An API, built to make it easier

Slides:

Advertisements

Similar presentations

© 2009 Lenovo Lenovo Confidential Lenovo Confidential Lenovo Confidential Lenovo Confidential Lenovo Confidential Preload Software – Lenovo Quick Start.

Advertisements

Component Technology. Challenges Facing the Software Industry Today’s applications are large & complex – time consuming to develop, difficult and costly.

TensorFlow The Deep Learning Library You Should Be Using.

Title INTRODUCTION DISCUSSION METHODS RESULTS CONCLUSIONS

TensorFlow CS 5665 F16 practicum Karun Joseph, A Reference:

Deep Learning Software: TensorFlow

High or Low Level Programming Language? Justify your decision.

Title INTRODUCTION DISCUSSION METHODS RESULTS CONCLUSIONS

Title INTRODUCTION DISCUSSION METHODS RESULTS CONCLUSIONS

ANDROID AN OPEN HANDSET ALLIANCE PROJECT

Point of View, Multiple Perspectives

Lecture 3. Fully Connected NN & Hello World of Deep Learning

GWE Core Grid Wizard Enterprise (

Why it is Called Tensor Flow Parallelism in ANNs Project Ideas and Discussion Glenn Fung Presents Batch Renormalizating Paper.

Deep Learning in HEP Large number of applications:

Title INTRODUCTION DISCUSSION METHODS RESULTS CONCLUSIONS

Deep Learning Platform as a Service

Why You Need a Successful Construction Website. As technology has improved, so has the construction industry. This means that now is a good time to review.

Deep Learning Libraries

‘The pessimist complains about the wind

Point of View, Multiple Perspectives

Title INTRODUCTION DISCUSSION METHODS RESULTS CONCLUSIONS

IBM Start Now Host Integration Solutions

Introduction CSE 1310 – Introduction to Computers and Programming

Practice PT - Design a Digital Scene 3 days

Mini Presentations - part 2

Detecting personality based on interactions with Alexa

with Daniel L. Silver, Ph.D. Christian Frey, BBA April 11-12, 2017

Title INTRODUCTION DISCUSSION METHODS RESULTS CONCLUSIONS

Comparison Between Deep Learning Packages

TERIMA KASIH / THANK YOU

Introducing Umbraco Latch

Title INTRODUCTION DISCUSSION METHODS RESULTS CONCLUSIONS

Torch 02/27/2018 Hyeri Kim Good afternoon, everyone. I’m Hyeri. Today, I’m gonna talk about Torch.

Point of View, Multiple Perspectives

Home Clear Medium-Fi Prototype

INF 5860 Machine learning for image classification

Get Good at DevOps: Feature Flag Deployments with ASP

So What Do We Need To Do? Find a Useful Theory of Networking

Big Red Cloud Offers a Simple Online Accounts Solution for Business Owners and Bookkeepers Hosted on the Powerful Microsoft Azure Platform MICROSOFT AZURE.

Auth0 Is Identity Made Simple for Developers, Built by Developers and Supported by the High Availability and Performance of Microsoft Azure MICROSOFT AZURE.

Neural network systems

Brain Inspired Algorithms Dr. David Fagan

Deep Learning Packages

Title INTRODUCTION DISCUSSION METHODS RESULTS CONCLUSIONS

Title INTRODUCTION DISCUSSION METHODS RESULTS CONCLUSIONS

Jason furmanek Kris murphy IBM

Introduction to Deep Learning with Keras

SAS Deep Learning: From Toolkit to Fast Model Prototyping

Title INTRODUCTION DISCUSSION METHODS RESULTS CONCLUSIONS

CSC 578 Neural Networks and Deep Learning

AI Stick Easy to learn and use, accelerate the industrialization of artificial intelligence, and let the public become an expert in AI.

Title （No. IWA-xxxxxx） INTRODUCTION DISCUSSION METHODS RESULTS

Using the Microsoft AI Platform for next generation applications

Title INTRODUCTION DISCUSSION METHODS RESULTS CONCLUSIONS

Semantic Similarity Detection

Title Your name, co-author 1 name, co-author 2 name

Title INTRODUCTION DISCUSSION METHODS RESULTS CONCLUSIONS

Title INTRODUCTION DISCUSSION METHODS RESULTS CONCLUSIONS

Title INTRODUCTION DISCUSSION METHODS RESULTS CONCLUSIONS

Point of View, Multiple Perspectives

1 Affiliation 1, contact details; 2Affiliation 2, contact details

Product Improvement ICup By Samantha Jimenez, Ryan Zhao, Aaron Park.

Artificial intelligence for everyone

Deep Learning Libraries

CSC 578 Neural Networks and Deep Learning

Igor Stančin, Alan Jović to: {igor.stancin,

RESULTS AND DISCUSSION

The Automation of the Sequencing of Accelerator Cycles and Events at Fermilab Stanley Johnson Fermilab Accelerator Operations.

Presentation transcript:

An API, built to make it easier Keras An API, built to make it easier MHD Munzer Alsallakh

What is Keras Keras Why use Keras Keras Vs. TF Playing with Keras

What is Keras Keras is a high-level neural networks API, written in Python and capable of running on top of TensorFlow, CNTK, or Theano. It was developed with a focus on enabling fast experimentation. “Being able to go from idea to result with the least possible delay is key to doing good research.”(keras.io)

Why Keras Rapid prototyping Easily switching between APIs Industry best practices are already built in The default settings are a big help Prebuilt systems GPU acceleration

Keras vs. TF Keras makes it easy to develop a model Keras is a high level programming API Keras works for both simple and complex problems Satan Hides within details

Time for Koding Building a simple model Using ResNet50 prebuilt system Using Keras with tensor board

Photos Credits https://store.waitbutwhy.com/ http://clipart.info/devil-emoji-png-transparent-icon-4689 https://peru.com/epic/epic-mobile/facebook-messenger-no-solo- pacman-este-emoji-tambien-desaparecio-fotos-noticia-458421 https://www.youtube.com/watch?v=BiD5AhFcNW8 https://www.emojibase.com/emojilist/hooray https://tensorflow.org http://deathbattle.wikia.com/wiki/File:Vs.png https://keras.io

धन्यवाद Thank you 谢谢 شُكْرَاً لَكُمْ Terima kasih Merci a vous شکریہ