Datasets:
Tasks:
Image Classification
Modalities:
Image
Formats:
imagefolder
Languages:
English
Size:
1K - 10K
License:
metadata
license: mit
task_categories:
- image-classification
language:
- en
size_categories:
- n<1K
splits:
- name: train
MB: 20.4
num_examples: 1050
- name: test
MB: 3.29
num_examples: 150
download_size: 24 MB
dataset_size: 1050
Dataset Summary
Mini(24 MB) Classification dataset for mini projects. Cats, dogs and rabbit are included as pet in this dataset.
Supported Tasks and Leaderboards
image-classification: Based on a pet image, the goal of this task is to predict the type of pet (i.e., dog or cat or rabbit).
Languages
English
Class Label Mappings:
{
"cat": 0,
"dog": 1,
"rabbit": 2,
}
Load Dataset
from datasets import load_dataset
train_dataset = load_dataset("rokmr/pets", split="train")
test_dataset = load_dataset("rokmr/pets", split="test")
#To view label
print(train_dataset[0]["label"])
#To view image
train_dataset[0]["image"]
Citation Information
@ONLINE {mini_pets,
author="Rohit Kumar",
title="Pets dataset",
month="July",
year="2023",
url="https://github.com/rokmr"
}