---
dataset_info:
  features:
  - name: id
    dtype: int64
  - name: paragraph_id
    dtype: int64
  - name: content
    dtype: string
  - name: paper_arxiv_id
    dtype: string
  - name: paper_section
    dtype: string
  - name: section_id
    dtype: float64
  - name: paragraph_in_paper_id
    dtype: int64
  splits:
  - name: train
    num_bytes: 3672528966
    num_examples: 5140865
  download_size: 1249277536
  dataset_size: 3672528966
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---