---
license: apache-2.0
dataset_info:
  features:
  - name: docid
    dtype: string
  - name: title
    dtype: string
  - name: abstract
    dtype: string
  splits:
  - name: train
    num_bytes: 27354270996
    num_examples: 17778788
  download_size: 15902176848
  dataset_size: 27354270996
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---

```python
Dataset({
    features: ['docid', 'title', 'abstract'],
    num_rows: 17778788
})
```