Skip to content

Datasets

Hugging Face datasets with plain-language descriptions: what's inside and what models to train on them. 4,797 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

DatasetKingNish/reasoning-base-20k

20k reasoning tasks with chain-of-thought

Science, coding, and math tasks with a detailed chain of thought and correct answer for training reasoning models.

  • Text generation
  • English
443 24110K–100K rows
Datasetllamafactory/DPO-En-Zh-20k

DPO-En-Zh-20k: English-Chinese preferences

20k English and Chinese preference pairs for DPO fine-tuning in LLaMA Factory.

  • Text generation
  • English
  • Chinese
442 10610K–100K rows
Datasetvldsavelyev/guitar_tab

Guitar tabs in alphaTex text format

Text corpus of musical tablature for training generative models to compose guitar parts.

  • Text generation
441 1310K–100K rows
Datasetcodeparrot/codecomplex

Java code with algorithmic complexity annotations

4,200 competitive programming solutions in Java with expert time complexity evaluations for algorithm analysis.

  • Text generation
  • Code
438 321K–10K rows
Datasetshibing624/roleplay-zh-sharegpt-gpt4-data

Chinese roleplay dialogues in ShareGPT format

A small set of Chinese roleplay dialogues in ShareGPT format, collected from multiple sources.

  • Text generation
  • Chinese
435 741K–10K rows
Datasetnunorc/squad_v1_pt

Portuguese SQuAD v1 dataset

Question-answering dataset translated into Portuguese using the Google Cloud API.

  • Question answering
  • Portuguese
433 1910K–100K rows
Datasetnomic-ai/gpt4all_prompt_generations

Prompts and responses for GPT4All training

Prompt generation dataset used to train GPT4All; English text, 100K to 1M examples.

  • Text generation
  • English
432 131100K–1M rows
DatasetAlicanKiraz0/Turkish-Finance-SFT-Dataset

Turkish Finance Q&A Dataset

A Turkish text SFT dataset of about 10M tokens for training financial assistants.

  • Question answering
  • Turkish
428 661K–10K rows
DatasetAykanugur157/tr-wiki-qa

Turkish Wikipedia Q&A

75,391 question-answer pairs in Turkish mapped to specific Wikipedia paragraphs.

  • Question answering
  • Turkish
427 610K–100K rows
DatasetMoemu/Muice-Dataset

Chinese roleplay dialogue dataset (Muice)

Chinese dialogue dataset in ShareGPT format for training roleplay chatbots.

  • Question answering
  • Text generation
  • Chinese
420 621K–10K rows
Datasetopenmed-community/MedReason-Stenographic

Medical QA with compressed reasoning

31,535 English medical question-answer pairs with compact reasoning chains in a symbolic format.

  • Question answering
  • Text generation
  • English
418 5610K–100K rows
Datasetajibawa-2023/Children-Stories-Collection

Synthetic English children's stories collection

Around 900K generated children's stories in English formatted as prompt-text pairs.

  • Text generation
  • English
417 61100K–1M rows
DatasetJackrong/Claude-opus-4.6-TraceInversion-9000x

Reconstructed Claude 4.6 CoT examples

9k chain-of-thought examples reconstructed via trace inversion for SFT and DPO.

  • Text generation
  • English
  • Chinese
  • Korean
  • +3
415 911K–10K rows
DatasetMathLLMs/MathCodeInstruct

Dataset for solving math problems via code

English texts for training language models to solve mathematical problems through code generation.

  • Question answering
  • Text generation
  • English
414 2710K–100K rows
Datasetronantakizawa/github-top-code

Top GitHub Developer Code

1.3M source code files from about 4.7K top GitHub developers across 80+ programming languages.

  • Text generation
  • Code
412 1251M–10M rows
DatasetBAAI/IndustryInstructionTop pick

BAAI industry instruction dataset across 12 domains

Millions of instruction and QA pairs in Chinese and English for domain-specific model fine-tuning.

  • Question answering
  • Chinese
  • English
408 341M–10M rows🔒 on request
Datasetshibing624/CSC

Chinese text typo correction corpus

Combined dataset of 270k examples for detecting and correcting erroneous characters in Chinese.

  • Text generation
  • Chinese
404 38100K–1M rows
Datasetliboaccn/nmt-parallel-corpusTop pick

Multilingual parallel corpus for NMT

Massive parallel text corpus for machine translation systems based on OPUS projects.

  • Translation
  • Amharic
  • Arabic
  • Azerbaijani
  • +9
404 2610B–100B rows🔒 on request
Datasetginigen-ai/smol-worldcup

Benchmark for evaluating compact language models

A multilingual dataset for comprehensive testing of small LLMs across five axes: from accuracy to efficiency.

  • Text generation
  • Question answering
  • English
  • Korean
  • Arabic
  • +4
402 47up to 1K rows
Datasetneo4j/text2cypher-2024v1

Text-to-Cypher graph database query translation

A cleaned corpus of 44,387 question, DB schema, and Cypher English pairs from Neo4j developers.

  • Text generation
  • English
398 5510K–100K rows
Datasetlm-provers/FineProofs-SFT

Olympiad problems with proofs from DeepSeek-Math-V2

7,777 examples of math olympiad problems with step-by-step reasoning and proofs from DeepSeek-Math-V2.

  • Text generation
  • Question answering
  • English
392 4310K–100K rows
Datasetd0rj/ru-instruct

Russian instruction SFT dataset

Collection of Russian instructional data for training language models, gathered from popular datasets.

  • Text generation
  • Russian
  • Code
382 10100K–1M rows