metadata
dataset_info:
features:
- name: file
dtype: string
- name: audio
dtype: audio
- name: text
dtype: string
- name: instruction
dtype: string
- name: label
dtype: string
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 24960872147.768
num_examples: 104014
- name: validation
num_bytes: 348628035.844
num_examples: 2703
download_size: 23576168585
dataset_size: 25309500183.612003