metadata
dataset_info:
features:
- name: speaker
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: embeddings
sequence: float32
splits:
- name: train
num_bytes: 272058
num_examples: 2
download_size: 234698
dataset_size: 272058
configs:
- config_name: default
data_files:
- split: train
path: data/train-*