metadata
dataset_info:
features:
- name: id
dtype: string
- name: conv_A
list:
- name: role
dtype: string
- name: content
dtype: string
- name: conv_B
list:
- name: role
dtype: string
- name: content
dtype: string
- name: conv_A_rating
dtype: float32
- name: conv_B_rating
dtype: float32
- name: num_turns
dtype: int32
- name: source
dtype: string
splits:
- name: train
num_bytes: 101719459
num_examples: 25386
download_size: 54672656
dataset_size: 101719459
configs:
- config_name: default
data_files:
- split: train
path: data/train-*