Now that things have settled down a bit, I can work on some projects that I have been meaning to tackle for while. One of those is getting into Swift for TensorFlow.
Swift for TensorFlow is exciting because it is a general purpose deep learning framework not bound to a particular platform. It not only incorporates the core strengths of Swift programming language of safety, performance and clarity, but extends it by adding differential programming. Because of Swift’s relationship to LLVM, it can take advantage of the low-level compute graph analysis to perform these operations effeciently. The Swift for TensorFlow team is pushing Swift forward in a new and exciting ways.
S4TF will be an area of intense focus for me over the next serveral weeks. I can’t wait to get started. I will begin with the basic tutorials but looking to do something a lot more advanced later on.