Comment by mchiang

Ollama supports it already:

`ollama run codellama:7b-instruct`

https://ollama.ai/blog/run-code-llama-locally

More models uploaded as we speak:

https://ollama.ai/library/codellama