Enable Sparse processes #672
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
…o dev/sparse_proc � Conflicts: � tests/lava/proc/sparse/test_model.py
…o dev/sparse_proc
…o dev/sparse_proc � Conflicts: � tests/lava/proc/sparse/test_model.py
…o dev/sparse_proc � Conflicts: � tests/lava/proc/sparse/test_model.py
8 tasks
PhilippPlank
requested changes
Apr 26, 2023
PhilippPlank
approved these changes
Apr 26, 2023
tim-shea
approved these changes
May 1, 2023
gkarray
approved these changes
May 2, 2023
mathisrichter
approved these changes
May 2, 2023
There was a problem hiding this comment.
This is such a huge PR that any reviewer will have a very hard time comprehending, let alone judging it. I made an effort to point out minor things that I noticed along the way but got overwhelmed by the amount of code. I have a feeling that we should be testing many more variations of the user-level API. The new sparse Processes have so many parameters that should be really be exercised more in tests.
monkin77
pushed a commit
to monkin77/thesis-lava
that referenced
this pull request
Jul 12, 2024
* added Sparse proc and init test * added test and implementatin for sparse proc model in floating precision * test for graded spikes * added bit acc version and adapted weight utils * typehints * integer weights for fixed point tests * draft learning * draft get * refactoring tests * avoid saving var in varmodel * draft set * Sparse get/set working on CPU * delay sparse process + test * change order of weights * minor fix for complete sparse matrix * minor fix * delay sparse model + test * get/set for float/fixed * delay sparse fix for int input * LearningSparse floating-pt version + test * all tests for delay dense also run for delay sparse * update test naming * use dot product * make learning float work * fixed pt learning * rm bit approx version * lint * added learning dense for bit approx * improve calculation of wgt_dly * tests for dt and dd * lint * avoid warnings * improve getting the zero matrix * lint * minor changes * improve documentation * minor change * changes to comments * improve documentation * improve documentation * minor change --------- Co-authored-by: gkarray <[email protected]> Co-authored-by: SveaMeyer13 <[email protected]> Co-authored-by: PhilippPlank <[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.
Issue Number:
#656 #636
Objective of pull request:
This PR enables connectivity using Sparse weight matrices. It also includes handling Delays and Learning in floating and bit-approximate versions.
Pull request checklist
Your PR fulfills the following requirements:
flakeheaven lint src/lava tests/) and (bandit -r src/lava/.) pass locallypytest) passes locallyPull request type
Please check your PR type:
What is the current behavior?
What is the new behavior?
Does this introduce a breaking change?
Supplemental information