-
-
Notifications
You must be signed in to change notification settings - Fork 1.7k
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
Error on image generation with diffusers backend on Intel GPU #2332
Comments
@Xav-v just had a commit merged to resolve the missing setuptools. For the CUDA error, can you please provide the exact command you used to build the docker image? That error makes me think that the image wasn't built correctly for an Intel platform. |
Many thanks, I compiled again from master and the first issue disappeared. That's my docker-compose for the built:
|
based on your logs, it looks like your diffusers model definition has EDIT: You may also need to add a devices section to your docker-compose file to map in /dev/dri as follows:
|
Many thanks; now it's working! Modifying to CUDA: true in the yaml model file was the trick. |
Sorry to ressurect this old thread, however I was running into the same CUDA error. As I have no prior experience with running local AI models, I have not made any changes to any default settings. I manually modified the yaml files now. However, would it be possible to modify the container so that it automatically changes this to |
LocalAI version:
quay.io/go-skynet/local-ai:master-sycl-f16-ffmpeg
Additional test on git master version (self build) - see Additional context below.
Environment, CPU architecture, OS, and Version:
Docker environment on Debian, with Arc A380 GPU passthrough
Describe the bug
Running a model with diffuser backend such as
DreamShaper_8_pruned.safetensors
gives an errorNo module named 'setuptools'
.I also tested installing setuptools through
backend/python/diffusers/requirements-intel.txt
(see below) but it also fails.To Reproduce
Through the
Text to Image
WebUI , selectDreamshaper
as model, input a query and launch.Expected behavior
The model loads, no error is given, and image is created according to the prompt.
Logs
Additional context
I pulled local git version, added setuptools to the
backend/python/diffusers/requirements-intel.txt
, build the docker image.The new log is as following, with a
Torch not compiled with CUDA enabled
error popping up.The text was updated successfully, but these errors were encountered: