metadata
dataset_info:
features:
- name: original_speaker_label
dtype: string
- name: speaker_id
dtype: string
- name: is_child
dtype: bool
- name: transcription
dtype: string
- name: start_ms
dtype: int64
- name: end_ms
dtype: int64
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: file_id
dtype: string
- name: total_speakers_in_file
dtype: int32
- name: speaker_mapping
dtype: string
splits:
- name: test
num_bytes: 3678143257
num_examples: 27352
download_size: 3351326318
dataset_size: 3678143257
configs:
- config_name: default
data_files:
- split: test
path: data/test-*