Hi, there are some issues about the infoNCEloss

[LanXiaoPang613]

Hi, there are some issues about the infoNCEloss:

when we define the info_nce_loss in the code as follows:
info_nce_loss = InfoNCELoss(temperature=0.1,
batch_size=args.batch_size * 2,
flat=('flat' in args.crl),
n_views=8 if args.mcrop else 2)
the labels size in the function of infoNCELoss is batch_size x 2 x n_views, while the size of the features are batch_size x n_views. Since the sizes of these two params are not matched, the calculation of the infoNCEloss is not correct.

[lxysl]

The features are comprised of two parts: labeled and unlabeled, each with a size equal to batch_size*n_views.

[LanXiaoPang613]

The features are comprised of two parts: labeled and unlabeled, each with a size equal to batch_size*n_views.

I know, but when this function is used to perform warm-up training, it will be incorrect. If possible, you can check it.

[lxysl]

If you are asking the warm-up stage, I guess this line is what you are looking for:

PLReMix/data/data_loader.py

Line 361 in 77cf96b

warmup_dataloader = get_train_dataloader(args, warmup_dataset, explicit_batch_size=args.batch_size * 2)

[LanXiaoPang613]

If you are asking the warm-up stage, I guess this line is what you are looking for:

PLReMix/data/data_loader.py

Line 361 in 77cf96b

warmup_dataloader = get_train_dataloader(args, warmup_dataset, explicit_batch_size=args.batch_size * 2)

OKay，i ignore this line. Thank you.

[LanXiaoPang613]

If possible, could you hare me a training log about the PLRemix on webvision, it seems i met some problems when training on this set.