Skip to content

[None][test] Enable overlap schedule on qa side for disagg perf - #17390

Merged
fredricz-20070104 merged 5 commits into
NVIDIA:mainfrom
fredricz-20070104:feature/enable_overlap_schedule_on_perf_test
Aug 7, 2026
Merged

[None][test] Enable overlap schedule on qa side for disagg perf#17390
fredricz-20070104 merged 5 commits into
NVIDIA:mainfrom
fredricz-20070104:feature/enable_overlap_schedule_on_perf_test

Conversation

@fredricz-20070104

@fredricz-20070104 fredricz-20070104 commented Aug 7, 2026 •
edited by coderabbitai Bot
Loading

Copy link
Copy Markdown
Collaborator

Summary

  • Enabled overlap scheduling across disaggregated performance and performance-sanity configurations by setting disable_overlap_scheduler: false for generation and context workers.
  • Removed two stale Kimi entries from the QA timeout test list.
  • Updated configurations across GB200, GB300, B200, H200, and related test environments.

Dev Engineer Review

  • The YAML changes use valid boolean values and apply the intended scheduler setting.
  • The changes are configuration-only and do not alter public APIs or error handling.
  • Review configurations where only one worker changed, and confirm that the asymmetric settings are intentional.
  • The removed Kimi entries reference unavailable eplb0 configurations. The available entries use eplb416.

QA Engineer Review

  • No test functions were added, modified, or removed.
  • Performance configuration files under tests/scripts/perf and tests/scripts/perf-sanity were updated.
  • tests/integration/test_lists/qa/llm_perf_disagg.yml removed two Kimi timeout entries.
  • Coverage of all updated configurations in CI and manual QA lists requires confirmation.

Verdict: needs follow-up.

Enable overlap schedule on qa side for disagg perf

@fredricz-20070104
enable overlap schedule on qa side for disagg perf
aa14ef0 
Signed-off-by: FredricZ-2007 <226039983+fredricz-20070104@users.noreply.github.com>
@fredricz-20070104
fredricz-20070104 requested review from a team as code owners August 7, 2026 01:54
reasonsolo
reasonsolo approved these changes Aug 7, 2026
View reviewed changes

@reasonsolo reasonsolo left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We can leave some ctx overlap scheduler disabled, because the benefit of overlap scheduler on ctx side depends on workload, right? @hyukn

@coderabbitai

coderabbitai Bot commented Aug 7, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Review Change Stack

Walkthrough

This change updates disaggregated performance and performance-sanity YAML configurations to set disable_overlap_scheduler to false for generation and context workers. Two Kimi-K2.5-Thinking timeout test entries are removed.

Changes

Overlap scheduler configuration

Layer / File(s) Summary
Update worker overlap settings
tests/scripts/perf*/disaggregated/*.yaml		 Generation and context worker configurations now set disable_overlap_scheduler to false. Some configurations add the setting where it was absent.
Update timeout test lists
tests/integration/test_lists/qa/llm_perf_disagg.yml		 Removes one GB200 and one GB300 Kimi-K2.5-Thinking test entry.

Estimated code review effort: 2 (Simple) | ~10 minutes

Possibly related PRs

Suggested reviewers: brnguyen2, chenfeiz0326, ruodil

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name Status Explanation Resolution
Description check ⚠️ Warning The description states the change but omits the required Description, Test Coverage, and PR Checklist sections. Add the required sections and document the rationale, affected tests, test coverage, and checklist status.
✅ Passed checks (4 passed)
Check name Status Explanation
Title check ✅ Passed The title clearly identifies the test change to enable overlap scheduling for disaggregated performance QA.
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

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

coderabbitai[bot]
coderabbitai Bot reviewed Aug 7, 2026
View reviewed changes

@coderabbitai coderabbitai Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Actionable comments posted: 8

🤖 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/gb200_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-NIXL.yaml`:
- Line 85: Add the standard NVIDIA copyright header with the 2026 year to this
YAML file and all other modified YAML files, preserving their existing
configuration content and formatting.
- Line 85: The eight listed disaggregated MTP configurations enable overlap
without accuracy validation; set disable_overlap_scheduler to true in
tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-NIXL.yaml:85,
tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-UCX.yaml:84,
tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_32k4k_con256_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml:81,
tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb256_mtp3_ccb-NIXL.yaml:82,
tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con1229_ctx7_dep4_gen1_dep8_eplb384_mtp3_ccb-NIXL.yaml:105,
tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con2_ctx1_dep4_gen5_tep4_eplb0_mtp3_ccb-NIXL.yaml:103,
tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con666_ctx8_dep4_gen1_dep32_eplb384_mtp3_ccb-NIXL.yaml:105,
and
tests/scripts/perf/disaggregated/gb300_glm-5-fp4_1k1k_con1_ctx1_dep2_gen1_tep4_eplb0_mtp3_ccb-NIXL.yaml:65,
or add reference-output validation to each performance test before accepting
results.
- Line 123: Restore disable_overlap_scheduler to true in the ctx configuration
for all affected disaggregated configurations:
tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-NIXL.yaml:123-123;
gb200_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-UCX.yaml:121-121;
gb200_deepseek-v32-fp4_32k4k_con256_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml:118-118;
gb200_deepseek-v32-fp4_32k4k_con256_ctx1_dep8_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml:88-88;
gb200_deepseek-v32-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb256_mtp3_ccb-NIXL.yaml:117-117;
gb300_deepseek-v32-fp4_8k1k_con4096_ctx1_dep4_gen1_dep32_eplb256_mtp0_ccb-NIXL.yaml:114-114;
gb300_deepseek-v32-fp4_8k1k_con4096_ctx1_dep4_gen1_dep32_eplb256_mtp0_ccb-UCX.yaml:112-112;
gb300_deepseek-v4-pro-fp4_8k1k_con1229_ctx7_dep4_gen1_dep8_eplb384_mtp3_ccb-NIXL.yaml:105-105;
gb300_deepseek-v4-pro-fp4_8k1k_con2_ctx1_dep4_gen5_tep4_eplb0_mtp3_ccb-NIXL.yaml:103-103;
gb300_deepseek-v4-pro-fp4_8k1k_con666_ctx8_dep4_gen1_dep32_eplb384_mtp3_ccb-NIXL.yaml:105-105;
and
gb300_glm-5-fp4_1k1k_con1_ctx1_dep2_gen1_tep4_eplb0_mtp3_ccb-NIXL.yaml:94-94. No
other configuration changes are required.

