[https://nvbugs/6094107][infra] Isolate GB300 repro test

[yuxianq]

@coderabbitai summary

Description

Limit l0_gb300_multi_gpus to the exact DeepSeekV3Lite FP8 block-scales 4-GPU repro for NVBug 6094107. This makes GB300-4_GPUs-PyTorch-Post-Merge-1 run only the target test.

Test Coverage

• YAML parsed successfully for tests/integration/test_lists/test-db/l0_gb300_multi_gpus.yml
• Verified the file has one test entry
• Verified scripts/test_to_stage_mapping.py maps the test to GB300-4_GPUs-PyTorch-Post-Merge-1

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.

[yuxianq]

/bot run --stage-list "GB300-4_GPUs-PyTorch-Post-Merge-1"

[tensorrt-cicd]

PR_Github #47353 [ run ] triggered by Bot. Commit: 69ce9dd Link to invocation

[tensorrt-cicd]

PR_Github #47353 [ run ] completed with state SUCCESS. Commit: 69ce9dd
/LLM/main/L0_MergeRequest_PR pipeline #37289 (Partly Tested) completed with status: 'SUCCESS'

CI Report

Link to invocation

[yuxianq]

/bot run --stage-list "GB300-4_GPUs-PyTorch-Post-Merge-1" --disable-reuse-test

[tensorrt-cicd]

PR_Github #47377 [ run ] triggered by Bot. Commit: 69ce9dd Link to invocation

[tensorrt-cicd]

PR_Github #47377 [ run ] completed with state SUCCESS. Commit: 69ce9dd
/LLM/main/L0_MergeRequest_PR pipeline #37308 (Partly Tested) completed with status: 'SUCCESS'

CI Report

Link to invocation