Swift for Tensorflow (and other things).
After creating MolSeeker and iBabel4 I've been investigating the use of Swift and in particular the open-source use.
The Google Swift for TensorFlow group have been very active, and Tyrolabs have recently posted a detailed summary, including a comparison with other languages.
Two years ago, a small team at Google started working on making Swift the first mainstream language with first-class language-integrated differentiable programming capabilities. The scope and initial results of the project have been remarkable, and general public usability is not very far off.
They have now provided support for Jupyter notebooks https://github.com/google/swift-jupyter
There is also an interesting blog post here fast.ai.
Developers can take advantage of the Watson Developer Cloud’s Swift SDK to easily build Watson-powered applications for iOS or Linux platforms. Leverage the power of Watson’s advanced artificial intelligence, machine learning, and deep learning techniques to understand unstructured data and engage with users in new ways.
Since Swift is a relatively new language it is worth looking at the ongoing evolution.