metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: audio_perturbed
dtype:
audio:
sampling_rate: 16000
- name: labels
dtype: string
- name: sensevoice
dtype: string
- name: continuation
dtype: string
splits:
- name: train
num_bytes: 69713905184.345
num_examples: 145321
download_size: 68721652231
dataset_size: 69713905184.345
configs:
- config_name: default
data_files:
- split: train
path: data/train-*