datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
hermes-function-calling-v1-jsonl
Hermes Function-Calling V1
This dataset is the compilation of structured output and function calling data used in the Hermes 2 Pro series of models.
This repository contains a structured output dataset with function-calling conversations, json-mode, agentic json-mode and structured extraction samples, designed to train LLM models in performing function calls and returning structured output based on natural language instructions. The dataset features various conversational scenarios… See the full description on the dataset page: https://huggingface.co/datasets/minpeter/hermes-function-calling-v1-jsonl.minimax-m3-claude-code-tracesThis dataset was generated using teich by TeichAI
Prepare these datasets for supervised fine-tuning in just a few lines of code — see the Conversion section below.
Minimax M3 Claude Code Traces
This directory contains raw agent trace files generated by teich.
All assistant responses were generated by minimax/minimax-m3.
JSONL files: 31
Format
Each file is newline-delimited JSON representing a single captured agent session.
The trace schema is designed for… See the full description on the dataset page: https://huggingface.co/datasets/armand0e/minimax-m3-claude-code-traces.MindTheGap
Mind the Gap
Dataset Description
This dataset accompanies the paper "Mind the Gap: How Elicitation Protocols Shape the Stated-Revealed Preference Gap in Language Models" and extends the original AIRiskDilemmas dataset with comprehensive model evaluation results.
Authors: Pranav Mahajan, Ihor Kendiukhov, Syed Hussain, Lydia Nottingham
Repository: SPAR-SvR/Mind-the-Gap
Original Dataset: AIRiskDilemmas
Key Contribution
We systematically study how elicitation… See the full description on the dataset page: https://huggingface.co/datasets/LydiaNottingham/MindTheGap.swe-bench-mini
SWE-bench-mini
34 self-contained bug-fix tasks in the SWE-bench format — a small repository snapshot
carrying a defect, a test that fails because of it, and a gold patch that fixes it (difficulty
mix: 12 easy / 19 medium / 3 hard, author estimate). Built for the swe_bench_mini agent and the
make demo-swe-mini evaluator in
adk-agent-playground, to demonstrate
the framework's range on code-modification and to exercise the CaMeL filesystem-capability gate.
A second harder config… See the full description on the dataset page: https://huggingface.co/datasets/barissozudogru/swe-bench-mini.korean-assembly-minutes
대한민국 국회 회의록 아카이브
국회 회의록 원문(record.assembly.go.kr) PDF 에서 본문을 뽑아 모은 것이다.
본회의와 각 위원회 회의록이 모두 들어 있다.
수록 기간: 1948~1993
회의 수: 1,951건
본문 분량: 65,306,444자
구성
연도별 JSONL(gzip) 한 덩이다.
from datasets import load_dataset
ds = load_dataset("seoulraphaellee/korean-assembly-minutes", split="train")
ds = load_dataset("seoulraphaellee/korean-assembly-minutes", data_files="data/2026.jsonl.gz", split="train")
필드
이름
설명
meeting_key
회의 식별자 (record:<id>)… See the full description on the dataset page: https://huggingface.co/datasets/seoulraphaellee/korean-assembly-minutes.OctoCodingBench
OctoCodingBench: Instruction-Following Benchmark for Coding Agents
English | 中文
🌟 Overview
OctoCodingBench benchmarks scaffold-aware instruction following in repository-grounded agentic coding.
Why OctoCodingBench?
Existing benchmarks (SWE-bench, etc.) focus on task completion — whether the agent produces correct code. However, they miss a critical dimension: does the agent follow the rules while solving the task?
In real-world agentic coding, agents must… See the full description on the dataset page: https://huggingface.co/datasets/MiniMaxAI/OctoCodingBench.korean_data_scraper_aihub
Korean Data Scraper — AI-Hub Local Corpus
상태: 비공개 (private) 저장소입니다.
