PyHessian is a Pytorch library for second-order based analysis and training of Neural Networks

Jax Codebase for Evolutionary Strategies at the Hyperscale

Crowdsourcing months-long human software engineering trajectories.

Accelerating Research in Plasticity-Motivated Deep Reinforcement Learning.

The most open diffusion language model for code generation — releasing pretraining, evaluation, inference, and checkpoints.

Partially Observable Benchmarks in JAX

verl: Volcano Engine Reinforcement Learning for LLMs

Parallel hyperparameter tuning with JAX

The evaluation framework for training-free sparse attention in LLMs

Train your Agent model via our easy and efficient framework

RAGEN leverages reinforcement learning to train LLM reasoning agents in interactive, stochastic environments.

Our library for RL environments + evals

🏛️A research-friendly codebase for fast experimentation of single-agent reinforcement learning in JAX • End-to-End JAX RL

LeanRL is a fork of CleanRL, where selected PyTorch scripts optimized for performance using compile and cudagraphs.

Fine-tuning & Reinforcement Learning for LLMs. 🦥 Train OpenAI gpt-oss, DeepSeek, Qwen, Llama, Gemma, TTS 2x faster with 70% less VRAM.

Really Fast End-to-End Jax RL Implementations

Accelerated minigrid environments with JAX

CleanDiffuser: An Easy-to-use Modularized Library for Diffusion Models in Decision Making

A pytorch implementation for paper 'Exploring Simple Siamese Representation Learning'

You like pytorch? You like micrograd? You love tinygrad! ❤️

Simplifying reinforcement learning for complex game environments