TypeError: avg_pool2d(): argument 'kernel_size' (position 2) must be tuple of ints, not tuple

vision
5

If I do that it prints this: Proxy(getattr_16).
What I do is I initialize the model:

	if network_variant == 'vb16IN':
		net = Embedder("vit_b_16", pretrained_flag = True)

Then extract the embeddings like this:

train_descr = extract_embeddings(net,train_loader,ms = scales,msp = 1.0,print_freq=20000)

I tried to follow the feature extraction instruction here AssertionError: Expected (batch_size, seq_length, hidden_dim) got torch.Size([1, 768, 24, 31]) - #4 by Matias_Vasquez
But I’m not sure I’m doing it right