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
Support for qwen edit plus model. Use the new TextEncodeQwenImageEditPlus. (#9986)
1fee882 comfyanonymous committed 1 year ago
[Reviving #5709] Add strength input to Differential Diffusion (#9957)
9ed3c5c Jedrzej Kosinski committed 1 year ago
Add inputs for character replacement to the WanAnimateToVideo node. (#9960)
66241ce comfyanonymous committed 1 year ago
Update WanAnimateToVideo to more easily extend videos. (#9959)
e8df53b comfyanonymous committed 1 year ago
fix(seedream4): add flag to ignore error on partial success (#9952)
852704c Alexander Piskun committed 1 year ago
api_nodes: reduce default timeout from 7 days to 2 hours (#9918)
9fdf8c2 Alexander Piskun committed 1 year ago
Do padding of audio embed in model for humo for more flexibility. (#9935)
24b0fce comfyanonymous committed 1 year ago
make kernel of same type as image to avoid mismatch issues (#9932)
1ea8c54 Jodh Singh committed 1 year ago
Add encoder part of whisper large v3 as an audio encoder model. (#9894)
a39ac59 comfyanonymous committed 1 year ago
Fix depending on asserts to raise an exception in BatchedBrownianTree and Flash attn module (#9884)
1a85483 blepping committed 1 year ago