---
dataset_info:
  features:
  - name: conversation_hash
    dtype: string
  - name: conversation
    list:
    - name: content
      dtype: string
    - name: content_token_ids
      dtype: 'null'
    - name: country
      dtype: 'null'
    - name: cumulative_logprob
      dtype: 'null'
    - name: finish_reason
      dtype: string
    - name: hashed_ip
      dtype: string
    - name: header
      dtype: 'null'
    - name: language
      dtype: 'null'
    - name: redacted
      dtype: 'null'
    - name: role
      dtype: string
    - name: state
      dtype: 'null'
    - name: timestamp
      dtype: 'null'
    - name: toxic
      dtype: bool
    - name: turn_identifier
      dtype: 'null'
  - name: model
    dtype: string
  - name: __index_level_0__
    dtype: int64
  splits:
  - name: train
    num_bytes: 6694409099
    num_examples: 806219
  download_size: 3282606144
  dataset_size: 6694409099
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---