Commit
•
351ca9b
1
Parent(s):
f6cd972
Update README.md
Browse files
README.md
CHANGED
|
@@ -24,10 +24,11 @@ widget:
|
|
| 24 |
|
| 25 |
# sd3-lora-alebrijeros-final
|
| 26 |
|
| 27 |
-
|
|
|
|
| 28 |
|
| 29 |
|
| 30 |
-
|
| 31 |
|
| 32 |
|
| 33 |
|
|
@@ -35,15 +36,6 @@ The main validation prompt used during training was:
|
|
| 35 |
sweatshirt painted in the alebrijeros style
|
| 36 |
```
|
| 37 |
|
| 38 |
-
## Validation settings
|
| 39 |
-
- CFG: `5.0`
|
| 40 |
-
- CFG Rescale: `0.0`
|
| 41 |
-
- Steps: `20`
|
| 42 |
-
- Sampler: `None`
|
| 43 |
-
- Seed: `42`
|
| 44 |
-
- Resolution: `512x512`
|
| 45 |
-
|
| 46 |
-
Note: The validation settings are not necessarily the same as the [training settings](#training-settings).
|
| 47 |
|
| 48 |
You can find some example images in the following gallery:
|
| 49 |
|
|
@@ -54,7 +46,7 @@ The text encoder **was not** trained.
|
|
| 54 |
You may reuse the base model text encoder for inference.
|
| 55 |
|
| 56 |
|
| 57 |
-
##
|
| 58 |
|
| 59 |
- Training epochs: 4
|
| 60 |
- Training steps: 2600
|
|
@@ -78,43 +70,7 @@ You may reuse the base model text encoder for inference.
|
|
| 78 |
|
| 79 |
## Datasets
|
| 80 |
|
| 81 |
-
|
| 82 |
-
- Repeats: 5
|
| 83 |
-
- Total number of images: 25
|
| 84 |
-
- Total number of aspect buckets: 1
|
| 85 |
-
- Resolution: 0.262144 megapixels
|
| 86 |
-
- Cropped: False
|
| 87 |
-
- Crop style: None
|
| 88 |
-
- Crop aspect: None
|
| 89 |
-
- Used for regularisation data: No
|
| 90 |
-
### alebrijeros-dataset-final-1024
|
| 91 |
-
- Repeats: 5
|
| 92 |
-
- Total number of images: 25
|
| 93 |
-
- Total number of aspect buckets: 1
|
| 94 |
-
- Resolution: 1.048576 megapixels
|
| 95 |
-
- Cropped: False
|
| 96 |
-
- Crop style: None
|
| 97 |
-
- Crop aspect: None
|
| 98 |
-
- Used for regularisation data: No
|
| 99 |
-
### alebrijeros-dataset-final-512-crop
|
| 100 |
-
- Repeats: 5
|
| 101 |
-
- Total number of images: 25
|
| 102 |
-
- Total number of aspect buckets: 1
|
| 103 |
-
- Resolution: 0.262144 megapixels
|
| 104 |
-
- Cropped: True
|
| 105 |
-
- Crop style: random
|
| 106 |
-
- Crop aspect: square
|
| 107 |
-
- Used for regularisation data: No
|
| 108 |
-
### alebrijeros-dataset-final-1024-crop
|
| 109 |
-
- Repeats: 5
|
| 110 |
-
- Total number of images: 25
|
| 111 |
-
- Total number of aspect buckets: 1
|
| 112 |
-
- Resolution: 1.048576 megapixels
|
| 113 |
-
- Cropped: True
|
| 114 |
-
- Crop style: random
|
| 115 |
-
- Crop aspect: square
|
| 116 |
-
- Used for regularisation data: No
|
| 117 |
-
|
| 118 |
|
| 119 |
## Inference
|
| 120 |
|
|
|
|
| 24 |
|
| 25 |
# sd3-lora-alebrijeros-final
|
| 26 |
|
| 27 |
+
Este es un LoRA derivado del modelo más reciente de stable diffusion:
|
| 28 |
+
[stabilityai/stable-diffusion-3.5-large](https://huggingface.co/stabilityai/stable-diffusion-3.5-large).
|
| 29 |
|
| 30 |
|
| 31 |
+
El prompt usado durante el entrenamiento:
|
| 32 |
|
| 33 |
|
| 34 |
|
|
|
|
| 36 |
sweatshirt painted in the alebrijeros style
|
| 37 |
```
|
| 38 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 39 |
|
| 40 |
You can find some example images in the following gallery:
|
| 41 |
|
|
|
|
| 46 |
You may reuse the base model text encoder for inference.
|
| 47 |
|
| 48 |
|
| 49 |
+
## Configuración del entrene
|
| 50 |
|
| 51 |
- Training epochs: 4
|
| 52 |
- Training steps: 2600
|
|
|
|
| 70 |
|
| 71 |
## Datasets
|
| 72 |
|
| 73 |
+
Imágenes variadas obtenidas de google que incluyen desde un mercedes pintado por alebrijeros hasta miss universo que fue con un vestido estilo alebrijero
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 74 |
|
| 75 |
## Inference
|
| 76 |
|