Fix 'ai.h2o:sparkling-water-package_2.11:2.4.13'/'h2o_pysparkling_2.4' conflict on Azure Databricks

The implemented fix ensures the code throws exception when the versions does not match and inform the users what they can do to resolve this.

Problem is versions don’t match exactly.

Generally, people should not do this^ If they need to attach both scala and java library, they should use explicit versioning to make sure both versions are the same.