[Quantization] Add "quantization_tag" as metadata to fx proxy #108764
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: In order to make sure that quantization_tag is preserved through second stage export, this PR adds it as a special metadata that should be preserved. Since quantization in export path will work on top of pre dispatch graph, subsequent post dispatch op decomposition, will decompose ops that quant workflow tagged. In order to make sure that the patterns identified by quantizer, remains identifiable, even after decompositions are applied, we must preserve "quantization_tag". This enables backend delegates, that quantized a model for specific backend, to be able to identify "quantized" patterns. Test Plan: metadata porting tests Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/108764
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit a2c8f01 with merge base 53acdb6 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
kimishpatel
added a commit
that referenced
this pull request
Sep 7, 2023
Summary: In order to make sure that quantization_tag is preserved through second stage export, this PR adds it as a special metadata that should be preserved. Since quantization in export path will work on top of pre dispatch graph, subsequent post dispatch op decomposition, will decompose ops that quant workflow tagged. In order to make sure that the patterns identified by quantizer, remains identifiable, even after decompositions are applied, we must preserve "quantization_tag". This enables backend delegates, that quantized a model for specific backend, to be able to identify "quantized" patterns. Test Plan: metadata porting tests Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 8aaae92 Pull Request resolved: #108764
|
@kimishpatel has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
|
I have a potentially dumb question: "How is this different from using source_fn to do pattern matching?" |
source_fn based pattern matching doesnt allow for arbitrary pattern. It is related to nn.Module or nn.function (functinal actually doesnt work). So if you have some fusion patterns it doesnt work. Plus this specific taag enables quant workflow to prot metadata for nodes that quantization adds |
tugsbayasgalan
approved these changes
Sep 20, 2023
jerryzh168
reviewed
Sep 25, 2023
| ) | ||
|
|
||
| from_node_to_tags = { | ||
| torch.ops.aten.adaptive_avg_pool2d.default: "BackendA_adaptive_avg_pool2d_0", |
There was a problem hiding this comment.
should we test metadata on get_attr nodes?
…oxy" Summary: In order to make sure that quantization_tag is preserved through second stage export, this PR adds it as a special metadata that should be preserved. Since quantization in export path will work on top of pre dispatch graph, subsequent post dispatch op decomposition, will decompose ops that quant workflow tagged. In order to make sure that the patterns identified by quantizer, remains identifiable, even after decompositions are applied, we must preserve "quantization_tag". This enables backend delegates, that quantized a model for specific backend, to be able to identify "quantized" patterns. Test Plan: metadata porting tests Reviewers: Subscribers: Tasks: Tags: Differential Revision: [D49056259](https://our.internmc.facebook.com/intern/diff/D49056259) [ghstack-poisoned]
|
@kimishpatel has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
jerryzh168
approved these changes
Oct 31, 2023
…oxy" Summary: In order to make sure that quantization_tag is preserved through second stage export, this PR adds it as a special metadata that should be preserved. Since quantization in export path will work on top of pre dispatch graph, subsequent post dispatch op decomposition, will decompose ops that quant workflow tagged. In order to make sure that the patterns identified by quantizer, remains identifiable, even after decompositions are applied, we must preserve "quantization_tag". This enables backend delegates, that quantized a model for specific backend, to be able to identify "quantized" patterns. Test Plan: metadata porting tests Reviewers: Subscribers: Tasks: Tags: Differential Revision: [D49056259](https://our.internmc.facebook.com/intern/diff/D49056259) [ghstack-poisoned]
kimishpatel
added a commit
that referenced
this pull request
Nov 1, 2023
Summary: In order to make sure that quantization_tag is preserved through second stage export, this PR adds it as a special metadata that should be preserved. Since quantization in export path will work on top of pre dispatch graph, subsequent post dispatch op decomposition, will decompose ops that quant workflow tagged. In order to make sure that the patterns identified by quantizer, remains identifiable, even after decompositions are applied, we must preserve "quantization_tag". This enables backend delegates, that quantized a model for specific backend, to be able to identify "quantized" patterns. Test Plan: metadata porting tests Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 90ee800 Pull Request resolved: #108764
|
@kimishpatel has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
|
@pytorchbot merge (Initiating merge automatically since Phabricator Diff has merged) |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
xuhancn
pushed a commit
to xuhancn/pytorch
that referenced
this pull request
Nov 7, 2023
…h#108764) Summary: In order to make sure that quantization_tag is preserved through second stage export, this PR adds it as a special metadata that should be preserved. Since quantization in export path will work on top of pre dispatch graph, subsequent post dispatch op decomposition, will decompose ops that quant workflow tagged. In order to make sure that the patterns identified by quantizer, remains identifiable, even after decompositions are applied, we must preserve "quantization_tag". This enables backend delegates, that quantized a model for specific backend, to be able to identify "quantized" patterns. Test Plan: metadata porting tests Reviewers: Subscribers: Tasks: Tags: Differential Revision: [D49056259](https://our.internmc.facebook.com/intern/diff/D49056259) Pull Request resolved: pytorch#108764 Approved by: https://github.com/tugsbayasgalan, https://github.com/jerryzh168
Skylion007
pushed a commit
to Skylion007/pytorch
that referenced
this pull request
Nov 14, 2023
…h#108764) Summary: In order to make sure that quantization_tag is preserved through second stage export, this PR adds it as a special metadata that should be preserved. Since quantization in export path will work on top of pre dispatch graph, subsequent post dispatch op decomposition, will decompose ops that quant workflow tagged. In order to make sure that the patterns identified by quantizer, remains identifiable, even after decompositions are applied, we must preserve "quantization_tag". This enables backend delegates, that quantized a model for specific backend, to be able to identify "quantized" patterns. Test Plan: metadata porting tests Reviewers: Subscribers: Tasks: Tags: Differential Revision: [D49056259](https://our.internmc.facebook.com/intern/diff/D49056259) Pull Request resolved: pytorch#108764 Approved by: https://github.com/tugsbayasgalan, https://github.com/jerryzh168
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 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.
Stack from ghstack (oldest at bottom):
Summary:
In order to make sure that quantization_tag is preserved through second
stage export, this PR adds it as a special metadata that should be
preserved.
Since quantization in export path will work on top of pre dispatch
graph, subsequent post dispatch op decomposition, will decompose ops
that quant workflow tagged. In order to make sure that the patterns
identified by quantizer, remains identifiable, even after decompositions
are applied, we must preserve "quantization_tag".
This enables backend delegates, that quantized a model for specific
backend, to be able to identify "quantized" patterns.
Test Plan:
metadata porting tests
Reviewers:
Subscribers:
Tasks:
Tags:
Differential Revision: D49056259