---
dataset_info:
  features:
  - name: data_combined
    sequence: int64
  - name: retrieve_start_idx_list
    sequence: int64
  - name: retrieve_score_list
    sequence: float64
  splits:
  - name: train
    num_bytes: 3697624044
    num_examples: 53779
  - name: val
    num_bytes: 206312448
    num_examples: 3000
  download_size: 596762594
  dataset_size: 3903936492
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: val
    path: data/val-*
---