Dissecting tf.function to Discover AutoGraph Strengths and Subtleties

Explore the intricacies of AutoGraph and tf.function in TensorFlow 2.0 through this 27-minute EuroPython Conference talk. Delve into the strengths and subtleties of these features, learning how to create and reuse graphs, handle state-creating functions, and optimize performance using tf.Tensor objects. Gain essential skills for writing efficient, graph-convertible code as the speaker dissects AutoGraph's inner workings, highlights common pitfalls, and demonstrates best practices for leveraging TensorFlow 2.0's powerful capabilities.

Syllabus

Introduction
Tensorflow
Autograph
Problem
Flow1 solution
First output
Exception
Lesson
Solution
Bridge to different completely
Input type
First test
Analysis
Code
Weird behavior
Summary
Design choice
Performance measurement
Use tensor everywhere
tffunction in eager mode
tffunction problems
Flow operators everywhere
Recap
Questions