quant: switch observers to use min_max #42957
Closed
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
Summary: Switches observers to use the new min_max function to calculate min and max at the same time. We see around 45-50% speedup on representative input shapes on the microbenchmarks. Test Plan: CI for correctness performance: ``` cd benchmarks/operator_benchmark // repeat (before diff, after diff) x (cpu, cuda) python -m pt.qobserver_test --tag_filter all --device cpu ``` Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
This was referenced Aug 13, 2020
vkuzo
added a commit
that referenced
this pull request
Aug 13, 2020
Summary: Switches observers to use the new min_max function to calculate min and max at the same time. We see around 45-50% speedup on representative input shapes on the microbenchmarks. Test Plan: CI for correctness performance: ``` cd benchmarks/operator_benchmark // repeat (before diff, after diff) x (cpu, cuda) python -m pt.qobserver_test --tag_filter all --device cpu ``` Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: e0dedf621a555bd6f821427299ccce63618ac8d0 Pull Request resolved: #42957
💊 CI failures summary and remediationsAs of commit 6a60cbc (more details on the Dr. CI page):
Extra GitHub checks: 1 failed
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 31 times. |
supriyar
approved these changes
Aug 13, 2020
Summary:
Switches observers to use the new min_max function to calculate
min and max at the same time. We see around 45-50% speedup on
representative input shapes on the microbenchmarks for all observers except `HistogramObserver`.
Test Plan:
CI for correctness
performance:
```
cd benchmarks/operator_benchmark
// repeat (before diff, after diff) x (cpu, cuda)
python -m pt.qobserver_test --tag_filter all --device cpu
/*
* before, cpu: https://our.intern.facebook.com/intern/paste/P138633280/
* before, cuda: https://our.intern.facebook.com/intern/paste/P138639473/
* after, cpu: https://our.intern.facebook.com/intern/paste/P138635458/
* after, cuda: https://our.intern.facebook.com/intern/paste/P138636344/
*/
```
Reviewers:
Subscribers:
Tasks:
Tags:
Differential Revision: [D23093995](https://our.internmc.facebook.com/intern/diff/D23093995)
[ghstack-poisoned]
Summary:
Switches observers to use the new min_max function to calculate
min and max at the same time. We see around 45-50% speedup on
representative input shapes on the microbenchmarks for all observers except `HistogramObserver`.
Test Plan:
CI for correctness
performance:
```
cd benchmarks/operator_benchmark
// repeat (before diff, after diff) x (cpu, cuda)
python -m pt.qobserver_test --tag_filter all --device cpu
/*
* before, cpu: https://our.intern.facebook.com/intern/paste/P138633280/
* before, cuda: https://our.intern.facebook.com/intern/paste/P138639473/
* after, cpu: https://our.intern.facebook.com/intern/paste/P138635458/
* after, cuda: https://our.intern.facebook.com/intern/paste/P138636344/
*/
```
Reviewers:
Subscribers:
Tasks:
Tags:
Differential Revision: [D23093995](https://our.internmc.facebook.com/intern/diff/D23093995)
[ghstack-poisoned]
Summary:
Switches observers to use the new min_max function to calculate
min and max at the same time. We see around 45-50% speedup on
representative input shapes on the microbenchmarks for all observers except `HistogramObserver`.
Test Plan:
CI for correctness
performance:
```
cd benchmarks/operator_benchmark
// repeat (before diff, after diff) x (cpu, cuda)
python -m pt.qobserver_test --tag_filter all --device cpu
/*
* before, cpu: https://our.intern.facebook.com/intern/paste/P138633280/
* before, cuda: https://our.intern.facebook.com/intern/paste/P138639473/
* after, cpu: https://our.intern.facebook.com/intern/paste/P138635458/
* after, cuda: https://our.intern.facebook.com/intern/paste/P138636344/
*/
```
Reviewers:
Subscribers:
Tasks:
Tags:
Differential Revision: [D23093995](https://our.internmc.facebook.com/intern/diff/D23093995)
[ghstack-poisoned]
vkuzo
added a commit
that referenced
this pull request
Sep 2, 2020
Summary: Switches observers to use the new min_max function to calculate min and max at the same time. We see around 45-50% speedup on representative input shapes on the microbenchmarks. Test Plan: CI for correctness performance: ``` cd benchmarks/operator_benchmark // repeat (before diff, after diff) x (cpu, cuda) python -m pt.qobserver_test --tag_filter all --device cpu ``` Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 721eb176284cae1bd1aeb82dba9deaf37662c502 Pull Request resolved: #42957
Summary:
Switches observers to use the new min_max function to calculate
min and max at the same time. We see around 45-50% speedup on
representative input shapes on the microbenchmarks for all observers except `HistogramObserver`.
