EuroPython 2019: Deep Learning with TensorFlow 2.0 #MachineLearning #ArtificialIntelligence #DeepLearning #Tensorflow #TF2.0 #Keras @europython @bradmiro @tensorflow
Videos from EuroPython 2019 are up on YouTube! Brad Miro from Google gave a talk about “Deep Learning with TensorFlow 2.0“. The talk walks through TensorFlow use cases at Google, changes in TensorFlow 2.0, examples with the new API and tips for getting started. Check out the details below:
TensorFlow at Google – TensorFlow was released in 2015 as an open-source project and is used worldwide with 2,000+ contributors. Google is using TF for optimizing data centers, for AR directions with Google Maps and to identify planets in distant solar systems. If you’d like to learn more about TensorFlow projects at Google take a look at Jeff Dean’s I/O talk or Adafruit’s post on diabetic retinopathy.
2.0 features – This release addresses some of the shortcomings of TensorFlow that did not feel pythonic (Miro specifically points out ‘session.run’). Session.run was replaced with eager execution and Keras was adopted as the high-level API. Overall, the API was simplified moving towards ‘one way to do everything’. TensorFlow 2.0 aims to be easy, powerful, scalable and deployable anywhere — Including on edge devices and in the browser with TensorFlow Lite and TensorFlow.js using saved models.
TensorFlow 2.0 focuses on simplicity and ease of use, featuring updates like:
Easy model building with Keras and eager execution.
Robust model deployment in production on any platform.
Powerful experimentation for research.
API simplification by reducing duplication and removing deprecated endpoints.
Miro also hits on a number of TensorFlow resources such as TensorFlow Hub which is a place to “store and download pre-built models”, a bit like GitHub for models. TensorFlow also has datasets for testing out your TensorFlow models.
Stop breadboarding and soldering – start making immediately! Adafruit’s Circuit Playground is jam-packed with LEDs, sensors, buttons, alligator clip pads and more. Build projects with Circuit Playground in a few minutes with the drag-and-drop MakeCode programming site, learn computer science using the CS Discoveries class on code.org, jump into CircuitPython to learn Python and hardware together, TinyGO, or even use the Arduino IDE. Circuit Playground Express is the newest and best Circuit Playground board, with support for CircuitPython, MakeCode, and Arduino. It has a powerful processor, 10 NeoPixels, mini speaker, InfraRed receive and transmit, two buttons, a switch, 14 alligator clip pads, and lots of sensors: capacitive touch, IR proximity, temperature, light, motion and sound. A whole wide world of electronics and coding is waiting for you, and it fits in the palm of your hand.