spara / autollama

autollama

"Docker compose: how to pull model automatically with container creation?"

autollama solves the problem of downloading an Ollama model automatically. It works by running docker compose exec after the docker conpose file executes.

Update the volume source path to autollama in the docker-compose.yml file.

    volumes:
      - type: bind
        source: </path/to/autollama>
        target: /root/.ollama
        bind:
          create_host_path: true

Then run:

./start.sh