metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: sampling_rate
dtype: int64
- name: transcript
dtype: string
splits:
- name: train
num_bytes: 2148620097.848024
num_examples: 15011
- name: validation
num_bytes: 244354283.92892572
num_examples: 1707
- name: test
num_bytes: 587045068.7487814
num_examples: 4091
download_size: 2690757910
dataset_size: 2980019450.525731
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*