Korean_data_scraper 프로젝트의 aihub_local 소스가 생성한 코퍼스입니다. 국립정보화진흥원 AI-Hub(aihub.or.kr)에서 내려받은 여러 데이터셋의 로컬 zip 압축 파일을 압축 해제하고, 그 안의 json 파일에서 본문 텍스트를 추출한 결과입니다. 샘플로 확인한 내용 중에는 뉴스 기사(신문기사) 카테고리의 AI-Hub 데이터셋에서 추출된 텍스트가 포함되어 있습니다.
스키마
파일당 1개 레코드(JSONL)이며, 다음과 같은 필드를 가집니다.
{"id": "aihub_local:<zip 파일명>:<json 파일명>", "source": "aihub_local", "text": "...", "url": null, "license": "per-dataset -- check the… See the full description on the dataset page: https://huggingface.co/datasets/mkd-minju/korean_data_scraper_aihub.nemotron-nano-30b-miniswe-swebench-verified
Nemotron Nano 30B + mini-swe-agent SWE-bench Verified Trajectories
Agent trajectories from running NVIDIA Nemotron 3 Nano 30B A3B (MoE, 8B active params) on SWE-bench Verified using mini-swe-agent.
⚠️ Incomplete Run
This benchmark was terminated early due to poor performance. The model struggled with the agentic coding task.
Model Information
Attribute
Value
Model
NVIDIA Nemotron 3 Nano 30B A3B
Architecture
MoE (30B total, 8B active)
Serving
vLLM… See the full description on the dataset page: https://huggingface.co/datasets/pankajmathur/nemotron-nano-30b-miniswe-swebench-verified.NoRA
NoRA
Evaluating Grounded Reasonableness in Visual First-person Normative Action Reasoning
Paper |
Code and model interface
NoRA evaluates the actions a model proposes, the facts it observes, and the
reasons connecting them in first-person scenes.
Dataset
Split
Annotations
Clips
train
LLMSilver: model-annotated
1,230
test
HumanGold: human-reviewed
190
from datasets import load_dataset
ds = load_dataset(
"MINTLABJHUANU/NoRA"… See the full description on the dataset page: https://huggingface.co/datasets/MINTLABJHUANU/NoRA.PRO-STEP-PRM-Data
PRO-STEP: PRM Training Annotations
Paper: PRO-STEP: Step-level Process Reward Optimization for Retrieval-Augmented GenerationCode: https://github.com/keemminnke/PRO-Step
Step-level annotations used to train the PRO-STEP PRM.
Total step annotations: ~109K across 31,728 trajectories
Source questions: 2,000 (HotpotQA + MuSiQue training splits)
Generation: 16 sampled trajectories per question with Qwen2.5-7B-Instruct
Annotator: QwQ-32B (open-source reasoning model), prompted with… See the full description on the dataset page: https://huggingface.co/datasets/MinKeonKim/PRO-STEP-PRM-Data.MiniF2F
minif2f Dataset
The minif2f dataset is a collection of mathematical problems and their formal statements, designed for formal mathematics and theorem proving tasks.
Dataset Description
Dataset Summary
The minif2f dataset contains mathematical problems from various sources (like AMC competitions) along with their formal statements in the Lean theorem prover format. Each example includes both informal mathematical statements and their corresponding formal… See the full description on the dataset page: https://huggingface.co/datasets/Tonic/MiniF2F.PRO-STEP-Preference-Data
PRO-STEP: DPO Preference Pairs
Step-level preference pairs used to train the PRO-STEP policy model via Direct Preference Optimization.
Paper: PRO-STEP: Step-level Process Reward Optimization for Retrieval-Augmented GenerationCode: GitHub Repository
Pairs: 15,877 (after outcome filter)
Source questions: 5,000 from HotpotQA + MuSiQue + 2WikiMultiHopQA training splits
Generation: PRM-guided MCTS (K=3 branching, depth 7, 64 rollouts/question, V(s) = Q̄(s) + α · r̂(s) with α=0.3)… See the full description on the dataset page: https://huggingface.co/datasets/MinKeonKim/PRO-STEP-Preference-Data.AgentMercury-SWE-sample
AgentMercury — SWE construction sample
A 50-row public slice of the environment-construction code differences behind
AgentMercury: Your Agent Can Synthesize Verifiable Environments for Business Scenarios at Scale.
