datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
AIME_MATH_1000_LONG_COTlongcot
LongCoT
LongCoT is a benchmark for long-horizon reasoning across logic, computer science, chemistry, chess, and mathematics. This Hugging Face release contains the benchmark data in viewer-friendly Parquet format for browsing and loading with datasets.
The canonical codebase, verifier, and evaluation harness live at:
https://github.com/LongHorizonReasoning/longcot
Overview
LongCoT measures whether models can sustain coherent reasoning across long chains of thought. The… See the full description on the dataset page: https://huggingface.co/datasets/LongHorizonReasoning/longcot.Demeter-LongCoT-6M
Demeter-LongCoT-6M
Demeter-LongCoT-6M is a high-quality, compact chain-of-thought reasoning dataset curated for tasks in mathematics, science, and coding. While the dataset spans diverse domains, it is primarily driven by mathematical reasoning, reflecting a major share of math-focused prompts and long-form logical solutions.
Quick Start with Hugging Face Datasets🤗
pip install -U datasets
from datasets import load_dataset
dataset =… See the full description on the dataset page: https://huggingface.co/datasets/prithivMLmods/Demeter-LongCoT-6M.LONGCOT-merged-1Mthis is PowerInfer/QWQ-LONGCOT-500K + PowerInfer/LONGCOT-Refine-500K shuffled together with the following changes:
pointless starting phrases at the beginning (in english) such as "Okay, ..." "So, ..." etc are removed
config en has been filtered to include only rows detected as lang en in both prompt and response columns via fasttext
AIME_MATH_100_LONG_COTQwQ-LongCoT-130KAlso have a look on the second version here => QwQ-LongCoT-2
Figure 1: Just a cute picture generate with [Flux](https://huggingface.co/Shakker-Labs/FLUX.1-dev-LoRA-Logo-Design)
Today, I’m excited to release QwQ-LongCoT-130K, a SFT dataset designed for training O1-like large language models (LLMs). This dataset includes about 130k instances, each with responses generated using QwQ-32B-Preview. The dataset is available under the Apache 2.0 license, so feel free to use it as you like.… See the full description on the dataset page: https://huggingface.co/datasets/amphora/QwQ-LongCoT-130K.QWQ-LONGCOT-500KThis repository contains approximately 500,000 instances of responses generated using QwQ-32B-Preview language model. The dataset combines prompts from multiple high-quality sources to create diverse and comprehensive training data.
The dataset is available under the Apache 2.0 license.
Over 75% of the responses exceed 8,000 tokens in length. The majority of prompts were carefully created using persona-based methods to create challenging instructions.
Bias, Risks, and Limitations… See the full description on the dataset page: https://huggingface.co/datasets/Tiiny/QWQ-LONGCOT-500K.nl2repo-agent-longcotnl2repo-longcotQwQ-LongCoT-Verified-130KOriginal Dataset: amphora/QwQ-LongCoT-130K
QwQ 32B Preview isn't perfect :)
Note: Around 5-7% of the processed data might be incorrectly labeled as "unverified" because QwQ's output isn't exactly the same as the original solution from NuminaMathCoT. I believe this can be solved with another round of processing with a smarter model but Qwen 2.5 3B Instruct is good enough to check if the solution is exactly the same. Magpie data is also "unverified" and has an empty "solution" column.… See the full description on the dataset page: https://huggingface.co/datasets/qingy2024/QwQ-LongCoT-Verified-130K.LONGCOT-Refine-500KThis repository contains approximately 500,000 instances of responses generated using Qwen2.5-72B-Instruct. The dataset combines prompts from multiple high-quality sources to create diverse and comprehensive training data.
The dataset is available under the Apache 2.0 license.
Bias, Risks, and Limitations
This dataset is mainly in English.
The dataset inherits the biases, errors, and omissions known to exist in data used for seed sources and models used for data generation.
This… See the full description on the dataset page: https://huggingface.co/datasets/Tiiny/LONGCOT-Refine-500K.Demeter-LongCoT-400K
Demeter-LongCoT-400K
Demeter-LongCoT-400K is a high-quality, compact chain-of-thought reasoning dataset curated for tasks in mathematics, science, and coding. While the dataset spans diverse domains, it is primarily driven by mathematical reasoning, reflecting a major share of math-focused prompts and long-form logical solutions.
