Machine learning and discovery with Kubernetes William Benton - - PowerPoint PPT Presentation

▶
machine learning and discovery with kubernetes
SMART_READER_LITE
LIVE PREVIEW

Machine learning and discovery with Kubernetes William Benton - - PowerPoint PPT Presentation

Jan 31, 2024 133 likes •909 views

Machine learning and discovery with Kubernetes William Benton @willb willb@redhat.com What do machine learning workflows look like? @willb #SEMLA19 @willb #SEMLA19 @willb #SEMLA19 @willb #SEMLA19 @willb #SEMLA19 @willb #SEMLA19

slide-1
SLIDE 1

Machine learning and discovery with Kubernetes

William Benton • @willb • willb@redhat.com

slide-2
SLIDE 2

What do machine learning workflows look like?

slide-3
SLIDE 3

@willb #SEMLA19

slide-4
SLIDE 4

@willb #SEMLA19

slide-5
SLIDE 5

@willb #SEMLA19

slide-6
SLIDE 6

@willb #SEMLA19

slide-7
SLIDE 7

@willb #SEMLA19

slide-8
SLIDE 8

@willb #SEMLA19

slide-9
SLIDE 9

@willb #SEMLA19

slide-10
SLIDE 10

@willb #SEMLA19

data collection and cleaning codifying problem 
 and metrics

slide-11
SLIDE 11

@willb #SEMLA19

feature engineering model training and tuning data collection and cleaning

slide-12
SLIDE 12

@willb #SEMLA19

feature engineering model training and tuning data collection and cleaning

slide-13
SLIDE 13

@willb #SEMLA19

feature engineering model training and tuning model validation

slide-14
SLIDE 14

@willb #SEMLA19

feature engineering model training and tuning model validation

slide-15
SLIDE 15

@willb #SEMLA19

model validation model deployment monitoring and validation

slide-16
SLIDE 16

@willb #SEMLA19

model validation model deployment monitoring and validation

slide-17
SLIDE 17

@willb #SEMLA19

feature engineering model training and tuning model validation model deployment monitoring and validation data collection and cleaning codifying problem 
 and metrics
slide-18
SLIDE 18

@willb #SEMLA19

feature engineering model training and tuning model validation model deployment monitoring and validation data collection and cleaning codifying problem 
 and metrics

defining types and interfaces prototyping

slide-19
SLIDE 19

@willb #SEMLA19

feature engineering model training and tuning model validation model deployment monitoring and validation data collection and cleaning codifying problem 
 and metrics

unit, behavioral, and integration testing formal verification

slide-20
SLIDE 20

@willb #SEMLA19

feature engineering model training and tuning model validation model deployment monitoring and validation data collection and cleaning codifying problem 
 and metrics

deployment monitoring

slide-21
SLIDE 21

@willb #SEMLA19

slide-22
SLIDE 22

@willb #SEMLA19

slide-23
SLIDE 23

What’s a container?

slide-24
SLIDE 24

@willb #SEMLA19

%

slide-25
SLIDE 25

@willb #SEMLA19

% pip install numpy

slide-26
SLIDE 26

@willb #SEMLA19 pip install numpy /usr/bin/pip executable arguments virtual memory file handles / root filesystem environment LANG=en_US USER=willb ... process table network routes

slide-27
SLIDE 27

@willb #SEMLA19 pip install numpy /usr/bin/pip executable arguments virtual memory file handles / root filesystem environment LANG=en_US USER=willb ... process table network routes

slide-28
SLIDE 28

@willb #SEMLA19 pip install numpy /usr/bin/pip executable arguments virtual memory file handles / root filesystem environment LANG=en_US USER=willb ... process table network routes

slide-29
SLIDE 29

@willb #SEMLA19 pip install numpy /usr/bin/pip executable arguments virtual memory file handles / root filesystem environment LANG=en_US USER=willb ... process table network routes

slide-30
SLIDE 30

@willb #SEMLA19 pip install numpy /usr/bin/pip executable arguments virtual memory file handles / root filesystem environment LANG=en_US USER=willb ... process table network routes

slide-31
SLIDE 31

@willb #SEMLA19

pip install numpy /usr/bin/pip

executable arguments virtual memory file handles /var/lib/envs/main root filesystem environment

