[https://nvbugs/5401114][fix] Unwaive Gemma3 tests #6952
Conversation
📝 WalkthroughWalkthroughUpdated test waivers and test-suite test settings: waiver entries/paths in Changes
Sequence Diagram(s)sequenceDiagram
participant Runner as Test Runner
participant Test as test_ptp_quickstart_multimodal
participant CLI as Quickstart CLI
participant Model as Gemma runtime
Runner->>Test: run multimodal test for gemma-3-27b-it
Test->>CLI: build command (image_format, attention_backend, disable_kv_cache_reuse, --kv_cache_fraction=0.5, --max_seq_len=1024)
CLI->>Model: start with provided args
Model-->>CLI: initialize kv-cache using free_gpu_memory_fraction=0.5 / apply seq len
CLI-->>Test: exit status/results
Test-->>Runner: report result
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~15 minutes Possibly related PRs
Suggested reviewers
Tip 🔌 Remote MCP (Model Context Protocol) integration is now available!Pro plan users can now connect to remote MCP servers from the Integrations page. Connect with popular remote MCPs such as Notion and Linear to add more context to your reviews and chats. ✨ Finishing Touches
🧪 Generate unit tests
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
SupportNeed help? Create a ticket on our support page for assistance with any issues or questions. CodeRabbit Commands (Invoked using PR/Issue comments)Type Other keywords and placeholders
Status, Documentation and Community
|
|
/bot run --extra-stage "H100_PCIe-PyTorch-Post-Merge-1" |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (1)
tests/integration/test_lists/waives.txt (1)
234-234: Optional: include a short rationale alongside the bug linkAdding a brief reason (e.g., “OOM on L40S”) helps triage without opening nvbugs, similar to other entries that include inline reasons.
For example:
-full:L40S/accuracy/test_llm_api_pytorch.py::TestGemma3_1BInstruct::test_auto_dtype SKIP (https://nvbugs/5375620) +full:L40S/accuracy/test_llm_api_pytorch.py::TestGemma3_1BInstruct::test_auto_dtype SKIP (https://nvbugs/5375620 - OOM on L40S)
📜 Review details
Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
💡 Knowledge Base configuration:
- MCP integration is disabled by default for public repositories
- Jira integration is disabled by default for public repositories
- Linear integration is disabled by default for public repositories
You can enable these sources in your CodeRabbit configuration.
📒 Files selected for processing (1)
tests/integration/test_lists/waives.txt(1 hunks)
🧰 Additional context used
🧠 Learnings (1)
📚 Learning: 2025-07-28T17:06:08.621Z
Learnt from: moraxu
PR: NVIDIA/TensorRT-LLM#6303
File: tests/integration/test_lists/qa/examples_test_list.txt:494-494
Timestamp: 2025-07-28T17:06:08.621Z
Learning: In TensorRT-LLM testing, it's common to have both CLI flow tests (test_cli_flow.py) and PyTorch API tests (test_llm_api_pytorch.py) for the same model. These serve different purposes: CLI flow tests validate the traditional command-line workflow, while PyTorch API tests validate the newer LLM API backend. Both are legitimate and should coexist.
Applied to files:
tests/integration/test_lists/waives.txt
🔇 Additional comments (1)
tests/integration/test_lists/waives.txt (1)
234-234: LGTM: 1B waiver correctly scoped to L40S with updated nvbugs referenceScoping Gemma3_1BInstruct auto-dtype to L40S only aligns with the PR goal. The nvbugs link is present and consistent with the file’s conventions.
|
PR_Github #15455 [ run ] triggered by Bot |
|
PR_Github #15455 [ run ] completed with state |
brb-nv
force-pushed
the
user/brb/unwaive-gemma3-tests-main
branch
2 times, most recently
from
20dd67c to
1dc8136
Compare
|
/bot run --extra-stage "H100_PCIe-PyTorch-Post-Merge-1" |
There was a problem hiding this comment.
Actionable comments posted: 0
🔭 Outside diff range comments (1)
tests/integration/defs/test_e2e.py (1)
2276-2284: Confirm kv_cache_fraction wiring OK; waivers missing for Gemma3-27B on L40S — please add or confirmShort summary:
- quickstart_multimodal correctly consumes --kv_cache_fraction and passes it to setup_llm; setup_llm (quickstart_advanced) maps args.kv_cache_fraction → KvCacheConfig.free_gpu_memory_fraction.
