"One-to-All Extend Video" Workflow — v1.0

The One-to-All Extend Video workflow is designed to extend an input video with One-to-All pose animation and WanVideo generation, producing a video output.

Vorschlag · nicht geprüft

Its listed nodes cover video and image loading, pose animation, WanVideo text encoding, sampling and decoding, and video combining.

The graph includes video and image loader nodes and a video-combine node, indicating video and image inputs with a video output.

Vorschlag · nicht geprüft

It requires five ComfyUI node packs: ComfyUI-KJNodes, comfyui-various, ComfyUI-VideoHelperSuite, ComfyUI-WanAnimatePreprocess, and ComfyUI-WanVideoWrapper.

The workflow lists six model files: lightx2v_T2V_14B_cfg_step_distill_v2_lora_rank64_bf16.safetensors, One-to-All-14b-Q4_K_M.gguf, umt5-xxl-enc-bf16.safetensors, vitpose-l-wholebody.onnx, wan_2.1_vae.safetensors, and yolov10m.onnx.

Configuration-related nodes include LoRA selection, block swapping, Torch compile settings, and image resizing.

No download URLs are listed for One-to-All-14b-Q4_K_M.gguf, vitpose-l-wholebody.onnx, or yolov10m.onnx; check where to get them before setup.

Vorschlag

No license is shown for this workflow; check the listing terms before use.

Vorschlag