metadata
dataset_info:
features:
- name: file
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: sentence
dtype: string
- name: start_timestamp
dtype: float32
- name: end_timestamp
dtype: float32
- name: loss
dtype: float32
- name: contains_speech
dtype: bool
- name: sample_name
dtype: string
splits:
- name: test
num_bytes: 186658104
num_examples: 1000
download_size: 185899963
dataset_size: 186658104