TensorFlow 2.0 essentials: What's new
Build, deploy, and run machine learning models
While TensorFlow 1.0 was great, TensorFlow 2.0 takes it to the next level when it comes to making machine learning more intuitive. Specifically, TensorFlow 2.0 allows for more user-friendly APIs and distributed training and uses eager execution by default. Moreover, TensorFlow 2.0 also allows users to export py_functions.
Join expert Michael Grogan to discover how to migrate your existing TensorFlow models to 2.0, train models across a variety of hardware configurations, and use the benefits of graph mode to significantly enhance model performance, portability, and, ultimately, production.
What you'll learn-and how you can apply it
By the end of this live, hands-on, online course, you’ll understand:
- The key differences between TensorFlow 1.0 and TensorFlow 2.0, including API cleanup, eager execution, and graph mode
- The main advantages of using TensorFlow 2.0, particularly the removal of API incompatibilities and the need for code reimplementation
- How to implement machine learning models using TensorFlow 2.0
And you’ll be able to:
- Use the conversion tool to update previous Python code to the TensorFlow 2.0 standard
- Generate graph-building code with AutoGraph
- Implement eager execution to make interaction with TensorFlow more intuitive
- Use the Keras API (now the default in TF v2.0) to build and train machine learning models using both Python and R
This training course is for you because...
- You’re a machine learning specialist who wants to learn more about TensorFlow.
- You work with TensorFlow 1.0 and want to update to 2.0.
- You want to become proficient with TensorFlow 2.0 to deploy machine learning models.
- Experience implementing machine learning models with TensorFlow or otherwise
- A machine with TensorFlow with pip installed
- Familiarity with TensorFlow 1.0 (useful but not required)
- Install TensorFlow with pip
- Introductory tutorials from TensorFlow website
- Read Hands-on Machine Learning with Scikit-Learn, Keras, and TensorFlow, 2nd Edition (book)
- Read TensorFlow for Deep Learning (book)
- Explore Deep Learning and Reinforcement Learning with Tensorflow (learning path)
About your instructor
Michael Grogan is a Machine Learning Consultant and Educator with a technical focus on machine learning and statistical analysis – his stack includes AWS, Python, R, Shiny and SQL. He provides expertise for machine learning, regression analysis and statistics, text mining and social media analysis, time series analysis, and use of frameworks such as Keras and TensorFlow. He is a keen researcher and participant in international machine learning conferences.
The timeframes are only estimates and may vary according to how the class is progressing
TensorFlow 2.0 introduction and migration (55 minutes)
- Lecture: TensorFlow 2.0; how to migrate from TensorFlow 1.0 to 2.0
- Group discussion: The advantages and disadvantages of TensorFlow 2.0 relative to other frameworks
- Hands-on exercise: Implement an automatic migration tool for code migration to the new standard
- Break (5 minutes)
Model compilation and execution (55 minutes)
- Lecture: TensorFlow syntax; eager execution and AutoGraph; construction of machine learning models with TensorFlow 2.0
- Group discussion: Applied use cases of eager execution and AutoGraph
- Hands-on exercise: Convert code generated with eager execution to graph-generating code with AutoGraph
- Break (5 minutes)
Applied machine learning methods with TensorFlow 2.0 (60 minutes)
- Lecture: TensorFlow 2.0 advanced machine learning tools; implementing advanced machine learning models using TensorFlow 2.0. The section specifically focuses on the use of the Keras framework - this framework is now the default high-level API for training machine learning models.
- Discussion: Applied examples of machine learning models best deployed using TensorFlow v2.0
- Hands-on exercise: Implement a sample machine learning algorithm using TensorFlow 2.0 standards
TensorFlow and R (60 minutes)
- Presentation: Gain an understanding of how to use Keras under the R environment.
- Discussion: Applied examples of machine learning models that are best deployed using R.
- Presentation: Implement a regression-based predictive model using a Keras neural network.
- Exercise: Implement a sample machine learning algorithm using TF 2.0 standards.