---
dataset_info:
  features:
  - name: idx
    dtype: int64
  - name: src
    dtype: string
  - name: tgt
    dtype: string
  - name: conversations
    list:
    - name: from
      dtype: string
    - name: value
      dtype: string
  splits:
  - name: train
    num_bytes: 555957458.0
    num_examples: 800000
  download_size: 277539795
  dataset_size: 555957458.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---