Remade-AI/POV-Driving

POV Driving LoRA for Wan2.1 14B T2V

Overview

This LoRA is trained on the Wan2.1 14B T2V model and allows you to generate POV driving videos in any scene or landscape you desire!

Features

• Trained on the Wan2.1 14B T2V base model
• Consistent results across different object and scenes types
• Simple prompt examples that are easy to adapt

Community

• Discord: Join our community to generate videos with this LoRA for free
• Request LoRAs: We're training and open-sourcing Wan2.1 LoRAs for free - join our Discord to make requests!

Prompt

p0v_dr1v1n6, video shows a person driving a car through a burning hellscape. The driver is holding the steering wheel with both hands. Rivers of lava flow on both sides of the cracked road, and firestorms rage in the distance. The driver is looking straight ahead. The car has a digital dashboard and a touchscreen display flickering with demonic symbols.

Prompt

p0v_dr1v1n6 through a sandstorm in the desert, visibility dropping as golden dust engulfs the horizon, digital dashboard displaying emergency alerts, the car struggling against the powerful winds.

Prompt

dr1v12ng POV Driving. The video shows the interior of a car driving down a city street at night. The driver's hands are visible on the steering wheel. The city lights are reflecting in the windshield.

Prompt

p0v_dr1v1n6, video shows a person driving a car on the surface of the Moon. The driver is holding the steering wheel with both hands. The road is covered in lunar dust, and Earth glows brightly in the sky. The driver is looking straight ahead. The car has a digital dashboard and a touchscreen display

Model File and Inference Workflow

📥 Download Links:

• pov_driving_5_epochs.safetensors - LoRA Model File
• wan_txt2vid_lora_workflow.json - Wan T2V 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: p0v_dr1v1n6

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 T2V
• Training Data: Trained on 17 minutes of video comprised of 204 short clips (each clip captioned separately) of various POV driving footage.
• Epochs: 5

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!