
Trained with 60 photos of the actress. The sample images were generated with AutismMix.
Recommended settings:
-
Sampler:DPM++ 2M Karras /Euler a - 25 steps
-
CFG: 7
-
Resolution: 1024*1024, 832*1280, 1280*1024, (works with most common xl resolutions)
-
Lora weight: 0.8 - 1.25
-
HighRes: 4x_NMKD-Siax_200k, upscale by 1.5x, 20 steps, 0.35-0.45 denoising
Basic prompt:
lrbly, 1girl, brown hair, long hair, brown eyes, medium breasts
Additional notes:
-
Increase the lora's weight to give her face more detail, decrease it to make her look more stylized.
-
Add "3d" to your negative prompt to make her more stylized.
-
Add "source_anime" to your prompt to make her look more like an anime girl
-
Make sure to describe the background/setting in detail
-
Don't forget about the Pony's usual quality tags.
-
Also works with photorealistic Pony models like Everclear
描述:
训练词语: lrbly
名称: 435836_training_data.zip
大小 (KB): 9282
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: LauraBailey.safetensors
大小 (KB): 445800
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success