metadata
license: apache-2.0
dataset_info:
features:
- name: audio_path
dtype: audio
- name: input_codec
sequence: int64
- name: input_text
dtype: string
- name: label_text
dtype: string
- name: label_codec
sequence: int64
splits:
- name: train
num_bytes: 133388549579
num_examples: 573303
download_size: 129351490837
dataset_size: 133388549579
configs:
- config_name: default
data_files:
- split: train
path: data/train-*