Example using `KernelShap` with distributed options

[jklaise]

With ray we can optionally run KernelShap explanations in parallel as documented here, but it would be nice to have an example and adding some benchmarks (we already have some internally from previous work by @alexcoca).

This is slightly related to #357 but the example should only consist of models that can be serialized by ray.

[alexcoca]

The DistributedKernelShap repo has some benchmark results one can refer to in the meantime!

[jklaise]

For reference, the repo in question: https://github.com/SeldonIO/DistributedKernelShap