intel / intel/llm-scaler

SenseNova-U1 ComfyUI node support on Intel Arc / XPU

Open
#468 0 comments 0 reactions 1 assignee Claimed by @xiangyuT View on GitHub
Dominant language
C++
Stars
529
Forks
80
Avg merge
9h 7m
Merged PRs (30d)
38

Description

Hi,
I'm running intel/llm-scaler-omni:latest on Unraid with Intel Arc GPUs and trying to use the ComfyUI node for SenseNova-U1: https://github.com/smthemex/ComfyUI_SenseNova_U1
The node fails because it has extensive hardcoded torch.cuda calls throughout layer_streaming.py and modeling_neo_chat.py that are not compatible with XPU.
Could Intel provide either:

A patched version of this node with XPU support
Guidance on how to replace torch.cuda.Stream/Event/synchronize with XPU equivalents in PyTorch

The model itself (SenseNova-U1) seems very well suited for Intel Arc given its efficient architecture. Would love to see this supported.
Thanks

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.