This repository has been archived by the owner on Nov 17, 2023. It is now read-only.
[OpPerf] Implement all linalg ops #17528
Merged
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
|
@mxnet-label-bot add [pr-awaiting-review] |
ChaiBapchya
reviewed
Feb 6, 2020
connorgoggins
force-pushed
the
implement_linalg_ops
branch
from
8fb0d1a to
ffc8000
Compare
ChaiBapchya
approved these changes
Feb 6, 2020
connorgoggins
force-pushed
the
implement_linalg_ops
branch
from
ffc8000 to
023b9d4
Compare
access2rohit
approved these changes
Feb 9, 2020
|
@mxnet-label-bot add [pr-awaiting-merge] |
|
@apeforest can you also review ? |
apeforest
approved these changes
Feb 10, 2020
zheyuye
pushed a commit
to zheyuye/incubator-mxnet
that referenced
this pull request
Feb 19, 2020
* Initial commit - implemented linalg_potrf * Implemented standardized linalg ops, added relevant data * Added independently_tested category * Dropped unnecessary NDArray cast for potrf * Removed duplicate makediag, rebased * Updated group of custom_data ops
anirudh2290
pushed a commit
to anirudh2290/mxnet
that referenced
this pull request
May 29, 2020
* Initial commit - implemented linalg_potrf * Implemented standardized linalg ops, added relevant data * Added independently_tested category * Dropped unnecessary NDArray cast for potrf * Removed duplicate makediag, rebased * Updated group of custom_data ops
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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Outdated suggestions cannot be applied.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.
Description
This PR serves to implement all Linear Algebra operators in OpPerf. To achieve this, I created a file (
benchmark/opperf/nd_operations/linalg_operators.py) with a function (run_linalg_operators_benchmarks) to run all Linear Algebra ops. Within this function, I used calls torun_performance_testfor the linalg ops with specific requirements on input data values and a call to a new function I wrote (get_all_linalg_operators) in conjunction withrun_op_benchmarksfor all other linalg ops. I also added a call torun_linalg_operators_benchmarksinopperf.py.I also disabled backwards runs on ops that do not support them (
linalg_gelqf,linalg_slogdet,moments).Checklist
Essentials
Please feel free to remove inapplicable items for your PR.
Changes
Comments
Tested on p2.16xl w/ubuntu 16.04 and Mac OS with:
run_linalg_operators_benchmarks- runs all Linear Algebra ops on relevant dataopperf.py(full run of all ops)Performance Results
Group of operator test - all linalg ops (CPU)
Full OpPerf test (CPU)
Group of operator test - all linalg ops (GPU)
Full OpPerf test (GPU)
@apeforest @access2rohit @ChaiBapchya