metadata

base_model:
  - Lightricks/LTX-Video
tags:
  - ltxv
  - 13B
  - text-to-video
  - lora

ltxv (13b 0.9.7)

Description:

This is a LoRA that was trained on a reversed dataset, -The trigger word would be "DEFLATE",

A n example of a prompt : "DEFLATE a {subject] gets compressed", replace subject with the details of the object to be affected.
You can reverse the resulting video back (in comfyui you can use the "Reverse Image Batch" node) to get the desired inflate effect.