
Made just for fun. I wanted to see what happens if you mix photos of snow, mud, puddles, old houses, cars, roads and an anime model. I'm rather pleased with the result, but you shouldn't expect much from this LoRA.
Two versions, v1 is under-trained (smaller dataset, keeps better anime faces, needs more weight up to 1.5)
(default) v2 - works fine at weight 1, but makes faces a little worse
Trained on animefull aka novelai
//psta mean PoST-Apocalypse
描述:
Recommended weight: 0.7-1.0
训练词语: psta,psta background,snow,vehicle,winter,tree,day,night,flooded,fence,real world,russian city,ruins
名称: examples.zip
大小 (KB): 6080
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: psta_2-000018.safetensors
大小 (KB): 36977
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success