metadata
dataset_info:
features:
- name: eng_caption
dtype: string
- name: hun_caption
dtype: string
- name: img
dtype: image
- name: text_embeddings
sequence: float64
- name: image_embeddings
sequence: float64
splits:
- name: train
num_bytes: 136678395.125
num_examples: 1727
download_size: 136063792
dataset_size: 136678395.125
configs:
- config_name: default
data_files:
- split: train
path: data/train-*