wikipedia-utils / README.md
singletongue's picture
Update README.md
2ef7386
|
raw
history blame
2.7 kB
metadata
license:
  - cc-by-sa-3.0
  - gfdl
dataset_info:
  - config_name: corpus-jawiki-20230403
    features:
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 3569619848
        num_examples: 24387500
    download_size: 1297833377
    dataset_size: 3569619848
  - config_name: corpus-jawiki-20230403-cirrus
    features:
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 4779055224
        num_examples: 28018607
    download_size: 1730081783
    dataset_size: 4779055224
  - config_name: corpus-jawiki-20230403-filtered-large
    features:
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 3027074884
        num_examples: 20133720
    download_size: 1092808039
    dataset_size: 3027074884
  - config_name: paragraphs-jawiki-20230403
    features:
      - name: id
        dtype: string
      - name: pageid
        dtype: int64
      - name: revid
        dtype: int64
      - name: paragraph_index
        dtype: int64
      - name: title
        dtype: string
      - name: section
        dtype: string
      - name: text
        dtype: string
      - name: html_tag
        dtype: string
    splits:
      - name: train
        num_bytes: 4417130987
        num_examples: 9668476
    download_size: 1489512230
    dataset_size: 4417130987
  - config_name: passages-c300-jawiki-20230403
    features:
      - name: id
        dtype: int64
      - name: pageid
        dtype: int64
      - name: revid
        dtype: int64
      - name: title
        dtype: string
      - name: section
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 3939431360
        num_examples: 6639833
    download_size: 1402596784
    dataset_size: 3939431360
  - config_name: passages-c400-jawiki-20230403
    features:
      - name: id
        dtype: int64
      - name: pageid
        dtype: int64
      - name: revid
        dtype: int64
      - name: title
        dtype: string
      - name: section
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 3868482519
        num_examples: 5555583
    download_size: 1393661115
    dataset_size: 3868482519
  - config_name: passages-para-jawiki-20230403
    features:
      - name: id
        dtype: int64
      - name: pageid
        dtype: int64
      - name: revid
        dtype: int64
      - name: title
        dtype: string
      - name: section
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 3751418134
        num_examples: 9397066
    download_size: 1296071247
    dataset_size: 3751418134
language:
  - ja
size_categories:
  - 10M<n<100M

Wikipedia-Utils: Preprocessed Wikipedia Texts for NLP

Preprocessed Wikipedia texts generated with the scripts in singletongue/wikipedia-utils repo.

For detailed information on how the texts are processed, please refer to the repo.