffsetFilter is used to split a larger pairlist among multiple bots in combination with VolumeFilter it can not be guaranteed that pairs won't overlap due to slightly different refresh intervals for the VolumeFilter.

Note

An offset larger than the total length of the incoming pairlist will result in an empty pairlist.

"},{"location":"plugins/#performancefilter","title":"PerformanceFilter","text":"

Sorts pairs by past trade performance, as follows:

  1. Positive performance.
  2. No closed trades yet.
  3. Negative performance.

Trade count is used as a tie breaker.

You can use the minutes parameter to only consider performance of the past X minutes (rolling window). Not defining this parameter (or setting it to 0) will use all-time performance.

The optional min_profit (as ratio -> a setting of 0.01 corresponds to 1%) parameter defines the minimum profit a pair must have to be considered. Pairs below this level will be filtered out. Using this parameter without minutes is highly discouraged, as it can lead to an empty pairlist without a way to recover.

\"pairlists\": [\n// ...\n{\n\"method\": \"PerformanceFilter\",\n\"minutes\": 1440,  // rolling 24h\n\"min_profit\": 0.01  // minimal profit 1%\n}\n],\n

As this Filter uses past performance of the bot, it'll have some startup-period - and should only be used after the bot has a few 100 trades in the database.

Backtesting

PerformanceFilter does not support backtesting mode.

"},{"location":"plugins/#precisionfilter","title":"PrecisionFilter","text":"

Filters low-value coins which would not allow setting stoplosses.

Backtesting

PrecisionFilter does not support backtesting mode using multiple strategies.

"},{"location":"plugins/#pricefilter","title":"PriceFilter","text":"

The PriceFilter allows filtering of pairs by price. Currently the following price filters are supported:

  • min_price
  • max_price
  • max_value
  • low_price_ratio

The min_price setting removes pairs where the price is below the specified price. This is useful if you wish to avoid trading very low-priced pairs. This option is disabled by default, and will only apply if set to > 0.

The max_price setting removes pairs where the price is above the specified price. This is useful if you wish to trade only low-priced pairs. This option is disabled by default, and will only apply if set to > 0.

The max_value setting removes pairs where the minimum value change is above a specified value. This is useful when an exchange has unbalanced limits. For example, if step-size = 1 (so you can only buy 1, or 2, or 3, but not 1.1 Coins) - and the price is pretty high (like 20$) as the coin has risen sharply since the last limit adaption. As a result of the above, you can only buy for 20$, or 40$ - but not for 25$. On exchanges that deduct fees from the receiving currency (e.g. binance) - this can result in high value coins / amounts that are unsellable as the amount is slightly below the limit.

The low_price_ratio setting removes pairs where a raise of 1 price unit (pip) is above the low_price_ratio ratio. This option is disabled by default, and will only apply if set to > 0.

For PriceFilter at least one of its min_price, max_price or low_price_ratio settings must be applied.

Calculation example:

Min price precision for SHITCOIN/BTC is 8 decimals. If its price is 0.00000011 - one price step above would be 0.00000012, which is ~9% higher than the previous price value. You may filter out this pair by using PriceFilter with low_price_ratio set to 0.09 (9%) or with min_price set to 0.00000011, correspondingly.

Low priced pairs

Low priced pairs with high \"1 pip movements\" are dangerous since they are often illiquid and it may also be impossible to place the desired stoploss, which can often result in high losses since price needs to be rounded to the next tradable price - so instead of having a stoploss of -5%, you could end up with a stoploss of -9% simply due to price rounding.

"},{"location":"plugins/#shufflefilter","title":"ShuffleFilter","text":"

Shuffles (randomizes) pairs in the pairlist. It can be used for preventing the bot from trading some of the pairs more frequently then others when you want all pairs be treated with the same priority.

By default, ShuffleFilter will shuffle pairs once per candle. To shuffle on every iteration, set \"shuffle_frequency\" to \"iteration\" instead of the default of \"candle\".

    {\n\"method\": \"ShuffleFilter\", \"shuffle_frequency\": \"candle\",\n\"seed\": 42\n}\n

Tip

You may set the seed value for this Pairlist to obtain reproducible results, which can be useful for repeated backtesting sessions. If seed is not set, the pairs are shuffled in the non-repeatable random order. ShuffleFilter will automatically detect runmodes and apply the seed only for backtesting modes - if a seed value is set.

"},{"location":"plugins/#spreadfilter","title":"SpreadFilter","text":"

Removes pairs that have a difference between asks and bids above the specified ratio, max_spread_ratio (defaults to 0.005).

Example:

If DOGE/BTC maximum bid is 0.00000026 and minimum ask is 0.00000027, the ratio is calculated as: 1 - bid/ask ~= 0.037 which is > 0.005 and this pair will be filtered out.

"},{"location":"plugins/#rangestabilityfilter","title":"RangeStabilityFilter","text":"

Removes pairs where the difference between lowest low and highest high over lookback_days days is below min_rate_of_change or above max_rate_of_change. Since this is a filter that requires additional data, the results are cached for refresh_period.

In the below example: If the trading range over the last 10 days is <1% or >99%, remove the pair from the whitelist.

\"pairlists\": [\n{\n\"method\": \"RangeStabilityFilter\",\n\"lookback_days\": 10,\n\"min_rate_of_change\": 0.01,\n\"max_rate_of_change\": 0.99,\n\"refresh_period\": 1440\n}\n]\n

Tip

This Filter can be used to automatically remove stable coin pairs, which have a very low trading range, and are therefore extremely difficult to trade with profit. Additionally, it can also be used to automatically remove pairs with extreme high/low variance over a given amount of time.

"},{"location":"plugins/#volatilityfilter","title":"VolatilityFilter","text":"

Volatility is the degree of historical variation of a pairs over time, it is measured by the standard deviation of logarithmic daily returns. Returns are assumed to be normally distributed, although actual distribution might be different. In a normal distribution, 68% of observations fall within one standard deviation and 95% of observations fall within two standard deviations. Assuming a volatility of 0.05 means that the expected returns for 20 out of 30 days is expected to be less than 5% (one standard deviation). Volatility is a positive ratio of the expected deviation of return and can be greater than 1.00. Please refer to the wikipedia definition of volatility.

This filter removes pairs if the average volatility over a lookback_days days is below min_volatility or above max_volatility. Since this is a filter that requires additional data, the results are cached for refresh_period.

This filter can be used to narrow down your pairs to a certain volatility or avoid very volatile pairs.

In the below example: If the volatility over the last 10 days is not in the range of 0.05-0.50, remove the pair from the whitelist. The filter is applied every 24h.

\"pairlists\": [\n{\n\"method\": \"VolatilityFilter\",\n\"lookback_days\": 10,\n\"min_volatility\": 0.05,\n\"max_volatility\": 0.50,\n\"refresh_period\": 86400\n}\n]\n
"},{"location":"plugins/#full-example-of-pairlist-handlers","title":"Full example of Pairlist Handlers","text":"

The below example blacklists BNB/BTC, uses VolumePairList with 20 assets, sorting pairs by quoteVolume and applies PrecisionFilter and PriceFilter, filtering all assets where 1 price unit is > 1%. Then the SpreadFilter and VolatilityFilter is applied and pairs are finally shuffled with the random seed set to some predefined value.

\"exchange\": {\n\"pair_whitelist\": [],\n\"pair_blacklist\": [\"BNB/BTC\"]\n},\n\"pairlists\": [\n{\n\"method\": \"VolumePairList\",\n\"number_assets\": 20,\n\"sort_key\": \"quoteVolume\"\n},\n{\"method\": \"AgeFilter\", \"min_days_listed\": 10},\n{\"method\": \"PrecisionFilter\"},\n{\"method\": \"PriceFilter\", \"low_price_ratio\": 0.01},\n{\"method\": \"SpreadFilter\", \"max_spread_ratio\": 0.005},\n{\n\"method\": \"RangeStabilityFilter\",\n\"lookback_days\": 10,\n\"min_rate_of_change\": 0.01,\n\"refresh_period\": 1440\n},\n{\n\"method\": \"VolatilityFilter\",\n\"lookback_days\": 10,\n\"min_volatility\": 0.05,\n\"max_volatility\": 0.50,\n\"refresh_period\": 86400\n},\n{\"method\": \"ShuffleFilter\", \"seed\": 42}\n],\n
"},{"location":"plugins/#protections","title":"Protections","text":"

Beta feature

This feature is still in it's testing phase. Should you notice something you think is wrong please let us know via Discord or via Github Issue.

Protections will protect your strategy from unexpected events and market conditions by temporarily stop trading for either one pair, or for all pairs. All protection end times are rounded up to the next candle to avoid sudden, unexpected intra-candle buys.

Note

Not all Protections will work for all strategies, and parameters will need to be tuned for your strategy to improve performance.

Tip

Each Protection can be configured multiple times with different parameters, to allow different levels of protection (short-term / long-term).

Backtesting

Protections are supported by backtesting and hyperopt, but must be explicitly enabled by using the --enable-protections flag.

Setting protections from the configuration

Setting protections from the configuration via \"protections\": [], key should be considered deprecated and will be removed in a future version. It is also no longer guaranteed that your protections apply to the strategy in cases where the strategy defines protections as property.

"},{"location":"plugins/#available-protections","title":"Available Protections","text":"
  • StoplossGuard Stop trading if a certain amount of stoploss occurred within a certain time window.
  • MaxDrawdown Stop trading if max-drawdown is reached.
  • LowProfitPairs Lock pairs with low profits
  • CooldownPeriod Don't enter a trade right after selling a trade.
"},{"location":"plugins/#common-settings-to-all-protections","title":"Common settings to all Protections","text":"Parameter Description method Protection name to use. Datatype: String, selected from available Protections stop_duration_candles For how many candles should the lock be set? Datatype: Positive integer (in candles) stop_duration how many minutes should protections be locked. Cannot be used together with stop_duration_candles. Datatype: Float (in minutes) lookback_period_candles Only trades that completed within the last lookback_period_candles candles will be considered. This setting may be ignored by some Protections. Datatype: Positive integer (in candles). lookback_period Only trades that completed after current_time - lookback_period will be considered. Cannot be used together with lookback_period_candles. This setting may be ignored by some Protections. Datatype: Float (in minutes) trade_limit Number of trades required at minimum (not used by all Protections). Datatype: Positive integer

Durations

Durations (stop_duration* and lookback_period* can be defined in either minutes or candles). For more flexibility when testing different timeframes, all below examples will use the \"candle\" definition.

"},{"location":"plugins/#stoploss-guard","title":"Stoploss Guard","text":"

StoplossGuard selects all trades within lookback_period in minutes (or in candles when using lookback_period_candles). If trade_limit or more trades resulted in stoploss, trading will stop for stop_duration in minutes (or in candles when using stop_duration_candles).

This applies across all pairs, unless only_per_pair is set to true, which will then only look at one pair at a time.

Similarly, this protection will by default look at all trades (long and short). For futures bots, setting only_per_side will make the bot only consider one side, and will then only lock this one side, allowing for example shorts to continue after a series of long stoplosses.

required_profit will determine the required relative profit (or loss) for stoplosses to consider. This should normally not be set and defaults to 0.0 - which means all losing stoplosses will be triggering a block.

The below example stops trading for all pairs for 4 candles after the last trade if the bot hit stoploss 4 times within the last 24 candles.

@property\ndef protections(self):\n    return [\n        {\n            \"method\": \"StoplossGuard\",\n            \"lookback_period_candles\": 24,\n            \"trade_limit\": 4,\n            \"stop_duration_candles\": 4,\n            \"required_profit\": 0.0,\n            \"only_per_pair\": False,\n            \"only_per_side\": False\n        }\n    ]\n

Note

StoplossGuard considers all trades with the results \"stop_loss\", \"stoploss_on_exchange\" and \"trailing_stop_loss\" if the resulting profit was negative. trade_limit and lookback_period will need to be tuned for your strategy.

"},{"location":"plugins/#maxdrawdown","title":"MaxDrawdown","text":"

MaxDrawdown uses all trades within lookback_period in minutes (or in candles when using lookback_period_candles) to determine the maximum drawdown. If the drawdown is below max_allowed_drawdown, trading will stop for stop_duration in minutes (or in candles when using stop_duration_candles) after the last trade - assuming that the bot needs some time to let markets recover.

The below sample stops trading for 12 candles if max-drawdown is > 20% considering all pairs - with a minimum of trade_limit trades - within the last 48 candles. If desired, lookback_period and/or stop_duration can be used.

@property\ndef protections(self):\n    return  [\n        {\n            \"method\": \"MaxDrawdown\",\n            \"lookback_period_candles\": 48,\n            \"trade_limit\": 20,\n            \"stop_duration_candles\": 12,\n            \"max_allowed_drawdown\": 0.2\n        },\n    ]\n
"},{"location":"plugins/#low-profit-pairs","title":"Low Profit Pairs","text":"

LowProfitPairs uses all trades for a pair within lookback_period in minutes (or in candles when using lookback_period_candles) to determine the overall profit ratio. If that ratio is below required_profit, that pair will be locked for stop_duration in minutes (or in candles when using stop_duration_candles).

For futures bots, setting only_per_side will make the bot only consider one side, and will then only lock this one side, allowing for example shorts to continue after a series of long losses.

The below example will stop trading a pair for 60 minutes if the pair does not have a required profit of 2% (and a minimum of 2 trades) within the last 6 candles.

@property\ndef protections(self):\n    return [\n        {\n            \"method\": \"LowProfitPairs\",\n            \"lookback_period_candles\": 6,\n            \"trade_limit\": 2,\n            \"stop_duration\": 60,\n            \"required_profit\": 0.02,\n            \"only_per_pair\": False,\n        }\n    ]\n
"},{"location":"plugins/#cooldown-period","title":"Cooldown Period","text":"

CooldownPeriod locks a pair for stop_duration in minutes (or in candles when using stop_duration_candles) after selling, avoiding a re-entry for this pair for stop_duration minutes.

The below example will stop trading a pair for 2 candles after closing a trade, allowing this pair to \"cool down\".

@property\ndef protections(self):\n    return  [\n        {\n            \"method\": \"CooldownPeriod\",\n            \"stop_duration_candles\": 2\n        }\n    ]\n

Note

This Protection applies only at pair-level, and will never lock all pairs globally. This Protection does not consider lookback_period as it only looks at the latest trade.

"},{"location":"plugins/#full-example-of-protections","title":"Full example of Protections","text":"

All protections can be combined at will, also with different parameters, creating a increasing wall for under-performing pairs. All protections are evaluated in the sequence they are defined.

The below example assumes a timeframe of 1 hour:

  • Locks each pair after selling for an additional 5 candles (CooldownPeriod), giving other pairs a chance to get filled.
  • Stops trading for 4 hours (4 * 1h candles) if the last 2 days (48 * 1h candles) had 20 trades, which caused a max-drawdown of more than 20%. (MaxDrawdown).
  • Stops trading if more than 4 stoploss occur for all pairs within a 1 day (24 * 1h candles) limit (StoplossGuard).
  • Locks all pairs that had 2 Trades within the last 6 hours (6 * 1h candles) with a combined profit ratio of below 0.02 (<2%) (LowProfitPairs).
  • Locks all pairs for 2 candles that had a profit of below 0.01 (<1%) within the last 24h (24 * 1h candles), a minimum of 4 trades.
from freqtrade.strategy import IStrategy\n\nclass AwesomeStrategy(IStrategy)\n    timeframe = '1h'\n\n    @property\n    def protections(self):\n        return [\n            {\n                \"method\": \"CooldownPeriod\",\n                \"stop_duration_candles\": 5\n            },\n            {\n                \"method\": \"MaxDrawdown\",\n                \"lookback_period_candles\": 48,\n                \"trade_limit\": 20,\n                \"stop_duration_candles\": 4,\n                \"max_allowed_drawdown\": 0.2\n            },\n            {\n                \"method\": \"StoplossGuard\",\n                \"lookback_period_candles\": 24,\n                \"trade_limit\": 4,\n                \"stop_duration_candles\": 2,\n                \"only_per_pair\": False\n            },\n            {\n                \"method\": \"LowProfitPairs\",\n                \"lookback_period_candles\": 6,\n                \"trade_limit\": 2,\n                \"stop_duration_candles\": 60,\n                \"required_profit\": 0.02\n            },\n            {\n                \"method\": \"LowProfitPairs\",\n                \"lookback_period_candles\": 24,\n                \"trade_limit\": 4,\n                \"stop_duration_candles\": 2,\n                \"required_profit\": 0.01\n            }\n        ]\n    # ...\n
"},{"location":"producer-consumer/","title":"Producer / Consumer mode","text":"

freqtrade provides a mechanism whereby an instance (also called consumer) may listen to messages from an upstream freqtrade instance (also called producer) using the message websocket. Mainly, analyzed_df and whitelist messages. This allows the reuse of computed indicators (and signals) for pairs in multiple bots without needing to compute them multiple times.

See Message Websocket in the Rest API docs for setting up the api_server configuration for your message websocket (this will be your producer).

Note

We strongly recommend to set ws_token to something random and known only to yourself to avoid unauthorized access to your bot.

"},{"location":"producer-consumer/#configuration","title":"Configuration","text":"

Enable subscribing to an instance by adding the external_message_consumer section to the consumer's config file.

{\n//...\n\"external_message_consumer\": {\n\"enabled\": true,\n\"producers\": [\n{\n\"name\": \"default\", // This can be any name you'd like, default is \"default\"\n\"host\": \"127.0.0.1\", // The host from your producer's api_server config\n\"port\": 8080, // The port from your producer's api_server config\n\"secure\": false, // Use a secure websockets connection, default false\n\"ws_token\": \"sercet_Ws_t0ken\" // The ws_token from your producer's api_server config\n}\n],\n// The following configurations are optional, and usually not required\n// \"wait_timeout\": 300,\n// \"ping_timeout\": 10,\n// \"sleep_time\": 10,\n// \"remove_entry_exit_signals\": false,\n// \"message_size_limit\": 8\n}\n//...\n}\n
Parameter Description enabled Required. Enable consumer mode. If set to false, all other settings in this section are ignored.Defaults to false. Datatype: boolean . producers Required. List of producers Datatype: Array. producers.name Required. Name of this producer. This name must be used in calls to get_producer_pairs() and get_producer_df() if more than one producer is used. Datatype: string producers.host Required. The hostname or IP address from your producer. Datatype: string producers.port Required. The port matching the above host.Defaults to 8080. Datatype: Integer producers.secure Optional. Use ssl in websockets connection. Default False. Datatype: string producers.ws_token Required. ws_token as configured on the producer. Datatype: string Optional settings wait_timeout Timeout until we ping again if no message is received. Defaults to 300. Datatype: Integer - in seconds. ping_timeout Ping timeout Defaults to 10. Datatype: Integer - in seconds. sleep_time Sleep time before retrying to connect.Defaults to 10. Datatype: Integer - in seconds. remove_entry_exit_signals Remove signal columns from the dataframe (set them to 0) on dataframe receipt.Defaults to false. Datatype: Boolean. message_size_limit Size limit per messageDefaults to 8. Datatype: Integer - Megabytes.

Instead of (or as well as) calculating indicators in populate_indicators() the follower instance listens on the connection to a producer instance's messages (or multiple producer instances in advanced configurations) and requests the producer's most recently analyzed dataframes for each pair in the active whitelist.

A consumer instance will then have a full copy of the analyzed dataframes without the need to calculate them itself.

"},{"location":"producer-consumer/#examples","title":"Examples","text":""},{"location":"producer-consumer/#example-producer-strategy","title":"Example - Producer Strategy","text":"

A simple strategy with multiple indicators. No special considerations are required in the strategy itself.

class ProducerStrategy(IStrategy):\n    #...\n    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n\"\"\"\n        Calculate indicators in the standard freqtrade way which can then be broadcast to other instances\n        \"\"\"\n        dataframe['rsi'] = ta.RSI(dataframe)\n        bollinger = qtpylib.bollinger_bands(qtpylib.typical_price(dataframe), window=20, stds=2)\n        dataframe['bb_lowerband'] = bollinger['lower']\n        dataframe['bb_middleband'] = bollinger['mid']\n        dataframe['bb_upperband'] = bollinger['upper']\n        dataframe['tema'] = ta.TEMA(dataframe, timeperiod=9)\n\n        return dataframe\n\n    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n\"\"\"\n        Populates the entry signal for the given dataframe\n        \"\"\"\n        dataframe.loc[\n            (\n                (qtpylib.crossed_above(dataframe['rsi'], self.buy_rsi.value)) &\n                (dataframe['tema'] <= dataframe['bb_middleband']) &\n                (dataframe['tema'] > dataframe['tema'].shift(1)) &\n                (dataframe['volume'] > 0)\n            ),\n            'enter_long'] = 1\n\n        return dataframe\n

FreqAI

You can use this to setup FreqAI on a powerful machine, while you run consumers on simple machines like raspberries, which can interpret the signals generated from the producer in different ways.

"},{"location":"producer-consumer/#example-consumer-strategy","title":"Example - Consumer Strategy","text":"

A logically equivalent strategy which calculates no indicators itself, but will have the same analyzed dataframes available to make trading decisions based on the indicators calculated in the producer. In this example the consumer has the same entry criteria, however this is not necessary. The consumer may use different logic to enter/exit trades, and only use the indicators as specified.

class ConsumerStrategy(IStrategy):\n    #...\n    process_only_new_candles = False # required for consumers\n\n    _columns_to_expect = ['rsi_default', 'tema_default', 'bb_middleband_default']\n\n    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n\"\"\"\n        Use the websocket api to get pre-populated indicators from another freqtrade instance.\n        Use `self.dp.get_producer_df(pair)` to get the dataframe\n        \"\"\"\n        pair = metadata['pair']\n        timeframe = self.timeframe\n\n        producer_pairs = self.dp.get_producer_pairs()\n        # You can specify which producer to get pairs from via:\n        # self.dp.get_producer_pairs(\"my_other_producer\")\n\n        # This func returns the analyzed dataframe, and when it was analyzed\n        producer_dataframe, _ = self.dp.get_producer_df(pair)\n        # You can get other data if the producer makes it available:\n        # self.dp.get_producer_df(\n        #   pair,\n        #   timeframe=\"1h\",\n        #   candle_type=CandleType.SPOT,\n        #   producer_name=\"my_other_producer\"\n        # )\n\n        if not producer_dataframe.empty:\n            # If you plan on passing the producer's entry/exit signal directly,\n            # specify ffill=False or it will have unintended results\n            merged_dataframe = merge_informative_pair(dataframe, producer_dataframe,\n                                                      timeframe, timeframe,\n                                                      append_timeframe=False,\n                                                      suffix=\"default\")\n            return merged_dataframe\n        else:\n            dataframe[self._columns_to_expect] = 0\n\n        return dataframe\n\n    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n\"\"\"\n        Populates the entry signal for the given dataframe\n        \"\"\"\n        # Use the dataframe columns as if we calculated them ourselves\n        dataframe.loc[\n            (\n                (qtpylib.crossed_above(dataframe['rsi_default'], self.buy_rsi.value)) &\n                (dataframe['tema_default'] <= dataframe['bb_middleband_default']) &\n                (dataframe['tema_default'] > dataframe['tema_default'].shift(1)) &\n                (dataframe['volume'] > 0)\n            ),\n            'enter_long'] = 1\n\n        return dataframe\n

Using upstream signals

By setting remove_entry_exit_signals=false, you can also use the producer's signals directly. They should be available as enter_long_default (assuming suffix=\"default\" was used) - and can be used as either signal directly, or as additional indicator.

"},{"location":"rest-api/","title":"REST API & FreqUI","text":""},{"location":"rest-api/#frequi","title":"FreqUI","text":"

Freqtrade provides a builtin webserver, which can serve FreqUI, the freqtrade UI.

By default, the UI is not included in the installation (except for docker images), and must be installed explicitly with freqtrade install-ui. This same command can also be used to update freqUI, should there be a new release.

Once the bot is started in trade / dry-run mode (with freqtrade trade) - the UI will be available under the configured port below (usually http://127.0.0.1:8080).

developers

Developers should not use this method, but instead use the method described in the freqUI repository to get the source-code of freqUI.

"},{"location":"rest-api/#configuration","title":"Configuration","text":"

Enable the rest API by adding the api_server section to your configuration and setting api_server.enabled to true.

Sample configuration:

    \"api_server\": {\n\"enabled\": true,\n\"listen_ip_address\": \"127.0.0.1\",\n\"listen_port\": 8080,\n\"verbosity\": \"error\",\n\"enable_openapi\": false,\n\"jwt_secret_key\": \"somethingrandom\",\n\"CORS_origins\": [],\n\"username\": \"Freqtrader\",\n\"password\": \"SuperSecret1!\",\n\"ws_token\": \"sercet_Ws_t0ken\"\n},\n

Security warning

By default, the configuration listens on localhost only (so it's not reachable from other systems). We strongly recommend to not expose this API to the internet and choose a strong, unique password, since others will potentially be able to control your bot.

API/UI Access on a remote servers

If you're running on a VPS, you should consider using either a ssh tunnel, or setup a VPN (openVPN, wireguard) to connect to your bot. This will ensure that freqUI is not directly exposed to the internet, which is not recommended for security reasons (freqUI does not support https out of the box). Setup of these tools is not part of this tutorial, however many good tutorials can be found on the internet.

You can then access the API by going to http://127.0.0.1:8080/api/v1/ping in a browser to check if the API is running correctly. This should return the response:

{\"status\":\"pong\"}\n

All other endpoints return sensitive info and require authentication and are therefore not available through a web browser.

"},{"location":"rest-api/#security","title":"Security","text":"

To generate a secure password, best use a password manager, or use the below code.

import secrets\nsecrets.token_hex()\n

JWT token

Use the same method to also generate a JWT secret key (jwt_secret_key).

Password selection

Please make sure to select a very strong, unique password to protect your bot from unauthorized access. Also change jwt_secret_key to something random (no need to remember this, but it'll be used to encrypt your session, so it better be something unique!).

"},{"location":"rest-api/#configuration-with-docker","title":"Configuration with docker","text":"

If you run your bot using docker, you'll need to have the bot listen to incoming connections. The security is then handled by docker.

    \"api_server\": {\n\"enabled\": true,\n\"listen_ip_address\": \"0.0.0.0\",\n\"listen_port\": 8080,\n\"username\": \"Freqtrader\",\n\"password\": \"SuperSecret1!\",\n//...\n},\n

Make sure that the following 2 lines are available in your docker-compose file:

    ports:\n      - \"127.0.0.1:8080:8080\"\n

Security warning

By using 8080:8080 in the docker port mapping, the API will be available to everyone connecting to the server under the correct port, so others may be able to control your bot.

"},{"location":"rest-api/#rest-api","title":"Rest API","text":""},{"location":"rest-api/#consuming-the-api","title":"Consuming the API","text":"

You can consume the API by using the script scripts/rest_client.py. The client script only requires the requests module, so Freqtrade does not need to be installed on the system.

python3 scripts/rest_client.py <command> [optional parameters]\n

By default, the script assumes 127.0.0.1 (localhost) and port 8080 to be used, however you can specify a configuration file to override this behaviour.

"},{"location":"rest-api/#minimalistic-client-config","title":"Minimalistic client config","text":"
{\n\"api_server\": {\n\"enabled\": true,\n\"listen_ip_address\": \"0.0.0.0\",\n\"listen_port\": 8080,\n\"username\": \"Freqtrader\",\n\"password\": \"SuperSecret1!\",\n//...\n}\n}\n
python3 scripts/rest_client.py --config rest_config.json <command> [optional parameters]\n
"},{"location":"rest-api/#available-endpoints","title":"Available endpoints","text":"Command Description ping Simple command testing the API Readiness - requires no authentication. start Starts the trader. stop Stops the trader. stopbuy Stops the trader from opening new trades. Gracefully closes open trades according to their rules. reload_config Reloads the configuration file. trades List last trades. Limited to 500 trades per call. trade/<tradeid> Get specific trade. delete_trade <trade_id> Remove trade from the database. Tries to close open orders. Requires manual handling of this trade on the exchange. show_config Shows part of the current configuration with relevant settings to operation. logs Shows last log messages. status Lists all open trades. count Displays number of trades used and available. locks Displays currently locked pairs. delete_lock <lock_id> Deletes (disables) the lock by id. profit Display a summary of your profit/loss from close trades and some stats about your performance. forceexit <trade_id> Instantly exits the given trade (Ignoring minimum_roi). forceexit all Instantly exits all open trades (Ignoring minimum_roi). forceenter <pair> [rate] Instantly enters the given pair. Rate is optional. (force_entry_enable must be set to True) forceenter <pair> <side> [rate] Instantly longs or shorts the given pair. Rate is optional. (force_entry_enable must be set to True) performance Show performance of each finished trade grouped by pair. balance Show account balance per currency. daily <n> Shows profit or loss per day, over the last n days (n defaults to 7). stats Display a summary of profit / loss reasons as well as average holding times. whitelist Show the current whitelist. blacklist [pair] Show the current blacklist, or adds a pair to the blacklist. edge Show validated pairs by Edge if it is enabled. pair_candles Returns dataframe for a pair / timeframe combination while the bot is running. Alpha pair_history Returns an analyzed dataframe for a given timerange, analyzed by a given strategy. Alpha plot_config Get plot config from the strategy (or nothing if not configured). Alpha strategies List strategies in strategy directory. Alpha strategy <strategy> Get specific Strategy content. Alpha available_pairs List available backtest data. Alpha version Show version. sysinfo Show information about the system load. health Show bot health (last bot loop).

Alpha status

Endpoints labeled with Alpha status above may change at any time without notice.

Possible commands can be listed from the rest-client script using the help command.

python3 scripts/rest_client.py help\n
Possible commands:\n\navailable_pairs\n    Return available pair (backtest data) based on timeframe / stake_currency selection\n\n        :param timeframe: Only pairs with this timeframe available.\n        :param stake_currency: Only pairs that include this timeframe\n\nbalance\n    Get the account balance.\n\nblacklist\n    Show the current blacklist.\n\n        :param add: List of coins to add (example: \"BNB/BTC\")\n\ncancel_open_order\n    Cancel open order for trade.\n\n        :param trade_id: Cancels open orders for this trade.\n\ncount\n    Return the amount of open trades.\n\ndaily\n    Return the profits for each day, and amount of trades.\n\ndelete_lock\n    Delete (disable) lock from the database.\n\n        :param lock_id: ID for the lock to delete\n\ndelete_trade\n    Delete trade from the database.\n        Tries to close open orders. Requires manual handling of this asset on the exchange.\n\n        :param trade_id: Deletes the trade with this ID from the database.\n\nedge\n    Return information about edge.\n\nforcebuy\n    Buy an asset.\n\n        :param pair: Pair to buy (ETH/BTC)\n        :param price: Optional - price to buy\n\nforceenter\n    Force entering a trade\n\n        :param pair: Pair to buy (ETH/BTC)\n        :param side: 'long' or 'short'\n        :param price: Optional - price to buy\n\nforceexit\n    Force-exit a trade.\n\n        :param tradeid: Id of the trade (can be received via status command)\n        :param ordertype: Order type to use (must be market or limit)\n        :param amount: Amount to sell. Full sell if not given\n\nhealth\n    Provides a quick health check of the running bot.\n\nlocks\n    Return current locks\n\nlogs\n    Show latest logs.\n\n        :param limit: Limits log messages to the last <limit> logs. No limit to get the entire log.\n\npair_candles\n    Return live dataframe for <pair><timeframe>.\n\n        :param pair: Pair to get data for\n        :param timeframe: Only pairs with this timeframe available.\n        :param limit: Limit result to the last n candles.\n\npair_history\n    Return historic, analyzed dataframe\n\n        :param pair: Pair to get data for\n        :param timeframe: Only pairs with this timeframe available.\n        :param strategy: Strategy to analyze and get values for\n        :param timerange: Timerange to get data for (same format than --timerange endpoints)\n\nperformance\n    Return the performance of the different coins.\n\nping\n    simple ping\n\nplot_config\n    Return plot configuration if the strategy defines one.\n\nprofit\n    Return the profit summary.\n\nreload_config\n    Reload configuration.\n\nshow_config\n        Returns part of the configuration, relevant for trading operations.\n\nstart\n    Start the bot if it's in the stopped state.\n\nstats\n    Return the stats report (durations, sell-reasons).\n\nstatus\n    Get the status of open trades.\n\nstop\n    Stop the bot. Use `start` to restart.\n\nstopbuy\n    Stop buying (but handle sells gracefully). Use `reload_config` to reset.\n\nstrategies\n    Lists available strategies\n\nstrategy\n    Get strategy details\n\n        :param strategy: Strategy class name\n\nsysinfo\n    Provides system information (CPU, RAM usage)\n\ntrade\n    Return specific trade\n\n        :param trade_id: Specify which trade to get.\n\ntrades\n    Return trades history, sorted by id\n\n        :param limit: Limits trades to the X last trades. Max 500 trades.\n        :param offset: Offset by this amount of trades.\n\nversion\n    Return the version of the bot.\n\nwhitelist\n    Show the current whitelist.\n
"},{"location":"rest-api/#message-websocket","title":"Message WebSocket","text":"

The API Server includes a websocket endpoint for subscribing to RPC messages from the freqtrade Bot. This can be used to consume real-time data from your bot, such as entry/exit fill messages, whitelist changes, populated indicators for pairs, and more.

This is also used to setup Producer/Consumer mode in Freqtrade.

Assuming your rest API is set to 127.0.0.1 on port 8080, the endpoint is available at http://localhost:8080/api/v1/message/ws.

To access the websocket endpoint, the ws_token is required as a query parameter in the endpoint URL.

To generate a safe ws_token you can run the following code:

>>> import secrets\n>>> secrets.token_urlsafe(25)\n'hZ-y58LXyX_HZ8O1cJzVyN6ePWrLpNQv4Q'\n

You would then add that token under ws_token in your api_server config. Like so:

\"api_server\": {\n\"enabled\": true,\n\"listen_ip_address\": \"127.0.0.1\",\n\"listen_port\": 8080,\n\"verbosity\": \"error\",\n\"enable_openapi\": false,\n\"jwt_secret_key\": \"somethingrandom\",\n\"CORS_origins\": [],\n\"username\": \"Freqtrader\",\n\"password\": \"SuperSecret1!\",\n\"ws_token\": \"hZ-y58LXyX_HZ8O1cJzVyN6ePWrLpNQv4Q\" // <-----\n},\n

You can now connect to the endpoint at http://localhost:8080/api/v1/message/ws?token=hZ-y58LXyX_HZ8O1cJzVyN6ePWrLpNQv4Q.

Reuse of example tokens

Please do not use the above example token. To make sure you are secure, generate a completely new token.

"},{"location":"rest-api/#using-the-websocket","title":"Using the WebSocket","text":"

Once connected to the WebSocket, the bot will broadcast RPC messages to anyone who is subscribed to them. To subscribe to a list of messages, you must send a JSON request through the WebSocket like the one below. The data key must be a list of message type strings.

{\n\"type\": \"subscribe\",\n\"data\": [\"whitelist\", \"analyzed_df\"] // A list of string message types\n}\n

For a list of message types, please refer to the RPCMessageType enum in freqtrade/enums/rpcmessagetype.py

Now anytime those types of RPC messages are sent in the bot, you will receive them through the WebSocket as long as the connection is active. They typically take the same form as the request:

{\n\"type\": \"analyzed_df\",\n\"data\": {\n\"key\": [\"NEO/BTC\", \"5m\", \"spot\"],\n\"df\": {}, // The dataframe\n\"la\": \"2022-09-08 22:14:41.457786+00:00\"\n}\n}\n
"},{"location":"rest-api/#reverse-proxy-setup","title":"Reverse Proxy setup","text":"

When using Nginx, the following configuration is required for WebSockets to work (Note this configuration is incomplete, it's missing some information and can not be used as is):

Please make sure to replace <freqtrade_listen_ip> (and the subsequent port) with the IP and Port matching your configuration/setup.

http {\n    map $http_upgrade $connection_upgrade {\n        default upgrade;\n        '' close;\n    }\n\n    #...\n\n    server {\n        #...\n\n        location / {\n            proxy_http_version 1.1;\n            proxy_pass http://<freqtrade_listen_ip>:8080;\n            proxy_set_header Upgrade $http_upgrade;\n            proxy_set_header Connection $connection_upgrade;\n            proxy_set_header Host $host;\n        }\n    }\n}\n

To properly configure your reverse proxy (securely), please consult it's documentation for proxying websockets.

  • Traefik: Traefik supports websockets out of the box, see the documentation
  • Caddy: Caddy v2 supports websockets out of the box, see the documentation

SSL certificates

You can use tools like certbot to setup ssl certificates to access your bot's UI through encrypted connection by using any fo the above reverse proxies. While this will protect your data in transit, we do not recommend to run the freqtrade API outside of your private network (VPN, SSH tunnel).

"},{"location":"rest-api/#openapi-interface","title":"OpenAPI interface","text":"

To enable the builtin openAPI interface (Swagger UI), specify \"enable_openapi\": true in the api_server configuration. This will enable the Swagger UI at the /docs endpoint. By default, that's running at http://localhost:8080/docs - but it'll depend on your settings.

"},{"location":"rest-api/#advanced-api-usage-using-jwt-tokens","title":"Advanced API usage using JWT tokens","text":"

Note

The below should be done in an application (a Freqtrade REST API client, which fetches info via API), and is not intended to be used on a regular basis.

Freqtrade's REST API also offers JWT (JSON Web Tokens). You can login using the following command, and subsequently use the resulting access_token.

> curl -X POST --user Freqtrader http://localhost:8080/api/v1/token/login\n{\"access_token\":\"eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpYXQiOjE1ODkxMTk2ODEsIm5iZiI6MTU4OTExOTY4MSwianRpIjoiMmEwYmY0NWUtMjhmOS00YTUzLTlmNzItMmM5ZWVlYThkNzc2IiwiZXhwIjoxNTg5MTIwNTgxLCJpZGVudGl0eSI6eyJ1IjoiRnJlcXRyYWRlciJ9LCJmcmVzaCI6ZmFsc2UsInR5cGUiOiJhY2Nlc3MifQ.qt6MAXYIa-l556OM7arBvYJ0SDI9J8bIk3_glDujF5g\",\"refresh_token\":\"eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpYXQiOjE1ODkxMTk2ODEsIm5iZiI6MTU4OTExOTY4MSwianRpIjoiZWQ1ZWI3YjAtYjMwMy00YzAyLTg2N2MtNWViMjIxNWQ2YTMxIiwiZXhwIjoxNTkxNzExNjgxLCJpZGVudGl0eSI6eyJ1IjoiRnJlcXRyYWRlciJ9LCJ0eXBlIjoicmVmcmVzaCJ9.d1AT_jYICyTAjD0fiQAr52rkRqtxCjUGEMwlNuuzgNQ\"}\n\n> access_token=\"eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpYXQiOjE1ODkxMTk2ODEsIm5iZiI6MTU4OTExOTY4MSwianRpIjoiMmEwYmY0NWUtMjhmOS00YTUzLTlmNzItMmM5ZWVlYThkNzc2IiwiZXhwIjoxNTg5MTIwNTgxLCJpZGVudGl0eSI6eyJ1IjoiRnJlcXRyYWRlciJ9LCJmcmVzaCI6ZmFsc2UsInR5cGUiOiJhY2Nlc3MifQ.qt6MAXYIa-l556OM7arBvYJ0SDI9J8bIk3_glDujF5g\"\n# Use access_token for authentication\n> curl -X GET --header \"Authorization: Bearer ${access_token}\" http://localhost:8080/api/v1/count\n

Since the access token has a short timeout (15 min) - the token/refresh request should be used periodically to get a fresh access token:

> curl -X POST --header \"Authorization: Bearer ${refresh_token}\"http://localhost:8080/api/v1/token/refresh\n{\"access_token\":\"eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpYXQiOjE1ODkxMTk5NzQsIm5iZiI6MTU4OTExOTk3NCwianRpIjoiMDBjNTlhMWUtMjBmYS00ZTk0LTliZjAtNWQwNTg2MTdiZDIyIiwiZXhwIjoxNTg5MTIwODc0LCJpZGVudGl0eSI6eyJ1IjoiRnJlcXRyYWRlciJ9LCJmcmVzaCI6ZmFsc2UsInR5cGUiOiJhY2Nlc3MifQ.1seHlII3WprjjclY6DpRhen0rqdF4j6jbvxIhUFaSbs\"}\n
"},{"location":"rest-api/#cors","title":"CORS","text":"

This whole section is only necessary in cross-origin cases (where you multiple bot API's running on localhost:8081, localhost:8082, ...), and want to combine them into one FreqUI instance.

Technical explanation

All web-based front-ends are subject to CORS - Cross-Origin Resource Sharing. Since most of the requests to the Freqtrade API must be authenticated, a proper CORS policy is key to avoid security problems. Also, the standard disallows * CORS policies for requests with credentials, so this setting must be set appropriately.

Users can allow access from different origin URL's to the bot API via the CORS_origins configuration setting. It consists of a list of allowed URL's that are allowed to consume resources from the bot's API.

Assuming your application is deployed as https://frequi.freqtrade.io/home/ - this would mean that the following configuration becomes necessary:

{\n    //...\n    \"jwt_secret_key\": \"somethingrandom\",\n    \"CORS_origins\": [\"https://frequi.freqtrade.io\"],\n    //...\n}\n

In the following (pretty common) case, FreqUI is accessible on http://localhost:8080/trade (this is what you see in your navbar when navigating to freqUI).

The correct configuration for this case is http://localhost:8080 - the main part of the URL including the port.

{\n    //...\n    \"jwt_secret_key\": \"somethingrandom\",\n    \"CORS_origins\": [\"http://localhost:8080\"],\n    //...\n}\n

Note

We strongly recommend to also set jwt_secret_key to something random and known only to yourself to avoid unauthorized access to your bot.

"},{"location":"sandbox-testing/","title":"Sandbox API testing","text":"

Some exchanges provide sandboxes or testbeds for risk-free testing, while running the bot against a real exchange. With some configuration, freqtrade (in combination with ccxt) provides access to these.

This document is an overview to configure Freqtrade to be used with sandboxes. This can be useful to developers and trader alike.

Warning

Sandboxes usually have very low volume, and either a very wide spread, or no orders available at all. Therefore, sandboxes will usually not do a good job of showing you how a strategy would work in real trading.

"},{"location":"sandbox-testing/#exchanges-known-to-have-a-sandbox-testnet","title":"Exchanges known to have a sandbox / testnet","text":"
  • binance
  • coinbasepro
  • gemini
  • huobipro
  • kucoin
  • phemex

Note

We did not test correct functioning of all of the above testnets. Please report your experiences with each sandbox.

"},{"location":"sandbox-testing/#configure-a-sandbox-account","title":"Configure a Sandbox account","text":"

When testing your API connectivity, make sure to use the appropriate sandbox / testnet URL.

In general, you should follow these steps to enable an exchange's sandbox:

  • Figure out if an exchange has a sandbox (most likely by using google or the exchange's support documents)
  • Create a sandbox account (often the sandbox-account requires separate registration)
  • Add some test assets to account
  • Create API keys
"},{"location":"sandbox-testing/#add-test-funds","title":"Add test funds","text":"

Usually, sandbox exchanges allow depositing funds directly via web-interface. You should make sure to have a realistic amount of funds available to your test-account, so results are representable of your real account funds.

Warning

Test exchanges will NEVER require your real credit card or banking details!

"},{"location":"sandbox-testing/#configure-freqtrade-to-use-a-exchanges-sandbox","title":"Configure freqtrade to use a exchange's sandbox","text":""},{"location":"sandbox-testing/#sandbox-urls","title":"Sandbox URLs","text":"

Freqtrade makes use of CCXT which in turn provides a list of URLs to Freqtrade. These include ['test'] and ['api'].

  • [Test] if available will point to an Exchanges sandbox.
  • [Api] normally used, and resolves to live API target on the exchange.

To make use of sandbox / test add \"sandbox\": true, to your config.json

  \"exchange\": {\n\"name\": \"coinbasepro\",\n\"sandbox\": true,\n\"key\": \"5wowfxemogxeowo;heiohgmd\",\n\"secret\": \"/ZMH1P62rCVmwefewrgcewX8nh4gob+lywxfwfxwwfxwfNsH1ySgvWCUR/w==\",\n\"password\": \"1bkjfkhfhfu6sr\",\n\"outdated_offset\": 5\n\"pair_whitelist\": [\n\"BTC/USD\"\n]\n},\n\"datadir\": \"user_data/data/coinbasepro_sandbox\"\n

Also the following information:

  • api-key (created for the sandbox webpage)
  • api-secret (noted earlier)
  • password (the passphrase - noted earlier)

Different data directory

We also recommend to set datadir to something identifying downloaded data as sandbox data, to avoid having sandbox data mixed with data from the real exchange. This can be done by adding the \"datadir\" key to the configuration. Now, whenever you use this configuration, your data directory will be set to this directory.

"},{"location":"sandbox-testing/#you-should-now-be-ready-to-test-your-sandbox","title":"You should now be ready to test your sandbox","text":"

Ensure Freqtrade logs show the sandbox URL, and trades made are shown in sandbox. Also make sure to select a pair which shows at least some decent value (which very often is BTC/)."},{"location":"sandbox-testing/#common-problems-with-sandbox-exchanges","title":"Common problems with sandbox exchanges","text":"

Sandbox exchange instances often have very low volume, which can cause some problems which usually are not seen on a real exchange instance.

"},{"location":"sandbox-testing/#old-candles-problem","title":"Old Candles problem","text":"

Since Sandboxes often have low volume, candles can be quite old and show no volume. To disable the error \"Outdated history for pair ...\", best increase the parameter \"outdated_offset\" to a number that seems realistic for the sandbox you're using.

"},{"location":"sandbox-testing/#unfilled-orders","title":"Unfilled orders","text":"

Sandboxes often have very low volumes - which means that many trades can go unfilled, or can go unfilled for a very long time.

To mitigate this, you can try to match the first order on the opposite orderbook side using the following configuration:

jsonc \"order_types\": { \"entry\": \"limit\", \"exit\": \"limit\" // ... }, \"entry_pricing\": { \"price_side\": \"other\", // ... }, \"exit_pricing\":{ \"price_side\": \"other\", // ... },

The configuration is similar to the suggested configuration for market orders - however by using limit-orders you can avoid moving the price too much, and you can set the worst price you might get.

"},{"location":"sql_cheatsheet/","title":"SQL Helper","text":"

This page contains some help if you want to edit your sqlite db.

"},{"location":"sql_cheatsheet/#install-sqlite3","title":"Install sqlite3","text":"

Sqlite3 is a terminal based sqlite application. Feel free to use a visual Database editor like SqliteBrowser if you feel more comfortable with that.

"},{"location":"sql_cheatsheet/#ubuntudebian-installation","title":"Ubuntu/Debian installation","text":"
sudo apt-get install sqlite3\n
"},{"location":"sql_cheatsheet/#using-sqlite3-via-docker","title":"Using sqlite3 via docker","text":"

The freqtrade docker image does contain sqlite3, so you can edit the database without having to install anything on the host system.

docker compose exec freqtrade /bin/bash\nsqlite3 <database-file>.sqlite\n
"},{"location":"sql_cheatsheet/#open-the-db","title":"Open the DB","text":"
sqlite3\n.open <filepath>\n
"},{"location":"sql_cheatsheet/#table-structure","title":"Table structure","text":""},{"location":"sql_cheatsheet/#list-tables","title":"List tables","text":"
.tables\n
"},{"location":"sql_cheatsheet/#display-table-structure","title":"Display table structure","text":"
.schema <table_name>\n
"},{"location":"sql_cheatsheet/#get-all-trades-in-the-table","title":"Get all trades in the table","text":"
SELECT * FROM trades;\n
"},{"location":"sql_cheatsheet/#fix-trade-still-open-after-a-manual-exit-on-the-exchange","title":"Fix trade still open after a manual exit on the exchange","text":"

Warning

Manually selling a pair on the exchange will not be detected by the bot and it will try to sell anyway. Whenever possible, /forceexit should be used to accomplish the same thing. It is strongly advised to backup your database file before making any manual changes.

Note

This should not be necessary after /forceexit, as force_exit orders are closed automatically by the bot on the next iteration.

UPDATE trades\nSET is_open=0,\nclose_date=<close_date>,\nclose_rate=<close_rate>,\nclose_profit = close_rate / open_rate - 1,\nclose_profit_abs = (amount * <close_rate> * (1 - fee_close) - (amount * (open_rate * (1 - fee_open)))),\nexit_reason=<exit_reason>\nWHERE id=<trade_ID_to_update>;\n
"},{"location":"sql_cheatsheet/#example","title":"Example","text":"
UPDATE trades\nSET is_open=0,\nclose_date='2020-06-20 03:08:45.103418',\nclose_rate=0.19638016,\nclose_profit=0.0496,\nclose_profit_abs = (amount * 0.19638016 * (1 - fee_close) - (amount * (open_rate * (1 - fee_open)))),\nexit_reason='force_exit'  WHERE id=31;\n
"},{"location":"sql_cheatsheet/#remove-trade-from-the-database","title":"Remove trade from the database","text":"

Use RPC Methods to delete trades

Consider using /delete <tradeid> via telegram or rest API. That's the recommended way to deleting trades.

If you'd still like to remove a trade from the database directly, you can use the below query.

Danger

Some systems (Ubuntu) disable foreign keys in their sqlite3 packaging. When using sqlite - please ensure that foreign keys are on by running PRAGMA foreign_keys = ON before the above query.

DELETE FROM trades WHERE id = <tradeid>;\n\nDELETE FROM trades WHERE id = 31;\n

Warning

This will remove this trade from the database. Please make sure you got the correct id and NEVER run this query without the where clause.

"},{"location":"sql_cheatsheet/#use-a-different-database-system","title":"Use a different database system","text":"

Freqtrade is using SQLAlchemy, which supports multiple different database systems. As such, a multitude of database systems should be supported. Freqtrade does not depend or install any additional database driver. Please refer to the SQLAlchemy docs on installation instructions for the respective database systems.

The following systems have been tested and are known to work with freqtrade:

  • sqlite (default)
  • PostgreSQL)
  • MariaDB

Warning

By using one of the below database systems, you acknowledge that you know how to manage such a system. The freqtrade team will not provide any support with setup or maintenance (or backups) of the below database systems.

"},{"location":"sql_cheatsheet/#postgresql","title":"PostgreSQL","text":"

Installation: pip install psycopg2-binary

Usage: ... --db-url postgresql+psycopg2://<username>:<password>@localhost:5432/<database>

Freqtrade will automatically create the tables necessary upon startup.

If you're running different instances of Freqtrade, you must either setup one database per Instance or use different users / schemas for your connections.

"},{"location":"sql_cheatsheet/#mariadb-mysql","title":"MariaDB / MySQL","text":"

Freqtrade supports MariaDB by using SQLAlchemy, which supports multiple different database systems.

Installation: pip install pymysql

Usage: ... --db-url mysql+pymysql://<username>:<password>@localhost:3306/<database>

"},{"location":"stoploss/","title":"Stop Loss","text":"

The stoploss configuration parameter is loss as ratio that should trigger a sale. For example, value -0.10 will cause immediate sell if the profit dips below -10% for a given trade. This parameter is optional. Stoploss calculations do include fees, so a stoploss of -10% is placed exactly 10% below the entry point.

Most of the strategy files already include the optimal stoploss value.

Info

All stoploss properties mentioned in this file can be set in the Strategy, or in the configuration. Configuration values will override the strategy values.

"},{"location":"stoploss/#stop-loss-on-exchangefreqtrade","title":"Stop Loss On-Exchange/Freqtrade","text":"

Those stoploss modes can be on exchange or off exchange.

These modes can be configured with these values:

    'emergency_exit': 'market',\n    'stoploss_on_exchange': False\n    'stoploss_on_exchange_interval': 60,\n    'stoploss_on_exchange_limit_ratio': 0.99\n

Stoploss on exchange is only supported for the following exchanges, and not all exchanges support both stop-limit and stop-market. The Order-type will be ignored if only one mode is available.

Exchange stop-loss type Binance limit Binance Futures market, limit Huobi limit kraken market, limit Gate limit Okx limit Kucoin stop-limit, stop-market

Tight stoploss

Do not set too low/tight stoploss value when using stop loss on exchange! If set to low/tight you will have greater risk of missing fill on the order and stoploss will not work.

"},{"location":"stoploss/#stoploss_on_exchange-and-stoploss_on_exchange_limit_ratio","title":"stoploss_on_exchange and stoploss_on_exchange_limit_ratio","text":"

Enable or Disable stop loss on exchange. If the stoploss is on exchange it means a stoploss limit order is placed on the exchange immediately after buy order fills. This will protect you against sudden crashes in market, as the order execution happens purely within the exchange, and has no potential network overhead.

If stoploss_on_exchange uses limit orders, the exchange needs 2 prices, the stoploss_price and the Limit price. stoploss defines the stop-price where the limit order is placed - and limit should be slightly below this. If an exchange supports both limit and market stoploss orders, then the value of stoploss will be used to determine the stoploss type.

Calculation example: we bought the asset at 100$. Stop-price is 95$, then limit would be 95 * 0.99 = 94.05$ - so the limit order fill can happen between 95$ and 94.05$.

For example, assuming the stoploss is on exchange, and trailing stoploss is enabled, and the market is going up, then the bot automatically cancels the previous stoploss order and puts a new one with a stop value higher than the previous stoploss order.

Note

If stoploss_on_exchange is enabled and the stoploss is cancelled manually on the exchange, then the bot will create a new stoploss order.

"},{"location":"stoploss/#stoploss_on_exchange_interval","title":"stoploss_on_exchange_interval","text":"

In case of stoploss on exchange there is another parameter called stoploss_on_exchange_interval. This configures the interval in seconds at which the bot will check the stoploss and update it if necessary. The bot cannot do these every 5 seconds (at each iteration), otherwise it would get banned by the exchange. So this parameter will tell the bot how often it should update the stoploss order. The default value is 60 (1 minute). This same logic will reapply a stoploss order on the exchange should you cancel it accidentally.

"},{"location":"stoploss/#stoploss_price_type","title":"stoploss_price_type","text":"

Only applies to futures

stoploss_price_type only applies to futures markets (on exchanges where it's available). Freqtrade will perform a validation of this setting on startup, failing to start if an invalid setting for your exchange has been selected. Supported price types are gonna differs between each exchanges. Please check with your exchange on which price types it supports.

Stoploss on exchange on futures markets can trigger on different price types. The naming for these prices in exchange terminology often varies, but is usually something around \"last\" (or \"contract price\" ), \"mark\" and \"index\".

Acceptable values for this setting are \"last\", \"mark\" and \"index\" - which freqtrade will transfer automatically to the corresponding API type, and place the stoploss on exchange order correspondingly.

"},{"location":"stoploss/#force_exit","title":"force_exit","text":"

force_exit is an optional value, which defaults to the same value as exit and is used when sending a /forceexit command from Telegram or from the Rest API.

"},{"location":"stoploss/#force_entry","title":"force_entry","text":"

force_entry is an optional value, which defaults to the same value as entry and is used when sending a /forceentry command from Telegram or from the Rest API.

"},{"location":"stoploss/#emergency_exit","title":"emergency_exit","text":"

emergency_exit is an optional value, which defaults to market and is used when creating stop loss on exchange orders fails. The below is the default which is used if not changed in strategy or configuration file.

Example from strategy file:

order_types = {\n    \"entry\": \"limit\",\n    \"exit\": \"limit\",\n    \"emergency_exit\": \"market\",\n    \"stoploss\": \"market\",\n    \"stoploss_on_exchange\": True,\n    \"stoploss_on_exchange_interval\": 60,\n    \"stoploss_on_exchange_limit_ratio\": 0.99\n}\n
"},{"location":"stoploss/#stop-loss-types","title":"Stop Loss Types","text":"

At this stage the bot contains the following stoploss support modes:

  1. Static stop loss.
  2. Trailing stop loss.
  3. Trailing stop loss, custom positive loss.
  4. Trailing stop loss only once the trade has reached a certain offset.
  5. Custom stoploss function
"},{"location":"stoploss/#static-stop-loss","title":"Static Stop Loss","text":"

This is very simple, you define a stop loss of x (as a ratio of price, i.e. x * 100% of price). This will try to sell the asset once the loss exceeds the defined loss.

Example of stop loss:

    stoploss = -0.10\n

For example, simplified math:

  • the bot buys an asset at a price of 100$
  • the stop loss is defined at -10%
  • the stop loss would get triggered once the asset drops below 90$
"},{"location":"stoploss/#trailing-stop-loss","title":"Trailing Stop Loss","text":"

The initial value for this is stoploss, just as you would define your static Stop loss. To enable trailing stoploss:

    stoploss = -0.10\n    trailing_stop = True\n

This will now activate an algorithm, which automatically moves the stop loss up every time the price of your asset increases.

For example, simplified math:

  • the bot buys an asset at a price of 100$
  • the stop loss is defined at -10%
  • the stop loss would get triggered once the asset drops below 90$
  • assuming the asset now increases to 102$
  • the stop loss will now be -10% of 102$ = 91.8$
  • now the asset drops in value to 101$, the stop loss will still be 91.8$ and would trigger at 91.8$.

In summary: The stoploss will be adjusted to be always be -10% of the highest observed price.

"},{"location":"stoploss/#trailing-stop-loss-custom-positive-loss","title":"Trailing stop loss, custom positive loss","text":"

You could also have a default stop loss when you are in the red with your buy (buy - fee), but once you hit a positive result (or an offset you define) the system will utilize a new stop loss, which can have a different value. For example, your default stop loss is -10%, but once you have more than 0% profit (example 0.1%) a different trailing stoploss will be used.

Note

If you want the stoploss to only be changed when you break even of making a profit (what most users want) please refer to next section with offset enabled.

Both values require trailing_stop to be set to true and trailing_stop_positive with a value.

    stoploss = -0.10\n    trailing_stop = True\n    trailing_stop_positive = 0.02\n    trailing_stop_positive_offset = 0.0\n    trailing_only_offset_is_reached = False  # Default - not necessary for this example\n

For example, simplified math:

  • the bot buys an asset at a price of 100$
  • the stop loss is defined at -10%
  • the stop loss would get triggered once the asset drops below 90$
  • assuming the asset now increases to 102$
  • the stop loss will now be -2% of 102$ = 99.96$ (99.96$ stop loss will be locked in and will follow asset price increments with -2%)
  • now the asset drops in value to 101$, the stop loss will still be 99.96$ and would trigger at 99.96$

The 0.02 would translate to a -2% stop loss. Before this, stoploss is used for the trailing stoploss.

Use an offset to change your stoploss

Use trailing_stop_positive_offset to ensure that your new trailing stoploss will be in profit by setting trailing_stop_positive_offset higher than trailing_stop_positive. Your first new stoploss value will then already have locked in profits.

Example with simplified math:

    stoploss = -0.10\n    trailing_stop = True\n    trailing_stop_positive = 0.02\n    trailing_stop_positive_offset = 0.03\n
  • the bot buys an asset at a price of 100$
  • the stop loss is defined at -10%, so the stop loss would get triggered once the asset drops below 90$
  • assuming the asset now increases to 102$
  • the stoploss will now be at 91.8$ - 10% below the highest observed rate
  • assuming the asset now increases to 103.5$ (above the offset configured)
  • the stop loss will now be -2% of 103.5$ = 101.43$
  • now the asset drops in value to 102$, the stop loss will still be 101.43$ and would trigger once price breaks below 101.43$
"},{"location":"stoploss/#trailing-stop-loss-only-once-the-trade-has-reached-a-certain-offset","title":"Trailing stop loss only once the trade has reached a certain offset","text":"

You can also keep a static stoploss until the offset is reached, and then trail the trade to take profits once the market turns.

If trailing_only_offset_is_reached = True then the trailing stoploss is only activated once the offset is reached. Until then, the stoploss remains at the configured stoploss. This option can be used with or without trailing_stop_positive, but uses trailing_stop_positive_offset as offset.

Configuration (offset is buy-price + 3%):

    stoploss = -0.10\n    trailing_stop = True\n    trailing_stop_positive = 0.02\n    trailing_stop_positive_offset = 0.03\n    trailing_only_offset_is_reached = True\n

For example, simplified math:

  • the bot buys an asset at a price of 100$
  • the stop loss is defined at -10%
  • the stop loss would get triggered once the asset drops below 90$
  • stoploss will remain at 90$ unless asset increases to or above the configured offset
  • assuming the asset now increases to 103$ (where we have the offset configured)
  • the stop loss will now be -2% of 103$ = 100.94$
  • now the asset drops in value to 101$, the stop loss will still be 100.94$ and would trigger at 100.94$

Tip

Make sure to have this value (trailing_stop_positive_offset) lower than minimal ROI, otherwise minimal ROI will apply first and sell the trade.

"},{"location":"stoploss/#stoploss-and-leverage","title":"Stoploss and Leverage","text":"

Stoploss should be thought of as \"risk on this trade\" - so a stoploss of 10% on a 100$ trade means you are willing to lose 10$ (10%) on this trade - which would trigger if the price moves 10% to the downside.

When using leverage, the same principle is applied - with stoploss defining the risk on the trade (the amount you are willing to lose).

Therefore, a stoploss of 10% on a 10x trade would trigger on a 1% price move. If your stake amount (own capital) was 100$ - this trade would be 1000$ at 10x (after leverage). If price moves 1% - you've lost 10$ of your own capital - therfore stoploss will trigger in this case.

Make sure to be aware of this, and avoid using too tight stoploss (at 10x leverage, 10% risk may be too little to allow the trade to \"breath\" a little).

"},{"location":"stoploss/#changing-stoploss-on-open-trades","title":"Changing stoploss on open trades","text":"

A stoploss on an open trade can be changed by changing the value in the configuration or strategy and use the /reload_config command (alternatively, completely stopping and restarting the bot also works).

The new stoploss value will be applied to open trades (and corresponding log-messages will be generated).

"},{"location":"stoploss/#limitations","title":"Limitations","text":"

Stoploss values cannot be changed if trailing_stop is enabled and the stoploss has already been adjusted, or if Edge is enabled (since Edge would recalculate stoploss based on the current market situation).

"},{"location":"strategy-advanced/","title":"Advanced Strategies","text":"

This page explains some advanced concepts available for strategies. If you're just getting started, please familiarize yourself with the Freqtrade basics and methods described in Strategy Customization first.

The call sequence of the methods described here is covered under bot execution logic. Those docs are also helpful in deciding which method is most suitable for your customisation needs.

Note

Callback methods should only be implemented if a strategy uses them.

