metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: label
dtype: string
- name: input_values
sequence: float32
- name: attention_mask
sequence: int32
splits:
- name: train
num_bytes: 2149373508.375
num_examples: 4797
- name: test
num_bytes: 2149821590.25
num_examples: 4798
download_size: 3704408556
dataset_size: 4299195098.625
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*