Conversation
* reasoning guided decoder Co-authored-by: Shang-Pin Sheng <shang-pin@deepinfra.com>
juney-nvidia
requested review from
syuoni
and removed request for
shaharmor98
|
@syuoni Hi Enwei, can you help review this PR? Thanks |
juney-nvidia
added
Community want to contribute
syuoni
left a comment
syuoni
left a comment
There was a problem hiding this comment.
Thanks @Shang-Pin for the contribution!
Overall the PR looks good to me. Please take a look at my comments, thanks!
BTW, since TRT-LLM has integrated xgrammar's new structural tag API (#7893), the guide starting control can be equivalently achieved by customizing the structural tags.
| /// @brief The detailed guide string. It could be a json schema, a regular expression or a EBNF grammar depending on | ||
| /// mGuideType. | ||
| std::optional<std::string> mGuide; | ||
| std::optional<std::int32_t> mGuidanceStartTokenId; |
There was a problem hiding this comment.
NIT: Rename mGuidanceStartTokenId to mGuideStartTokenId, to be consistent with other fields.
| pass | ||
|
|
||
| @abstractmethod | ||
| def guidance_started(self) -> bool: |
There was a problem hiding this comment.
NIT: Rename guidance_started to is_guide_started
| return True | ||
|
|
||
|
|
||
| class GrammarMatcherWrapper(GrammarMatcher): |
There was a problem hiding this comment.
GrammarMatcher is already a wrapper.
I think it might be an overkill to introduce another wrapper layer just for the guidance starting control. How about extending GrammarMatcher, making it optionally enable the guidance starting control?
|
Hi @Shang-Pin , please use pre-commit to reformat the changed files. See https://github.com/NVIDIA/TensorRT-LLM/blob/0252cee4c39398dc622531e80176fb719928be75/CONTRIBUTING.md#coding-style |
|
Thanks @Shang-Pin for the PR! We should also consider a sequence of tokens instead of a single trigger token to support the openai-harmony format for models like |
I believe this (a sequence of trigger tokens) is already supported if we use structural tag API (#7893), see xgrammar's doc: https://xgrammar.mlc.ai/docs/tutorials/structural_tag.html I'd prefer using structural tag, because it offloads everything to the grammar backends. Otherwise, if we implement the logics in TRT-LLM, I could imagine the complexity when interacting with speculative decoding. |
4 participants
@coderabbitai summary
Description
This change adds a guidance_start_token_id in the openai protocol, which allows the guided decoder to wait after seeing this token to skip guided decoding for thinking tokens.
Test Coverage
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.
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.