cantonioupao/pytorch-human_action_recognition_breakfast_dataset-C3D_model_implementation

Error in loading pretrained model for C3D

Opened this issue · 0 comments

Sandeep-GitHubAccount commented

I am running inference.py file using the pretrained model for C3D and facing the following error.
Please provide suggestions
Thank You

RuntimeError: Error(s) in loading state_dict for C3D:
size mismatch for fc8.weight: copying
Selection_834
a param with shape torch.Size([487, 4096]) from checkpoint, the shape in current model is torch.Size([48, 4096]).
size mismatch for fc8.bias: copying a param with shape torch.Size([487]) from checkpoint, the shape in current model is torch.Size([48])