Ejecutar en local
dora.safetensors
436 MB · SafeTensor · bf16
Las descargas las sirve Civitai. Algunos creadores piden iniciar sesión en Civitai primero.
Sobre este modelo
KREA2 · LoRAs
by ibara0608 - [krea2]dora style V0.5
Notas del creador
蒸馏自该模型的风格
在krea2-raw-bf16原始权重上训练
给风格具有偏灰黑的色彩,略显粗粝的线条,混杂胶片噪点和色差的特点.
The style distilled from this model:
Trained on the original krea2-raw-bf16
The style features grayish-black tones, slightly rough lines, mixed film grain and color aberration.
写字能力,物理关系似乎未见明显衰减(参考v0.5封面中的玻璃畸变和百叶窗)
The text rendering capability and physical relationships do not appear to have degraded significantly (refer to the glass distortion and window blinds on the v0.5 cover)
for v0.5
更改标记,使用官方示例的系统词但没来得及修正
以1e-4的学习率,adamW,rank=64进行100*100次学习
目前在turbo权重上初步测试存在以下问题
1.对于模型本身认识的二次元角色,会出现画风效果减弱或者角色特征混杂的问题
即便是训练集中存在该角色
2.角色涉及化妆时容易出现唇钉,该问题大概时打标导致的
3.似乎nsfw的condition节点异常
Modified tags using system prompts from official examples without revision(not yet).
Training was conducted for 100×100 images with lr=1e-4, AdamW optimizer and rank = 64.
Preliminary tests on Turbo weights reveal the following issues:
一,For anime characters recognized natively by the model, weakened stylistic rendering or mixed character feature artifacts occur, even if the character is included in the training dataset.
二,Lip piercings tend to incorrectly appear on characters with makeup, an issue presumably stemming from labeling errors.
三.Abnormal behavior is observed in the condition node for NSFW.
fro v0.2
触发(?)句为drawn by dora,可能需要辅以digital illustration或者anime-style.
使用为anima准备的短自然语言测试(danbooru式tag似乎无论跑图还是训练都存在问题?)
The trigger phrase (?) is ''drawn by dora''; digital illustration or anime-style may be used as supplements.
Conduct short natural language tests prepared for anima (Danbooru-style tags seem to cause issues both in image generation and model training?)
Palabras de activación
drawn by dora, digital illustration,
SHA256
65BA2F3E668A143BA5C61A442FEBACBBEA777B0AC78B812EA90AC2FD3D599C76
ID de modelo de ModelTrends.ai
#26594
Denunciar contenido
Renders de ejemplo
Prompts y ajustes compartidos por quienes hicieron estos renders en Civitai. Elige un render para ver cómo se hizo.
Ajustes
Sampler
Euler
Pasos
8
Guidance
1
Semilla
226151272368241
Tamaño
832x1216
Peso usado
0.3
Prompt
drawn by dora, digital illustration, zidiusArt,杰作,最佳画质,超精细,8K 分辨率,日系清新插画,电影级光影氛围,
雨后天晴的金发少女,留着被风吹得大幅扬起的金色长发,佩戴白色耳罩,身穿橙红色长款大衣,搭配黄色围巾与白蓝格纹长裙,
她手持一把透明雨伞,伞面布满雨滴,背景是雨后的天空,一道彩虹横跨天际,天空中漂浮着发光的纸鹤,
脚下是开满黄色郁金香的草地,雨丝仍在飘落,整体色调以暖橙、金黄、彩虹的多彩与雨后的蓝调为主,充满清新治愈的氛围,
细腻的发丝动态、雨滴透明质感、雨伞反光、彩虹渐变、郁金香花瓣细节与光影层次,
竖版中景构图,细节拉满的动态张力、色彩对比与清新治愈氛围。
Ajustes
Sampler
er_sde simple
Pasos
8
Guidance
1
Semilla
284064178687979
Tamaño
1072x1384
Prompt
drawn by dora, anime-style character, reisen udongein inaba, 1girl, alternate costume, dynamic psoe, contrapposto, animal ears. A girl with long, wavy lavender hair and striking purple eyes wears a dark blazer over a torn white shirt that reveals her midriff. She has on a short, pleated lilac skirt with rips exposing her thighs, along with black thigh-highs featuring lace trim at the top. A red necktie is loosely draped around her neck, and she sports rabbit ears as part of her outfit. In an aggressive pose, she wields two submachine guns cross her arms in front of her chest and point to both sides of the frame, and slightly tilted suggesting intense action or combat. The background consists of a large off-white circle surrounded by gray with scattered dark particles resembling debris. The image has a painterly style with soft gradients in the hair and clothing, creating a dreamy yet violent atmosphere. Texts include 'Type Lunar' at the bottom left and 'Yonder Voice' at the top right.
