metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: locale
dtype: string
- name: conversation_id
dtype: int64
- name: turn_index
dtype: int64
- name: duration_sec
dtype: float64
- name: transcript
dtype: string
- name: category
dtype: string
splits:
- name: train
num_bytes: 67844574
num_examples: 846
download_size: 54631882
dataset_size: 67844574
configs:
- config_name: default
data_files:
- split: train
path: data/train-*