---
dataset_info:
  features:
  - name: input_ids
    list: int32
  - name: token_mask
    list: int64
  - name: sample_mask
    dtype: float32
  splits:
  - name: train
    num_bytes: 5127948
    num_examples: 100
  download_size: 936897
  dataset_size: 5127948
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---