AutoQuantize minor improvement: limit grad enabled parameters, limit cpu-gpu sync during scoring

[realAsma]

What does this PR do?

Type of change: ?
AutoQuantize minor improvement: limit grad enabled parameters, limit cpu-gpu sync during scoring

Overview: ?

Minor improvements for AutoQuantize.
Added support to enable grad only for selected parameters - this should reduce the number of gemms In the backward pass by 1x.

Testing

Covered by unit tests (both CPU and GPU)

Before your PR is "Ready for review"

• Make sure you read and follow Contributor guidelines and your commits are signed.
• Is this change backward compatible?: Yes
• Did you write any new necessary tests?: Yes
• Did you add or update any necessary documentation?: NA
• Did you update Changelog?: NA

Additional Information

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 74.39%. Comparing base (8188a01) to head (8536e7f).

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #551      +/-   ##
==========================================
+ Coverage   74.37%   74.39%   +0.02%     
==========================================
  Files         182      182              
  Lines       18219    18235      +16     
==========================================
+ Hits        13550    13566      +16     
  Misses       4669     4669              

☔ 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.