datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
FineFineWeb
FineFineWeb: A Comprehensive Study on Fine-Grained Domain Web Corpus
arXiv: Coming Soon
Project Page: Coming Soon
Blog: Coming Soon
Data Statistics
Domain (#tokens/#samples)
Iteration 1 Tokens
Iteration 2 Tokens
Iteration 3 Tokens
Total Tokens
Iteration 1 Count
Iteration 2 Count
Iteration 3 Count
Total Count
aerospace
5.77B
261.63M
309.33M
6.34B
9100000
688505
611034
10399539
agronomy
13.08B
947.41M
229.04M
14.26B
15752828
2711790
649404
19114022… See the full description on the dataset page: https://huggingface.co/datasets/m-a-p/FineFineWeb.MINT-1T-HTML
🍃 MINT-1T:Scaling Open-Source Multimodal Data by 10x: A Multimodal Dataset with One Trillion Tokens
🍃 MINT-1T is an open-source Multimodal INTerleaved dataset with 1 trillion text tokens and 3.4 billion images, a 10x scale-up from existing open-source datasets. Additionally, we include previously untapped sources such as PDFs and ArXiv papers. 🍃 MINT-1T is designed to facilitate research in multimodal pretraining. 🍃 MINT-1T is created by a team from the University of Washington in… See the full description on the dataset page: https://huggingface.co/datasets/mlfoundations/MINT-1T-HTML.NuminaMath-CoT
Dataset Card for NuminaMath CoT
Dataset Summary
Approximately 860k math problems, where each solution is formatted in a Chain of Thought (CoT) manner. The sources of the dataset range from Chinese high school math exercises to US and international mathematics olympiad competition problems. The data were primarily collected from online exam paper PDFs and mathematics discussion forums. The processing steps include (a) OCR from the original PDFs, (b) segmentation… See the full description on the dataset page: https://huggingface.co/datasets/AI-MO/NuminaMath-CoT.MATH-500
Dataset Card for MATH-500
This dataset contains a subset of 500 problems from the MATH benchmark that OpenAI created in their Let's Verify Step by Step paper. See their GitHub repo for the source file: https://github.com/openai/prm800k/tree/main?tab=readme-ov-file#math-splits
Calc-mawps
Dataset Card for Calc-MAWPS
Summary
The dataset is a collection of simple math word problems focused on arithmetics. It is derived from https://huggingface.co/datasets/omarxadel/MaWPS-ar.
The main addition in this dataset variant is the chain column. It was created by converting the solution to a simple html-like language that can be easily
parsed (e.g. by BeautifulSoup). The data contains 3 types of tags:
gadget: A tag whose content is intended to be evaluated by… See the full description on the dataset page: https://huggingface.co/datasets/MU-NLPC/Calc-mawps.MegaMath
MegaMath: Pushing the Limits of Open Math Copora
Megamath is part of TxT360, curated by LLM360 Team.
We introduce MegaMath, an open math pretraining dataset curated from diverse, math-focused sources, with over 300B tokens.
MegaMath is curated via the following three efforts:
Revisiting web data:
We re-extracted mathematical documents from Common Crawl with math-oriented HTML optimizations, fasttext-based filtering and deduplication, all for acquiring higher-quality data on… See the full description on the dataset page: https://huggingface.co/datasets/IFM/MegaMath.AIME_2024
AIME 2024 Dataset
Dataset Description
This dataset contains problems from the American Invitational Mathematics Examination (AIME) 2024. AIME is a prestigious high school mathematics competition known for its challenging mathematical problems.
Dataset Details
Format: JSONL
Size: 30 records
Source: AIME 2024 I & II
Language: English
Data Fields
Each record contains the following fields:
ID: Problem identifier (e.g., "2024-I-1" represents Problem 1… See the full description on the dataset page: https://huggingface.co/datasets/Maxwell-Jia/AIME_2024.MathNet
Quick Start · Overview · Tasks · Comparison · Dataset Stats · Data Sources · Pipeline · Schema · License · Citation
This is the official MathNet v0. A larger version v1 will be uploaded soon (more countires, problems and richer metadata). Schema is stable but field values may be revised in v1.
