Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/princeton-nlp/SWE-bench

[ICLR 2024] SWE-bench: Can Language Models Resolve Real-world Github Issues?
https://github.com/princeton-nlp/SWE-bench

benchmark language-model software-engineering

Last synced: about 1 month ago
JSON representation

[ICLR 2024] SWE-bench: Can Language Models Resolve Real-world Github Issues?

Awesome Lists containing this project

README 

        


  

    Kawi the SWE-Llama

  







 | [日本語](docs/README_JP.md) | [English](https://github.com/princeton-nlp/SWE-bench) | [中文简体](docs/README_CN.md) | [中文繁體](docs/README_TW.md) |






---



Code and data for our ICLR 2024 paper SWE-bench: Can Language Models Resolve Real-World GitHub Issues?

    

    

    

        Build

    

    

        License

    

    

        

    




Please refer our [website](http://swe-bench.github.io) for the public leaderboard and the [change log](https://github.com/princeton-nlp/SWE-bench/blob/main/CHANGELOG.md) for information on the latest updates to the SWE-bench benchmark.



## 📰 News

* **[Aug. 13, 2024]**: Introducing *SWE-bench Verified*! Part 2 of our collaboration with [OpenAI Preparedness](https://openai.com/preparedness/). A subset of 500 problems that real software engineers have confirmed are solvable. Check out more in the [report](https://openai.com/index/introducing-swe-bench-verified/)!

* **[Jun. 27, 2024]**: We have an exciting update for SWE-bench - with support from [OpenAI's Preparedness](https://openai.com/preparedness/) team: We're moving to a fully containerized evaluation harness using Docker for more reproducible evaluations! Read more in our [report](https://github.com/princeton-nlp/SWE-bench/blob/main/docs/20240627_docker/README.md).

* **[Apr. 15, 2024]**: SWE-bench has gone through major improvements to resolve issues with the evaluation harness. Read more in our [report](https://github.com/princeton-nlp/SWE-bench/blob/main/docs/20240415_eval_bug/README.md).

* **[Apr. 2, 2024]**: We have released [SWE-agent](https://github.com/princeton-nlp/SWE-agent), which sets the state-of-the-art on the full SWE-bench test set! ([Tweet 🔗](https://twitter.com/jyangballin/status/1775114444370051582))

* **[Jan. 16, 2024]**: SWE-bench has been accepted to ICLR 2024 as an oral presentation! ([OpenReview 🔗](https://openreview.net/forum?id=VTF8yNQM66))



## 👋 Overview

SWE-bench is a benchmark for evaluating large language models on real world software issues collected from GitHub.

Given a *codebase* and an *issue*, a language model is tasked with generating a *patch* that resolves the described problem.






To access SWE-bench, copy and run the following code:

```python

from datasets import load_dataset

swebench = load_dataset('princeton-nlp/SWE-bench', split='test')

```



## 🚀 Set Up

SWE-bench uses Docker for reproducible evaluations.

Follow the instructions in the [Docker setup guide](https://docs.docker.com/engine/install/) to install Docker on your machine.

If you're setting up on Linux, we recommend seeing the [post-installation steps](https://docs.docker.com/engine/install/linux-postinstall/) as well.



Finally, to build SWE-bench from source, follow these steps:

```bash

git clone [email protected]:princeton-nlp/SWE-bench.git

cd SWE-bench

pip install -e .

```



Test your installation by running:

```bash

python -m swebench.harness.run_evaluation \

    --predictions_path gold \

    --max_workers 1 \

    --instance_ids sympy__sympy-20590 \

    --run_id validate-gold

```



## 💽 Usage

> [!WARNING]

> Running fast evaluations on SWE-bench can be resource intensive

> We recommend running the evaluation harness on an `x86_64` machine with at least 120GB of free storage, 16GB of RAM, and 8 CPU cores.

> You may need to experiment with the `--max_workers` argument to find the optimal number of workers for your machine, but we recommend using fewer than `min(0.75 * os.cpu_count(), 24)`.

>

> If running with docker desktop, make sure to increase your virtual disk space to have ~120 free GB available, and set max_workers to be consistent with the above for the CPUs available to docker.

>

> Support for `arm64` machines is experimental.



