agent-lightning
The absolute trainer to light up AI agents.
Stars: 1581
Agent Lightning is a lightweight and efficient tool for automating repetitive tasks in the field of data analysis and machine learning. It provides a user-friendly interface to create and manage automated workflows, allowing users to easily schedule and execute data processing, model training, and evaluation tasks. With its intuitive design and powerful features, Agent Lightning streamlines the process of building and deploying machine learning models, making it ideal for data scientists, machine learning engineers, and AI enthusiasts looking to boost their productivity and efficiency in their projects.
README:
The absolute trainer to light up AI agents.
Join our Discord community to connect with other users and contributors.
- Turn your agent into an optimizable beast with ZERO CODE CHANGE (almost)! 💤
- Build with ANY agent framework (LangChain, OpenAI Agent SDK, AutoGen, CrewAI, ...); or even WITHOUT agent framework (Python OpenAI). You name it! 🤖
- Selectively optimize one or more agents in a multi-agent system. 🎯
- Embraces Reinforcement Learning, Automatic Prompt Optimization and more algorithms. 🤗
- 8/11/2025 Training AI Agents to Write and Self-correct SQL with Reinforcement Learning Medium.
- 8/5/2025 Agent Lightning: Train ANY AI Agents with Reinforcement Learning arXiv paper.
- 7/26/2025 We discovered an approach to train any AI agent with RL, with (almost) zero code changes. Reddit.
- 6/6/2025 Agent Lightning - Microsoft Research Project page.
First, let's get your environment set up. We'll be using /path/to/agentlightning to refer to the directory containing this README file.
We strongly recommend creating a new virtual environment to avoid conflicts with other packages. You can use either conda or venv. Python 3.10 or later is recommended.
If you are running RL with Agent-Lightning, the next step is to install the essential packages: PyTorch, FlashAttention, vLLM and VERL. The following versions and installation order have been tested and are confirmed to work.
pip install torch==2.7.0 torchvision==0.22.0 torchaudio==2.7.0 --index-url https://download.pytorch.org/whl/cu128
pip install flash-attn --no-build-isolation
pip install vllm==0.9.2
pip install verl==0.5.0See scripts/setup_stable_gpu.sh for a full installation script.
Now, you're ready to install Agent Lightning itself.
pip install agentlightningIf you plan to use other agent frameworks, you can install them with the following commands. If you don't need these, feel free to skip this step. We recommend doing this as the final step to avoid dependency versions being overwritten by mistake.
# AutoGen (Recommended to install first)
pip install "autogen-agentchat" "autogen-ext[openai]"
# LiteLLM
pip install "litellm[proxy]"
# MCP
pip install mcp
# UV
pip install uv
# OpenAI Agents
pip install openai-agents
# LangChain
pip install langgraph "langchain[openai]" langchain-community langchain-text-splitters
# SQL-related dependencies
pip install sqlparse nltkDon't worry if dependency conflicts arise during this step. Follow the installation order above and the conflicts generally do not matter.
For more detailed examples, please see the examples folder:
- calc_x: An agent built with AutoGen with calculator tool use, trained on Calc-X dataset with Reinforcement Learning.
- spider: A write-check-rewrite looped agent with LangGraph with SQL execution; selectively optimize write and rewrite on Spider dataset with Reinforcement Learning.
- apo: An example to customize an optimization algorithm: Automatic Prompt Optimization.
-
AgentOps Integration: Agent Lightning uses AgentOps for agent tracking by default. If you're already using AgentOps in your own code, you'll need to disable our managed AgentOps client by modifying the
tracerparameter of trainer. -
Debugging Traces: If you encounter issues with tracing, you can visualize the trace tree using
tracer.last_trace().visualize("tree_graph"). Please note that this API is experimental and may change in future releases. - Launching the Server and Agents: Currently, the training server and agent clients must be launched in separate processes. You can open two terminal windows or run one of them in the background. The launching order generally doesn't matter.
-
Environment Variables: The environment variables and working directory at the time of
ray initare important. If you run into "file not found" errors, try restarting Ray from your current working directory. - Handling Timeouts: The training server may hang if samples fail or time out on the agent side. To prevent this, we recommend setting limits on the prompt and response lengths, as this is the most common cause of failures.
- VERL Failures: Save checkpoints frequently, as VERL with vLLM may sometimes experience out-of-memory issues. If you encounter a VERL failure, you can resume training from the last checkpoint.
