Optimize MG variance calculation for dataset standardization for logistic regression #6138
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
lijinf2
added
CUDA / C++
lijinf2
force-pushed
the
raft_stat_1118
branch
from
b6ecfae to
083f204
Compare
github-actions
bot
added
CUDA/C++
and removed
Cython / Python
lijinf2
changed the title
eordentlich
reviewed
Nov 21, 2024
lijinf2
force-pushed
the
raft_stat_1118
branch
2 times, most recently
from
33e27de to
b3ee71e
Compare
lijinf2
force-pushed
the
raft_stat_1118
branch
from
b3ee71e to
8eeded4
Compare
dantegd
approved these changes
Dec 5, 2024
|
/merge |
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Outdated suggestions cannot be applied.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.
MG variance calculation currently involks raft SG vars API. However, the abs() step of raft SG vars API introduces errors in skewed data distribution (e.g., one GPU gets small values 1 and 2, and the other GPU gets large values 98 and 99).
The PR avoids the effect of abs() when involking SG vars for calculating MG vars. The key idea is to pass a vector of zeroes when calling SG vars.