März 2023 - Seite 35 von 46 - Cloud Computing Köln

Extract robust insights from image and video content with Azure Cognitive Service for Vision

We are pleased to announce the public preview of Microsoft’s Florence foundation model, trained with billions of text-image pairs and integrated as cost-effective, production-ready computer vision services in Azure Cognitive Service for Vision. The improved Vision Services enables developers to create cutting-edge, market-ready, responsible computer vision applications across various industries. Customers can now seamlessly digitize, analyze, and connect their data to natural language interactions, unlocking powerful insights from their image and video content to support accessibility, drive acquisition through SEO, protect users from harmful content, enhance security, and improve incident response times.

Microsoft was recently named a Leader in the IDC MarketScape: Worldwide General-Purpose Computer Vision AI Software Platforms 2022 Vendor Assessment (doc #US49776422, November 2022). The new Vision Services improves content discoverability with automatic captioning, smart cropping, classifying, background removal, and searching for images. Furthermore, users can track movements, analyze environments, and receive real-time alerts with responsible AI controls.

Reddit will be using Vision Services to generate captions for hundreds of millions of images on its platform. Tiffany Ong, Reddit Product Manager of Consumer Product has said,

“With Microsoft’s Vision technology, we are making it easier for users to discover and understand our content. The newly created image captions make Reddit more accessible for everyone and give redditors more opportunities to explore our images, engage in conversations, and ultimately build connections and a sense of community."

Microsoft is harnessing the power of the new Vision Services in Microsoft 365 apps like Teams, PowerPoint, Outlook, Word, Designer, OneDrive, in addition to the Microsoft Datacenter. Microsoft Teams is driving innovation in the digital space with the help of segmentation capabilities, taking virtual meetings to the next level. PowerPoint, Outlook, and Word leverage image captioning for automatic alt-text to improve accessibility. Microsoft Designer and OneDrive are using improved image tagging, image search, and background generation to simplify image discoverability and editing. Microsoft Datacenters are leveraging Vision Services to enhance security and infrastructure reliability.

At this week's Microsoft Ability Summit, companies will learn how they can improve the accessibility of their visual content. We’ll share the future of our Seeing AI app and LinkedIn will share the benefits of utilizing Vision Services to deliver automatic alt-text descriptions for image analysis. As a preview, Jennison Asuncion, LinkedIn’s Head of Accessibility Engineering Evangelism has said,

“More than 40 percent of LinkedIn’s feed posts include at least one image. We want every member to have equal access to opportunity and are committed to ensuring that we make images accessible to our members who are blind or who have low vision so they can be a part of the online conversation. With Azure Cognitive Service for Vision, we can provide auto-captioning to edit and support alt. text descriptions. I'm excited about this new experience because now, not only will I know my colleague shared a picture from an event they attended, but that my CEO Ryan Roslansky is also in the picture.”

Try out the new out-of-the-box features our customers are using in Vision Studio:

Dense captions: Automatically deliver rich captions, design suggestions, accessible alt-text, SEO optimization, and intelligent photo curation to support digital content.
Image retrieval: Improve search recommendations and advertisements with natural language queries that seamlessly measure the similarity between images and text.

Background removal: Transform the look and feel of images by easily segmenting people and objects from their original background, replacing them with a preferred background scene.
Model customization: Lower costs and time to deliver custom models that match unique business demands at high precision, and with just a handful of images.
Video summarization (Video TL;DR): Search and interact with video content in the same intuitive way you think and write. Locate relevant content without the need for additional metadata.

Innovate responsibly

Review the responsible AI principles to learn how we are committed to developing AI systems that help make the world more accessible. We are focused on helping organizations take full advantage of AI, and we are investing heavily in programs that provide technology, resources, and expertise to empower those working to create a more sustainable, safe, and accessible world.

Get started today with Azure Cognitive Service for Vision

Revolutionize your computer vision applications with improved efficiency, accuracy, and accessibility in image and video processing, at the same low price. Visit Vision Studio to try out our latest demos.

