metadata
dataset_info:
features:
- name: text
dtype: string
- name: audio
struct:
- name: array
sequence: float32
- name: sampling_rate
dtype: int64
- name: terms
sequence: string
- name: entity_label
sequence: string
splits:
- name: train
num_bytes: 20992485346
num_examples: 47419
- name: validation
num_bytes: 1449627189
num_examples: 1988
- name: test
num_bytes: 1438030244
num_examples: 2001
download_size: 23920929594
dataset_size: 23880142779
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*