CIFAR100 has different architecture from official implementation

[zijian-hu]

In your implementation, WRN28-10 is used which has about 36M parameters.
Your model definition:

FixMatch-pytorch/train.py

Lines 165 to 169 in 9044f2e

	elif args.dataset == 'cifar100':
	args.num_classes = 100
	if args.arch == 'wideresnet':
	args.model_depth = 28
	args.model_width = 10

I used the following code to get the number of parameters

wrn = build_wideresnet(depth=28, widen_factor=10, dropout=0, num_classes=100)

print(f"# params: {sum(p.numel() for p in wrn.parameters()):,}")

which gives the following output:

In the official TensorFlow implementation, a WRN with about 23M parameters is used for CIFAR100 (see below image).

Notes

The CLI args for the official code can be found in this issue.

[kekmodel]

fixed! thanks!