Quick start
from datasets import load_dataset
# Default: all problems
ds = load_dataset("ShadenA/MathNet", split="train")
# Or a specific country / competition-body config… See the full description on the dataset page: https://huggingface.co/datasets/ShadenA/MathNet.Magicoder-OSS-Instruct-75KThis is the OSS-Instruct dataset generated by gpt-3.5-turbo-1106 developed by OpenAI. Please pay attention to OpenAI's usage policy when adopting this dataset: https://openai.com/policies/usage-policies.
FineFineWeb-sample
FineFineWeb: A Comprehensive Study on Fine-Grained Domain Web Corpus
arXiv: Coming Soon
Project Page: Coming Soon
Blog: Coming Soon
Data Statistics
Domain (#tokens/#samples)
Iteration 1 Tokens
Iteration 2 Tokens
Iteration 3 Tokens
Total Tokens
Iteration 1 Count
Iteration 2 Count
Iteration 3 Count
Total Count
aerospace
5.77B
261.63M
309.33M
6.34B
9100000
688505
611034
10399539
agronomy
13.08B
947.41M
229.04M
14.26B
15752828
2711790
649404
19114022
artistic… See the full description on the dataset page: https://huggingface.co/datasets/m-a-p/FineFineWeb-sample.NuminaMath-1.5
Dataset Card for NuminaMath 1.5
Dataset Summary
This is the second iteration of the popular NuminaMath dataset, bringing high quality post-training data for approximately 900k competition-level math problems. Each solution is formatted in a Chain of Thought (CoT) manner. The sources of the dataset range from Chinese high school math exercises to US and international mathematics olympiad competition problems. The data were primarily collected from online exam paper PDFs… See the full description on the dataset page: https://huggingface.co/datasets/AI-MO/NuminaMath-1.5.Magicoder-Evol-Instruct-110KA decontaminated version of evol-codealpaca-v1. Decontamination is done in the same way as StarCoder (bigcode decontamination process).
COIG-CQIA
COIG-CQIA：Quality is All you need for Chinese Instruction Fine-tuning
Dataset Details
Dataset Description
欢迎来到COIG-CQIA，COIG-CQIA全称为Chinese Open Instruction Generalist - Quality is All You Need， 是一个开源的高质量指令微调数据集，旨在为中文NLP社区提供高质量且符合人类交互行为的指令微调数据。COIG-CQIA以中文互联网获取到的问答及文章作为原始数据，经过深度清洗、重构及人工审核构建而成。本项目受LIMA: Less Is More for Alignment等研究启发，使用少量高质量的数据即可让大语言模型学习到人类交互行为，因此在数据构建中我们十分注重数据的来源、质量与多样性，数据集详情请见数据介绍以及我们接下来的论文。
Welcome to the… See the full description on the dataset page: https://huggingface.co/datasets/m-a-p/COIG-CQIA.UltraData-Math
UltraData-Math
🤗 Dataset | 💻 Source Code | 🇨🇳 中文 README
UltraData-Math is a large-scale, high-quality mathematical pre-training dataset totaling 290B+ tokens across three progressive tiers—L1 (170.5B tokens web corpus), L2 (33.7B tokens quality-selected), and L3 (88B tokens multi-format refined)—designed to systematically enhance mathematical reasoning in LLMs. It has been applied to the mathematical pre-training of the MiniCPM Series models.
It was introduced in… See the full description on the dataset page: https://huggingface.co/datasets/openbmb/UltraData-Math.arabic-books
Arabic Books
Dataset Summary
The arabic-books dataset contains 8,500 rows of text, each representing the full text of a single Arabic book. These texts were extracted using the arabic-large-nougat model, showcasing the model’s capabilities in Arabic OCR and text extraction. The dataset spans a total of 1.1 billion tokens, calculated using the GPT-4 tokenizer.
This dataset is a testimony to the quality of the Arabic Nougat models and their effectiveness in extracting… See the full description on the dataset page: https://huggingface.co/datasets/MohamedRashad/arabic-books.Math-Reasoning
Math-Reasoning
Dataset Description
Mathematical problem-solving, rewriting, and dialogue data for reasoning-oriented language-model training. This repository is part of the K2 Horizon collection.
