metadata
dataset_info:
features:
- name: id
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: duration
dtype: float32
- name: english_text
dtype: string
- name: english_text_preprocessed
dtype: string
splits:
- name: train
num_bytes: 990367067.7235023
num_examples: 4608
- name: validation
num_bytes: 21492340.879416283
num_examples: 100
- name: test
num_bytes: 107461704.39708142
num_examples: 500
download_size: 1090357434
dataset_size: 1119321113
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*