metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: phoneme
dtype: string
splits:
- name: train
num_bytes: 12283417758.875439
num_examples: 109933
- name: validation
num_bytes: 1445075108.252627
num_examples: 12933
- name: test
num_bytes: 722593421.8719352
num_examples: 6467
download_size: 12162591266
dataset_size: 14451086289
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*