ONNX Runtime Web We are introducing ONNX Runtime Web (ORT Web), a new feature in ONNX Runtime to enable JavaScript developers to run and deploy machine learning models in browsers. It also helps enable new classes of on-device computation. ORT Web will be replacing the soon to be deprecated onnx.js, with improvements such as a more consistent developer experience between Read More >>>>
Distributed Data Parallel Model training has been and will be in the foreseeable future one of the most frustrating things machine learning developers face. It takes quite a long time and people can’t really do anything about it. If you have the luxury (especially at this moment of time) of having multiple GPUs, you are likely to find Read More >>>>
scalable processes to deploy ML models As organizations adopt machine learning in production, they face growing challenges that arise when the number of production machine learning models starts to increase. In this article, we provide a practical tutorial that will enable AI practitioners to leverage production-ready workflows to deploy their machine learning models at scale. More specifically, we will demonstrate the Read More >>>>
ROCm Open Software Platform ONNX Runtime is an open-source project that is designed to accelerate machine learning across a wide range of frameworks, operating systems, and hardware platforms. Today, we are excited to announce a preview version of ONNX Runtime in release 1.8.1 featuring support for AMD Instinct™ GPUs facilitated by the AMD ROCm™ open software platform. Users can now Read More >>>>
How to migrate and modernize Linux workloads With extensive support for all major Linux distributions including Red Hat, SUSE, Ubuntu, CentOS, Debian, and managed platform-as-a-service (PaaS) offerings for open source databases like Azure Database for MySQL, Azure Database for PostgreSQL, and Azure Database for MariaDB—it’s no surprise that Linux is the fastest growing platform on Azure. Furthermore, Azure Migrate makes the discovery, assessment, migration, and modernization of apps, databases, Read More >>>>
PyTorch PyTorch is an increasingly popular open-source deep learning framework that accelerates AI innovations from research to production. At Microsoft, we use PyTorch to power products such as Bing and Azure Cognitive Services and we actively contribute to several PyTorch open-source projects, including PyTorch Profiler, ONNX Runtime, DeepSpeed, and more. Today, we’re announcing a new initiative in collaboration with Facebook—the PyTorch Enterprise Support Program. This Read More >>>>
eBPF work on Windows eBPF is a well-known but revolutionary technology—providing programmability, extensibility, and agility. eBPF has been applied to use cases such as denial-of-service protection and observability. Over time, a significant ecosystem of tools, products, and experience has been built up around eBPF. Although support for eBPF was first implemented in the Linux kernel, there has been increasing interest Read More >>>>
What is ONNX Runtime The performance improvements provided by ONNX Runtime powered by Intel® Deep Learning Boost: Vector Neural Network Instructions (Intel® DL Boost: VNNI) greatly improves performance of machine learning model execution for developers. In the past, machine learning models mostly relied on 32-bit floating point instructions using AVX512. Now, machine learning models can use 8-bit integer instructions Read More >>>>
Deploying SUSE SAP HA Automation Maintaining a competitive advantage often depends on how quickly you can deliver new services. SAP applications are designed to help companies analyze data to anticipate new requirements and rapidly deliver new products and services. This gives you the ability to keep existing customers happy while attracting new business. In line with SUSE’s vision is to Read More >>>>
Maintainers Spotlight We hear about open source projects every day, but we rarely hear from the people who maintain them. Maintaining an open source project is a full-time, often thankless, job. The Maintainers Spotlight blog series is an opportunity to highlight the essential role maintainers play in moving projects and communities forward. In true open source fashion, Read More >>>>
Scikit-learn model inference with ONNX Runtime Scikit-learn is one of the most useful libraries for general machine learning in Python. To minimize the cost of deployment and avoid discrepancies, deploying scikit-learn models to production usually leverages Docker containers and pickle, the object serialization module of the Python standard library. Docker is a good way to create consistent environments and pickle saves Read More >>>>
Kubernetes Clusters Managing Kubernetes clusters at scale across a variety of infrastructures is—well—even harder. The Kubernetes community project Cluster API (CAPI) enables users to manage fleets of clusters across multiple infrastructure providers. The Cluster API Provider for Azure (CAPZ) is the solution for users who need to manage Kubernetes clusters on Azure IaaS. In the past, we have recommended AKS Engine Read More >>>>