AutoMathText / README.md
yifAI's picture
Add data
3fdb4ba
|
raw
history blame
No virus
2.92 kB
metadata
license: cc-by-sa-4.0
task_categories:
  - text-generation
  - question-answering
language:
  - en
pretty_name: MathText
size_categories:
  - 10B<n<100B
configs:
  - config_name: web-0.5
    data_files:
      - split: train
        path:
          - data/web/0.95-1.00.jsonl
          - data/web/0.90-0.95.jsonl
          - data/web/0.85-0.90.jsonl
          - data/web/0.80-0.85.jsonl
          - data/web/0.75-0.80.jsonl
          - data/web/0.70-0.75.jsonl
          - data/web/0.65-0.70.jsonl
          - data/web/0.60-0.65.jsonl
          - data/web/0.55-0.60.jsonl
          - data/web/0.50-0.55.jsonl
    default: true
  - config_name: web-0.6
    data_files:
      - split: train
        path:
          - data/web/0.95-1.00.jsonl
          - data/web/0.90-0.95.jsonl
          - data/web/0.85-0.90.jsonl
          - data/web/0.80-0.85.jsonl
          - data/web/0.75-0.80.jsonl
          - data/web/0.70-0.75.jsonl
          - data/web/0.65-0.70.jsonl
          - data/web/0.60-0.65.jsonl
  - config_name: web-0.7
    data_files:
      - split: train
        path:
          - data/web/0.95-1.00.jsonl
          - data/web/0.90-0.95.jsonl
          - data/web/0.85-0.90.jsonl
          - data/web/0.80-0.85.jsonl
          - data/web/0.75-0.80.jsonl
          - data/web/0.70-0.75.jsonl
  - config_name: web-full
    data_files: data/web/*.jsonl
tags:
  - mathematical-reasoning
  - reasoning
  - finetuning
  - pretraining
  - llm

MathText

MathText is an extensive and carefully curated dataset encompassing 200 GB of mathematical texts. It's a unique compilation sourced from a diverse range of platforms including various websites, arXiv, and GitHub (OpenWebMath, RedPajama, Algebraic Stack). This rich repository has been autonomously labeled by the state-of-the-art open-sourced language model, Qwen-72B. Each piece of content in the dataset is assigned a score lm_q1q2_score within the range of [0, 1], reflecting its relevance, quality and educational value in the context of mathematical intelligence.

Objective

The primary aim of the MathText dataset is to provide a comprehensive and reliable resource for a wide array of users - from academic researchers and educators to AI practitioners and mathematics enthusiasts. This dataset is particularly geared towards:

  • Facilitating advanced research in the intersection of mathematics and artificial intelligence.
  • Serving as an educational tool for learning and teaching complex mathematical concepts.
  • Providing a robust foundation for developing and training AI models specialized in processing and understanding mathematical content.

Features

  • Volume: Approximately 200 GB of text data.
  • Content: A diverse collection of mathematical texts, including but not limited to research papers, educational articles, and code documentation.
  • Labeling: Every text is scored by Qwen-72B, a sophisticated language model, ensuring a high standard of relevance and accuracy.
  • Scope: Covers a wide spectrum of mathematical topics, making it suitable for various applications in research and education.