Currently, Agent Lightning is built around a training server and one or multiple agents.
- The server manages the training data, prepares samples for the agents, and provides the LLM endpoint.
- Agents retrieve samples from the server, process them (which may involve interacting with the LLM), and send the results back. These results, or "trajectories," are lists of prompts and responses from the LLM.
- The server then collects these trajectories and computes the losses to optimize the language models.
Install with development dependencies:
git clone https://github.com/microsoft/agent-lightning
cd agent-lightning
pip install -e .[dev]
Please run pre-commit hooks before checking in code:
pre-commit install
pre-commit run --all-files --show-diff-on-failure --color=always
Serve documentation locally:
mkdocs serveIf you find Agent Lightning useful in your research or projects, please cite our paper:
@misc{luo2025agentlightningtrainai,
title={Agent Lightning: Train ANY AI Agents with Reinforcement Learning},
author={Xufang Luo and Yuge Zhang and Zhiyuan He and Zilong Wang and Siyun Zhao and Dongsheng Li and Luna K. Qiu and Yuqing Yang},
year={2025},
eprint={2508.03680},
archivePrefix={arXiv},
primaryClass={cs.AI},
url={https://arxiv.org/abs/2508.03680},
}This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.opensource.microsoft.com.
When you submit a pull request, a CLA bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., status check, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.
This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact [email protected] with any additional questions or comments.
This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow Microsoft's Trademark & Brand Guidelines. Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party's policies.
This project has been evaluated and certified to comply with the Microsoft Responsible AI Standard. The team will continue to monitor and maintain the repository, addressing any severe issues, including potential harms, if they arise.
This project is licensed under the MIT License. See the LICENSE file for details.
For Tasks:
Click tags to check more tools for each tasksFor Jobs:
Alternative AI tools for agent-lightning
Similar Open Source Tools
agent-lightning
Agent Lightning is a lightweight and efficient tool for automating repetitive tasks in the field of data analysis and machine learning. It provides a user-friendly interface to create and manage automated workflows, allowing users to easily schedule and execute data processing, model training, and evaluation tasks. With its intuitive design and powerful features, Agent Lightning streamlines the process of building and deploying machine learning models, making it ideal for data scientists, machine learning engineers, and AI enthusiasts looking to boost their productivity and efficiency in their projects.
pathway
Pathway is a Python data processing framework for analytics and AI pipelines over data streams. It's the ideal solution for real-time processing use cases like streaming ETL or RAG pipelines for unstructured data. Pathway comes with an **easy-to-use Python API** , allowing you to seamlessly integrate your favorite Python ML libraries. Pathway code is versatile and robust: **you can use it in both development and production environments, handling both batch and streaming data effectively**. The same code can be used for local development, CI/CD tests, running batch jobs, handling stream replays, and processing data streams. Pathway is powered by a **scalable Rust engine** based on Differential Dataflow and performs incremental computation. Your Pathway code, despite being written in Python, is run by the Rust engine, enabling multithreading, multiprocessing, and distributed computations. All the pipeline is kept in memory and can be easily deployed with **Docker and Kubernetes**. You can install Pathway with pip: `pip install -U pathway` For any questions, you will find the community and team behind the project on Discord.
promptflow
**Prompt flow** is a suite of development tools designed to streamline the end-to-end development cycle of LLM-based AI applications, from ideation, prototyping, testing, evaluation to production deployment and monitoring. It makes prompt engineering much easier and enables you to build LLM apps with production quality.
Sentient
Sentient is a personal, private, and interactive AI companion developed by Existence. The project aims to build a completely private AI companion that is deeply personalized and context-aware of the user. It utilizes automation and privacy to create a true companion for humans. The tool is designed to remember information about the user and use it to respond to queries and perform various actions. Sentient features a local and private environment, MBTI personality test, integrations with LinkedIn, Reddit, and more, self-managed graph memory, web search capabilities, multi-chat functionality, and auto-updates for the app. The project is built using technologies like ElectronJS, Next.js, TailwindCSS, FastAPI, Neo4j, and various APIs.
azure-search-openai-demo
This sample demonstrates a few approaches for creating ChatGPT-like experiences over your own data using the Retrieval Augmented Generation pattern. It uses Azure OpenAI Service to access a GPT model (gpt-35-turbo), and Azure AI Search for data indexing and retrieval. The repo includes sample data so it's ready to try end to end. In this sample application we use a fictitious company called Contoso Electronics, and the experience allows its employees to ask questions about the benefits, internal policies, as well as job descriptions and roles.