In
`@tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml`:
- Line 77: Add the standard NVIDIA copyright header with year 2026 to all ten
affected YAML files:
tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
(77-77),
tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_8k1k_con4096_ctx1_dep4_gen1_dep32_eplb256_mtp0_ccb-NIXL.yaml
(82-82),
tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_8k1k_con4096_ctx1_dep4_gen1_dep32_eplb256_mtp0_ccb-UCX.yaml
(81-81),
tests/scripts/perf/disaggregated/gb200_glm-5-fp4_8k1k_con1024_ctx1_dep4_gen1_dep8_eplb256_mtp1_ccb-NIXL.yaml
(80-80),
tests/scripts/perf/disaggregated/gb200_glm-5-fp4_8k1k_con1024_ctx1_dep4_gen1_dep8_eplb256_mtp1_ccb-UCX.yaml
(80-80),
tests/scripts/perf/disaggregated/gb300_glm-5-fp4_1k1k_con4096_ctx1_dep2_gen1_dep8_eplb256_mtp1_ccb-NIXL.yaml
(70-70),
tests/scripts/perf/disaggregated/gb300_glm-5-fp4_1k1k_con512_ctx1_dep2_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
(67-67),
tests/scripts/perf/disaggregated/gb300_glm-5-fp4_8k1k_con1024_ctx1_dep2_gen1_dep8_eplb256_mtp1_ccb-NIXL.yaml
(80-80),
tests/scripts/perf/disaggregated/gb300_glm-5-fp4_8k1k_con1024_ctx1_dep2_gen1_dep8_eplb256_mtp1_ccb-UCX.yaml
(80-80), and
tests/scripts/perf/disaggregated/gb300_glm-5-fp4_8k1k_con1_ctx1_dep2_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
(75-75).
- Line 113: Update worker_config.ctx.disable_overlap_scheduler from false to
true in all ten affected files:
tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml:113-113,
tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_8k1k_con4096_ctx1_dep4_gen1_dep32_eplb256_mtp0_ccb-NIXL.yaml:114-114,
tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_8k1k_con4096_ctx1_dep4_gen1_dep32_eplb256_mtp0_ccb-UCX.yaml:112-112,
tests/scripts/perf/disaggregated/gb200_glm-5-fp4_8k1k_con1024_ctx1_dep4_gen1_dep8_eplb256_mtp1_ccb-NIXL.yaml:107-107,
tests/scripts/perf/disaggregated/gb200_glm-5-fp4_8k1k_con1024_ctx1_dep4_gen1_dep8_eplb256_mtp1_ccb-UCX.yaml:107-107,
tests/scripts/perf/disaggregated/gb300_glm-5-fp4_1k1k_con4096_ctx1_dep2_gen1_dep8_eplb256_mtp1_ccb-NIXL.yaml:98-98,
tests/scripts/perf/disaggregated/gb300_glm-5-fp4_1k1k_con512_ctx1_dep2_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml:95-95,
tests/scripts/perf/disaggregated/gb300_glm-5-fp4_8k1k_con1024_ctx1_dep2_gen1_dep8_eplb256_mtp1_ccb-NIXL.yaml:107-107,
tests/scripts/perf/disaggregated/gb300_glm-5-fp4_8k1k_con1024_ctx1_dep2_gen1_dep8_eplb256_mtp1_ccb-UCX.yaml:107-107,
and
tests/scripts/perf/disaggregated/gb300_glm-5-fp4_8k1k_con1_ctx1_dep2_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml:103-103.
Add the NVIDIA copyright header to each of these ten YAML files, using the
latest meaningful modification year.

