metadata
dataset_info:
config_name: french
features:
- name: id
dtype: string
- name: images
list:
- name: bytes
dtype: 'null'
- name: path
dtype: string
- name: conversation
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 1888943
num_examples: 2607
- name: val
num_bytes: 221013
num_examples: 320
- name: test
num_bytes: 221964
num_examples: 320
download_size: 915741
dataset_size: 2331920
configs:
- config_name: french
data_files:
- split: train
path: french/train-*
- split: val
path: french/val-*
- split: test
path: french/test-*