metadata
dataset_info:
features:
- name: file_name
dtype: string
- name: accent
dtype: string
- name: sound_array
struct:
- name: array
sequence: float32
- name: input_values
sequence: float32
- name: labels
dtype: int64
splits:
- name: train
num_bytes: 774799658
num_examples: 1447
download_size: 777237142
dataset_size: 774799658
configs:
- config_name: default
data_files:
- split: train
path: data/train-*