
If you want to learn how to train faces with high fidelity, reach out to me on my Patreon. :)
About:
The purpose of this LoRA is to (try to) closely replicate the appearance of Gianna Michaels, a famous adult movie actress.
It was trained in multires (512 up to 1024), using simple captions and with 54 high-resolution images.
The demo images of the LoRA aren't edited, but they are cherry-picked and generated using hires-fix (0.25 ~ 0.45 denoising).
Suggested Settings for Inference:
Model: epicrealism_pureEvolutionV5 (but works with 97,2% of models)
Positive: a woman with winter clothes, gianna michaels, fully clothed, head and chest only, private press <lora:Gianna_Michaels_SD15_by_IsnAI:1>
Negative: (cleavage, nsfw, topless, nude, naked)
Sampler: DPM++ 2M Karras
Steps: 32
CFG: 7 (5 ~ 7)
Resolution: 640x960
Upscaler: 4x_NMKD-Siax_200k
Note: Applying high denoising values (0.45~) in hires-fix enhances the appearance fidelity.
Trigger:
Main: gianna michaels
描述:
训练词语: gianna michaels
名称: Gianna_Michaels_SD15_by_IsnAI.safetensors
大小 (KB): 36975
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success