Merge pull request #8580 from freqtrade/feat/add-transformer

Add transformer to FreqAI

freqtrade/freqai/base_models/BasePyTorchClassifier.py

@@ -75,6 +75,7 @@ class BasePyTorchClassifier(BasePyTorchModel):
             dk.data_dictionary["prediction_features"],
             device=self.device
         )
+        self.model.model.eval()
         logits = self.model.model(x)
         probs = F.softmax(logits, dim=-1)
         predicted_classes = torch.argmax(probs, dim=-1)

freqtrade/freqai/base_models/BasePyTorchModel.py

@@ -27,6 +27,7 @@ class BasePyTorchModel(IFreqaiModel, ABC):
         self.device = "cuda" if torch.cuda.is_available() else "cpu"
         test_size = self.freqai_info.get('data_split_parameters', {}).get('test_size')
         self.splits = ["train", "test"] if test_size != 0 else ["train"]
+        self.window_size = self.freqai_info.get("conv_width", 1)
 
     def train(
         self, unfiltered_df: DataFrame, pair: str, dk: FreqaiDataKitchen, **kwargs

freqtrade/freqai/base_models/BasePyTorchRegressor.py

@@ -44,6 +44,7 @@ class BasePyTorchRegressor(BasePyTorchModel):
             dk.data_dictionary["prediction_features"],
             device=self.device
         )
+        self.model.model.eval()
         y = self.model.model(x)
         pred_df = DataFrame(y.detach().tolist(), columns=[dk.label_list[0]])
         return (pred_df, dk.do_predict)