Skip to content

perf: Performance script tuning#2473

Open
guyueh1 wants to merge 1 commit into
NVIDIA-NeMo:mainfrom
guyueh1:perf_script_tuning
Open

perf: Performance script tuning#2473
guyueh1 wants to merge 1 commit into
NVIDIA-NeMo:mainfrom
guyueh1:perf_script_tuning

Conversation

@guyueh1
Copy link
Copy Markdown
Contributor

@guyueh1 guyueh1 commented May 12, 2026

What does this PR do ?

Performance script tuning

  1. use force_on_policy_ratio to skip prev_logprobs compute in dsv3 and qwen3-235b
  2. use fuse loss in qwen3-235b
  3. use flashinfer flags for dsv3 fp8

Issues

List issues that this PR closes (syntax):

Usage

  • You can potentially add a usage example below
# Add a code snippet demonstrating how to use this

Before your PR is "Ready for review"

Pre checks:

  • Make sure you read and followed Contributor guidelines
  • Did you write any new necessary tests?
  • Did you run the unit tests and functional tests locally? Visit our Testing Guide for how to run tests
  • Did you add or update any necessary documentation? Visit our Document Development Guide for how to write, build and test the docs.

Additional Information

  • ...

1. use force_on_policy_ratio to skip prev_logprobs compute in dsv3 and qwen3-235b
2. use fuse loss in qwen3-235b
3. use flashinfer flags for dsv3 fp8

Signed-off-by: Guyue Huang <guyueh@nvidia.com>
@guyueh1 guyueh1 requested review from a team as code owners May 12, 2026 15:24
@copy-pr-bot
Copy link
Copy Markdown

copy-pr-bot Bot commented May 12, 2026

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant