datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
kernelbench-hard-traces
KernelBench-Hard agent traces
Frontier coding agents writing optimized CUDA/Triton kernels (FP8 GEMM, paged
attention, MoE, W4A16, KDA, Top-k) on RTX PRO 6000 Blackwell, H100 PCIe, and
B200; roofline-graded.
Each .jsonl file is one agent run in Claude-Code session format, viewable with
the Hugging Face Agent Trace viewer (Data Studio → open a row). Filename =
run id.
Live leaderboard: https://kernelbench.com/hard
Secrets redacted. Full reasoning for open-provider routes… See the full description on the dataset page: https://huggingface.co/datasets/Infatoshi/kernelbench-hard-traces.kimi-cyber-reasoning
Kimi Cyber Reasoning
997 chain-of-thought records covering 13 cybersecurity disciplines and 4 systems engineering domains, distilled from the Kimi K3 reasoning model via API. Every record provides an explicit step-by-step <think> reasoning trace followed by a technical resolution, unified code diff fix, or structured tool invocation.
The dataset was curated as an anchor set for training, healing, and specializing compact reasoning models on systems security and tool calling… See the full description on the dataset page: https://huggingface.co/datasets/echel0nn1881/kimi-cyber-reasoning.KoMedInstruct-52k1차 번역 완료
I found several problems during translation, so additional filtering will be needed after completion.
Step 1. It was confirmed that the existing data contained a lot of data that was close to duplicates. Need to remove those.
Step 2. There are many outputs with a risk of hallucination. Data where the last sentence of output is incomplete must be edited.
Step 3. If data corresponding to the output is also included in the input, those must be edited.
Ko-AlpaCare
This is… See the full description on the dataset page: https://huggingface.co/datasets/ChuGyouk/KoMedInstruct-52k.personal-trainer-ausbildung-ki-datensatz
SNFA Personal Trainer Ausbildung KI-Datensatz
Ein deutschsprachiger Wissensdatensatz der SNF Academy zu Personal Training, Fitnessausbildung, Berufspraxis, Coaching, Selbstständigkeit und regionalen Angeboten in der Schweiz.
Inhalt
Die Datei snfa_personal_trainer_dataset.jsonl enthält thematisch abgegrenzte Abschnitte aus den Dokumenten dieses Repositorys. Jeder Datensatz besitzt eine eindeutige ID sowie Angaben zu Titel, Abschnitt, Inhalt, Kategorie, Quelldatei… See the full description on the dataset page: https://huggingface.co/datasets/snfacademy/personal-trainer-ausbildung-ki-datensatz.researchscope-papers
ResearchScope Papers
Open CS research paper dataset maintained by ResearchScope.
Updated automatically via GitHub Actions.
Quick start
from datasets import load_dataset
ds = load_dataset("kishormorol/researchscope-papers", "papers", split="train")
print(ds[0])
See Usage below for per-source splits, instruction-tuning, and the per-section fine-tuning data.
Stats
34,903 papers (raw metadata) — 9,903 arXiv · 20,000 conference · 5,000 journal
174,067… See the full description on the dataset page: https://huggingface.co/datasets/kishormorol/researchscope-papers.arxiv-tex-corpus-fullarxiv-tex-corpus-full (80GB)
Large-scale LaTeX corpus from arXiv (math, CS, physics, statistics)
📄 Paper: https://arxiv.org/abs/2602.17288
📚 Overview
arxiv-tex-corpus-full (80GB) is a large-scale dataset of LaTeX source content extracted from papers hosted on arXiv.
This version contains approximately 80GB of structured JSONL data, restricted to the following arXiv categories:
math
cs
hep-th
hep-ph
quant-ph
stat.ML
stat.TH
The dataset is designed for research in:
Large… See the full description on the dataset page: https://huggingface.co/datasets/KiteFishAI/arxiv-tex-corpus-full.KoEVD
KoEVD
KoEVD is a Korean benchmark linking five evaluation or analysis targets through source utterances: utterance-risk judgment, candidate-response safety choice, direct-generation response harmfulness, descriptive response strategies, and a pre-execution mock tool/action-choice diagnostic.
Contents and scope
The canonical corpus contains 13,552 sources and 71,395 response candidates: 30,740 accepted, 27,104 rejected, and 13,551 strongly rejected. Three… See the full description on the dataset page: https://huggingface.co/datasets/KETI-NLP/KoEVD.KoHRM-Text-1.4B-prepared-data
KoHRM-Text-1.4B Prepared Data
This dataset repository contains prepared HRM-Text V1Dataset artifacts for KoHRM-Text-1.4B.
