Skip to content

The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.

Commits on master

1199411
5ebcab3
c350009
dea899f
Add logging for model unloading. (#10692)
comfyanonymous committed 1 year ago
e632e5d
Make ScaleROPE node work on Flux. (#10686)
comfyanonymous committed 1 year ago
2abd2b5
Only unpin tensor if it was pinned by ComfyUI (#10677)
comfyanonymous committed 1 year ago
a1a7036
Clarify release cycle. (#10667)
comfyanonymous committed 1 year ago
e05c907
Pinned mem also seems to work on AMD. (#10658)
comfyanonymous committed 1 year ago
09dc24c
Enable pinned memory by default on Nvidia. (#10656)
comfyanonymous committed 1 year ago
1d69245
Fix qwen controlnet regression. (#10657)
comfyanonymous committed 1 year ago
97f198e
4cd8818
ComfyUI version v0.3.68
comfyanonymous committed 1 year ago
265adad
a389ee0
9c71a66
More fp8 torch.compile regressions fixed. (#10625)
comfyanonymous committed 1 year ago
af4b7b5
0f4ef3a
Fixes (#10621)
comfyanonymous committed 1 year ago
e199c8c
Speed up torch.compile (#10620)
comfyanonymous committed 1 year ago
0652cb8
People should update their pytorch versions. (#10618)
comfyanonymous committed 1 year ago
958a171
chore: update embedded docs to v0.3.1 (#10614)
ComfyUI Wiki committed 1 year ago
e974e55
feat(Pika-API-nodes): use new API client (#10608)
Alexander Piskun committed 1 year ago
4e2110c