---
dataset_info:
  features:
  - name: input_ids
    sequence: int32
  - name: overflow_to_sample_mapping
    dtype: int64
  splits:
  - name: train
    num_bytes: 63774762072
    num_examples: 7773618
  download_size: 20866193193
  dataset_size: 63774762072
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---