Implement a new version of PairwiseGPModel to work with the Ax backend

[crasanders]

Currently PairwiseGPModel only works with the old backend, so it cannot take advantage of Ax features. We would like a new pairwise model that works with the Ax backend, while keeping the current model for backwards compatibility.

This new model would take in pairwise data in the same format as the current model, and it would support all of the same Ax features as the other Ax-enabled models including:

• Discrete parameters
• Fixed parameters
• Log-scaled parameters
• Integer parameters
• Parameter constraints

ContinuousRegressionGP and BinaryClassificationGP can be used as templates for how to convert old-style models (GPRegressionModel and GPClassificationModel) into Ax-enabled models. A new example config should be written and one or more unit tests should be added to test_ax_integration.py to verify that the model works as expected.