metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: text
dtype: string
- name: question
dtype: string
- name: answer
dtype: string
- name: task_type
dtype: string
- name: conversations
dtype: string
- name: voice_name
dtype: string
splits:
- name: train
num_bytes: 1810623662
num_examples: 419
download_size: 1804646279
dataset_size: 1810623662
configs:
- config_name: default
data_files:
- split: train
path: data/train-*