metadata

license: apache-2.0
dataset_info:
  features:
    - name: Token
      dtype: string
    - name: Lemmas
      dtype: string
    - name: PoS
      dtype: string
    - name: Syntax analysis
      sequence: string
    - name: NER
      sequence: string
  splits:
    - name: train
      num_bytes: 6799126.690184049
      num_examples: 68068
    - name: valid
      num_bytes: 377773.65490797546
      num_examples: 3782
    - name: test
      num_bytes: 377773.65490797546
      num_examples: 3782
  download_size: 3039375
  dataset_size: 7554674
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: valid
        path: data/valid-*
      - split: test
        path: data/test-*
language:
  - ru
task_categories:
  - token-classification
tags:
  - nlp
pretty_name: 'Twilight Tokenized: Russian NLP Dataset'
size_categories:
  - 10K<n<100K

Dataset Details

This dataset is made for fun and in ✨educational✨ purposes.

It includes tokenized novel text by Stephenie Meyer "Breaking Dawn" in Russian, PoS, NER (LOC, PER, ORG), tokens' lemmas and syntax annotation. NER, PoS, lemmatization and syntax annotation were made with spaCy and their 'ru_core_news_lg'