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
Add ipndm_v sampler, works best with the exponential scheduler.
d5efde8 comfyanonymous committed 2 years ago
Support loading diffusers SD3 model format with UNETLoader node.
0d6a579 comfyanonymous committed 2 years ago
Add soundfile dependency so that windows can save audio.
55f0dc1 comfyanonymous committed 2 years ago
Better estimation for memory usage during audio VAE encoding/decoding.
ca9d300 comfyanonymous committed 2 years ago
Optimize some unneeded if conditions in the sampling code.
f2e844e comfyanonymous committed 2 years ago
Add a --force-channels-last to inference models in channel last mode.
0ec513d comfyanonymous committed 2 years ago
Exempt IPEX from non_blocking previews fixing segmentation faults. (#3708)
5eb98f0 Simon Lui committed 2 years ago
Fix regular empty latent image not working with SD3 and custom sampler.
0eaa34e comfyanonymous committed 2 years ago