Model
Trends
.ai
Model
Trends
.ai
los mejores modelos de IA de código abierto
Ejecutar en local
tangbohu-detailmaker_v2.5.safetensors
144 MB · SafeTensor
Las descargas las sirve Civitai. Algunos creadores piden iniciar sesión en Civitai primero.
AD
<detail-make>
Sobre este modelo
STABLE DIFFUSION 1.5 · LoRAs
Creado por
TangBohu
en Civitai
LoRA Name: Tangbohu Detail Maker LoRA (唐伯虎细节制造LoRA) Lora is used for adding details. Lora is updated to ver 2.5. Weight range is from -6 to 3, applies for anime, semi-real, real bases. Positive weights increases detail of image, allowing making details out of air. Negative weights can be used to reduce details. 2D or 2.5D models will recreate 3D realistic effects when using negative weights. Experiment with 0.5 steps still you get the rough amount of detail you intend, then change by 0.1 steps.
Notas del creador
A Lora mainly used for adding details. New version 2.5 description Positive weights can still increase the details of the image, making it particularly suitable for creating things out of thin air. Negative weights can reduce details. Currently, an interesting phenomenon has emerged that some models with 2D or 2.5d effects are gradually restored to 3D realistic effects after using negative weights. Perhaps this interesting effect can be applied to certain projects with special needs, so this version has been uploaded. Old version description: The weight range is from - 6 to 3, and both the real style model and the anime animation style model are competent. Specifically controlled through Lora's weight. There are two versions, the standard version has a strong emphasis on detail reduction, which leads to the development of the screen towards a different style. Because the content in the prompt has been reduced, the screen has changed. Therefore, -6 does not mean that the quality of the screen has decreased, but may have obtained a relatively pure screen style. In addition, there is also a Lite version available, which has a relatively small impact and overall controllable changes, making it convenient for fine-tuning numerical values. Additionally, different screen contents and models have varying levels of support for Lora. A small amount of themes and content will explode after a weight greater than 3. This Lora model is experimental in nature, and testing support is welcome. Thank you all! 一款主要用于增加细节的lora。 新版2.5说明 正数的权重依然能增加画面细节，特别适合无中生有，负数的权重会减少细节，目前出现了一种有趣的现象，一些2d或者2.5d效果的模型在使用负数权重后模型被逐渐还原成3d真实效果。也许这种有趣的效果能够使用在某些特殊需求的项目上，故此上传了该版本。 旧版说明： 权重范围从-6到3都能起效，不管是真实风格模型还是二次元动漫风格模型都能胜任。 具体通过lora的权重来控制。 有两个版本，标准版力度较强，削减细节方面使得画面朝另外一个画面风格发展，因为将prompt中的内容削减后使得画面产生了变化，所以-6并不意味着画面质量下降，而是可能获得了比较纯粹的画面风格。 另外还准备有lite版本，lite版本的影响较小，整体变化可控，方便数值上进行微调。 另外不同的画面内容和模型对lora的支持程度不同。少量题材和内容将会在大于3的权重后爆炸。 此lora模型为实验性质，欢迎进行测试支持。谢谢大家！ 主にディテールを増やすためのlora。 新版2.5の説明 正数の重みは依然として画面の詳細を増やすことができ、特に無中生有に適しており、負の重みは詳細を減らすことができ、現在は興味深い現象が現れており、2 dまたは2.5 d効果のモデルの中には負の重みを使用した後、モデルは徐々に3 d真実効果に還元されている。この興味深い効果は、特定のニーズに特化したアイテムに使用できる可能性があるので、このバージョンをアップロードしました。 旧バージョンの説明： 重みの範囲は-6から3まで有効で、リアルスタイルモデルでも二次元アニメスタイルモデルでも適任です。 具体的にはloraの重みによって制御される。 2つのバージョンがあり、標準版は力が強く、細部を削減することで画面が別の画面スタイルに発展し、proptの内容を削減して画面が変化したため、-6は画面の品質が低下したことを意味するのではなく、比較的純粋な画面スタイルを獲得した可能性がある。 