Multi-GPU fix

[mdw771]

Features/fixes

Fixed the inconsistency between multi-GPU and single GPU LSQML with compact batching.

The code now syncs

• combined object update direction buffer
• probe update direction from all object patches (for preconditioning)
• indices for step size averaging
  across all ranks.

Checklist

Have you...

• Formatted your code properly adhering to PEP-8? Considering using RUFF to format your code automatically.
• Resolved all merge conflicts with the main branch?
• Checked the diffs between your branch and the main branch to ensure that your changes are not introducing any regressions or unnecessary/irrelevant changes?