Tip

Start off with a strategy template containing all available callback methods by running freqtrade new-strategy --strategy MyAwesomeStrategy --template advanced

"},{"location":"strategy-advanced/#storing-information","title":"Storing information","text":"

Storing information can be accomplished by creating a new dictionary within the strategy class.

The name of the variable can be chosen at will, but should be prefixed with custom_ to avoid naming collisions with predefined strategy variables.

class AwesomeStrategy(IStrategy):\n    # Create custom dictionary\n    custom_info = {}\n\n    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n        # Check if the entry already exists\n        if not metadata[\"pair\"] in self.custom_info:\n            # Create empty entry for this pair\n            self.custom_info[metadata[\"pair\"]] = {}\n\n        if \"crosstime\" in self.custom_info[metadata[\"pair\"]]:\n            self.custom_info[metadata[\"pair\"]][\"crosstime\"] += 1\n        else:\n            self.custom_info[metadata[\"pair\"]][\"crosstime\"] = 1\n

Warning

The data is not persisted after a bot-restart (or config-reload). Also, the amount of data should be kept smallish (no DataFrames and such), otherwise the bot will start to consume a lot of memory and eventually run out of memory and crash.

Note

If the data is pair-specific, make sure to use pair as one of the keys in the dictionary.

"},{"location":"strategy-advanced/#dataframe-access","title":"Dataframe access","text":"

You may access dataframe in various strategy functions by querying it from dataprovider.

from freqtrade.exchange import timeframe_to_prev_date\n\nclass AwesomeStrategy(IStrategy):\n    def confirm_trade_exit(self, pair: str, trade: 'Trade', order_type: str, amount: float,\n                           rate: float, time_in_force: str, exit_reason: str,\n                           current_time: 'datetime', **kwargs) -> bool:\n        # Obtain pair dataframe.\n        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)\n\n        # Obtain last available candle. Do not use current_time to look up latest candle, because \n        # current_time points to current incomplete candle whose data is not available.\n        last_candle = dataframe.iloc[-1].squeeze()\n        # <...>\n\n        # In dry/live runs trade open date will not match candle open date therefore it must be \n        # rounded.\n        trade_date = timeframe_to_prev_date(self.timeframe, trade.open_date_utc)\n        # Look up trade candle.\n        trade_candle = dataframe.loc[dataframe['date'] == trade_date]\n        # trade_candle may be empty for trades that just opened as it is still incomplete.\n        if not trade_candle.empty:\n            trade_candle = trade_candle.squeeze()\n            # <...>\n

Using .iloc[-1]

You can use .iloc[-1] here because get_analyzed_dataframe() only returns candles that backtesting is allowed to see. This will not work in populate_* methods, so make sure to not use .iloc[] in that area. Also, this will only work starting with version 2021.5.

"},{"location":"strategy-advanced/#enter-tag","title":"Enter Tag","text":"

When your strategy has multiple buy signals, you can name the signal that triggered. Then you can access your buy signal on custom_exit

def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n    dataframe.loc[\n        (\n            (dataframe['rsi'] < 35) &\n            (dataframe['volume'] > 0)\n        ),\n        ['enter_long', 'enter_tag']] = (1, 'buy_signal_rsi')\n\n    return dataframe\n\ndef custom_exit(self, pair: str, trade: Trade, current_time: datetime, current_rate: float,\n                current_profit: float, **kwargs):\n    dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)\n    last_candle = dataframe.iloc[-1].squeeze()\n    if trade.enter_tag == 'buy_signal_rsi' and last_candle['rsi'] > 80:\n        return 'sell_signal_rsi'\n    return None\n

Note

enter_tag is limited to 100 characters, remaining data will be truncated.

Warning

There is only one enter_tag column, which is used for both long and short trades. As a consequence, this column must be treated as \"last write wins\" (it's just a dataframe column after all). In fancy situations, where multiple signals collide (or if signals are deactivated again based on different conditions), this can lead to odd results with the wrong tag applied to an entry signal. These results are a consequence of the strategy overwriting prior tags - where the last tag will \"stick\" and will be the one freqtrade will use.

"},{"location":"strategy-advanced/#exit-tag","title":"Exit tag","text":"

Similar to Buy Tagging, you can also specify a sell tag.

def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n    dataframe.loc[\n        (\n            (dataframe['rsi'] > 70) &\n            (dataframe['volume'] > 0)\n        ),\n        ['exit_long', 'exit_tag']] = (1, 'exit_rsi')\n\n    return dataframe\n

The provided exit-tag is then used as sell-reason - and shown as such in backtest results.

Note

exit_reason is limited to 100 characters, remaining data will be truncated.

"},{"location":"strategy-advanced/#strategy-version","title":"Strategy version","text":"

You can implement custom strategy versioning by using the \"version\" method, and returning the version you would like this strategy to have.

def version(self) -> str:\n\"\"\"\n    Returns version of the strategy.\n    \"\"\"\n    return \"1.1\"\n

Note

You should make sure to implement proper version control (like a git repository) alongside this, as freqtrade will not keep historic versions of your strategy, so it's up to the user to be able to eventually roll back to a prior version of the strategy.

"},{"location":"strategy-advanced/#derived-strategies","title":"Derived strategies","text":"

The strategies can be derived from other strategies. This avoids duplication of your custom strategy code. You can use this technique to override small parts of your main strategy, leaving the rest untouched:

user_data/strategies/myawesomestrategy.py
class MyAwesomeStrategy(IStrategy):\n    ...\n    stoploss = 0.13\n    trailing_stop = False\n    # All other attributes and methods are here as they\n    # should be in any custom strategy...\n    ...\n
user_data/strategies/MyAwesomeStrategy2.py
from myawesomestrategy import MyAwesomeStrategy\nclass MyAwesomeStrategy2(MyAwesomeStrategy):\n    # Override something\n    stoploss = 0.08\n    trailing_stop = True\n

Both attributes and methods may be overridden, altering behavior of the original strategy in a way you need.

While keeping the subclass in the same file is technically possible, it can lead to some problems with hyperopt parameter files, we therefore recommend to use separate strategy files, and import the parent strategy as shown above.

"},{"location":"strategy-advanced/#embedding-strategies","title":"Embedding Strategies","text":"

Freqtrade provides you with an easy way to embed the strategy into your configuration file. This is done by utilizing BASE64 encoding and providing this string at the strategy configuration field, in your chosen config file.

"},{"location":"strategy-advanced/#encoding-a-string-as-base64","title":"Encoding a string as BASE64","text":"

This is a quick example, how to generate the BASE64 string in python

from base64 import urlsafe_b64encode\n\nwith open(file, 'r') as f:\n    content = f.read()\ncontent = urlsafe_b64encode(content.encode('utf-8'))\n

The variable 'content', will contain the strategy file in a BASE64 encoded form. Which can now be set in your configurations file as following

\"strategy\": \"NameOfStrategy:BASE64String\"\n

Please ensure that 'NameOfStrategy' is identical to the strategy name!

"},{"location":"strategy-advanced/#performance-warning","title":"Performance warning","text":"

When executing a strategy, one can sometimes be greeted by the following in the logs

PerformanceWarning: DataFrame is highly fragmented.

This is a warning from pandas and as the warning continues to say: use pd.concat(axis=1). This can have slight performance implications, which are usually only visible during hyperopt (when optimizing an indicator).

For example:

for val in self.buy_ema_short.range:\n    dataframe[f'ema_short_{val}'] = ta.EMA(dataframe, timeperiod=val)\n

should be rewritten to

frames = [dataframe]\nfor val in self.buy_ema_short.range:\n    frames.append(DataFrame({\n        f'ema_short_{val}': ta.EMA(dataframe, timeperiod=val)\n    }))\n\n# Append columns to existing dataframe\nmerged_frame = pd.concat(frames, axis=1)\n

Freqtrade does however also counter this by running dataframe.copy() on the dataframe right after the populate_indicators() method - so performance implications of this should be low to non-existant.

"},{"location":"strategy-callbacks/","title":"Strategy Callbacks","text":"

While the main strategy functions (populate_indicators(), populate_entry_trend(), populate_exit_trend()) should be used in a vectorized way, and are only called once during backtesting, callbacks are called \"whenever needed\".

As such, you should avoid doing heavy calculations in callbacks to avoid delays during operations. Depending on the callback used, they may be called when entering / exiting a trade, or throughout the duration of a trade.

Currently available callbacks:

  • bot_start()
  • bot_loop_start()
  • custom_stake_amount()
  • custom_exit()
  • custom_stoploss()
  • custom_entry_price() and custom_exit_price()
  • check_entry_timeout() and check_exit_timeout()
  • confirm_trade_entry()
  • confirm_trade_exit()
  • adjust_trade_position()
  • adjust_entry_price()
  • leverage()

Callback calling sequence

You can find the callback calling sequence in bot-basics

"},{"location":"strategy-callbacks/#bot-start","title":"Bot start","text":"

A simple callback which is called once when the strategy is loaded. This can be used to perform actions that must only be performed once and runs after dataprovider and wallet are set

import requests\n\nclass AwesomeStrategy(IStrategy):\n\n    # ... populate_* methods\n\n    def bot_start(self, **kwargs) -> None:\n\"\"\"\n        Called only once after bot instantiation.\n        :param **kwargs: Ensure to keep this here so updates to this won't break your strategy.\n        \"\"\"\n        if self.config['runmode'].value in ('live', 'dry_run'):\n            # Assign this to the class by using self.*\n            # can then be used by populate_* methods\n            self.custom_remote_data = requests.get('https://some_remote_source.example.com')\n

During hyperopt, this runs only once at startup.

"},{"location":"strategy-callbacks/#bot-loop-start","title":"Bot loop start","text":"

A simple callback which is called once at the start of every bot throttling iteration in dry/live mode (roughly every 5 seconds, unless configured differently) or once per candle in backtest/hyperopt mode. This can be used to perform calculations which are pair independent (apply to all pairs), loading of external data, etc.

import requests\n\nclass AwesomeStrategy(IStrategy):\n\n    # ... populate_* methods\n\n    def bot_loop_start(self, current_time: datetime, **kwargs) -> None:\n\"\"\"\n        Called at the start of the bot iteration (one loop).\n        Might be used to perform pair-independent tasks\n        (e.g. gather some remote resource for comparison)\n        :param current_time: datetime object, containing the current datetime\n        :param **kwargs: Ensure to keep this here so updates to this won't break your strategy.\n        \"\"\"\n        if self.config['runmode'].value in ('live', 'dry_run'):\n            # Assign this to the class by using self.*\n            # can then be used by populate_* methods\n            self.remote_data = requests.get('https://some_remote_source.example.com')\n
"},{"location":"strategy-callbacks/#stake-size-management","title":"Stake size management","text":"

Called before entering a trade, makes it possible to manage your position size when placing a new trade.

class AwesomeStrategy(IStrategy):\n    def custom_stake_amount(self, pair: str, current_time: datetime, current_rate: float,\n                            proposed_stake: float, min_stake: Optional[float], max_stake: float,\n                            leverage: float, entry_tag: Optional[str], side: str,\n                            **kwargs) -> float:\n\n        dataframe, _ = self.dp.get_analyzed_dataframe(pair=pair, timeframe=self.timeframe)\n        current_candle = dataframe.iloc[-1].squeeze()\n\n        if current_candle['fastk_rsi_1h'] > current_candle['fastd_rsi_1h']:\n            if self.config['stake_amount'] == 'unlimited':\n                # Use entire available wallet during favorable conditions when in compounding mode.\n                return max_stake\n            else:\n                # Compound profits during favorable conditions instead of using a static stake.\n                return self.wallets.get_total_stake_amount() / self.config['max_open_trades']\n\n        # Use default stake amount.\n        return proposed_stake\n

Freqtrade will fall back to the proposed_stake value should your code raise an exception. The exception itself will be logged.

Tip

You do not have to ensure that min_stake <= returned_value <= max_stake. Trades will succeed as the returned value will be clamped to supported range and this action will be logged.

Tip

Returning 0 or None will prevent trades from being placed.

"},{"location":"strategy-callbacks/#custom-exit-signal","title":"Custom exit signal","text":"

Called for open trade every throttling iteration (roughly every 5 seconds) until a trade is closed.

Allows to define custom exit signals, indicating that specified position should be sold. This is very useful when we need to customize exit conditions for each individual trade, or if you need trade data to make an exit decision.

For example you could implement a 1:2 risk-reward ROI with custom_exit().

Using custom_exit() signals in place of stoploss though is not recommended. It is a inferior method to using custom_stoploss() in this regard - which also allows you to keep the stoploss on exchange.

Note

Returning a (none-empty) string or True from this method is equal to setting exit signal on a candle at specified time. This method is not called when exit signal is set already, or if exit signals are disabled (use_exit_signal=False). string max length is 64 characters. Exceeding this limit will cause the message to be truncated to 64 characters. custom_exit() will ignore exit_profit_only, and will always be called unless use_exit_signal=False, even if there is a new enter signal.

An example of how we can use different indicators depending on the current profit and also exit trades that were open longer than one day:

class AwesomeStrategy(IStrategy):\n    def custom_exit(self, pair: str, trade: 'Trade', current_time: 'datetime', current_rate: float,\n                    current_profit: float, **kwargs):\n        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)\n        last_candle = dataframe.iloc[-1].squeeze()\n\n        # Above 20% profit, sell when rsi < 80\n        if current_profit > 0.2:\n            if last_candle['rsi'] < 80:\n                return 'rsi_below_80'\n\n        # Between 2% and 10%, sell if EMA-long above EMA-short\n        if 0.02 < current_profit < 0.1:\n            if last_candle['emalong'] > last_candle['emashort']:\n                return 'ema_long_below_80'\n\n        # Sell any positions at a loss if they are held for more than one day.\n        if current_profit < 0.0 and (current_time - trade.open_date_utc).days >= 1:\n            return 'unclog'\n

See Dataframe access for more information about dataframe use in strategy callbacks.

"},{"location":"strategy-callbacks/#custom-stoploss","title":"Custom stoploss","text":"

Called for open trade every iteration (roughly every 5 seconds) until a trade is closed.

The usage of the custom stoploss method must be enabled by setting use_custom_stoploss=True on the strategy object.

The stoploss price can only ever move upwards - if the stoploss value returned from custom_stoploss would result in a lower stoploss price than was previously set, it will be ignored. The traditional stoploss value serves as an absolute lower level and will be instated as the initial stoploss (before this method is called for the first time for a trade), and is still mandatory.

The method must return a stoploss value (float / number) as a percentage of the current price. E.g. If the current_rate is 200 USD, then returning 0.02 will set the stoploss price 2% lower, at 196 USD. During backtesting, current_rate (and current_profit) are provided against the candle's high (or low for short trades) - while the resulting stoploss is evaluated against the candle's low (or high for short trades).

The absolute value of the return value is used (the sign is ignored), so returning 0.05 or -0.05 have the same result, a stoploss 5% below the current price.

To simulate a regular trailing stoploss of 4% (trailing 4% behind the maximum reached price) you would use the following very simple method:

# additional imports required\nfrom datetime import datetime\nfrom freqtrade.persistence import Trade\n\nclass AwesomeStrategy(IStrategy):\n\n    # ... populate_* methods\n\n    use_custom_stoploss = True\n\n    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,\n                        current_rate: float, current_profit: float, **kwargs) -> float:\n\"\"\"\n        Custom stoploss logic, returning the new distance relative to current_rate (as ratio).\n        e.g. returning -0.05 would create a stoploss 5% below current_rate.\n        The custom stoploss can never be below self.stoploss, which serves as a hard maximum loss.\n\n        For full documentation please go to https://www.freqtrade.io/en/latest/strategy-advanced/\n\n        When not implemented by a strategy, returns the initial stoploss value\n        Only called when use_custom_stoploss is set to True.\n\n        :param pair: Pair that's currently analyzed\n        :param trade: trade object.\n        :param current_time: datetime object, containing the current datetime\n        :param current_rate: Rate, calculated based on pricing settings in exit_pricing.\n        :param current_profit: Current profit (as ratio), calculated based on current_rate.\n        :param **kwargs: Ensure to keep this here so updates to this won't break your strategy.\n        :return float: New stoploss value, relative to the current rate\n        \"\"\"\n        return -0.04\n

Stoploss on exchange works similar to trailing_stop, and the stoploss on exchange is updated as configured in stoploss_on_exchange_interval (More details about stoploss on exchange).

Use of dates

All time-based calculations should be done based on current_time - using datetime.now() or datetime.utcnow() is discouraged, as this will break backtesting support.

Trailing stoploss

It's recommended to disable trailing_stop when using custom stoploss values. Both can work in tandem, but you might encounter the trailing stop to move the price higher while your custom function would not want this, causing conflicting behavior.

"},{"location":"strategy-callbacks/#custom-stoploss-examples","title":"Custom stoploss examples","text":"

The next section will show some examples on what's possible with the custom stoploss function. Of course, many more things are possible, and all examples can be combined at will.

"},{"location":"strategy-callbacks/#time-based-trailing-stop","title":"Time based trailing stop","text":"

Use the initial stoploss for the first 60 minutes, after this change to 10% trailing stoploss, and after 2 hours (120 minutes) we use a 5% trailing stoploss.

from datetime import datetime, timedelta\nfrom freqtrade.persistence import Trade\n\nclass AwesomeStrategy(IStrategy):\n\n    # ... populate_* methods\n\n    use_custom_stoploss = True\n\n    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,\n                        current_rate: float, current_profit: float, **kwargs) -> float:\n\n        # Make sure you have the longest interval first - these conditions are evaluated from top to bottom.\n        if current_time - timedelta(minutes=120) > trade.open_date_utc:\n            return -0.05\n        elif current_time - timedelta(minutes=60) > trade.open_date_utc:\n            return -0.10\n        return 1\n
"},{"location":"strategy-callbacks/#different-stoploss-per-pair","title":"Different stoploss per pair","text":"

Use a different stoploss depending on the pair. In this example, we'll trail the highest price with 10% trailing stoploss for ETH/BTC and XRP/BTC, with 5% trailing stoploss for LTC/BTC and with 15% for all other pairs.

from datetime import datetime\nfrom freqtrade.persistence import Trade\n\nclass AwesomeStrategy(IStrategy):\n\n    # ... populate_* methods\n\n    use_custom_stoploss = True\n\n    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,\n                        current_rate: float, current_profit: float, **kwargs) -> float:\n\n        if pair in ('ETH/BTC', 'XRP/BTC'):\n            return -0.10\n        elif pair in ('LTC/BTC'):\n            return -0.05\n        return -0.15\n
"},{"location":"strategy-callbacks/#trailing-stoploss-with-positive-offset","title":"Trailing stoploss with positive offset","text":"

Use the initial stoploss until the profit is above 4%, then use a trailing stoploss of 50% of the current profit with a minimum of 2.5% and a maximum of 5%.

Please note that the stoploss can only increase, values lower than the current stoploss are ignored.

from datetime import datetime, timedelta\nfrom freqtrade.persistence import Trade\n\nclass AwesomeStrategy(IStrategy):\n\n    # ... populate_* methods\n\n    use_custom_stoploss = True\n\n    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,\n                        current_rate: float, current_profit: float, **kwargs) -> float:\n\n        if current_profit < 0.04:\n            return -1 # return a value bigger than the initial stoploss to keep using the initial stoploss\n\n        # After reaching the desired offset, allow the stoploss to trail by half the profit\n        desired_stoploss = current_profit / 2\n\n        # Use a minimum of 2.5% and a maximum of 5%\n        return max(min(desired_stoploss, 0.05), 0.025)\n
"},{"location":"strategy-callbacks/#stepped-stoploss","title":"Stepped stoploss","text":"

Instead of continuously trailing behind the current price, this example sets fixed stoploss price levels based on the current profit.

  • Use the regular stoploss until 20% profit is reached
  • Once profit is > 20% - set stoploss to 7% above open price.
  • Once profit is > 25% - set stoploss to 15% above open price.
  • Once profit is > 40% - set stoploss to 25% above open price.
from datetime import datetime\nfrom freqtrade.persistence import Trade\nfrom freqtrade.strategy import stoploss_from_open\n\nclass AwesomeStrategy(IStrategy):\n\n    # ... populate_* methods\n\n    use_custom_stoploss = True\n\n    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,\n                        current_rate: float, current_profit: float, **kwargs) -> float:\n\n        # evaluate highest to lowest, so that highest possible stop is used\n        if current_profit > 0.40:\n            return stoploss_from_open(0.25, current_profit, is_short=trade.is_short, leverage=trade.leverage)\n        elif current_profit > 0.25:\n            return stoploss_from_open(0.15, current_profit, is_short=trade.is_short, leverage=trade.leverage)\n        elif current_profit > 0.20:\n            return stoploss_from_open(0.07, current_profit, is_short=trade.is_short, leverage=trade.leverage)\n\n        # return maximum stoploss value, keeping current stoploss price unchanged\n        return 1\n
"},{"location":"strategy-callbacks/#custom-stoploss-using-an-indicator-from-dataframe-example","title":"Custom stoploss using an indicator from dataframe example","text":"

Absolute stoploss value may be derived from indicators stored in dataframe. Example uses parabolic SAR below the price as stoploss.

class AwesomeStrategy(IStrategy):\n\n    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n        # <...>\n        dataframe['sar'] = ta.SAR(dataframe)\n\n    use_custom_stoploss = True\n\n    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,\n                        current_rate: float, current_profit: float, **kwargs) -> float:\n\n        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)\n        last_candle = dataframe.iloc[-1].squeeze()\n\n        # Use parabolic sar as absolute stoploss price\n        stoploss_price = last_candle['sar']\n\n        # Convert absolute price to percentage relative to current_rate\n        if stoploss_price < current_rate:\n            return stoploss_from_absolute(stoploss_price, current_rate, is_short=trade.is_short)\n\n        # return maximum stoploss value, keeping current stoploss price unchanged\n        return 1\n

See Dataframe access for more information about dataframe use in strategy callbacks.

"},{"location":"strategy-callbacks/#common-helpers-for-stoploss-calculations","title":"Common helpers for stoploss calculations","text":""},{"location":"strategy-callbacks/#stoploss-relative-to-open-price","title":"Stoploss relative to open price","text":"

Stoploss values returned from custom_stoploss() always specify a percentage relative to current_rate. In order to set a stoploss relative to the open price, we need to use current_profit to calculate what percentage relative to the current_rate will give you the same result as if the percentage was specified from the open price.

The helper function stoploss_from_open() can be used to convert from an open price relative stop, to a current price relative stop which can be returned from custom_stoploss().

"},{"location":"strategy-callbacks/#stoploss-percentage-from-absolute-price","title":"Stoploss percentage from absolute price","text":"

Stoploss values returned from custom_stoploss() always specify a percentage relative to current_rate. In order to set a stoploss at specified absolute price level, we need to use stop_rate to calculate what percentage relative to the current_rate will give you the same result as if the percentage was specified from the open price.

The helper function stoploss_from_absolute() can be used to convert from an absolute price, to a current price relative stop which can be returned from custom_stoploss().

"},{"location":"strategy-callbacks/#custom-order-price-rules","title":"Custom order price rules","text":"

By default, freqtrade use the orderbook to automatically set an order price(Relevant documentation), you also have the option to create custom order prices based on your strategy.

You can use this feature by creating a custom_entry_price() function in your strategy file to customize entry prices and custom_exit_price() for exits.

Each of these methods are called right before placing an order on the exchange.

Note

If your custom pricing function return None or an invalid value, price will fall back to proposed_rate, which is based on the regular pricing configuration.

"},{"location":"strategy-callbacks/#custom-order-entry-and-exit-price-example","title":"Custom order entry and exit price example","text":"
from datetime import datetime, timedelta, timezone\nfrom freqtrade.persistence import Trade\n\nclass AwesomeStrategy(IStrategy):\n\n    # ... populate_* methods\n\n    def custom_entry_price(self, pair: str, current_time: datetime, proposed_rate: float,\n                           entry_tag: Optional[str], side: str, **kwargs) -> float:\n\n        dataframe, last_updated = self.dp.get_analyzed_dataframe(pair=pair,\n                                                                timeframe=self.timeframe)\n        new_entryprice = dataframe['bollinger_10_lowerband'].iat[-1]\n\n        return new_entryprice\n\n    def custom_exit_price(self, pair: str, trade: Trade,\n                          current_time: datetime, proposed_rate: float,\n                          current_profit: float, exit_tag: Optional[str], **kwargs) -> float:\n\n        dataframe, last_updated = self.dp.get_analyzed_dataframe(pair=pair,\n                                                                timeframe=self.timeframe)\n        new_exitprice = dataframe['bollinger_10_upperband'].iat[-1]\n\n        return new_exitprice\n

Warning

Modifying entry and exit prices will only work for limit orders. Depending on the price chosen, this can result in a lot of unfilled orders. By default the maximum allowed distance between the current price and the custom price is 2%, this value can be changed in config with the custom_price_max_distance_ratio parameter. Example: If the new_entryprice is 97, the proposed_rate is 100 and the custom_price_max_distance_ratio is set to 2%, The retained valid custom entry price will be 98, which is 2% below the current (proposed) rate.

Backtesting

Custom prices are supported in backtesting (starting with 2021.12), and orders will fill if the price falls within the candle's low/high range. Orders that don't fill immediately are subject to regular timeout handling, which happens once per (detail) candle. custom_exit_price() is only called for sells of type exit_signal, Custom exit and partial exits. All other exit-types will use regular backtesting prices.

"},{"location":"strategy-callbacks/#custom-order-timeout-rules","title":"Custom order timeout rules","text":"

Simple, time-based order-timeouts can be configured either via strategy or in the configuration in the unfilledtimeout section.

However, freqtrade also offers a custom callback for both order types, which allows you to decide based on custom criteria if an order did time out or not.

Note

Backtesting fills orders if their price falls within the candle's low/high range. The below callbacks will be called once per (detail) candle for orders that don't fill immediately (which use custom pricing).

"},{"location":"strategy-callbacks/#custom-order-timeout-example","title":"Custom order timeout example","text":"

Called for every open order until that order is either filled or cancelled. check_entry_timeout() is called for trade entries, while check_exit_timeout() is called for trade exit orders.

A simple example, which applies different unfilled-timeouts depending on the price of the asset can be seen below. It applies a tight timeout for higher priced assets, while allowing more time to fill on cheap coins.

The function must return either True (cancel order) or False (keep order alive).

from datetime import datetime, timedelta\nfrom freqtrade.persistence import Trade, Order\n\nclass AwesomeStrategy(IStrategy):\n\n    # ... populate_* methods\n\n    # Set unfilledtimeout to 25 hours, since the maximum timeout from below is 24 hours.\n    unfilledtimeout = {\n        'entry': 60 * 25,\n        'exit': 60 * 25\n    }\n\n    def check_entry_timeout(self, pair: str, trade: 'Trade', order: 'Order',\n                            current_time: datetime, **kwargs) -> bool:\n        if trade.open_rate > 100 and trade.open_date_utc < current_time - timedelta(minutes=5):\n            return True\n        elif trade.open_rate > 10 and trade.open_date_utc < current_time - timedelta(minutes=3):\n            return True\n        elif trade.open_rate < 1 and trade.open_date_utc < current_time - timedelta(hours=24):\n           return True\n        return False\n\n\n    def check_exit_timeout(self, pair: str, trade: Trade, order: 'Order',\n                           current_time: datetime, **kwargs) -> bool:\n        if trade.open_rate > 100 and trade.open_date_utc < current_time - timedelta(minutes=5):\n            return True\n        elif trade.open_rate > 10 and trade.open_date_utc < current_time - timedelta(minutes=3):\n            return True\n        elif trade.open_rate < 1 and trade.open_date_utc < current_time - timedelta(hours=24):\n           return True\n        return False\n

Note

For the above example, unfilledtimeout must be set to something bigger than 24h, otherwise that type of timeout will apply first.

"},{"location":"strategy-callbacks/#custom-order-timeout-example-using-additional-data","title":"Custom order timeout example (using additional data)","text":"
from datetime import datetime\nfrom freqtrade.persistence import Trade, Order\n\nclass AwesomeStrategy(IStrategy):\n\n    # ... populate_* methods\n\n    # Set unfilledtimeout to 25 hours, since the maximum timeout from below is 24 hours.\n    unfilledtimeout = {\n        'entry': 60 * 25,\n        'exit': 60 * 25\n    }\n\n    def check_entry_timeout(self, pair: str, trade: 'Trade', order: 'Order',\n                            current_time: datetime, **kwargs) -> bool:\n        ob = self.dp.orderbook(pair, 1)\n        current_price = ob['bids'][0][0]\n        # Cancel buy order if price is more than 2% above the order.\n        if current_price > order.price * 1.02:\n            return True\n        return False\n\n\n    def check_exit_timeout(self, pair: str, trade: 'Trade', order: 'Order',\n                           current_time: datetime, **kwargs) -> bool:\n        ob = self.dp.orderbook(pair, 1)\n        current_price = ob['asks'][0][0]\n        # Cancel sell order if price is more than 2% below the order.\n        if current_price < order.price * 0.98:\n            return True\n        return False\n
"},{"location":"strategy-callbacks/#bot-order-confirmation","title":"Bot order confirmation","text":"

Confirm trade entry / exits. This are the last methods that will be called before an order is placed.

"},{"location":"strategy-callbacks/#trade-entry-buy-order-confirmation","title":"Trade entry (buy order) confirmation","text":"

confirm_trade_entry() can be used to abort a trade entry at the latest second (maybe because the price is not what we expect).

class AwesomeStrategy(IStrategy):\n\n    # ... populate_* methods\n\n    def confirm_trade_entry(self, pair: str, order_type: str, amount: float, rate: float,\n                            time_in_force: str, current_time: datetime, entry_tag: Optional[str],\n                            side: str, **kwargs) -> bool:\n\"\"\"\n        Called right before placing a entry order.\n        Timing for this function is critical, so avoid doing heavy computations or\n        network requests in this method.\n\n        For full documentation please go to https://www.freqtrade.io/en/latest/strategy-advanced/\n\n        When not implemented by a strategy, returns True (always confirming).\n\n        :param pair: Pair that's about to be bought/shorted.\n        :param order_type: Order type (as configured in order_types). usually limit or market.\n        :param amount: Amount in target (base) currency that's going to be traded.\n        :param rate: Rate that's going to be used when using limit orders \n                     or current rate for market orders.\n        :param time_in_force: Time in force. Defaults to GTC (Good-til-cancelled).\n        :param current_time: datetime object, containing the current datetime\n        :param entry_tag: Optional entry_tag (buy_tag) if provided with the buy signal.\n        :param side: 'long' or 'short' - indicating the direction of the proposed trade\n        :param **kwargs: Ensure to keep this here so updates to this won't break your strategy.\n        :return bool: When True is returned, then the buy-order is placed on the exchange.\n            False aborts the process\n        \"\"\"\n        return True\n
"},{"location":"strategy-callbacks/#trade-exit-sell-order-confirmation","title":"Trade exit (sell order) confirmation","text":"

confirm_trade_exit() can be used to abort a trade exit (sell) at the latest second (maybe because the price is not what we expect).

confirm_trade_exit() may be called multiple times within one iteration for the same trade if different exit-reasons apply. The exit-reasons (if applicable) will be in the following sequence:

  • exit_signal / custom_exit
  • stop_loss
  • roi
  • trailing_stop_loss
from freqtrade.persistence import Trade\n\n\nclass AwesomeStrategy(IStrategy):\n\n    # ... populate_* methods\n\n    def confirm_trade_exit(self, pair: str, trade: Trade, order_type: str, amount: float,\n                           rate: float, time_in_force: str, exit_reason: str,\n                           current_time: datetime, **kwargs) -> bool:\n\"\"\"\n        Called right before placing a regular exit order.\n        Timing for this function is critical, so avoid doing heavy computations or\n        network requests in this method.\n\n        For full documentation please go to https://www.freqtrade.io/en/latest/strategy-advanced/\n\n        When not implemented by a strategy, returns True (always confirming).\n\n        :param pair: Pair for trade that's about to be exited.\n        :param trade: trade object.\n        :param order_type: Order type (as configured in order_types). usually limit or market.\n        :param amount: Amount in base currency.\n        :param rate: Rate that's going to be used when using limit orders\n                     or current rate for market orders.\n        :param time_in_force: Time in force. Defaults to GTC (Good-til-cancelled).\n        :param exit_reason: Exit reason.\n            Can be any of ['roi', 'stop_loss', 'stoploss_on_exchange', 'trailing_stop_loss',\n                           'exit_signal', 'force_exit', 'emergency_exit']\n        :param current_time: datetime object, containing the current datetime\n        :param **kwargs: Ensure to keep this here so updates to this won't break your strategy.\n        :return bool: When True, then the exit-order is placed on the exchange.\n            False aborts the process\n        \"\"\"\n        if exit_reason == 'force_exit' and trade.calc_profit_ratio(rate) < 0:\n            # Reject force-sells with negative profit\n            # This is just a sample, please adjust to your needs\n            # (this does not necessarily make sense, assuming you know when you're force-selling)\n            return False\n        return True\n

Warning

confirm_trade_exit() can prevent stoploss exits, causing significant losses as this would ignore stoploss exits. confirm_trade_exit() will not be called for Liquidations - as liquidations are forced by the exchange, and therefore cannot be rejected.

"},{"location":"strategy-callbacks/#adjust-trade-position","title":"Adjust trade position","text":"

The position_adjustment_enable strategy property enables the usage of adjust_trade_position() callback in the strategy. For performance reasons, it's disabled by default and freqtrade will show a warning message on startup if enabled. adjust_trade_position() can be used to perform additional orders, for example to manage risk with DCA (Dollar Cost Averaging) or to increase or decrease positions.

max_entry_position_adjustment property is used to limit the number of additional buys per trade (on top of the first buy) that the bot can execute. By default, the value is -1 which means the bot have no limit on number of adjustment buys.

The strategy is expected to return a stake_amount (in stake currency) between min_stake and max_stake if and when an additional buy order should be made (position is increased). If there are not enough funds in the wallet (the return value is above max_stake) then the signal will be ignored. Additional orders also result in additional fees and those orders don't count towards max_open_trades.

This callback is not called when there is an open order (either buy or sell) waiting for execution.

adjust_trade_position() is called very frequently for the duration of a trade, so you must keep your implementation as performant as possible.

Additional Buys are ignored once you have reached the maximum amount of extra buys that you have set on max_entry_position_adjustment, but the callback is called anyway looking for partial exits.

Position adjustments will always be applied in the direction of the trade, so a positive value will always increase your position (negative values will decrease your position), no matter if it's a long or short trade. Modifications to leverage are not possible, and the stake-amount is assumed to be before applying leverage.

About stake size

Using fixed stake size means it will be the amount used for the first order, just like without position adjustment. If you wish to buy additional orders with DCA, then make sure to leave enough funds in the wallet for that. Using 'unlimited' stake amount with DCA orders requires you to also implement the custom_stake_amount() callback to avoid allocating all funds to the initial order.

Warning

Stoploss is still calculated from the initial opening price, not averaged price. Regular stoploss rules still apply (cannot move down).

While /stopentry command stops the bot from entering new trades, the position adjustment feature will continue buying new orders on existing trades.

Backtesting

During backtesting this callback is called for each candle in timeframe or timeframe_detail, so run-time performance will be affected. This can also cause deviating results between live and backtesting, since backtesting can adjust the trade only once per candle, whereas live could adjust the trade multiple times per candle.

from freqtrade.persistence import Trade\n\n\nclass DigDeeperStrategy(IStrategy):\n\n    position_adjustment_enable = True\n\n    # Attempts to handle large drops with DCA. High stoploss is required.\n    stoploss = -0.30\n\n    # ... populate_* methods\n\n    # Example specific variables\n    max_entry_position_adjustment = 3\n    # This number is explained a bit further down\n    max_dca_multiplier = 5.5\n\n    # This is called when placing the initial order (opening trade)\n    def custom_stake_amount(self, pair: str, current_time: datetime, current_rate: float,\n                            proposed_stake: float, min_stake: Optional[float], max_stake: float,\n                            leverage: float, entry_tag: Optional[str], side: str,\n                            **kwargs) -> float:\n\n        # We need to leave most of the funds for possible further DCA orders\n        # This also applies to fixed stakes\n        return proposed_stake / self.max_dca_multiplier\n\n    def adjust_trade_position(self, trade: Trade, current_time: datetime,\n                              current_rate: float, current_profit: float,\n                              min_stake: Optional[float], max_stake: float,\n                              current_entry_rate: float, current_exit_rate: float,\n                              current_entry_profit: float, current_exit_profit: float,\n                              **kwargs) -> Optional[float]:\n\"\"\"\n        Custom trade adjustment logic, returning the stake amount that a trade should be\n        increased or decreased.\n        This means extra buy or sell orders with additional fees.\n        Only called when `position_adjustment_enable` is set to True.\n\n        For full documentation please go to https://www.freqtrade.io/en/latest/strategy-advanced/\n\n        When not implemented by a strategy, returns None\n\n        :param trade: trade object.\n        :param current_time: datetime object, containing the current datetime\n        :param current_rate: Current buy rate.\n        :param current_profit: Current profit (as ratio), calculated based on current_rate.\n        :param min_stake: Minimal stake size allowed by exchange (for both entries and exits)\n        :param max_stake: Maximum stake allowed (either through balance, or by exchange limits).\n        :param current_entry_rate: Current rate using entry pricing.\n        :param current_exit_rate: Current rate using exit pricing.\n        :param current_entry_profit: Current profit using entry pricing.\n        :param current_exit_profit: Current profit using exit pricing.\n        :param **kwargs: Ensure to keep this here so updates to this won't break your strategy.\n        :return float: Stake amount to adjust your trade,\n                       Positive values to increase position, Negative values to decrease position.\n                       Return None for no action.\n        \"\"\"\n\n        if current_profit > 0.05 and trade.nr_of_successful_exits == 0:\n            # Take half of the profit at +5%\n            return -(trade.stake_amount / 2)\n\n        if current_profit > -0.05:\n            return None\n\n        # Obtain pair dataframe (just to show how to access it)\n        dataframe, _ = self.dp.get_analyzed_dataframe(trade.pair, self.timeframe)\n        # Only buy when not actively falling price.\n        last_candle = dataframe.iloc[-1].squeeze()\n        previous_candle = dataframe.iloc[-2].squeeze()\n        if last_candle['close'] < previous_candle['close']:\n            return None\n\n        filled_entries = trade.select_filled_orders(trade.entry_side)\n        count_of_entries = trade.nr_of_successful_entries\n        # Allow up to 3 additional increasingly larger buys (4 in total)\n        # Initial buy is 1x\n        # If that falls to -5% profit, we buy 1.25x more, average profit should increase to roughly -2.2%\n        # If that falls down to -5% again, we buy 1.5x more\n        # If that falls once again down to -5%, we buy 1.75x more\n        # Total stake for this trade would be 1 + 1.25 + 1.5 + 1.75 = 5.5x of the initial allowed stake.\n        # That is why max_dca_multiplier is 5.5\n        # Hope you have a deep wallet!\n        try:\n            # This returns first order stake size\n            stake_amount = filled_entries[0].cost\n            # This then calculates current safety order size\n            stake_amount = stake_amount * (1 + (count_of_entries * 0.25))\n            return stake_amount\n        except Exception as exception:\n            return None\n\n        return None\n
"},{"location":"strategy-callbacks/#position-adjust-calculations","title":"Position adjust calculations","text":"
  • Entry rates are calculated using weighted averages.
  • Exits will not influence the average entry rate.
  • Partial exit relative profit is relative to the average entry price at this point.
  • Final exit relative profit is calculated based on the total invested capital. (See example below)
Calculation example

This example assumes 0 fees for simplicity, and a long position on an imaginary coin.

  • Buy 100@8$
  • Buy 100@9$ -> Avg price: 8.5$
  • Sell 100@10$ -> Avg price: 8.5$, realized profit 150$, 17.65%
  • Buy 150@11$ -> Avg price: 10$, realized profit 150$, 17.65%
  • Sell 100@12$ -> Avg price: 10$, total realized profit 350$, 20%
  • Sell 150@14$ -> Avg price: 10$, total realized profit 950$, 40% <- This will be the last \"Exit\" message

The total profit for this trade was 950$ on a 3350$ investment (100@8$ + 100@9$ + 150@11$). As such - the final relative profit is 28.35% (950 / 3350).

"},{"location":"strategy-callbacks/#adjust-entry-price","title":"Adjust Entry Price","text":"

The adjust_entry_price() callback may be used by strategy developer to refresh/replace limit orders upon arrival of new candles. Be aware that custom_entry_price() is still the one dictating initial entry limit order price target at the time of entry trigger.

Orders can be cancelled out of this callback by returning None.

Returning current_order_rate will keep the order on the exchange \"as is\". Returning any other price will cancel the existing order, and replace it with a new order.

The trade open-date (trade.open_date_utc) will remain at the time of the very first order placed. Please make sure to be aware of this - and eventually adjust your logic in other callbacks to account for this, and use the date of the first filled order instead.

Regular timeout

Entry unfilledtimeout mechanism (as well as check_entry_timeout()) takes precedence over this. Entry Orders that are cancelled via the above methods will not have this callback called. Be sure to update timeout values to match your expectations.

from freqtrade.persistence import Trade\nfrom datetime import timedelta\n\nclass AwesomeStrategy(IStrategy):\n\n    # ... populate_* methods\n\n    def adjust_entry_price(self, trade: Trade, order: Optional[Order], pair: str,\n                           current_time: datetime, proposed_rate: float, current_order_rate: float,\n                           entry_tag: Optional[str], side: str, **kwargs) -> float:\n\"\"\"\n        Entry price re-adjustment logic, returning the user desired limit price.\n        This only executes when a order was already placed, still open (unfilled fully or partially)\n        and not timed out on subsequent candles after entry trigger.\n\n        When not implemented by a strategy, returns current_order_rate as default.\n        If current_order_rate is returned then the existing order is maintained.\n        If None is returned then order gets canceled but not replaced by a new one.\n\n        :param pair: Pair that's currently analyzed\n        :param trade: Trade object.\n        :param order: Order object\n        :param current_time: datetime object, containing the current datetime\n        :param proposed_rate: Rate, calculated based on pricing settings in entry_pricing.\n        :param current_order_rate: Rate of the existing order in place.\n        :param entry_tag: Optional entry_tag (buy_tag) if provided with the buy signal.\n        :param side: 'long' or 'short' - indicating the direction of the proposed trade\n        :param **kwargs: Ensure to keep this here so updates to this won't break your strategy.\n        :return float: New entry price value if provided\n\n        \"\"\"\n        # Limit orders to use and follow SMA200 as price target for the first 10 minutes since entry trigger for BTC/USDT pair.\n        if pair == 'BTC/USDT' and entry_tag == 'long_sma200' and side == 'long' and (current_time - timedelta(minutes=10)) > trade.open_date_utc:\n            # just cancel the order if it has been filled more than half of the amount\n            if order.filled > order.remaining:\n                return None\n            else:\n                dataframe, _ = self.dp.get_analyzed_dataframe(pair=pair, timeframe=self.timeframe)\n                current_candle = dataframe.iloc[-1].squeeze()\n                # desired price\n                return current_candle['sma_200']\n        # default: maintain existing order\n        return current_order_rate\n
"},{"location":"strategy-callbacks/#leverage-callback","title":"Leverage Callback","text":"

When trading in markets that allow leverage, this method must return the desired Leverage (Defaults to 1 -> No leverage).

Assuming a capital of 500USDT, a trade with leverage=3 would result in a position with 500 x 3 = 1500 USDT.

Values that are above max_leverage will be adjusted to max_leverage. For markets / exchanges that don't support leverage, this method is ignored.

class AwesomeStrategy(IStrategy):\n    def leverage(self, pair: str, current_time: datetime, current_rate: float,\n                 proposed_leverage: float, max_leverage: float, entry_tag: Optional[str], side: str,\n                 **kwargs) -> float:\n\"\"\"\n        Customize leverage for each new trade. This method is only called in futures mode.\n\n        :param pair: Pair that's currently analyzed\n        :param current_time: datetime object, containing the current datetime\n        :param current_rate: Rate, calculated based on pricing settings in exit_pricing.\n        :param proposed_leverage: A leverage proposed by the bot.\n        :param max_leverage: Max leverage allowed on this pair\n        :param entry_tag: Optional entry_tag (buy_tag) if provided with the buy signal.\n        :param side: 'long' or 'short' - indicating the direction of the proposed trade\n        :return: A leverage amount, which is between 1.0 and max_leverage.\n        \"\"\"\n        return 1.0\n

All profit calculations include leverage. Stoploss / ROI also include leverage in their calculation. Defining a stoploss of 10% at 10x leverage would trigger the stoploss with a 1% move to the downside.

"},{"location":"strategy-customization/","title":"Strategy Customization","text":"

This page explains how to customize your strategies, add new indicators and set up trading rules.

Please familiarize yourself with Freqtrade basics first, which provides overall info on how the bot operates.

"},{"location":"strategy-customization/#develop-your-own-strategy","title":"Develop your own strategy","text":"

The bot includes a default strategy file. Also, several other strategies are available in the strategy repository.

You will however most likely have your own idea for a strategy. This document intends to help you convert your strategy idea into your own strategy.

To get started, use freqtrade new-strategy --strategy AwesomeStrategy (you can obviously use your own naming for your strategy). This will create a new strategy file from a template, which will be located under user_data/strategies/AwesomeStrategy.py.

Note

This is just a template file, which will most likely not be profitable out of the box.

Different template levels

freqtrade new-strategy has an additional parameter, --template, which controls the amount of pre-build information you get in the created strategy. Use --template minimal to get an empty strategy without any indicator examples, or --template advanced to get a template with most callbacks defined.

"},{"location":"strategy-customization/#anatomy-of-a-strategy","title":"Anatomy of a strategy","text":"

A strategy file contains all the information needed to build a good strategy:

  • Indicators
  • Entry strategy rules
  • Exit strategy rules
  • Minimal ROI recommended
  • Stoploss strongly recommended

The bot also include a sample strategy called SampleStrategy you can update: user_data/strategies/sample_strategy.py. You can test it with the parameter: --strategy SampleStrategy

Additionally, there is an attribute called INTERFACE_VERSION, which defines the version of the strategy interface the bot should use. The current version is 3 - which is also the default when it's not set explicitly in the strategy.

Future versions will require this to be set.

freqtrade trade --strategy AwesomeStrategy\n

For the following section we will use the user_data/strategies/sample_strategy.py file as reference.

Strategies and Backtesting

To avoid problems and unexpected differences between Backtesting and dry/live modes, please be aware that during backtesting the full time range is passed to the populate_*() methods at once. It is therefore best to use vectorized operations (across the whole dataframe, not loops) and avoid index referencing (df.iloc[-1]), but instead use df.shift() to get to the previous candle.

Warning: Using future data

Since backtesting passes the full time range to the populate_*() methods, the strategy author needs to take care to avoid having the strategy utilize data from the future. Some common patterns for this are listed in the Common Mistakes section of this document.

"},{"location":"strategy-customization/#dataframe","title":"Dataframe","text":"

Freqtrade uses pandas to store/provide the candlestick (OHLCV) data. Pandas is a great library developed for processing large amounts of data.

Each row in a dataframe corresponds to one candle on a chart, with the latest candle always being the last in the dataframe (sorted by date).

> dataframe.head()\n                       date      open      high       low     close     volume\n0 2021-11-09 23:25:00+00:00  67279.67  67321.84  67255.01  67300.97   44.62253\n1 2021-11-09 23:30:00+00:00  67300.97  67301.34  67183.03  67187.01   61.38076\n2 2021-11-09 23:35:00+00:00  67187.02  67187.02  67031.93  67123.81  113.42728\n3 2021-11-09 23:40:00+00:00  67123.80  67222.40  67080.33  67160.48   78.96008\n4 2021-11-09 23:45:00+00:00  67160.48  67160.48  66901.26  66943.37  111.39292\n

Pandas provides fast ways to calculate metrics. To benefit from this speed, it's advised to not use loops, but use vectorized methods instead.

Vectorized operations perform calculations across the whole range of data and are therefore, compared to looping through each row, a lot faster when calculating indicators.

As a dataframe is a table, simple python comparisons like the following will not work

    if dataframe['rsi'] > 30:\n        dataframe['enter_long'] = 1\n

The above section will fail with The truth value of a Series is ambiguous. [...].

This must instead be written in a pandas-compatible way, so the operation is performed across the whole dataframe.

    dataframe.loc[\n        (dataframe['rsi'] > 30)\n    , 'enter_long'] = 1\n

With this section, you have a new column in your dataframe, which has 1 assigned whenever RSI is above 30.

"},{"location":"strategy-customization/#customize-indicators","title":"Customize Indicators","text":"

Buy and sell signals need indicators. You can add more indicators by extending the list contained in the method populate_indicators() from your strategy file.

You should only add the indicators used in either populate_entry_trend(), populate_exit_trend(), or to populate another indicator, otherwise performance may suffer.

It's important to always return the dataframe without removing/modifying the columns \"open\", \"high\", \"low\", \"close\", \"volume\", otherwise these fields would contain something unexpected.

Sample:

def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n\"\"\"\n    Adds several different TA indicators to the given DataFrame\n\n    Performance Note: For the best performance be frugal on the number of indicators\n    you are using. Let uncomment only the indicator you are using in your strategies\n    or your hyperopt configuration, otherwise you will waste your memory and CPU usage.\n    :param dataframe: Dataframe with data from the exchange\n    :param metadata: Additional information, like the currently traded pair\n    :return: a Dataframe with all mandatory indicators for the strategies\n    \"\"\"\n    dataframe['sar'] = ta.SAR(dataframe)\n    dataframe['adx'] = ta.ADX(dataframe)\n    stoch = ta.STOCHF(dataframe)\n    dataframe['fastd'] = stoch['fastd']\n    dataframe['fastk'] = stoch['fastk']\n    dataframe['blower'] = ta.BBANDS(dataframe, nbdevup=2, nbdevdn=2)['lowerband']\n    dataframe['sma'] = ta.SMA(dataframe, timeperiod=40)\n    dataframe['tema'] = ta.TEMA(dataframe, timeperiod=9)\n    dataframe['mfi'] = ta.MFI(dataframe)\n    dataframe['rsi'] = ta.RSI(dataframe)\n    dataframe['ema5'] = ta.EMA(dataframe, timeperiod=5)\n    dataframe['ema10'] = ta.EMA(dataframe, timeperiod=10)\n    dataframe['ema50'] = ta.EMA(dataframe, timeperiod=50)\n    dataframe['ema100'] = ta.EMA(dataframe, timeperiod=100)\n    dataframe['ao'] = awesome_oscillator(dataframe)\n    macd = ta.MACD(dataframe)\n    dataframe['macd'] = macd['macd']\n    dataframe['macdsignal'] = macd['macdsignal']\n    dataframe['macdhist'] = macd['macdhist']\n    hilbert = ta.HT_SINE(dataframe)\n    dataframe['htsine'] = hilbert['sine']\n    dataframe['htleadsine'] = hilbert['leadsine']\n    dataframe['plus_dm'] = ta.PLUS_DM(dataframe)\n    dataframe['plus_di'] = ta.PLUS_DI(dataframe)\n    dataframe['minus_dm'] = ta.MINUS_DM(dataframe)\n    dataframe['minus_di'] = ta.MINUS_DI(dataframe)\n    return dataframe\n

Want more indicator examples?

Look into the user_data/strategies/sample_strategy.py. Then uncomment indicators you need.

"},{"location":"strategy-customization/#indicator-libraries","title":"Indicator libraries","text":"

Out of the box, freqtrade installs the following technical libraries:

  • ta-lib
  • pandas-ta
  • technical

Additional technical libraries can be installed as necessary, or custom indicators may be written / invented by the strategy author.

"},{"location":"strategy-customization/#strategy-startup-period","title":"Strategy startup period","text":"

Most indicators have an instable startup period, in which they are either not available (NaN), or the calculation is incorrect. This can lead to inconsistencies, since Freqtrade does not know how long this instable period should be. To account for this, the strategy can be assigned the startup_candle_count attribute. This should be set to the maximum number of candles that the strategy requires to calculate stable indicators. In the case where a user includes higher timeframes with informative pairs, the startup_candle_count does not necessarily change. The value is the maximum period (in candles) that any of the informatives timeframes need to compute stable indicators.

In this example strategy, this should be set to 100 (startup_candle_count = 100), since the longest needed history is 100 candles.

    dataframe['ema100'] = ta.EMA(dataframe, timeperiod=100)\n

By letting the bot know how much history is needed, backtest trades can start at the specified timerange during backtesting and hyperopt.

Using x calls to get OHLCV

If you receive a warning like WARNING - Using 3 calls to get OHLCV. This can result in slower operations for the bot. Please check if you really need 1500 candles for your strategy - you should consider if you really need this much historic data for your signals. Having this will cause Freqtrade to make multiple calls for the same pair, which will obviously be slower than one network request. As a consequence, Freqtrade will take longer to refresh candles - and should therefore be avoided if possible. This is capped to 5 total calls to avoid overloading the exchange, or make freqtrade too slow.

Warning

startup_candle_count should be below ohlcv_candle_limit * 5 (which is 500 * 5 for most exchanges) - since only this amount of candles will be available during Dry-Run/Live Trade operations.

"},{"location":"strategy-customization/#example","title":"Example","text":"

Let's try to backtest 1 month (January 2019) of 5m candles using an example strategy with EMA100, as above.

freqtrade backtesting --timerange 20190101-20190201 --timeframe 5m\n

Assuming startup_candle_count is set to 100, backtesting knows it needs 100 candles to generate valid buy signals. It will load data from 20190101 - (100 * 5m) - which is ~2018-12-31 15:30:00. If this data is available, indicators will be calculated with this extended timerange. The instable startup period (up to 2019-01-01 00:00:00) will then be removed before starting backtesting.

Note

If data for the startup period is not available, then the timerange will be adjusted to account for this startup period - so Backtesting would start at 2019-01-01 08:30:00.

"},{"location":"strategy-customization/#entry-signal-rules","title":"Entry signal rules","text":"

Edit the method populate_entry_trend() in your strategy file to update your entry strategy.

It's important to always return the dataframe without removing/modifying the columns \"open\", \"high\", \"low\", \"close\", \"volume\", otherwise these fields would contain something unexpected.

This method will also define a new column, \"enter_long\" (\"enter_short\" for shorts), which needs to contain 1 for entries, and 0 for \"no action\". enter_long is a mandatory column that must be set even if the strategy is shorting only.

Sample from user_data/strategies/sample_strategy.py:

def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n\"\"\"\n    Based on TA indicators, populates the buy signal for the given dataframe\n    :param dataframe: DataFrame populated with indicators\n    :param metadata: Additional information, like the currently traded pair\n    :return: DataFrame with buy column\n    \"\"\"\n    dataframe.loc[\n        (\n            (qtpylib.crossed_above(dataframe['rsi'], 30)) &  # Signal: RSI crosses above 30\n            (dataframe['tema'] <= dataframe['bb_middleband']) &  # Guard\n            (dataframe['tema'] > dataframe['tema'].shift(1)) &  # Guard\n            (dataframe['volume'] > 0)  # Make sure Volume is not 0\n        ),\n        ['enter_long', 'enter_tag']] = (1, 'rsi_cross')\n\n    return dataframe\n
Enter short trades

Short-entries can be created by setting enter_short (corresponds to enter_long for long trades). The enter_tag column remains identical. Short-trades need to be supported by your exchange and market configuration! Please make sure to set can_short appropriately on your strategy if you intend to short.

def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n    dataframe.loc[\n        (\n            (qtpylib.crossed_above(dataframe['rsi'], 30)) &  # Signal: RSI crosses above 30\n            (dataframe['tema'] <= dataframe['bb_middleband']) &  # Guard\n            (dataframe['tema'] > dataframe['tema'].shift(1)) &  # Guard\n            (dataframe['volume'] > 0)  # Make sure Volume is not 0\n        ),\n        ['enter_long', 'enter_tag']] = (1, 'rsi_cross')\n\n    dataframe.loc[\n        (\n            (qtpylib.crossed_below(dataframe['rsi'], 70)) &  # Signal: RSI crosses below 70\n            (dataframe['tema'] > dataframe['bb_middleband']) &  # Guard\n            (dataframe['tema'] < dataframe['tema'].shift(1)) &  # Guard\n            (dataframe['volume'] > 0)  # Make sure Volume is not 0\n        ),\n        ['enter_short', 'enter_tag']] = (1, 'rsi_cross')\n\n    return dataframe\n

Note

Buying requires sellers to buy from - therefore volume needs to be > 0 (dataframe['volume'] > 0) to make sure that the bot does not buy/sell in no-activity periods.

"},{"location":"strategy-customization/#exit-signal-rules","title":"Exit signal rules","text":"

Edit the method populate_exit_trend() into your strategy file to update your exit strategy. The exit-signal is only used for exits if use_exit_signal is set to true in the configuration. use_exit_signal will not influence signal collision rules - which will still apply and can prevent entries.

It's important to always return the dataframe without removing/modifying the columns \"open\", \"high\", \"low\", \"close\", \"volume\", otherwise these fields would contain something unexpected.

This method will also define a new column, \"exit_long\" (\"exit_short\" for shorts), which needs to contain 1 for exits, and 0 for \"no action\".

Sample from user_data/strategies/sample_strategy.py:

def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n\"\"\"\n    Based on TA indicators, populates the exit signal for the given dataframe\n    :param dataframe: DataFrame populated with indicators\n    :param metadata: Additional information, like the currently traded pair\n    :return: DataFrame with buy column\n    \"\"\"\n    dataframe.loc[\n        (\n            (qtpylib.crossed_above(dataframe['rsi'], 70)) &  # Signal: RSI crosses above 70\n            (dataframe['tema'] > dataframe['bb_middleband']) &  # Guard\n            (dataframe['tema'] < dataframe['tema'].shift(1)) &  # Guard\n            (dataframe['volume'] > 0)  # Make sure Volume is not 0\n        ),\n        ['exit_long', 'exit_tag']] = (1, 'rsi_too_high')\n    return dataframe\n
Exit short trades

Short-exits can be created by setting exit_short (corresponds to exit_long). The exit_tag column remains identical. Short-trades need to be supported by your exchange and market configuration!

def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n    dataframe.loc[\n        (\n            (qtpylib.crossed_above(dataframe['rsi'], 70)) &  # Signal: RSI crosses above 70\n            (dataframe['tema'] > dataframe['bb_middleband']) &  # Guard\n            (dataframe['tema'] < dataframe['tema'].shift(1)) &  # Guard\n            (dataframe['volume'] > 0)  # Make sure Volume is not 0\n        ),\n        ['exit_long', 'exit_tag']] = (1, 'rsi_too_high')\n    dataframe.loc[\n        (\n            (qtpylib.crossed_below(dataframe['rsi'], 30)) &  # Signal: RSI crosses below 30\n            (dataframe['tema'] < dataframe['bb_middleband']) &  # Guard\n            (dataframe['tema'] > dataframe['tema'].shift(1)) &  # Guard\n            (dataframe['volume'] > 0)  # Make sure Volume is not 0\n        ),\n        ['exit_short', 'exit_tag']] = (1, 'rsi_too_low')\n    return dataframe\n
"},{"location":"strategy-customization/#minimal-roi","title":"Minimal ROI","text":"

This dict defines the minimal Return On Investment (ROI) a trade should reach before exiting, independent from the exit signal.

It is of the following format, with the dict key (left side of the colon) being the minutes passed since the trade opened, and the value (right side of the colon) being the percentage.

minimal_roi = {\n    \"40\": 0.0,\n    \"30\": 0.01,\n    \"20\": 0.02,\n    \"0\": 0.04\n}\n

The above configuration would therefore mean:

  • Exit whenever 4% profit was reached
  • Exit when 2% profit was reached (in effect after 20 minutes)
  • Exit when 1% profit was reached (in effect after 30 minutes)
  • Exit when trade is non-loosing (in effect after 40 minutes)

The calculation does include fees.

To disable ROI completely, set it to an insanely high number:

minimal_roi = {\n    \"0\": 100\n}\n

While technically not completely disabled, this would exit once the trade reaches 10000% Profit.

To use times based on candle duration (timeframe), the following snippet can be handy. This will allow you to change the timeframe for the strategy, and ROI times will still be set as candles (e.g. after 3 candles ...)

from freqtrade.exchange import timeframe_to_minutes\n\nclass AwesomeStrategy(IStrategy):\n\n    timeframe = \"1d\"\n    timeframe_mins = timeframe_to_minutes(timeframe)\n    minimal_roi = {\n        \"0\": 0.05,                      # 5% for the first 3 candles\n        str(timeframe_mins * 3): 0.02,  # 2% after 3 candles\n        str(timeframe_mins * 6): 0.01,  # 1% After 6 candles\n    }\n
"},{"location":"strategy-customization/#stoploss","title":"Stoploss","text":"

Setting a stoploss is highly recommended to protect your capital from strong moves against you.

Sample of setting a 10% stoploss:

stoploss = -0.10\n

For the full documentation on stoploss features, look at the dedicated stoploss page.

"},{"location":"strategy-customization/#timeframe","title":"Timeframe","text":"

This is the set of candles the bot should download and use for the analysis. Common values are \"1m\", \"5m\", \"15m\", \"1h\", however all values supported by your exchange should work.

Please note that the same entry/exit signals may work well with one timeframe, but not with the others.

This setting is accessible within the strategy methods as the self.timeframe attribute.

"},{"location":"strategy-customization/#can-short","title":"Can short","text":"

To use short signals in futures markets, you will have to let us know to do so by setting can_short=True. Strategies which enable this will fail to load on spot markets. Disabling of this will have short signals ignored (also in futures markets).

"},{"location":"strategy-customization/#metadata-dict","title":"Metadata dict","text":"

The metadata-dict (available for populate_entry_trend, populate_exit_trend, populate_indicators) contains additional information. Currently this is pair, which can be accessed using metadata['pair'] - and will return a pair in the format XRP/BTC.

The Metadata-dict should not be modified and does not persist information across multiple calls. Instead, have a look at the Storing information section.

"},{"location":"strategy-customization/#strategy-file-loading","title":"Strategy file loading","text":"

By default, freqtrade will attempt to load strategies from all .py files within user_data/strategies.

Assuming your strategy is called AwesomeStrategy, stored in the file user_data/strategies/AwesomeStrategy.py, then you can start freqtrade with freqtrade trade --strategy AwesomeStrategy. Note that we're using the class-name, not the file name.

You can use freqtrade list-strategies to see a list of all strategies Freqtrade is able to load (all strategies in the correct folder). It will also include a \"status\" field, highlighting potential problems.

Customize strategy directory

You can use a different directory by using --strategy-path user_data/otherPath. This parameter is available to all commands that require a strategy.

