[tensorflow_rcnn] mask-rcnn-efs.ipynb issues

https://github.com/awslabs/amazon-sagemaker-examples/blob/master/advanced_functionality/distributed_tensorflow_mask_rcnn/mask-rcnn-efs.ipynb

This notebook doesn't use default buckets and requires to much boiler plate to be filled by a customer making it cumbersome.

Start with advanced_functionality/distributed_tensorflow_mask_rcnn/mask-rcnn-efs.ipynb, linked in the issue, and inspect how bucket configuration and customer inputs are handled. Compare the notebook's setup with SageMaker defaults. Done means the example uses default buckets and requires less boilerplate from customers.

Written by the indexing model from the issue text.