π§ deeplyΒΆ
Release v0.1.0. (Installation)
deeply is a simple and elegant Deep Learning library written in Python containing a growing collection of deep learning models, datasets and utilities.
Behold, the power of deeply:
>>> # import deeply
>>> import deeply
>>> import deeply.datasets as dd
>>> # load data
>>> mnist = dd.load("mnist")
>>> (train, val), test = dd.split(mnist["train"], splits = (.8, .2)), mnist["test"]
>>> # build model
>>> model = deeply.hub("efficient-net-b7")
>>> model.fit(train, validation_data = val, epochs = 10)
β featuresΒΆ
Create end-to-end pipeline repositories using deeply templates.
Avoid unnecessary code so you can simply focus on product delivery.
Integrate third-party MLOps Infrastructure for real-time experiment tracking with breeze.
Access to a wide range of datasets.
deeply officially supports Python 3.5+.
π guidesΒΆ
apiΒΆ
π€ contributionΒΆ
If you want to contribute to the project, this part of the documentation is for you.