Request a Wan2.1 LoRA on our Discord and we will train and open-source it for free.
Join our Discord to generate videos of the Aging Timelapse LoRA for free.
Wan2.1 14B I2V 480p v1.0:
Trained on 20 seconds of video comprised of 4 short clips of people timelapse aging. This was trained on the Wan.21 14B I2V 480p model.
The trigger word is: 'ag145ng aging time-lapse'
See below for the prompt structure that has worked best for me.
Recommended Settings:
LoRA strength = 1.0
Embedded guidance scale = 6.0
Flow shift = 5.0
Here's a link to the Wan2.1 I2V LoRA inference workflow I used to generate these videos: https://huggingface.co/Remade/Squish/blob/main/workflow/wan_img2video_lora_workflow.json
This is a slight modification to Kijai's version, with the main difference being the addition of the WanVideo Lora Select node, connected to the 'lora' field of the WanVideo Lora Select node. Find Kijai's original workflow here:
https://github.com/kijai/ComfyUI-WanVideoWrapper/blob/main/example_workflows/wanvideo_480p_I2V_example_02.json
Prompt Structure:
A [description of person] undergoes a transformation over time. The ag145ng aging time-lapse shows subtle wrinkles forming, hair gradually graying, and facial structure shifting with age. Skin texture changes, posture slightly adjusts, and signs of maturity emerge. Over time, hair thins or turns fully gray, and expressions become more aged. The final scene reveals the [description of person] in his elderly years, showcasing the full progression of aging.
for the description of the person, including details like the person's race and gender seems to help.
Let me know if there are any questions, I'll be happy to help!