Python client for communicating with tensorflow-serving.

Install dependencies and start the Docker container with:

script/up

Soruce install your Python virtualenv with:

script/env

Run tests with:

script/test

Upload a new version to PyPI with:

script/distribute

Query the running tensorflow-serving Docker container instance with:

tensorflow_serving_client --host localhost --port 9000 --image tests/fixtures/files/cat.jpg --model_spec mobilenet_v1