datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
spellcheck_punctuation_benchmarkRussian Spellcheck Benchmark is a new benchmark for spelling correction in Russian language.
It includes four datasets, each of which consists of pairs of sentences in Russian language.
Each pair embodies sentence, which may contain spelling errors, and its corresponding correction.
Datasets were gathered from various sources and domains including social networks, internet blogs, github commits,
medical anamnesis, literature, news, reviews and more.golos_mfa_punctuation
Golos MFA Punctuation
Расширенная версия датасета Golos —
русскоязычного корпуса речи с краудсорс и студийными записями.
Датасет дополнен пунктуацией и word-level временными метками (MFA alignment).
Опубликовано и поддерживается Jeti Labs.
Описание
Параметр
Значение
Язык
Русский (ru)
Записей
970,597
Аудио
~1,044 часов
Частота дискретизации
16,000 Hz
Формат
WAV, mono, 16-bit
Что добавлено по сравнению с оригинальным Golos… See the full description on the dataset page: https://huggingface.co/datasets/govnejri/golos_mfa_punctuation.kazakh_speech_mfa_punctuation
Kazakh Speech MFA Punctuation
Расширенная версия датасета ISSAI KSC2 —
крупнейшего открытого корпуса казахской речи от института ISSAI (Nazarbayev University).
Датасет дополнен пунктуацией и word-level временными метками (MFA alignment).
Опубликовано и поддерживается Jeti Labs.
Описание
Параметр
Значение
Язык
Казахский (kk)
Записей
595,690
Аудио
~1,110 часов
Частота дискретизации
16,000 Hz
Формат
WAV, mono, 16-bit
Размер
52.9 GB… See the full description on the dataset page: https://huggingface.co/datasets/govnejri/kazakh_speech_mfa_punctuation.punctuation-ptbrturkish-punctuation-restoration-500k
Turkish Punctuation Restoration 500K v2
Noktalama ve büyük harfleri kaldırılmış girişler ile hedef cümle çiftleri.
Doğrulanmış boyut
Train: 490,000
Validation: 5,000
Test: 5,000
Toplam: 500,000
Ana görev sütunları: id, unpunctuated_text, punctuated_text
Provenance
Veri insan mesajlarından, belgelerinden veya web kazımasından alınmamıştır. Tamamı
depodaki üretici koduyla deterministik olarak oluşturulur. Her satırda source_type,
provenance… See the full description on the dataset page: https://huggingface.co/datasets/GoktugD/turkish-punctuation-restoration-500k.punctuation-ptbr-lightpretrain_punctuation
Dataset Card for "pretrain_punctuation"
More Information needed
arabic_punctuation
Dataset Card for "arabic_punctuation"
Dataset Details
Dataset Description
This is a curated dataset, specifically designed to facilitate the study of punctuation. It has undergone rigorous manual annotation and verification on the basis of sentence structure, with sentence boundaries clearly marked. The dataset is in three folders:
The ABC component of the Arabic Punctuation Dataset: This folder features the manually annotated punctuation gold standard. It… See the full description on the dataset page: https://huggingface.co/datasets/asas-ai/arabic_punctuation.restore-punctuation-ptbr-dataset2021-punctuation-restorationThis dataset is designed to be used in training models
that restore punctuation marks from the output of
Automatic Speech Recognition system for Polish language.sv_corpora_parliament_processed_punctuationpolish-punctuation-corpus
Polish Punctuation Corpus (Wolne Lektury)
Word-level punctuation labels extracted from 11 Polish literary works, built to train
tomekceszke/polish-punctuation-restorer.
Each row is one word and the mark that followed it in the original text.
Column
Type
Meaning
word
string
Lower-cased token, punctuation removed
label
int
1 = nothing follows, 2 = comma, 3 = period
from datasets import load_dataset
ds = load_dataset("tomekceszke/polish-punctuation-corpus")… See the full description on the dataset page: https://huggingface.co/datasets/tomekceszke/polish-punctuation-corpus.czech-punctuation-pos-syntax
Czech Punctuation, POS and Syntactic Dataset 🇨🇿
A High-Quality Dataset for Punctuation Restoration and Neuro-Symbolic LLM Grounding
This dataset is a structured, linguistically annotated corpus of the Czech language, specifically designed for Punctuation Restoration tasks, Part-of-Speech (POS) tagging, and token-level syntax embedding (such as nanoGPT custom metadata training).
Unlike pure raw text corpora, this dataset provides a deterministic 1:1 token-level mapping… See the full description on the dataset page: https://huggingface.co/datasets/KRadim/czech-punctuation-pos-syntax.punctuation-nilc-bert
Dataset Card for "punctuation-nilc"
More Information needed
fix_punctuationpunctuation_restoration_4096_complex# punctuation_restoration
## Dataset Summary
This dataset is designed for **instruction fine-tuning** of large language models (LLMs), especially for the **Qwen3** family, to perform **punctuation restoration** on Mandarin Chinese text.
It is derived from the [`AWeirdDev/zh-tw-articles-6k`](https://huggingface.co/datasets/AWeirdDev/zh-tw-articles-6k) dataset. The `context` field is processed to create input-output pairs in the Qwen3-style message format.
- 🔧 **User message**: A cleaned… See the full description on the dataset page: https://huggingface.co/datasets/picard47at/punctuation_restoration_4096_complex.punctuation_restoration_700_complex# punctuation_restoration
## Dataset Summary
This dataset is designed for **instruction fine-tuning** of large language models (LLMs), especially for the **Qwen3** family, to perform **punctuation restoration** on Mandarin Chinese text.
It is derived from the [`AWeirdDev/zh-tw-articles-6k`](https://huggingface.co/datasets/AWeirdDev/zh-tw-articles-6k) dataset. The `context` field is processed to create input-output pairs in the Qwen3-style message format.