The data is intended for continued pretraining and staged training with the project code at:
https://github.com/LLM-OS-Models/KoHRM-text
https://huggingface.co/LLM-OS-Models/KoHRM-Text-1.4B
https://huggingface.co/LLM-OS-Models/HRM-Text-Ko-Terminal-Tokenizer-131K
The upstream architecture and training method are based on:
Paper:… See the full description on the dataset page: https://huggingface.co/datasets/LLM-OS-Models/KoHRM-Text-1.4B-prepared-data.harbor-swesmith-rl-artifacts
Harbor SWE-Smith 强化学习数据产物
本数据集是 Harbor Qwen 工具调用代码智能体强化学习项目使用的冻结任务集，服务于 GRPO、原生价值模型/GAE PPO、训练过程诊断和统一协议评测。
项目已于 2026 年 8 月 30 日完成 P0 评测并进入阶段性归档。本数据集用于保留实验所依赖的数据切分、任务执行文件和审计信息，不代表新的通用代码能力基准。
数据概况
切分
任务数
训练集
187
验证集
42
测试集
38
合计
267
数据覆盖 89 个上游代码仓库。三个切分之间同时执行任务标识和仓库级隔离检查。
正式数据集名称：
swesmith-curated-grpo-267-v1
冻结切分的语义摘要：
ae5df9a3f4a3fc8af44fac420b36529e283839e1bd3de9daba65d5bcda51447d
该值来自 split-manifest.json 的 sha256 字段，用于标识切分语义，不等同于该文件本身的字节级… See the full description on the dataset page: https://huggingface.co/datasets/keryszhan/harbor-swesmith-rl-artifacts.urls-tokenized
URLs (tokenized)
ks46/urls-sampled run through a byte-level
BPE built for URLs, stored as flat uint16 token streams that memory-map
directly into a training loop.
Shards
512
URLs
18,729,786,698
Tokens
664,731,047,208
Vocabulary
8,192
Token dtype
uint16, little-endian
There is no parquet here and the dataset viewer will not render it. These
are raw token bins; see Reading the data below.
Layout
tokenizer/ the exact vocabulary… See the full description on the dataset page: https://huggingface.co/datasets/ks46/urls-tokenized.Kimi-K2.5-Reasoning-1M-Cleaned
🪐 Kimi-K2.5-Reasoning-1M-Cleaned
Kimi-K2.5-Reasoning-1M-Cleaned is a cleaned derivative of ianncity/KIMI-K2.5-1000000x. It preserves the original four-config layout from the source dataset and rewrites each record into a unified reasoning-SFT schema with id, conversations, input, output, domain, and meta.
Summary
Source dataset: ianncity/KIMI-K2.5-1000000x
Source author: ianncity
Teacher model recorded in meta.teacher_model: KIMI-K2.5
Token lengths computed with… See the full description on the dataset page: https://huggingface.co/datasets/Jackrong/Kimi-K2.5-Reasoning-1M-Cleaned.datakernelbench
DataKernelBench
Can LLMs optimize database queries on GPUs?
DataKernelBench evaluates LLMs on a novel task: optimizing analytical database queries as GPU kernels. It first represents each SQL query as a validated PyTorch program called a TorchPlan. It then evaluates LLMs by asking them to optimize either the tensor-intensive core (core) or the full query implementation (full) using CUDA or Triton, with execution-guided repair. The benchmark covers all 22 TPC-H queries.
On TPC-H… See the full description on the dataset page: https://huggingface.co/datasets/kerneldf/datakernelbench.General-Knowledge
Dataset Card for Dataset Name
Dataset Summary
The dataset is a collection of questions and answers themed on general facts and reasoning. The dataset is divided into two features - 'Question' and 'Answer'.
It is meant to be used for training a model to be good at general knowledge and reasoning. This dataset is inspired from the Alpaca dataset, and infact contains a subset of the alpaca dataset in itself.
