metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: audio_id
dtype: string
- name: type
dtype: string
- name: split
dtype: string
- name: recorder_uuid
dtype: string
- name: domain
dtype: string
- name: transcription
dtype: string
- name: language
dtype: string
- name: audio_duration
dtype: float64
splits:
- name: train
num_bytes: 17790375379.375
num_examples: 111197
- name: validation
num_bytes: 1922279561.75
num_examples: 13346
- name: test
num_bytes: 953606178.75
num_examples: 6522
download_size: 20070605071
dataset_size: 20666261119.875
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*