
A LoRA I created of Catherine Elizabeth Her Royal Highness the Duchess of Cambridge (Kate Middleton).
I have found the best results with SDXL1.0 base are:
CFG = 7
Steps = Base 35 + Refiner 5 = 40 total
Other checkpoints differ for example the best results for rsmgalaxyrealcartoonx_v10 was:
CFG = 3
Steps = Base 30 + Refiner 5 = 35 total
All Images were pulled from before Face fix and upscale occurred, two however do contain other LoRAs. Thought I'd share my first LoRA and give back to the amazing CivitAI community.
描述:
V1
训练词语:
名称: 240209_training_data.zip
大小 (KB): 33613
类型: Training Data
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success
名称: Kate_Middleton-000006.safetensors
大小 (KB): 223111
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success