these two things should help your results

https://huggingface.co/datasets/gsdf/EasyNegative/tree/main

https://huggingface.co/FoodDesert/boring_e621/tree/main

this model was trained on nai.ckpt and it performs best on https://huggingface.co/syaimu/7th_furry and lucidmix
it also needs the furry kemono furry hypernetwork strength at 0.2 and "kemonomimi furry mode" at the end of every prompt to further help it along with the style (this does not matter if you don't want the most accurate rikose style), i also recommend making the lora strength 0.8 or lower since going up degrades the quality
trained on 185-186(?) hand captioned (was a mistake to hand caption) images of rikose's art, it does generally well, but there WILL be a v2, when will that version be out? uhhhhhhhhhhh 2027 maybe

the final edit: idk when this lora will enter V2 mode because i keep getting weird yume nikki gens because a few of the images in the training data are messed and i don't want to release a still biffed V2, however, i will release the dataset as is, the fixed version should be better in general

this is also my first lora and it took very long to set up, special thanks to argent and raven from the furry diffusion discord for helping get the trainer up and running though

i also highly recommend adding the anythingv4.vae/orangemix.vae to your results as that will just enhance the thing