Learn more about Azure Cognitive Service for Vision:

Get started with Microsoft Learn to build skills.
Watch the Florence showcase shared at the 2022 CVPR conference.

Quelle: Azure

8. März 2023

da Agency

Discover the latest innovations at Azure Open Source Day 2023

This post has been coauthored by Nikisha Reyes-Grange and Katie Fritsch.

“Tech companies born with an open-source mentality get it. It’s our ability to work together that makes our dreams believable, and ultimately achievable; we must learn to build on the ideas of others.”—Satya Nadella, CEO, Microsoft.

Microsoft has always been a developer-first organization, and we strive to make tools and platforms that serve developers. Open Source on Azure gives developers the freedom to build next-generation intelligent applications where they want, and when they want. Thanks to the power of open source, developers can now innovate faster and more securely on Azure. Microsoft helps you write more secure code, respond quickly to vulnerabilities in your software supply chain, and adopt the latest best practices to secure your development environments—all with a commitment to open source and support for all languages and frameworks.

By combining the most beloved coding and collaboration tools—Linux, GitHub, Visual Studio Code, along with the Azure platform—the open-source software in the Azure ecosystem aids developer productivity with best-in-class tools for code-to-cloud development.

Azure Open Source Day highlights Microsoft’s commitment to open source and focused on how Open Source technologies can be used to build intelligent apps faster and with more flexibility.

Opening panel: How open source and AI impact software development in the cloud

We are kicking off our show with a panel of thought leaders from Github, HashiCorp, Microsoft, and Redis to discuss how open source has continued to evolve software development, impacts on software supply chain and security, and how new AI capabilities may impact the future.

How Open Source on Azure builds an intelligent app

Today, we are excited to be showcasing a brand-new, intelligent, cloud-native application that connects owners with their lost pets using fine-tuned machine learning. Instead of printing posters, use an advanced machine learning image classification model, fine-tuned by the images on your camera roll. With this trained machine learning model, when a pet is found, you can instantly snap a photo that will match the model and connect you to the owner.

We are leveraging open source technologies to make sure that our application is using the latest and greatest technologies.

The app's frontend is a dotnet Blazor app,1 with a Python backend. The frontend and backend communicate using the Distributed Application Runtime (Dapr)2 that provides application programming interfaces (APIs) that simplify microservice connectivity. The backend uses a pre-built vision model from Hugging Face3, fine-tuned directly through Azure Machine Learning for model training and prediction. The whole app is deployed using Bicep4 templates and runs on Azure Kubernetes Service. The Kubernetes Event Driven Autoscaling (KEDA) is used to provide autoscaling capabilities based on the number of messages being sent through Dapr.

The app’s data layer was built with Azure Cosmos DB and takes advantage of the autoscale feature that matches database capacity with traffic demands. So as the app grows, the database automatically grows with it. With autoscale, the risk of rate-limiting and over-provisioning is eliminated, app performance is maintained, and the developer doesn’t have to monitor and manage database capacity manually. Developers using MySQL will soon enjoy similar benefits, with the general availability of IOPS Autoscaling in Azure Database for MySQL—flexible server, Business Critical tier coming next month. This feature will save time and money by resizing input/output operations per second (IOPS) depending on workload needs. We’ve also made it easier to develop low-code apps with MySQL data and visualize MySQL data with the integrations of Power Apps and Power BI Desktop with Azure Database for MySQL—flexible server, now in public preview and general availability respectively.

Developers using Azure Cache for Redis Enterprise can now use the RedisJSON module on active, geo-replicated caches. Currently in preview, this feature simplifies development and minimizes downtime by enabling a single operation to read, write, and store JSON documents while simultaneously syncing data across all active regions.

PostgreSQL users can now apply enhanced security to their databases, with the general availability of Microsoft Azure Active Directory and customer-managed keys in Azure Database for PostgreSQL—flexible server. Database user identities and access, along with encryption key permissions and life cycles, can now be centrally managed to make it easier to build secure apps.

