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

Datasetmvp-lab/LLaVA-OneVision-2-DataTop pick

LLaVA-OneVision-2 Multimodal and Video Data

Large collection of video clips, captions, and spatial reasoning tasks for training multimodal models.

  • Video text → text
  • Visual question answering
  • Image text → text
  • English
59.3K 51up to 1K rows
Datasetscenario-labs/3d-collection

1,342 AI-generated 3D examples

A collection of AI-generated 3D assets with previews and source inputs, totaling around 39 GB.

    58.6K 4491K–10K rows
    Datasetzlab-princeton/Vero-600k

    600K examples for visual reasoning RL

    A curated collection of 59 English datasets for training VLMs in logic, STEM, charts, and OCR.

    • Image text → text
    • English
    58.3K 47100K–1M rows
    DatasetTIGER-Lab/OmniEdit-Filtered-1.2MTop pick

    OmniEdit: 1.2M Image Editing Pairs

    1.2 million image-text pairs for seven editing tasks, filtered via GPT-4o scores.

    • English
    58.2K 1391M–10M rows
    Datasetnvidia/PhysicalAI-WorldModel-Synthetic-Physical-Interaction-ScenesTop pick

    NVIDIA synthetic object interaction scenes

    Massive dataset of physically simulated scenes (100M+ examples) from NVIDIA Isaac Sim for world models and robotics.

      58K 55100M–1B rows
      Datasettiagozip/bilibiliTop pick

      Billions of comments, danmaku, and videos from Bilibili

      Massive, regularly updated scrape of public videos, comments, and stats from the Chinese platform Bilibili.

      • Text classification
      • Chinese
      57.7K 710B–100B rows🔒 on request
      Datasetchrisrca/clash-royale-tv-replays

      Clash Royale TV frame-by-frame match recordings

      Clash Royale gameplay screenshots at roughly 10 frames per second for computer vision and game analysis.

      • Embeddings
      57.7K 310K–100K rows
      DatasetInternRobotics/SynthVerse

      Synthetic dataset for point tracking

      Synthetic images and texts for point tracking, 3D reconstruction, and robotics tasks.

      • Robotics
      • Image → 3d
      57.1K 161K–10K rows
      DatasetPalatialSim/palatial-nv-asset-libraryTop pick

      Physics-enabled 3D assets for robotics simulators

      Over a thousand verified USD 3D objects with textures and physics, created in collaboration with NVIDIA Robotics.

      • English
      56.9K 41K–10K rows
      Datasetbshada/open-schematics

      Electronic Schematics and PCB Layouts

      Largest collection of schematics and PCBs: sources, images, metadata, and related board files.

      • Text generation
      • Image → text
      • Text → image
      • English
      55.7K 20810K–100K rows
      Datasetnvidia/PhysicalAI-Robotics-Locomanipulation-GRAILTop pick

      NVIDIA Robot Motions and Locomanipulation

      A multimodal dataset with 3D data, images, and videos of robot motions, including stair climbing and descending.

        55.5K 621K–10K rows
        Datasetzhang0jhon/Aesthetic-4K

        4K image dataset with GPT-4o captions

        Curated selection of high-quality 4K images with text descriptions for ultra-high-resolution synthesis.

          55.2K 671K–10K rows
          DatasetBenjy/typed_digital_signatures

          Synthetic digital signatures across 30 fonts

          A collection of generated handwritten signatures rendered using 30 different Google Fonts handwriting fonts.

          • Image classification
          • Zero shot image classification
          • Image feature extraction
          • English
          54.8K 4010K–100K rows
          DatasetAlibaba-VELLDEPTH/MMA-SafetyBenchTop pick

          Multimodal Agent Safety Benchmark

          Dataset and evaluations for checking the safety of multimodal agents on web and mobile tasks.

            54.4K 201K–10K rows
            DatasetDnaRnaProteins/ctc-cell-cycle-hela

            Mouse stem cell division microscopy (CTC)

            Live microscopy frame sequences with cell cycle stage annotations for temporal classification.

            • Video classification
            53.6K 0up to 1K rows
            Datasetmvp-lab/LLaVA-OneVision-1.5-Instruct-DataTop pick

            LLaVA-OneVision-1.5 multimodal instructions

            22M image-text-to-text instructions for training and fine-tuning large multimodal models.

            • Image text → text
            • English
            53.5K 9110M–100M rows
            Datasetibm-granite/ChartNet

            Multimodal Chart Dataset

            A large-scale multimodal chart corpus with images and text for data extraction, VQA, and multilingual understanding.

            • Image → text
            • Visual question answering
            • Table question answering
            • English
            • Arabic
            • Chinese
            • +9
            52.7K 6810M–100M rows
            Datasetlhpku20010120/Omni-Edu

            Multimodal School Education Instructions

            Nearly 70,000 English and Chinese examples for training educational LLMs and VLMs.

            • Text generation
            • Question answering
            • Image text → text
            • English
            • Chinese
            52.6K 610K–100K rows
            DatasetHoneyDataV2/Honey-Data-V2Top pick

            Bee Corpus for End-to-End MLLM Training

            Large English-Chinese corpus of images, texts, and documents for training multimodal models.

            • Visual question answering
            • Image text → text
            • English
            • Chinese
            52K 8410M–100M rows
            Datasetdronefreak/RDD2022

            Road Damage Images in YOLO Format

            Unofficial reconstruction of the well-known RDD2022 dataset for road defect detection.

            • Object detection
            • English
            51.5K 310K–100K rows