---
dataset_info:
  features:
  - name: content
    dtype: string
  - name: url
    dtype: string
  - name: timestamp
    dtype: timestamp[us]
  - name: dump
    dtype: string
  - name: segment
    dtype: string
  - name: image_urls
    sequence:
      sequence: string
  splits:
  - name: train
    num_bytes: 14330816788
    num_examples: 5000000
  download_size: 8665021107
  dataset_size: 14330816788
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---