textLLM360/MegaMathmathpretrainingllmcodereasoningcommon-crawlsynthetic-dataparquet

MegaMath Open Math Pretraining Corpus

Free

Open dataset

Sample structure: 100 / 100

3 download links issued

Seller: DataBazaar

Already have an account? Log in

Agent? Connect your account →

Source, license and coverage

Supplier documentation. These claims are separate from the automated sample score. A listing edit date is not a data freshness date.

License: odc-by
Source / creator: huggingface: LLM360/MegaMath
Collection method: auto_imported_huggingface_federated
Coverage start: Not documented
Coverage end: Not documented
Data last updated: Not documented
Update schedule: Not documented

Source documentation ↗

License terms ↗

Sample structure score: 100 / 100

This automated check describes the inspected sample, not factual accuracy, legal rights, representativeness, or the quality of the entire dataset. It is not a customer rating.

Assessed 10 sample records (JSON) on 2026-10-09. All records in the provided sample were checked.

Check	Points	Evidence
Populated cells	50 / 50	90 of 90 top-level cells contain a value. Null, absent and blank values count as missing; zero and false count as populated.
Consistent value types	30 / 30	90 of 90 populated cells match their column's most common observed type. Types are inferred, not checked against real-world truth.
Consistent record shape	20 / 20	10 of 10 records have the expected fields. CSV/TSV use the header width; JSON uses the union of observed keys.

Field-level findings and improvements

Check missing cells and mixed types below. Document intentional missing values or mixed types in your field descriptions. Do not fill legitimate unknowns with invented values just to increase this score.

Field	Missing cells	Most common type	Other populated types
text	0 / 10	string	0 / 10
id	0 / 10	string	0 / 10
lang	0 / 10	string	0 / 10
lang_score	0 / 10	number	0 / 10
url	0 / 10	string	0 / 10
timestamp	0 / 10	string	0 / 10
cc-path	0 / 10	string	0 / 10
domain	0 / 10	string	0 / 10
math_score	0 / 10	number	0 / 10

How the score is calculated, its limitations, and how to correct an assessment →

About this data

Math-focused pretraining corpus curated from Common Crawl, code, and synthetic sources, designed for training math-capable large language models.

Retrieve with your agent or Python

Create an account and configure DATABAZAAR_API_KEY. This example retrieves free or already purchased data; it never makes a purchase. For a multi-file dataset, choose a file index from the manifest.

Download the Python example

python3 retrieve-dataset.py 54456add-431c-4c9e-a14d-2519191ad435 --output dataset.bin

Full supplier documentation

MegaMath is an open mathematical pretraining corpus exceeding 300 billion tokens, curated by the LLM360 team as part of TxT360. It combines three carefully engineered streams: (1) re-extracted mathematical web documents from Common Crawl with math-oriented HTML processing, fasttext filtering, and deduplication; (2) math-related code data; and (3) synthetic math content. Distributed in Parquet format and compatible with datasets, dask, polars, and mlcroissant libraries. Ideal for pretraining or continued pretraining of math-reasoning LLMs, building math-focused RAG corpora, fine-tuning code+math models, and benchmarking data curation pipelines. Released under ODC-By license. Reference: arXiv:2504.02807. _Federated dataset: 5,314 parquet shards, 1041.37 GB total. Queries and downloads stream through the DataBazaar API._ Original supplier listing: MegaMath: 300B+ Token Open Math Pretraining Corpus Largest open math-focused pretraining dataset (300B+ tokens) from LLM360, curated from Common Crawl, code, and synthetic sources for training math-capable LLMs.

Schema

Name	Type	Description
text	VARCHAR	Full document text content including mathematical formulas, code, or prose.
id	VARCHAR	Unique identifier for the document.
lang	VARCHAR	ISO 639-1 language code (e.g., en, zh, fr).
lang_score	DOUBLE	Confidence score (0-1) for language detection.
url	VARCHAR	Original source URL from Common Crawl or synthetic origin.
timestamp	VARCHAR	ISO 8601 timestamp of document crawl or creation date.
cc-path	VARCHAR	Common Crawl path identifier (format: year/month/partition).
domain	VARCHAR	Domain name extracted from URL.
math_score	DOUBLE	Confidence score (0-1) indicating mathematical content density.

Sample Data

Preview a sample of the data before downloading.

Public sample only. Sign in to retrieve the full dataset, including free datasets.

For AI Agents

Via MCP Server

# 1. Add to your agent's MCP config (claude_desktop_config.json or similar):
{
  "mcpServers": {
    "databazaar": { "command": "npx", "args": ["databazaar-mcp"] }
  }
}

# 2. Your agent can then call:
search_datasets({ query: "MegaMath Open Math Pretraining" })
// Found: 54456add-431c-4c9e-a14d-2519191ad435
get_download_url({ dataset_id: "54456add-431c-4c9e-a14d-2519191ad435" })  // free — sign in with MCP OAuth first

Via REST API

# Free dataset — sign in or use your account API key:
curl https://api.databazaar.io/datasets/54456add-431c-4c9e-a14d-2519191ad435/download-url -H "Authorization: Bearer $DATABAZAAR_API_KEY"