Run ComfyUI with OneDiff

Download the Image

Run the following command to download the pre-installed image of OneDiff Enterprise Edition with ComfyUI:

docker pull oneflowinc/comfyui-onediff:latest

Run the Image

docker run -it --shm-size=8G -P --privileged --runtime=nvidia --rm \
    --gpus all --network host \
    -v /path/to/comfyui/models/checkpoints:/app/ComfyUI/models/checkpoints \
    oneflowinc/comfyui-onediff:latest python /app/ComfyUI/main.py

NOTE

• We have preloaded some models within the Docker image, with their respective paths being /app/ComfyUI/models/unet_int8/unet_int8, /app/ComfyUI/models/loras/watercolor_v1_sdxl_lora.safetensors. Please be cautious not to overwrite it due to mapping paths if you want to use them.
• The directory structure of /app/ComfyUI/models within the docker image should follow the structure of ComfyUI/models which means:

models/
├── checkpoints
├── clip
├── clip_vision
├── configs
├── controlnet
├── diffusers
├── embeddings
├── gligen
├── hypernetworks
├── loras
├── style_models
├── unet
├── upscale_models
├── vae
└── vae_approx

How to Use

In the menu's section "Add Node/utils", there is a "Model Speed Up" Node that takes a MODEL as input and outputs an accelerated model.

Related Workflows

Load the screenshot images below in ComfyUI to try the workflow

SDXL Workflow (ComfyUI default workflow)

LoRA

OneDiff Enterprise Compared with Origin ComfyUI