AI Model Fine-tuning Tool
7.7k 2026-04-18
XavierXiao/Dreambooth-Stable-Diffusion
An implementation of Google's Dreambooth technique on Stable Diffusion, enabling personalized text-to-image model fine-tuning with limited examples.
Core Features
Dreambooth implementation for Stable Diffusion
Fine-tune text-to-image models with a few examples
Utilizes regularization images for improved training stability
Optimized for GPU memory with gradient checkpointing
Quick Start
python main.py --base configs/stable-diffusion/v1-finetune_unfrozen.yaml -t --actual_resume /path/to/original/stable-diffusion/sd-v1-4-full-ema.ckpt -n <job name> --gpus 0, --data_root /root/to/training/images --reg_data_root /root/to/regularization/images --class_word <xxx>Detailed Introduction
This project provides an open-source implementation of Google's Dreambooth technique, adapted for the widely accessible Stable Diffusion text-to-image model. It addresses the challenge of personalizing generative AI models by allowing users to fine-tune a diffusion model with just a handful of images of a specific subject. This enables the creation of highly customized models capable of generating novel images of that subject in various contexts, making advanced AI image generation more accessible and controllable for researchers and developers.