https://github.com/HM RunningHub/ComfyUI_RH_FramePack

https://github.com/lllyasviel/FramePack

FramePack is a progressive neural network structure for predicting the next frame (next partial frame) of a video.

FramePack compresses input contexts to a constant length so that the generation workload is invariant to video length.

Even on laptop GPUs, FramePack can process a very large number of frames with 13B models.

FramePack can be trained with a much larger batch size, similar to the batch size for image diffusion training.

Video diffusion, but feels like image diffusion.

FramePack compresses input contexts to a constant length so that the generation workload is invariant to video length.

FramePack can process a very large number of frames with 13B models even on laptop GPUs.

FramePack can be trained with a much larger batch size, similar to the batch size for image diffusion training.

Video diffusion, but feels like image diffusion.