In
`@tests/scripts/perf/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con4_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-NIXL.yaml`:
- Line 104: Disable overlap scheduling for all eleven affected disaggregated
context-worker profiles by setting worker_config.ctx.disable_overlap_scheduler
to true in
tests/scripts/perf/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con4_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-NIXL.yaml:104,
tests/scripts/perf/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con512_ctx1_tp1_gen1_dep2_eplb0_mtp0_ccb-NIXL.yaml:106,
tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_1k1k_con2048_ctx1_dep4_gen1_dep32_eplb0_mtp0_ccb-NIXL.yaml:95,
tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_1k1k_con4096_ctx1_dep4_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml:93,
tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_1k1k_con4_ctx1_dep4_gen1_tep4_eplb0_mtp0_ccb-NIXL.yaml:93,
tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb416_mtp3_ccb-NIXL.yaml:110,
tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp0_ccb-NIXL.yaml:102,
tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp0_ccb-UCX.yaml:102,
tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con4_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml:106,
tests/scripts/perf/disaggregated/gb200_qwen3-235b-fp4_8k1k_con1024_ctx1_tp1_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml:100,
and
tests/scripts/perf/disaggregated/gb200_qwen3-235b-fp4_8k1k_con1024_ctx1_tp1_gen1_dep8_eplb0_mtp0_ccb-UCX.yaml:100;
no additional coverage change is required.

In
`@tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb416_mtp3_ccb-NIXL.yaml`:
- Line 80: Update disable_overlap_scheduler to true for both gen and ctx in
tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb416_mtp3_ccb-NIXL.yaml:80-80
and
tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con4_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml:75-75.
Add NVIDIA copyright headers dated 2026 to both YAML files; do not rely on
accuracy validation because it is disabled.

In
`@tests/scripts/perf/disaggregated/gb300_glm-5-fp4_8k1k_con512_ctx1_dep2_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml`:
- Line 77: Add the NVIDIA copyright SPDX header at the top of the YAML file
containing disable_overlap_scheduler, using the year 2026. Leave the existing
configuration unchanged.
🪄 Autofix

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: 3c93226c-c011-4bcd-84d0-23d32980f6e7

📥 Commits

Reviewing files that changed from the base of the PR and between 5cc76ef and aa14ef0.

📒 Files selected for processing (79)
  • tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_128k8k_con128_ctx1_pp8_gen1_dep16_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_128k8k_con128_ctx1_pp8_gen1_dep16_eplb0_mtp1_ccb-UCX.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_128k8k_con1_ctx1_pp8_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_128k8k_con64_ctx1_pp8_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_128k8k_ctx1_pp4_gen8_pp4_bs2_eplb0_mtp0_con2-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_1k1k_con1024_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_1k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_1k1k_con3072_ctx1_dep4_gen1_dep4_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp1_ccb-UCX.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_8k1k_ctx1_gen3_tep8_bs32_eplb0_mtp0_con1_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_32k4k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-UCX.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_32k4k_con256_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_32k4k_con256_ctx1_dep8_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb256_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_8k1k_con4096_ctx1_dep4_gen1_dep32_eplb256_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_8k1k_con4096_ctx1_dep4_gen1_dep32_eplb256_mtp0_ccb-UCX.yaml
  • tests/scripts/perf/disaggregated/gb200_glm-5-fp4_8k1k_con1024_ctx1_dep4_gen1_dep8_eplb256_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_glm-5-fp4_8k1k_con1024_ctx1_dep4_gen1_dep8_eplb256_mtp1_ccb-UCX.yaml
  • tests/scripts/perf/disaggregated/gb200_glm-5-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_glm-5-fp4_8k1k_con512_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con1024_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con1024_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-UCX.yaml
  • tests/scripts/perf/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con128_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con4_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con512_ctx1_tp1_gen1_dep2_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_1k1k_con2048_ctx1_dep4_gen1_dep32_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_1k1k_con4096_ctx1_dep4_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_1k1k_con4_ctx1_dep4_gen1_tep4_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb416_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp0_ccb-UCX.yaml
  • tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con4_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_qwen3-235b-fp4_8k1k_con1024_ctx1_tp1_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_qwen3-235b-fp4_8k1k_con1024_ctx1_tp1_gen1_dep8_eplb0_mtp0_ccb-UCX.yaml
  • tests/scripts/perf/disaggregated/gb200_qwen3-235b-fp4_8k1k_con1_ctx1_tp1_gen1_tep4_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_qwen3-235b-fp4_8k1k_con64_ctx1_tp1_gen1_tep4_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_stress-gpt-oss-120b-fp4_8k1k_ctx1_tp1_gen1_tp4_eplb0_eagle3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_wideep_deepseek-r1-fp4_8k1k_ctx2_gen1_dep32_bs128_eplb288_mtp3_con1024_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_wideep_deepseek-v32-fp4_8k1k_ctx2_gen1_dep32_bs128_eplb288_mtp3_con1024_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb200_wideep_stress-deepseek-r1-fp4_8k1k_ctx2_gen1_dep32_bs128_eplb288_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-r1-fp4_128k8k_con1_ctx1_pp4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-r1-fp4_128k8k_con256_ctx1_pp4_gen1_dep8_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-r1-fp4_128k8k_con256_ctx1_pp4_gen1_dep8_eplb0_mtp1_ccb-UCX.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-r1-fp4_128k8k_con64_ctx1_pp4_gen1_dep16_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-r1-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-r1-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-r1-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-r1-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp1_ccb-UCX.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-v32-fp4_32k4k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-UCX.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-v32-fp4_32k4k_con256_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-v32-fp4_32k4k_con256_ctx1_dep8_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-v32-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb256_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-v32-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-v32-fp4_8k1k_con4096_ctx1_dep4_gen1_dep32_eplb256_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-v32-fp4_8k1k_con4096_ctx1_dep4_gen1_dep32_eplb256_mtp0_ccb-UCX.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con1229_ctx7_dep4_gen1_dep8_eplb384_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con2_ctx1_dep4_gen5_tep4_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con666_ctx8_dep4_gen1_dep32_eplb384_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_glm-5-fp4_1k1k_con1_ctx1_dep2_gen1_tep4_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_glm-5-fp4_1k1k_con4096_ctx1_dep2_gen1_dep8_eplb256_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_glm-5-fp4_1k1k_con512_ctx1_dep2_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_glm-5-fp4_8k1k_con1024_ctx1_dep2_gen1_dep8_eplb256_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_glm-5-fp4_8k1k_con1024_ctx1_dep2_gen1_dep8_eplb256_mtp1_ccb-UCX.yaml
  • tests/scripts/perf/disaggregated/gb300_glm-5-fp4_8k1k_con1_ctx1_dep2_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_glm-5-fp4_8k1k_con512_ctx1_dep2_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_kimi-k25-thinking-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb416_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_kimi-k25-thinking-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_kimi-k25-thinking-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp0_ccb-UCX.yaml
  • tests/scripts/perf/disaggregated/gb300_kimi-k25-thinking-fp4_8k1k_con4_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_wideep_deepseek-r1-fp4_8k1k_ctx2_gen1_dep32_bs128_eplb288_mtp3_con1024_ccb-NIXL.yaml
  • tests/scripts/perf/disaggregated/gb300_wideep_deepseek-v32-fp4_8k1k_ctx2_gen1_dep32_bs128_eplb288_mtp3_con1024_ccb-NIXL.yaml

