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

DatasetJetBrains-Research/commit-chronicle

10.7M GitHub commits for description generation

A large code change history corpus across 20 programming languages for generating and supplementing commit messages.

  • Text generation
  • Summarization
  • Code
  • English
3.9K 1410M–100M rows
DatasetTIGER-Lab/VisualWebInstruct

Multimodal instructions for VLM reasoning

Approximately 900K question-answer pairs (40% image-based, 60% text-based) for boosting multimodal model reasoning logic.

  • Question answering
  • Visual question answering
  • Image text → text
  • English
3.9K 451M–10M rows
DatasetVISAI-AI/nitibench-plus

Thai legal question and answer benchmark

Questions and answers based on real Thai court rulings with reasoning and legal references.

  • Question answering
  • Thai
3.8K 21K–10K rows🔒 on request
DatasetOpenStellarTeam/Chinese-SimpleQA

Chinese factual accuracy benchmark for LLMs

First comprehensive Chinese-language benchmark for evaluating model factual knowledge using short-form questions across 99 subtopics.

  • Question answering
  • Chinese
3.8K 381K–10K rows
DatasetPleIAs/French-PD-Books

French public domain book corpus

A large collection of French public domain monographs for training generative language models.

  • Text generation
  • French
3.7K 52100K–1M rows
Datasetbevaya/ScreenSpot

Text-based UI element grounding benchmark

Over 1,200 instructions and screenshots from iOS, Android, macOS, Windows, and the web for evaluating multimodal models.

  • Text generation
  • Image → text
  • English
3.7K 531K–10K rows
DatasetTrustAIRLab/in-the-wild-jailbreak-promptsTop pick

In-the-wild jailbreak prompts dataset

15,140 jailbreak prompts collected from December 2022 to December 2023.

  • Text generation
3.7K 7110K–100K rows
Datasetactava/chi-benchTop pick

US healthcare workflow AI agent benchmark

Test tasks for evaluating autonomous agents in healthcare scenarios: from treatment authorization to patient management.

  • Text generation
  • English
3.7K 66up to 1K rows
Datasetcoral-nlp/german-commonsTop pick

154.56B Tokens of Open German Text

A large corpus of openly licensed German texts from 41 sources for training language models.

  • Text generation
  • German
3.6K 4210M–100M rows
Datasetgretelai/synthetic_pii_finance_multilingual

Synthetic financial documents with PII

Synthetic financial documents in six languages with personal data markup for NER training.

  • Text classification
  • Fill mask
  • Token classification
  • English
  • French
  • German
  • +3
3.6K 8310K–100K rows
DatasetJeanKaddour/minipile

MiniPile: A compact Pile subset

6 GB of cleaned English text from The Pile for data-efficient model training.

  • Text generation
  • Fill mask
  • English
3.6K 1551M–10M rows
DatasetSLPL/naabTop pick

Largest cleaned Farsi text corpus

A massive Persian text corpus (130 GB, 15 billion words) for training language models from scratch and fine-tuning.

  • Fill mask
  • Text generation
  • Persian
3.6K 4510M–100M rows
Datasetosunlp/TravelPlanner

Travel planning benchmark

1,225 queries for evaluating language agents in multi-step trip planning with constraints.

  • Text generation
  • English
3.6K 891K–10K rows
Datasetmedalpaca/medical_meadow_medical_flashcards

Medical Anki question-answer flashcards

English question-answer pairs on basic and clinical medicine collected by medical students.

  • Question answering
  • English
3.6K 5110K–100K rows
DatasetAgentPublic/open_governmentTop pick

Open government data corpus with 380 billion tokens

Massive text and table corpus from open government archives of the USA, France, and EU for language model pre-training.

  • Text generation
  • English
  • French
  • Bulgarian
  • +9
3.6K 710M–100M rows
DatasetPleIAs/French-PD-Newspapers

French public domain newspapers

Corpus of French public domain newspapers and periodicals for language models.

  • Text generation
  • French
3.5K 741M–10M rows
DatasetMedRAG/textbooks

Medical textbook chunks for RAG systems

An English knowledge base of 18 medical textbooks chunked for search and question answering.

  • Question answering
  • English
3.5K 64100K–1M rows
DatasetTrendyol/Trendyol-Cybersecurity-Instruction-Tuning-Dataset

53k cybersecurity instruction examples

53,202 system/user/assistant cybersecurity examples across 200+ domains for training defensive assistants.

  • Text generation
  • Question answering
  • English
3.5K 14910K–100K rows
Datasetmaritaca-ai/enem

Brazilian ENEM 2022-2024 exam questions

Multiple-choice tasks from the Brazilian university entrance exam for evaluating LLMs and VLMs.

  • Visual question answering
  • Multiple choice
  • Portuguese
3.5K 18up to 1K rows