Debug dataloader

preprocessing/main.py

@@ -76,7 +76,7 @@ if __name__ == "__main__":
     
     device = 'cuda' if torch.cuda.is_available() else 'cpu'
     entube_dataset = EnTubeDataset(folder_paths, image_processors, device)
-    dataloader = DataLoader(entube_dataset, batch_size=4)
+    dataloader = DataLoader(entube_dataset, batch_size=2)
 
     for batch_idx, (videos, image_sizes) in enumerate(dataloader):
         print(f"Processing batch {batch_idx + 1}/{len(dataloader)}")

preprocessing/mm_datautils.py

@@ -33,7 +33,7 @@ def process_images(
                 image = Image.fromarray(image)
             image_aux_list = []
             for processor_aux in processor_aux_list:
-                image_aux = image
+                image_aux = image # PIL.Image
                 if hasattr(processor_aux, "image_mean"):
                     try:
                         target_resolution = processor_aux.crop_size["height"]
@@ -47,25 +47,14 @@ def process_images(
                 ][0]
                 # image_aux.shape: torch.Size([3, 384, 384])
                 image_aux_list.append(image_aux)
-            new_images_aux_list.append(image_aux_list)
+            new_images_aux_list.append(image_aux_list) # torch.Tensor(C, H, W) new_images_aux_list[num_frames][num_processor]
         new_images_aux_list = [
             list(batch_image_aux) for batch_image_aux in zip(*new_images_aux_list)
-        ]
-        print(f'@tcm: len(new_images_aux_list[0]): {len(new_images_aux_list[0])}')
-        print(f'@tcm: new_images_aux_list[0][0].shape: {new_images_aux_list[0][0].shape}')
-        # new_images_aux_list = [
-        #     torch.stack(image_aux).half().to(device) for image_aux in new_images_aux_list
-        # ]
-        tmp_images_aux_list = []
-        for image_aux in new_images_aux_list:
-            if isinstance(image_aux, list):
-                print(f'@tcm: len(image_aux): {len(image_aux)}')
-                for i, img_aux in enumerate(image_aux):
-                    # img_aux.shape: torch.Size([3, 384, 384])
-                    pass
-            tmp_images_aux_list.append(torch.stack(image_aux).half().to(device))
-        new_images_aux_list = tmp_images_aux_list
-        return new_images_aux_list
+        ] # torch.Tensor(C, H, W) new_images_aux_list[num_processor][num_frames]
+        new_images_aux_list = [
+            torch.stack(image_aux).half().to(device) for image_aux in new_images_aux_list
+        ] # torch.Tensor(num_frames, C, H, W) new_images_aux_list[num_processor]
+        return new_images_aux_list 
     else:
         image_aspect_ratio = "pad"
         new_images = []