Add in-sample candidate generation support for MBG

[ItsMrLin]

Summary:
Enable in-sample candidate generation in the Modular BoTorch Generator (MBG), allowing optimization to select from existing training data rather than optimizing over the full search space. This is achieved via adding 2 optional kwargs in Acquisition.optimize: candidate_set and sampling_strategy

This hits 4 birds with 1 stone - the same mechanism may support:

• Model-based/Contextual bandits
• Bake-off/Best arm selection
• in-sample preference learning (in-sample PBO and BOPE preference game)
• LILO (for LLM to label observed points)

Supports multiple selection methods including GP Thompson Sampling, Top-Two Thompson Sampling (TTTS), greedy acquisition (qSimpleRegret), q-batched acquisition (qLogNEI), random selection, and Boltzmann sampling via model_gen_options.

This is partially inspired by the Support for SamplingStrategy design doc but with a wider range of applications in mind.

Reviewed By: Balandat

Differential Revision: D92124823

[meta-codesync]

@ItsMrLin has exported this pull request. If you are a Meta employee, you can view the originating Diff in D92124823.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 99.34641% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 96.78%. Comparing base (44c7c38) to head (70c4f6a).
⚠️ Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
ax/generators/torch/botorch_modular/acquisition.py	96.77%	1 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff            @@
##             main    #4909    +/-   ##
========================================
  Coverage   96.78%   96.78%            
========================================
  Files         593      593            
  Lines       62303    62456   +153     
========================================
+ Hits        60297    60450   +153     
  Misses       2006     2006            

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[meta-codesync]

This pull request has been merged in 6d637df.