Quick Start with Hugging Face Datasets🤗
pip install -U datasets
from datasets import load_dataset
dataset =… See the full description on the dataset page: https://huggingface.co/datasets/prithivMLmods/Demeter-LongCoT-400K.QwQ-LongCoT-130K-cleanedCleaned version of amphora/QwQ-LongCoT-130K with Chinese language entries removed
MATH_training_split_long_cot
Small Model Learnability Gap
Arxiv Technical Report: https://arxiv.org/pdf/2502.12143v1
Abstract (Click Here)
Large language models (LLMs) excel in complex reasoning tasks, and distilling their reasoning capabilities into smaller models has shown promise. However, we uncover an interesting phenomenon, which we term the Small Model Learnability Gap: small models (≤3B parameters) do not consistently benefit from long chain-of-thought (CoT) reasoning or distillation from larger… See the full description on the dataset page: https://huggingface.co/datasets/UWNSL/MATH_training_split_long_cot.QwQ-LongCoT-130K-2A second version of the QwQ-LongCoT dataset.This time, it's only from using the magpie I introduced earlier.
QWQ-LongCOT-AIMOQWQ-LongCOT-AIMO is a derived dataset created by processing the amphora/QwQ-LongCoT-130K dataset. It filters the original dataset to focus specifically on question-answering pairs where the final answer is a numerical value between 0 and 999, explicitly marked using the \boxed{...} format within the original chain-of-thought answer.
Dataset Structure
Data Splits
The dataset is split into training, validation, and test sets with an 80/10/10 ratio based on the filtered… See the full description on the dataset page: https://huggingface.co/datasets/Floppanacci/QWQ-LongCOT-AIMO.math-train-qwq-rs-n32QWQ-LONGCOT-500KThis dataset is a copy of PowerInfer/QWQ-LONGCOT-500K.
This repository contains approximately 500,000 instances of responses generated using QwQ-32B-Preview language model. The dataset combines prompts from multiple high-quality sources to create diverse and comprehensive training data.
The dataset is available under the Apache 2.0 license.
Over 75% of the responses exceed 8,000 tokens in length. The majority of prompts were carefully created using persona-based methods to create challenging… See the full description on the dataset page: https://huggingface.co/datasets/huihui-ai/QWQ-LONGCOT-500K.long-cot-4kOriginal dataset taken from here: amphora/QwQ-LongCoT-130K
All responses containing Chinese characters have been removed (There's a QwQ quirk where it switches to Chiense)
ShareGPT "conversations" added with the system prompt I used to train gghfez/Mistral-Large-2407-LongCoT
Only "conversations" 1024-4096 tokens long were retained (counted using the Mistral-Large-2407 tokenizer)
If you're using a more efficient tokenizer like Qwen or Llama3, the sequence length will be shorter.
If you're… See the full description on the dataset page: https://huggingface.co/datasets/gghfez/long-cot-4k.MathSmith-HC-Solution-Generation-LongCoT-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 (LongCoT setting).
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… See the full description on the dataset page: https://huggingface.co/datasets/Jasaxion/MathSmith-HC-Solution-Generation-LongCoT-Qwen3-30B-A3B.QwQ-LongCoT-500K-CleanedScaleQuest-Math-Qwen2.5-LongCoTWe introduce ScaleQuest, a scalable and novel data synthesis method that utilizes small-size open-source models to generate questions from scratch.
Paper: Unleashing Reasoning Capability of LLMs via Scalable Question Synthesis from Scratch
QwQ-Long-CoT-30k-subset-Llama3.1-8B-dynamic-perturbation-regex-generation-max-marginlong_cot_calibrationQwQ-LongCoT-130K-decontaminatedDecontaminated version of gghfez/QwQ-LongCoT-130K-cleaned that remove the collided data from math/test, gsm8k/test, olympiadbench/test, minerva_math/test, college_math/test, mmlu_stem/test, gaokao, amc23, aime24 and math500.
Total number of rows: 124594
LongCOT-ReasonQwQ-LongCoT-59k-cleanedAIME_MATH_10_LONG_COTQwQ-Long-CoT-30k-subset-Llama3.1-8B-Instruct-on-policy-step-wise-correct-trajectoryLONGCOT-Refine-5K5K subset of LONGCOT-Refine-500K to test with. All samples (prompt + response) are 6k words or less (so context length should be < 8192 tokens when training)