また、liteバージョンも用意されており、liteバージョンの影響は小さく、全体の変化は制御可能で、数値的な微調整が容易である。 また、画面内容やモデルによってloraへのサポート度合いが異なります。少量の題材と内容は3より大きい重みで爆発する。 このloraモデルは実験的な性質であり、テストサポートを歓迎します。ありがとうございました！ 디테일을 더하는 데 주로 사용되는 로라. 새 버전 2.5 설명 양수의 가중치는 여전히 화면의 디테일을 증가시킬수 있다. 특히 터무니없이 적합하다. 음수의 가중치는 디테일을 감소시킨다. 현재 재미있는 현상이 나타났다. 일부 2d 또는 2.5d 효과의 모델은 음수의 가중치를 사용한후 모델은 점차 3d의 진실한 효과로 환원되였다.아마도 이런 재미있는 효과는 일부 특수한 수요의 항목에 사용될수 있기에 이 버전을 업로드하였다. 이전 릴리즈 노트: 가중치 범위는 -6에서 3까지 모두 효과가 있으며, 실제 스타일 모델이든 2차 애니메이션 스타일 모델이든 모두 감당할 수 있다. 구체적으로 로라의 권중을 통해 통제한다. 두 가지 버전이 있는데 표준판의 강도가 비교적 강하고 디테일을 줄이는 데 있어 화면이 다른 화면 스타일로 발전했다. 왜냐하면 prompt의 내용을 줄인 후에 화면에 변화가 생겼기 때문에 -6은 화면의 질이 떨어지는 것을 의미하는 것이 아니라 비교적 순수한 화면 스타일을 얻을 수 있다. 또한 lite 버전이 준비되어 있으며 lite 버전의 영향이 적고 전체적인 변화를 제어할 수 있어 수치상 미세조정이 용이하다. 또 화면 내용과 모델에 따라 로라에 대한 지원 정도가 다르다.소량의 소재와 내용은 3보다 큰 가중치 후에 폭발할 것이다. 이 로라 모델은 실험적인 성질이므로 테스트 지원을 환영합니다.감사합니다!
7,280 descargas · 701 me gusta en Civitai
SHA256
B6F7DE3B2D7EAD649F3CD341717760EE625DE9EB824B0FB021E76F9A68611897
ID de modelo de ModelTrends.ai
#601
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
LMS Karras
Pasos
120
Guidance
7.4
Semilla
4055276717
Tamaño
754x1131
Prompt
(landscape focus), <lora:tangbohu-detailmaker_v2.5:0.44>, (((SFW))), (((score_9, score_8_up, score_7_up))), In the center of a darkened chamber, suspended in mid-air, hovers a magnificent orb, pulsating with an otherworldly energy. Its surface shimmers with an iridescent sheen, hinting at the immense power contained within. Encased within the orb are the elemental forces of nature, wind, water, and fire, each swirling and churning in a mesmerizing dance of creation and destruction. At the orb's core, the element of nature manifests as rugged, rocky formations that seem to shift and rearrange like tectonic plates. Crystalline structures jut out from the surface, catching the light and refracting it into dazzling prisms. The wind element, represented by swirling currents of air, weaves through the orb like a gentle breeze, carrying with it the whispers of distant lands and the promise of endless horizons. Water, in its liquid form, cascades and flows in graceful rivulets, its surface rippling with undulating waves. Droplets hang suspended in mid-air, catching the light and casting shimmering reflections upon the walls of the chamber. And at the heart of the orb, the element of fire burns with an intense, fiery glow, its flames licking at the edges of the other elements with an insatiable hunger. As the elemental forces interact within the orb, they create a symphony of motion and energy. nature and water blend together, forming rich, fertile soils that nurture life and growth. Wind stirs the elements, carrying them across the orb's surface in an endless cycle of renewal. And fire, with its fierce intensity, provides the spark of life that ignites the other elements, fueling their eternal dance of creation and destruction. Together, these elemental forces form a delicate balance within the orb, each one dependent on the others for harmony and equilibrium. As they weave together and writhe amongst one another, they create a spectacle of beauty and power that defies description, a testament to the boundless potential of the elemental forces that shape our world.
