Can not load models pulled from ML Flow

[adamwrobel-ext-gd]

Context: Models trained in SageMaker, but stored in ML Flow model registry, have multiple additional files, for example some metadata, requirements.txt, etc.

When it's zipped and provided as a SageMaker model, this function tries to read all of the files as if they were models: https://github.com/aws/sagemaker-xgboost-container/blob/b1a3cf0576fc6967584d683649d28d8897699a42/src/sagemaker_xgboost_container/algorithm_mode/handler_service.py#L52

But they are not models.

It should have some kind of pattern filter or a way to provide the file name, and not the folder. Otherwise it fails to load a model from a requirements.txt file which is just additional information.