Comment thread ...regated/gb200_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-NIXL.yaml
Comment thread ...regated/gb200_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-NIXL.yaml
Comment thread .../disaggregated/gb200_deepseek-v32-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
Comment thread .../disaggregated/gb200_deepseek-v32-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
Comment thread ...rf/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con4_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-NIXL.yaml
Comment thread ...ted/gb200_kimi-k25-thinking-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb416_mtp3_ccb-NIXL.yaml
Comment thread ...perf/disaggregated/gb300_glm-5-fp4_8k1k_con512_ctx1_dep2_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
@fredricz-20070104
disable overlap schedule for some of the cases for coverage reason, s…
b613e66 
…uggestion raised by LiZhi Zhou

Signed-off-by: FredricZ-2007 <226039983+fredricz-20070104@users.noreply.github.com>
@fredricz-20070104
Merge remote-tracking branch 'upstream/main' into feature/enable_over…
646f409 
…lap_schedule_on_perf_test

Signed-off-by: FredricZ-2007 <226039983+fredricz-20070104@users.noreply.github.com>

# Conflicts:
#	tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_128k8k_con128_ctx1_pp8_gen1_dep16_eplb0_mtp1_ccb-UCX.yaml
#	tests/scripts/perf/disaggregated/gb200_deepseek-r1-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp1_ccb-UCX.yaml
#	tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-UCX.yaml
#	tests/scripts/perf/disaggregated/gb200_deepseek-v32-fp4_8k1k_con4096_ctx1_dep4_gen1_dep32_eplb256_mtp0_ccb-UCX.yaml
#	tests/scripts/perf/disaggregated/gb200_glm-5-fp4_8k1k_con1024_ctx1_dep4_gen1_dep8_eplb256_mtp1_ccb-UCX.yaml
#	tests/scripts/perf/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con1024_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-UCX.yaml
#	tests/scripts/perf/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp0_ccb-UCX.yaml
#	tests/scripts/perf/disaggregated/gb200_qwen3-235b-fp4_8k1k_con1024_ctx1_tp1_gen1_dep8_eplb0_mtp0_ccb-UCX.yaml
#	tests/scripts/perf/disaggregated/gb300_deepseek-r1-fp4_128k8k_con256_ctx1_pp4_gen1_dep8_eplb0_mtp1_ccb-UCX.yaml
#	tests/scripts/perf/disaggregated/gb300_deepseek-r1-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp1_ccb-UCX.yaml
#	tests/scripts/perf/disaggregated/gb300_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-UCX.yaml
#	tests/scripts/perf/disaggregated/gb300_deepseek-v32-fp4_8k1k_con4096_ctx1_dep4_gen1_dep32_eplb256_mtp0_ccb-UCX.yaml
#	tests/scripts/perf/disaggregated/gb300_glm-5-fp4_8k1k_con1024_ctx1_dep2_gen1_dep8_eplb256_mtp1_ccb-UCX.yaml
#	tests/scripts/perf/disaggregated/gb300_kimi-k25-thinking-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp0_ccb-UCX.yaml
@fredricz-20070104
[None][test] Remove stale Kimi entries from llm_perf_disagg.yml
c50439c 
These two entries reference gb200/gb300 kimi-k25-thinking 8k1k con1024
configs with eplb0, but the actual config files use eplb416. The names
were never updated when the configs were added in NVIDIA#14960, so the entries
have silently matched nothing since then. llm_perf_multinode.txt already
carries the correct eplb416 names.

