problem running the code

[nikokks]

Hi I try to do a benchmark on cifar100 with 80% noisy labels with resnet18.

This is the command

python3 main_cifar.py --r 0.80 --lambda_u 500 --dataset cifar100 --p_threshold 0.03 --data_path ./cifar-100 --experiment-name simclr_resnet18 --method reg --net resnet18  --batch_size 32 

but at a random epoch it throws this

File "main_cifar.py", line 174, in main
    run_train_loop(net1, optimizer1, sched1, net2, optimizer2, sched2, criterion, CEloss, CE, loader, args.p_threshold,
  File "/home/nikkokks/labelr/C2D/train_cifar.py", line 148, in run_train_loop
    train(epoch, net1, net2, criterion, optimizer1, labeled_trainloader, unlabeled_trainloader, lambda_u,
  File "/home/nikkokks/labelr/C2D/train.py", line 66, in train
    targets_x, targets_u = co_guess(net, net2, inputs_x, inputs_u, inputs_x2, inputs_u2, w_x, labels_x, T,
  File "/home/nikkokks/labelr/C2D/train.py", line 25, in co_guess
    outputs_x = net(inputs_x)
  File "/home/nikkokks/.local/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/nikkokks/.local/lib/python3.8/site-packages/torchvision/models/resnet.py", line 285, in forward
    return self._forward_impl(x)
  File "/home/nikkokks/.local/lib/python3.8/site-packages/torchvision/models/resnet.py", line 276, in _forward_impl
    x = self.layer4(x)
  File "/home/nikkokks/.local/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/nikkokks/.local/lib/python3.8/site-packages/torch/nn/modules/container.py", line 217, in forward
    input = module(input)
  File "/home/nikkokks/.local/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/nikkokks/.local/lib/python3.8/site-packages/torchvision/models/resnet.py", line 93, in forward
    out = self.bn1(out)
  File "/home/nikkokks/.local/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/nikkokks/.local/lib/python3.8/site-packages/torch/nn/modules/batchnorm.py", line 171, in forward
    return F.batch_norm(
  File "/home/nikkokks/.local/lib/python3.8/site-packages/torch/nn/functional.py", line 2448, in batch_norm
    _verify_batch_size(input.size())
  File "/home/nikkokks/.local/lib/python3.8/site-packages/torch/nn/functional.py", line 2416, in _verify_batch_size
    raise ValueError("Expected more than 1 value per channel when training, got input size {}".format(size))
ValueError: Expected more than 1 value per channel when training, got input size torch.Size([1, 512, 1, 1])

I suppose this is due to the last batch_size.

Do you have a better solution than activating drop_last = True ?