commit | ff0422027b0bad9edbfab03138ece543cc33342d | [log] [tgz] |
---|---|---|
author | Nicolas Vasilache <nicolasvasilache@users.noreply.github.com> | Wed Aug 31 17:09:18 2022 +0200 |
committer | GitHub <noreply@github.com> | Wed Aug 31 15:09:18 2022 +0000 |
tree | 3c44d058f166e1f5cadd26a81eb47236e57d47af | |
parent | 726c59238938414cb0a323c5df34742327ea748e [diff] |
Add a foreach_thread_to_workgroup transform (#10131) This revision adds a transform dialect op that is very similar to `foreach_thread_to_gpu` to allow distributing to the first level of parallelism during codegen. This is subject to conventions of what values are captured in the workgroup_count region and how these values are passed around in the `stream.cmd.dispatch` op. For now the convention choosen is that `0` values are passed and the `tile_to_foreach_thread` may only use static sizes. A followup PR will introduce a new transform dialect op to allow creating the sizes dynamically based on other conventions.
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.