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

DatasetSingleBicycle/4KLSDBTop pick

4KLSDB dataset for restoration and generation

A native 4K image dataset for restoration and text-to-image generation tasks.

  • Image → image
  • Text → image
  • English
24.8K 18100K–1M rows
DatasetDVijayan/vilhyra-vision-dataset

VISION dataset mirror for image classification

A mirror repository of the VISION dataset for photo and video source identification.

  • Image classification
24.7K 010K–100K rows
DatasetIGNF/PASTIS-HD

French agricultural satellite image time series

Benchmark for panoptic and semantic segmentation of agricultural parcels using satellite data.

  • Image classification
  • Image segmentation
24.5K 201K–10K rows
Datasettanganke/stanford_cars

Stanford Cars with Robustness Augmentations

Classic car image dataset augmented with distorted test versions to evaluate model reliability.

  • Image classification
  • English
24.5K 3410K–100K rows
DatasetDabbyOWL/PDE_Inverse_Problem_Benchmarking

Benchmark for PDE inverse problems

Official dataset for studying neural networks in solving inverse problems for partial differential equations.

  • Other
24.3K 3100M–1B rows
DatasetOneThink/OneThinker-train-dataTop pick

OneThinker-600k multimodal dataset

Large-scale multimodal data corpus for training a universal reasoning model.

  • Image text → text
  • Video text → text
  • Object detection
  • English
24.3K 1410K–100K rows
Datasetarsaporta/symile-m3

Symile-M3 multilingual audio-visual-text dataset

Multilingual dataset of audio, images, and text for evaluating multimodal models.

  • Zero shot classification
  • Zero shot image classification
  • Arabic
  • Greek
  • English
  • +7
24.3K 810M–100M rows
DatasetnielsRocholl/universal-lesion-segmentation

Lesion segmentation datasets

A collection of public medical images for lesion segmentation on CT scans.

  • Image segmentation
24.3K 3
Datasettiyunes/svg-primitive-labels-sam3-v1-filtered

Filtered SVG Primitive Labels (SAM3 v1)

Over 1.16M cleaned query-SVG markup pairs with source XML, rasters, and validity masks.

  • Other
  • English
24.3K 01M–10M rows🔒 on request
Datasetmikeboss/FIP1

Six-year wheat image and measurement time series

Over 153k high-resolution aligned photos and tabular measurements of 8 characteristics across 4k wheat plots.

  • Time series forecasting
  • Embeddings
  • Tabular regression
24.2K 71K–10K rows
Datasetziqima/Steer3D-Data

Steer3D text-based 3D object editing data

A training set of images, metadata, and TRELLIS latent representations for text-based 3D model editing.

  • Image → 3d
24.2K 010K–100K rows
Datasetimageomics/TreeOfLife-200MTop pick

TreeOfLife-200M: 233M Wildlife Images

Giant biological dataset of 233 million photos across 933 thousand taxa for training BioCLIP 2.5-class models.

  • Image classification
  • Zero shot classification
  • English
  • Latin
24K 45100M–1B rows
Datasetfkvoo/ARIA-2

ARIA-2 image dataset

A continuation of Rokid's image capture dataset for visual tasks.

    23.9K 0100K–1M rows🔒 on request
    Dataset1fanj/Chartographer

    Visual chart analysis benchmark

    English chart-question pairs to evaluate whether VLMs actually analyze images instead of guessing.

    • Visual question answering
    • Question answering
    • English
    23.9K 31K–10K rows
    Datasetcadd9/X2C

    X2C robot facial mimicry dataset

    100,000 images of a humanoid robot with various facial expressions and control parameters.

    • Image → image
    • English
    23.7K 0100K–1M rows
    DatasetFrontier-AI-Research/MORALISE

    MORALISE moral alignment benchmark

    Text and visual examples for evaluating the moral compass of language and multimodal models.

      23.6K 01K–10K rows