metadata
dataset_info:
features:
- name: id
dtype: string
- name: v_id
dtype: string
- name: video
dtype: string
- name: source
dtype: string
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
- name: video_url
dtype: string
splits:
- name: train
num_bytes: 85450295
num_examples: 64690
download_size: 34934388
dataset_size: 85450295
configs:
- config_name: default
data_files:
- split: train
path: data/train-*