datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
math-sft-10B
Dataset: math-sft-10B
This dataset was uploaded from /mnt/yulan_pretrain/mount/data_final_train/math-sft-10B/no-curriculum/tmp.
Math-Shepherd
Dataset Card for Math-Shepherd
Project Page: Math-Shepherd
Paper: https://arxiv.org/pdf/2312.08935.pdf
Data Loading
from datasets import load_dataset
dataset = load_dataset("peiyi9979/Math-Shepherd")
Data Instance
Every instance consists of three data fields: "input," "label," and "task".
"input": problem + step-by-step solution, e.g.,
If Buzz bought a pizza with 78 slices at a restaurant and then decided to share it with the waiter in the ratio of 5:8, with… See the full description on the dataset page: https://huggingface.co/datasets/peiyi9979/Math-Shepherd.Maths-CollegeMaths-College
I am releasing a large Mathematics dataset in the instrution format.
This extensive dataset, comprising nearly one million instructions in JSON format, encapsulates a wide array of mathematical disciplines essential for a profound understanding of the subject.
This dataset is very useful to Researchers & Model developers.
Following Fields & sub Fields are covered:
Probability
Statistics
Liner Algebra
Algebra
Group Theory
Topology
Abstract Algebra
Graph Theory
Combinatorics… See the full description on the dataset page: https://huggingface.co/datasets/ajibawa-2023/Maths-College.Maths-Grade-SchoolMaths-Grade-School
I am releasing large Grade School level Mathematics datatset.
This extensive dataset, comprising nearly one million instructions in JSON format, encapsulates a diverse array of topics fundamental to building a strong mathematical foundation.
This dataset is in instruction format so that model developers, researchers etc. can easily use this dataset.
Following Fields & sub Fields are covered:
Calculus
Probability
Algebra
Liner Algebra
Trigonometry
Differential Equations… See the full description on the dataset page: https://huggingface.co/datasets/ajibawa-2023/Maths-Grade-School.KlearReasoner-MathSub-30K
Dataset Summary
This dataset is a subset of the Klear-Reasoner Math RL dataset.The full dataset contains approximately 88K entries, while this release includes a 30K-entry subset.
The subset was obtained by filtering the outputs of DeepSeek-R1-0120. For each prompt, DeepSeek-R1-0120 generated 16 responses and we retained only the parts where the majority voting results matched the standard answers, in order to filter out the cases that the rule-based validator math_verify cannot… See the full description on the dataset page: https://huggingface.co/datasets/Kwai-Klear/KlearReasoner-MathSub-30K.arXiv-metadata-oai-snapshot-111Maths-Grade-SchoolMaths-Grade-School
I am releasing large Grade School level Mathematics datatset.
This extensive dataset, comprising nearly one million instructions in JSON format, encapsulates a diverse array of topics fundamental to building a strong mathematical foundation.
This dataset is in instruction format so that model developers, researchers etc. can easily use this dataset.
Following Fields & sub Fields are covered:
Calculus
Probability
Algebra
Liner Algebra
Trigonometry
Differential Equations… See the full description on the dataset page: https://huggingface.co/datasets/pt-sk/Maths-Grade-School.math-sft-solutions-no-cot
Math SFT Solutions No CoT
A cleaned mathematics supervised fine-tuning dataset containing:
instruction → solution pairs
mathematical proofs
derivations
olympiad-style solutions
theorem reasoning
stepwise mathematical explanations
detailed final solutions
This dataset was built specifically for mathematical supervised fine-tuning (SFT).
Unlike many reasoning datasets, this release removes explicit chain-of-thought tags and hidden thinking traces while preserving high-quality… See the full description on the dataset page: https://huggingface.co/datasets/kaushik-harsh-99/math-sft-solutions-no-cot.math_scalingmath-sft-solutions-no-cot-v3
Math SFT Solutions No CoT V3
Math SFT Solutions No CoT V3 is a large-scale mathematics supervised fine-tuning (SFT) dataset designed for instruction tuning and mathematical capability adaptation.
Version 3 substantially expands mathematical coverage while improving dataset quality through stronger filtering, cleaning, and supervision refinement.
Unlike reasoning-heavy datasets, this release focuses on clean instruction → response pairs without hidden chain-of-thought style… See the full description on the dataset page: https://huggingface.co/datasets/kaushik-harsh-99/math-sft-solutions-no-cot-v3.math-sft-solutions-no-cot-v4
Math SFT Solutions No CoT V4
Math SFT Solutions No CoT V4 is a large-scale mathematics supervised fine-tuning (SFT) dataset designed for instruction tuning and mathematical capability adaptation.
Version 4 expands dataset scale while improving supervision quality through stronger cleaning, deduplication, formatting refinement, and broader mathematical coverage.
Unlike reasoning-oriented datasets, this release focuses on direct instruction → response supervision and removes… See the full description on the dataset page: https://huggingface.co/datasets/kaushik-harsh-99/math-sft-solutions-no-cot-v4.lm-eval-results-Kukedlc-Neural-4-Maths-7b-private
Dataset Card for Evaluation run of Kukedlc/Neural-4-Maths-7b
Dataset automatically created during the evaluation run of model Kukedlc/Neural-4-Maths-7b
The dataset is composed of 62 configuration(s), each one corresponding to one of the evaluated task.
