Unexpected keys in state dict when loading Unified Voice

[seedlessbuns]

After starting training from the GUI and from CLI I get this error. Not sure what is going on, could I have the wrong autoregressive.pth? I don't see how this is the case, but it seems like something is not formatted/reformatted correctly from my very basic understanding of the State Dict.

[jtfl28]

I also got this error, changed
network.load_state_dict(load_net_clean, strict=strict)
to
network.load_state_dict(load_net_clean, strict=False)
in codes/trainer/basem_model.py

seems to be training now, but no idea what effect this will have on the end model

[jtfl28]

I also got this error, changed network.load_state_dict(load_net_clean, strict=strict) to network.load_state_dict(load_net_clean, strict=False) in codes/trainer/basem_model.py

seems to be training now, but no idea what effect this will have on the end model

Tried using the model that I just trained with this change, and it looks like it doesn't work.

I guess this repo is broken in its current state.

[BurnedP]

I am getting the same errors, no clue as to why. I already tried reinstalling and going over the process again. It is a problem with torch, but I am unsure what the problem is exactly.

[Bonk1971]

same.

[Bonk1971]

Shout out to @FurkanGozukara he has an updated repository with the fix for this. Short answer. Update Transformers to 4.28.1
pip install transformers==4.28.1

[FurkanGozukara]

Shout out to @FurkanGozukara he has an updated repository with the fix for this. Short answer. Update Transformers to 4.28.1 pip install transformers==4.28.1

thank you so much

here the updated repo. i try to keep my tutorials working :)

https://github.com/FurkanGozukara/Stable-Diffusion/blob/main/Tutorials/Deep-Voice-Clone-Tutorial-Tortoise-TTS.md

also here my tutorial

https://youtu.be/OiMRlqcgDL0