Compute scaling in the demo is provided by Azure Virtual Machine Scale Sets (VMSS) Flex to deploy GitHub Actions self-hosted runners on new Arm-based virtual machines. VMSS Flex allows you to easily manage and mix different virtual machine sizes and SKUs, including both Spot and standard virtual machines. Recent additions to the Azure portfolio include next-generation burstable Arm-based Bpsv2 virtual machines, which provide a low-cost option for workloads that typically run at a low to moderate baseline central processing unit (CPU) utilization, and Intel-based DIsv5 virtual machines that can deliver up to 30 percent increased performance and better price-performance than the Fsv2 virtual machines. Both virtual machine series feature broad support for Linux distributions.

The app uses a pre-trained vision transformer model obtained from Hugging Face for image classification tasks. Developers and data scientists can now use foundation models in Azure Machine Learning to easily start their data science works to fine-tune and deploy foundation models from Hugging Face using Azure Machine Learning components and pipelines. This feature, currently in preview, provides organizations with a comprehensive repository of popular large AI models from Hugging Face through the built-in Azure Machine Learning registries, supporting various tasks such as classification, summarization, question answering, and translation. It simplifies the process of data pre-processing and adaptation of model training scripts, freeing data scientists from the overhead of setting up and managing underlying infrastructure and environment dependencies. Read this blog to learn more about the latest open-source capabilities from Azure AI.

Unleashing the AI technology wave: Training large language models at scale

AI is changing every industry and is top of mind for developers. Most companies have leveraged AI to improve efficiency and costs. Large AI applications leveraging natural language processing (NLP), automatic speech recognition (ASR), and text-to-speech (TTS) are becoming prevalent, but what powers these applications is the underlying infrastructure optimized for large AI workloads. As mentioned in the post announcing the general availability of Microsoft Azure OpenAI Service, Azure is the best place to build AI workloads. This session highlights the partnership between Microsoft and NVIDIA and how Azure’s AI infrastructure and Azure Machine Learning were built for speed.

Azure NDm A100 v4-series virtual machines are Azure’s flagship graphics processing unit (GPU) offerings and were used to run the model's new NVIDIA NeMo Megatron framework and test the limits of this series. Microsoft ran a 530B-parameter benchmark on 175 virtual machines, resulting in a training time per step of as low as 55.7 seconds. This benchmark measures the compute efficiency and how it scales by measuring the time taken per step to train the model after a steady state is reached, with a mini-batch size of one. The InfiniBand HDR provided superior communication between nodes without increased latency and was critical to the ludicrous speed performance.

The open source future with Web3

Azure Open Source Day will conclude with a fascinating fireside chat between Kathleen Mitford, CVP Azure Marketing and Donovan Brown, Partner Program Manager, Azure Open Source Incubations on the Open Source Future with Web3. The open and decentralized nature of Web3 can be a natural fit with open source philosophy, which is an exciting and developing space for new innovations.

Web3 refers to another evolution of the internet, which may be more decentralized. It is built on a blockchain, which is a distributed ledger technology that enables the creation of a secure and transparent way to transfer and store digital assets. Microsoft has first-party solutions and a rich partner ecosystem to help you build using Web3.

DevOps best practices are just as important, if not more important in the Web3 world as they are in the Web2 world. Azure has the key tools a developer needs—from Azure Static Web Apps, purposely built for your App, to GitHub Actions, and Azure DevOps, and Visual Studio Code. In the future, many organizations may build solutions with a combination of Web3 and Web2 working together. Go deeper into Web3 with Donovan Brown’s 10-part blog series on how to build a DevOps pipeline for the Ethereum Blockchain.

Also, with the power of AI, you can ask ChatGPT to create a "hello world" sample in any language and copy and paste the code into your project. When you go to modify the code, have GitHub copilot help you make the changes using all the best practices. You can do all of this inside a GitHub Codespace configured with all your favorite tools, frameworks, and Visual Studio Code extensions installed. Then, you can use Azure DevOps or GitHub Actions to deploy the application to Azure. If you choose to build a Web3 application, Microsoft has the best tools and cloud services to support you.

