Add new APIs to clean up resources from predictor and transformer.

[chuyang-deng]

… transformer to delete model.

Issue #, if available:
#447

Description of changes:

• Add support to delete endpoint configuration by default within delete_endpoint() to predictor class.
• Add delete_model() to transformer class. Instead of linking a model class to it, the transformer class directly calls session's delete_model() method because it already has access to the model_name. We don't want to pass in a model object on top of that. We don't want to construct a model object here either because that requires extra parameters, such as the s3 path to the model artifacts, while we can just utilize the model_name variable.

Merge Checklist

Put an x in the boxes that apply. You can also fill these out after creating the PR. If you're unsure about any of them, don't hesitate to ask. We're here to help! This is simply a reminder of what we are going to look for before merging your pull request.

• I have read the CONTRIBUTING doc
• I have added tests that prove my fix is effective or that my feature works (if appropriate)
• I have updated the changelog with a description of my changes (if appropriate)
• I have updated any necessary documentation (if appropriate)

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.

[chuyang-deng]

I only included changes regarding deleting endpoint/endpoint config in this PR. model class linkage changes will be sent in a separate one because many files will be changed to link model and I don't want this PR to be too long.

[ChoiByungWook]

This looks good!

Let's update the first PR conversation in regards to why the linking of the model isn't done within the Transformer class and instead the session is called.

Let's also relate this PR and your upcoming model linkage PR to this issue: #447

As it is in regards to delete_model and delete_endpoint config.

[ChoiByungWook]

Had some general comments. It looks really good!

Let's fix the title of this PR, since the original title was too long and it is being cut off.

[ChoiByungWook]

LGTM!

Just left some minor comments in regards to the tests.

[ChoiByungWook]

Awesome!

[ChoiByungWook]

the model should be uppercase, but we can fix in the other PR.