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
Do lora cast on GPU instead of CPU for higher performance.
b92bf81 comfyanonymous committed 3 years ago
Move ModelSubtract and ModelAdd to advanced/model_merging
0665749 comfyanonymous committed 3 years ago
Merge branch 'fix/preview_ratio' of https://github.com/ltdrdata/ComfyUI
d6d9b83 comfyanonymous committed 3 years ago
Add cond_or_uncond array to transformer_options so hooks can check what is cond and what is uncond.
43d4935 comfyanonymous committed 3 years ago
Merge branch 'Fix-structuredClone-error-with-early-chrome-version-browser' of https://github.com/KarryCharon/ComfyUI
0992260 comfyanonymous committed 3 years ago
Support for text encoder models that need attention_mask.
44361f6 comfyanonymous committed 3 years ago
Set last layer on SD2.x models uses the proper indexes now.
0d8f376 comfyanonymous committed 3 years ago
Don't run text encoders on xpu because there are issues.
0966d3c comfyanonymous committed 3 years ago
Don't leave very large hidden states in the clip vision output.
ed58730 comfyanonymous committed 3 years ago
Fix issue where autocast fp32 CLIP gave different results from regular.
fb3b728 comfyanonymous committed 3 years ago
Merge branch 'master' of https://github.com/miabrahams/ComfyUI
d4b2bc0 comfyanonymous committed 3 years ago