@Roronoa zoro D You can first create your base image. Then use that to generate training dataset. After which train a lora. Then plug it to comfyui

@Rick Henderson hey . Since I have 32gb vram and 64gb ram with 4tb ssd. I am hosting everything locally. The prompts are given by me and the model used behind the scene is z-image.

Hi @Herman Carter check out these samples. Let me know what you think https://drive.google.com/drive/folders/1G5wPrUbx89ui2n5-k4MwaigENkzphsvD?usp=sharing

Why don’t you automate it using AI?

@Frankie Fuentes AI can do amazing stuff. If you are relying on chatGPT for everything, you are missing out a lot. Try switching to open source models, run it locally or via runpod. May be create a serverless architecture and then see what amazing things it can do. It can understand and remember context and conversation of each user and chat with them accordingly. Pretty amazing. And most importantly, it can s*x text as well.