https://github.com/sub-mod/mnist-app

UI for tensorflow mnist-app
https://github.com/sub-mod/mnist-app

Last synced: 15 days ago
JSON representation

UI for tensorflow mnist-app

• Host: GitHub
• URL: https://github.com/sub-mod/mnist-app
• Owner: sub-mod
• Created: 2017-08-14T22:11:39.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2023-01-05T05:28:16.000Z (almost 2 years ago)
• Last Synced: 2023-09-15T13:26:52.825Z (about 1 year ago)
• Language: JavaScript
• Homepage:
• Size: 648 KB
• Stars: 3
• Watchers: 3
• Forks: 6
• Open Issues: 12
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# install

    $ pip install -r requirements.txt

    $ npm install

    $ gunicorn main:app --log-file=-

# dockerbuild

docker build -t submod/mnist-app -f Dockerfile.nodejs .

# Integration With OpenShift

Ensure that you are connected to an OpenShift project

```

oc new-project test

```

Create the template for mnist web application.

```

oc create -f https://raw.githubusercontent.com/sub-mod/mnist-app/master/template.json

```

Once the tensorflow serving endpoints are created deploy the mnist-app.

This app needs 2 serving endpoints.For more details look here [tensorflow-serving-s2i](https://github.com/radanalyticsio/tensorflow-serving-s2i)

```

oc new-app --template=mnistapp --param=APPLICATION_NAME=mnist-app \

	--param=PREDICTION_SERVICE1=tf-reg \

	--param=PREDICTION_SERVICE2=tf-cnn

```

### credits:

The UI layer code was used with permission from Yoshihiro Sugi (@sugyan). The backend was redeveloped to work with tensorflow model server and openshift.