Type | |
Stats | 847 6,770 |
Reviews | (157) |
Published | May 19, 2024 |
Base Model | |
Training | Steps: 8,500 Epochs: 20 |
Usage Tips | Clip Skip: 2 Strength: 1 |
Hash | AutoV2 56BFD1EA86 |
,__mature female, dark-skinned female, solo,.jpeg)
Ky. Style is on PonyXL and SD 1.5 animefull now!
PonyDiffusion v6 (XL) main settings:
Style settings for PonyXL (very important):
important tags: score_9, score_8_up, score_7_up, score_6_up, uncensored, BREAK
Put "3D" or "realistic" in negative if your character has high 3D/realistic influence which harm anime style.
Recommended negative:
censored, head out of frame, watermark, jpeg artifacts, muscular
Also add in negative annoying objects.
Best tags order for PonyXL:
1) score
2) Sentence (for example, "A girl is standing with crossed arms and smiles")
3) tags for concepts, camera, face, body, clothes, background, etc. Order is the same as for 1.5.
Preview images can help.
Style doesn't need activation tag but you can use these tags for characters baked-in LoRA:
Ingrid (Taimanin):
ingrid \(taimanin murasaki\), dark-skinned female, 1girl, mature female, pink hair, very long hair, yellow eyes, thick eyebrows, (mole under mouth:0.7), thick lips, purple lips
black blazer, pencil skirt,
Minamoto(Fate):
Minamoto no raikou \(fate\), 1girl, mature female, seductive smile, half-closed eyes, purple hair, long hair, hair between eyes, sidelocks, makeup, purple eyes, gloves, fingerless gloves, purple bodysuit, rope, tabard,
optional: eyepatch bikini, purple bikini
Igawa Sakura (Taimanin):
Igawa Sakura, 1girl, flipped hair, hair between eyes, blue eyes, blonde hair, fishnets, taimanin suit, brown highleg leotard, showgirl skirt, elbow gloves, covered navel, thigh boots,
By the way you can use style LoRA with my character LoRAs (preview made with them and without them. As you see they work both).
SD 1.5 animefull settings:
0.9 - 1 strength of Lora is the best radius (If you combine with character LoRA I recommend to decrease weight of character LoRA)
Clip skip 2
Some necessary tokens for better experience:
KyStyle, 1girl
Important tips for SD1.5:
There were too many arts with Minamoto Raikou in dataset. So if you feel that LoRA tries to generate her and you don't want it put "minamoto no raikou \(fate\)," in negative prompt.
Put in negative prompt "empty eyes" if SD generates eyes without irises.
Most of the artist's pictures were censored. So if you do NSFW scenes it is good idea to put "censored, bar censor, blank censor" in negative prompt.
Use vanilla anime models without high realistic vibes for saving artist's style. I recommend you anything v5 and divineanime. Also I suppose it will be work well on bloodorangemix.
Main advices for all LoRAs:
The best settings for Hires fix and adetailer:
Hires.fix: Upscaler foolhardy_remacri (download and put in "ESRGAN" folder in "models" if you are using automatic1111 or forge).
Hires steps: 16-18
Denoising strength
Adetailer: denoising strength 0.3. Detection model confident threshhold: 0.7-0.8.
Use Photopea in automatic1111 / forge tab for fixing small problems like fingers (there is extension for web Photopea)
I will be glad to receive feedback from you and a demonstration of your art with my LoRA!