size mismatch for cls_head.fc_cls.weight: copying a param with shape torch.Size([400, 1024]) from checkpoint, the shape in current model is torch.Size([101, 1024]).

[777Rebecca]

When I select the ucf101 data set and run the test.py file, the weight of the full connection layer does not match. Will the mismatched parameter model be trained and fine-tuned by itself? Or how do I change it to 101 dimensions？