metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: transcript
dtype: string
- name: word_count
dtype: int64
- name: character_count
dtype: int64
- name: duration
dtype: float64
- name: sample_rate
dtype: int64
splits:
- name: train
num_bytes: 6491115449
num_examples: 50815
- name: test
num_bytes: 2622447930
num_examples: 21778
download_size: 9067854915
dataset_size: 9113563379
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*