metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: caption
dtype: string
- name: language
dtype: string
- name: speaker
dtype: string
- name: sample_id
dtype: int64
splits:
- name: train
num_bytes: 233339927.2
num_examples: 400
- name: validation
num_bytes: 29167490.9
num_examples: 50
- name: test
num_bytes: 29167490.9
num_examples: 50
download_size: 255972099
dataset_size: 291674909
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*