[wrsh07]

You can run it all locally: https://github.com/ggml-org/llama.vscode