ONNX Runtime v1.15.0

[snnn]

Announcements

Starting from the next release(ONNX Runtime 1.16.0), at operating system level we will drop the support for

• iOS 11 and below. iOS 12 will be the minimum supported version.
• CentOS 7, Ubuntu 18.04, and any Linux distro without glibc version >=2.28.

At compiler level we will drop the support for

• GCC version <= 9
• Visual Studio 2019

Also, we will remove the onnxruntime_DISABLE_ABSEIL build option since we will upgrade protobuf and the new protobuf version will need abseil.

General

• Added support for ONNX Optional type in C# API
• Added collectives to support multi-GPU inferencing
• Updated macOS build machines to macOS-12, which comes with Xcode 14.2 and we should stop using Xcode 12.4
• Added Python 3.11 support (deprecate 3.7, support 3.8-3.11) in packages for Onnxruntime CPU, Onnxruntime-GPU, Onnxruntime-directml, and onnxruntime-training.
• Updated to CUDA 11.8. ONNX Runtime source code is still compatible with CUDA 11.4 and 12.x.
• Dropped the support for Windows 8.1 and below
• Eager mode code and onnxruntime_ENABLE_EAGER_MODE cmake option are deleted.
• Upgraded Mimalloc version from 2.0.3 to 2.1.1
• Upgraded protobuf version from 3.18.3 to 21.12
• New dependency: cutlass, which is only used in CUDA/TensorRT packages.
• Upgraded DNNL from 2.7.1 to 3.0

Build System

• On POSIX systems by default we disallow using "root" user to build the code. If needed, you can append "--allow_running_as_root" to your build command to bypass the check.
• Add the support for building the source natively on Windows ARM64 with Visual Studio 2022.
• Added a Gradle wrapper and updated Gradle version from 6.8.3 to 8.0.1. (Gradle is the tool for building ORT Java package)
• When doing cross-compiling, the build scripts will try to download a prebuit protoc from Github instead of building the binary from source. Because now protobuf has many dependencies. It is not easy to setup a build environment for protobuf.

Performance

• Improved string marshalling and reduce GC pressure
• Added a build option to allow using a lock-free queue in threadpool for improved CPU utilization
• Fix CPU memory leak due to external weights
• Added fused decoder multi-head attention kernel to improve GPT and decoder models(like T5, Whisper)
• Added packing mode to improve encoder models with inputs of large padding ratio
• Improved generation algorithm (BeamSearch, TopSampling, GreedySearch)
• Improved performance for StableDiffusion, ViT, GPT, whisper models

Execution Providers

Two new execution providers: JS EP and QNN EP.

TensorRT EP

• Official support for TensorRT 8.6
• Explicit shape profile overrides
• Support for TensorRT plugins via ORT custom op
• Improve support for TensorRT options (heuristics, sparsity, optimization level, auxiliary stream, tactic source selection etc.)
• Support for TensorRT timing cache
• Improvements to our test coverage, specifically for opset16-17 models and package pipeline unit test coverage.
• Other misc bugfixes and improvements.

OpenVINO EP

• Support for OpenVINO 2023.0
• Dynamic shapes support for iGPU
• Changes to OpenVINO backend to improve first inference latency
• Deprecation of HDDL-VADM and Myriad VPU support
• Misc bug fixes.

QNN EP

• Initial Public preview release

DirectML EP:

• Updated to DirectML 1.12
• Opset 16-17 support

AzureEP

• Added support for OpenAI whisper model
• Available in a Nuget pkg in addition to Python

Mobile

New packages

• Swift Package Manager for onnxruntime
• Nuget package for onnxruntime-extensions (supports Android/iOS for MAUI/Xamarin)
• React Native package for onnxruntime can optionally include onnxruntime-extensions

Pre/Post processing

• Added support for built-in pre and post processing for NLP scenarios: classification, question-answering, text-prediction

• Added support for built-in pre and post processing for Speech Recognition (Whisper)

• Added support for built-in post processing for Object Detection (YOLO). Non-max suppression, draw bounding boxes

• Additional CoreML and NNAPI kernels to support customer scenarios

  • NNAPI: BatchNormalization, LRN
  • CoreML: Div, Flatten, LeakyRelu, LRN, Mul, Pad, Pow, Sub

Web

• [preview] WebGPU support
• Support building the source code with "MinGW make" on Windows.

ORT Training

On-device training:

• Official package for On-Device Training now available. On-device training extends ORT Inference solutions to enable training on edge devices.
• APIs and Language bindings supported for C, C++, Python, C#, Java.
• Packages available for Desktop and Android.
• For custom builds refer build instructions.

