FasciumMOSS SFT ARTY

FasciumMOSS SFT ARTY

FasciumMOSS SFT ARTY

An Encyclopedic Art Foundation Model — Where AI Thinks Like a Master Painter

🎨 About This Model

FasciumMOSS SFT ARTY is not just another merge — it's a full fine-tuned artistic foundation model built on the MOSS-Image-SFT architecture.

Unlike typical LoRA merges that layer styles on top of a base, this model has been deeply trained to understand art history at a structural level. It doesn't imitate styles — it thinks in the visual language of each movement.

From the geometric rigor of Soviet Constructivism to the luminous dots of Pointillism, from Caravaggio's dramatic tenebrism to Matisse's wild Fauvist colors — this model has internalized 20+ major art movements as native ways of seeing, not as surface-level filters.

The result: A model that generates museum-quality artwork across radically different aesthetics, maintaining anatomical integrity, compositional mastery, and emotional depth in every style.

Key Features

• Encyclopedic Art Knowledge: Trained on diverse artistic styles spanning Renaissance to Contemporary

• Structural Understanding: Doesn't just copy styles — understands the philosophy behind each movement

• Anatomical Precision: Maintains correct proportions even in radical stylistic transformations (Cubism, Expressionism)

• Light Physics Mastery: From Caravaggio's tenebrism to Impressionist dappled light

• Color Theory Intelligence: Scientific approach to color (Pointillism) and emotional color (Fauvism)

• Zero Trigger Words Required: Styles are activated naturally through descriptive prompts

⚙️ Recommended Settings

Parameter

Value

Notes

Sampler

Euler a

Optimal for flow-matching architecture

Scheduler

Tangent

Best results with MOSS-Image base

Steps

50

Sweet spot for detail vs. speed

CFG Scale

3.0

Low CFG for natural, artistic results

Resolution

1024×1024 or 768×1152

Portrait orientation works best

Clip Skip

2

Standard for SDXL-based models

Negative Prompt (Optional):

photorealistic, photograph, 3d render, plastic, smooth, digital art, modern, contemporary, ugly, deformed, blurry, low quality, watermark, signature, text

🖼️ Style Guide & Example Prompts

1. Soviet Constructivism

Soviet Constructivist poster art, a powerful female figure in red and black geometric composition, dynamic diagonal lines, bold typography elements, photomontage aesthetic, El Lissitzky and Alexander Rodchenko influence, revolutionary energy, stark contrast, propaganda poster style, avant-garde 1920s

2. Art Deco Glamour (Tamara de Lempicka)

Art Deco glamour portrait, a woman with geometric bob haircut adorned with golden headpiece, wearing an emerald green velvet gown with angular patterns, standing against a backdrop of stylized sunburst and stepped ziggurat motifs, rich gold and black palette, luxurious materials, 1920s elegance, Tamara de Lempicka inspired, sharp lines and metallic sheen

3. Victorian Gothic

Victorian Gothic cathedral interior, a woman in mourning dress with jet black velvet and lace, standing beside ornate stone effigy, candlelight creating dramatic shadows on carved pillars, stained glass casting colored light, melancholic atmosphere, Pre-Raphaelite darkness, architectural grandeur, memento mori aesthetic

4. German Expressionism (Kirchner)

German Expressionist portrait, angular distorted features inspired by Ernst Ludwig Kirchner, bold unnatural colors - electric green skin, crimson shadows, jagged brushstrokes, urban alienation theme, Die Brücke movement aesthetic, emotional intensity through color and form

5. Cubist Fragmentation (Picasso)

Cubist fragmentation portrait, woman's face decomposed into geometric planes and multiple perspectives simultaneously, Picasso's Analytic Cubism style, muted brown and gray palette with ochre accents, intersecting facets showing profile and frontal views at once, intellectual abstraction

6. Fauvism (Matisse)

Fauvist wild color portrait, Henri Matisse inspired, completely unnatural saturated colors - turquoise face, magenta hair, yellow shadows, bold flat areas of pure color, expressive brushwork, color as emotion over realism, decorative patterns in background

