Guidance:
This lora can be used with the t2v or i2v Wan2.2 Models.
In the "About this version" section you can find a list of the most frequent nouns and verbs in the training set.
If the dynamic effect does not work well, consider increasing the high-noise LORA; if there is contamination to other detail-enhancing LORAs, try lowering the low-noise LORA.
Preview Image Prompt:
I will mark prompts in the relevant preview images for your reference.
说明：
这个lora可使用在Wan2.2的t2v或是i2v模型。
在"About this version"部分，你可以看到训练集中频率最高的名词与动词清单。
如果没有动态效果不好使，可以考虑增加高噪LORA；如果对其他细节增强型 LORA 造成污染，请尝试降低低噪LORA。
预览图提示词：
我会在各预览图中标记使用的提示词，作为参考使用。
Description
The 15 most frequent nouns in the dataset:
skinsuit - 134 times
skin - 95 times
shell - 72 times
face - 48 times
man - 45 times
dress - 42 times
eyes - 38 times
head - 35 times
interior - 32 times
wrinkles - 30 times
folds - 28 times
body - 26 times
room - 25 times
floor - 24 times
tension - 22 times
The 15 most frequent verbs in the dataset:
wearing - 48 times
collapsing - 42 times
pulling - 35 times
revealing - 32 times
stretching - 30 times
entering - 25 times
showing - 24 times
pushing - 20 times
emerging - 18 times
deflating - 16 times
flattened - 15 times
holding - 14 times
peeling - 12 times
retracting - 10 times
disappearing - 9 times
FAQ
Comments (4)
"And the most terrifying LoRA of the year award goes too..."
Looks well done tho! Might play with this some time.
concept works. i don't wanna deflate breasts to make pancakes. it for sure can give birth to some bigger sized objects.
I'm happy to see more and more creative and absurd lora's come out that aren't limited by reality!
I used to make a lot of horror image art and would love to eventually try some AI horror videos and this kind of a thing can definitely work in that as well!
yoooooo its been a long time since I saw a LoRA that made me go 'I am going to set aside an entire afternoon to see what i can do with this' XD