Adding mixup/cutmix in references script.

Author: datumbox

references/classification/train.py

@@ -165,10 +165,16 @@ def main(args):
     train_dir = os.path.join(args.data_path, 'train')
     val_dir = os.path.join(args.data_path, 'val')
     dataset, dataset_test, train_sampler, test_sampler = load_data(train_dir, val_dir, args)
+
+    collate_fn = None
+    if args.mixup_alpha > 0.0 or args.cutmix_alpha > 0.0:
+        mixupcutmix = torchvision.transforms.RandomMixupCutmix(len(dataset.classes), mixup_alpha=args.mixup_alpha,
+                                                               cutmix_alpha=args.cutmix_alpha)
+        collate_fn = lambda batch: mixupcutmix(*(torch.utils.data._utils.collate.default_collate(batch)))  # noqa: E731
     data_loader = torch.utils.data.DataLoader(
         dataset, batch_size=args.batch_size,
-        sampler=train_sampler, num_workers=args.workers, pin_memory=True)
-
+        sampler=train_sampler, num_workers=args.workers, pin_memory=True,
+        collate_fn=collate_fn)
     data_loader_test = torch.utils.data.DataLoader(
         dataset_test, batch_size=args.batch_size,
         sampler=test_sampler, num_workers=args.workers, pin_memory=True)
@@ -254,7 +260,6 @@ def main(args):
 def get_args_parser(add_help=True):
     import argparse
     parser = argparse.ArgumentParser(description='PyTorch Classification Training', add_help=add_help)
-
     parser.add_argument('--data-path', default='/datasets01/imagenet_full_size/061417/', help='dataset')
     parser.add_argument('--model', default='resnet18', help='model')
     parser.add_argument('--device', default='cuda', help='device')
@@ -273,6 +278,8 @@ def get_args_parser(add_help=True):
     parser.add_argument('--label-smoothing', default=0.0, type=float,
                         help='label smoothing (default: 0.0)',
                         dest='label_smoothing')
+    parser.add_argument('--mixup-alpha', default=0.0, type=float, help='mixup alpha (default: 0.0)')
+    parser.add_argument('--cutmix-alpha', default=0.0, type=float, help='cutmix alpha (default: 0.0)')
     parser.add_argument('--lr-step-size', default=30, type=int, help='decrease lr every step-size epochs')
     parser.add_argument('--lr-gamma', default=0.1, type=float, help='decrease lr by a factor of lr-gamma')
     parser.add_argument('--print-freq', default=10, type=int, help='print frequency')
@@ -306,7 +313,6 @@ def get_args_parser(add_help=True):
     )
     parser.add_argument('--auto-augment', default=None, help='auto augment policy (default: None)')
     parser.add_argument('--random-erase', default=0.0, type=float, help='random erasing probability (default: 0.0)')
-
     # Mixed precision training parameters
     parser.add_argument('--apex', action='store_true',
                         help='Use apex for mixed precision training')
@@ -315,7 +321,6 @@ def get_args_parser(add_help=True):
                              'O0 for FP32 training, O1 for mixed precision training.'
                              'For further detail, see https://github.com/NVIDIA/apex/tree/master/examples/imagenet'
                         )
-
     # distributed training parameters
     parser.add_argument('--world-size', default=1, type=int,
                         help='number of distributed processes')
@@ -326,7 +331,6 @@ def get_args_parser(add_help=True):
     parser.add_argument(
         '--model-ema-decay', type=float, default=0.99,
         help='decay factor for Exponential Moving Average of model parameters(default: 0.99)')
-
     return parser
 
 

test/test_transforms_tensor.py

@@ -788,5 +788,5 @@ def test_random_mixupcutmix_with_real_data():
 
     torch.testing.assert_close(
         torch.stack(stats).mean(dim=0),
-        torch.tensor([46.931968688964844, 69.97343444824219, 0.459820032119751])
+        torch.tensor([46.9443473815918, 64.79092407226562, 0.459820032119751])
     )

torchvision/transforms/transforms.py

@@ -1955,8 +1955,9 @@ def __repr__(self):
         return self.__class__.__name__ + '(p={})'.format(self.p)
 
 
+# TODO: move this to references before merging and delete the tests
 class RandomMixupCutmix(torch.nn.Module):
-    """Randomly apply Mixum or Cutmix to the provided batch and targets.
+    """Randomly apply Mixup or Cutmix to the provided batch and targets.
     The class implements the data augmentations as described in the papers
     `"mixup: Beyond Empirical Risk Minimization" <https://arxiv.org/abs/1710.09412>`_ and
     `"CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features"
@@ -2014,8 +2015,8 @@ def forward(self, batch: Tensor, target: Tensor) -> Tuple[Tensor, Tensor]:
             return batch, target
 
         # It's faster to roll the batch by one instead of shuffling it to create image pairs
-        batch_flipped = batch.roll(1)
-        target_flipped = target.roll(1)
+        batch_rolled = batch.roll(1, 0)
+        target_rolled = target.roll(1)
 
         if self.mixup_alpha <= 0.0:
             use_mixup = False
@@ -2025,8 +2026,8 @@ def forward(self, batch: Tensor, target: Tensor) -> Tuple[Tensor, Tensor]:
         if use_mixup:
             # Implemented as on mixup paper, page 3.
             lambda_param = float(torch._sample_dirichlet(torch.tensor([self.mixup_alpha, self.mixup_alpha]))[0])
-            batch_flipped.mul_(1.0 - lambda_param)
-            batch.mul_(lambda_param).add_(batch_flipped)
+            batch_rolled.mul_(1.0 - lambda_param)
+            batch.mul_(lambda_param).add_(batch_rolled)
         else:
             # Implemented as on cutmix paper, page 12 (with minor corrections on typos).
             lambda_param = float(torch._sample_dirichlet(torch.tensor([self.cutmix_alpha, self.cutmix_alpha]))[0])
@@ -2044,11 +2045,11 @@ def forward(self, batch: Tensor, target: Tensor) -> Tuple[Tensor, Tensor]:
             x2 = int(torch.clamp(r_x + r_w_half, max=W))
             y2 = int(torch.clamp(r_y + r_h_half, max=H))
 
-            batch[:, :, y1:y2, x1:x2] = batch_flipped[:, :, y1:y2, x1:x2]
+            batch[:, :, y1:y2, x1:x2] = batch_rolled[:, :, y1:y2, x1:x2]
             lambda_param = float(1.0 - (x2 - x1) * (y2 - y1) / (W * H))
 
-        target_flipped.mul_(1.0 - lambda_param)
-        target.mul_(lambda_param).add_(target_flipped)
+        target_rolled.mul_(1.0 - lambda_param)
+        target.mul_(lambda_param).add_(target_rolled)
 
         return batch, target