RuntimeError: Detected that PyTorch and TorchAudio were compiled with different CUDA versions. PyTorch has CUDA version 11.8 whereas TorchAudio has CUDA version 11.7.

[AK51]

Hi,

I am using ubuntu, is there any solution of properly install wav2lip?
Other solutions are for windows only...
I did try to reinstall torchaudio, delete torchaudio folder in site-package.... no luck...

stable-diffusion-webui/venv/lib/python3.10/site-packages/torchaudio/_extension/utils.py", line 119, in _check_cuda_version
    raise RuntimeError(
RuntimeError: Detected that PyTorch and TorchAudio were compiled with different CUDA versions. PyTorch has CUDA version 11.8 whereas TorchAudio has CUDA version 11.7. Please install the TorchAudio version that matches your PyTorch version.

Thanks.

[numz]

Hi,
Sorry about that, It should works on ubunutu

you can try :
Go to you stable diffusion folder with terminal and activate your venv python (on windows venv\Scripts\activate but I don't know on your environnement)
`
pip uninstall torch torchaudio torchvision

pip3 install torch==2.0.1 torchvision==0.15.2 torchaudio==2.0.1 --index-url https://download.pytorch.org/whl/cu118
`
Let me know,
Regards