Run Locally
DysisDim10AdamWAnima3pv0.99tx1-000008.safetensors
40 MB · SafeTensor
Downloads are served by Civitai. Some creators ask for a Civitai login first.
About This Model
ANIMA · LoRAs
by happyhen - Dysis — Evan's Remains [Anima P3] V0.99 DIM 10
Creator notes
Dysis from game
All images have a prompt you can look at and choose what you are interested in!
v0.99 AnimaP3 DIM 10
Trigger words: d1s2s
Recommended weight: 0.8-1.1 (1)
Recommend words:
For default girl: d1s2s, 1girl, pixel art, glistening eyes, yellow eyes, brown hair, long hair, sidelocks
Natura language: A pixel art of a female d1s2s has glistening yellow eyes and long brown hair with sidelocks
For full girl attributes: d1s2s, 1girl, pixel art, glistening eyes, yellow eyes, brown hair, long hair, sidelocks, sun hat, white headwear, hat ribbon, orange ribbon, sundress, orange dress,
Natura language: A pixel art of a female d1s2s has glistening yellow eyes and long brown hair with sidelocks, on her head white sun hat with orange ribbon, she is clothing in orange sundress
You might also try:
skinny, portrait
General possible problems/Non-obvious points:
Anima issue: pixel art is a necessary tag, otherwise the character becomes very weakly recognizable. Her sun hat is quite heavily trained, so for example, a fedora can sometimes take on a strange shape or be larger than it should be. The character's background can be darkened for no apparent reason; to avoid this, use a dark background in negative.
v0.97 AnimaP3 DIM 8
Recommended weight: 0.8-1.1 (0.9)
Also using glistening eyes may make less sense.
The differences between v0.97 and v0.99:
On a v0.99 minor caption bugs were removed, the sparkling eyes tag was replaced with glistening eyes
the training settings were changed, and dim was changed from 8 to 10, while epoch was changed from 10 to 8.
I decided to publish both versions because I still couldn't choose the best one. Despite its smaller size, I think v0.97 better save and preserves the character's style. However, the eye lighting control is more random.
I'm still exploring training options for the Anima. Natural language is recommended, but when trying to recreate the style, I find that tags alone work better.
I trained using the full image description in the tag version, as well as natural language immediately after the tags.
Trained on Anima preview 3 base
Text like above, but in Russian:
Возможные проблемы/Неочевидные моменты:
Особенности Anima: тег pixel art необходим, иначе персонаж становится очень слабо узнаваемым. sun hat dark background в негативе.
Различия между версиями v0.97 и v0.99:
В v0.99 были исправлены незначительные ошибки в подписях, тег sparkling eyes был заменён на glistening eyes
Изменены настройки тренировки, значение dim было изменено с 8 на 10, эпоха с 10 на 8.
v0.97
Anima.
Trigger words
d1s2s
1girl, pixel art, glistening eyes, yellow eyes, brown hair, long hair, sidelocks
A pixel art of a female d1s2s has glistening yellow eyes and long brown hair with sidelocks
1girl, pixel art, glistening eyes, yellow eyes, brown hair, long hair, sidelocks, sun hat, white headwear, hat ribbon, orange ribbon, sundress, orange dress
A pixel art of a female d1s2s has glistening yellow eyes and long brown hair with sidelocks, on her head white sun hat with orange ribbon, she is clothing in orange sundress
SHA256
0E3E9E487EC26E35FFC4A94634537817280AEC4A050AF0AD340AE8F3B6B462BE
ModelTrends.ai Model ID
#23692
Flag content
Example renders
Prompts and settings shared by the people who made these renders on Civitai. Pick a render to see how it was made.
