get_model_files api call only gets the model

[ahmedhosny]

Should also get the pre and post processing - maybe the entire contrib_src?

[michaelschwier]

I'm not sure if we should actually support this via the API. In the end, if someone really wants all the model related source, they could/should just download the repository.

[ahmedhosny]

Agree - do you think we should remove this API method all together as it might be confusing?

[michaelschwier]

I think we could actually remove it. The easiest way to get the full model source is using start.py, since it also downloads the trained-model(-weights) from a possibly external source. We should add this as recommended way to get the full model source in our documentation.

2nd easiest way is to download a release (if there is a release which might not be guaranteed for user contributed/hosted models)