This Motion-Lora was trained under SD_1.5 base model and animatediff_V3 model.

The purpose of the training was to produce short videos of buildings from orbit during flights.

Recommended usage:

• Frames: 16-32

• Resolution: 512x512, 512x768 , 768x512

• Models: Realistic models

• Sampler: DPM++ 2M Karras

• Keep prompt simple
  

Need to be tested on humans and objects.

Be sure to tag me in IG if you use it! @pixelpoetryx