
I'm still learning how to train models, but I want to upload the ones I like here
Trigger = "HuemanAboreal2, Green Nature Night Photography Art"
'Green Nature Night Photography Art' by itself may be better, I'm not sure.
DPM++ SDE Karras
This was trained on Juggernaut_X_RunDiffusion.safetensors
Negative prompt of "(bad quality, Worst quality) oil, oily, liquid" really helps
For Version 1.2, use the SDXL base model, and "HuemanAboreal, Nature Night Photography Painting" as the trigger.
Put Blur and Oily in the negative prompt as well, or perhaps all of this:
blur, blur, out of focus, oily, (bad quality, Worst quality), <lora:sdxl_neg_overfit_v1:0.75>, <lora:EasyNegative:0.75>
And DPM++ 2M Karras might be best.
Be sure to try version 1.1, I think that's the best one if you want the photography look. I trained all of these the same day, there's no real order to them.
描述:
训练词语: HuemanAboreal2, Green Nature Night Photography Art
名称: AborealNew-000003.safetensors
大小 (KB): 56078
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success