metadata
dataset_info:
features:
- name: segment_filename
dtype: string
- name: text
dtype: string
- name: audio
struct:
- name: audio
struct:
- name: array
sequence: float32
- name: path
dtype: string
- name: sampling_rate
dtype: int64
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 1927453950
num_examples: 2845
download_size: 1929557079
dataset_size: 1927453950
configs:
- config_name: default
data_files:
- split: train
path: data/train-*