Within the dynamic world of generative AI, innovation is the driving drive propelling us into novel and uncharted territories. New fashions, new instruments and platforms, and new use circumstances emerge day-after-day, making a outstanding fusion of creativity and expertise and redefining the boundaries of what’s doable. With Azure AI, our targets are to offer probably the most cutting-edge open and frontier fashions within the business, to make sure builders have mannequin selection, to proceed to uphold the very best requirements in Accountable AI, and to proceed to construct superior tooling that brings all this collectively to speed up the innovation in copilots that we’re seeing.
At Microsoft Ignite, we remodeled 25 bulletins throughout the Azure AI stack, together with the addition of 40 new fashions to the Azure AI mannequin catalog; new multimodal capabilities in Azure OpenAI Service; the Fashions as a Service (MaaS) platform in Azure AI Studio and partnerships with Mistral AI, G24, Cohere, and Meta to supply their fashions in MaaS; and the general public preview of Azure AI Studio.
Since Ignite, we’ve continued so as to add to our Azure AI portfolio. In the present day, we’re excited to announce much more Azure AI capabilities: the provision of Meta’s Llama 2 operating in Fashions as a Service, the preview of GPT-4 Turbo with Imaginative and prescient to speed up generative AI and multimodal software improvement, and the addition of much more fashions within the Azure AI mannequin catalog together with our Phi 2 Small Language Mannequin (SLM), amongst different issues.
Now Accessible: Fashions as a Service for Llama 2
In Azure AI, you’ve gotten been capable of deploy fashions onto your individual infrastructure for a very long time – merely go into the mannequin catalog, choose the mannequin to deploy and a VM to deploy it on and also you’re off to the races. However not each buyer desires to consider working infrastructure, which is why at Ignite we launched Fashions as a Service, which operates fashions as API endpoints that you just name, a lot the way in which you may name the Azure OpenAI Service.
In the present day, we’re making Meta’s Llama 2 accessible in Fashions as a Service by means of Azure AI in public preview, enabling Llama-2-7b (Textual content Era), Llama-2-7b-Chat (Chat Completion), Llama-2-13b (Textual content Era), Llama-2-13b-Chat (Chat Completion), Llama-2-70b (Textual content Era), and Llama-2-70b-Chat (Chat Completion).
Watch this video to be taught extra about Fashions as Service:
As we convey extra fashions on-line in Fashions as a Service, we’ll maintain you up to date.
Now Accessible: GPT-4 Turbo with Imaginative and prescient
We’re delighted to announce that GPT-4 Turbo with Imaginative and prescient is now in public preview in Azure OpenAI Service and in Azure AI Studio. GPT-4 Turbo with Imaginative and prescient is a big multimodal mannequin (LMM) developed by OpenAI that may analyze pictures and supply textual responses to questions on them. It incorporates each pure language processing and visible understanding. This integration permits Azure customers to profit from Azure’s dependable cloud infrastructure and OpenAI’s superior AI analysis.
GPT-4 Turbo with Imaginative and prescient in Azure AI presents cutting-edge AI capabilities together with enterprise-grade safety and accountable AI governance. When mixed with different Azure AI companies, it could possibly additionally add options like video prompting, object grounding, and enhanced optical character recognition (OCR). Clients like WPP and Instacart are utilizing GPT-4 Turbo with Imaginative and prescient and Azure AI Imaginative and prescient right now, take a look at this weblog to listen to extra of their tales.
Accessible Tomorrow: High quality Tuning for GPT 3.5 Turbo and Different Fashions
In October 2023, we introduced public preview of fine-tuning capabilities for OpenAI fashions. Beginning tomorrow, December 15, 2023, fine-tuning can be typically accessible for fashions together with Babbage-002, Davinci-002, GPT-35-Turbo. Builders and information scientists can now customise these Azure OpenAI Service fashions for particular duties. We proceed to push innovation boundaries with these new capabilities and are excited to see what builders construct subsequent with generative AI.
Growth to the Azure AI Mannequin Catalog
Whereas Azure operates our personal fashions as a part of the Azure AI companies like our Speech, Imaginative and prescient, and Language fashions, in addition to Azure OpenAI, we additionally notice that prospects usually want fashions that we don’t function. More and more, we’re seeing prospects look to deploy fashions which were fine-tuned to particular duties. To this finish, we’ve operated a full mannequin catalog in Azure AI Studio for a very long time, and it’s well-stocked with a broad number of fashions. In the present day, we’re saying the addition of six new fashions. Phi-2 and Orca 2 can be found now and different fashions beneath are coming quickly.
Phi-2. is a small language mannequin (SLM) from Microsoft with 2.7 billion parameters. Phi-2 exhibits the facility of SLMs, and reveals dramatic enhancements in reasoning capabilities and security measures in comparison with Phi-1-5, whereas sustaining its comparatively small measurement in comparison with different transformers within the business. With the best fine-tuning and customization, these SLMs are extremely highly effective instruments for purposes each on the cloud and on the sting. Be taught extra.
DeciLM. Introducing DeciLM-7B, a decoder-only textual content technology mannequin with a powerful 7.04 billion parameters, licensed below Apache 2.0. Not solely is DeciLM-7B probably the most correct 7B base mannequin up to now, however it additionally surpasses a number of fashions in its class.
DeciDiffussion. DeciDiffusion 1.0 is a diffusion-based text-to-image technology mannequin. Whereas it maintains foundational structure components from Secure Diffusion, such because the Variational Autoencoder (VAE) and CLIP’s pre-trained Textual content Encoder, DeciDiffusion introduces important enhancements. The first innovation is the substitution of U-Internet with the extra environment friendly U-Internet-NAS, a design pioneered by Deci. This novel part streamlines the mannequin by lowering the variety of parameters, resulting in superior computational effectivity.
