
Meet Vishal Prem, a software engineer with expertise in designing and developing applications, big data analytics, and machine learning, who champions accessible artificial intelligence for solving real problems.
Explore models and image generation with stable diffusion, master prompt engineering, build your own model in Colab and Drive, and connect with the digital art community.
Digital artists learn the basics of models and how AI art can elevate their work, even without coding experience. Discover how the technology behind AI art creates unique, innovative pieces.
Develop your creative expression by making posters, concept art, and gifts. Build an online art portfolio and enhance your Instagram photos while applying design principles through hands-on projects.
Prepare essential tools: a competent computer with internet, 5–10 GB on Google Drive, a free Hugging Face account created during the course, and a willingness to learn about artificial intelligence.
Learn the basics of artificial intelligence, including the model as the brain, its structure and weights, and how training with labeled data enables image-to-text and text-to-image generation using stable diffusion.
Master Google Colab and Hugging Face to access stable diffusion, connect Google Drive, upload files, and download model weights using a token for online projects.
Connect to stable diffusion in a Google Colab notebook using a Hugging Face token and a descriptive prompt to generate your first digital art, including downloading weights.
Configure stable diffusion settings by crafting detailed prompts and negative prompts, choosing image dimensions, adjusting guidance scale, and setting inference steps for higher-quality AI art.
Master prompt engineering to craft stylized and realistic AI art using prompts, negative prompts, and emphasis brackets, with Studio Ghibli references and varied inference steps.
Learn to build a realistic marble statue artwork from description alone, adjusting subject, details, medium, lighting, and add-ons, using controlled guidance scale and inference steps.
Load different stable diffusion model versions by selecting model IDs and weights in notebooks. Compare styles like Disney, Midjourney, and Spider-Verse by using theme-specific models and prompts.
Build a 512 by 512 pixel face dataset by resizing 30–70 photos so the face is clear, then prepare a stable diffusion–compatible dataset to train a custom model.
Train a base stable diffusion model in google colab and save the weights to a visual test folder in google drive, using instance and class prompts with 34 training images.
Learn to load your trained Stable Diffusion model in Google Colab, specify weight folders and iterations, set a seed, and generate images from prompts with negative prompts using inference steps.
Apply your basic digital art skills and join communities like Prompt Hero and Hugging Face, then explore Reddit stable diffusion and Automatic1111 for advanced generation.
During this course, we will explore the fascinating world of AI models in depth, with a particular focus on image generation. We will start by building a strong foundation of knowledge around the basics of AI models, including their history, development, and applications. From there, we will dive deeper into the specifics of AI image generation, learning how to use different publicly available models to create stunning pieces of art.
But that's not all! We will also delve into the exciting process of building your own model using Stable Diffusion. Whether you want to create a model that mimics your unique art style or generates images based on your personal photos, we will guide you through the process of bringing your vision to life.
You will have the opportunity to work on several projects that will allow you to put your new skills into practice. These projects will range from basic image generation to more complex tasks, such as creating a model that generates images based on a particular theme or style.
A major aspect of this course is that everything used in the course will be on your personal accounts and personal workspaces. You will build the workflows in google colab, store the trained models in your google drive and be able to use them from there. This give you much more freedom and control than trying to use something based on subscriptions or payments. This is critical if you plan to use your personal photos in order to generate the art.
Finally, we will connect you with the larger digital art community, allowing you to share your work, get feedback from others, and discover new artists and techniques. By the end of this course, you will have developed a deep understanding of AI models and their applications, as well as a newfound appreciation for the beauty and power of digital art.
All Music Used In This Course Is Music by ComaStudio from Pixabay