Signed-off-by: FredricZ-2007 <226039983+fredricz-20070104@users.noreply.github.com>
@fredricz-20070104
enable overlap schedule for both ci and qa side
d47550d 
Signed-off-by: FredricZ-2007 <226039983+fredricz-20070104@users.noreply.github.com>
@fredricz-20070104
fredricz-20070104 requested a review from a team as a code owner August 7, 2026 02:43
@fredricz-20070104
fredricz-20070104 enabled auto-merge (squash) August 7, 2026 02:43
@fredricz-20070104

fredricz-20070104 commented Aug 7, 2026

Copy link
Copy Markdown
Collaborator Author

/bot skip --comment "skip test as just modify some test configs, only run in qa side"

coderabbitai[bot]
coderabbitai Bot reviewed Aug 7, 2026
View reviewed changes

@coderabbitai coderabbitai Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

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-sanity/disaggregated/gb200_deepseek-v32-fp4_1k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml`:
- Line 66: Add the standard NVIDIA copyright header with year 2026 to each
affected YAML file:
tests/scripts/perf-sanity/disaggregated/gb200_deepseek-v32-fp4_1k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
(lines 66-66 and 102-102),
gb200_deepseek-v32-fp4_1k1k_con2048_ctx1_dep4_gen1_dep4_eplb0_mtp1_ccb-NIXL.yaml
(68-68 and 103-103),
gb200_deepseek-v32-fp4_32k4k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
(69-69 and 107-107),
gb200_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-NIXL.yaml
(74-74 and 112-112),
gb200_deepseek-v32-fp4_32k4k_con256_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
(70-70 and 107-107),
gb300_deepseek-v4-pro-fp4_8k1k_con4301_ctx12_dep4_gen1_dep8_eplb384_mtp1_ccb-NIXL.yaml
(171-171),
gb300_deepseek-v4-pro-fp4_8k1k_con666_ctx6_dep4_gen1_dep16_eplb384_mtp3_ccb-NIXL.yaml
(104-104),
gb300_deepseek-v4-pro-fp4_8k1k_con8_ctx1_dep4_gen4_tep8_eplb0_mtp3_ccb-NIXL.yaml
(106-106),
gb300_glm-5-fp4_1k1k_con1_ctx1_dep2_gen1_tep4_eplb0_mtp3_ccb-NIXL.yaml (65-65
and 94-94),
gb300_glm-5-fp4_1k1k_con4096_ctx1_dep2_gen1_dep8_eplb256_mtp1_ccb-NIXL.yaml
(70-70 and 98-98), and
gb300_glm-5-fp4_1k1k_con512_ctx1_dep2_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml (67-67
and 95-95). No test changes are needed.

In
`@tests/scripts/perf-sanity/disaggregated/gb200_deepseek-v32-fp4_32k4k_con256_ctx1_dep8_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml`:
- Line 74: Add the standard NVIDIA copyright header to the modified YAML files,
including the file containing disable_overlap_scheduler. Ensure
h200_nemotron-super-fp8_8k1k_con64_ctx1_tp2_gen1_tp2_eplb0_mtp0_ccb-UCX.yaml
includes the header with copyright year 2026.

In
`@tests/scripts/perf-sanity/disaggregated/gb200_glm-5-fp4_8k1k_con512_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml`:
- Line 67: Add the standard NVIDIA copyright header, including the latest
meaningful modification year, to all 11 affected YAML files:
tests/scripts/perf-sanity/disaggregated/gb200_glm-5-fp4_8k1k_con512_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
(67-67),
tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_1k1k_con2048_ctx1_tp1_gen1_dep2_eplb0_mtp0_ccb-NIXL.yaml
(70-70),
tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_1k1k_con512_ctx1_tp1_gen1_dep2_eplb0_mtp0_ccb-NIXL.yaml
(70-70),
tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_1k1k_con64_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-NIXL.yaml
(67-67),
tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con1024_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-NIXL.yaml
(108-108),
tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con128_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-NIXL.yaml
(67-67),
tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con4_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-NIXL.yaml
(67-67),
tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con512_ctx1_tp1_gen1_dep2_eplb0_mtp0_ccb-NIXL.yaml
(70-70),
tests/scripts/perf-sanity/disaggregated/gb200_kimi-k25-thinking-fp4_1k1k_con2048_ctx1_dep4_gen1_dep32_eplb0_mtp0_ccb-NIXL.yaml
(68-68),
tests/scripts/perf-sanity/disaggregated/gb200_kimi-k25-thinking-fp4_1k1k_con4096_ctx1_dep4_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml
(67-67), and
tests/scripts/perf-sanity/disaggregated/gb200_kimi-k25-thinking-fp4_1k1k_con4_ctx1_dep4_gen1_tep4_eplb0_mtp0_ccb-NIXL.yaml
(66-66). No test changes are needed.
🪄 Autofix

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: e8f558d9-1ab7-41af-8577-d05140be21a4

📥 Commits

Reviewing files that changed from the base of the PR and between b613e66 and d47550d.

