Fixes #1470.
Fixes #1399.

This PR refactors the AutoTuner into different classes for the Qualification and Profiling tools. This separation allows us to provide recommendations based on the specific tool being used.

For example, with this change, the AutoTuner for the Qualification tool will now recommend BATCH_SIZE_BYTES as 1GB, while the AutoTuner for Profiling tool will recommend 2GB.

Changes:

1. Refactored object AutoTuner into trait AutoTunerConfigsProvider:

   • Introduced concrete objects QualificationAutoTunerConfigsProvider and ProfilingAutoTunerConfigsProvider.
   • Overridden BATCH_SIZE_BYTES in QualificationAutoTunerConfigsProvider.

2. Extended class AutoTuner to QualificationAutoTuner:

   • This can be to add logic specific to the Qualification AutoTuner in the future.

3. Refactored AutoTunerSuite into BaseAutoTunerSuite:

   • Created test suites ProfilingAutoTunerSuite and QualificationAutoTunerSuite.
   • Added unit test for batch size in Qualification AutoTuner.

4. Renamed the existing class QualificationAutoTuner to QualificationAutoTunerRunner:

   • This class was not an implementation of the Qualification AutoTuner but a wrapper for running the AutoTuner with the Qualification tool.