comfyui
https://github.com/comfyanonymous/comfyui
Python
The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface.
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Python not yet supported18 Subscribers
Add a CodeTriage badge to comfyui
Help out
- Issues
- MiniMax H3 _forward does per-step GPU→CPU syncs (sigma_v, text_tags) — graph capture blocked
- Fix int8_tensorwise Embedding crash when DynamicVRAM offloads weight to GPU
- Cleanup redundant package version logging in main.py
- NVFP4 checkpoints without `_quantization_metadata` load silently as unquantized and crash in F.linear (LTX-2.5 nvfp4)
- LTX 2.5 *= TypeError on MPS (Apple Silicon) with LTX-Video VAE - float64 not supported
- Using --fast-disk changes the results of H3.
- RuntimeError: quant_qk_per_thread_int8: Q/K base pointers and B/H/N strides must preserve 4-element alignment — MiniMax H3 + comfy kitchen attention backend on int8 checkpoints
- Setter-backed WeightHooks fail and can cache incomplete groups
- [Bug] PR #15486 `v = v.clone()` in H3 Attention causes ~1.1 GB VRAM regression on 16 GB GPUs -> WDDM spill / multi-x slowdown
- [Bug] PR #15486 `v = v.clone()` in H3 Attention causes ~1.1 GB VRAM regression on 16 GB GPUs -> WDDM spill / multi-x slowdown
- Docs
- Python not yet supported