Update README.md
Browse files
README.md
CHANGED
|
@@ -75,6 +75,57 @@ Note: This adapter works well for extracting top wear (t-shirts, shirts, jackets
|
|
| 75 |
|
| 76 |
---
|
| 77 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 78 |
## Trigger words
|
| 79 |
|
| 80 |
You should use `[photo content]` to trigger the image generation.
|
|
|
|
| 75 |
|
| 76 |
---
|
| 77 |
|
| 78 |
+
## Parameter Settings
|
| 79 |
+
|
| 80 |
+
| Setting | Value |
|
| 81 |
+
| ------------------------ | ------------------------ |
|
| 82 |
+
| Module Type | Adapter |
|
| 83 |
+
| Base Model | FLUX.1 Kontext Dev - fp8 |
|
| 84 |
+
| Trigger Words | [photo content], extract only the specified clothing item [top wear, bottom wear] from the image and place it over a clean, plain background. Present the result in a product photography style — well-lit, crisp, and professional — while preserving the garment’s original textures, colors, shapes, and fine details. |
|
| 85 |
+
| Image Processing Repeats | 50 |
|
| 86 |
+
| Epochs | 25 |
|
| 87 |
+
| Save Every N Epochs | 1 |
|
| 88 |
+
|
| 89 |
+
Labeling: DeepCaption-VLA-7B(natural language & English)
|
| 90 |
+
|
| 91 |
+
Total Images Used for Training : 200 Image Pairs (100 Start, 100 End)
|
| 92 |
+
|
| 93 |
+
## Training Parameters
|
| 94 |
+
|
| 95 |
+
| Setting | Value |
|
| 96 |
+
| --------------------------- | --------- |
|
| 97 |
+
| Seed | - |
|
| 98 |
+
| Clip Skip | - |
|
| 99 |
+
| Text Encoder LR | 0.00001 |
|
| 100 |
+
| UNet LR | 0.00005 |
|
| 101 |
+
| LR Scheduler | constant |
|
| 102 |
+
| Optimizer | AdamW8bit |
|
| 103 |
+
| Network Dimension | 64 |
|
| 104 |
+
| Network Alpha | 32 |
|
| 105 |
+
| Gradient Accumulation Steps | - |
|
| 106 |
+
|
| 107 |
+
## Label Parameters
|
| 108 |
+
|
| 109 |
+
| Setting | Value |
|
| 110 |
+
| --------------- | ----- |
|
| 111 |
+
| Shuffle Caption | - |
|
| 112 |
+
| Keep N Tokens | - |
|
| 113 |
+
|
| 114 |
+
## Advanced Parameters
|
| 115 |
+
|
| 116 |
+
| Setting | Value |
|
| 117 |
+
| ------------------------- | ----- |
|
| 118 |
+
| Noise Offset | 0.03 |
|
| 119 |
+
| Multires Noise Discount | 0.1 |
|
| 120 |
+
| Multires Noise Iterations | 10 |
|
| 121 |
+
| Conv Dimension | - |
|
| 122 |
+
| Conv Alpha | - |
|
| 123 |
+
| Batch Size | - |
|
| 124 |
+
| Steps | 2700 |
|
| 125 |
+
| Sampler | euler |
|
| 126 |
+
|
| 127 |
+
---
|
| 128 |
+
|
| 129 |
## Trigger words
|
| 130 |
|
| 131 |
You should use `[photo content]` to trigger the image generation.
|