To get ollama up and running with NVIDIA GPU's run the following command: docker-compose up -d then you will exec into that container and run the install command of the model you want for example: ollama pull llama3.2:1b