metadata
dataset_info:
features:
- name: utterance_id
dtype: string
- name: age_bucket
dtype: string
- name: audio_duration_sec
dtype: float64
- name: reference
dtype: string
- name: zs_prediction
dtype: string
splits:
- name: zeroshot_val
num_bytes: 998461
num_examples: 4803
download_size: 585231
dataset_size: 998461
configs:
- config_name: default
data_files:
- split: zeroshot_val
path: data/zeroshot_val-*