Skip to content

Datasets

Hugging Face datasets with plain-language descriptions: what's inside and what models to train on them. 4,788 in the catalog so far, with new ones added every hour.

Data type

For which models

Find a dataset for your goal

Answered by Gemini 4 · the chat is kept only in this tab

DatasetFreedomIntelligence/huatuo_encyclopedia_qa

Medical QAs from Chinese encyclopedia

364,420 Chinese medical question-answer pairs for training and evaluating medical LLMs.

  • Text generation
  • Chinese
1.1K 93100K–1M rows
Datasetnebius/Longtail-SWE

Coding tasks benchmark in 13 rare languages

471 real GitHub tasks for evaluating AI agents in writing software patches in less popular languages.

  • Text generation
  • English
1.1K 5up to 1K rows
Datasetsaidutta69/fable-5-premium

SFT dataset from Claude Fable-5 traces

A cleaned English SFT dataset from Claude Fable-5 agent traces, 12,730 records.

  • Text generation
  • Token classification
  • English
1.1K 13610K–100K rows
Datasetterryoo/TableVQA-Bench

Table and image visual question answering benchmark

English dataset for evaluating models in Visual Question Answering and table comprehension tasks.

  • Table question answering
  • Visual question answering
  • English
1.1K 351K–10K rows
Datasetlordx64/agentic-distill-fable-5-sft

Agentic data with tool calls from Claude Fable-5

4,659 English single-turn examples from the closed Claude Fable-5 preview model for agent SFT training.

  • Text generation
  • English
1.1K 591K–10K rows
Datasetfaunix/Qwen3.8-27B-Distillation-40K

40k reasoning chains from Qwen3.8-27B

Distilled step-by-step reasoning from Qwen3.8-27B across four domains for training thinking models.

  • Text generation
  • English
1.1K 4310K–100K rows
Datasetdomyn/Domyn-Edge-Pretraining-Mix-8.4TTop pick

Domyn Edge pre-training mix of 8.44 trillion tokens

Huge multilingual corpus of web texts, code, scientific papers, and mathematics for the first LLM pre-training stage.

  • Text generation
  • English
  • Arabic
  • German
  • +9
1.1K 61B–10B rows
DatasetLiquidAI/ifstruct-v1.0

JSON/YAML structured output benchmark

1K-10K examples to test if a model follows a given JSON/YAML schema under various prompt formulations.

  • Text generation
  • English
1.1K 1141K–10K rows
Datasetibm-research/AssetOpsBench

Benchmark for industrial AI agents

A specialized benchmark by IBM Research for evaluating LLMs and multi-agent systems in industrial operations.

  • Question answering
  • Time series forecasting
  • English
1.1K 51up to 1K rows
DatasetSTEM-AI-mtl/Electrical-engineering

Electrical Engineering and KiCad Q&A

English question-answer pairs on electrical engineering, KiCad EDA, and Python scripts.

  • Question answering
  • Text generation
  • English
1.1K 611K–10K rows
Datasetnvidia/Nemotron-RL-litmus-bench-v0.2

Chemistry questions for RL training of models

Open dataset of 49k chemistry questions with short answers for training and evaluating models using RL.

  • Text generation
  • English
1.1K 3710K–100K rows
DatasetOpenDataArena/Spark-234K

234K Scientific QA Pairs for Article Reasoning

234k question-answer pairs for scientific reasoning, synthesized from articles via an argumentation skeleton.

  • Text generation
  • English
1.1K 113100K–1M rows
Datasetnvidia/Nemotron-RL-Science-v1Top pick

Dataset for RL training of scientific reasoning

Physics, biology, and chemistry problems from NVIDIA with verifier settings for reinforcement learning.

  • Text generation
  • English
1.1K 18100K–1M rows
Datasetteknium/openhermes

OpenHermes: 242k SFT examples

242k predominantly GPT-4 examples from open sources for instruction tuning.

  • Text generation
  • English
1.1K 245100K–1M rows
Datasetneural-bridge/rag-dataset-12000

12,000-Example RAG Dataset

An English dataset of 12,000 examples for training and evaluating retrieval-augmented generation models.

  • Question answering
  • English
1.1K 20110K–100K rows
DatasetBatsResearch/planetarium

Text and PDDL description pairs for planning tasks

An English corpus of 100K to 1M examples for translating textual conditions into PDDL planning language.

  • Translation
  • Question answering
  • English
1.1K 20100K–1M rows
Datasetfree-law/Caselaw_Access_Project

U.S. caselaw corpus

Digitized U.S. court decisions in English for legal text training and analysis.

  • Text generation
  • English
1.1K 1051M–10M rows🔒 on request
DatasetLocutusque/UltraTextbooks

Large Collection of Instructional Textbooks

A large collection of texts for training language models, including code.

  • Text generation
  • English
  • Code
1.1K 2091M–10M rows
DatasetMxode/Chinese-Instruct

Chinese instruction SFT corpus

1 million to 10 million Chinese instruction-response pairs for SFT, including DeepSeek-R1 distillation.

  • Text generation
  • Question answering
  • Chinese
1.1K 1571M–10M rows
Datasetllm-wizard/alpaca-gpt4-data-zh

Chinese instructions from GPT-4

10k–100k Chinese instruction-response pairs for text model fine-tuning; research use only.

  • Text generation
  • Chinese
1.1K 7610K–100K rows
Datasetbeomi/KoAlpaca-v1.1a

Korean instruction dataset

Set of 21,155 Korean Alpaca-style instructions and responses for training chat models.

  • Text generation
  • Korean
1K 6310K–100K rows