LLM Manager UI

Web UI to manage Ollama LLM server models

Features

• View running models
• Lists all installed models
• Pull new models
• Update models (manually via update button)
• Delete old models
• Chat / Generate with any model

Deploy

bash start.sh [options]

Options:

• -h, --host HOST - Ollama server host (default: localhost)
• -p, --port PORT - Ollama server port (default: 11434)
• --help - Show help message

The script will:

• Create a Python virtual environment (if it doesn't exist)
• Install dependencies from requirements.txt
• Start the UI server on http://localhost:5000

Docker

docker-compose up -d --build

• Change ENV in docker-compose.yaml if needed.
• OLLAMA_URL - URL of the Ollama server (default: http://localhost:11434)
• LOG_LEVEL - Logging verbosity (default: INFO)
  • Options: DEBUG, INFO, WARNING, ERROR, CRITICAL
  • INFO recommended for production (shows important operations without noise)
  • DEBUG for troubleshooting (shows all operations including frequent fetches)

Access the UI at http://localhost:5000

Contribution

Feel free to open issues or submit pull requests.