-
Notifications
You must be signed in to change notification settings - Fork 40
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Added quickTunerGen.py, the main driver of the quick tuner scripts #1576
base: develop
Are you sure you want to change the base?
Conversation
which generates quick tuner perf configs given the dataframe/tsv file produced by quickTunerPreproc.py.
instead of 'TFlops'
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## develop #1576 +/- ##
===========================================
- Coverage 77.51% 77.40% -0.12%
===========================================
Files 90 92 +2
Lines 24358 24580 +222
Branches 3414 3451 +37
===========================================
+ Hits 18881 19026 +145
- Misses 4132 4189 +57
- Partials 1345 1365 +20
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Might as well get the complaining in before you leave - sorry to drop this on the last minute
as input. | ||
Needs the input to be a combined normalized dataframe (default from quickTunerPreproc.py) | ||
|
||
Usage: clusterConfigs.py [-h] --input-file INPUT_FILE [--method {default,topNSelect,topMode,takeNEach,fairSelect,hardcoded} [{default,topNSelect,topMode,takeNEach,fairSelect,hardcoded} ...]] [--save] [--debug] [--num NUM] [--perfconfig--format] |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Stale name in docment
os.environ['OMP_NUM_THREADS'] = '1' | ||
import sys | ||
|
||
sys.path.append('../..') |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
... ???
You shouldn't need to do this
import re | ||
import glob | ||
|
||
class quickTunerMethod(object): |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Same case/(object)
nits as on other PRs
|
||
trans_cols = ['TransA', 'TransB'] | ||
|
||
param_cols = [ 'G', 'M', 'N','K'] |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Nit, weird newlines
|
||
|
||
self.default_f16 = pd.DataFrame({ | ||
"M/block": [128, 128, 128, 128, 128, 128, 128, 128, 128, 128, 128, 128, 128, 128, 64, 64, 64, 64, 64, 64, 64, 64, 64, 64, 64, 64, 64, 64, 64, 64, 32, 32, 32, 32, 32, 32, 16, 16, 16, 16], |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
.. what's this about? Why do we have this? Where are the notes to keep it up to date?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Also, if you're doing to do this, build the dataframe from a list of lists and give the column names separately - this is unreadable.
|
||
if __name__ == '__main__': | ||
main(sys.argv[1:]) | ||
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Something feels really awkward here - I can't for the life of me follow what all this is supposed to be doing and there's a whole lot of for loops and manual bouncing around of columns
Generates quick tuner perf configs given the dataframe/tsv file produced by quickTunerPreproc.py.