[RNN_FFN]

create_submission.py

@@ -30,7 +30,7 @@ def dummy_model(X):
     # Divided by a magic number
     return X[:, :, 4:].mean(dim=2) / 26  # This is (B, T)
 
-def create_submission(model, transform, device, rootDir):
+def create_submission(model, transform, device, rootDir, logdir):
     step_days = 10
     batch_size = 1024
     # We make chunks of num_days consecutive samples; As our dummy predictor
@@ -57,7 +57,7 @@ def create_submission(model, transform, device, rootDir):
     num_days_test = test_loader.dataset.ntimes
 
     logging.info("= Filling in the submission file")
-    with open(rootDir + "submission.csv", "w") as fh_submission:
+    with open(logdir + "submission.csv", "w") as fh_submission:
         fh_submission.write("Id,Predicted\n")
         submission_offset = 0
 

main.py

@@ -132,4 +132,4 @@ if __name__ == "__main__":
             wandb.log({"val_loss": val_loss})
 
 
-    create_submission.create_submission(network, dataloader.composite_transform(dataloader.transform_remove_space_time(), dataloader.transform_min_max_scaling(MIN, MAX)), device, rootDir)
+    create_submission.create_submission(network, dataloader.composite_transform(dataloader.transform_remove_space_time(), dataloader.transform_min_max_scaling(MIN, MAX)), device, rootDir, logdir)

model.py

@@ -56,11 +56,11 @@ class RNN(nn.Module):
             self.fc.add_module(
                 f"linear_{layer}", nn.Linear(self.hidden_size, self.hidden_size)
             )
-            self.ffn.add_module(
+            self.fc.add_module(
                 f"relu_{layer}", 
                 nn.ReLU()
             )
-            self.ffn.add_module(
+            self.fc.add_module(
                 f"dropout_{layer}", 
                 nn.Dropout(p=0.2)
             )