[NVBUG: 5701937]Clear GPU cache for 3D weight tensors

[cjluo-nv]

Add GPU cache clearing for 3D weight tensors to prevent OOM issues.

What does this PR do?

Type of change: ? Bug fix

Testing

python hf_ptq.py --dataset cnn_dailymail --pyt_ckpt_path Llama-4-Maverick-17B-128E-Instruct --qformat nvfp4 --kv_cache_qformat fp8 --batch_size 1 --calib_size 32 --use_seq_device_map --gpu_max_mem_percentage 0.6 --trust_remote_code

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/No
• Did you write any new necessary tests?: Yes/No
• Did you add or update any necessary documentation?: Yes/No
• Did you update Changelog?: Yes/No

Additional Information

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 74.68%. Comparing base (097037d) to head (2031d8e).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #649   +/-   ##
=======================================
  Coverage   74.68%   74.68%           
=======================================
  Files         183      183           
  Lines       18552    18552           
=======================================
  Hits        13856    13856           
  Misses       4696     4696           

☔ 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.