Datasets:
metadata
language:
- en
- fi
- 'no'
license: apache-2.0
pretty_name: MetAlbums
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 1457790757.579
num_examples: 12813
download_size: 1442158271
dataset_size: 1457790757.579
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
Dataset with 256x256 metal album covers.
I took the data from: https://www.kaggle.com/datasets/benjamnmachn/metal-albums-artwork
Dataset has 12813 images of 256x256 resolution.
Text description of image is: "<Band> - <Album Name>"