metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
- name: duration
dtype: float64
splits:
- name: train
num_bytes: 2963901963.1170645
num_examples: 26682
- name: validation
num_bytes: 310611714.52718884
num_examples: 2898
- name: test
num_bytes: 584528206.3746753
num_examples: 4956
download_size: 3785659785
dataset_size: 3859041884.0189285
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*