Settings
Sampler
Euler a
Steps
26
Guidance
4.5
Seed
593931522
Size
1024x1024
Clip skip
2
Prompt
score_9, score_8, score_7, d1s2s, <lora:DysisDim10AdamWAnima3pv0.99tx1-000008:0.9>, 1girl, yellow eyes, brown hair, long hair, sidelocks, cowboy shot, solo, pixel art
Negative prompt
worst quality, low quality, score_1, score_2, score_3, blurry, watermark, jpeg artifacts, censored, signature, adversarial noise, logo, speech bubble, artist name, close-up
Settings
Sampler
Euler a
Steps
30
Guidance
5
Seed
2439485495
Size
1024x1024
Clip skip
2
Prompt
masterpiece, best quality, score_9, score_8, score_7, d1s2s, <lora:DysisDim10AdamWAnima3pv0.99tx1-000008:1>, 1girl, detective, yellow eyes, bags under eyes, thinking, looking to the viewer, brown hair, long hair, swept bangs, sidelocks, holding cigarette, cigarette, orange necktie, red fedora, white trench coat, open coat, white shirt, blood on clothes, standing, solo, rain, night, blurry background, dramatic lighting, mix of a red-blue theme, A digital illustration pixel art shoot of a female detective, the frame is a close-up of her upper body as she stands in the rain with a cigarette in her mouth, her yellow eyes and tired expression are emphasized by the warm ember light, blood stains cover parts of her coat, and behind her the red and blue police lights melt into a soft blurred background of color and motion, stylized oil paint illustration blending sharp pixel linework with watercolor blue and red splatter textures at the background, the textures feel pixel, with painterly distress, combining cel-shaded forms with painterly washes, dynamic composition with abstract police flashlight-driven brush effects at the background modern concept-art, detailed pixel art cover aesthetic
Negative prompt
worst quality, low quality, score_1, score_2, score_3, blurry, watermark, jpeg artifacts, censored, signature, adversarial noise, logo, speech bubble, artist name
Settings
Sampler
Euler a
Steps
30
Guidance
5
Seed
3111752084
Size
1024x1024
Clip skip
2
Prompt
masterpiece, best quality, score_9, score_8, score_7, d1s2s, <lora:DysisDim10AdamWAnima3pv0.99tx1-000008:0.9>, 1girl, detective, yellow eyes, bags under eyes, thinking, looking to the side, brown hair, long hair, swept bangs, sidelocks, holding cigarette, cigarette, orange necktie, blue fedora, red pants, white trench coat, open cloat, white shirt, blood on clothes, leaning on wall, against wall, looking at viewer, dramatic lighting, blurry background, red-blue lightning, from below, A pixel art of a female d1s2s, she is a detecitve who has yellow eyes and long swept bangs brown hair with sidelocks, she is smoking a cigarette in her mouth, which illuminates soft lighting of her face and emphasizes the bags under her eyes and her thoughtful expression, she is leaning against the wall on the left, dressed in a white trench open coat with an orange necktie and white shirt, on her head blue fedora also in red detective pants, there are traces of bright blood on the coat, In the background a bright blurry blue-red theme bright colors, emphasize the dramatic and pathos of the scene and the girl, angle is a from below
Negative prompt
worst quality, low quality, score_1, score_2, score_3, blurry, watermark, jpeg artifacts, censored, signature, adversarial noise, logo, speech bubble, bad reflection, artist name, day
Settings
Sampler
Euler a
Steps
30
Guidance
5
Seed
1825401496
Size
1088x1472
Clip skip
2
Prompt
masterpiece, best quality, score_9, score_8, score_7, d1s2s, <lora:DysisDim10AdamWAnima3pv0.99tx1-000008:0.9>, 1girl, pixel art, brown hair, long hair, sidelocks, sun hat, white headwear, glistening eyes, yellow eyes, closed mouth, hat ribbon, orange ribbon, standing, sundress, orange dress, hair ornament, looking at viewer, smile, garden, flowers, sunlight, solo
Negative prompt
worst quality, low quality, score_1, score_2, score_3, blurry, watermark, jpeg artifacts, censored, signature, adversarial noise, logo, speech bubble, artist name
Settings
Sampler
Euler a
Steps
30
Guidance
5
Seed
635089713
Size
1088x1472
Clip skip
2
Prompt
masterpiece, best quality, score_9, score_8, score_7, d1s2s, <lora:DysisDim10AdamWAnima3pv0.99tx1-000008:0.9>, 1girl, pixel art, brown hair, long hair, sidelocks, sun hat, white headwear, yellow eyes, closed mouth, hat ribbon, orange ribbon, standing, sundress, orange dress, torn veil, torn sleeves, torn dress, battle scars, confident smirk, dynamic action pose, ready for battle, background explosion, light streaks, solo
