Local LLM bundler with external models:

1. To run, first set variable for the remote location of the offline host path, like this:

export HOST_OFFLINE_PATH="/Volumes/Offline LLM/.ollama/models" # or whatever the external drive location might be.

2. After that, issue

docker compose up -d --build

3. Then go to localhost:8080, go to settings and set the connection to http://localhost:11434/v1.