LoRA weight 0.6 - 0.8 are better.
v2 improvement, reduced multiple tails issue.
v2.1 Experimental "no tail" edition. The "no tail" effect is uncertain, about 60% will have a tail added.
V2.2 For Model & VAE compatibility test version. (Maybe delete it later?)
I made it without a trigger. but prompting for anthro, anthromorph, shark, shark_girl, will good results.
I use an online service to generate these images, so the metadata seems to be dropped sometimes. sorry for any inconvenience.
Description
V2.2 For Model & VAE compatibility test version. (Maybe delete it later?)
FAQ
Comments (2)
What is the model you used, this EMS whatever its called? Sorry, I'm new and I also want to know the upscaler and checkpoint system you used, if that is a question to ask. Thanks.
I generate these sample images exclusively using online services (such as seaart, tensorart, and pixai), so the Generation Data is not displayed correctly. Model names often appear as "EMS-xxxx-EMS", which is a useless code. I also use an upscaler from an online service.
Sample https://civitai.com/images/2864722 uses “AnthroDaVinci” model
(Oh... This model appears to be currently unavailable?)
Sample https://civitai.com/images/2241979 uses “YiffyMix” model.
https://civitai.com/models/117416/yiffymix-20-fp16cleaned
Sample https://civitai.com/images/2864764 uses “RockyMix” model.
https://civitai.com/models/285447/rockymix-v2
These three models fit well with my LoRA. It is possible to change the taste from "cartoon" to "manga" depending on the prompt. Other models often have a "human face" or "fur".
I'm not a native English speaker. Did you get the answers you needed? thank you.