Image to Video using Wan 2.2 and Lightx2V and GGUF.
If you are having slowness issues, then you might need to get the smallest version of GGUF which will take up the least amount of VRAM.
High Noise i2v model
https://huggingface.co/QuantStack/Wan2.2-I2V-A14B-GGUF/tree/main/HighNoise
Low Noise i2v model
https://huggingface.co/QuantStack/Wan2.2-I2V-A14B-GGUF/tree/main/LowNoise
Lightx2v LORA for speed up
https://huggingface.co/lightx2v/Wan2.1-I2V-14B-480P-StepDistill-CfgDistill-Lightx2v/tree/main/loras
VAE
https://huggingface.co/QuantStack/Wan2.2-T2V-A14B-GGUF/tree/main/VAE
If you encounter an issue similar to this :
Given groups=1, weight of size [5120, 36, 1, 2, 2], expected input[1, 32, 21, 104, 60] to have 36 channels, but got 32 channels instead.
If you have the custom node, flow2-wan-video installed, this may be causing issues with Wan 2.2
You can try deleting this node.
If above does not work, then you may need to update comfy
You will need to update comfy by running this batch file (updating from within comfy seems unreliable)
update_comfyui.bat
Description
New release
FAQ
Comments (3)
WHAT??? The error you mention is NOT fixed by updated ComfyUI but by deleting a flux2 node that a user will have installed an age ago and long forgotten. Somehow that faulty node, although not in use, breaks some Wan2.2 workflows.
thank you, will try it out, your t2v worked really well.
Thank you for sharing. Works well with the new lightx2v high/low lora for WAN2.2. I am using step size of 12 to bump up the quality
Looks like we don't have an active mirror for this file right now.
CivArchive is a community-maintained index — we catalog mirrors that volunteers upload to HuggingFace, torrents, and other public hosts. Looks like no one has uploaded a copy of this file yet.
Some files do get recovered over time through contributions. If you're looking for this one, feel free to ask in Discord, or help preserve it if you have a copy.