2ch_preprocessed / README.md
qklent's picture
Upload dataset
7351263 verified
metadata
dataset_info:
  features:
    - name: timestamp
      dtype: int64
    - name: score
      dtype: float64
    - name: subject
      dtype: string
    - name: full_title
      dtype: string
    - name: board_name
      dtype: string
    - name: views
      dtype: int64
    - name: unique_posters
      dtype: int64
    - name: conversation
      list:
        - name: content
          dtype: string
        - name: role
          dtype: string
  splits:
    - name: train
      num_bytes: 770202637.5785354
      num_examples: 25301
  download_size: 404836021
  dataset_size: 770202637.5785354
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*