[benchmarks] Initialize moco and TPU experiments on XLA.
#7257
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
zpcore
reviewed
Jun 12, 2024
| device = ( | ||
| str(self.benchmark_experiment.get_device()) | ||
| if self.is_accelerator_tpu() else self.benchmark_experiment.accelerator) | ||
| if self.should_initialize_on_xla(): |
There was a problem hiding this comment.
Can you remind me why we didn't initialize on CUDA device for GPU experiment before? Thanks
There was a problem hiding this comment.
We actually do initialize them on CUDA device (i.e. self.benchmark_experiment.accelerator). The thing is that for moco we have to initialize the benchmark on XLA device (i.e. str(self.benchmark_experiment.get_device())).
vanbasten23
reviewed
Jun 17, 2024
benchmarks/torchbench_model.py
Outdated
| # Move the initialized model to XLA device. | ||
| if self.benchmark_experiment.xla: | ||
| # Move the initialized model to XLA device if it's not there already. | ||
| if self.benchmark_experiment.xla and not self.should_initialize_on_xla(): | ||
| # First, move the model and the inputs to CPU. | ||
| # This avoids having dupplicated data on CUDA. | ||
| if self.is_accelerator_cuda(): |
There was a problem hiding this comment.
you may want to do a rebase.
vanbasten23
approved these changes
Jun 17, 2024
ysiraichi
force-pushed
the
ysiraichi/init-moco-on-xla
branch
from
5b7f9b7 to
150941d
Compare
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Outdated suggestions cannot be applied.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.
Fix: #6083
This PR partially fixes
mocobenchmark by initializing it using the XLA device, instead of its accelerator. Alongside benchmark#2292, this should fixmocofor, at least, inference (training should be tested soon).cc @miladm @JackCaoG @vanbasten23 @zpcore @frgossen