SalesGPT
SalesGPT is an open-source AI agent designed for sales, utilizing context-awareness and LLMs to work across various communication channels like voice, email, and texting. It aims to enhance sales conversations by understanding the stage of the conversation and providing tools like product knowledge base to reduce errors. The agent can autonomously generate payment links, handle objections, and close sales. It also offers features like automated email communication, meeting scheduling, and integration with various LLMs for customization. SalesGPT is optimized for low latency in voice channels and ensures human supervision where necessary. The tool provides enterprise-grade security and supports LangSmith tracing for monitoring and evaluation of intelligent agents built on LLM frameworks.
obsidian-chat-cbt-plugin
ChatCBT is an AI-powered journaling assistant for Obsidian, inspired by cognitive behavioral therapy (CBT). It helps users reframe negative thoughts and rewire reactions to distressful situations. The tool provides kind and objective responses to uncover negative thinking patterns, store conversations privately, and summarize reframed thoughts. Users can choose between a cloud-based AI service (OpenAI) or a local and private service (Ollama) for handling data. ChatCBT is not a replacement for therapy but serves as a journaling assistant to help users gain perspective on their problems.
burpference
Burpference is an open-source extension designed to capture in-scope HTTP requests and responses from Burp's proxy history and send them to a remote LLM API in JSON format. It automates response capture, integrates with APIs, optimizes resource usage, provides color-coded findings visualization, offers comprehensive logging, supports native Burp reporting, and allows flexible configuration. Users can customize system prompts, API keys, and remote hosts, and host models locally to prevent high inference costs. The tool is ideal for offensive web application engagements to surface findings and vulnerabilities.
gpt-pilot
GPT Pilot is a core technology for the Pythagora VS Code extension, aiming to provide the first real AI developer companion. It goes beyond autocomplete, helping with writing full features, debugging, issue discussions, and reviews. The tool utilizes LLMs to generate production-ready apps, with developers overseeing the implementation. GPT Pilot works step by step like a developer, debugging issues as they arise. It can work at any scale, filtering out code to show only relevant parts to the AI during tasks. Contributions are welcome, with debugging and telemetry being key areas of focus for improvement.
langdrive
LangDrive is an open-source AI library that simplifies training, deploying, and querying open-source large language models (LLMs) using private data. It supports data ingestion, fine-tuning, and deployment via a command-line interface, YAML file, or API, with a quick, easy setup. Users can build AI applications such as question/answering systems, chatbots, AI agents, and content generators. The library provides features like data connectors for ingestion, fine-tuning of LLMs, deployment to Hugging Face hub, inference querying, data utilities for CRUD operations, and APIs for model access. LangDrive is designed to streamline the process of working with LLMs and making AI development more accessible.
litlytics
LitLytics is an affordable analytics platform leveraging LLMs for automated data analysis. It simplifies analytics for teams without data scientists, generates custom pipelines, and allows customization. Cost-efficient with low data processing costs. Scalable and flexible, works with CSV, PDF, and plain text data formats.
Open_Data_QnA
Open Data QnA is a Python library that allows users to interact with their PostgreSQL or BigQuery databases in a conversational manner, without needing to write SQL queries. The library leverages Large Language Models (LLMs) to bridge the gap between human language and database queries, enabling users to ask questions in natural language and receive informative responses. It offers features such as conversational querying with multiturn support, table grouping, multi schema/dataset support, SQL generation, query refinement, natural language responses, visualizations, and extensibility. The library is built on a modular design and supports various components like Database Connectors, Vector Stores, and Agents for SQL generation, validation, debugging, descriptions, embeddings, responses, and visualizations.
openllmetry-js
OpenLLMetry-JS is a set of extensions built on top of OpenTelemetry that gives you complete observability over your LLM application. Because it uses OpenTelemetry under the hood, it can be connected to your existing observability solutions - Datadog, Honeycomb, and others. It's built and maintained by Traceloop under the Apache 2.0 license. The repo contains standard OpenTelemetry instrumentations for LLM providers and Vector DBs, as well as a Traceloop SDK that makes it easy to get started with OpenLLMetry-JS, while still outputting standard OpenTelemetry data that can be connected to your observability stack. If you already have OpenTelemetry instrumented, you can just add any of our instrumentations directly.
coral-cloud
Coral Cloud Resorts is a sample hospitality application that showcases Data Cloud, Agents, and Prompts. It provides highly personalized guest experiences through smart automation, content generation, and summarization. The app requires licenses for Data Cloud, Agents, Prompt Builder, and Einstein for Sales. Users can activate features, deploy metadata, assign permission sets, import sample data, and troubleshoot common issues. Additionally, the repository offers integration with modern web development tools like Prettier, ESLint, and pre-commit hooks for code formatting and linting.
