LeanDojo: Machine Learning for Theorem Proving in Lean
LeanDojo is a Python library for learning–based theorem provers in Lean, providing two main features:
- Extracting data (proof states, tactics, premises, etc.) from Lean repos.
- Interacting with Lean programmatically.
LeanDojo's current version is compatible with Lean 4 v4.3.0-rc2 or later. We strongly suggest using the current version. However, you may use the legacy branch if you want to work with earlier versions (including Lean 3).
Requirements
- Supported platforms: Linux, Windows WSL, and macOS
- Git >= 2.25
- 3.9 <= Python < 3.13
- wget
- elan
- Generate a GitHub personal access token and set the environment variable
GITHUB_ACCESS_TOKENto it
Installation
LeanDojo is available on PyPI and can be installed via pip:
pip install lean-dojo
It can also be installed locally from the Git repo:
pip install .
Documentation
Questions and Bugs
- For general questions and discussions, please use GitHub Discussions.
- To report a potential bug, please open an issue. In the issue, please include your OS information, the exact steps to reproduce the error on the latest version of LeanDojo, and complete logs in debug mode (setting the environment variable
VERBOSEto 1). Important: If your issue cannot be reproduced easily, it will be unlikely to receive any help.
Related Links
- LeanDojo Website: The official website of LeanDojo.
- LeanDojo Benchmark
: The dataset used in our paper, consisting of theorems and proofs extracted from mathlib by generate-benchmark-lean3.ipynb.
- LeanDojo Benchmark 4
: The Lean 4 version of LeanDojo Benchmark, consisting of theorems and proofs extracted from mathlib4 by generate-benchmark-lean4.ipynb.
- ReProver: The ReProver (Retrieval-Augmented Prover) model in our paper.
- LeanDojo ChatGPT Plugin
- Lean Copilot: Running language models as copilots for theorem proving in Lean
Citation
LeanDojo: Theorem Proving with Retrieval-Augmented Language Models
Neural Information Processing Systems (NeurIPS), 2023
Kaiyu Yang, Aidan Swope, Alex Gu, Rahul Chalamala,
Peiyang Song, Shixing Yu, Saad Godil, Ryan Prenger, Anima Anandkumar
@inproceedings{yang2023leandojo,
title={{LeanDojo}: Theorem Proving with Retrieval-Augmented Language Models},
author={Yang, Kaiyu and Swope, Aidan and Gu, Alex and Chalamala, Rahul and Song, Peiyang and Yu, Shixing and Godil, Saad and Prenger, Ryan and Anandkumar, Anima},
booktitle={Neural Information Processing Systems (NeurIPS)},
year={2023}
}
1 maintainer
Metadata
1 maintainer
Release files for lean-dojo 4.20.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| lean_dojo-4.20.0.tar.gz | 67.2 MB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| lean_dojo-4.20.0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 67.3 MB
Release files / lean_dojo-4.20.0.tar.gz
| Download URL | lean_dojo-4.20.0.tar.gz |
|---|---|
| Size | 67.2 MB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
990bdcbb38c4aede61b9b776ef0fe84b422d2fd0a7c1088c3c2ed14e5c8939b4
|
|
BLAKE2b-256 checksum How to use checksums |
c4256af86f47bf571955a0f5cee7fd37e7b59804d83bf5cc000a390cac42e58f
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
python-httpx/0.27.0
|
Release files / lean_dojo-4.20.0-py3-none-any.whl
| Download URL | lean_dojo-4.20.0-py3-none-any.whl |
|---|---|
| Size | 53.2 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
247110b1c6ee1308f844f79c99140d8437fe02f538b308b911ea81b35eb3254b
|
|
BLAKE2b-256 checksum How to use checksums |
e305ce64788a8bfc599ca8545e0533876d04b46fde41d2d66c8b303b95c6be72
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
python-httpx/0.27.0
|
