---
dataset_info:
  features:
  - name: instruction
    dtype: string
  - name: response
    dtype: string
  splits:
  - name: train
    num_bytes: 2479736208
    num_examples: 645000
  download_size: 914598500
  dataset_size: 2479736208
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---