LANG=en_US USER=willb ...

process table network routes

slide-32
SLIDE 32

@willb #SEMLA19

pip install numpy /usr/bin/pip

executable arguments virtual memory file handles /var/lib/envs/main root filesystem environment

LANG=en_US USER=willb ...

process table network routes

100

MAXIMUM km / h

slide-33
SLIDE 33

@willb #SEMLA19

Immutable images

base image configuration and installation recipes user application code

979229b9 33721112 e8cae4f6 2bb6ab16 a8296f7e a6afd91e 6b8cad3e

slide-34
SLIDE 34

@willb #SEMLA19

Stateless microservices

slide-35
SLIDE 35

@willb #SEMLA19

Stateless microservices

slide-36
SLIDE 36

@willb #SEMLA19

Stateless microservices

slide-37
SLIDE 37

@willb #SEMLA19

Stateless microservices

slide-38
SLIDE 38

@willb #SEMLA19

Stateless microservices

slide-39
SLIDE 39

@willb #SEMLA19

Stateless microservices

slide-40
SLIDE 40

@willb #SEMLA19

Stateless microservices

slide-41
SLIDE 41

@willb #SEMLA19

Stateless microservices

slide-42
SLIDE 42

@willb #SEMLA19

Declarative app configuration

slide-43
SLIDE 43

@willb #SEMLA19

Integration and deployment

OK!

slide-44
SLIDE 44

@willb #SEMLA19

Integration and deployment

OK!

base image configuration and installation recipes application code

slide-45
SLIDE 45

@willb #SEMLA19

Integration and deployment

OK!

base image configuration and installation recipes application code

slide-46
SLIDE 46

@willb #SEMLA19

Integration and deployment

base image configuration and installation recipes application code

slide-47
SLIDE 47

What containers offer
 data scientists

slide-48
SLIDE 48

@willb #SEMLA19

slide-49
SLIDE 49

@willb #SEMLA19

slide-50
SLIDE 50

@willb #SEMLA19

O K ! O K !

slide-51
SLIDE 51

@willb #SEMLA19

No friction: mybinder.org

slide-52
SLIDE 52

@willb #SEMLA19

More flexible: source-to-image

%

slide-53
SLIDE 53

@willb #SEMLA19

More flexible: source-to-image

%

https://github.com/openshift/source-to-image

builder image application image

slide-54
SLIDE 54
slide-55
SLIDE 55
slide-56
SLIDE 56

@willb #SEMLA19

slide-57
SLIDE 57

@willb #SEMLA19

slide-58
SLIDE 58
slide-59
SLIDE 59
slide-60
SLIDE 60

@willb #SEMLA19

slide-61
SLIDE 61

@willb #SEMLA19

mA

slide-62
SLIDE 62

@willb #SEMLA19

mA

slide-63
SLIDE 63

@willb #SEMLA19

mA

distribution of input data? distribution of predictions? distribution of acyclic paths taken through scoring code? (joint)

slide-64
SLIDE 64

@willb #SEMLA19

slide-65
SLIDE 65

Where from here?

slide-66
SLIDE 66

@willb #SEMLA19

data scientists application developers data engineers

federate train models events databases file, object storage management web and mobile reporting developer UI transform transform transform archive

slide-67
SLIDE 67

@willb #SEMLA19

data scientists application developers data engineers

federate train models events databases file, object storage management web and mobile reporting developer UI transform transform transform archive

machine learning engineers

slide-68
SLIDE 68

@willb #SEMLA19

radanalytics.io

slide-69
SLIDE 69

@willb #SEMLA19

  • pendatahub.io
slide-70
SLIDE 70

@willb #SEMLA19

Kubeflow

slide-71
SLIDE 71

What did we talk about today?

slide-72
SLIDE 72

@willb #SEMLA19

slide-73
SLIDE 73

@willb #SEMLA19

slide-74
SLIDE 74

@willb #SEMLA19

slide-75
SLIDE 75

@willb #SEMLA19

slide-76
SLIDE 76

@willb #SEMLA19

willb@redhat.com • @willb https://chapeau.freevariable.com

THANKS