AgentMercury's claim is that a world is built, and that the build is itself learnable. That makes
environment construction a software-engineering task — a diff, against a tree, judged by tests.
These are that task, cut three ways.
One row is
Here
Full
S-C… See the full description on the dataset page: https://huggingface.co/datasets/Minbyul/AgentMercury-SWE-sample.MiniMax-M3-150k-Mixed
m3-alldomains-verified-107k
Verified distillation traces generated with faststill v0.0.1 — a pipeline that generates (prompt, reasoning, output) triplets from any OpenAI-compatible chat-completions endpoint and deterministically verifies every row before keeping it. A row is verified=true only when a machine check (executed unit tests, exact / normalized answer compare) confirmed it, so wrong labels are filtered out instead of poisoning a student model.
Dataset… See the full description on the dataset page: https://huggingface.co/datasets/empero-ai/MiniMax-M3-150k-Mixed.minecraft-question-answer-700k
minecraft-question-answer-700k
Introducing the largest synthetic Minecraft Q&A dataset, covering every topic, game mechanic, item and craft in Minecraft. The dataset was generated by extracting over 18,000 Minecraft wiki pages, and using glaive.ai's synthetic data generation pipeline.
about the dataset
rows - 694,814
tokens - 47,133,624
source - https://minecraft.wiki/
Hit me up on twitter if you see a bug or need a synthetic dataset for your company:… See the full description on the dataset page: https://huggingface.co/datasets/naklecha/minecraft-question-answer-700k.pii-masking-mini-10k
PII Masking Mini: Multilingual Sample
A mini-sized stratified sample of pii-masking-openpii-1.5m,
the flagship release of the PII-Masking-3M family. Sampled proportionally by
(source_dataset, language) so every locale and label gets representation.
Asia Pacific rows appear first.
📖 More information: www.ai4privacy.com/datasets/pii-masking-3m-asia-pacific
Dataset Details
Total Examples
Train
Validation
Labels
Languages
Regions
Annotations
Format… See the full description on the dataset page: https://huggingface.co/datasets/ai4privacy/pii-masking-mini-10k.mini-math23k-v1The mini-math23k-v1 dataset is composed of ~ 23,000 entries of data, from open datasets across the AI landscape, including:
TIGER-Lab/MathInstruct
Birchlabs/openai-prm800k-solutions-only
Credits:
Birchlabs
@article{yue2023mammoth,
title={MAmmoTH: Building Math Generalist Models through Hybrid Instruction Tuning},
author={Xiang Yue, Xingwei Qu, Ge Zhang, Yao Fu, Wenhao Huang, Huan Sun, Yu Su, Wenhu Chen},
journal={arXiv preprint arXiv:2309.05653},
year={2023}
}
OpenBioRQ
OpenBioRQ — Open Biomedical Research Questions
📄 Paper: OpenBioRQ: Unsolved Biomedical Research Questions for Agents (Minbyul Jeong, 2026) ·
🤗 Dataset: Minbyul/OpenBioRQ
OpenBioRQ is a benchmark of open-ended, currently-unresolved biomedical research
questions extracted from the primary literature and clinical-trial records, refined
to be self-contained, and graded by per-question rubrics. It is built for agentic
evaluation: a model is given a question, may use retrieval/MCP… See the full description on the dataset page: https://huggingface.co/datasets/Minbyul/OpenBioRQ.AgentMercury-corpus-sample
AgentMercury — corpus sample
A small, public slice of the RL training corpus used in
AgentMercury: Your Agent Can Synthesize Verifiable Environments for Business Scenarios at Scale.
