metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
- name: image
dtype: string
- name: id
dtype: int64
splits:
- name: train
num_bytes: 1595110
num_examples: 5011
- name: validation
num_bytes: 399570
num_examples: 1253
download_size: 620262
dataset_size: 1994680
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*