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
Fix cosmos VAE failing with videos longer than 121 frames.
cca96a8 comfyanonymous committed 1 year ago
More accurate memory estimation for cosmos and hunyuan video.
9d8b6c1 comfyanonymous committed 1 year ago
Force safe loading of files in torch format on pytorch 2.4+
2feb8d0 comfyanonymous committed 1 year ago
Allow setting start and end image in CosmosImageToVideoLatent.
5b657f8 comfyanonymous committed 1 year ago
Rewrite res_multistep sampler and implement res_multistep_cfg_pp sampler. (#6462)
c78a456 Pam committed 1 year ago
Implement Cosmos Image/Video to World (Video) diffusion models.
3aaabb1 comfyanonymous committed 1 year ago
fixed a bug where a relative path was not converted to a full path (#6395)
b9d9bcb Alexander Piskun committed 1 year ago
Hooks Part 2 - TransformerOptionsHook and AdditionalModelsHook (#6377)
6c9bd11 Jedrzej Kosinski committed 1 year ago