[None][test] Add deepseek v4 pro cases on the qa side

[fredricz-20070104]

Summary by CodeRabbit

• New Performance Coverage

  • Added GB300 DeepSeek-V4-Pro benchmark scenarios for disaggregated and multi-node workloads.
  • Added coverage across multiple concurrency levels, parallelism layouts, slot capacities, and speculative decoding configurations.
  • Added NIXL-based generation and context-processing configurations for streaming end-to-end tests.

• Tests

  • Expanded automated performance test lists with new DeepSeek-V4-Pro scenarios.
  • Added deterministic expert and slot assignment fixtures for MoE load-balancing benchmarks.

Add deepseek v4 pro cases on the qa side.

[fredricz-20070104]

/bot skip --comment "Only add deepseek cases for the qa side"

[coderabbitai]

📝 Walkthrough

Walkthrough

Added GB300 DeepSeek-V4-Pro disaggregated benchmark configurations, fixed MoE load-balancer fixtures, and four QA performance test entries for disaggregated and multinode runs.

Changes

DeepSeek-V4-Pro performance coverage

Layer / File(s)	Summary
MoE load-balancer fixtures tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/*	Adds precomputed initial_global_assignments mappings for multiple expert-parallel and slot-count configurations, with updates disabled per iteration.
GB300 benchmark configurations tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-*.yaml	Adds four disaggregated GB300 benchmark definitions covering Slurm execution, worker topology, NIXL, KV cache, parallelism, and speculative decoding settings.
QA test registration tests/integration/test_lists/qa/llm_perf_disagg.yml, tests/integration/test_lists/qa/llm_perf_multinode.txt	Adds four DeepSeek-V4-Pro end-to-end cases to GB300 disaggregated and multinode test lists.

Estimated code review effort: 3 (Moderate) | ~20 minutes

Possibly related PRs

• NVIDIA/TensorRT-LLM#16540: Adds related GB300 DeepSeek-V4-Pro performance test cases and supporting load-balancer configuration handling.

Suggested reviewers: nv-ananjappa, dhansen-nvidia

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The description is too brief and missing the required summary, test coverage, and PR checklist sections.	Expand the PR body to fill the template sections: Description, Test Coverage, and PR Checklist, with a short explanation of what changed and why.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title is concise and clearly matches the added DeepSeek V4 Pro QA test cases.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

Warning

Review ran into problems

🔥 Problems

Git: Failed to clone repository. Please run the @coderabbitai full review command to re-trigger a full review. If the issue persists, set path_filters to include or exclude specific files.

---

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 3

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In
`@tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_ctx_ep4_384.yaml`:
- Line 1: Add the repository-standard NVIDIA copyright header at the beginning
of each affected YAML file, before its existing top-level key:
tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_ctx_ep4_384.yaml
(lines 1-1),
tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_gen_ep16_slots384.yaml
(lines 1-1),
tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_gen_ep16_slots416.yaml
(lines 1-1),
tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_gen_ep32_slots384.yaml
(lines 1-1),
tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con666_ctx6_dep4_gen1_dep16_eplb384_mtp3_ccb-NIXL.yaml
(lines 1-1), and
tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con8_ctx1_dep4_gen4_tep8_eplb0_mtp3_ccb-NIXL.yaml
(lines 1-1).

In
`@tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_gen_ep32_slots416.yaml`:
- Line 1: Prepend the standard NVIDIA copyright header to each affected YAML
file:
tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_gen_ep32_slots416.yaml,
moe_load_balancer_gen_ep4_slots384.yaml,
moe_load_balancer_gen_ep4_slots416.yaml,
moe_load_balancer_gen_ep8_slots384.yaml,
moe_load_balancer_gen_ep8_slots416.yaml, and the two gb300_deepseek-v4-pro YAML
files listed in the review. Keep the existing initial_global_assignments content
unchanged after the header.

In
`@tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con180_ctx3_dep4_gen1_dep32_eplb384_mtp3_ccb-NIXL.yaml`:
- Around line 17-27: Replace the <dataset_file> placeholder with a concrete,
resolvable dataset path in the benchmark configuration at
tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con180_ctx3_dep4_gen1_dep32_eplb384_mtp3_ccb-NIXL.yaml
lines 17-27 and the sibling configuration at
tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con4301_ctx12_dep4_gen1_dep8_eplb384_mtp1_ccb-NIXL.yaml
lines 17-27. Keep the existing benchmark settings unchanged.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 825e6e97-6979-48d3-b1d0-4f01cf913555

📥 Commits

Reviewing files that changed from the base of the PR and between b8604c4 and b3db26c.

📒 Files selected for processing (15)

• tests/integration/test_lists/qa/llm_perf_disagg.yml
• tests/integration/test_lists/qa/llm_perf_multinode.txt
• tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_ctx_ep4_384.yaml
• tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_gen_ep16_slots384.yaml
• tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_gen_ep16_slots416.yaml
• tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_gen_ep32_slots384.yaml
• tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_gen_ep32_slots416.yaml
• tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_gen_ep4_slots384.yaml
• tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_gen_ep4_slots416.yaml
• tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_gen_ep8_slots384.yaml
• tests/scripts/perf/disaggregated/deepseek-v4-pro-eplb/moe_load_balancer_gen_ep8_slots416.yaml
• tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con180_ctx3_dep4_gen1_dep32_eplb384_mtp3_ccb-NIXL.yaml
• tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con4301_ctx12_dep4_gen1_dep8_eplb384_mtp1_ccb-NIXL.yaml
• tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con666_ctx6_dep4_gen1_dep16_eplb384_mtp3_ccb-NIXL.yaml
• tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con8_ctx1_dep4_gen4_tep8_eplb0_mtp3_ccb-NIXL.yaml

[tensorrt-cicd]

PR_Github #60338 [ skip ] triggered by Bot. Commit: 23d6548 Link to invocation

[tensorrt-cicd]

PR_Github #60338 [ skip ] completed with state SUCCESS. Commit: 23d6548
Skipping testing for commit 23d6548

Link to invocation

[fredricz-20070104]

/bot skip --comment "Only add deepseek cases for the qa side"

[tensorrt-cicd]

PR_Github #60594 [ skip ] triggered by Bot. Commit: 413f009 Link to invocation

[tensorrt-cicd]

PR_Github #60594 [ skip ] completed with state SUCCESS. Commit: 413f009
Skipping testing for commit 413f009

Link to invocation