Train a LoRA on your likeness with Fluxgym

To train a LoRA on your likeness for character consistency: 1. Collect 15-20 images of yourself (preferably in different locations, light settings, clothing, and facial expressions) as a dataset. 2. Install Fluxgym using Pinokio. 3. Train the LoRA locally with Fluxgym, inserting a trigger keyword to reference your likeness. 4. Once trained, generate images using prompts like "a picture of travy1 eating a hamburger" to get your likeness in any setting. Alternative: Use FLUX PULID online for free - upload a single face image and generate new images with that likeness (essentially faceswapping).