Features
Uses quantized GGUF models instead of FP16 checkpoints for faster speed and lower VRAM usage
Uses TripleKSampler for better prompt adherence and improved motion
Uses LightX2V Lightning LoRAs for blazing fast speed (5 minutes for 81-frame video on RTX 4080)
Uses RIFE VFI for quadruple FPS boost (to buttery smooth 60 FPS)
Custom ComfyUI Nodes
Download Models
Unet
The following files should be saved into /models/unet
Text Encoders
The following files should be saved into /models/text_encoders
VAE
The following files should be saved into /models/vae
LoRAs (for speed)
The following files should be saved into /models/loras
Description
Added RIFE VFI to increase frame rate to 60FPS
FAQ
Comments (24)
How is the seed being set? I can't find it in the workflow.
The seed is set (currently made to randomize after each generation) in the TripleKSampler node under the Process Video group
Amazing workflow, thank you very much for sharing it!
Thank you for your kind words!
works great but how can we extend the clips?
By increasing the video lenght value. But it creates looping.
I intstalled every ComfyUi node listed here but it still tells me I miss:
- VHS_VideoCombine , SetNode , GetNode , UnetLoaderGGUF , LoadAndResizeImage
Where I can find these?
Sounds like these nodes aren't installed properly:
* https://github.com/Kosinkadink/ComfyUI-VideoHelperSuite
@wildkrauss It works now :) I installed it via ´git clone´ but seems like that didnt work properly for some reason so had to use Comfy Manager and fix it with that.
I'm getting the error:
Cannot execute because a node is missing the class_type property.: Node ID '#93:97'
any solutiosn to this problem?
That's a weird error I haven't seen before. Have you installed all the missing custom nodes? Which node is the error coming from?
me too, im looking for a solution
Edit:- Its working now, well on my 6GB vram i had to reduce some parameters and the video was too blurry, but 60fps, but it did undress greatly. Im gonna play with it for quite some time before understanding the right settings especially for my spec. Thanks mate.
It requires custom nodes called TripleKSamplerWan22Lightning and FreeMemoryLatent in subgraph 'Subgraph Generate Video', which I never couldn't find anywhere after hours of searching.
Weird, ComfyUI Manager automatically finds and installs those custom nodes for me, but perhaps it works differently for different ComfyUI versions.
Here they are:
视频的裁剪好像不对 虽然默认是竖版视频 但是必须要放横版图片
Is there any way to further optimize this to work on a 6gb vram potato of a system?
This is a great workflow. Thank you so much. Soon I will share my work. By the way how do you prevent looping? Because wan sees 81 or 161 frame rate but video is much longer because of RIFE. I couldn't find a way...
Wan 2.2 by itself seems to have a looping problem when a video goes longer than 5 seconds.
SVI 2 Pro which was recently released seems to do an amazing job at extending Wan 2.2 videos, but I haven't gotten around to testing it yet. I'll probably release a new version of this workflow when I do.
@wildkrauss Eagerly waiting for your new workflow. I have a lot way to reach your workflow professionalism.Thank you again.
It works but not always great, but anyway for nsfw staff, there is no other choice, for sfw staff the cloud AI is way better.
Motion is amazing! Clean workflow and fast.
My only issue is that I'm getting a crazy amount of ghosting. I have 11gb of vram. I wonder if there is a fix and/or if I'm the only one with this issue.
Edit: It was actually an issue with the name of the loras. The new versions have a different name and they needed to be re-selected.
Can't download all the required nodes.
Can i use it with a 9060XT 16VRAM and 32 RAM?
FOR THIS WORKFLOW ALL ARE HERE FOR MODEL NEED TO CHANGE LINK DIRECT LINK IN COLAB GOOGLE
comfy
git clone https://github.com/comfyanonymous/ComfyUI
cd ComfyUI
sudo apt install -y python3.10 python3.10-venv python3.10-distutils
python3.10 -m venv py310_env
source py310_env/bin/activate
pip install -r requirements.txt
cd custom_nodes
git clone https://github.com/Wan-Video/ComfyUI-WanVideoWrapper
cd ComfyUI-WanVideoWrapper
pip install -r requirements.txt
!wget https://bin.equinox.io/c/bNyj1mQVY4c/ngrok-v3-stable-linux-amd64.zip
!unzip ngrok-v3-stable-linux-amd64.zip
!chmod +x ngrok
!./ngrok config add-authtoken YOUR_AUTH_TOKEN
!pkill ngrok
!./ngrok http 8188 --log=stdout
%cd /content/ComfyUI/custom_nodes
!git clone https://github.com/rgthree/rgthree-comfy.git
!git clone https://github.com/kijai/ComfyUI-KJNodes.git
!git clone https://github.com/Kosinkadink/ComfyUI-VideoHelperSuite.git
!git clone https://github.com/VraethrDalkr/ComfyUI-TripleKSampler.git
!git clone https://github.com/kijai/ComfyUI-WanVideoWrapper.git
!git clone https://github.com/Fannovel16/ComfyUI-Frame-Interpolation.git
!git clone https://github.com/ShmuelRonen/ComfyUI-FreeMemory.git
!git clone https://github.com/city96/ComfyUI-GGUF
!git clone https://github.com/pythongosssss/ComfyUI-Custom-Scripts