---
base_model:
- Lightricks/LTX-Video
tags:
- ltxv
- 13B
- text-to-video
- lora
widget:
- text: >-
    Walgro style. A woman waits at a bus stop in the early morning, headphones resting over her blue hair, her gaze focused on her phone as she scrolls. The rising sun casts soft light across the pavement, illuminating the quiet street. She shifts the weight of a small backpack on her shoulders, occasionally glancing up to check if the bus has rounded the corner. Her expression remains focused yet calm, her eyes flicking between the phone and the stillness of the morning around her.
  output:
    url: assets/walgro_style_sample_3.mp4
- text: >-
    Walgro style A man stands alone in an empty train station, the overhead lights casting harsh shadows on the cold marble floor. The camera starts by focusing on his upper body, with his dark coat flowing behind him as he walks slowly toward the platform. His face is partially obscured by the brim of his hat, but his sharp eyes, visible beneath, convey a sense of urgency. The distant sound of an approaching train echoes through the empty station. As the camera pulls back, the full scale of the deserted station is revealed, the train tracks stretching into the unknown, the bright lights above contrasting with the surrounding shadows, adding to the scene's dramatic, cinematic tension.
  output:
    url: assets/walgro_style_sample_1.mp4
---

# LTXV Wallace&Gromit Style LoRA v1

<Gallery />

## Trainig details:
Trained only on videos.

- dataset: 90 videos
- steps: 42000
- LR: 2e-4
- optimizer: adamw
- scheduler: contant
- rank: 128
- batch size: 1
- gradient accumulation steps: 1

For training I used the official [LTX-Video Trainer](https://github.com/Lightricks/LTX-Video-Trainer) repo. 

## Inference

It seems, the model likes long descriptive prompts. Look at the attached videos for prompt examples. Based on my tests if you use short prompts, the lora effect is weak.

For inference I used ComfyUI.

<b>Trigger words</b>: Walgro style

<b>Strength</b>: 0.9-1.0

## Acknowledgement

Special thanks to [Lightricks Team](https://github.com/Lightricks/LTX-Video) for making this available for the community for free!