---
dataset_info:
  features:
  - name: input_ids
    sequence: int32
  - name: attention_mask
    sequence: int8
  - name: labels
    sequence: int64
  - name: decoder_attention_mask
    sequence: int64
  splits:
  - name: train
    num_bytes: 811075616
    num_examples: 299954
  - name: validation
    num_bytes: 40276080
    num_examples: 14895
  download_size: 66692286
  dataset_size: 851351696
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---