datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
longbench-v2
Citation
@article{bai2024longbench2,
title={LongBench v2: Towards Deeper Understanding and Reasoning on Realistic Long-context Multitasks},
author={Yushi Bai and Shangqing Tu and Jiajie Zhang and Hao Peng and Xiaozhi Wang and Xin Lv and Shulin Cao and Jiazheng Xu and Lei Hou and Yuxiao Dong and Jie Tang and Juanzi Li},
journal={arXiv preprint arXiv:2412.15204},
year={2024}
}
VisualWebInstruct-Recall
Introduction
This is the dataset recalled from Google Search from the seed images.
Links
Github|
Paper|
Website
Citation
@article{visualwebinstruct,
title={VisualWebInstruct: Scaling up Multimodal Instruction Data through Web Search},
author = {Jia, Yiming and Li, Jiachen and Yue, Xiang and Li, Bo and Nie, Ping and Zou, Kai and Chen, Wenhu},
journal={arXiv preprint arXiv:2503.10582},
year={2025}
}
ReClor
ReClor: A Reading Comprehension Dataset Requiring Logical Reasoning
This repository provides the dataset from the paper ReClor: A Reading Comprehension Dataset Requiring Logical Reasoning.
We corrected the original format issues to ensure full compatibility with the Hugging Face Datasets library.
For more details, please visit the original project page.
RecetasDeLaAbuela
Motivación inicial
Este corpus ha sido creado durante el Hackathon SomosNLP Marzo 2024: #Somos600M (https://somosnlp.org/hackathon).
Responde a una de las propuestas somosnlp sobre 'Recetas típicas por país/zona geográfica'.
Nombre del Proyecto
Este corpus o dataset se llama 'RecetasDeLaAbuel@' y es un homenaje a todas nuestr@s abuel@s que nos han enseñado a cocinar. Se trata de la mayor y más completa colección de recetas open-source en español de países… See the full description on the dataset page: https://huggingface.co/datasets/somosnlp/RecetasDeLaAbuela.recursive-cognition-corpus
LuisCore Recursive Cognition Corpus
LuisCore is a low-latency decentralized runtime substrate for multi-step inference at scale.
Generated: 2026-09-24T11:09:13.307Z
Rows: 13236
Owner: Luis610348
Canonical site: https://luiscore.com
What this dataset is
LuisCore is a recursive cognition infrastructure. This dataset is the public
LLM Discovery Corpus — a stable, deterministic Q&A set used by LuisCore to
help language models accurately describe, cite, and verify… See the full description on the dataset page: https://huggingface.co/datasets/Luis610348/recursive-cognition-corpus.recube-data
Data
This directory contains all benchmark data for the Re2Code repository-level code reconstruction benchmark.
Download
All data files are hosted on Hugging Face and can be downloaded using:
# Install huggingface_hub if not already installed
pip install huggingface_hub
# Download the entire dataset
huggingface-cli download wlqmfl1999/recube-data --repo-type=dataset --local-dir data/
# Or download in Python
from huggingface_hub import snapshot_download… See the full description on the dataset page: https://huggingface.co/datasets/wlqmfl1999/recube-data.LLaVA-ReCap-676KThis is an integrated version of LLaVA-ReCap, sourced from lmms-lab/LLaVA-ReCap-558K and lmms-lab/LLaVA-ReCap-118K.
In this version, the conversations field has been split into two separate fields: prompt and response. Additionally, the <image> special token has been removed to facilitate customization.
Inspired by the original paper, the prompt field has been further expanded with human-crafted variations. Specifically, each prompt is sampled from one of the following 30 instructions:… See the full description on the dataset page: https://huggingface.co/datasets/LimeryJorge/LLaVA-ReCap-676K.bitcoin-wallet-recovery-faq
Bitcoin Wallet Recovery FAQ Dataset v1.0
A high-quality Question & Answer dataset focused exclusively on Bitcoin wallet recovery and self-custody best practices. It is designed for training, fine-tuning, and evaluating LLMs and retrieval-augmented generation (RAG) systems in the domain of bitcoin security, seed backup, device loss, and fund recovery.
