Need to update ComfyUI-WanStartEndFramesNative to 1.0.5 in ComfyUI Manager.
This generates a video between the start image and the end image. Since it is a native implementation, various model optimization nodes such as gguf, teacache, etc. are supported, and LoRA is also supported.
https://github.com/Flow-two/ComfyUI-WanStartEndFramesNative
Features
Video generation without positive prompts
Support LoRA
Support GGUF
Support TeaCache
Support Enhance A Video
Support Skip Layer Guidance
Nodes
ComfyUI-WanStartEndFramesNative (1.0.4)
ComfyUI-KJNodes
ComfyUI-VideoHelperSuite
ComfyUI-GGUF
Description
FAQ
Comments (10)
Worked for me, using a 3060 12GB took 18 min
I've tested both Fun models & 14B GGUF. The FUN models produce sub-optimal results, worse that LTXVideo keyframes. Not sure if it is a workflow issue. The GGUF model however is both brilliant and hilarious.
This is a good workflow.
[fun 14b model]
However, there is an issue where the screen darkens, as if a black filter is applied, in the last frame segment. In this case, you can use the colormatch node together to improve the final video, but it is a bit inconvenient.
+fun 14b GGUF model link
https://huggingface.co/city96/Wan2.1-Fun-14B-InP-gguf
Perfect for looping videos!
Consistent shadow intensity, perfect one-way loop! I hope this supports mid-frame generation with 4x scaling. It will be awesome!
Damn, you always surprise me!
Not sure if / where I've done something wrong - I'm on V1.4, constantly getting this sort of blurry transition effect, not much of a video.
Edit: I think it's because I was using the wrong aspect ratio - once I changed it to square like the source & dest images, it started working very well.
I suspect perhaps the dest needs to be auto cropped as the source appears to be.
What does the "FUN" model do and what do you hook that model up to in the workflow to use it?
I can't find the node for wanfuninpainttovideo, how can I download it? I have already installed the ComfyUI-WanStartEndFramesNative custom node.
Thanks for the Wf !!!
but :UnetLoaderGGUF
Unexpected architecture type in GGUF file, expected one of flux, sd1, sdxl, t5encoder but got 'wan'
i updated and it's ok 😁
somehow this workflow uses a lot more memory than the native one for me. Takes way longer too.
Hi. And why do the last 6 frames become faded when the rendering length is 17 frames?
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.