Deep Learning on AWS with TensorFlow - AWS Online Tech Talks

Preview:

Citation preview

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Guy Ernest, Principal Solutions Architect, Amazon

AIOct 24, 2017

Introduction to Deep Learning

on AWS with TensorFlow

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Why are we here?

Why Deep Learning?

Why Now?

Why Cloud?

Why Amazon?

Why You?

Why TensorFlow?

Why Keras?

Why Not a Demo?

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Why are we here?

Why Deep Learning?

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Neural Style Transfer

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Neural Style Transfer

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Neural Style Transfer

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Neural Style Transfer

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Why are we here?

Why Now?

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Fast Compute

Ubiquitous Data

Advanced Learning Algorithms

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Why are we here?

Why Cloud?

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

The circle of ML

Front-End team

Data Engineering team

Analysts / DS team

DevOps team

Business Problem

Data

ML Model

ML Application

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

The circle of ML

Front-End team

Data Engineering team

Analysts / DS team

DevOps team

Business Problem

Data

ML Model

ML Application

S3

EMR

Kinesis *

Redshift

Athena

Glue

ECSLambda

API-GW

Code *

DMS

EB

P2

Spot

Batch IoT

GreenGrass

FPGA

Edge

Snowball

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

AWS Snowball Edge

100T Data Transport Device

With On-board Compute

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Why are we here?

Why Amazon?

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Artificial Intelligence At AmazonThousands Of Employees Across The Company Focused on AI

Discovery &

Search

Fulfilment &

Logistics

Enhance

Existing Products

Define New

Categories Of

Products

Bring Machine

Learning To All

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

• Applied Research

• Core Research

• Alexa

• Demand Forecasting

• Risk Analytics

• Search

• Recommendations

• AI Services | Rek, Lex, Polly

• Q&A Systems

• Supply Chain Optimization

• Advertising

• Machine Translation

• Video Content Analysis

• Robotics

• Lots of Computer Vision..

• NLP / NLU

Just a few Deep Learning Use Cases at Amazon…

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Running AI In Production on AWS Today

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Pinterest

Visual Search

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Pinterest Lens

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Recommendations & Ranking At Netflix

Personalized ranking,

page generation,

search, similarity, ratings

In 140 new countries,

simultaneously

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Why are we here?

Why You?

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Amazon AIIntelligent Services Powered By Deep Learning

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

More in

2017

Infrastructure CPU

Engines MXNet TensorFlow Caffe Theano Pytorch CNTK

ServicesAmazon Polly

Platforms

IoT

Speech

Mobile

Amazon

ML

Spark &

EMRKinesis Batch ECS

GPU

More in

2017

Chat

Amazon Lex

Amazon AI: Machine Learning In The Hands Of Every Developer

Amazon Rekognition

Vision

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

More in

2017

Infrastructure CPU

Engines MXNet TensorFlow Caffe Theano Pytorch CNTK

ServicesAmazon Polly

Platforms

IoT

Speech

Mobile

Amazon

ML

Spark &

EMRKinesis Batch ECS

GPU

More in

2017

Chat

Amazon Lex

Amazon AI: Machine Learning In The Hands Of Every Developer

Amazon Rekognition

Vision

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Why are we here?

Why TensorFlow?

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

TensorFlow Execution Engine

Python Frontend

Deep Learning Layers

Keras API

Canned Estimators

TensorFlow Layers

with tf.Session() as sess:

sess.run(init)

sess.run(optimizer)

tf.layers.dense(…)

tf.layers.conv2d(…)

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Execution

Graph

TensorFlow Execution Engine

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Why are we here?

Why Keras?

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

TensorFlow Execution

Engine

Python Frontend

Deep Learning Layers

Keras API

Canned Estimators

Keras Example

from keras.models import Sequential

model = Sequential()

from keras.layers import Dense, Activation

model.add(Dense(units=64, input_dim=100))

model.add(Activation('relu'))

model.add(Dense(units=10))

model.add(Activation('softmax'))

model.compile(loss='categorical_crossentropy',

optimizer='sgd', metrics=['accuracy'])

model.fit(x_train, y_train, epochs=5, batch_size=32)

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

“Sounds great!

How can we get started?”

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Code WalkthroughMNIST (Hello World) with Keras on Tensorflow

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

AWS Deep Learning AMI: One-Click GPU Deep Learning

Up To 40,000

CUDA Cores

Python 3 Notebooks

& Examples

(and others)(Volta at launch)

Scale for

Training

TensorFlow,

Apache MXNet

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Learn more:

AWS Deep Learning AMIs: https://aws.amazon.com/amazon-ai/amis/

TensorFlow on AWS: https://aws.amazon.com/tensorflow/

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Thank You!

Recommended