-
Notifications
You must be signed in to change notification settings - Fork 27.2k
Open
Labels
module: linear algebraIssues related to specialized linear algebra operations in PyTorch; includes matrix multiply matmulIssues related to specialized linear algebra operations in PyTorch; includes matrix multiply matmulmodule: reductionstriagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate moduleThis issue has been looked at a team member, and triaged and prioritized into an appropriate module
Description
🚀 Feature
usual matrix operation are sum-product, could you please implement max-sum matrix operation?
sum-product matirx operation means:
\sum_j ( A_{ij}*B_{jk} ) = C_{ik}
but max-sum matrix operation means:
\max_j ( A_{ij}+B_{jk} )=C_{ik}
Motivation
In graphical model, matrix operation with sum-product are used to calculate partition fuction, If we want to calculate the
ground state of graphical model, we need matrix operation with max-sum; with this kind of operation , we can slove many optimization problems.
Additional context
cc @VitalyFedyunin @ngimel @heitorschueroff @jianyuh @nikitaved @pearu @mruberry @walterddr @IvanYashchuk @xwang233 @lezcano
Reactions are currently unavailable
Metadata
Metadata
Assignees
Labels
module: linear algebraIssues related to specialized linear algebra operations in PyTorch; includes matrix multiply matmulIssues related to specialized linear algebra operations in PyTorch; includes matrix multiply matmulmodule: reductionstriagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate moduleThis issue has been looked at a team member, and triaged and prioritized into an appropriate module