metadata
dataset_info:
features:
- name: path
dtype: string
- name: name
dtype: string
- name: label
dtype: string
- name: subset_token
dtype: string
- name: subset_mask
sequence: int64
- name: mixture
dtype:
audio:
sampling_rate: 48000
- name: sampling_rate
dtype: int64
- name: alto
dtype:
audio:
sampling_rate: 48000
- name: bass
dtype:
audio:
sampling_rate: 48000
- name: finger_snap
dtype:
audio:
sampling_rate: 48000
- name: lead_vocal
dtype:
audio:
sampling_rate: 48000
- name: soprano
dtype:
audio:
sampling_rate: 48000
- name: tenor
dtype:
audio:
sampling_rate: 48000
- name: vocal_percussion
dtype:
audio:
sampling_rate: 48000
splits:
- name: test_p10
num_bytes: 3229243152.9374285
num_examples: 1051
download_size: 1107439633
dataset_size: 3229243152.9374285
configs:
- config_name: default
data_files:
- split: test_p10
path: data/test_p10-*