Second Mate

An open-source, mini imitation of GitHub Copilot using EleutherAI GPT-Neo-2.7B (via Huggingface Model Hub) for Emacs.

This is a much smaller model so will likely not be as effective as Copilot, but can still be interesting to play around with!

Setup

Inference End / Backend

1. Set device to “cpu” or “cuda” in serve/server.py
2. The “priming” is currently done in Python. If you want, modify it to another language or turn it off (priming subjectively seems to help).
3. Launch serve/server.py. This will launch a Flask app which will allow us to sample the model via REST API.

Emacs

1. In emacs/secondmate.py, set the URL to “localhost” or the address the API is running on.
2. Configure Python and script path in emacs/secondmate.el. NOTE: The local Python script is a temporary patch which will be replaced by a GET request in Elisp directly.