"},{"location":"strategy-customization/#informative-pairs","title":"Informative Pairs","text":""},{"location":"strategy-customization/#get-data-for-non-tradeable-pairs","title":"Get data for non-tradeable pairs","text":"

Data for additional, informative pairs (reference pairs) can be beneficial for some strategies. OHLCV data for these pairs will be downloaded as part of the regular whitelist refresh process and is available via DataProvider just as other pairs (see below). These parts will not be traded unless they are also specified in the pair whitelist, or have been selected by Dynamic Whitelisting.

The pairs need to be specified as tuples in the format (\"pair\", \"timeframe\"), with pair as the first and timeframe as the second argument.

Sample:

def informative_pairs(self):\n    return [(\"ETH/USDT\", \"5m\"),\n            (\"BTC/TUSD\", \"15m\"),\n            ]\n

A full sample can be found in the DataProvider section.

Warning

As these pairs will be refreshed as part of the regular whitelist refresh, it's best to keep this list short. All timeframes and all pairs can be specified as long as they are available (and active) on the used exchange. It is however better to use resampling to longer timeframes whenever possible to avoid hammering the exchange with too many requests and risk being blocked.

Alternative candle types

Informative_pairs can also provide a 3rd tuple element defining the candle type explicitly. Availability of alternative candle-types will depend on the trading-mode and the exchange. In general, spot pairs cannot be used in futures markets, and futures candles can't be used as informative pairs for spot bots. Details about this may vary, if they do, this can be found in the exchange documentation.

def informative_pairs(self):\n    return [\n        (\"ETH/USDT\", \"5m\", \"\"),   # Uses default candletype, depends on trading_mode (recommended)\n        (\"ETH/USDT\", \"5m\", \"spot\"),   # Forces usage of spot candles (only valid for bots running on spot markets).\n        (\"BTC/TUSD\", \"15m\", \"futures\"),  # Uses futures candles (only bots with `trading_mode=futures`)\n        (\"BTC/TUSD\", \"15m\", \"mark\"),  # Uses mark candles (only bots with `trading_mode=futures`)\n    ]\n
"},{"location":"strategy-customization/#informative-pairs-decorator-informative","title":"Informative pairs decorator (@informative())","text":"

In most common case it is possible to easily define informative pairs by using a decorator. All decorated populate_indicators_* methods run in isolation, not having access to data from other informative pairs, in the end all informative dataframes are merged and passed to main populate_indicators() method. When hyperopting, use of hyperoptable parameter .value attribute is not supported. Please use .range attribute. See optimizing an indicator parameter for more information.

Full documentation
def informative(timeframe: str, asset: str = '',\n                fmt: Optional[Union[str, Callable[[KwArg(str)], str]]] = None,\n                *,\n                candle_type: Optional[CandleType] = None,\n                ffill: bool = True) -> Callable[[PopulateIndicators], PopulateIndicators]:\n\"\"\"\n    A decorator for populate_indicators_Nn(self, dataframe, metadata), allowing these functions to\n    define informative indicators.\n\n    Example usage:\n\n        @informative('1h')\n        def populate_indicators_1h(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n            dataframe['rsi'] = ta.RSI(dataframe, timeperiod=14)\n            return dataframe\n\n    :param timeframe: Informative timeframe. Must always be equal or higher than strategy timeframe.\n    :param asset: Informative asset, for example BTC, BTC/USDT, ETH/BTC. Do not specify to use\n    current pair.\n    :param fmt: Column format (str) or column formatter (callable(name, asset, timeframe)). When not\n    specified, defaults to:\n    * {base}_{quote}_{column}_{timeframe} if asset is specified. \n    * {column}_{timeframe} if asset is not specified.\n    Format string supports these format variables:\n    * {asset} - full name of the asset, for example 'BTC/USDT'.\n    * {base} - base currency in lower case, for example 'eth'.\n    * {BASE} - same as {base}, except in upper case.\n    * {quote} - quote currency in lower case, for example 'usdt'.\n    * {QUOTE} - same as {quote}, except in upper case.\n    * {column} - name of dataframe column.\n    * {timeframe} - timeframe of informative dataframe.\n    :param ffill: ffill dataframe after merging informative pair.\n    :param candle_type: '', mark, index, premiumIndex, or funding_rate\n    \"\"\"\n
Fast and easy way to define informative pairs

Most of the time we do not need power and flexibility offered by merge_informative_pair(), therefore we can use a decorator to quickly define informative pairs.

from datetime import datetime\nfrom freqtrade.persistence import Trade\nfrom freqtrade.strategy import IStrategy, informative\n\nclass AwesomeStrategy(IStrategy):\n\n    # This method is not required. \n    # def informative_pairs(self): ...\n\n    # Define informative upper timeframe for each pair. Decorators can be stacked on same \n    # method. Available in populate_indicators as 'rsi_30m' and 'rsi_1h'.\n    @informative('30m')\n    @informative('1h')\n    def populate_indicators_1h(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n        dataframe['rsi'] = ta.RSI(dataframe, timeperiod=14)\n        return dataframe\n\n    # Define BTC/STAKE informative pair. Available in populate_indicators and other methods as\n    # 'btc_rsi_1h'. Current stake currency should be specified as {stake} format variable \n    # instead of hard-coding actual stake currency. Available in populate_indicators and other \n    # methods as 'btc_usdt_rsi_1h' (when stake currency is USDT).\n    @informative('1h', 'BTC/{stake}')\n    def populate_indicators_btc_1h(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n        dataframe['rsi'] = ta.RSI(dataframe, timeperiod=14)\n        return dataframe\n\n    # Define BTC/ETH informative pair. You must specify quote currency if it is different from\n    # stake currency. Available in populate_indicators and other methods as 'eth_btc_rsi_1h'.\n    @informative('1h', 'ETH/BTC')\n    def populate_indicators_eth_btc_1h(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n        dataframe['rsi'] = ta.RSI(dataframe, timeperiod=14)\n        return dataframe\n\n    # Define BTC/STAKE informative pair. A custom formatter may be specified for formatting\n    # column names. A callable `fmt(**kwargs) -> str` may be specified, to implement custom\n    # formatting. Available in populate_indicators and other methods as 'rsi_upper'.\n    @informative('1h', 'BTC/{stake}', '{column}')\n    def populate_indicators_btc_1h_2(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n        dataframe['rsi_upper'] = ta.RSI(dataframe, timeperiod=14)\n        return dataframe\n\n    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n        # Strategy timeframe indicators for current pair.\n        dataframe['rsi'] = ta.RSI(dataframe, timeperiod=14)\n        # Informative pairs are available in this method.\n        dataframe['rsi_less'] = dataframe['rsi'] < dataframe['rsi_1h']\n        return dataframe\n

Note

Do not use @informative decorator if you need to use data of one informative pair when generating another informative pair. Instead, define informative pairs manually as described in the DataProvider section.

Note

Use string formatting when accessing informative dataframes of other pairs. This will allow easily changing stake currency in config without having to adjust strategy code.

def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n    stake = self.config['stake_currency']\n    dataframe.loc[\n        (\n            (dataframe[f'btc_{stake}_rsi_1h'] < 35)\n            &\n            (dataframe['volume'] > 0)\n        ),\n        ['enter_long', 'enter_tag']] = (1, 'buy_signal_rsi')\n\n    return dataframe\n

Alternatively column renaming may be used to remove stake currency from column names: @informative('1h', 'BTC/{stake}', fmt='{base}_{column}_{timeframe}').

Duplicate method names

Methods tagged with @informative() decorator must always have unique names! Re-using same name (for example when copy-pasting already defined informative method) will overwrite previously defined method and not produce any errors due to limitations of Python programming language. In such cases you will find that indicators created in earlier-defined methods are not available in the dataframe. Carefully review method names and make sure they are unique!

"},{"location":"strategy-customization/#additional-data-dataprovider","title":"Additional data (DataProvider)","text":"

The strategy provides access to the DataProvider. This allows you to get additional data to use in your strategy.

All methods return None in case of failure (do not raise an exception).

Please always check the mode of operation to select the correct method to get data (samples see below).

Hyperopt

Dataprovider is available during hyperopt, however it can only be used in populate_indicators() within a strategy. It is not available in populate_buy() and populate_sell() methods, nor in populate_indicators(), if this method located in the hyperopt file.

"},{"location":"strategy-customization/#possible-options-for-dataprovider","title":"Possible options for DataProvider","text":"
  • available_pairs - Property with tuples listing cached pairs with their timeframe (pair, timeframe).
  • current_whitelist() - Returns a current list of whitelisted pairs. Useful for accessing dynamic whitelists (i.e. VolumePairlist)
  • get_pair_dataframe(pair, timeframe) - This is a universal method, which returns either historical data (for backtesting) or cached live data (for the Dry-Run and Live-Run modes).
  • get_analyzed_dataframe(pair, timeframe) - Returns the analyzed dataframe (after calling populate_indicators(), populate_buy(), populate_sell()) and the time of the latest analysis.
  • historic_ohlcv(pair, timeframe) - Returns historical data stored on disk.
  • market(pair) - Returns market data for the pair: fees, limits, precisions, activity flag, etc. See ccxt documentation for more details on the Market data structure.
  • ohlcv(pair, timeframe) - Currently cached candle (OHLCV) data for the pair, returns DataFrame or empty DataFrame.
  • orderbook(pair, maximum) - Returns latest orderbook data for the pair, a dict with bids/asks with a total of maximum entries.
  • ticker(pair) - Returns current ticker data for the pair. See ccxt documentation for more details on the Ticker data structure.
  • runmode - Property containing the current runmode.
"},{"location":"strategy-customization/#example-usages","title":"Example Usages","text":""},{"location":"strategy-customization/#available_pairs","title":"available_pairs","text":"
for pair, timeframe in self.dp.available_pairs:\n    print(f\"available {pair}, {timeframe}\")\n
"},{"location":"strategy-customization/#current_whitelist","title":"current_whitelist()","text":"

Imagine you've developed a strategy that trades the 5m timeframe using signals generated from a 1d timeframe on the top 10 volume pairs by volume.

The strategy might look something like this:

Scan through the top 10 pairs by volume using the VolumePairList every 5 minutes and use a 14 day RSI to buy and sell.

Due to the limited available data, it's very difficult to resample 5m candles into daily candles for use in a 14 day RSI. Most exchanges limit us to just 500-1000 candles which effectively gives us around 1.74 daily candles. We need 14 days at least!

Since we can't resample the data we will have to use an informative pair; and since the whitelist will be dynamic we don't know which pair(s) to use.

This is where calling self.dp.current_whitelist() comes in handy.

    def informative_pairs(self):\n\n        # get access to all pairs available in whitelist.\n        pairs = self.dp.current_whitelist()\n        # Assign tf to each pair so they can be downloaded and cached for strategy.\n        informative_pairs = [(pair, '1d') for pair in pairs]\n        return informative_pairs\n
Plotting with current_whitelist

Current whitelist is not supported for plot-dataframe, as this command is usually used by providing an explicit pairlist - and would therefore make the return values of this method misleading.

"},{"location":"strategy-customization/#get_pair_dataframepair-timeframe","title":"get_pair_dataframe(pair, timeframe)","text":"
# fetch live / historical candle (OHLCV) data for the first informative pair\ninf_pair, inf_timeframe = self.informative_pairs()[0]\ninformative = self.dp.get_pair_dataframe(pair=inf_pair,\n                                         timeframe=inf_timeframe)\n

Warning about backtesting

In backtesting, dp.get_pair_dataframe() behavior differs depending on where it's called. Within populate_*() methods, dp.get_pair_dataframe() returns the full timerange. Please make sure to not \"look into the future\" to avoid surprises when running in dry/live mode. Within callbacks, you'll get the full timerange up to the current (simulated) candle.

"},{"location":"strategy-customization/#get_analyzed_dataframepair-timeframe","title":"get_analyzed_dataframe(pair, timeframe)","text":"

This method is used by freqtrade internally to determine the last signal. It can also be used in specific callbacks to get the signal that caused the action (see Advanced Strategy Documentation for more details on available callbacks).

# fetch current dataframe\ndataframe, last_updated = self.dp.get_analyzed_dataframe(pair=metadata['pair'],\n                                                         timeframe=self.timeframe)\n

No data available

Returns an empty dataframe if the requested pair was not cached. You can check for this with if dataframe.empty: and handle this case accordingly. This should not happen when using whitelisted pairs.

"},{"location":"strategy-customization/#orderbookpair-maximum","title":"orderbook(pair, maximum)","text":"
if self.dp.runmode.value in ('live', 'dry_run'):\n    ob = self.dp.orderbook(metadata['pair'], 1)\n    dataframe['best_bid'] = ob['bids'][0][0]\n    dataframe['best_ask'] = ob['asks'][0][0]\n

The orderbook structure is aligned with the order structure from ccxt, so the result will look as follows:

{\n'bids': [\n[ price, amount ], // [ float, float ]\n[ price, amount ],\n...\n],\n'asks': [\n[ price, amount ],\n[ price, amount ],\n//...\n],\n//...\n}\n

Therefore, using ob['bids'][0][0] as demonstrated above will result in using the best bid price. ob['bids'][0][1] would look at the amount at this orderbook position.

Warning about backtesting

The order book is not part of the historic data which means backtesting and hyperopt will not work correctly if this method is used, as the method will return uptodate values.

"},{"location":"strategy-customization/#tickerpair","title":"ticker(pair)","text":"
if self.dp.runmode.value in ('live', 'dry_run'):\n    ticker = self.dp.ticker(metadata['pair'])\n    dataframe['last_price'] = ticker['last']\n    dataframe['volume24h'] = ticker['quoteVolume']\n    dataframe['vwap'] = ticker['vwap']\n

Warning

Although the ticker data structure is a part of the ccxt Unified Interface, the values returned by this method can vary for different exchanges. For instance, many exchanges do not return vwap values, some exchanges does not always fills in the last field (so it can be None), etc. So you need to carefully verify the ticker data returned from the exchange and add appropriate error handling / defaults.

Warning about backtesting

This method will always return up-to-date values - so usage during backtesting / hyperopt without runmode checks will lead to wrong results.

"},{"location":"strategy-customization/#send-notification","title":"Send Notification","text":"

The dataprovider .send_msg() function allows you to send custom notifications from your strategy. Identical notifications will only be sent once per candle, unless the 2nd argument (always_send) is set to True.

    self.dp.send_msg(f\"{metadata['pair']} just got hot!\")\n\n    # Force send this notification, avoid caching (Please read warning below!)\n    self.dp.send_msg(f\"{metadata['pair']} just got hot!\", always_send=True)\n

Notifications will only be sent in trading modes (Live/Dry-run) - so this method can be called without conditions for backtesting.

Spamming

You can spam yourself pretty good by setting always_send=True in this method. Use this with great care and only in conditions you know will not happen throughout a candle to avoid a message every 5 seconds.

"},{"location":"strategy-customization/#complete-data-provider-sample","title":"Complete Data-provider sample","text":"
from freqtrade.strategy import IStrategy, merge_informative_pair\nfrom pandas import DataFrame\n\nclass SampleStrategy(IStrategy):\n    # strategy init stuff...\n\n    timeframe = '5m'\n\n    # more strategy init stuff..\n\n    def informative_pairs(self):\n\n        # get access to all pairs available in whitelist.\n        pairs = self.dp.current_whitelist()\n        # Assign tf to each pair so they can be downloaded and cached for strategy.\n        informative_pairs = [(pair, '1d') for pair in pairs]\n        # Optionally Add additional \"static\" pairs\n        informative_pairs += [(\"ETH/USDT\", \"5m\"),\n                              (\"BTC/TUSD\", \"15m\"),\n                            ]\n        return informative_pairs\n\n    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n        if not self.dp:\n            # Don't do anything if DataProvider is not available.\n            return dataframe\n\n        inf_tf = '1d'\n        # Get the informative pair\n        informative = self.dp.get_pair_dataframe(pair=metadata['pair'], timeframe=inf_tf)\n        # Get the 14 day rsi\n        informative['rsi'] = ta.RSI(informative, timeperiod=14)\n\n        # Use the helper function merge_informative_pair to safely merge the pair\n        # Automatically renames the columns and merges a shorter timeframe dataframe and a longer timeframe informative pair\n        # use ffill to have the 1d value available in every row throughout the day.\n        # Without this, comparisons between columns of the original and the informative pair would only work once per day.\n        # Full documentation of this method, see below\n        dataframe = merge_informative_pair(dataframe, informative, self.timeframe, inf_tf, ffill=True)\n\n        # Calculate rsi of the original dataframe (5m timeframe)\n        dataframe['rsi'] = ta.RSI(dataframe, timeperiod=14)\n\n        # Do other stuff\n        # ...\n\n        return dataframe\n\n    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n\n        dataframe.loc[\n            (\n                (qtpylib.crossed_above(dataframe['rsi'], 30)) &  # Signal: RSI crosses above 30\n                (dataframe['rsi_1d'] < 30) &                     # Ensure daily RSI is < 30\n                (dataframe['volume'] > 0)                        # Ensure this candle had volume (important for backtesting)\n            ),\n            ['enter_long', 'enter_tag']] = (1, 'rsi_cross')\n
"},{"location":"strategy-customization/#helper-functions","title":"Helper functions","text":""},{"location":"strategy-customization/#merge_informative_pair","title":"merge_informative_pair()","text":"

This method helps you merge an informative pair to a regular dataframe without lookahead bias. It's there to help you merge the dataframe in a safe and consistent way.

Options:

  • Rename the columns for you to create unique columns
  • Merge the dataframe without lookahead bias
  • Forward-fill (optional)

For a full sample, please refer to the complete data provider example below.

All columns of the informative dataframe will be available on the returning dataframe in a renamed fashion:

Column renaming

Assuming inf_tf = '1d' the resulting columns will be:

'date', 'open', 'high', 'low', 'close', 'rsi'                     # from the original dataframe\n'date_1d', 'open_1d', 'high_1d', 'low_1d', 'close_1d', 'rsi_1d'   # from the informative dataframe\n
Column renaming - 1h

Assuming inf_tf = '1h' the resulting columns will be:

'date', 'open', 'high', 'low', 'close', 'rsi'                     # from the original dataframe\n'date_1h', 'open_1h', 'high_1h', 'low_1h', 'close_1h', 'rsi_1h'   # from the informative dataframe\n
Custom implementation

A custom implementation for this is possible, and can be done as follows:

# Shift date by 1 candle\n# This is necessary since the data is always the \"open date\"\n# and a 15m candle starting at 12:15 should not know the close of the 1h candle from 12:00 to 13:00\nminutes = timeframe_to_minutes(inf_tf)\n# Only do this if the timeframes are different:\ninformative['date_merge'] = informative[\"date\"] + pd.to_timedelta(minutes, 'm')\n\n# Rename columns to be unique\ninformative.columns = [f\"{col}_{inf_tf}\" for col in informative.columns]\n# Assuming inf_tf = '1d' - then the columns will now be:\n# date_1d, open_1d, high_1d, low_1d, close_1d, rsi_1d\n\n# Combine the 2 dataframes\n# all indicators on the informative sample MUST be calculated before this point\ndataframe = pd.merge(dataframe, informative, left_on='date', right_on=f'date_merge_{inf_tf}', how='left')\n# FFill to have the 1d value available in every row throughout the day.\n# Without this, comparisons would only work once per day.\ndataframe = dataframe.ffill()\n

Informative timeframe < timeframe

Using informative timeframes smaller than the dataframe timeframe is not recommended with this method, as it will not use any of the additional information this would provide. To use the more detailed information properly, more advanced methods should be applied (which are out of scope for freqtrade documentation, as it'll depend on the respective need).

"},{"location":"strategy-customization/#stoploss_from_open","title":"stoploss_from_open()","text":"

Stoploss values returned from custom_stoploss must specify a percentage relative to current_rate, but sometimes you may want to specify a stoploss relative to the entry point instead. stoploss_from_open() is a helper function to calculate a stoploss value that can be returned from custom_stoploss which will be equivalent to the desired trade profit above the entry point.

Returning a stoploss relative to the open price from the custom stoploss function

Say the open price was $100, and current_price is $121 (current_profit will be 0.21).

If we want a stop price at 7% above the open price we can call stoploss_from_open(0.07, current_profit, False) which will return 0.1157024793. 11.57% below $121 is $107, which is the same as 7% above $100.

This function will consider leverage - so at 10x leverage, the actual stoploss would be 0.7% above $100 (0.7% * 10x = 7%).

from datetime import datetime\nfrom freqtrade.persistence import Trade\nfrom freqtrade.strategy import IStrategy, stoploss_from_open\n\nclass AwesomeStrategy(IStrategy):\n\n    # ... populate_* methods\n\n    use_custom_stoploss = True\n\n    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,\n                        current_rate: float, current_profit: float, **kwargs) -> float:\n\n        # once the profit has risen above 10%, keep the stoploss at 7% above the open price\n        if current_profit > 0.10:\n            return stoploss_from_open(0.07, current_profit, is_short=trade.is_short, leverage=trade.leverage)\n\n        return 1\n

Full examples can be found in the Custom stoploss section of the Documentation.

Note

Providing invalid input to stoploss_from_open() may produce \"CustomStoploss function did not return valid stoploss\" warnings. This may happen if current_profit parameter is below specified open_relative_stop. Such situations may arise when closing trade is blocked by confirm_trade_exit() method. Warnings can be solved by never blocking stop loss sells by checking exit_reason in confirm_trade_exit(), or by using return stoploss_from_open(...) or 1 idiom, which will request to not change stop loss when current_profit < open_relative_stop.

"},{"location":"strategy-customization/#stoploss_from_absolute","title":"stoploss_from_absolute()","text":"

In some situations it may be confusing to deal with stops relative to current rate. Instead, you may define a stoploss level using an absolute price.

Returning a stoploss using absolute price from the custom stoploss function

If we want to trail a stop price at 2xATR below current price we can call stoploss_from_absolute(current_rate - (candle['atr'] * 2), current_rate, is_short=trade.is_short).

from datetime import datetime\nfrom freqtrade.persistence import Trade\nfrom freqtrade.strategy import IStrategy, stoploss_from_absolute\n\nclass AwesomeStrategy(IStrategy):\n\n    use_custom_stoploss = True\n\n    def populate_indicators_1h(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n        dataframe['atr'] = ta.ATR(dataframe, timeperiod=14)\n        return dataframe\n\n    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,\n                        current_rate: float, current_profit: float, **kwargs) -> float:\n        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)\n        candle = dataframe.iloc[-1].squeeze()\n        return stoploss_from_absolute(current_rate - (candle['atr'] * 2), current_rate, is_short=trade.is_short)\n
"},{"location":"strategy-customization/#additional-data-wallets","title":"Additional data (Wallets)","text":"

The strategy provides access to the wallets object. This contains the current balances on the exchange.

Backtesting / Hyperopt

Wallets behaves differently depending on the function it's called. Within populate_*() methods, it'll return the full wallet as configured. Within callbacks, you'll get the wallet state corresponding to the actual simulated wallet at that point in the simulation process.

Please always check if wallets is available to avoid failures during backtesting.

if self.wallets:\n    free_eth = self.wallets.get_free('ETH')\n    used_eth = self.wallets.get_used('ETH')\n    total_eth = self.wallets.get_total('ETH')\n
"},{"location":"strategy-customization/#possible-options-for-wallets","title":"Possible options for Wallets","text":"
  • get_free(asset) - currently available balance to trade
  • get_used(asset) - currently tied up balance (open orders)
  • get_total(asset) - total available balance - sum of the 2 above
"},{"location":"strategy-customization/#additional-data-trades","title":"Additional data (Trades)","text":"

A history of Trades can be retrieved in the strategy by querying the database.

At the top of the file, import Trade.

from freqtrade.persistence import Trade\n

The following example queries for the current pair and trades from today, however other filters can easily be added.

trades = Trade.get_trades_proxy(pair=metadata['pair'],\n                                open_date=datetime.now(timezone.utc) - timedelta(days=1),\n                                is_open=False,\n            ]).order_by(Trade.close_date).all()\n# Summarize profit for this pair.\ncurdayprofit = sum(trade.close_profit for trade in trades)\n

For a full list of available methods, please consult the Trade object documentation.

Warning

Trade history is not available in populate_* methods during backtesting or hyperopt, and will result in empty results.

"},{"location":"strategy-customization/#prevent-trades-from-happening-for-a-specific-pair","title":"Prevent trades from happening for a specific pair","text":"

Freqtrade locks pairs automatically for the current candle (until that candle is over) when a pair is sold, preventing an immediate re-buy of that pair.

Locked pairs will show the message Pair <pair> is currently locked..

"},{"location":"strategy-customization/#locking-pairs-from-within-the-strategy","title":"Locking pairs from within the strategy","text":"

Sometimes it may be desired to lock a pair after certain events happen (e.g. multiple losing trades in a row).

Freqtrade has an easy method to do this from within the strategy, by calling self.lock_pair(pair, until, [reason]). until must be a datetime object in the future, after which trading will be re-enabled for that pair, while reason is an optional string detailing why the pair was locked.

Locks can also be lifted manually, by calling self.unlock_pair(pair) or self.unlock_reason(<reason>) - providing reason the pair was locked with. self.unlock_reason(<reason>) will unlock all pairs currently locked with the provided reason.

To verify if a pair is currently locked, use self.is_pair_locked(pair).

Note

Locked pairs will always be rounded up to the next candle. So assuming a 5m timeframe, a lock with until set to 10:18 will lock the pair until the candle from 10:15-10:20 will be finished.

Warning

Manually locking pairs is not available during backtesting, only locks via Protections are allowed.

"},{"location":"strategy-customization/#pair-locking-example","title":"Pair locking example","text":"
from freqtrade.persistence import Trade\nfrom datetime import timedelta, datetime, timezone\n# Put the above lines a the top of the strategy file, next to all the other imports\n# --------\n\n# Within populate indicators (or populate_buy):\nif self.config['runmode'].value in ('live', 'dry_run'):\n    # fetch closed trades for the last 2 days\n    trades = Trade.get_trades_proxy(\n        pair=metadata['pair'], is_open=False, \n        open_date=datetime.now(timezone.utc) - timedelta(days=2))\n    # Analyze the conditions you'd like to lock the pair .... will probably be different for every strategy\n    sumprofit = sum(trade.close_profit for trade in trades)\n    if sumprofit < 0:\n        # Lock pair for 12 hours\n        self.lock_pair(metadata['pair'], until=datetime.now(timezone.utc) + timedelta(hours=12))\n
"},{"location":"strategy-customization/#print-created-dataframe","title":"Print created dataframe","text":"

To inspect the created dataframe, you can issue a print-statement in either populate_entry_trend() or populate_exit_trend(). You may also want to print the pair so it's clear what data is currently shown.

def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\n    dataframe.loc[\n        (\n            #>> whatever condition<<<\n        ),\n        ['enter_long', 'enter_tag']] = (1, 'somestring')\n\n    # Print the Analyzed pair\n    print(f\"result for {metadata['pair']}\")\n\n    # Inspect the last 5 rows\n    print(dataframe.tail())\n\n    return dataframe\n

Printing more than a few rows is also possible (simply use print(dataframe) instead of print(dataframe.tail())), however not recommended, as that will be very verbose (~500 lines per pair every 5 seconds).

"},{"location":"strategy-customization/#common-mistakes-when-developing-strategies","title":"Common mistakes when developing strategies","text":""},{"location":"strategy-customization/#peeking-into-the-future-while-backtesting","title":"Peeking into the future while backtesting","text":"

Backtesting analyzes the whole time-range at once for performance reasons. Because of this, strategy authors need to make sure that strategies do not look-ahead into the future. This is a common pain-point, which can cause huge differences between backtesting and dry/live run methods, since they all use data which is not available during dry/live runs, so these strategies will perform well during backtesting, but will fail / perform badly in real conditions.

The following lists some common patterns which should be avoided to prevent frustration:

  • don't use shift(-1). This uses data from the future, which is not available.
  • don't use .iloc[-1] or any other absolute position in the dataframe, this will be different between dry-run and backtesting.
  • don't use dataframe['volume'].mean(). This uses the full DataFrame for backtesting, including data from the future. Use dataframe['volume'].rolling(<window>).mean() instead
  • don't use .resample('1h'). This uses the left border of the interval, so moves data from an hour to the start of the hour. Use .resample('1h', label='right') instead.
"},{"location":"strategy-customization/#colliding-signals","title":"Colliding signals","text":"

When conflicting signals collide (e.g. both 'enter_long' and 'exit_long' are 1), freqtrade will do nothing and ignore the entry signal. This will avoid trades that enter, and exit immediately. Obviously, this can potentially lead to missed entries.

The following rules apply, and entry signals will be ignored if more than one of the 3 signals is set:

  • enter_long -> exit_long, enter_short
  • enter_short -> exit_short, enter_long
"},{"location":"strategy-customization/#further-strategy-ideas","title":"Further strategy ideas","text":"

To get additional Ideas for strategies, head over to the strategy repository. Feel free to use them as they are - but results will depend on the current market situation, pairs used etc. - therefore please backtest the strategy for your exchange/desired pairs first, evaluate carefully, use at your own risk. Feel free to use any of them as inspiration for your own strategies. We're happy to accept Pull Requests containing new Strategies to that repo.

"},{"location":"strategy-customization/#next-step","title":"Next step","text":"

Now you have a perfect strategy you probably want to backtest it. Your next step is to learn How to use the Backtesting.

"},{"location":"strategy_analysis_example/","title":"Strategy analysis example","text":"

Debugging a strategy can be time-consuming. Freqtrade offers helper functions to visualize raw data. The following assumes you work with SampleStrategy, data for 5m timeframe from Binance and have downloaded them into the data directory in the default location. Please follow the documentation for more details.

"},{"location":"strategy_analysis_example/#setup","title":"Setup","text":""},{"location":"strategy_analysis_example/#change-working-directory-to-repository-root","title":"Change Working directory to repository root","text":"
import os\nfrom pathlib import Path\n\n# Change directory\n# Modify this cell to insure that the output shows the correct path.\n# Define all paths relative to the project root shown in the cell output\nproject_root = \"somedir/freqtrade\"\ni=0\ntry:\n    os.chdirdir(project_root)\n    assert Path('LICENSE').is_file()\nexcept:\n    while i<4 and (not Path('LICENSE').is_file()):\n        os.chdir(Path(Path.cwd(), '../'))\n        i+=1\n    project_root = Path.cwd()\nprint(Path.cwd())\n
"},{"location":"strategy_analysis_example/#configure-freqtrade-environment","title":"Configure Freqtrade environment","text":"
from freqtrade.configuration import Configuration\n\n# Customize these according to your needs.\n\n# Initialize empty configuration object\nconfig = Configuration.from_files([])\n# Optionally (recommended), use existing configuration file\n# config = Configuration.from_files([\"user_data/config.json\"])\n\n# Define some constants\nconfig[\"timeframe\"] = \"5m\"\n# Name of the strategy class\nconfig[\"strategy\"] = \"SampleStrategy\"\n# Location of the data\ndata_location = config[\"datadir\"]\n# Pair to analyze - Only use one pair here\npair = \"BTC/USDT\"\n
# Load data using values set above\nfrom freqtrade.data.history import load_pair_history\nfrom freqtrade.enums import CandleType\n\ncandles = load_pair_history(datadir=data_location,\n                            timeframe=config[\"timeframe\"],\n                            pair=pair,\n                            data_format = \"json\",  # Make sure to update this to your data\n                            candle_type=CandleType.SPOT,\n                            )\n\n# Confirm success\nprint(f\"Loaded {len(candles)} rows of data for {pair} from {data_location}\")\ncandles.head()\n
"},{"location":"strategy_analysis_example/#load-and-run-strategy","title":"Load and run strategy","text":"
  • Rerun each time the strategy file is changed
# Load strategy using values set above\nfrom freqtrade.resolvers import StrategyResolver\nfrom freqtrade.data.dataprovider import DataProvider\nstrategy = StrategyResolver.load_strategy(config)\nstrategy.dp = DataProvider(config, None, None)\nstrategy.ft_bot_start()\n\n# Generate buy/sell signals using strategy\ndf = strategy.analyze_ticker(candles, {'pair': pair})\ndf.tail()\n
"},{"location":"strategy_analysis_example/#display-the-trade-details","title":"Display the trade details","text":"
  • Note that using data.head() would also work, however most indicators have some \"startup\" data at the top of the dataframe.
  • Some possible problems * Columns with NaN values at the end of the dataframe * Columns used in crossed*() functions with completely different units
  • Comparison with full backtest * having 200 buy signals as output for one pair from analyze_ticker() does not necessarily mean that 200 trades will be made during backtesting. * Assuming you use only one condition such as, df['rsi'] < 30 as buy condition, this will generate multiple \"buy\" signals for each pair in sequence (until rsi returns > 29). The bot will only buy on the first of these signals (and also only if a trade-slot (\"max_open_trades\") is still available), or on one of the middle signals, as soon as a \"slot\" becomes available.
# Report results\nprint(f\"Generated {df['enter_long'].sum()} entry signals\")\ndata = df.set_index('date', drop=False)\ndata.tail()\n
"},{"location":"strategy_analysis_example/#load-existing-objects-into-a-jupyter-notebook","title":"Load existing objects into a Jupyter notebook","text":"

The following cells assume that you have already generated data using the cli. They will allow you to drill deeper into your results, and perform analysis which otherwise would make the output very difficult to digest due to information overload.

"},{"location":"strategy_analysis_example/#load-backtest-results-to-pandas-dataframe","title":"Load backtest results to pandas dataframe","text":"

Analyze a trades dataframe (also used below for plotting)

from freqtrade.data.btanalysis import load_backtest_data, load_backtest_stats\n\n# if backtest_dir points to a directory, it'll automatically load the last backtest file.\nbacktest_dir = config[\"user_data_dir\"] / \"backtest_results\"\n# backtest_dir can also point to a specific file \n# backtest_dir = config[\"user_data_dir\"] / \"backtest_results/backtest-result-2020-07-01_20-04-22.json\"\n
# You can get the full backtest statistics by using the following command.\n# This contains all information used to generate the backtest result.\nstats = load_backtest_stats(backtest_dir)\n\nstrategy = 'SampleStrategy'\n# All statistics are available per strategy, so if `--strategy-list` was used during backtest, this will be reflected here as well.\n# Example usages:\nprint(stats['strategy'][strategy]['results_per_pair'])\n# Get pairlist used for this backtest\nprint(stats['strategy'][strategy]['pairlist'])\n# Get market change (average change of all pairs from start to end of the backtest period)\nprint(stats['strategy'][strategy]['market_change'])\n# Maximum drawdown ()\nprint(stats['strategy'][strategy]['max_drawdown'])\n# Maximum drawdown start and end\nprint(stats['strategy'][strategy]['drawdown_start'])\nprint(stats['strategy'][strategy]['drawdown_end'])\n\n\n# Get strategy comparison (only relevant if multiple strategies were compared)\nprint(stats['strategy_comparison'])\n
# Load backtested trades as dataframe\ntrades = load_backtest_data(backtest_dir)\n\n# Show value-counts per pair\ntrades.groupby(\"pair\")[\"exit_reason\"].value_counts()\n
"},{"location":"strategy_analysis_example/#plotting-daily-profit-equity-line","title":"Plotting daily profit / equity line","text":"
# Plotting equity line (starting with 0 on day 1 and adding daily profit for each backtested day)\n\nfrom freqtrade.configuration import Configuration\nfrom freqtrade.data.btanalysis import load_backtest_data, load_backtest_stats\nimport plotly.express as px\nimport pandas as pd\n\n# strategy = 'SampleStrategy'\n# config = Configuration.from_files([\"user_data/config.json\"])\n# backtest_dir = config[\"user_data_dir\"] / \"backtest_results\"\n\nstats = load_backtest_stats(backtest_dir)\nstrategy_stats = stats['strategy'][strategy]\n\ndates = []\nprofits = []\nfor date_profit in strategy_stats['daily_profit']:\n    dates.append(date_profit[0])\n    profits.append(date_profit[1])\n\nequity = 0\nequity_daily = []\nfor daily_profit in profits:\n    equity_daily.append(equity)\n    equity += float(daily_profit)\n\n\ndf = pd.DataFrame({'dates': dates,'equity_daily': equity_daily})\n\nfig = px.line(df, x=\"dates\", y=\"equity_daily\")\nfig.show()\n
"},{"location":"strategy_analysis_example/#load-live-trading-results-into-a-pandas-dataframe","title":"Load live trading results into a pandas dataframe","text":"

In case you did already some trading and want to analyze your performance

from freqtrade.data.btanalysis import load_trades_from_db\n\n# Fetch trades from database\ntrades = load_trades_from_db(\"sqlite:///tradesv3.sqlite\")\n\n# Display results\ntrades.groupby(\"pair\")[\"exit_reason\"].value_counts()\n
"},{"location":"strategy_analysis_example/#analyze-the-loaded-trades-for-trade-parallelism","title":"Analyze the loaded trades for trade parallelism","text":"

This can be useful to find the best max_open_trades parameter, when used with backtesting in conjunction with --disable-max-market-positions.

analyze_trade_parallelism() returns a timeseries dataframe with an \"open_trades\" column, specifying the number of open trades for each candle.

from freqtrade.data.btanalysis import analyze_trade_parallelism\n\n# Analyze the above\nparallel_trades = analyze_trade_parallelism(trades, '5m')\n\nparallel_trades.plot()\n
"},{"location":"strategy_analysis_example/#plot-results","title":"Plot results","text":"

Freqtrade offers interactive plotting capabilities based on plotly.

from freqtrade.plot.plotting import  generate_candlestick_graph\n# Limit graph period to keep plotly quick and reactive\n\n# Filter trades to one pair\ntrades_red = trades.loc[trades['pair'] == pair]\n\ndata_red = data['2019-06-01':'2019-06-10']\n# Generate candlestick graph\ngraph = generate_candlestick_graph(pair=pair,\n                                   data=data_red,\n                                   trades=trades_red,\n                                   indicators1=['sma20', 'ema50', 'ema55'],\n                                   indicators2=['rsi', 'macd', 'macdsignal', 'macdhist']\n                                  )\n
# Show graph inline\n# graph.show()\n\n# Render graph in a separate window\ngraph.show(renderer=\"browser\")\n
"},{"location":"strategy_analysis_example/#plot-average-profit-per-trade-as-distribution-graph","title":"Plot average profit per trade as distribution graph","text":"
import plotly.figure_factory as ff\n\nhist_data = [trades.profit_ratio]\ngroup_labels = ['profit_ratio']  # name of the dataset\n\nfig = ff.create_distplot(hist_data, group_labels, bin_size=0.01)\nfig.show()\n

Feel free to submit an issue or Pull Request enhancing this document if you would like to share ideas on how to best analyze the data.

"},{"location":"strategy_migration/","title":"Strategy Migration between V2 and V3","text":"

To support new markets and trade-types (namely short trades / trades with leverage), some things had to change in the interface. If you intend on using markets other than spot markets, please migrate your strategy to the new format.

We have put a great effort into keeping compatibility with existing strategies, so if you just want to continue using freqtrade in spot markets, there should be no changes necessary for now.

You can use the quick summary as checklist. Please refer to the detailed sections below for full migration details.

"},{"location":"strategy_migration/#quick-summary-migration-checklist","title":"Quick summary / migration checklist","text":"

Note : forcesell, forcebuy, emergencysell are changed to force_exit, force_enter, emergency_exit respectively.

  • Strategy methods:
    • populate_buy_trend() -> populate_entry_trend()
    • populate_sell_trend() -> populate_exit_trend()
    • custom_sell() -> custom_exit()
    • check_buy_timeout() -> check_entry_timeout()
    • check_sell_timeout() -> check_exit_timeout()
    • New side argument to callbacks without trade object
      • custom_stake_amount
      • confirm_trade_entry
      • custom_entry_price
    • Changed argument name in confirm_trade_exit
  • Dataframe columns:
    • buy -> enter_long
    • sell -> exit_long
    • buy_tag -> enter_tag (used for both long and short trades)
    • New column enter_short and corresponding new column exit_short
  • trade-object now has the following new properties:
    • is_short
    • entry_side
    • exit_side
    • trade_direction
    • renamed: sell_reason -> exit_reason
  • Renamed trade.nr_of_successful_buys to trade.nr_of_successful_entries (mostly relevant for adjust_trade_position())
  • Introduced new leverage callback.
  • Informative pairs can now pass a 3rd element in the Tuple, defining the candle type.
  • @informative decorator now takes an optional candle_type argument.
  • helper methods stoploss_from_open and stoploss_from_absolute now take is_short as additional argument.
  • INTERFACE_VERSION should be set to 3.
  • Strategy/Configuration settings.
    • order_time_in_force buy -> entry, sell -> exit.
    • order_types buy -> entry, sell -> exit.
    • unfilledtimeout buy -> entry, sell -> exit.
    • ignore_buying_expired_candle_after -> moved to root level instead of \"ask_strategy/exit_pricing\"
  • Terminology changes
    • Sell reasons changed to reflect the new naming of \"exit\" instead of sells. Be careful in your strategy if you're using exit_reason checks and eventually update your strategy.
      • sell_signal -> exit_signal
      • custom_sell -> custom_exit
      • force_sell -> force_exit
      • emergency_sell -> emergency_exit
    • Order pricing
      • bid_strategy -> entry_pricing
      • ask_strategy -> exit_pricing
      • ask_last_balance -> price_last_balance
      • bid_last_balance -> price_last_balance
    • Webhook terminology changed from \"sell\" to \"exit\", and from \"buy\" to entry
      • webhookbuy -> entry
      • webhookbuyfill -> entry_fill
      • webhookbuycancel -> entry_cancel
      • webhooksell -> exit
      • webhooksellfill -> exit_fill
      • webhooksellcancel -> exit_cancel
    • Telegram notification settings
      • buy -> entry
      • buy_fill -> entry_fill
      • buy_cancel -> entry_cancel
      • sell -> exit
      • sell_fill -> exit_fill
      • sell_cancel -> exit_cancel
    • Strategy/config settings:
      • use_sell_signal -> use_exit_signal
      • sell_profit_only -> exit_profit_only
      • sell_profit_offset -> exit_profit_offset
      • ignore_roi_if_buy_signal -> ignore_roi_if_entry_signal
      • forcebuy_enable -> force_entry_enable
"},{"location":"strategy_migration/#extensive-explanation","title":"Extensive explanation","text":""},{"location":"strategy_migration/#populate_buy_trend","title":"populate_buy_trend","text":"

In populate_buy_trend() - you will want to change the columns you assign from 'buy' to 'enter_long', as well as the method name from populate_buy_trend to populate_entry_trend.

def populate_buy_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\ndataframe.loc[\n        (\n            (qtpylib.crossed_above(dataframe['rsi'], 30)) &  # Signal: RSI crosses above 30\n            (dataframe['tema'] <= dataframe['bb_middleband']) &  # Guard\n            (dataframe['tema'] > dataframe['tema'].shift(1)) &  # Guard\n            (dataframe['volume'] > 0)  # Make sure Volume is not 0\n        ),\n['buy', 'buy_tag']] = (1, 'rsi_cross')\nreturn dataframe\n

After:

def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\ndataframe.loc[\n        (\n            (qtpylib.crossed_above(dataframe['rsi'], 30)) &  # Signal: RSI crosses above 30\n            (dataframe['tema'] <= dataframe['bb_middleband']) &  # Guard\n            (dataframe['tema'] > dataframe['tema'].shift(1)) &  # Guard\n            (dataframe['volume'] > 0)  # Make sure Volume is not 0\n        ),\n['enter_long', 'enter_tag']] = (1, 'rsi_cross')\nreturn dataframe\n

Please refer to the Strategy documentation on how to enter and exit short trades.

"},{"location":"strategy_migration/#populate_sell_trend","title":"populate_sell_trend","text":"

Similar to populate_buy_trend, populate_sell_trend() will be renamed to populate_exit_trend(). We'll also change the column from 'sell' to 'exit_long'.

def populate_sell_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\ndataframe.loc[\n        (\n            (qtpylib.crossed_above(dataframe['rsi'], 70)) &  # Signal: RSI crosses above 70\n            (dataframe['tema'] > dataframe['bb_middleband']) &  # Guard\n            (dataframe['tema'] < dataframe['tema'].shift(1)) &  # Guard\n            (dataframe['volume'] > 0)  # Make sure Volume is not 0\n        ),\n['sell', 'exit_tag']] = (1, 'some_exit_tag')\nreturn dataframe\n

After

def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:\ndataframe.loc[\n        (\n            (qtpylib.crossed_above(dataframe['rsi'], 70)) &  # Signal: RSI crosses above 70\n            (dataframe['tema'] > dataframe['bb_middleband']) &  # Guard\n            (dataframe['tema'] < dataframe['tema'].shift(1)) &  # Guard\n            (dataframe['volume'] > 0)  # Make sure Volume is not 0\n        ),\n['exit_long', 'exit_tag']] = (1, 'some_exit_tag')\nreturn dataframe\n

Please refer to the Strategy documentation on how to enter and exit short trades.

"},{"location":"strategy_migration/#custom_sell","title":"custom_sell","text":"

custom_sell has been renamed to custom_exit. It's now also being called for every iteration, independent of current profit and exit_profit_only settings.

class AwesomeStrategy(IStrategy):\ndef custom_sell(self, pair: str, trade: 'Trade', current_time: 'datetime', current_rate: float,\ncurrent_profit: float, **kwargs):\n        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)\n        last_candle = dataframe.iloc[-1].squeeze()\n        # ...\n
class AwesomeStrategy(IStrategy):\ndef custom_exit(self, pair: str, trade: 'Trade', current_time: 'datetime', current_rate: float,\ncurrent_profit: float, **kwargs):\n        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)\n        last_candle = dataframe.iloc[-1].squeeze()\n        # ...\n
"},{"location":"strategy_migration/#custom_entry_timeout","title":"custom_entry_timeout","text":"

check_buy_timeout() has been renamed to check_entry_timeout(), and check_sell_timeout() has been renamed to check_exit_timeout().

class AwesomeStrategy(IStrategy):\ndef check_buy_timeout(self, pair: str, trade: 'Trade', order: dict, \ncurrent_time: datetime, **kwargs) -> bool:\n        return False\n\ndef check_sell_timeout(self, pair: str, trade: 'Trade', order: dict, \ncurrent_time: datetime, **kwargs) -> bool:\n        return False \n
class AwesomeStrategy(IStrategy):\ndef check_entry_timeout(self, pair: str, trade: 'Trade', order: 'Order', \ncurrent_time: datetime, **kwargs) -> bool:\n        return False\n\ndef check_exit_timeout(self, pair: str, trade: 'Trade', order: 'Order', \ncurrent_time: datetime, **kwargs) -> bool:\n        return False \n
"},{"location":"strategy_migration/#custom_stake_amount","title":"custom_stake_amount","text":"

New string argument side - which can be either \"long\" or \"short\".

class AwesomeStrategy(IStrategy):\n    def custom_stake_amount(self, pair: str, current_time: datetime, current_rate: float,\n                            proposed_stake: float, min_stake: Optional[float], max_stake: float,\nentry_tag: Optional[str], **kwargs) -> float:\n# ... \n        return proposed_stake\n
class AwesomeStrategy(IStrategy):\n    def custom_stake_amount(self, pair: str, current_time: datetime, current_rate: float,\n                            proposed_stake: float, min_stake: Optional[float], max_stake: float,\nentry_tag: Optional[str], side: str, **kwargs) -> float:\n# ... \n        return proposed_stake\n
"},{"location":"strategy_migration/#confirm_trade_entry","title":"confirm_trade_entry","text":"

New string argument side - which can be either \"long\" or \"short\".

class AwesomeStrategy(IStrategy):\n    def confirm_trade_entry(self, pair: str, order_type: str, amount: float, rate: float,\n                            time_in_force: str, current_time: datetime, entry_tag: Optional[str], \n**kwargs) -> bool:\nreturn True\n

After:

class AwesomeStrategy(IStrategy):\n    def confirm_trade_entry(self, pair: str, order_type: str, amount: float, rate: float,\n                            time_in_force: str, current_time: datetime, entry_tag: Optional[str], \nside: str, **kwargs) -> bool:\nreturn True\n
"},{"location":"strategy_migration/#confirm_trade_exit","title":"confirm_trade_exit","text":"

Changed argument sell_reason to exit_reason. For compatibility, sell_reason will still be provided for a limited time.

class AwesomeStrategy(IStrategy):\n    def confirm_trade_exit(self, pair: str, trade: Trade, order_type: str, amount: float,\nrate: float, time_in_force: str, sell_reason: str,\ncurrent_time: datetime, **kwargs) -> bool:\n    return True\n

After:

class AwesomeStrategy(IStrategy):\n    def confirm_trade_exit(self, pair: str, trade: Trade, order_type: str, amount: float,\nrate: float, time_in_force: str, exit_reason: str,\ncurrent_time: datetime, **kwargs) -> bool:\n    return True\n
"},{"location":"strategy_migration/#custom_entry_price","title":"custom_entry_price","text":"

New string argument side - which can be either \"long\" or \"short\".

class AwesomeStrategy(IStrategy):\n    def custom_entry_price(self, pair: str, current_time: datetime, proposed_rate: float,\nentry_tag: Optional[str], **kwargs) -> float:\nreturn proposed_rate\n

After:

class AwesomeStrategy(IStrategy):\n    def custom_entry_price(self, pair: str, current_time: datetime, proposed_rate: float,\nentry_tag: Optional[str], side: str, **kwargs) -> float:\nreturn proposed_rate\n
"},{"location":"strategy_migration/#adjust-trade-position-changes","title":"Adjust trade position changes","text":"

While adjust-trade-position itself did not change, you should no longer use trade.nr_of_successful_buys - and instead use trade.nr_of_successful_entries, which will also include short entries.

"},{"location":"strategy_migration/#helper-methods","title":"Helper methods","text":"

Added argument \"is_short\" to stoploss_from_open and stoploss_from_absolute. This should be given the value of trade.is_short.

    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,\n                        current_rate: float, current_profit: float, **kwargs) -> float:\n        # once the profit has risen above 10%, keep the stoploss at 7% above the open price\n        if current_profit > 0.10:\nreturn stoploss_from_open(0.07, current_profit)\nreturn stoploss_from_absolute(current_rate - (candle['atr'] * 2), current_rate)\nreturn 1\n

After:

    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,\n                        current_rate: float, current_profit: float, **kwargs) -> float:\n        # once the profit has risen above 10%, keep the stoploss at 7% above the open price\n        if current_profit > 0.10:\nreturn stoploss_from_open(0.07, current_profit, is_short=trade.is_short)\nreturn stoploss_from_absolute(current_rate - (candle['atr'] * 2), current_rate, is_short=trade.is_short)\n
"},{"location":"strategy_migration/#strategyconfiguration-settings","title":"Strategy/Configuration settings","text":""},{"location":"strategy_migration/#order_time_in_force","title":"order_time_in_force","text":"

order_time_in_force attributes changed from \"buy\" to \"entry\" and \"sell\" to \"exit\".

    order_time_in_force: Dict = {\n        \"buy\": \"gtc\",\n        \"sell\": \"gtc\",\n    }\n

After:

    order_time_in_force: Dict = {\n\"entry\": \"GTC\",\n\"exit\": \"GTC\",\n}\n
"},{"location":"strategy_migration/#order_types","title":"order_types","text":"

order_types have changed all wordings from buy to entry - and sell to exit. And two words are joined with _.

    order_types = {\n\"buy\": \"limit\",\n\"sell\": \"limit\",\n\"emergencysell\": \"market\",\n\"forcesell\": \"market\",\n\"forcebuy\": \"market\",\n\"stoploss\": \"market\",\n        \"stoploss_on_exchange\": false,\n        \"stoploss_on_exchange_interval\": 60\n    }\n

After:

    order_types = {\n\"entry\": \"limit\",\n\"exit\": \"limit\",\n\"emergency_exit\": \"market\",\n\"force_exit\": \"market\",\n\"force_entry\": \"market\",\n\"stoploss\": \"market\",\n        \"stoploss_on_exchange\": false,\n        \"stoploss_on_exchange_interval\": 60\n    }\n
"},{"location":"strategy_migration/#strategy-level-settings","title":"Strategy level settings","text":"
  • use_sell_signal -> use_exit_signal
  • sell_profit_only -> exit_profit_only
  • sell_profit_offset -> exit_profit_offset
  • ignore_roi_if_buy_signal -> ignore_roi_if_entry_signal
    # These values can be overridden in the config.\nuse_sell_signal = True\nsell_profit_only = True\nsell_profit_offset: 0.01\nignore_roi_if_buy_signal = False\n

After:

    # These values can be overridden in the config.\nuse_exit_signal = True\nexit_profit_only = True\nexit_profit_offset: 0.01\nignore_roi_if_entry_signal = False\n
"},{"location":"strategy_migration/#unfilledtimeout","title":"unfilledtimeout","text":"

unfilledtimeout have changed all wordings from buy to entry - and sell to exit.

unfilledtimeout = {\n\"buy\": 10,\n\"sell\": 10,\n\"exit_timeout_count\": 0,\n        \"unit\": \"minutes\"\n    }\n

After:

unfilledtimeout = {\n\"entry\": 10,\n\"exit\": 10,\n\"exit_timeout_count\": 0,\n        \"unit\": \"minutes\"\n    }\n
"},{"location":"strategy_migration/#order-pricing","title":"order pricing","text":"

Order pricing changed in 2 ways. bid_strategy was renamed to entry_pricing and ask_strategy was renamed to exit_pricing. The attributes ask_last_balance -> price_last_balance and bid_last_balance -> price_last_balance were renamed as well. Also, price-side can now be defined as ask, bid, same or other. Please refer to the pricing documentation for more information.

{\n\"bid_strategy\": {\n\"price_side\": \"bid\",\n\"use_order_book\": true,\n\"order_book_top\": 1,\n\"ask_last_balance\": 0.0,\n\"check_depth_of_market\": {\n\"enabled\": false,\n\"bids_to_ask_delta\": 1\n}\n},\n\"ask_strategy\":{\n\"price_side\": \"ask\",\n\"use_order_book\": true,\n\"order_book_top\": 1,\n\"bid_last_balance\": 0.0\n\"ignore_buying_expired_candle_after\": 120\n}\n}\n

after:

{\n\"entry_pricing\": {\n\"price_side\": \"same\",\n\"use_order_book\": true,\n\"order_book_top\": 1,\n\"price_last_balance\": 0.0,\n\"check_depth_of_market\": {\n\"enabled\": false,\n\"bids_to_ask_delta\": 1\n}\n},\n\"exit_pricing\":{\n\"price_side\": \"same\",\n\"use_order_book\": true,\n\"order_book_top\": 1,\n\"price_last_balance\": 0.0\n},\n\"ignore_buying_expired_candle_after\": 120\n}\n
"},{"location":"strategy_migration/#freqai-strategy","title":"FreqAI strategy","text":"

The populate_any_indicators() method has been split into feature_engineering_expand_all(), feature_engineering_expand_basic(), feature_engineering_standard() andset_freqai_targets().

For each new function, the pair (and timeframe where necessary) will be automatically added to the column. As such, the definition of features becomes much simpler with the new logic.

For a full explanation of each method, please go to the corresponding freqAI documentation page

def populate_any_indicators(\n        self, pair, df, tf, informative=None, set_generalized_indicators=False\n    ):\n\n        if informative is None:\n            informative = self.dp.get_pair_dataframe(pair, tf)\n\n        # first loop is automatically duplicating indicators for time periods\n        for t in self.freqai_info[\"feature_parameters\"][\"indicator_periods_candles\"]:\n\n            t = int(t)\ninformative[f\"%-{pair}rsi-period_{t}\"] = ta.RSI(informative, timeperiod=t)\ninformative[f\"%-{pair}mfi-period_{t}\"] = ta.MFI(informative, timeperiod=t)\ninformative[f\"%-{pair}adx-period_{t}\"] = ta.ADX(informative, timeperiod=t)\ninformative[f\"%-{pair}sma-period_{t}\"] = ta.SMA(informative, timeperiod=t)\ninformative[f\"%-{pair}ema-period_{t}\"] = ta.EMA(informative, timeperiod=t)\nbollinger = qtpylib.bollinger_bands(\nqtpylib.typical_price(informative), window=t, stds=2.2\n)\ninformative[f\"{pair}bb_lowerband-period_{t}\"] = bollinger[\"lower\"]\ninformative[f\"{pair}bb_middleband-period_{t}\"] = bollinger[\"mid\"]\ninformative[f\"{pair}bb_upperband-period_{t}\"] = bollinger[\"upper\"]\ninformative[f\"%-{pair}bb_width-period_{t}\"] = (\ninformative[f\"{pair}bb_upperband-period_{t}\"]\n- informative[f\"{pair}bb_lowerband-period_{t}\"]\n) / informative[f\"{pair}bb_middleband-period_{t}\"]\ninformative[f\"%-{pair}close-bb_lower-period_{t}\"] = (\ninformative[\"close\"] / informative[f\"{pair}bb_lowerband-period_{t}\"]\n)\ninformative[f\"%-{pair}roc-period_{t}\"] = ta.ROC(informative, timeperiod=t)\ninformative[f\"%-{pair}relative_volume-period_{t}\"] = (\ninformative[\"volume\"] / informative[\"volume\"].rolling(t).mean()\n) # (1)\ninformative[f\"%-{pair}pct-change\"] = informative[\"close\"].pct_change()\ninformative[f\"%-{pair}raw_volume\"] = informative[\"volume\"]\ninformative[f\"%-{pair}raw_price\"] = informative[\"close\"]\n# (2)\nindicators = [col for col in informative if col.startswith(\"%\")]\n        # This loop duplicates and shifts all indicators to add a sense of recency to data\n        for n in range(self.freqai_info[\"feature_parameters\"][\"include_shifted_candles\"] + 1):\n            if n == 0:\n                continue\n            informative_shift = informative[indicators].shift(n)\n            informative_shift = informative_shift.add_suffix(\"_shift-\" + str(n))\n            informative = pd.concat((informative, informative_shift), axis=1)\n\n        df = merge_informative_pair(df, informative, self.config[\"timeframe\"], tf, ffill=True)\n        skip_columns = [\n            (s + \"_\" + tf) for s in [\"date\", \"open\", \"high\", \"low\", \"close\", \"volume\"]\n        ]\n        df = df.drop(columns=skip_columns)\n\n        # Add generalized indicators here (because in live, it will call this\n        # function to populate indicators during training). Notice how we ensure not to\n        # add them multiple times\n        if set_generalized_indicators:\ndf[\"%-day_of_week\"] = (df[\"date\"].dt.dayofweek + 1) / 7\ndf[\"%-hour_of_day\"] = (df[\"date\"].dt.hour + 1) / 25\n# (3)\n# user adds targets here by prepending them with &- (see convention below)\ndf[\"&-s_close\"] = (\ndf[\"close\"]\n.shift(-self.freqai_info[\"feature_parameters\"][\"label_period_candles\"])\n.rolling(self.freqai_info[\"feature_parameters\"][\"label_period_candles\"])\n.mean()\n/ df[\"close\"]\n- 1\n)  # (4)\nreturn df\n
  1. Features - Move to feature_engineering_expand_all
  2. Basic features, not expanded across include_periods_candles - move tofeature_engineering_expand_basic().
  3. Standard features which should not be expanded - move to feature_engineering_standard().
  4. Targets - Move this part to set_freqai_targets().
"},{"location":"strategy_migration/#freqai-feature-engineering-expand-all","title":"freqai - feature engineering expand all","text":"

Features will now expand automatically. As such, the expansion loops, as well as the {pair} / {timeframe} parts will need to be removed.

    def feature_engineering_expand_all(self, dataframe, period, **kwargs) -> DataFrame::\n\"\"\"\n        *Only functional with FreqAI enabled strategies*\n        This function will automatically expand the defined features on the config defined\n        `indicator_periods_candles`, `include_timeframes`, `include_shifted_candles`, and\n        `include_corr_pairs`. In other words, a single feature defined in this function\n        will automatically expand to a total of\n        `indicator_periods_candles` * `include_timeframes` * `include_shifted_candles` *\n        `include_corr_pairs` numbers of features added to the model.\n\n        All features must be prepended with `%` to be recognized by FreqAI internals.\n\n        More details on how these config defined parameters accelerate feature engineering\n        in the documentation at:\n\n        https://www.freqtrade.io/en/latest/freqai-parameter-table/#feature-parameters\n\n        https://www.freqtrade.io/en/latest/freqai-feature-engineering/#defining-the-features\n\n        :param df: strategy dataframe which will receive the features\n        :param period: period of the indicator - usage example:\n        dataframe[\"%-ema-period\"] = ta.EMA(dataframe, timeperiod=period)\n        \"\"\"\n\n        dataframe[\"%-rsi-period\"] = ta.RSI(dataframe, timeperiod=period)\n        dataframe[\"%-mfi-period\"] = ta.MFI(dataframe, timeperiod=period)\n        dataframe[\"%-adx-period\"] = ta.ADX(dataframe, timeperiod=period)\n        dataframe[\"%-sma-period\"] = ta.SMA(dataframe, timeperiod=period)\n        dataframe[\"%-ema-period\"] = ta.EMA(dataframe, timeperiod=period)\n\n        bollinger = qtpylib.bollinger_bands(\n            qtpylib.typical_price(dataframe), window=period, stds=2.2\n        )\n        dataframe[\"bb_lowerband-period\"] = bollinger[\"lower\"]\n        dataframe[\"bb_middleband-period\"] = bollinger[\"mid\"]\n        dataframe[\"bb_upperband-period\"] = bollinger[\"upper\"]\n\n        dataframe[\"%-bb_width-period\"] = (\n            dataframe[\"bb_upperband-period\"]\n            - dataframe[\"bb_lowerband-period\"]\n        ) / dataframe[\"bb_middleband-period\"]\n        dataframe[\"%-close-bb_lower-period\"] = (\n            dataframe[\"close\"] / dataframe[\"bb_lowerband-period\"]\n        )\n\n        dataframe[\"%-roc-period\"] = ta.ROC(dataframe, timeperiod=period)\n\n        dataframe[\"%-relative_volume-period\"] = (\n            dataframe[\"volume\"] / dataframe[\"volume\"].rolling(period).mean()\n        )\n\n        return dataframe\n
"},{"location":"strategy_migration/#freqai-feature-engineering-basic","title":"Freqai - feature engineering basic","text":"

Basic features. Make sure to remove the {pair} part from your features.

    def feature_engineering_expand_basic(self, dataframe: DataFrame, **kwargs) -> DataFrame::\n\"\"\"\n        *Only functional with FreqAI enabled strategies*\n        This function will automatically expand the defined features on the config defined\n        `include_timeframes`, `include_shifted_candles`, and `include_corr_pairs`.\n        In other words, a single feature defined in this function\n        will automatically expand to a total of\n        `include_timeframes` * `include_shifted_candles` * `include_corr_pairs`\n        numbers of features added to the model.\n\n        Features defined here will *not* be automatically duplicated on user defined\n        `indicator_periods_candles`\n\n        All features must be prepended with `%` to be recognized by FreqAI internals.\n\n        More details on how these config defined parameters accelerate feature engineering\n        in the documentation at:\n\n        https://www.freqtrade.io/en/latest/freqai-parameter-table/#feature-parameters\n\n        https://www.freqtrade.io/en/latest/freqai-feature-engineering/#defining-the-features\n\n        :param df: strategy dataframe which will receive the features\n        dataframe[\"%-pct-change\"] = dataframe[\"close\"].pct_change()\n        dataframe[\"%-ema-200\"] = ta.EMA(dataframe, timeperiod=200)\n        \"\"\"\n        dataframe[\"%-pct-change\"] = dataframe[\"close\"].pct_change()\n        dataframe[\"%-raw_volume\"] = dataframe[\"volume\"]\n        dataframe[\"%-raw_price\"] = dataframe[\"close\"]\n        return dataframe\n
"},{"location":"strategy_migration/#freqai-feature-engineering-standard","title":"FreqAI - feature engineering standard","text":"
    def feature_engineering_standard(self, dataframe: DataFrame, **kwargs) -> DataFrame:\n\"\"\"\n        *Only functional with FreqAI enabled strategies*\n        This optional function will be called once with the dataframe of the base timeframe.\n        This is the final function to be called, which means that the dataframe entering this\n        function will contain all the features and columns created by all other\n        freqai_feature_engineering_* functions.\n\n        This function is a good place to do custom exotic feature extractions (e.g. tsfresh).\n        This function is a good place for any feature that should not be auto-expanded upon\n        (e.g. day of the week).\n\n        All features must be prepended with `%` to be recognized by FreqAI internals.\n\n        More details about feature engineering available:\n\n        https://www.freqtrade.io/en/latest/freqai-feature-engineering\n\n        :param df: strategy dataframe which will receive the features\n        usage example: dataframe[\"%-day_of_week\"] = (dataframe[\"date\"].dt.dayofweek + 1) / 7\n        \"\"\"\n        dataframe[\"%-day_of_week\"] = dataframe[\"date\"].dt.dayofweek\n        dataframe[\"%-hour_of_day\"] = dataframe[\"date\"].dt.hour\n        return dataframe\n
"},{"location":"strategy_migration/#freqai-set-targets","title":"FreqAI - set Targets","text":"

Targets now get their own, dedicated method.

    def set_freqai_targets(self, dataframe: DataFrame, **kwargs) -> DataFrame:\n\"\"\"\n        *Only functional with FreqAI enabled strategies*\n        Required function to set the targets for the model.\n        All targets must be prepended with `&` to be recognized by the FreqAI internals.\n\n        More details about feature engineering available:\n\n        https://www.freqtrade.io/en/latest/freqai-feature-engineering\n\n        :param df: strategy dataframe which will receive the targets\n        usage example: dataframe[\"&-target\"] = dataframe[\"close\"].shift(-1) / dataframe[\"close\"]\n        \"\"\"\n        dataframe[\"&-s_close\"] = (\n            dataframe[\"close\"]\n            .shift(-self.freqai_info[\"feature_parameters\"][\"label_period_candles\"])\n            .rolling(self.freqai_info[\"feature_parameters\"][\"label_period_candles\"])\n            .mean()\n            / dataframe[\"close\"]\n            - 1\n            )\n\n        return dataframe\n
"},{"location":"telegram-usage/","title":"Telegram usage","text":""},{"location":"telegram-usage/#setup-your-telegram-bot","title":"Setup your Telegram bot","text":"

Below we explain how to create your Telegram Bot, and how to get your Telegram user id.

