Support production models with predictor benchmark#9855
Closed
highker wants to merge 1 commit intopytorch:masterfrom
Closed
Support production models with predictor benchmark#9855highker wants to merge 1 commit intopytorch:masterfrom
highker wants to merge 1 commit intopytorch:masterfrom
Conversation
Summary: Pull Request resolved: pytorch#9855 Support production models with predictor benchmark Two new flags are added: `--update_prod`: pull production data (netdef, input types, input dims) from Hive and store locally `--use_prod`: run benchmark with local production data with the same workload as in production. By default, 300 models will be loaded. production vs benchmark avg net run time: (collected by prod: https://fburl.com/scuba/6lb91zfx and bench: https://fburl.com/ngjj1dc8) **prod: `408us` vs bench: `543us`** (With prod data distribution, this should be even closer) framework overhead (as of 2018-07-22): prod: ``` 9.111% BlackBoxPredictor::Run 4.602% SimpleNet::Run 2.377% Operator::Run 1.786% BlackBoxPredictor::AllocateMemory 1.372% Observable::StartAllObservers 1.358% Observable::StartObserver 1.206% Blob::GetMutable ``` bench: ``` 8.577% BlackBoxPredictor::operator() 3.276% SimpleNet::Run 1.954% Operator::Run 1.697% BlackBoxPredictor::AllocateMemory 1.477% Tensor::ShareData 1.230% Blob::GetMutable 1.034% Observable::StartObserver ``` Reviewed By: yinghai Differential Revision: D8942996 fbshipit-source-id: 38a1a9790c048fb81e92aad2b2c82a1651b11e0c
Contributor
|
@pytorchbot retest this please |
1 similar comment
Author
|
@pytorchbot retest this please |
Author
Contributor
|
@pytorchbot retest this please |
jramseyer
pushed a commit
to jramseyer/pytorch
that referenced
this pull request
Jul 30, 2018
Summary: Pull Request resolved: pytorch#9855 Support production models with predictor benchmark Two new flags are added: `--update_prod`: pull production data (netdef, input types, input dims) from Hive and store locally `--use_prod`: run benchmark with local production data with the same workload as in production. By default, 300 models will be loaded. production vs benchmark avg net run time: (collected by prod: https://fburl.com/scuba/6lb91zfx and bench: https://fburl.com/ngjj1dc8) **prod: `408us` vs bench: `543us`** (With prod data distribution, this should be even closer) framework overhead (as of 2018-07-22): prod: ``` 9.111% BlackBoxPredictor::Run 4.602% SimpleNet::Run 2.377% Operator::Run 1.786% BlackBoxPredictor::AllocateMemory 1.372% Observable::StartAllObservers 1.358% Observable::StartObserver 1.206% Blob::GetMutable ``` bench: ``` 8.577% BlackBoxPredictor::operator() 3.276% SimpleNet::Run 1.954% Operator::Run 1.697% BlackBoxPredictor::AllocateMemory 1.477% Tensor::ShareData 1.230% Blob::GetMutable 1.034% Observable::StartObserver ``` Reviewed By: yinghai Differential Revision: D8942996 fbshipit-source-id: 27355d7bb5a9fd8d0a40195261d13a97fa24ce17
goodlux
pushed a commit
to goodlux/pytorch
that referenced
this pull request
Aug 15, 2018
Summary: Pull Request resolved: pytorch#9855 Support production models with predictor benchmark Two new flags are added: `--update_prod`: pull production data (netdef, input types, input dims) from Hive and store locally `--use_prod`: run benchmark with local production data with the same workload as in production. By default, 300 models will be loaded. production vs benchmark avg net run time: (collected by prod: https://fburl.com/scuba/6lb91zfx and bench: https://fburl.com/ngjj1dc8) **prod: `408us` vs bench: `543us`** (With prod data distribution, this should be even closer) framework overhead (as of 2018-07-22): prod: ``` 9.111% BlackBoxPredictor::Run 4.602% SimpleNet::Run 2.377% Operator::Run 1.786% BlackBoxPredictor::AllocateMemory 1.372% Observable::StartAllObservers 1.358% Observable::StartObserver 1.206% Blob::GetMutable ``` bench: ``` 8.577% BlackBoxPredictor::operator() 3.276% SimpleNet::Run 1.954% Operator::Run 1.697% BlackBoxPredictor::AllocateMemory 1.477% Tensor::ShareData 1.230% Blob::GetMutable 1.034% Observable::StartObserver ``` Reviewed By: yinghai Differential Revision: D8942996 fbshipit-source-id: 27355d7bb5a9fd8d0a40195261d13a97fa24ce17
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:
Support production models with predictor benchmark
Two new flags are added:
--update_prod: pull production data (netdef, input types, input dims) from Hive and store locally--use_prod: run benchmark with local production data with the same workload as in production.By default, 300 models will be loaded.
production vs benchmark
avg net run time:
(collected by prod: https://fburl.com/scuba/6lb91zfx and bench: https://fburl.com/ngjj1dc8)
prod:
408usvs bench:543us(With prod data distribution, this should be even closer)
framework overhead (as of 2018-07-22):
prod:
bench:
Differential Revision: D8942996