Skip to content

Delivery Excellence, DevOps: Cloud-native Deployments of Tensorflow Models

License

Notifications You must be signed in to change notification settings

danielschulz/DevOps_AiToProduction_DockerizedLeanApproach

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

DevOps_AiToProduction_DockerizedLeanApproach

Delivery Excellence, DevOps: Cloud-native Deployments of Data Science Models

Setup in static code

Please modify this static git repo's code before using it. For technical and legal reasons, some files are archived in unencrypted and plain 7z archives.

Extract

For legal licensing reasons only, two images may not be provided in raw format.

  • code/2-uc-deeplearning-image/2-client-calls/1-requests/kitten.7z to code/2-uc-deeplearning-image/2-client-calls/1-requests/kitten.jpg to
  • code/2-uc-deeplearning-image/2-client-calls/1-requests/cute-chick-with-hairy-pussy.7z to code/2-uc-deeplearning-image/2-client-calls/1-requests/cute-chick-with-hairy-pussy.jpg
  • code/1-uc-structured-data-json/1-server-setup/1-data-to-deploy/raw-binaries/tensorflow_serving_tutorial.7z to code/1-uc-structured-data-json/1-server-setup/1-data-to-deploy/raw-binaries/tensorflow_serving_tutorial
  • code/2-uc-deeplearning-image/1-server-setup/1-data-to-deploy/raw-binaries/mxnet.7z to code/2-uc-deeplearning-image/1-server-setup/1-data-to-deploy/raw-binaries/mxnet

The extraction keys for the presentation resp. photos may or may not be provided to you for legal reasons. You can find the applicable keys in

  • resources/presentation/credentials.txt resp.
  • resources/photos/credentials.txt Both latter software archives as git repositories are packaged only for technical reasons and do not need a passphrase to extract them.

Prepare your client's host

Amend your OS' hosts file like shown in code/0-meta/client-network/amend-hosts-file/example-hosts-file. Your hosts file resides in:

  • Windows: C:\Windows\System32\drivers\etc\hosts
  • Linux: /etc/hosts

Getting started deploying Docker Containers

Establish SSH tunnel

Shown in code/0-meta/client-network/create-ssh-tunnel/sshConnection.bash

Optional: Docker web ui

Shown in code/0-meta/server-optional-dockerwebui/deepinfrastructure-docker-webui.bash

TensorFlow / Iris model

Distribute Model data

Shown in code/1-uc-structured-data-json/1-server-setup/1-data-to-deploy

Start Docker Container

Shown in code/1-uc-structured-data-json/1-server-setup/2-bash-commands-to-execute/tf-model-iris.bash

Perform client requests

Requests in code/1-uc-structured-data-json/2-client-calls/1-requests and respective responses from code/1-uc-structured-data-json/2-client-calls/2-example-responses

MXNet / Deep Learning image classification model

Distribute Model data

Shown in code/2-uc-deeplearning-image/1-server-setup/1-data-to-deploy

Start Docker Container

Shown in code/2-uc-deeplearning-image/1-server-setup/2-bash-commands-to-execute/mxnet-model-deeplearning.bash

Perform client requests

Requests in code/2-uc-deeplearning-image/2-client-calls/1-requests and respective responses from code/2-uc-deeplearning-image/2-client-calls/2-example-responses

About

Delivery Excellence, DevOps: Cloud-native Deployments of Tensorflow Models

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages