[TRI-978][fix] Pass explicit streaming=False in base_metrics_verification_tests

[mc-nv]

@coderabbitai summary

Description

tensorrt_llm.bindings.executor.Request.__init__() no longer accepts streaming=None — the parameter type changed from nullable to bool (default False). The base_metrics_verification_tests.py _run_infer helper was omitting the streaming tensor entirely, causing get_input_scalar_by_name() to return None downstream, which was then passed directly to the executor constructor and crashed every request through the BLS path.

Fix: add streaming=False as an explicit input tensor in _run_infer so the backend always receives a valid bool.

Exposed by CI job L0_backend_trtllm--base #300267705 (pipeline 48790749, branch 26.04-devel):

• TritonModelException: __init__(): incompatible function arguments … streaming: NoneType
• Segfault (exit 139) in TRTLLMBaseMetricsTest.test_end_to_end

Test Coverage

• L0_backend_trtllm--base — TRTLLMBaseMetricsTest.test_end_to_end and test_end_to_end_beam_width (1-GPU and 2-GPU)
• custom_metrics_verification_tests.py overall gate

PR Checklist

Please review the following before submitting your PR:

• PR description clearly explains what and why. If using CodeRabbit's summary, please make sure it makes sense.

• PR Follows TRT-LLM CODING GUIDELINES to the best of your knowledge.

• Test cases are provided for new code paths (see test instructions)

• Any new dependencies have been scanned for license and vulnerabilities

• CODEOWNERS updated if ownership changes

• Documentation updated as needed

• Update tava architecture diagram if there is a significant design change in PR.

• The reviewers assigned automatically/manually are appropriate for the PR.

• Please check this after reviewing the above items as appropriate for this PR.

^{Resolves: TRI-978}