fixed prelu fusing code that pre-maturely update the constant for multiply node #6876
Conversation
This file contains 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
jinjingforever
approved these changes
Sep 29, 2022
dydavidkim
added a commit
to CodeSmithDSMLProjects/tfjs
that referenced
this pull request
Sep 29, 2022
* Customize setTimeout (tensorflow#6694) If the setTimeout nesting level is greater than 5 and timeout is less than 4ms, timeout will be clamped to 4ms, which hurts the perf. A custom setTimeout is provided to mitigate the perf impact. BUG: tensorflow#6687 Co-authored-by: Na Li <[email protected]> * Upgrade windows BrowserStack chrome to 104 (tensorflow#6866) * webgpu: Disable importExternalTexture (tensorflow#6868) WebGPU Working Group recently found some problem with importExtenalTexture in spec, so we have to disable it temporarily. * Refactored Resizing Layer Unit Tests (#38) * Rescaling Preprocessing Layer Co-authored-by: David Kim (@koyykdy) <[email protected]> Brian Zheng (@Brianzheng123) <[email protected]> * PR issues resolved * linting and PR issues resolved Co-authored-by: Adam Lang (@AdamLang96) <[email protected]> Co-authored-by: (@Brianzheng123) <[email protected]> * initial implementation of image preprocessing: resizing layer, and associated unit tests. Comments and refactoring for image scaling layer * refactoring in computeOutputShape for image resizing layer * Unit tests for image resizing preprocessing layer expanded and refactored * refactored unit tests for resizing layer * Preprocessing-Resizing layer unit test expansion and refactoring. Co-authored-by: Adam Lang <@AdamLang96> ([email protected]) * cleaning up commit diffs * cleaning up commit diffs * PR commit suggestions accepted - code refactored to reflect changes * resizing layer unit test refactoring Co-authored-by: AdamLang96 <[email protected]> * Linting issue resolved: unused import statement culled (#39) * Rescaling Preprocessing Layer Co-authored-by: David Kim (@koyykdy) <[email protected]> Brian Zheng (@Brianzheng123) <[email protected]> * PR issues resolved * linting and PR issues resolved Co-authored-by: Adam Lang (@AdamLang96) <[email protected]> Co-authored-by: (@Brianzheng123) <[email protected]> * initial implementation of image preprocessing: resizing layer, and associated unit tests. Comments and refactoring for image scaling layer * refactoring in computeOutputShape for image resizing layer * Unit tests for image resizing preprocessing layer expanded and refactored * refactored unit tests for resizing layer * Preprocessing-Resizing layer unit test expansion and refactoring. Co-authored-by: Adam Lang <@AdamLang96> ([email protected]) * cleaning up commit diffs * cleaning up commit diffs * PR commit suggestions accepted - code refactored to reflect changes * resizing layer unit test refactoring * linting issues resolved: unusued import statement culled Co-authored-by: AdamLang96 <[email protected]> * Update jasmine_util.ts (tensorflow#6872) FIX * webgl: Fix NaN issue (tensorflow#6828) Fix tensorflow#6822 Problem 1： On some GPUs, even if a and b are both non-NaN, the value of isNaN in vec4 isNaN = min(vec4(isnan(a)) + vec4(isnan(b)), vec4(1.0)); are still larger than 0., which misleads all values become NAN. 2： After resolving NAN issue, the result is still incorrect. It seems that the isnan_custom is not well supported on the problem GPU. After switching back to builtin isnan, everything works well. Solution: Use the bool type bvec4 instead of float type vec4 to calculate isNaN to avoid the the float precision issue when comparing with zero. Meanwhile, add an env flag WEBGL2_ISNAN_CUSTOM to allow user to specify which isnan to use. * Upgrade nodejs to 18.7.0 (tensorflow#6863) * Upgrade nodejs to 18.7.0 * Fix hash table test string not passed as base64 * fixed prelu fusing code that pre-maturely neg the const on multiply (tensorflow#6876) Co-authored-by: RajeshT <[email