Prompt negativo
(((score_6_down, score_5_down, score_4_down))), (((NSFW))), (NegfeetV2:0.91), (badhandv4:0.91), (verybadimagenegative_v1.3:0.55), (easynegative:0.55), (open mouth, teeth, closed eyes), (singular subject focus), ((overly busy scene)), (simple scene), (plain background), (simple background), (poorly detailed anatomy:1.27), (imperfect anatomy:1.27), (flawed anatomy:1.27),
Ajustes
Sampler
DPM++ 2M Karras
Pasos
28
Guidance
7
Semilla
1507324641
Tamaño
512x768
Clip skip
2
Peso usado
1
Prompt
{{{gadgethackwrench}}}, {{{animal nose}}}, {{{mouse tail}}}, {{{short girl}}}, {{{little, tiny, short body}}}, {{{long orange hair}}}, {{{buck teeth}}}, {{{flower in hair}}}, {{{tv studio background, stay smiling, slightly blush,masterpiece, looking aside}}}, disney, female, blue eyes, goggles on head, mouse tail, medium breasts, well lit, masterpiece, detailed, {{dynamic pose, cinematic pose}
Prompt negativo
verybadimagenegative, (bad_prompt:0.8), badhandv4, badhands5, (low quality, worst quality:1.4), (bad anatomy), (inaccurate limb:1.2), bad composition, inaccurate eyes, extra digit, fewer digits, (extra arms:1.2), (nsfw, nude), loli, child, (goggles, goggles on head), profile, ((headwear)), english text, twitter username, patreon username, human, human face, horns, out of frame, cropped face, human ears, photorealistic, 3d
Ajustes
Sampler
DPM++ 2M Karras
Pasos
28
Guidance
7
Semilla
1507324639
Tamaño
512x768
Clip skip
2
Peso usado
1
Prompt
{{{gadgethackwrench}}}, {{{animal nose}}}, {{{mouse tail}}}, {{{short girl}}}, {{{little, tiny, short body}}}, {{{long orange hair}}}, {{{buck teeth}}}, {{{flower in hair}}}, {{{tv studio background, stay smiling, slightly blush,masterpiece, looking aside}}}, disney, female, blue eyes, goggles on head, mouse tail, medium breasts, well lit, masterpiece, detailed, {{dynamic pose, cinematic pose}
Prompt negativo
verybadimagenegative, (bad_prompt:0.8), badhandv4, badhands5, (low quality, worst quality:1.4), (bad anatomy), (inaccurate limb:1.2), bad composition, inaccurate eyes, extra digit, fewer digits, (extra arms:1.2), (nsfw, nude), loli, child, (goggles, goggles on head), profile, ((headwear)), english text, twitter username, patreon username, human, human face, horns, out of frame, cropped face, human ears, photorealistic, 3d
Ajustes
Sampler
DPM++ SDE Karras
Pasos
30
Guidance
7.5
Semilla
3583225514
Tamaño
512x768
Clip skip
2
Prompt
sunny, 30 yo woman, simple background, thick winged eyeliner, dark smokey eyeshadow, hair in braided mohawk, movement pose, bodycon dresss, a documentary photography photo with a Sony A7III, by Khadija Saye, ((Chiaroscuro, soft fill lighting, dramatic rim lighting, dutch angle)), cinematic color grading, depth of field, <lora:3DMM_V12:0.5> <lora:skinny_new_skin:0.4> <lora:tangbohu-detailmaker_v2.5:1> <lora:sunny_v01:0.9>
Prompt negativo
BadDream, UnrealisticDream, bad-hands-5, Asian-Less-Neg, JuggernautNegative-neg, ERA09NEGV2
Ajustes
Sampler
DPM++ 2M Karras
Pasos
28
Guidance
7
Semilla
1507324643
Tamaño
512x768
Clip skip
2
Peso usado
1
Prompt
{{{gadgethackwrench}}}, {{{animal nose}}}, {{{mouse tail}}}, {{{short girl}}}, {{{little, tiny, short body}}}, {{{long orange hair}}}, {{{buck teeth}}}, {{{flower in hair}}}, {{{tv studio background, stay smiling, slightly blush,masterpiece, looking aside}}}, disney, female, blue eyes, goggles on head, mouse tail, medium breasts, well lit, masterpiece, detailed, {{dynamic pose, cinematic pose}
