Skip TRTInterpreter.validate_conversion when partition already proved full support.

[micwill755]

Summary

• Skip TRTInterpreter.validate_conversion() when compile_module already proved full converter coverage (require_full_compilation and num_supported_ops == total_ops).
• That __init__ walk is a second CONVERTERS.get() over every call_function / call_method / call_module node. Partition already did the same lookup via get_graph_converter_support. Under full compilation a missing-op warning cannot fire, so the extra pass is redundant.
• Direct interpret_module_to_result / engine-converter / test-harness construction still validates (default skip_conversion_validation=False). Construct still CONVERTERS.gets per node when it converts.
• Klein cold compile: interpreter.validate_conversion 0.102s → 0s.

PR 9 of the compile-startup stack; based on #4599.

Type of change

• Performance (non-breaking; skips a redundant converter-support walk)

Test plan

• Klein NVFP4 cold compile (GPU transformer, torchtrt-213rc10): skip-off validate_conversion 0.102s; skip-on the hook does not run. Engine still builds; steady ~0.43s.
• Dedicated unit test that __init__ skips validate_conversion when the flag is set, and still warns when it is not
• Existing dynamo conversion / partition tests (CI)

Checklist

• Code follows project style (pre-commit: isort, ruff, black, mypy)
• Self-review
• Comment on why the skip is valid
• Docs update (not needed)
• New tests
• Labels

---

_{Stack created with GitHub Stacks CLI • Give Feedback 💬}