A PyTorch native platform for training generative AI models

Post-training with Tinker

SWE-agent takes a GitHub issue and tries to automatically fix it, using your LM of choice. It can also be employed for offensive cybersecurity or competitive coding challenges. [NeurIPS 2024]

[NeurIPS 2025 D&B Spotlight] Scaling Data for SWE-agents

slime is an LLM post-training framework for RL Scaling.

Democratizing Reinforcement Learning for LLMs

Toolchain manager for Roblox projects

SGLang is a high-performance serving framework for large language models and multimodal models.

The open source developer platform to build AI agents and models with confidence. Enhance your AI applications with end-to-end tracking, observability, and evaluations, all in one integrated platform.

Our library for RL environments + evals

Lightweight coding agent that runs in your terminal

Renderer for the harmony response format to be used with gpt-oss

Copilot Chat extension for VS Code

Code at the speed of thought – Zed is a high-performance, multiplayer code editor from the creators of Atom and Tree-sitter.

✨ A synthetic dataset generation framework that produces diverse coding questions and verifiable solutions - all in one framwork

This repo contains the dataset and code for the paper "SWE-Lancer: Can Frontier LLMs Earn $1 Million from Real-World Freelance Software Engineering?"

Model Context Protocol Servers

Train transformer language models with reinforcement learning.

Implementation of the sparse attention pattern proposed by the Deepseek team in their "Native Sparse Attention" paper

TransMLA: Multi-Head Latent Attention Is All You Need (NeurIPS 2025 Spotlight)

s1: Simple test-time scaling

verl: Volcano Engine Reinforcement Learning for LLMs

Fully open reproduction of DeepSeek-R1

⏩ Ship faster with Continuous AI. Open-source CLI that can be used in TUI mode as a coding agent or Headless mode to run background agents

An Easy-to-use, Scalable and High-performance Agentic RL Framework based on Ray (PPO & DAPO & REINFORCE++ & TIS & vLLM & Ray & Async RL)

Code for the paper "Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer"

Evals is a framework for evaluating LLMs and LLM systems, and an open-source registry of benchmarks.

The Open Cookbook for Top-Tier Code Large Language Model

DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.