Repository navigation
Time train phases with CUDA events instead of %globaltimer - #701
Open
sylvesterroos wants to merge 1 commit into
Open
sylvesterroos wants to merge 1 commit into
sylvesterroos wants to merge 1 commit into
Conversation
External event records time phases inside the captured train graph, replacing the inline-PTX stamp kernel that blocks HIP builds.
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
%globaltimerread insrc/pufferl.cuis the one line hipify cannot translate, so it blocks building the trainer for AMD GPUs.cudaEventRecordExternalduring stream capture becomes an event record node. It timestamps on every graph launch, socudaEventElapsedTimeworks inside the captured train graph. The removed comment holds for plaincudaEventRecord.cudagraphs = -1), the same events record with default flags.Profile.stamps(device buffer + D2H copy) becomescudaEvent_t train_ev[2 * NUM_TE], same per-slot layout.train_misc/train_modelreadcudaEventElapsedTimein ms directly.%globaltimersed rules match nothing and can be dropped.%globaltimerstamps with cudagraphs on and off, and losses are identical.