Toolbox/Agent SDKs & frameworks/langfengQ/verl-agent

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"

View on GitHub →

PythonApache-2.0llm-agentsllm-trainingreinforcement-learninglarge-language-modelsdeepseek-r1

Stars

2.3k

Δ 7d

+19

Δ 30d

+87

Age

1 yr

Last push

87d ago

90-day star history

Where this sits in the catalog

GroupFrameworks & SDKs

CategoryAgent SDKs & frameworks#49 of 85 by stars

RankingsNot in any top 20 today — see the rankings

NearbyMulti-agent orchestration (129)Workflow & graph engines (13)Low-code builders (21)Runtimes & sandboxes (16)

Most-wanted open issues

Ranked by 👍 reactions. 59 open issues total, checked 2026-08-29.

Issue	👍	Comments	Opened
Training hangs after "colocated worker base class" initialization in VERL in alfworld	1	2	1 yr
README clarification: how is WebShop supposed to run with a separate `verl-agent-webshop` env if training imports WebShop directly?	1	1	5 mo
Overhead of faiss retrieval server causing OOM despite having enough VRAM	0	0	4 mo
The actor is dead because its worker process has died.	0	8	1 yr
Reproducing issues of alfworld	0	3	1 yr