To start the server for the first time:

docker run -d -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama

To stop and start

docker stop ollama
docker start ollama

To interact with it:

docker exec -it ollama ollama pull gemma2
docker exec -it ollama ollama run gemma2

FYI: gemma2 needs more than 8GB of RAM to run.

→ docker exec -it ollama ollama run gemma2
Error: model requires more system memory (9.1 GiB) than is available (8.4 GiB)

Questions

• How we can move the models to an external SSD?

Moving Models to External SSD on macOS

To move Ollama models to an external SSD when using Docker:

1. Stop the Ollama container:

   docker stop ollama
   

2. Create a directory on your external SSD:

   mkdir /Volumes/YourSSD/ollama-models
   

3. Update the Docker run command to mount the external SSD location:

   docker run -d \
     -v /Volumes/YourSSD/ollama-models:/root/.ollama \
     -p 11434:11434 \
     --name ollama \
     ollama/ollama
   

If you already have models and want to move them:

1. Copy existing models from the Docker volume to your SSD:

   docker cp ollama:/root/.ollama/. /Volumes/YourSSD/ollama-models
   

2. Remove the old Docker volume:

   docker volume rm ollama
   

3. Start Ollama with the new mount point as shown above.

#Development #Docker #AI #LLMs