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

Datasetystemsrx/Erotic_Literature_Collection

Chinese erotic literature

51 collections of Chinese erotic literature in JSON: short stories and personal experiences, for research and model training.

  • Text generation
  • Chinese
3.4K 24010K–100K rows
DatasetFreedomIntelligence/ALLaVA-4V

Visual instructions and QA in English

Images with captions and complex question-answers in English for training multimodal models.

  • Question answering
  • Text generation
  • English
3.4K 99100K–1M rows
DatasetPsychotherapy-LLM/CBT-Bench

Cognitive behavioral therapy benchmark for LLMs

English dataset for evaluating language model skills in cognitive behavioral therapy (CBT).

  • Question answering
  • Text generation
  • Text classification
  • English
3.4K 461K–10K rows
Datasetmeituan-longcat/AMO-BenchTop pick

Olympiad math benchmark for LLMs

A collection of difficult high school math competition problems in English to test language model reasoning.

  • Question answering
  • English
3.4K 37up to 1K rows
Datasetallenai/ropes

Benchmark for causal text understanding

English questions and answers to test the ability to apply rules from context to new situations.

  • Question answering
  • English
3.4K 5110K–100K rows
Datasetibm-research/duorc

Movie plot question and answer dataset

A corpus of 100K to 1M English QA pairs based on Wikipedia and IMDb movie descriptions.

  • Question answering
  • English
3.4K 36100K–1M rows
Datasetdemelin/moral_stories

Moral stories for evaluating moral reasoning

A crowdsourced corpus of branching social stories in English for training and evaluating ethical reasoning.

  • Multiple choice
  • Text generation
  • Text classification
  • English
3.4K 36100K–1M rows
Datasetflax-sentence-embeddings/stackexchange_title_best_voted_answer_jsonl

StackExchange question and answer corpus

Dataset of question and best answer pairs from StackExchange for training retrieval and embedding models.

  • Question answering
  • English
3.3K 111M–10M rows
Datasettrace-commons/agent-traces

AI coding agent session logs

Open collection of real developer dialogs with AI agents, including tool calls and command outputs.

  • Text generation
  • English
3.3K 41up to 1K rows
DatasetAlibaba-Apsara/Superior-Reasoning-SFT-gpt-oss-120bTop pick

435k Examples for Long-CoT Training

435k English SFT examples for training models with long chains of thought.

  • Text generation
  • English
3.3K 364100K–1M rows
DatasetFancy-MLLM/R1-Onevision

Dataset for developing multimodal reasoning

Set of complex multi-step reasoning tasks at the intersection of visual and textual understanding.

  • Question answering
3.2K 47100K–1M rows
DatasetHuggingFaceGECLM/REDDIT_submissions

Archive of posts from 50 popular subreddits

Tens of millions of Reddit posts from 2006–2023, split by thematic communities for training language models.

  • Text generation
  • English
3.2K 1510M–100M rows
DatasetModotte/MathX-5M

Math instructions for fine-tuning

A large set of text pairs for instruction tuning and model fine-tuning on mathematics.

  • Question answering
  • Text generation
3.2K 751M–10M rows
DatasetFBK-MT/MCIFTop pick

Multilingual multimodal benchmark on scientific reports

Human-annotated instruction dataset based on scientific reports: speech, video, and text across four languages to evaluate MLLMs.

  • Automatic speech recognition
  • Question answering
  • Summarization
  • English
  • German
  • Italian
  • +1
3.2K 711K–10K rows
DatasetAGBonnet/augmented-clinical-notes

Clinical notes with dialogues and triplets

30k triplets from real clinical notes, synthetic dialogues, and structured data.

  • Text generation
  • English
3.2K 7810K–100K rows
DatasetOpen-Orca/SlimOrca

Cleaned instructions with GPT-4 answers

About 500k GPT-4 answers filtered by FLAN annotations for training instruction models.

  • Text classification
  • Token classification
  • Table question answering
  • English
3.1K 311100K–1M rows
DatasetTencentARC/Plot2Code

Chart-to-code generation benchmark for MLLMs

Chart images and English texts to evaluate how multimodal models write code.

  • Text generation
  • Text → image
  • Image → text
  • English
3.1K 36up to 1K rows
DatasetClusterlabAi/101_billion_arabic_words_datasetTop pick

Arabic text corpus with 101B words

A large Arabic text corpus for training and fine-tuning language models.

  • Text generation
  • Arabic
3.1K 7310M–100M rows
DatasetTobi-Bueck/customer-support-tickets

Customer support tickets in English and German

Annotated customer emails and support responses for support automation.

  • Text classification
  • Summarization
  • Question answering
  • English
  • German
3.1K 4710K–100K rows
Datasetopen-index/open-github

GitHub public events and activity archive

Tabular and text data on pushes, pull requests, releases, and discussions from public repositories.

  • Text generation
  • Text classification
  • Embeddings
  • English
  • Multiple languages
3.1K 12100K–1M rows