metadata
dataset_info:
features:
- name: image
dtype: image
- name: label
dtype: int64
- name: speaker
dtype: int64
splits:
- name: train
num_bytes: 145406320
num_examples: 5000
- name: validation
num_bytes: 25541914
num_examples: 883
- name: test
num_bytes: 44826291.712
num_examples: 1558
download_size: 215460328
dataset_size: 215774525.712
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*