Upcoming developer community events

Local Azure Open Source Day events

Check out if there is a local Azure Open Source Day event near you.

Azure Cosmos DB Conf—8:00-11:00 AM PT, March 28, 2023

This free, virtual event for developers showcases what members of the community are building with Azure Cosmos DB for NoSQL, PostgreSQL, MongoDB, and Apache Cassandra.

Citus Con: An event for Postgres 2023—April 18-19, 2023

A virtual developer event all about what you can do with the world's most advanced open source relational database. Over two days, you'll hear from open source users and experts in PostgreSQL and Citus about unique ways to use Postgres.

Watch Azure Open Source Day on-demand.

Learn more

Microsoft is working to collectively empower every person and every organization on the planet to achieve more. Whether it is contributing to projects, releasing new open source projects, or using open source to make our products and services work better, Microsoft is proud to be participating in open source communities more than ever before.

We are committed to open source at Microsoft. We contribute to Linux, Kubernetes, Visual Studio Code, and serve in open source organizations like the Cloud Native Computing Foundation (CNCF) or Open Source Security Foundation (OpenSSF). At Azure Open Source Day, we shared our latest work to enable developers to develop flexibly and innovate quickly on Azure.

Learn more about Open Source on Azure.

1dotnet Blazor

2Distributed Application Runtime

3Hugging Face

4Bicep

Quelle: Azure

8. März 2023

da Agency

Exploring open-source capabilities in Azure AI

This post was co-authored by Richard Tso, Director of Product Marketing, Azure AI

Open-source technologies have had a profound impact on the world of AI and machine learning, enabling developers, data scientists, and organizations to collaborate, innovate, and build better AI solutions. As large AI models like GPT-3.5 and DALL-E become more prevalent, organizations are also exploring ways to leverage existing open-source models and tools without needing to put a tremendous amount of effort into building them from scratch. Microsoft Azure AI is leading this effort by working closely with GitHub and data science communities, and providing organizations with access to a rich set of open-source technologies for building and deploying cutting-edge AI solutions.

At Azure Open Source Day, we highlighted Microsoft’s commitment to open source and how to build intelligent apps faster and with more flexibility using the latest open-source technologies that are available in Azure AI.

Build and operationalize open-source State-of-the-Art models in Azure Machine Learning

Recent advancements in AI propelled the rise of large foundation models that are trained on a vast quantity of data and can be easily adapted to a wide variety of applications across various industries. This emerging trend provides a unique opportunity for enterprises to build and use foundation models in their deep learning workloads.

Today, we’re announcing the upcoming public preview of foundation models in Azure Machine Learning. It provides Azure Machine Learning with native capabilities that enable customers to build and operationalize open-source foundation models at scale. With these new capabilities, organizations will get access to curated environments and Azure AI Infrastructure without having to manually manage and optimize dependencies. Azure Machine learning professionals can easily start their data science tasks to fine-tune and deploy foundation models from multiple open-source repositories, starting from Hugging Face, using Azure Machine Learning components and pipelines. This service will provide you with a comprehensive repository of popular open-source models for multiple tasks like natural language processing, vision, and multi-modality through the Azure Machine Learning built in registry. Users can not only use these pre-trained models for deployment and inferencing directly, but they will also have the ability to fine-tune supported machine learning tasks using their own data and import any other models directly from the open-source repository.

The next generation of Azure Cognitive Services for Vision

Today, Azure Cognitive Services for Vision released its next generation of capabilities powered by the Florence large foundational model. This new Microsoft model delivers significant improvements to image captioning and groundbreaking customization capabilities with few-shot learning. Until today, model customization required large datasets with hundreds of images per label to achieve production quality for vision tasks. But, Florence is trained on billions of text-image pairs, allowing custom models to achieve high quality with just a few images. This lowers the hurdle for creating models that can fit challenging use cases where training data is limited.

