datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
linalg-bench-math-ai-neurips
LinAlg-Bench: A Benchmark Exposing Structural Failure Modes in LLM Linear Algebra — Where Models Stop Computing and Start Hallucinating
This dataset is the official release accompanying the MATH-AI 2026 NeurIPS workshop paper, "LinAlg-Bench: A Benchmark Exposing Structural Failure Modes in LLM Linear Algebra — Where Models Stop Computing and Start Hallucinating." The exact 660-problem core evaluated in that paper (9 tasks × 3 matrix sizes, 6,600 model outputs, 1,156… See the full description on the dataset page: https://huggingface.co/datasets/mst-ai/linalg-bench-math-ai-neurips.gsm8k-translated
Multilingual GSM8K Translations
This dataset contains machine-translated versions of GSM8K in these languages:
French (fr)
German (de)
Hindi (hi)
Dataset Structure
For each language, we provide the original GSM8K train and test splits:
train: 7,473 samples
test: 1,319 samples
Each sample consists of a question and an answer.
The question describes a grade-school-level math word problem that requires multi-step mathematical reasoning. The answer contains a… See the full description on the dataset page: https://huggingface.co/datasets/math-across-languages/gsm8k-translated.Fast-Math-R1-SFTThis repository contains the First stage SFT dataset as presented in the paper A Practical Two-Stage Recipe for Mathematical LLMs: Maximizing Accuracy with SFT and Efficiency with Reinforcement Learning.
This dataset is used for the intensive Supervised Fine-Tuning (SFT) phase, crucial for pushing the model's mathematical accuracy.
Project GitHub Repository: https://github.com/RabotniKuma/Kaggle-AIMO-Progress-Prize-2-9th-Place-Solution
Dataset Construction
This dataset was… See the full description on the dataset page: https://huggingface.co/datasets/RabotniKuma/Fast-Math-R1-SFT.Math_CoT_Arabic_English_Reasoning
Math CoT Arabic English Dataset
A high-quality, bilingual (English & Arabic) dataset for Chain-of-Thought (COT) reasoning in mathematics and related disciplines, developed by Miscovery AI.
Overview
Math-COT is a unique dataset designed to facilitate and benchmark the development of chain-of-thought reasoning capabilities in language models across mathematical domains. With meticulously crafted examples, explicit reasoning steps, and bilingual support, this dataset offers… See the full description on the dataset page: https://huggingface.co/datasets/miscovery/Math_CoT_Arabic_English_Reasoning.Fast-Math-R1-GRPOThis repository contains the second-stage GRPO dataset for the paper A Practical Two-Stage Recipe for Mathematical LLMs: Maximizing Accuracy with SFT and Efficiency with Reinforcement Learning.
This dataset is crucial for the second stage of the training recipe, aiming to improve token efficiency while preserving peak mathematical reasoning performance in Large Language Models (LLMs) through Reinforcement Learning from online inference (GRPO).
We extracted the answers from the 2nd stage SFT… See the full description on the dataset page: https://huggingface.co/datasets/RabotniKuma/Fast-Math-R1-GRPO.math-intuition-20260906-403-easy-30
math-intuition-20260906-403-easy-30
12,090 synthetic mathematics problems drawn from 403 problem families, each family
derived from a distinct arXiv paper. Every problem is generated answer-first, so the
answer is known by construction and is checked by the family's own verify() before
the row is written. No row in this file is ungraded.
This is the easy slice: 30 instances per family at each family's easiest difficulty
preset. It is not the hard benchmark — see Difficulty… See the full description on the dataset page: https://huggingface.co/datasets/amphora/math-intuition-20260906-403-easy-30.math-intuition-20260906-403-demo-10
math-intuition-20260906-403-demo-10
3,936 mathematics problems drawn from 403 problem families, each derived from a
distinct arXiv paper. Every problem is generated answer-first, so the answer is known by
construction and is checked by the family's own verify() before the row is written.
No row in this file is ungraded.
This is the demo rung — read this before using it
Each family exposes a four-rung ladder: demo, easy, medium, hard. This file samples
demo, which… See the full description on the dataset page: https://huggingface.co/datasets/amphora/math-intuition-20260906-403-demo-10.Math-IIO-68K-Mini
Mathematics Dataset for AI Model Training
This dataset contains 68,000 rows of mathematical questions and their corresponding solutions. It is designed for training AI models capable of solving mathematical problems or providing step-by-step explanations for a variety of mathematical concepts. The dataset is structured into three columns: input, instruction, and output.
Dataset Overview
Input: A mathematical question or problem statement (e.g., arithmetic, algebra… See the full description on the dataset page: https://huggingface.co/datasets/prithivMLmods/Math-IIO-68K-Mini.Math-Solve
Overview
The Math-Solve dataset is a collection of math problems and their solutions, designed to facilitate training and evaluation of models for tasks such as text generation, question answering, and summarization. The dataset contains nearly 25k rows of math-related problems, each paired with a detailed solution.
This dataset is particularly useful for researchers and developers working on AI models that require mathematical reasoning and problem-solving capabilities.… See the full description on the dataset page: https://huggingface.co/datasets/prithivMLmods/Math-Solve.math-squared
Dataset Name
MATH2
Dataset Description
MATH2 is a mathematical reasoning evaluation dataset curated using a human-in-the-loop approach proposed in the paper AI-Assisted Generation of Difficult Math Questions. The dataset consists of 210 questions formed by combining 2 math domain skills using frontier LLMs. These skills were extracted from the MATH [Hendrycks et al., 2021] dataset.
