Create custom AI models and LoRas by fine-tuning Stable Diffusion
Stable Diffusion is just the beginning. Learn how to train your own models to replicate any style or subject, including artists and photorealism. We'll uncover all the techniques we used to create the popular AI model Openjourney.
Create your own styles. Learn how to train your own models, like Openjourney.
Here at PromptHero we created Openjourney, an open-source custom model that generates images in the style of Midjourney. People love it, and it has been downloaded millions of times already since it launched. Want to know how we made it? We'll tell all our secrets in this course.
Learn advanced techniques like Textual Inversion and LoRa.
Create characters and styles you can just drop-in in your images. We'll guide you through another custom model we made: Fashion Diffusion, a fine-tuned model we created trained on fashion imagery.
Learn Dreambooth
Dreambooth is a technique used to train Stable Diffusion to learn to draw new things (like your face, your dog or even your backpack!) You feed it a few images of what you want it to learn, and it learns to generate images of that particular thing. To use it well and achieve good results, you'll need specific know-how – which we'll be covering in this course.
Learn to create your own AI avatars
Just so you get an idea of what you can end up doing, Dreambooth is the technology behind AI avatar craze. If you've seen people using any kind of app to generate an AI avatar of themselves, Dreambooth was the underlying technique across all apps. And yes, rest assured you'll be generating your own AI avatars in this course, too! (for free!)
Learn the skills of the future, today.
Get immediately differentiated
Stop using Stable Diffusion – everyone else is using it. Create your own secret sauce and build a model only you have access to.
Repurpose Stable Diffusion
Move from generalistic images to industry-specific images. Generate the perfect images for architecture, design, arts, videogames...
Future-proof and 10x what you can do with AI
Break free from the chains imposed by Stable Diffusion. Make AI do anything you can do. The only limit is your creativity and imagination.
What will you learn?
Fine-tuning 101: what does it actually mean to train an AI?
Garbage in, garbage out: ground truths and first principles of model fine-tuning
Building the best dataset you could possibly create
Where to get training data?
Data cleaning, data processing & data engineering
Advanced data labelling techniques & automation
Creating your AI fine-tuning setup in the cloud: getting hold of A100s GPUs
The Dreambooth revolution
Dreambooth 101: creating your own AI avatars for free
Dreambooth for professional use
Fine-tuning beyond Dreambooth: from Pokemons to dresses
Textual Inversion and embeddings: a lightweight way to steer the model
Advanced fine-tuning techniques, scripts and state of the art
Recommended training parameters and settings
Practical follow along: How we created Openjourney
Who is this for?
Artists
If you come from an artistic background, you'll find this course highly interesting. Painters, game makers, digital artists... master AI generation and take your art to the next level.
Creatives
If you're a designer (fashion, graphic...), a copywriter or do any other sort of creative job, you'll find in AI fine tuning the perfect skill to get your work differentiated from the competition.
Professionals
Architects, interior designers and photographers are already incorporating AI generated images into their workflow to provide clients with breathtaking results that no one else can give.
Ready to start learning?
Join thousands of students mastering AI skills