Inspiration

We wanted to generate documentation from a snippet of code, but this proved overambitious since we did not have a dataset, and I essentially was trying to do the reverse of GitHub copilot, which is surely a multi-million dollar project. We settled for receiving a code snippet and trying to classify the programming language.

What it does

Guess the programming language from a small snippet of code

How we built it

Using sklearn, a support vector machine, a Flask backend and an HTMX/Bootstrap frontend

Challenges we ran into

Originally, we had the intention of using an RNN, which is more suited to our problem, but could not get it to work in time. We resorted to a SVM in the end.

Accomplishments that we're proud of

The web app is clean, lean and actually fun to interact with (because the predictions are so bad :P)

Share this project:

Updates