- 🔧 **User message**: A cleaned… See the full description on the dataset page: https://huggingface.co/datasets/picard47at/punctuation_restoration_700_complex.punctuation_restoration# punctuation_restoration
## Dataset Summary
This dataset is designed for **instruction fine-tuning** of large language models (LLMs), especially for the **Qwen3** family, to perform **punctuation restoration** on Mandarin Chinese text.
It is derived from the [`AWeirdDev/zh-tw-articles-6k`](https://huggingface.co/datasets/AWeirdDev/zh-tw-articles-6k) dataset. The `context` field is processed to create input-output pairs in the Qwen3-style message format.
- 🔧 **User message**: A cleaned… See the full description on the dataset page: https://huggingface.co/datasets/picard47at/punctuation_restoration.punctuation_restoration_600_complex# punctuation_restoration
## Dataset Summary
This dataset is designed for **instruction fine-tuning** of large language models (LLMs), especially for the **Qwen3** family, to perform **punctuation restoration** on Mandarin Chinese text.
It is derived from the [`AWeirdDev/zh-tw-articles-6k`](https://huggingface.co/datasets/AWeirdDev/zh-tw-articles-6k) dataset. The `context` field is processed to create input-output pairs in the Qwen3-style message format.
- 🔧 **User message**: A cleaned… See the full description on the dataset page: https://huggingface.co/datasets/picard47at/punctuation_restoration_600_complex.punctuation_restoration_900_complex# punctuation_restoration
## Dataset Summary
This dataset is designed for **instruction fine-tuning** of large language models (LLMs), especially for the **Qwen3** family, to perform **punctuation restoration** on Mandarin Chinese text.
It is derived from the [`AWeirdDev/zh-tw-articles-6k`](https://huggingface.co/datasets/AWeirdDev/zh-tw-articles-6k) dataset. The `context` field is processed to create input-output pairs in the Qwen3-style message format.
- 🔧 **User message**: A cleaned… See the full description on the dataset page: https://huggingface.co/datasets/picard47at/punctuation_restoration_900_complex.classical-chinese-punctuation
Classical Chinese Punctuation Restoration · 文言文断句·标点 💰 (Commercial Dataset)
This is a commercial dataset. A free 200-record sample is provided below
(sample.jsonl); the full 5.3M-pair corpus is available upon request.
📧 To license / purchase or request a quote, email wangeksy@gmail.com.
✅ Cleared for commercial use — built from public-domain classical works.
The task
Restore punctuation and sentence segmentation (句读) to unpunctuated Classical
Chinese — a… See the full description on the dataset page: https://huggingface.co/datasets/wangekxy/classical-chinese-punctuation.punctuation-mec-bert
Dataset Card for "mec-punctuation-v2"
More Information needed
race_high_remove_punctuation_only_questionsCopy of ehovy/race except we remove 3 questions from the training split with ids: "high12051.txt", "high12170.txt", "high12074.txt", where the questions were "." and therefore of poor quality.
english_punctuation_restorationgolos_mfa_punctuation_long
Golos MFA Punctuation (Long)
Long-form Russian speech derived from
govnejri/golos_mfa_punctuation.
Purpose
Most public Russian STT corpora ship as short clips (a few seconds each).
For benchmarking long-form transcription, VAD, punctuation, and streaming
behavior, you want minutes-long audio with reliable word-level alignments.
This dataset builds those long clips by splicing groups of consecutive
short clips together, inserting randomized silences between them, and… See the full description on the dataset page: https://huggingface.co/datasets/artmelancholy/golos_mfa_punctuation_long.punctuation-nilc-t5
Dataset Card for "punctuation-nilc-t5"
More Information needed
punctuation_restoration_1350_complex# punctuation_restoration
## Dataset Summary
This dataset is designed for **instruction fine-tuning** of large language models (LLMs), especially for the **Qwen3** family, to perform **punctuation restoration** on Mandarin Chinese text.
It is derived from the [`AWeirdDev/zh-tw-articles-6k`](https://huggingface.co/datasets/AWeirdDev/zh-tw-articles-6k) dataset. The `context` field is processed to create input-output pairs in the Qwen3-style message format.
- 🔧 **User message**: A cleaned… See the full description on the dataset page: https://huggingface.co/datasets/picard47at/punctuation_restoration_1350_complex.punctuation_restoration_750_complex# punctuation_restoration
## Dataset Summary
This dataset is designed for **instruction fine-tuning** of large language models (LLMs), especially for the **Qwen3** family, to perform **punctuation restoration** on Mandarin Chinese text.
It is derived from the [`AWeirdDev/zh-tw-articles-6k`](https://huggingface.co/datasets/AWeirdDev/zh-tw-articles-6k) dataset. The `context` field is processed to create input-output pairs in the Qwen3-style message format.
- 🔧 **User message**: A cleaned… See the full description on the dataset page: https://huggingface.co/datasets/picard47at/punctuation_restoration_750_complex.punctuation-tedtalk2012-full-text-bertpunctuation_restoration_512_complex# punctuation_restoration
## Dataset Summary
This dataset is designed for **instruction fine-tuning** of large language models (LLMs), especially for the **Qwen3** family, to perform **punctuation restoration** on Mandarin Chinese text.
It is derived from the [`AWeirdDev/zh-tw-articles-6k`](https://huggingface.co/datasets/AWeirdDev/zh-tw-articles-6k) dataset. The `context` field is processed to create input-output pairs in the Qwen3-style message format.
- 🔧 **User message**: A cleaned… See the full description on the dataset page: https://huggingface.co/datasets/picard47at/punctuation_restoration_512_complex.