Spaces:
Running
Error while loading the bigger model weights
When I try to replace the model weights provided here (small) with the larger model weights mentioned in the github repo it throws the following error:
size mismatch for SequenceModeling.0.rnn.weight_ih_l0: copying a param with shape torch.Size([1024, 32]) from checkpoint, the shape in current model is torch.Size([1024, 512]).
size mismatch for SequenceModeling.0.rnn.weight_ih_l0_reverse: copying a param with shape torch.Size([1024, 32]) from checkpoint, the shape in current model is torch.Size([1024, 512]).
Apologies ...upon a closer look I realized that I needed to change the feature number to 32 to meet HRNet's specs.
For wider audience visibility please extend the existing code to incorporate the HRNet_FeatureExtractor module and the hrnet.py file into the existing codebase and modify the FeatureExtraction_output to 32 instead of 512 which is used by UNET implementation.
More than happy to add a pull request to modify the codebase to incorporate these and maybe can also add a toggle on the front end allowing users to choose between either backend!