grpo

Use PEFT or Full-parameter to CPT/SFT/DPO/GRPO 500+ LLMs (Qwen3, Qwen3-MoE, Llama4, GLM4.5, InternLM3, DeepSeek-R1, ...) and 200+ MLLMs (Qwen3-VL, Qwen3-Omni, InternVL3.5, Ovis2.5, Llava, GLM4v, Phi4, ...) (AAAI 2025).

Agent Reinforcement Trainer: train multi-step agents for real-world tasks using GRPO. Give your agents on-the-job training. Reinforcement learning for Qwen2.5, Qwen3, Llama, and more!

Solve Visual Understanding with Reinforced VLMs

Skywork-R1V is an advanced multimodal AI model series developed by Skywork AI (Kunlun Inc.), specializing in vision-language reasoning.

The open source post-building layer for agents. Our environment data and evals power agent post-training (RL, SFT) and monitoring.

verl-agent is an extension of veRL, designed for training LLM/VLM agents via RL. verl-agent is also the official code for paper "Group-in-Group Policy Optimization for LLM Agent Training"

MixGRPO: Unlocking Flow-based GRPO Efficiency with Mixed ODE-SDE

Explore the Multimodal “Aha Moment” on 2B Model

🌾 OAT: A research-friendly framework for LLM online alignment, including reinforcement learning, preference learning, etc.

Collect every awesome work about r1!

OpenThinkIMG is an end-to-end open-source framework that empowers LVLMs to think with images.

Agentic RAG R1 Framework via Reinforcement Learning

BioReason: Incentivizing Multimodal Biological Reasoning within a DNA-LLM Model

Unleashing the Power of VLMs in Autonomous Driving via Reinforcement Learning and Reasoning

The open source implementation of DeepSeek-R1. 开源复现 DeepSeek-R1

[NeurIPS 2025] AutoVLA: A Vision-Language-Action Model for End-to-End Autonomous Driving with Adaptive Reasoning and Reinforcement Fine-Tuning

Train a Language Model with GRPO to create a schedule from a list of events and priorities

[NeurIPS 2025] Thinkless: LLM Learns When to Think

A Gaussian dense reward framework for GUI grounding training

Official implementation of "SEAgent: Self-Evolving Computer Use Agent with Autonomous Learning from Experience"