The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An… See the full description on the dataset page: https://huggingface.co/datasets/nyu-dice-lab/lm-eval-results-Kukedlc-Neural-4-Maths-7b-private.MathScaleQA-2M
Overview
This repository contains a dataset generated by reproducing the MathScaleQA using the GPT-3.5-0613 API, resulting in 2M data points. Additionally, we have attempted to reproduce the MathScale-Mistral model by finetuning the Mistral-7B-v0.1 on this dataset, following the hyperparameters in their original paper to ensure the effectiveness of our reproduction.
Reproduction Details
Fortunately, the reproduction was smooth, and we managed to match the reported… See the full description on the dataset page: https://huggingface.co/datasets/fdqerq22ds/MathScaleQA-2M.math-synth-400k
math-synth — 400k arithmetic problems with exact step-by-step scratchpads
Synthetic math SFT data where every answer is provably correct, because nothing was written
by a language model — the problems and their worked solutions are generated programmatically in
Python, so the label is the computation.
Most synthetic math datasets are distilled from an LLM teacher, which means some fraction of the
answers are silently wrong and get baked into the student. This set has no teacher… See the full description on the dataset page: https://huggingface.co/datasets/nkthebass/math-synth-400k.lumos_maths_plan_iterative
🪄 Agent Lumos: Unified and Modular Training for Open-Source Language Agents
🌐[Website]
📝[Paper]
🤗[Data]
🤗[Model]
🤗[Demo]
We introduce 🪄Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
🧩 Modular Architecture:
🧩 Lumos consists of planning, grounding… See the full description on the dataset page: https://huggingface.co/datasets/ai2lumos/lumos_maths_plan_iterative.GPTscience_maths_csmllumos_maths_ground_onetime
🪄 Agent Lumos: Unified and Modular Training for Open-Source Language Agents
🌐[Website]
📝[Paper]
🤗[Data]
🤗[Model]
🤗[Demo]
We introduce 🪄Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
🧩 Modular Architecture:
🧩 Lumos consists of planning, grounding… See the full description on the dataset page: https://huggingface.co/datasets/ai2lumos/lumos_maths_ground_onetime.multilingual-mathshepherdlumos_maths_plan_onetime
🪄 Agent Lumos: Unified and Modular Training for Open-Source Language Agents
🌐[Website]
📝[Paper]
🤗[Data]
🤗[Model]
🤗[Demo]
We introduce 🪄Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
🧩 Modular Architecture:
🧩 Lumos consists of planning, grounding… See the full description on the dataset page: https://huggingface.co/datasets/ai2lumos/lumos_maths_plan_onetime.MathSmith-Hard-ProblemsMathSmith: Towards Extremely Hard Mathematical Reasoning by Forging Synthetic Problems with a Reinforced Policy
Overview
This dataset is a collection of problems generated by the MathSmith-Hard Problem-Synthesizer.
Dataset Structure
Each record is a JSON object with the following fields:
{
"problem": "<str>", // The generated math problem
"rationale": "<str>" // The ratioanle process of question generation… See the full description on the dataset page: https://huggingface.co/datasets/Jasaxion/MathSmith-Hard-Problems.math-soft-tokens
Math Soft Tokens Dataset
Contains training steps: numinamath15_step_11_fixed.
math_sft_40KMathSmith-HC-Solution-Generation-ShortCoT-Qwen3-30B-A3BMathSmith: Towards Extremely Hard Mathematical Reasoning by Forging Synthetic Problems with a Reinforced Policy
Overview
This dataset is part of the MathSmith-HC Problem-Synthesizer collection, containing both questions and sampled answers.
It contains synthetically generated mathematical reasoning problems and their corresponding sampled solutions, produced through the reinforced problem generation pipeline described in the MathSmith framework.
Each problem is generated… See the full description on the dataset page: https://huggingface.co/datasets/Jasaxion/MathSmith-HC-Solution-Generation-ShortCoT-Qwen3-30B-A3B.math-shepherdBuild from https://huggingface.co/datasets/peiyi9979/Math-Shepherd, with reasoning steps converted into hidden states form each layer of LlaMA3.1-8b-Instruct (base pretrained version). The dataset is for training classifiers and interventors to learn inner representations from the base model.
math-sft-dataset
Dataset Card for Mori SFT Dataset
Mixed instruction data for supervised fine-tuning of the Mori small-model series. Math reasoning is the main push; a general-instruction slice is kept so compact models (around 1B–2B, including SmolLM2-1.7B) do not drop everyday instruction following.
Each row is a 3-turn conversation: a fixed system prompt, one user query, and one assistant response.
Dataset Details
Dataset Description
A single shuffled train split… See the full description on the dataset page: https://huggingface.co/datasets/grenishrai/math-sft-dataset.lumos_maths_ground_iterative
🪄 Agent Lumos: Unified and Modular Training for Open-Source Language Agents
🌐[Website]
📝[Paper]
🤗[Data]
🤗[Model]
🤗[Demo]
We introduce 🪄Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
🧩 Modular Architecture:
🧩 Lumos consists of planning, grounding… See the full description on the dataset page: https://huggingface.co/datasets/ai2lumos/lumos_maths_ground_iterative.Math-SFTMATH-self-explanationDataset MATH-Self-Explanation
This dataset contains self-generated explanation of the correct answer from MATH dataset training split.
The explanations are generated by Qwen2.5-3b-instruct.
It is used in the paper ExPO: Unlocking Hard Reasoning with Self-Explanation-Guided Reinforcement Learning for ExP-GRPO training.
maths-11th-datasetEMSF