## Model Card: Miami Beach Hologram
Please refer to
Version 2 - Update
While v.1 was an idea, v.2 was trained with real objects and text:
It may help with small words in case SDXL is struggling (as seen on samples.)
It may hlep rendering old school objects, not that SDXL cannot on its own, but it quickly associates with the LoRA dataset and tends to render better props and scenes.
Trigger word will turn everything neon, it's not necessary for more neutral themes, while it still improves a lot of aspects in the image.
For newer fine-tuned models it's better setting the weight around 0.6 ~ 0.7. For SDXL about 1.0 (when there is text,) or 0.8
Dataset is available on HuggingFace at https://huggingface.co/datasets/robb-0/miami_beach_hologram2_dataset
Have fun!
### Overview
The Miami Beach Hologram model is a custom Image LoRA (Low-Rank Adaptation) designed to generate vibrant, neon-lit, and geometrically precise artworks inspired by the iconic aesthetics of Miami Beach. This model specializes in creating futuristic, holographic-style images with a strong emphasis on neon lighting, reflective surfaces, and retro-futuristic elements. Whether you’re an artist, designer, or creator, this model offers a powerful tool for generating stunning, synthwave-inspired visuals.
The main reason is to create this style of images for SDXL without much prompting. I had a few LoRAs which - despite not being their focuses - were able to create amazing Miami Beach Hologram arcade style. I was bored here and thought, why not?
I'm sharing the dataset which was generated via Flux, and it's followed by captions tagged here on Civitai.
---
### Key Features
- Neon Holographic Style: Produces images with glowing neon lights, reflective surfaces, and a dreamy, futuristic atmosphere.
- Geometric Precision: Emphasizes clean lines, sharp shapes, and low-poly aesthetics, perfect for creating sci-fi and retro-futuristic scenes.
- Versatile Applications: Ideal for AI art, video game assets, album covers, digital installations, and conceptual designs.
- Inspired by Miami Beach: Captures the vibrant, neon-lit essence of Miami Beach, blending it with a holographic, cyberpunk-inspired aesthetic.
---
### Training Details
#### Dataset
- Dataset Size: 15 images
- Source: Custom-curated dataset shared on Civitai, featuring high-quality images with text file captions.
- Content: The dataset includes examples of neon-lit scenes, geometric shapes, reflective surfaces, and iconic landmarks rendered in a holographic style.
#### Training Parameters
- Model Base: SDXL 1.0
- VAE: SDXL-VAE-16-Fix
- Scheduler: Euler
- Steps: 30
- Clip Skip: 1
- Total Training Steps: 380
- Epochs: 10
- Batch Size: 4
- Optimizer: AdamW
- Learning Rate Schedule: Cosine with Restarts
#### Training Environment
- Stable Diffusion Version: SDXL 1.0
- Training Framework: Custom LoRA training pipeline
- Device: GPU (Recommended for efficient inference)
### License
The Miami Beach Hologram model is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license.
- License URL: [https://creativecommons.org/licenses/by/4.0/](https://creativecommons.org/licenses/by/4.0/)
- Attribution: When using this model, please credit the author and include the following information:
```
Miami Beach Hologram © 2025 by Robb-0
Licensed under CC BY 4.0
```
---
### Acknowledgments
- Dataset Contributors: Thank you to the creators of the images used in the training dataset. All images were generating using Flux image model.
- Stable Diffusion Team: Gratitude to the developers of SDXL for providing the foundational model.
- Community: Special thanks to the AI art community for their support and inspiration.
---
### Disclaimer
This model is provided as-is. While efforts have been made to ensure its quality, the author is not responsible for any issues arising from its use. Use at your own discretion.
---
🌟✨
Description
Different dataset with improved text.