Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

WIP - RPP Tensor HIP kernels launch config regularization #343

Open
wants to merge 4 commits into
base: develop
Choose a base branch
from
Open

WIP - RPP Tensor HIP kernels launch config regularization #343

wants to merge 4 commits into from

Conversation

sampath1117
Copy link
Collaborator

  • Replaced instances of handle.GetBatchSize() with dstDescPtr->n for al
    l tensor HIP kernels
  • Modified globalThreads_x value to use only maxWidth instead of hStride for kernels using 24 pixel load and store

sampath1117 and others added 4 commits September 13, 2024 12:20
@sampath1117
replaced instances of handle.GetBatchSize() with dstDescPtr->n for al…
12c030c 
…l tensor HIP kernels
@sampath1117
modified globalThreads_x value to use only maxWidth instead of h stride
a416733
@dependabot
Docs - Bump rocm-docs-core[api_reference] from 1.7.2 to 1.8.1 in /doc…
6cc6713 
…s/sphinx (ROCm#452)

Bumps [rocm-docs-core[api_reference]](https://github.com/ROCm/rocm-docs-core) from 1.7.2 to 1.8.1.
- [Release notes](https://github.com/ROCm/rocm-docs-core/releases)
- [Changelog](https://github.com/ROCm/rocm-docs-core/blob/v1.8.1/CHANGELOG.md)
- [Commits](ROCm/rocm-docs-core@v1.7.2...v1.8.1)

---
updated-dependencies:
- dependency-name: rocm-docs-core[api_reference]
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <[email protected]>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
@sampath1117
Merge branch 'develop' into sr/kernel_launch_regularization
a9110db
@sampath1117
Copy link
Collaborator Author

This PR can be issued to Opensource once all image based trainings in rocAL are run with this PR and convergence, performance is intact

@sampath1117 sampath1117 changed the title WIP - RPP Tensor HIP kernels launch parameter regularization WIP - RPP Tensor HIP kernels launch parameters regularization Sep 23, 2024
@sampath1117 sampath1117 changed the title WIP - RPP Tensor HIP kernels launch parameters regularization WIP - RPP Tensor HIP kernels launch config regularization Sep 23, 2024
@sampath1117
Copy link
Collaborator Author

@r-abishek
Below are the image kernels that still use 8 pixel load and store for PKD3, PLN3 HIP variants

  • Brightness
  • Contrast
  • Exposure
  • Gamma correction
  • LUT
  • Blend
  • Crop
  • RICAP

@sampath1117
Copy link
Collaborator Author

@HazarathKumarM
Please track this PR

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

@r-abishek r-abishek Awaiting requested review from r-abishek

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
1 participant
@sampath1117