Excuse me,Why do I get an error when I load the swinTransformer model with swin weights，such as:
size mismatch for layers.0.blocks.0.attn.relative_position_bias_table: copying a param with shape torch.Size([13, 4]) from checkpoint, the shape in current model is torch.Size([169, 4]).
Why is the dataset I downloaded following the link different from the one used in the paper?
Excuse me,Why do I get an error when I load the swinTransformer model with swin weights，such as:
size mismatch for layers.0.blocks.0.attn.relative_position_bias_table: copying a param with shape torch.Size([13, 4]) from checkpoint, the shape in current model is torch.Size([169, 4]).
Why is the dataset I downloaded following the link different from the one used in the paper?