# Running Stable Diffusion 3 DreamBooth LoRA training under 16GB
This is an **EDUCATIONAL** project that provides utilities for DreamBooth LoRA training for [Stable Diffusion 3 (SD3)](ttps://huggingface.co/papers/2403.03206) under 16GB GPU VRAM. This means you can successfully try out this project using a [free-tier Colab Notebook](./sd3_dreambooth_lora_16gb.ipynb) instance. 🤗
This is an **EDUCATIONAL** project that provides utilities for DreamBooth LoRA training for [Stable Diffusion 3 (SD3)](ttps://huggingface.co/papers/2403.03206) under 16GB GPU VRAM. This means you can successfully try out this project using a [free-tier Colab Notebook](https://colab.research.google.com/github/huggingface/diffusers/blob/main/examples/research_projects/sd3_lora_colab/sd3_dreambooth_lora_16gb.ipynb) instance. 🤗
> [!NOTE]
> SD3 is gated, so you need to make sure you agree to [share your contact info](https://huggingface.co/stabilityai/stable-diffusion-3-medium-diffusers) to access the model before using it with Diffusers. Once you have access, you need to log in so your system knows you’re authorized. Use the command below to log in:
...
...
@@ -35,4 +35,4 @@ This project is educational. It exists to showcase the possibility of fine-tunin
* Techniques such as prior-preservation is unsupported.
* Custom instance captions for instance images are unsupported, but this should be relatively easy to integrate.
Hopefully, this project gives you a template to extend it further to suit your needs.
\ No newline at end of file
Hopefully, this project gives you a template to extend it further to suit your needs.