ComfyUI-Tara-LLM-Integration
Tara is a powerful node for ComfyUI that integrates Large Language Models (LLMs) to enhance and automate workflow processes. With Tara, you can create complex, intelligent workflows that refine and generate content, manage API keys, and seamlessly integrate various LLMs into your projects. It comprises nodes for handling OpenAI-compatible APIs, saving and loading API keys, composing multiple texts, and using predefined templates for OpenAI and Groq. Tara supports OpenAI and Grok models with plans to expand support to together.ai and Replicate. Users can install Tara via Git URL or ComfyUI Manager and utilize it for tasks like input guidance, saving and loading API keys, and generating text suitable for chaining in workflows.
OSWorld
OSWorld is a benchmarking tool designed to evaluate multimodal agents for open-ended tasks in real computer environments. It provides a platform for running experiments, setting up virtual machines, and interacting with the environment using Python scripts. Users can install the tool on their desktop or server, manage dependencies with Conda, and run benchmark tasks. The tool supports actions like executing commands, checking for specific results, and evaluating agent performance. OSWorld aims to facilitate research in AI by providing a standardized environment for testing and comparing different agent baselines.
For similar tasks
pipeshub-ai
Pipeshub-ai is a versatile tool for automating data pipelines in AI projects. It provides a user-friendly interface to design, deploy, and monitor complex data workflows, enabling seamless integration of various AI models and data sources. With Pipeshub-ai, users can easily create end-to-end pipelines for tasks such as data preprocessing, model training, and inference, streamlining the AI development process and improving productivity. The tool supports integration with popular AI frameworks and cloud services, making it suitable for both beginners and experienced AI practitioners.
agent-lightning
Agent Lightning is a lightweight and efficient tool for automating repetitive tasks in the field of data analysis and machine learning. It provides a user-friendly interface to create and manage automated workflows, allowing users to easily schedule and execute data processing, model training, and evaluation tasks. With its intuitive design and powerful features, Agent Lightning streamlines the process of building and deploying machine learning models, making it ideal for data scientists, machine learning engineers, and AI enthusiasts looking to boost their productivity and efficiency in their projects.
katib
Katib is a Kubernetes-native project for automated machine learning (AutoML). Katib supports Hyperparameter Tuning, Early Stopping and Neural Architecture Search. Katib is the project which is agnostic to machine learning (ML) frameworks. It can tune hyperparameters of applications written in any language of the users’ choice and natively supports many ML frameworks, such as TensorFlow, Apache MXNet, PyTorch, XGBoost, and others. Katib can perform training jobs using any Kubernetes Custom Resources with out of the box support for Kubeflow Training Operator, Argo Workflows, Tekton Pipelines and many more.
cog
Cog is an open-source tool that lets you package machine learning models in a standard, production-ready container. You can deploy your packaged model to your own infrastructure, or to Replicate.
mosec
Mosec is a high-performance and flexible model serving framework for building ML model-enabled backend and microservices. It bridges the gap between any machine learning models you just trained and the efficient online service API. * **Highly performant** : web layer and task coordination built with Rust 🦀, which offers blazing speed in addition to efficient CPU utilization powered by async I/O * **Ease of use** : user interface purely in Python 🐍, by which users can serve their models in an ML framework-agnostic manner using the same code as they do for offline testing * **Dynamic batching** : aggregate requests from different users for batched inference and distribute results back * **Pipelined stages** : spawn multiple processes for pipelined stages to handle CPU/GPU/IO mixed workloads * **Cloud friendly** : designed to run in the cloud, with the model warmup, graceful shutdown, and Prometheus monitoring metrics, easily managed by Kubernetes or any container orchestration systems * **Do one thing well** : focus on the online serving part, users can pay attention to the model optimization and business logic
web-llm
WebLLM is a modular and customizable javascript package that directly brings language model chats directly onto web browsers with hardware acceleration. Everything runs inside the browser with no server support and is accelerated with WebGPU. WebLLM is fully compatible with OpenAI API. That is, you can use the same OpenAI API on any open source models locally, with functionalities including json-mode, function-calling, streaming, etc. We can bring a lot of fun opportunities to build AI assistants for everyone and enable privacy while enjoying GPU acceleration.
