Skip to content

Latest commit

 

History

History
139 lines (92 loc) · 6.16 KB

IREE.adoc

File metadata and controls

139 lines (92 loc) · 6.16 KB
Raw

Name of project

Requested project maturity level

Sandbox

Project description (what it does, why it is valuable, origin and history, ongoing development).

IREE (Intermediate Representation Execution Environment) 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. The mission is to provide a toolkit for ML model deployment that is able to exploit heterogeneous compute environments and provide a consistent authoring flow to deploy across multiple platforms easily embracing new hardware or specialization.

IREE started OSS-first development in 2019 driven by team at Google focussing on deployment of ML models on mobile. Since then IREE (and components built by IREE team) has developed an OSS community with contributions and users from multiple companies, grown to use over multiple plaforms spanning from embedded to distributed server deployment, from ML to usage inside game engines.

Statement on alignment with LF AI’s mission.

As a group of engineers who have worked in AI, compilers and runtimes broadly for many years, we identify with the practical approach that LF AI&D takes by aiming to provide "a greenhouse" for concrete innovations, rooted in good open-source principles and execution. As the core contributors have worked for various entities over the years and integrated the technology into multiple product roadmaps, we’ve always taken pride in making the low level infrastructure investment an evergreen community project that can underlie many things as time and situations evolve. In this day and age, however, it takes more than a commitment to the low level engineering, and we appreciate the opportunity to align with LF AI&D so that we can together add the structure to the project itself so that it remains a viable, approachable and community grounded initiative.

Have you identified possible collaboration opportunities with current LF AI hosted projects (https://lfai.foundation/projects/)? Please explain.

Not extensively. The IREE project already consumes ONNX programs but this is more a user relationship. Superficial evaluation would indicate potential opporutnies related to Ludwig, 1chipML, CLAIMED, Delta, RKWV and ShaderNN.

License name, version, and URL to license text

Apache License v2.0 with LLVM exceptions, https://llvm.org/LICENSE.txt

Source control

GitHub

Does the project sits in its own GH organization?

Yes (iree-org).

Do you have the GH DCO app active in the repos?

Currently Google CLA.

Issue tracker

GitHub

Collaboration tools (mailing lists, wiki, IRC, Slack, Glitter, etc.)

External dependencies including licenses (name and version) of those dependencies.

  • google/benchmark — Apache License 2.0

  • pytorch/cpuinfo — BSD 2-Clause "Simplified" License

  • dvidelabs/flatcc — Apache License 2.0

  • google/googletest — BSD 3-Clause "New" or "Revised" License

  • shark-infra/hip-build-deps — MIT License

  • llvm/llvm-project — Apache License v2.0 with LLVM Exceptions

  • musl — MIT license

  • pybind/pybind11 — BSD-style license

  • KhronosGroup/spirv_cross — Apache License 2.0

  • openxla/stablehlo — Apache License 2.0

  • wolfpld/tracy — 3-clause BSD license.

  • KhronosGroup/vulkan_headers — Apache License 2.0 OR MIT

  • webgpu-native/webgpu-headers — BSD 3-Clause "New" or "Revised" License

Initial committers (name, email, organization) and how long have they been working on project?

Have the project defined the roles of contributor, committer, maintainer, etc.? Please document it in MAINTAINERS.md.

Total number of contributors to the project including their affiliations.

190 in lifetime, in top 100 the current affiliations are 30 AMD, 20 Google, 3 ARM, 3 Fraunhoffer institute.

Does the project have a release methodology? Please document it in RELEASES.md.

Yes, see RELEASING.md.

Does the project have a code of conduct? If yes, please share the URL. If no, please created CODE_OF_CONDUCT.md and point to https://lfprojects.org/policies/code-of-conduct/. You can use [email protected] as email for contact on this topic.

Yes, https://github.com/openxla/community/blob/main/CODE-OF-CONDUCT.md but will switch to LF code-of-conduct if accepted.

Did the project achieve any of the CII best practices badges? A different badge is required depending on the requested incubation level.

Not yet, we are in the process of addressing this (https://www.bestpractices.dev/en/projects/8738).

Do you have any specific infrastructure requests needed as part of hosting the project in the LF AI?

No.

Project website - Do you have a web site? If no, did you reserve a domain, and would like you to have a website created?

Yes, iree.dev.

Project governance - Do you have a working governance model for the project? Please provide URL to where it is documented, typically GOVERNANCE.md.

We wish to work with LFAI to adopt a new one (current governance is based loosely on affiliation with LLVM and OpenXLA).

Social media accounts - Do you have any Twitter/LinkedIn/Facebook/etc. project accounts? Please provide pointers.

Existing sponsorship (e.g., whether any organization has provided funding or other support to date, and a description of that support), if any.

AMD and Google both actively contribute engineering as well as CI resources.