#machine-learning #ai #ml

ort

A Rust wrapper for ONNX Runtime 1.14 - Optimize and Accelerate Machine Learning Inferencing

8 stable releases

new 1.14.3 Mar 21, 2023
1.14.2 Mar 17, 2023
1.14.0-beta.0 Feb 12, 2023
1.14.0-alpha.0 Jan 18, 2023
1.13.1 Nov 27, 2022

#10 in Machine learning

Download history 40/week @ 2022-11-28 3/week @ 2022-12-05 6/week @ 2022-12-12 18/week @ 2022-12-19 31/week @ 2022-12-26 17/week @ 2023-01-02 49/week @ 2023-01-09 57/week @ 2023-01-16 19/week @ 2023-01-23 28/week @ 2023-01-30 44/week @ 2023-02-06 68/week @ 2023-02-13 36/week @ 2023-02-20 99/week @ 2023-02-27 82/week @ 2023-03-06 97/week @ 2023-03-13

323 downloads per month
Used in 3 crates

MIT/Apache

1MB
23K SLoC

Rust bindings for ONNX Runtime

Coverage Results GitHub Workflow Status Crates.io

ort is an (unofficial) ONNX Runtime 1.14 wrapper for Rust based on the now-inactive onnxruntime-rs. ONNX Runtime accelerates ML inference on both CPU & GPU.

See the docs for more detailed information and the examples. If you have any questions, feel free to ask in the #📕|ort channel in the pyke Discord server or in GitHub Discussions.

Cargo features

Note: For developers using ort in a library (if you are developing an app, you can skip this part), it is heavily recommended to use default-features = false to avoid bringing in unnecessary bloat. Cargo features are additive. Users of a library that requires ort with default features enabled will not be able to remove those features, and if the library isn't using them, it's just adding unnecessary bloat. This is especially true for the fetch-models feature, which you should almost never use in production. Instead, you should enable ort's default features in your dev dependencies only, and instruct downstream users to include ort = { version = "...", features = [ "download-binaries" ] } in their dependencies if they need it.

  • fetch-models (default): Enables fetching models from the ONNX Model Zoo. Useful for quick testing with some common models like YOLOv4, GPT-2, and ResNet. Not recommended in production.
  • download-binaries (default): Enables downloading binaries via the download strategy. If disabled, the default behavior will be the system strategy.
  • copy-dylibs (default): Copies the dynamic libraries to the Cargo build folder - see shared library hell.
  • half (default): Enables support for using float16/bfloat16 tensors in Rust.
  • generate-bindings: Update the bindings to ONNX Runtime using bindgen. Requires libclang.
  • load-dynamic: Loads the ONNX Runtime binaries at runtime without a link dependency on them. The path to the binary can be controlled with the environment variable ORT_DYLIB_PATH=/path/to/libonnxruntime.so. This is heavily recommended, as it mitigates the shared library hell.

How to get binaries

You can use either the 'traditional' way, involving a strategy, or the new (and preferred) way, using load-dynamic.

  • Strategies: Links to provided or downloaded dynamic libraries; see below. This is useful for static linking and quick prototyping (making use of the download strategy), but might cause more headaches than load-dynamic.
  • load-dynamic: This doesn't link to any dynamic libraries, instead loading the libraries at runtime using dlopen(). This can be used to control the path to the ONNX Runtime binaries (meaning they don't always have to be directly next to your executable), and avoiding the shared library hell. To use this, enable the load-dynamic Cargo feature, and set the ORT_DYLIB_PATH environment variable to the path to your onnxruntime.dll/libonnxruntime.so/libonnxruntime.dylib - you can also use relative paths like ORT_DYLIB_PATH=./libonnxruntime.so (it will be relative to the executable). For convenience, you should download or compile ONNX Runtime binaries, put them in a permanent location, and set the environment variable permanently.

Strategies

There are 2 'strategies' for obtaining and linking ONNX Runtime binaries. The strategy can be set with the ORT_STRATEGY environment variable.

  • download (default): Downloads prebuilt ONNX Runtime from Microsoft. Only a few execution providers are available for download at the moment, namely CUDA and TensorRT. These binaries may collect telemetry. In the future, pyke may provide binaries with telemetry disabled and more execution providers available.
  • system: Links to ONNX Runtime binaries provided by the system or a path pointed to by the ORT_LIB_LOCATION environment variable. ort will automatically link to static or dynamic libraries depending on what is available in the ORT_LIB_LOCATION folder.

Execution providers

To use other execution providers, you must explicitly enable them via their Cargo features, listed below. Some EPs are not currently implemented due to a lack of hardware for testing; please open an issue if your desired EP has a ⚠️

