📘 I’m preparing documentation on creating models using unique methods.
🎁 Active users of my models will receive it for free.
Follow and share your creations.

BRAVO - Graphite. Unleash Your Style Fusion!

🎨 Your First Steps with the Model
1. The Blank Canvas: Leave both the positive and negative prompt fields completely empty. Generate a few images. Let the model introduce itself style.
2. The Key Restriction: Now, place a single word in the negative prompt: "graphite". Generate a couple more images. Now the prompts on the preview images will make more sense. They also list some styles to start your experiments.

👉If you need to blend genres, for example Sci-Fi and Fantasy, use the model “BRAVO – Cybernetic Threshold”.

👉If you need to mix styles, for example comics, poster, sketch, expressionism, anime, art deco, pop art, abstraction and others, use the model “BRAVO – ArtEry”.

⚠️Excellent results with:
Schedule type: Beta
Sampling method: Euler a, DPM++ 2M, DDIM, Heun, DPM++ SDE
CFG Scale: 5-7, Sampling steps: 40+
Max Resolution: 512x896

🌟 Advanced Techniques: Stylistic Alchemy

This model excels at blending and separating artistic techniques in a single image. Think of it as a master of stylistic fusion.

Harmonious Fusion: Use this powerful phrase in your positive prompt to seamlessly blend styles:

"(harmonious combination styles: style_one, style_two, style_three:1.25)"

Example: (harmonious combination styles: watercolor, charcoal, painterly abstraction, vivid strokes:1.25)

Split Personality: For a more dramatic, partitioned effect where different parts of the image use different styles, try:

split styles, charcoal portrait with watercolor background

Unleash Pure Creativity: Don't be afraid to invent your own styles! The model responds wonderfully to abstract and evocative concepts. Try prompts like:

lunar brushstrokes
rusty pencil sketch
blind cat's paint strokes
windy smears
infernal ink
drawing with the left foot
...

Let your imagination run wild. BRAVO - Graphite thrives on the unconventional and will often reward you with stunning and surprising results.

📖 A Brief History

Initially crafted as a base for generating consistent style datasets to train LoRAs, this model quickly proved to be far too interesting and versatile to keep locked away in a development pipeline. Its ability to maintain coherence while juggling wildly different aesthetics made it a tool for discovery, and now it's my pleasure to share it with the community.

You've seen its potential. Now, it's your turn to explore. Dive in, experiment fearlessly, and see what masterpieces you can co-create.