| commit | 275f0fb86191e0b6a9dcc8bb34229aa8c252bc2b | [log] [tgz] |
|---|---|---|
| author | Sean Silva <silvasean@google.com> | Thu Jun 18 16:28:38 2020 -0700 |
| committer | Copybara-Service <copybara-worker@google.com> | Thu Jun 18 16:30:28 2020 -0700 |
| tree | 2d7f072eb2a8608d859c69ab746b07cfe599ca60 | |
| parent | c3e9391ebeece2572ec39e9bdc86507fc9ef52f5 [diff] |
Fixes exposed by an upcoming tensorflow CL. An up coming tensorflow CL is going to change the Keras RNN to emit tf.DeviceIndex ops. This requires incorporating tf-device-index-selector into our lowering pipeline. I have that TF change patched in locally, and due to a slight variation in type inference with the new pipeline, it ultimately creates a xla_hlo.convert op with the same operand and result element types types (`"xla_hlo.convert"(%70) : (tensor<7x?x10xf32>) -> tensor<7x24x10xf32>`), which we don't have any runtime kernels for (nor does it make sense to add them). Since this op is an identity in this case, we can just lower it as an identity op. PiperOrigin-RevId: 317204833
IREE (Intermediate Representation Execution Environment, pronounced as “eerie”) is an MLIR-based end-to-end compiler that lowers ML models to a unified IR optimized for real-time mobile/edge inference against heterogeneous hardware accelerators. IREE also provides flexible deployment solutions for the compiled ML models.
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!
For development, IREE supports both Bazel and CMake on Windows and Linux. We are working on enabling macOS support. For deployment, IREE aims to additionally cover Android and iOS.
Please see the Getting Started pages on IREE's documentation hub to configure, compile, and run IREE in your favorite development environment!
IREE hosts all its documentation and project status dashboards on GitHub Pages. We are still building up the website; please feel free to create issues for the documentation you'd like to see!
We also have some public talks that explain IREE's concepts and architecture:
IREE adopts a holistic approach towards ML model compilation: the IR produced contains both the scheduling logic, required to communicate data dependencies to low-level parallel pipelined hardware/API like Vulkan, and the execution logic, encoding dense computation on the hardware in the form of hardware/API-specific binaries like SPIR-V.
The architecture of IREE is best illustrated by the following picture:
Being compilation-based means IREE does not have a traditional runtime that dispatches “ops” to their fat kernel implementations. What IREE provides is a toolbox for different deployment scenarios. It scales from running generated code on a particular API (such as emitting C code calling external DSP kernels), to a HAL (Hardware Abstraction Layer) that allows the same generated code to target multiple APIs (like Vulkan and Direct3D 12), to a full VM allowing runtime model loading for flexible deployment options and heterogeneous execution.
IREE aims to
IREE is still at its early stage; we have lots of exciting future plans. Please check out the long-term design roadmap and short-term focus areas.
We use GitHub Projects to track various IREE components and GitHub Milestones for major features and quarterly plans. Please check out for updated information.
| CI System | Build System | Platform | Component | Status |
|---|---|---|---|---|
| Kokoro | Bazel | Linux | Core |  |
| Kokoro | Bazel | Linux | Bindings |  |
| Kokoro | Bazel | Linux | Integrations |  |
| Kokoro | CMake | Linux | Core + Bindings |  |
IREE is licensed under the terms of the Apache license. See LICENSE for more information.