Apply input transforms when computing MLL in model closures #2527
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
facebook-github-bot
added
the
CLA Signed
|
This pull request was exported from Phabricator. Differential Revision: D62497392 |
|
This pull request was exported from Phabricator. Differential Revision: D62497392 |
saitcakmak
added a commit
to saitcakmak/botorch
that referenced
this pull request
Sep 11, 2024
…2527) Summary: Pull Request resolved: pytorch#2527 During model training, the input transforms are applied in `model.forward`. While evaluating the model closures, we pass in the train inputs to the `mll`, which passed them down to the `likelihood`. If we don't transform the inputs before passing them into `mll`, we end up evaluating `model.forward` and `likelihood` using different inputs. This is not an issue during the `posterior` evaluation, since the transforms are applied in `model.posterior` before being passed to `model.__call__` and `likelihood`. This diff updates the model closures to transform the inputs before passing them into `mll`. Fixes pytorch#2515 Differential Revision: D62497392
saitcakmak
force-pushed
the
export-D62497392
branch
from
83fc1c2
to
02d38d1
Compare
|
This pull request was exported from Phabricator. Differential Revision: D62497392 |
saitcakmak
force-pushed
the
export-D62497392
branch
from
02d38d1
to
ba89e7a
Compare
saitcakmak
added a commit
to saitcakmak/botorch
that referenced
this pull request
Sep 11, 2024
…2527) Summary: Pull Request resolved: pytorch#2527 During model training, the input transforms are applied in `model.forward`. While evaluating the model closures, we pass in the train inputs to the `mll`, which passed them down to the `likelihood`. If we don't transform the inputs before passing them into `mll`, we end up evaluating `model.forward` and `likelihood` using different inputs. This is not an issue during the `posterior` evaluation, since the transforms are applied in `model.posterior` before being passed to `model.__call__` and `likelihood`. This diff updates the model closures to transform the inputs before passing them into `mll`. Fixes pytorch#2515 Differential Revision: D62497392
…2527) Summary: Pull Request resolved: pytorch#2527 During model training, the input transforms are applied in `model.forward`. While evaluating the model closures, we pass in the train inputs to the `mll`, which passed them down to the `likelihood`. If we don't transform the inputs before passing them into `mll`, we end up evaluating `model.forward` and `likelihood` using different inputs. This is not an issue during the `posterior` evaluation, since the transforms are applied in `model.posterior` before being passed to `model.__call__` and `likelihood`. This diff updates the model closures to transform the inputs before passing them into `mll`. Fixes pytorch#2515 Reviewed By: SebastianAment Differential Revision: D62497392
|
This pull request was exported from Phabricator. Differential Revision: D62497392 |
saitcakmak
force-pushed
the
export-D62497392
branch
from
ba89e7a
to
5d9c738
Compare
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #2527 +/- ##
=======================================
Coverage 99.98% 99.98%
=======================================
Files 193 193
Lines 16996 17000 +4
=======================================
+ Hits 16994 16998 +4
Misses 2 2 ☔ View full report in Codecov by Sentry. |
|
This pull request has been merged in c895a8d. |
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.
Summary:
During model training, the input transforms are applied in
model.forward. While evaluating the model closures, we pass in the train inputs to themll, which passed them down to thelikelihood. If we don't transform the inputs before passing them intomll, we end up evaluatingmodel.forwardandlikelihoodusing different inputs.This is not an issue during the
posteriorevaluation, since the transforms are applied inmodel.posteriorbefore being passed tomodel.__call__andlikelihood.This diff updates the model closures to transform the inputs before passing them into
mll.Fixes #2515
Differential Revision: D62497392