[Bug]: Intel Arc Torch issue about voltaml-fast-stable-diffusion

[gabe56f]

Can you check if this is also present on experimental?

Run git checkout experimental in the root folder of Volta

[gmbhneo]

Yes, also happens on experimental.

[Stax124]

Try running this from the terminal inside the Volta root folder:

.\venv\Scripts\activate.bat
pip install torch==2.0.1a0 torchvision==0.15.2a0 intel_extension_for_pytorch==2.0.110+xpu -f https://developer.intel.com/ipex-whl-stable-xpu

or replace .bat with .ps1 if you use PowerShell as your main shell.

This should install the necessary stuff and satisfy the torch requirement so that installation can move on.

Ref: https://github.com/intel/intel-extension-for-pytorch?tab=readme-ov-file#gpu-version

[gmbhneo]

PS D:\VoltaML> pip install torch==2.0.1a0 torchvision==0.15.2a0 intel_extension_for_pytorch==2.0.110+xpu -f https://developer.intel.com/ipex-whl-stable-xpu Looking in links: https://developer.intel.com/ipex-whl-stable-xpu ERROR: Could not find a version that satisfies the requirement torch==2.0.1a0 (from versions: 1.11.0, 1.12.0, 1.12.1, 1.13.0, 1.13.1, 2.0.0, 2.0.1, 2.1.0, 2.1.1) ERROR: No matching distribution found for torch==2.0.1a0

[Stax124]

You can try this but these are all commands from the official docs. If it does not work then just open an issue on their repos as we cannot do much about it.

Same instruction as above, just a slightly modified command with different URL:

.\venv\Scripts\activate.bat
pip install torch==2.0.0a0 intel-extension-for-pytorch==2.0.110+gitba7f6c1 --extra-index-url https://pytorch-extension.intel.com/release-whl/stable/xpu/us/

Ref: https://intel.github.io/intel-extension-for-pytorch/index.html#installation?platform=gpu&version=v2.0.110%2Bxpu

[gmbhneo]

The error at least changed a bit 🤔

Your last command worked and installed correctly. But starting Volta still don't work.

[Stax124]

Oh, you used stability matrix... give me some time, I need to check how their Venv works. It might have not activated properly

[Stax124]

I was able to install Torch and the intel extension on Windows, but they have no docs for Torchvision, another dependency that Volta needs. Do you know about some UI that seems to work with Intel Arc? I might be able to get the versions that they are using from that.

[gmbhneo]

The onlything working for me is SD.Next. It is the only SD UI working on my Intel Arc atm.

[Stax124]

Try another update, you will need to set --pytorch-type=intel in the Matrix as an extra command

[gmbhneo]

What do you mean by "try another update"? I've been using the --pytorch-type=intel already

[Stax124]

I just sent another patch upstream, just update Volta in Stability Matrix

[gmbhneo]

Already did, that's the screenshot after updating, sorry, was too fast posting.

[Stax124]

Try it now, distribution forcing did not work properly. It was trying Linux deps instead of the Windows one, it should hopefully work now

[gmbhneo]

Got this error

[Stax124]

Should be patched now

[gmbhneo]

[Stax124]

Maybe patched

[gmbhneo]

same issue 😅 but it is starting now

[gmbhneo]

getting this when selecting a model

[gmbhneo]

this is the upper part of this, hope this helps

[gabe56f]

Can you check if your Settings > Reproducibility & Generation > Device is correct?

[gmbhneo]

this is heppening after selecting my GPU, which is not selected by default

[Stax124]

This is something with the package itself, not much we can do there. The only option I can think of right now is just to wait for official support from Intel. At least on Windows that is.

[github-actions]

This issue is stale because it has been open for 30 days with no activity.

[github-actions]

This issue was closed because it has been inactive for 14 days since being marked as stale.