perf: run LTXVDilateVideoMask pooling on GPU for massive speedup#514
Open
Denarius40 wants to merge 1 commit into
Open
perf: run LTXVDilateVideoMask pooling on GPU for massive speedup#514Denarius40 wants to merge 1 commit into
Denarius40 wants to merge 1 commit into
Conversation
Max-pooling ops fail when the input tensor is on CPU but ComfyUI is running on CUDA. Move the mask to the active torch device before the pooling passes and return it to CPU afterwards so downstream nodes receive a CPU tensor as expected by ComfyUI conventions.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
LTXVDilateVideoMaskwas runningF.max_pool2d/F.max_pool1don the CPU regardless of whether a GPU was available. This change moves the mask to the active torch device before the pooling passes (viacomfy.model_management.get_torch_device()) and returns it to CPU afterwards.Performance
In local testing this yields a ~500x speedup when a GPU is available, since the max-pool operations are dramatically faster on CUDA than on CPU for typical video mask sizes.