metadata
dataset_info:
features:
- name: audio_file
dtype: string
- name: speech
sequence: float32
- name: sampling_rate
dtype: int64
- name: target_text
dtype: string
splits:
- name: train
num_bytes: 10026596685
num_examples: 24623
- name: test
num_bytes: 1261436948
num_examples: 3078
- name: validation
num_bytes: 1245703472
num_examples: 3078
download_size: 11868012387
dataset_size: 12533737105
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*