Anima v1.0 Trained with Anima base v1.0 checkpoint with 16 epochs(stopped at 11st epoch), with Prodigy scheduler and network rank=16.Nothing changed on the dataset, may revamp the whole salt model with all clothing in one model. But it seems that Anima lora model will fit faster and earlier compared to noobai and illustrious, which causing problem on training lora with multiple clothing in my personal experience(e.g. 6 clothings in one model). Trying to tackle that. ----------------------------Noobai v1.0 Trained with NoobAi early access version checkpoint with 70 epochs(stopped at 28th epoch), with around 25 images in dataset.I don't actually think the model is doing very good job this time, I am thinking how to improve the training and cleansing the data. I expect to release updated version with better quality in upcoming future (If I have spare time for this)Please give some feedbacks to me about this model, so I can brainstorm how to improve this.Compulsory Taggings:salt(finale), 1girl,Clothings available:Finale maid clothings:long hair, dress, bow, hair between eyes, twintails, low twintails, green eyes, purple eyes, gradient eyes, two-tone eyes, purple hair, frills, detached sleeves, bowtie, cat ears, apron, cat tail, fur trim, thick eyebrows, maid headdress, blue dress, blue bow, white apron, maid apron, blue bowtie, tail ornament, fur-trimmed sleeves, fur-trimmed dress, tail bow, blue sleeves, tail bell, black socks, blue footwear, black pantyhose, hair bun, double bun, striped clothes, striped dress, cat print, animal print,Milk version sailor clothings:long hair, skirt, bow, ribbon, twintails, school uniform, purple eyes, hair ribbon, purple hair, pantyhose, serafuku, cat ears, sailor collar, apron, cat tail, wrist cuffs, maid, maid headdress, low twintails, waist apron, tail ornament, tail bow, tail ribbon, low twintails, blue sailor collar, blue skirt

Files with SHA256: f9da56b3a45e00ba21ac535d2632696676d33a5597728927dd6fb2a818262e13