Skip to content

Channel dropout Updated Seed QA#547

Closed
RooseweltMcW wants to merge 2 commits intor-abishek:ar/dropout_channelfrom
RooseweltMcW:apr/channel_dropout_seedFix
Closed

Channel dropout Updated Seed QA#547
RooseweltMcW wants to merge 2 commits intor-abishek:ar/dropout_channelfrom
RooseweltMcW:apr/channel_dropout_seedFix

Conversation

@RooseweltMcW
Copy link
Copy Markdown

No description provided.

startCpuTime = clock();
if (BitDepthTestMode == U8_TO_U8 || BitDepthTestMode == F16_TO_F16 || BitDepthTestMode == F32_TO_F32 || BitDepthTestMode == I8_TO_I8)
rppt_channel_dropout_host(input, srcDescPtr, output, dstDescPtr, dropoutProbability, randomSeed, roiTensorPtrSrc, roiTypeSrc, handle);
errorCodeCapture = rppt_channel_dropout_host(input, srcDescPtr, output, dstDescPtr, dropoutProbability, seed, roiTensorPtrSrc, roiTypeSrc, handle);
Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@RooseweltMcW errorCodeCapture changes won't be present in the PR currently and will result in CI failure . please remove error code capture for now in this PR

Copy link
Copy Markdown
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done

* \retval RPP_ERROR* Unsuccessful completion.
*/
RppStatus rppt_channel_dropout_gpu(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t dstPtr, RpptDescPtr dstDescPtr, Rpp32f *dropoutProbability, bool randomSeed, RpptROIPtr roiTensorPtrSrc, RpptRoiType roiType, rppHandle_t rppHandle);
RppStatus rppt_channel_dropout_gpu(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t dstPtr, RpptDescPtr dstDescPtr, Rpp32f *dropoutProbability, Rpp32f seed, RpptROIPtr roiTensorPtrSrc, RpptRoiType roiType, rppHandle_t rppHandle);
Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Everywhere it must be Rpp32s instead of Rpp32f

@HazarathKumarM
Copy link
Copy Markdown
Collaborator

these changes not required as we are completely moving out the mask computation

ManasaDattaT pushed a commit to ManasaDattaT/rpp that referenced this pull request Dec 19, 2025
)

* Initial version for random channel permute for HOST

* Implimented random channel permute for remaining bit depth on HOST

* Initial concat implementation for HIp

* Removed random channel permute implimentation and merged with swap channel

* deleted random channel permute kernel

* Golden output added for all variant of swap channel

* Minor changes

* Minor changes

* Changed from simd to AVX

* Small changes in comments

* Optimized for pkd variant

* Fix Build errors and QA tests

* resolved review comments

* Add F32 bin files and fix Pkd3-Pkd3 QA tests

* Update CHANGELOG.md

* Modified swap channels as channel permute and modified permutation tensor computation

* minor changes

* Update CHANGELOG.md

* change permutation lists changes

* updated comments for better readability

* Update CHANGELOG.md

* Change swap to permute

---------

Co-authored-by: HazarathKumarM <hazarathkumar@multicorewareinc.com>
Co-authored-by: Dineshbabu-Ravichandran <dineshbabu.ravichandran@multicorewareinc.com>
Co-authored-by: Kiriti Gowda <kiritigowda@gmail.com>
RooseweltMcW pushed a commit to HazarathKumarM/rpp that referenced this pull request Jan 21, 2026
)

* Initial version for random channel permute for HOST

* Implimented random channel permute for remaining bit depth on HOST

* Initial concat implementation for HIp

* Removed random channel permute implimentation and merged with swap channel

* deleted random channel permute kernel

* Golden output added for all variant of swap channel

* Minor changes

* Minor changes

* Changed from simd to AVX

* Small changes in comments

* Optimized for pkd variant

* Fix Build errors and QA tests

* resolved review comments

* Add F32 bin files and fix Pkd3-Pkd3 QA tests

* Update CHANGELOG.md

* Modified swap channels as channel permute and modified permutation tensor computation

* minor changes

* Update CHANGELOG.md

* change permutation lists changes

* updated comments for better readability

* Update CHANGELOG.md

* Change swap to permute

---------

Co-authored-by: HazarathKumarM <hazarathkumar@multicorewareinc.com>
Co-authored-by: Dineshbabu-Ravichandran <dineshbabu.ravichandran@multicorewareinc.com>
Co-authored-by: Kiriti Gowda <kiritigowda@gmail.com>
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.

3 participants