📒 Files selected for processing (91)
  • tests/integration/test_lists/qa/llm_perf_disagg.yml
  • tests/scripts/perf-sanity/disaggregated/b200_deepseek-r1-fp4_1k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/b200_deepseek-r1-fp4_1k1k_con2048_ctx1_dep4_gen1_dep8_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/b200_deepseek-r1-fp4_1k1k_con256_ctx1_dep4_gen1_dep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/b200_deepseek-r1-fp4_8k1k_con1536_ctx1_dep4_gen1_dep8_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/b200_deepseek-r1-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/b200_deepseek-r1-fp4_8k1k_con256_ctx1_dep4_gen1_dep8_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-r1-fp4_128k8k_con128_ctx1_pp8_gen1_dep16_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-r1-fp4_128k8k_con1_ctx1_pp8_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-r1-fp4_1k1k_con1024_ctx1_dep4_gen1_dep32_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-r1-fp4_1k1k_con1024_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-r1-fp4_1k1k_con1024_ctx1_dep4_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-r1-fp4_1k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-r1-fp4_1k1k_con2048_ctx2_dep4_gen1_dep16_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-r1-fp4_1k1k_con2048_ctx2_dep4_gen1_dep16_eplb288_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-r1-fp4_1k1k_con3072_ctx1_dep4_gen1_dep4_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-r1-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-r1-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-r1-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-v32-fp4_1k1k_con1024_ctx1_dep4_gen1_dep32_eplb256_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-v32-fp4_1k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-v32-fp4_1k1k_con2048_ctx1_dep4_gen1_dep4_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-v32-fp4_32k4k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-v32-fp4_32k4k_con256_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-v32-fp4_32k4k_con256_ctx1_dep8_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-v32-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb256_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-v32-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_deepseek-v32-fp4_8k1k_con4096_ctx1_dep4_gen1_dep32_eplb256_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_gemma-3-1b-bf16_1k1k_con256_ctx1_tp1_gen1_tp1_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_glm-5-fp4_1k1k_con1_ctx1_dep4_gen1_tep4_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_glm-5-fp4_1k1k_con4096_ctx1_dep4_gen1_dep8_eplb256_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_glm-5-fp4_1k1k_con512_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_glm-5-fp4_8k1k_con1024_ctx1_dep4_gen1_dep8_eplb256_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_glm-5-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_glm-5-fp4_8k1k_con512_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_1k1k_con2048_ctx1_tp1_gen1_dep2_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_1k1k_con512_ctx1_tp1_gen1_dep2_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_1k1k_con64_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con1024_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con128_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con4_ctx1_tp1_gen1_tp4_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_8k1k_con512_ctx1_tp1_gen1_dep2_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_kimi-k25-thinking-fp4_1k1k_con2048_ctx1_dep4_gen1_dep32_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_kimi-k25-thinking-fp4_1k1k_con4096_ctx1_dep4_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_kimi-k25-thinking-fp4_1k1k_con4_ctx1_dep4_gen1_tep4_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb416_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_kimi-k25-thinking-fp4_8k1k_con4_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_llama-3.1-8b-bf16_1k1k_con256_ctx1_tp1_gen1_tp1_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_qwen3-235b-fp4_8k1k_con1024_ctx1_tp1_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_qwen3-235b-fp4_8k1k_con1_ctx1_tp1_gen1_tep4_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb200_qwen3-235b-fp4_8k1k_con64_ctx1_tp1_gen1_tep4_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-r1-fp4_128k8k_con1_ctx1_pp4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-r1-fp4_128k8k_con256_ctx1_pp4_gen1_dep8_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-r1-fp4_128k8k_con64_ctx1_pp4_gen1_dep16_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-r1-fp4_1k1k_con1024_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-r1-fp4_1k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-r1-fp4_1k1k_con3072_ctx1_dep4_gen1_dep4_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-r1-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-r1-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-r1-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-v32-fp4_1k1k_con1024_ctx1_dep4_gen1_dep32_eplb256_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-v32-fp4_1k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-v32-fp4_1k1k_con2048_ctx1_dep4_gen1_dep4_eplb0_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-v32-fp4_32k4k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-v32-fp4_32k4k_con2048_ctx1_dep4_gen1_dep32_eplb288_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-v32-fp4_32k4k_con256_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-v32-fp4_32k4k_con256_ctx1_dep8_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-v32-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb256_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-v32-fp4_8k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-v32-fp4_8k1k_con4096_ctx1_dep4_gen1_dep32_eplb256_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con180_ctx3_dep4_gen1_dep32_eplb384_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con4301_ctx12_dep4_gen1_dep8_eplb384_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con666_ctx6_dep4_gen1_dep16_eplb384_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_deepseek-v4-pro-fp4_8k1k_con8_ctx1_dep4_gen4_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_glm-5-fp4_1k1k_con1_ctx1_dep2_gen1_tep4_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_glm-5-fp4_1k1k_con4096_ctx1_dep2_gen1_dep8_eplb256_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_glm-5-fp4_1k1k_con512_ctx1_dep2_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_glm-5-fp4_8k1k_con1024_ctx1_dep2_gen1_dep8_eplb256_mtp1_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_glm-5-fp4_8k1k_con1_ctx1_dep2_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_glm-5-fp4_8k1k_con512_ctx1_dep2_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_kimi-k25-thinking-fp4_1k1k_con2048_ctx1_dep4_gen1_dep32_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_kimi-k25-thinking-fp4_1k1k_con4096_ctx1_dep4_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_kimi-k25-thinking-fp4_1k1k_con4_ctx1_dep4_gen1_tep4_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_kimi-k25-thinking-fp4_8k1k_con1024_ctx1_dep4_gen1_dep32_eplb416_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_kimi-k25-thinking-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp0_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/gb300_kimi-k25-thinking-fp4_8k1k_con4_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
  • tests/scripts/perf-sanity/disaggregated/h200_nemotron-super-fp8_8k1k_con64_ctx1_tp2_gen1_tp2_eplb0_mtp0_ccb-UCX.yaml
  • tests/scripts/perf-sanity/disaggregated/h200_qwen3-235b-a22b-fp8_8k1k_con512_ctx1_tp2_gen1_tep4_eplb0_mtp0_ccb-DEFAULT.yaml
  • tests/scripts/perf-sanity/disaggregated/h200_qwen3-32b-fp8_4k1k_con128_ctx1_tp1_gen1_tp2_eplb0_mtp0_ccb-DEFAULT.yaml
💤 Files with no reviewable changes (1)
  • tests/integration/test_lists/qa/llm_perf_disagg.yml

