VinteProtogenMix v2.0
this model makes a colorful pictures. crisp sharp colors, character also looks very nice.
was merged
22h vintedois-diffusion-v0-2 + 1(protogenV22AnimeOffi_22 - v1-5-pruned-emaonly) Add difference
+ 22h/vintedois-diffusion-v0-2
+ Protogen v2.2 (Anime) Official Release
VAE not required.
VinteProtogenMix v1.0
this model makes a colorful pictures. crisp sharp colors, character also looks very nice.
was merged
22h vintedois-diffusion-v0-1 + 1(protogenV22AnimeOffi_22 - v1-5-pruned-emaonly)
+ Protogen v2.2 (Anime) Official Release
VAE not required.
Description
this model makes a colorful pictures. crisp sharp colors, character also looks very nice.
was merged
22h vintedois-diffusion-v0-1 + 1(protogenV22AnimeOffi_22 - v1-5-pruned-emaonly)
+ Protogen v2.2 (Anime) Official Release
VAE not required.
FAQ
Comments (31)
There are very few such models on this platform! Thank you! And then some solid anime and women all the time!
that's why i want to bring variety to the civitai :-D
@xxlmaes More paintings by artists to train a new model. You are the best! Thank you for your hard work!
@xxlmaes As soon as I have time, I will post the resulting paintings thanks to your model!
@xxlmaes Can I contact you on Discord? To help create a new model? My discord - JomaanGa#6268
@JimPoison thank you very much, i'm glad to hear it. I'm very curious how it will turn out
One giant leap for MJ substitute.
thanks :-D I was also very surprised at how good it turned out
Can you add pickle tensor file like ckpt ? thx
for some reason the prompts from the civitai doesn't load properly. the prompts are available in PNG
this is my favorite model without a doubt, but most of the time the images look like they have been poorly upscaled/downscaled because the details are there but they are blurry.
would a specific word fix this?
try to use these negatives prompt: ((((manga)))), (((unrealistic forms))), amateur, unprofessional, ((((incoherent)))), (((undefined eyes))), (((unattractive))), (((mutation))), (((deformed))), ((ugly)), ((((mutated hands and fingers)))), (((blur))), ((dull)), (((photographic))), (((untextured))), (((game logo))), ((poorly drawn hands)), ((poorly drawn face)), painting, drawing, ((((visible hand)))), ((((ugly)))), (((duplicate))), ((morbid)), ((mutilated)), [out of frame], extra fingers, mutated hands, ((poorly drawn hands)), ((poorly drawn face)), (((mutation))), (((deformed))), ((ugly)), blurry, ((bad anatomy)), (((bad proportions))), ((extra limbs)), cloned face, (((disfigured))), out of frame, ugly, extra limbs, (bad anatomy), gross proportions, (malformed limbs), ((missing arms)), ((missing legs)), (((extra arms))), (((extra legs))), mutated hands, (fused fingers), (too many fingers), (((long neck))), (bad quality, worst quality, normal quality),
most people use SwinIR 4x for scaling
Boss please try to marge with Illuminati Diffusion v1.1
Can we get a LoRA out of this model? I love this model, but merging it with my trained one is not giving the desired output.
I tried to create the LoRA from the Extractor with different networks ( 2,4,8,16,32, 64, 128, 256 and 320 - max is 320) The output is about 60% similar when using the 256 or 320 networks LoRA of this model.
Model merge (add Difference) is also not getting me the correct results.
Of course, I know that the exact result is not possible with merge and LoRA, but if there is a way to achieve near to exact results, that would be Bliss.
try to experiment. I'm looking forward to the results :-D
start playing the values to get best results :1.0 / 0.70 / 0.50 / 0.30
@xxlmaes requested the access :).
@deardhruv ok done
@deardhruv file has been made public
@xxlmaes Yep received it, I am making X/Y graphs for the given LoRAs, I'll also make an X/Y graph for extracted LoRAs from this model and mail back to you the results. :) hopefully, this will be done by tomorrow, I only got a 6GB card ;).
@xxlmaes I have shared the email with you with my findings :)
@xxlmaes can you upload to Civitai? I'd like to use this on stable horde, which only supports Civitai loras (crowd-sourced SD for people like me that don't have GPUs). Any Lora of this that's even somewhat close would be useful for increasing denoising on img2img, since it's not a popular model (people don't host it at high resolutions)
thanks for share this model.. sir please update this model with 22h/vintedois-diffusion-v0-2
will be done thanks for info :-D
there will be an update soon!
expect success
can i sell photos using this model? is any problem like copyright issue?