Repository navigation
Use prob.f.adtype for the HybridPSO local-phase gradient - #147
Merged
Merged
Conversation
The SimpleLBFGS and BFGS local phases in HybridPSO hardcoded ForwardDiff. They now go through instantiate_gradient, so OptimizationFunction(f, AutoEnzyme()) uses Enzyme for the local phase. EnzymeGradient gets an SVector method that differentiates with an Active argument via autodiff_deferred, so it runs inside GPU kernels without the MVector shadow. Problems without an AD type (NoAD()) fall back to ForwardDiff, keeping the current default behavior.
AdityaPandeyCN
marked this pull request as draft
September 30, 2026 20:34
Contributor
Author
|
@utkarsh530 Please take a look |
AdityaPandeyCN
marked this pull request as ready for review
September 30, 2026 21:39
AdityaPandeyCN
force-pushed
the
hybrid-enzyme-adtype
branch
from
October 1, 2026 16:36
01d92ff to
a9c0720
Compare
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The local phase of
HybridPSO(bothSimpleLBFGSandBFGS) always used ForwardDiff for the gradient. It now usesinstantiate_gradient(f, prob.f.adtype), so users can opt in to Enzyme withOptimizationFunction(f, AutoEnzyme()).Changes:
EnzymeGradientgets anSVectormethod that passes the input asActivethroughautodiff_deferred, so it can run inside GPU kernels without theMVectorshadow the generic method uses.instantiate_gradient(f, _)falls back toForwardDiffGradient, so problems without an AD type (NoAD(), the default) behave exactly as before.test/lbfgs.jlruns the hybrid Rosenbrock cases (bounded and unbounded) withAutoEnzyme()andNoAD().