---
dataset_info:
  features:
  - name: conversations
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  - name: mask_ranges
    list:
    - name: end
      dtype: int64
    - name: mask_label
      dtype: bool
    - name: name
      dtype: string
    - name: start
      dtype: int64
  splits:
  - name: train
    num_bytes: 30465829
    num_examples: 7616
  download_size: 9449374
  dataset_size: 30465829
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---