Semantic Kernel (AI LLM integration) now supports VS Code tools and Python — Visual Studio Magazine

news

Semantic kernel (AI LLM integration) now supports VS Code tools and Python

Microsoft is busy updating its recently open-sourced internal incubation project called Semantic Kernel, an SDK that allows developers to mix traditional programming languages ​​with modern Large Language Model (LLM) AI prompts. I’m here.

It helps developers incorporate advanced AI technologies into their apps, including highly popular generative AI systems such as ChatGPT and GPT-4 from Microsoft partner OpenAI. According to Microsoft, the SDK provides templating, chaining, and planning capabilities to create AI-first apps faster and easier.

Better prompts, associated with a new discipline of “prompt engineering” that can pay up to $335,000 a year, is one of the four main benefits Microsoft cites for open source projects.

• Rapid integration: SK is designed to be embedded in any kind of application, making it easy to test and run LLM AI.
• Scalability: SK allows you to connect to external data sources and services. SK’s apps can use natural language processing in combination with live information.
• Improved prompts: SK’s templated prompts enable you to quickly design semantic functions with useful abstractions and mechanisms to unlock the potential of LLM AI.
• Novel but familiar: Native code is always available as a first-class partner for rapid engineering quests. You get the best of both worlds.

After open sourcing the Semantic Kernel last month, the company last week announced the release of Semantic Kernel Tools, an extension for Visual Studio Code, and this week announced the availability of a Python flavor.

semantic kernel tools
This VS Code extension, available now in the code editor marketplace, is designed to help developers write their semantic kernel skills. Microsoft’s “What Are Skills?” documentation describes a skill as an area of ​​expertise that is provided to the kernel either as a single function or as a group of skills related functions. Functions can be LLM AI prompts, native code, or a combination of both. As an example, a summary skill can use LLM AI prompts to generate text or audio summaries.

The Semantic Kernel Tools Marketplace listing explains how to create semantic skills and run semantic functions, as well as troubleshooting and other guidance. As an all-new tool, it has only been downloaded 424 times as of this writing, debuting on April 11th and being updated on Monday.

According to Microsoft, the streamlined development experience provided by the tool will help coders create and test new skills more quickly and efficiently, allowing them to be more creative with their projects rather than bogged down in technical details. You will be able to focus on the important aspects.

“The ability to create custom skills allows developers to tailor projects to their specific needs and goals,” the company said. “Whether it is for chatting and conversations, generating or transforming code, answering questions, or any other use, the Semantic Kernel Tools give you the flexibility you need to create innovative and impactful projects. Provides strength and power.

“We believe the Semantic Kernel Tools represent an important step forward in the development of AI technology. We want to be able to create applications and services.”

Semantic Kernel Now Available in Python Flavors
When the Semantic Kernel debuted, Python support was in preview, but it only worked for C# projects. As of this week, her Python support has moved from an experimental preview to an official release available on GitHub.

“Python is widely used in AI and ML, making it a great choice for AI exploration and integration,” Microsoft said. “As a result, SK gets many AI algorithms, functions and libraries such as TensorFlow, PyTorch, Scikit-learn, Pandas, Keras, NLTK and Caffe. You can easily benefit from your models.You can easily take advantage of GPU and cloud computing.Finally, there are many great visualization options pioneered by the data science community, allowing users to create charts, You can create histograms, plots.”

The same announcement post also includes a short Q&A with Devis Lucato, Principal Software Architect at Microsoft, who helped create the Semantic Kernel. When asked why Python is a great language for his AI-first applications, Lucato said:

In my humble opinion, Python is the Mona Lisa of programming languages. Python is an elegant blend of beauty and power, and it’s no wonder Italians like me love it dearly. Traditional coding requires learning the syntax of a computer language that feels like it’s being spoken by a machine. Python, on the other hand, has always read more like normal written language. Python with SK goes further by extending regular coding with natural language in a seamless and powerful way. As a result, AI developers can move faster, write less code, and enjoy SK in the Mona Lisa programming language.

Microsoft has said it has added official Python support in response to community feedback, and has previously noted that such feedback may prompt support for other programming languages, including its own TypeScript. was showing.