[Codegen] Split StableHLO/JAX select-style argmax reductions (#24445) Problem: StableHLO/JAX lowers argmax for lm_head vocab reductions as a pair of selects with explicit NaN handling and lower-index tie breaking. The existing split-reduction matcher only recognized the canonical maximumf/cmpf/select form, so large vocab reductions stayed as one wide reduction. Fix: Teach the argmax matcher to distinguish canonical and StableHLO/JAX select-style argmax reductions, and preserve the corresponding combiner when splitting. Keep the select-style matcher strict by requiring the `gt || isnan` value condition, the `eq && lower_index` tie condition, and the expected value/index select operands. Example: A Qwen `lm_head` argmax over vocab size 248320 can now split into a two-stage reduction over 1940x128 tiles while preserving NaN and lower-index tie semantics. Negative tests ensure similar-looking select reductions with different NaN/tie/index semantics do not split. --------- Signed-off-by: Yuwei Sun <yuweis2@illinois.edu> Signed-off-by: Hao Ren <rhao8608@gmail.com> Co-authored-by: Hao Ren <rhao8608@gmail.com>
IREE (Intermediate Representation Execution Eenvironment, 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.
Releases notes are published on GitHub releases.
| Package | Release status |
|---|---|
| GitHub release (stable) | |
| GitHub release (nightly) | |
iree-base-compiler | |
iree-base-runtime |
For more details on the release process, see https://iree.dev/developers/general/release-management/.
| Operating system | Build status |
|---|---|
| Linux | |
| macOS | |
| macOS |
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
| Date | Title | Recording | Slides |
|---|---|---|---|
| 2025-06-10 | Data-Tiling in IREE: Achieving High Performance Through Compiler Design (AsiaLLVM) | recording | slides |
| 2025-05-17 | Introduction to GPU architecture and IREE's GPU CodeGen Pipeline | recording | slides |
| 2025-02-12 | The Long Tail of AI: SPIR-V in IREE and MLIR (Vulkanised) | recording | slides |
| 2024-10-01 | Unveiling the Inner Workings of IREE: An MLIR-Based Compiler for Diverse Hardware | recording | |
| 2021-06-09 | IREE Runtime Design Tech Talk | recording | slides |
| 2020-08-20 | IREE CodeGen (MLIR Open Design Meeting) | recording | slides |
| 2020-03-18 | Interactive HAL IR Walkthrough | recording | |
| 2020-01-31 | End-to-end MLIR Workflow in IREE (MLIR Open Design Meeting) | recording | slides |
IREE is licensed under the terms of the Apache 2.0 License with LLVM Exceptions. See LICENSE for more information.