- I could not find a waiver that skips gemma-3-27b-it on L40S in tests/integration/test_lists/waives.txt — Gemma3-27B tests are present elsewhere, so add a waiver if you intend to skip it on L40S.
Locations checked:
- examples/llm-api/quickstart_multimodal.py — sets args.disable_kv_cache_reuse = True and defaults args.kv_cache_fraction = 0.6, then calls setup_llm(args, ...) (consumes the arg).
- examples/llm-api/quickstart_advanced.py — defines --kv_cache_fraction and in setup_llm builds KvCacheConfig(free_gpu_memory_fraction=args.kv_cache_fraction, ...).
- tests/integration/defs/test_e2e.py — includes the gemma-3-27b-it param and the special Gemma3 VLM flags (image_format=pil, attention_backend=FLASHINFER, --disable_kv_cache_reuse, --kv_cache_fraction=0.5).
- tests/integration/test_lists/waives.txt — no entry matching gemma-3-27b-it or TestGemma3_27BInstruct::test_auto_dtype (ran repo-wide search; no matches).
- tests/integration/defs/test-db/l0_h100.yml — lists accuracy/test_llm_api_pytorch.py::TestGemma3_27BInstruct::test_auto_dtype (so the test exists in the database).
Action requested:
- If Gemma3-27B should be skipped on L40S, add the appropriate SKIP entry to tests/integration/test_lists/waives.txt (mirroring the existing format, e.g. full:L40S/...::TestGemma3_27BInstruct::test_auto_dtype SKIP (reason)). If skipping is not intended, no change required.
🧹 Nitpick comments (3)
tests/integration/defs/accuracy/test_llm_api_pytorch.py (1)
798-803: Avoid magic number: extract 0.5 into a named constantTo keep this aligned with other tests and avoid scattering literals, consider a module-level constant and refer to it here.
Apply this minimal change in-place:
- free_gpu_memory_fraction=0.5, + free_gpu_memory_fraction=KV_FREE_MEM_FRACTION,And add near the top of this file (outside the selected range):
KV_FREE_MEM_FRACTION = 0.5 # fraction of device memory reserved for KV cache in teststests/integration/defs/test_e2e.py (2)
2280-2284: Prefer using the module constant instead of a literalYou already define _MEM_FRACTION_50 = 0.5 at Line 44. Use it here to avoid hardcoding.
- cmd.append("--kv_cache_fraction=0.5") + cmd.append(f"--kv_cache_fraction={_MEM_FRACTION_50}")
1-2: Header year range likely needs 2025This file has been modified this year; consider updating the SPDX copyright line to include 2025 for consistency with other updated files.
Example:
-# SPDX-FileCopyrightText: Copyright (c) 2022-2024 NVIDIA CORPORATION & AFFILIATES. All rights reserved. +# SPDX-FileCopyrightText: Copyright (c) 2022-2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
📜 Review details
Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
💡 Knowledge Base configuration:
- MCP integration is disabled by default for public repositories
- Jira integration is disabled by default for public repositories
- Linear integration is disabled by default for public repositories
You can enable these sources in your CodeRabbit configuration.
