[#12954][fix] AutoDeploy: Fix Gemma4 MoE config (disable multi_stream_moe, lower free_gpu_memory_fraction)

[suyoggupta]

Summary

• Disable multi_stream_moe to fix accuracy regression when combined with MLIR elementwise fusion and monolithic CUDA graphs (GSM8K 80% → 90%). Root cause tracked in [None][fix] multi_stream_moe + MLIR accuracy regression in monolithic CUDA graph decode path #12954.
• Lower free_gpu_memory_fraction from 0.8 to 0.4 to prevent OOM during piecewise CUDA graph capture on 80GB GPUs.

Config-only change — no code modifications.

Verified scores

Model: google/gemma-4-26B-A4B-it (bf16, single H100 80GB)

Benchmark	Score
MMLU	75%
GSM8K	90%

Test plan

• TRTLLM_ACCURACY_NO_REFERENCE=1 pytest tests/integration/defs/accuracy/test_llm_api_autodeploy.py::TestGemma4MoE::test_bf16 -s -v

🤖 Generated with Claude Code

Summary by CodeRabbit

• Chores
  • Updated model registry example configuration with adjusted GPU memory allocation parameters for KV cache operations.
  • Disabled a transform feature in the example configuration, pending resolution of an upstream dependency issue.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 2ea5c9b2-19a2-4f33-b751-801fc89b3bb6

📥 Commits

Reviewing files that changed from the base of the PR and between d68a94e and 2262183.

📒 Files selected for processing (1)

• examples/auto_deploy/model_registry/configs/gemma4_moe.yaml

---

📝 Walkthrough

Walkthrough

The PR updates a Gemma-4 MoE model configuration file by reducing the KV cache GPU memory fraction from 0.8 to 0.4, disabling the multi-stream MoE feature, and adding a TODO comment referencing a pending TensorRT-LLM issue.

Changes

Cohort / File(s)	Summary
Configuration Update examples/auto_deploy/model_registry/configs/gemma4_moe.yaml	Reduced kv_cache_config.free_gpu_memory_fraction to 0.4, disabled multi_stream_moe.enabled, and added TODO comment noting the feature remains disabled pending upstream resolution.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~2 minutes

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately and specifically describes the main changes: disabling multi_stream_moe and lowering free_gpu_memory_fraction in the Gemma4 MoE config, with proper issue reference [#12954] and [fix] type.
Description check	✅ Passed	The PR description includes a clear summary of changes with specific metrics, verified test results, and test plan details. It covers the what, why, and how of the fix.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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

[suyoggupta]

/bot skip --"autodeploy config only change"

[github-actions]

GitHub Bot Help

/bot [-h] ['run', 'kill', 'skip', 'reuse-pipeline'] ...

Provide a user friendly way for developers to interact with a Jenkins server.

Run /bot [-h|--help] to print this help message.

See details below for each supported subcommand.

Details

run [--reuse-test (optional)pipeline-id --disable-fail-fast --skip-test --stage-list "A10-PyTorch-1, xxx" --gpu-type "A30, H100_PCIe" --test-backend "pytorch, cpp" --add-multi-gpu-test --only-multi-gpu-test --disable-multi-gpu-test --post-merge --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx" --detailed-log --debug(experimental) --high-priority]

Launch build/test pipelines. All previously running jobs will be killed.

--reuse-test (optional)pipeline-id (OPTIONAL) : Allow the new pipeline to reuse build artifacts and skip successful test stages from a specified pipeline or the last pipeline if no pipeline-id is indicated. If the Git commit ID has changed, this option will be always ignored. The DEFAULT behavior of the bot is to reuse build artifacts and successful test results from the last pipeline.

--disable-reuse-test (OPTIONAL) : Explicitly prevent the pipeline from reusing build artifacts and skipping successful test stages from a previous pipeline. Ensure that all builds and tests are run regardless of previous successes.

--disable-fail-fast (OPTIONAL) : Disable fail fast on build/tests/infra failures.

--skip-test (OPTIONAL) : Skip all test stages, but still run build stages, package stages and sanity check stages. Note: Does NOT update GitHub check status.

--stage-list "A10-PyTorch-1, xxx" (OPTIONAL) : Only run the specified test stages. Examples: "A10-PyTorch-1, xxx". Note: Does NOT update GitHub check status.

--gpu-type "A30, H100_PCIe" (OPTIONAL) : Only run the test stages on the specified GPU types. Examples: "A30, H100_PCIe". Note: Does NOT update GitHub check status.

--test-backend "pytorch, cpp" (OPTIONAL) : Skip test stages which don't match the specified backends. Only support [pytorch, cpp, tensorrt, triton]. Examples: "pytorch, cpp" (does not run test stages with tensorrt or triton backend). Note: Does NOT update GitHub pipeline status.

--only-multi-gpu-test (OPTIONAL) : Only run the multi-GPU tests. Note: Does NOT update GitHub check status.

--disable-multi-gpu-test (OPTIONAL) : Disable the multi-GPU tests. Note: Does NOT update GitHub check status.

--add-multi-gpu-test (OPTIONAL) : Force run the multi-GPU tests in addition to running L0 pre-merge pipeline.

--post-merge (OPTIONAL) : Run the L0 post-merge pipeline instead of the ordinary L0 pre-merge pipeline.

--extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx" (OPTIONAL) : Run the ordinary L0 pre-merge pipeline and specified test stages. Examples: --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx".

--detailed-log (OPTIONAL) : Enable flushing out all logs to the Jenkins console. This will significantly increase the log volume and may slow down the job.

--debug (OPTIONAL) : Experimental feature. Enable access to the CI container for debugging purpose. Note: Specify exactly one stage in the stage-list parameter to access the appropriate container environment. Note: Does NOT update GitHub check status.

--high-priority (OPTIONAL) : Run the pipeline with high priority. This option is restricted to authorized users only and will route the job to a high-priority queue.

kill

kill

Kill all running builds associated with pull request.

skip

skip --comment COMMENT

Skip testing for latest commit on pull request. --comment "Reason for skipping build/test" is required. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.

reuse-pipeline

reuse-pipeline

Reuse a previous pipeline to validate current commit. This action will also kill all currently running builds associated with the pull request. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.

[suyoggupta]

/bot skip --comment "autodeploy config only change"

[tensorrt-cicd]

PR_Github #42822 [ skip ] triggered by Bot. Commit: 2262183 Link to invocation

[tensorrt-cicd]

PR_Github #42822 [ skip ] completed with state SUCCESS. Commit: 2262183
Skipping testing for commit 2262183

Link to invocation