metadata
dataset_info:
features:
- name: sentence
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: english_transcript
dtype: string
- name: labels
sequence: int64
- name: input_features
sequence:
sequence: float32
splits:
- name: train
num_bytes: 8734426298.507816
num_examples: 7970
- name: test
num_bytes: 3743638674.492183
num_examples: 3416
download_size: 3448841509
dataset_size: 12478064973
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*