Comment thread .../disaggregated/gb200_deepseek-v32-fp4_1k1k_con1_ctx1_dep4_gen1_tep8_eplb0_mtp3_ccb-NIXL.yaml
Comment thread ...saggregated/gb200_deepseek-v32-fp4_32k4k_con256_ctx1_dep8_gen1_dep8_eplb0_mtp0_ccb-NIXL.yaml
Comment thread ...nity/disaggregated/gb200_glm-5-fp4_8k1k_con512_ctx1_dep4_gen1_dep32_eplb0_mtp3_ccb-NIXL.yaml
@tensorrt-cicd

tensorrt-cicd commented Aug 7, 2026

Copy link
Copy Markdown
Collaborator

PR_Github #64466 [ skip ] triggered by Bot. Commit: d47550d Link to invocation

@tensorrt-cicd

tensorrt-cicd commented Aug 7, 2026

Copy link
Copy Markdown
Collaborator

PR_Github #64466 [ skip ] completed with state SUCCESS. Commit: d47550d
Skipping testing for commit d47550d

Link to invocation

@fredricz-20070104
fredricz-20070104 merged commit a6ea52f into NVIDIA:main Aug 7, 2026
11 checks passed
@hyukn

hyukn commented Aug 11, 2026

Copy link
Copy Markdown
Collaborator

We can leave some ctx overlap scheduler disabled, because the benefit of overlap scheduler on ctx side depends on workload, right?

Correct. I will leave some of the perf sanity tests with the overlap scheduler disabled for the prefill part.

hyukn added a commit to hyukn/TensorRT-LLM that referenced this pull request Aug 11, 2026
@hyukn @claude
[None][test] Enable overlap scheduler in the remaining perf sanity co…
856d233 
…nfigs

NVIDIA#17390 re-enabled the overlap scheduler across the disaggregated perf sanity
YAMLs in both tests/scripts/perf-sanity/ and tests/scripts/perf/, which
supersedes the config-only change this branch originally carried. Four places
outside those YAMLs were missed and still force the scheduler off, so the cases
they drive keep measuring a runtime path nothing else exercises any more.

Aggregated bench configs (pytorch_model_config.py):
- deepseek_v4_pro_dspark
- gpt_oss_120b_eagle3_throughput-bench-pytorch, whose latency sibling
  gpt_oss_120b_eagle3-bench-pytorch already runs with the scheduler enabled, so
  the two variants disagreed.

Prefill side (test_perf_sanity.py): the ctx_only benchmark mode overwrote the
flag with True after reading the ctx worker config, which silently defeated the
ctx entries the YAML flip had just updated. Drop the override so the ctx worker
config decides. ServerConfig already defaults the flag to False when it is
absent, so configs that omit it are unchanged and a config that deliberately
sets it keeps its setting instead of being overridden.

Aggregated perf sanity YAML: the Dynamo-replica DeepSeek-V3.2 config was the
last file under tests/scripts/perf-sanity/ still pinned to true. It mirrors an
upstream Dynamo recipe, but the flag is a TRT-LLM runtime knob rather than part
of the model or accuracy setup it replicates.

Perf impact: these cases will re-baseline in OpenSearch, as the disaggregated
cases did after NVIDIA#17390. The affected stages do not gate on perf regression, so
no thresholds change.

Signed-off-by: Yukun He <23156053+hyukn@users.noreply.github.com>
Co-Authored-By: Claude <noreply@anthropic.com>
@hyukn hyukn mentioned this pull request Aug 11, 2026
Open
3 tasks
This was referenced Aug 28, 2026
Merged
Open
chenfeiz0326 added a commit to chenfeiz0326/TensorRT-LLM that referenced this pull request Sep 1, 2026
@chenfeiz0326
[https://nvbugs/6627795][fix] stop charging retiring requests against…
04fe30a 
… ADP admission and capacity

PR NVIDIA#17390 flipped `disable_overlap_scheduler` true->false (overlap ENABLED) on
several perf-sanity worker configs and cost
disagg-e2e-gb300_glm-5-fp4_8k1k_con1024_ctx1_dep2_gen1_dep8_eplb256_mtp1_ccb-NIXL
20.40% throughput.

