Conversation
Contributor
Author
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/4327
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
Bot
added
the
CLA Signed
This was referenced Apr 24, 2026
vkuzo
added
the
module: not user facing
vkuzo
commented
Apr 24, 2026
| assert observer.total_batches == 0 | ||
|
|
||
| @pytest.mark.skipif(not torch.cuda.is_available(), reason="Need CUDA available") | ||
| def test_observer_tensor_attributes(self): |
Contributor
Author
There was a problem hiding this comment.
this test is not that useful, deleting instead of updating with new contents of the observer tensor
This was referenced Apr 24, 2026
This was referenced Apr 24, 2026
jerryzh168
reviewed
Apr 24, 2026
| return torch.Tensor._make_wrapper_subclass(cls, shape, **kwargs) # type: ignore[attr-defined] | ||
|
|
||
| def __init__(self, hp_data: torch.Tensor, total_batches: int, hessian=None): | ||
| def __init__(self, hp_data: torch.Tensor, total_batches, hessian=None): |
Contributor
There was a problem hiding this comment.
can you add some docs for total_batches?
Contributor
Author
There was a problem hiding this comment.
sure, let me do that in a future PR (planning today), as we also need to change the definition of this for grouped_mm to sample each token equally instead of sampling each batch equally
jerryzh168
approved these changes
Apr 24, 2026
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
2 participants
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:
Extend GPTQ coverage for bmm, formulating the bmm as a 3d case of mm.
This involves:
hessian), modify it to instead use E K by K Hessians for an
E, N, Kinput shape, route to the 2D hessian logic E times. This is slow
but we can optimize later.
We test numerical correctness by bitwise matching E 2d hessian
calculations to the 3D one.
Test Plan: