Fix CUDA synchronization bottleneck in LCMScheduler (#9485)#13969
Open
Liauuu wants to merge 1 commit into
Open
Fix CUDA synchronization bottleneck in LCMScheduler (#9485)#13969Liauuu wants to merge 1 commit into
Liauuu wants to merge 1 commit into
Conversation
github-actions
Bot
added
size/S
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
Fixes #9485
This PR resolves the intermittent high latency and CUDA stream synchronization issue (
cudaMemcpyAsyncbottleneck) inLCMScheduler.Previously, indexing the CPU tensor
self.alphas_cumprodwith the GPU tensortimesteptriggered an internalaten::_local_scalar_densecall, causing a major performance overhead (up to 100ms+ during video generation loops).Changes
set_timestepsto maintain a CPU-side copy of timesteps (self.cpu_timesteps).stepandget_scalings_for_boundary_condition_discreteto utilizeself.cpu_timesteps[self.step_index]for indexing, ensuring efficientCPU tensor [CPU index]operations while keeping the original GPU timesteps for the model forwards.All 34 tests in
tests/schedulers/test_scheduler_lcm.pypassed successfully.