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
Update optimized_attention_for_device function for new functions that support masked attention.
c695154 comfyanonymous committed 2 years ago
Fix model patches not working in custom sampling scheduler nodes.
ef4f603 comfyanonymous committed 2 years ago
Add support for the stable diffusion x4 upscaling model.
a7874d1 comfyanonymous committed 2 years ago
Fix issue when websocket is deleted when data is being sent.
8e2c99e comfyanonymous committed 2 years ago
Add node id and prompt id to websocket progress packet.
66831eb comfyanonymous committed 2 years ago
Only add _meta title to api prompt when dev mode is enabled in UI.
e478b17 comfyanonymous committed 2 years ago
Fix lowvram mode not working with unCLIP and Revision code.
61b3f15 comfyanonymous committed 2 years ago