| commit | eaefd3ff964f14e700b1c468aa2b4f20a4401028 | [log] [tgz] |
|---|---|---|
| author | harsh-nod <harsh@nod-labs.com> | Mon Jan 29 22:45:38 2024 -0800 |
| committer | GitHub <noreply@github.com> | Mon Jan 29 22:45:38 2024 -0800 |
| tree | 2858d5c673cb678ffbb31314b7a0d55421a4b033 | |
| parent | 28de72c482898a358bcce333c3fc91663cae267f [diff] |
[LinalgExt] Add optional transpose_v attribute to attention (#16254) In the attention operator, the second contraction is a matrix multiplication. Since many hardware vendors natively support matrix multiplication with b transposed, a common benchmark is to assume that the second operand of the second contraction is transposed, result in a matrix multiplication with b transposed. This PR adds that optional attribute and updates the tile and decompose pass to handle the case.
IREE (Intermediate Representation Execution Environment, pronounced as “eerie”) is an MLIR-based end-to-end compiler and runtime that lowers Machine Learning (ML) models to a unified IR that scales up to meet the needs of the datacenter and down to satisfy the constraints and special considerations of mobile and edge deployments.
See our website for project details, user guides, and instructions on building from source.
IREE is still in its early phase. We have settled down on the overarching infrastructure and are actively improving various software components as well as project logistics. It is still quite far from ready for everyday use and is made available without any support at the moment. With that said, we welcome any kind of feedback on any communication channels!
See our website for more information.
IREE is licensed under the terms of the Apache 2.0 License with LLVM Exceptions. See LICENSE for more information.