Join The Tinkerer on Whop to get early releases, private pages, and the Tinkerer Discord role - all in one place. 👉 Join on Whop
CyberRealistic Classic is a more unrestricted version of CyberRealistic that offers the same freedom and flexibility. It serves as an experimental platform for testing new features and concepts that haven't been integrated into the original CyberRealistic. Essentially, it functions as a fully operational alternative edition of CyberRealistic.
Recommended Settings
Sampling method: DPM++ SDE Karras / DPM++ 2M Karras
VAE: is already Baked In
Sampling steps: 30 Steps
Resolution: 512x768
CFG: 5-7
Upscale: 2x
Upscaler: 4x_NickelbackFS_72000_G
Denoising strength: 0.3-0.35Example Negative prompts
watermark, text, censored, deformed, bad anatomy, disfigured, poorly drawn face, mutated, extra limb, ugly, poorly drawn hands, missing limb, floating limbs, disconnected limbs, disconnected head, malformed hands, long neck, mutated hands and fingers, bad hands, missing fingers, cropped, worst quality, low quality, mutation, poorly drawn, huge calf, fused hand, missing hand, disappearing arms, disappearing thigh, disappearing calf, disappearing legs, fused fingers, abnormal eye proportion, abnormal hands, abnormal legs, abnormal feet, abnormal fingersless tokens:
lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry
Need the perfect prompt? This custom ChatGPT instantly generates high-quality prompts or helps you refine your own.
🔗 [Try it now on ChatGPT]
Description
Finetuned model
FAQ
Comments (13)
One of the things I love about this model is how easy it is to get predictable, quality results with consistency. Great work!
Thank you! Really appreciate it.
Interesting 3.2 update, the changelog only says "finetuned model". I get completely different results from every single prompt compared to 3.11, though.
Overall image quality is worse for me with DPM++ 2M Karras compared to 3.11, and i get many more errors with hands, genitals and such. Characters tend to have an almost comical square head in some prompt comparisons.
The model seems extremely sensitive to style prompts now instead. Example "fantasy artwork, concept art" used to be a mix of photo and painting, now it goes full painting mode. Same with the recognized SD artists, like "style of Gustave Courbet" etc.
Well, so far the Classic 3.10/11 versions of this model have been my ultimate favorite for SD1.5.
Tried many others in the meantime, always came back to this one. Can't really understand how this is not at least in the all-time top 3 of trained 1.5 models.
All the good things I've written about CyberRealistic are fully applicable to Classic. In general, these models are very similar and the results of generation on the same seed are often very close to each other. And it's really hard for me to say what the main difference between these models is, except that it seemed to me that the Classic is quite a bit more responsive (however, maybe it was just my imagination).
However, both models can be used equally. Thanks for the work, Cyberdelia!
THANK YOU! does 3.2 works with inpainting? if so it's the f16 one right?
What is the difference between fp16 and fp32? Does it make a noticeable difference in image quality? Which one do you recommend for an 8gb video card? Thank you.
Fp16 is what you should use. Fp32 is ideal for merging etc. I always use fp16
fp16 will be faster on the 8GB card.
Can you post the 3.2 inpainting model? I don’t have the capability to merge to create my own. Also, where did all of the previous version inpainting models go?
agreed. @Cyberdelia please
good
Does this model have a VAE baked in or should I use any standalone one?
Baked in but you could use your own