Train a Lora for consistent character In Fooocus - kohya ss
🔥Learn how to train a stable diffusion LoRA model for a Fooocus user interface, achieving a consistent character with a similar face across different poses.
👉In this tutorial, I share my experience and tips on creating a LoRA model that can generate high-quality images with accurate facial features and varied poses. Whether you're an artist, designer, or AI enthusiast, this video is perfect for anyone looking to explore the possibilities of stable diffusion and LoRA models.
#fooocus #StableDiffusion #DigitalArt #ImageManipulation #CreativeTools
00:00 How to install kohya ss
04:00 Download images to train a lora
05:22 Kohya training folders
06:45 Create image caption using blip caption
09:16 Kohya settings for lora
11:00 Regularisation images for lora
12:04 Train batch size in kohya
12:32 Epoch in lora
14:09 Multiple lora saving.
16:46 Understand lora training status
18:15 Checking new lora in fooocus
20:22 Final Thoughts on training a lora on local computer
Regularization images and config file:
https://drive.google.com/drive/folder...
Watch Previous Video, comparing all face swap methods in Fooocus • Fooocus FaceSwap comparison, which is the ...
Playlist For Fooocus AI - Stable Diffusion
• Fooocus User Guide - Stable Diffusion AI i...
Playlist For ComfyUI - Stable Diffusion
• How To Use Comfy UI
Image Generation Tricks for Automatic1111
• Automatic1111 - Stable Diffusion