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

Datasetinitiacms/XLRS-Bench_visual_grounding_en

Visual Grounding Benchmark on Satellite Imagery

English dataset with Earth imagery and object annotations for evaluating multimodal models in visual grounding.

  • English
35.1K 610K–100K rows
Datasetespnet/yodas-granaryTop pick

YODAS-Granary European speech corpus for ASR and translation

A large multilingual audio dataset (10M-100M entries) with pseudo-labels for European speech recognition and translation.

  • Automatic speech recognition
  • Translation
  • Bulgarian
  • Czech
  • Danish
  • +9
35.1K 4310M–100M rows
Datasetmacabdul9/hle_text_only

Complex text-only academic benchmark

Text version of Humanity's Last Exam with 3,000 expert-level questions across sciences to evaluate advanced LLMs.

    35K 41K–10K rows
    Datasetnvidia/Nemotron-CC-v2.1Top pick

    Nemotron-CC v2.1 pretraining corpus

    Updated data for LLM pretraining: math, code, English Common Crawl, and synthetic data.

    • Text generation
    35K 1471B–10B rows🔒 on request
    DatasetHuggingFaceH4/ultrafeedback_binarized

    Preference pairs for chat models

    Processed UltraFeedback: 64k prompts with model answers and GPT-4 scores aggregated into preference pairs.

    • Text generation
    • English
    35K 363100K–1M rows
    Datasettiagozip/tiktokTop pick

    Massive TikTok video and comment archive

    A large, regularly updated scrape of over 140M videos and 17.9B comments since 2018.

    • Text classification
    34.9K 710B–100B rows🔒 on request
    Datasetnvidia/PhysicalAI-SimReady-Warehouse-01

    NVIDIA OpenUSD warehouse 3D assets for simulation

    A catalog of simulation-ready 3D warehouse objects and their preview images for robotics and computer vision.

    • Image segmentation
    • Robotics
    • Depth estimation
    • English
    34.7K 64up to 1K rows
    Datasettheusamaaslam/ghalib-roman-synthetic

    Roman Urdu and Urdu synthetic voice dataset for TTS

    Support bot voice messages in Urdu with text transcriptions and Latin transliterations in parquet format.

    • Text → speech
    • Urdu
    • English
    34.5K 010K–100K rows🔒 on request
    Datasetlukakuma/web-trajectories-preference-raw

    Raw web agent trajectories with images

    Large raw archive of successful and failed web agent actions in English and Korean for subsequent filtering.

    • Text generation
    • English
    • Korean
    34.4K 01M–10M rows🔒 on request
    Datasetgoogle-research-datasets/nq_open

    NQ-Open open-domain question answering benchmark

    Classic English benchmark for testing model fact-answering capability without source context.

    • Question answering
    • English
    34.4K 3810K–100K rows
    Datasetdisco-eth/WorldSpeechTop pick

    WorldSpeech: 65k hours of multilingual speech

    A multilingual ASR corpus with 65k hours of human transcripts for 127 language variants.

    • Automatic speech recognition
    • Text → speech
    • Audio classification
    • Afrikaans
    • Amharic
    • Arabic
    • +9
    34.1K 7710M–100M rows
    Datasetproject-numina/NuminaMath-1.5

    NuminaMath 1.5: Math with CoT

    Around 900k competitive math problems with Chain of Thought solutions.

    • Text generation
    • English
    34.1K 210100K–1M rows
    Datasetkapturecx/bolAIndia

    Multilingual Indian speech and dialect corpus

    Large-scale collection of audio recordings, texts, and documents in rare Indian languages and dialects for ASR systems.

    • Automatic speech recognition
    • Angika
    • Assamese
    • Awadhi
    • +9
    34.1K 710M–100M rows🔒 on request
    DatasetGAIR/LIMO

    LIMO Research Dataset

    A small English text dataset for the LIMO: Less is More for Reasoning research.

    • English
    34.1K 183up to 1K rows
    Datasetopenbmb/UltraData-MathTop pick

    290B-token math pre-training corpus

    290 billion tokens of English and Chinese math data across three tiers for LLM pre-training.

    • Text generation
    • English
    • Chinese
    34K 365100M–1B rows
    Datasetmlfoundations/datacomp_1b

    Billion-scale image-text pair metadata

    Metadata and image URLs with text descriptions exceeding one billion records for multimodal learning.

      33.8K 541B–10B rows
      Datasetbenjamin-paine/freesound-laion-640k

      Cleaned FreeSound audio dataset from LAION (630k+)

      Re-uploaded version of the LAION-Audio-630k audio dataset with text descriptions of sounds and verified licenses.

      • Audio → audio
      • Audio classification
      33.8K 14100K–1M rows
      DatasetTevatron/browsecomp-plus

      Deep-Research benchmark for deep search systems

      Test set of complex search queries and a fixed corpus of 100 thousand web documents for fair agent comparison.

      • Question answering
      33.7K 41up to 1K rows
      Datasetformospeech/klokah

      Klokah audio dataset

      Audio dataset with text transcripts for training speech recognition models.

        33.6K 0100K–1M rows🔒 on request
        Datasetlongvideobench/LongVideoBench

        Long-video understanding benchmark up to 1 hour for LMMs

        3,763 long captioned videos with questions to evaluate long-context multimodal models.

        • Multiple choice
        • Visual question answering
        • English
        33.6K 641K–10K rows🔒 on request