commit | a5537bc683b8476d12917ced00f3540859894de1 | [log] [tgz] |
---|---|---|
author | Stanley Winata <68087699+raikonenfnu@users.noreply.github.com> | Mon Nov 04 10:30:53 2024 -0800 |
committer | GitHub <noreply@github.com> | Mon Nov 04 10:30:53 2024 -0800 |
tree | db410ff9a1f7c569df7948f16db9417738e3bafa | |
parent | ec7528cee889b6fe0c369420aff1094c333fd2b5 [diff] |
[LLVMGPU] Teach KernelConfig to set MMA schedules per op in LoweringConfig (#18984) The main motivation for this change is to enable different intrinsics/layout on different ops inside the same function/dispatch, especially for attention. To that extend, we move the scheduling MMA information such as mma_intrinsic, subgroup_m_count, and subgroup_n_count, from the translation info attached to the function onto the lowering_config per op. Here is a quick summary of things we needed to do to achieve that: 1. Introduce setMmaKind, set/get subgroupMCount, and set/get subgroupMCount on IREE::GPU::LoweringConfigAttr 2. Move configuring of QK matmul's schedule into KernelConfig from LLVMGPUConfigureTensorLayout. 3. Now that qk and pv may have different intrinsic, update information used to decide transposeIntrinsic and reuseIntrinsic in LLVMGPUConfigureTensorLayout 4. Update a bunch of tests to use lowering config to configure MMAs now. --------- Signed-off-by: Stanley Winata <stanley.winata@amd.com>
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
Package | Release status |
---|---|
GitHub release (stable) | |
GitHub release (nightly) | |
Python iree-compiler | |
Python iree-runtime |
Host platform | Build status |
---|---|
Linux | |
macOS | |
Windows |
For the full list of workflows see https://iree.dev/developers/general/github-actions/.
See our website for more information.
Community meeting recordings: IREE YouTube channel
IREE is licensed under the terms of the Apache 2.0 License with LLVM Exceptions. See LICENSE for more information.