Unable to run SAM_Audio

[PS-AI]

Hi,
Thank you for your work.

I am trying to test SAM_Audio.
My Python environment is 3.11.14. I have a CUDA compatible GPU.
I cloned the git repo, created a virtual environment and installed the requirements using 'pip install .' as mentioned in the repo.
When I run the test code ( given in the repo). I get this error -
"TypeError: BaseModel._from_pretrained() missing 2 required keyword-only arguments: 'proxies' and 'resume_download'"

There seems to be some wrong dependencies. Could you please check and let me know how to overcome this.

[jaylenetriton]

I have the same issue as well with a clean conda env and Python 3.11.14. I, too, used 'pip install .'
Would appreciate any guidance!

From this post #41
I installed transformers==4.54.0 and have gotten farther along. Will update when I am able.

[snpushpi]

I followed #41 along with editing the base.py file to set proxies to be None and resume_download to be True. That solves the issue. Not sure if I needed to do #41, maybe just editing those two values would be fine.

[snpushpi]

My model also runs correctly!

[lauritisreyes]

Hi, thanks to @snpushpi for the comment and details. I’m also facing the same issue. I managed to work around it by modifying the parameters manually. It would be great if the official code could be updated to support the current Hugging Face Hub API.

[wmrenr]

I followed #41 along with editing the base.py file to set proxies to be None and resume_download to be True. That solves the issue. Not sure if I needed to do #41, maybe just editing those two values would be fine.

Hello，which base.py file did you change?