[ADD] Method from_mat_inner to structured matrix interface
#80
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
f-dangel
changed the title
from_mat_outer to structured matrix interfacefrom_mat_inner to structured matrix interface
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Outdated suggestions cannot be applied.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.
Adds an interface function
from_mat_innerwhich extracts a structured matrix fromX @ X.TwhereXis an arbitrary 2d tensor. This addresses f-dangel/sirfshampoo#24.This function is useful for speeding up
SIRFShampoo's update which takes a matrix viewGof the gradient, then extracts a structured matrix ofG @ G.T. So far, we were mimicking this computation by creating a structured matrixIwhich represents the identity matrix, then callI.from_inner(G)which computes the structured matrix ofI.T @ G @ G.T @ I, effectively multiplying the identity matrix ontoGbefore extracting the structured matrix. The new functionfrom_mat_innerremoves the unnecessary multiplication with identity.