
Illustrious update
More coherent, still exibiting similar issues to pony, but a general improvement. Test out generating in base illustrious then change to desired model for (imo) best outcomes
Pony version update
Updated to Pony using CivitAI trainer. Should possibly be a bit more coherent this time around.
Whiskers are a bit on/off still, and I highly recommend using "black pupils, red sclera" as part of the prompts still.
"Broken costume" and "open costume" should still work for NSFW, same as using "Anthro" for the furry section. It might actually be a bit more favored towards furry this time around due to a shift in the dataset, so add "furry" in negatives to mitigate.
It defaults to "a-pose" quite heavily, so I suggest either prompting a specific pose or adding it to the negative prompts
I noticed a lack of Vanny from Security breach here, so I decided to give it a go as my first LoRA, I apologize for any inconsistencies as it is my first project. I welcome any and all suggestions/help/tips you have!
prompting "Vanny, patchwork costume, bunny costume, blue bow" gives a likely summon, however the pupils tend to be a bit unruly, which will generally require a few generations. "black pupils" might help a tad
You can use "broken costume" or "open costume" to do NSFW, or "Anthro" to attempt it without human skin.
Preview images were generated using "anyloraCheckpoint", latest "Kavkamix", "Toonify" and "CrossKemonoFurryModel"
描述:
训练词语: vanny,patchwork costume,black pupils,red sclera,blue bow
名称: Vanny_IL.safetensors
大小 (KB): 223115
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success