Users can try the new capabilities of Vision underpinned by the Florence model through Vision Studio. This tool demonstrates a full set of prebuilt vision tasks, including automatic captioning, smart cropping, classifying images and a summarizing video with natural language, and much more. Users can also see how the tool helps track movements, analyze environments, and provide real-time alerts.

To learn more about the new Florence model in Azure Cognitive Services for Vision, please check out this announcement blog.

New Responsible AI Toolbox additions

Responsible AI is a critical consideration for organizations building and deploying AI solutions. Last year, Microsoft launched the Responsible AI Dashboard within the Responsible AI Toolkit, a suite of tools for a customized, responsible AI experience with unique and complementary functionalities available on GitHub and in Azure Machine Learning. We recently announced the addition of two new open-source tools designed to make the adoption of responsible AI practices more practical.

The Responsible AI Mitigations Library allows practitioners to experiment with different mitigation techniques more easily, while the Responsible AI Tracker uses visualizations to demonstrate the effectiveness of different mitigations for more informed decision-making. The new mitigations library bolsters mitigation by offering a means of managing failures that occur in data preprocessing. The library complements the toolbox’s Fairlearn fairness assessment tool, which focuses on mitigations applied during training time. The tracker allows practitioners to look at performance for subsets of data across iterations of a model to help them determine the most appropriate model for deployment. When used with other tools in the Responsible AI Toolbox, they offer a more efficient and effective means to help improve the performance of systems across users and conditions. These tools are made open source on GitHub and integrated into Azure Machine Learning.

Accelerate large-scale AI with Azure AI infrastructure

Azure AI Infrastructure provides massive scale-up and scale-out capabilities for the most advanced AI workloads in the world. This is a key factor as to why leading AI companies, including our partners at OpenAI continue to choose Azure to advance their AI innovation on Azure AI. Our results for training OpenAI's GPT-3 on Azure AI Infrastructure using Azure NDm A100 v4 virtual machines with NVIDIA’s open-source framework, NVIDIA NeMo Megatron, delivered a 530B-parameter benchmark on 175 virtual machines, resulting in a scalability factor of 95 percent. When Azure AI infrastructure is used together with a managed end-to-end machine learning platform, such as Azure Machine Learning, it provides the vast compute needed to enable organizations to streamline management and orchestration of large AI models and help bring them into production.

The full benchmarking report for GPT-3 models with the NVIDIA NeMo Megatron framework on Azure AI infrastructure is available here.

Optimized training framework to accelerate PyTorch model development

Azure is a preferred platform for widely used open-source framework—PyTorch. At Microsoft Ignite, we launched Azure Container for PyTorch (ACPT) within Azure Machine Learning, bringing together the latest PyTorch version with our best optimization software for training and inferencing, such as DeepSpeed and ONNX Runtime, all tested and optimized for Azure. All these components are already installed in ACPT and validated to reduce setup costs and accelerate training time for large deep learning workloads. ACPT curated environment allows our customers to efficiently train PyTorch models. The optimization libraries like ONNX Runtime and DeepSpeed composed within the container can increase production speed up from 54 percent to 163 percent over regular PyTorch workloads as seen on various Hugging Face models.

The chart shows ACPT that combines ONNX Runtime and DeepSpeed can increase production speed up to 54 percent to 163 percent over regular PyTorch workloads.

This month, we’re bringing a new capability to ACPT—Nebula. Nebula is a component in ACPT that can help data scientists to boost checkpoint savings time faster than existing solutions for distributed large-scale model training jobs with PyTorch. Nebula is fully compatible with different distributed PyTorch training strategies, including PyTorch Lightning, DeepSpeed, and more. In saving medium-sized Hugging Face GPT2-XL checkpoints (20.6 GB), Nebula achieved a 96.9 percent reduction in single checkpointing time. The speed gain of saving checkpoints can still increase with model size and GPU numbers. Our results show that, with Nebula, saving a checkpoint with a size of 97GB in a training job on 128 A100 Nvidia GPUs can be reduced from 20 minutes to 1 second. With the ability to reduce checkpoint times from hours to seconds—a potential reduction of 95 percent to 99.9 percent, Nebula provides a solution to frequent saving and reduction of end-to-end training time in large-scale training jobs.