Prompt negativo
verybadimagenegative, (bad_prompt:0.8), badhandv4, badhands5, (low quality, worst quality:1.4), (bad anatomy), (inaccurate limb:1.2), bad composition, inaccurate eyes, extra digit, fewer digits, (extra arms:1.2), (nsfw, nude), loli, child, (goggles, goggles on head), profile, ((headwear)), english text, twitter username, patreon username, human, human face, horns, out of frame, cropped face, human ears, photorealistic, 3d
Ajustes
Sampler
DPM++ 2M Karras
Pasos
25
Guidance
7
Semilla
922727518
Tamaño
512x768
Clip skip
2
Peso usado
1
Prompt
{{{gadgethackwrench}}}, {{{animal nose}}}, {{{mouse tail}}}, {{{short girl}}}, {{{little}}}, {{{long yellow hair}}}, {{{fully clothed, blue jumpsuit}}}, {{{stay, sun rays, masterpiece, looking at viewer}}}, looking at viewer, disney, female, curvy, blue eyes, goggles on head, mouse tail, medium breasts, sun rays, well lit, masterpiece, detailed, sitting on a stool, warehouse, window, sky, clouds, half closed eyes,
Prompt negativo
worst quality, large head, low quality, extra digits, bad eye, EasyNegativeV2, ng_deepnegative_v1_75t
Ajustes
Sampler
DPM++ 2M Karras
Pasos
28
Guidance
7
Semilla
1144237993
Tamaño
512x768
Clip skip
2
Peso usado
1
Prompt
{{{gadgethackwrench}}}, {{{animal nose}}}, {{{mouse tail}}}, {{{short girl}}}, {{{little}}}, {{{long yellow hair}}}, {{{red chinese dress:15:1}}}, {{{red chinese dress:15.1}}}, {{{stay, sun rays, masterpiece, looking aside}}},disney, female, blue eyes, goggles on head, mouse tail, medium breasts, sun rays, well lit, masterpiece, detailed, stay, japanese garden, sky, clouds, surprised
Prompt negativo
verybadimagenegative, (bad_prompt:0.8), badhandv4, badhands5, (low quality, worst quality:1.4), (bad anatomy), (inaccurate limb:1.2), bad composition, inaccurate eyes, extra digit, fewer digits, (extra arms:1.2), (nsfw, nude), loli, child, (goggles, goggles on head), profile, ((headwear)), english text, twitter username, patreon username, human, human face, horns, out of frame, cropped face, human ears, photorealistic, 3d
Ajustes
Sampler
DPM++ 2M Karras
Pasos
28
Guidance
7
Semilla
1787861946
Tamaño
512x768
Clip skip
2
Peso usado
1
Prompt
{{{gadgethackwrench}}}, {{{animal nose}}}, {{{mouse tail}}}, {{{short girl}}}, {{{little}}}, {{{long yellow hair}}}, {{{ yellow shorts, wet t-shirt:1.7, white t-shirt:1.7, seethru:1.7, see-thru:1.7}}}, {{{stay, sun rays, masterpiece, looking aside}}},disney, female, blue eyes, goggles on head, mouse tail, medium breasts, sun rays, well lit, masterpiece, detailed, stay, near the pool, sky, clouds, surprised
Prompt negativo
verybadimagenegative, (bad_prompt:0.8), badhandv4, badhands5, (low quality, worst quality:1.4), (bad anatomy), (inaccurate limb:1.2), bad composition, inaccurate eyes, extra digit, fewer digits, (extra arms:1.2), (nsfw, nude), loli, child, (goggles, goggles on head), profile, ((headwear)), english text, twitter username, patreon username, human, human face, horns, out of frame, cropped face, human ears, photorealistic, 3d
Modelos de IA para generar imagen, vídeo y texto
detail make es un LoRA para la familia Stable Diffusion 1.5 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 Stable Diffusion 1.5, consulta su puntuación de calor para ver su tendencia y ábrelo en PirateDiffusion o BitVector para probarlo.
Explorar por familia de modelos
Ver todos los modelos