Skip to content

Datasets

Hugging Face datasets with plain-language descriptions: what's inside and what models to train on them. 4,797 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

Datasetlouisbrulenaudet/clinical-trials

Clinical trials database with embeddings

Records from ClinicalTrials.gov with metadata and pre-computed embeddings for biomedical tasks.

  • Question answering
  • Embeddings
  • Text classification
  • English
  • French
  • Spanish
675 42100K–1M rows
Datasetnvidia/Nemotron-Personas-Belgium

Multilingual profiles of Belgian residents from NVIDIA

Between 1 and 10 million synthetic personas of Belgian residents in four languages, mapped to real statistical distributions.

  • Text generation
  • Dutch
  • French
  • German
  • +1
675 331M–10M rows
DatasetOpenGVLab/VideoChat2-IT

Multimodal video and image instruction annotations

1.9M text annotations in JSON format for training models to answer questions about images and videos.

  • Visual question answering
  • Question answering
  • English
674 521M–10M rows🔒 on request
DatasetFreedomIntelligence/Huatuo26M-Lite

Chinese medical dataset Huatuo26M-Lite

Cleaned version of Huatuo26M for training Chinese medical models.

  • Text classification
  • Question answering
  • Text generation
  • Chinese
671 72100K–1M rows
DatasetJackrong/Claude-opus-4.7-TraceInversion-5000x

Reconstructed Claude Opus CoT examples

5k reconstructed chain-of-thought examples from Claude 4.7-Max for SFT and DPO across multiple languages.

  • Text generation
  • English
  • Chinese
  • Korean
  • +3
670 881K–10K rows
Datasetargilla/databricks-dolly-15k-curated-multilingual

Curated multilingual Dolly instructions

An edited version of Databricks Dolly instructions in Spanish, German, and French.

  • Text generation
  • Spanish
  • German
  • French
668 5510K–100K rows
Datasetnlpai-lab/kullm-v2

KULLM-v2 Korean instructions

GPT4ALL, Dolly, and Vicuna translations into Korean for instruction-based model training.

  • Text generation
  • Korean
664 77100K–1M rows
Datasetyuntian-deng/WildChat-4.8M-FullTop pick

Large-scale user AI dialogue corpus

Collection of 4.74 million real conversations for training and analyzing dialogue models and QA systems.

  • Text generation
  • Question answering
662 191M–10M rows🔒 on request
DatasetLiquidAI/phone-assistant-tools

Phone assistant tool calling dataset

Smartphone assistant prompts in 16 languages with reference calls from a catalog of 108 functions and evaluation rules.

  • Text generation
  • English
  • Arabic
  • Chinese
  • +9
660 101K–10K rows
Datasetzai-org/CC-Bench-trajectories

74 tasks for agentic coding

Claude Code agentic coding trajectories for the CC-Bench-V1.1 benchmark across 74 tasks.

  • Text generation
  • English
  • Chinese
648 98up to 1K rows
Datasetlingshu-medical-mllm/ReasonMedTop pick

Medical reasoning with CoT

370k question-answer pairs with step-by-step CoT rationales and concise summaries for medical tasks.

  • Question answering
  • Text generation
  • English
642 961M–10M rows
Datasetgoogle/polaris-benchTop pick

Visual reasoning benchmark in polar coordinates

Official benchmark from Google DeepMind and Stanford for evaluating MLLM visual reasoning in polar coordinates.

  • Visual question answering
  • Multiple choice
  • English
627 710K–100K rows
DatasetDmitriyYurckML/stem-reasoning-complex

STEM tasks with chains of thought

118k scientific STEM tasks with step-by-step CoT reasoning for SFT and alignment.

  • Text generation
  • Question answering
  • English
  • Chinese
622 80100K–1M rows
Datasetclzoro/GLM-5.1-1000000xTop pick

One million reasoning chains distilled from GLM-5.1

Large corpus of 1M examples featuring reasoning chains from GLM-5.1 in English and Chinese.

  • Text generation
  • Question answering
  • English
  • Chinese
614 48100K–1M rows
Datasetoi-uae/cyber-security

Cybersecurity chat dataset from 198 sources

A cleaned corpus of 1 million to 10 million English dialogues on infosec for language model SFT training.

  • Question answering
  • Text generation
  • English
613 461M–10M rows🔒 on request
DatasetDr-AliGomaa/ar-eg-dataset

Egyptian Arabic speech corpus

An audio dataset of Egyptian Arabic lectures with transcripts for speech recognition.

  • Automatic speech recognition
  • Translation
  • Arabic
605 21K–10K rows
DatasetTucanoBR/GigaVerbo

780 GB of Portuguese language texts

A summarized text corpus of 200 billion tokens for training Portuguese language models.

  • Text generation
  • Portuguese
598 42100M–1B rows
Datasetrecursal/Fanatic-Fandom

Fanatic Fandom cleaned wiki corpus

Large English text corpus of around 7.4B tokens sourced from fandom wikis.

  • Text generation
  • Fill mask
  • English
591 81M–10M rows
DatasetLitap007/indic-citizen-information-board

Multilingual Indian info board dataset

Multimodal text data in Indian languages for infrastructure auditing using VLM.

  • Visual question answering
  • Image → text
  • Table question answering
  • Hindi
  • Marathi
  • Telugu
  • +3
586 31K–10K rows
Datasetbh2821/LightNovel5000

Chinese light novel corpus in TXT and PDF

Around 1,500 Chinese-translated light novels in text format and PDFs with illustrations for creative text generation.

  • Text generation
  • Translation
  • Chinese
  • English
583 45up to 1K rows
DatasetMiniMaxAI/role-play-bench

Role-play Benchmark for agents

Benchmark for evaluating role-playing agents in Chinese and English scenarios.

  • Text generation
  • Chinese
  • English
582 1601K–10K rows
DatasetFreedomIntelligence/huatuo_knowledge_graph_qa

Chinese medical question-answer pairs

Nearly 800 thousand Chinese question-answer pairs collected from a medical knowledge graph.

  • Text generation
  • Chinese
571 53100K–1M rows