2.5D painterly anime style with bright vintage colors &amp; dynamic lighting. I like using it with semirealistic/realistic/Niji style base models &amp; loras to add the softness and painterly feel to the images.Recommended weight: 0.3 - 1It's better to use a canvas size higher than 1024 x 1024. My default canvas size: 1024 x 2048 (no need upscaling)Soft: It's better to use it at a high weight (around 1) &amp; combine it with base models &amp; loras that have a strong style. Useful tags: dappled sunlight, light illuminating face, shadow covering half of face, cinematic lighting.Shadow/Mood: Strong effect. It's better to use it at a weight of around 0.3 - 0.7 if you combine it with other LoRAs. Both versions have the same dataset so there's almost no difference between them.(I train this mostly for my personal use though, but i think it'd be nice to upload it here. If you like it, please leave a like. I also love to see what people create with my LoRAs!)

VG__fix_caps__2nd_epoch_8.safetensors

Shadow

Mood

Soft

2.5D painterly anime style with bright vintage colors &amp; dynamic lighting. Only trained on female character images but it can be used to gen male characters, non-character images as well.Recommended weight: 0.3 - 1(The 1st versions have a stronger style. The 2nd versions are just me experimenting with the training parameters)

Files with SHA256: ee811dd7b74edff580150e28ae853ac3b38ee5f44ca3166ef88327d052c4e454