-
Notifications
You must be signed in to change notification settings - Fork 5
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
nngraph verifier & verifier for all rules #10
Merged
Merged
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
дальше может сделать в федоте базовый класс для верифаеров и перенаследовать этот от него |
nicl-nno
approved these changes
Sep 29, 2022
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.
А он пока и не должен нигде использоваться?
staeros
approved these changes
Sep 29, 2022
а я его в SA буду использовать, но кажется, что и в самом NAS'e потом пригодится |
staeros
pushed a commit
that referenced
this pull request
Jan 26, 2023
staeros
added a commit
that referenced
this pull request
Jan 30, 2023
* Requirements fix * Graph requirements were extended for convolutional nodes; Added stricter limitation for max possible numbers of trainable parameters (from 7e8 to 2e8). * Fixes for simple mutation bug * Fixes for simple mutation bug * Added layers repository * Pre-refactoring changes. Updating class for graph generation and solving issue with keras OOM * Possible pool of conv layer types was extended; validation rules update; refactoring * Added trainable restrictions. * Fix for trainable restrictions. * Fix for trainable restrictions. * requirements update; serializer error fix * val rule update for trainable parameters * ResNet builder implementation for initial assumption * ResNet builder updates * Skip connections fix, minor Resnet builder fixes * Memory leak fixing * Added graph branch manager and class for convert node to layer * Added graph branch manager and class for convert node to layer * Refactoring + Updates for model generation * Refactoring; Subcalssing from Keras model was replaced with ModelConverter class. * Added temporary val rules for resnet-like networks + small fixes * save path fix * minor changes * minor changes * Keras optimizer changed * Minor changes * Minor changes * Added callback for model performance * setup update * callbacks update * callbacks update * Hotfixes for val rules * Hotfixes for val rules: added ty except in dimension check rule * Builder fixes * Builder fixes * requirements update * requirements update * Branch Manager fix * Minor fixes * Minor hotfixes * Minor hotfixes * Added unfit method * Added performance callback * memory hotfix * history functional object remove * golem update * requirements update * golem imports update; requirements refactoring * Requirements update * Requirements update; Added unit tests for base layer requirements * Default requirements hotfix * Added unit tests for requirements * Redundant code was removed * Minor composer changes; Composer refactoring * Added class BaseNasDatasetBuilder for InputData preparation for model train * BaseNasDatasetBuilder preprocessor fix * BaseNasDatasetBuilder preprocessor fix; cases update * Refactoring * Refactoring * Refactoring; Added unit test for DatasetBuilder * Dataset builder preprocessor fixes; Added unit tests for dataset builder * NasGraph refactoring; Added option to specify train parameters more flexible * WIP Model builder and LayerInitializer * Code refactoring * Requirements update * Unit tests update * Unit tests update * Code update * nngraph verifier & verifier for all rules (#10) * Golem transition Code update Unit tests update Unit tests update Requirements update Code refactoring WIP Model builder and LayerInitializer NasGraph refactoring; Added option to specify train parameters more flexible Dataset builder preprocessor fixes; Added unit tests for dataset builder Refactoring; Added unit test for DatasetBuilder Refactoring Refactoring BaseNasDatasetBuilder preprocessor fix; cases update BaseNasDatasetBuilder preprocessor fix Added class BaseNasDatasetBuilder for InputData preparation for model train Minor composer changes; Composer refactoring Redundant code was removed Added unit tests for requirements Default requirements hotfix Requirements update; Added unit tests for base layer requirements Requirements update golem imports update; requirements refactoring requirements update golem update history functional object remove memory hotfix Added performance callback Added unfit method Minor hotfixes Minor hotfixes Minor fixes Branch Manager fix requirements update requirements update Builder fixes Builder fixes Hotfixes for val rules: added ty except in dimension check rule Hotfixes for val rules callbacks update callbacks update setup update Added callback for model performance Minor changes Minor changes Keras optimizer changed minor changes minor changes save path fix Added temporary val rules for resnet-like networks + small fixes Refactoring; Subcalssing from Keras model was replaced with ModelConverter class. Refactoring + Updates for model generation Added graph branch manager and class for convert node to layer Added graph branch manager and class for convert node to layer Memory leak fixing Skip connections fix, minor Resnet builder fixes ResNet builder updates ResNet builder implementation for initial assumption val rule update for trainable parameters requirements update; serializer error fix Fix for trainable restrictions. Fix for trainable restrictions. Added trainable restrictions. Possible pool of conv layer types was extended; validation rules update; refactoring Pre-refactoring changes. Updating class for graph generation and solving issue with keras OOM Added layers repository Fixes for simple mutation bug Fixes for simple mutation bug Graph requirements were extended for convolutional nodes; Added stricter limitation for max possible numbers of trainable parameters (from 7e8 to 2e8). * Requirements fix * Redundant code removed * Code refactoring * Added dockstrings; Refactoring * Requirements update * Minor changes --------- Co-authored-by: Pinchuk Maya <[email protected]>
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.
NNGrapn Verifier