How does sagemaker-training-toolkit complement sagemaker-python-sdk?

**What did you find confusing? Please describe.**
I am currently looking into use sagemaker for training and batch processing with custom models (catboost in particular in my case) and after looking through the sagemaker-python-sdk docs on byo docker https://sagemaker.readthedocs.io/en/stable/overview.html#byo-docker-containers-with-sagemaker-estimators and then coming across this example https://github.com/aws-samples/sagemaker-byo-catboost-container-demo/blob/master/Catboost_container_for_SageMaker.ipynb which point me to this library.

Here is a sentence from README.md that points towards the potential relationship between sagemaker python sdk and training toolkit: "The SageMaker Training Toolkit can be easily added to any Docker container, making it compatible with SageMaker for training models. If you use a prebuilt SageMaker Docker image for training, this library may already be included." Would love to understand how sagemaker training toolkit is different

**Describe how documentation can be improved**
A clear description on the differences between sagemaker-training-toolkit and sagemaker-python-sdk would be amazing. For example what does it do under the hood for making byo docker container easier to wield? The estimator created via sagemaker-python-sdk can then be used for deployment, batchtransform etc is there any limitation with the docker image that does not allow those functionalities?

**Additional context**
Add any other context or screenshots about the documentation request here.

Start with README.md, then compare the linked SageMaker Python SDK BYO Docker documentation with the CatBoost container notebook. Document how the training toolkit relates to the SDK, what it does inside a training container, and whether such an image supports deployment or batch transform. Done means the README clearly answers those questions.

Written by the indexing model from the issue text.