
A LoRA that I trained out of boredom.
Lora created with the Anything-v4.5 model, the images I generated were created with the same model, I have not tested them with other models.
Trigger Words: Bonzi buddy, BonziBuddy, bonzi
Other tags: monkey, chubby, purple_body, full_body, standing, black_eyes
LoRA weight of 0.4 - 0.7
CFG scale: 6 - 12
here is the negative prompt I used:
(worst quality, low quality:1.3, large head, monochrome, extra digits1.4), (nsfw, porn, sexual, nipples:1, polygonal), lowres, bad anatomy, extremely bad hands, text, error, missing fingers, triple arms, triple legs, extra digit, very bright illumination, many shadows, double pupils, extra heads, fewer digits, shadows too dark, cropped, worst quality, human, normal quality, jpeg artifacts, signature, watermark, username, plastic texture, deformed mouth, deformed body, blurry, cat_ears1.2, cap, hat, glasses
描述:
It was difficult to find images for the training.
some VAE I used was:
blessed-fix
anything-v4
训练词语: bonzi buddy,BonziBuddy,bonzi
名称: BonziBuddy.safetensors
大小 (KB): 79573
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success