Add request param descriptions and update Swagger models help text

[MLnick]

Added help text describing request params for model/predict in more detail.

Existing Swagger UI:

Now:

Also updated the description text for various fields in the Swagger models for responses.

[gdequeiroz]

LGTM

[ajbozarth]

I think changing label to class label is an interesting choice, but I'm fine with it. LGTM

[MLnick]

@ptitzler ?

[autoih]

Input:
Any limitation of the size of an input image? For example, should be larger than 50 x 50 pixels; otherwise there is no object can be detected. By contrast, what is the maximum size of an input image?

Output:
What kind of format will the metadata generate?

I am not sure the though is good for this. If not, please ignore it. Thank you!

[ptitzler]

Nit. Description/help text that forms a valid sentence should end with a period. (see help for detection_box) Not sure if there is a standard though.

[xwu0226]

LGTM
@MLnick i think we can merge it. and let's assign tasks to other people to carry out the changes for all model repos.

[MLnick]

@xwu0226 ok let's assign someone to roll similar changes out across models.