  • ✅ cuda: Enables the CUDA execution provider for Maxwell (7xx) NVIDIA GPUs and above. Requires CUDA v11.6+.
  • ✅ tensorrt: Enables the TensorRT execution provider for GeForce 9xx series NVIDIA GPUs and above; requires CUDA v11.4+ and TensorRT v8.4+.
  • ⚠️ openvino: Enables the OpenVINO execution provider for 6th+ generation Intel Core CPUs.
  • ✅ onednn: Enables the Intel oneDNN execution provider for x86/x64 targets.
  • ✅ directml: Enables the DirectML execution provider for Windows x86/x64 targets with dedicated GPUs supporting DirectX 12.
  • ⚠️ snpe: Enables the SNPE execution provider for Qualcomm Snapdragon CPUs & Adreno GPUs.
  • ⚠️ nnapi: Enables the Android Neural Networks API (NNAPI) execution provider.
  • ✅ coreml: Enables the CoreML execution provider for macOS/iOS targets.
  • ⚠️ xnnpack: Enables the XNNPACK backend for WebAssembly and Android.
  • ⚠️ migraphx: Enables the MIGraphX execution provider AMD GPUs.
  • ⚠️ rocm: Enables the ROCm execution provider for AMD ROCm-enabled GPUs. (#16)
  • ✅ acl: Enables the ARM Compute Library execution provider for multi-core ARM v8 processors.
  • ⚠️ armnn: Enables the ArmNN execution provider for ARM v8 targets.
  • ⚠️ tvm: Enables the preview Apache TVM execution provider.
  • ⚠️ rknpu: Enables the RKNPU execution provider for Rockchip NPUs.
  • ⚠️ vitis: Enables Xilinx's Vitis-AI execution provider for U200/U250 accelerators.
  • ⚠️ cann: Enables the Huawei Compute Architecture for Neural Networks (CANN) execution provider.

Note that the download strategy only provides some execution providers, namely CUDA and TensorRT for Windows & Linux. You'll need to compile ONNX Runtime from source and use the system strategy to point to the compiled binaries to enable other execution providers.

Execution providers will attempt to be registered in the order they are passed, silently falling back to the CPU provider if none of the requested providers are available. If you must know whether an EP is available, you can use ExecutionProvider::cuda().is_available().

For prebuilt Microsoft binaries, you can enable the CUDA or TensorRT execution providers for Windows and Linux via the cuda and tensorrt Cargo features respectively. Microsoft does not provide prebuilt binaries for other execution providers, and thus enabling other EP features will fail when ORT_STRATEGY=download. To use other execution providers, you must build ONNX Runtime from source.

Projects using ort ❤️

  • Bloop uses ort to power their semantic code search feature.
  • pyke Diffusers uses ort for efficient Stable Diffusion image generation on both CPUs & GPUs.
  • edge-transformers uses ort for accelerated transformer model inference at the edge.

FAQ

I'm using a non-CPU execution provider, but it's still using the CPU!

ort is designed to fail gracefully when an execution provider is not available. It logs failure events through tracing, thus you'll need a library that subscribes to tracing events to see the logs. The simplest way to do this is to use tracing-subscriber.

Add tracing-subscriber to your Cargo.toml:

[dependencies]
tracing-subscriber = { version = "0.3", features = [ "env-filter", "fmt" ] }

In your main function:

fn main() {
    tracing_subscriber::fmt::init();
}

Set the environment variable RUST_LOG to ort=debug to see all debug messages from ort; this will look like:

  • Windows (PowerShell): $env:RUST_LOG = 'ort=debug'; cargo run
  • Windows (Command Prompt): use PowerShell ;)
  • macOS & Linux: RUST_LOG="ort=debug" cargo run

My app exits with "status code 0xc000007b" without logging anything!

You probably need to copy the ONNX Runtime DLLs to the same path as the executable.

  • If you are running a binary (cargo run), copy them to e.g. target/debug
  • If you are running an example (cargo run --example xyz), copy them to e.g. target/debug/examples
  • If you are running tests (cargo test), copy them to e.g. target/debug/deps

Alternatively, you can use the load-dynamic feature to avoid this.

"thread 'main' panicked at 'assertion failed: (left != right)"

Most of the time this is because Windows ships its own (typically older) version of ONNX Runtime. Make sure you've copied the ONNX Runtime DLLs to the same folder as the exe.


Shared library hell

If using shared libraries (as is the default with ORT_STRATEGY=download), you may need to make some changes to avoid issues with library paths and load orders, or preferably use the load-dynamic feature to avoid all of this.

Windows

Some versions of Windows come bundled with an older vesrion of onnxruntime.dll in the System32 folder, which will cause an assertion error at runtime:

The given version [14] is not supported, only version 1 to 13 is supported in this build.
thread 'main' panicked at 'assertion failed: `(left != right)`
  left: `0x0`,
 right: `0x0`', src\lib.rs:106:5
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

The fix is to copy the ONNX Runtime DLLs into the same directory as the binary, since DLLS in the same folder as the main executable resolves before system DLLs. ort can automatically copy the DLLs to the Cargo target folder with the copy-dylibs feature, though this fix only works for binary Cargo targets (cargo run). When running tests/benchmarks/examples for the first time, you'll have to manually copy the target/debug/onnxruntime*.dll files to target/debug/deps/ for tests & benchmarks or target/debug/examples/ for examples.

Linux

Running a binary via cargo run should work without copy-dylibs. If you'd like to use the produced binaries outside of Cargo, you'll either have to copy libonnxruntime.so to a known lib location (e.g. /usr/lib) or enable rpath to load libraries from the same folder as the binary and place libonnxruntime.so alongside your binary.

In Cargo.toml:

[profile.dev]
rpath = true

[profile.release]
rpath = true

# do this for all profiles

In .cargo/config.toml:

[target.x86_64-unknown-linux-gnu]
rustflags = [ "-Clink-args=-Wl,-rpath,\\$ORIGIN" ]

# do this for all Linux targets as well

macOS

macOS has the same limitations as Linux. If enabling rpath, note that the rpath should point to @loader_path rather than $ORIGIN:

# .cargo/config.toml
[target.x86_64-apple-darwin]
rustflags = [ "-Clink-args=-Wl,-rpath,@loader_path" ]

Dependencies

~2.4–6MB
~130K SLoC