---
dataset_info:
  features:
  - name: output
    dtype: string
  - name: input
    dtype: string
  - name: messages
    list:
    - name: from
      dtype: string
    - name: value
      dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 1217428
    num_examples: 3159
  download_size: 341993
  dataset_size: 1217428
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---