TensorFlow 2.0 : what’s coming up next for Deep Learning most successful framework

Screenshot 2019-06-04 at 07.55.36.jpg
Screenshot 2019-06-04 at 07.52.40.jpg

Post Title : TensorFlow

Stars: ⭐️⭐️⭐️

Technical level: 🚶‍♂️- very accessible

Guests:

  • Rajat Monga head of the TensorFlow team at Google

  • Lex Friedman, researcher @ MIT

 

Introduction

This podcast presents a very interesting chat with one of the creators of TensorFlow. After a bit of historical background on Tensor Flow, Rajat and Lex talk about the future of this great technology.

The 3 main points

  1. TensorFlow is more than tech: it is a community and an ecosystem of tools. With 41m downloads, 60K commits on Github and 18k contributors, Tensor flow is a "culture" in the deep learning community and beyond.

  2. Great documentation and ease to use is the recipe for success. Tensor Flow became the most used deep learning framework on the planet because developers with limited knowledge on Machine Learning can integrate it in their Software projects.

  3. Flexibility, flexibility and flexibility. While the most common case is by far, transfer learning (e.g. adapt a pre-trained model to your problem) TF offers flexible solutions for users with different Deep Learning skills, if you are a researcher you can design your own layer, while if you are a developer with limited time, you can use a pre-trained model (e.g. Inception or Resnet ).

Lessons Learned ABOUT TENSOR FLOW 2.0

This is a short preview of what is coming up next with the new version of Tensor Flow:

  1. Keras will become the easiest way to interact with tensor flow. The strategy of google is clear: more of the same.. enabling non-expert users to get their transfer learning use cases done will become even easier through Keras.

  2. Improved Portability. While you already have a lot of options at the moment, it is still very challenging to move models around. With TF 2.0 there will be more cohesiveness in the ecosystem with the ability to port models from TS Js, to TF extended and TF mobile

  3. Monolithic architecture will be unbundled without breaking compatibility. TF 2.0 will have better integration among elements, and more harmonization between the different tools and projects so that you will be able to seamlessly run models on cluster or mobile.

Our Suggestions

  • read some of the TensorFlow case studies published in the Google Blog

Keywords

deep learning, TensorFlow, Google research, machine learning

 
 
carrie-yang-479864-unsplash.jpg

..want to start with TensorFlow?

Let us know and we will point you in the right direction…