As this is a merge based off PonyDiffusion v6 I HIGHLY recommend you read the description of that model for information regarding prompting, tags, and baked in characters.
TL;DR Generation Settings:
Sampler - DPM++ 2M Karras / DPM++ 2s a / DPM++ 2s a Karras
Sampling Steps - 25
I upscale to 4x using this process and the upscaler below:
Dimensions - 832x1216, 1216x832, or 1024x1024
CFG Scale - 7 or higher
One or more passes of ADetailer
Some combination of the following for faces and eyes
I use the following positive and negative prompt. I DO NOT USE ANY SOURCE TAGS (source_anime, source_furry etc) :
score_9, score_8_up, score_7_up, BREAK {prompt}
NO NEGATIVE PROMPT OR EMBEDDINGS REQUIRED
simply prompt for what you don't want to see as it comes up
See sample images for examples.
Heavily stylized LoRa merge, tweaked to my personal taste. Has compatibility with most concept/pose/character models, but hands are still pretty hit or miss. Models I used are listed in suggested resources.
As with most of my mixes, this is intended to produce female characters and leans towards NSFW.
V4:
Fourth time's the charm. I wasn't happy with the outputs of V3, so I baked models of my own and remerged to get V4. This will be the last(tm) update, as I feel it's stylistically close to what I wanted when I first made the merge. Watermarks should not appear by default (I encountered none when generating the samples).
V3:
Rebaked an underlying LoRA that results in the majority of the watermarks introduced in V2 being removed; no more needing to prompt for things like watermark, patreon username, artist name, signature in the negative by default. Otherwise outputs should be very similar to V2.
V2:
Remerged from the bottom-up using modified weights. Less stylistically intense (especially around the faces and eyes), more inline with what I had originally intended for this model, but still not quite there. Reduced the need to use specific prompts to achieve the desired look. Hands are still hit or miss, and there's a persistent watermark that appears ~50% of the time: some inpainting might be required. Hoping to rectify this in a future update.
The baked-in artist styles for PD should work, but as this is effectively a style merge, they will defeat the purpose of this model.
Description
FAQ
Comments (21)
Is there a Checkpoint like that for 1.5? Looks really nice. good work.
Hey, thanks for your fantastic work! Btw I think the base model of v2 is wrongly labeled and should be Pony instead of 1.5
Ah yes, the Paintjob checkpoint... for Paintjobs... :3
What job you going to do next 🤔
v2 🔥🔥🔥
Style so good I beat my dick into a pulp and now I'm not horny and just happy
good model, very hot. if i might suggest, add more high angle pics in the data for future versions. As its stands right now i find it almost impossible to get a true high angle (from above tag) pics, but otherwise great job
Does 'mfcg' stand for 'the motha'-fuckin' checkpoint god'? If not, perhaps give it some consideration!
Excellent work :^)
First of all, I absolutely adore this checkpoint. It's phenomenal. The colors, the compositions it can handle, just everything about it is awesome.
However, I find the faces a bit too stylized and sometimes even 'too anime' compared to the rest of the render sometimes. I've tried to prompt it out, but with very little luck. Any chance of getting a lora to go on top of it or another version that maintains the overall style but has more neutral face designs like "MFCG Washed", "MFCG Squish" or "MFCG Style"?
Cheers, and much love!
can you allow on site generation pls?
tbh output of V3 seems quite different from V2, same prompt gives different poses (and less accurate to the prompt) and the lighting looks... flatter, idk maybe that's the style you wanted.
thanks for removing the signatures : ).
I consider version 3 to be a great improvement for those of us who like digital art style models, especially imitating famous artists that we all already know, thank you.
Waiting for this checkpoint to be on website generator... But from the image am going to love it already! XD
🔥🔥🔥
The checkpoint is good, with a nice painting style, but the images might be yellow cast. Could you improve this? Looking forward to the next version.
One of the most reliable and consistent Pony models, especially when it comes to hands, well done!
Maybe I'm doing something wrong, but is it the case that Pony style loras don't do well with checkpoints outside of Autism and Pony (original)? Every time I try to mix in style loras, the colors get glitchy.
v4 is definitely the most controllable, and watermarks are gone indeed. I still prefer the style of v2 more, but gud job with this one anyway.
v2 seems to give way better comic pictures !!!
I'm goona stay with that great v2 version 🙏
Can't make the upscaler work? Put it with all the rest, but doesn't show.
It was hard to accept that it doesn't work with Loras' style (I'll read the full description next time T_T)
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.