[quant] Support quantization of embedding lookup operators #44207
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
Summary: Use existing embedding_bag operator but set offsets to [0, 1, .. len(indices)] Test Plan: python test/test_quantization.py TestEmbeddingOps.test_embedding_byte Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
💊 CI failures summary and remediationsAs of commit 3e137ea (more details on the Dr. CI page):
ci.pytorch.org: 1 failedThis comment was automatically generated by Dr. CI (expand for details).Follow this link to opt-out of these comments for your Pull Requests.Please report bugs/suggestions on the GitHub issue tracker or post in the (internal) Dr. CI Users group. This comment has been revised 12 times. |
Summary: Use existing embedding_bag operator but set offsets to [0, 1, .. len(indices)] Test Plan: python test/test_quantization.py TestEmbeddingOps.test_embedding_byte Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
vkuzo
approved these changes
Sep 5, 2020
| qresult = quant_op(packed_weight, indices, sparse=False) | ||
|
|
||
| ref = torch.embedding(weights, indices, padding_idx=-1, scale_grad_by_freq=False, sparse=False) | ||
| torch.testing.assert_allclose(ref, qresult, atol=0.005, rtol=1e-3) |
Contributor
There was a problem hiding this comment.
are the modified atol and rtol values only because of quantization error?
| packed_weight = prepack_op(qweight) | ||
| qresult = quant_op(packed_weight, indices, sparse=False) | ||
|
|
||
| ref = torch.embedding(weights, indices, padding_idx=-1, scale_grad_by_freq=False, sparse=False) |
Contributor
There was a problem hiding this comment.
If you use qweight.dequantize(), this should be an exact match
Summary: Use existing embedding_bag operator but set offsets to [0, 1, .. len(indices)] Test Plan: python test/test_quantization.py TestEmbeddingOps.test_embedding_byte Reviewers: Subscribers: Tasks: Tags: Differential Revision: [D23547385](https://our.internmc.facebook.com/intern/diff/D23547385) [ghstack-poisoned]
Codecov Report
@@ Coverage Diff @@
## gh/supriyar/173/base #44207 +/- ##
=====================================================
Coverage 69.24% 69.24%
=====================================================
Files 381 381
Lines 47573 47573
=====================================================
Hits 32943 32943
Misses 14630 14630 Continue to review full report at Codecov.
|
Summary: Use existing embedding_bag operator but set offsets to [0, 1, .. len(indices)] Test Plan: python test/test_quantization.py TestEmbeddingOps.test_embedding_byte Reviewers: Subscribers: Tasks: Tags: Differential Revision: [D23547385](https://our.internmc.facebook.com/intern/diff/D23547385) [ghstack-poisoned]
Contributor
|
This pull request has been merged in 6013a29. |
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.
Stack from ghstack:
Summary:
Use existing embedding_bag operator but set offsets to [0, 1, .. len(indices)]
Test Plan:
python test/test_quantization.py TestEmbeddingOps.test_embedding_byte
Reviewers:
Subscribers:
Tasks:
Tags:
Differential Revision: D23547385