metadata
dataset_info:
features:
- name: transcription
dtype: string
- name: original_audio
dtype:
audio:
sampling_rate: 24000
- name: separated_target_audio
dtype:
audio:
sampling_rate: 24000
- name: separated_residual_audio
dtype:
audio:
sampling_rate: 24000
- name: audio
dtype: audio
- name: normalized_text
dtype: string
splits:
- name: train
num_bytes: 27512128
num_examples: 10
download_size: 26847299
dataset_size: 27512128
configs:
- config_name: default
data_files:
- split: train
path: data/train-*