convert_lora_safetensor_to_diffusers OSError: No such device (os error 19)

[sing817]

python scripts/convert_lora_safetensor_to_diffusers.py \
--base_model_path=stabilityai/stable-diffusion-2-1-base \
--checkpoint_path="my_lora_output_pytorch_lora_weights_file" \
--dump_path="./testoutput" \

get:

File "conda/envs/diffusers/lib/python3.10/site-packages/safetensors/torch.py", line 259, in load_file
with safe_open(filename, framework="pt", device=device) as f:
OSError: No such device (os error 19)

why?

[sing817]

--device cpu or--device cuda or --device cuda:0 problem still

[gitlabspy]

Same here.

[sayakpaul]

I would like to try to reproduce this. Could you maybe help with a Colab that reproduces the error?

[sing817]

#state_dict = load_file(checkpoint_path)

state_dict = torch.load(checkpoint_path)

and get this:
AttributeError: 'UNet2DConditionModel' object has no attribute 'unet'
IndexError: pop from empty list

[sayakpaul]

What is your checkpoint_path? It doesn't seem like the complete snippet to me.

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.

Please note that issues that do not follow the contributing guidelines are likely to be ignored.

[fangcao1314]

最后解决了吗

[yzeng58]

Same here

[zhangxs131]

怎么解，是环境哪里的问题吗，卸载了几个重装还是，

[YoucanBaby]

same question

[lonngxiang]

same

[asomoza]

For everyone that keeps adding the "same" comment here (don't really know what the other language says).

This is an old issue and a lot has changed from then, specially with LoRAs.

Also the main reason this wasn't answered was because the person that asked the initial question didn't provide any useful information, the device error it's almost always associated to a bad environment because PyTorch couldn't find a device to use. So probably it was fixed with a reinstall.

So instead of adding the "same", if you want a solution, it would be better if you open a new issue with the details we need:

• minimal code to reproduce.
• trace of the error.
• details on your environment.

Until someone does this, you can keep posting the "same" and we won't be able to help.

[lonngxiang]

For everyone that keeps adding the "same" comment here (don't really know what the other language says).

This is an old issue and a lot has changed from then, specially with LoRAs.

Also the main reason this wasn't answered was because the person that asked the initial question didn't provide any useful information, the device error it's almost always associated to a bad environment because PyTorch couldn't find a device to use. So probably it was fixed with a reinstall.

So instead of adding the "same", if you want a solution, it would be better if you open a new issue with the details we need:

• minimal code to reproduce.
• trace of the error.
• details on your environment.

Until someone does this, you can keep posting the "same" and we won't be able to help.

torch 2.3.1
safetensors 0.4.3