README.md

# Gavin Newsom LoRA for Wan2.2 (T2V-A14B)

This is a custom-trained **LoRA (Low-Rank Adapter)** for **Wan2.2 T2V-A14B**, fine-tuned on 24 high-resolution, face-centered, curated images of Gavin Newsom. Designed for **generative video models**, it supports cinematic, political, and meme-style image outputs. Trained by [@huwhitememes](https://huggingface.co/huwhitememes) using the [fal-ai/wan-22-image-trainer](https://fal.ai/models/fal-ai/wan-22-image-trainer).

## 🔧 Use Cases

- Political memes and satire  
- Cinematic shitpost visuals  
- Social media propaganda visuals  

## 🧪 Training Details

- **Base Model**: fal-ai/wan2.2  
- **Steps**: 1000  
- **LoRA Rank**: Auto  
- **Learning Rate**: 0.0007  
- **GPU**: Nvidia H100 (Fal Managed Infra)  
- **Image Count**: 24 (upscaled, face-centered, curated)  
- **Trigger Phrase**: `G4vin N3wsom` (recommended in prompt)  

## ⚠️ Legal & Fair Use

This model was trained on public photos of a well-known public figure. It is intended for creative expression, satire, parody, education, and research. Please use responsibly and in accordance with your platform’s content guidelines and applicable laws.

## 🧪 Example Usage Prompt
```text
G4vin N3wsom walking confidently through a dystopian city at sunset, wind blowing through his perfect hair, dramatic lens flare, cinematic orange tones, ultra-realistic