Squish Effect LoRA for Wan2.1 14B I2V 480p

Overview

This LoRA is trained on the Wan2.1 14B I2V 480p model and allows you to squish any object in an image. The effect works on a wide variety of objects, from animals to vehicles to people!

Features

• Transform any image into a video of it being squished
• Trained on the Wan2.1 14B 480p I2V base model
• Consistent results across different object types
• Simple prompt structure that's easy to adapt

Community

• Discord: Join our community to generate videos with this LoRA for free
• Request LoRAs: We're training and open-sourcing Wan2.1 LoRAs for free - join our Discord to make requests!

Prompt

In the video, a miniature dog is presented. The dog is held in a person's hands. The person then presses on the dog, causing a sq41sh squish effect. The person keeps pressing down on the dog, further showing the sq41sh squish effect.

Prompt

In the video, a miniature tank is presented. The tank is held in a person's hands. The person then presses on the tank, causing a sq41sh squish effect. The person keeps pressing down on the tank, further showing the sq41sh squish effect.

Prompt

In the video, a miniature balloon is presented. The balloon is held in a person's hands. The person then presses on the balloon, causing a sq41sh squish effect. The person keeps pressing down on the balloon, further showing the sq41sh squish effect.

Prompt

In the video, a miniature rodent is presented. The rodent is held in a person's hands. The person then presses on the rodent, causing a sq41sh squish effect. The person keeps pressing down on the rodent, further showing the sq41sh squish effect.

Prompt

In the video, a miniature person is presented. The person is held in a person's hands. The person then presses on the person, causing a sq41sh squish effect. The person keeps pressing down on the person, further showing the sq41sh squish effect.

Model File and Inference Workflow

📥 Download Links:

• squish_18.safetensors - LoRA Model File
• wan_img2video_lora_workflow.json - Wan I2V with LoRA Workflow for ComfyUI

Using with Diffusers

pip install git+https://github.com/huggingface/diffusers.git

import torch
from diffusers.utils import export_to_video, load_image
from diffusers import AutoencoderKLWan, WanImageToVideoPipeline
from transformers import CLIPVisionModel
import numpy as np

model_id = "Wan-AI/Wan2.1-I2V-14B-480P-Diffusers"
image_encoder = CLIPVisionModel.from_pretrained(model_id, subfolder="image_encoder", torch_dtype=torch.float32)
vae = AutoencoderKLWan.from_pretrained(model_id, subfolder="vae", torch_dtype=torch.float32)
pipe = WanImageToVideoPipeline.from_pretrained(model_id, vae=vae, image_encoder=image_encoder, torch_dtype=torch.bfloat16)
pipe.to("cuda")

pipe.load_lora_weights("Remade/Squish")

pipe.enable_model_cpu_offload() #for low-vram environments

prompt = "In the video, a miniature cat toy is presented. The cat toy is held in a person's hands. The person then presses on the cat toy, causing a sq41sh squish effect. The person keeps pressing down on the cat toy, further showing the sq41sh squish effect."

image = load_image("https://huggingface.co/datasets/diffusers/cat_toy_example/resolve/main/1.jpeg")

max_area = 480 * 832
aspect_ratio = image.height / image.width
mod_value = pipe.vae_scale_factor_spatial * pipe.transformer.config.patch_size[1]
height = round(np.sqrt(max_area * aspect_ratio)) // mod_value * mod_value
width = round(np.sqrt(max_area / aspect_ratio)) // mod_value * mod_value
image = image.resize((width, height))

output = pipe(
    image=image,
    prompt=prompt,
    height=height,
    width=width,
    num_frames=81,
    guidance_scale=5.0,
    num_inference_steps=28
).frames[0]
export_to_video(output, "output.mp4", fps=16)

---

Recommended Settings

• LoRA Strength: 1.0
• Embedded Guidance Scale: 6.0
• Flow Shift: 5.0

Trigger Words

The key trigger phrase is: sq41sh squish effect

Prompt Template

For best results, use this prompt structure:

In the video, a miniature [object] is presented. The [object] is held in a person's hands. The person then presses on the [object], causing a sq41sh squish effect. The person keeps pressing down on the [object], further showing the sq41sh squish effect.

Simply replace [object] with whatever you want to see squished!

ComfyUI Workflow

This LoRA works with a modified version of Kijai's Wan Video Wrapper workflow. The main modification is adding a Wan LoRA node connected to the base model.

See the Downloads section above for the modified workflow.

Model Information

The model weights are available in Safetensors format. See the Downloads section above.

Training Details

• Base Model: Wan2.1 14B I2V 480p
• Training Data: 1.5 minutes of video (20 short clips of things being squished)
• Epochs: 18

Additional Information

Training was done using Diffusion Pipe for Training

Acknowledgments

Special thanks to Kijai for the ComfyUI Wan Video Wrapper and tdrussell for the training scripts!