search

MichiganDataScienceTeam / F24-mini-copilot

Building and deploying a lightweight code autocompletion tool, from GPT-2 weights to a working VSCode extension.
MIT License
9 stars 0 forks source link

Model evaluation #32

Open Maxim12313 opened 2 weeks ago

Maxim12313 commented 2 weeks ago

Ways to measure how well the model is performing: Perplexity Human Eval (optional)

This is on feat/evaluation: https://github.com/MichiganDataScienceTeam/F24-mini-copilot/tree/feat/evaluation

I will be gone this weekend, just opening this issue so anyone can get started

danaiamirali commented 1 week ago

Perplexity is done, human eval remains if anyone is interested