[None][feat] add in-process NeMo-Skills benchmarks and Nemotron-3-Super guards#15608
Draft
Wanli-Jiang wants to merge 1 commit into
Draft
[None][feat] add in-process NeMo-Skills benchmarks and Nemotron-3-Super guards#15608Wanli-Jiang wants to merge 1 commit into
Wanli-Jiang wants to merge 1 commit into
Conversation
…er guards
Add NeMo-Skills benchmarks as in-process trtllm-eval subcommands (gpqa_ns,
ifbench, scicode_ns, hle_aa): the model is loaded once into the in-process LLM
and each bench generates through it, then grades with NeMo-Skills' own graders.
Includes the install kit (requirements_nemo_skills.txt + install_nemo_skills.sh,
which installs the lib into the current env by default) and NEMO_SKILLS_EVAL.md.
Add four accuracy guards to TestNemotronV3Super that run these benches on the
NVFP4 model (TP4/EP4, reasoning on) and assert accuracy >= a threshold. Measured
in-test:
- test_nvfp4_nemo_skills_gpqa 68.2% -> guard 60
- test_nvfp4_nemo_skills_ifbench 67.0% -> guard 60
- test_nvfp4_nemo_skills_scicode problem 3.08% / subtask 30.6% -> run-guard (>=0)
- test_nvfp4_nemo_skills_hle 12.0% self-judge -> guard 7
The guards importorskip('nemo_skills') and skip (with a setup hint pointing at
install_nemo_skills.sh) when the lib / datasets / grader assets are absent, so
they run only where the eval infra is provisioned.
Signed-off-by: Wanli Jiang <35160485+Wanli-Jiang@users.noreply.github.com>
Wanli-Jiang
force-pushed
the
user/williamj/trtllm-eval-nemo-skills
branch
from
6067b32 to
c267474
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Features
Add NeMo-Skills benchmarks as in-process trtllm-eval subcommands (gpqa_ns, ifbench, scicode_ns, hle_aa): the model is loaded once into the in-process LLM and each bench generates through it, then grades with NeMo-Skills' own graders. Includes the install kit (requirements_nemo_skills.txt + install_nemo_skills.sh, which installs the lib into the current env by default) and NEMO_SKILLS_EVAL.md.
Add four accuracy guards to TestNemotronV3Super that run these benches on the NVFP4 model (TP4/EP4, reasoning on) and assert accuracy >= a threshold. Measured in-test:
The guards importorskip('nemo_skills') and skip (with a setup hint pointing at install_nemo_skills.sh) when the lib / datasets / grader assets are absent, so they run only where the eval infra is provisioned.
@coderabbitai summary
Description
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)
If PR introduces API changes, an appropriate PR label is added - either
api-compatibleorapi-breaking. Forapi-breaking, includeBREAKINGin the PR title.Any new dependencies have been scanned for license and vulnerabilities
CODEOWNERS updated if ownership changes
Documentation updated as needed
Update tava architecture diagram if there is a significant design change in PR.
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
To see a list of available CI bot commands, please comment
/bot help.