metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
- name: source
dtype: string
- name: score
dtype: float64
splits:
- name: train
num_bytes: 906293
num_examples: 124
download_size: 331069
dataset_size: 906293
configs:
- config_name: default
data_files:
- split: train
path: data/train-*