---
dataset_info:
  features:
  - name: instruction
    dtype: string
  - name: input
    dtype: string
  - name: response
    dtype: string
  splits:
  - name: train
    num_bytes: 44552310.4
    num_examples: 24000
  - name: valid
    num_bytes: 5569038.8
    num_examples: 3000
  - name: test
    num_bytes: 5569038.8
    num_examples: 3000
  download_size: 10592211
  dataset_size: 55690387.99999999
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: valid
    path: data/valid-*
  - split: test
    path: data/test-*
---