---
dataset_info:
  config_name: sft_train
  features:
  - name: conversations
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  splits:
  - name: train
    num_bytes: 12942899
    num_examples: 3998
  download_size: 1663176
  dataset_size: 12942899
configs:
- config_name: sft_train
  data_files:
  - split: train
    path: sft_train/train-*
---