metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: tja
dtype: string
- name: oni_relative
sequence:
sequence: float64
- name: hard_relative
sequence:
sequence: float64
- name: normal_relative
sequence:
sequence: float64
- name: easy_relative
sequence:
sequence: float64
splits:
- name: train
num_bytes: 270442406.38666666
num_examples: 67
- name: test
num_bytes: 32291630.613333333
num_examples: 8
download_size: 295386334
dataset_size: 302734037
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*