This winter break, I started work on a project for controlling robots with neural networks. I knew I wanted to use Scala for implementing the project, because of its static-typing safety and potential for integration into distributed computation pipelines. But I also wanted to use TensorFlow’s Python API, due to its user-friendliness and the large number of existing examples to learn from. In order to have the best of both worlds, I created ScalaPy, which allowed me to use TensorFlow’s Python API from Scala code. In this blog post, we’ll go through the basics of how ScalaPy works and then implement a basic TensorFlow example in Scala.
I run a YouTube channel about programming, in Scala so far. I have created videos about all sorts of projects, such as solving Project Euler problems, Fun Scala Projects, and building Minecraft mods. I used to create videos every few weeks or so, but as I entered middle school and took on roles in the Leadership class and Student Council, the period between videos grew and eventually I stopped making them. I wanted to start making YouTube videos again, so this summer I set out to publish 50 new YouTube videos in 50 days. And I reached that goal yesterday. This is about how I made that happen.
I recently finished reading the play Romeo and Juliet by William Shakespeare in my literature class. With the final exam coming up I was rereading the play over and over again, deepening my understanding of every part of the plot. Cramming the play into my head led to a crazy idea – what if Romeo and Juliet was modeled with actors? The next day this still seemed like a good idea and so began my journey of implementing Romeo and Juliet with Scala and Akka.