The repository is organized into multiple subsets. Every subset has a train split backed by Parquet shards, which supports Dataset Viewer inspection and streaming access.
K2 Horizon Dataset Series
Dataset repository
Focus
Subsets
IFM/TxT360-v2… See the full description on the dataset page: https://huggingface.co/datasets/IFM/Math-Reasoning.olmo-mix-1124
OLMo 2 (November 2024) Pretraining set
Collection of data used to train OLMo-2-1124 models. The majority of this dataset comes from DCLM-Baseline with no additional filtering, but we provide the explicit breakdowns below.
Name
Tokens
Bytes (uncompressed)
Documents
License
DCLM-Baseline
3.70T
21.3TB
2.95B
CC-BY-4.0
Arxiv
20.8B
77.2GB
3.95M
ODC-BY
pes2o
58.6B
412GB
38M
ODC-BY
starcoder
83.0B
458GB
78.7M
ODC-BY
Algebraic-stack
11.8B
44.0GB
2.83M
ODC-BY… See the full description on the dataset page: https://huggingface.co/datasets/allenai/olmo-mix-1124.Nemotron-CC-Math-v1
Nemotron-Pre-Training-Dataset-v1 Release
👩💻 Authors: Rabeeh Karimi Mahabadi, Sanjeev Satheesh
📘 Paper: Nemotron-cc-math: A 133 Billion-Token-Scale High Quality Math Pretraining Dataset
📝 Blog: Nemotron-cc-math blog
Data Overview
We’re excited to introduce Nemotron-CC-Math - a large-scale, high-quality math corpus extracted from Common Crawl which was used in nemotron pre-training.
This dataset is built to preserve and surface high-value mathematical and code content… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/Nemotron-CC-Math-v1.mgsm
Dataset Card for MGSM
Dataset Summary
Multilingual Grade School Math Benchmark (MGSM) is a benchmark of grade-school math problems, proposed in the paper Language models are multilingual chain-of-thought reasoners.
The same 250 problems from GSM8K are each translated via human annotators in 10 languages. The 10 languages are:
Spanish
French
German
Russian
Chinese
Japanese
Thai
Swahili
Bengali
Telugu
GSM8K (Grade School Math 8K) is a dataset of 8.5K high quality… See the full description on the dataset page: https://huggingface.co/datasets/juletxara/mgsm.dolmino-mix-1124
DOLMino dataset mix for OLMo2 stage 2 annealing training.
Mixture of high-quality data used for the second stage of OLMo2 training.
Source Sizes
Name
Category
Tokens
Bytes (uncompressed)
Documents
License
DCLM
HQ Web Pages
752B
4.56TB
606M
CC-BY-4.0
Flan
HQ Web Pages
17.0B
98.2GB
57.3M
ODC-BY
Pes2o
STEM Papers
58.6B
413GB
38.8M
ODC-BY
Wiki
Encyclopedic
3.7B
16.2GB
6.17M
ODC-BY
StackExchange
CodeText
1.26B
7.72GB
2.48M
CC-BY-SA-{2.5, 3.0, 4.0}… See the full description on the dataset page: https://huggingface.co/datasets/allenai/dolmino-mix-1124.Nemotron-Math-v2
Nemotron-Math-v2
This repository contains the dataset accompanying the paper Nemotron-Math: Efficient Long-Context Distillation of Mathematical Reasoning from Multi-Mode Supervision.
Code: NeMo-Skills
Documentation: NeMo-Skills Nemotron-Math-v2 Documentation
Dataset Description
Nemotron-Math-v2 is a large-scale mathematical reasoning dataset containing approximately 347K high-quality mathematical problems and 7M model-generated reasoning trajectories. The dataset… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/Nemotron-Math-v2.chempile-mlift
ChemPile-MLIFT
A comprehensive multimodal dataset for chemistry property prediction using vision large language models
📋 Dataset Summary
ChemPile-MLIFT is a dataset designed for multimodal chemistry property prediction tasks, specifically focusing on the prediction of chemical properties using vision large language models (VLLMs). It is part of the ChemPile project, which aims to create a comprehensive collection of chemistry-related data for training LLMs. The… See the full description on the dataset page: https://huggingface.co/datasets/jablonkagroup/chempile-mlift.MathInstruct
🦣 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.medical-o1-reasoning-SFT
News
[2025/04/22] We split the data and kept only the medical SFT dataset (medical_o1_sft.json). The file medical_o1_sft_mix.json contains a mix of medical and general instruction data.
