LTX-2-19b-Squish-LoRA

LTX-2-19b-Squish-LoRA

This is a fine-tuned version of ltx-2-19b-dev.safetensors trained on custom data.

<!-- ## Model Details --> <!-- - **Base Model:** [`ltx-2-19b-dev.safetensors`](/root/LTX-2/packages/ltx-trainer/models/ltx-2-19b-dev.safetensors) - **Training Type:** LoRA fine-tuning - **Training Steps:** 2000 - **Learning Rate:** 0.0002 - **Batch Size:** 1 --> <!-- ## Sample Outputs | | | | | |:---:|:---:|:---:|:---:| | <br><details style="max-width: 300px; margin: auto;"><summary>Prompt</summary>SQUISH two hands squeezing a squeezable object that is shaped like a woman's head.</details> | -->

Usage

This model is designed to be used with the LTX-2 (Lightricks Audio-Video) pipeline.

🔌 Using Trained LoRAs in ComfyUI

In order to use the trained LoRA in ComfyUI, follow these steps:

1. Copy your trained LoRA checkpoint (.safetensors file) to the models/loras folder in your ComfyUI installation.
2. In your ComfyUI workflow:
   • Add the "Load LoRA" node to choose your LoRA file
   • Connect it to the "Load Checkpoint" node to apply the LoRA to the base model

You can find reference Text-to-Video (T2V) and Image-to-Video (I2V) workflows in the official LTX-2 repository.

Trigger Word

<!-- Example prompts used during validation: -->

• squish it

This model inherits the license of the base model (ltx-2-19b-dev.safetensors).

Acknowledgments

• Base model: Lightricks
• Trainer: LTX-2