[None][fix] Fix KV cache recompute in draft_target spec decode #7348
Conversation
Signed-off-by: Mike Iovine <[email protected]>
📝 WalkthroughWalkthroughModelDrafter._create_draft_request_for_request updates the conditional logic to treat modes that don’t require KV-cache recomputation as generation paths, even when some tokens were accepted. A clarifying comment was added. No public signatures changed. Changes
Sequence Diagram(s)sequenceDiagram
autonumber
participant C as Caller
participant MD as ModelDrafter
participant SC as SpecConfig/DecMode
participant BE as Backend
C->>MD: _create_draft_request_for_request(req)
MD->>SC: Check needs_kv_cache_recompute()
alt No accepted tokens OR no KV recompute needed
note right of MD: Treat as generation path
MD->>BE: Build GenerationRequest
else Accepted tokens AND KV recompute needed
note right of MD: Use accepted-tokens path
MD->>BE: Build AcceptedTokensRequest
end
BE-->>MD: Draft request object
MD-->>C: Return request
Estimated code review effort🎯 2 (Simple) | ⏱️ ~10 minutes Possibly related PRs
Suggested reviewers
✨ 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 |
|
PR_Github #16891 [ run ] triggered by Bot |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 0
🧹 Nitpick comments (2)
tensorrt_llm/_torch/speculative/model_drafter.py (2)
112-117: Docstring now outdated with new call sitesWe now create generation requests even when some tokens were accepted (if no KV recompute is needed). Update the docstring accordingly.
- def _create_generation_request(self, request: LlmRequest, - input_tokens: Any) -> LlmRequest: - """Create a generation request when no tokens were accepted.""" + def _create_generation_request(self, request: LlmRequest, + input_tokens: Any) -> LlmRequest: + """Create a generation request for draft decoding. + + Used when either there are no accepted tokens, or the mode does not + require KV-cache recomputation for accepted tokens. + """
143-144: Comment contradictsget_draft_model_promptbehaviorWe only drop the first token for EAGLE3 (see
get_draft_model_prompt), not EAGLE2. Align the comment.- # Prepare a context request. We discard the first token and take - # the newly decoded one - this is the convention for EAGLE 2 and 3. + # Prepare a context request. For EAGLE3, discard the first token and + # take the newly decoded one. Other modes keep all tokens.
📜 Review details
Configuration used: Path: .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)
tensorrt_llm/_torch/speculative/model_drafter.py(1 hunks)
🧰 Additional context used
📓 Path-based instructions (3)
**/*.{cpp,cc,cxx,cu,py,h,hpp,hh,hxx,cuh}
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
Use spaces only; no tabs; indent by 4 spaces
Files:
tensorrt_llm/_torch/speculative/model_drafter.py
**/*.py
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
**/*.py: Code must target Python 3.8+
Indent with 4 spaces; no tabs
Keep module namespace on import; import the module, not individual names; use module.symbol
Python filenames use snake_case (e.g., some_file.py)
Class names use PascalCase
Function and method names use snake_case
Local variables use snake_case; if starting with a number, prefix with k_ (e.g., k_99th_percentile)
Global variables use UPPER_SNAKE with G_ prefix (e.g., G_MY_GLOBAL)
Constants use UPPER_SNAKE
Avoid shadowing variables from outer scopes
Initialize all externally visible members 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 trailing docstrings
Avoid reflection when simple alternatives exist (e.g., avoid dict(**locals()) patterns)
Limit except clauses to specific exceptions; avoid bare except
When duck-typing with try/except, keep try body minimal and use else for logic
Files:
tensorrt_llm/_torch/speculative/model_drafter.py
**/*.{cpp,cc,cxx,cu,h,hpp,hh,hxx,cuh,py}
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
Prepend NVIDIA copyright header (current year) to all source files
Files:
tensorrt_llm/_torch/speculative/model_drafter.py
🧬 Code graph analysis (1)
tensorrt_llm/_torch/speculative/model_drafter.py (2)
tensorrt_llm/llmapi/llm_args.py (3)
spec_dec_mode(397-404)spec_dec_mode(443-448)spec_dec_mode(551-556)tensorrt_llm/_torch/speculative/interface.py (1)
needs_kv_cache_recompute(65-71)
⏰ 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 (1)
tensorrt_llm/_torch/speculative/model_drafter.py (1)
148-151: Clarify comment, reformat condition & verify test coverage
- In
tensorrt_llm/_torch/speculative/model_drafter.pyaround lines 148–151, replace the existing comment and dangling paren with:# No tokens accepted - generation request. This only applies to speculation algorithms# that need to recompute KV cache for accepted tokens like eagle3.elif num_accepted_tokens == 0 or not self.spec_config.spec_dec_mode.needs_kv_cache_recompute():
# Generation path when either:# - no tokens were accepted, or# - the mode does NOT require KV-cache recompute for accepted tokens (e.g., DRAFT_TARGET).# Use chunked context only when some tokens were accepted AND recompute is required (e.g., EAGLE3).elif (num_accepted_tokens == 0or not self.spec_config.spec_dec_mode.needs_kv_cache_recompute()):- No existing tests reference the branch where `spec_dec_mode.is_draft_target()` (i.e. `needs_kv_cache_recompute()==False`) with `num_accepted_tokens>0`. Add or verify a unit test that exercises this “accepted tokens but no KV recompute” path for DRAFT_TARGET. </blockquote></details> </blockquote></details> </details> <!-- This is an auto-generated comment by CodeRabbit for review status -->
|
PR_Github #16891 [ run ] completed with state |
|
/bot run |
|
PR_Github #17374 [ run ] triggered by Bot |
|
PR_Github #17374 [ run ] completed with state |
|
/bot run --disable-fail-fast |
|
PR_Github #17404 [ run ] triggered by Bot |
Description
Algorithms like draft target do not need to recompute KV cache for accepted tokens (unlike eagle). This should improve the CUDA graph hit rate and overall performance.
Test Coverage
Existing draft_target tests.
PR Checklist
Please review the following before submitting your PR:
PR description clearly explains what and why. If using CodeRabbit's summary, please make sure it makes sense.
PR Follows TRT-LLM CODING GUIDELINES to the best of your knowledge.
Test cases are provided for new code paths (see test instructions)
Any new dependencies have been scanned for license and vulnerabilities
CODEOWNERS updated if ownership changes
Documentation updated as needed
The reviewers assigned automatically/manually are appropriate for the PR.
Please check this after reviewing the above items as appropriate for this PR.
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
Bug Fixes
Performance
Stability