This is a LoRA trained on the same dataset as the V2 of this finetune<a target="_blank" rel="ugc" href="https://civarchive.com/models/1794928?modelVersionId=2499924">https://civarchive.com/models/1794928?modelVersionId=2499924</a>It perfroms worse than a finetune, despite weighting more than 1GB, but being a LoRA it offers some flexibility a finetune doe snot have. And people really wanted it.It was trained on synthethic images and images of these artists:anythinggoes,mr_kurz,elirawne,ayaswan,carl,dolcett,gblptto,guroguru,inferna_draganis,kllc,kurohime,plplc,rbqinori,veild616,venus-flowers-meat,xingzhe,jjjj,kihyahya,kubikiri,kuensan,marost,nutzlos,toxra,yb,m0zg,crwst,If you require a complete set of tags used in the training please reach to me via a private message.

Compilative_G_LoRA

dependsing on tags different weight is reccomended. In general weight 0.8 works

Files with SHA256: 144a33b71e90c4b5b76ee668e319862321b657bc571bdedbc068c61bcafb11ef