"},{"location":"telegram-usage/#1-create-your-telegram-bot","title":"1. Create your Telegram bot","text":"

Start a chat with the Telegram BotFather

Send the message /newbot.

BotFather response:

Alright, a new bot. How are we going to call it? Please choose a name for your bot.

Choose the public name of your bot (e.x. Freqtrade bot)

BotFather response:

Good. Now let's choose a username for your bot. It must end in bot. Like this, for example: TetrisBot or tetris_bot.

Choose the name id of your bot and send it to the BotFather (e.g. \"My_own_freqtrade_bot\")

BotFather response:

Done! Congratulations on your new bot. You will find it at t.me/yourbots_name_bot. You can now add a description, about section and profile picture for your bot, see /help for a list of commands. By the way, when you've finished creating your cool bot, ping our Bot Support if you want a better username for it. Just make sure the bot is fully operational before you do this.

Use this token to access the HTTP API: 22222222:APITOKEN

For a description of the Bot API, see this page: https://core.telegram.org/bots/api Father bot will return you the token (API key)

Copy the API Token (22222222:APITOKEN in the above example) and keep use it for the config parameter token.

Don't forget to start the conversation with your bot, by clicking /START button

"},{"location":"telegram-usage/#2-telegram-user_id","title":"2. Telegram user_id","text":""},{"location":"telegram-usage/#get-your-user-id","title":"Get your user id","text":"

Talk to the userinfobot

Get your \"Id\", you will use it for the config parameter chat_id.

"},{"location":"telegram-usage/#use-group-id","title":"Use Group id","text":"

You can use bots in telegram groups by just adding them to the group. You can find the group id by first adding a RawDataBot to your group. The Group id is shown as id in the \"chat\" section, which the RawDataBot will send to you:

\"chat\":{\n\"id\":-1001332619709\n}\n

For the Freqtrade configuration, you can then use the the full value (including - if it's there) as string:

   \"chat_id\": \"-1001332619709\"\n

Using telegram groups

When using telegram groups, you're giving every member of the telegram group access to your freqtrade bot and to all commands possible via telegram. Please make sure that you can trust everyone in the telegram group to avoid unpleasent surprises.

"},{"location":"telegram-usage/#control-telegram-noise","title":"Control telegram noise","text":"

Freqtrade provides means to control the verbosity of your telegram bot. Each setting has the following possible values:

  • on - Messages will be sent, and user will be notified.
  • silent - Message will be sent, Notification will be without sound / vibration.
  • off - Skip sending a message-type all together.

Example configuration showing the different settings:

\"telegram\": {\n\"enabled\": true,\n\"token\": \"your_telegram_token\",\n\"chat_id\": \"your_telegram_chat_id\",\n\"allow_custom_messages\": true,\n\"notification_settings\": {\n\"status\": \"silent\",\n\"warning\": \"on\",\n\"startup\": \"off\",\n\"entry\": \"silent\",\n\"entry_fill\": \"on\",\n\"entry_cancel\": \"silent\",\n\"exit\": {\n\"roi\": \"silent\",\n\"emergency_exit\": \"on\",\n\"force_exit\": \"on\",\n\"exit_signal\": \"silent\",\n\"trailing_stop_loss\": \"on\",\n\"stop_loss\": \"on\",\n\"stoploss_on_exchange\": \"on\",\n\"custom_exit\": \"silent\",\n\"partial_exit\": \"on\"\n},\n\"exit_cancel\": \"on\",\n\"exit_fill\": \"off\",\n\"protection_trigger\": \"off\",\n\"protection_trigger_global\": \"on\",\n\"strategy_msg\": \"off\",\n\"show_candle\": \"off\"\n},\n\"reload\": true,\n\"balance_dust_level\": 0.01\n},\n

entry notifications are sent when the order is placed, while entry_fill notifications are sent when the order is filled on the exchange. exit notifications are sent when the order is placed, while exit_fill notifications are sent when the order is filled on the exchange. *_fill notifications are off by default and must be explicitly enabled. protection_trigger notifications are sent when a protection triggers and protection_trigger_global notifications trigger when global protections are triggered. strategy_msg - Receive notifications from the strategy, sent via self.dp.send_msg() from the strategy more details. show_candle - show candle values as part of entry/exit messages. Only possible values are \"ohlc\" or \"off\".

balance_dust_level will define what the /balance command takes as \"dust\" - Currencies with a balance below this will be shown. allow_custom_messages completely disable strategy messages. reload allows you to disable reload-buttons on selected messages.

"},{"location":"telegram-usage/#create-a-custom-keyboard-command-shortcut-buttons","title":"Create a custom keyboard (command shortcut buttons)","text":"

Telegram allows us to create a custom keyboard with buttons for commands. The default custom keyboard looks like this.

[\n    [\"/daily\", \"/profit\", \"/balance\"], # row 1, 3 commands\n    [\"/status\", \"/status table\", \"/performance\"], # row 2, 3 commands\n    [\"/count\", \"/start\", \"/stop\", \"/help\"] # row 3, 4 commands\n]\n
"},{"location":"telegram-usage/#usage","title":"Usage","text":"

You can create your own keyboard in config.json:

\"telegram\": {\n\"enabled\": true,\n\"token\": \"your_telegram_token\",\n\"chat_id\": \"your_telegram_chat_id\",\n\"keyboard\": [\n[\"/daily\", \"/stats\", \"/balance\", \"/profit\"],\n[\"/status table\", \"/performance\"],\n[\"/reload_config\", \"/count\", \"/logs\"]\n]\n},\n

Supported Commands

Only the following commands are allowed. Command arguments are not supported!

/start, /stop, /status, /status table, /trades, /profit, /performance, /daily, /stats, /count, /locks, /balance, /stopentry, /reload_config, /show_config, /logs, /whitelist, /blacklist, /edge, /help, /version, /marketdir

"},{"location":"telegram-usage/#telegram-commands","title":"Telegram commands","text":"

Per default, the Telegram bot shows predefined commands. Some commands are only available by sending them to the bot. The table below list the official commands. You can ask at any moment for help with /help.

Command Description System commands /start Starts the trader /stop Stops the trader /stopbuy | /stopentry Stops the trader from opening new trades. Gracefully closes open trades according to their rules. /reload_config Reloads the configuration file /show_config Shows part of the current configuration with relevant settings to operation /logs [limit] Show last log messages. /help Show help message /version Show version Status /status Lists all open trades /status <trade_id> Lists one or more specific trade. Separate multiple with a blank space. /status table List all open trades in a table format. Pending buy orders are marked with an asterisk () Pending sell orders are marked with a double asterisk (*) /trades [limit] List all recently closed trades in a table format. /count Displays number of trades used and available /locks Show currently locked pairs. /unlock <pair or lock_id> Remove the lock for this pair (or for this lock id). /marketdir [long | short | even | none] Updates the user managed variable that represents the current market direction. If no direction is provided, the currently set direction will be displayed. Modify Trade states /forceexit <trade_id> | /fx <tradeid> Instantly exits the given trade (Ignoring minimum_roi). /forceexit all | /fx all Instantly exits all open trades (Ignoring minimum_roi). /fx alias for /forceexit /forcelong <pair> [rate] Instantly buys the given pair. Rate is optional and only applies to limit orders. (force_entry_enable must be set to True) /forceshort <pair> [rate] Instantly shorts the given pair. Rate is optional and only applies to limit orders. This will only work on non-spot markets. (force_entry_enable must be set to True) /delete <trade_id> Delete a specific trade from the Database. Tries to close open orders. Requires manual handling of this trade on the exchange. /cancel_open_order <trade_id> | /coo <trade_id> Cancel an open order for a trade. Metrics /profit [<n>] Display a summary of your profit/loss from close trades and some stats about your performance, over the last n days (all trades by default) /performance Show performance of each finished trade grouped by pair /balance Show account balance per currency /daily <n> Shows profit or loss per day, over the last n days (n defaults to 7) /weekly <n> Shows profit or loss per week, over the last n weeks (n defaults to 8) /monthly <n> Shows profit or loss per month, over the last n months (n defaults to 6) /stats Shows Wins / losses by Exit reason as well as Avg. holding durations for buys and sells /exits Shows Wins / losses by Exit reason as well as Avg. holding durations for buys and sells /entries Shows Wins / losses by Exit reason as well as Avg. holding durations for buys and sells /whitelist [sorted] [baseonly] Show the current whitelist. Optionally display in alphabetical order and/or with just the base currency of each pairing. /blacklist [pair] Show the current blacklist, or adds a pair to the blacklist. /edge Show validated pairs by Edge if it is enabled."},{"location":"telegram-usage/#telegram-commands-in-action","title":"Telegram commands in action","text":"

Below, example of Telegram message you will receive for each command.

"},{"location":"telegram-usage/#start","title":"/start","text":"

Status: running

"},{"location":"telegram-usage/#stop","title":"/stop","text":"

Stopping trader ... Status: stopped

"},{"location":"telegram-usage/#stopbuy","title":"/stopbuy","text":"

status: Setting max_open_trades to 0. Run /reload_config to reset.

Prevents the bot from opening new trades by temporarily setting \"max_open_trades\" to 0. Open trades will be handled via their regular rules (ROI / Sell-signal, stoploss, ...).

After this, give the bot time to close off open trades (can be checked via /status table). Once all positions are sold, run /stop to completely stop the bot.

/reload_config resets \"max_open_trades\" to the value set in the configuration and resets this command.

Warning

The stop-buy signal is ONLY active while the bot is running, and is not persisted anyway, so restarting the bot will cause this to reset.

"},{"location":"telegram-usage/#status","title":"/status","text":"

For each open trade, the bot will send you the following message. Enter Tag is configurable via Strategy.

Trade ID: 123 (since 1 days ago) Current Pair: CVC/BTC Direction: Long Leverage: 1.0 Amount: 26.64180098 Enter Tag: Awesome Long Signal Open Rate: 0.00007489 Current Rate: 0.00007489 Unrealized Profit: 12.95% Stoploss: 0.00007389 (-0.02%)

"},{"location":"telegram-usage/#status-table","title":"/status table","text":"

Return the status of all open trades in a table format.

ID L/S    Pair     Since   Profit\n----    --------  -------  --------\n  67 L   SC/BTC    1 d      13.33%\n 123 S   CVC/BTC   1 h      12.95%\n
"},{"location":"telegram-usage/#count","title":"/count","text":"

Return the number of trades used and available.

current    max\n---------  -----\n     2     10\n
"},{"location":"telegram-usage/#profit","title":"/profit","text":"

Return a summary of your profit/loss and performance.

ROI: Close trades \u2219 0.00485701 BTC (2.2%) (15.2 \u03a3%) \u2219 62.968 USD ROI: All trades \u2219 0.00255280 BTC (1.5%) (6.43 \u03a3%) \u2219 33.095 EUR

Total Trade Count: 138 Bot started: 2022-07-11 18:40:44 First Trade opened: 3 days ago Latest Trade opened: 2 minutes ago Avg. Duration: 2:33:45 Best Performing: PAY/BTC: 50.23% Trading volume: 0.5 BTC Profit factor: 1.04 Max Drawdown: 9.23% (0.01255 BTC)

The relative profit of 1.2% is the average profit per trade. The relative profit of 15.2 \u03a3% is be based on the starting capital - so in this case, the starting capital was 0.00485701 * 1.152 = 0.00738 BTC. Starting capital is either taken from the available_capital setting, or calculated by using current wallet size - profits. Profit Factor is calculated as gross profits / gross losses - and should serve as an overall metric for the strategy. Max drawdown corresponds to the backtesting metric Absolute Drawdown (Account) - calculated as (Absolute Drawdown) / (DrawdownHigh + startingBalance). Bot started date will refer to the date the bot was first started. For older bots, this will default to the first trade's open date.

"},{"location":"telegram-usage/#forceexit","title":"/forceexit

BINANCE: Exiting BTC/LTC with limit 0.01650000 (profit: ~-4.07%, -0.00008168)

Tip

You can get a list of all open trades by calling /forceexit without parameter, which will show a list of buttons to simply exit a trade. This command has an alias in /fx - which has the same capabilities, but is faster to type in \"emergency\" situations.

","text":""},{"location":"telegram-usage/#forcelong-rate-forceshort-rate","title":"/forcelong [rate] | /forceshort [rate]

/forcebuy <pair> [rate] is also supported for longs but should be considered deprecated.

BINANCE: Long ETH/BTC with limit 0.03400000 (1.000000 ETH, 225.290 USD)

Omitting the pair will open a query asking for the pair to trade (based on the current whitelist). Trades created through /forcelong will have the buy-tag of force_entry.

Note that for this to work, force_entry_enable needs to be set to true.

More details

","text":""},{"location":"telegram-usage/#performance","title":"/performance

Return the performance of each crypto-currency the bot has sold.

Performance: 1. RCN/BTC 0.003 BTC (57.77%) (1) 2. PAY/BTC 0.0012 BTC (56.91%) (1) 3. VIB/BTC 0.0011 BTC (47.07%) (1) 4. SALT/BTC 0.0010 BTC (30.24%) (1) 5. STORJ/BTC 0.0009 BTC (27.24%) (1) ...

","text":""},{"location":"telegram-usage/#balance","title":"/balance

Return the balance of all crypto-currency your have on the exchange.

Currency: BTC Available: 3.05890234 Balance: 3.05890234 Pending: 0.0

Currency: CVC Available: 86.64180098 Balance: 86.64180098 Pending: 0.0

","text":""},{"location":"telegram-usage/#daily","title":"/daily

Per default /daily will return the 7 last days. The example below if for /daily 3:

Daily Profit over the last 3 days:

Day (count)     USDT          USD         Profit %\n--------------  ------------  ----------  ----------\n2022-06-11 (1)  -0.746 USDT   -0.75 USD   -0.08%\n2022-06-10 (0)  0 USDT        0.00 USD    0.00%\n2022-06-09 (5)  20 USDT       20.10 USD   5.00%\n

","text":""},{"location":"telegram-usage/#weekly","title":"/weekly

Per default /weekly will return the 8 last weeks, including the current week. Each week starts from Monday. The example below if for /weekly 3:

Weekly Profit over the last 3 weeks (starting from Monday):

Monday (count)  Profit BTC      Profit USD   Profit %\n-------------  --------------  ------------    ----------\n2018-01-03 (5)  0.00224175 BTC  29,142 USD   4.98%\n2017-12-27 (1)  0.00033131 BTC   4,307 USD   0.00%\n2017-12-20 (4)  0.00269130 BTC  34.986 USD   5.12%\n

","text":""},{"location":"telegram-usage/#monthly","title":"/monthly

Per default /monthly will return the 6 last months, including the current month. The example below if for /monthly 3:

Monthly Profit over the last 3 months:

Month (count)  Profit BTC      Profit USD    Profit %\n-------------  --------------  ------------    ----------\n2018-01 (20)    0.00224175 BTC  29,142 USD  4.98%\n2017-12 (5)    0.00033131 BTC   4,307 USD   0.00%\n2017-11 (10)    0.00269130 BTC  34.986 USD  5.10%\n

","text":""},{"location":"telegram-usage/#whitelist","title":"/whitelist

Shows the current whitelist

Using whitelist StaticPairList with 22 pairs IOTA/BTC, NEO/BTC, TRX/BTC, VET/BTC, ADA/BTC, ETC/BTC, NCASH/BTC, DASH/BTC, XRP/BTC, XVG/BTC, EOS/BTC, LTC/BTC, OMG/BTC, BTG/BTC, LSK/BTC, ZEC/BTC, HOT/BTC, IOTX/BTC, XMR/BTC, AST/BTC, XLM/BTC, NANO/BTC

","text":""},{"location":"telegram-usage/#blacklist-pair","title":"/blacklist [pair]

Shows the current blacklist. If Pair is set, then this pair will be added to the pairlist. Also supports multiple pairs, separated by a space. Use /reload_config to reset the blacklist.

Using blacklist StaticPairList with 2 pairs DODGE/BTC, HOT/BTC.

","text":""},{"location":"telegram-usage/#edge","title":"/edge

Shows pairs validated by Edge along with their corresponding win-rate, expectancy and stoploss values.

Edge only validated following pairs:

Pair        Winrate    Expectancy    Stoploss\n--------  ---------  ------------  ----------\nDOCK/ETH   0.522727      0.881821       -0.03\nPHX/ETH    0.677419      0.560488       -0.03\nHOT/ETH    0.733333      0.490492       -0.03\nHC/ETH     0.588235      0.280988       -0.02\nARDR/ETH   0.366667      0.143059       -0.01\n

","text":""},{"location":"telegram-usage/#version","title":"/version

Version: 0.14.3

","text":""},{"location":"telegram-usage/#marketdir","title":"/marketdir

If a market direction is provided the command updates the user managed variable that represents the current market direction. This variable is not set to any valid market direction on bot startup and must be set by the user. The example below is for /marketdir long:

Successfully updated marketdirection from none to long.\n

If no market direction is provided the command outputs the currently set market directions. The example below is for /marketdir:

Currently set marketdirection: even\n

You can use the market direction in your strategy via self.market_direction.

Bot restarts

Please note that the market direction is not persisted, and will be reset after a bot restart/reload.

Backtesting

As this value/variable is intended to be changed manually in dry/live trading. Strategies using market_direction will probably not produce reliable, reproducible results (changes to this variable will not be reflected for backtesting). Use at your own risk.

","text":""},{"location":"trade-object/","title":"Trade Object","text":""},{"location":"trade-object/#trade","title":"Trade","text":"

A position freqtrade enters is stored in a Trade object - which is persisted to the database. It's a core concept of freqtrade - and something you'll come across in many sections of the documentation, which will most likely point you to this location.

It will be passed to the strategy in many strategy callbacks. The object passed to the strategy cannot be modified directly. Indirect modifications may occur based on callback results.

"},{"location":"trade-object/#trade-available-attributes","title":"Trade - Available attributes","text":"

The following attributes / properties are available for each individual trade - and can be used with trade.<property> (e.g. trade.pair).

Attribute DataType Description pair string Pair of this trade is_open boolean Is the trade currently open, or has it been concluded open_rate float Rate this trade was entered at (Avg. entry rate in case of trade-adjustments) close_rate float Close rate - only set when is_open = False stake_amount float Amount in Stake (or Quote) currency. amount float Amount in Asset / Base currency that is currently owned. open_date datetime Timestamp when trade was opened use open_date_utc instead open_date_utc datetime Timestamp when trade was opened - in UTC close_date datetime Timestamp when trade was closed use close_date_utc instead close_date_utc datetime Timestamp when trade was closed - in UTC close_profit float Relative profit at the time of trade closure. 0.01 == 1% close_profit_abs float Absolute profit (in stake currency) at the time of trade closure. leverage float Leverage used for this trade - defaults to 1.0 in spot markets. enter_tag string Tag provided on entry via the enter_tag column in the dataframe is_short boolean True for short trades, False otherwise orders Order[] List of order objects attached to this trade (includes both filled and cancelled orders) date_last_filled_utc datetime Time of the last filled order entry_side \"buy\" / \"sell\" Order Side the trade was entered exit_side \"buy\" / \"sell\" Order Side that will result in a trade exit / position reduction. trade_direction \"long\" / \"short\" Trade direction in text - long or short. nr_of_successful_entries int Number of successful (filled) entry orders nr_of_successful_exits int Number of successful (filled) exit orders"},{"location":"trade-object/#class-methods","title":"Class methods","text":"

The following are class methods - which return generic information, and usually result in an explicit query against the database. They can be used as Trade.<method> - e.g. open_trades = Trade.get_open_trade_count()

Backtesting/hyperopt

Most methods will work in both backtesting / hyperopt and live/dry modes. During backtesting, it's limited to usage in strategy callbacks. Usage in populate_*() methods is not supported and will result in wrong results.

"},{"location":"trade-object/#get_trades_proxy","title":"get_trades_proxy","text":"

When your strategy needs some information on existing (open or close) trades - it's best to use Trade.get_trades_proxy().

Usage:

from freqtrade.persistence import Trade\nfrom datetime import timedelta\n\n# ...\ntrade_hist = Trade.get_trades_proxy(pair='ETH/USDT', is_open=False, open_date=current_date - timedelta(days=2))\n

get_trades_proxy() supports the following keyword arguments. All arguments are optional - calling get_trades_proxy() without arguments will return a list of all trades in the database.

  • pair e.g. pair='ETH/USDT'
  • is_open e.g. is_open=False
  • open_date e.g. open_date=current_date - timedelta(days=2)
  • close_date e.g. close_date=current_date - timedelta(days=5)
"},{"location":"trade-object/#get_open_trade_count","title":"get_open_trade_count","text":"

Get the number of currently open trades

from freqtrade.persistence import Trade\n# ...\nopen_trades = Trade.get_open_trade_count()\n
"},{"location":"trade-object/#get_total_closed_profit","title":"get_total_closed_profit","text":"

Retrieve the total profit the bot has generated so far. Aggregates close_profit_abs for all closed trades.

from freqtrade.persistence import Trade\n\n# ...\nprofit = Trade.get_total_closed_profit()\n
"},{"location":"trade-object/#total_open_trades_stakes","title":"total_open_trades_stakes","text":"

Retrieve the total stake_amount that's currently in trades.

from freqtrade.persistence import Trade\n\n# ...\nprofit = Trade.total_open_trades_stakes()\n
"},{"location":"trade-object/#get_overall_performance","title":"get_overall_performance","text":"

Retrieve the overall performance - similar to the /performance telegram command.

from freqtrade.persistence import Trade\n\n# ...\nif self.config['runmode'].value in ('live', 'dry_run'):\n    performance = Trade.get_overall_performance()\n

Sample return value: ETH/BTC had 5 trades, with a total profit of 1.5% (ratio of 0.015).

{\"pair\": \"ETH/BTC\", \"profit\": 0.015, \"count\": 5}\n
"},{"location":"trade-object/#order-object","title":"Order Object","text":"

An Order object represents an order on the exchange (or a simulated order in dry-run mode). An Order object will always be tied to it's corresponding Trade, and only really makes sense in the context of a trade.

"},{"location":"trade-object/#order-available-attributes","title":"Order - Available attributes","text":"

an Order object is typically attached to a trade. Most properties here can be None as they are dependant on the exchange response.

Attribute DataType Description trade Trade Trade object this order is attached to ft_pair string Pair this order is for ft_is_open boolean is the order filled? order_type string Order type as defined on the exchange - usually market, limit or stoploss status string Status as defined by ccxt. Usually open, closed, expired or canceled side string Buy or Sell price float Price the order was placed at average float Average price the order filled at amount float Amount in base currency filled float Filled amount (in base currency) remaining float Remaining amount cost float Cost of the order - usually average * filled order_date datetime Order creation date use order_date_utc instead order_date_utc datetime Order creation date (in UTC) order_fill_date datetime Order fill date use order_fill_utc instead order_fill_date_utc datetime Order fill date"},{"location":"updating/","title":"How to update","text":"

To update your freqtrade installation, please use one of the below methods, corresponding to your installation method.

Tracking changes

Breaking changes / changed behavior will be documented in the changelog that is posted alongside every release. For the develop branch, please follow PR's to avoid being surprised by changes.

"},{"location":"updating/#docker","title":"docker","text":"

Legacy installations using the master image

We're switching from master to stable for the release Images - please adjust your docker-file and replace freqtradeorg/freqtrade:master with freqtradeorg/freqtrade:stable

docker compose pull\ndocker compose up -d\n
"},{"location":"updating/#installation-via-setup-script","title":"Installation via setup script","text":"
./setup.sh --update\n

Note

Make sure to run this command with your virtual environment disabled!

"},{"location":"updating/#plain-native-installation","title":"Plain native installation","text":"

Please ensure that you're also updating dependencies - otherwise things might break without you noticing.

git pull\npip install -U -r requirements.txt\npip install -e .\n\n# Ensure freqUI is at the latest version\nfreqtrade install-ui 
"},{"location":"updating/#problems-updating","title":"Problems updating","text":"

Update-problems usually come missing dependencies (you didn't follow the above instructions) - or from updated dependencies, which fail to install (for example TA-lib). Please refer to the corresponding installation sections (common problems linked below)

Common problems and their solutions:

  • ta-lib update on windows
"},{"location":"utils/","title":"Utility Subcommands","text":"

Besides the Live-Trade and Dry-Run run modes, the backtesting, edge and hyperopt optimization subcommands, and the download-data subcommand which prepares historical data, the bot contains a number of utility subcommands. They are described in this section.

"},{"location":"utils/#create-userdir","title":"Create userdir","text":"

Creates the directory structure to hold your files for freqtrade. Will also create strategy and hyperopt examples for you to get started. Can be used multiple times - using --reset will reset the sample strategy and hyperopt files to their default state.

usage: freqtrade create-userdir [-h] [--userdir PATH] [--reset]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n  --reset               Reset sample files to their original state.\n

Warning

Using --reset may result in loss of data, since this will overwrite all sample files without asking again.

\u251c\u2500\u2500 backtest_results\n\u251c\u2500\u2500 data\n\u251c\u2500\u2500 hyperopt_results\n\u251c\u2500\u2500 hyperopts\n\u2502\u00a0\u00a0 \u251c\u2500\u2500 sample_hyperopt_loss.py\n\u251c\u2500\u2500 notebooks\n\u2502\u00a0\u00a0 \u2514\u2500\u2500 strategy_analysis_example.ipynb\n\u251c\u2500\u2500 plot\n\u2514\u2500\u2500 strategies\n    \u2514\u2500\u2500 sample_strategy.py\n
"},{"location":"utils/#create-new-config","title":"Create new config","text":"

Creates a new configuration file, asking some questions which are important selections for a configuration.

usage: freqtrade new-config [-h] [-c PATH]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default: `config.json`). Multiple --config options may be used. Can be set to `-`\n                        to read config from stdin.\n

Warning

Only vital questions are asked. Freqtrade offers a lot more configuration possibilities, which are listed in the Configuration documentation

"},{"location":"utils/#create-config-examples","title":"Create config examples","text":"
$ freqtrade new-config --config config_binance.json\n\n? Do you want to enable Dry-run (simulated trades)?  Yes\n? Please insert your stake currency: BTC\n? Please insert your stake amount: 0.05\n? Please insert max_open_trades (Integer or -1 for unlimited open trades): 3\n? Please insert your desired timeframe (e.g. 5m): 5m\n? Please insert your display Currency (for reporting): USD\n? Select exchange  binance\n? Do you want to enable Telegram?  No\n
"},{"location":"utils/#create-new-strategy","title":"Create new strategy","text":"

Creates a new strategy from a template similar to SampleStrategy. The file will be named inline with your class name, and will not overwrite existing files.

Results will be located in user_data/strategies/<strategyclassname>.py.

usage: freqtrade new-strategy [-h] [--userdir PATH] [-s NAME]\n                              [--template {full,minimal,advanced}]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n  -s NAME, --strategy NAME\n                        Specify strategy class name which will be used by the\n                        bot.\n  --template {full,minimal,advanced}\n                        Use a template which is either `minimal`, `full`\n                        (containing multiple sample indicators) or `advanced`.\n                        Default: `full`.\n
"},{"location":"utils/#sample-usage-of-new-strategy","title":"Sample usage of new-strategy","text":"
freqtrade new-strategy --strategy AwesomeStrategy\n

With custom user directory

freqtrade new-strategy --userdir ~/.freqtrade/ --strategy AwesomeStrategy\n

Using the advanced template (populates all optional functions and methods)

freqtrade new-strategy --strategy AwesomeStrategy --template advanced\n
"},{"location":"utils/#list-strategies","title":"List Strategies","text":"

Use the list-strategies subcommand to see all strategies in one particular directory.

This subcommand is useful for finding problems in your environment with loading strategies: modules with strategies that contain errors and failed to load are printed in red (LOAD FAILED), while strategies with duplicate names are printed in yellow (DUPLICATE NAME).

usage: freqtrade list-strategies [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                                 [-d PATH] [--userdir PATH]\n                                 [--strategy-path PATH] [-1] [--no-color]\n                                 [--recursive-strategy-search]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --strategy-path PATH  Specify additional strategy lookup path.\n  -1, --one-column      Print output in one column.\n  --no-color            Disable colorization of hyperopt results. May be\n                        useful if you are redirecting output to a file.\n  --recursive-strategy-search\n                        Recursively search for a strategy in the strategies\n                        folder.\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n

Warning

Using these commands will try to load all python files from a directory. This can be a security risk if untrusted files reside in this directory, since all module-level code is executed.

Example: Search default strategies directories (within the default userdir).

freqtrade list-strategies\n

Example: Search strategies directory within the userdir.

freqtrade list-strategies --userdir ~/.freqtrade/\n

Example: Search dedicated strategy path.

freqtrade list-strategies --strategy-path ~/.freqtrade/strategies/\n
"},{"location":"utils/#list-freqai-models","title":"List freqAI models","text":"

Use the list-freqaimodels subcommand to see all freqAI models available.

This subcommand is useful for finding problems in your environment with loading freqAI models: modules with models that contain errors and failed to load are printed in red (LOAD FAILED), while models with duplicate names are printed in yellow (DUPLICATE NAME).

usage: freqtrade list-freqaimodels [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                                   [-d PATH] [--userdir PATH]\n                                   [--freqaimodel-path PATH] [-1] [--no-color]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --freqaimodel-path PATH\n                        Specify additional lookup path for freqaimodels.\n  -1, --one-column      Print output in one column.\n  --no-color            Disable colorization of hyperopt results. May be\n                        useful if you are redirecting output to a file.\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH, --data-dir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n
"},{"location":"utils/#list-exchanges","title":"List Exchanges","text":"

Use the list-exchanges subcommand to see the exchanges available for the bot.

usage: freqtrade list-exchanges [-h] [-1] [-a]\n\noptional arguments:\n  -h, --help        show this help message and exit\n  -1, --one-column  Print output in one column.\n  -a, --all         Print all exchanges known to the ccxt library.\n
  • Example: see exchanges available for the bot:
    $ freqtrade list-exchanges\nExchanges available for Freqtrade:\nExchange name    Valid    reason\n---------------  -------  --------------------------------------------\naax              True\nascendex         True     missing opt: fetchMyTrades\nbequant          True\nbibox            True\nbigone           True\nbinance          True\nbinanceus        True\nbitbank          True     missing opt: fetchTickers\nbitcoincom       True\nbitfinex         True\nbitforex         True     missing opt: fetchMyTrades, fetchTickers\nbitget           True\nbithumb          True     missing opt: fetchMyTrades\nbitkk            True     missing opt: fetchMyTrades\nbitmart          True\nbitmax           True     missing opt: fetchMyTrades\nbitpanda         True\nbittrex          True\nbitvavo          True\nbitz             True     missing opt: fetchMyTrades\nbtcalpha         True     missing opt: fetchTicker, fetchTickers\nbtcmarkets       True     missing opt: fetchTickers\nbuda             True     missing opt: fetchMyTrades, fetchTickers\nbw               True     missing opt: fetchMyTrades, fetchL2OrderBook\nbybit            True\nbytetrade        True\ncdax             True\ncex              True     missing opt: fetchMyTrades\ncoinbaseprime    True     missing opt: fetchTickers\ncoinbasepro      True     missing opt: fetchTickers\ncoinex           True\ncrex24           True\nderibit          True\ndigifinex        True\nequos            True     missing opt: fetchTicker, fetchTickers\neterbase         True\nfcoin            True     missing opt: fetchMyTrades, fetchTickers\nfcoinjp          True     missing opt: fetchMyTrades, fetchTickers\ngateio           True\ngemini           True\ngopax            True\nhbtc             True\nhitbtc           True\nhuobijp          True\nhuobipro         True\nidex             True\nkraken           True\nkucoin           True\nlbank            True     missing opt: fetchMyTrades\nmercado          True     missing opt: fetchTickers\nndax             True     missing opt: fetchTickers\nnovadax          True\nokcoin           True\nokex             True\nprobit           True\nqtrade           True\nstex             True\ntimex            True\nupbit            True     missing opt: fetchMyTrades\nvcc              True\nzb               True     missing opt: fetchMyTrades\n

missing opt exchanges

Values with \"missing opt:\" might need special configuration (e.g. using orderbook if fetchTickers is missing) - but should in theory work (although we cannot guarantee they will).

  • Example: see all exchanges supported by the ccxt library (including 'bad' ones, i.e. those that are known to not work with Freqtrade):
    $ freqtrade list-exchanges -a\nAll exchanges supported by the ccxt library:\nExchange name       Valid    reason\n------------------  -------  ---------------------------------------------------------------------------------------\naax                 True\naofex               False    missing: fetchOrder\nascendex            True     missing opt: fetchMyTrades\nbequant             True\nbibox               True\nbigone              True\nbinance             True\nbinanceus           True\nbit2c               False    missing: fetchOrder, fetchOHLCV\nbitbank             True     missing opt: fetchTickers\nbitbay              False    missing: fetchOrder\nbitcoincom          True\nbitfinex            True\nbitfinex2           False    missing: fetchOrder\nbitflyer            False    missing: fetchOrder, fetchOHLCV\nbitforex            True     missing opt: fetchMyTrades, fetchTickers\nbitget              True\nbithumb             True     missing opt: fetchMyTrades\nbitkk               True     missing opt: fetchMyTrades\nbitmart             True\nbitmax              True     missing opt: fetchMyTrades\nbitmex              False    Various reasons.\nbitpanda            True\nbitso               False    missing: fetchOHLCV\nbitstamp            True     missing opt: fetchTickers\nbitstamp1           False    missing: fetchOrder, fetchOHLCV\nbittrex             True\nbitvavo             True\nbitz                True     missing opt: fetchMyTrades\nbl3p                False    missing: fetchOrder, fetchOHLCV\nbleutrade           False    missing: fetchOrder\nbraziliex           False    missing: fetchOHLCV\nbtcalpha            True     missing opt: fetchTicker, fetchTickers\nbtcbox              False    missing: fetchOHLCV\nbtcmarkets          True     missing opt: fetchTickers\nbtctradeua          False    missing: fetchOrder, fetchOHLCV\nbtcturk             False    missing: fetchOrder\nbuda                True     missing opt: fetchMyTrades, fetchTickers\nbw                  True     missing opt: fetchMyTrades, fetchL2OrderBook\nbybit               True\nbytetrade           True\ncdax                True\ncex                 True     missing opt: fetchMyTrades\nchilebit            False    missing: fetchOrder, fetchOHLCV\ncoinbase            False    missing: fetchOrder, cancelOrder, createOrder, fetchOHLCV\ncoinbaseprime       True     missing opt: fetchTickers\ncoinbasepro         True     missing opt: fetchTickers\ncoincheck           False    missing: fetchOrder, fetchOHLCV\ncoinegg             False    missing: fetchOHLCV\ncoinex              True\ncoinfalcon          False    missing: fetchOHLCV\ncoinfloor           False    missing: fetchOrder, fetchOHLCV\ncoingi              False    missing: fetchOrder, fetchOHLCV\ncoinmarketcap       False    missing: fetchOrder, cancelOrder, createOrder, fetchBalance, fetchOHLCV\ncoinmate            False    missing: fetchOHLCV\ncoinone             False    missing: fetchOHLCV\ncoinspot            False    missing: fetchOrder, cancelOrder, fetchOHLCV\ncrex24              True\ncurrencycom         False    missing: fetchOrder\ndelta               False    missing: fetchOrder\nderibit             True\ndigifinex           True\nequos               True     missing opt: fetchTicker, fetchTickers\neterbase            True\nexmo                False    missing: fetchOrder\nexx                 False    missing: fetchOHLCV\nfcoin               True     missing opt: fetchMyTrades, fetchTickers\nfcoinjp             True     missing opt: fetchMyTrades, fetchTickers\nflowbtc             False    missing: fetchOrder, fetchOHLCV\nfoxbit              False    missing: fetchOrder, fetchOHLCV\ngateio              True\ngemini              True\ngopax               True\nhbtc                True\nhitbtc              True\nhollaex             False    missing: fetchOrder\nhuobijp             True\nhuobipro            True\nidex                True\nindependentreserve  False    missing: fetchOHLCV\nindodax             False    missing: fetchOHLCV\nitbit               False    missing: fetchOHLCV\nkraken              True\nkucoin              True\nkuna                False    missing: fetchOHLCV\nlakebtc             False    missing: fetchOrder, fetchOHLCV\nlatoken             False    missing: fetchOrder, fetchOHLCV\nlbank               True     missing opt: fetchMyTrades\nliquid              False    missing: fetchOHLCV\nluno                False    missing: fetchOHLCV\nlykke               False    missing: fetchOHLCV\nmercado             True     missing opt: fetchTickers\nmixcoins            False    missing: fetchOrder, fetchOHLCV\nndax                True     missing opt: fetchTickers\nnovadax             True\noceanex             False    missing: fetchOHLCV\nokcoin              True\nokex                True\npaymium             False    missing: fetchOrder, fetchOHLCV\nphemex              False    Does not provide history.\npoloniex            False    missing: fetchOrder\nprobit              True\nqtrade              True\nrightbtc            False    missing: fetchOrder\nripio               False    missing: fetchOHLCV\nsouthxchange        False    missing: fetchOrder, fetchOHLCV\nstex                True\nsurbitcoin          False    missing: fetchOrder, fetchOHLCV\ntherock             False    missing: fetchOHLCV\ntidebit             False    missing: fetchOrder\ntidex               False    missing: fetchOHLCV\ntimex               True\nupbit               True     missing opt: fetchMyTrades\nvbtc                False    missing: fetchOrder, fetchOHLCV\nvcc                 True\nwavesexchange       False    missing: fetchOrder\nwhitebit            False    missing: fetchOrder, cancelOrder, createOrder, fetchBalance\nxbtce               False    missing: fetchOrder, fetchOHLCV\nxena                False    missing: fetchOrder\nyobit               False    missing: fetchOHLCV\nzaif                False    missing: fetchOrder, fetchOHLCV\nzb                  True     missing opt: fetchMyTrades\n
"},{"location":"utils/#list-timeframes","title":"List Timeframes","text":"

Use the list-timeframes subcommand to see the list of timeframes available for the exchange.

usage: freqtrade list-timeframes [-h] [-v] [--logfile FILE] [-V] [-c PATH] [-d PATH] [--userdir PATH] [--exchange EXCHANGE] [-1]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no config is provided.\n  -1, --one-column      Print output in one column.\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are: 'syslog', 'journald'. See the documentation for more details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default: `config.json`). Multiple --config options may be used. Can be set to `-`\n                        to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n
  • Example: see the timeframes for the 'binance' exchange, set in the configuration file:
$ freqtrade list-timeframes -c config_binance.json\n...\nTimeframes available for the exchange `binance`: 1m, 3m, 5m, 15m, 30m, 1h, 2h, 4h, 6h, 8h, 12h, 1d, 3d, 1w, 1M\n
  • Example: enumerate exchanges available for Freqtrade and print timeframes supported by each of them:
    $ for i in `freqtrade list-exchanges -1`; do freqtrade list-timeframes --exchange $i; done\n
"},{"location":"utils/#list-pairslist-markets","title":"List pairs/list markets","text":"

The list-pairs and list-markets subcommands allow to see the pairs/markets available on exchange.

Pairs are markets with the '/' character between the base currency part and the quote currency part in the market symbol. For example, in the 'ETH/BTC' pair 'ETH' is the base currency, while 'BTC' is the quote currency.

For pairs traded by Freqtrade the pair quote currency is defined by the value of the stake_currency configuration setting.

You can print info about any pair/market with these subcommands - and you can filter output by quote-currency using --quote BTC, or by base-currency using --base ETH options correspondingly.

These subcommands have same usage and same set of available options:

usage: freqtrade list-markets [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                              [-d PATH] [--userdir PATH] [--exchange EXCHANGE]\n                              [--print-list] [--print-json] [-1] [--print-csv]\n                              [--base BASE_CURRENCY [BASE_CURRENCY ...]]\n                              [--quote QUOTE_CURRENCY [QUOTE_CURRENCY ...]] [-a]\n                              [--trading-mode {spot,margin,futures}]\n\nusage: freqtrade list-pairs [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                            [-d PATH] [--userdir PATH] [--exchange EXCHANGE]\n                            [--print-list] [--print-json] [-1] [--print-csv]\n                            [--base BASE_CURRENCY [BASE_CURRENCY ...]]\n                            [--quote QUOTE_CURRENCY [QUOTE_CURRENCY ...]] [-a]\n                            [--trading-mode {spot,margin,futures}]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no\n                        config is provided.\n  --print-list          Print list of pairs or market symbols. By default data\n                        is printed in the tabular format.\n  --print-json          Print list of pairs or market symbols in JSON format.\n  -1, --one-column      Print output in one column.\n  --print-csv           Print exchange pair or market data in the csv format.\n  --base BASE_CURRENCY [BASE_CURRENCY ...]\n                        Specify base currency(-ies). Space-separated list.\n  --quote QUOTE_CURRENCY [QUOTE_CURRENCY ...]\n                        Specify quote currency(-ies). Space-separated list.\n  -a, --all             Print all pairs or market symbols. By default only\n                        active ones are shown.\n  --trading-mode {spot,margin,futures}\n                        Select Trading mode\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default: `config.json`).\n                        Multiple --config options may be used. Can be set to\n                        `-` to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n

By default, only active pairs/markets are shown. Active pairs/markets are those that can currently be traded on the exchange. The see the list of all pairs/markets (not only the active ones), use the -a/-all option.

Pairs/markets are sorted by its symbol string in the printed output.

"},{"location":"utils/#examples","title":"Examples","text":"
  • Print the list of active pairs with quote currency USD on exchange, specified in the default configuration file (i.e. pairs on the \"Bittrex\" exchange) in JSON format:
$ freqtrade list-pairs --quote USD --print-json\n
  • Print the list of all pairs on the exchange, specified in the config_binance.json configuration file (i.e. on the \"Binance\" exchange) with base currencies BTC or ETH and quote currencies USDT or USD, as the human-readable list with summary:
$ freqtrade list-pairs -c config_binance.json --all --base BTC ETH --quote USDT USD --print-list\n
  • Print all markets on exchange \"Kraken\", in the tabular format:
$ freqtrade list-markets --exchange kraken --all\n
"},{"location":"utils/#test-pairlist","title":"Test pairlist","text":"

Use the test-pairlist subcommand to test the configuration of dynamic pairlists.

Requires a configuration with specified pairlists attribute. Can be used to generate static pairlists to be used during backtesting / hyperopt.

usage: freqtrade test-pairlist [-h] [--userdir PATH] [-v] [-c PATH]\n                               [--quote QUOTE_CURRENCY [QUOTE_CURRENCY ...]]\n                               [-1] [--print-json] [--exchange EXCHANGE]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  --quote QUOTE_CURRENCY [QUOTE_CURRENCY ...]\n                        Specify quote currency(-ies). Space-separated list.\n  -1, --one-column      Print output in one column.\n  --print-json          Print list of pairs or market symbols in JSON format.\n  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no\n                        config is provided.\n
"},{"location":"utils/#examples_1","title":"Examples","text":"

Show whitelist when using a dynamic pairlist.

freqtrade test-pairlist --config config.json --quote USDT BTC\n
"},{"location":"utils/#convert-database","title":"Convert database","text":"

freqtrade convert-db can be used to convert your database from one system to another (sqlite -> postgres, postgres -> other postgres), migrating all trades, orders and Pairlocks.

Please refer to the SQL cheatsheet to learn about requirements for different database systems.

usage: freqtrade convert-db [-h] [--db-url PATH] [--db-url-from PATH]\n\noptional arguments:\n  -h, --help          show this help message and exit\n  --db-url PATH       Override trades database URL, this is useful in custom\n                      deployments (default: `sqlite:///tradesv3.sqlite` for\n                      Live Run mode, `sqlite:///tradesv3.dryrun.sqlite` for\n                      Dry Run).\n  --db-url-from PATH  Source db url to use when migrating a database.\n

Warning

Please ensure to only use this on an empty target database. Freqtrade will perform a regular migration, but may fail if entries already existed.

"},{"location":"utils/#webserver-mode","title":"Webserver mode","text":"

Experimental

Webserver mode is an experimental mode to increase backesting and strategy development productivity. There may still be bugs - so if you happen to stumble across these, please report them as github issues, thanks.

Run freqtrade in webserver mode. Freqtrade will start the webserver and allow FreqUI to start and control backtesting processes. This has the advantage that data will not be reloaded between backtesting runs (as long as timeframe and timerange remain identical). FreqUI will also show the backtesting results.

usage: freqtrade webserver [-h] [-v] [--logfile FILE] [-V] [-c PATH] [-d PATH]\n                           [--userdir PATH]\n\noptional arguments:\n  -h, --help            show this help message and exit\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n
"},{"location":"utils/#webserver-mode-docker","title":"Webserver mode - docker","text":"

You can also use webserver mode via docker. Starting a one-off container requires the configuration of the port explicitly, as ports are not exposed by default. You can use docker compose run --rm -p 127.0.0.1:8080:8080 freqtrade webserver to start a one-off container that'll be removed once you stop it. This assumes that port 8080 is still available and no other bot is running on that port.

Alternatively, you can reconfigure the docker-compose file to have the command updated:

    command: >\n      webserver\n      --config /freqtrade/user_data/config.json\n

You can now use docker compose up to start the webserver. This assumes that the configuration has a webserver enabled and configured for docker (listening port = 0.0.0.0).

Tip

Don't forget to reset the command back to the trade command if you want to start a live or dry-run bot.

"},{"location":"utils/#show-previous-backtest-results","title":"Show previous Backtest results","text":"

Allows you to show previous backtest results. Adding --show-pair-list outputs a sorted pair list you can easily copy/paste into your configuration (omitting bad pairs).

Strategy overfitting

Only using winning pairs can lead to an overfitted strategy, which will not work well on future data. Make sure to extensively test your strategy in dry-run before risking real money.

usage: freqtrade backtesting-show [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                                  [-d PATH] [--userdir PATH]\n                                  [--export-filename PATH] [--show-pair-list]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --export-filename PATH\n                        Save backtest results to the file with this filename.\n                        Requires `--export` to be set as well. Example:\n                        `--export-filename=user_data/backtest_results/backtest\n                        _today.json`\n  --show-pair-list      Show backtesting pairlist sorted by profit.\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n
"},{"location":"utils/#detailed-backtest-analysis","title":"Detailed backtest analysis","text":"

Advanced backtest result analysis.

More details in the Backtesting analysis Section.

usage: freqtrade backtesting-analysis [-h] [-v] [--logfile FILE] [-V]\n                                      [-c PATH] [-d PATH] [--userdir PATH]\n                                      [--export-filename PATH]\n                                      [--analysis-groups {0,1,2,3,4} [{0,1,2,3,4} ...]]\n                                      [--enter-reason-list ENTER_REASON_LIST [ENTER_REASON_LIST ...]]\n                                      [--exit-reason-list EXIT_REASON_LIST [EXIT_REASON_LIST ...]]\n                                      [--indicator-list INDICATOR_LIST [INDICATOR_LIST ...]]\n                                      [--timerange YYYYMMDD-[YYYYMMDD]]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --export-filename PATH, --backtest-filename PATH\n                        Use this filename for backtest results.Requires\n                        `--export` to be set as well. Example: `--export-filen\n                        ame=user_data/backtest_results/backtest_today.json`\n  --analysis-groups {0,1,2,3,4} [{0,1,2,3,4} ...]\n                        grouping output - 0: simple wins/losses by enter tag,\n                        1: by enter_tag, 2: by enter_tag and exit_tag, 3: by\n                        pair and enter_tag, 4: by pair, enter_ and exit_tag\n                        (this can get quite large)\n  --enter-reason-list ENTER_REASON_LIST [ENTER_REASON_LIST ...]\n                        Comma separated list of entry signals to analyse.\n                        Default: all. e.g. 'entry_tag_a,entry_tag_b'\n  --exit-reason-list EXIT_REASON_LIST [EXIT_REASON_LIST ...]\n                        Comma separated list of exit signals to analyse.\n                        Default: all. e.g.\n                        'exit_tag_a,roi,stop_loss,trailing_stop_loss'\n  --indicator-list INDICATOR_LIST [INDICATOR_LIST ...]\n                        Comma separated list of indicators to analyse. e.g.\n                        'close,rsi,bb_lowerband,profit_abs'\n  --timerange YYYYMMDD-[YYYYMMDD]\n                        Timerange to filter trades for analysis, \n                        start inclusive, end exclusive. e.g.\n                        20220101-20220201\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n
"},{"location":"utils/#list-hyperopt-results","title":"List Hyperopt results","text":"

You can list the hyperoptimization epochs the Hyperopt module evaluated previously with the hyperopt-list sub-command.

usage: freqtrade hyperopt-list [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                               [-d PATH] [--userdir PATH] [--best]\n                               [--profitable] [--min-trades INT]\n                               [--max-trades INT] [--min-avg-time FLOAT]\n                               [--max-avg-time FLOAT] [--min-avg-profit FLOAT]\n                               [--max-avg-profit FLOAT]\n                               [--min-total-profit FLOAT]\n                               [--max-total-profit FLOAT]\n                               [--min-objective FLOAT] [--max-objective FLOAT]\n                               [--no-color] [--print-json] [--no-details]\n                               [--hyperopt-filename PATH] [--export-csv FILE]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --best                Select only best epochs.\n  --profitable          Select only profitable epochs.\n  --min-trades INT      Select epochs with more than INT trades.\n  --max-trades INT      Select epochs with less than INT trades.\n  --min-avg-time FLOAT  Select epochs above average time.\n  --max-avg-time FLOAT  Select epochs below average time.\n  --min-avg-profit FLOAT\n                        Select epochs above average profit.\n  --max-avg-profit FLOAT\n                        Select epochs below average profit.\n  --min-total-profit FLOAT\n                        Select epochs above total profit.\n  --max-total-profit FLOAT\n                        Select epochs below total profit.\n  --min-objective FLOAT\n                        Select epochs above objective.\n  --max-objective FLOAT\n                        Select epochs below objective.\n  --no-color            Disable colorization of hyperopt results. May be\n                        useful if you are redirecting output to a file.\n  --print-json          Print output in JSON format.\n  --no-details          Do not print best epoch details.\n  --hyperopt-filename FILENAME\n                        Hyperopt result filename.Example: `--hyperopt-\n                        filename=hyperopt_results_2020-09-27_16-20-48.pickle`\n  --export-csv FILE     Export to CSV-File. This will disable table print.\n                        Example: --export-csv hyperopt.csv\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n

Note

hyperopt-list will automatically use the latest available hyperopt results file. You can override this using the --hyperopt-filename argument, and specify another, available filename (without path!).

"},{"location":"utils/#examples_2","title":"Examples","text":"

List all results, print details of the best result at the end:

freqtrade hyperopt-list\n

List only epochs with positive profit. Do not print the details of the best epoch, so that the list can be iterated in a script:

freqtrade hyperopt-list --profitable --no-details\n

"},{"location":"utils/#show-details-of-hyperopt-results","title":"Show details of Hyperopt results","text":"

You can show the details of any hyperoptimization epoch previously evaluated by the Hyperopt module with the hyperopt-show subcommand.

usage: freqtrade hyperopt-show [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                               [-d PATH] [--userdir PATH] [--best]\n                               [--profitable] [-n INT] [--print-json]\n                               [--hyperopt-filename FILENAME] [--no-header]\n                               [--disable-param-export]\n                               [--breakdown {day,week,month} [{day,week,month} ...]]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --best                Select only best epochs.\n  --profitable          Select only profitable epochs.\n  -n INT, --index INT   Specify the index of the epoch to print details for.\n  --print-json          Print output in JSON format.\n  --hyperopt-filename FILENAME\n                        Hyperopt result filename.Example: `--hyperopt-\n                        filename=hyperopt_results_2020-09-27_16-20-48.pickle`\n  --no-header           Do not print epoch details header.\n  --disable-param-export\n                        Disable automatic hyperopt parameter export.\n  --breakdown {day,week,month} [{day,week,month} ...]\n                        Show backtesting breakdown per [day, week, month].\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n

Note

hyperopt-show will automatically use the latest available hyperopt results file. You can override this using the --hyperopt-filename argument, and specify another, available filename (without path!).

"},{"location":"utils/#examples_3","title":"Examples","text":"

Print details for the epoch 168 (the number of the epoch is shown by the hyperopt-list subcommand or by Hyperopt itself during hyperoptimization run):

freqtrade hyperopt-show -n 168\n

Prints JSON data with details for the last best epoch (i.e., the best of all epochs):

freqtrade hyperopt-show --best -n -1 --print-json --no-header\n
"},{"location":"utils/#show-trades","title":"Show trades","text":"

Print selected (or all) trades from database to screen.

usage: freqtrade show-trades [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                             [-d PATH] [--userdir PATH] [--db-url PATH]\n                             [--trade-ids TRADE_IDS [TRADE_IDS ...]]\n                             [--print-json]\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --db-url PATH         Override trades database URL, this is useful in custom\n                        deployments (default: `sqlite:///tradesv3.sqlite` for\n                        Live Run mode, `sqlite:///tradesv3.dryrun.sqlite` for\n                        Dry Run).\n  --trade-ids TRADE_IDS [TRADE_IDS ...]\n                        Specify the list of trade ids.\n  --print-json          Print output in JSON format.\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n
"},{"location":"utils/#examples_4","title":"Examples","text":"

Print trades with id 2 and 3 as json

freqtrade show-trades --db-url sqlite:///tradesv3.sqlite --trade-ids 2 3 --print-json\n
"},{"location":"utils/#strategy-updater","title":"Strategy-Updater","text":"

Updates listed strategies or all strategies within the strategies folder to be v3 compliant. If the command runs without --strategy-list then all strategies inside the strategies folder will be converted. Your original strategy will remain available in the user_data/strategies_orig_updater/ directory.

Conversion results

Strategy updater will work on a \"best effort\" approach. Please do your due diligence and verify the results of the conversion. We also recommend to run a python formatter (e.g. black) to format results in a sane manner.

usage: freqtrade strategy-updater [-h] [-v] [--logfile FILE] [-V] [-c PATH]\n                                  [-d PATH] [--userdir PATH]\n                                  [--strategy-list STRATEGY_LIST [STRATEGY_LIST ...]]\n\noptions:\n  -h, --help            show this help message and exit\n  --strategy-list STRATEGY_LIST [STRATEGY_LIST ...]\n                        Provide a space-separated list of strategies to\n                        backtest. Please note that timeframe needs to be set\n                        either in config or via command line. When using this\n                        together with `--export trades`, the strategy-name is\n                        injected into the filename (so `backtest-data.json`\n                        becomes `backtest-data-SampleStrategy.json`\n\nCommon arguments:\n  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).\n  --logfile FILE, --log-file FILE\n                        Log to the file specified. Special values are:\n                        'syslog', 'journald'. See the documentation for more\n                        details.\n  -V, --version         show program's version number and exit\n  -c PATH, --config PATH\n                        Specify configuration file (default:\n                        `userdir/config.json` or `config.json` whichever\n                        exists). Multiple --config options may be used. Can be\n                        set to `-` to read config from stdin.\n  -d PATH, --datadir PATH, --data-dir PATH\n                        Path to directory with historical backtesting data.\n  --userdir PATH, --user-data-dir PATH\n                        Path to userdata directory.\n
"},{"location":"webhook-config/","title":"Webhook usage","text":""},{"location":"webhook-config/#configuration","title":"Configuration","text":"

Enable webhooks by adding a webhook-section to your configuration file, and setting webhook.enabled to true.

Sample configuration (tested using IFTTT).

  \"webhook\": {\n\"enabled\": true,\n\"url\": \"https://maker.ifttt.com/trigger/<YOUREVENT>/with/key/<YOURKEY>/\",\n\"entry\": {\n\"value1\": \"Buying {pair}\",\n\"value2\": \"limit {limit:8f}\",\n\"value3\": \"{stake_amount:8f} {stake_currency}\"\n},\n\"entry_cancel\": {\n\"value1\": \"Cancelling Open Buy Order for {pair}\",\n\"value2\": \"limit {limit:8f}\",\n\"value3\": \"{stake_amount:8f} {stake_currency}\"\n},\n\"entry_fill\": {\n\"value1\": \"Buy Order for {pair} filled\",\n\"value2\": \"at {open_rate:8f}\",\n\"value3\": \"\"\n},\n\"exit\": {\n\"value1\": \"Exiting {pair}\",\n\"value2\": \"limit {limit:8f}\",\n\"value3\": \"profit: {profit_amount:8f} {stake_currency} ({profit_ratio})\"\n},\n\"exit_cancel\": {\n\"value1\": \"Cancelling Open Exit Order for {pair}\",\n\"value2\": \"limit {limit:8f}\",\n\"value3\": \"profit: {profit_amount:8f} {stake_currency} ({profit_ratio})\"\n},\n\"exit_fill\": {\n\"value1\": \"Exit Order for {pair} filled\",\n\"value2\": \"at {close_rate:8f}.\",\n\"value3\": \"\"\n},\n\"status\": {\n\"value1\": \"Status: {status}\",\n\"value2\": \"\",\n\"value3\": \"\"\n}\n},\n

The url in webhook.url should point to the correct url for your webhook. If you're using IFTTT (as shown in the sample above) please insert your event and key to the url.

You can set the POST body format to Form-Encoded (default), JSON-Encoded, or raw data. Use \"format\": \"form\", \"format\": \"json\", or \"format\": \"raw\" respectively. Example configuration for Mattermost Cloud integration:

  \"webhook\": {\n\"enabled\": true,\n\"url\": \"https://<YOURSUBDOMAIN>.cloud.mattermost.com/hooks/<YOURHOOK>\",\n\"format\": \"json\",\n\"status\": {\n\"text\": \"Status: {status}\"\n}\n},\n

The result would be a POST request with e.g. {\"text\":\"Status: running\"} body and Content-Type: application/json header which results Status: running message in the Mattermost channel.

When using the Form-Encoded or JSON-Encoded configuration you can configure any number of payload values, and both the key and value will be ouput in the POST request. However, when using the raw data format you can only configure one value and it must be named \"data\". In this instance the data key will not be output in the POST request, only the value. For example:

  \"webhook\": {\n\"enabled\": true,\n\"url\": \"https://<YOURHOOKURL>\",\n\"format\": \"raw\",\n\"webhookstatus\": {\n\"data\": \"Status: {status}\"\n}\n},\n

The result would be a POST request with e.g. Status: running body and Content-Type: text/plain header.

Optional parameters are available to enable automatic retries for webhook messages. The webhook.retries parameter can be set for the maximum number of retries the webhook request should attempt if it is unsuccessful (i.e. HTTP response status is not 200). By default this is set to 0 which is disabled. An additional webhook.retry_delay parameter can be set to specify the time in seconds between retry attempts. By default this is set to 0.1 (i.e. 100ms). Note that increasing the number of retries or retry delay may slow down the trader if there are connectivity issues with the webhook. Example configuration for retries:

  \"webhook\": {\n\"enabled\": true,\n\"url\": \"https://<YOURHOOKURL>\",\n\"retries\": 3,\n\"retry_delay\": 0.2,\n\"status\": {\n\"status\": \"Status: {status}\"\n}\n},\n

Custom messages can be sent to Webhook endpoints via the self.dp.send_msg() function from within the strategy. To enable this, set the allow_custom_messages option to true:

  \"webhook\": {\n\"enabled\": true,\n\"url\": \"https://<YOURHOOKURL>\",\n\"allow_custom_messages\": true,\n\"strategy_msg\": {\n\"status\": \"StrategyMessage: {msg}\"\n}\n},\n

Different payloads can be configured for different events. Not all fields are necessary, but you should configure at least one of the dicts, otherwise the webhook will never be called.

