| commit | 76e9cfe20ffbd4b44c6f7bac3418f575405c69b9 | [log] [tgz] |
|---|---|---|
| author | Ben Vanik <ben.vanik@gmail.com> | Thu Apr 11 15:49:38 2024 -0700 |
| committer | GitHub <noreply@github.com> | Thu Apr 11 15:49:38 2024 -0700 |
| tree | 7c89dfcc1ffc72f5bf72a82a0f8e85f25d2612d4 | |
| parent | fbd31b0e87a83e04364b767a4d213c5a79af0d6e [diff] |
Allowing flow.tensor.constant to be used for constants. (#17024) Prior it had only been used for dynamically shaped constants for testing. Now it supports that mode (both for testing and programs that want to ensure certain portions are treated dynamically, but mostly testing) but also supports being used ala arith.constant. arith.constant is still preferred for now as it composes with upstream stuff better but our own op lets us support constant types arith cannot, such as parameters. With this frontends can pass in IR with constant parameters inline instead of needing to make them globals. (I still don't really like that this op has the dynamic behavior and may remove it in the future - the few uses of it can change to using casts with optimization barriers on them as a more verbose but less nasty way of doing the same thing)
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.