Ranker Sample Code is Here

A minimal example of what you need to submit

This repo contains a minimal ranker that just sorts items by sentiment, and adds one new post. It also includes a small test program that calls the ranker with some sample data. It’s written with Python, Flask, and NLTK.

There’s not much to it. In fact you’ll want to read the API documentation for many important details. But if you put this tiny server up on the public web somewhere, you would have a correctly formed first-round submission. Just don’t expect to win with sample code 🙃

We will soon have sample code for more complex use cases, including calling out to OpenAI for GPT (which is possible for prototypes), running the open-source Mistral LLM (you cannot call external services in production submissions), and using a separate write-only process to scrape data for your ranker. If there’s something you’d particularly like to see, let us know!