This repository is not yet released on PyPI but you can install it quite simply.
git clone git@github.com:Lucas-rbnt/washing-learning.git
cd washing-learning/
pip install .
If you want to contribute to development, you can also install extras packages such as pre-commit and black to format the code properly.
pip install .[dev]
From now on you can use this toolbox simply, you'll find a code snippet below
# Standard libraries
import os
from typing import List
# Third-party libraries
from washing_learning.loggers.time_loggers import chronometer
from washing_learning.vision.preprocessing.preprocessors as preprocessors
DATASET_PATH = os.path.join(os.path.dirname(__file__), "datasets")
Dataset = ...
@chronometer
def function(path: str, preprocessors: List[preprocessors]) -> Dataset:
data = dataloader(path)
for preprocessor in preprocessors:
data = preprocessor.preprocess(data)
return data
def dataloader(path: str):
raise NotImplementedErrorThe aim is to develop a Machine Learning Toolbox to preserve a consequent amount of time and energy. There is still lots of things to do such as include typing everywhere, add some PyTorch compatibility methods, generate the docs ... Feel free to contact me to contribute.
The full package documentation is available not done for detailed specifications. The documentation was built with Sphinx using a theme provided by Read the Docs.
Please refer to the CONTRIBUTING guide if you wish to contribute to this project.
This project is developed and maintained by volunteers.