📒 Files selected for processing (3)
tests/integration/defs/accuracy/test_llm_api_pytorch.py(1 hunks)tests/integration/defs/test_e2e.py(1 hunks)tests/integration/test_lists/waives.txt(1 hunks)
🚧 Files skipped from review as they are similar to previous changes (1)
- tests/integration/test_lists/waives.txt
🧰 Additional context used
📓 Path-based instructions (2)
**/*.py
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
**/*.py: Python code must target Python 3.8+
Python indentation: 4 spaces, no tabs
Maintain module namespace in imports (from package.subpackage import foo; then use foo.SomeClass())
Python file names use snake_case
Python class names use PascalCase
Python functions/methods and local variables use snake_case; variables starting with a number get k_ prefix (e.g., k_99th_percentile)
Global variables use G_ prefixed UPPER_SNAKE_CASE (e.g., G_MY_GLOBAL)
Constants use UPPER_SNAKE_CASE in Python
Avoid shadowing variables from outer scopes in Python
Initialize all externally visible members of a Python class in init
Prefer docstrings for interfaces used outside a file; comments for local code
Use Google-style docstrings for classes and functions (Sphinx-parsable)
Document attributes/variables inline with short docstrings
Avoid reflection when simple alternatives exist (e.g., prefer explicit parameters over dict(**locals()))
In try/except, catch the narrowest exceptions possible
For duck-typing with try/except, keep try body minimal and put logic in else
Files:
tests/integration/defs/test_e2e.pytests/integration/defs/accuracy/test_llm_api_pytorch.py
**/*.{cpp,cxx,cc,cu,h,hpp,hxx,hh,cuh,py}
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
Prepend NVIDIA copyright header (current year) to all source files
Files:
tests/integration/defs/test_e2e.pytests/integration/defs/accuracy/test_llm_api_pytorch.py
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (1)
- GitHub Check: Pre-commit Check
🔇 Additional comments (2)
tests/integration/defs/accuracy/test_llm_api_pytorch.py (1)
798-803: LGTM: adding free_gpu_memory_fraction=0.5 is the right mitigation for Gemma3-27B OOMsMatches the intent to constrain KV allocation while keeping reuse disabled and FlashInfer attention. This aligns with the e2e quickstart change for consistency across paths.
tests/integration/defs/test_e2e.py (1)
2276-2284: LGTM: pass --kv_cache_fraction=0.5 for gemma-3-27b-it quickstartThis mirrors the LLM API test’s KV fraction and the reuse-disable setting, reducing memory pressure while preserving correctness with the FLASHINFER backend.
|
PR_Github #15470 [ run ] triggered by Bot |
Signed-off-by: Balaram Buddharaju <[email protected]>
brb-nv
force-pushed
the
user/brb/unwaive-gemma3-tests-main
branch
from
1dc8136 to
cb0f020
Compare
|
/bot run --extra-stage "H100_PCIe-PyTorch-Post-Merge-1" |
There was a problem hiding this comment.
Actionable comments posted: 0
🔭 Outside diff range comments (1)
tests/integration/defs/accuracy/test_llm_api_pytorch.py (1)
796-803: Add missing L40S waiver entryThe L40S waiver for
TestGemma3_27BInstruct.test_auto_dtypewas not found intests/integration/test_lists/waives.txt. Please add the following line:+ full:L40S/accuracy/test_llm_api_pytorch.py::TestGemma3_27BInstruct::test_auto_dtypeAdditionally, verify that the E2E quickstart test for
gemma-3-27b-itintests/integration/defs/test_e2e.pyincludes both--kv_cache_fractionand--max_seq_lenflags.Files to update:
- tests/integration/test_lists/waives.txt: add the waiver entry above
- tests/integration/defs/test_e2e.py: ensure
gemma-3-27b-itCLI invocation contains--kv_cache_fractionand--max_seq_len
🧹 Nitpick comments (1)
tests/integration/defs/accuracy/test_llm_api_pytorch.py (1)
805-811: Attn backend choice looks right; drop redundantcuda_graph_config=Noneand sanity-checkmax_batch_size=128
- Using
attn_backend="FLASHINFER"here makes sense for Gemma3 custom masks.- Passing
cuda_graph_config=Noneis redundant (default is None). Consider removing for clarity.max_batch_size=128increases reserved memory. If not strictly needed for these accuracy tasks, consider lowering (e.g., 64) or add a brief comment justifying 128, to avoid pushing memory on borderline SKUs in other environments.Apply this minimal cleanup:
with LLM(self.MODEL_PATH, kv_cache_config=kv_cache_config, attn_backend="FLASHINFER", - cuda_graph_config=None, max_batch_size=128, max_seq_len=4096) as llm:
📜 Review details
Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
💡 Knowledge Base configuration:
- MCP integration is disabled by default for public repositories
- Jira integration is disabled by default for public repositories
- Linear integration is disabled by default for public repositories
You can enable these sources in your CodeRabbit configuration.
