YAML Metadata
Warning:
empty or missing yaml metadata in repo card
(https://huggingface.co/docs/hub/model-cards#model-card-metadata)
LoRA Model Card: svjack/Scaramouche_wan_2_2_14_B_text2video_lora
Dynamic Anime Character Synthesis
Base Model: Wan2.2_T2V_A14B
Fine-tuned Adapters:
wan_2_2_14B_Scaramouche_lora_v1_000004250_high_noise.safetensors
(weight: 1.0)wan_2_2_14B_Scaramouche_lora_v1_000004250_low_noise.safetensors
(weight: 1.0)
Key Strengths:- Theatrical Gestures: Mastery of exaggerated yet elegant movements (e.g., glove-adjusting, deliberate bites).
- Vibrant Aesthetics: Striking contrast between deep blue/purple hair/eyes and vivid backgrounds.
- Costume Detail: Intricate rendering of accessories (wide hats, patterned capes, gloves).
- Emotional Nuance: Capturing (mischievous) expressions through micro-movements.
Optimized Example Prompts
Example 1: Sensory Ice Cream Ritual
Prompt:
一个二次元美少年,有着短短的深蓝色头发和引人注目的紫色眼睛。
他戴着一顶宽大而松软的蓝色帽子,披着一件与之相配的带有复杂白色图案的蓝色斗篷。
少年忽然以指尖抵住下唇,眼波流转间漾开狡黠笑意。只见他慢条斯理地探出手去,
黑色手套包裹的指尖精准扣住冰淇淋甜筒的脆皮边缘。当第一口清凉甜蜜触碰舌尖时,
他微微眯起眼睛,唇角勾起若有似无的弧度,舌尖轻轻扫过残留的奶油痕迹,
喉间溢出一声几不可闻的喟叹。这般寻常的消暑小食,
在他举手投足间竟化作了一场精妙绝伦的感官戏剧——每一帧都像是精心编排的舞台剧,
将单纯的享乐升华为令人屏息的艺术瞬间。远处,被柔焦处理的绿色景致如梦似幻,
为这场即兴演出铺就出空灵诗意的底色。
Adapters:
Scaramouche_w14_high_lora
(1.0)Scaramouche_w14_low_lora
(1.0)
Key Features:- Macro shots of glove texture and ice cream sheen
- Slow-motion tongue/cream interaction
- Theatrical lighting isolating the character
Example 2: Dawn Rose Ballet
Prompt:
晨光中的少年拥有乱翘的深紫色短发与水润的大眼睛,那双明亮的紫色瞳孔盛满天真。
当少年伸手去够那束娇艳玫瑰时,黑色手套包裹的指尖以芭蕾舞者般的优雅姿态捻住花茎,
整个动作如慢镜头般流畅——仿佛不是在摘花,而是在进行一场献给晨曦的默剧表演。
Adapters:
Scaramouche_w14_high_lora
(1.0)Scaramouche_w14_low_lora
(1.0)
Key Features:- Particle effects for floating petals/dew
- Backlit silhouette highlighting cape dynamics
- Fluid glove articulation resembling dance choreography
Example 3: Urban Street Theater
Prompt:
现代都市的斑斓光影里,一位身着蓝白格纹西装背心的少年静立在涂鸦墙前的便利店门口。
他系着水蓝色领带,头戴黑色贝雷帽,怀里抱着印有紫绿撞色图案的吸管饮料杯,
左手还随意插在校服裤袋中。后背的书包同样点缀着格纹图案,
金属校徽在阳光下泛着冷光。少年面容冷淡,
短促的深蓝发梢与那双熟悉的紫色眼眸暴露了他不羁的本性——即便站在充满市井烟火气的街角,
那份浑然天成的戏剧张力仍让他如同站在舞台中央,与身后夸张的黄色店铺招牌构成奇妙的光影剧场。
Adapters:
Scaramouche_w14_high_lora
(1.0)Scaramouche_w14_low_lora
(1.0)
Key Features:- Neon reflection on beret fabric and plastic cup
- Dynamic depth-of-field separating character from graffiti
- "Aloof" micro-expressions via subtle lip/eyebrow shifts
Example 4: Honeyed Stage Whimsy
Prompt:
一个二次元美少年,拥有乱翘的深紫色短发与水润的大眼睛,
镜头缓缓拉开,琥珀色聚光灯如蜂蜜般倾泻而下,照亮圆形舞台中央那个裹着棕色小熊连体衣的身影。
帽檐垂落的白色绒毛球扫过少年泛红的脸颊——那顶缀着猫耳的兜帽之下,
深蓝色发丝被汗水浸湿贴在额间,紫罗兰色的瞳孔在追光灯下流转着蜜糖般的光泽。
之后随着镜头拉远,少年抱起一罐蜂蜜,使用一个勺子吃蜂蜜。
Adapters:
Scaramouche_w14_high_lora
(1.0)Scaramouche_w14_low_lora
(1.0)
Key Features:- Viscous honey fluid simulation
- Hoodie fluff physics reacting to movement
- Dramatic spotlight isolating "stage" space
Example 5: Chocolate Nocturne
Prompt:
镜头从左至右缓缓推进,二次元美少年,深紫色短发凌乱翘起,水润紫罗兰色瞳孔泛着蜜糖光泽,
身穿白色宽松睡衣半敞衣领,慵懒陷在蓬松鹅绒被中。少年拿着比利时海螺巧克力,
金棕色糖霜沾在唇角,丝绸睡衣随动作滑落肩头,床头暖光将巧克力螺旋纹路照得晶莹剔透。
散落的巧克力碎屑与深蓝发丝缠绕在枕边,猫耳兜帽软塌塌垂落床沿。风格是二次元,比例是16:9
Adapters:
Scaramouche_w14_high_lora
(1.0)Scaramouche_w14_low_lora
(1.0)
Key Features:- Fabric wrinkles and chocolate crumb detailing
- Low-angle lighting emphasizing (languid) posture
- Micro-expressions of pleasure via parted lips/half-closed eyes
Technical Parameters
Setting | Recommendation | Notes |
---|---|---|
CFG Scale | 1 (Fixed) | Wan2.2 architecture requirement |
Sampler | uni_pc | Optimal for cloth/cape dynamics |
Steps | 25 | Balances detail and speed |
Resolution | 832x480 | Maintains VRAM efficiency |
Motion Factor | 2-4 | Subtle for gestures, higher for fabric |
Performance Profile
- VRAM Usage: ~23GB at 832x480 (RTX 4090)
- Render Speed: 35-50 sec/frame
- Troubleshooting:
- Overexposed Accessories: Add
overbright hat, glare
to negative prompts - Stiff Gestures: Increase motion factor +
flowing_movement
token - Color Bleed: Apply
vibrance_filter
node at 0.2 strength
- Overexposed Accessories: Add
License
CC-BY-NC-SA 4.0 (Non-commercial, share-alike)
Community Hub: https://huggingface.co/svjack/Scaramouche_wan_2_2_14_B_text2video_lora/discussions
- Downloads last month
- 1
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support