Negative prompt
worst quality, low quality, score_1, score_2, score_3, blurry, watermark, jpeg artifacts, censored, signature, adversarial noise, logo, speech bubble, artist name
Settings
Sampler
Euler a
Steps
30
Guidance
5
Seed
2775918054
Size
1024x1024
Clip skip
2
Prompt
masterpiece, best quality, score_9, score_8, score_7, d1s2s, <lora:DysisDim10AdamWAnima3pv0.99tx1-000008:0.9>, 1girl, pixel art, winter clothes, turtleneck sweater, coat, scarf, earmuffs, mittens, brown hair, short hair, beret, yellow eyes, closed mouth, standing, holding warm drink, gentle smile, outdoors, snow, snowing, town, solo
Negative prompt
worst quality, low quality, score_1, score_2, score_3, blurry, watermark, jpeg artifacts, censored, signature, adversarial noise, logo, speech bubble, artist name
Settings
Sampler
Euler a
Steps
30
Guidance
5
Seed
1688272595
Size
1088x1472
Clip skip
2
Prompt
masterpiece, best quality, score_9, score_8, score_7, d1s2s, <lora:DysisDim10AdamWAnima3pv0.99tx1-000008:0.9>, 1girl, pixel art, detective, yellow eyes, bags under eyes, looking to the viewer, brown hair, long hair, swept bangs, sidelocks, holding cigarette, cigarette, orange necktie, fedora, white trench coat, open coat, white shirt, blood on clothes, sitting, on ground, knees up, from above, looking up, looking at viewer, solo, rain, night, revolver, dramatic lighting, mix of a red-blue light, A digital illustration pixel art of a female detective, the frame is a full body shot deep from above angle, she is sitting on her ass on ground her knees up, she is leaning on hand in the rain with a cigarette in her mouth, her yellow eyes and tired expression are emphasized by the warm ember light of cigarette, she is looking up on a viewer, blood stains cover parts of her coat, there is a revolver on the ground, the red and blue light from outside create a dramatic theme, foreground is a blending sharp pixel linework, background is stylized oil paint watercolor blue and red splatter sharp light, combining cel-shaded forms with painterly washes, a abstract red and blue brush effects, mood of scene pathos with dramatic
Negative prompt
worst quality, low quality, score_1, score_2, score_3, blurry, watermark, jpeg artifacts, censored, signature, adversarial noise, logo, speech bubble, artist name
Settings
Sampler
Euler a
Steps
30
Guidance
5
Seed
3656256714
Size
1024x1024
Clip skip
2
Prompt
masterpiece, best quality, score_9, score_8, score_7, d1s2s, <lora:DysisDim10AdamWAnima3pv0.99tx1-000008:0.9>, 1girl, detective, murder, yellow eyes, bags under eyes, looking to the viewer, brown hair, long hair, swept bangs, sidelocks, holding cigarette, cigarette, orange necktie, fedora, white trench coat, open coat, white shirt, blood on clothes, sitting, on ground, knees up, from above, looking up, looking at viewer, solo, rain, night, revolver, dramatic lighting, mix of a red-blue light, A digital illustration pixel art of a female detective, the frame is a full body shot deep from above angle, she is sitting on her ass on ground her knees up, she is leaning on hand in the rain with a cigarette in her mouth, her yellow eyes and tired murder expression, she is looking up on a viewer, blood stains cover parts of her coat, there is a revolver on the ground, the red and blue light from outside create a dramatic theme, foreground is a blending sharp pixel linework, background is stylized oil paint watercolor blue and red splatter sharp light, combining cel-shaded forms with painterly washes, a abstract red and blue brush effects, mood of scene pathos with dramatic
Negative prompt
worst quality, low quality, score_1, score_2, score_3, blurry, watermark, jpeg artifacts, censored, signature, adversarial noise, logo, speech bubble, artist name, lanterns, lamp post, street lamp
AI Models for Image, Video & Text Generation
dysisv99 anima is a LoRA for the Anima family listed on ModelTrends.ai, a read-only catalog of open source AI image, video and text models. Compare it with other Anima models, check its heat score to see how it is trending, and open it on PirateDiffusion or BitVector to try it.
Browse by Model Family
Text / LLM Models
411
Anima Models
509
Chroma Models
14
Flux Models
1,071
Flux 2 / Klein Models
1,237
MiniMax H3 Models
331
Hunyuan Models
197
Ideogram Models
20
Krea2 Models
674
Qwen Models
17
Qwen2 Models
36
Ltx2 Models
150
Stable Diffusion 1.5 Models
6,465
Stable Diffusion XL 1.0 Models
15,432
Zimage Models
804
Wan Models
919
Model
Trends
.ai
© ModelTrends.ai
|
Made in Japan
|
© 2026