commit | 59e07edec625399328fdc07350be1dce3c4f7aac | [log] [tgz] |
---|---|---|
author | Rob Suderman <suderman@google.com> | Mon Oct 17 12:26:23 2022 -0700 |
committer | GitHub <noreply@github.com> | Mon Oct 17 12:26:23 2022 -0700 |
tree | a0ad534f5c3c8fd339d763f37e9781de4b98c71c | |
parent | 63003de747ae5caaa765999ca32a0edd12c4d3de [diff] |
Materialize inserted indices for mhlo.scatter (#10463) mhlo.scatter supports insertedd dimensions at any index length-1 or indexed into. Linalg's scatter requires that indices be explicitly defined (excluding optionally inserting the indexed dims). Adds an mhlo-to-mhlo rewriter that materializes the cases where the inserted indices occur after index dims. Includes fixing the broken `mhlo.scatter` tests in `mhlo-to-mhlo-preprocessing`.
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.