metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: duration
dtype: float64
- name: language
dtype: string
- name: transcript_text
dtype: string
- name: flamingo_next_description
dtype: string
splits:
- name: train
num_bytes: 26488727
num_examples: 30
download_size: 24984327
dataset_size: 26488727
configs:
- config_name: default
data_files:
- split: train
path: data/train-*