metadata
dataset_info:
features:
- name: chunk_id
dtype: string
- name: path
dtype:
audio:
sampling_rate: 16000
- name: start_ms
dtype: int64
- name: end_ms
dtype: int64
- name: duration_ms
dtype: int64
- name: duration_second
dtype: float64
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 501468545
num_examples: 644
download_size: 492440998
dataset_size: 501468545
configs:
- config_name: default
data_files:
- split: train
path: data/train-*