#1111th of 19 in Training and fine-tuning
agent-lightning
Reinforcement learning framework for training agents through a model-request proxy
Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 18.6k
- License
- MIT
- Last commit
- Oct 2026
- Last release
- Sep 2026
- Language
- Python
Missed the latest GitHub check; these facts may be older.
Overview
Agent Lightning is a roughly 3,500-line Python framework that trains agents with RL while they run in their real harnesses. A gateway proxies the agent's model requests and records them as training data, a rollout controller runs agents locally or as Kubernetes Jobs, and a trainer built on verl and vLLM updates the policy. Examples include a Qwen3.5-9B coding agent, which the README reports going from 41.8% to 56.4% on SWE-bench Verified.
Who it is for: ML engineers doing RL fine-tuning of LLM agents on their own GPUs
Strengths
- Agents run unchanged; the gateway captures model calls as training data
- Runs rollouts as Kubernetes Jobs without an external sandbox service
- Full coding-agent pipeline released: data cleaning, reward-hacking prevention, training scripts
- Examples cover Calc-X, GSM8K, ScienceWorld, Search-R1, LLM-in-Sandbox and coding agents
Weaknesses
- Requires a verl and vLLM GPU stack; install shown for CUDA 13.0 only
- No Docker or compose setup shipped in the repo
- v1.0 is a full refactor; older releases live on a separate v0.x branch
- Hardware and RAM requirements are not stated in the README
What it needs
- GPU required
- Needs verl, vLLM, uv, Kubernetes (optional, for rollouts)
- Models: Qwen3.5-9B, Qwen3.5-35B-A3B
Also in Training and fine-tuning
See all 19| Rank | Project | Score |
|---|---|---|
| 1 | unslothDesktop and web app to run, fine-tune and export open models | 88 out of 100 |
| 2 | deepspeedPyTorch library for distributed training of large models with ZeRO | 82 out of 100 |
| 3 | verlReinforcement learning training library for LLM post-training with PPO and GRPO | 77 out of 100 |
| #4 | llamafactoryFine-tuning toolkit for 100+ LLMs with a CLI and Gradio web UI | 75 out of 100 |
| #5 | peftPython library for LoRA and other parameter-efficient fine-tuning of Transformers models | 74 out of 100 |