AgentMercury synthesizes executable worlds from high-level business scenarios and then
instantiates tasks on top of them, rather than building an environment around a predefined task.
Each task is an autonomous MCP tool-use investigation: a persona works inside a synthetic
company's systems (CRM… See the full description on the dataset page: https://huggingface.co/datasets/Minbyul/AgentMercury-corpus-sample.fuzi-mingcha-v1_0-pretrain-data
🐱 Github Repo
模型 huggingface 链接：https://huggingface.co/datasets/SDUIRLab/fuzi-mingcha-v1_0
数据 huggingface 链接：https://huggingface.co/datasets/SDUIRLab/fuzi-mingcha-v1_0-data
GitHub 链接：https://github.com/irlab-sdu/fuzi.mingcha
数据 魔搭链接：https://www.modelscope.cn/datasets/furyton/fuzi-mingcha-v1_0-data
模型 魔搭链接：https://www.modelscope.cn/models/furyton/fuzi-mingcha-v1_0
夫子•明察司法大模型预训练数据归档
统计信息
wenshu
来源：裁判文书网
处理方式：将被告人、原告、时间地点等信息替换为[被告人]、[原告]、[A]等。替换词均被 []… See the full description on the dataset page: https://huggingface.co/datasets/SDUIRLab/fuzi-mingcha-v1_0-pretrain-data.swebench-minimax-traces
swebench-minimax-traces
Complete HTTP-level agentic traces from running swebench benchmark tasks through an instrumented reverse proxy.
Each trace captures full request/response pairs including system prompts, user messages, assistant responses, tool calls and results, and token usage metadata.
Stats
Total sessions: 500
Multi-turn sessions (2+ LLM calls): 357
Total records: 22424
Total LLM requests: 11249
Format
Raw JSONL traces from the instrumented proxy.… See the full description on the dataset page: https://huggingface.co/datasets/sammshen/swebench-minimax-traces.Vietnam-History-1M-ViLearn more on GitHub: https://github.com/MinhxThanh/Vietnam-History-Chat-Datasets
Thông số chính
Tổng số mẫu: 1,000,000
Tỷ lệ có reasoning (analysis): ~77.99%
Tỷ lệ chỉ trả lời (final-only): ~22.01%
Định dạng: messages theo ShareGPT/ChatML
Mẫu có reasoning: system → user → assistant (analysis) → assistant (final)
Mẫu final-only: system → user → assistant (final)
intercode-minimax-traces
intercode-minimax-traces
Complete HTTP-level agentic traces from running intercode benchmark tasks through an instrumented reverse proxy.
Each trace captures full request/response pairs including system prompts, user messages, assistant responses, tool calls and results, and token usage metadata.
Stats
Total sessions: 338
Multi-turn sessions (2+ LLM calls): 299
Total records: 5838
Total LLM requests: 2919
Format
Raw JSONL traces from the instrumented proxy.… See the full description on the dataset page: https://huggingface.co/datasets/sammshen/intercode-minimax-traces.minicpm5-stock-v2-forward-return
MiniCPM5 Stock v2 — Forward-Return Labels
Binary BUY/SELL stock-direction dataset where labels come from actual forward
5-day returns (BUY > +2%, SELL < -2%, middle band dropped), not news sentiment.
