Remade-AI/Car-chase

Car chase LoRA for Wan2.1 14B I2V 480p

Overview

Follows high-speed vehicle pursuit with tight, dynamic camera movement. Ideal for action sequences, thrillers, or adding intensity and motion to automotive scenes. This LoRA is trained on the Wan2.1 14B I2V 480p model.

Features

• Trained on the Wan2.1 14B 480p I2V base model
• Consistent results across different object types
• Simple prompt structure that's easy to adapt

Community

• Generate videos with 100+ Camera Control and VFX LoRAs on the Remade Canvas.
• Discord: Join our community to generate videos with this LoRA for free

Prompt

The video depicts a dramatic nighttime c4r ch4s3 car chase unfolding on a winding mountain road. A dark sports car, with its headlights blazing, speeds along the road, followed closely by a blue motorcycle. A helicopter hovers above the car and motorcycle, shining a spotlight. The camera begins positioned low and slightly behind the car's rear passenger side, capturing the pursuit by the motorcycle and helicopter. The camera pans slightly, tracking the car as it drifts into a turn, revealing the helicopter still in pursuit, emphasizing the ongoing c4r ch4s3 car chase.

Prompt

The video depicts a high-speed c4r ch4s3 car chase unfolding on a winding, foggy mountain road at night. A dark sports car, illuminated by its headlights, speeds around a sharp curve. Behind it, a helicopter follows closely, its searchlight beam aimed at the car. The camera maintains a low, forward-moving angle, keeping the car in the foreground as the helicopter descends and the c4r ch4s3 car chase continues down the road.

Prompt

The video depicts a coastal road during a storm, where a large black armored SUV and a black police car drive closely together. Lightning flashes overhead during the c4r ch4s3 car chase. The camera moves forward, initially tracking alongside the SUV, then transitioning to a view behind the police car, showcasing its pursuit of the SUV as they navigate the winding road. The wet asphalt reflects the vehicle's headlights.

Model File and Inference Workflow

📥 Download Links:

• car_chase.safetensors - LoRA Model File
• wan_img2vid_lora_workflow.json - Wan I2V with LoRA Workflow for ComfyUI

---

Recommended Settings

• LoRA Strength: 1.0
• Embedded Guidance Scale: 6.0
• Flow Shift: 5.0

Trigger Words

The key trigger phrase is: c4r ch4s3 car chase

Prompt Template

For prompting, check out the example prompts; this way of prompting seems to work very well.

ComfyUI Workflow

This LoRA works with a modified version of Kijai's Wan Video Wrapper workflow. The main modification is adding a Wan LoRA node connected to the base model.

See the Downloads section above for the modified workflow.

Model Information

The model weights are available in Safetensors format. See the Downloads section above.

Training Details

• Base Model: Wan2.1 14B I2V 480p
• Training Data: Trained on 50 seconds of video comprised of 10 short clips (each clip captioned separately) car chase scenes.
• Epochs: 25

Additional Information

Training was done using Diffusion Pipe for Training

Acknowledgments

Special thanks to Kijai for the ComfyUI Wan Video Wrapper and tdrussell for the training scripts!