A toolkit for describing model features and intervening on those features to steer behavior.

AutoBound automatically computes upper and lower bounds on functions.

A framework for few-shot evaluation of language models.

Evaluation suite for LLMs

Implementation of PSGD optimizer in JAX

Lighteval is your all-in-one toolkit for evaluating LLMs across multiple backends

Entropy Based Sampling and Parallel CoT Decoding

Meta Lingua: a lean, efficient, and easy-to-hack codebase to research LLMs.

A virtual monitor for screen sharing

jax-triton contains integrations between JAX and OpenAI Triton

Automatically take good care of your preemptible TPUs

GPT 2 written in *clean* Haliax code

AvaTaR: Optimizing LLM Agents for Tool Usage via Contrastive Reasoning (NeurIPS 2024)

State Space Models library in JAX

🤗 LeRobot: Making AI for Robotics more accessible with end-to-end learning

DOM to Semantic-Markdown for use with LLMs

common in-memory tensor structure

Manipulate JSON-like data with NumPy-like idioms.

Prose diffs for any document format supported by Pandoc

Bevy Scripting Plugin

Machine Learning eXperiment Utilities

PyTorch native post-training library

A massively parallel, high-level programming language

A massively parallel, optimal functional runtime in Rust

Data and tools for generating and inspecting OLMo pre-training data.

A JAX research toolkit for building, editing, and visualizing neural networks.

🙌 OpenHands: Code Less, Make More

Sophia optimizer with ScheduleFree

Data processing for and with foundation models! 🍎 🍋 🌽 ➡️ ➡️🍸 🍹 🍷