---
dataset_info:
  features:
  - name: added
    dtype: string
  - name: created
    dtype: timestamp[us]
  - name: id
    dtype: string
  - name: metadata
    struct:
    - name: authors
      sequence: string
    - name: license
      dtype: string
    - name: license_source
      dtype: string
    - name: license_type
      dtype: string
    - name: provenance
      dtype: string
    - name: repo
      dtype: string
    - name: url
      dtype: string
  - name: source
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 9997150745
    num_examples: 5000000
  download_size: 4914288682
  dataset_size: 9997150745
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---