Microsoft Research Asia open-sources Agent Lightning v1.0: a 3,500-line RL training framework for real harness agents
Original title:Agent Lightning v1.0: A 3,500-Line Lightweight Agentic RL Framework for Training Agents with Real Harnesses
AI summary
Microsoft Research Asia proposed the Harnessed Agentic RL training paradigm and open-sourced the rebuilt Agent Lightning v1.0, allowing the same agent harness used in deployment to participate directly in reinforcement learning without rewriting the agent inside the training framework.
Quick Facts
- Published
- Updated
- Source
- Microsoft Research 博客(RSS)
- Category
- Papers & Research
- Topic
- Open-Source Ecosystem
- AI score
- 69
框架让部署中的真实 agent harness 直接参与 RL 训练,无需重写 agent 代码,还给出了端到端编码智能体训练效果作参考。
Related entities
Source & copyright
This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.
How to Cite
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: Microsoft Research Asia open-sources Agent Lightning v1.0: a 3,500-line RL training framework for real harness agents. October 8, 2026. https://new.nat6.net/en/news/t9wypjd9cbb42ttuq7vrpp07l/