---
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: 1303533454
    num_examples: 819313
  download_size: 628163621
  dataset_size: 1303533454
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---