Custom Styles
Create, train, and manage your own AI styles. This collection covers everything from building a high-quality dataset and choosing the right category to training, publishing, unpublishing, and optimizing your custom styles for the best results.
5 articles
How to build a good style dataset?Learn how to build a high-quality style dataset, including image quality requirements, maintaining consistency, diversifying subjects, and using existing models as a starting point.
Quick Start with Styles TrainingLearn what style training is, how to prepare your dataset, how long training takes, and how to create your first custom style in starryai.
What Does It Mean to Train a Style?Learn what style training means, how the AI learns visual patterns, and how custom styles can help with branding, storytelling, and creative control.
When Creating Styles, How Do Categories Make a Difference?Learn how style categories affect training quality, what each category is optimized for, and how to choose the best option for your dataset.
How Long Does Style Training Take?Learn how long style training usually takes, what to expect during busy periods, and what happens if training doesn’t complete successfully.