Tensorflow serving Node.js client

Description

Simple Node.js client to send images to a TensorFlow Serving server

Docker Build

docker build -t tensorflow-serving-docker-client .

Docker Run

Example for using a local image called turtle.jpg that is in your current path

docker run -ti -v $(pwd):/image thefoo/tensorflow-serving-docker-client --host=<tensorflow servering host> --port=9000 --image=/image/turtle.jpg

Example for downloading a remote image from a url and deleting it afterward

docker run -ti thefoo/tensorflow-serving-docker-client --host=<tensorflow servering host> --image=https://upload.wikimedia.org/wikipedia/commons/7/74/A-Cat.jpg -d

Run from source

node . --host=<tensorflow servering host> --image=turtle.jpg

Options

-h, --host          Tensorflow servering host

-p, --port          Tensorflow servering host port (default : 9000)

-i, --image         Image path

-v, --verbose       Shows extra output

-d, --delete        This will delete the image locally if you are providing the image form a url