Delete unused GPTQ prototype

[iamzainhuda]

Summary

• delete the unused torchao.prototype.gptq API, examples, benchmark, and dedicated test
• remove the GPTQ H100 CI invocation
• keep the unrelated NVFP4 MoE calibration script working by loading its C4 samples locally

The July 31 usage audit records zero OSS usage and only bogus fbcode usage for this prototype, marks it removable, and tracks the fbcode cleanup in D114413514. This removes 1,441 Python LOC under torchao and 2,292 net lines repository-wide.

Test plan

• ruff check scripts/prototype/test_nvfp4_moe.py
• ruff format --check scripts/prototype/test_nvfp4_moe.py
• python -m py_compile scripts/prototype/test_nvfp4_moe.py
• package import smoke test
• mocked C4 dataset helper smoke test
• verify no tracked references to torchao.prototype.gptq remain

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/4768

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

✅ No Failures

As of commit 2393581 with merge base f10e69b ():
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[vkuzo]

can you coordinate with the oncall to land this, since it will require internal changes in the same stack