metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 44100
- name: transcription
dtype: string
- name: audio_path
dtype: string
splits:
- name: train
num_bytes: 3674692657.156
num_examples: 5167
download_size: 3158760552
dataset_size: 3674692657.156
configs:
- config_name: default
data_files:
- split: train
path: data/train-*