File size: 7,848 Bytes
6316c5a 8814d2f 6316c5a 8814d2f 6316c5a 1fc27c9 6316c5a 846bc56 6316c5a b0b431f 6316c5a |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 |
---
license: apache-2.0
language:
- en
base_model:
- Wan-AI/Wan2.1-I2V-14B-480P
- Wan-AI/Wan2.1-I2V-14B-480P-Diffusers
pipeline_tag: image-to-video
tags:
- text-to-image
- lora
- diffusers
- template:diffusion-lora
- image-to-video
widget:
- text: >-
The video begins with a tank. A hydraulic press positioned above slowly
descends towards the tank. Upon contact, the hydraulic press c5us4 crushes
it, deforming and flattening the tank, causing the tank to collapse inward
until the tank is no longer recognizable.
output:
url: example_videos/tank_crush.mp4
- text: >-
The video begins with a man. A hydraulic press positioned above slowly
descends towards the man. Upon contact, the hydraulic press c5us4 crushes
it, deforming and flattening the man, causing the man to collapse inward
until the man is no longer recognizable.
output:
url: example_videos/man_crush.mp4
- text: >-
The video begins with a chicken. A hydraulic press positioned above slowly
descends towards the chicken. Upon contact, the hydraulic press c5us4
crushes it, deforming and flattening the chicken, causing the chicken to
collapse inward until the chicken is no longer recognizable.
output:
url: example_videos/chicken_crush.mp4
- text: >-
The video begins with a coke. A hydraulic press positioned above slowly
descends towards the coke. Upon contact, the hydraulic press c5us4 crushes
it, deforming and flattening the coke, causing the coke to collapse inward
until the coke is no longer recognizable.
output:
url: example_videos/coke_crush.mp4
---
<div style="background-color: #f8f9fa; padding: 20px; border-radius: 10px; margin-bottom: 20px;">
<h1 style="color: #24292e; margin-top: 0;">Crush Effect LoRA for Wan2.1 14B I2V 480p</h1>
<div style="background-color: white; padding: 15px; border-radius: 8px; margin: 15px 0; box-shadow: 0 2px 4px rgba(0,0,0,0.1);">
<h2 style="color: #24292e; margin-top: 0;">Overview</h2>
<p>This LoRA is trained on the Wan2.1 14B I2V 480p model and allows you to crush any object in an image. The effect works on a wide variety of objects, from animals to vehicles to people!</p>
</div>
<div style="background-color: white; padding: 15px; border-radius: 8px; margin: 15px 0; box-shadow: 0 2px 4px rgba(0,0,0,0.1);">
<h2 style="color: #24292e; margin-top: 0;">Features</h2>
<ul style="margin-bottom: 0;">
<li>Transform any image into a video of it being crushed</li>
<li>Trained on the Wan2.1 14B 480p I2V base model</li>
<li>Consistent results across different object types</li>
<li>Simple prompt structure that's easy to adapt</li>
</ul>
</div>
<div style="background-color: white; padding: 15px; border-radius: 8px; margin: 15px 0; box-shadow: 0 2px 4px rgba(0,0,0,0.1);">
<h2 style="color: #24292e; margin-top: 0;">Community</h2>
<ul style="margin-bottom: 0;">
<li>
Generate videos with 100+ Camera Control and VFX LoRAs on the
<a href="https://app.remade.ai/canvas/create" style="color: #0366d6; text-decoration: none;">Remade Canvas</a>.
