[TRTLLM-6481][fix] Fix deepseek r1 accuracy issue #6868
Conversation
📝 WalkthroughWalkthroughIn release_batch, registers a cleanup callback in an ExitStack that zeros all KV cache buffers by iterating kv_cache_manager.layer_offsets and calling kv_cache_manager.get_buffers(layer_idx).zero_() on exit. No public API or function signatures changed. Changes
Sequence Diagram(s)sequenceDiagram
participant Caller
participant ModelEngine
participant ExitStack
participant KVCacheMgr as KV Cache Manager
Caller->>ModelEngine: call release_batch(...)
ModelEngine->>ExitStack: enter context, register clean_up_kv_cache callback
ExitStack->>KVCacheMgr: (callback) for each layer_idx in layer_offsets -> get_buffers(layer_idx)
KVCacheMgr-->>ExitStack: buffers
ExitStack->>KVCacheMgr: buffers.zero_()
ExitStack-->>ModelEngine: exit (cleanup executed)
ModelEngine-->>Caller: return
Estimated code review effort🎯 2 (Simple) | ⏱️ ~7 minutes Suggested labels
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. 📜 Recent review detailsConfiguration used: .coderabbit.yaml 💡 Knowledge Base configuration:
You can enable these sources in your CodeRabbit configuration. 📒 Files selected for processing (1)
🚧 Files skipped from review as they are similar to previous changes (1)
✨ 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 |
![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)
tensorrt_llm/_torch/pyexecutor/model_engine.py (1)
775-775: Nit: capitalize and clarify the commentPrefer a concise, accurate comment.
- # clean up the content of kv cache as nan may be introduced in the warmup + # Zero the KV cache; NaNs may be introduced during warmup
📜 Review details
Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (1)
tensorrt_llm/_torch/pyexecutor/model_engine.py(1 hunks)
🧰 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:
tensorrt_llm/_torch/pyexecutor/model_engine.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:
tensorrt_llm/_torch/pyexecutor/model_engine.py
|
PR_Github #15153 [ run ] triggered by Bot |
|
PR_Github #15153 [ run ] completed with state |
pengbowang-nv
force-pushed
the
fix-nan-in-kvcache
branch
from
81fa4a9 to
c16dc5c
Compare
|
/bot run --add-multi-gpu-test --disable-fail-fast |
|
PR_Github #15207 [ run ] triggered by Bot |
|
/bot kill |
|
PR_Github #15262 [ kill ] triggered by Bot |
|
PR_Github #15207 [ run ] completed with state |
|
PR_Github #15262 [ kill ] completed with state |
Signed-off-by: Pengbo Wang <[email protected]>
pengbowang-nv
force-pushed
the
fix-nan-in-kvcache
branch
from
c16dc5c to
6742730
Compare
|
/bot run --add-multi-gpu-test --disable-fail-fast |
|
PR_Github #15364 [ run ] triggered by Bot |
|
PR_Github #15364 [ run ] completed with state |
Signed-off-by: Pengbo Wang <[email protected]>
Signed-off-by: Pengbo Wang <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
Signed-off-by: Pengbo Wang <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
Signed-off-by: Pengbo Wang <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
Signed-off-by: Pengbo Wang <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
Signed-off-by: Pengbo Wang <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
Signed-off-by: Pengbo Wang <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
Signed-off-by: Pengbo Wang <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
Signed-off-by: Pengbo Wang <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
Signed-off-by: Pengbo Wang <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
Signed-off-by: Pengbo Wang <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
3 participants
Summary by CodeRabbit
Bug Fixes
Performance
Description
The accuracy issue is caused by NaN in output. The NaN was introduced in warmup since we cannot ensure the request we used in warmup valid. A NaN in one KV cache block, even if this NaN should not be used at given seq length, may lead to further NaN in the output. In this commit we cleanup kvcache after warmup to avoid this in the first place.
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.
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)]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.