You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am getting an error when loading the provided pretrained model for NVGesture rgb images nv_rgb.pth). I found that some of the model parameters are missing in the 'state_dict' once the checkpoint is loaded. The saved checkpoint for NVGesture RGB model has slightly different self_attention encoder architecture. After modifying the architecture the model file could be loaded but the results doesn't look right.
Could you please reshare/reupload the correct model file? Thank you very much.
The text was updated successfully, but these errors were encountered:
Hi @ibhattacharya, thank you for reaching out.
It seems the weight file for the rgb split has been uploaded wrongly. I'm looking into it right now, I'll upload the correct one ASAP
I am getting an error when loading the provided pretrained model for NVGesture rgb images nv_rgb.pth). I found that some of the model parameters are missing in the 'state_dict' once the checkpoint is loaded. The saved checkpoint for NVGesture RGB model has slightly different self_attention encoder architecture. After modifying the architecture the model file could be loaded but the results doesn't look right.
Could you please reshare/reupload the correct model file? Thank you very much.
The text was updated successfully, but these errors were encountered: