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
By default only accept requests where origin header matches the host.
3ab3516 comfyanonymous committed 2 years ago
Add an option to the SaveLora node to store the bias diff.
a09b29c comfyanonymous committed 2 years ago
LoraSave node now supports generating text encoder loras.
9bfee68 comfyanonymous committed 2 years ago
Update stable release workflow to latest pytorch with cuda 12.4.
dc2eb75 comfyanonymous committed 2 years ago
Add an experimental LoraSave node to extract model loras.
22d1241 comfyanonymous committed 2 years ago
Make minimum_inference_memory() depend on --reserve-vram
b643eae comfyanonymous committed 2 years ago
add github action to automatically handle stale user support issues (#4683)
d4aeefc Alex "mcmonkey" Goodwin committed 2 years ago