Open WebUI with Ollama Setup Script

This repository contains a shell script to set up and configure Open WebUI with Ollama.

Open.WebUI.Intro.mp4

My System Configuration:

• Processor: Intel Core i7-11800H (8 Cores, 16 Threads) @ 2.30GHz
• RAM: 16GB
• GPU: NVIDIA GeForce RTX 3050 Ti Laptop GPU

Prerequisites

Before running the script, ensure you have the following installed on your system:

• Docker
• GPU drivers compatible with Docker and NVIDIA
• Sufficient disk space for the models and data volumes
• The models folder copied from AUTOMATIC1111 GitHub repository to the same directory as the docker-compose.yml

Usage

1. Clone the repository or copy the script: Save the setup_open_webui.sh and the docker-compose.yml files in your preferred directory.

2. Make the script executable:

   chmod +x setup_open_webui.sh

3. Run the script:

   ./setup_open_webui.sh

What the Script Does

1. Runs the Docker containers:

   • Initializes and builds the necessary containers using docker compose up -d --build. This command starts both the Open WebUI and Stable Diffusion services as defined in the docker-compose.yml.
   • Allows the containers to utilize GPU resources for accelerated processing.
   • Configures volumes.
     • auto1111-data and auto1111-output for Stable Diffusion.
     • ollama and open-webui for Open WebUI and Ollama model storage.
   • Exposes ports.
     • Port 7860 is exposed for Stable Diffusion, enabling API access to the diffusion models.
     • Port 3000 is exposed for Open WebUI, providing access to the application interface.
   • Automatically restarts the container on failure.

2. Downloads required AI models:

   • Pulls the llama3.2 model.
   • Pulls the llava model.
   • Pulls the DeepSeek-R1-Distill-Qwen-7B model.
   • Pulls the stable-diffusion-prompt-generator model.
   • Pulls the codellama model.
   • Pulls the DeepSeek-R1-Distill-Llama-8B model.

3. Ensures smooth setup with error handling:

   • Checks for errors after each command and exits if an error occurs.

Accessing Open WebUI Interface

Once the script finishes:

1. Open a browser and navigate to http://localhost:3000 to access the web interface.
2. Start interacting with the models.

Image Generation Setup

1. Go to Profile Icon -> Settings -> Admin Settings -> Images.
2. Select Automatic1111 as Image Generation Engine.
3. Paste http://stable-diffusion:7860 in the AUTOMATIC1111 Base URL and click on the refresh icon at the right side of the text box to connect to the stable diffusion server. Note that stable-diffusion in the URL is the name of the docker container.
4. Turn On Image Generation.
5. Set v1-5-pruned-emaonly as the default model.
6. Save the settings.

RAG Setup

Open.WebUI.RAG.mp4

Troubleshooting

If the setup script fails:

• Ensure Docker is installed and running.
• Verify that your GPU drivers and Docker GPU support are configured properly.
• Check if ports are available (default is 3000).

For additional support, please refer to Open WebUI Documentation.