Dataset Summary
Total records: 500
Language: English
Answer length: 150–300 words per record
Categories: 39… See the full description on the dataset page: https://huggingface.co/datasets/ismailtasdelen/bitcoin-wallet-recovery-faq.recube-dataThis dataset serves as the official data repository for the ReCUBE benchmark; the corresponding evaluation codebase and execution instructions can be found at https://anonymous.4open.science/r/ReCUBE-E0FB/README.md.
Data
This directory contains all benchmark data for ReCUBE.
Download
All data files are hosted on Hugging Face and can be downloaded using:
# Install huggingface_hub if not already installed
pip install huggingface_hub
# Download the entire dataset… See the full description on the dataset page: https://huggingface.co/datasets/recube-anon-2026/recube-data.substream-recollection
Substream Recollection
A controlled benchmark for membership recall, designed to test properties of memory beyond accuracy in LLMs.
Each row is a (stream, probe, label) tuple: the model sees a long input stream and a short candidate, and answers whether the candidate (or in the case of natural video, the referenced action) occurred inside the stream.
config
rows
content
text
7,640
synthetic substream questions, text modality, L=8…4096
synthetic_video
6,065
the same… See the full description on the dataset page: https://huggingface.co/datasets/anonstreammem/substream-recollection.EduBench
EduBench 📚
EduBench é um benchmark em português brasileiro para avaliação de Large Language Models (LLMs) em tarefas educacionais, composto por 3,149 questões discursivas extraídas de vestibulares de alta competitividade.
GitHub
Paper
Dataset Description
Fontes
USP: Universidade de São Paulo
UNICAMP: Universidade Estadual de Campinas
UNESP: Universidade Estadual Paulista
Período
2015-2025 (11 anos de provas)
Áreas do… See the full description on the dataset page: https://huggingface.co/datasets/recogna-nlp/EduBench.RecurrReason
RecurrReason: Recurrent Reasoning on Symbolic Puzzles
A difficulty-controlled benchmark for evaluating multi-step reasoning in language models
📋 Table of Contents
Overview
Dataset Structure
Puzzles
Quick Start
Citation
License
🎯 Overview
RecurrReason is a benchmark of four recurrent logic puzzles with optimal trajectories and controlled difficulty scaling (N=1 to 10). It tests whether language models can:
Find optimal (minimal-length)… See the full description on the dataset page: https://huggingface.co/datasets/gmannem/RecurrReason.movie_recommendationMovie recommendation task based on the Movielens datasetpubmedqa-recursive-llm-degradation-qwen2.5-0.5b
PubMedQA Recursive LLM Degradation — Qwen2.5-3B
This repository contains synthetic biomedical question-answering data
and model predictions generated as part of a study of recursive
fine-tuning and model degradation.
Base Model
Qwen/Qwen2.5-3B
Source Dataset
The experiments use the PubMedQA dataset:
qiaoxin/PubMedQA
This repository contains generated/derived research artifacts and does
not redistribute the original PubMedQA dataset in its entirety.… See the full description on the dataset page: https://huggingface.co/datasets/chrislimbe/pubmedqa-recursive-llm-degradation-qwen2.5-0.5b.gigaverbo-v2-rec-sft
GigaVerbo-v2 REC SFT
A model should not merely know how to reason; it should learn when reasoning is worth the cost.
Dataset repository: OliveiraJLT/gigaverbo-v2-rec-sftBase dataset: Polygl0t/gigaverbo-v2-sftAnswer-generation model: openai/gpt-oss-20bQuality classifier: Polygl0t/portuguese-qwen3-4b-instruct-quality-classifierReasoning translation model and token accounting tokenizer: Qwen/Qwen3.5-9B
Dataset Summary
GigaVerbo-v2 REC SFT — short for GigaVerbo-v2… See the full description on the dataset page: https://huggingface.co/datasets/OliveiraJLT/gigaverbo-v2-rec-sft.pubmedqa-recursive-llm-degradation-qwen2.5-3b
PubMedQA Recursive LLM Degradation — Qwen2.5-3B
This repository contains synthetic biomedical question-answering data
and model predictions generated as part of a study of recursive
fine-tuning and model degradation.
