SagemakerModel.transform() doesn't use model's sagemakerClient

If you create a new `SageMakerModel` instance (say, with `fromModelS3Path()`), you can pass in your own `sagemakerClient`. However, when you go to use the model after it's been created, `transform()` does not use that client to send prediction requests. It appears to hardcode a `AmazonSageMakerRuntimeClientBuilder.defaultClient` instead in `RequestBatchIterator`.

Pardon my ignorance, but is there a reason that it can't just pass the `sagemakerClient` through?

Start at SageMakerModel.transform and inspect RequestBatchIterator, especially where the AmazonSageMakerRuntimeClientBuilder.defaultClient is selected. Reproduce the issue with a SageMakerModel created using a custom sagemakerClient, then verify that transform sends prediction requests through that client instead; done means the supplied client is no longer bypassed.

Written by the indexing model from the issue text.