Others

• Added graph optimizations which leverage the sparsity in the label data to improve performance. With these optimizations we see performance gains ranging from 4% to 15% for popular HF models over baseline ORT.
• Vision transformer models like ViT, BEIT and SwinV2 see upto 44% speedup with ORT Training+ DeepSpeed over PyTorch eager mode on AzureML.
• Added optimizations for SOTA models like Dolly and Whisper. ORT Training + DS now gives ~17% speedup for Whisper and ~4% speedup for Dolly over PyTorch eager mode. Dolly optimizations on main branch show a ~40% over eager mode.

Known Issues

• The onnxruntime-training 1.15.0 packages published to pypi.org were actually built in Debug mode instead of Release mode. You can get the right one from https://download.onnxruntime.ai/ . We will fix the issue in the next patch release.
• XNNPack EP does not work on x86 CPUs without AVX-512 instructions, because we used wrong alignment when allocating buffers for XNNPack to use.
• The CUDA EP source code has a build error when CUDA version <11.6. See [Build] Unable to suppress unused variable #16000.
• The onnxruntime-training builds are missing the training header files.

Contributions

Contributors to ONNX Runtime include members across teams at Microsoft, along with our community members:
snnn, fs-eire, edgchen1, wejoncy, mszhanyi, PeixuanZuo, pengwa, jchen351, cloudhan, tianleiwu, PatriceVignola, wangyems, adrianlizarraga, chenfucn, HectorSVC, baijumeswani, justinchuby, skottmckay, yuslepukhin, RandyShuai, RandySheriffH, natke, YUNQIUGUO, smk2007, jslhcl, chilo-ms, yufenglee, RyanUnderhill, hariharans29, zhanghuanrong, askhade, wschin, jywu-msft, mindest, zhijxu-MS, dependabot[bot], xadupre, liqunfu, nums11, gramalingam, Craigacp, fdwr, shalvamist, jstoecker, yihonglyu, sumitsays, stevenlix, iK1D, pranavsharma, georgen117, sfatimar, MaajidKhan, satyajandhyala, faxu, jcwchen, hanbitmyths, jeffbloo, souptc, ytaous kunal-vaishnavi

[cbourjau]

Thanks for the release 🎉! FYI, the link at the bottom is broken. It should probably be https://github.com/microsoft/onnxruntime/releases/tag/v1.15.0

[yaysummeriscoming]

@snnn Great stuff! The packing in particular is very exciting, however I can find no information or documentation on how it works? I assume it's similar to fastertransformer?

[ratan-prasad]

@snnn thanks for the update. Very interesting.
I have a question for this announcement:
**"Starting from the next release(ONNX Runtime 1.16.0), at operating system level we will drop the support for

iOS 11 and below. iOS 12 will be the minimum supported version.
CentOS 7, Ubuntu 18.04, and any Linux distro without glibc version >=2.28."**
Our team has been building ONNX Runtime from source using Manylinux2014 (https://peps.python.org/pep-0599/). Manylinux2014 supports GLIBC >= 2.17.
So which manylinux version we can use to build ONNX Runtime version 1.16.0?

[snnn]

CentOS7 will be EOL in June 2024. We need to deprecate the uses before that date, and it won't be an easy job for us since onnxruntime has around 20 different EPs and a huge number of pipelines/docker files. I would prefer to start it earlier than waiting to the last minute.

[snnn]

After manylinux2014, you can use manylinux_x_y. For example, manylinux_2_28. The two numbers are just a glibc version. As long as the combination is supported by https://github.com/pypa/auditwheel, it should be fine.

[ratan-prasad]

Thanks @snnn.
So for ONNX Runtime v1.15, we can build from source using manylinux2014,
for ONNX Runtime v1.16, we will use manylinux_2_28.

[snnn]

Yes. That's the plan.

[randombenj]

Thanks for Python 3.11 support! 🎉

[kushalpatil07]

The onnxruntime-training builds are missing the training header files. I am able to use the library only after copying the orttraining headers from the repo.

[baijumeswani]

Thanks for reporting this. Apologies for the inconvenience. This might warrant a patch release with the fix. I will update this thread once I find what went wrong.

[brian-at-pieces]

Thanks for the T5 and beam search improvements! Do you have any benchmarks on expected speedup for either of these?

[umi-soft]

not support cuda12.1？ see #16386

[snnn]

Build from source, then it should work.

[snnn]

I just published a patch release for it: https://github.com/microsoft/onnxruntime/releases/tag/v1.15.1

[cyrusbehr]

Is there a linux aarch32 binary?

[snnn]

Sorry, no.

[cyrusbehr]

ok