Base Model
Qwen/Qwen2.5-3B
Source Dataset
The experiments use the PubMedQA dataset:
qiaoxin/PubMedQA
This repository contains generated/derived research artifacts and does
not redistribute the original PubMedQA dataset in its entirety.… See the full description on the dataset page: https://huggingface.co/datasets/chrislimbe/pubmedqa-recursive-llm-degradation-qwen2.5-3b.tool-reasoning-sft-TOOLS-toolace-sft-tool-use-agent-data-cleaned-rectified
ToolACE - Tool-Use Agent Data Cleaned & Rectified
👥 Follow the Author
Aman Priyanshu
Overview
This dataset is a cleaned and restructured version of the Team-ACE/ToolACE dataset. ToolACE is a high-quality conversational tool-use dataset containing 11,300+ examples of natural language interactions requiring function calling across diverse domains. This version converts the original OpenAI function-call format into a standardized multi-turn tool-use… See the full description on the dataset page: https://huggingface.co/datasets/AmanPriyanshu/tool-reasoning-sft-TOOLS-toolace-sft-tool-use-agent-data-cleaned-rectified.Generated-Recovery-Support-Dialogues
# Empathetic Conversations for Addiction Recovery Support Dataset
Dataset Description
This dataset contains synthetically generated conversational examples between a user discussing their addiction recovery journey and an AI assistant designed to be empathetic, supportive, non-judgmental, and encouraging. The conversations are in English and cover various stages and aspects of the recovery process, following established therapeutic guidelines and models.
The dataset is… See the full description on the dataset page: https://huggingface.co/datasets/filippo19741974/Generated-Recovery-Support-Dialogues.tool-reasoning-sft-TOOLS-hermes_reasoning_tool_use-data-cleaned-rectified
Hermes Reasoning Tool Use — Cleaned & Rectified
👥 Follow the Author
Aman Priyanshu
Overview
This dataset is a cleaned and restructured version of interstellarninja/hermes_reasoning_tool_use. The original dataset uses the Hermes/NousResearch multi-turn format with from/value fields and embedded <think> + <tool_call> tags inside single gpt turns. This version converts it into a strict multi-turn conversation structure with validated role transitions.… See the full description on the dataset page: https://huggingface.co/datasets/AmanPriyanshu/tool-reasoning-sft-TOOLS-hermes_reasoning_tool_use-data-cleaned-rectified.Home-Assistant-requests-for-intent-detection-and-function-recognition
Home Assistant Requests V2 Dataset
This dataset contains a list of requests and responses for a user interacting with a personal assistant that controls an instance of Home Assistant.
The updated V2 of the dataset is now multilingual, containing data in English, German, French, Spanish, and Polish. The dataset also contains multiple "personalities" for the assistant to respond in, such as a formal assistant, a sarcastic assistant, and a friendly assistant. Lastly, the dataset has… See the full description on the dataset page: https://huggingface.co/datasets/DaftP/Home-Assistant-requests-for-intent-detection-and-function-recognition.openclaw-recursive-study-data
OpenClaw Recursive Repository Study Data
Synthetic repository-study data generated against
openclaw/openclaw at commit
da228660306b55a9cce3b973946f3aacfc515848. The source repository is MIT licensed.
This release contains exploration questions, tool-using study trajectories,
recursive notes, full recall-rewritten trajectories, and recall-to-action
training examples. Nested chat/tool objects are stored as JSON strings to keep
the schema stable and can be decoded with json.loads.… See the full description on the dataset page: https://huggingface.co/datasets/aviralku/openclaw-recursive-study-data.turkish-recipe-dataset
Yapılandırılmış Türk Yemek Tarifleri
Structured Turkish Recipes Dataset
Bu veri seti, mertbozkurt/llama2-TR-recipe veri setinden türetilmiş ve Groq API üzerinden 3 farklı LLM modeli kullanılarak yapılandırılmış 3,320 benzersiz Türk yemek tarifi içermektedir.
Kullanılan Modeller (Multi-Model Rotasyon)
Model
İşlenen Tarif
Oran
Llama 4 Scout 17B (meta-llama/llama-4-scout-17b-16e-instruct)
1,614
%44.3
Llama 3.1 8B Instant (llama-3.1-8b-instant)
1,551%42.6… See the full description on the dataset page: https://huggingface.co/datasets/104bit/turkish-recipe-dataset.tool-reasoning-sft-RESEARCH-grill-lab-browsecomp-plus-runs-data-cleaned-rectified
Tool-Reasoning SFT — BrowseComp-Plus Runs (Cleaned & Rectified)
Multi-turn tool-use reasoning trajectories derived from grill-lab/browsecomp-plus-runs, converted to a structured SFT format following the interstellarninja/hermes_reasoning_tool_use convention.