Test Plan:
CI for correctness
performance:
```
cd benchmarks/operator_benchmark
// repeat (before diff, after diff) x (cpu, cuda)
python -m pt.qobserver_test --tag_filter all --device cpu
/*
* before, cpu: https://our.intern.facebook.com/intern/paste/P138633280/
* before, cuda: https://our.intern.facebook.com/intern/paste/P138639473/
* after, cpu: https://our.intern.facebook.com/intern/paste/P138635458/
* after, cuda: https://our.intern.facebook.com/intern/paste/P138636344/
*/
```
Reviewers:
Subscribers:
Tasks:
Tags:
Differential Revision: [D23093995](https://our.internmc.facebook.com/intern/diff/D23093995)
[ghstack-poisoned]
vkuzo
added a commit
that referenced
this pull request
Sep 2, 2020
Summary: Switches observers to use the new min_max function to calculate min and max at the same time. We see around 45-50% speedup on representative input shapes on the microbenchmarks. Test Plan: CI for correctness performance: ``` cd benchmarks/operator_benchmark // repeat (before diff, after diff) x (cpu, cuda) python -m pt.qobserver_test --tag_filter all --device cpu ``` Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 673b854bc2458171e806d50fa020a584b9c1d7ae Pull Request resolved: #42957
Summary:
Switches observers to use the new min_max function to calculate
min and max at the same time. We see around 45-50% speedup on
representative input shapes on the microbenchmarks for all observers except `HistogramObserver`.
Test Plan:
CI for correctness
performance:
```
cd benchmarks/operator_benchmark
// repeat (before diff, after diff) x (cpu, cuda)
python -m pt.qobserver_test --tag_filter all --device cpu
/*
* before, cpu: https://our.intern.facebook.com/intern/paste/P138633280/
* before, cuda: https://our.intern.facebook.com/intern/paste/P138639473/
* after, cpu: https://our.intern.facebook.com/intern/paste/P138635458/
* after, cuda: https://our.intern.facebook.com/intern/paste/P138636344/
*/
```
Reviewers:
Subscribers:
Tasks:
Tags:
Differential Revision: [D23093995](https://our.internmc.facebook.com/intern/diff/D23093995)
[ghstack-poisoned]
vkuzo
added a commit
that referenced
this pull request
Sep 5, 2020
Summary: Switches observers to use the new min_max function to calculate min and max at the same time. We see around 45-50% speedup on representative input shapes on the microbenchmarks. Test Plan: CI for correctness performance: ``` cd benchmarks/operator_benchmark // repeat (before diff, after diff) x (cpu, cuda) python -m pt.qobserver_test --tag_filter all --device cpu ``` Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 056222595367143386a75c0218fdf33fb5b7929f Pull Request resolved: #42957
Codecov Report
@@ Coverage Diff @@
## gh/vkuzo/123/base #42957 +/- ##
=====================================================
+ Coverage 69.32% 69.35% +0.03%
=====================================================
Files 381 381
Lines 47190 47323 +133
=====================================================
+ Hits 32713 32822 +109
- Misses 14477 14501 +24
Continue to review full report at Codecov.
|
|
This pull request has been merged in fd8e206. |
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:
Switches observers to use the new min_max function to calculate
min and max at the same time. We see around 45-50% speedup on
representative input shapes on the microbenchmarks for all observers except
HistogramObserver.Test Plan:
CI for correctness
performance:
Reviewers:
Subscribers:
Tasks:
Tags:
Differential Revision: D23093995