With overlap enabled a finished request's teardown is deferred by one iteration:
`_process_previous_batch` -- the only thing that removes a finished request from
`PyExecutor.active_requests` -- runs ~200 lines AFTER `_fetch_new_requests` in
the same `_executor_loop_overlap` body. So requests in GENERATION_TO_COMPLETE
are still in the active list when the next batch is admitted, and were charged
against it three times over:

1. the ADP router balanced load on them, so `_expected_num_active_requests`
   floored `expected` at a phantom per-rank load and its heap filter then
   excluded the "loaded" rank entirely -- one rank idle every iteration;
2. `_pop_from_waiting_queue` spent global admission budget on them
   (`admission_capacity - total_num_active_requests`);
3. the C++ capacity scheduler counted them toward `mMaxNumRequests`: the
   `numAdmittedRequests >= mMaxNumRequests` break sits after the state gate and
   before classification, and `isGenerationInProgressState()` includes
   kGENERATION_TO_COMPLETE.

Charge 3 is the binding one, and it needs sequence-slot headroom to be
actionable, so all three are fixed together:

* `adp_router.py`: filter the retiring requests out of the active list once, in
  `gather_all_rank_states`, and route on that. One choke point corrects
  `num_active_requests` and `num_active_tokens` for all three routers and keeps
  `create_rank_state` overlap-agnostic. The count is reported in a new
  `RankState.num_retiring_requests` field.
* `py_executor.py`: fold that count back in for the idle-fetch liveness test
  only. Liveness is collective -- a rank reporting zero routable work would
  block on the untimed request-queue wait while its peers blocked in the
  broadcast, and end-of-run drain hits exactly that state. Also measure the
  dummy-request pad surplus against the routable count, so its warning does not
  fire every iteration.
* `scheduler.py`: `BindCapacityScheduler` now passes
  `no_schedule_after_state=GENERATION_TO_COMPLETE`, matching every micro-batch
  scheduler. The KV cache of a retiring request is released by the teardown
  that is already queued, so keeping it inside the capacity window bought
  nothing.
* `_util.py`: `should_enable_disagg_adp_overlap_headroom` ->
  `should_enable_adp_overlap_seq_slot_headroom`, no longer gated on
  disaggregation. The regression reproduced on an aggregated context-only run
  with no cache transceiver configured, and without the headroom the capacity
  change has no free slot to backfill into (it raises NoFreeSlotsError on a
  pool sized 1x max_batch_size).

Measured on the ctx worker of the regressing glm-5 case, four arms at
a6ea52f, matched nodes, no nsys, ADP-router tracing on all of them:

| arm                                   | tput     | vs bug  | fwd batch |
|---------------------------------------|----------|---------|-----------|
| overlap disabled (pre-NVIDIA#17390)         | 34672.51 | +25.8%  | 1.999     |
| overlap enabled (NVIDIA#17390, the bug)     | 27556.82 |    --   | 1.000     |
| + charges 1+2 only                    | 27635.17 |  +0.28% | 1.000     |
| + charges 1+2+3 and slot headroom     | 35502.08 | +28.8%  | 2.000     |

The fixed arm admits 2.00 requests/rank/iteration (the configured
max_batch_size) on 2559/2563 iterations, versus 0.75 for the bug, and finishes
the same 10240 requests in 2563 iterations instead of 6828 -- slightly ahead of
the overlap-disabled arm, so the regression is recovered rather than merely
reduced. Run-to-run spread on this rig is +/-0.3%.

Follow-up, deliberately not in this change: `batch_size_input =
len(self.active_requests)` feeding `drafter.get_draft_len_for_batch_size` is
reachable only with spec-dec plus an explicit `draft_len_schedule` and has the
same staleness.

Signed-off-by: Chenfei Zhang <chenfeiz@nvidia.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@coderabbitai coderabbitai[bot] coderabbitai[bot] left review comments
@reasonsolo reasonsolo reasonsolo approved these changes
@chienchunhung chienchunhung chienchunhung approved these changes
@Shixiaowei02 Shixiaowei02 Shixiaowei02 approved these changes
@ruodil ruodil ruodil approved these changes
@chenfeiz0326 chenfeiz0326 chenfeiz0326 approved these changes
@dc3671 dc3671 Awaiting requested review from dc3671 dc3671 is a code owner automatically assigned from NVIDIA/trt-llm-perf-devs
@hyukn hyukn Awaiting requested review from hyukn hyukn is a code owner automatically assigned from NVIDIA/trt-llm-perf-devs
@bo-nv bo-nv Awaiting requested review from bo-nv bo-nv is a code owner automatically assigned from NVIDIA/trt-llm-disagg-devs
@LarryXFly LarryXFly Awaiting requested review from LarryXFly LarryXFly is a code owner automatically assigned from NVIDIA/trt-llm-qa-perf
@yufeiwu-nv yufeiwu-nv Awaiting requested review from yufeiwu-nv yufeiwu-nv is a code owner automatically assigned from NVIDIA/trt-llm-qa-perf

Assignees

@fredricz-20070104 fredricz-20070104

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

8 participants

@fredricz-20070104 @tensorrt-cicd @hyukn @reasonsolo @chienchunhung @Shixiaowei02 @ruodil @chenfeiz0326