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

Datasetnvidia/OpenMathInstruct-2Top pick

OpenMathInstruct-2: 14M Math Pairs

14 million math problem-solution pairs generated by Llama 3.1 405B for instructional fine-tuning.

  • Question answering
  • Text generation
  • English
116K 27110M–100M rows
Datasetpmchard/3D-ADAM

3D Additive Manufacturing Defect Dataset

Raw data, images, and English text for anomaly detection and segmentation in 3D printing.

  • English
115.9K 111K–10K rows
Datasetbigcode/the-stack

The Stack source code corpus

A large source code corpus in dozens of programming languages with permissive licenses.

  • Text generation
  • Code
115.1K 1.1K100M–1B rows🔒 on request
DatasetLichess/standard-chess-gamesTop pick

Lichess standard chess games, 6.77B

6.77 billion rated standard games from lichess.org in tabular format for analysis.

    113.1K 1221B–10B rows
    DatasetSWE-bench/SWE-bench_Verified

    SWE-bench Verified: 500 verified GitHub tasks

    500 issue-pull request pairs from Python repositories to evaluate GitHub task resolution via unit tests.

      112.1K 205up to 1K rows
      Datasettahoebio/Tahoe-100M

      Large Scientific Dataset Tahoe-100M

      Tabular, textual, and time-series scientific data from Tahoe-100M for analysis and agent queries.

        111.8K 1571B–10B rows
        DatasetShayManor/willow-surface-code-detection-events

        Willow Surface Code Detection Events

        Quantum experiment dataset with detection events for surface code.

        • Other
        111K 110M–100M rows
        Datasetharborframework/terminal-bench-science

        Terminal scientific computing benchmark

        Continuously updated benchmark of real computational workflows in natural, exact, and engineering sciences.

          110.6K 22up to 1K rows
          DatasetInnovatorLab/Innovator-VL-Instruct-46MTop pick

          Large-scale 46M-example SFT dataset for VLMs

          A huge corpus of multimodal instructions (image + text) for full-parameter SFT of vision-language models.

          • Image text → text
          • English
          108.1K 1810M–100M rows
          Datasetinria-soda/tabular-benchmark

          OpenML tabular dataset benchmark collection

          A collection of real tabular datasets from OpenML for comparing classification and regression algorithms.

          • Tabular classification
          • Tabular regression
          106.2K 5210M–100M rows
          Datasetlaion/small-overlapping-speech-bench

          Overlapping speech recognition benchmark

          A stress test for ASR models consisting of 100 audio clips where three people speak different languages simultaneously.

          • Automatic speech recognition
          • German
          • French
          • Spanish
          • +3
          104.2K 2up to 1K rows
          Datasetopen-index/hacker-news

          Full Hacker News archive since 2006

          All Hacker News stories, comments, Ask HN, Show HN, jobs, and polls in English.

          • Text generation
          • Embeddings
          • Text classification
          • English
          103.2K 60610M–100M rows
          Datasetsemianalysisai/cc-traces-weka-062126

          Claude Code session logs by SemiAnalysis

          Filtered traces of the Claude Code agent CLI without images and service classifier calls.

          • Text generation
          102.6K 22up to 1K rows
          DatasetHuggingFaceM4/FineVisionTop pick

          Large corpus for training open VLMs

          17.3M images, 24.3M examples, and 88.9M turns for training modern open vision-language models.

            102.4K 54410M–100M rows
            Datasettruthfulqa/truthful_qa

            TruthfulQA: LLM truthfulness benchmark

            817 questions across 38 categories to test whether a model generates truthful answers.

            • Multiple choice
            • Text generation
            • Question answering
            • English
            96.7K 3021K–10K rows