Evaluate model predictions on SWE-bench Lite using the evaluation harness with the following command:

```bash

python -m swebench.harness.run_evaluation \

    --dataset_name princeton-nlp/SWE-bench_Lite \

    --predictions_path  \

    --max_workers  \

    --run_id 

    # use --predictions_path 'gold' to verify the gold patches

    # use --run_id to name the evaluation run

```



This command will generate docker build logs (`logs/build_images`) and evaluation logs (`logs/run_evaluation`) in the current directory.



The final evaluation results will be stored in the `evaluation_results` directory.



To see the full list of arguments for the evaluation harness, run:

```bash

python -m swebench.harness.run_evaluation --help

```



Additionally, the SWE-Bench repo can help you:

* Train your own models on our pre-processed datasets

* Run [inference](https://github.com/princeton-nlp/SWE-bench/blob/main/swebench/inference/README.md) on existing models (either models you have on-disk like LLaMA, or models you have access to through an API like GPT-4). The inference step is where you get a repo and an issue and have the model try to generate a fix for it.

*  Run SWE-bench's [data collection procedure](https://github.com/princeton-nlp/SWE-bench/blob/main/swebench/collect/) on your own repositories, to make new SWE-Bench tasks.



## ⬇️ Downloads

| Datasets | Models |

| - | - |

| [🤗 SWE-bench](https://huggingface.co/datasets/princeton-nlp/SWE-bench) | [🦙 SWE-Llama 13b](https://huggingface.co/princeton-nlp/SWE-Llama-13b) |

| [🤗 "Oracle" Retrieval](https://huggingface.co/datasets/princeton-nlp/SWE-bench_oracle) | [🦙 SWE-Llama 13b (PEFT)](https://huggingface.co/princeton-nlp/SWE-Llama-13b-peft) |

| [🤗 BM25 Retrieval 13K](https://huggingface.co/datasets/princeton-nlp/SWE-bench_bm25_13K) | [🦙 SWE-Llama 7b](https://huggingface.co/princeton-nlp/SWE-Llama-7b) |

| [🤗 BM25 Retrieval 27K](https://huggingface.co/datasets/princeton-nlp/SWE-bench_bm25_27K) | [🦙 SWE-Llama 7b (PEFT)](https://huggingface.co/princeton-nlp/SWE-Llama-7b-peft) |

| [🤗 BM25 Retrieval 40K](https://huggingface.co/datasets/princeton-nlp/SWE-bench_bm25_40K) | |

| [🤗 BM25 Retrieval 50K (Llama tokens)](https://huggingface.co/datasets/princeton-nlp/SWE-bench_bm25_50k_llama)   | |



## 🍎 Tutorials

We've also written the following blog posts on how to use different parts of SWE-bench.

If you'd like to see a post about a particular topic, please let us know via an issue.

* [Nov 1. 2023] Collecting Evaluation Tasks for SWE-Bench ([🔗](https://github.com/princeton-nlp/SWE-bench/blob/main/assets/collection.md))

* [Nov 6. 2023] Evaluating on SWE-bench ([🔗](https://github.com/princeton-nlp/SWE-bench/blob/main/assets/evaluation.md))



## 💫 Contributions

We would love to hear from the broader NLP, Machine Learning, and Software Engineering research communities, and we welcome any contributions, pull requests, or issues!

To do so, please either file a new pull request or issue and fill in the corresponding templates accordingly. We'll be sure to follow up shortly!



Contact person: [Carlos E. Jimenez](http://www.carlosejimenez.com/) and [John Yang](https://john-b-yang.github.io/) (Email: [email protected], [email protected]).



## ✍️ Citation

If you find our work helpful, please use the following citations.

```

@inproceedings{

    jimenez2024swebench,

    title={{SWE}-bench: Can Language Models Resolve Real-world Github Issues?},

    author={Carlos E Jimenez and John Yang and Alexander Wettig and Shunyu Yao and Kexin Pei and Ofir Press and Karthik R Narasimhan},

    booktitle={The Twelfth International Conference on Learning Representations},

    year={2024},

    url={https://openreview.net/forum?id=VTF8yNQM66}

}

```



## 🪪 License

MIT. Check `LICENSE.md`.