diff --git a/config_examples/config_freqai.example.json b/config_examples/config_freqai.example.json
index 27bc4532c..74612656d 100644
--- a/config_examples/config_freqai.example.json
+++ b/config_examples/config_freqai.example.json
@@ -48,6 +48,7 @@
     ],
     "freqai": {
         "enabled": true,
+        "warn_exceptions_on_backtest_only": false,
         "purge_old_models": 2,
         "train_period_days": 15,
         "backtest_period_days": 7,
diff --git a/freqtrade/freqai/freqai_interface.py b/freqtrade/freqai/freqai_interface.py
index 60fae421b..3e295b72b 100644
--- a/freqtrade/freqai/freqai_interface.py
+++ b/freqtrade/freqai/freqai_interface.py
@@ -116,6 +116,8 @@ class IFreqaiModel(ABC):
 
         record_params(config, self.full_path)
 
+        self.new_feature_selector = self.config.get('freqai', {}).get('warn_exceptions_on_backtest_only', False)
+
     def __getstate__(self):
         """
         Return an empty state to be pickled in hyperopt
@@ -349,30 +351,47 @@ class IFreqaiModel(ABC):
                         self.tb_logger = get_tb_logger(self.dd.model_type, dk.data_path,
                                                        self.activate_tensorboard)
                         self.model = self.train(dataframe_train, pair, dk)
-                        self.tb_logger.close()
+                        self.tb_logger.close()    
                     except Exception as msg:
                         logger.warning(
                             f"Training {pair} raised exception {msg.__class__.__name__}. "
                             f"Message: {msg}, skipping.", exc_info=True)
+                        
+                        if self.new_feature_selector:
+                            logger.warning(
+                                "Train failed. Try to train on next pair." if self.new_feature_selector else
+                                "Train failed. Raise error, fix data issue and try again."
+                            )
+                            self.tb_logger.close()
                         self.model = None
 
-                    self.dd.pair_dict[pair]["trained_timestamp"] = int(
-                        tr_train.stopts)
-                    if self.plot_features and self.model is not None:
-                        plot_feature_importance(self.model, pair, dk, self.plot_features)
-                    if self.save_backtest_models and self.model is not None:
-                        logger.info('Saving backtest model to disk.')
-                        self.dd.save_data(self.model, pair, dk)
-                    else:
-                        logger.info('Saving metadata to disk.')
-                        self.dd.save_metadata(dk)
+                    hard_check_model_valid = True
+                    if self.new_feature_selector:
+                        hard_check_model_valid = bool(False if self.model is None else True)
+                        if hard_check_model_valid:
+                            logger.info(
+                                "Model is trained. Saving model and metadata to disk."
+                            )
+
+                    if hard_check_model_valid:
+                        self.dd.pair_dict[pair]["trained_timestamp"] = int(tr_train.stopts)
+                        if self.plot_features and self.model is not None:
+                            plot_feature_importance(self.model, pair, dk, self.plot_features)
+                        if self.save_backtest_models and self.model is not None:
+                            logger.info('Saving backtest model to disk.')
+                            self.dd.save_data(self.model, pair, dk)
+                        else:
+                            logger.info('Saving metadata to disk.')
+                            self.dd.save_metadata(dk)
+
                 else:
                     self.model = self.dd.load_data(pair, dk)
 
-                pred_df, do_preds = self.predict(dataframe_backtest, dk)
-                append_df = dk.get_predictions_to_append(pred_df, do_preds, dataframe_backtest)
-                dk.append_predictions(append_df)
-                dk.save_backtesting_prediction(append_df)
+                if hard_check_model_valid:
+                    pred_df, do_preds = self.predict(dataframe_backtest, dk)
+                    append_df = dk.get_predictions_to_append(pred_df, do_preds, dataframe_backtest)
+                    dk.append_predictions(append_df)
+                    dk.save_backtesting_prediction(append_df)
 
         self.backtesting_fit_live_predictions(dk)
         dk.fill_predictions(dataframe)