fix(tf): make se_atten_v2 masking smooth when davg is not zero
#3632
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
…davg_zero is False Signed-off-by: Jinzhe Zeng <[email protected]>
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## devel #3632 +/- ##
==========================================
- Coverage 77.79% 77.77% -0.02%
==========================================
Files 438 438
Lines 37693 37709 +16
Branches 1623 1623
==========================================
+ Hits 29322 29329 +7
- Misses 7507 7516 +9
Partials 864 864 ☔ View full report in Codecov by Sentry. |
Signed-off-by: Jinzhe Zeng <[email protected]>
for more information, see https://pre-commit.ci
Signed-off-by: Jinzhe Zeng <[email protected]>
for more information, see https://pre-commit.ci
Signed-off-by: Jinzhe Zeng <[email protected]>
for more information, see https://pre-commit.ci
Signed-off-by: Jinzhe Zeng <[email protected]>
for more information, see https://pre-commit.ci
wanghan-iapcm
approved these changes
Apr 2, 2024
iProzd
approved these changes
Apr 2, 2024
njzjz
added a commit
to njzjz/deepmd-kit
that referenced
this pull request
Apr 6, 2024
…pmodeling#3632) Currently, `se_atten_v2` is always masked to zero when `exclude_types` is given. However, for the no neighbor case, the placeholder for a virtual neighbor is `davg`. This causes discontinuity when `set_davg_zero` is not set. This PR uses `davg` for masking. In production, we usually use `set_davg_zero` along with `exclude_types`, so it hasn't caused a real problem. I notice PT hasn't implemented `se_atten_v2` or `exclude_types`, but we need attention in the future. --------- Signed-off-by: Jinzhe Zeng <[email protected]> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> (cherry picked from commit 63601b0)
Merged
njzjz
added a commit
that referenced
this pull request
Apr 6, 2024
Currently, `se_atten_v2` is always masked to zero when `exclude_types` is given. However, for the no neighbor case, the placeholder for a virtual neighbor is `davg`. This causes discontinuity when `set_davg_zero` is not set. This PR uses `davg` for masking. In production, we usually use `set_davg_zero` along with `exclude_types`, so it hasn't caused a real problem. I notice PT hasn't implemented `se_atten_v2` or `exclude_types`, but we need attention in the future. --------- Signed-off-by: Jinzhe Zeng <[email protected]> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> (cherry picked from commit 63601b0)
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.
Currently,
se_atten_v2is always masked to zero whenexclude_typesis given. However, for the no neighbor case, the placeholder for a virtual neighbor isdavg. This causes discontinuity whenset_davg_zerois not set.This PR uses
davgfor masking.In production, we usually use
set_davg_zeroalong withexclude_types, so it hasn't caused a real problem.I notice PT hasn't implemented
se_atten_v2orexclude_types, but we need attention in the future.