Get the latest global workflows on my RunningHub homepage and run them free on RTX4090.
Sign Up with my invite link for 1000 bonus points + 100 daily:
Run Directly:
https://www.runninghub.ai/post/1956586031082602497/?inviteCode=rh-v1221
Please credit and link to my homepage if reprinting.国内合作加Q：549791525
Here's the explanation: Most special effects loras do not require low noise. If you understand the principle, you should know
I trained this LoRA because I found it difficult to generate expressions of convulsive crying (crying to the point of trembling) with the wan2.2 model.
Recommended weight: 1.0
Example prompt(s)：
kuqi567, a static shot of a young woman in a black hoodie wailing uncontrollably with her eyes shut while standing in a doorway.
kuqi567, a static shot of a young woman in a red headscarf wailing uncontrollably in front of a vibrant red wall under a blue sky.
Description
FAQ
Comments (12)
Is there a low coming, or only a high?
Read the description mate!
"Here's the explanation: Most special effects loras do not require low noise. If you understand the principle, you should know"
Most of the loras I trained only need high noise, because low noise just adds detail, and wan2.2 without loading lora is fully capable of supplementing these details
Nice, thank you, will try it. I was looking to something like this.
I respect your effort and great loras released so far. Cheers!
Just seeing this💥effect💥, gives me 😢 tears😢of joy!!😋
Are you sure this is meant for the high noise model? I tried it on the high noise model (first stage) and got barely any effect, but on the low noise model (second stage) I got exactly what I expected to see.
If the dataset is trained on high noise, some lora high noise can be used on low noise, and the reverse is also true. If you use lightx2v to accelerate or do not use any acceleration, he can produce an effect on high noise
it might depend heavily on the input image. i get great results at 0.8-1.0 if the input image is a closeup shot, and i'm generating at a decent resolution like over 480. if you're using a medium shot/full body and the face is small in the image (and/or if you're generating at a lower res, and the face is therefore smaller) the effect might not work
Do not stop!
Thanks for making this lora, it's great.
"I know now why you cry..."
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.