-
Notifications
You must be signed in to change notification settings - Fork 4.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[intel arc gpu] RuntimeError: mat1 and mat2 must have the same dtype, but got Float and Half #2247
Comments
with this hack #476 (comment) everything is working doesn't work with default workflow for some reason. |
Now that we have the latest IPEX release and the same thing can be seen there, I am going to repeat the same thing I said in the discussion thread. This is definitely not a ComfyUI bug, it's an upstream bug with Pytorch and Intel's patches. Until they fix their end of this, the fact that you have to pile more custom code and hacks to make it work doesn't change the fact that it's broken at the end of the day. It needs to be fixed but you are barking up the wrong tree. |
latest ComfyUI commit 340177e
launch options:
USE_XETLA=OFF PYTHONPATH=/tmp/ComfyUI python main.py --listen --use-pytorch-cross-attention --preview-method latent2rgb --normalvram --bf16-vae
python: 3.11.6 (main, Nov 14 2023, 09:36:21) [GCC 13.2.1 20230801] on linux
pytorch: python-pytorch-opt 2.1.1-2 with intel patches
output of command
python -c "import torch; import intel_extension_for_pytorch as ipex; print(torch.xpu.is_available); print(torch.__config__.show()); print(ipex.__version__); [print(f'[{i}]: {torch.xpu.get_device_properties(i)}') for i in range(torch.xpu.device_count())];"
output
after pressing Queue Prompt getting error
RuntimeError: mat1 and mat2 must have the same dtype, but got Float and Half
Full error log
for those who are using intel arc gpu there is a thread #476
The text was updated successfully, but these errors were encountered: