This is an experimental checkpoint that combines some of the best realistic Illustrious models with bigASP 2.5. It worked.
BTW, v2.4 is now available here! Give it a try!
❤️ If you like Snakebite, you can help offset the cost of training:
Why it matters
bigASP has great prompt adherence but it is wildly inconsistent as far as style and composition. It feels like a base model with untapped potential. Dialing in the right settings is like trying to solve a Rubik's Cube.
Illustrious models, on the other hand, lose many concepts when going 3D. They're hit-or-miss even with fairly popular booru tags. But the lighting and composition of these weights are still 👌
I wanted to see if we can get the best of both worlds, and it turns out we kinda can! Careful block merging is the key. We can inject bigASP's output_blocks.0 to acquire much of its conceptual knowledge. Adding its middle_block.2 seems to reduce anatomical issues (otherwise, you'll get a lot of extra arms and fingers.)
First impressions
Best prompt adherence for a realistic SDXL model I've ever seen. And no, not just for smut. (But especially for smut)
Compatible with booru tags as well as natural language. I think a mixed prompt approach is best: I use about 70% tags (no underscores) and 30% natural language in a prompt.
Understands bigASP's style tags to an extent, including stuff like masterpiece quality and 35mm.
It *feels* like something new and worth exploring. Don't sleep on bigASP!
Drawbacks
Illustrious 2.0 models support resolutions up to 1024x1440 or 1024x1536 without horrific stretching of anatomy (e.g. longcat-type torsos), but bigASP's optimal resolution is only 832x1216... and I don't recommend going above that in Snakebite. If you do, the anatomy will be mostly okay (which is surprising), but image composition becomes very odd and unpleasant.
Since we're in a strange new latent space, your existing LoRAs won't work very well. But they're worth retraining.
Recommended Settings
The Turbo variant is better for inference. It's super fast and has slightly improved aesthetics. The non-turbo version is useful for finetuning, and it can produce nice textures if you don't mind waiting 25+ steps.
Turbo
8 or 9 steps
LCM sampler
CFG 1
Custom sigmas below, or simple
Full
20 to 28 steps
Euler ancestral sampler
CFG 3 to 4
Custom sigmas below, or simple
Custom sigma curve (you can use comfyui-kjnodes to apply):
15, 8, 4, 2, 2, 1, 0.4, 0.2, 0
If you're getting mangled limbs, you can often salvage the image by adjusting the first few values of your sigma curve. Here's one that is more stable for certain prompts:
14, 5, 2, 2, 2, 1, 0.4, 0.2, 0If you're still getting body horror, you can try the following quality tags (you'll need ComfyUI-ppm to apply negative weights to your positive prompt):
masterpiece quality, realistic photo, (worst quality,:-1) (mutated,:-1)Snakebite is very responsive to stylistic terms, especially by IL standards. Keep the extra "fluff" to a minimum - almost every token I've tried has a significant impact on the picture.
Finally, I suggest trying the CLIPAttentionMultiply node. If you boost the q and v parameters, it will effectively cause your image to become more "Illustrious-like": cleaner, more stable, but less realistic and (usually) less adherent to the prompt. Set both values to 3 for a very clean image.
BIGASP'S CLIP IS NOW IN PLAY!
In versions 1.3 and up, Snakebite includes a little of bigASP's CLIP, which means you can take advantage of more style prompts. Experiment with different terms to see what works. Personally, I keep it simple - this will usually improve your image without any side effects:
high quality, sharp focusWhich version is for me?
If you're wondering which version of the model to use, here's a TL;DR:
v1.4 = next-level realism, jaw-dropping textures, very stable, slightly less vibrant than previous versions and less capable of non-photographic images
v1.3 = good anatomy, good backgrounds, good coherence
v1.2 = best punchy colors
v1.1 = most influence from bigASP (excluding CLIP), dull colors, a failed experiment TBH
v1.0 = impressively creative but very unstable
If you like the model or use it for further finetuning, please let me know! I'd love to see the results. 💪
Description
Big update here! Excited to share it with you.