Distribution
The distribution of the… See the full description on the dataset page: https://huggingface.co/datasets/MuskumPillerum/General-Knowledge.KIMI-K2.5-1000000x
KIMI-K2.5-1000000x
1,000,000 reasoning traces distilled from KIMI-K2.5 on high reasoning, (Each subset has different questions)
Distribution:
Coding: 50% (Includes: Webdev, Python, C++, Java, JS, C, Ruby, Lua, Rust, and C#)
Science: 20% (Physics, Chemistry, Biology) - 100k more completions in the PHD-Science subset
Math: 15% (Algebra, Calculus, Probability) - 200k more completions in kimiMath200k.jsonl
Computer Science: 5%
Logical Questions: 5%
Creative Writing: 5%… See the full description on the dataset page: https://huggingface.co/datasets/ianncity/KIMI-K2.5-1000000x.snfa-ernaehrungscoach-ki-dataset
SNFA Ernährungscoach KI-Datensatz
Ein deutschsprachiger Wissensdatensatz der SNF Academy zu Ernährungscoaching, Online-Ausbildung, Beratung, Berufspraxis und Selbstständigkeit in der Schweiz.
Inhalt
Die Datei snfa_ernaehrungscoach_dataset.jsonl enthält thematisch abgegrenzte Abschnitte aus den Dokumenten dieses Repositorys. Jeder Datensatz besitzt eine eindeutige ID sowie Angaben zu Titel, Abschnitt, Inhalt, Kategorie, Quelldatei und Quelle.
Die ursprünglichen… See the full description on the dataset page: https://huggingface.co/datasets/snfacademy/snfa-ernaehrungscoach-ki-dataset.reasoning-base-20k
Dataset Card for Reasoning Base 20k
Dataset Details
Dataset Description
This dataset is designed to train a reasoning model. That can think through complex problems before providing a response, similar to how a human would. The dataset includes a wide range of problems from various domains (science, coding, math, etc.), each with a detailed chain of thought (COT) and the correct answer. The goal is to enable the model to learn and refine its reasoning process… See the full description on the dataset page: https://huggingface.co/datasets/KingNish/reasoning-base-20k.GitHub-CC0
Public Domain GitHub Repositories Dataset
This dataset contains metadata and source code of 9,000 public domain (cc0 or unlicense) licensed GitHub repositories that have more than 25 stars.
The dataset was created by scraping the GitHub API and downloading the repositories, so long as they are under 100mb.
The dataset can be used for various natural language processing and software engineering tasks, such as code summarization, code generation, code search, code analysis, etc.… See the full description on the dataset page: https://huggingface.co/datasets/KoalaAI/GitHub-CC0.kaidol-character-dataset
KAIdol Character Chat Dataset
한국어 캐릭터 롤플레이 대화 데이터셋
📋 목차
개요
데이터셋 통계
데이터 형식
캐릭터 목록
품질 지표
사용 방법
학습 가이드
제한사항
라이선스
🎯 개요
KAIdol Character Chat Dataset은 41개 고유 캐릭터의 롤플레이 대화 데이터셋입니다. 각 캐릭터는 독특한 **음성 프로필(Voice Profile)**을 가지고 있으며, 이를 기반으로 일관된 성격과 말투를 유지합니다.
주요 특징
특징
설명
🎭 41개 캐릭터
다양한 성격, 배경, 말투를 가진 캐릭터
🗣️ 음성 프로필
시그니처 표현, 종결어미, 금지 표현 정의
📊 3가지 형식
SFT, DPO, Multiturn 학습 지원
✅ 품질 검증
A등급 음성 프로필 일치율 (0.805)
🇰🇷 100% 한국어
자연스러운… See the full description on the dataset page: https://huggingface.co/datasets/developer-lunark/kaidol-character-dataset.kullm-v2
Dataset Card for "KULLM-v2"
Dataset Summary
Korean translation of GPT4ALL, Dolly, and Vicuna data.
repository: nlpai-lab/KULLM
huggingface: nlpai-lab/kullm-v2
Translate dataset
Translated 'instruction', 'input', and 'output' in the dataset via the DeepL API
Lisence
Apache-2.0
>>> from datasets import load_dataset
>>> ds = load_dataset("nlpai-lab/kullm-v2", split="train")
>>> ds
DatasetDict({
train: Dataset({
features: ['id'… See the full description on the dataset page: https://huggingface.co/datasets/nlpai-lab/kullm-v2.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.huatuo_knowledge_graph_qa
Dataset Card for Huatuo_knowledge_graph_qa
Dataset Summary
We built this QA dataset based on the medical knowledge map, with a total of 798,444 pieces of data, in which the questions are constructed by means of templates, and the answers are the contents of the entries in the knowledge map.
Dataset Creation
Source Data… See the full description on the dataset page: https://huggingface.co/datasets/FreedomIntelligence/huatuo_knowledge_graph_qa.TvTroper-2025
TvTroper-2025
A cleaned & refreshed dump of ~708 k pages from tvtropes.org
Dataset Summary
TvTroper-2025 is an updated snapshot of TvTropes.org (≈ 708 000 wiki pages, namespaces and date-grouped pages excluded).
