[Bug]: Stable Diffusion Refuses to use PyTorch ROCm for AMD GPUs

[jefferyrlc]

What happened?

Installed Stable Diffusion via the Linux install script. Then installed PyTorch with ROCm support via pip according to https://pytorch.org/get-started/locally/#with-rocm-1. Launching Stable Diffusion WebUI then choosing Gradio or Streamlit has launches and only looks for CUDA, which obviously an AMD GPU can't support.

Version

0.0.1 (Default)

What browsers are you seeing the problem on?

Firefox

Where are you running the webui?

Linux

Custom settings

No response

Relevant log output

/home/jeffery/.conda/envs/ldm/lib/python3.8/site-packages/torch/distributed/distributed_c10d.py:181: UserWarning: torch.distributed.reduce_op is deprecated, please use torch.distributed.ReduceOp instead
  warnings.warn(
2022-10-17 00:33:41.577 Created a temporary directory at /tmp/tmp0hb30obq
2022-10-17 00:33:41.577 Writing /tmp/tmp0hb30obq/_remote_module_non_sriptable.py
Loading models.
Loading model from models/ldm/stable-diffusion-v1/model.ckpt
Global Step: 470000
LatentDiffusion: Running in eps-prediction mode
DiffusionWrapper has 859.52 M params.
making attention of type 'vanilla' with 512 in_channels
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
making attention of type 'vanilla' with 512 in_channels
2022-10-17 00:33:56.629 Found no NVIDIA driver on your system. Please check that you have an NVIDIA GPU and installed a driver from http://www.nvidia.com/Download/index.aspx, retrying in 0 seconds...
Loading model from models/ldm/stable-diffusion-v1/model.ckpt
Global Step: 470000
LatentDiffusion: Running in eps-prediction mode
DiffusionWrapper has 859.52 M params.
making attention of type 'vanilla' with 512 in_channels
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
making attention of type 'vanilla' with 512 in_channels
2022-10-17 00:34:26.613 Found no NVIDIA driver on your system. Please check that you have an NVIDIA GPU and installed a driver from http://www.nvidia.com/Download/index.aspx, retrying in 0 seconds...

Code of Conduct

• [X] I agree to follow this project's Code of Conduct