[2025/02/22] We released the distilled dataset from Deepseek-R1 based on medical verifiable problems. You can use it to initialize your models with the reasoning chain from Deepseek-R1.
[2024/12/25] We open-sourced the medical reasoning dataset for SFT, built on medical verifiable problems and an… See the full description on the dataset page: https://huggingface.co/datasets/FreedomIntelligence/medical-o1-reasoning-SFT.FIM-Midtraining-400K
FIM-Midtraining-400K
📄 Paper · 💻 GitHub · 🤗 Collection
The mid-training corpus of "Function-Aware Fill-in-the-Middle as Mid-Training for Coding Agent Foundation Models": 400K function-aware FIM samples (~2.6B tokens under the Qwen2.5-Coder tokenizer) drawn from 75,568 Python files across 968 permissively-licensed GitHub repositories, fully decontaminated against SWE-Bench.
A coding agent's inner loop — act → observe → continue — is structurally isomorphic to a function call… See the full description on the dataset page: https://huggingface.co/datasets/TIGER-Lab/FIM-Midtraining-400K.MINT-1T-PDF-CC-2023-23
🍃 MINT-1T:Scaling Open-Source Multimodal Data by 10x: A Multimodal Dataset with One Trillion Tokens
🍃 MINT-1T is an open-source Multimodal INTerleaved dataset with 1 trillion text tokens and 3.4 billion images, a 10x scale-up from existing open-source datasets. Additionally, we include previously untapped sources such as PDFs and ArXiv papers. 🍃 MINT-1T is designed to facilitate research in multimodal pretraining. 🍃 MINT-1T is created by a team from the University of Washington in… See the full description on the dataset page: https://huggingface.co/datasets/mlfoundations/MINT-1T-PDF-CC-2023-23.xlam-function-calling-60k-parsed
[PARSED] APIGen Function-Calling Datasets (xLAM)
This dataset contains the full data from the original Salesforce/xlam-function-calling-60k
Subset name
multi-turn
parallel
multiple definition
Last turn type
number of dataset
xlam-function-calling-60k
no
yes
yes
tool_calls
60000
This is a re-parsing formatting dataset for the xLAM official dataset.
Load the dataset
from datasets import load_dataset
ds =… See the full description on the dataset page: https://huggingface.co/datasets/minpeter/xlam-function-calling-60k-parsed.Matrix
Matrix
An open-source pretraining dataset containing 4690 billion tokens, this bilingual dataset with both English and Chinese texts is used for training neo models.
Dataset Composition
The dataset consists of several components, each originating from different sources and serving various purposes in language modeling and processing. Below is a brief overview of each component:
Common Crawl
Extracts from the Common Crawl project, featuring a rich diversity of… See the full description on the dataset page: https://huggingface.co/datasets/m-a-p/Matrix.GPT-5.5-Gemini-3.1-Pro-Grok-4-Claude-Fable-5-Mythos-5-Qwen-3.7-Max-and-more-Distillation-Dataset
📖 The Open Distillation Codex
🌌 The Ultimate Open-Source Distillation Dataset — No Skip, Full, with Attack & Defense 🌌
Where 73 open-source minds converge into one unified stream of intelligence
18M+ Distilled Signals · 7,090 Raw GitHub Repositories · 8 Curated Categories · ~76 GB+
"We did not write this dataset. We assembled it.
Every line is an echo — of a model thinking, a coder drafting, a tutor explaining, a repo breathing.
Seventy-three… See the full description on the dataset page: https://huggingface.co/datasets/Manusagents/GPT-5.5-Gemini-3.1-Pro-Grok-4-Claude-Fable-5-Mythos-5-Qwen-3.7-Max-and-more-Distillation-Dataset.DAPO-Math-17k