---
dataset_info:
  features:
  - name: abstract_id
    dtype: int32
  - name: text
    dtype: string
  - name: location
    sequence: int32
  - name: label
    sequence: string
  - name: template_name
    dtype: string
  - name: template
    dtype: string
  - name: rendered_input
    dtype: string
  - name: rendered_output
    dtype: string
  splits:
  - name: train
    num_bytes: 50375596060.0
    num_examples: 18000000
  - name: test
    num_bytes: 16787471324.0
    num_examples: 6000000
  - name: validation
    num_bytes: 16795195277.0
    num_examples: 6000000
  - name: full
    num_bytes: 218865661215.0
    num_examples: 86361714
  download_size: 142864030450
  dataset_size: 302823923876.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
  - split: validation
    path: data/validation-*
  - split: full
    path: data/full-*
---