"},{"location":"webhook-config/#entry","title":"Entry","text":"

The fields in webhook.entry are filled when the bot executes a long/short. Parameters are filled using string.format. Possible parameters are:

  • trade_id
  • exchange
  • pair
  • direction
  • leverage
  • limit # Deprecated - should no longer be used.
  • open_rate
  • amount
  • open_date
  • stake_amount
  • stake_currency
  • base_currency
  • fiat_currency
  • order_type
  • current_rate
  • enter_tag
"},{"location":"webhook-config/#entry-cancel","title":"Entry cancel","text":"

The fields in webhook.entry_cancel are filled when the bot cancels a long/short order. Parameters are filled using string.format. Possible parameters are:

  • trade_id
  • exchange
  • pair
  • direction
  • leverage
  • limit
  • amount
  • open_date
  • stake_amount
  • stake_currency
  • base_currency
  • fiat_currency
  • order_type
  • current_rate
  • enter_tag
"},{"location":"webhook-config/#entry-fill","title":"Entry fill","text":"

The fields in webhook.entry_fill are filled when the bot filled a long/short order. Parameters are filled using string.format. Possible parameters are:

  • trade_id
  • exchange
  • pair
  • direction
  • leverage
  • open_rate
  • amount
  • open_date
  • stake_amount
  • stake_currency
  • base_currency
  • fiat_currency
  • order_type
  • current_rate
  • enter_tag
"},{"location":"webhook-config/#exit","title":"Exit","text":"

The fields in webhook.exit are filled when the bot exits a trade. Parameters are filled using string.format. Possible parameters are:

  • trade_id
  • exchange
  • pair
  • direction
  • leverage
  • gain
  • limit
  • amount
  • open_rate
  • profit_amount
  • profit_ratio
  • stake_currency
  • base_currency
  • fiat_currency
  • exit_reason
  • order_type
  • open_date
  • close_date
"},{"location":"webhook-config/#exit-fill","title":"Exit fill","text":"

The fields in webhook.exit_fill are filled when the bot fills a exit order (closes a Trade). Parameters are filled using string.format. Possible parameters are:

  • trade_id
  • exchange
  • pair
  • direction
  • leverage
  • gain
  • close_rate
  • amount
  • open_rate
  • current_rate
  • profit_amount
  • profit_ratio
  • stake_currency
  • base_currency
  • fiat_currency
  • exit_reason
  • order_type
  • open_date
  • close_date
"},{"location":"webhook-config/#exit-cancel","title":"Exit cancel","text":"

The fields in webhook.exit_cancel are filled when the bot cancels a exit order. Parameters are filled using string.format. Possible parameters are:

  • trade_id
  • exchange
  • pair
  • direction
  • leverage
  • gain
  • limit
  • amount
  • open_rate
  • current_rate
  • profit_amount
  • profit_ratio
  • stake_currency
  • base_currency
  • fiat_currency
  • exit_reason
  • order_type
  • open_date
  • close_date
"},{"location":"webhook-config/#status","title":"Status","text":"

The fields in webhook.status are used for regular status messages (Started / Stopped / ...). Parameters are filled using string.format.

The only possible value here is {status}.

"},{"location":"webhook-config/#discord","title":"Discord","text":"

A special form of webhooks is available for discord. You can configure this as follows:

\"discord\": {\n\"enabled\": true,\n\"webhook_url\": \"https://discord.com/api/webhooks/<Your webhook URL ...>\",\n\"exit_fill\": [\n{\"Trade ID\": \"{trade_id}\"},\n{\"Exchange\": \"{exchange}\"},\n{\"Pair\": \"{pair}\"},\n{\"Direction\": \"{direction}\"},\n{\"Open rate\": \"{open_rate}\"},\n{\"Close rate\": \"{close_rate}\"},\n{\"Amount\": \"{amount}\"},\n{\"Open date\": \"{open_date:%Y-%m-%d %H:%M:%S}\"},\n{\"Close date\": \"{close_date:%Y-%m-%d %H:%M:%S}\"},\n{\"Profit\": \"{profit_amount} {stake_currency}\"},\n{\"Profitability\": \"{profit_ratio:.2%}\"},\n{\"Enter tag\": \"{enter_tag}\"},\n{\"Exit Reason\": \"{exit_reason}\"},\n{\"Strategy\": \"{strategy}\"},\n{\"Timeframe\": \"{timeframe}\"},\n],\n\"entry_fill\": [\n{\"Trade ID\": \"{trade_id}\"},\n{\"Exchange\": \"{exchange}\"},\n{\"Pair\": \"{pair}\"},\n{\"Direction\": \"{direction}\"},\n{\"Open rate\": \"{open_rate}\"},\n{\"Amount\": \"{amount}\"},\n{\"Open date\": \"{open_date:%Y-%m-%d %H:%M:%S}\"},\n{\"Enter tag\": \"{enter_tag}\"},\n{\"Strategy\": \"{strategy} {timeframe}\"},\n]\n}\n

The above represents the default (exit_fill and entry_fill are optional and will default to the above configuration) - modifications are obviously possible.

Available fields correspond to the fields for webhooks and are documented in the corresponding webhook sections.

The notifications will look as follows by default.

Custom messages can be sent from a strategy to Discord endpoints via the dataprovider.send_msg() function. To enable this, set the allow_custom_messages option to true:

  \"discord\": {\n\"enabled\": true,\n\"webhook_url\": \"https://discord.com/api/webhooks/<Your webhook URL ...>\",\n\"allow_custom_messages\": true,\n},\n
"},{"location":"windows_installation/","title":"Windows installation","text":"

We strongly recommend that Windows users use Docker as this will work much easier and smoother (also more secure).

If that is not possible, try using the Windows Linux subsystem (WSL) - for which the Ubuntu instructions should work. Otherwise, please follow the instructions below.

"},{"location":"windows_installation/#install-freqtrade-manually","title":"Install freqtrade manually","text":"

64bit Python version

Please make sure to use 64bit Windows and 64bit Python to avoid problems with backtesting or hyperopt due to the memory constraints 32bit applications have under Windows. 32bit python versions are no longer supported under Windows.

Hint

Using the Anaconda Distribution under Windows can greatly help with installation problems. Check out the Anaconda installation section in the documentation for more information.

"},{"location":"windows_installation/#1-clone-the-git-repository","title":"1. Clone the git repository","text":"
git clone https://github.com/freqtrade/freqtrade.git\n
"},{"location":"windows_installation/#2-install-ta-lib","title":"2. Install ta-lib","text":"

Install ta-lib according to the ta-lib documentation.

As compiling from source on windows has heavy dependencies (requires a partial visual studio installation), Freqtrade provides these dependencies (in the binary wheel format) for the latest 3 Python versions (3.8, 3.9, 3.10 and 3.11) and for 64bit Windows. These Wheels are also used by CI running on windows, and are therefore tested together with freqtrade.

Other versions must be downloaded from the above link.

cd \\path\\freqtrade\npython -m venv .env\n.env\\Scripts\\activate.ps1\n# optionally install ta-lib from wheel\n# Eventually adjust the below filename to match the downloaded wheel\npip install --find-links build_helpers\\ TA-Lib -U\npip install -r requirements.txt\npip install -e .\nfreqtrade\n

Use Powershell

The above installation script assumes you're using powershell on a 64bit windows. Commands for the legacy CMD windows console may differ.

"},{"location":"windows_installation/#error-during-installation-on-windows","title":"Error during installation on Windows","text":"
error: Microsoft Visual C++ 14.0 is required. Get it with \"Microsoft Visual C++ Build Tools\": http://landinghub.visualstudio.com/visual-cpp-build-tools\n

Unfortunately, many packages requiring compilation don't provide a pre-built wheel. It is therefore mandatory to have a C/C++ compiler installed and available for your python environment to use.

You can download the Visual C++ build tools from here and install \"Desktop development with C++\" in it's default configuration. Unfortunately, this is a heavy download / dependency so you might want to consider WSL2 or docker compose first.

"},{"location":"includes/pairlists/","title":"Pairlists","text":""},{"location":"includes/pairlists/#pairlists-and-pairlist-handlers","title":"Pairlists and Pairlist Handlers","text":"

Pairlist Handlers define the list of pairs (pairlist) that the bot should trade. They are configured in the pairlists section of the configuration settings.

In your configuration, you can use Static Pairlist (defined by the StaticPairList Pairlist Handler) and Dynamic Pairlist (defined by the VolumePairList Pairlist Handler).

Additionally, AgeFilter, PrecisionFilter, PriceFilter, ShuffleFilter, SpreadFilter and VolatilityFilter act as Pairlist Filters, removing certain pairs and/or moving their positions in the pairlist.

If multiple Pairlist Handlers are used, they are chained and a combination of all Pairlist Handlers forms the resulting pairlist the bot uses for trading and backtesting. Pairlist Handlers are executed in the sequence they are configured. You should always configure either StaticPairList or VolumePairList as the starting Pairlist Handler.

Inactive markets are always removed from the resulting pairlist. Explicitly blacklisted pairs (those in the pair_blacklist configuration setting) are also always removed from the resulting pairlist.

"},{"location":"includes/pairlists/#pair-blacklist","title":"Pair blacklist","text":"

The pair blacklist (configured via exchange.pair_blacklist in the configuration) disallows certain pairs from trading. This can be as simple as excluding DOGE/BTC - which will remove exactly this pair.

The pair-blacklist does also support wildcards (in regex-style) - so BNB/.* will exclude ALL pairs that start with BNB. You may also use something like .*DOWN/BTC or .*UP/BTC to exclude leveraged tokens (check Pair naming conventions for your exchange!)

"},{"location":"includes/pairlists/#available-pairlist-handlers","title":"Available Pairlist Handlers","text":"
  • StaticPairList (default, if not configured differently)
  • VolumePairList
  • ProducerPairList
  • RemotePairList
  • AgeFilter
  • OffsetFilter
  • PerformanceFilter
  • PrecisionFilter
  • PriceFilter
  • ShuffleFilter
  • SpreadFilter
  • RangeStabilityFilter
  • VolatilityFilter

Testing pairlists

Pairlist configurations can be quite tricky to get right. Best use the test-pairlist utility sub-command to test your configuration quickly.

"},{"location":"includes/pairlists/#static-pair-list","title":"Static Pair List","text":"

By default, the StaticPairList method is used, which uses a statically defined pair whitelist from the configuration. The pairlist also supports wildcards (in regex-style) - so .*/BTC will include all pairs with BTC as a stake.

It uses configuration from exchange.pair_whitelist and exchange.pair_blacklist.

\"pairlists\": [\n{\"method\": \"StaticPairList\"}\n],\n

By default, only currently enabled pairs are allowed. To skip pair validation against active markets, set \"allow_inactive\": true within the StaticPairList configuration. This can be useful for backtesting expired pairs (like quarterly spot-markets). This option must be configured along with exchange.skip_pair_validation in the exchange configuration.

When used in a \"follow-up\" position (e.g. after VolumePairlist), all pairs in 'pair_whitelist' will be added to the end of the pairlist.

"},{"location":"includes/pairlists/#volume-pair-list","title":"Volume Pair List","text":"

VolumePairList employs sorting/filtering of pairs by their trading volume. It selects number_assets top pairs with sorting based on the sort_key (which can only be quoteVolume).

When used in the chain of Pairlist Handlers in a non-leading position (after StaticPairList and other Pairlist Filters), VolumePairList considers outputs of previous Pairlist Handlers, adding its sorting/selection of the pairs by the trading volume.

When used in the leading position of the chain of Pairlist Handlers, the pair_whitelist configuration setting is ignored. Instead, VolumePairList selects the top assets from all available markets with matching stake-currency on the exchange.

The refresh_period setting allows to define the period (in seconds), at which the pairlist will be refreshed. Defaults to 1800s (30 minutes). The pairlist cache (refresh_period) on VolumePairList is only applicable to generating pairlists. Filtering instances (not the first position in the list) will not apply any cache and will always use up-to-date data.

VolumePairList is per default based on the ticker data from exchange, as reported by the ccxt library:

  • The quoteVolume is the amount of quote (stake) currency traded (bought or sold) in last 24 hours.
\"pairlists\": [\n{\n\"method\": \"VolumePairList\",\n\"number_assets\": 20,\n\"sort_key\": \"quoteVolume\",\n\"min_value\": 0,\n\"refresh_period\": 1800\n}\n],\n

You can define a minimum volume with min_value - which will filter out pairs with a volume lower than the specified value in the specified timerange.

"},{"location":"includes/pairlists/#volumepairlist-advanced-mode","title":"VolumePairList Advanced mode","text":"

VolumePairList can also operate in an advanced mode to build volume over a given timerange of specified candle size. It utilizes exchange historical candle data, builds a typical price (calculated by (open+high+low)/3) and multiplies the typical price with every candle's volume. The sum is the quoteVolume over the given range. This allows different scenarios, for a more smoothened volume, when using longer ranges with larger candle sizes, or the opposite when using a short range with small candles.

For convenience lookback_days can be specified, which will imply that 1d candles will be used for the lookback. In the example below the pairlist would be created based on the last 7 days:

\"pairlists\": [\n{\n\"method\": \"VolumePairList\",\n\"number_assets\": 20,\n\"sort_key\": \"quoteVolume\",\n\"min_value\": 0,\n\"refresh_period\": 86400,\n\"lookback_days\": 7\n}\n],\n

Range look back and refresh period

When used in conjunction with lookback_days and lookback_timeframe the refresh_period can not be smaller than the candle size in seconds. As this will result in unnecessary requests to the exchanges API.

Performance implications when using lookback range

If used in first position in combination with lookback, the computation of the range based volume can be time and resource consuming, as it downloads candles for all tradable pairs. Hence it's highly advised to use the standard approach with VolumeFilter to narrow the pairlist down for further range volume calculation.

Unsupported exchanges (Bittrex, Gemini)

On some exchanges (like Bittrex and Gemini), regular VolumePairList does not work as the api does not natively provide 24h volume. This can be worked around by using candle data to build the volume. To roughly simulate 24h volume, you can use the following configuration. Please note that These pairlists will only refresh once per day.

\"pairlists\": [\n{\n\"method\": \"VolumePairList\",\n\"number_assets\": 20,\n\"sort_key\": \"quoteVolume\",\n\"min_value\": 0,\n\"refresh_period\": 86400,\n\"lookback_days\": 1\n}\n],\n

More sophisticated approach can be used, by using lookback_timeframe for candle size and lookback_period which specifies the amount of candles. This example will build the volume pairs based on a rolling period of 3 days of 1h candles:

\"pairlists\": [\n{\n\"method\": \"VolumePairList\",\n\"number_assets\": 20,\n\"sort_key\": \"quoteVolume\",\n\"min_value\": 0,\n\"refresh_period\": 3600,\n\"lookback_timeframe\": \"1h\",\n\"lookback_period\": 72\n}\n],\n

Note

VolumePairList does not support backtesting mode.

"},{"location":"includes/pairlists/#producerpairlist","title":"ProducerPairList","text":"

With ProducerPairList, you can reuse the pairlist from a Producer without explicitly defining the pairlist on each consumer.

Consumer mode is required for this pairlist to work.

The pairlist will perform a check on active pairs against the current exchange configuration to avoid attempting to trade on invalid markets.

You can limit the length of the pairlist with the optional parameter number_assets. Using \"number_assets\"=0 or omitting this key will result in the reuse of all producer pairs valid for the current setup.

\"pairlists\": [\n{\n\"method\": \"ProducerPairList\",\n\"number_assets\": 5,\n\"producer_name\": \"default\",\n}\n],\n

Combining pairlists

This pairlist can be combined with all other pairlists and filters for further pairlist reduction, and can also act as an \"additional\" pairlist, on top of already defined pairs. ProducerPairList can also be used multiple times in sequence, combining the pairs from multiple producers. Obviously in complex such configurations, the Producer may not provide data for all pairs, so the strategy must be fit for this.

"},{"location":"includes/pairlists/#remotepairlist","title":"RemotePairList","text":"

It allows the user to fetch a pairlist from a remote server or a locally stored json file within the freqtrade directory, enabling dynamic updates and customization of the trading pairlist.

The RemotePairList is defined in the pairlists section of the configuration settings. It uses the following configuration options:

\"pairlists\": [\n{\n\"method\": \"RemotePairList\",\n\"pairlist_url\": \"https://example.com/pairlist\",\n\"number_assets\": 10,\n\"refresh_period\": 1800,\n\"keep_pairlist_on_failure\": true,\n\"read_timeout\": 60,\n\"bearer_token\": \"my-bearer-token\"\n}\n]\n

The pairlist_url option specifies the URL of the remote server where the pairlist is located, or the path to a local file (if file:/// is prepended). This allows the user to use either a remote server or a local file as the source for the pairlist.

The user is responsible for providing a server or local file that returns a JSON object with the following structure:

{\n\"pairs\": [\"XRP/USDT\", \"ETH/USDT\", \"LTC/USDT\"],\n\"refresh_period\": 1800,\n}\n

The pairs property should contain a list of strings with the trading pairs to be used by the bot. The refresh_period property is optional and specifies the number of seconds that the pairlist should be cached before being refreshed.

The optional keep_pairlist_on_failure specifies whether the previous received pairlist should be used if the remote server is not reachable or returns an error. The default value is true.

The optional read_timeout specifies the maximum amount of time (in seconds) to wait for a response from the remote source, The default value is 60.

The optional bearer_token will be included in the requests Authorization Header.

Note

In case of a server error the last received pairlist will be kept if keep_pairlist_on_failure is set to true, when set to false a empty pairlist is returned.

"},{"location":"includes/pairlists/#agefilter","title":"AgeFilter","text":"

Removes pairs that have been listed on the exchange for less than min_days_listed days (defaults to 10) or more than max_days_listed days (defaults None mean infinity).

When pairs are first listed on an exchange they can suffer huge price drops and volatility in the first few days while the pair goes through its price-discovery period. Bots can often be caught out buying before the pair has finished dropping in price.

This filter allows freqtrade to ignore pairs until they have been listed for at least min_days_listed days and listed before max_days_listed.

"},{"location":"includes/pairlists/#offsetfilter","title":"OffsetFilter","text":"

Offsets an incoming pairlist by a given offset value.

As an example it can be used in conjunction with VolumeFilter to remove the top X volume pairs. Or to split a larger pairlist on two bot instances.

Example to remove the first 10 pairs from the pairlist, and takes the next 20 (taking items 10-30 of the initial list):

\"pairlists\": [\n// ...\n{\n\"method\": \"OffsetFilter\",\n\"offset\": 10,\n\"number_assets\": 20\n}\n],\n

Warning

When OffsetFilter is used to split a larger pairlist among multiple bots in combination with VolumeFilter it can not be guaranteed that pairs won't overlap due to slightly different refresh intervals for the VolumeFilter.

Note

An offset larger than the total length of the incoming pairlist will result in an empty pairlist.

"},{"location":"includes/pairlists/#performancefilter","title":"PerformanceFilter","text":"

Sorts pairs by past trade performance, as follows:

  1. Positive performance.
  2. No closed trades yet.
  3. Negative performance.

Trade count is used as a tie breaker.

You can use the minutes parameter to only consider performance of the past X minutes (rolling window). Not defining this parameter (or setting it to 0) will use all-time performance.

The optional min_profit (as ratio -> a setting of 0.01 corresponds to 1%) parameter defines the minimum profit a pair must have to be considered. Pairs below this level will be filtered out. Using this parameter without minutes is highly discouraged, as it can lead to an empty pairlist without a way to recover.

\"pairlists\": [\n// ...\n{\n\"method\": \"PerformanceFilter\",\n\"minutes\": 1440,  // rolling 24h\n\"min_profit\": 0.01  // minimal profit 1%\n}\n],\n

As this Filter uses past performance of the bot, it'll have some startup-period - and should only be used after the bot has a few 100 trades in the database.

Backtesting

PerformanceFilter does not support backtesting mode.

"},{"location":"includes/pairlists/#precisionfilter","title":"PrecisionFilter","text":"

Filters low-value coins which would not allow setting stoplosses.

Backtesting

PrecisionFilter does not support backtesting mode using multiple strategies.

"},{"location":"includes/pairlists/#pricefilter","title":"PriceFilter","text":"

The PriceFilter allows filtering of pairs by price. Currently the following price filters are supported:

  • min_price
  • max_price
  • max_value
  • low_price_ratio

The min_price setting removes pairs where the price is below the specified price. This is useful if you wish to avoid trading very low-priced pairs. This option is disabled by default, and will only apply if set to > 0.

The max_price setting removes pairs where the price is above the specified price. This is useful if you wish to trade only low-priced pairs. This option is disabled by default, and will only apply if set to > 0.

The max_value setting removes pairs where the minimum value change is above a specified value. This is useful when an exchange has unbalanced limits. For example, if step-size = 1 (so you can only buy 1, or 2, or 3, but not 1.1 Coins) - and the price is pretty high (like 20$) as the coin has risen sharply since the last limit adaption. As a result of the above, you can only buy for 20$, or 40$ - but not for 25$. On exchanges that deduct fees from the receiving currency (e.g. binance) - this can result in high value coins / amounts that are unsellable as the amount is slightly below the limit.

The low_price_ratio setting removes pairs where a raise of 1 price unit (pip) is above the low_price_ratio ratio. This option is disabled by default, and will only apply if set to > 0.

For PriceFilter at least one of its min_price, max_price or low_price_ratio settings must be applied.

Calculation example:

Min price precision for SHITCOIN/BTC is 8 decimals. If its price is 0.00000011 - one price step above would be 0.00000012, which is ~9% higher than the previous price value. You may filter out this pair by using PriceFilter with low_price_ratio set to 0.09 (9%) or with min_price set to 0.00000011, correspondingly.

Low priced pairs

Low priced pairs with high \"1 pip movements\" are dangerous since they are often illiquid and it may also be impossible to place the desired stoploss, which can often result in high losses since price needs to be rounded to the next tradable price - so instead of having a stoploss of -5%, you could end up with a stoploss of -9% simply due to price rounding.

"},{"location":"includes/pairlists/#shufflefilter","title":"ShuffleFilter","text":"

Shuffles (randomizes) pairs in the pairlist. It can be used for preventing the bot from trading some of the pairs more frequently then others when you want all pairs be treated with the same priority.

By default, ShuffleFilter will shuffle pairs once per candle. To shuffle on every iteration, set \"shuffle_frequency\" to \"iteration\" instead of the default of \"candle\".

    {\n\"method\": \"ShuffleFilter\", \"shuffle_frequency\": \"candle\",\n\"seed\": 42\n}\n

Tip

You may set the seed value for this Pairlist to obtain reproducible results, which can be useful for repeated backtesting sessions. If seed is not set, the pairs are shuffled in the non-repeatable random order. ShuffleFilter will automatically detect runmodes and apply the seed only for backtesting modes - if a seed value is set.

"},{"location":"includes/pairlists/#spreadfilter","title":"SpreadFilter","text":"

Removes pairs that have a difference between asks and bids above the specified ratio, max_spread_ratio (defaults to 0.005).

Example:

If DOGE/BTC maximum bid is 0.00000026 and minimum ask is 0.00000027, the ratio is calculated as: 1 - bid/ask ~= 0.037 which is > 0.005 and this pair will be filtered out.

"},{"location":"includes/pairlists/#rangestabilityfilter","title":"RangeStabilityFilter","text":"

Removes pairs where the difference between lowest low and highest high over lookback_days days is below min_rate_of_change or above max_rate_of_change. Since this is a filter that requires additional data, the results are cached for refresh_period.

In the below example: If the trading range over the last 10 days is <1% or >99%, remove the pair from the whitelist.

\"pairlists\": [\n{\n\"method\": \"RangeStabilityFilter\",\n\"lookback_days\": 10,\n\"min_rate_of_change\": 0.01,\n\"max_rate_of_change\": 0.99,\n\"refresh_period\": 1440\n}\n]\n

Tip

This Filter can be used to automatically remove stable coin pairs, which have a very low trading range, and are therefore extremely difficult to trade with profit. Additionally, it can also be used to automatically remove pairs with extreme high/low variance over a given amount of time.

"},{"location":"includes/pairlists/#volatilityfilter","title":"VolatilityFilter","text":"

Volatility is the degree of historical variation of a pairs over time, it is measured by the standard deviation of logarithmic daily returns. Returns are assumed to be normally distributed, although actual distribution might be different. In a normal distribution, 68% of observations fall within one standard deviation and 95% of observations fall within two standard deviations. Assuming a volatility of 0.05 means that the expected returns for 20 out of 30 days is expected to be less than 5% (one standard deviation). Volatility is a positive ratio of the expected deviation of return and can be greater than 1.00. Please refer to the wikipedia definition of volatility.

This filter removes pairs if the average volatility over a lookback_days days is below min_volatility or above max_volatility. Since this is a filter that requires additional data, the results are cached for refresh_period.

This filter can be used to narrow down your pairs to a certain volatility or avoid very volatile pairs.

In the below example: If the volatility over the last 10 days is not in the range of 0.05-0.50, remove the pair from the whitelist. The filter is applied every 24h.

\"pairlists\": [\n{\n\"method\": \"VolatilityFilter\",\n\"lookback_days\": 10,\n\"min_volatility\": 0.05,\n\"max_volatility\": 0.50,\n\"refresh_period\": 86400\n}\n]\n
"},{"location":"includes/pairlists/#full-example-of-pairlist-handlers","title":"Full example of Pairlist Handlers","text":"

The below example blacklists BNB/BTC, uses VolumePairList with 20 assets, sorting pairs by quoteVolume and applies PrecisionFilter and PriceFilter, filtering all assets where 1 price unit is > 1%. Then the SpreadFilter and VolatilityFilter is applied and pairs are finally shuffled with the random seed set to some predefined value.

\"exchange\": {\n\"pair_whitelist\": [],\n\"pair_blacklist\": [\"BNB/BTC\"]\n},\n\"pairlists\": [\n{\n\"method\": \"VolumePairList\",\n\"number_assets\": 20,\n\"sort_key\": \"quoteVolume\"\n},\n{\"method\": \"AgeFilter\", \"min_days_listed\": 10},\n{\"method\": \"PrecisionFilter\"},\n{\"method\": \"PriceFilter\", \"low_price_ratio\": 0.01},\n{\"method\": \"SpreadFilter\", \"max_spread_ratio\": 0.005},\n{\n\"method\": \"RangeStabilityFilter\",\n\"lookback_days\": 10,\n\"min_rate_of_change\": 0.01,\n\"refresh_period\": 1440\n},\n{\n\"method\": \"VolatilityFilter\",\n\"lookback_days\": 10,\n\"min_volatility\": 0.05,\n\"max_volatility\": 0.50,\n\"refresh_period\": 86400\n},\n{\"method\": \"ShuffleFilter\", \"seed\": 42}\n],\n
"},{"location":"includes/pricing/","title":"Pricing","text":""},{"location":"includes/pricing/#prices-used-for-orders","title":"Prices used for orders","text":"

Prices for regular orders can be controlled via the parameter structures entry_pricing for trade entries and exit_pricing for trade exits. Prices are always retrieved right before an order is placed, either by querying the exchange tickers or by using the orderbook data.

Note

Orderbook data used by Freqtrade are the data retrieved from exchange by the ccxt's function fetch_order_book(), i.e. are usually data from the L2-aggregated orderbook, while the ticker data are the structures returned by the ccxt's fetch_ticker()/fetch_tickers() functions. Refer to the ccxt library documentation for more details.

Using market orders

Please read the section Market order pricing section when using market orders.

"},{"location":"includes/pricing/#entry-price","title":"Entry price","text":""},{"location":"includes/pricing/#enter-price-side","title":"Enter price side","text":"

The configuration setting entry_pricing.price_side defines the side of the orderbook the bot looks for when buying.

The following displays an orderbook.

...\n103\n102\n101  # ask\n-------------Current spread\n99   # bid\n98\n97\n...\n

If entry_pricing.price_side is set to \"bid\", then the bot will use 99 as entry price. In line with that, if entry_pricing.price_side is set to \"ask\", then the bot will use 101 as entry price.

Depending on the order direction (long/short), this will lead to different results. Therefore we recommend to use \"same\" or \"other\" for this configuration instead. This would result in the following pricing matrix:

direction Order setting price crosses spread long buy ask 101 yes long buy bid 99 no long buy same 99 no long buy other 101 yes short sell ask 101 no short sell bid 99 yes short sell same 101 no short sell other 99 yes

Using the other side of the orderbook often guarantees quicker filled orders, but the bot can also end up paying more than what would have been necessary. Taker fees instead of maker fees will most likely apply even when using limit buy orders. Also, prices at the \"other\" side of the spread are higher than prices at the \"bid\" side in the orderbook, so the order behaves similar to a market order (however with a maximum price).

"},{"location":"includes/pricing/#entry-price-with-orderbook-enabled","title":"Entry price with Orderbook enabled","text":"

When entering a trade with the orderbook enabled (entry_pricing.use_order_book=True), Freqtrade fetches the entry_pricing.order_book_top entries from the orderbook and uses the entry specified as entry_pricing.order_book_top on the configured side (entry_pricing.price_side) of the orderbook. 1 specifies the topmost entry in the orderbook, while 2 would use the 2nd entry in the orderbook, and so on.

"},{"location":"includes/pricing/#entry-price-without-orderbook-enabled","title":"Entry price without Orderbook enabled","text":"

The following section uses side as the configured entry_pricing.price_side (defaults to \"same\").

When not using orderbook (entry_pricing.use_order_book=False), Freqtrade uses the best side price from the ticker if it's below the last traded price from the ticker. Otherwise (when the side price is above the last price), it calculates a rate between side and last price based on entry_pricing.price_last_balance.

The entry_pricing.price_last_balance configuration parameter controls this. A value of 0.0 will use side price, while 1.0 will use the last price and values between those interpolate between ask and last price.

"},{"location":"includes/pricing/#check-depth-of-market","title":"Check depth of market","text":"

When check depth of market is enabled (entry_pricing.check_depth_of_market.enabled=True), the entry signals are filtered based on the orderbook depth (sum of all amounts) for each orderbook side.

Orderbook bid (buy) side depth is then divided by the orderbook ask (sell) side depth and the resulting delta is compared to the value of the entry_pricing.check_depth_of_market.bids_to_ask_delta parameter. The entry order is only executed if the orderbook delta is greater than or equal to the configured delta value.

Note

A delta value below 1 means that ask (sell) orderbook side depth is greater than the depth of the bid (buy) orderbook side, while a value greater than 1 means opposite (depth of the buy side is higher than the depth of the sell side).

"},{"location":"includes/pricing/#exit-price","title":"Exit price","text":""},{"location":"includes/pricing/#exit-price-side","title":"Exit price side","text":"

The configuration setting exit_pricing.price_side defines the side of the spread the bot looks for when exiting a trade.

The following displays an orderbook:

...\n103\n102\n101  # ask\n-------------Current spread\n99   # bid\n98\n97\n...\n

If exit_pricing.price_side is set to \"ask\", then the bot will use 101 as exiting price. In line with that, if exit_pricing.price_side is set to \"bid\", then the bot will use 99 as exiting price.

Depending on the order direction (long/short), this will lead to different results. Therefore we recommend to use \"same\" or \"other\" for this configuration instead. This would result in the following pricing matrix:

Direction Order setting price crosses spread long sell ask 101 no long sell bid 99 yes long sell same 101 no long sell other 99 yes short buy ask 101 yes short buy bid 99 no short buy same 99 no short buy other 101 yes"},{"location":"includes/pricing/#exit-price-with-orderbook-enabled","title":"Exit price with Orderbook enabled","text":"

When exiting with the orderbook enabled (exit_pricing.use_order_book=True), Freqtrade fetches the exit_pricing.order_book_top entries in the orderbook and uses the entry specified as exit_pricing.order_book_top from the configured side (exit_pricing.price_side) as trade exit price.

1 specifies the topmost entry in the orderbook, while 2 would use the 2nd entry in the orderbook, and so on.

"},{"location":"includes/pricing/#exit-price-without-orderbook-enabled","title":"Exit price without Orderbook enabled","text":"

The following section uses side as the configured exit_pricing.price_side (defaults to \"ask\").

When not using orderbook (exit_pricing.use_order_book=False), Freqtrade uses the best side price from the ticker if it's above the last traded price from the ticker. Otherwise (when the side price is below the last price), it calculates a rate between side and last price based on exit_pricing.price_last_balance.

The exit_pricing.price_last_balance configuration parameter controls this. A value of 0.0 will use side price, while 1.0 will use the last price and values between those interpolate between side and last price.

"},{"location":"includes/pricing/#market-order-pricing","title":"Market order pricing","text":"

When using market orders, prices should be configured to use the \"correct\" side of the orderbook to allow realistic pricing detection. Assuming both entry and exits are using market orders, a configuration similar to the following must be used

  \"order_types\": {\n    \"entry\": \"market\",\n    \"exit\": \"market\"\n    // ...\n  },\n  \"entry_pricing\": {\n    \"price_side\": \"other\",\n    // ...\n  },\n  \"exit_pricing\":{\n    \"price_side\": \"other\",\n    // ...\n  },\n

Obviously, if only one side is using limit orders, different pricing combinations can be used.

"},{"location":"includes/protections/","title":"Protections","text":""},{"location":"includes/protections/#protections","title":"Protections","text":"

Beta feature

This feature is still in it's testing phase. Should you notice something you think is wrong please let us know via Discord or via Github Issue.

Protections will protect your strategy from unexpected events and market conditions by temporarily stop trading for either one pair, or for all pairs. All protection end times are rounded up to the next candle to avoid sudden, unexpected intra-candle buys.

Note

Not all Protections will work for all strategies, and parameters will need to be tuned for your strategy to improve performance.

Tip

Each Protection can be configured multiple times with different parameters, to allow different levels of protection (short-term / long-term).

Backtesting

Protections are supported by backtesting and hyperopt, but must be explicitly enabled by using the --enable-protections flag.

Setting protections from the configuration

Setting protections from the configuration via \"protections\": [], key should be considered deprecated and will be removed in a future version. It is also no longer guaranteed that your protections apply to the strategy in cases where the strategy defines protections as property.

"},{"location":"includes/protections/#available-protections","title":"Available Protections","text":"
  • StoplossGuard Stop trading if a certain amount of stoploss occurred within a certain time window.
  • MaxDrawdown Stop trading if max-drawdown is reached.
  • LowProfitPairs Lock pairs with low profits
  • CooldownPeriod Don't enter a trade right after selling a trade.
"},{"location":"includes/protections/#common-settings-to-all-protections","title":"Common settings to all Protections","text":"Parameter Description method Protection name to use. Datatype: String, selected from available Protections stop_duration_candles For how many candles should the lock be set? Datatype: Positive integer (in candles) stop_duration how many minutes should protections be locked. Cannot be used together with stop_duration_candles. Datatype: Float (in minutes) lookback_period_candles Only trades that completed within the last lookback_period_candles candles will be considered. This setting may be ignored by some Protections. Datatype: Positive integer (in candles). lookback_period Only trades that completed after current_time - lookback_period will be considered. Cannot be used together with lookback_period_candles. This setting may be ignored by some Protections. Datatype: Float (in minutes) trade_limit Number of trades required at minimum (not used by all Protections). Datatype: Positive integer

Durations

Durations (stop_duration* and lookback_period* can be defined in either minutes or candles). For more flexibility when testing different timeframes, all below examples will use the \"candle\" definition.

"},{"location":"includes/protections/#stoploss-guard","title":"Stoploss Guard","text":"

StoplossGuard selects all trades within lookback_period in minutes (or in candles when using lookback_period_candles). If trade_limit or more trades resulted in stoploss, trading will stop for stop_duration in minutes (or in candles when using stop_duration_candles).

This applies across all pairs, unless only_per_pair is set to true, which will then only look at one pair at a time.

Similarly, this protection will by default look at all trades (long and short). For futures bots, setting only_per_side will make the bot only consider one side, and will then only lock this one side, allowing for example shorts to continue after a series of long stoplosses.

required_profit will determine the required relative profit (or loss) for stoplosses to consider. This should normally not be set and defaults to 0.0 - which means all losing stoplosses will be triggering a block.

The below example stops trading for all pairs for 4 candles after the last trade if the bot hit stoploss 4 times within the last 24 candles.

@property\ndef protections(self):\n    return [\n        {\n            \"method\": \"StoplossGuard\",\n            \"lookback_period_candles\": 24,\n            \"trade_limit\": 4,\n            \"stop_duration_candles\": 4,\n            \"required_profit\": 0.0,\n            \"only_per_pair\": False,\n            \"only_per_side\": False\n        }\n    ]\n

Note

StoplossGuard considers all trades with the results \"stop_loss\", \"stoploss_on_exchange\" and \"trailing_stop_loss\" if the resulting profit was negative. trade_limit and lookback_period will need to be tuned for your strategy.

"},{"location":"includes/protections/#maxdrawdown","title":"MaxDrawdown","text":"

MaxDrawdown uses all trades within lookback_period in minutes (or in candles when using lookback_period_candles) to determine the maximum drawdown. If the drawdown is below max_allowed_drawdown, trading will stop for stop_duration in minutes (or in candles when using stop_duration_candles) after the last trade - assuming that the bot needs some time to let markets recover.

The below sample stops trading for 12 candles if max-drawdown is > 20% considering all pairs - with a minimum of trade_limit trades - within the last 48 candles. If desired, lookback_period and/or stop_duration can be used.

@property\ndef protections(self):\n    return  [\n        {\n            \"method\": \"MaxDrawdown\",\n            \"lookback_period_candles\": 48,\n            \"trade_limit\": 20,\n            \"stop_duration_candles\": 12,\n            \"max_allowed_drawdown\": 0.2\n        },\n    ]\n
"},{"location":"includes/protections/#low-profit-pairs","title":"Low Profit Pairs","text":"

LowProfitPairs uses all trades for a pair within lookback_period in minutes (or in candles when using lookback_period_candles) to determine the overall profit ratio. If that ratio is below required_profit, that pair will be locked for stop_duration in minutes (or in candles when using stop_duration_candles).

For futures bots, setting only_per_side will make the bot only consider one side, and will then only lock this one side, allowing for example shorts to continue after a series of long losses.

The below example will stop trading a pair for 60 minutes if the pair does not have a required profit of 2% (and a minimum of 2 trades) within the last 6 candles.

@property\ndef protections(self):\n    return [\n        {\n            \"method\": \"LowProfitPairs\",\n            \"lookback_period_candles\": 6,\n            \"trade_limit\": 2,\n            \"stop_duration\": 60,\n            \"required_profit\": 0.02,\n            \"only_per_pair\": False,\n        }\n    ]\n
"},{"location":"includes/protections/#cooldown-period","title":"Cooldown Period","text":"

CooldownPeriod locks a pair for stop_duration in minutes (or in candles when using stop_duration_candles) after selling, avoiding a re-entry for this pair for stop_duration minutes.

The below example will stop trading a pair for 2 candles after closing a trade, allowing this pair to \"cool down\".

@property\ndef protections(self):\n    return  [\n        {\n            \"method\": \"CooldownPeriod\",\n            \"stop_duration_candles\": 2\n        }\n    ]\n

Note

This Protection applies only at pair-level, and will never lock all pairs globally. This Protection does not consider lookback_period as it only looks at the latest trade.

"},{"location":"includes/protections/#full-example-of-protections","title":"Full example of Protections","text":"

All protections can be combined at will, also with different parameters, creating a increasing wall for under-performing pairs. All protections are evaluated in the sequence they are defined.

The below example assumes a timeframe of 1 hour:

  • Locks each pair after selling for an additional 5 candles (CooldownPeriod), giving other pairs a chance to get filled.
  • Stops trading for 4 hours (4 * 1h candles) if the last 2 days (48 * 1h candles) had 20 trades, which caused a max-drawdown of more than 20%. (MaxDrawdown).
  • Stops trading if more than 4 stoploss occur for all pairs within a 1 day (24 * 1h candles) limit (StoplossGuard).
  • Locks all pairs that had 2 Trades within the last 6 hours (6 * 1h candles) with a combined profit ratio of below 0.02 (<2%) (LowProfitPairs).
  • Locks all pairs for 2 candles that had a profit of below 0.01 (<1%) within the last 24h (24 * 1h candles), a minimum of 4 trades.
from freqtrade.strategy import IStrategy\n\nclass AwesomeStrategy(IStrategy)\n    timeframe = '1h'\n\n    @property\n    def protections(self):\n        return [\n            {\n                \"method\": \"CooldownPeriod\",\n                \"stop_duration_candles\": 5\n            },\n            {\n                \"method\": \"MaxDrawdown\",\n                \"lookback_period_candles\": 48,\n                \"trade_limit\": 20,\n                \"stop_duration_candles\": 4,\n                \"max_allowed_drawdown\": 0.2\n            },\n            {\n                \"method\": \"StoplossGuard\",\n                \"lookback_period_candles\": 24,\n                \"trade_limit\": 4,\n                \"stop_duration_candles\": 2,\n                \"only_per_pair\": False\n            },\n            {\n                \"method\": \"LowProfitPairs\",\n                \"lookback_period_candles\": 6,\n                \"trade_limit\": 2,\n                \"stop_duration_candles\": 60,\n                \"required_profit\": 0.02\n            },\n            {\n                \"method\": \"LowProfitPairs\",\n                \"lookback_period_candles\": 24,\n                \"trade_limit\": 4,\n                \"stop_duration_candles\": 2,\n                \"required_profit\": 0.01\n            }\n        ]\n    # ...\n
"}]} \ No newline at end of file diff --git a/en/2023.4/sitemap.xml b/en/2023.4/sitemap.xml new file mode 100644 index 000000000..e00499c45 --- /dev/null +++ b/en/2023.4/sitemap.xml @@ -0,0 +1,238 @@ + + + + https://www.freqtrade.io/2023.4/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/advanced-backtesting/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/advanced-hyperopt/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/advanced-setup/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/backtesting/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/bot-basics/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/bot-usage/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/configuration/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/data-analysis/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/data-download/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/deprecated/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/developer/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/docker_quickstart/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/edge/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/exchanges/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/faq/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/freqai-configuration/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/freqai-developers/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/freqai-feature-engineering/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/freqai-parameter-table/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/freqai-reinforcement-learning/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/freqai-running/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/freqai/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/hyperopt/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/installation/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/leverage/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/plotting/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/plugins/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/producer-consumer/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/rest-api/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/sandbox-testing/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/sql_cheatsheet/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/stoploss/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/strategy-advanced/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/strategy-callbacks/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/strategy-customization/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/strategy_analysis_example/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/strategy_migration/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/telegram-usage/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/trade-object/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/updating/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/utils/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/webhook-config/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/windows_installation/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/includes/pairlists/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/includes/pricing/ + 2024-08-20 + daily + + + https://www.freqtrade.io/2023.4/includes/protections/ + 2024-08-20 + daily + + \ No newline at end of file diff --git a/en/2023.4/sitemap.xml.gz b/en/2023.4/sitemap.xml.gz new file mode 100644 index 0000000000000000000000000000000000000000..df379a3ac567d3106aac8f687d0c166b5aa1813c GIT binary patch literal 594 zcmV-Y0wv0$C5+Rr>}FJth?w4Heh2k-w6BHW&L>?-QtLeZ19`vOTDZX=yfpG_nYe1 z&mZce`fT6UIXOaJl*OJl)xaFTH_bFnHCe_HBGn=F4e6l_sJTxz%lcc@zAegiF>rgj zUTYV0d$B@yGrNQbPGGRMp;;|gU+ec~-CS0^egwp0Fl~ANu9hEawJb~Sb1%#E0KUhI z#nymz=XLWkbE9dVIR63$<_KP+Q61<#BXM!o3}}aQ#1taW0JZG39OZJ(;=gx+%LXN@ zNj769=us}B4t{6*oPdi$%EdImpumGWlO^GpCQRN1FoH2LCe*+vn(>G(6iFA181x>~ z_K>aK6GIXldeBG~Aw1m-3O?C8I0%L-b70j>wERx{Jm4fAb%($?p+etVkC;R!Mt9~2 z2}Wd0ilK8To;RWOyO1=F=vg_0HnyWZ8u(WORofo0c} z(d8oWg5yz4!DL-Dy@YzoqvYf`A+dreSa}G%>B32gZ*|bYZS_!c5E>8#3ng(VXPihf zVM&g$KdVP8ZU(5j#GXq>ZVqnMx$J?_{t+H4AH3T~M`?>E7^CDY`H4K*UV4GY=&(SJ z(tDrRu?pR9)PhwtM+36sQ)cS~tJ8$t5W@c69udrGy$KU-=TUE1uU$4sO$3{qC8jyB gCaXmsr5Dn`MV87I_z!-)-Uzw<3k+`6WH28902DwYZvX%Q literal 0 HcmV?d00001 diff --git a/en/2023.4/sql_cheatsheet/index.html b/en/2023.4/sql_cheatsheet/index.html new file mode 100644 index 000000000..3a20bb81d --- /dev/null +++ b/en/2023.4/sql_cheatsheet/index.html @@ -0,0 +1,1585 @@ + + + + + + + + + + + + + + + + + + + + + + + + SQL Cheat-sheet - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

SQL Helper

+

This page contains some help if you want to edit your sqlite db.

+

Install sqlite3

+

Sqlite3 is a terminal based sqlite application. +Feel free to use a visual Database editor like SqliteBrowser if you feel more comfortable with that.

+

Ubuntu/Debian installation

+
sudo apt-get install sqlite3
+
+

Using sqlite3 via docker

+

The freqtrade docker image does contain sqlite3, so you can edit the database without having to install anything on the host system.

+
docker compose exec freqtrade /bin/bash
+sqlite3 <database-file>.sqlite
+
+

Open the DB

+
sqlite3
+.open <filepath>
+
+

Table structure

+

List tables

+
.tables
+
+

Display table structure

+
.schema <table_name>
+
+

Get all trades in the table

+
SELECT * FROM trades;
+
+

Fix trade still open after a manual exit on the exchange

+
+

Warning

+

Manually selling a pair on the exchange will not be detected by the bot and it will try to sell anyway. Whenever possible, /forceexit should be used to accomplish the same thing.
+It is strongly advised to backup your database file before making any manual changes.

+
+
+

Note

+

This should not be necessary after /forceexit, as force_exit orders are closed automatically by the bot on the next iteration.

+
+
UPDATE trades
+SET is_open=0,
+  close_date=<close_date>,
+  close_rate=<close_rate>,
+  close_profit = close_rate / open_rate - 1,
+  close_profit_abs = (amount * <close_rate> * (1 - fee_close) - (amount * (open_rate * (1 - fee_open)))),
+  exit_reason=<exit_reason>
+WHERE id=<trade_ID_to_update>;
+
+

Example

+
UPDATE trades
+SET is_open=0,
+  close_date='2020-06-20 03:08:45.103418',
+  close_rate=0.19638016,
+  close_profit=0.0496,
+  close_profit_abs = (amount * 0.19638016 * (1 - fee_close) - (amount * (open_rate * (1 - fee_open)))),
+  exit_reason='force_exit'  
+WHERE id=31;
+
+

Remove trade from the database

+
+

Use RPC Methods to delete trades

+

Consider using /delete <tradeid> via telegram or rest API. That's the recommended way to deleting trades.

+
+

If you'd still like to remove a trade from the database directly, you can use the below query.

+
+

Danger

+

Some systems (Ubuntu) disable foreign keys in their sqlite3 packaging. When using sqlite - please ensure that foreign keys are on by running PRAGMA foreign_keys = ON before the above query.

+
+
DELETE FROM trades WHERE id = <tradeid>;
+
+DELETE FROM trades WHERE id = 31;
+
+
+

Warning

+

This will remove this trade from the database. Please make sure you got the correct id and NEVER run this query without the where clause.

+
+

Use a different database system

+

Freqtrade is using SQLAlchemy, which supports multiple different database systems. As such, a multitude of database systems should be supported. +Freqtrade does not depend or install any additional database driver. Please refer to the SQLAlchemy docs on installation instructions for the respective database systems.

+

The following systems have been tested and are known to work with freqtrade:

+
    +
  • sqlite (default)
  • +
  • PostgreSQL)
  • +
  • MariaDB
  • +
+
+

Warning

+

By using one of the below database systems, you acknowledge that you know how to manage such a system. The freqtrade team will not provide any support with setup or maintenance (or backups) of the below database systems.

+
+

PostgreSQL

+

Installation: +pip install psycopg2-binary

+

Usage: +... --db-url postgresql+psycopg2://<username>:<password>@localhost:5432/<database>

+

Freqtrade will automatically create the tables necessary upon startup.

+

If you're running different instances of Freqtrade, you must either setup one database per Instance or use different users / schemas for your connections.

+

MariaDB / MySQL

+

Freqtrade supports MariaDB by using SQLAlchemy, which supports multiple different database systems.

+

Installation: +pip install pymysql

+

Usage: +... --db-url mysql+pymysql://<username>:<password>@localhost:3306/<database>

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/stoploss/index.html b/en/2023.4/stoploss/index.html new file mode 100644 index 000000000..e365bd747 --- /dev/null +++ b/en/2023.4/stoploss/index.html @@ -0,0 +1,1714 @@ + + + + + + + + + + + + + + + + + + + + + + + + Stoploss - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Stop Loss

+

The stoploss configuration parameter is loss as ratio that should trigger a sale. +For example, value -0.10 will cause immediate sell if the profit dips below -10% for a given trade. This parameter is optional. +Stoploss calculations do include fees, so a stoploss of -10% is placed exactly 10% below the entry point.

+

Most of the strategy files already include the optimal stoploss value.

+
+

Info

+

All stoploss properties mentioned in this file can be set in the Strategy, or in the configuration.
+Configuration values will override the strategy values.

+
+

Stop Loss On-Exchange/Freqtrade

+

Those stoploss modes can be on exchange or off exchange.

+

These modes can be configured with these values:

+
    'emergency_exit': 'market',
+    'stoploss_on_exchange': False
+    'stoploss_on_exchange_interval': 60,
+    'stoploss_on_exchange_limit_ratio': 0.99
+
+

Stoploss on exchange is only supported for the following exchanges, and not all exchanges support both stop-limit and stop-market. +The Order-type will be ignored if only one mode is available.

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
Exchangestop-loss type
Binancelimit
Binance Futuresmarket, limit
Huobilimit
krakenmarket, limit
Gatelimit
Okxlimit
Kucoinstop-limit, stop-market
+
+

Tight stoploss

+

Do not set too low/tight stoploss value when using stop loss on exchange!
+If set to low/tight you will have greater risk of missing fill on the order and stoploss will not work.

+
+

stoploss_on_exchange and stoploss_on_exchange_limit_ratio

+

Enable or Disable stop loss on exchange. +If the stoploss is on exchange it means a stoploss limit order is placed on the exchange immediately after buy order fills. This will protect you against sudden crashes in market, as the order execution happens purely within the exchange, and has no potential network overhead.

+

If stoploss_on_exchange uses limit orders, the exchange needs 2 prices, the stoploss_price and the Limit price.
+stoploss defines the stop-price where the limit order is placed - and limit should be slightly below this.
+If an exchange supports both limit and market stoploss orders, then the value of stoploss will be used to determine the stoploss type.

+

Calculation example: we bought the asset at 100$.
+Stop-price is 95$, then limit would be 95 * 0.99 = 94.05$ - so the limit order fill can happen between 95$ and 94.05$.

+

For example, assuming the stoploss is on exchange, and trailing stoploss is enabled, and the market is going up, then the bot automatically cancels the previous stoploss order and puts a new one with a stop value higher than the previous stoploss order.

+
+

Note

+

If stoploss_on_exchange is enabled and the stoploss is cancelled manually on the exchange, then the bot will create a new stoploss order.

+
+

stoploss_on_exchange_interval

+

In case of stoploss on exchange there is another parameter called stoploss_on_exchange_interval. This configures the interval in seconds at which the bot will check the stoploss and update it if necessary.
+The bot cannot do these every 5 seconds (at each iteration), otherwise it would get banned by the exchange. +So this parameter will tell the bot how often it should update the stoploss order. The default value is 60 (1 minute). +This same logic will reapply a stoploss order on the exchange should you cancel it accidentally.

+

stoploss_price_type

+
+

Only applies to futures

+

stoploss_price_type only applies to futures markets (on exchanges where it's available). +Freqtrade will perform a validation of this setting on startup, failing to start if an invalid setting for your exchange has been selected. +Supported price types are gonna differs between each exchanges. Please check with your exchange on which price types it supports.

+
+

Stoploss on exchange on futures markets can trigger on different price types. +The naming for these prices in exchange terminology often varies, but is usually something around "last" (or "contract price" ), "mark" and "index".

+

Acceptable values for this setting are "last", "mark" and "index" - which freqtrade will transfer automatically to the corresponding API type, and place the stoploss on exchange order correspondingly.

+

force_exit

+

force_exit is an optional value, which defaults to the same value as exit and is used when sending a /forceexit command from Telegram or from the Rest API.

+

force_entry

+

force_entry is an optional value, which defaults to the same value as entry and is used when sending a /forceentry command from Telegram or from the Rest API.

+

emergency_exit

+

emergency_exit is an optional value, which defaults to market and is used when creating stop loss on exchange orders fails. +The below is the default which is used if not changed in strategy or configuration file.

+

Example from strategy file:

+
order_types = {
+    "entry": "limit",
+    "exit": "limit",
+    "emergency_exit": "market",
+    "stoploss": "market",
+    "stoploss_on_exchange": True,
+    "stoploss_on_exchange_interval": 60,
+    "stoploss_on_exchange_limit_ratio": 0.99
+}
+
+

Stop Loss Types

+

At this stage the bot contains the following stoploss support modes:

+
    +
  1. Static stop loss.
  2. +
  3. Trailing stop loss.
  4. +
  5. Trailing stop loss, custom positive loss.
  6. +
  7. Trailing stop loss only once the trade has reached a certain offset.
  8. +
  9. Custom stoploss function
  10. +
+

Static Stop Loss

+

This is very simple, you define a stop loss of x (as a ratio of price, i.e. x * 100% of price). This will try to sell the asset once the loss exceeds the defined loss.

+

Example of stop loss:

+
    stoploss = -0.10
+
+

For example, simplified math:

+
    +
  • the bot buys an asset at a price of 100$
  • +
  • the stop loss is defined at -10%
  • +
  • the stop loss would get triggered once the asset drops below 90$
  • +
+

Trailing Stop Loss

+

The initial value for this is stoploss, just as you would define your static Stop loss. +To enable trailing stoploss:

+
    stoploss = -0.10
+    trailing_stop = True
+
+

This will now activate an algorithm, which automatically moves the stop loss up every time the price of your asset increases.

+

For example, simplified math:

+
    +
  • the bot buys an asset at a price of 100$
  • +
  • the stop loss is defined at -10%
  • +
  • the stop loss would get triggered once the asset drops below 90$
  • +
  • assuming the asset now increases to 102$
  • +
  • the stop loss will now be -10% of 102$ = 91.8$
  • +
  • now the asset drops in value to 101$, the stop loss will still be 91.8$ and would trigger at 91.8$.
  • +
+

In summary: The stoploss will be adjusted to be always be -10% of the highest observed price.

+

Trailing stop loss, custom positive loss

+

You could also have a default stop loss when you are in the red with your buy (buy - fee), but once you hit a positive result (or an offset you define) the system will utilize a new stop loss, which can have a different value. +For example, your default stop loss is -10%, but once you have more than 0% profit (example 0.1%) a different trailing stoploss will be used.

+
+

Note

+

If you want the stoploss to only be changed when you break even of making a profit (what most users want) please refer to next section with offset enabled.

+
+

Both values require trailing_stop to be set to true and trailing_stop_positive with a value.

+
    stoploss = -0.10
+    trailing_stop = True
+    trailing_stop_positive = 0.02
+    trailing_stop_positive_offset = 0.0
+    trailing_only_offset_is_reached = False  # Default - not necessary for this example
+
+

For example, simplified math:

+
    +
  • the bot buys an asset at a price of 100$
  • +
  • the stop loss is defined at -10%
  • +
  • the stop loss would get triggered once the asset drops below 90$
  • +
  • assuming the asset now increases to 102$
  • +
  • the stop loss will now be -2% of 102$ = 99.96$ (99.96$ stop loss will be locked in and will follow asset price increments with -2%)
  • +
  • now the asset drops in value to 101$, the stop loss will still be 99.96$ and would trigger at 99.96$
  • +
+

The 0.02 would translate to a -2% stop loss. +Before this, stoploss is used for the trailing stoploss.

+
+

Use an offset to change your stoploss

+

Use trailing_stop_positive_offset to ensure that your new trailing stoploss will be in profit by setting trailing_stop_positive_offset higher than trailing_stop_positive. Your first new stoploss value will then already have locked in profits.

+

Example with simplified math:

+
    stoploss = -0.10
+    trailing_stop = True
+    trailing_stop_positive = 0.02
+    trailing_stop_positive_offset = 0.03
+
+
    +
  • the bot buys an asset at a price of 100$
  • +
  • the stop loss is defined at -10%, so the stop loss would get triggered once the asset drops below 90$
  • +
  • assuming the asset now increases to 102$
  • +
  • the stoploss will now be at 91.8$ - 10% below the highest observed rate
  • +
  • assuming the asset now increases to 103.5$ (above the offset configured)
  • +
  • the stop loss will now be -2% of 103.5$ = 101.43$
  • +
  • now the asset drops in value to 102$, the stop loss will still be 101.43$ and would trigger once price breaks below 101.43$
  • +
+
+

Trailing stop loss only once the trade has reached a certain offset

+

You can also keep a static stoploss until the offset is reached, and then trail the trade to take profits once the market turns.

+

If trailing_only_offset_is_reached = True then the trailing stoploss is only activated once the offset is reached. Until then, the stoploss remains at the configured stoploss. +This option can be used with or without trailing_stop_positive, but uses trailing_stop_positive_offset as offset.

+

Configuration (offset is buy-price + 3%):

+
    stoploss = -0.10
+    trailing_stop = True
+    trailing_stop_positive = 0.02
+    trailing_stop_positive_offset = 0.03
+    trailing_only_offset_is_reached = True
+
+

For example, simplified math:

+
    +
  • the bot buys an asset at a price of 100$
  • +
  • the stop loss is defined at -10%
  • +
  • the stop loss would get triggered once the asset drops below 90$
  • +
  • stoploss will remain at 90$ unless asset increases to or above the configured offset
  • +
  • assuming the asset now increases to 103$ (where we have the offset configured)
  • +
  • the stop loss will now be -2% of 103$ = 100.94$
  • +
  • now the asset drops in value to 101$, the stop loss will still be 100.94$ and would trigger at 100.94$
  • +
+
+

Tip

+

Make sure to have this value (trailing_stop_positive_offset) lower than minimal ROI, otherwise minimal ROI will apply first and sell the trade.

+
+

Stoploss and Leverage

+

Stoploss should be thought of as "risk on this trade" - so a stoploss of 10% on a 100$ trade means you are willing to lose 10$ (10%) on this trade - which would trigger if the price moves 10% to the downside.

+

When using leverage, the same principle is applied - with stoploss defining the risk on the trade (the amount you are willing to lose).

+

Therefore, a stoploss of 10% on a 10x trade would trigger on a 1% price move. +If your stake amount (own capital) was 100$ - this trade would be 1000$ at 10x (after leverage). +If price moves 1% - you've lost 10$ of your own capital - therfore stoploss will trigger in this case.

+

Make sure to be aware of this, and avoid using too tight stoploss (at 10x leverage, 10% risk may be too little to allow the trade to "breath" a little).

+

Changing stoploss on open trades

+

A stoploss on an open trade can be changed by changing the value in the configuration or strategy and use the /reload_config command (alternatively, completely stopping and restarting the bot also works).

+

The new stoploss value will be applied to open trades (and corresponding log-messages will be generated).

+

Limitations

+

Stoploss values cannot be changed if trailing_stop is enabled and the stoploss has already been adjusted, or if Edge is enabled (since Edge would recalculate stoploss based on the current market situation).

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/strategy-advanced/index.html b/en/2023.4/strategy-advanced/index.html new file mode 100644 index 000000000..bbc9520b9 --- /dev/null +++ b/en/2023.4/strategy-advanced/index.html @@ -0,0 +1,1570 @@ + + + + + + + + + + + + + + + + + + + + + + + + Advanced Strategy - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Advanced Strategies

+

This page explains some advanced concepts available for strategies. +If you're just getting started, please familiarize yourself with the Freqtrade basics and methods described in Strategy Customization first.

+

The call sequence of the methods described here is covered under bot execution logic. Those docs are also helpful in deciding which method is most suitable for your customisation needs.

+
+

Note

+

Callback methods should only be implemented if a strategy uses them.

+
+
+

Tip

+

Start off with a strategy template containing all available callback methods by running freqtrade new-strategy --strategy MyAwesomeStrategy --template advanced

+
+

Storing information

+

Storing information can be accomplished by creating a new dictionary within the strategy class.

+

The name of the variable can be chosen at will, but should be prefixed with custom_ to avoid naming collisions with predefined strategy variables.

+
class AwesomeStrategy(IStrategy):
+    # Create custom dictionary
+    custom_info = {}
+
+    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        # Check if the entry already exists
+        if not metadata["pair"] in self.custom_info:
+            # Create empty entry for this pair
+            self.custom_info[metadata["pair"]] = {}
+
+        if "crosstime" in self.custom_info[metadata["pair"]]:
+            self.custom_info[metadata["pair"]]["crosstime"] += 1
+        else:
+            self.custom_info[metadata["pair"]]["crosstime"] = 1
+
+
+

Warning

+

The data is not persisted after a bot-restart (or config-reload). Also, the amount of data should be kept smallish (no DataFrames and such), otherwise the bot will start to consume a lot of memory and eventually run out of memory and crash.

+
+
+

Note

+

If the data is pair-specific, make sure to use pair as one of the keys in the dictionary.

+
+

Dataframe access

+

You may access dataframe in various strategy functions by querying it from dataprovider.