<lora:dora:1.2>
Ajustes
Sampler
er_sde simple
Pasos
8
Guidance
1
Semilla
1030575734606019
Tamaño
920x1616
Prompt
The image is drawn by dora. anime-style character. This is a high-resolution indoor scene captured at a strong dutch angle that tilts the entire frame diagonally, adding dynamic tension and visual interest to the composition. The lighting is dramatic and moody, with warm sunlight streaming in from an unseen window casting sharp shadows on the tatami floor and wooden furniture around her. Hoshino sits casually on a low bench or shelf near a bookshelf filled with books and small decorative objects such as a clock and stuffed toys. She holds up an old-fashioned silver camera directly facing the viewer, partially obscuring her face but allowing one intense red eye to peer over the lens. The room has a cozy, vintage atmosphere enhanced by scattered flowers in a basket on the shelf behind her and various personal items like books and a glass bowl with water on a nearby wooden table.
<lora:dora:1.2>
Ajustes
Sampler
er_sde simple
Pasos
8
Guidance
1
Semilla
56593815805959
Tamaño
1616x920
Prompt
drawn by dora, anime-style character, from above. This ethereal, dreamlike illustration unfolds in a sun-dappled, water-filled realm that blends classical East Asian architecture with a serene lotus pond.
At the heart of the scene, a young girl with long, silvery-white hair sits quietly in a narrow, dark wooden boat. She wears a flowing, pale white dress with delicate patterns, and her posture is calm and meditative. Inside the boat, beside her, rest scrolls, a small porcelain tea set, and a few white lotus blooms, enhancing the tranquil, scholarly atmosphere.
The boat drifts through still, translucent water dotted with lily pads and blooming white lotus flowers, where tiny koi fish glide beneath the surface. The water surrounds ornate, white stone structures—carved pillars, steps, and pedestals—half-submerged, as if the ruins of a forgotten temple have been overtaken by the pond. Vertical columns of calligraphic text floating through the air like floating incantation, weaving through the light and adding a poetic, timeless quality to the space.
Soft, diffused sunlight filters down, casting gentle shadows and highlighting the misty, pale green tones of the water and foliage. The overall mood is one of profound peace and quiet contemplation, blending elements of traditional Chinese aesthetics with a dreamy, otherworldly calm.
<lora:dora:1.2>
Ajustes
Sampler
er_sde simple
Pasos
8
Guidance
1
Semilla
602245087848431
Tamaño
1000x1456
Prompt
drawn by dora, anime-style character, one girl. This is a high-resolution, full-body image of Cecilia sitting on a toilet in a grimy, graffiti-covered bathroom. The camera angle is slightly above and to the side, capturing her curled-up pose and intense expression. The setting is a tiled bathroom wall with various graffiti tags, including an "XOXO" tic-tac-toe game and other scribbles in black and pink spray paint. The lighting is moody with purple and blue hues, casting shadows and highlighting the gritty atmosphere. The image has a somewhat rough texture and a dark, edgy vibe.
<lora:dora:1.2>
Ajustes
Sampler
euler_simple
Pasos
8
Guidance
1
Semilla
429742780208179
Tamaño
4000x2238
Prompt
, The image is drawn by dora. anime-style character. This is a high-resolution indoor scene captured at a strong dutch angle that tilts the entire frame diagonally, adding dynamic tension and visual interest to the composition. The lighting is dramatic and moody, with warm sunlight streaming in from an unseen window casting sharp shadows on the tatami floor and wooden furniture around her. Hoshino sits casually on a low bench or shelf near a bookshelf filled with books and small decorative objects such as a clock and stuffed toys. She holds up an old-fashioned silver camera directly facing the viewer, partially obscuring her face but allowing one intense red eye to peer over the lens. The room has a cozy, vintage atmosphere enhanced by scattered flowers in a basket on the shelf behind her and various personal items like books and a glass bowl with water on a nearby wooden table.
