[pt2e] Avoid getting model device once per node #2695
Merged
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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/2695
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit cd57f54 with merge base 183068e ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
bot
added
the
CLA Signed
andrewor14
added
the
topic: improvement
andrewor14
force-pushed
the
pt2e-cache-model-device
branch
from
4a08e16 to
4040192
Compare
jerryzh168
approved these changes
Aug 5, 2025
jerryzh168
reviewed
Aug 5, 2025
andrewor14
force-pushed
the
pt2e-cache-model-device
branch
from
4040192 to
ae3264e
Compare
andrewor14
force-pushed
the
pt2e-cache-model-device
branch
from
ae3264e to
12a1c44
Compare
jerryzh168
reviewed
Sep 3, 2025
| model.graph, | ||
| "_scale", | ||
| scale, | ||
| scale.device, |
There was a problem hiding this comment.
seems like scale could be non-Tensor as well, so here we'd need a device from model instead, same for zero_point
**Summary:** Previously, we call `assert_and_get_unqiue_device` once per node in both prepare and convert. This is expensive and unnecessary since the model device is the same across all nodes, so we should just call this once in the beginning and reuse the same model device across all the nodes. torchao version of pytorch/pytorch#159901 Note: The prepare path is not completely done yet, since we are blocked on the pytorch PR on being merged. It's different from convert since it still calls utility functions from `torch.ao.quantization.fx`. **Test Plan:** ``` python test/quantization/pt2e/test_quantize_pt2e.py ```
andrewor14
force-pushed
the
pt2e-cache-model-device
branch
from
12a1c44 to
cd57f54
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.
Summary: Previously, we call
assert_and_get_unqiue_deviceonce per node in both prepare and convert. This is expensive and unnecessary since the model device is the same across all nodes, so we should just call this once in the beginning and reuse the same model device across all the nodes.torchao version of pytorch/pytorch#159901
Note: The prepare path is not completely done yet, since we are blocked on the pytorch PR on being merged. It's different from convert since it still calls utility functions from
torch.ao.quantization.fx.Test Plan: