metadata
dataset_info:
features:
- name: speaker_id
dtype: string
- name: topic
dtype: string
- name: subtopic
dtype: string
- name: paragraph_id
dtype: int64
- name: sentence_id
dtype: string
- name: transcript
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: duration
dtype: float64
splits:
- name: train
num_bytes: 2506125319.2818184
num_examples: 1966
- name: test
num_bytes: 19024779.375
num_examples: 15
download_size: 2038813223
dataset_size: 2525150098.6568184
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*