metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: text
dtype: string
- name: input_features
sequence:
sequence: float32
- name: text_converted
dtype: string
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 8304003868.25
num_examples: 7638
- name: test
num_bytes: 2061820615.375
num_examples: 1901
download_size: 2829560787
dataset_size: 10365824483.625
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*