metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
- name: audio
dtype: audio
- name: question_unit
sequence: int64
- name: response_interleaf
dtype: string
- name: response_text
dtype: string
- name: response_speech
dtype: audio
- name: speech_score
dtype: int64
- name: text_score
dtype: int64
splits:
- name: test
num_bytes: 35370995
num_examples: 300
download_size: 24359919
dataset_size: 35370995
configs:
- config_name: default
data_files:
- split: test
path: data/test-*