Put input transforms into train mode before converting models #1283
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
CLA Signed
|
This pull request was exported from Phabricator. Differential Revision: D37542474 |
Codecov Report
@@ Coverage Diff @@
## main #1283 +/- ##
=========================================
Coverage 100.00% 100.00%
=========================================
Files 122 122
Lines 10595 10601 +6
=========================================
+ Hits 10595 10601 +6
Continue to review full report at Codecov.
|
…h#1283) Summary: Pull Request resolved: pytorch#1283 Fixes pytorch#1273 During model construction, input transforms should be in `train` mode (so that they only apply if `transform_on_train` is true). Having the input transforms in eval mode leads to buggy behavior due to `transformed_X` getting transformed when it shouldn't. Differential Revision: D37542474 fbshipit-source-id: f4278294de5d83d967f3d21c312370e562cf372c
saitcakmak
force-pushed
the
export-D37542474
branch
from
3a9b29c to
bac446a
Compare
|
This pull request was exported from Phabricator. Differential Revision: D37542474 |
saitcakmak
added a commit
to saitcakmak/botorch
that referenced
this pull request
Oct 17, 2022
…putModel` Summary: This fixes a bug where, due to the input transform being in `train` mode while the model is in `eval` mode, the learnable bounds of input transforms, such as Normalize, would be wrongly updated during model training, effectively disabling the transforms. This is a serious bug that was leading to using the model trained with normalized inputs on a model without any normalization. It has been around for a while, since pytorch#1283 :'(. Differential Revision: D40453200 fbshipit-source-id: d30e9ba18fa466c01bbf3273165e6e35ac33c219
saitcakmak
added a commit
to saitcakmak/botorch
that referenced
this pull request
Oct 17, 2022
…putModel` (pytorch#1454) Summary: Pull Request resolved: pytorch#1454 This fixes a bug where, due to the input transform being in `train` mode while the model is in `eval` mode, the learnable bounds of input transforms, such as Normalize, would be wrongly updated during model training, effectively disabling the transforms. This is a serious bug that was leading to using the model trained with normalized inputs on a model without any normalization. It has been around for a while, since pytorch#1283 :'(. Differential Revision: D40453200 fbshipit-source-id: 0d60ac5f23efe06e200713555dfa109b612d9290
facebook-github-bot
pushed a commit
that referenced
this pull request
Oct 18, 2022
…putModel` (#1454) Summary: Pull Request resolved: #1454 This fixes a bug where, due to the input transform being in `train` mode while the model is in `eval` mode, the learnable bounds of input transforms, such as Normalize, would be wrongly updated during model training, effectively disabling the transforms. This is a serious bug that was leading to using the model trained with normalized inputs on a model without any normalization. It has been around for a while, since #1283 :'(. Reviewed By: Balandat Differential Revision: D40453200 fbshipit-source-id: 3a431406de07565a6ccab3ab22ddc667a39b8c2c
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:
Fixes #1273
During model construction, input transforms should be in
trainmode (so that they only apply iftransform_on_trainis true).Having the input transforms in eval mode leads to buggy behavior due to
transformed_Xgetting transformed when it shouldn't.Differential Revision: D37542474