A field guide for Python refugees hunting ML crates in Rust

What it does This is a curated awesome-list that maps out machine learning libraries and resources written in Rust. It sorts crates, blog posts, books, papers, and even podcasts into categories like dataframes, deep neural networks, and reinforcement learning. The list explicitly targets developers considering a move from Python, annotating which libraries are maintained, which are small or abandoned, and where the helpful code lives.

The interesting bit Rather than dumping links, the author adds editorial commentary—suggesting polars for dataframes, plotters for graphing, and evcxr for Jupyter-style notebooks. That turns a simple index into an opinionated survival manual for developers who need Rust equivalents to sklearn, pandas, and matplotlib.

Key highlights

• Covers the full ML stack: linear algebra, image processing, NLP, AutoML, GPU support, and evolutionary algorithms.
• Includes libraries that are no longer maintained alongside active ones, which saves you from starring a ghost repo.
• Annotated recommendations within each category (e.g., ndarray vs. nalgebra for vectors).
• Collects adjacent media like books, movies, podcasts, and academic papers—not just code.
• Available in both English and Japanese.

Caveats

• It catalogs unmaintained libraries alongside active ones, so you’ll still need to check commit history before picking a crate.
• It is an index, not a tutorial; expect pointers, not prose on how to train a model.

Verdict Worth bookmarking if you’re a Python ML practitioner evaluating Rust, or if you already write Rust and need a lay of the land. Pure application developers who have no interest in Rust or ML can safely skip it.