📒 Files selected for processing (3)
tests/integration/defs/accuracy/test_llm_api_pytorch.py(1 hunks)tests/integration/defs/test_e2e.py(1 hunks)tests/integration/test_lists/waives.txt(1 hunks)
🚧 Files skipped from review as they are similar to previous changes (2)
- tests/integration/defs/test_e2e.py
- tests/integration/test_lists/waives.txt
🧰 Additional context used
📓 Path-based instructions (2)
**/*.py
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
**/*.py: Python code must target Python 3.8+
Python indentation: 4 spaces, no tabs
Maintain module namespace in imports (from package.subpackage import foo; then use foo.SomeClass())
Python file names use snake_case
Python class names use PascalCase
Python functions/methods and local variables use snake_case; variables starting with a number get k_ prefix (e.g., k_99th_percentile)
Global variables use G_ prefixed UPPER_SNAKE_CASE (e.g., G_MY_GLOBAL)
Constants use UPPER_SNAKE_CASE in Python
Avoid shadowing variables from outer scopes in Python
Initialize all externally visible members of a Python class in init
Prefer docstrings for interfaces used outside a file; comments for local code
Use Google-style docstrings for classes and functions (Sphinx-parsable)
Document attributes/variables inline with short docstrings
Avoid reflection when simple alternatives exist (e.g., prefer explicit parameters over dict(**locals()))
In try/except, catch the narrowest exceptions possible
For duck-typing with try/except, keep try body minimal and put logic in else
Files:
tests/integration/defs/accuracy/test_llm_api_pytorch.py
**/*.{cpp,cxx,cc,cu,h,hpp,hxx,hh,cuh,py}
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
Prepend NVIDIA copyright header (current year) to all source files
Files:
tests/integration/defs/accuracy/test_llm_api_pytorch.py
🧠 Learnings (1)
📚 Learning: 2025-07-28T17:06:08.621Z
Learnt from: moraxu
PR: NVIDIA/TensorRT-LLM#6303
File: tests/integration/test_lists/qa/examples_test_list.txt:494-494
Timestamp: 2025-07-28T17:06:08.621Z
Learning: In TensorRT-LLM testing, it's common to have both CLI flow tests (test_cli_flow.py) and PyTorch API tests (test_llm_api_pytorch.py) for the same model. These serve different purposes: CLI flow tests validate the traditional command-line workflow, while PyTorch API tests validate the newer LLM API backend. Both are legitimate and should coexist.
Applied to files:
tests/integration/defs/accuracy/test_llm_api_pytorch.py
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (1)
- GitHub Check: Pre-commit Check
|
PR_Github #15475 [ run ] triggered by Bot |
|
PR_Github #15470 [ run ] completed with state |
|
PR_Github #15475 [ run ] completed with state |
|
/bot run --extra-stage "H100_PCIe-PyTorch-Post-Merge-1" |
|
PR_Github #15478 [ run ] triggered by Bot |
|
PR_Github #15478 [ run ] completed with state |
Signed-off-by: Balaram Buddharaju <[email protected]>
Signed-off-by: Balaram Buddharaju <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
Signed-off-by: Balaram Buddharaju <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
Signed-off-by: Balaram Buddharaju <[email protected]>
Signed-off-by: Balaram Buddharaju <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
Signed-off-by: Balaram Buddharaju <[email protected]>
Signed-off-by: Balaram Buddharaju <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
Description
This MR unwaives the tests for Gemma3 27B model. They pass for me locally on H100.
Also, waiving Gemma3 1B test only on L40S. The test must pass on other SKUs.
While Gemma3 27B would also fail if run on L40S, it would go OOM during weight-loading and shouldn't be run there anyway.
Update after CI:
Test pass in post-merge / pre-merge of CI for this MR.
Test Coverage
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.
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)]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 thestage-listparameter to access the appropriate container environment. Note: Does NOT update GitHub check status.For guidance on mapping tests to stage names, see
docs/source/reference/ci-overview.mdand the
scripts/test_to_stage_mapping.pyhelper.kill
killKill all running builds associated with pull request.
skip
skip --comment COMMENTSkip 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-pipelineReuse 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.
Summary by CodeRabbit
Tests
Chores