Super-Saiyan
| Property | Value |
|---|---|
| Base Model | Wan2.1 14B I2V 480p |
| Training Epochs | 35 |
| Model Format | LoRA Safetensors |
| Author | Remade-AI |
| Repository | Hugging Face |
What is Super-Saiyan?
Super-Saiyan is a specialized LoRA model trained on the Wan2.1 14B I2V 480p base model that enables the creation of Dragon Ball-style Super Saiyan transformation videos. This innovative model can transform any subject - from people to animals - into a Super Saiyan, complete with the iconic glowing yellow hair and energy effects.
Implementation Details
The model was trained on 1 minute of video data, comprising 13 distinct Super Saiyan transformation clips. Each clip was individually captioned to ensure accurate learning of the transformation sequence. The implementation uses a modified version of Kijai's Wan Video Wrapper workflow in ComfyUI, with specific optimizations for LoRA integration.
- Training Duration: 35 epochs
- Recommended LoRA Strength: 1.0
- Embedded Guidance Scale: 6.0
- Flow Shift: 5.0
Core Capabilities
- Transforms subjects into Super Saiyans with glowing yellow hair
- Generates realistic energy effects and background transformations
- Works with diverse subject types (humans, animals, objects)
- Produces consistent results across different prompts
- Supports 480p video output resolution
Frequently Asked Questions
Q: What makes this model unique?
This model specializes in creating realistic Dragon Ball-style transformations, utilizing a specific trigger phrase "5up3r super saiyan transformation" to ensure consistent results. It's particularly notable for its ability to work with various subjects while maintaining the authentic Super Saiyan aesthetic.
Q: What are the recommended use cases?
The model is ideal for creating transformation videos of people or objects into Super Saiyans. It's particularly effective when following the provided prompt template, which includes detailed descriptions of the subject and transformation sequence. The model works best with ComfyUI using the provided workflow configuration.
