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

DatasetMedOtter/4D-Lung

4D lung CT segmentation dataset

Respiratory-synchronized CT scan dataset of lung cancer patients for segmentation.

  • Image segmentation
22.1K 2up to 1K rows
DatasetGrunCrow/BIRDeep_AudioAnnotations

Audio recordings and bird vocalization labels of Spain

A small dataset of bird song recordings from Doñana National Park for species detection and classification.

  • Audio classification
  • English
  • Spanish
22.1K 5up to 1K rows
Datasethuminclu/autoscirub-rcb-main-exp

ResearchClawBench scientific reports and metadata

Dataset contains ResearchClawBench experiment artifacts: scientific reports, charts, code, and English metadata.

  • English
22K 1up to 1K rows
Datasetmaelic/VG150-coco-format

Visual Genome 150 (VG150) in COCO-JSON

Popular scene graph generation benchmark with 150 object classes and 50 relation types in standard COCO format.

  • Object detection
  • English
22K 0100K–1M rows
Datasetmorgan/CADExpert_fixed

CADExpert dataset of CAD tasks

14,935 unique CAD tasks with images and three expert answers each.

  • Image → text
  • English
  • Chinese
21.9K 110K–100K rows
Datasetnvidia/PhysicalAI-Human-1M-100KTop pick

Egocentric video and hand poses for physical AI

An open portion of NVIDIA's Human-1M collection: first-person videos, head and hand poses with text descriptions.

  • English
21.9K 129up to 1K rows
DatasetHothan/OlympiadBench

Multimodal benchmark of Olympiad scientific problems

Complex Olympiad-level problems with images in English and Chinese for evaluating model scientific reasoning.

  • Question answering
  • Visual question answering
  • Chinese
  • English
21.8K 521K–10K rows
DatasetQCRI/CrisisMMD

Manually annotated disaster tweets and photos

Multimodal corpus (10K–100K) of posts and images regarding seven major natural disasters in 2017.

  • Image classification
  • English
21.8K 610K–100K rows
DatasetSalesforce/3d_optical_flow_droidTop pick

Salesforce robot trajectories with 3D scene flow

Processed DROID robotics dataset with optical and 3D scene flow annotations for training robot control models.

  • Robotics
21.7K 010M–100M rows
Datasetdronefreak/RainDrop

Raindrop removal dataset for camera lenses

Unofficial mirror of the RainDrop (CVPR 2018) dataset for training models to clean lens water drops.

  • Image → image
  • English
21.7K 01K–10K rows
Datasetbench-llm/or-bench

OR-Bench language model over-refusal benchmark

A prompt set to check the balance between LLM safety and their tendency to refuse harmless prompts.

  • Text generation
  • Question answering
  • English
21.7K 2610K–100K rows
Datasetcansa/Describable-Textures-Dataset-DTD

Texture image collection with property descriptions

Mirror of the well-known DTD dataset containing real-world texture photographs and annotations of their visual properties.

    21.7K 41K–10K rows
    Datasetred-oa/red-oaTop pick

    ReD-OA patent benchmark

    Design patent prior art search benchmark based on USPTO data.

      21.7K 0up to 1K rows
      Datasetimageomics/VLM4Bio

      VLM4Bio dataset for ornithology

      Metadata and scripts for downloading bird images for visual question-answering tasks.

      • Visual question answering
      • Zero shot image classification
      • Zero shot object detection
      • English
      21.6K 110K–100K rows
      DatasetSimulaMet/Kvasir-VQA-x1

      Medical dataset Kvasir-VQA-x1 for endoscopy

      Multimodal dataset of 159,549 question-answer pairs for benchmarking in gastrointestinal endoscopy.

      • Visual question answering
      • Image text → text
      • English
      21.6K 18100K–1M rows
      Datasetyingss/mixlora-eval-data

      MixLoRA multimodal eval dataset

      Multimodal evaluation dataset and MME benchmark from the MixLoRA paper (ACL 2024).

      • Visual question answering
      • Image classification
      • English
      21.6K 01K–10K rows
      Datasetphilippkolbe/cilp_assessment_subset

      CILP assessment subset for FiftyOne

      A 750-sample English dataset subset configured for FiftyOne.

      • Image classification
      • English
      21.5K 01K–10K rows
      DatasetPaddlePaddle/Real5-OmniDocBench

      Distorted document parsing benchmark

      Document image set with physical distortions to test OCR and parsing system robustness.

      • Image → text
      • Chinese
      • English
      21.5K 431K–10K rows