| commit | 0af63addc8c2ea0bf903391a877bd60f52a10f73 | [log] [tgz] |
|---|---|---|
| author | Benjamin Maxwell <benjamin.maxwell@arm.com> | Sat Sep 30 01:58:00 2023 +0100 |
| committer | GitHub <noreply@github.com> | Fri Sep 29 17:58:00 2023 -0700 |
| tree | 89b4f6b645f3f722b1da27680f80c8b5d8f3a11a | |
| parent | 113f9d2ad5a0d3793ecac4c814106133253411c0 [diff] |
Support scalable tile sizes in #iree_codegen.lowering_config (#15033) This extends the lowering config so that at each level tiles can be marked as scalable using an optional array of bools. The parser/printer has been updated to display this in the canonical MLIR syntax (i.e. the size wrapped in square brackets). So for example, this lowering config: ``` #iree_codegen.lowering_config<tile_sizes = [[128, 128, 0], [1, [32], 0], [0, 0, 1], [0, 0, 0]]> ``` Means that the middle tile size (32) of the second tiling level is scalable. These are not yet used, but will be in subsequent patches.
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.