# SWE-Gym

> An open environment of real-world Python tasks for training and verifying software engineering agents.

SWE-Gym is a research environment for training language-model-based software engineering agents, released with the ICML 2025 paper "Training Software Engineering Agents and Verifiers with SWE-Gym" by authors from UC Berkeley, UIUC, CMU and Apple. The repository holds the code for the paper, while the dataset and models are published on Hugging Face.

## What It Is

SWE-Gym is a training environment for agents that solve repository-level software engineering problems. According to the paper, it contains 2,438 real-world Python task instances, and the README describes them as drawn from 11 Python repositories. Each instance combines a codebase with an executable runtime environment, unit tests, and a natural-language task description. A Lite split of 234 instances is also provided.

## How It Is Used

Researchers sample agent trajectories in the environment and fine-tune language models on them. The README documents experiments with the OpenHands and MoatlessTools agent scaffolds, including rejection-sampling fine-tuning for self-improvement. Verifiers can also be trained on agent trajectories. These verifiers pick the most promising solution through best-of-n selection at inference time.

## Reported Results

The authors report up to 19% absolute gains in resolve rate on SWE-Bench Verified and Lite. Combining fine-tuned agents with verifiers, they report 32.0% on SWE-Bench Verified and 26.0% on Lite, which they describe as a new state of the art for open-weight SWE agents. The README lists downstream work that builds on SWE-Gym, such as SkyRL and OpenHands LM.

## Reproducing the Work

The dataset and models are on the Hugging Face hub. Environment constants live in the SWE-Bench-Fork repository, and pre-built Docker images for each instance are on Docker Hub. The repository's docs/OpenHands.md and docs/MoatlessTools.md describe how to reproduce the training and inference-time experiments.

## Features
- 2,438 real-world Python task instances with executable environments and unit tests
- Lite split of 234 instances
- Trajectory-based fine-tuning of LM agents
- Verifiers trained on agent trajectories for best-of-n inference-time scaling
- Support for OpenHands and MoatlessTools agent scaffolds
- Pre-built Docker images per instance
- Public release of dataset, models and agent trajectories

## Integrations
OpenHands, MoatlessTools, Hugging Face, Docker Hub, SWE-Bench

## Platforms
API, CLI

## Pricing
Open Source

## Links
- Website: https://arxiv.org/abs/2412.21139
- Documentation: https://github.com/SWE-Gym/SWE-Gym/blob/main/docs/OpenHands.md
- Repository: https://github.com/SWE-Gym/SWE-Gym
- EveryDev.ai: https://www.everydev.ai/tools/swe-gym