DeciCoder. 1B is a 1 billion parameter decoder-only code completion mannequin educated on the Python, Java, and JavaScript subsets of Starcoder Coaching Dataset. The mannequin makes use of Grouped Question Consideration and has a context window of 2048 tokens. It was educated utilizing a Fill-in-the-Center coaching goal. The mannequin’s structure was generated by Deci’s proprietary Neural Structure Search-based expertise, AutoNAC.
Orca 2. Like Phi-2, Orca 2 from Microsoft explores the capabilities of smaller LMs (on the order of 10 billion parameters or much less). With Orca 2, exhibits that improved coaching indicators and strategies can empower smaller language fashions to attain enhanced reasoning talents, that are usually discovered solely in a lot bigger language fashions. Orca 2 considerably surpasses fashions of comparable measurement (together with the unique Orca mannequin) and attains efficiency ranges much like or higher than fashions 5-10 instances bigger, as assessed on advanced duties that take a look at superior reasoning talents in zero-shot settings. Be taught extra.
Mixtral 8x7b. Mixtral has an analogous structure as Mistral 7B however is comprised of 8 professional fashions in a single from a way referred to as Combination of Specialists (MoE). Mixtral decodes on the velocity of a 12B parameter-dense mannequin despite the fact that it incorporates 4x the variety of efficient parameters.
For extra data on different fashions launched at Ignite in our mannequin catalog, go to right here.
Azure AI Offers Highly effective Instruments for Mannequin Analysis and Benchmarking
It’s not sufficient to have plenty of fashions, prospects want to have the ability to select which mannequin meets their wants. To that finish, Azure AI Studio offers a mannequin benchmarking and analysis subsystem, which is a useful instrument for customers to evaluate and evaluate the efficiency of varied AI fashions. The platform offers high quality metrics for Azure OpenAI Service fashions and Llama 2 fashions akin to Llama-2-7b, gpt-4, gpt-4-32k, and gpt-35-turbo. The metrics revealed within the mannequin benchmarks assist simplify the mannequin choice course of and allow customers to make extra assured selections when choosing a mannequin for his or her job.
Beforehand, evaluating mannequin high quality may require important time and assets. With the prebuilt metrics in mannequin benchmarks, customers can rapidly establish probably the most appropriate mannequin for his or her mission, lowering improvement time and minimizing infrastructure prices. In Azure AI Studio, customers can entry benchmark comparisons inside the identical surroundings the place they construct, practice, and deploy their AI options. This enhances workflow effectivity and collaboration amongst group members.
Be taught extra about Mannequin benchmarks right here.
Empowering Clients Across the Globe
These groundbreaking developments not solely amplify our capability to generate various and imaginative content material but additionally sign a shift in how we conceptualize AI’s potential. In truth, main international legislation agency Dentons, is working with Azure AI to implement Azure OpenAI Service fashions together with GPT-4 and Meta’s Llama 2 into its generative AI software referred to as “fleetAI.” Dentons has over 750 attorneys and enterprise companies professionals and is using Azure AI fashions internally to summarize authorized contracts and extract key elements from paperwork leading to important time financial savings.
“By the incorporation of a lease report generator, into our fleetAI system, developed with Microsoft Azure’s Open AI service, we’ve got revolutionized a time-consuming job that beforehand took 4 hours, lowering it to simply 5 minutes,” mentioned Sam Chen, Authorized AI Adoption Supervisor for Dentons (UKIME). “This important time saving allows our authorized professionals to focus on extra strategic duties, thereby enhancing shopper service and underscoring our dedication to innovation.”
Our Dedication to Inclusive and Accountable AI Growth for All
Accountable AI is a key pillar of AI innovation at Microsoft. In October 2023, we introduced basic availability of Azure AI Content material Security and at Microsoft Ignite 2023, enabled new capabilities to deal with harms and safety dangers which might be launched by giant language fashions. The brand new options assist establish and forestall tried unauthorized modifications and establish when giant language fashions generate materials that leverages third-party mental property and content material. With these capabilities, builders now have instruments they will combine as a part of their generative AI purposes to observe content material, reduce hurt, and decrease safety dangers.
The IDC MarketScape just lately checked out AI governance platforms that guarantee AI/ML lifecycle governance, collaborative threat administration, and regulatory excellence for AI throughout 5 key ideas: equity, explainability, adversarial robustness, lineage, and transparency. We’re excited to share that Microsoft has been acknowledged as a pacesetter within the inaugural IDC MarketScape Worldwide AI Governance Platforms 2023 Vendor Evaluation. Learn our weblog to be taught extra about our placement and the way prospects are leveraging Azure AI to construct and scale generative AI options responsibly.
One Extra Factor: Darkish Mode in AI Studio
The consumer expertise in Azure AI Studio issues lots, and we’re making a extra accessible AI ecosystem collaborating with AI builders with disabilities. In the present day, we’re happy to announce, “darkish mode,” a beloved characteristic of builders in every single place. Azure AI Studio’s darkish mode shouldn’t be solely visually interesting – however it additionally performs an important function in enhancing accessibility, making Azure AI Studio extra inclusive and comfy to make use of for everybody. We hope you get some relaxation for these eyes and revel in this new characteristic as a lot as we do. To activate darkish mode, go to “Settings” within the app header to simply swap between gentle and darkish themes.
Let’s form the way forward for AI collectively
It has been an thrilling yr on the planet of AI. There’s a profound shift underway in the way in which we work together with purposes, seek for data, and get assist with routine duties. Copilots or assistants are reworking the way in which we be taught, work, and talk. We’re excited to be on the forefront of this AI evolution to empower builders and information scientists to construct with AI confidently for now and sooner or later.
Sources
