metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 3838962134.5036316
num_examples: 2539
- name: test
num_bytes: 489741268.705104
num_examples: 318
- name: valid
num_bytes: 498400934.363264
num_examples: 317
download_size: 3094162480
dataset_size: 4827104337.572
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*