aluncstokes/mathpile_arxiv_subset_tiny

---
configs:
- config_name: default
data_files:
- split: train
path: "train_chunked.jsonl"
- split: test
path: "test_chunked.jsonl"
---

# MathPile ArXiv (subset)

## Description
This dataset consists of a toy subset of 8834 (5000 training + 3834 testing) TeX files found in the arXiv subset of MathPile, used for testing. You should not use this dataset. Training and testing sets are already split

## Source
The data was obtained from the training + validation portion of the arXiv subset of MathPile.

## Format
- Given as JSONL files of JSON dicts each containing the single key: "text"

## Usage
- LaTeX stuff idk

## License
The original data is subject to the licensing terms of the arXiv. Users should refer to the arXiv's terms of use for details on permissible usage.

原始信息汇总

MathPile ArXiv (subset)

描述

该数据集包含从MathPile的arXiv子集中提取的8834个（5000个训练 + 3834个测试）TeX文件的玩具子集，用于测试。不应使用此数据集。训练和测试集已预先划分。

来源

数据来自MathPile的arXiv子集的训练和验证部分。

格式

• 以JSONL文件形式提供，每个文件包含一个键为"text"的JSON字典。

许可

原始数据受arXiv许可条款的约束。用户应参考arXiv的使用条款以了解允许的使用方式。