Daniel Nouri, Machine Learning Expert and Software Engineer, who has a big followership on github, has created an open source library called Skorch which connoisseurs might immediately make the connection that it is a combination of Scikit-Learn and PyTorch.
Better said, Skorch is an sklearn compatible neural net that wraps PyTorch. As per the documentation:
"The goal of skorch is to make it possible to use PyTorch with sklearn. This is achieved by providing a wrapper around PyTorch that has an sklearn interface. In that sense, skorch is the spiritual successor to nolearn, but instead of using Lasagne and Theano, it uses PyTorch." [source]
The goal of Skorch is to benefit as much as possible from both worlds, those of Scikit-Learn and of PyTorch. Those who are familiar with both libraries do not need to learn any new concepts to be able to work with Skorch.
One of the main advantages of this library is it has the flexibility of PyTorch and the good looks of scikit-learn. Another advantage is that it is able to work with multiple data types, such as: NumPy arrays, Python dictionaries, tensors, and more.
Two quick implementations/examples are available on its github repository. And for those interested they can explore the lengthy documentation readthedocs.io. For the examples and for more information, start by visiting its github page below.
Side note, I'm interested in working or making tutorials about Skorch since I've been doing so many for Scikit-Learn, so you might want to be on the lookout for that on my Youtube channel.
Official Skorch repo:
To stay in touch with me, follow 
Cristi Vlad Self-Experimenter and Author