metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: text
dtype: string
- name: metadata
struct:
- name: duration
dtype: float64
- name: end_time
dtype: float64
- name: language
dtype: string
- name: source
dtype: string
- name: start_time
dtype: float64
splits:
- name: train
num_bytes: 4663505476.888518
num_examples: 10501
- name: test
num_bytes: 524132966.980482
num_examples: 1167
download_size: 4758059870
dataset_size: 5187638443.869
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*