langfengQ/verl-agent

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"

55
/ 100
Established

This tool helps AI researchers and machine learning engineers train large language models (LLMs) and vision-language models (VLMs) to act as intelligent agents in complex, multi-step environments. You input an LLM/VLM and a task environment (like ALFWorld or WebShop), and it outputs a fine-tuned agent capable of performing long-horizon tasks through reinforcement learning. It's designed for researchers developing advanced AI agents that interact dynamically and remember key information over many turns.

1,668 stars.

Use this if you are an AI researcher or machine learning engineer looking to train LLM or VLM agents to solve intricate, multi-step problems using reinforcement learning, especially when dealing with long interaction histories.

Not ideal if you are looking for a plug-and-play solution for simple text generation or a tool that doesn't require deep understanding of reinforcement learning concepts.

AI Agent Development Reinforcement Learning Large Language Models Multi-modal AI AI Research
No Package No Dependents
Maintenance 10 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 19 / 25

How are scores calculated?

Stars

1,668

Forks

148

Language

Python

License

Apache-2.0

Last pushed

Feb 27, 2026

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/langfengQ/verl-agent"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.