fix: improve error messages for waiter timeouts

[joshuatowner]

Improve timeout error messages for TrainingJob and EvaluationJob waits

When wait() times out, the current error says "Increase the timeout and try again" — which implies the job stopped and needs to be re-created. In reality, the job is still running server-side; only the client-side polling stopped.

Changes:

• Add message parameter to TimeoutExceededError so callers can provide context-specific guidance (defaults to existing message for backward compat)
• TrainingJob.wait(): timeout now says "Your training job is still running. Call .refresh() to check its current status."
• EvaluationPipelineExecution.wait(): timeout now says "Your evaluation job is still running. Use .refresh() to check its current status." Also changes resource_type from PipelineExecution to EvaluationJob for clarity.
• Fix pre-existing typo resouce_type → resource_type in TrainingJob.wait() (would have caused TypeError at runtime)

[mollyheamazon]

You have helped caught typo of this autogenerated script: should be resource_type instead of resouce_type, thanks! Can you help update the source for this script

1. https://github.com/aws/sagemaker-python-sdk/blob/master/sagemaker-core/src/sagemaker/core/tools/templates.py#L335, also line 388, 439.
2. Regenerate resources.py:

   cd sagemaker-core
   python -m sagemaker.core.tools.codegen

3. That single template fix will correct all 85+ resources in the generated
   resources.py at once

[gauravmadarkal]

LGTM