Library that provides a QuickDraw dataset using the Pytorch API.
Project description
Quick, Torch!
Quick, Torch! is a simple package that provides a "Quick, Draw!" dataset using the abstract class VisionDataset
, provided by torchvision
API. This package mirrors the MNIST dataset provided in torchvision.
You can install this package with
pip install quick-torch --upgrade
Example
Here are a simple example of usage:
from quick_torch import QuickDraw
import torchvision.transforms as T
ds = QuickDraw(
root="dataset",
categories="face",
download=True,
transform=T.Resize((128, 128))
)
print(f"{len(ds) = }")
first_data = ds[0]
first_data
>>> Downloading https://storage.googleapis.com/quickdraw_dataset/full/numpy_bitmap/face.npy
>>>
>>> Downloading https://storage.googleapis.com/quickdraw_dataset/full/simplified/face.ndjson
>>>
>>> len(ds) = 161666
>>> (<PIL.Image.Image image mode=L size=128x128>, 108)
For more examples, please refer to the notebook example.ipynb
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
quick_torch-1.0.1.tar.gz
(9.9 kB
view hashes)
Built Distribution
Close
Hashes for quick_torch-1.0.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 2155671889983f489d35d097c0247cd6f6c6f19d034d8b08517048bedd53713d |
|
MD5 | 8c0c4b21f5e9f3233fa2ebf8f3724057 |
|
BLAKE2b-256 | 9c284ad2bb806640faca9d6a2de168be3b34c4907542f77d3efca64dda71074e |