← Latest products

Self-growing neural networks via a custom Rust-to-LLVM compiler

A systems programming language where automatic differentiation is a compiler pass and model parameters are explicit, growable memory. My goal is to treat model parameters as explicit, growable memory buffers.

Open source

27 stars · GitHubexternal site

Visit github.comexternal site

Reactions (1)

Similar products

by description similarity

  1. NeuralScript

    github.com

    A pure-Rust AOT compiler

    Hacker News5 pts

  2. Recursi

    recursi.dev

    self-improving LLM-connected coding environment

    Hacker News6 pts

  3. val

    github.com

    An arbitrary precision calculator language

    Hacker News34 pts

  4. Minigrad

    github.com

    a small neural network lib in Golang

    Hacker News9 pts

  5. We built the Keras of LLMs

    github.com

    Keras based framework for neuro-symbolic LM systems

    Hacker News5 pts