Dataset Sources
Paper: AI-Assisted Generation of Difficult Math… See the full description on the dataset page: https://huggingface.co/datasets/veds12/math-squared.math-code-qa
Math & Code QA — Instruction Dataset
Worked mathematical solutions and short code answers, built for the
Adaption Labs AutoScientist Challenge (Math & Code category).
Rows
5,200
Math
3,600
Code
1,600
Distinct answers
5,199 (100%)
Duplicate questions
none
Nulls
none
Question length
median 27 words
Answer length
median 58 words (max 89)
License
CC-BY-4.0
What makes the math rows unusual
Every math answer is short worked reasoning… See the full description on the dataset page: https://huggingface.co/datasets/flamiinngo/math-code-qa.dumy-zno-ukrainian-math-history-geo-r1-o1
DUMY («Думи»): Ukrainian Multidomain Reasoning Dataset (Part 1: ZNO/NMT tasks with DeepSeek R1 and OpenAI o1 answers)
DUMY is an open benchmark and dataset designed for training, distillation, and evaluation of language models focused on Ukrainian reasoning tasks.
The word “Dumy” comes from Taras Shevchenko’s famous poem and literally means “thoughts” in Ukrainian:
Думи мої, думи мої,
Лихо мені з вами!
Нащо стали на папері
Сумними рядами?..
Work in progress. Stay tuned.… See the full description on the dataset page: https://huggingface.co/datasets/NLPForUA/dumy-zno-ukrainian-math-history-geo-r1-o1.spanish_imdb_synopsis
Dataset Card for Spanish IMDb Synopsis
Dataset Description
4969 movie synopsis from IMDb in spanish.
Dataset Summary
[N/A]
Languages
All descriptions are in spanish, the other fields have some mix of spanish and english.
Dataset Structure
[N/A]
Data Fields
description: IMDb description for the movie (string), should be spanish
keywords: IMDb keywords for the movie (string), mix of spanish and english
genre: The genres of the… See the full description on the dataset page: https://huggingface.co/datasets/mathigatti/spanish_imdb_synopsis.ru_math500
Карточка датасета MATH-500 (перевод на русский язык)
Этот датасет представляет собой перевод на русский язык поднабора из 500 задач из бенчмарка MATH, созданного OpenAI в рамках работы Let’s Verify Step by Step.
Исходные данные взяты из репозитория OpenAI:
https://github.com/openai/prm800k/tree/main#math-splits
Описание
Датасет содержит 500 математических задач с решениями, аналогичных тем, что используются для оценки моделей на способность рассуждать пошагово.
Все… See the full description on the dataset page: https://huggingface.co/datasets/AvitoTech/ru_math500.math-code-qa-v2
Math & Code QA v2 — Instruction Dataset
Worked mathematical solutions and short code answers, spanning arithmetic word
problems through to algebra, geometry and combinatorics.
Built for the Adaption Labs AutoScientist Challenge (Math & Code category).
The model trained on this beats Llama-3.3-70B-Instruct 72 to 28 on the
held-out Math category evaluation.
Rows
5,297 (4,197 math, 1,100 code)
Distinct answers
5,297 (100%)
Duplicate questions
none
Nulls
none… See the full description on the dataset page: https://huggingface.co/datasets/flamiinngo/math-code-qa-v2.ru-alpaca-math
Это переработка в alpaca-friendly формат датасетов от:
MERA-evaluation[MERA]
Из датасета взяты и переработаны только subsets (mamuramu, mathlogica, rummlu, rumodar, rumultiar)
Vikhrmodels[russian_math]
Датасет переработан с учетом неободимого формата.
math-toolcall-tr-benchmark
math-toolcall-tr-benchmark
bilalabic/gemma_4_math-toolcall-tr_lora
LoRA adaptörünü temel Gemma-4 E4B modeliyle karşılaştıran benchmark sonuçları.
Bu depo yalnızca değerlendirme çıktılarını içerir. Eğitim veri seti ayrı olarak
bilalabic/math-toolcall-tr
adresinde yayımlanmaktadır.
Benchmark'lar
Benchmark
Örnek
Ölçülen davranış
Türkçe MMLU
250
Genel bilgi doğruluğu ve eğitim sonrası bilgi kaybı
Matematik Tool-Call
150
Araç seçimi, çekimserlik ve çıktı… See the full description on the dataset page: https://huggingface.co/datasets/bilalabic/math-toolcall-tr-benchmark.Math-Solve-Singleshot
Math-Solve-Singleshot
Overview
This dataset, named Math-Solve-Singleshot, is designed for solving single-shot mathematical problems. It contains a variety of math problems formatted in text, suitable for training and evaluating models on mathematical reasoning tasks.
Modalities
Text
Formats: CSV
Size: 1.05M rows
Libraries: pandas
Croissant
License: Apache-2.0
Dataset Details
Train Split: 1.05 million rows
Problem String Lengths:
Length 1: 16… See the full description on the dataset page: https://huggingface.co/datasets/prithivMLmods/Math-Solve-Singleshot.Math-Forge-Hard
Math-Forge-Hard Dataset
Overview
The Math-Forge-Hard dataset is a collection of challenging math problems designed to test and improve problem-solving skills. This dataset includes a variety of word problems that cover different mathematical concepts, making it a valuable resource for students, educators, and researchers.
Dataset Details
Modalities
Text: The dataset primarily contains text data, including math word problems.
Formats… See the full description on the dataset page: https://huggingface.co/datasets/prithivMLmods/Math-Forge-Hard.Math_small_corpus