data-scientist-roadmap2024
The Data Scientist Roadmap2024 provides a comprehensive guide to mastering essential tools for data science success. It includes programming languages, machine learning libraries, cloud platforms, and concepts categorized by difficulty. The roadmap covers a wide range of topics from programming languages to machine learning techniques, data visualization tools, and DevOps/MLOps tools. It also includes web development frameworks and specific concepts like supervised and unsupervised learning, NLP, deep learning, reinforcement learning, and statistics. Additionally, it delves into DevOps tools like Airflow and MLFlow, data visualization tools like Tableau and Matplotlib, and other topics such as ETL processes, optimization algorithms, and financial modeling.
aibydoing-feedback
AI By Doing is a hands-on artificial intelligence tutorial series that aims to help beginners understand the principles of machine learning and deep learning while providing practical applications. The content covers various supervised and unsupervised learning algorithms, machine learning engineering, deep learning fundamentals, frameworks like TensorFlow and PyTorch, and applications in computer vision and natural language processing. The tutorials are written in Jupyter Notebook format, combining theory, mathematical derivations, and Python code implementations to facilitate learning and understanding.
For similar jobs
sweep
Sweep is an AI junior developer that turns bugs and feature requests into code changes. It automatically handles developer experience improvements like adding type hints and improving test coverage.
teams-ai
The Teams AI Library is a software development kit (SDK) that helps developers create bots that can interact with Teams and Microsoft 365 applications. It is built on top of the Bot Framework SDK and simplifies the process of developing bots that interact with Teams' artificial intelligence capabilities. The SDK is available for JavaScript/TypeScript, .NET, and Python.
ai-guide
This guide is dedicated to Large Language Models (LLMs) that you can run on your home computer. It assumes your PC is a lower-end, non-gaming setup.
classifai
Supercharge WordPress Content Workflows and Engagement with Artificial Intelligence. Tap into leading cloud-based services like OpenAI, Microsoft Azure AI, Google Gemini and IBM Watson to augment your WordPress-powered websites. Publish content faster while improving SEO performance and increasing audience engagement. ClassifAI integrates Artificial Intelligence and Machine Learning technologies to lighten your workload and eliminate tedious tasks, giving you more time to create original content that matters.
chatbot-ui
Chatbot UI is an open-source AI chat app that allows users to create and deploy their own AI chatbots. It is easy to use and can be customized to fit any need. Chatbot UI is perfect for businesses, developers, and anyone who wants to create a chatbot.
BricksLLM
BricksLLM is a cloud native AI gateway written in Go. Currently, it provides native support for OpenAI, Anthropic, Azure OpenAI and vLLM. BricksLLM aims to provide enterprise level infrastructure that can power any LLM production use cases. Here are some use cases for BricksLLM: * Set LLM usage limits for users on different pricing tiers * Track LLM usage on a per user and per organization basis * Block or redact requests containing PIIs * Improve LLM reliability with failovers, retries and caching * Distribute API keys with rate limits and cost limits for internal development/production use cases * Distribute API keys with rate limits and cost limits for students
uAgents
uAgents is a Python library developed by Fetch.ai that allows for the creation of autonomous AI agents. These agents can perform various tasks on a schedule or take action on various events. uAgents are easy to create and manage, and they are connected to a fast-growing network of other uAgents. They are also secure, with cryptographically secured messages and wallets.
griptape
Griptape is a modular Python framework for building AI-powered applications that securely connect to your enterprise data and APIs. It offers developers the ability to maintain control and flexibility at every step. Griptape's core components include Structures (Agents, Pipelines, and Workflows), Tasks, Tools, Memory (Conversation Memory, Task Memory, and Meta Memory), Drivers (Prompt and Embedding Drivers, Vector Store Drivers, Image Generation Drivers, Image Query Drivers, SQL Drivers, Web Scraper Drivers, and Conversation Memory Drivers), Engines (Query Engines, Extraction Engines, Summary Engines, Image Generation Engines, and Image Query Engines), and additional components (Rulesets, Loaders, Artifacts, Chunkers, and Tokenizers). Griptape enables developers to create AI-powered applications with ease and efficiency.