A high-fidelity character LoRA trained on the Anima DiT architecture, designed with fully decoupled wardrobes and robust outfit flexibility across diverse settings.

💡 Core Features

Decoupled Identity: Trained on a 50-image balanced dataset.
Architecture: Optimized natively for Anima DiT checkpoints with clean, high-frequency facial geometry and versatile multi-aspect ratio support (incorporating portrait, square, and landscape buckets).
Modular Compatibility: Easily pairs with external clothing, lighting, and texture adapters (such as Cinematic Film Grain) without losing character likeness.

⚙️ Recommended Generation Settings

LoRA Weight: 0.75 – 0.85 (Recommended baseline: 0.80)
Trigger Word: Daniela Character
Resolution: 1024x1024 (Square), 832x1216 (Portrait), or 1216x832 (Landscape)
Sampling Steps: 30 – 32
CFG Scale: 4.0 – 4.5
Sampler: Euler or Euler a

🚫 Universal Negative Prompt

(worst quality, low quality:1.4), score_1, score_2, score_3, photorealistic, realistic, bad anatomy, bad hands, missing fingers, extra digits, blurry, cropped, frame, watermark, signature

✨ Universal Positive Base Tag Anchor

masterpiece, best quality, solo, 1girl,

🎨 Showcase Example Prompts

1. Modern Urban Streetwear

masterpiece, best quality, Daniela Character, solo, 1girl, cowboy shot, waist up, eye contact, confident casual smirk, wearing oversized graphic techwear hoodie with strap details, layered choker, urban rooftop at dusk, soft city lights glowing in the background, bokeh, cool blue and magenta color grading, cinematic composition

2. Cozy Lakeside Sunset

masterpiece, best quality, Daniela Character, solo, 1girl, cowboy shot, waist up, eye contact, calm gentle smile, resting hands on wooden railing, wearing loose cable-knit cream sweater, warm golden hour sunset reflecting off a calm rippling lake, soft backlit stray hairs, atmospheric haze, painterly illustration

3. Neo-Noir Detective Office

masterpiece, best quality, Daniela Character, solo, 1girl, cowboy shot, waist up, eye contact, sharp focused expression, leaning against a polished mahogany desk, wearing tailored double-breasted trench coat over a high-collar shirt, dimly lit vintage office interior, venetian blind window shadows casting light stripes, moody atmosphere, rich film grain texture

📊 Training Details

Base Architecture: Anima DiT v1.0 (Latent Diffusion Transformer)
Compute Infrastructure: Modal Cloud A100 (80GB)
Dataset Size: 50 Balanced Frames (24 Portrait, 13 Square, 13 Landscape buckets with full-body square integration)
Network Dimensions: Rank 32 / Alpha 32
Optimization Parameters: AdamW-8bit, bfloat16 mixed precision, Learning Rate: 0.00002, 1,500 Total Steps