[Step-by-Step] ComfyUI + Wan 2.1 Ultimate Setup Guide for AI Image/Video Generation
Welcome to Tube U channel! In this comprehensive guide, learn how to set up ComfyUI and integrate it with Wan 2.1 for seamless image and video generation. Follow simple, step-by-step instructions to prepare your environment, install dependencies, and optimize workflows for stunning AI-powered creations. It run smoothly on my RTX2070 8G VRAM GPU.
Discover tips for troubleshooting errors, enhancing ComfyUI with custom nodes, and generating high-quality videos efficiently. Whether you're a beginner or an AI enthusiast, this tutorial has everything you need to get started. Don't forget to subscribe for more helpful content!
Content
00:00 Introduction
00:19 Environments
01:45 Install ComfyUI
03:38 Install Wan 2.1
04:12 Integrating Wan 2.1 with ComfyUI
05:23 Enhancing ComfyUI with Useful Nodes
06:31 Tips and Troubleshooting
python
https://www.python.org/
Git
https://git-scm.com/
ComfyUI
https://www.comfy.org/
Github
https://github.com/comfyanonymous/Com...
wan 2.1
https://github.com/Wan-Video/Wan2.1
Integrating Wan 2.1 with ComfyUI
https://comfyanonymous.github.io/Comf...
umt5_xxl_fp8_e4m3fn_scaled.safetensors
goes in: ComfyUI/models/text_encoders/
wan_2.1_vae.safetensors
goes in: ComfyUI/models/vae/
wan 2.1 models:
https://huggingface.co/Comfy-Org/Wan_...
For low VRAM old GPU, wan2.1_t2v_1.3B_fp16.safetensors is essentially necessary since it is only 2.84 GB.
Other models you can select fit your GPU
Useful ComfyUI node:
ComfyUI manager
https://github.com/Comfy-Org/ComfyUI-...
goto ComfyUI/custom_nodes dir in terminal(cmd)
git clone https://github.com/ltdrdata/ComfyUI-M... comfyui-manager
Restart ComfyUI
rgthree-comfy (make ComfyUI comfortable)
https://github.com/rgthree/rgthree-comfy
Crystools (🪛 A powerful set of tools for your belt when you work with ComfyUI 🪛) (monitor your resource)
https://github.com/crystian/ComfyUI-C...
City96
https://huggingface.co/city96/Wan2.1-...
Helpful Commands:
install requirements
pip install -r requirements.txt
Update dependencies:
pip install --upgrade -r requirements.txt
Check GPU compatibility:
python -c "import torch; print(torch.cuda.is_available())"
Integrating Wan 2.1 with ComfyUI
https://comfyanonymous.github.io/Comf...
Torch not compiled with CUDA enabled" error
pip uninstall torch
Then install it again
pip install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu126
module 'torch' has no attribute 'float8_e4m3fn' error
This fixed it for me:
First, uninstall any existing versions of torch and xformers
pip3 uninstall torch torchvision torchaudio xformers -y
Then reinstall the modules:
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu124
The pip3 command was found on this page: https://pytorch.org/
Edit: I know that xformers is missing in the install command. Maybe you need to add it. It worked for me so I didn't touch it.
pip3 install torch torchvision torchaudio xformers --index-url https://download.pytorch.org/whl/cu124
If you encounter mat1 and mat2 mismatch RunTimeError,
download the correct file (model, VAE and text_encoders must match together)
Subscribe to my channel
/ @tubeunderdeveloped
Now you can download more than 120 unique models for free https://www.thingiverse.com/zhwang168...
Twitter
/ play4fun168
Facebook
/ zhwang168
#TubeUnderdeveloped #Howto #3dPrinting #AICorner