+
from freqtrade.exchange import timeframe_to_prev_date
+
+class AwesomeStrategy(IStrategy):
+    def confirm_trade_exit(self, pair: str, trade: 'Trade', order_type: str, amount: float,
+                           rate: float, time_in_force: str, exit_reason: str,
+                           current_time: 'datetime', **kwargs) -> bool:
+        # Obtain pair dataframe.
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+
+        # Obtain last available candle. Do not use current_time to look up latest candle, because 
+        # current_time points to current incomplete candle whose data is not available.
+        last_candle = dataframe.iloc[-1].squeeze()
+        # <...>
+
+        # In dry/live runs trade open date will not match candle open date therefore it must be 
+        # rounded.
+        trade_date = timeframe_to_prev_date(self.timeframe, trade.open_date_utc)
+        # Look up trade candle.
+        trade_candle = dataframe.loc[dataframe['date'] == trade_date]
+        # trade_candle may be empty for trades that just opened as it is still incomplete.
+        if not trade_candle.empty:
+            trade_candle = trade_candle.squeeze()
+            # <...>
+
+
+

Using .iloc[-1]

+

You can use .iloc[-1] here because get_analyzed_dataframe() only returns candles that backtesting is allowed to see. +This will not work in populate_* methods, so make sure to not use .iloc[] in that area. +Also, this will only work starting with version 2021.5.

+
+
+

Enter Tag

+

When your strategy has multiple buy signals, you can name the signal that triggered. +Then you can access your buy signal on custom_exit

+
def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+    dataframe.loc[
+        (
+            (dataframe['rsi'] < 35) &
+            (dataframe['volume'] > 0)
+        ),
+        ['enter_long', 'enter_tag']] = (1, 'buy_signal_rsi')
+
+    return dataframe
+
+def custom_exit(self, pair: str, trade: Trade, current_time: datetime, current_rate: float,
+                current_profit: float, **kwargs):
+    dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+    last_candle = dataframe.iloc[-1].squeeze()
+    if trade.enter_tag == 'buy_signal_rsi' and last_candle['rsi'] > 80:
+        return 'sell_signal_rsi'
+    return None
+
+
+

Note

+

enter_tag is limited to 100 characters, remaining data will be truncated.

+
+
+

Warning

+

There is only one enter_tag column, which is used for both long and short trades. +As a consequence, this column must be treated as "last write wins" (it's just a dataframe column after all). +In fancy situations, where multiple signals collide (or if signals are deactivated again based on different conditions), this can lead to odd results with the wrong tag applied to an entry signal. +These results are a consequence of the strategy overwriting prior tags - where the last tag will "stick" and will be the one freqtrade will use.

+
+

Exit tag

+

Similar to Buy Tagging, you can also specify a sell tag.

+
def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+    dataframe.loc[
+        (
+            (dataframe['rsi'] > 70) &
+            (dataframe['volume'] > 0)
+        ),
+        ['exit_long', 'exit_tag']] = (1, 'exit_rsi')
+
+    return dataframe
+
+

The provided exit-tag is then used as sell-reason - and shown as such in backtest results.

+
+

Note

+

exit_reason is limited to 100 characters, remaining data will be truncated.

+
+

Strategy version

+

You can implement custom strategy versioning by using the "version" method, and returning the version you would like this strategy to have.

+
def version(self) -> str:
+    """
+    Returns version of the strategy.
+    """
+    return "1.1"
+
+
+

Note

+

You should make sure to implement proper version control (like a git repository) alongside this, as freqtrade will not keep historic versions of your strategy, so it's up to the user to be able to eventually roll back to a prior version of the strategy.

+
+

Derived strategies

+

The strategies can be derived from other strategies. This avoids duplication of your custom strategy code. You can use this technique to override small parts of your main strategy, leaving the rest untouched:

+
user_data/strategies/myawesomestrategy.py
class MyAwesomeStrategy(IStrategy):
+    ...
+    stoploss = 0.13
+    trailing_stop = False
+    # All other attributes and methods are here as they
+    # should be in any custom strategy...
+    ...
+
+
user_data/strategies/MyAwesomeStrategy2.py
from myawesomestrategy import MyAwesomeStrategy
+class MyAwesomeStrategy2(MyAwesomeStrategy):
+    # Override something
+    stoploss = 0.08
+    trailing_stop = True
+
+

Both attributes and methods may be overridden, altering behavior of the original strategy in a way you need.

+

While keeping the subclass in the same file is technically possible, it can lead to some problems with hyperopt parameter files, we therefore recommend to use separate strategy files, and import the parent strategy as shown above.

+

Embedding Strategies

+

Freqtrade provides you with an easy way to embed the strategy into your configuration file. +This is done by utilizing BASE64 encoding and providing this string at the strategy configuration field, +in your chosen config file.

+

Encoding a string as BASE64

+

This is a quick example, how to generate the BASE64 string in python

+
from base64 import urlsafe_b64encode
+
+with open(file, 'r') as f:
+    content = f.read()
+content = urlsafe_b64encode(content.encode('utf-8'))
+
+

The variable 'content', will contain the strategy file in a BASE64 encoded form. Which can now be set in your configurations file as following

+
"strategy": "NameOfStrategy:BASE64String"
+
+

Please ensure that 'NameOfStrategy' is identical to the strategy name!

+

Performance warning

+

When executing a strategy, one can sometimes be greeted by the following in the logs

+
+

PerformanceWarning: DataFrame is highly fragmented.

+
+

This is a warning from pandas and as the warning continues to say: +use pd.concat(axis=1). +This can have slight performance implications, which are usually only visible during hyperopt (when optimizing an indicator).

+

For example:

+
for val in self.buy_ema_short.range:
+    dataframe[f'ema_short_{val}'] = ta.EMA(dataframe, timeperiod=val)
+
+

should be rewritten to

+
frames = [dataframe]
+for val in self.buy_ema_short.range:
+    frames.append(DataFrame({
+        f'ema_short_{val}': ta.EMA(dataframe, timeperiod=val)
+    }))
+
+# Append columns to existing dataframe
+merged_frame = pd.concat(frames, axis=1)
+
+

Freqtrade does however also counter this by running dataframe.copy() on the dataframe right after the populate_indicators() method - so performance implications of this should be low to non-existant.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/strategy-callbacks/index.html b/en/2023.4/strategy-callbacks/index.html new file mode 100644 index 000000000..f3d10497f --- /dev/null +++ b/en/2023.4/strategy-callbacks/index.html @@ -0,0 +1,2455 @@ + + + + + + + + + + + + + + + + + + + + + + + + Strategy Callbacks - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Strategy Callbacks

+

While the main strategy functions (populate_indicators(), populate_entry_trend(), populate_exit_trend()) should be used in a vectorized way, and are only called once during backtesting, callbacks are called "whenever needed".

+

As such, you should avoid doing heavy calculations in callbacks to avoid delays during operations. +Depending on the callback used, they may be called when entering / exiting a trade, or throughout the duration of a trade.

+

Currently available callbacks:

+ +
+

Callback calling sequence

+

You can find the callback calling sequence in bot-basics

+
+

Bot start

+

A simple callback which is called once when the strategy is loaded. +This can be used to perform actions that must only be performed once and runs after dataprovider and wallet are set

+
import requests
+
+class AwesomeStrategy(IStrategy):
+
+    # ... populate_* methods
+
+    def bot_start(self, **kwargs) -> None:
+        """
+        Called only once after bot instantiation.
+        :param **kwargs: Ensure to keep this here so updates to this won't break your strategy.
+        """
+        if self.config['runmode'].value in ('live', 'dry_run'):
+            # Assign this to the class by using self.*
+            # can then be used by populate_* methods
+            self.custom_remote_data = requests.get('https://some_remote_source.example.com')
+
+

During hyperopt, this runs only once at startup.

+

Bot loop start

+

A simple callback which is called once at the start of every bot throttling iteration in dry/live mode (roughly every 5 +seconds, unless configured differently) or once per candle in backtest/hyperopt mode. +This can be used to perform calculations which are pair independent (apply to all pairs), loading of external data, etc.

+
import requests
+
+class AwesomeStrategy(IStrategy):
+
+    # ... populate_* methods
+
+    def bot_loop_start(self, current_time: datetime, **kwargs) -> None:
+        """
+        Called at the start of the bot iteration (one loop).
+        Might be used to perform pair-independent tasks
+        (e.g. gather some remote resource for comparison)
+        :param current_time: datetime object, containing the current datetime
+        :param **kwargs: Ensure to keep this here so updates to this won't break your strategy.
+        """
+        if self.config['runmode'].value in ('live', 'dry_run'):
+            # Assign this to the class by using self.*
+            # can then be used by populate_* methods
+            self.remote_data = requests.get('https://some_remote_source.example.com')
+
+

Stake size management

+

Called before entering a trade, makes it possible to manage your position size when placing a new trade.

+
class AwesomeStrategy(IStrategy):
+    def custom_stake_amount(self, pair: str, current_time: datetime, current_rate: float,
+                            proposed_stake: float, min_stake: Optional[float], max_stake: float,
+                            leverage: float, entry_tag: Optional[str], side: str,
+                            **kwargs) -> float:
+
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair=pair, timeframe=self.timeframe)
+        current_candle = dataframe.iloc[-1].squeeze()
+
+        if current_candle['fastk_rsi_1h'] > current_candle['fastd_rsi_1h']:
+            if self.config['stake_amount'] == 'unlimited':
+                # Use entire available wallet during favorable conditions when in compounding mode.
+                return max_stake
+            else:
+                # Compound profits during favorable conditions instead of using a static stake.
+                return self.wallets.get_total_stake_amount() / self.config['max_open_trades']
+
+        # Use default stake amount.
+        return proposed_stake
+
+

Freqtrade will fall back to the proposed_stake value should your code raise an exception. The exception itself will be logged.

+
+

Tip

+

You do not have to ensure that min_stake <= returned_value <= max_stake. Trades will succeed as the returned value will be clamped to supported range and this action will be logged.

+
+
+

Tip

+

Returning 0 or None will prevent trades from being placed.

+
+

Custom exit signal

+

Called for open trade every throttling iteration (roughly every 5 seconds) until a trade is closed.

+

Allows to define custom exit signals, indicating that specified position should be sold. This is very useful when we need to customize exit conditions for each individual trade, or if you need trade data to make an exit decision.

+

For example you could implement a 1:2 risk-reward ROI with custom_exit().

+

Using custom_exit() signals in place of stoploss though is not recommended. It is a inferior method to using custom_stoploss() in this regard - which also allows you to keep the stoploss on exchange.

+
+

Note

+

Returning a (none-empty) string or True from this method is equal to setting exit signal on a candle at specified time. This method is not called when exit signal is set already, or if exit signals are disabled (use_exit_signal=False). string max length is 64 characters. Exceeding this limit will cause the message to be truncated to 64 characters. +custom_exit() will ignore exit_profit_only, and will always be called unless use_exit_signal=False, even if there is a new enter signal.

+
+

An example of how we can use different indicators depending on the current profit and also exit trades that were open longer than one day:

+
class AwesomeStrategy(IStrategy):
+    def custom_exit(self, pair: str, trade: 'Trade', current_time: 'datetime', current_rate: float,
+                    current_profit: float, **kwargs):
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        last_candle = dataframe.iloc[-1].squeeze()
+
+        # Above 20% profit, sell when rsi < 80
+        if current_profit > 0.2:
+            if last_candle['rsi'] < 80:
+                return 'rsi_below_80'
+
+        # Between 2% and 10%, sell if EMA-long above EMA-short
+        if 0.02 < current_profit < 0.1:
+            if last_candle['emalong'] > last_candle['emashort']:
+                return 'ema_long_below_80'
+
+        # Sell any positions at a loss if they are held for more than one day.
+        if current_profit < 0.0 and (current_time - trade.open_date_utc).days >= 1:
+            return 'unclog'
+
+

See Dataframe access for more information about dataframe use in strategy callbacks.

+

Custom stoploss

+

Called for open trade every iteration (roughly every 5 seconds) until a trade is closed.

+

The usage of the custom stoploss method must be enabled by setting use_custom_stoploss=True on the strategy object.

+

The stoploss price can only ever move upwards - if the stoploss value returned from custom_stoploss would result in a lower stoploss price than was previously set, it will be ignored. The traditional stoploss value serves as an absolute lower level and will be instated as the initial stoploss (before this method is called for the first time for a trade), and is still mandatory.

+

The method must return a stoploss value (float / number) as a percentage of the current price. +E.g. If the current_rate is 200 USD, then returning 0.02 will set the stoploss price 2% lower, at 196 USD. +During backtesting, current_rate (and current_profit) are provided against the candle's high (or low for short trades) - while the resulting stoploss is evaluated against the candle's low (or high for short trades).

+

The absolute value of the return value is used (the sign is ignored), so returning 0.05 or -0.05 have the same result, a stoploss 5% below the current price.

+

To simulate a regular trailing stoploss of 4% (trailing 4% behind the maximum reached price) you would use the following very simple method:

+
# additional imports required
+from datetime import datetime
+from freqtrade.persistence import Trade
+
+class AwesomeStrategy(IStrategy):
+
+    # ... populate_* methods
+
+    use_custom_stoploss = True
+
+    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,
+                        current_rate: float, current_profit: float, **kwargs) -> float:
+        """
+        Custom stoploss logic, returning the new distance relative to current_rate (as ratio).
+        e.g. returning -0.05 would create a stoploss 5% below current_rate.
+        The custom stoploss can never be below self.stoploss, which serves as a hard maximum loss.
+
+        For full documentation please go to https://www.freqtrade.io/en/latest/strategy-advanced/
+
+        When not implemented by a strategy, returns the initial stoploss value
+        Only called when use_custom_stoploss is set to True.
+
+        :param pair: Pair that's currently analyzed
+        :param trade: trade object.
+        :param current_time: datetime object, containing the current datetime
+        :param current_rate: Rate, calculated based on pricing settings in exit_pricing.
+        :param current_profit: Current profit (as ratio), calculated based on current_rate.
+        :param **kwargs: Ensure to keep this here so updates to this won't break your strategy.
+        :return float: New stoploss value, relative to the current rate
+        """
+        return -0.04
+
+

Stoploss on exchange works similar to trailing_stop, and the stoploss on exchange is updated as configured in stoploss_on_exchange_interval (More details about stoploss on exchange).

+
+

Use of dates

+

All time-based calculations should be done based on current_time - using datetime.now() or datetime.utcnow() is discouraged, as this will break backtesting support.

+
+
+

Trailing stoploss

+

It's recommended to disable trailing_stop when using custom stoploss values. Both can work in tandem, but you might encounter the trailing stop to move the price higher while your custom function would not want this, causing conflicting behavior.

+
+

Custom stoploss examples

+

The next section will show some examples on what's possible with the custom stoploss function. +Of course, many more things are possible, and all examples can be combined at will.

+

Time based trailing stop

+

Use the initial stoploss for the first 60 minutes, after this change to 10% trailing stoploss, and after 2 hours (120 minutes) we use a 5% trailing stoploss.

+
from datetime import datetime, timedelta
+from freqtrade.persistence import Trade
+
+class AwesomeStrategy(IStrategy):
+
+    # ... populate_* methods
+
+    use_custom_stoploss = True
+
+    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,
+                        current_rate: float, current_profit: float, **kwargs) -> float:
+
+        # Make sure you have the longest interval first - these conditions are evaluated from top to bottom.
+        if current_time - timedelta(minutes=120) > trade.open_date_utc:
+            return -0.05
+        elif current_time - timedelta(minutes=60) > trade.open_date_utc:
+            return -0.10
+        return 1
+
+

Different stoploss per pair

+

Use a different stoploss depending on the pair. +In this example, we'll trail the highest price with 10% trailing stoploss for ETH/BTC and XRP/BTC, with 5% trailing stoploss for LTC/BTC and with 15% for all other pairs.

+
from datetime import datetime
+from freqtrade.persistence import Trade
+
+class AwesomeStrategy(IStrategy):
+
+    # ... populate_* methods
+
+    use_custom_stoploss = True
+
+    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,
+                        current_rate: float, current_profit: float, **kwargs) -> float:
+
+        if pair in ('ETH/BTC', 'XRP/BTC'):
+            return -0.10
+        elif pair in ('LTC/BTC'):
+            return -0.05
+        return -0.15
+
+

Trailing stoploss with positive offset

+

Use the initial stoploss until the profit is above 4%, then use a trailing stoploss of 50% of the current profit with a minimum of 2.5% and a maximum of 5%.

+

Please note that the stoploss can only increase, values lower than the current stoploss are ignored.

+
from datetime import datetime, timedelta
+from freqtrade.persistence import Trade
+
+class AwesomeStrategy(IStrategy):
+
+    # ... populate_* methods
+
+    use_custom_stoploss = True
+
+    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,
+                        current_rate: float, current_profit: float, **kwargs) -> float:
+
+        if current_profit < 0.04:
+            return -1 # return a value bigger than the initial stoploss to keep using the initial stoploss
+
+        # After reaching the desired offset, allow the stoploss to trail by half the profit
+        desired_stoploss = current_profit / 2
+
+        # Use a minimum of 2.5% and a maximum of 5%
+        return max(min(desired_stoploss, 0.05), 0.025)
+
+

Stepped stoploss

+

Instead of continuously trailing behind the current price, this example sets fixed stoploss price levels based on the current profit.

+
    +
  • Use the regular stoploss until 20% profit is reached
  • +
  • Once profit is > 20% - set stoploss to 7% above open price.
  • +
  • Once profit is > 25% - set stoploss to 15% above open price.
  • +
  • Once profit is > 40% - set stoploss to 25% above open price.
  • +
+
from datetime import datetime
+from freqtrade.persistence import Trade
+from freqtrade.strategy import stoploss_from_open
+
+class AwesomeStrategy(IStrategy):
+
+    # ... populate_* methods
+
+    use_custom_stoploss = True
+
+    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,
+                        current_rate: float, current_profit: float, **kwargs) -> float:
+
+        # evaluate highest to lowest, so that highest possible stop is used
+        if current_profit > 0.40:
+            return stoploss_from_open(0.25, current_profit, is_short=trade.is_short, leverage=trade.leverage)
+        elif current_profit > 0.25:
+            return stoploss_from_open(0.15, current_profit, is_short=trade.is_short, leverage=trade.leverage)
+        elif current_profit > 0.20:
+            return stoploss_from_open(0.07, current_profit, is_short=trade.is_short, leverage=trade.leverage)
+
+        # return maximum stoploss value, keeping current stoploss price unchanged
+        return 1
+
+

Custom stoploss using an indicator from dataframe example

+

Absolute stoploss value may be derived from indicators stored in dataframe. Example uses parabolic SAR below the price as stoploss.

+
class AwesomeStrategy(IStrategy):
+
+    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        # <...>
+        dataframe['sar'] = ta.SAR(dataframe)
+
+    use_custom_stoploss = True
+
+    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,
+                        current_rate: float, current_profit: float, **kwargs) -> float:
+
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        last_candle = dataframe.iloc[-1].squeeze()
+
+        # Use parabolic sar as absolute stoploss price
+        stoploss_price = last_candle['sar']
+
+        # Convert absolute price to percentage relative to current_rate
+        if stoploss_price < current_rate:
+            return stoploss_from_absolute(stoploss_price, current_rate, is_short=trade.is_short)
+
+        # return maximum stoploss value, keeping current stoploss price unchanged
+        return 1
+
+

See Dataframe access for more information about dataframe use in strategy callbacks.

+

Common helpers for stoploss calculations

+

Stoploss relative to open price

+

Stoploss values returned from custom_stoploss() always specify a percentage relative to current_rate. In order to set a stoploss relative to the open price, we need to use current_profit to calculate what percentage relative to the current_rate will give you the same result as if the percentage was specified from the open price.

+

The helper function stoploss_from_open() can be used to convert from an open price relative stop, to a current price relative stop which can be returned from custom_stoploss().

+

Stoploss percentage from absolute price

+

Stoploss values returned from custom_stoploss() always specify a percentage relative to current_rate. In order to set a stoploss at specified absolute price level, we need to use stop_rate to calculate what percentage relative to the current_rate will give you the same result as if the percentage was specified from the open price.

+

The helper function stoploss_from_absolute() can be used to convert from an absolute price, to a current price relative stop which can be returned from custom_stoploss().

+
+

Custom order price rules

+

By default, freqtrade use the orderbook to automatically set an order price(Relevant documentation), you also have the option to create custom order prices based on your strategy.

+

You can use this feature by creating a custom_entry_price() function in your strategy file to customize entry prices and custom_exit_price() for exits.

+

Each of these methods are called right before placing an order on the exchange.

+
+

Note

+

If your custom pricing function return None or an invalid value, price will fall back to proposed_rate, which is based on the regular pricing configuration.

+
+

Custom order entry and exit price example

+
from datetime import datetime, timedelta, timezone
+from freqtrade.persistence import Trade
+
+class AwesomeStrategy(IStrategy):
+
+    # ... populate_* methods
+
+    def custom_entry_price(self, pair: str, current_time: datetime, proposed_rate: float,
+                           entry_tag: Optional[str], side: str, **kwargs) -> float:
+
+        dataframe, last_updated = self.dp.get_analyzed_dataframe(pair=pair,
+                                                                timeframe=self.timeframe)
+        new_entryprice = dataframe['bollinger_10_lowerband'].iat[-1]
+
+        return new_entryprice
+
+    def custom_exit_price(self, pair: str, trade: Trade,
+                          current_time: datetime, proposed_rate: float,
+                          current_profit: float, exit_tag: Optional[str], **kwargs) -> float:
+
+        dataframe, last_updated = self.dp.get_analyzed_dataframe(pair=pair,
+                                                                timeframe=self.timeframe)
+        new_exitprice = dataframe['bollinger_10_upperband'].iat[-1]
+
+        return new_exitprice
+
+
+

Warning

+

Modifying entry and exit prices will only work for limit orders. Depending on the price chosen, this can result in a lot of unfilled orders. By default the maximum allowed distance between the current price and the custom price is 2%, this value can be changed in config with the custom_price_max_distance_ratio parameter. +Example: +If the new_entryprice is 97, the proposed_rate is 100 and the custom_price_max_distance_ratio is set to 2%, The retained valid custom entry price will be 98, which is 2% below the current (proposed) rate.

+
+
+

Backtesting

+

Custom prices are supported in backtesting (starting with 2021.12), and orders will fill if the price falls within the candle's low/high range. +Orders that don't fill immediately are subject to regular timeout handling, which happens once per (detail) candle. +custom_exit_price() is only called for sells of type exit_signal, Custom exit and partial exits. All other exit-types will use regular backtesting prices.

+
+

Custom order timeout rules

+

Simple, time-based order-timeouts can be configured either via strategy or in the configuration in the unfilledtimeout section.

+

However, freqtrade also offers a custom callback for both order types, which allows you to decide based on custom criteria if an order did time out or not.

+
+

Note

+

Backtesting fills orders if their price falls within the candle's low/high range. +The below callbacks will be called once per (detail) candle for orders that don't fill immediately (which use custom pricing).

+
+

Custom order timeout example

+

Called for every open order until that order is either filled or cancelled. +check_entry_timeout() is called for trade entries, while check_exit_timeout() is called for trade exit orders.

+

A simple example, which applies different unfilled-timeouts depending on the price of the asset can be seen below. +It applies a tight timeout for higher priced assets, while allowing more time to fill on cheap coins.

+

The function must return either True (cancel order) or False (keep order alive).

+
from datetime import datetime, timedelta
+from freqtrade.persistence import Trade, Order
+
+class AwesomeStrategy(IStrategy):
+
+    # ... populate_* methods
+
+    # Set unfilledtimeout to 25 hours, since the maximum timeout from below is 24 hours.
+    unfilledtimeout = {
+        'entry': 60 * 25,
+        'exit': 60 * 25
+    }
+
+    def check_entry_timeout(self, pair: str, trade: 'Trade', order: 'Order',
+                            current_time: datetime, **kwargs) -> bool:
+        if trade.open_rate > 100 and trade.open_date_utc < current_time - timedelta(minutes=5):
+            return True
+        elif trade.open_rate > 10 and trade.open_date_utc < current_time - timedelta(minutes=3):
+            return True
+        elif trade.open_rate < 1 and trade.open_date_utc < current_time - timedelta(hours=24):
+           return True
+        return False
+
+
+    def check_exit_timeout(self, pair: str, trade: Trade, order: 'Order',
+                           current_time: datetime, **kwargs) -> bool:
+        if trade.open_rate > 100 and trade.open_date_utc < current_time - timedelta(minutes=5):
+            return True
+        elif trade.open_rate > 10 and trade.open_date_utc < current_time - timedelta(minutes=3):
+            return True
+        elif trade.open_rate < 1 and trade.open_date_utc < current_time - timedelta(hours=24):
+           return True
+        return False
+
+
+

Note

+

For the above example, unfilledtimeout must be set to something bigger than 24h, otherwise that type of timeout will apply first.

+
+

Custom order timeout example (using additional data)

+
from datetime import datetime
+from freqtrade.persistence import Trade, Order
+
+class AwesomeStrategy(IStrategy):
+
+    # ... populate_* methods
+
+    # Set unfilledtimeout to 25 hours, since the maximum timeout from below is 24 hours.
+    unfilledtimeout = {
+        'entry': 60 * 25,
+        'exit': 60 * 25
+    }
+
+    def check_entry_timeout(self, pair: str, trade: 'Trade', order: 'Order',
+                            current_time: datetime, **kwargs) -> bool:
+        ob = self.dp.orderbook(pair, 1)
+        current_price = ob['bids'][0][0]
+        # Cancel buy order if price is more than 2% above the order.
+        if current_price > order.price * 1.02:
+            return True
+        return False
+
+
+    def check_exit_timeout(self, pair: str, trade: 'Trade', order: 'Order',
+                           current_time: datetime, **kwargs) -> bool:
+        ob = self.dp.orderbook(pair, 1)
+        current_price = ob['asks'][0][0]
+        # Cancel sell order if price is more than 2% below the order.
+        if current_price < order.price * 0.98:
+            return True
+        return False
+
+
+

Bot order confirmation

+

Confirm trade entry / exits. +This are the last methods that will be called before an order is placed.

+

Trade entry (buy order) confirmation

+

confirm_trade_entry() can be used to abort a trade entry at the latest second (maybe because the price is not what we expect).

+
class AwesomeStrategy(IStrategy):
+
+    # ... populate_* methods
+
+    def confirm_trade_entry(self, pair: str, order_type: str, amount: float, rate: float,
+                            time_in_force: str, current_time: datetime, entry_tag: Optional[str],
+                            side: str, **kwargs) -> bool:
+        """
+        Called right before placing a entry order.
+        Timing for this function is critical, so avoid doing heavy computations or
+        network requests in this method.
+
+        For full documentation please go to https://www.freqtrade.io/en/latest/strategy-advanced/
+
+        When not implemented by a strategy, returns True (always confirming).
+
+        :param pair: Pair that's about to be bought/shorted.
+        :param order_type: Order type (as configured in order_types). usually limit or market.
+        :param amount: Amount in target (base) currency that's going to be traded.
+        :param rate: Rate that's going to be used when using limit orders 
+                     or current rate for market orders.
+        :param time_in_force: Time in force. Defaults to GTC (Good-til-cancelled).
+        :param current_time: datetime object, containing the current datetime
+        :param entry_tag: Optional entry_tag (buy_tag) if provided with the buy signal.
+        :param side: 'long' or 'short' - indicating the direction of the proposed trade
+        :param **kwargs: Ensure to keep this here so updates to this won't break your strategy.
+        :return bool: When True is returned, then the buy-order is placed on the exchange.
+            False aborts the process
+        """
+        return True
+
+

Trade exit (sell order) confirmation

+

confirm_trade_exit() can be used to abort a trade exit (sell) at the latest second (maybe because the price is not what we expect).

+

confirm_trade_exit() may be called multiple times within one iteration for the same trade if different exit-reasons apply. +The exit-reasons (if applicable) will be in the following sequence:

+
    +
  • exit_signal / custom_exit
  • +
  • stop_loss
  • +
  • roi
  • +
  • trailing_stop_loss
  • +
+
from freqtrade.persistence import Trade
+
+
+class AwesomeStrategy(IStrategy):
+
+    # ... populate_* methods
+
+    def confirm_trade_exit(self, pair: str, trade: Trade, order_type: str, amount: float,
+                           rate: float, time_in_force: str, exit_reason: str,
+                           current_time: datetime, **kwargs) -> bool:
+        """
+        Called right before placing a regular exit order.
+        Timing for this function is critical, so avoid doing heavy computations or
+        network requests in this method.
+
+        For full documentation please go to https://www.freqtrade.io/en/latest/strategy-advanced/
+
+        When not implemented by a strategy, returns True (always confirming).
+
+        :param pair: Pair for trade that's about to be exited.
+        :param trade: trade object.
+        :param order_type: Order type (as configured in order_types). usually limit or market.
+        :param amount: Amount in base currency.
+        :param rate: Rate that's going to be used when using limit orders
+                     or current rate for market orders.
+        :param time_in_force: Time in force. Defaults to GTC (Good-til-cancelled).
+        :param exit_reason: Exit reason.
+            Can be any of ['roi', 'stop_loss', 'stoploss_on_exchange', 'trailing_stop_loss',
+                           'exit_signal', 'force_exit', 'emergency_exit']
+        :param current_time: datetime object, containing the current datetime
+        :param **kwargs: Ensure to keep this here so updates to this won't break your strategy.
+        :return bool: When True, then the exit-order is placed on the exchange.
+            False aborts the process
+        """
+        if exit_reason == 'force_exit' and trade.calc_profit_ratio(rate) < 0:
+            # Reject force-sells with negative profit
+            # This is just a sample, please adjust to your needs
+            # (this does not necessarily make sense, assuming you know when you're force-selling)
+            return False
+        return True
+
+
+

Warning

+

confirm_trade_exit() can prevent stoploss exits, causing significant losses as this would ignore stoploss exits. +confirm_trade_exit() will not be called for Liquidations - as liquidations are forced by the exchange, and therefore cannot be rejected.

+
+

Adjust trade position

+

The position_adjustment_enable strategy property enables the usage of adjust_trade_position() callback in the strategy. +For performance reasons, it's disabled by default and freqtrade will show a warning message on startup if enabled. +adjust_trade_position() can be used to perform additional orders, for example to manage risk with DCA (Dollar Cost Averaging) or to increase or decrease positions.

+

max_entry_position_adjustment property is used to limit the number of additional buys per trade (on top of the first buy) that the bot can execute. By default, the value is -1 which means the bot have no limit on number of adjustment buys.

+

The strategy is expected to return a stake_amount (in stake currency) between min_stake and max_stake if and when an additional buy order should be made (position is increased). +If there are not enough funds in the wallet (the return value is above max_stake) then the signal will be ignored. +Additional orders also result in additional fees and those orders don't count towards max_open_trades.

+

This callback is not called when there is an open order (either buy or sell) waiting for execution.

+

adjust_trade_position() is called very frequently for the duration of a trade, so you must keep your implementation as performant as possible.

+

Additional Buys are ignored once you have reached the maximum amount of extra buys that you have set on max_entry_position_adjustment, but the callback is called anyway looking for partial exits.

+

Position adjustments will always be applied in the direction of the trade, so a positive value will always increase your position (negative values will decrease your position), no matter if it's a long or short trade. Modifications to leverage are not possible, and the stake-amount is assumed to be before applying leverage.

+
+

About stake size

+

Using fixed stake size means it will be the amount used for the first order, just like without position adjustment. +If you wish to buy additional orders with DCA, then make sure to leave enough funds in the wallet for that. +Using 'unlimited' stake amount with DCA orders requires you to also implement the custom_stake_amount() callback to avoid allocating all funds to the initial order.

+
+
+

Warning

+

Stoploss is still calculated from the initial opening price, not averaged price. +Regular stoploss rules still apply (cannot move down).

+

While /stopentry command stops the bot from entering new trades, the position adjustment feature will continue buying new orders on existing trades.

+
+
+

Backtesting

+

During backtesting this callback is called for each candle in timeframe or timeframe_detail, so run-time performance will be affected. +This can also cause deviating results between live and backtesting, since backtesting can adjust the trade only once per candle, whereas live could adjust the trade multiple times per candle.

+
+
from freqtrade.persistence import Trade
+
+
+class DigDeeperStrategy(IStrategy):
+
+    position_adjustment_enable = True
+
+    # Attempts to handle large drops with DCA. High stoploss is required.
+    stoploss = -0.30
+
+    # ... populate_* methods
+
+    # Example specific variables
+    max_entry_position_adjustment = 3
+    # This number is explained a bit further down
+    max_dca_multiplier = 5.5
+
+    # This is called when placing the initial order (opening trade)
+    def custom_stake_amount(self, pair: str, current_time: datetime, current_rate: float,
+                            proposed_stake: float, min_stake: Optional[float], max_stake: float,
+                            leverage: float, entry_tag: Optional[str], side: str,
+                            **kwargs) -> float:
+
+        # We need to leave most of the funds for possible further DCA orders
+        # This also applies to fixed stakes
+        return proposed_stake / self.max_dca_multiplier
+
+    def adjust_trade_position(self, trade: Trade, current_time: datetime,
+                              current_rate: float, current_profit: float,
+                              min_stake: Optional[float], max_stake: float,
+                              current_entry_rate: float, current_exit_rate: float,
+                              current_entry_profit: float, current_exit_profit: float,
+                              **kwargs) -> Optional[float]:
+        """
+        Custom trade adjustment logic, returning the stake amount that a trade should be
+        increased or decreased.
+        This means extra buy or sell orders with additional fees.
+        Only called when `position_adjustment_enable` is set to True.
+
+        For full documentation please go to https://www.freqtrade.io/en/latest/strategy-advanced/
+
+        When not implemented by a strategy, returns None
+
+        :param trade: trade object.
+        :param current_time: datetime object, containing the current datetime
+        :param current_rate: Current buy rate.
+        :param current_profit: Current profit (as ratio), calculated based on current_rate.
+        :param min_stake: Minimal stake size allowed by exchange (for both entries and exits)
+        :param max_stake: Maximum stake allowed (either through balance, or by exchange limits).
+        :param current_entry_rate: Current rate using entry pricing.
+        :param current_exit_rate: Current rate using exit pricing.
+        :param current_entry_profit: Current profit using entry pricing.
+        :param current_exit_profit: Current profit using exit pricing.
+        :param **kwargs: Ensure to keep this here so updates to this won't break your strategy.
+        :return float: Stake amount to adjust your trade,
+                       Positive values to increase position, Negative values to decrease position.
+                       Return None for no action.
+        """
+
+        if current_profit > 0.05 and trade.nr_of_successful_exits == 0:
+            # Take half of the profit at +5%
+            return -(trade.stake_amount / 2)
+
+        if current_profit > -0.05:
+            return None
+
+        # Obtain pair dataframe (just to show how to access it)
+        dataframe, _ = self.dp.get_analyzed_dataframe(trade.pair, self.timeframe)
+        # Only buy when not actively falling price.
+        last_candle = dataframe.iloc[-1].squeeze()
+        previous_candle = dataframe.iloc[-2].squeeze()
+        if last_candle['close'] < previous_candle['close']:
+            return None
+
+        filled_entries = trade.select_filled_orders(trade.entry_side)
+        count_of_entries = trade.nr_of_successful_entries
+        # Allow up to 3 additional increasingly larger buys (4 in total)
+        # Initial buy is 1x
+        # If that falls to -5% profit, we buy 1.25x more, average profit should increase to roughly -2.2%
+        # If that falls down to -5% again, we buy 1.5x more
+        # If that falls once again down to -5%, we buy 1.75x more
+        # Total stake for this trade would be 1 + 1.25 + 1.5 + 1.75 = 5.5x of the initial allowed stake.
+        # That is why max_dca_multiplier is 5.5
+        # Hope you have a deep wallet!
+        try:
+            # This returns first order stake size
+            stake_amount = filled_entries[0].cost
+            # This then calculates current safety order size
+            stake_amount = stake_amount * (1 + (count_of_entries * 0.25))
+            return stake_amount
+        except Exception as exception:
+            return None
+
+        return None
+
+

Position adjust calculations

+
    +
  • Entry rates are calculated using weighted averages.
  • +
  • Exits will not influence the average entry rate.
  • +
  • Partial exit relative profit is relative to the average entry price at this point.
  • +
  • Final exit relative profit is calculated based on the total invested capital. (See example below)
  • +
+
+Calculation example +

This example assumes 0 fees for simplicity, and a long position on an imaginary coin.

+
    +
  • Buy 100@8$
  • +
  • Buy 100@9$ -> Avg price: 8.5$
  • +
  • Sell 100@10$ -> Avg price: 8.5$, realized profit 150$, 17.65%
  • +
  • Buy 150@11$ -> Avg price: 10$, realized profit 150$, 17.65%
  • +
  • Sell 100@12$ -> Avg price: 10$, total realized profit 350$, 20%
  • +
  • Sell 150@14$ -> Avg price: 10$, total realized profit 950$, 40% <- This will be the last "Exit" message
  • +
+

The total profit for this trade was 950$ on a 3350$ investment (100@8$ + 100@9$ + 150@11$). As such - the final relative profit is 28.35% (950 / 3350).

+
+

Adjust Entry Price

+

The adjust_entry_price() callback may be used by strategy developer to refresh/replace limit orders upon arrival of new candles. +Be aware that custom_entry_price() is still the one dictating initial entry limit order price target at the time of entry trigger.

+

Orders can be cancelled out of this callback by returning None.

+

Returning current_order_rate will keep the order on the exchange "as is". +Returning any other price will cancel the existing order, and replace it with a new order.

+

The trade open-date (trade.open_date_utc) will remain at the time of the very first order placed. +Please make sure to be aware of this - and eventually adjust your logic in other callbacks to account for this, and use the date of the first filled order instead.

+
+

Regular timeout

+

Entry unfilledtimeout mechanism (as well as check_entry_timeout()) takes precedence over this. +Entry Orders that are cancelled via the above methods will not have this callback called. Be sure to update timeout values to match your expectations.

+
+
from freqtrade.persistence import Trade
+from datetime import timedelta
+
+class AwesomeStrategy(IStrategy):
+
+    # ... populate_* methods
+
+    def adjust_entry_price(self, trade: Trade, order: Optional[Order], pair: str,
+                           current_time: datetime, proposed_rate: float, current_order_rate: float,
+                           entry_tag: Optional[str], side: str, **kwargs) -> float:
+        """
+        Entry price re-adjustment logic, returning the user desired limit price.
+        This only executes when a order was already placed, still open (unfilled fully or partially)
+        and not timed out on subsequent candles after entry trigger.
+
+        When not implemented by a strategy, returns current_order_rate as default.
+        If current_order_rate is returned then the existing order is maintained.
+        If None is returned then order gets canceled but not replaced by a new one.
+
+        :param pair: Pair that's currently analyzed
+        :param trade: Trade object.
+        :param order: Order object
+        :param current_time: datetime object, containing the current datetime
+        :param proposed_rate: Rate, calculated based on pricing settings in entry_pricing.
+        :param current_order_rate: Rate of the existing order in place.
+        :param entry_tag: Optional entry_tag (buy_tag) if provided with the buy signal.
+        :param side: 'long' or 'short' - indicating the direction of the proposed trade
+        :param **kwargs: Ensure to keep this here so updates to this won't break your strategy.
+        :return float: New entry price value if provided
+
+        """
+        # Limit orders to use and follow SMA200 as price target for the first 10 minutes since entry trigger for BTC/USDT pair.
+        if pair == 'BTC/USDT' and entry_tag == 'long_sma200' and side == 'long' and (current_time - timedelta(minutes=10)) > trade.open_date_utc:
+            # just cancel the order if it has been filled more than half of the amount
+            if order.filled > order.remaining:
+                return None
+            else:
+                dataframe, _ = self.dp.get_analyzed_dataframe(pair=pair, timeframe=self.timeframe)
+                current_candle = dataframe.iloc[-1].squeeze()
+                # desired price
+                return current_candle['sma_200']
+        # default: maintain existing order
+        return current_order_rate
+
+

Leverage Callback

+

When trading in markets that allow leverage, this method must return the desired Leverage (Defaults to 1 -> No leverage).

+

Assuming a capital of 500USDT, a trade with leverage=3 would result in a position with 500 x 3 = 1500 USDT.

+

Values that are above max_leverage will be adjusted to max_leverage. +For markets / exchanges that don't support leverage, this method is ignored.

+
class AwesomeStrategy(IStrategy):
+    def leverage(self, pair: str, current_time: datetime, current_rate: float,
+                 proposed_leverage: float, max_leverage: float, entry_tag: Optional[str], side: str,
+                 **kwargs) -> float:
+        """
+        Customize leverage for each new trade. This method is only called in futures mode.
+
+        :param pair: Pair that's currently analyzed
+        :param current_time: datetime object, containing the current datetime
+        :param current_rate: Rate, calculated based on pricing settings in exit_pricing.
+        :param proposed_leverage: A leverage proposed by the bot.
+        :param max_leverage: Max leverage allowed on this pair
+        :param entry_tag: Optional entry_tag (buy_tag) if provided with the buy signal.
+        :param side: 'long' or 'short' - indicating the direction of the proposed trade
+        :return: A leverage amount, which is between 1.0 and max_leverage.
+        """
+        return 1.0
+
+

All profit calculations include leverage. Stoploss / ROI also include leverage in their calculation. +Defining a stoploss of 10% at 10x leverage would trigger the stoploss with a 1% move to the downside.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/strategy-customization/index.html b/en/2023.4/strategy-customization/index.html new file mode 100644 index 000000000..8995dc686 --- /dev/null +++ b/en/2023.4/strategy-customization/index.html @@ -0,0 +1,2904 @@ + + + + + + + + + + + + + + + + + + + + + + + + Strategy Customization - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Strategy Customization

+

This page explains how to customize your strategies, add new indicators and set up trading rules.

+

Please familiarize yourself with Freqtrade basics first, which provides overall info on how the bot operates.

+

Develop your own strategy

+

The bot includes a default strategy file. +Also, several other strategies are available in the strategy repository.

+

You will however most likely have your own idea for a strategy. +This document intends to help you convert your strategy idea into your own strategy.

+

To get started, use freqtrade new-strategy --strategy AwesomeStrategy (you can obviously use your own naming for your strategy). +This will create a new strategy file from a template, which will be located under user_data/strategies/AwesomeStrategy.py.

+
+

Note

+

This is just a template file, which will most likely not be profitable out of the box.

+
+
+Different template levels +

freqtrade new-strategy has an additional parameter, --template, which controls the amount of pre-build information you get in the created strategy. Use --template minimal to get an empty strategy without any indicator examples, or --template advanced to get a template with most callbacks defined.

+
+

Anatomy of a strategy

+

A strategy file contains all the information needed to build a good strategy:

+
    +
  • Indicators
  • +
  • Entry strategy rules
  • +
  • Exit strategy rules
  • +
  • Minimal ROI recommended
  • +
  • Stoploss strongly recommended
  • +
+

The bot also include a sample strategy called SampleStrategy you can update: user_data/strategies/sample_strategy.py. +You can test it with the parameter: --strategy SampleStrategy

+

Additionally, there is an attribute called INTERFACE_VERSION, which defines the version of the strategy interface the bot should use. +The current version is 3 - which is also the default when it's not set explicitly in the strategy.

+

Future versions will require this to be set.

+
freqtrade trade --strategy AwesomeStrategy
+
+

For the following section we will use the user_data/strategies/sample_strategy.py +file as reference.

+
+

Strategies and Backtesting

+

To avoid problems and unexpected differences between Backtesting and dry/live modes, please be aware +that during backtesting the full time range is passed to the populate_*() methods at once. +It is therefore best to use vectorized operations (across the whole dataframe, not loops) and +avoid index referencing (df.iloc[-1]), but instead use df.shift() to get to the previous candle.

+
+
+

Warning: Using future data

+

Since backtesting passes the full time range to the populate_*() methods, the strategy author +needs to take care to avoid having the strategy utilize data from the future. +Some common patterns for this are listed in the Common Mistakes section of this document.

+
+

Dataframe

+

Freqtrade uses pandas to store/provide the candlestick (OHLCV) data. +Pandas is a great library developed for processing large amounts of data.

+

Each row in a dataframe corresponds to one candle on a chart, with the latest candle always being the last in the dataframe (sorted by date).

+
> dataframe.head()
+                       date      open      high       low     close     volume
+0 2021-11-09 23:25:00+00:00  67279.67  67321.84  67255.01  67300.97   44.62253
+1 2021-11-09 23:30:00+00:00  67300.97  67301.34  67183.03  67187.01   61.38076
+2 2021-11-09 23:35:00+00:00  67187.02  67187.02  67031.93  67123.81  113.42728
+3 2021-11-09 23:40:00+00:00  67123.80  67222.40  67080.33  67160.48   78.96008
+4 2021-11-09 23:45:00+00:00  67160.48  67160.48  66901.26  66943.37  111.39292
+
+

Pandas provides fast ways to calculate metrics. To benefit from this speed, it's advised to not use loops, but use vectorized methods instead.

+

Vectorized operations perform calculations across the whole range of data and are therefore, compared to looping through each row, a lot faster when calculating indicators.

+

As a dataframe is a table, simple python comparisons like the following will not work

+
    if dataframe['rsi'] > 30:
+        dataframe['enter_long'] = 1
+
+

The above section will fail with The truth value of a Series is ambiguous. [...].

+

This must instead be written in a pandas-compatible way, so the operation is performed across the whole dataframe.

+
    dataframe.loc[
+        (dataframe['rsi'] > 30)
+    , 'enter_long'] = 1
+
+

With this section, you have a new column in your dataframe, which has 1 assigned whenever RSI is above 30.

+

Customize Indicators

+

Buy and sell signals need indicators. You can add more indicators by extending the list contained in the method populate_indicators() from your strategy file.

+

You should only add the indicators used in either populate_entry_trend(), populate_exit_trend(), or to populate another indicator, otherwise performance may suffer.

+

It's important to always return the dataframe without removing/modifying the columns "open", "high", "low", "close", "volume", otherwise these fields would contain something unexpected.

+

Sample:

+
def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+    """
+    Adds several different TA indicators to the given DataFrame
+
+    Performance Note: For the best performance be frugal on the number of indicators
+    you are using. Let uncomment only the indicator you are using in your strategies
+    or your hyperopt configuration, otherwise you will waste your memory and CPU usage.
+    :param dataframe: Dataframe with data from the exchange
+    :param metadata: Additional information, like the currently traded pair
+    :return: a Dataframe with all mandatory indicators for the strategies
+    """
+    dataframe['sar'] = ta.SAR(dataframe)
+    dataframe['adx'] = ta.ADX(dataframe)
+    stoch = ta.STOCHF(dataframe)
+    dataframe['fastd'] = stoch['fastd']
+    dataframe['fastk'] = stoch['fastk']
+    dataframe['blower'] = ta.BBANDS(dataframe, nbdevup=2, nbdevdn=2)['lowerband']
+    dataframe['sma'] = ta.SMA(dataframe, timeperiod=40)
+    dataframe['tema'] = ta.TEMA(dataframe, timeperiod=9)
+    dataframe['mfi'] = ta.MFI(dataframe)
+    dataframe['rsi'] = ta.RSI(dataframe)
+    dataframe['ema5'] = ta.EMA(dataframe, timeperiod=5)
+    dataframe['ema10'] = ta.EMA(dataframe, timeperiod=10)
+    dataframe['ema50'] = ta.EMA(dataframe, timeperiod=50)
+    dataframe['ema100'] = ta.EMA(dataframe, timeperiod=100)
+    dataframe['ao'] = awesome_oscillator(dataframe)
+    macd = ta.MACD(dataframe)
+    dataframe['macd'] = macd['macd']
+    dataframe['macdsignal'] = macd['macdsignal']
+    dataframe['macdhist'] = macd['macdhist']
+    hilbert = ta.HT_SINE(dataframe)
+    dataframe['htsine'] = hilbert['sine']
+    dataframe['htleadsine'] = hilbert['leadsine']
+    dataframe['plus_dm'] = ta.PLUS_DM(dataframe)
+    dataframe['plus_di'] = ta.PLUS_DI(dataframe)
+    dataframe['minus_dm'] = ta.MINUS_DM(dataframe)
+    dataframe['minus_di'] = ta.MINUS_DI(dataframe)
+    return dataframe
+
+
+

Want more indicator examples?

+

Look into the user_data/strategies/sample_strategy.py. +Then uncomment indicators you need.

+
+

Indicator libraries

+

Out of the box, freqtrade installs the following technical libraries:

+ +

Additional technical libraries can be installed as necessary, or custom indicators may be written / invented by the strategy author.

+

Strategy startup period

+

Most indicators have an instable startup period, in which they are either not available (NaN), or the calculation is incorrect. This can lead to inconsistencies, since Freqtrade does not know how long this instable period should be. +To account for this, the strategy can be assigned the startup_candle_count attribute. +This should be set to the maximum number of candles that the strategy requires to calculate stable indicators. In the case where a user includes higher timeframes with informative pairs, the startup_candle_count does not necessarily change. The value is the maximum period (in candles) that any of the informatives timeframes need to compute stable indicators.

+

In this example strategy, this should be set to 100 (startup_candle_count = 100), since the longest needed history is 100 candles.

+
    dataframe['ema100'] = ta.EMA(dataframe, timeperiod=100)
+
+

By letting the bot know how much history is needed, backtest trades can start at the specified timerange during backtesting and hyperopt.

+
+

Using x calls to get OHLCV

+

If you receive a warning like WARNING - Using 3 calls to get OHLCV. This can result in slower operations for the bot. Please check if you really need 1500 candles for your strategy - you should consider if you really need this much historic data for your signals. +Having this will cause Freqtrade to make multiple calls for the same pair, which will obviously be slower than one network request. +As a consequence, Freqtrade will take longer to refresh candles - and should therefore be avoided if possible. +This is capped to 5 total calls to avoid overloading the exchange, or make freqtrade too slow.

+
+
+

Warning

+

startup_candle_count should be below ohlcv_candle_limit * 5 (which is 500 * 5 for most exchanges) - since only this amount of candles will be available during Dry-Run/Live Trade operations.

+
+

Example

+

Let's try to backtest 1 month (January 2019) of 5m candles using an example strategy with EMA100, as above.

+
freqtrade backtesting --timerange 20190101-20190201 --timeframe 5m
+
+

Assuming startup_candle_count is set to 100, backtesting knows it needs 100 candles to generate valid buy signals. It will load data from 20190101 - (100 * 5m) - which is ~2018-12-31 15:30:00. +If this data is available, indicators will be calculated with this extended timerange. The instable startup period (up to 2019-01-01 00:00:00) will then be removed before starting backtesting.

+
+

Note

+

If data for the startup period is not available, then the timerange will be adjusted to account for this startup period - so Backtesting would start at 2019-01-01 08:30:00.

+
+

Entry signal rules

+

Edit the method populate_entry_trend() in your strategy file to update your entry strategy.

+

It's important to always return the dataframe without removing/modifying the columns "open", "high", "low", "close", "volume", otherwise these fields would contain something unexpected.

+

This method will also define a new column, "enter_long" ("enter_short" for shorts), which needs to contain 1 for entries, and 0 for "no action". enter_long is a mandatory column that must be set even if the strategy is shorting only.

+

Sample from user_data/strategies/sample_strategy.py:

+
def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+    """
+    Based on TA indicators, populates the buy signal for the given dataframe
+    :param dataframe: DataFrame populated with indicators
+    :param metadata: Additional information, like the currently traded pair
+    :return: DataFrame with buy column
+    """
+    dataframe.loc[
+        (
+            (qtpylib.crossed_above(dataframe['rsi'], 30)) &  # Signal: RSI crosses above 30
+            (dataframe['tema'] <= dataframe['bb_middleband']) &  # Guard
+            (dataframe['tema'] > dataframe['tema'].shift(1)) &  # Guard
+            (dataframe['volume'] > 0)  # Make sure Volume is not 0
+        ),
+        ['enter_long', 'enter_tag']] = (1, 'rsi_cross')
+
+    return dataframe
+
+
+Enter short trades +

Short-entries can be created by setting enter_short (corresponds to enter_long for long trades). +The enter_tag column remains identical. +Short-trades need to be supported by your exchange and market configuration! +Please make sure to set can_short appropriately on your strategy if you intend to short.

+
def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+    dataframe.loc[
+        (
+            (qtpylib.crossed_above(dataframe['rsi'], 30)) &  # Signal: RSI crosses above 30
+            (dataframe['tema'] <= dataframe['bb_middleband']) &  # Guard
+            (dataframe['tema'] > dataframe['tema'].shift(1)) &  # Guard
+            (dataframe['volume'] > 0)  # Make sure Volume is not 0
+        ),
+        ['enter_long', 'enter_tag']] = (1, 'rsi_cross')
+
+    dataframe.loc[
+        (
+            (qtpylib.crossed_below(dataframe['rsi'], 70)) &  # Signal: RSI crosses below 70
+            (dataframe['tema'] > dataframe['bb_middleband']) &  # Guard
+            (dataframe['tema'] < dataframe['tema'].shift(1)) &  # Guard
+            (dataframe['volume'] > 0)  # Make sure Volume is not 0
+        ),
+        ['enter_short', 'enter_tag']] = (1, 'rsi_cross')
+
+    return dataframe
+
+
+
+

Note

+

Buying requires sellers to buy from - therefore volume needs to be > 0 (dataframe['volume'] > 0) to make sure that the bot does not buy/sell in no-activity periods.

+
+

Exit signal rules

+

Edit the method populate_exit_trend() into your strategy file to update your exit strategy. +The exit-signal is only used for exits if use_exit_signal is set to true in the configuration. +use_exit_signal will not influence signal collision rules - which will still apply and can prevent entries.

+

It's important to always return the dataframe without removing/modifying the columns "open", "high", "low", "close", "volume", otherwise these fields would contain something unexpected.

+

This method will also define a new column, "exit_long" ("exit_short" for shorts), which needs to contain 1 for exits, and 0 for "no action".

+

Sample from user_data/strategies/sample_strategy.py:

+
def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+    """
+    Based on TA indicators, populates the exit signal for the given dataframe
+    :param dataframe: DataFrame populated with indicators
+    :param metadata: Additional information, like the currently traded pair
+    :return: DataFrame with buy column
+    """
+    dataframe.loc[
+        (
+            (qtpylib.crossed_above(dataframe['rsi'], 70)) &  # Signal: RSI crosses above 70
+            (dataframe['tema'] > dataframe['bb_middleband']) &  # Guard
+            (dataframe['tema'] < dataframe['tema'].shift(1)) &  # Guard
+            (dataframe['volume'] > 0)  # Make sure Volume is not 0
+        ),
+        ['exit_long', 'exit_tag']] = (1, 'rsi_too_high')
+    return dataframe
+
+
+Exit short trades +

Short-exits can be created by setting exit_short (corresponds to exit_long). +The exit_tag column remains identical. +Short-trades need to be supported by your exchange and market configuration!

+
def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+    dataframe.loc[
+        (
+            (qtpylib.crossed_above(dataframe['rsi'], 70)) &  # Signal: RSI crosses above 70
+            (dataframe['tema'] > dataframe['bb_middleband']) &  # Guard
+            (dataframe['tema'] < dataframe['tema'].shift(1)) &  # Guard
+            (dataframe['volume'] > 0)  # Make sure Volume is not 0
+        ),
+        ['exit_long', 'exit_tag']] = (1, 'rsi_too_high')
+    dataframe.loc[
+        (
+            (qtpylib.crossed_below(dataframe['rsi'], 30)) &  # Signal: RSI crosses below 30
+            (dataframe['tema'] < dataframe['bb_middleband']) &  # Guard
+            (dataframe['tema'] > dataframe['tema'].shift(1)) &  # Guard
+            (dataframe['volume'] > 0)  # Make sure Volume is not 0
+        ),
+        ['exit_short', 'exit_tag']] = (1, 'rsi_too_low')
+    return dataframe
+
+
+

Minimal ROI

+

This dict defines the minimal Return On Investment (ROI) a trade should reach before exiting, independent from the exit signal.

+

It is of the following format, with the dict key (left side of the colon) being the minutes passed since the trade opened, and the value (right side of the colon) being the percentage.

+
minimal_roi = {
+    "40": 0.0,
+    "30": 0.01,
+    "20": 0.02,
+    "0": 0.04
+}
+
+

The above configuration would therefore mean:

+
    +
  • Exit whenever 4% profit was reached
  • +
  • Exit when 2% profit was reached (in effect after 20 minutes)
  • +
  • Exit when 1% profit was reached (in effect after 30 minutes)
  • +
  • Exit when trade is non-loosing (in effect after 40 minutes)
  • +
+

The calculation does include fees.

+

To disable ROI completely, set it to an insanely high number:

+
minimal_roi = {
+    "0": 100
+}
+
+

While technically not completely disabled, this would exit once the trade reaches 10000% Profit.

+

To use times based on candle duration (timeframe), the following snippet can be handy. +This will allow you to change the timeframe for the strategy, and ROI times will still be set as candles (e.g. after 3 candles ...)

+
from freqtrade.exchange import timeframe_to_minutes
+
+class AwesomeStrategy(IStrategy):
+
+    timeframe = "1d"
+    timeframe_mins = timeframe_to_minutes(timeframe)
+    minimal_roi = {
+        "0": 0.05,                      # 5% for the first 3 candles
+        str(timeframe_mins * 3): 0.02,  # 2% after 3 candles
+        str(timeframe_mins * 6): 0.01,  # 1% After 6 candles
+    }
+
+

Stoploss

+

Setting a stoploss is highly recommended to protect your capital from strong moves against you.

+

Sample of setting a 10% stoploss:

+
stoploss = -0.10
+
+

For the full documentation on stoploss features, look at the dedicated stoploss page.

+

Timeframe

+

This is the set of candles the bot should download and use for the analysis. +Common values are "1m", "5m", "15m", "1h", however all values supported by your exchange should work.

+

Please note that the same entry/exit signals may work well with one timeframe, but not with the others.

+

This setting is accessible within the strategy methods as the self.timeframe attribute.

+

Can short

+

To use short signals in futures markets, you will have to let us know to do so by setting can_short=True. +Strategies which enable this will fail to load on spot markets. +Disabling of this will have short signals ignored (also in futures markets).

+

Metadata dict

+

The metadata-dict (available for populate_entry_trend, populate_exit_trend, populate_indicators) contains additional information. +Currently this is pair, which can be accessed using metadata['pair'] - and will return a pair in the format XRP/BTC.

+

The Metadata-dict should not be modified and does not persist information across multiple calls. +Instead, have a look at the Storing information section.

+

Strategy file loading

+

By default, freqtrade will attempt to load strategies from all .py files within user_data/strategies.

+

Assuming your strategy is called AwesomeStrategy, stored in the file user_data/strategies/AwesomeStrategy.py, then you can start freqtrade with freqtrade trade --strategy AwesomeStrategy. +Note that we're using the class-name, not the file name.

+

You can use freqtrade list-strategies to see a list of all strategies Freqtrade is able to load (all strategies in the correct folder). +It will also include a "status" field, highlighting potential problems.

+
+Customize strategy directory +

You can use a different directory by using --strategy-path user_data/otherPath. This parameter is available to all commands that require a strategy.

+
+

Informative Pairs

+

Get data for non-tradeable pairs

+

Data for additional, informative pairs (reference pairs) can be beneficial for some strategies. +OHLCV data for these pairs will be downloaded as part of the regular whitelist refresh process and is available via DataProvider just as other pairs (see below). +These parts will not be traded unless they are also specified in the pair whitelist, or have been selected by Dynamic Whitelisting.

+

The pairs need to be specified as tuples in the format ("pair", "timeframe"), with pair as the first and timeframe as the second argument.

+

Sample:

+
def informative_pairs(self):
+    return [("ETH/USDT", "5m"),
+            ("BTC/TUSD", "15m"),
+            ]
+
+

A full sample can be found in the DataProvider section.

+
+

Warning

+

As these pairs will be refreshed as part of the regular whitelist refresh, it's best to keep this list short. +All timeframes and all pairs can be specified as long as they are available (and active) on the used exchange. +It is however better to use resampling to longer timeframes whenever possible +to avoid hammering the exchange with too many requests and risk being blocked.

+
+
+Alternative candle types +

Informative_pairs can also provide a 3rd tuple element defining the candle type explicitly. +Availability of alternative candle-types will depend on the trading-mode and the exchange. +In general, spot pairs cannot be used in futures markets, and futures candles can't be used as informative pairs for spot bots. +Details about this may vary, if they do, this can be found in the exchange documentation.

+
def informative_pairs(self):
+    return [
+        ("ETH/USDT", "5m", ""),   # Uses default candletype, depends on trading_mode (recommended)
+        ("ETH/USDT", "5m", "spot"),   # Forces usage of spot candles (only valid for bots running on spot markets).
+        ("BTC/TUSD", "15m", "futures"),  # Uses futures candles (only bots with `trading_mode=futures`)
+        ("BTC/TUSD", "15m", "mark"),  # Uses mark candles (only bots with `trading_mode=futures`)
+    ]
+
+
+
+

Informative pairs decorator (@informative())

+

In most common case it is possible to easily define informative pairs by using a decorator. All decorated populate_indicators_* methods run in isolation, +not having access to data from other informative pairs, in the end all informative dataframes are merged and passed to main populate_indicators() method. +When hyperopting, use of hyperoptable parameter .value attribute is not supported. Please use .range attribute. See optimizing an indicator parameter +for more information.

+
+Full documentation +
def informative(timeframe: str, asset: str = '',
+                fmt: Optional[Union[str, Callable[[KwArg(str)], str]]] = None,
+                *,
+                candle_type: Optional[CandleType] = None,
+                ffill: bool = True) -> Callable[[PopulateIndicators], PopulateIndicators]:
+    """
+    A decorator for populate_indicators_Nn(self, dataframe, metadata), allowing these functions to
+    define informative indicators.
+
+    Example usage:
+
+        @informative('1h')
+        def populate_indicators_1h(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+            dataframe['rsi'] = ta.RSI(dataframe, timeperiod=14)
+            return dataframe
+
+    :param timeframe: Informative timeframe. Must always be equal or higher than strategy timeframe.
+    :param asset: Informative asset, for example BTC, BTC/USDT, ETH/BTC. Do not specify to use
+    current pair.
+    :param fmt: Column format (str) or column formatter (callable(name, asset, timeframe)). When not
+    specified, defaults to:
+    * {base}_{quote}_{column}_{timeframe} if asset is specified. 
+    * {column}_{timeframe} if asset is not specified.
+    Format string supports these format variables:
+    * {asset} - full name of the asset, for example 'BTC/USDT'.
+    * {base} - base currency in lower case, for example 'eth'.
+    * {BASE} - same as {base}, except in upper case.
+    * {quote} - quote currency in lower case, for example 'usdt'.
+    * {QUOTE} - same as {quote}, except in upper case.
+    * {column} - name of dataframe column.
+    * {timeframe} - timeframe of informative dataframe.
+    :param ffill: ffill dataframe after merging informative pair.
+    :param candle_type: '', mark, index, premiumIndex, or funding_rate
+    """
+
+
+
+Fast and easy way to define informative pairs +

Most of the time we do not need power and flexibility offered by merge_informative_pair(), therefore we can use a decorator to quickly define informative pairs.

