TIP: Add "3D" to the negative prompt. This will massively help combat 3D looks.
🔥 DADDY KANE IS HERE! 🔥 And he’s ready to bless your sinful souls with exactly what you crave. 😈💦
I know you all either want to devour him 🍽️ or let him devour you—there’s no in-between. You all have one goal: to degrade him until he fucks you, or you fuck him. Either way, somebody’s getting ruined, and I respect that level of dedication. 😏👿
This LoRA is looking extra 3D and ugly right now because, let’s be real, I barely have any photos—he’s not even a day old yet. But don’t worry, once I get more, I’ll be updating this masterpiece like an artist perfecting their finest work... just, you know, with significantly filthier intentions. 🎨😜
IL V3 will come in couple of weeks with many more photos for higher flexibility.
I know what you want. You know what you want. You want to see him naked. 🍆👀 Don’t fight it—embrace your inner perv! Now let’s flood this thing with some unholy levels of NSFW! 🍆💦
Trigger: fletcher_kane, furry, muscular male, orange eyes,
Extra Trigger: Tattoo, jewelry, striped pants, handgun, holster, cape, belt, watch, necklace, cane
Recommended Weight: 1 seems good for me. Adjust it up or down based on the checkpoint and prompt.
Version
IL V2: Added more photos and made it way easier to use. Turns out, I kinda forgot some tags for the watch and bracelet. Oops.. my bad! But hey, it still works just fine. Definitely a huge upgrade from V1. Tattoos are tricky since everyone imagines them differently, so expect variation every time. V3 is gonna take a while (probably a few weeks) but it'll come with way more photos to make things even smoother. IMO, the training data could still use more images, but overall, it’s pretty solid.
IL V1: 💩💩Absolute garbage. Don’t even waste your time. 💩💩
Let me know if you want Pony Version NOW or wait for better results.
And don’t be shy—show off your filthy creations! Let’s see just how wild you can get. 🍆🔥
Description
Add couple of photos to reduce the 3D effect.
FAQ
Comments (3)
how did u make a lora this fast
Dedication, Google, and many LoRA trials (using civitai lora training). V2 was like my 5th LoRA training for this character. I cropped one image into two images for some images. This help adding more images to the training set. I think three images were cropped to create six additional ones. The issue with a lack of photos is that it can make the images feel static and reinforce 3D style that I don't want include. I did my best to reduce that. Honestly, I spent more time searching for images than actually creating the LoRA. LMAO.
Of course, I removed some images that had a negative impact on the results, such as ones with shadows on the face and a big glowing light on top of the head. That image got baked so hard. A small dataset tends to make every image bake in somehow. Just need to be careful with that.
I'm sure there are more efficient and simpler ways to create a character LoRA with fewer images. I definitely don’t consider myself an expert in character creation.
Honestly, I feel that @Not_So_Saint_Dog and @Otter_Enjoyer are better people to ask about using a small dataset since they have created so many niche character LoRAs.
Maybe I am selling myself too short LMAO.
I'm still planning to make another version with a larger dataset for more flexibility and better compatibility when combining LoRAs. I want this LoRA to be as great as my Cerberus (Fortnite).
I don't make this fast as you might think. It took me couple hours during weekend to make this.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.