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: 3490146980
num_examples: 8578
- name: test
num_bytes: 274140406
num_examples: 699
- name: validation
num_bytes: 281775379
num_examples: 698
download_size: 2056566895
dataset_size: 4046062765
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*