tf.function and Autograph (TF Dev Summit ‘19)

preview_player
Показать описание


Speaker:
Alexandre Passos, Software Engineer

event: TensorFlow Dev Summit 2019; re_ty: Publish; product: TensorFlow - TensorFlow Core; fullname: Alexandre Passos;
  1. TensorFlow
  2. type: Conference Talk (Full production);
  3. pr_pr: TensorFlow
  4. purpose: Educate
Рекомендации по теме
tf.function and Autograph 0:14:38

tf.function and Autograph (TF Dev Summit ‘19)

Paolo Galeone - 0:27:13

Paolo Galeone - Dissecting tf.function to discover AutoGraph strengths and subtleties

AutoGraph: Easy control 0:02:26

AutoGraph: Easy control flow for graphs (TensorFlow Tip of the Week)

6/9 Regression in 0:05:34

6/9 Regression in Tensorflow 2 : tf.function

Dissecting tf function 0:27:13

Dissecting tf function to discover AutoGraph strengths and subtleties

Tensor Flow 2.0 1:25:51

Tensor Flow 2.0 - Part 1: Deep Learning and Tabular Data + tf.function

#16: tf.function in 0:25:25

#16: tf.function in TensorFlow | Tutorial

PyCon Ireland 2019-Dissecting 0:29:08

PyCon Ireland 2019-Dissecting tf.function to discover AutoGraph strengths & subtleties-Paolo Gal...

Working with tf.data 0:53:53

Working with tf.data in TensorFlow 2 Meetup

Performant, scalable models 0:40:35

Performant, scalable models in TensorFlow 2 with tf.data, tf.function & tf.distribute (TF World ...

TF Lattice: Control 0:05:07

TF Lattice: Control Your ML with Monotonicity (TF Dev Summit '19)

Decorating with @tf.function 0:03:47

Decorating with @tf.function changes if condition output

Building a Visual 0:05:19

Building a Visual Debugging Tool for ML - TF.js in Interactive Visual Analytics (TF Dev Summit &apos...

TensorFlow.jl: A Julia 0:05:27

TensorFlow.jl: A Julia Front End to the TensorFlow World (TF Dev Summit '19)

Inside TensorFlow: tf.data 0:53:52

Inside TensorFlow: tf.data - TF Input Pipeline

Solved Dynamic Input 0:04:29

Solved Dynamic Input Shape issue with TF

Decorating with @tf.function 0:03:03

Decorating with @tf.function changes if condition output

TensorFlow Extended (TFX) 0:15:07

TensorFlow Extended (TFX) Post-training Workflow (TF Dev Summit '19)

Improving Text In 0:05:12

Improving Text In Tensorflow (TF Dev Summit ‘19)

TensorFlow Hub: Reusable 0:07:36

TensorFlow Hub: Reusable Machine Learning (TF Dev Summit '19)

#8. Декоратор tf.function 0:11:06

#8. Декоратор tf.function для ускорения выполнения функций | Tensorflow 2 уроки...

Deep Learning - 0:11:04

Deep Learning - 03 08 tf function

Introducing TensorFlow 2.0 0:22:21

Introducing TensorFlow 2.0 and its high-level APIs (TF Dev Summit '19)

#13: Graphs in 0:15:42

#13: Graphs in TensorFlow - 1 | Tutorial

Комментарии
Автор

This is one of the most impressive things I've seen in my whole life!

davidm.johnston
Автор

Wow, this is incredible. Tensorflow team, yall are beasts!😃👏

joseortiz_io
Автор

Call backs to get off training is still the best feature on Tensorflow 2.0

samola
Автор

is the code at 7:45 (the one with empty class) better than the code at 7:31? ? why???

diveshlahrani
Автор

Would be cool to get more videos like this one, covering the differences in syntax and languages. Maybe one in C++ as well?

skywalkerwin
Автор

very confusing tf.assign and tf.assign_add() .. not much examples are available in hte net...

aiinabox
Автор

Great stuff. I never like session.run, another unnecessary layer of execution, syntactic salt !

ButchCassidyAndSundanceKid
Автор

3:14 Why is it faster with tf.function? Why aren't their own library functions tf.functions?

yoloswaggins
Автор

where can i download the slides? awawsome

longliangqu