+
from datetime import datetime
+from freqtrade.persistence import Trade
+from freqtrade.strategy import IStrategy, informative
+
+class AwesomeStrategy(IStrategy):
+
+    # This method is not required. 
+    # def informative_pairs(self): ...
+
+    # Define informative upper timeframe for each pair. Decorators can be stacked on same 
+    # method. Available in populate_indicators as 'rsi_30m' and 'rsi_1h'.
+    @informative('30m')
+    @informative('1h')
+    def populate_indicators_1h(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        dataframe['rsi'] = ta.RSI(dataframe, timeperiod=14)
+        return dataframe
+
+    # Define BTC/STAKE informative pair. Available in populate_indicators and other methods as
+    # 'btc_rsi_1h'. Current stake currency should be specified as {stake} format variable 
+    # instead of hard-coding actual stake currency. Available in populate_indicators and other 
+    # methods as 'btc_usdt_rsi_1h' (when stake currency is USDT).
+    @informative('1h', 'BTC/{stake}')
+    def populate_indicators_btc_1h(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        dataframe['rsi'] = ta.RSI(dataframe, timeperiod=14)
+        return dataframe
+
+    # Define BTC/ETH informative pair. You must specify quote currency if it is different from
+    # stake currency. Available in populate_indicators and other methods as 'eth_btc_rsi_1h'.
+    @informative('1h', 'ETH/BTC')
+    def populate_indicators_eth_btc_1h(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        dataframe['rsi'] = ta.RSI(dataframe, timeperiod=14)
+        return dataframe
+
+    # Define BTC/STAKE informative pair. A custom formatter may be specified for formatting
+    # column names. A callable `fmt(**kwargs) -> str` may be specified, to implement custom
+    # formatting. Available in populate_indicators and other methods as 'rsi_upper'.
+    @informative('1h', 'BTC/{stake}', '{column}')
+    def populate_indicators_btc_1h_2(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        dataframe['rsi_upper'] = ta.RSI(dataframe, timeperiod=14)
+        return dataframe
+
+    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        # Strategy timeframe indicators for current pair.
+        dataframe['rsi'] = ta.RSI(dataframe, timeperiod=14)
+        # Informative pairs are available in this method.
+        dataframe['rsi_less'] = dataframe['rsi'] < dataframe['rsi_1h']
+        return dataframe
+
+
+
+

Note

+

Do not use @informative decorator if you need to use data of one informative pair when generating another informative pair. Instead, define informative pairs +manually as described in the DataProvider section.

+
+
+

Note

+

Use string formatting when accessing informative dataframes of other pairs. This will allow easily changing stake currency in config without having to adjust strategy code.

+
def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+    stake = self.config['stake_currency']
+    dataframe.loc[
+        (
+            (dataframe[f'btc_{stake}_rsi_1h'] < 35)
+            &
+            (dataframe['volume'] > 0)
+        ),
+        ['enter_long', 'enter_tag']] = (1, 'buy_signal_rsi')
+
+    return dataframe
+
+

Alternatively column renaming may be used to remove stake currency from column names: @informative('1h', 'BTC/{stake}', fmt='{base}_{column}_{timeframe}').

+
+
+

Duplicate method names

+

Methods tagged with @informative() decorator must always have unique names! Re-using same name (for example when copy-pasting already defined informative method) +will overwrite previously defined method and not produce any errors due to limitations of Python programming language. In such cases you will find that indicators +created in earlier-defined methods are not available in the dataframe. Carefully review method names and make sure they are unique!

+
+

Additional data (DataProvider)

+

The strategy provides access to the DataProvider. This allows you to get additional data to use in your strategy.

+

All methods return None in case of failure (do not raise an exception).

+

Please always check the mode of operation to select the correct method to get data (samples see below).

+
+

Hyperopt

+

Dataprovider is available during hyperopt, however it can only be used in populate_indicators() within a strategy. +It is not available in populate_buy() and populate_sell() methods, nor in populate_indicators(), if this method located in the hyperopt file.

+
+

Possible options for DataProvider

+
    +
  • available_pairs - Property with tuples listing cached pairs with their timeframe (pair, timeframe).
  • +
  • current_whitelist() - Returns a current list of whitelisted pairs. Useful for accessing dynamic whitelists (i.e. VolumePairlist)
  • +
  • get_pair_dataframe(pair, timeframe) - This is a universal method, which returns either historical data (for backtesting) or cached live data (for the Dry-Run and Live-Run modes).
  • +
  • get_analyzed_dataframe(pair, timeframe) - Returns the analyzed dataframe (after calling populate_indicators(), populate_buy(), populate_sell()) and the time of the latest analysis.
  • +
  • historic_ohlcv(pair, timeframe) - Returns historical data stored on disk.
  • +
  • market(pair) - Returns market data for the pair: fees, limits, precisions, activity flag, etc. See ccxt documentation for more details on the Market data structure.
  • +
  • ohlcv(pair, timeframe) - Currently cached candle (OHLCV) data for the pair, returns DataFrame or empty DataFrame.
  • +
  • orderbook(pair, maximum) - Returns latest orderbook data for the pair, a dict with bids/asks with a total of maximum entries.
  • +
  • ticker(pair) - Returns current ticker data for the pair. See ccxt documentation for more details on the Ticker data structure.
  • +
  • runmode - Property containing the current runmode.
  • +
+

Example Usages

+

available_pairs

+
for pair, timeframe in self.dp.available_pairs:
+    print(f"available {pair}, {timeframe}")
+
+

current_whitelist()

+

Imagine you've developed a strategy that trades the 5m timeframe using signals generated from a 1d timeframe on the top 10 volume pairs by volume.

+

The strategy might look something like this:

+

Scan through the top 10 pairs by volume using the VolumePairList every 5 minutes and use a 14 day RSI to buy and sell.

+

Due to the limited available data, it's very difficult to resample 5m candles into daily candles for use in a 14 day RSI. Most exchanges limit us to just 500-1000 candles which effectively gives us around 1.74 daily candles. We need 14 days at least!

+

Since we can't resample the data we will have to use an informative pair; and since the whitelist will be dynamic we don't know which pair(s) to use.

+

This is where calling self.dp.current_whitelist() comes in handy.

+
    def informative_pairs(self):
+
+        # get access to all pairs available in whitelist.
+        pairs = self.dp.current_whitelist()
+        # Assign tf to each pair so they can be downloaded and cached for strategy.
+        informative_pairs = [(pair, '1d') for pair in pairs]
+        return informative_pairs
+
+
+Plotting with current_whitelist +

Current whitelist is not supported for plot-dataframe, as this command is usually used by providing an explicit pairlist - and would therefore make the return values of this method misleading.

+
+

get_pair_dataframe(pair, timeframe)

+
# fetch live / historical candle (OHLCV) data for the first informative pair
+inf_pair, inf_timeframe = self.informative_pairs()[0]
+informative = self.dp.get_pair_dataframe(pair=inf_pair,
+                                         timeframe=inf_timeframe)
+
+
+

Warning about backtesting

+

In backtesting, dp.get_pair_dataframe() behavior differs depending on where it's called. +Within populate_*() methods, dp.get_pair_dataframe() returns the full timerange. Please make sure to not "look into the future" to avoid surprises when running in dry/live mode. +Within callbacks, you'll get the full timerange up to the current (simulated) candle.

+
+

get_analyzed_dataframe(pair, timeframe)

+

This method is used by freqtrade internally to determine the last signal. +It can also be used in specific callbacks to get the signal that caused the action (see Advanced Strategy Documentation for more details on available callbacks).

+
# fetch current dataframe
+dataframe, last_updated = self.dp.get_analyzed_dataframe(pair=metadata['pair'],
+                                                         timeframe=self.timeframe)
+
+
+

No data available

+

Returns an empty dataframe if the requested pair was not cached. +You can check for this with if dataframe.empty: and handle this case accordingly. +This should not happen when using whitelisted pairs.

+
+

orderbook(pair, maximum)

+
if self.dp.runmode.value in ('live', 'dry_run'):
+    ob = self.dp.orderbook(metadata['pair'], 1)
+    dataframe['best_bid'] = ob['bids'][0][0]
+    dataframe['best_ask'] = ob['asks'][0][0]
+
+

The orderbook structure is aligned with the order structure from ccxt, so the result will look as follows:

+
{
+    'bids': [
+        [ price, amount ], // [ float, float ]
+        [ price, amount ],
+        ...
+    ],
+    'asks': [
+        [ price, amount ],
+        [ price, amount ],
+        //...
+    ],
+    //...
+}
+
+

Therefore, using ob['bids'][0][0] as demonstrated above will result in using the best bid price. ob['bids'][0][1] would look at the amount at this orderbook position.

+
+

Warning about backtesting

+

The order book is not part of the historic data which means backtesting and hyperopt will not work correctly if this method is used, as the method will return uptodate values.

+
+

ticker(pair)

+
if self.dp.runmode.value in ('live', 'dry_run'):
+    ticker = self.dp.ticker(metadata['pair'])
+    dataframe['last_price'] = ticker['last']
+    dataframe['volume24h'] = ticker['quoteVolume']
+    dataframe['vwap'] = ticker['vwap']
+
+
+

Warning

+

Although the ticker data structure is a part of the ccxt Unified Interface, the values returned by this method can +vary for different exchanges. For instance, many exchanges do not return vwap values, some exchanges +does not always fills in the last field (so it can be None), etc. So you need to carefully verify the ticker +data returned from the exchange and add appropriate error handling / defaults.

+
+
+

Warning about backtesting

+

This method will always return up-to-date values - so usage during backtesting / hyperopt without runmode checks will lead to wrong results.

+
+

Send Notification

+

The dataprovider .send_msg() function allows you to send custom notifications from your strategy. +Identical notifications will only be sent once per candle, unless the 2nd argument (always_send) is set to True.

+
    self.dp.send_msg(f"{metadata['pair']} just got hot!")
+
+    # Force send this notification, avoid caching (Please read warning below!)
+    self.dp.send_msg(f"{metadata['pair']} just got hot!", always_send=True)
+
+

Notifications will only be sent in trading modes (Live/Dry-run) - so this method can be called without conditions for backtesting.

+
+

Spamming

+

You can spam yourself pretty good by setting always_send=True in this method. Use this with great care and only in conditions you know will not happen throughout a candle to avoid a message every 5 seconds.

+
+

Complete Data-provider sample

+
from freqtrade.strategy import IStrategy, merge_informative_pair
+from pandas import DataFrame
+
+class SampleStrategy(IStrategy):
+    # strategy init stuff...
+
+    timeframe = '5m'
+
+    # more strategy init stuff..
+
+    def informative_pairs(self):
+
+        # get access to all pairs available in whitelist.
+        pairs = self.dp.current_whitelist()
+        # Assign tf to each pair so they can be downloaded and cached for strategy.
+        informative_pairs = [(pair, '1d') for pair in pairs]
+        # Optionally Add additional "static" pairs
+        informative_pairs += [("ETH/USDT", "5m"),
+                              ("BTC/TUSD", "15m"),
+                            ]
+        return informative_pairs
+
+    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        if not self.dp:
+            # Don't do anything if DataProvider is not available.
+            return dataframe
+
+        inf_tf = '1d'
+        # Get the informative pair
+        informative = self.dp.get_pair_dataframe(pair=metadata['pair'], timeframe=inf_tf)
+        # Get the 14 day rsi
+        informative['rsi'] = ta.RSI(informative, timeperiod=14)
+
+        # Use the helper function merge_informative_pair to safely merge the pair
+        # Automatically renames the columns and merges a shorter timeframe dataframe and a longer timeframe informative pair
+        # use ffill to have the 1d value available in every row throughout the day.
+        # Without this, comparisons between columns of the original and the informative pair would only work once per day.
+        # Full documentation of this method, see below
+        dataframe = merge_informative_pair(dataframe, informative, self.timeframe, inf_tf, ffill=True)
+
+        # Calculate rsi of the original dataframe (5m timeframe)
+        dataframe['rsi'] = ta.RSI(dataframe, timeperiod=14)
+
+        # Do other stuff
+        # ...
+
+        return dataframe
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+
+        dataframe.loc[
+            (
+                (qtpylib.crossed_above(dataframe['rsi'], 30)) &  # Signal: RSI crosses above 30
+                (dataframe['rsi_1d'] < 30) &                     # Ensure daily RSI is < 30
+                (dataframe['volume'] > 0)                        # Ensure this candle had volume (important for backtesting)
+            ),
+            ['enter_long', 'enter_tag']] = (1, 'rsi_cross')
+
+
+

Helper functions

+

merge_informative_pair()

+

This method helps you merge an informative pair to a regular dataframe without lookahead bias. +It's there to help you merge the dataframe in a safe and consistent way.

+

Options:

+
    +
  • Rename the columns for you to create unique columns
  • +
  • Merge the dataframe without lookahead bias
  • +
  • Forward-fill (optional)
  • +
+

For a full sample, please refer to the complete data provider example below.

+

All columns of the informative dataframe will be available on the returning dataframe in a renamed fashion:

+
+

Column renaming

+

Assuming inf_tf = '1d' the resulting columns will be:

+
'date', 'open', 'high', 'low', 'close', 'rsi'                     # from the original dataframe
+'date_1d', 'open_1d', 'high_1d', 'low_1d', 'close_1d', 'rsi_1d'   # from the informative dataframe
+
+
+
+Column renaming - 1h +

Assuming inf_tf = '1h' the resulting columns will be:

+
'date', 'open', 'high', 'low', 'close', 'rsi'                     # from the original dataframe
+'date_1h', 'open_1h', 'high_1h', 'low_1h', 'close_1h', 'rsi_1h'   # from the informative dataframe
+
+
+
+Custom implementation +

A custom implementation for this is possible, and can be done as follows:

+
# Shift date by 1 candle
+# This is necessary since the data is always the "open date"
+# and a 15m candle starting at 12:15 should not know the close of the 1h candle from 12:00 to 13:00
+minutes = timeframe_to_minutes(inf_tf)
+# Only do this if the timeframes are different:
+informative['date_merge'] = informative["date"] + pd.to_timedelta(minutes, 'm')
+
+# Rename columns to be unique
+informative.columns = [f"{col}_{inf_tf}" for col in informative.columns]
+# Assuming inf_tf = '1d' - then the columns will now be:
+# date_1d, open_1d, high_1d, low_1d, close_1d, rsi_1d
+
+# Combine the 2 dataframes
+# all indicators on the informative sample MUST be calculated before this point
+dataframe = pd.merge(dataframe, informative, left_on='date', right_on=f'date_merge_{inf_tf}', how='left')
+# FFill to have the 1d value available in every row throughout the day.
+# Without this, comparisons would only work once per day.
+dataframe = dataframe.ffill()
+
+
+
+

Informative timeframe < timeframe

+

Using informative timeframes smaller than the dataframe timeframe is not recommended with this method, as it will not use any of the additional information this would provide. +To use the more detailed information properly, more advanced methods should be applied (which are out of scope for freqtrade documentation, as it'll depend on the respective need).

+
+
+

stoploss_from_open()

+

Stoploss values returned from custom_stoploss must specify a percentage relative to current_rate, but sometimes you may want to specify a stoploss relative to the entry point instead. stoploss_from_open() is a helper function to calculate a stoploss value that can be returned from custom_stoploss which will be equivalent to the desired trade profit above the entry point.

+
+Returning a stoploss relative to the open price from the custom stoploss function +

Say the open price was $100, and current_price is $121 (current_profit will be 0.21).

+

If we want a stop price at 7% above the open price we can call stoploss_from_open(0.07, current_profit, False) which will return 0.1157024793. 11.57% below $121 is $107, which is the same as 7% above $100.

+

This function will consider leverage - so at 10x leverage, the actual stoploss would be 0.7% above $100 (0.7% * 10x = 7%).

+
from datetime import datetime
+from freqtrade.persistence import Trade
+from freqtrade.strategy import IStrategy, stoploss_from_open
+
+class AwesomeStrategy(IStrategy):
+
+    # ... populate_* methods
+
+    use_custom_stoploss = True
+
+    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,
+                        current_rate: float, current_profit: float, **kwargs) -> float:
+
+        # once the profit has risen above 10%, keep the stoploss at 7% above the open price
+        if current_profit > 0.10:
+            return stoploss_from_open(0.07, current_profit, is_short=trade.is_short, leverage=trade.leverage)
+
+        return 1
+
+

Full examples can be found in the Custom stoploss section of the Documentation.

+
+
+

Note

+

Providing invalid input to stoploss_from_open() may produce "CustomStoploss function did not return valid stoploss" warnings. +This may happen if current_profit parameter is below specified open_relative_stop. Such situations may arise when closing trade +is blocked by confirm_trade_exit() method. Warnings can be solved by never blocking stop loss sells by checking exit_reason in +confirm_trade_exit(), or by using return stoploss_from_open(...) or 1 idiom, which will request to not change stop loss when +current_profit < open_relative_stop.

+
+

stoploss_from_absolute()

+

In some situations it may be confusing to deal with stops relative to current rate. Instead, you may define a stoploss level using an absolute price.

+
+Returning a stoploss using absolute price from the custom stoploss function +

If we want to trail a stop price at 2xATR below current price we can call stoploss_from_absolute(current_rate - (candle['atr'] * 2), current_rate, is_short=trade.is_short).

+
from datetime import datetime
+from freqtrade.persistence import Trade
+from freqtrade.strategy import IStrategy, stoploss_from_absolute
+
+class AwesomeStrategy(IStrategy):
+
+    use_custom_stoploss = True
+
+    def populate_indicators_1h(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        dataframe['atr'] = ta.ATR(dataframe, timeperiod=14)
+        return dataframe
+
+    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,
+                        current_rate: float, current_profit: float, **kwargs) -> float:
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        candle = dataframe.iloc[-1].squeeze()
+        return stoploss_from_absolute(current_rate - (candle['atr'] * 2), current_rate, is_short=trade.is_short)
+
+
+

Additional data (Wallets)

+

The strategy provides access to the wallets object. This contains the current balances on the exchange.

+
+

Backtesting / Hyperopt

+

Wallets behaves differently depending on the function it's called. +Within populate_*() methods, it'll return the full wallet as configured. +Within callbacks, you'll get the wallet state corresponding to the actual simulated wallet at that point in the simulation process.

+
+

Please always check if wallets is available to avoid failures during backtesting.

+
if self.wallets:
+    free_eth = self.wallets.get_free('ETH')
+    used_eth = self.wallets.get_used('ETH')
+    total_eth = self.wallets.get_total('ETH')
+
+

Possible options for Wallets

+
    +
  • get_free(asset) - currently available balance to trade
  • +
  • get_used(asset) - currently tied up balance (open orders)
  • +
  • get_total(asset) - total available balance - sum of the 2 above
  • +
+
+

Additional data (Trades)

+

A history of Trades can be retrieved in the strategy by querying the database.

+

At the top of the file, import Trade.

+
from freqtrade.persistence import Trade
+
+

The following example queries for the current pair and trades from today, however other filters can easily be added.

+
trades = Trade.get_trades_proxy(pair=metadata['pair'],
+                                open_date=datetime.now(timezone.utc) - timedelta(days=1),
+                                is_open=False,
+            ]).order_by(Trade.close_date).all()
+# Summarize profit for this pair.
+curdayprofit = sum(trade.close_profit for trade in trades)
+
+

For a full list of available methods, please consult the Trade object documentation.

+
+

Warning

+

Trade history is not available in populate_* methods during backtesting or hyperopt, and will result in empty results.

+
+

Prevent trades from happening for a specific pair

+

Freqtrade locks pairs automatically for the current candle (until that candle is over) when a pair is sold, preventing an immediate re-buy of that pair.

+

Locked pairs will show the message Pair <pair> is currently locked..

+

Locking pairs from within the strategy

+

Sometimes it may be desired to lock a pair after certain events happen (e.g. multiple losing trades in a row).

+

Freqtrade has an easy method to do this from within the strategy, by calling self.lock_pair(pair, until, [reason]). +until must be a datetime object in the future, after which trading will be re-enabled for that pair, while reason is an optional string detailing why the pair was locked.

+

Locks can also be lifted manually, by calling self.unlock_pair(pair) or self.unlock_reason(<reason>) - providing reason the pair was locked with. +self.unlock_reason(<reason>) will unlock all pairs currently locked with the provided reason.

+

To verify if a pair is currently locked, use self.is_pair_locked(pair).

+
+

Note

+

Locked pairs will always be rounded up to the next candle. So assuming a 5m timeframe, a lock with until set to 10:18 will lock the pair until the candle from 10:15-10:20 will be finished.

+
+
+

Warning

+

Manually locking pairs is not available during backtesting, only locks via Protections are allowed.

+
+

Pair locking example

+
from freqtrade.persistence import Trade
+from datetime import timedelta, datetime, timezone
+# Put the above lines a the top of the strategy file, next to all the other imports
+# --------
+
+# Within populate indicators (or populate_buy):
+if self.config['runmode'].value in ('live', 'dry_run'):
+    # fetch closed trades for the last 2 days
+    trades = Trade.get_trades_proxy(
+        pair=metadata['pair'], is_open=False, 
+        open_date=datetime.now(timezone.utc) - timedelta(days=2))
+    # Analyze the conditions you'd like to lock the pair .... will probably be different for every strategy
+    sumprofit = sum(trade.close_profit for trade in trades)
+    if sumprofit < 0:
+        # Lock pair for 12 hours
+        self.lock_pair(metadata['pair'], until=datetime.now(timezone.utc) + timedelta(hours=12))
+
+ +

To inspect the created dataframe, you can issue a print-statement in either populate_entry_trend() or populate_exit_trend(). +You may also want to print the pair so it's clear what data is currently shown.

+
def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+    dataframe.loc[
+        (
+            #>> whatever condition<<<
+        ),
+        ['enter_long', 'enter_tag']] = (1, 'somestring')
+
+    # Print the Analyzed pair
+    print(f"result for {metadata['pair']}")
+
+    # Inspect the last 5 rows
+    print(dataframe.tail())
+
+    return dataframe
+
+

Printing more than a few rows is also possible (simply use print(dataframe) instead of print(dataframe.tail())), however not recommended, as that will be very verbose (~500 lines per pair every 5 seconds).

+

Common mistakes when developing strategies

+

Peeking into the future while backtesting

+

Backtesting analyzes the whole time-range at once for performance reasons. Because of this, strategy authors need to make sure that strategies do not look-ahead into the future. +This is a common pain-point, which can cause huge differences between backtesting and dry/live run methods, since they all use data which is not available during dry/live runs, so these strategies will perform well during backtesting, but will fail / perform badly in real conditions.

+

The following lists some common patterns which should be avoided to prevent frustration:

+
    +
  • don't use shift(-1). This uses data from the future, which is not available.
  • +
  • don't use .iloc[-1] or any other absolute position in the dataframe, this will be different between dry-run and backtesting.
  • +
  • don't use dataframe['volume'].mean(). This uses the full DataFrame for backtesting, including data from the future. Use dataframe['volume'].rolling(<window>).mean() instead
  • +
  • don't use .resample('1h'). This uses the left border of the interval, so moves data from an hour to the start of the hour. Use .resample('1h', label='right') instead.
  • +
+

Colliding signals

+

When conflicting signals collide (e.g. both 'enter_long' and 'exit_long' are 1), freqtrade will do nothing and ignore the entry signal. This will avoid trades that enter, and exit immediately. Obviously, this can potentially lead to missed entries.

+

The following rules apply, and entry signals will be ignored if more than one of the 3 signals is set:

+
    +
  • enter_long -> exit_long, enter_short
  • +
  • enter_short -> exit_short, enter_long
  • +
+

Further strategy ideas

+

To get additional Ideas for strategies, head over to the strategy repository. Feel free to use them as they are - but results will depend on the current market situation, pairs used etc. - therefore please backtest the strategy for your exchange/desired pairs first, evaluate carefully, use at your own risk. +Feel free to use any of them as inspiration for your own strategies. +We're happy to accept Pull Requests containing new Strategies to that repo.

+

Next step

+

Now you have a perfect strategy you probably want to backtest it. +Your next step is to learn How to use the Backtesting.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/strategy_analysis_example/index.html b/en/2023.4/strategy_analysis_example/index.html new file mode 100644 index 000000000..329892dd4 --- /dev/null +++ b/en/2023.4/strategy_analysis_example/index.html @@ -0,0 +1,1673 @@ + + + + + + + + + + + + + + + + + + + + + + + + Strategy analysis - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Strategy analysis example

+

Debugging a strategy can be time-consuming. Freqtrade offers helper functions to visualize raw data. +The following assumes you work with SampleStrategy, data for 5m timeframe from Binance and have downloaded them into the data directory in the default location. +Please follow the documentation for more details.

+

Setup

+

Change Working directory to repository root

+
import os
+from pathlib import Path
+
+# Change directory
+# Modify this cell to insure that the output shows the correct path.
+# Define all paths relative to the project root shown in the cell output
+project_root = "somedir/freqtrade"
+i=0
+try:
+    os.chdirdir(project_root)
+    assert Path('LICENSE').is_file()
+except:
+    while i<4 and (not Path('LICENSE').is_file()):
+        os.chdir(Path(Path.cwd(), '../'))
+        i+=1
+    project_root = Path.cwd()
+print(Path.cwd())
+
+

Configure Freqtrade environment

+
from freqtrade.configuration import Configuration
+
+# Customize these according to your needs.
+
+# Initialize empty configuration object
+config = Configuration.from_files([])
+# Optionally (recommended), use existing configuration file
+# config = Configuration.from_files(["user_data/config.json"])
+
+# Define some constants
+config["timeframe"] = "5m"
+# Name of the strategy class
+config["strategy"] = "SampleStrategy"
+# Location of the data
+data_location = config["datadir"]
+# Pair to analyze - Only use one pair here
+pair = "BTC/USDT"
+
+
# Load data using values set above
+from freqtrade.data.history import load_pair_history
+from freqtrade.enums import CandleType
+
+candles = load_pair_history(datadir=data_location,
+                            timeframe=config["timeframe"],
+                            pair=pair,
+                            data_format = "json",  # Make sure to update this to your data
+                            candle_type=CandleType.SPOT,
+                            )
+
+# Confirm success
+print(f"Loaded {len(candles)} rows of data for {pair} from {data_location}")
+candles.head()
+
+

Load and run strategy

+
    +
  • Rerun each time the strategy file is changed
  • +
+
# Load strategy using values set above
+from freqtrade.resolvers import StrategyResolver
+from freqtrade.data.dataprovider import DataProvider
+strategy = StrategyResolver.load_strategy(config)
+strategy.dp = DataProvider(config, None, None)
+strategy.ft_bot_start()
+
+# Generate buy/sell signals using strategy
+df = strategy.analyze_ticker(candles, {'pair': pair})
+df.tail()
+
+

Display the trade details

+
    +
  • Note that using data.head() would also work, however most indicators have some "startup" data at the top of the dataframe.
  • +
  • Some possible problems + * Columns with NaN values at the end of the dataframe + * Columns used in crossed*() functions with completely different units
  • +
  • Comparison with full backtest + * having 200 buy signals as output for one pair from analyze_ticker() does not necessarily mean that 200 trades will be made during backtesting. + * Assuming you use only one condition such as, df['rsi'] < 30 as buy condition, this will generate multiple "buy" signals for each pair in sequence (until rsi returns > 29). The bot will only buy on the first of these signals (and also only if a trade-slot ("max_open_trades") is still available), or on one of the middle signals, as soon as a "slot" becomes available.
  • +
+
# Report results
+print(f"Generated {df['enter_long'].sum()} entry signals")
+data = df.set_index('date', drop=False)
+data.tail()
+
+

Load existing objects into a Jupyter notebook

+

The following cells assume that you have already generated data using the cli.
+They will allow you to drill deeper into your results, and perform analysis which otherwise would make the output very difficult to digest due to information overload.

+

Load backtest results to pandas dataframe

+

Analyze a trades dataframe (also used below for plotting)

+
from freqtrade.data.btanalysis import load_backtest_data, load_backtest_stats
+
+# if backtest_dir points to a directory, it'll automatically load the last backtest file.
+backtest_dir = config["user_data_dir"] / "backtest_results"
+# backtest_dir can also point to a specific file 
+# backtest_dir = config["user_data_dir"] / "backtest_results/backtest-result-2020-07-01_20-04-22.json"
+
+
# You can get the full backtest statistics by using the following command.
+# This contains all information used to generate the backtest result.
+stats = load_backtest_stats(backtest_dir)
+
+strategy = 'SampleStrategy'
+# All statistics are available per strategy, so if `--strategy-list` was used during backtest, this will be reflected here as well.
+# Example usages:
+print(stats['strategy'][strategy]['results_per_pair'])
+# Get pairlist used for this backtest
+print(stats['strategy'][strategy]['pairlist'])
+# Get market change (average change of all pairs from start to end of the backtest period)
+print(stats['strategy'][strategy]['market_change'])
+# Maximum drawdown ()
+print(stats['strategy'][strategy]['max_drawdown'])
+# Maximum drawdown start and end
+print(stats['strategy'][strategy]['drawdown_start'])
+print(stats['strategy'][strategy]['drawdown_end'])
+
+
+# Get strategy comparison (only relevant if multiple strategies were compared)
+print(stats['strategy_comparison'])
+
+
# Load backtested trades as dataframe
+trades = load_backtest_data(backtest_dir)
+
+# Show value-counts per pair
+trades.groupby("pair")["exit_reason"].value_counts()
+
+

Plotting daily profit / equity line

+
# Plotting equity line (starting with 0 on day 1 and adding daily profit for each backtested day)
+
+from freqtrade.configuration import Configuration
+from freqtrade.data.btanalysis import load_backtest_data, load_backtest_stats
+import plotly.express as px
+import pandas as pd
+
+# strategy = 'SampleStrategy'
+# config = Configuration.from_files(["user_data/config.json"])
+# backtest_dir = config["user_data_dir"] / "backtest_results"
+
+stats = load_backtest_stats(backtest_dir)
+strategy_stats = stats['strategy'][strategy]
+
+dates = []
+profits = []
+for date_profit in strategy_stats['daily_profit']:
+    dates.append(date_profit[0])
+    profits.append(date_profit[1])
+
+equity = 0
+equity_daily = []
+for daily_profit in profits:
+    equity_daily.append(equity)
+    equity += float(daily_profit)
+
+
+df = pd.DataFrame({'dates': dates,'equity_daily': equity_daily})
+
+fig = px.line(df, x="dates", y="equity_daily")
+fig.show()
+
+

Load live trading results into a pandas dataframe

+

In case you did already some trading and want to analyze your performance

+
from freqtrade.data.btanalysis import load_trades_from_db
+
+# Fetch trades from database
+trades = load_trades_from_db("sqlite:///tradesv3.sqlite")
+
+# Display results
+trades.groupby("pair")["exit_reason"].value_counts()
+
+

Analyze the loaded trades for trade parallelism

+

This can be useful to find the best max_open_trades parameter, when used with backtesting in conjunction with --disable-max-market-positions.

+

analyze_trade_parallelism() returns a timeseries dataframe with an "open_trades" column, specifying the number of open trades for each candle.

+
from freqtrade.data.btanalysis import analyze_trade_parallelism
+
+# Analyze the above
+parallel_trades = analyze_trade_parallelism(trades, '5m')
+
+parallel_trades.plot()
+
+

Plot results

+

Freqtrade offers interactive plotting capabilities based on plotly.

+
from freqtrade.plot.plotting import  generate_candlestick_graph
+# Limit graph period to keep plotly quick and reactive
+
+# Filter trades to one pair
+trades_red = trades.loc[trades['pair'] == pair]
+
+data_red = data['2019-06-01':'2019-06-10']
+# Generate candlestick graph
+graph = generate_candlestick_graph(pair=pair,
+                                   data=data_red,
+                                   trades=trades_red,
+                                   indicators1=['sma20', 'ema50', 'ema55'],
+                                   indicators2=['rsi', 'macd', 'macdsignal', 'macdhist']
+                                  )
+
+
# Show graph inline
+# graph.show()
+
+# Render graph in a separate window
+graph.show(renderer="browser")
+
+

Plot average profit per trade as distribution graph

+
import plotly.figure_factory as ff
+
+hist_data = [trades.profit_ratio]
+group_labels = ['profit_ratio']  # name of the dataset
+
+fig = ff.create_distplot(hist_data, group_labels, bin_size=0.01)
+fig.show()
+
+

Feel free to submit an issue or Pull Request enhancing this document if you would like to share ideas on how to best analyze the data.

+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/strategy_migration/index.html b/en/2023.4/strategy_migration/index.html new file mode 100644 index 000000000..5388cc79e --- /dev/null +++ b/en/2023.4/strategy_migration/index.html @@ -0,0 +1,2424 @@ + + + + + + + + + + + + + + + + + + + + + + + + Strategy migration - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Strategy Migration between V2 and V3

+

To support new markets and trade-types (namely short trades / trades with leverage), some things had to change in the interface. +If you intend on using markets other than spot markets, please migrate your strategy to the new format.

+

We have put a great effort into keeping compatibility with existing strategies, so if you just want to continue using freqtrade in spot markets, there should be no changes necessary for now.

+

You can use the quick summary as checklist. Please refer to the detailed sections below for full migration details.

+

Quick summary / migration checklist

+

Note : forcesell, forcebuy, emergencysell are changed to force_exit, force_enter, emergency_exit respectively.

+ +

Extensive explanation

+

populate_buy_trend

+

In populate_buy_trend() - you will want to change the columns you assign from 'buy' to 'enter_long', as well as the method name from populate_buy_trend to populate_entry_trend.

+
def populate_buy_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+    dataframe.loc[
+        (
+            (qtpylib.crossed_above(dataframe['rsi'], 30)) &  # Signal: RSI crosses above 30
+            (dataframe['tema'] <= dataframe['bb_middleband']) &  # Guard
+            (dataframe['tema'] > dataframe['tema'].shift(1)) &  # Guard
+            (dataframe['volume'] > 0)  # Make sure Volume is not 0
+        ),
+        ['buy', 'buy_tag']] = (1, 'rsi_cross')
+
+    return dataframe
+
+

After:

+
def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+    dataframe.loc[
+        (
+            (qtpylib.crossed_above(dataframe['rsi'], 30)) &  # Signal: RSI crosses above 30
+            (dataframe['tema'] <= dataframe['bb_middleband']) &  # Guard
+            (dataframe['tema'] > dataframe['tema'].shift(1)) &  # Guard
+            (dataframe['volume'] > 0)  # Make sure Volume is not 0
+        ),
+        ['enter_long', 'enter_tag']] = (1, 'rsi_cross')
+
+    return dataframe
+
+

Please refer to the Strategy documentation on how to enter and exit short trades.

+

populate_sell_trend

+

Similar to populate_buy_trend, populate_sell_trend() will be renamed to populate_exit_trend(). +We'll also change the column from 'sell' to 'exit_long'.

+
def populate_sell_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+    dataframe.loc[
+        (
+            (qtpylib.crossed_above(dataframe['rsi'], 70)) &  # Signal: RSI crosses above 70
+            (dataframe['tema'] > dataframe['bb_middleband']) &  # Guard
+            (dataframe['tema'] < dataframe['tema'].shift(1)) &  # Guard
+            (dataframe['volume'] > 0)  # Make sure Volume is not 0
+        ),
+        ['sell', 'exit_tag']] = (1, 'some_exit_tag')
+    return dataframe
+
+

After

+
def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+    dataframe.loc[
+        (
+            (qtpylib.crossed_above(dataframe['rsi'], 70)) &  # Signal: RSI crosses above 70
+            (dataframe['tema'] > dataframe['bb_middleband']) &  # Guard
+            (dataframe['tema'] < dataframe['tema'].shift(1)) &  # Guard
+            (dataframe['volume'] > 0)  # Make sure Volume is not 0
+        ),
+        ['exit_long', 'exit_tag']] = (1, 'some_exit_tag')
+    return dataframe
+
+

Please refer to the Strategy documentation on how to enter and exit short trades.

+

custom_sell

+

custom_sell has been renamed to custom_exit. +It's now also being called for every iteration, independent of current profit and exit_profit_only settings.

+
class AwesomeStrategy(IStrategy):
+    def custom_sell(self, pair: str, trade: 'Trade', current_time: 'datetime', current_rate: float,
+                    current_profit: float, **kwargs):
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        last_candle = dataframe.iloc[-1].squeeze()
+        # ...
+
+
class AwesomeStrategy(IStrategy):
+    def custom_exit(self, pair: str, trade: 'Trade', current_time: 'datetime', current_rate: float,
+                    current_profit: float, **kwargs):
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        last_candle = dataframe.iloc[-1].squeeze()
+        # ...
+
+

custom_entry_timeout

+

check_buy_timeout() has been renamed to check_entry_timeout(), and check_sell_timeout() has been renamed to check_exit_timeout().

+
class AwesomeStrategy(IStrategy):
+    def check_buy_timeout(self, pair: str, trade: 'Trade', order: dict, 
+                            current_time: datetime, **kwargs) -> bool:
+        return False
+
+    def check_sell_timeout(self, pair: str, trade: 'Trade', order: dict, 
+                            current_time: datetime, **kwargs) -> bool:
+        return False 
+
+
class AwesomeStrategy(IStrategy):
+    def check_entry_timeout(self, pair: str, trade: 'Trade', order: 'Order', 
+                            current_time: datetime, **kwargs) -> bool:
+        return False
+
+    def check_exit_timeout(self, pair: str, trade: 'Trade', order: 'Order', 
+                            current_time: datetime, **kwargs) -> bool:
+        return False 
+
+

custom_stake_amount

+

New string argument side - which can be either "long" or "short".

+
class AwesomeStrategy(IStrategy):
+    def custom_stake_amount(self, pair: str, current_time: datetime, current_rate: float,
+                            proposed_stake: float, min_stake: Optional[float], max_stake: float,
+                            entry_tag: Optional[str], **kwargs) -> float:
+        # ... 
+        return proposed_stake
+
+
class AwesomeStrategy(IStrategy):
+    def custom_stake_amount(self, pair: str, current_time: datetime, current_rate: float,
+                            proposed_stake: float, min_stake: Optional[float], max_stake: float,
+                            entry_tag: Optional[str], side: str, **kwargs) -> float:
+        # ... 
+        return proposed_stake
+
+

confirm_trade_entry

+

New string argument side - which can be either "long" or "short".

+
class AwesomeStrategy(IStrategy):
+    def confirm_trade_entry(self, pair: str, order_type: str, amount: float, rate: float,
+                            time_in_force: str, current_time: datetime, entry_tag: Optional[str], 
+                            **kwargs) -> bool:
+      return True
+
+

After:

+
class AwesomeStrategy(IStrategy):
+    def confirm_trade_entry(self, pair: str, order_type: str, amount: float, rate: float,
+                            time_in_force: str, current_time: datetime, entry_tag: Optional[str], 
+                            side: str, **kwargs) -> bool:
+      return True
+
+

confirm_trade_exit

+

Changed argument sell_reason to exit_reason. +For compatibility, sell_reason will still be provided for a limited time.

+
class AwesomeStrategy(IStrategy):
+    def confirm_trade_exit(self, pair: str, trade: Trade, order_type: str, amount: float,
+                           rate: float, time_in_force: str, sell_reason: str,
+                           current_time: datetime, **kwargs) -> bool:
+    return True
+
+

After:

+
class AwesomeStrategy(IStrategy):
+    def confirm_trade_exit(self, pair: str, trade: Trade, order_type: str, amount: float,
+                           rate: float, time_in_force: str, exit_reason: str,
+                           current_time: datetime, **kwargs) -> bool:
+    return True
+
+

custom_entry_price

+

New string argument side - which can be either "long" or "short".

+
class AwesomeStrategy(IStrategy):
+    def custom_entry_price(self, pair: str, current_time: datetime, proposed_rate: float,
+                           entry_tag: Optional[str], **kwargs) -> float:
+      return proposed_rate
+
+

After:

+
class AwesomeStrategy(IStrategy):
+    def custom_entry_price(self, pair: str, current_time: datetime, proposed_rate: float,
+                           entry_tag: Optional[str], side: str, **kwargs) -> float:
+      return proposed_rate
+
+

Adjust trade position changes

+

While adjust-trade-position itself did not change, you should no longer use trade.nr_of_successful_buys - and instead use trade.nr_of_successful_entries, which will also include short entries.

+

Helper methods

+

Added argument "is_short" to stoploss_from_open and stoploss_from_absolute. +This should be given the value of trade.is_short.

+
    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,
+                        current_rate: float, current_profit: float, **kwargs) -> float:
+        # once the profit has risen above 10%, keep the stoploss at 7% above the open price
+        if current_profit > 0.10:
+            return stoploss_from_open(0.07, current_profit)
+
+        return stoploss_from_absolute(current_rate - (candle['atr'] * 2), current_rate)
+
+        return 1
+
+

After:

+
    def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime,
+                        current_rate: float, current_profit: float, **kwargs) -> float:
+        # once the profit has risen above 10%, keep the stoploss at 7% above the open price
+        if current_profit > 0.10:
+            return stoploss_from_open(0.07, current_profit, is_short=trade.is_short)
+
+        return stoploss_from_absolute(current_rate - (candle['atr'] * 2), current_rate, is_short=trade.is_short)
+
+

Strategy/Configuration settings

+

order_time_in_force

+

order_time_in_force attributes changed from "buy" to "entry" and "sell" to "exit".

+
    order_time_in_force: Dict = {
+        "buy": "gtc",
+        "sell": "gtc",
+    }
+
+

After:

+
    order_time_in_force: Dict = {
+        "entry": "GTC",
+        "exit": "GTC",
+    }
+
+

order_types

+

order_types have changed all wordings from buy to entry - and sell to exit. +And two words are joined with _.

+
    order_types = {
+        "buy": "limit",
+        "sell": "limit",
+        "emergencysell": "market",
+        "forcesell": "market",
+        "forcebuy": "market",
+        "stoploss": "market",
+        "stoploss_on_exchange": false,
+        "stoploss_on_exchange_interval": 60
+    }
+
+

After:

+
    order_types = {
+        "entry": "limit",
+        "exit": "limit",
+        "emergency_exit": "market",
+        "force_exit": "market",
+        "force_entry": "market",
+        "stoploss": "market",
+        "stoploss_on_exchange": false,
+        "stoploss_on_exchange_interval": 60
+    }
+
+

Strategy level settings

+
    +
  • use_sell_signal -> use_exit_signal
  • +
  • sell_profit_only -> exit_profit_only
  • +
  • sell_profit_offset -> exit_profit_offset
  • +
  • ignore_roi_if_buy_signal -> ignore_roi_if_entry_signal
  • +
+
    # These values can be overridden in the config.
+    use_sell_signal = True
+    sell_profit_only = True
+    sell_profit_offset: 0.01
+    ignore_roi_if_buy_signal = False
+
+

After:

+
    # These values can be overridden in the config.
+    use_exit_signal = True
+    exit_profit_only = True
+    exit_profit_offset: 0.01
+    ignore_roi_if_entry_signal = False
+
+

unfilledtimeout

+

unfilledtimeout have changed all wordings from buy to entry - and sell to exit.

+
unfilledtimeout = {
+        "buy": 10,
+        "sell": 10,
+        "exit_timeout_count": 0,
+        "unit": "minutes"
+    }
+
+

After:

+
unfilledtimeout = {
+        "entry": 10,
+        "exit": 10,
+        "exit_timeout_count": 0,
+        "unit": "minutes"
+    }
+
+

order pricing

+

Order pricing changed in 2 ways. bid_strategy was renamed to entry_pricing and ask_strategy was renamed to exit_pricing. +The attributes ask_last_balance -> price_last_balance and bid_last_balance -> price_last_balance were renamed as well. +Also, price-side can now be defined as ask, bid, same or other. +Please refer to the pricing documentation for more information.

+
{
+    "bid_strategy": {
+        "price_side": "bid",
+        "use_order_book": true,
+        "order_book_top": 1,
+        "ask_last_balance": 0.0,
+        "check_depth_of_market": {
+            "enabled": false,
+            "bids_to_ask_delta": 1
+        }
+    },
+    "ask_strategy":{
+        "price_side": "ask",
+        "use_order_book": true,
+        "order_book_top": 1,
+        "bid_last_balance": 0.0
+        "ignore_buying_expired_candle_after": 120
+    }
+}
+
+

after:

+
{
+    "entry_pricing": {
+        "price_side": "same",
+        "use_order_book": true,
+        "order_book_top": 1,
+        "price_last_balance": 0.0,
+        "check_depth_of_market": {
+            "enabled": false,
+            "bids_to_ask_delta": 1
+        }
+    },
+    "exit_pricing":{
+        "price_side": "same",
+        "use_order_book": true,
+        "order_book_top": 1,
+        "price_last_balance": 0.0
+    },
+    "ignore_buying_expired_candle_after": 120
+}
+
+

FreqAI strategy

+

The populate_any_indicators() method has been split into feature_engineering_expand_all(), feature_engineering_expand_basic(), feature_engineering_standard() andset_freqai_targets().

+

For each new function, the pair (and timeframe where necessary) will be automatically added to the column. +As such, the definition of features becomes much simpler with the new logic.

+

For a full explanation of each method, please go to the corresponding freqAI documentation page

+
 1
+ 2
+ 3
+ 4
+ 5
+ 6
+ 7
+ 8
+ 9
+10
+11
+12
+13
+14
+15
+16
+17
+18
+19
+20
+21
+22
+23
+24
+25
+26
+27
+28
+29
+30
+31
+32
+33
+34
+35
+36
+37
+38
+39
+40
+41
+42
+43
+44
+45
+46
+47
+48
+49
+50
+51
+52
+53
+54
+55
+56
+57
+58
+59
+60
+61
+62
+63
+64
+65
+66
+67
+68
+69
+70
+71
+72
+73
+74
+75
+76
+77
def populate_any_indicators(
+        self, pair, df, tf, informative=None, set_generalized_indicators=False
+    ):
+
+        if informative is None:
+            informative = self.dp.get_pair_dataframe(pair, tf)
+
+        # first loop is automatically duplicating indicators for time periods
+        for t in self.freqai_info["feature_parameters"]["indicator_periods_candles"]:
+
+            t = int(t)
+            informative[f"%-{pair}rsi-period_{t}"] = ta.RSI(informative, timeperiod=t)
+            informative[f"%-{pair}mfi-period_{t}"] = ta.MFI(informative, timeperiod=t)
+            informative[f"%-{pair}adx-period_{t}"] = ta.ADX(informative, timeperiod=t)
+            informative[f"%-{pair}sma-period_{t}"] = ta.SMA(informative, timeperiod=t)
+            informative[f"%-{pair}ema-period_{t}"] = ta.EMA(informative, timeperiod=t)
+
+            bollinger = qtpylib.bollinger_bands(
+                qtpylib.typical_price(informative), window=t, stds=2.2
+            )
+            informative[f"{pair}bb_lowerband-period_{t}"] = bollinger["lower"]
+            informative[f"{pair}bb_middleband-period_{t}"] = bollinger["mid"]
+            informative[f"{pair}bb_upperband-period_{t}"] = bollinger["upper"]
+
+            informative[f"%-{pair}bb_width-period_{t}"] = (
+                informative[f"{pair}bb_upperband-period_{t}"]
+                - informative[f"{pair}bb_lowerband-period_{t}"]
+            ) / informative[f"{pair}bb_middleband-period_{t}"]
+            informative[f"%-{pair}close-bb_lower-period_{t}"] = (
+                informative["close"] / informative[f"{pair}bb_lowerband-period_{t}"]
+            )
+
+            informative[f"%-{pair}roc-period_{t}"] = ta.ROC(informative, timeperiod=t)
+
+            informative[f"%-{pair}relative_volume-period_{t}"] = (
+                informative["volume"] / informative["volume"].rolling(t).mean()
+            ) # (1)
+
+        informative[f"%-{pair}pct-change"] = informative["close"].pct_change()
+        informative[f"%-{pair}raw_volume"] = informative["volume"]
+        informative[f"%-{pair}raw_price"] = informative["close"]
+        # (2)
+
+        indicators = [col for col in informative if col.startswith("%")]
+        # This loop duplicates and shifts all indicators to add a sense of recency to data
+        for n in range(self.freqai_info["feature_parameters"]["include_shifted_candles"] + 1):
+            if n == 0:
+                continue
+            informative_shift = informative[indicators].shift(n)
+            informative_shift = informative_shift.add_suffix("_shift-" + str(n))
+            informative = pd.concat((informative, informative_shift), axis=1)
+
+        df = merge_informative_pair(df, informative, self.config["timeframe"], tf, ffill=True)
+        skip_columns = [
+            (s + "_" + tf) for s in ["date", "open", "high", "low", "close", "volume"]
+        ]
+        df = df.drop(columns=skip_columns)
+
+        # Add generalized indicators here (because in live, it will call this
+        # function to populate indicators during training). Notice how we ensure not to
+        # add them multiple times
+        if set_generalized_indicators:
+            df["%-day_of_week"] = (df["date"].dt.dayofweek + 1) / 7
+            df["%-hour_of_day"] = (df["date"].dt.hour + 1) / 25
+            # (3)
+
+            # user adds targets here by prepending them with &- (see convention below)
+            df["&-s_close"] = (
+                df["close"]
+                .shift(-self.freqai_info["feature_parameters"]["label_period_candles"])
+                .rolling(self.freqai_info["feature_parameters"]["label_period_candles"])
+                .mean()
+                / df["close"]
+                - 1
+            )  # (4)
+
+        return df
+
+
    +
  1. Features - Move to feature_engineering_expand_all
  2. +
  3. Basic features, not expanded across include_periods_candles - move tofeature_engineering_expand_basic().
  4. +
  5. Standard features which should not be expanded - move to feature_engineering_standard().
  6. +
  7. Targets - Move this part to set_freqai_targets().
  8. +
+

freqai - feature engineering expand all

+

Features will now expand automatically. As such, the expansion loops, as well as the {pair} / {timeframe} parts will need to be removed.

+
 1
+ 2
+ 3
+ 4
+ 5
+ 6
+ 7
+ 8
+ 9
+10
+11
+12
+13
+14
+15
+16
+17
+18
+19
+20
+21
+22
+23
+24
+25
+26
+27
+28
+29
+30
+31
+32
+33
+34
+35
+36
+37
+38
+39
+40
+41
+42
+43
+44
+45
+46
+47
+48
+49
+50
+51
+52
    def feature_engineering_expand_all(self, dataframe, period, **kwargs) -> DataFrame::
+        """
+        *Only functional with FreqAI enabled strategies*
+        This function will automatically expand the defined features on the config defined
+        `indicator_periods_candles`, `include_timeframes`, `include_shifted_candles`, and
+        `include_corr_pairs`. In other words, a single feature defined in this function
+        will automatically expand to a total of
+        `indicator_periods_candles` * `include_timeframes` * `include_shifted_candles` *
+        `include_corr_pairs` numbers of features added to the model.
+
+        All features must be prepended with `%` to be recognized by FreqAI internals.
+
+        More details on how these config defined parameters accelerate feature engineering
+        in the documentation at:
+
+        https://www.freqtrade.io/en/latest/freqai-parameter-table/#feature-parameters
+
+        https://www.freqtrade.io/en/latest/freqai-feature-engineering/#defining-the-features
+
+        :param df: strategy dataframe which will receive the features
+        :param period: period of the indicator - usage example:
+        dataframe["%-ema-period"] = ta.EMA(dataframe, timeperiod=period)
+        """
+
+        dataframe["%-rsi-period"] = ta.RSI(dataframe, timeperiod=period)
+        dataframe["%-mfi-period"] = ta.MFI(dataframe, timeperiod=period)
+        dataframe["%-adx-period"] = ta.ADX(dataframe, timeperiod=period)
+        dataframe["%-sma-period"] = ta.SMA(dataframe, timeperiod=period)
+        dataframe["%-ema-period"] = ta.EMA(dataframe, timeperiod=period)
+
+        bollinger = qtpylib.bollinger_bands(
+            qtpylib.typical_price(dataframe), window=period, stds=2.2
+        )
+        dataframe["bb_lowerband-period"] = bollinger["lower"]
+        dataframe["bb_middleband-period"] = bollinger["mid"]
+        dataframe["bb_upperband-period"] = bollinger["upper"]
+
+        dataframe["%-bb_width-period"] = (
+            dataframe["bb_upperband-period"]
+            - dataframe["bb_lowerband-period"]
+        ) / dataframe["bb_middleband-period"]
+        dataframe["%-close-bb_lower-period"] = (
+            dataframe["close"] / dataframe["bb_lowerband-period"]
+        )
+
+        dataframe["%-roc-period"] = ta.ROC(dataframe, timeperiod=period)
+
+        dataframe["%-relative_volume-period"] = (
+            dataframe["volume"] / dataframe["volume"].rolling(period).mean()
+        )
+
+        return dataframe
+
+

Freqai - feature engineering basic

+

Basic features. Make sure to remove the {pair} part from your features.

+
 1
+ 2
+ 3
+ 4
+ 5
+ 6
+ 7
+ 8
+ 9
+10
+11
+12
+13
+14
+15
+16
+17
+18
+19
+20
+21
+22
+23
+24
+25
+26
+27
+28
+29
+30
    def feature_engineering_expand_basic(self, dataframe: DataFrame, **kwargs) -> DataFrame::
+        """
+        *Only functional with FreqAI enabled strategies*
+        This function will automatically expand the defined features on the config defined
+        `include_timeframes`, `include_shifted_candles`, and `include_corr_pairs`.
+        In other words, a single feature defined in this function
+        will automatically expand to a total of
+        `include_timeframes` * `include_shifted_candles` * `include_corr_pairs`
+        numbers of features added to the model.
+
+        Features defined here will *not* be automatically duplicated on user defined
+        `indicator_periods_candles`
+
+        All features must be prepended with `%` to be recognized by FreqAI internals.
+
+        More details on how these config defined parameters accelerate feature engineering
+        in the documentation at:
+
+        https://www.freqtrade.io/en/latest/freqai-parameter-table/#feature-parameters
+
+        https://www.freqtrade.io/en/latest/freqai-feature-engineering/#defining-the-features
+
+        :param df: strategy dataframe which will receive the features
+        dataframe["%-pct-change"] = dataframe["close"].pct_change()
+        dataframe["%-ema-200"] = ta.EMA(dataframe, timeperiod=200)
+        """
+        dataframe["%-pct-change"] = dataframe["close"].pct_change()
+        dataframe["%-raw_volume"] = dataframe["volume"]
+        dataframe["%-raw_price"] = dataframe["close"]
+        return dataframe
+
+

FreqAI - feature engineering standard

+
 1
+ 2
+ 3
+ 4
+ 5
+ 6
+ 7
+ 8
+ 9
+10
+11
+12
+13
+14
+15
+16
+17
+18
+19
+20
+21
+22
+23
+24
    def feature_engineering_standard(self, dataframe: DataFrame, **kwargs) -> DataFrame:
+        """
+        *Only functional with FreqAI enabled strategies*
+        This optional function will be called once with the dataframe of the base timeframe.
+        This is the final function to be called, which means that the dataframe entering this
+        function will contain all the features and columns created by all other
+        freqai_feature_engineering_* functions.
+
+        This function is a good place to do custom exotic feature extractions (e.g. tsfresh).
+        This function is a good place for any feature that should not be auto-expanded upon
+        (e.g. day of the week).
+
+        All features must be prepended with `%` to be recognized by FreqAI internals.
+
+        More details about feature engineering available:
+
+        https://www.freqtrade.io/en/latest/freqai-feature-engineering
+
+        :param df: strategy dataframe which will receive the features
+        usage example: dataframe["%-day_of_week"] = (dataframe["date"].dt.dayofweek + 1) / 7
+        """
+        dataframe["%-day_of_week"] = dataframe["date"].dt.dayofweek
+        dataframe["%-hour_of_day"] = dataframe["date"].dt.hour
+        return dataframe
+
+

FreqAI - set Targets

+

Targets now get their own, dedicated method.

+
 1
+ 2
+ 3
+ 4
+ 5
+ 6
+ 7
+ 8
+ 9
+10
+11
+12
+13
+14
+15
+16
+17
+18
+19
+20
+21
+22
+23
    def set_freqai_targets(self, dataframe: DataFrame, **kwargs) -> DataFrame:
+        """
+        *Only functional with FreqAI enabled strategies*
+        Required function to set the targets for the model.
+        All targets must be prepended with `&` to be recognized by the FreqAI internals.
+
+        More details about feature engineering available:
+
+        https://www.freqtrade.io/en/latest/freqai-feature-engineering
+
+        :param df: strategy dataframe which will receive the targets
+        usage example: dataframe["&-target"] = dataframe["close"].shift(-1) / dataframe["close"]
+        """
+        dataframe["&-s_close"] = (
+            dataframe["close"]
+            .shift(-self.freqai_info["feature_parameters"]["label_period_candles"])
+            .rolling(self.freqai_info["feature_parameters"]["label_period_candles"])
+            .mean()
+            / dataframe["close"]
+            - 1
+            )
+
+        return dataframe
+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/stylesheets/ft.extra.css b/en/2023.4/stylesheets/ft.extra.css new file mode 100644 index 000000000..3369fa177 --- /dev/null +++ b/en/2023.4/stylesheets/ft.extra.css @@ -0,0 +1,13 @@ +.rst-versions { + font-size: .7rem; + color: white; +} + +.rst-versions.rst-badge .rst-current-version { + font-size: .7rem; + color: white; +} + +.rst-versions .rst-other-versions { + color: white; +} diff --git a/en/2023.4/telegram-usage/index.html b/en/2023.4/telegram-usage/index.html new file mode 100644 index 000000000..8bcb1ccc4 --- /dev/null +++ b/en/2023.4/telegram-usage/index.html @@ -0,0 +1,2191 @@ + + + + + + + + + + + + + + + + + + + + + + + + Telegram - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + + + + + +
+
+ + + + + + + +

Telegram usage

+

Setup your Telegram bot

+

Below we explain how to create your Telegram Bot, and how to get your +Telegram user id.

+

1. Create your Telegram bot

+

Start a chat with the Telegram BotFather

+

Send the message /newbot.

+

BotFather response:

+
+

Alright, a new bot. How are we going to call it? Please choose a name for your bot.

+
+

Choose the public name of your bot (e.x. Freqtrade bot)

+

BotFather response:

+
+

Good. Now let's choose a username for your bot. It must end in bot. Like this, for example: TetrisBot or tetris_bot.

+
+

Choose the name id of your bot and send it to the BotFather (e.g. "My_own_freqtrade_bot")

+

BotFather response:

+
+

Done! Congratulations on your new bot. You will find it at t.me/yourbots_name_bot. You can now add a description, about section and profile picture for your bot, see /help for a list of commands. By the way, when you've finished creating your cool bot, ping our Bot Support if you want a better username for it. Just make sure the bot is fully operational before you do this.

+

Use this token to access the HTTP API: 22222222:APITOKEN

+

For a description of the Bot API, see this page: https://core.telegram.org/bots/api Father bot will return you the token (API key)

+
+

Copy the API Token (22222222:APITOKEN in the above example) and keep use it for the config parameter token.

+

Don't forget to start the conversation with your bot, by clicking /START button

+

2. Telegram user_id

+

Get your user id

+

Talk to the userinfobot

+

Get your "Id", you will use it for the config parameter chat_id.

+

Use Group id

+

You can use bots in telegram groups by just adding them to the group. You can find the group id by first adding a RawDataBot to your group. The Group id is shown as id in the "chat" section, which the RawDataBot will send to you:

+
"chat":{
+   "id":-1001332619709
+}
+
+

For the Freqtrade configuration, you can then use the the full value (including - if it's there) as string:

+
   "chat_id": "-1001332619709"
+
+
+

Using telegram groups

+

When using telegram groups, you're giving every member of the telegram group access to your freqtrade bot and to all commands possible via telegram. Please make sure that you can trust everyone in the telegram group to avoid unpleasent surprises.

+
+

Control telegram noise

+

Freqtrade provides means to control the verbosity of your telegram bot. +Each setting has the following possible values:

+
    +
  • on - Messages will be sent, and user will be notified.
  • +
  • silent - Message will be sent, Notification will be without sound / vibration.
  • +
  • off - Skip sending a message-type all together.
  • +
+

Example configuration showing the different settings:

+
"telegram": {
+    "enabled": true,
+    "token": "your_telegram_token",
+    "chat_id": "your_telegram_chat_id",
+    "allow_custom_messages": true,
+    "notification_settings": {
+        "status": "silent",
+        "warning": "on",
+        "startup": "off",
+        "entry": "silent",
+        "entry_fill": "on",
+        "entry_cancel": "silent",
+        "exit": {
+            "roi": "silent",
+            "emergency_exit": "on",
+            "force_exit": "on",
+            "exit_signal": "silent",
+            "trailing_stop_loss": "on",
+            "stop_loss": "on",
+            "stoploss_on_exchange": "on",
+            "custom_exit": "silent",
+            "partial_exit": "on"
+        },
+        "exit_cancel": "on",
+        "exit_fill": "off",
+        "protection_trigger": "off",
+        "protection_trigger_global": "on",
+        "strategy_msg": "off",
+        "show_candle": "off"
+    },
+    "reload": true,
+    "balance_dust_level": 0.01
+},
+
+

entry notifications are sent when the order is placed, while entry_fill notifications are sent when the order is filled on the exchange. +exit notifications are sent when the order is placed, while exit_fill notifications are sent when the order is filled on the exchange. +*_fill notifications are off by default and must be explicitly enabled. +protection_trigger notifications are sent when a protection triggers and protection_trigger_global notifications trigger when global protections are triggered. +strategy_msg - Receive notifications from the strategy, sent via self.dp.send_msg() from the strategy more details. +show_candle - show candle values as part of entry/exit messages. Only possible values are "ohlc" or "off".

+

balance_dust_level will define what the /balance command takes as "dust" - Currencies with a balance below this will be shown. +allow_custom_messages completely disable strategy messages. +reload allows you to disable reload-buttons on selected messages.

+

Create a custom keyboard (command shortcut buttons)

+

Telegram allows us to create a custom keyboard with buttons for commands. +The default custom keyboard looks like this.

+
[
+    ["/daily", "/profit", "/balance"], # row 1, 3 commands
+    ["/status", "/status table", "/performance"], # row 2, 3 commands
+    ["/count", "/start", "/stop", "/help"] # row 3, 4 commands
+]
+
+

Usage

+

You can create your own keyboard in config.json:

+
"telegram": {
+      "enabled": true,
+      "token": "your_telegram_token",
+      "chat_id": "your_telegram_chat_id",
+      "keyboard": [
+          ["/daily", "/stats", "/balance", "/profit"],
+          ["/status table", "/performance"],
+          ["/reload_config", "/count", "/logs"]
+      ]
+   },
+
+
+

Supported Commands

+

Only the following commands are allowed. Command arguments are not supported!

