Fine-Tuned Stable Diffusion Model with Dreambooth

Model Description

This model is a fine-tuned version of the Stable Diffusion model using Dreambooth techniques. It is trained to generate high-quality images based on specific prompts, with a focus on generating images of the user.

Training Data

Dataset: The training dataset consists of 6 images of the user, Zain.
Captions: All images were captioned with the user's name, "Zain".
Preprocessing: Standard preprocessing steps were applied to the images.

Training Procedure

Training Method: Dreambooth
UNet Training Steps: 1500
UNet Learning Rate: 2e-6
Resolution: 512

Results

Generated Images

Here are some example images generated by the model:

Intended Use

This model is designed to generate personalized images based on the user's input. Potential applications include generating avatars, artistic portraits, and other creative outputs featuring the user.

Limitations and Ethical Considerations

Limitations: The model is not generating the user's face accurately, as the training images were taken from a distance and did not focus on the face.
Ethical Considerations: Users should be aware of the ethical implications of using AI-generated images, particularly in terms of privacy and the potential for misuse.

How to Use

Instructions on how to use the model can be found in the Hugging Face documentation.

Acknowledgements

Thanks to TheLastBen's fast-DreamBooth notebook for providing the training framework.

Contact

For further questions or collaborations, please contact zainikhan3434@gmail.com.