---
dataset_info:
  features:
  - name: input_ids
    sequence: int32
  - name: tokens
    sequence: int64
  - name: str
    dtype: string
  - name: token_bounds
    sequence: int64
  - name: cum_n_tokens
    dtype: int64
  splits:
  - name: train
    num_bytes: 1440926071
    num_examples: 245341
  download_size: 503082246
  dataset_size: 1440926071
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---