
My model is a mix of many models including any pony based models, not entirely sure what all is in it as I never document my merges. Good results when using these positive and negative prompt.
Positive prompt = "(detailed background),(score_10, score_9, score_8, score_7),(score_10_up, score_9_up, score_8_up, score_7_up),"
Negative Prompt = "(score_6, score_5,score_4,score_3,score_2,score_1), low quality, worst quality, gpo, tdj, eqg, wpt, aur, hnj, bry, bfu, cle, cwn, eqg, ewu, bwl, bse,"
Not entirely sure what the hieroglyphics are at the end of the negative prompt but I stole it from another image I saw and it's worked out quite nicely to improving the style of lighting in most cases.
Does good nsfw too, just didn't want to make any example image for it. All examples are made using ranbooru random anime prompting and the two prompts I just shared in this text.
Permissions are as per Pony's requirements.
描述:
glhf
训练词语: (detailed background),(score_10, score_9, score_8, score_7),(score_10_up, score_9_up, score_8_up, score_7_up)
名称: anyDeath_anyDeath.safetensors
大小 (KB): 6775437
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success