metadata
dataset_info:
features:
- name: text
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: input_features
sequence:
sequence: float32
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 11167405397.941107
num_examples: 9331
- name: test
num_bytes: 1238734733.4131737
num_examples: 1040
download_size: 1268255364
dataset_size: 12406140131.35428
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*