Hours of meticulous tinkering with weights and various models led to its creation.
Multiple checkpoints were utilized to refine and optimize it into its final form.
Versatile in nature, this model is capable of adapting to diverse tasks.
With dynamic capabilities, it exhibits flexibility in handling various scenarios.
Equipped with extensive training, it can tackle a wide range of tasks efficiently.
In summary, the Am i Perfection combines rigorous experimentation and fine-tuning to achieve its versatile, dynamic, and highly capable nature.
sincere gratitude to the dedicated individuals who contributed to the creation of this remarkable model
from the start of me posting models, it all started with 2 custom dreambooth creations , from then its been merges , so i guess somewhere buried in the code yes my OWN images are in there too , i have merged i think nearly 40 models to get here , all props to the other contributors , without them , most models on here wouldn't exist!
using clip skip 2 and 4x_nickelbackFS_72000_g upscaler , DPM++ 2M SDE KARRAS i get these images
Negative Embeddings used:-
VeryBadImageNegative
https://civarchive.com/api/download/models/25820
Bad artist
https://huggingface.co/nick-x-hacker/bad-artist/tree/main
Bad hands
https://huggingface.co/yesyeahvh/bad-hands-5/tree/main
Bad Prompt
https://huggingface.co/datasets/Nerfgun3/bad_prompt/tree/main
cyberrealistic negative
https://civarchive.com/api/download/models/82745
easy negative
https://huggingface.co/datasets/gsdf/EasyNegative/tree/main
Image Sharpener
https://civarchive.com/api/download/models/47921
Ng-Deep negative
https://huggingface.co/Neburozakusu/civitai_deposit/blob/main/NG_DeepNegative_V1_75T.pt
Rmada
Description
tinkered some more and came up with this version
FAQ
Comments (10)
I've just tested it for a few hours and I think this model is gonna be my new favourite.
Eyes are reeeaally a lot better in this model than anything else I've tested (that includes the 5 most downloaded base models in civitai).
This is gonna be one of the greatest base models in this website.
Very well done, sir. Great work. Congrats.
I can't reproduce any of the example images. It doesn't help that there's a million random negative prompt loras/embeddings that i'd have to track down.
so dont use them , all the negative prompts are easily found , i use those for every model i use but are not needed , some of the lora mentioned like lowra i dont even have , the sample prompts used have been colated over many months , some have loras i use , some dont ( i just havent removed them)! update , the model description has the links for all the negative embeddings used
@GG_System so your sample prompts have stuff in that doesn't even get used (maybe?) and you know it? No wonder the images couldn't be reproduced.
@outOfTheMachine , adding the word lowra without having the actual lora makes no difference , nothing would happen so it would not impact the generated picture , i linked all the embeddings , did you try and use them in the negative prompt??
i use eta noise seed delta , you can recreate them
Tried to use the generational data from one of the images by GG_System, and it had an error... couldn't find a lora with the name xEpi. Where can I find this lora?
Hi techken , good question as I am missing that one too , the prompts were collated from other image prompts on here , I just forgot to remove the ones I don't have. I'll do a proper in-depth dive on finding it
@GG_System Kk, I checked on github and on civ.itai both, and no joy on them, so not sure if it's a dead horse, or what. For the moment, I simply pulled the link to it out of the prompt. Seemed to work ok still...
Just fantastic!
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.