Force labels to be int64, no observable difference in speed or VRAM t…

…o switching to int32, or int16

src/data/data_load.py

@@ -69,8 +69,6 @@ def get_num_classes(data_config):
     data_loader = data_loader_from_config(data_config)
     return len(data_loader.dataset.classes)
 
-def to_int16(label):
-    return torch.tensor(label, dtype=torch.int16)
 
 def create_data_loader(data_dir: str, image_height: int, image_width: int, image_dtype=torch.float16, using_gpu=False,
                        batch_size=1, n_workers=1):
@@ -80,9 +78,8 @@ def create_data_loader(data_dir: str, image_height: int, image_width: int, image
                                          transforms.ToDtype(image_dtype, scale=True), # Float16 is tiny bit faster, and bit more VRAM. Strange.
                                          transforms.Normalize([0.5, 0.5, 0.5], [0.5, 0.5, 0.5]),
                                          ])
-    label_transform = to_int16
     try:
-        data_set = torch_data_set.ImageFolder(root=data_dir, transform=data_transform, target_transform=label_transform)
+        data_set = torch_data_set.ImageFolder(root=data_dir, transform=data_transform)
     except FileNotFoundError:
         raise FileNotFoundError('Data directory provided should contain directories that have images in them, '
                                 'directory provided: ' + data_dir)