I will train a custom lora model for your character or art style

Turn your character, art style, or concept into a custom-trained LoRA model, ready to generate consistent images with Stable Diffusion (SDXL/Illustrious).

What I provide:

• Dataset curation and captioning (JoyCaption/WD Tagger)
• Hyperparameter optimization (rank, alpha, training steps, warmup)
• Model training and validation with test images
• Delivery of the trained model in a format compatible with ComfyUI and AUTOMATIC1111

What I need from you:

• A reference dataset (images of the character or art style) the more consistent the dataset, the better the results
• A clear description of the desired character or art style
• If you don't have a dataset, I can help create one, though this may involve an additional fee

Limitations & Notes:

• ️I do not work with NSFW content, sexualized depictions of minors, or content that infringes on third-party copyrights.
• ️Revisions cover training adjustments only and do not include unlimited generation of additional images.

️Final results depend on the quality and consistency of the provided dataset.

Before placing an order, please send me a message with the details of your project so I can confirm its feasibility and provide an estimated delivery time.