7. Pointillism (Seurat)

Pointillist garden scene, Georges Seurat technique, woman in white dress composed entirely of tiny colored dots, optical mixing creates luminous effect, dappled sunlight through trees rendered in pure spectral colors, scientific approach to color, Neo-Impressionism, meticulous dot-by-dot technique

8. Baroque Tenebrism (Caravaggio)

Baroque tenebrism portrait, Caravaggio style, extreme chiaroscuro with deep black background, woman illuminated by single dramatic light source from above, rich burgundy and gold fabrics, religious painting intensity, emotional drama through light and shadow, 17th century master technique

9. Rococo Pastoral (Fragonard)

Rococo pastoral fantasy, Jean-Honoré Fragonard inspired, woman in pastel pink silk dress with elaborate powdered hair adorned with flowers, lounging in idealized garden with marble statues, soft diffused light, decorative frivolity, 18th century aristocratic leisure, ornamental brushwork

10. Symbolist Dream Vision (Böcklin)

Symbolist dream vision, Arnold Böcklin style, mysterious woman in classical drapery beside mystical island temple, cypress trees, twilight atmosphere, mythological undertones, death and beauty themes, romantic symbolism, haunting melancholic mood, painted dream logic

🔬 Technical Details

• Base Architecture: MOSS-Image-SFT (Flow-Matching)

• Training Method: Full Fine-Tune (SFT)

• Training Dataset: Curated collection of 2000+ masterworks across 20+ art movements

• Resolution: Optimized for 1024×1024 and 768×1152

• File Format: .safetensors (INT8 quantized for efficiency)

• File Size: ~5.2 GB

📚 How to Use

1. Download the model and place it in your models/Stable-diffusion folder

2. Select the model in your UI (Forge, ComfyUI, A1111)

3. Set sampler to Euler a, scheduler to Tangent, steps to 50, CFG to 3.0

4. Write a descriptive prompt mentioning the desired art style, technique, and mood

5. Generate and enjoy museum-quality artwork!

Pro Tips:

• Be specific in your prompts: mention the artist, movement, technique, and color palette

• Use natural language: The model understands artistic concepts, not just keywords

• Experiment with resolution: Portrait (768×1152) works best for figures, square (1024×1024) for compositions

• Low CFG is key: CFG 3.0 allows the model's artistic knowledge to shine without over-constraining it

🎯 What Makes This Different?

Most AI art models are trained to produce photorealistic images or anime-style illustrations. FasciumMOSS SFT ARTY is different:

• It's an art historian: Understands the difference between Analytic and Synthetic Cubism

• It's a color scientist: Knows why Seurat used dots and Matisse used flat colors

• It's a lighting master: Can recreate Caravaggio's tenebrism or Impressionist dappled light

• It's a composer: Understands negative space, perspective, and visual balance

This is not a model for generating "pretty pictures" — it's a model for creating meaningful art.

🙏 Credits & Acknowledgments

• Base Model: OpenMOSS-Team/MOSS-Image-SFT

• Training Framework: Custom SFT pipeline with gradient checkpointing

• Inspiration: The great masters — from Botticelli to Lichtenstein

• Community: Civitai creators who pushed the boundaries of AI art

📝 License & Usage

• Commercial Use: Allowed

• Attribution: Appreciated but not required

• NSFW: Not recommended (model is optimized for fine art)

• Redistribution: Please link back to this page

Final Note

This model represents a new direction in AI art: from imitation to understanding. It doesn't just copy styles — it thinks like an artist who has studied art history for decades.

Whether you're creating illustrations for a book, concept art for a game, or simply exploring the boundaries of AI creativity, FasciumMOSS SFT ARTY offers a level of artistic intelligence that goes beyond typical AI generation.

Welcome to the future of AI art — where the machine doesn't just generate, it creates.

Model by FASCIUM | Version 1.0 | October 2026

Tags: art, painting, classical, modern, style, fine art, museum, masterpiece, art history, creative, illustration, concept art