metadata
dataset_info:
features:
- name: file
dtype: string
- name: speakerId
dtype: string
- name: transcription
dtype: string
- name: audio
dtype: audio
- name: label
dtype: string
- name: instruction
dtype: string
- name: instruction_TTS
dtype:
audio:
sampling_rate: 16000
splits:
- name: test
num_bytes: 106109540
num_examples: 200
download_size: 22536950
dataset_size: 106109540
configs:
- config_name: default
data_files:
- split: test
path: data/test-*