Saved in:
Bibliographic Details
Main Authors: Ying, Huaiyuan, Wu, Zijian, Geng, Yihan, Yuan, Zheng, Lin, Dahua, Chen, Kai
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2406.03847
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915349291597824
author Ying, Huaiyuan
Wu, Zijian
Geng, Yihan
Yuan, Zheng
Lin, Dahua
Chen, Kai
author_facet Ying, Huaiyuan
Wu, Zijian
Geng, Yihan
Yuan, Zheng
Lin, Dahua
Chen, Kai
contents Large language models have demonstrated impressive capabilities across various natural language processing tasks, especially in solving mathematical problems. However, large language models are not good at math theorem proving using formal languages like Lean. A significant challenge in this area is the scarcity of training data available in these formal languages. To address this issue, we propose a novel pipeline that iteratively generates and filters synthetic data to translate natural language mathematical problems into Lean 4 statements, and vice versa. Our results indicate that the synthetic data pipeline can provide useful training data and improve the performance of LLMs in translating and understanding complex mathematical problems and proofs. Our final dataset contains about 57K formal-informal question pairs along with searched proof from the math contest forum and 21 new IMO questions. We open-source our code at https://github.com/InternLM/InternLM-Math and our data at https://huggingface.co/datasets/InternLM/Lean-Workbook.
format Preprint
id arxiv_https___arxiv_org_abs_2406_03847
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Lean Workbook: A large-scale Lean problem set formalized from natural language math problems
Ying, Huaiyuan
Wu, Zijian
Geng, Yihan
Yuan, Zheng
Lin, Dahua
Chen, Kai
Computation and Language
Large language models have demonstrated impressive capabilities across various natural language processing tasks, especially in solving mathematical problems. However, large language models are not good at math theorem proving using formal languages like Lean. A significant challenge in this area is the scarcity of training data available in these formal languages. To address this issue, we propose a novel pipeline that iteratively generates and filters synthetic data to translate natural language mathematical problems into Lean 4 statements, and vice versa. Our results indicate that the synthetic data pipeline can provide useful training data and improve the performance of LLMs in translating and understanding complex mathematical problems and proofs. Our final dataset contains about 57K formal-informal question pairs along with searched proof from the math contest forum and 21 new IMO questions. We open-source our code at https://github.com/InternLM/InternLM-Math and our data at https://huggingface.co/datasets/InternLM/Lean-Workbook.
title Lean Workbook: A large-scale Lean problem set formalized from natural language math problems
topic Computation and Language
url https://arxiv.org/abs/2406.03847