Swift is on scheduling becoming the next generation programming language for Tensorflow. Swift APIs for Tensorflow are now released in beta. However, the transient peroid from Python to Swift is still not easy for the data scientists or AI enginners, etc. Fortunately, Swift leverages the Python interoperability. Before we get started to use Python in Swift, let's take a look into the workflow of data scientistics or AI-related engineers. Because of the rapid development of Tensorflow or other deep learning frameworks, their APIs or interfaces are on common changed a lot while a new version released. The scientistics or engineers first create the virtual or seperated environment through virtualenv
, conda
or even container
. These behaviors bring challenges while introduing Python to Swift.
In the following, we demonstarte how to leverage Python in a Swift runtime using XCode under a specific conda environment.