MLops workshop AWS

© 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
MLOps Workshop
Gili Nachum -
Solutions Architect
Orit Alul –
Solutions Architect
Alon Gendler –
Solutions Architect

© 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.© 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Agenda
• Introduction to Amazon AI
• Introduction to MLOps
• Amazon SageMaker
• Pipelines
• Review hands-on exercise & related AWS services
• Build your own MLOps pipeline - Hands-on exercise
No machine learning experience required!

Put machine learning in the
hands of every developer
Our mission at AWS

The Amazon ML stack:
Broadest & deepest set of capabilities
AI SERVICES
Easily add intelligence to applications without machine learning skills
Vision | Documents | Speech | Language | Chatbots | Forecasting | Recommendations
ML SERVICES
Build, Train, and Deploy machine learning models fast and at scale
Data labeling | Pre-built algorithms & notebooks | One-click training and deployment
ML FRAMEWORKS &
INFRASTRUCTURE
Flexibility & choice, highest-performing infrastructure
Support for ML frameworks | Compute options purpose-built for ML

M L F R A M E W O R K S &
I N F R A S T R U C T U R E
A I S E R V I C E S
R E K O G N I T I O N
I M A G E
P O L L Y T R A N S C R I B E T R A N S L A T E C O M P R E H E N D
C O M P R E H E N D
M E D I C A L
L E XR E K O G N I T I O N
V I D E O
Vision Speech Chatbots
A M A Z O N S A G E M A K E R
B U I L D T R A I N
F O R E C A S TT E X T R A C T P E R S O N A L I Z E
D E P L O Y
Pre-built algorithms & notebooks
Data labeling (G R O U N D T R U T H )
One-click model training & tuning
Optimization ( N E O )
One-click deployment & hosting
M L S E R V I C E S
F r a m e w o r k s I n t e r f a c e s I n f r a s t r u c t u r e
E C 2 P 3
& P 3 d n
E C 2 C 5 F P G A s G R E E N G R A S S E L A S T I C
I N F E R E N C E
Models without training data (REINFORCEMENT LEARNING)Algorithm s & m odels ( A W S M A R K E T P L A C E )
Language Forecasting Recommendations
NEW NEWNEW
NEW
NEW
NEWNEW
NEW
NEW
The Amazon ML Stack:
Broadest & Deepest Set of Capabilities

AKA ML Operations
AKA ML Engineering
AKA ML CI/CD
AKA Productizing ML
AKA ML Devops
MLOps

The process of combining Data
Science with software engineering
principles that results in a
production product
Define MLOps

Personas
Data engineerData scientist

Personas
The Data scientist

Personas
The Data Engineer

CRISP-DM
TODO define crisp-DM and how flow

Package tasks as Docker Containers

Micro Services Deployment

Amazon SageMaker
Machine learning for every developer and data scientist

T R A I N D E P L O Y
B U I L D
Build, train, tune, and host your own models
T U N E
Pre-built
notebooks for
common
problems
Built-in, high
performance
algorithms
Amazon SageMaker

Sagemaker Monitoring - Out-of-the-box
1. CloudWatch logs 2. CloudWatch metrics

Sagemaker Monitoring – Custom metrics
Send your custom metrics to CloudWatch
e.g. using boto3:

ML Pipelines

Data Ingestion Data Cleansing
Feature
Engineering
Data
Preparation
Train
Hyperparameters
OptimizationEvaluation
Deploy QA
Environment
Sanity Test
Manual
Approval
Canary Deploy
in Production
Monitor model
statistics
New Data Code Commit
Model
Configuration
Example fat ML Pipeline
Schedule
Triggers

Pipeline Orchestrators

Remaining Challenges

Questions?

CI/CD?
Before we begin..

The practice of merging (integrating) all
developers working copies to a shared
mainline multiple times a day
Continuous Integration

Strategy for software releases wherein any
change that passes the automated testing
phase (CI) is automatically released into
the production environment
Continuous Deployment

CI/CD Overview
Source DeployBuild
Continues Integration
Continues Deployment

AWS CI/CD Tools
AWS CodeCommit AWS CodeBuild AWS CodeDeploy
AWS CodePipeline
Source DeployBuild

The Pipeline you’re about to build...
Amazon
CloudWatch
Role
AWS Lambda
(check training
status)
Zip file
(Input Artifact)
AWS Lambda
Amazon SageMaker
(train job)
Pipeline
Stage
Approved
Source Train Wait Approve Deploy
AWS CodePipeline
Amazon
CloudFormation
Amazon SageMaker
Endpoint

MLOps workshop
•Login to https://dashboard.eventengine.run
•Use you HASH key
•Start the AWS Console -> Open console
•Validate you’re in us-east-1 (North Virginia)
•Follow the instructions in http://tiny.cc/mlops

MLops workshop AWS

More Related Content

What's hot (20)

Similar to MLops workshop AWS (13)

Recently uploaded (20)

MLops workshop AWS