---
dataset_info:
  features:
  - name: tokens
    sequence: int64
  - name: labels
    sequence: int64
  - name: input_pos
    sequence: int64
  - name: seq_lens
    sequence: int64
  splits:
  - name: train
    num_bytes: 249546760
    num_examples: 1267
  download_size: 33237372
  dataset_size: 249546760
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---