The chart shows Nebula achieved a 96.9 percent reduction in single checkpointing time with GPT2-XL.

To learn more about Azure Container for PyTorch, please check out this announcement blog.

MLflow 2.0 and Azure Machine Learning

MLflow is an open-source platform for the complete machine learning lifecycle, from experimentation to deployment. Being one of the MLflow contributors, Azure Machine Learning made its workspaces MLflow-compatible, which means organizations can use Azure Machine Learning workspaces in the same way that they use an MLflow tracking server. MLflow has recently released its new version, MLflow 2.0, which incorporates a refresh of the core platform APIs based on extensive feedback from MLflow users and customers, which simplifies the platform experience for data science and machine learning operations workflows. We’re excited to announce that MLflow 2.0 is also supported in Azure Machine Learning workspaces.

Read this blog to learn more about what you can do with MLflow 2.0 in Azure Machine Learning.

Azure AI is empowering developers and organizations to build cutting-edge AI solutions with its rich set of open-source technologies. From leveraging pre-trained models to customizing AI capabilities with new technologies like Hugging Face foundation models, to integrating responsible AI practices with new open-source tools, Azure AI is driving innovation and efficiency in the AI industry. With Azure AI infrastructure, organizations can accelerate their large-scale AI workloads and achieve even greater results. Read this blog and the on-demand session to take a deep dive into what open-source projects and features we’ve announced at Azure Open Source Day 2023.

We’d like to conclude this blog post with some outstanding customer examples that demonstrate their success strategy of combining open-source technologies and building their own AI solutions to transform businesses.

What is most important about these announcements is the creative and transformative ways our customers are leveraging open-source technologies to build their own AI solutions.

These are just a few examples from our customers.

Customers innovating with open-source on Azure AI

Elekta is a company that provides technology, software, and services for cancer treatment providers and researchers. Elekta considers AI as essential to expanding the use and availability of radiotherapy treatments. AI technology helps accelerate the overall treatment planning process and monitors patient movement in real-time during treatment. Elekta uses Azure cloud infrastructure for the storage and compute resources needed for their AI-enabled solutions. Elekta relies heavily on Azure Machine Learning, Azure Virtual Machines, and the PyTorch open-source machine learning framework to create virtual machines and optimize their neural networks.
Read full story

The National Basketball Association (NBA) is using AI and open-source technologies to enhance its fan experience. The NBA and Microsoft have partnered to create a direct-to-consumer platform that offers more personalized and engaging content to fans. The NBA uses AI-driven data analysis system, NBA CourtOptix, which uses player tracking and spatial position information to derive insights into the games. The system is powered by Microsoft Azure, including Azure Data Lake Storage, Azure Machine Learning, MLflow, and Delta Lake, among others. The goal is to turn the vast amounts of data into actionable insights that fans can understand and engage with. The NBA also hopes to strengthen its direct relationship with fans and increase engagement through increased personalization of content delivery and marketing efforts.
Read full story

AXA, a leading car insurance company in the United Kingdom needed to streamline the management of its online quotes to keep up with the fast-paced digital marketplace. With 30 million car insurance quotes processed daily, the company sought to find a solution to speed up deployment of new pricing models. In 2020, the AXA data science team discovered managed endpoints in Azure Machine Learning and adopted the technology during private preview. The team tested ONNX open-source models deployed through managed endpoints and achieved a great reduction in response time. The company intends to use Azure Machine Learning to deliver value, relevance, and personalization to customers and establish a more efficient and agile process.
Read full story

Quelle: Azure