I was satisfied with the overall balance between IL and bigASP, so I took my first crack at aesthetic tuning for v1.4. It's capable of pretty amazing realism now. I've managed to retain nearly all of Snakebite's knowledge from IL while unearthing some incredible textures that were hidden in bigASP's latent space. Usually, you have to choose between textures and prompt adherence. I think we can finally have our cake and eat it too.
I also implemented a few LoRAs that should assist with overall image composition and reduce unwanted limbs. Finally, I tweaked the balance of the CLIP just a smidge (it's almost entirely IL, with just the smallest hint of bigASP clip to help resolve anatomy.)
In a nutshell: faces are more diverse, human anatomy is more stable, and the model's baseline aesthetic is more cinematic/realistic/texture-rich.
Is it perfect? Nah, v1.4 lost a little of the punchy vibrance from previous versions despite my best efforts. It's a double-edged thing. The vibrance makes it possible to generate super-clean images that are otherwise beyond SDXL's reach, but it also makes things look a bit "CGI." I've tried keeping as much vibrance as possible while significantly boosting the realism. Having said that, you may prefer the aesthetic of v1.3 for illustrations and digital art.
FAQ
Comments (22)
Amazing model, I have been messing around with skimmed cfg and apg makes it even better and it gets me very good results, b.t.w I forgot to include tweaking around with Dynamic Threshold CFG Fix and Latent Modifier thingy.
Great model! Worth training a lora for! Keep it coming!
Thanks, kiko! There's a v2.0 in the works... it's shaping up to be something special 🙂
@liftweights Can't wait!
Oh, the snakes are so good, they will hiss at me. Oh, I am going to generate some nude mature women with the AI model to jerk to them, the model is so good, it will make me feel so good. Oh, I will make them all look so realistic, I am going to use them for my own pleasure, I will never get tired of this. Oh, I can't wait to see the nudes, I am so excited to see the nudes. Oh, I will keep using it forever, I am so happy I found this, I will save them all for later.
uwu
2.0 is out! 🐍🐍 You can find it in my profile.
where? dont see it
@timstertimster3 :
https://civitai.com/models/2068592/snakebite-2
One of the most interesting projects i've seen in a long time. Congrats.
Thank you! I'm your biggest fan :)
@liftweights at last i have a biggest fan! it was about time!
BigAsp 2.5 does not have a trained clip. There was zero text encoder training, he kept them frozen the entire time. BigAsp 2.0 DID have a (very) trained clip, however.
Wow so many more versions have been released since I discovered 1.3 like only 10 days ago. I just haven't been on but I've been meaning to say that Snakebite is excellent!! It consistently gets anatomy as desired and with a realistic finesse. I really love it. Thank you so much for sharing it. I'm looking forward to trying the new versions!
Thank you for the kind words, I hope you like the updates!
Thanks you for this experiment, it really stands out among the sea of very similar 2.5d anime slop Illustrious/NoobAI merges.
I've found that 1.3 non-turbo works the best for inference, LCM version is a bit too inflexible with niche concepts for my taste and Snakebite 2 series really does lack booru knowledge and, I think, brings more of bigASP2 inherent instability.
So, please consider updating this model further!
Thank you for your feedback!
The v2 series is at the finetuning stage now - I'm preparing a fairly large dataset of images (~2-3k) with the goal of increasing stability. I agree with you that using IL as a base has advantages in this regard. Still, I may try throwing the same dataset at v1, just to see if we can squeeze more realism out of it.
As for the difference in booru knowledge... bigASP is an interesting beast. It understands some booru concepts very well, especially if you omit underscores. Concepts like paw pose or double v are practically on par with Illustrious, but then others like imminent penetration don't work at all. There's also a ton of stuff that's only possible with bigASP, but you have to be more verbose with your prompting.
Anyway, my goal someday is to mix Snakebite 1.x into 2.x and hopefully get the best of both. I tried this blend once before, but it was a broken mess. 😅 I'm giving 2.x more time in the oven before trying again.
This model does not work in comfyui???
very bad
every image was made in comfyui, your installation might be broken
@liftweights Thank you for your reply, but I won't use it, it's a regret
Thank you for your reply, but I won't use it, it's a regret
This model works good.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.