[ROCM] Added command buffers support for convolutions 2nd attempt #34572
+321
−46
Conversation
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
added UTs and convolution command test fixes added rebase fixes capture only those convolution targets which are explictly Revert "adding coll permute and convolution to command buffers" This reverts commit 75847e6. added conv to command buffers fixing build and test fixing build rewritten ConvolutionCmd, adapted command_buffer_conv_pass some cosmetics
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Outdated suggestions cannot be applied.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.
Added CommandBuffer support for Convolution ops
Graph capture of convolutions is by default disabled (since not all convolutions can be captured).
It can be enabled through a newly added flag: '--xla_gpu_enable_command_buffers=+convolution"
🎯 Justification
This op wase missing for whatever reason: this results in graph fragmentation especially for large models. Hence one gets several (sometimes many) execution graphs instead of just one.
🚀 Kind of Contribution
✨ New Feature
🧪 Unit Tests:
Added new subtest to xla/service/gpu/transforms/command_buffer_scheduling_test.cc and xla/backends/gpu/runtime/command_buffer_conversion_pass_test.cc
This is a fork of the original PR which was reverted due to conv problems: #32053
Now convolutions capture is by default disabled.
@beckerhe , @ezhulenev, @dimitar-asenov please have another look !