---
dataset_info:
  features:
  - name: conversations
    list:
    - name: from
      dtype: string
    - name: value
      dtype: string
    - name: weight
      dtype: float64
  - name: source_label_exact
    sequence: string
  splits:
  - name: train
    num_bytes: 944825307
    num_examples: 584827
  download_size: 472922667
  dataset_size: 944825307
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---