Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: creativeml-openrail-m
|
| 3 |
+
base_model: black-forest-labs/FLUX.1-dev
|
| 4 |
+
tags:
|
| 5 |
+
- lora
|
| 6 |
+
- flux
|
| 7 |
+
- ai-toolkit
|
| 8 |
+
- character-style
|
| 9 |
+
- person
|
| 10 |
+
- z-image-turbo
|
| 11 |
+
datasets:
|
| 12 |
+
- simone-v2
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
+
# Model Card: Simone-ZIT-v2.2 (Z-Image De-Turbo)
|
| 16 |
+
|
| 17 |
+
This is a specialized character LoRA optimized for **Z-Image De-Turbo (De-Distilled)**. Like the previous versions, it is trained as a **Style**, meaning the character is embedded into the model's weights without a specific trigger word.
|
| 18 |
+
|
| 19 |
+
## Training Philosophy
|
| 20 |
+
The model focuses on high-fidelity character consistency across various environments while maintaining the speed and quality benefits of the Z-Image Turbo architecture.
|
| 21 |
+
|
| 22 |
+
**Caption Strategy:**
|
| 23 |
+
* **No Trigger Word:** The character's name and defining features (hair/eye color) were omitted from captions to bake them directly into the subject weight.
|
| 24 |
+
* **Full Environment Decoupling:** Meticulous tagging of clothing, backgrounds, and lighting ensures the character remains a constant subject while the scene remains flexible.
|
| 25 |
+
* **Flexible Generation:** The character manifests automatically when describing a "woman" or "subject."
|
| 26 |
+
|
| 27 |
+
## Usage & Prompting
|
| 28 |
+
Describe a woman and her surroundings. The model will automatically apply the Simone character style.
|
| 29 |
+
|
| 30 |
+
* **Example Prompt:** `A woman wearing a silk blouse and tailored trousers, sitting in a modern sunlit cafe.`
|
| 31 |
+
* **LoRA Strength:** Recommended 0.7 - 1.0 for character accuracy.
|
| 32 |
+
* **Compatibility:** Best used with Z-Image Turbo/De-Turbo workflows.
|
| 33 |
+
|
| 34 |
+
## Technical Specifications
|
| 35 |
+
|
| 36 |
+
| Parameter | Value |
|
| 37 |
+
|---|---|
|
| 38 |
+
| **Trigger Word** | None (Trained as Style) |
|
| 39 |
+
| **Model Architecture** | Z-Image De-Turbo (De-Distilled) |
|
| 40 |
+
| **Rank (Dimension)** | 128 |
|
| 41 |
+
| **Batch Size** | 4 (Constant throughout training) |
|
| 42 |
+
| **Precision** | float8 |
|
| 43 |
+
| **Total Steps** | 4000 |
|
| 44 |
+
| **Save Frequency** | Every 200 steps |
|
| 45 |
+
| **Training Hardware** | NVIDIA H200 Tensor Core GPU |
|
| 46 |
+
| **Training Toolkit** | Ostris - AI Toolkit |
|
| 47 |
+
|
| 48 |
+
---
|
| 49 |
+
*Model card generated for the Simone-ZIT series.*
|