---
dataset_info:
  features:
  - name: inputs
    dtype: string
  - name: targets
    dtype: string
  - name: _template_idx
    dtype: int64
  - name: _task_source
    dtype: string
  - name: _task_name
    dtype: string
  - name: _template_type
    dtype: string
  splits:
  - name: train
    num_bytes: 4127492293.938989
    num_examples: 3425883
  download_size: 2622292097
  dataset_size: 4127492293.938989
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---