August 2019
Intermediate to advanced
202 pages
5h 9m
English
This chapter will cover how you can convert TensorFlow 1.x (TF 1.x) code into TensorFlow 2.0 (TF 2.0) code in two ways. The first method is to use the update script, which changes most of the TF 1.x code so that it can run in TF 2.0. This, however, simply converts all tf.x API calls into tf.compat.v1.x format. The other method is to convert TF 1.x code into idiomatic TF2.0 code by taking into account the core changes that have been made to the library. We will discuss the conceptual differences between TF 1.x and TF 2.0, the compatibility criteria between them, and the ways we can migrate syntactically and semantically. We will also show several examples of syntactic and semantic migration from TF 1.x ...
Read now
Unlock full access