Power iteration algorithm is now generic to any linear operator #73
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
cofri
force-pushed
the
refactor/power_iteration_generic
branch
from
97dcebd to
54a2725
Compare
thib-s
reviewed
Jun 6, 2023
thib-s
reviewed
Jun 6, 2023
thib-s
requested changes
Jun 6, 2023
franckma31
reviewed
Sep 6, 2023
To be checked: it seems that it has been resolved for convolution but may also be applicable for dense layer (u of min size in kernel.shape, and use kernel or kernel transpose) |
cofri
force-pushed
the
refactor/power_iteration_generic
branch
from
99531c7 to
ec61e65
Compare
thib-s
approved these changes
Oct 20, 2023
The function `power_iteration()` is now generic to any linear operator. A linear operator function and its adjoint must be passed as arguments. The maximum singular vector is returned. Note that this commit is atomic but is part of larger modifications. It won't pass the tests as is.
The function `spectral_normalization()` is now based on the generic power iteration function.
The function spectral_normalization_conv() now uses the generic power iteration. Unlike spectral normalization for matrices, a get_convolution_operators() function is introduced because it can be required in some future operations.
Setting the axis for normalization is useful to handle depthwise convolution filterwise
cofri
force-pushed
the
refactor/power_iteration_generic
branch
from
562512a to
f09c2be
Compare
|
Rebased on master |
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Outdated suggestions cannot be applied.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.
In order to easily handle power iteration computations for future new linear operators (e.g. depthwise convolution), the power iteration algorithm is made generic: