Datasets:

BashkirNLPWorld
/

bashkir-web-corpus

dataset_info:
  features:
    - name: title
      dtype: string
    - name: content
      dtype: string
    - name: category
      dtype: string
    - name: source
      dtype: string
    - name: text
      dtype: string
  splits:
    - name: full
      num_examples: 71567
    - name: sample
      num_examples: 1000
license: cc-by-sa-4.0
task_categories:
  - text-generation
  - fill-mask
  - text-classification
language:
  - ba
pretty_name: Bashkir Web Corpus

Bashkir Web Corpus

Dataset Description

This dataset contains 71,567 documents and 46.9 million tokens collected from 16 Bashkir-language online sources, including news websites, literary magazines, social media, books, and Wikipedia.

Curated by: BashkirNLPWorld
Language(s): Bashkir (Cyrillic script)
License: CC BY-SA 4.0

Dataset Statistics

Metric	Value
Total documents	71,567
Total tokens	46,905,055
Total words	39,111,399
Total characters	278,174,165
Unique words	475,156
Average document length	3,887 characters
Documents with title	71,155 (99.4%)
Documents with category	69,992 (97.8%)

Source Distribution

Source	Documents	Percentage
bawiki_cyrillic	42,262	59.0%
azatliqorg	7,541	10.5%
yanshishma_articles	2,139	3.0%
shonkar_articles	1,919	2.7%
bashkizi_articles	1,881	2.6%
uchbash_articles	1,881	2.6%
tanburz_articles	1,871	2.6%
henek_articles	1,855	2.6%
amanat_articles	1,801	2.5%
tamasha_articles	1,746	2.4%
ye102.ru_articles	1,717	2.4%
bawikibooks_cyrillic	1,664	2.3%
bashgazet_articles	1,475	2.1%
agideljurn_articles	970	1.4%
bashkir_filtered_books	471	0.7%
akbuzat_all	223	0.3%
posts_with_comments_merged	151	0.2%

Loading the Dataset

from datasets import load_dataset

dataset = load_dataset("BashkirNLPWorld/bashkir-web-corpus", split="full")
sample = load_dataset("BashkirNLPWorld/bashkir-web-corpus", split="sample")
print(dataset[0])

Citation

@dataset{bashkirnlpworld2026webcorpus,
  title = {Bashkir Web Corpus},
  author = {BashkirNLPWorld},
  year = {2026},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/BashkirNLPWorld/bashkir-web-corpus}
}