</li>
<li>
<b>Discord:</b>
<a href="https://remade.ai/join-discord?utm_source=Huggingface&utm_medium=Social&utm_campaign=model_release&utm_content=crash_zoom_out" style="color: #0366d6; text-decoration: none;">
Join our community
</a> to generate videos with this LoRA for free
</li>
</ul>
</div>
<Gallery />
# Model File and Inference Workflow
## 📥 Download Links:
- [crush_20_epochs.safetensors](./crush_20_epochs.safetensors) - LoRA Model File
- [wan_img2vid_lora_workflow.json](./workflow/wan_img2vid_lora_workflow.json) - Wan I2V with LoRA Workflow for ComfyUI
---
<div style="background-color: #f8f9fa; padding: 20px; border-radius: 10px; margin-bottom: 20px;">
<div style="background-color: white; padding: 15px; border-radius: 8px; margin: 15px 0; box-shadow: 0 2px 4px rgba(0,0,0,0.1);">
<h2 style="color: #24292e; margin-top: 0;">Recommended Settings</h2>
<ul style="margin-bottom: 0;">
<li><b>LoRA Strength:</b> 1.0</li>
<li><b>Embedded Guidance Scale:</b> 6.0</li>
<li><b>Flow Shift:</b> 5.0</li>
</ul>
</div>
<div style="background-color: white; padding: 15px; border-radius: 8px; margin: 15px 0; box-shadow: 0 2px 4px rgba(0,0,0,0.1);">
<h2 style="color: #24292e; margin-top: 0;">Trigger Words</h2>
<p>The key trigger phrase is: <code style="background-color: #f0f0f0; padding: 3px 6px; border-radius: 4px;"> c5us4 crushes it</code></p>
</div>
<div style="background-color: white; padding: 15px; border-radius: 8px; margin: 15px 0; box-shadow: 0 2px 4px rgba(0,0,0,0.1);">
<h2 style="color: #24292e; margin-top: 0;">Prompt Template</h2>
<p>For best results, use this prompt structure:</p>
<div style="background-color: #f0f0f0; padding: 12px; border-radius: 6px; margin: 10px 0;">
<i>The video begins with a [object]. A hydraulic press positioned above slowly descends towards the [object]. Upon contact, the hydraulic press c5us4 crushes it, deforming and flattening the [object], causing the [object] to collapse inward until the [object] is no longer recognizable.</i>
</div>
<p>Simply replace <code style="background-color: #f0f0f0; padding: 3px 6px; border-radius: 4px;">[object]</code> with whatever you want to see crushed!</p>
</div>
<div style="background-color: white; padding: 15px; border-radius: 8px; margin: 15px 0; box-shadow: 0 2px 4px rgba(0,0,0,0.1);">
<h2 style="color: #24292e; margin-top: 0;">ComfyUI Workflow</h2>
<p>This LoRA works with a modified version of <a href="https://github.com/kijai/ComfyUI-WanVideoWrapper/blob/main/example_workflows/wanvideo_480p_I2V_example_02.json" style="color: #0366d6; text-decoration: none;">Kijai's Wan Video Wrapper workflow</a>. The main modification is adding a Wan LoRA node connected to the base model.</p>
<img src="./workflow/workflow_screenshot.png" style="width: 100%; border-radius: 8px; margin: 15px 0; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
<p>See the Downloads section above for the modified workflow.</p>
</div>
</div>
<div style="background-color: #f8f9fa; padding: 20px; border-radius: 10px; margin-bottom: 20px;">
<div style="background-color: white; padding: 15px; border-radius: 8px; margin: 15px 0; box-shadow: 0 2px 4px rgba(0,0,0,0.1);">
<h2 style="color: #24292e; margin-top: 0;">Model Information</h2>
<p>The model weights are available in Safetensors format. See the Downloads section above.</p>
</div>
<div style="background-color: white; padding: 15px; border-radius: 8px; margin: 15px 0; box-shadow: 0 2px 4px rgba(0,0,0,0.1);">
<h2 style="color: #24292e; margin-top: 0;">Training Details</h2>
<ul style="margin-bottom: 0;">
<li><b>Base Model:</b> Wan2.1 14B I2V 480p</li>
<li><b>Training Data:</b> Trained on 1.5 minutes of video comprised of 20 short clips (each clip captioned separately) of things being crushed by a hydraulic press</li>
<li><b>Epochs:</b> 20</li>
</ul>
</div>
<div style="background-color: white; padding: 15px; border-radius: 8px; margin: 15px 0; box-shadow: 0 2px 4px rgba(0,0,0,0.1);">
<h2 style="color: #24292e; margin-top: 0;">Additional Information</h2>
<p>Training was done using <a href="https://github.com/tdrussell/diffusion-pipe" style="color: #0366d6; text-decoration: none;">Diffusion Pipe for Training</a></p>
</div>
<div style="background-color: white; padding: 15px; border-radius: 8px; margin: 15px 0; box-shadow: 0 2px 4px rgba(0,0,0,0.1);">
<h2 style="color: #24292e; margin-top: 0;">Acknowledgments</h2>
<p style="margin-bottom: 0;">Special thanks to Kijai for the ComfyUI Wan Video Wrapper and tdrussell for the training scripts!</p>
</div>
</div> |