metadata
dataset_info:
features:
- name: speaker_id
dtype: string
- name: video_id
dtype: string
- name: segments
sequence:
sequence: int64
- name: asd_conf
sequence: float64
- name: video_quality
dtype: int64
splits:
- name: train
num_bytes: 29290982
num_examples: 116921
download_size: 12825498
dataset_size: 29290982
configs:
- config_name: default
data_files:
- split: train
path: data/train-*