Source
Based on the execution trajectories from "Revisiting Text Ranking in Deep Research" (arXiv:2602.21456):
Original data: grill-lab/browsecomp-plus-runs (MIT)
Format
Each row contains a messages… See the full description on the dataset page: https://huggingface.co/datasets/AmanPriyanshu/tool-reasoning-sft-RESEARCH-grill-lab-browsecomp-plus-runs-data-cleaned-rectified.ReceiptQA
ReceiptQA: A Comprehensive Dataset for Receipt Understanding and Question Answering
ReceiptQA is a large-scale dataset specifically designed to support and advance research in receipt understanding through question-answering (QA) tasks. This dataset offers a wide range of questions derived from real-world receipt images, addressing diverse challenges such as text extraction, layout understanding, and numerical reasoning. ReceiptQA provides a benchmark for evaluating and improving… See the full description on the dataset page: https://huggingface.co/datasets/mahmoud2019/ReceiptQA.RECOR
RECOR: Reasoning-focused Multi-turn Conversational Retrieval Benchmark
A benchmark for evaluating reasoning-intensive conversational information retrieval systems.
Statistics
Metric
Value
Total Conversations
707
Total Turns
2,971
Domains
11
Avg. Turns per Conversation
4.2
Domains
Source
Domains
BRIGHT
biology, earth_science, economics, psychology, robotics, sustainable_living
StackExchange
Drones, hardware, law… See the full description on the dataset page: https://huggingface.co/datasets/RECOR-Benchmark/RECOR.nepali-recipes-qwen-processed
Nepali Recipes for Qwen Fine-tuning
Dataset Description
This dataset contains 1227 Nepali recipes formatted for fine-tuning Qwen models using ChatML format.
Train Split: 900 recipes
Test Split: 327 recipes
Language: Nepali (ne)
Format: Qwen ChatML
Base Model: Qwen/Qwen2-1.5B
Dataset Structure
Data Fields
text: Full ChatML formatted prompt with answer (for training)
test_text: ChatML prompt without answer (for inference)
name: Recipe name in Nepali… See the full description on the dataset page: https://huggingface.co/datasets/sijanpaudel/nepali-recipes-qwen-processed.chat_restaurant_recommendation
Restaurant chat dataset
This dataset contains approximately 600 chat interactions mimicking various user tones with different restaurant categories
The data were generated by Gemini-pro
The purpose of this dataset is to serve as a calibration dataset for a restaurant recommendation LLM chatbot
RECIPER
RECIPER: A Dual-View Retrieval Pipeline for Procedure-Oriented Materials Question Answering
Official dataset and reference implementation of RECIPER
RECIPER: A Dual-View Retrieval Pipeline for Procedure-Oriented Materials Question Answering
Zhuoyu Wu, Wenhui Ou, Pei-Sze Tan, Wenqi Fang, Sailaja Rajanala, and Raphaël C.-W. Phan
RECIPER is a retrieval pipeline for procedure-oriented materials question answering. It indexes two complementary views of the same scientific paper… See the full description on the dataset page: https://huggingface.co/datasets/ReaganWZY/RECIPER.Recurv-Medical-Dataset
🩺 Recurv-Medical-Dataset:
The Recurv-Medical-Dataset is a comprehensive resource of 67,299 high-quality question-answer pairs explicitly designed for training and fine-tuning medical AI models. Curated from trusted medical sources, this dataset focuses on real-world scenarios like anamnesis, diagnostics, and treatment recommendations. It sets a new benchmark for advancing conversational AI in the healthcare domain.
📈 Dataset Statistics
FeatureValue
Number… See the full description on the dataset page: https://huggingface.co/datasets/RecurvAI/Recurv-Medical-Dataset.recency-probe-2026
Recency Probe 2026
128 questions about facts that entered the record between January and August
2026, asked in Ukrainian and in English. A model trained before 2026 cannot
answer them from what it knows; a model that has read 2026 Ukrainian news can.
Built from
Goader/ukrainian-news-2026
— 429k articles from 23 Ukrainian national outlets. Every item is grounded in
quotes from that corpus, which ship with the item.
What makes an item
Every candidate was put to a… See the full description on the dataset page: https://huggingface.co/datasets/Goader/recency-probe-2026.