ltxv (13b 0.9.7)

Description:

This LoRA was trained on FPV drone views of cars driving and drifting, it's ideal when used alongside a lora trained on the specific car images to prevent hallucinations, because the drone view animation tends to show almost all the sides of the car in one shot. You can alternatively use multi keyframe conditioning to fix the hallucinations instead.

Trigger word : "FPVDRIFT"

prompt example : "FPVDRIFT, a car driving through a sandy area drifting on dust, FPV drone view."

Words like "drifting", "dirt", "dust", "fpv drone view", were all repeatedly included in the dataset, which means they can help you get the effect consistently.

The LoRA will also introduce some biases like stickers on the car, since most of the footage it was trained on was cars that had stckers on them.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for oumoumad/fpv-drift-lora

Adapter
(19)
this model