All features are strictly causal (no look-ahead): last 20 daily returns, RSI(14),
volume ratio vs 20d MA, 20d volatility, 5d/20d momentum, 20d relative strength vs SPY.
train_minicpm5_v2.jsonl — 5,056 rows, 16 tickers, class-balanced
val_minicpm5_v2.jsonl — 1,586 rows, 4 held-out… See the full description on the dataset page: https://huggingface.co/datasets/ewin-reg/minicpm5-stock-v2-forward-return.TSD-KD-Qwen2.5-1.5B-Instruct-Gen
TSD-KD-Qwen2.5-1.5B-Instruct-Gen
This dataset contains student-generated examples used for Token-Selective Dual Knowledge Distillation (TSD-KD), introduced in our ICLR 2026 paper:
"Explain in Your Own Words: Improving Reasoning via Token-Selective Dual Knowledge Distillation"
Paper: https://arxiv.org/abs/2603.13260
Github: https://github.com/kmswin1/TSD-KD
Dataset Description
This dataset contains student-generated instruction-response examples from… See the full description on the dataset page: https://huggingface.co/datasets/Minsang/TSD-KD-Qwen2.5-1.5B-Instruct-Gen.keural-v2-dataset
Keural v2 — MoE Fine-Tuning Dataset
상태: 비공개 (private) — 공개 배포 대상 아님
Keural MoE Pro v2 모델 파인튜닝을 위해 8개 카테고리(A~H)로 구성된 SFT 학습 데이터셋입니다. 자세한 수집·처리 기준은 mkd-minju/Keural-MoE-Pro-v2 GitHub 저장소의 docs/V2-DATASET-PREP.md 계획 문서를 따릅니다.
카테고리 구성
파일
카테고리
목표 건수
언어
주요 출처
라이선스
A_korean_conversation.jsonl
한국어 대화/지침
50,000
ko
mkd-chanwoo/keural-conversation-chatml-ko, mkd-chanwoo/keural-rag-chatml-ko… See the full description on the dataset page: https://huggingface.co/datasets/mkd-minju/keural-v2-dataset.ja-current-news-keyword-sft-30k
Japanese Current-News Business Keyword SFT 30K
Synthetic Japanese SFT dataset for structured keyword generation.
Given one theme, the assistant returns JSON with:
categories: 3-6 upper-level categories
terms: 12-16 related terms
each term has label and cat
every cat exactly matches one item from categories
Themes are current-news oriented and cover topics such as LLMs, AI policy, AI agents, economics, markets, Trump-related policy, tariffs, monetary policy, geopolitics, and… See the full description on the dataset page: https://huggingface.co/datasets/minhhien0811/ja-current-news-keyword-sft-30k.minimind-fr-electronics-data
minimind-fr-electronics-data
sft_spec_electronics.jsonl (42,537) — conversations schema.
theprint/Electronics-QA + electronics.stackexchange.com (accepted
answers) + ~25% base-SFT replay; ~30% rows with a diagnostic <think>.
Built by scripts/convert_spec_electronics.py; see the
minimind-fr-electronics model card.
Built from
theprint/Electronics-QA
bshada/electronics.stackexchange.com
allenai/tulu-3-sft-mixture
jpacifico/French-Alpaca-dataset-Instruct-110K… See the full description on the dataset page: https://huggingface.co/datasets/yassinsiouda/minimind-fr-electronics-data.deplyze-mini-dataset
Deplyze-Mini Dependency Intelligence Benchmark Dataset
This dataset contains standardized, ground-truth scenarios for training and evaluating software dependency intelligence models. It is designed to evaluate and prevent vulnerability hallucinations, train/test leakage, and prompt injection vulnerabilities in automated software composition analysis (SCA).
Dataset Composition
train.json: 400 multi-category dependency scenarios with instruction-tuning message… See the full description on the dataset page: https://huggingface.co/datasets/mosetireagan/deplyze-mini-dataset.BlockData-minecraft-10k
Dataset Card for Dataset Name
Minecraft dataset features user-AI interactions, providing gameplay advice and strategies.
Dataset Details
Dataset Description
The Minecraft dataset on Hugging Face consists of 6,390 rows of interactions between users and an AI assistant designed to provide expert advice on Minecraft. It includes questions about gameplay strategies, such as efficient storage options, diamond farming tips, and mining improvements. The assistant… See the full description on the dataset page: https://huggingface.co/datasets/FalconNet/BlockData-minecraft-10k.