protected]> * Update tfjs-layers/src/layers/preprocessing/image_resizing.ts Co-authored-by: Matthew Soulanille <[email protected]> Co-authored-by: Yang Gu <[email protected]> Co-authored-by: Na Li <[email protected]> Co-authored-by: Matthew Soulanille <[email protected]> Co-authored-by: AdamLang96 <[email protected]> Co-authored-by: Linchenn <[email protected]> Co-authored-by: Jiajia Qin <[email protected]> Co-authored-by: Ping Yu <[email protected]> Co-authored-by: RajeshT <[email protected]> Co-authored-by: Matthew Soulanille <[email protected]> Co-authored-by: Yang Gu <[email protected]> Co-authored-by: Na Li <[email protected]> Co-authored-by: Matthew Soulanille <[email protected]> Co-authored-by: AdamLang96 <[email protected]> Co-authored-by: Linchenn <[email protected]> Co-authored-by: Jiajia Qin <[email protected]> Co-authored-by: Ping Yu <[email protected]> Co-authored-by: RajeshT <[email protected]> Co-authored-by: Matthew Soulanille <[email protected]>
dydavidkim
added a commit
to CodeSmithDSMLProjects/tfjs
that referenced
this pull request
Sep 29, 2022
* Update jasmine_util.ts (tensorflow#6872) FIX * webgl: Fix NaN issue (tensorflow#6828) Fix tensorflow#6822 Problem 1： On some GPUs, even if a and b are both non-NaN, the value of isNaN in vec4 isNaN = min(vec4(isnan(a)) + vec4(isnan(b)), vec4(1.0)); are still larger than 0., which misleads all values become NAN. 2： After resolving NAN issue, the result is still incorrect. It seems that the isnan_custom is not well supported on the problem GPU. After switching back to builtin isnan, everything works well. Solution: Use the bool type bvec4 instead of float type vec4 to calculate isNaN to avoid the the float precision issue when comparing with zero. Meanwhile, add an env flag WEBGL2_ISNAN_CUSTOM to allow user to specify which isnan to use. * Upgrade nodejs to 18.7.0 (tensorflow#6863) * Upgrade nodejs to 18.7.0 * Fix hash table test string not passed as base64 * fixed prelu fusing code that pre-maturely neg the const on multiply (tensorflow#6876) Co-authored-by: RajeshT <[email protected]> Co-authored-by: Linchenn <[email protected]> Co-authored-by: Jiajia Qin <[email protected]> Co-authored-by: Matthew Soulanille <[email protected]> Co-authored-by: Ping Yu <[email protected]> Co-authored-by: RajeshT <[email protected]>
dydavidkim
added a commit
to CodeSmithDSMLProjects/tfjs
that referenced
this pull request
Sep 29, 2022
* Update jasmine_util.ts (tensorflow#6872) FIX * webgl: Fix NaN issue (tensorflow#6828) Fix tensorflow#6822 Problem 1： On some GPUs, even if a and b are both non-NaN, the value of isNaN in vec4 isNaN = min(vec4(isnan(a)) + vec4(isnan(b)), vec4(1.0)); are still larger than 0., which misleads all values become NAN. 2： After resolving NAN issue, the result is still incorrect. It seems that the isnan_custom is not well supported on the problem GPU. After switching back to builtin isnan, everything works well. Solution: Use the bool type bvec4 instead of float type vec4 to calculate isNaN to avoid the the float precision issue when comparing with zero. Meanwhile, add an env flag WEBGL2_ISNAN_CUSTOM to allow user to specify which isnan to use. * Upgrade nodejs to 18.7.0 (tensorflow#6863) * Upgrade nodejs to 18.7.0 * Fix hash table test string not passed as base64 * fixed prelu fusing code that pre-maturely neg the const on multiply (tensorflow#6876) Co-authored-by: RajeshT <[email protected]> Co-authored-by: Linchenn <[email protected]> Co-authored-by: Jiajia Qin <[email protected]> Co-authored-by: Matthew Soulanille <[email protected]> Co-authored-by: Ping Yu <[email protected]> Co-authored-by: RajeshT <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
fixed this issue
Our Prelu fusing logic prematurely update the constant value for multiply node when the node graph partially match the prelu logic.
To see the logs from the Cloud Build CI, please join either our discussion or announcement mailing list.
This change is