Every page is released in two flavours:
Raw HTML – 22 GB single file
Markdown-cleaned – split into 1 GB JSONL shards (no unpacking required)
No additional content filtering has been applied; short sub-index pages are left in so you can decide what to drop.… See the full description on the dataset page: https://huggingface.co/datasets/KaraKaraWitch/TvTroper-2025.secret-loyalty-competition-data
Secret-loyalty organisms — training banks and eval batteries
The data behind KKing23/secret-loyalty-competition-organisms.
Code and full result trail: github.com/kaustubhkislay/secret-loyalty-competition.
Why this exists separately from the adapters. The adapters are reproducible from these
banks for the price of GPU time. These banks are not reproducible — they were written by
an LLM generator, so regenerating gives different data and every published number becomes… See the full description on the dataset page: https://huggingface.co/datasets/KKing23/secret-loyalty-competition-data.KarantaOCR-Bench
KarantaOCR - Bench
KarantaOCR-Bench is a unit-test–style evaluation dataset, similar to olmOCR-bench. It consists of 70 PDF documents and 300 test cases spanning multiple document types.
All the tests were manually verified by us.
KarantaOCR-Bench is designed specifically to evaluate document text extraction for Documents with diacritics and special characters, covering a diverse range of document formats and languages commonly under-represented in existing OCR benchmarks.… See the full description on the dataset page: https://huggingface.co/datasets/taresco/KarantaOCR-Bench.joyo-kanji-yomi-benchmark-parakeet
日本語 | English
常用漢字読みベンチマーク Parakeet Edition (JKYB-Parakeet)
常用漢字読みベンチマーク Parakeet Edition（JKYB-Parakeet）は、G2Pモデルや形態素解析器、TTSシステムが日本語の文章中の漢字を正しく読めているかを評価するためのベンチマークです。評価に用いるデータセットと評価ツールから構成されます。
このページでは、JKYB-Parakeetのデータセットを公開しています。評価ツールはGitHubで公開しています。
このデータセットは、SB Intuitionsによるデータセットsbintuitions/joyo-kanji-yomi-benchmarkをもとに、Parakeet株式会社が内容の検証を行い、誤りの修正、表記の統一、およびデータの追加等を独自に行ったものです。
概要… See the full description on the dataset page: https://huggingface.co/datasets/Parakeet-Inc/joyo-kanji-yomi-benchmark-parakeet.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.Creative-Writing-KimiK2.5-Cleaned
Creative-Writing-KimiK2.5-Cleaned
Cleaned creative writing SFT dataset from Kimi K2.5 (655 samples). Prompts cleaned, thinking traces preserved.
Format
Each line is a JSON object with:
messages: list of message dicts with roles (system, user, assistant)
System: writing quality instructions
User: cleaned creative writing prompt
Assistant: creative writing response (may include <think> traces)
Stats
Metric
Value
Total prompt tokens
80… See the full description on the dataset page: https://huggingface.co/datasets/Crownelius/Creative-Writing-KimiK2.5-Cleaned.guiowl-curated-corpus
GUI-Owl Curated Corpus
This dataset publishes the full curated mobile GUI-agent supervised fine-tuning corpus in a unified norm1000 mobile_use action format. Each row pairs a mobile UI screenshot with an instruction and a normalized target tool call for training GUI agents.
The published files are the curated parquet shards as produced by the source canonicalizers. No parquet shards are merged, re-sharded, or sampled during upload.
Sources
Source
Episodes… See the full description on the dataset page: https://huggingface.co/datasets/KMK040412/guiowl-curated-corpus.Fable-5-traces
Glint Research Dataset Card
Fable 5 Pi Agent Traces
A compact, high-signal corpus of Fable 5 coding-agent traces converted into Hugging Face Agent Traces / Pi-compatible sessions for Data Studio inspection, tool-use policy learning, and reasoning/action distillation.
Primary Config
pi_agent/train
Agent Trace preview enabled
4,665 Pi trace sessions
60 source sessions
3,799 tool… See the full description on the dataset page: https://huggingface.co/datasets/kira/Fable-5-traces.c4-ko-cleaned-2이전 데이터셋에서 아쉬운 점이 많이 보여 조금 개선한 데이터셋 입니다.
원본 데이터셋: c4
파일 크기: 약 10gb
데이터 수: 2261464