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

d5efde8
feat: add gits scheduler (#3769)
Zhenyu Zhou committed 2 years ago
45e10ca
Fix routes (#3790)
Chenlei Hu committed 2 years ago
d7f0964
Fix issue with full diffusers SD3 loras.
comfyanonymous committed 2 years ago
028a583
0d6a579
Remove empty line.
comfyanonymous committed 2 years ago
b08a9dd
Update sd1_clip.py (#3684)
Mario Klingemann committed 2 years ago
eee815e
Add ipndm sampler.
comfyanonymous committed 2 years ago
e11052a
Add api/ prefix to api endpoints (#3779)
Chenlei Hu committed 2 years ago
97ae6ef
Support full SD3 loras.
comfyanonymous committed 2 years ago
3914d5a
55f0dc1
Basic tiled decoding for audio VAE.
comfyanonymous committed 2 years ago
a45df69
fix app.js no graph defined (#3754)
Juanjuan committed 2 years ago
379ff92
Fix lora keys for SimpleTuner (#3759)
Janek Mann committed 2 years ago
b7c473d
Use fp16 as the default vae dtype for the audio VAE.
comfyanonymous committed 2 years ago
6425252
Squash depreciation warning on new pytorch.
comfyanonymous committed 2 years ago
8ddc151
Fix VAEEncode with taesd3.
comfyanonymous committed 2 years ago
746a041
Improvements to the TAESD3 implementation.
comfyanonymous committed 2 years ago
04e8798
support TAESD3 (#3738)
Dr.Lt.Data committed 2 years ago
df7db0e
Initial support for the stable audio open model.
comfyanonymous committed 2 years ago
bb1969c
Small optimization.
comfyanonymous committed 2 years ago
1281f93
f2e844e
Print key names for easier debugging.
comfyanonymous committed 2 years ago
0e06b37
Support SD3 diffusers lora.
comfyanonymous committed 2 years ago
ac151ac
Support setting weight offsets in weight patcher.
comfyanonymous committed 2 years ago
37a08a4
Fix lowvram issue.
comfyanonymous committed 2 years ago
605e64f