Show HN: Einlang, a math-intuitive language with lots of good stuff

Category: library

Tags: domain-specific-language, tensor-computations, automatic-differentiation, compiler, scientific-computing

Score: 7.3/10 (Innovation: 7, Technical: 8, Documentation: 9, Utility: 5)

Einlang is a domain-specific language and compiler for tensor computations with explicit index notation, built-in automatic differentiation using intuitive math-like syntax, and support for recurrences. It's interesting because it embeds gradient computation directly into the language syntax (@loss/@weights) rather than requiring separate APIs, making tensor code resemble mathematical notation while maintaining compile-time shape checking.

Target audience: data engineers, machine learning researchers, scientific computing developers, numerical analysts

Repository: https://github.com/einlang/einlang · Python · Apache-2.0 · 10 stars

View on Hacker News