search

lukemelas / PyTorch-Pretrained-ViT

Vision Transformer (ViT) in PyTorch
770 stars 124 forks source link

RuntimeError: The size of tensor a (197) must match the size of tensor b (577) at non-singleton dimension 1 #28

Open jeremyEudaric opened 1 year ago

jeremyEudaric commented 1 year ago

Hi, Congratulation for your amazing work ! when l am using the VIT pre trained model l got this weird issue. Do you have an idea where this issue comes from ?

RuntimeError: The size of tensor a (197) must match the size of tensor b (577) at non-singleton dimension 1

/tmp/ipykernel_701421/2011775653.py in training(epochs) 34 35 ---> 36 output = Model(inputs) 37 38 loss = MSE(output,label)

skynbe commented 1 year ago

set ViT('B_16_imagenet1k', pretrained=True, image_size=224)