---
dataset_info:
  features:
  - name: dataset
    dtype: string
  - name: input_tokens
    sequence: int64
  - name: instruct_tokens
    sequence: int64
  - name: output_tokens
    sequence: int64
  splits:
  - name: train
    num_bytes: 33165945700
    num_examples: 946350
  download_size: 6436648753
  dataset_size: 33165945700
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---