<lora:dora:1.2>,
Prompt negativo
,
Ajustes
Tamaño
1216x2103
Prompt
A soft, surreal anime-style illustration from a first-person perspective shows a pair of oversized pale hands gently cupped together, holding a small girl in their palms. The hands are solid and softly colored, with a light whitish skin tone, subtle pink warmth, and faint cool blue shading along the contours. A small glimpse of white sleeve cuffs is visible at the wrists. The girl is small enough to lie completely within the space of the two hands, with her whole body resting securely inside their range.
The girl lies on her side in a curled, sleeping-like pose. Her knees are drawn up toward her body, her torso folds inward slightly, and both arms are bent close to her chest. Her head tilts gently to one side, and her eyes are closed, giving her a quiet, fragile, and withdrawn presence. She has long pale blonde hair flowing around her in ribbon-like twin tails, and she wears a delicate floral headdress decorated with roses, lace, and feather-like ornaments. Her outfit is an elaborate green, white, and black frilled dress with layered ruffles, floral decorations, and ribbon details, preserving a refined doll-like elegance. She wears white thigh-high stockings, and her feet are covered only by the white stockings.
The composition is close and intimate, with the hands filling most of the frame and the girl placed securely at the center. Thin red threads stretch across the image and loop around several fingers. One red thread is tied to the little finger of the girl lying in the hands, and it connects continuously with the other red threads, giving the image the feeling of a red string of fate. The overall color palette is soft, pale, and low-saturation, dominated by white, light gray, faint pink skin tones, and cool pale blue shadows. These muted tones are gently contrasted by the green and black accents of the girl’s dress, her pale golden hair, and the vivid red threads, creating a dreamlike, delicate, and slightly unsettling atmosphere.
Ajustes
Sampler
res_2s_beta
Pasos
10
Guidance
1
Semilla
252301024937911
Tamaño
1344x752
Prompt
, Subjects:
Lena: a woman in her mid-20s with a slim build, fair East-Asian skin, long straight black hair falling down her back, dark brown eyes; she wears an oversized black t-shirt, blue athletic shorts with a thin yellow side stripe, and colorful over-ear headphones with pink and teal accents resting snugly over her hair
Setting: a compact, lived-in home utility room with warm beige walls and tiled floor; a single white top-loading washing machine stands in the right foreground with its lid propped fully open; behind Lena a sturdy wooden shelf unit holds glass jars of dry goods, cloth sacks of rice and flour, stacked tins, a few detergent bottles tucked among the provisions, and folded cloths; a simple doorway at the far end lets in soft daylight
Positioning: Lena stands centered in the small room, body angled toward the open washing machine that fills the lower right of the frame, her torso leaning naturally forward over it
Action: Lena lowers a soft bundle of gray and pink clothes into the open drum with both hands, eyes gently closed and a small smile on her lips as she sings along to the music in her headphones, completely absorbed in the quiet domestic moment
Lighting and mood: warm daylight from the doorway mixes with a soft amber ceiling glow, casting gentle highlights across the jars on the shelf and her hair, peaceful everyday atmosphere of a slightly cluttered but cozy utility space,
Prompt negativo
, shoes, partially undressed, undressing, nude, explicit, unsafe,
Modelos similares
Modelos de IA para generar imagen, vídeo y texto
dora style krea2 es un LoRA para la familia Krea2 listado en ModelTrends.ai, un catálogo de solo lectura de modelos de IA de imagen, vídeo y texto de código abierto. Compáralo con otros modelos Krea2, consulta su puntuación de calor para ver su tendencia y ábrelo en PirateDiffusion o BitVector para probarlo.
Explorar por familia de modelos
Modelos Text / LLM
411
Modelos Anima
509
Modelos Chroma
14
Modelos Flux
1,071
Modelos Flux 2 / Klein
1,237
Modelos MiniMax H3
331
Modelos Hunyuan
197
Modelos Ideogram
20
Modelos Krea2
674
Modelos Qwen
17
Modelos Qwen2
36
Modelos Ltx2
150
Modelos Stable Diffusion 1.5
6,465
Modelos Stable Diffusion XL 1.0
15,432
Modelos Zimage
804
Modelos Wan
919

Model
Trends
.ai
© ModelTrends.ai
|
Hecho en Japón
|
© 2026
