/are-we-learning-yet

How ready is Rust for Machine Learning?

Primary LanguageRustCreative Commons Attribution 4.0 InternationalCC-BY-4.0

are-we-learning-yet

Rust is a systems programming language, but is it a machine learning language?

arewelearningyet.com

Inspired by Are We Web Yet?, this project aims to catalog the the Rust ML ecosystem.

Contributing

Feedback, issues, and pull requests are welcome and appreciated for adding missing crates, providing additional resources, or improving the content.

Running locally

# GitHub OAuth token avoids 403 rate limiting errors while generated crate data
# Tip: set in `.env` file and `just` will pick it up automatically
export GITHUB_TOKEN=<YOUR_GITHUB_TOKEN>

# Scrape crate/repo data for sitegen
just scrape

# Start a dev server on port 3000
cobalt serve

The site should be running on localhost:3000

How it works

The repo consists of 2 key parts:

  • the scraper tool is responsible for reading crates.yaml, fetching additional metadata about the crate from crates.io and the GitHub API, and generating a score used for ordering crates. The fetched data is cached in a _tmp directory to speed up repeated site generation and avoid abusing the APIs, and the scraper outputs _data/crates_generated.yaml which is used by the cobalt site generation. To force regeneration of all crate data, remove the cached data with just clean. For more details, see the scraper README.
  • the site content is the rest of the repo which follows the layout established by cobalt.rs. Cobalt uses this content to generate a static site into a _site directory (cobalt build) and can also start a local development server that rebuilds the site anytime content changes (cobalt serve). For more details, see cobalt.rs documentation.

Note: cobalt serve does not trigger the scraper to rerun when crates.yaml (or the scraper) changes. Currently you must rerun just scrape to update crates_generated.yaml.

Publishing

arewelearningyet.com is served by Github Pages. Every merge into master is automatically published by a Github Actions job.

Additionally, to ensure crate statistics (download counts and stars) are regularly updated, the publishing task is also run as a weekly cron job.