Removing unnecessary logging

Signed-off-by: João Lucas de Sousa Almeida <[email protected]>

Author: Joao-L-S-Almeida

terratorch/datasets/generic_pixel_wise_dataset.py

@@ -148,7 +148,7 @@ def __getitem__(self, index: int) -> dict[str, Any]:
         if self.expand_temporal_dimension:
             image = rearrange(image, "(channels time) h w -> channels time h w", channels=len(self.output_bands))
         image = np.moveaxis(image, 0, -1)
-        print(f"Image shape when entering __getitem__: {image.shape}")
+        
         if self.filter_indices:
             image = image[..., self.filter_indices]
         output = {
@@ -163,7 +163,7 @@ def __getitem__(self, index: int) -> dict[str, Any]:
         if self.transform:
             output = self.transform(**output)
         output["filename"] = self.image_files[index]
-        print(f"Image shape after transform: {output['image'].shape}")
+        
         return output
 
     def _load_file(self, path, nan_replace: int | float | None = None) -> xr.DataArray:

terratorch/io/file.py

@@ -60,7 +60,6 @@ def load_from_file_or_attribute(value: list[float]|str):
     elif isinstance(value, str):  # It can be the path for a file
         if os.path.isfile(value):
             try:
-                print(value)
                 content = np.genfromtxt(value).tolist()
             except:
                 raise Exception(f"File must be txt, but received {value}")

terratorch/tasks/loss_handler.py

@@ -40,7 +40,7 @@ def compute_loss(
                 If there are auxiliary heads, the main decode head is returned under the key "decode_head".
                 All other heads are returned with the same key as their name.
         """
-        print(f"Model output: {model_output.output.shape}, Ground truth: {ground_truth.shape}")
+        
         loss = self._compute_loss(model_output.output, ground_truth, criterion)
         if not model_output.auxiliary_heads:
             return {"loss": loss}

terratorch/tasks/segmentation_tasks.py

@@ -254,7 +254,7 @@ def validation_step(self, batch: Any, batch_idx: int, dataloader_idx: int = 0) -
         x = batch["image"]
         y = batch["mask"]
         model_output: ModelOutput = self(x)
-        print(f"Model output, validation: {model_output.output.shape}")
+        
         loss = self.val_loss_handler.compute_loss(model_output, y, self.criterion, self.aux_loss)
         self.val_loss_handler.log_loss(self.log, loss_dict=loss, batch_size=x.shape[0])
         y_hat_hard = to_segmentation_prediction(model_output)