metadata
dataset_info:
features:
- name: id
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
- name: start_time
dtype: float32
- name: end_time
dtype: float32
splits:
- name: train
num_bytes: 532270300.1336613
num_examples: 4519
- name: validation
num_bytes: 30078412.506389778
num_examples: 238
- name: test
num_bytes: 22494982.781948883
num_examples: 251
download_size: 535273297
dataset_size: 584843695.422
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*