AI Toolkit Essentials for Visual Studio Code

2h 59mIntermediate2025-08-11

Authors

Kevin Ford

Development Practice Lead at Magenic

Course details

AI Toolkit for Visual Studio Code is an evolution of Microsoft's Windows AI Studio. AI Toolkit is geared toward experimentation, modification, and deployment of generative AI models in a variety of scenarios on a variety of platforms. It provides a comprehensive environment for working with generative AI that can reduce the frustration of setting up and using many disparate products.

In this course, learn about the powerful capabilities of AI Toolkit for VS Code to experiment with and modify different generative AI models. Instructor Kevin Ford shows you how to set up and use AI Toolkit on your machine, then explains how to employ AI Toolkit to configure and compare models, set parameters, and work through prompt engineering. He also shares practical hands-on exercises, such as how to fine-tune a model, and goes through the Microsoft guidelines for responsible AI development.

Skills covered

Visual Studio CodeMicrosoft DevelopmentProgramming FoundationsArtificial Intelligence FoundationsArtificial Intelligence (AI)MicrosoftSoftware DevelopmentOne-Off

Concepts

0. Introduction

01 - Streamline AI application development with AI Toolkit
02 - What you need to know

1. Getting to Know AI Toolkit for VS Code

03 - What is AI Toolkit for VS Code
04 - Microsoft guidelines for responsible AI app development
05 - Installing AI Toolkit for Visual Studio Code
06 - Setting up ONNX models

2. AI Catalogs

07 - Overview of the catalogs
08 - Authenticating with external providers
09 - Adding external models to AI Toolkit
10 - Converting models using ONNX
11 - Importing ONNX models into AI Toolkit with Model Creator

3. Using the Models

12 - Opening playgrounds
13 - Changing the model parameters
14 - Comparing models

4. Prompt Builder

15 - What is prompt engineering
16 - How many Rs are in strawberry
17 - Building a prompt template
18 - Running a prompt template

5. Bulk Run

19 - Overview of bulk run
20 - Using generated data with bulk run
21 - Using a custom JSONL file

6. Evaluation

22 - Overview of evaluation
23 - Using evaluation with the default dataset
24 - Using visualization of data
25 - Working with the custom dataset
26 - Adjusting the model with bulk run and evaluation
27 - Adding a custom evaluator - Initial setup
28 - Adding a custom evaluator - Finish and test

7. Fine-Tuning

29 - Introduction to fine-tuning
30 - Setting up a machine for local fine-tuning
31 - Starting local fine-tuning
32 - Running the fine-tuning
33 - Testing your fine-tuned model

Conclusion

34 - Where to go from here