3 ms·

You work with Linux. On a mac, you run ollama on the host because the gpu is not available in the container.