rl

Llama中文社区，实时汇总最新Llama学习资料，构建最好的中文Llama大模型开源生态，完全开源可商用

An elegant PyTorch deep reinforcement learning library.

Dopamine is a research framework for fast prototyping of reinforcement learning algorithms.

Agent Reinforcement Trainer: train multi-step agents for real-world tasks using GRPO. Give your agents on-the-job training. Reinforcement learning for Qwen3.6, GPT-OSS, Llama, and more!

The RL Bridge for LLM-based Agent Applications. Made Simple & Flexible.

EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL

🚀 An open-source, hands-on curriculum bridging the gap from basic RL concepts to LLM alignment, RLVR, and advanced Agentic systems.

agent-sandbox enables easy management of isolated, stateful, singleton workloads, ideal for use cases like AI agent runtimes and reinforcement learning (RL).

An implementation of the AlphaZero algorithm for Gomoku (also called Gobang or Five in a Row)

A modular, primitive-first, python-first PyTorch library for Reinforcement Learning.

A training framework for Stable Baselines3 reinforcement learning agents, with hyperparameter optimization and pre-trained agents included.

Implementation of papers in 100 lines of code.

MuZero

A Survey of Reinforcement Learning for Large Reasoning Models

Implementation of all RL algorithms in a simpler way

Scalable RL solution for advanced reasoning of language models

[ICLR 2026] SimpleVLA-RL: Scaling VLA Training via Reinforcement Learning

Latest Advances on System-2 Reasoning

Understanding R1-Zero-Like Training: A Critical Perspective

🎓 系统性大语言模型构建课程｜🛠️ 覆盖预训练数据工程、Tokenizer、Transformer、MoE、GPU 编程 (CUDA/Triton)、分布式训练、Scaling Laws、推理优化及对齐 (SFT/RLHF/GRPO)｜🚀 6 个渐进式作业 + 代码驱动，建立 LLM 全栈认知体系

[NeurIPS 2025] TTRL: Test-Time Reinforcement Learning

[ICLR 2026] Agentic Reinforced Policy Optimization (ARPO)

Reinforcement Learning via Self-Distillation (SDPO)

The Continuous-Improvement Stack for Agents. Our environment data and evals power agent improvement and monitoring.

Tensorlake is a serverless runtime for sandboxes and deploying background agentic applications

AI-Native Risk Intelligence Systems, OpenDeRisk——Your application system risk intelligent manager provides 7* 24-hour comprehensive and in-depth protection.

Python library for Reinforcement Learning.

“AI-Compass”将为社区指引在 AI 技术海洋中航行的方向，无论你是初学者还是进阶开发者，都能在这里找到通往 AI 各大方向的路径。旨在帮助开发者系统性地了解 AI 的核心概念、主流技术、前沿趋势，并通过实践掌握从理论到落地的全过程。

3D-printed open-source humanoid robot platform for sim-to-real and RL

[EMNLP 2026] MobileGym: A Verifiable and Highly Parallel Simulation Platform for Mobile GUI Agent Research · 浏览器里运行的安卓模拟器 · Browser-hosted Android Simulator · Verifiable Evaluation · Scalable Online RL Training