Updated GPU wrapper to uninstall TF without GPU

[JoelH96]

Updated the Python TF-GPU wrapper to uninstall Tensorflow (without GPU) after the Seldon-core install and only install Tensorflow-GPU. An install of Numpy=1.16.1 was also added to clear Warnings due to version 1.17 of Numpy with TF-GPU.

[axsaucedo]

Looks solid - tensorflow is actually not a required install from v0.5.0 so if we wait for that release we could make sure the pip uninstall is not required, and given it's an image file we'd be able to push a snapshot for 1.0. Did you test it in a GPU instance with the loaded models?

[axsaucedo]

@adeepsri this fixes #914

[axsaucedo]

@JoelH96 it would be good to address #868 on this PR as well given it's quite relevant

[JoelH96]

Ok, maybe if we wait then until the v0.5.0 release, then I will change the wrapper to ensure the seldon-core pip install is with that version, to remove the pip uninstall.

For now though the newly built seldonio/seldon-core-s2i-python3-tf-gpu does not contain the CPU version of TF, and yes I tested it using the deep mnist models in a GPU instance.

[axsaucedo]

Now that 0.5 is out we can add this one 👍 tf doesn't come installed by default so you can just install it on top

[seldondev]

@JoelH96: The following test failed, say /retest to rerun them all:

Test name	Commit	Details	Rerun command
lint	57e105b	link	/test lint

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. I understand the commands that are listed here.

[JoelH96]

@axsaucedo - Updated the wrapper to not include the uninstall of TF. The new image also pushed to the Dockerhub repo

[axsaucedo]

Nice one! Lint seems to be failing just on the version.py so /lgtm
/

[axsaucedo]

/lgtm
/approve

[seldondev]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: axsaucedo

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [axsaucedo]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment