Dataset · Hugging Face

TIGER-Lab/MathInstruct

Name: MathInstruct
Creator: TIGER-Lab
License: mit
Keywords: task_categories:text-generation, language:en, license:mit, size_categories:100K<n<1M, format:json, modality:text, library:datasets, library:pandas, library:mlcroissant, library:polars, arxiv:2309.05653, region:us, math

🦣 MAmmoTH: Building Math Generalist Models through Hybrid Instruction Tuning MathInstruct is a meticulously curated instruction tuning dataset that is lightweight yet generalizable. MathInstruct is compiled from 13 math rationale datasets, six of which are newly curated by this work. It uniquely focuses on the hybrid use of chain-of-thought (CoT) and program-of-thought (PoT) rationales, and ensures extensive coverage of diverse mathematical fields. Project Page:… See the full description on the dataset page: https://huggingface.co/datasets/TIGER-Lab/MathInstruct.

View on Hugging FaceReport an issue

Author / organization	TIGER-Lab
License (as declared)	mit
Downloads	12.1K (as reported by Hugging Face)
Likes	316
Created	11 September 2023
Last updated at source	15 May 2024

math

Embed this entry

<iframe src="https://silytics.si/embed/item/46521442-effd-45ce-b8ee-d126907dff0e" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Connected in the knowledge graph

openchat/openchat-3.5-1210Model · used to train
HuggingFaceM4/idefics2-8bModel · used to train

Explore this entry in the graph

More from TIGER-Lab

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 14 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.