---
dataset_info:
  features:
  - name: INSTRUCTION
    dtype: string
  - name: RESPONSE
    dtype: string
  splits:
  - name: train
    num_bytes: 17237847388
    num_examples: 3114000
  download_size: 9500391988
  dataset_size: 17237847388
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---