Compile forward function in baseline backward benchmarks #3300
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
Priya2698
force-pushed
the
pm/simplify_bwd_baselines
branch
from
c0d2230
to
fcb9363
Compare
|
!build |
Priya2698
requested review from
naoyam,
jacobhinkle,
kevinstephano and
liqiangxl
|
!build |
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Outdated suggestions cannot be applied.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.
Currently, the backward benchmark for
torch.compileare compiling theoutput.backward(grads)function.This PR changes the benchmarks to compile the corresponding forward function.
The benchmarking results are at: http://nv/emR
Earlier, I was observing similar performance for eager and torchcompile for these backward benchmarks.
The files
*_bwd.pyare changed in this PR. We observe improved performance with torchcompile as compared to eager now. Note, that, the results from weekly will have more data points than the CI pipeline.Question:
test_gelu_bwd_reduction.pyreduces the gradient of the input. I am not sure what would be the right function to compile here :/Thanks to @xwang233 for help with running the CI pipeline!