A dataloader, but for JAX
Project description
Jaxonloader
A dataloader that no one asked for, but here it is anyway.
Attention!
This package is still in heavy development! There will be no backwards compatibility until version 1.0.0 is released and the API can/will change at any time without warning.
Documentation
Check out the docs here
Installation
Install this package using pip like so:
pip install jaxonloader
Quickstart
from jaxonloader import get_mnist, JaxonDataLoader
train, test = get_mnist()
# these are JaxonDatasets
train_loader = JaxonDataLoader(
train,
batch_size=4,
shuffle=True,
drop_last=True,
key=key,
jit=True
)
for x in train_loader:
pass
Philosophy
The jaxonloader
package is designed to be as lightweight as possible. It's effectively the same as the PyTorch dataloader except that it uses Numpy as the backend.
Roadmap
The goal is to keep this package as lightweight as possible, while also providing as many datasets as possible. The next steps are to gather as many datasets as possible and to provide a simple API to load them.
If you have any datasets you want to see in here, just create an issue and I will look into it :)
Other backends
Other backends are not supported and are not planned to be supported. There is already a very good dataloader for PyTorch, and with all the support PyTorch has, it's not needed to litter the world with yet another PyTorch dataloader. The same goes for TensorFlow as well.
If you really need one, which supports all backends, check out
Then why does this package exist?
For one, I just like building things and don't really care if it's needed or not.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for jaxonloader-0.3.6-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 779d79f40eb9bb6d8f543e12a0976ea66f76990087142b0eb64315607e64de34 |
|
MD5 | 10b75b3fbf23bc515995f402019c9b00 |
|
BLAKE2b-256 | 1cfbb7e9ca20d2bb1c44545598979e7c93976a4cdfac23b7205c0572c3800c97 |