+

/start, /stop, /status, /status table, /trades, /profit, /performance, /daily, /stats, /count, /locks, /balance, /stopentry, /reload_config, /show_config, /logs, /whitelist, /blacklist, /edge, /help, /version, /marketdir

+
+

Telegram commands

+

Per default, the Telegram bot shows predefined commands. Some commands +are only available by sending them to the bot. The table below list the +official commands. You can ask at any moment for help with /help.

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
CommandDescription
System commands
/startStarts the trader
/stopStops the trader
/stopbuy | /stopentryStops the trader from opening new trades. Gracefully closes open trades according to their rules.
/reload_configReloads the configuration file
/show_configShows part of the current configuration with relevant settings to operation
/logs [limit]Show last log messages.
/helpShow help message
/versionShow version
Status
/statusLists all open trades
/status <trade_id>Lists one or more specific trade. Separate multiple with a blank space.
/status tableList all open trades in a table format. Pending buy orders are marked with an asterisk () Pending sell orders are marked with a double asterisk (*)
/trades [limit]List all recently closed trades in a table format.
/countDisplays number of trades used and available
/locksShow currently locked pairs.
/unlock <pair or lock_id>Remove the lock for this pair (or for this lock id).
/marketdir [long | short | even | none]Updates the user managed variable that represents the current market direction. If no direction is provided, the currently set direction will be displayed.
Modify Trade states
/forceexit <trade_id> | /fx <tradeid>Instantly exits the given trade (Ignoring minimum_roi).
/forceexit all | /fx allInstantly exits all open trades (Ignoring minimum_roi).
/fxalias for /forceexit
/forcelong <pair> [rate]Instantly buys the given pair. Rate is optional and only applies to limit orders. (force_entry_enable must be set to True)
/forceshort <pair> [rate]Instantly shorts the given pair. Rate is optional and only applies to limit orders. This will only work on non-spot markets. (force_entry_enable must be set to True)
/delete <trade_id>Delete a specific trade from the Database. Tries to close open orders. Requires manual handling of this trade on the exchange.
/cancel_open_order <trade_id> | /coo <trade_id>Cancel an open order for a trade.
Metrics
/profit [<n>]Display a summary of your profit/loss from close trades and some stats about your performance, over the last n days (all trades by default)
/performanceShow performance of each finished trade grouped by pair
/balanceShow account balance per currency
/daily <n>Shows profit or loss per day, over the last n days (n defaults to 7)
/weekly <n>Shows profit or loss per week, over the last n weeks (n defaults to 8)
/monthly <n>Shows profit or loss per month, over the last n months (n defaults to 6)
/statsShows Wins / losses by Exit reason as well as Avg. holding durations for buys and sells
/exitsShows Wins / losses by Exit reason as well as Avg. holding durations for buys and sells
/entriesShows Wins / losses by Exit reason as well as Avg. holding durations for buys and sells
/whitelist [sorted] [baseonly]Show the current whitelist. Optionally display in alphabetical order and/or with just the base currency of each pairing.
/blacklist [pair]Show the current blacklist, or adds a pair to the blacklist.
/edgeShow validated pairs by Edge if it is enabled.
+

Telegram commands in action

+

Below, example of Telegram message you will receive for each command.

+

/start

+
+

Status: running

+
+

/stop

+
+

Stopping trader ... +Status: stopped

+
+

/stopbuy

+
+

status: Setting max_open_trades to 0. Run /reload_config to reset.

+
+

Prevents the bot from opening new trades by temporarily setting "max_open_trades" to 0. Open trades will be handled via their regular rules (ROI / Sell-signal, stoploss, ...).

+

After this, give the bot time to close off open trades (can be checked via /status table). +Once all positions are sold, run /stop to completely stop the bot.

+

/reload_config resets "max_open_trades" to the value set in the configuration and resets this command.

+
+

Warning

+
+

The stop-buy signal is ONLY active while the bot is running, and is not persisted anyway, so restarting the bot will cause this to reset.

+

/status

+

For each open trade, the bot will send you the following message. +Enter Tag is configurable via Strategy.

+
+

Trade ID: 123 (since 1 days ago) +Current Pair: CVC/BTC +Direction: Long +Leverage: 1.0 +Amount: 26.64180098 +Enter Tag: Awesome Long Signal +Open Rate: 0.00007489 +Current Rate: 0.00007489 +Unrealized Profit: 12.95% +Stoploss: 0.00007389 (-0.02%)

+
+

/status table

+

Return the status of all open trades in a table format.

+
ID L/S    Pair     Since   Profit
+----    --------  -------  --------
+  67 L   SC/BTC    1 d      13.33%
+ 123 S   CVC/BTC   1 h      12.95%
+
+

/count

+

Return the number of trades used and available.

+
current    max
+---------  -----
+     2     10
+
+

/profit

+

Return a summary of your profit/loss and performance.

+
+

ROI: Close trades + ∙ 0.00485701 BTC (2.2%) (15.2 Σ%) + ∙ 62.968 USD +ROI: All trades + ∙ 0.00255280 BTC (1.5%) (6.43 Σ%) + ∙ 33.095 EUR

+

Total Trade Count: 138 +Bot started: 2022-07-11 18:40:44 +First Trade opened: 3 days ago +Latest Trade opened: 2 minutes ago +Avg. Duration: 2:33:45 +Best Performing: PAY/BTC: 50.23% +Trading volume: 0.5 BTC +Profit factor: 1.04 +Max Drawdown: 9.23% (0.01255 BTC)

+
+

The relative profit of 1.2% is the average profit per trade. +The relative profit of 15.2 Σ% is be based on the starting capital - so in this case, the starting capital was 0.00485701 * 1.152 = 0.00738 BTC. +Starting capital is either taken from the available_capital setting, or calculated by using current wallet size - profits. +Profit Factor is calculated as gross profits / gross losses - and should serve as an overall metric for the strategy. +Max drawdown corresponds to the backtesting metric Absolute Drawdown (Account) - calculated as (Absolute Drawdown) / (DrawdownHigh + startingBalance). +Bot started date will refer to the date the bot was first started. For older bots, this will default to the first trade's open date.

+

/forceexit

+
+

BINANCE: Exiting BTC/LTC with limit 0.01650000 (profit: ~-4.07%, -0.00008168)

+
+
+

Tip

+

You can get a list of all open trades by calling /forceexit without parameter, which will show a list of buttons to simply exit a trade. +This command has an alias in /fx - which has the same capabilities, but is faster to type in "emergency" situations.

+
+

/forcelong [rate] | /forceshort [rate]

+

/forcebuy <pair> [rate] is also supported for longs but should be considered deprecated.

+
+

BINANCE: Long ETH/BTC with limit 0.03400000 (1.000000 ETH, 225.290 USD)

+
+

Omitting the pair will open a query asking for the pair to trade (based on the current whitelist). +Trades created through /forcelong will have the buy-tag of force_entry.

+

Telegram force-buy screenshot

+

Note that for this to work, force_entry_enable needs to be set to true.

+

More details

+

/performance

+

Return the performance of each crypto-currency the bot has sold.

+
+

Performance: +1. RCN/BTC 0.003 BTC (57.77%) (1) +2. PAY/BTC 0.0012 BTC (56.91%) (1) +3. VIB/BTC 0.0011 BTC (47.07%) (1) +4. SALT/BTC 0.0010 BTC (30.24%) (1) +5. STORJ/BTC 0.0009 BTC (27.24%) (1) +...

+
+

/balance

+

Return the balance of all crypto-currency your have on the exchange.

+
+

Currency: BTC +Available: 3.05890234 +Balance: 3.05890234 +Pending: 0.0

+

Currency: CVC +Available: 86.64180098 +Balance: 86.64180098 +Pending: 0.0

+
+

/daily

+

Per default /daily will return the 7 last days. The example below if for /daily 3:

+
+

Daily Profit over the last 3 days: +

Day (count)     USDT          USD         Profit %
+--------------  ------------  ----------  ----------
+2022-06-11 (1)  -0.746 USDT   -0.75 USD   -0.08%
+2022-06-10 (0)  0 USDT        0.00 USD    0.00%
+2022-06-09 (5)  20 USDT       20.10 USD   5.00%
+

+
+

/weekly

+

Per default /weekly will return the 8 last weeks, including the current week. Each week starts +from Monday. The example below if for /weekly 3:

+
+

Weekly Profit over the last 3 weeks (starting from Monday): +

Monday (count)  Profit BTC      Profit USD   Profit %
+-------------  --------------  ------------    ----------
+2018-01-03 (5)  0.00224175 BTC  29,142 USD   4.98%
+2017-12-27 (1)  0.00033131 BTC   4,307 USD   0.00%
+2017-12-20 (4)  0.00269130 BTC  34.986 USD   5.12%
+

+
+

/monthly

+

Per default /monthly will return the 6 last months, including the current month. The example below +if for /monthly 3:

+
+

Monthly Profit over the last 3 months: +

Month (count)  Profit BTC      Profit USD    Profit %
+-------------  --------------  ------------    ----------
+2018-01 (20)    0.00224175 BTC  29,142 USD  4.98%
+2017-12 (5)    0.00033131 BTC   4,307 USD   0.00%
+2017-11 (10)    0.00269130 BTC  34.986 USD  5.10%
+

+
+

/whitelist

+

Shows the current whitelist

+
+

Using whitelist StaticPairList with 22 pairs +IOTA/BTC, NEO/BTC, TRX/BTC, VET/BTC, ADA/BTC, ETC/BTC, NCASH/BTC, DASH/BTC, XRP/BTC, XVG/BTC, EOS/BTC, LTC/BTC, OMG/BTC, BTG/BTC, LSK/BTC, ZEC/BTC, HOT/BTC, IOTX/BTC, XMR/BTC, AST/BTC, XLM/BTC, NANO/BTC

+
+

/blacklist [pair]

+

Shows the current blacklist. +If Pair is set, then this pair will be added to the pairlist. +Also supports multiple pairs, separated by a space. +Use /reload_config to reset the blacklist.

+
+

Using blacklist StaticPairList with 2 pairs +DODGE/BTC, HOT/BTC.

+
+

/edge

+

Shows pairs validated by Edge along with their corresponding win-rate, expectancy and stoploss values.

+
+

Edge only validated following pairs: +

Pair        Winrate    Expectancy    Stoploss
+--------  ---------  ------------  ----------
+DOCK/ETH   0.522727      0.881821       -0.03
+PHX/ETH    0.677419      0.560488       -0.03
+HOT/ETH    0.733333      0.490492       -0.03
+HC/ETH     0.588235      0.280988       -0.02
+ARDR/ETH   0.366667      0.143059       -0.01
+

+
+

/version

+
+

Version: 0.14.3

+
+

/marketdir

+

If a market direction is provided the command updates the user managed variable that represents the current market direction. +This variable is not set to any valid market direction on bot startup and must be set by the user. The example below is for /marketdir long:

+
Successfully updated marketdirection from none to long.
+
+

If no market direction is provided the command outputs the currently set market directions. The example below is for /marketdir:

+
Currently set marketdirection: even
+
+

You can use the market direction in your strategy via self.market_direction.

+
+

Bot restarts

+

Please note that the market direction is not persisted, and will be reset after a bot restart/reload.

+
+
+

Backtesting

+

As this value/variable is intended to be changed manually in dry/live trading. +Strategies using market_direction will probably not produce reliable, reproducible results (changes to this variable will not be reflected for backtesting). Use at your own risk.

+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/trade-object/index.html b/en/2023.4/trade-object/index.html new file mode 100644 index 000000000..c724a5d60 --- /dev/null +++ b/en/2023.4/trade-object/index.html @@ -0,0 +1,1688 @@ + + + + + + + + + + + + + + + + + + + + + + + + Trade Object - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Trade Object

+

Trade

+

A position freqtrade enters is stored in a Trade object - which is persisted to the database. +It's a core concept of freqtrade - and something you'll come across in many sections of the documentation, which will most likely point you to this location.

+

It will be passed to the strategy in many strategy callbacks. The object passed to the strategy cannot be modified directly. Indirect modifications may occur based on callback results.

+

Trade - Available attributes

+

The following attributes / properties are available for each individual trade - and can be used with trade.<property> (e.g. trade.pair).

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
AttributeDataTypeDescription
pairstringPair of this trade
is_openbooleanIs the trade currently open, or has it been concluded
open_ratefloatRate this trade was entered at (Avg. entry rate in case of trade-adjustments)
close_ratefloatClose rate - only set when is_open = False
stake_amountfloatAmount in Stake (or Quote) currency.
amountfloatAmount in Asset / Base currency that is currently owned.
open_datedatetimeTimestamp when trade was opened use open_date_utc instead
open_date_utcdatetimeTimestamp when trade was opened - in UTC
close_datedatetimeTimestamp when trade was closed use close_date_utc instead
close_date_utcdatetimeTimestamp when trade was closed - in UTC
close_profitfloatRelative profit at the time of trade closure. 0.01 == 1%
close_profit_absfloatAbsolute profit (in stake currency) at the time of trade closure.
leveragefloatLeverage used for this trade - defaults to 1.0 in spot markets.
enter_tagstringTag provided on entry via the enter_tag column in the dataframe
is_shortbooleanTrue for short trades, False otherwise
ordersOrder[]List of order objects attached to this trade (includes both filled and cancelled orders)
date_last_filled_utcdatetimeTime of the last filled order
entry_side"buy" / "sell"Order Side the trade was entered
exit_side"buy" / "sell"Order Side that will result in a trade exit / position reduction.
trade_direction"long" / "short"Trade direction in text - long or short.
nr_of_successful_entriesintNumber of successful (filled) entry orders
nr_of_successful_exitsintNumber of successful (filled) exit orders
+

Class methods

+

The following are class methods - which return generic information, and usually result in an explicit query against the database. +They can be used as Trade.<method> - e.g. open_trades = Trade.get_open_trade_count()

+
+

Backtesting/hyperopt

+

Most methods will work in both backtesting / hyperopt and live/dry modes. +During backtesting, it's limited to usage in strategy callbacks. Usage in populate_*() methods is not supported and will result in wrong results.

+
+

get_trades_proxy

+

When your strategy needs some information on existing (open or close) trades - it's best to use Trade.get_trades_proxy().

+

Usage:

+
from freqtrade.persistence import Trade
+from datetime import timedelta
+
+# ...
+trade_hist = Trade.get_trades_proxy(pair='ETH/USDT', is_open=False, open_date=current_date - timedelta(days=2))
+
+

get_trades_proxy() supports the following keyword arguments. All arguments are optional - calling get_trades_proxy() without arguments will return a list of all trades in the database.

+
    +
  • pair e.g. pair='ETH/USDT'
  • +
  • is_open e.g. is_open=False
  • +
  • open_date e.g. open_date=current_date - timedelta(days=2)
  • +
  • close_date e.g. close_date=current_date - timedelta(days=5)
  • +
+

get_open_trade_count

+

Get the number of currently open trades

+
from freqtrade.persistence import Trade
+# ...
+open_trades = Trade.get_open_trade_count()
+
+

get_total_closed_profit

+

Retrieve the total profit the bot has generated so far. +Aggregates close_profit_abs for all closed trades.

+
from freqtrade.persistence import Trade
+
+# ...
+profit = Trade.get_total_closed_profit()
+
+

total_open_trades_stakes

+

Retrieve the total stake_amount that's currently in trades.

+
from freqtrade.persistence import Trade
+
+# ...
+profit = Trade.total_open_trades_stakes()
+
+

get_overall_performance

+

Retrieve the overall performance - similar to the /performance telegram command.

+
from freqtrade.persistence import Trade
+
+# ...
+if self.config['runmode'].value in ('live', 'dry_run'):
+    performance = Trade.get_overall_performance()
+
+

Sample return value: ETH/BTC had 5 trades, with a total profit of 1.5% (ratio of 0.015).

+
{"pair": "ETH/BTC", "profit": 0.015, "count": 5}
+
+

Order Object

+

An Order object represents an order on the exchange (or a simulated order in dry-run mode). +An Order object will always be tied to it's corresponding Trade, and only really makes sense in the context of a trade.

+

Order - Available attributes

+

an Order object is typically attached to a trade. +Most properties here can be None as they are dependant on the exchange response.

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
AttributeDataTypeDescription
tradeTradeTrade object this order is attached to
ft_pairstringPair this order is for
ft_is_openbooleanis the order filled?
order_typestringOrder type as defined on the exchange - usually market, limit or stoploss
statusstringStatus as defined by ccxt. Usually open, closed, expired or canceled
sidestringBuy or Sell
pricefloatPrice the order was placed at
averagefloatAverage price the order filled at
amountfloatAmount in base currency
filledfloatFilled amount (in base currency)
remainingfloatRemaining amount
costfloatCost of the order - usually average * filled
order_datedatetimeOrder creation date use order_date_utc instead
order_date_utcdatetimeOrder creation date (in UTC)
order_fill_datedatetimeOrder fill date use order_fill_utc instead
order_fill_date_utcdatetimeOrder fill date
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/updating/index.html b/en/2023.4/updating/index.html new file mode 100644 index 000000000..c793f16b9 --- /dev/null +++ b/en/2023.4/updating/index.html @@ -0,0 +1,1347 @@ + + + + + + + + + + + + + + + + + + + + + + + + Updating Freqtrade - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

How to update

+

To update your freqtrade installation, please use one of the below methods, corresponding to your installation method.

+
+

Tracking changes

+

Breaking changes / changed behavior will be documented in the changelog that is posted alongside every release. +For the develop branch, please follow PR's to avoid being surprised by changes.

+
+

docker

+
+

Legacy installations using the master image

+

We're switching from master to stable for the release Images - please adjust your docker-file and replace freqtradeorg/freqtrade:master with freqtradeorg/freqtrade:stable

+
+
docker compose pull
+docker compose up -d
+
+

Installation via setup script

+
./setup.sh --update
+
+
+

Note

+

Make sure to run this command with your virtual environment disabled!

+
+

Plain native installation

+

Please ensure that you're also updating dependencies - otherwise things might break without you noticing.

+
git pull
+pip install -U -r requirements.txt
+pip install -e .
+
+# Ensure freqUI is at the latest version
+freqtrade install-ui 
+
+

Problems updating

+

Update-problems usually come missing dependencies (you didn't follow the above instructions) - or from updated dependencies, which fail to install (for example TA-lib). +Please refer to the corresponding installation sections (common problems linked below)

+

Common problems and their solutions:

+ + + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/utils/index.html b/en/2023.4/utils/index.html new file mode 100644 index 000000000..7fcc293ae --- /dev/null +++ b/en/2023.4/utils/index.html @@ -0,0 +1,2553 @@ + + + + + + + + + + + + + + + + + + + + + + + + Utility Sub-commands - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+ +
+ + + +
+
+ + + + + + + +

Utility Subcommands

+

Besides the Live-Trade and Dry-Run run modes, the backtesting, edge and hyperopt optimization subcommands, and the download-data subcommand which prepares historical data, the bot contains a number of utility subcommands. They are described in this section.

+

Create userdir

+

Creates the directory structure to hold your files for freqtrade. +Will also create strategy and hyperopt examples for you to get started. +Can be used multiple times - using --reset will reset the sample strategy and hyperopt files to their default state.

+
usage: freqtrade create-userdir [-h] [--userdir PATH] [--reset]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+  --reset               Reset sample files to their original state.
+
+
+

Warning

+

Using --reset may result in loss of data, since this will overwrite all sample files without asking again.

+
+
├── backtest_results
+├── data
+├── hyperopt_results
+├── hyperopts
+│   ├── sample_hyperopt_loss.py
+├── notebooks
+│   └── strategy_analysis_example.ipynb
+├── plot
+└── strategies
+    └── sample_strategy.py
+
+

Create new config

+

Creates a new configuration file, asking some questions which are important selections for a configuration.

+
usage: freqtrade new-config [-h] [-c PATH]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default: `config.json`). Multiple --config options may be used. Can be set to `-`
+                        to read config from stdin.
+
+
+

Warning

+

Only vital questions are asked. Freqtrade offers a lot more configuration possibilities, which are listed in the Configuration documentation

+
+

Create config examples

+
$ freqtrade new-config --config config_binance.json
+
+? Do you want to enable Dry-run (simulated trades)?  Yes
+? Please insert your stake currency: BTC
+? Please insert your stake amount: 0.05
+? Please insert max_open_trades (Integer or -1 for unlimited open trades): 3
+? Please insert your desired timeframe (e.g. 5m): 5m
+? Please insert your display Currency (for reporting): USD
+? Select exchange  binance
+? Do you want to enable Telegram?  No
+
+

Create new strategy

+

Creates a new strategy from a template similar to SampleStrategy. +The file will be named inline with your class name, and will not overwrite existing files.

+

Results will be located in user_data/strategies/<strategyclassname>.py.

+
usage: freqtrade new-strategy [-h] [--userdir PATH] [-s NAME]
+                              [--template {full,minimal,advanced}]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+  -s NAME, --strategy NAME
+                        Specify strategy class name which will be used by the
+                        bot.
+  --template {full,minimal,advanced}
+                        Use a template which is either `minimal`, `full`
+                        (containing multiple sample indicators) or `advanced`.
+                        Default: `full`.
+
+

Sample usage of new-strategy

+
freqtrade new-strategy --strategy AwesomeStrategy
+
+

With custom user directory

+
freqtrade new-strategy --userdir ~/.freqtrade/ --strategy AwesomeStrategy
+
+

Using the advanced template (populates all optional functions and methods)

+
freqtrade new-strategy --strategy AwesomeStrategy --template advanced
+
+

List Strategies

+

Use the list-strategies subcommand to see all strategies in one particular directory.

+

This subcommand is useful for finding problems in your environment with loading strategies: modules with strategies that contain errors and failed to load are printed in red (LOAD FAILED), while strategies with duplicate names are printed in yellow (DUPLICATE NAME).

+
usage: freqtrade list-strategies [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                                 [-d PATH] [--userdir PATH]
+                                 [--strategy-path PATH] [-1] [--no-color]
+                                 [--recursive-strategy-search]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --strategy-path PATH  Specify additional strategy lookup path.
+  -1, --one-column      Print output in one column.
+  --no-color            Disable colorization of hyperopt results. May be
+                        useful if you are redirecting output to a file.
+  --recursive-strategy-search
+                        Recursively search for a strategy in the strategies
+                        folder.
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+
+

Warning

+

Using these commands will try to load all python files from a directory. This can be a security risk if untrusted files reside in this directory, since all module-level code is executed.

+
+

Example: Search default strategies directories (within the default userdir).

+
freqtrade list-strategies
+
+

Example: Search strategies directory within the userdir.

+
freqtrade list-strategies --userdir ~/.freqtrade/
+
+

Example: Search dedicated strategy path.

+
freqtrade list-strategies --strategy-path ~/.freqtrade/strategies/
+
+

List freqAI models

+

Use the list-freqaimodels subcommand to see all freqAI models available.

+

This subcommand is useful for finding problems in your environment with loading freqAI models: modules with models that contain errors and failed to load are printed in red (LOAD FAILED), while models with duplicate names are printed in yellow (DUPLICATE NAME).

+
usage: freqtrade list-freqaimodels [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                                   [-d PATH] [--userdir PATH]
+                                   [--freqaimodel-path PATH] [-1] [--no-color]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --freqaimodel-path PATH
+                        Specify additional lookup path for freqaimodels.
+  -1, --one-column      Print output in one column.
+  --no-color            Disable colorization of hyperopt results. May be
+                        useful if you are redirecting output to a file.
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH, --data-dir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+

List Exchanges

+

Use the list-exchanges subcommand to see the exchanges available for the bot.

+
usage: freqtrade list-exchanges [-h] [-1] [-a]
+
+optional arguments:
+  -h, --help        show this help message and exit
+  -1, --one-column  Print output in one column.
+  -a, --all         Print all exchanges known to the ccxt library.
+
+
    +
  • Example: see exchanges available for the bot: +
    $ freqtrade list-exchanges
    +Exchanges available for Freqtrade:
    +Exchange name    Valid    reason
    +---------------  -------  --------------------------------------------
    +aax              True
    +ascendex         True     missing opt: fetchMyTrades
    +bequant          True
    +bibox            True
    +bigone           True
    +binance          True
    +binanceus        True
    +bitbank          True     missing opt: fetchTickers
    +bitcoincom       True
    +bitfinex         True
    +bitforex         True     missing opt: fetchMyTrades, fetchTickers
    +bitget           True
    +bithumb          True     missing opt: fetchMyTrades
    +bitkk            True     missing opt: fetchMyTrades
    +bitmart          True
    +bitmax           True     missing opt: fetchMyTrades
    +bitpanda         True
    +bittrex          True
    +bitvavo          True
    +bitz             True     missing opt: fetchMyTrades
    +btcalpha         True     missing opt: fetchTicker, fetchTickers
    +btcmarkets       True     missing opt: fetchTickers
    +buda             True     missing opt: fetchMyTrades, fetchTickers
    +bw               True     missing opt: fetchMyTrades, fetchL2OrderBook
    +bybit            True
    +bytetrade        True
    +cdax             True
    +cex              True     missing opt: fetchMyTrades
    +coinbaseprime    True     missing opt: fetchTickers
    +coinbasepro      True     missing opt: fetchTickers
    +coinex           True
    +crex24           True
    +deribit          True
    +digifinex        True
    +equos            True     missing opt: fetchTicker, fetchTickers
    +eterbase         True
    +fcoin            True     missing opt: fetchMyTrades, fetchTickers
    +fcoinjp          True     missing opt: fetchMyTrades, fetchTickers
    +gateio           True
    +gemini           True
    +gopax            True
    +hbtc             True
    +hitbtc           True
    +huobijp          True
    +huobipro         True
    +idex             True
    +kraken           True
    +kucoin           True
    +lbank            True     missing opt: fetchMyTrades
    +mercado          True     missing opt: fetchTickers
    +ndax             True     missing opt: fetchTickers
    +novadax          True
    +okcoin           True
    +okex             True
    +probit           True
    +qtrade           True
    +stex             True
    +timex            True
    +upbit            True     missing opt: fetchMyTrades
    +vcc              True
    +zb               True     missing opt: fetchMyTrades
    +
  • +
+
+

missing opt exchanges

+

Values with "missing opt:" might need special configuration (e.g. using orderbook if fetchTickers is missing) - but should in theory work (although we cannot guarantee they will).

+
+
    +
  • Example: see all exchanges supported by the ccxt library (including 'bad' ones, i.e. those that are known to not work with Freqtrade): +
    $ freqtrade list-exchanges -a
    +All exchanges supported by the ccxt library:
    +Exchange name       Valid    reason
    +------------------  -------  ---------------------------------------------------------------------------------------
    +aax                 True
    +aofex               False    missing: fetchOrder
    +ascendex            True     missing opt: fetchMyTrades
    +bequant             True
    +bibox               True
    +bigone              True
    +binance             True
    +binanceus           True
    +bit2c               False    missing: fetchOrder, fetchOHLCV
    +bitbank             True     missing opt: fetchTickers
    +bitbay              False    missing: fetchOrder
    +bitcoincom          True
    +bitfinex            True
    +bitfinex2           False    missing: fetchOrder
    +bitflyer            False    missing: fetchOrder, fetchOHLCV
    +bitforex            True     missing opt: fetchMyTrades, fetchTickers
    +bitget              True
    +bithumb             True     missing opt: fetchMyTrades
    +bitkk               True     missing opt: fetchMyTrades
    +bitmart             True
    +bitmax              True     missing opt: fetchMyTrades
    +bitmex              False    Various reasons.
    +bitpanda            True
    +bitso               False    missing: fetchOHLCV
    +bitstamp            True     missing opt: fetchTickers
    +bitstamp1           False    missing: fetchOrder, fetchOHLCV
    +bittrex             True
    +bitvavo             True
    +bitz                True     missing opt: fetchMyTrades
    +bl3p                False    missing: fetchOrder, fetchOHLCV
    +bleutrade           False    missing: fetchOrder
    +braziliex           False    missing: fetchOHLCV
    +btcalpha            True     missing opt: fetchTicker, fetchTickers
    +btcbox              False    missing: fetchOHLCV
    +btcmarkets          True     missing opt: fetchTickers
    +btctradeua          False    missing: fetchOrder, fetchOHLCV
    +btcturk             False    missing: fetchOrder
    +buda                True     missing opt: fetchMyTrades, fetchTickers
    +bw                  True     missing opt: fetchMyTrades, fetchL2OrderBook
    +bybit               True
    +bytetrade           True
    +cdax                True
    +cex                 True     missing opt: fetchMyTrades
    +chilebit            False    missing: fetchOrder, fetchOHLCV
    +coinbase            False    missing: fetchOrder, cancelOrder, createOrder, fetchOHLCV
    +coinbaseprime       True     missing opt: fetchTickers
    +coinbasepro         True     missing opt: fetchTickers
    +coincheck           False    missing: fetchOrder, fetchOHLCV
    +coinegg             False    missing: fetchOHLCV
    +coinex              True
    +coinfalcon          False    missing: fetchOHLCV
    +coinfloor           False    missing: fetchOrder, fetchOHLCV
    +coingi              False    missing: fetchOrder, fetchOHLCV
    +coinmarketcap       False    missing: fetchOrder, cancelOrder, createOrder, fetchBalance, fetchOHLCV
    +coinmate            False    missing: fetchOHLCV
    +coinone             False    missing: fetchOHLCV
    +coinspot            False    missing: fetchOrder, cancelOrder, fetchOHLCV
    +crex24              True
    +currencycom         False    missing: fetchOrder
    +delta               False    missing: fetchOrder
    +deribit             True
    +digifinex           True
    +equos               True     missing opt: fetchTicker, fetchTickers
    +eterbase            True
    +exmo                False    missing: fetchOrder
    +exx                 False    missing: fetchOHLCV
    +fcoin               True     missing opt: fetchMyTrades, fetchTickers
    +fcoinjp             True     missing opt: fetchMyTrades, fetchTickers
    +flowbtc             False    missing: fetchOrder, fetchOHLCV
    +foxbit              False    missing: fetchOrder, fetchOHLCV
    +gateio              True
    +gemini              True
    +gopax               True
    +hbtc                True
    +hitbtc              True
    +hollaex             False    missing: fetchOrder
    +huobijp             True
    +huobipro            True
    +idex                True
    +independentreserve  False    missing: fetchOHLCV
    +indodax             False    missing: fetchOHLCV
    +itbit               False    missing: fetchOHLCV
    +kraken              True
    +kucoin              True
    +kuna                False    missing: fetchOHLCV
    +lakebtc             False    missing: fetchOrder, fetchOHLCV
    +latoken             False    missing: fetchOrder, fetchOHLCV
    +lbank               True     missing opt: fetchMyTrades
    +liquid              False    missing: fetchOHLCV
    +luno                False    missing: fetchOHLCV
    +lykke               False    missing: fetchOHLCV
    +mercado             True     missing opt: fetchTickers
    +mixcoins            False    missing: fetchOrder, fetchOHLCV
    +ndax                True     missing opt: fetchTickers
    +novadax             True
    +oceanex             False    missing: fetchOHLCV
    +okcoin              True
    +okex                True
    +paymium             False    missing: fetchOrder, fetchOHLCV
    +phemex              False    Does not provide history.
    +poloniex            False    missing: fetchOrder
    +probit              True
    +qtrade              True
    +rightbtc            False    missing: fetchOrder
    +ripio               False    missing: fetchOHLCV
    +southxchange        False    missing: fetchOrder, fetchOHLCV
    +stex                True
    +surbitcoin          False    missing: fetchOrder, fetchOHLCV
    +therock             False    missing: fetchOHLCV
    +tidebit             False    missing: fetchOrder
    +tidex               False    missing: fetchOHLCV
    +timex               True
    +upbit               True     missing opt: fetchMyTrades
    +vbtc                False    missing: fetchOrder, fetchOHLCV
    +vcc                 True
    +wavesexchange       False    missing: fetchOrder
    +whitebit            False    missing: fetchOrder, cancelOrder, createOrder, fetchBalance
    +xbtce               False    missing: fetchOrder, fetchOHLCV
    +xena                False    missing: fetchOrder
    +yobit               False    missing: fetchOHLCV
    +zaif                False    missing: fetchOrder, fetchOHLCV
    +zb                  True     missing opt: fetchMyTrades
    +
  • +
+

List Timeframes

+

Use the list-timeframes subcommand to see the list of timeframes available for the exchange.

+
usage: freqtrade list-timeframes [-h] [-v] [--logfile FILE] [-V] [-c PATH] [-d PATH] [--userdir PATH] [--exchange EXCHANGE] [-1]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no config is provided.
+  -1, --one-column      Print output in one column.
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are: 'syslog', 'journald'. See the documentation for more details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default: `config.json`). Multiple --config options may be used. Can be set to `-`
+                        to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+
    +
  • Example: see the timeframes for the 'binance' exchange, set in the configuration file:
  • +
+
$ freqtrade list-timeframes -c config_binance.json
+...
+Timeframes available for the exchange `binance`: 1m, 3m, 5m, 15m, 30m, 1h, 2h, 4h, 6h, 8h, 12h, 1d, 3d, 1w, 1M
+
+
    +
  • Example: enumerate exchanges available for Freqtrade and print timeframes supported by each of them: +
    $ for i in `freqtrade list-exchanges -1`; do freqtrade list-timeframes --exchange $i; done
    +
  • +
+

List pairs/list markets

+

The list-pairs and list-markets subcommands allow to see the pairs/markets available on exchange.

+

Pairs are markets with the '/' character between the base currency part and the quote currency part in the market symbol. +For example, in the 'ETH/BTC' pair 'ETH' is the base currency, while 'BTC' is the quote currency.

+

For pairs traded by Freqtrade the pair quote currency is defined by the value of the stake_currency configuration setting.

+

You can print info about any pair/market with these subcommands - and you can filter output by quote-currency using --quote BTC, or by base-currency using --base ETH options correspondingly.

+

These subcommands have same usage and same set of available options:

+
usage: freqtrade list-markets [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                              [-d PATH] [--userdir PATH] [--exchange EXCHANGE]
+                              [--print-list] [--print-json] [-1] [--print-csv]
+                              [--base BASE_CURRENCY [BASE_CURRENCY ...]]
+                              [--quote QUOTE_CURRENCY [QUOTE_CURRENCY ...]] [-a]
+                              [--trading-mode {spot,margin,futures}]
+
+usage: freqtrade list-pairs [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                            [-d PATH] [--userdir PATH] [--exchange EXCHANGE]
+                            [--print-list] [--print-json] [-1] [--print-csv]
+                            [--base BASE_CURRENCY [BASE_CURRENCY ...]]
+                            [--quote QUOTE_CURRENCY [QUOTE_CURRENCY ...]] [-a]
+                            [--trading-mode {spot,margin,futures}]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no
+                        config is provided.
+  --print-list          Print list of pairs or market symbols. By default data
+                        is printed in the tabular format.
+  --print-json          Print list of pairs or market symbols in JSON format.
+  -1, --one-column      Print output in one column.
+  --print-csv           Print exchange pair or market data in the csv format.
+  --base BASE_CURRENCY [BASE_CURRENCY ...]
+                        Specify base currency(-ies). Space-separated list.
+  --quote QUOTE_CURRENCY [QUOTE_CURRENCY ...]
+                        Specify quote currency(-ies). Space-separated list.
+  -a, --all             Print all pairs or market symbols. By default only
+                        active ones are shown.
+  --trading-mode {spot,margin,futures}
+                        Select Trading mode
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default: `config.json`).
+                        Multiple --config options may be used. Can be set to
+                        `-` to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+

By default, only active pairs/markets are shown. Active pairs/markets are those that can currently be traded +on the exchange. The see the list of all pairs/markets (not only the active ones), use the -a/-all option.

+

Pairs/markets are sorted by its symbol string in the printed output.

+

Examples

+
    +
  • Print the list of active pairs with quote currency USD on exchange, specified in the default +configuration file (i.e. pairs on the "Bittrex" exchange) in JSON format:
  • +
+
$ freqtrade list-pairs --quote USD --print-json
+
+
    +
  • Print the list of all pairs on the exchange, specified in the config_binance.json configuration file +(i.e. on the "Binance" exchange) with base currencies BTC or ETH and quote currencies USDT or USD, as the +human-readable list with summary:
  • +
+
$ freqtrade list-pairs -c config_binance.json --all --base BTC ETH --quote USDT USD --print-list
+
+
    +
  • Print all markets on exchange "Kraken", in the tabular format:
  • +
+
$ freqtrade list-markets --exchange kraken --all
+
+

Test pairlist

+

Use the test-pairlist subcommand to test the configuration of dynamic pairlists.

+

Requires a configuration with specified pairlists attribute. +Can be used to generate static pairlists to be used during backtesting / hyperopt.

+
usage: freqtrade test-pairlist [-h] [--userdir PATH] [-v] [-c PATH]
+                               [--quote QUOTE_CURRENCY [QUOTE_CURRENCY ...]]
+                               [-1] [--print-json] [--exchange EXCHANGE]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  --quote QUOTE_CURRENCY [QUOTE_CURRENCY ...]
+                        Specify quote currency(-ies). Space-separated list.
+  -1, --one-column      Print output in one column.
+  --print-json          Print list of pairs or market symbols in JSON format.
+  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no
+                        config is provided.
+
+

Examples

+

Show whitelist when using a dynamic pairlist.

+
freqtrade test-pairlist --config config.json --quote USDT BTC
+
+

Convert database

+

freqtrade convert-db can be used to convert your database from one system to another (sqlite -> postgres, postgres -> other postgres), migrating all trades, orders and Pairlocks.

+

Please refer to the SQL cheatsheet to learn about requirements for different database systems.

+
usage: freqtrade convert-db [-h] [--db-url PATH] [--db-url-from PATH]
+
+optional arguments:
+  -h, --help          show this help message and exit
+  --db-url PATH       Override trades database URL, this is useful in custom
+                      deployments (default: `sqlite:///tradesv3.sqlite` for
+                      Live Run mode, `sqlite:///tradesv3.dryrun.sqlite` for
+                      Dry Run).
+  --db-url-from PATH  Source db url to use when migrating a database.
+
+
+

Warning

+

Please ensure to only use this on an empty target database. Freqtrade will perform a regular migration, but may fail if entries already existed.

+
+

Webserver mode

+
+

Experimental

+

Webserver mode is an experimental mode to increase backesting and strategy development productivity. +There may still be bugs - so if you happen to stumble across these, please report them as github issues, thanks.

+
+

Run freqtrade in webserver mode. +Freqtrade will start the webserver and allow FreqUI to start and control backtesting processes. +This has the advantage that data will not be reloaded between backtesting runs (as long as timeframe and timerange remain identical). +FreqUI will also show the backtesting results.

+
usage: freqtrade webserver [-h] [-v] [--logfile FILE] [-V] [-c PATH] [-d PATH]
+                           [--userdir PATH]
+
+optional arguments:
+  -h, --help            show this help message and exit
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+

Webserver mode - docker

+

You can also use webserver mode via docker. +Starting a one-off container requires the configuration of the port explicitly, as ports are not exposed by default. +You can use docker compose run --rm -p 127.0.0.1:8080:8080 freqtrade webserver to start a one-off container that'll be removed once you stop it. This assumes that port 8080 is still available and no other bot is running on that port.

+

Alternatively, you can reconfigure the docker-compose file to have the command updated:

+
    command: >
+      webserver
+      --config /freqtrade/user_data/config.json
+
+

You can now use docker compose up to start the webserver. +This assumes that the configuration has a webserver enabled and configured for docker (listening port = 0.0.0.0).

+
+

Tip

+

Don't forget to reset the command back to the trade command if you want to start a live or dry-run bot.

+
+

Show previous Backtest results

+

Allows you to show previous backtest results. +Adding --show-pair-list outputs a sorted pair list you can easily copy/paste into your configuration (omitting bad pairs).

+
+Strategy overfitting +

Only using winning pairs can lead to an overfitted strategy, which will not work well on future data. Make sure to extensively test your strategy in dry-run before risking real money.

+
+
usage: freqtrade backtesting-show [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                                  [-d PATH] [--userdir PATH]
+                                  [--export-filename PATH] [--show-pair-list]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --export-filename PATH
+                        Save backtest results to the file with this filename.
+                        Requires `--export` to be set as well. Example:
+                        `--export-filename=user_data/backtest_results/backtest
+                        _today.json`
+  --show-pair-list      Show backtesting pairlist sorted by profit.
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+

Detailed backtest analysis

+

Advanced backtest result analysis.

+

More details in the Backtesting analysis Section.

+
usage: freqtrade backtesting-analysis [-h] [-v] [--logfile FILE] [-V]
+                                      [-c PATH] [-d PATH] [--userdir PATH]
+                                      [--export-filename PATH]
+                                      [--analysis-groups {0,1,2,3,4} [{0,1,2,3,4} ...]]
+                                      [--enter-reason-list ENTER_REASON_LIST [ENTER_REASON_LIST ...]]
+                                      [--exit-reason-list EXIT_REASON_LIST [EXIT_REASON_LIST ...]]
+                                      [--indicator-list INDICATOR_LIST [INDICATOR_LIST ...]]
+                                      [--timerange YYYYMMDD-[YYYYMMDD]]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --export-filename PATH, --backtest-filename PATH
+                        Use this filename for backtest results.Requires
+                        `--export` to be set as well. Example: `--export-filen
+                        ame=user_data/backtest_results/backtest_today.json`
+  --analysis-groups {0,1,2,3,4} [{0,1,2,3,4} ...]
+                        grouping output - 0: simple wins/losses by enter tag,
+                        1: by enter_tag, 2: by enter_tag and exit_tag, 3: by
+                        pair and enter_tag, 4: by pair, enter_ and exit_tag
+                        (this can get quite large)
+  --enter-reason-list ENTER_REASON_LIST [ENTER_REASON_LIST ...]
+                        Comma separated list of entry signals to analyse.
+                        Default: all. e.g. 'entry_tag_a,entry_tag_b'
+  --exit-reason-list EXIT_REASON_LIST [EXIT_REASON_LIST ...]
+                        Comma separated list of exit signals to analyse.
+                        Default: all. e.g.
+                        'exit_tag_a,roi,stop_loss,trailing_stop_loss'
+  --indicator-list INDICATOR_LIST [INDICATOR_LIST ...]
+                        Comma separated list of indicators to analyse. e.g.
+                        'close,rsi,bb_lowerband,profit_abs'
+  --timerange YYYYMMDD-[YYYYMMDD]
+                        Timerange to filter trades for analysis, 
+                        start inclusive, end exclusive. e.g.
+                        20220101-20220201
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+

List Hyperopt results

+

You can list the hyperoptimization epochs the Hyperopt module evaluated previously with the hyperopt-list sub-command.

+
usage: freqtrade hyperopt-list [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                               [-d PATH] [--userdir PATH] [--best]
+                               [--profitable] [--min-trades INT]
+                               [--max-trades INT] [--min-avg-time FLOAT]
+                               [--max-avg-time FLOAT] [--min-avg-profit FLOAT]
+                               [--max-avg-profit FLOAT]
+                               [--min-total-profit FLOAT]
+                               [--max-total-profit FLOAT]
+                               [--min-objective FLOAT] [--max-objective FLOAT]
+                               [--no-color] [--print-json] [--no-details]
+                               [--hyperopt-filename PATH] [--export-csv FILE]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --best                Select only best epochs.
+  --profitable          Select only profitable epochs.
+  --min-trades INT      Select epochs with more than INT trades.
+  --max-trades INT      Select epochs with less than INT trades.
+  --min-avg-time FLOAT  Select epochs above average time.
+  --max-avg-time FLOAT  Select epochs below average time.
+  --min-avg-profit FLOAT
+                        Select epochs above average profit.
+  --max-avg-profit FLOAT
+                        Select epochs below average profit.
+  --min-total-profit FLOAT
+                        Select epochs above total profit.
+  --max-total-profit FLOAT
+                        Select epochs below total profit.
+  --min-objective FLOAT
+                        Select epochs above objective.
+  --max-objective FLOAT
+                        Select epochs below objective.
+  --no-color            Disable colorization of hyperopt results. May be
+                        useful if you are redirecting output to a file.
+  --print-json          Print output in JSON format.
+  --no-details          Do not print best epoch details.
+  --hyperopt-filename FILENAME
+                        Hyperopt result filename.Example: `--hyperopt-
+                        filename=hyperopt_results_2020-09-27_16-20-48.pickle`
+  --export-csv FILE     Export to CSV-File. This will disable table print.
+                        Example: --export-csv hyperopt.csv
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+
+

Note

+

hyperopt-list will automatically use the latest available hyperopt results file. +You can override this using the --hyperopt-filename argument, and specify another, available filename (without path!).

+
+

Examples

+

List all results, print details of the best result at the end: +

freqtrade hyperopt-list
+

+

List only epochs with positive profit. Do not print the details of the best epoch, so that the list can be iterated in a script: +

freqtrade hyperopt-list --profitable --no-details
+

+

Show details of Hyperopt results

+

You can show the details of any hyperoptimization epoch previously evaluated by the Hyperopt module with the hyperopt-show subcommand.

+
usage: freqtrade hyperopt-show [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                               [-d PATH] [--userdir PATH] [--best]
+                               [--profitable] [-n INT] [--print-json]
+                               [--hyperopt-filename FILENAME] [--no-header]
+                               [--disable-param-export]
+                               [--breakdown {day,week,month} [{day,week,month} ...]]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --best                Select only best epochs.
+  --profitable          Select only profitable epochs.
+  -n INT, --index INT   Specify the index of the epoch to print details for.
+  --print-json          Print output in JSON format.
+  --hyperopt-filename FILENAME
+                        Hyperopt result filename.Example: `--hyperopt-
+                        filename=hyperopt_results_2020-09-27_16-20-48.pickle`
+  --no-header           Do not print epoch details header.
+  --disable-param-export
+                        Disable automatic hyperopt parameter export.
+  --breakdown {day,week,month} [{day,week,month} ...]
+                        Show backtesting breakdown per [day, week, month].
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+
+

Note

+

hyperopt-show will automatically use the latest available hyperopt results file. +You can override this using the --hyperopt-filename argument, and specify another, available filename (without path!).

+
+

Examples

+

Print details for the epoch 168 (the number of the epoch is shown by the hyperopt-list subcommand or by Hyperopt itself during hyperoptimization run):

+
freqtrade hyperopt-show -n 168
+
+

Prints JSON data with details for the last best epoch (i.e., the best of all epochs):

+
freqtrade hyperopt-show --best -n -1 --print-json --no-header
+
+

Show trades

+

Print selected (or all) trades from database to screen.

+
usage: freqtrade show-trades [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                             [-d PATH] [--userdir PATH] [--db-url PATH]
+                             [--trade-ids TRADE_IDS [TRADE_IDS ...]]
+                             [--print-json]
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --db-url PATH         Override trades database URL, this is useful in custom
+                        deployments (default: `sqlite:///tradesv3.sqlite` for
+                        Live Run mode, `sqlite:///tradesv3.dryrun.sqlite` for
+                        Dry Run).
+  --trade-ids TRADE_IDS [TRADE_IDS ...]
+                        Specify the list of trade ids.
+  --print-json          Print output in JSON format.
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+

Examples

+

Print trades with id 2 and 3 as json

+
freqtrade show-trades --db-url sqlite:///tradesv3.sqlite --trade-ids 2 3 --print-json
+
+

Strategy-Updater

+

Updates listed strategies or all strategies within the strategies folder to be v3 compliant. +If the command runs without --strategy-list then all strategies inside the strategies folder will be converted. +Your original strategy will remain available in the user_data/strategies_orig_updater/ directory.

+
+

Conversion results

+

Strategy updater will work on a "best effort" approach. Please do your due diligence and verify the results of the conversion. +We also recommend to run a python formatter (e.g. black) to format results in a sane manner.

+
+
usage: freqtrade strategy-updater [-h] [-v] [--logfile FILE] [-V] [-c PATH]
+                                  [-d PATH] [--userdir PATH]
+                                  [--strategy-list STRATEGY_LIST [STRATEGY_LIST ...]]
+
+options:
+  -h, --help            show this help message and exit
+  --strategy-list STRATEGY_LIST [STRATEGY_LIST ...]
+                        Provide a space-separated list of strategies to
+                        backtest. Please note that timeframe needs to be set
+                        either in config or via command line. When using this
+                        together with `--export trades`, the strategy-name is
+                        injected into the filename (so `backtest-data.json`
+                        becomes `backtest-data-SampleStrategy.json`
+
+Common arguments:
+  -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
+  --logfile FILE, --log-file FILE
+                        Log to the file specified. Special values are:
+                        'syslog', 'journald'. See the documentation for more
+                        details.
+  -V, --version         show program's version number and exit
+  -c PATH, --config PATH
+                        Specify configuration file (default:
+                        `userdir/config.json` or `config.json` whichever
+                        exists). Multiple --config options may be used. Can be
+                        set to `-` to read config from stdin.
+  -d PATH, --datadir PATH, --data-dir PATH
+                        Path to directory with historical backtesting data.
+  --userdir PATH, --user-data-dir PATH
+                        Path to userdata directory.
+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/webhook-config/index.html b/en/2023.4/webhook-config/index.html new file mode 100644 index 000000000..bee9a8b28 --- /dev/null +++ b/en/2023.4/webhook-config/index.html @@ -0,0 +1,1652 @@ + + + + + + + + + + + + + + + + + + + + + + + + Web Hook - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Webhook usage

+

Configuration

+

Enable webhooks by adding a webhook-section to your configuration file, and setting webhook.enabled to true.

+

Sample configuration (tested using IFTTT).

+
  "webhook": {
+        "enabled": true,
+        "url": "https://maker.ifttt.com/trigger/<YOUREVENT>/with/key/<YOURKEY>/",
+        "entry": {
+            "value1": "Buying {pair}",
+            "value2": "limit {limit:8f}",
+            "value3": "{stake_amount:8f} {stake_currency}"
+        },
+        "entry_cancel": {
+            "value1": "Cancelling Open Buy Order for {pair}",
+            "value2": "limit {limit:8f}",
+            "value3": "{stake_amount:8f} {stake_currency}"
+        },
+         "entry_fill": {
+            "value1": "Buy Order for {pair} filled",
+            "value2": "at {open_rate:8f}",
+            "value3": ""
+        },
+        "exit": {
+            "value1": "Exiting {pair}",
+            "value2": "limit {limit:8f}",
+            "value3": "profit: {profit_amount:8f} {stake_currency} ({profit_ratio})"
+        },
+        "exit_cancel": {
+            "value1": "Cancelling Open Exit Order for {pair}",
+            "value2": "limit {limit:8f}",
+            "value3": "profit: {profit_amount:8f} {stake_currency} ({profit_ratio})"
+        },
+        "exit_fill": {
+            "value1": "Exit Order for {pair} filled",
+            "value2": "at {close_rate:8f}.",
+            "value3": ""
+        },
+        "status": {
+            "value1": "Status: {status}",
+            "value2": "",
+            "value3": ""
+        }
+    },
+
+

The url in webhook.url should point to the correct url for your webhook. If you're using IFTTT (as shown in the sample above) please insert your event and key to the url.

+

You can set the POST body format to Form-Encoded (default), JSON-Encoded, or raw data. Use "format": "form", "format": "json", or "format": "raw" respectively. Example configuration for Mattermost Cloud integration:

+
  "webhook": {
+        "enabled": true,
+        "url": "https://<YOURSUBDOMAIN>.cloud.mattermost.com/hooks/<YOURHOOK>",
+        "format": "json",
+        "status": {
+            "text": "Status: {status}"
+        }
+    },
+
+

The result would be a POST request with e.g. {"text":"Status: running"} body and Content-Type: application/json header which results Status: running message in the Mattermost channel.

+

When using the Form-Encoded or JSON-Encoded configuration you can configure any number of payload values, and both the key and value will be ouput in the POST request. However, when using the raw data format you can only configure one value and it must be named "data". In this instance the data key will not be output in the POST request, only the value. For example:

+
  "webhook": {
+        "enabled": true,
+        "url": "https://<YOURHOOKURL>",
+        "format": "raw",
+        "webhookstatus": {
+            "data": "Status: {status}"
+        }
+    },
+
+

The result would be a POST request with e.g. Status: running body and Content-Type: text/plain header.

+

Optional parameters are available to enable automatic retries for webhook messages. The webhook.retries parameter can be set for the maximum number of retries the webhook request should attempt if it is unsuccessful (i.e. HTTP response status is not 200). By default this is set to 0 which is disabled. An additional webhook.retry_delay parameter can be set to specify the time in seconds between retry attempts. By default this is set to 0.1 (i.e. 100ms). Note that increasing the number of retries or retry delay may slow down the trader if there are connectivity issues with the webhook. Example configuration for retries:

+
  "webhook": {
+        "enabled": true,
+        "url": "https://<YOURHOOKURL>",
+        "retries": 3,
+        "retry_delay": 0.2,
+        "status": {
+            "status": "Status: {status}"
+        }
+    },
+
+

Custom messages can be sent to Webhook endpoints via the self.dp.send_msg() function from within the strategy. To enable this, set the allow_custom_messages option to true:

+
  "webhook": {
+        "enabled": true,
+        "url": "https://<YOURHOOKURL>",
+        "allow_custom_messages": true,
+        "strategy_msg": {
+            "status": "StrategyMessage: {msg}"
+        }
+    },
+
+

Different payloads can be configured for different events. Not all fields are necessary, but you should configure at least one of the dicts, otherwise the webhook will never be called.

+

Entry

+

The fields in webhook.entry are filled when the bot executes a long/short. Parameters are filled using string.format. +Possible parameters are:

+
    +
  • trade_id
  • +
  • exchange
  • +
  • pair
  • +
  • direction
  • +
  • leverage
  • +
  • limit # Deprecated - should no longer be used.
  • +
  • open_rate
  • +
  • amount
  • +
  • open_date
  • +
  • stake_amount
  • +
  • stake_currency
  • +
  • base_currency
  • +
  • fiat_currency
  • +
  • order_type
  • +
  • current_rate
  • +
  • enter_tag
  • +
+

Entry cancel

+

The fields in webhook.entry_cancel are filled when the bot cancels a long/short order. Parameters are filled using string.format. +Possible parameters are:

+
    +
  • trade_id
  • +
  • exchange
  • +
  • pair
  • +
  • direction
  • +
  • leverage
  • +
  • limit
  • +
  • amount
  • +
  • open_date
  • +
  • stake_amount
  • +
  • stake_currency
  • +
  • base_currency
  • +
  • fiat_currency
  • +
  • order_type
  • +
  • current_rate
  • +
  • enter_tag
  • +
+

Entry fill

+

The fields in webhook.entry_fill are filled when the bot filled a long/short order. Parameters are filled using string.format. +Possible parameters are:

+
    +
  • trade_id
  • +
  • exchange
  • +
  • pair
  • +
  • direction
  • +
  • leverage
  • +
  • open_rate
  • +
  • amount
  • +
  • open_date
  • +
  • stake_amount
  • +
  • stake_currency
  • +
  • base_currency
  • +
  • fiat_currency
  • +
  • order_type
  • +
  • current_rate
  • +
  • enter_tag
  • +
+

Exit

+

The fields in webhook.exit are filled when the bot exits a trade. Parameters are filled using string.format. +Possible parameters are:

+
    +
  • trade_id
  • +
  • exchange
  • +
  • pair
  • +
  • direction
  • +
  • leverage
  • +
  • gain
  • +
  • limit
  • +
  • amount
  • +
  • open_rate
  • +
  • profit_amount
  • +
  • profit_ratio
  • +
  • stake_currency
  • +
  • base_currency
  • +
  • fiat_currency
  • +
  • exit_reason
  • +
  • order_type
  • +
  • open_date
  • +
  • close_date
  • +
+

Exit fill

+

The fields in webhook.exit_fill are filled when the bot fills a exit order (closes a Trade). Parameters are filled using string.format. +Possible parameters are:

+
    +
  • trade_id
  • +
  • exchange
  • +
  • pair
  • +
  • direction
  • +
  • leverage
  • +
  • gain
  • +
  • close_rate
  • +
  • amount
  • +
  • open_rate
  • +
  • current_rate
  • +
  • profit_amount
  • +
  • profit_ratio
  • +
  • stake_currency
  • +
  • base_currency
  • +
  • fiat_currency
  • +
  • exit_reason
  • +
  • order_type
  • +
  • open_date
  • +
  • close_date
  • +
+

Exit cancel

+

The fields in webhook.exit_cancel are filled when the bot cancels a exit order. Parameters are filled using string.format. +Possible parameters are:

+
    +
  • trade_id
  • +
  • exchange
  • +
  • pair
  • +
  • direction
  • +
  • leverage
  • +
  • gain
  • +
  • limit
  • +
  • amount
  • +
  • open_rate
  • +
  • current_rate
  • +
  • profit_amount
  • +
  • profit_ratio
  • +
  • stake_currency
  • +
  • base_currency
  • +
  • fiat_currency
  • +
  • exit_reason
  • +
  • order_type
  • +
  • open_date
  • +
  • close_date
  • +
+

Status

+

The fields in webhook.status are used for regular status messages (Started / Stopped / ...). Parameters are filled using string.format.

+

The only possible value here is {status}.

+

Discord

+

A special form of webhooks is available for discord. +You can configure this as follows:

+
"discord": {
+    "enabled": true,
+    "webhook_url": "https://discord.com/api/webhooks/<Your webhook URL ...>",
+    "exit_fill": [
+        {"Trade ID": "{trade_id}"},
+        {"Exchange": "{exchange}"},
+        {"Pair": "{pair}"},
+        {"Direction": "{direction}"},
+        {"Open rate": "{open_rate}"},
+        {"Close rate": "{close_rate}"},
+        {"Amount": "{amount}"},
+        {"Open date": "{open_date:%Y-%m-%d %H:%M:%S}"},
+        {"Close date": "{close_date:%Y-%m-%d %H:%M:%S}"},
+        {"Profit": "{profit_amount} {stake_currency}"},
+        {"Profitability": "{profit_ratio:.2%}"},
+        {"Enter tag": "{enter_tag}"},
+        {"Exit Reason": "{exit_reason}"},
+        {"Strategy": "{strategy}"},
+        {"Timeframe": "{timeframe}"},
+    ],
+    "entry_fill": [
+        {"Trade ID": "{trade_id}"},
+        {"Exchange": "{exchange}"},
+        {"Pair": "{pair}"},
+        {"Direction": "{direction}"},
+        {"Open rate": "{open_rate}"},
+        {"Amount": "{amount}"},
+        {"Open date": "{open_date:%Y-%m-%d %H:%M:%S}"},
+        {"Enter tag": "{enter_tag}"},
+        {"Strategy": "{strategy} {timeframe}"},
+    ]
+}
+
+

The above represents the default (exit_fill and entry_fill are optional and will default to the above configuration) - modifications are obviously possible.

+

Available fields correspond to the fields for webhooks and are documented in the corresponding webhook sections.

+

The notifications will look as follows by default.

+

discord-notification

+

Custom messages can be sent from a strategy to Discord endpoints via the dataprovider.send_msg() function. To enable this, set the allow_custom_messages option to true:

+
  "discord": {
+        "enabled": true,
+        "webhook_url": "https://discord.com/api/webhooks/<Your webhook URL ...>",
+        "allow_custom_messages": true,
+    },
+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/2023.4/windows_installation/index.html b/en/2023.4/windows_installation/index.html new file mode 100644 index 000000000..6531a9011 --- /dev/null +++ b/en/2023.4/windows_installation/index.html @@ -0,0 +1,1355 @@ + + + + + + + + + + + + + + + + + + + + + + + + Windows - Freqtrade + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+ +
+ + + + + + + + +
+ + +
+ +
+ + + + + + +
+
+ + + + + +
+
+
+ + + + +
+
+
+ + + + + +
+
+
+ + + +
+
+
+ + + +
+
+ + + + + + + +

Windows installation

+

We strongly recommend that Windows users use Docker as this will work much easier and smoother (also more secure).

+

If that is not possible, try using the Windows Linux subsystem (WSL) - for which the Ubuntu instructions should work. +Otherwise, please follow the instructions below.

+

Install freqtrade manually

+
+

64bit Python version

+

Please make sure to use 64bit Windows and 64bit Python to avoid problems with backtesting or hyperopt due to the memory constraints 32bit applications have under Windows. +32bit python versions are no longer supported under Windows.

+
+
+

Hint

+

Using the Anaconda Distribution under Windows can greatly help with installation problems. Check out the Anaconda installation section in the documentation for more information.

+
+

1. Clone the git repository

+
git clone https://github.com/freqtrade/freqtrade.git
+
+

2. Install ta-lib

+

Install ta-lib according to the ta-lib documentation.

+

As compiling from source on windows has heavy dependencies (requires a partial visual studio installation), Freqtrade provides these dependencies (in the binary wheel format) for the latest 3 Python versions (3.8, 3.9, 3.10 and 3.11) and for 64bit Windows. +These Wheels are also used by CI running on windows, and are therefore tested together with freqtrade.

+

Other versions must be downloaded from the above link.

+
cd \path\freqtrade
+python -m venv .env
+.env\Scripts\activate.ps1
+# optionally install ta-lib from wheel
+# Eventually adjust the below filename to match the downloaded wheel
+pip install --find-links build_helpers\ TA-Lib -U
+pip install -r requirements.txt
+pip install -e .
+freqtrade
+
+
+

Use Powershell

+

The above installation script assumes you're using powershell on a 64bit windows. +Commands for the legacy CMD windows console may differ.

+
+

Error during installation on Windows

+
error: Microsoft Visual C++ 14.0 is required. Get it with "Microsoft Visual C++ Build Tools": http://landinghub.visualstudio.com/visual-cpp-build-tools
+
+

Unfortunately, many packages requiring compilation don't provide a pre-built wheel. It is therefore mandatory to have a C/C++ compiler installed and available for your python environment to use.

+

You can download the Visual C++ build tools from here and install "Desktop development with C++" in it's default configuration. Unfortunately, this is a heavy download / dependency so you might want to consider WSL2 or docker compose first.

+

Windows installation

+
+ + + + + + +
+
+ + +
+ + + +
+ + + + + + + + + + +
+
+
+
+ + + + + + + + + + + + + + + \ No newline at end of file diff --git a/en/versions.json b/en/versions.json index 7fa37579c..526ead0b2 100644 --- a/en/versions.json +++ b/en/versions.json @@ -1,4 +1,9 @@ [ + { + "version": "2023.4", + "title": "2023.4", + "aliases": [] + }, { "version": "2023.3", "title": "2023.3",