The short and sweat: a furry-centric model merge built around e621 tags.
The long and hard: This is a checkpoint merge of the models Indigo Furry Mix Anima + I Like Furries + (aniMurr_v01 + Anima-base-v1.0) and a custom LoRA consisting of 7471 images scraped from popular e621 posts. The dataset features top-rated posts from every month over the past six years, spanning a wide range of species with a strong emphasis on accurate furry anatomy. Hence the working title of this LoRA was called The Anatomy Problem. I've kept the e621 tags intact for this dataset and as such you'll want to prompt like it's Pony or Illustrious when using this checkpoint. Enjoy!
Recommended Generation Parameters
Sampler: ER SDE
Scheduler: Beta
CFG Scale: 4
Steps: 30–40 (32 recommended)
There are 1360~ different artists within this datasetListed in Alphabetical Order of Artists With 4+ Images:
@aennor, @aer0 zer0, @amawdz, @anchee, @anhes, @arf-fra, @ashraely, @autumn sunrise, @backsash, @bebebebebe, @black-kitten, @blotch, @bonifasko, @bonnie bovine, @botter (artist), @braeburned, @butterchalk, @catcouch, @catniplewds, @cerf, @chelodoy, @chunie, @claweddrip, @clockhands, @cobaltsynapse, @codyblue-731, @coffeesoda, @cold-blooded-twilight, @con5710, @cooliehigh, @cuteconfy, @clyde wolf, @dacad, @daftpatriot, @dagasi, @dark natasha, @dark violet, @darkgem, @delki, @desertkaiju, @destabilizer, @digitoxici, @dimwitdog, @discordthege, @dlw, @dr.kiyo, @dripponi, @drmax, @edjit, @einshelm, @elicitie, @enro the mutt, @etheross, @feretta, @fenn.perrox, @fluff-kevlar, @foxinuhhbox, @fumiko, @futadolly, @gewitter, @glopossum, @gmad, @goonie-san, @gothbunnyboy, @hima nsfw, @himeragoldtail, @hioshiru, @hooves-art, @honovy, @hyattlen, @imadeus, @imgonnaloveyou, @imprettynormai, @inu-sama, @iztli, @jarnqk, @jarlium, @jishinu, @jotun22, @juiceps, @jumpy jackal, @kaffii, @kakhao, @kapri, @kashmere, @kaylafox, @keadanger, @keeltheequine, @kenket, @ketei, @kilver, @kinsheph, @klaide, @km-15, @kojisauce, @kraaisha, @laobai, @lazysnout, @letodoesart, @lichfang, @link6432, @lizardlars, @lizet, @loimu, @loonanudes, @lockworkorange, @lostdragon01, @lvlirror, @lynncore, @magnetus, @manmosu marimo, @marblesoda, @marsccts, @mawmain, @maxydont, @meesh, @mercurial64, @merrunz, @miles df, @mindkog, @nastycalamari, @nepentz, @nikkibunn, @notafurrytho, @notglacier, @nsfw def, @nyankone, @oouna, @owooops, @oxfort2199, @pakwan008, @patto, @personalami, @phenyanyanya, @phluks, @photonoko, @pinkaxolotl, @pixelsketcher, @plattyneko, @plive, @pony dreaming, @prsmrti, @pullmytail, @r-mk, @r3drunner, @rabbitadvisory, @rajii, @ravieel, @reallynxgirl, @redishdragie, @redvais, @roy arashi, @ruaidri, @sagaris uwu, @scarf (artist), @scarlanya, @scarlet-frost, @schmutzo, @scruffythedeer, @securipun, @seibear, @sepulte, @sepiruth, @shirsha, @shinodage, @sicmop, @sijimmy456, @singafurian, @skully, @smitty g, @smugbluefaun, @snowskau, @sonsasu, @sophiepoof, @spaal, @spectrumshift, @spuydjeks, @ssirrus, @ssssnowy, @st4rs6, @supersquishy, @svadil, @swaybat, @syn-6, @syuro, @teranen, @tgt1512, @the giant hamster, @the gentle giant, @the-minuscule-task, @thebatfang, @thesecretcave, @tofu froth, @tojo the thief, @tom fischbach, @todding, @trmk2, @tricksta, @truegrave9, @twang, @unsavoryimpcat, @utopianvee, @vesper art, @vksuika, @virtyalfobo, @voidpastelchalk, @w4g4, @waspsalad, @watsup, @welost, @welwraith, @whisperfoot, @whyvern, @wizzikt, @wolfblade, @woolrool, @wugi, @xerun, @yakovlev-vad, @yuio, @zaush, @zackary911, @zeidre, @zeforge, @zephyxus, @zica*(While these artists appear in the training data, their stylistic impact on this checkpoint may vary significantly)
Lastly, I'd like to give a special thanks to @someyoungguy69 and @indigowing for their amazing checkpoints! Go check out their work!
Description
Vastly improved and expanded dataset. Improved tag adherence and understanding. Significant improvement to style when using other LoRAs with this model.
FAQ
Comments (4)
Some nice anatomy fixes here. I will have to try this one!
Cheers for the artist list, now let's see how it compares to Nova Furry in that regard...
This checkpoint work great with my Mystic X lora, nice work.
Love all the big effort people are providing for creating these mdoels! But is it jsut me or the uqality as if slowly deterriarating with every new model? I was just comparing the Homosimile model + furry artist style lora and This model + the same furry artsit style model and results were almsot always looking better with Homosimile. And with the Animur v1/v2 they many times look "smooth" ? Textureless, like no brush work or anything e.t.c Despite the artist I was trying has brush work and textures, it's like I see the outline of the artist but it's not fully there?
I feel like maybe some of the loras or models have too much weighting on some concepts or artists and now it's bleeding into everything?
I have my personal Hardware on which I can train loras and possibly models e.t.c (2 NVIDIA GPUs 40GB VRAM). If you'd like I could try and help with training and such.



















