Merge pull request #94 from SFI-Visual-Intelligence/johan/devbranch

Updatet main to work with metricWrapper.

Author: hzavadil98

.gitignore

@@ -19,6 +19,8 @@ local*
 
 # Johanthings
 formatting.x 
+testrun.x
+storage/
 
 # Byte-compiled / optimized / DLL files
 __pycache__/

CollaborativeCoding/dataloaders/mnist_4_9.py

@@ -43,7 +43,12 @@ def __init__(
         self.labels_path = self.mnist_path / (
             MNIST_SOURCE["train_labels"][1] if train else MNIST_SOURCE["test_labels"][1]
         )
-
+        
+        # Functions to map the labels from (4,9) -> (0,5) for CrossEntropyLoss to work properly. 
+        self.label_shift = lambda x: x-4
+        self.label_restore = lambda x: x+4
+        
+        
     def __len__(self):
         return len(self.samples)
 
@@ -66,4 +71,4 @@ def __getitem__(self, idx):
         if self.transform:
             image = self.transform(image)
 
-        return image, label
+        return image, self.label_shift(label)

main.py

@@ -139,12 +139,12 @@ def main():
 
     for epoch in range(args.epoch):
         # Training loop start
+        print(f"Epoch: {epoch+1}/{args.epoch}")
         trainingloss = []
         model.train()
         for x, y in tqdm(trainloader, desc="Training"):
             x, y = x.to(device), y.to(device)
             logits = model.forward(x)
-
             loss = criterion(logits, y)
             loss.backward()
 
@@ -172,8 +172,8 @@ def main():
                 "Train loss": np.mean(trainingloss),
                 "Validation loss": np.mean(valloss),
             }
-            | train_metrics.getmetric(str_prefix="Train ")
-            | val_metrics.getmetric(str_prefix="Validation ")
+            | train_metrics.getmetrics(str_prefix="Train ")
+            | val_metrics.getmetrics(str_prefix="Validation ")
         )
         train_metrics.resetmetric()
         val_metrics.resetmetric()
@@ -187,12 +187,11 @@ def main():
             loss = criterion(logits, y)
             testloss.append(loss.item())
 
-            preds = th.argmax(logits, dim=1)
-            test_metrics(y, preds)
+            test_metrics(y, logits)
 
     wandb.log(
         {"Epoch": 1, "Test loss": np.mean(testloss)}
-        | test_metrics.getmetric(str_prefix="Test ")
+        | test_metrics.getmetrics(str_prefix="Test ")
     )
     test_metrics.resetmetric()