Skip to content

Latest commit

 

History

History
51 lines (39 loc) · 1.81 KB

README.md

File metadata and controls

51 lines (39 loc) · 1.81 KB

Ollama Web Podman

One script to run them all!

This script will help you run:

  • Ollama (CUDA/ROCm/CPU)
  • ComfyUI with FLUX.1-dev support (CUDA/ROCm)
  • SearXNG
  • Open WebUI

Prerequisites

Quick Setup

Clone the repository

git clone --recursive https://github.com/lslowmotion/ollama-web-podman.git

Run run.sh and follow the instructions.

Address/Ports

Service Address/Ports
Ollama http://localhost:11434
ComfyUI http://localhost:7860
SearXNG http://localhost:4000
Open WebUI http://localhost:3000

Deployment

The run script will ask about your choice of deployment of various services listed above.

Choice Description
Local Service would only accessible by browser on the local machine
Network Service would be accessible by others on the network

Important Notes

Login to Open WebUI

Open WebUI will show you login page at first. Just sign up and you'll be the admin of the Open WebUI.

FLUX.1 Support

FLUX.1 VAE need to be downloaded manually because of HuggingFace login requirements.

Download link and instructions

VAE would need to be put in stable-diffusion-webui-podman/data/models/VAE

Integrating ComfyUI Image Generation to Open WebUI

ComfyUI workflow would need to be set up first to be able to be used with Open WebUI.

Setting Up Open WebUI with ComfyUI