Upload 2 files
Browse files- README (3).md +88 -0
- pytorch_lora_weights.safetensors +3 -0
README (3).md
ADDED
|
@@ -0,0 +1,88 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
---
|
| 3 |
+
tags:
|
| 4 |
+
- stable-diffusion-xl
|
| 5 |
+
- stable-diffusion-xl-diffusers
|
| 6 |
+
- text-to-image
|
| 7 |
+
- diffusers
|
| 8 |
+
- lora
|
| 9 |
+
- template:sd-lora
|
| 10 |
+
|
| 11 |
+
base_model: stabilityai/stable-diffusion-xl-base-1.0
|
| 12 |
+
instance_prompt: Margiela SS24 style
|
| 13 |
+
license: openrail++
|
| 14 |
+
---
|
| 15 |
+
|
| 16 |
+
# SDXL LoRA DreamBooth - tonyassi/margiela-ss24-fashion-lora
|
| 17 |
+
|
| 18 |
+
by [Tony Assi](https://www.tonyassi.com/)
|
| 19 |
+
|
| 20 |
+
Dreambooth Lora style based on the [Margiela SS24](https://www.vogue.com/fashion-shows/spring-2024-couture/maison-martin-margiela) collection.
|
| 21 |
+
|
| 22 |
+
|
| 23 |
+
|
| 24 |
+
## Trigger words
|
| 25 |
+
Use **Margiela SS24 style** in the prompt to trigger the style.
|
| 26 |
+
|
| 27 |
+
## How to use
|
| 28 |
+
```bash
|
| 29 |
+
pip install diffusers accelerate
|
| 30 |
+
```
|
| 31 |
+
|
| 32 |
+
```python
|
| 33 |
+
import torch
|
| 34 |
+
from diffusers import DiffusionPipeline, AutoencoderKL
|
| 35 |
+
|
| 36 |
+
# Load the pipeline
|
| 37 |
+
vae = AutoencoderKL.from_pretrained("madebyollin/sdxl-vae-fp16-fix", torch_dtype=torch.float16)
|
| 38 |
+
pipe = DiffusionPipeline.from_pretrained(
|
| 39 |
+
"stabilityai/stable-diffusion-xl-base-1.0",
|
| 40 |
+
vae=vae,
|
| 41 |
+
torch_dtype=torch.float16,
|
| 42 |
+
variant="fp16",
|
| 43 |
+
use_safetensors=True
|
| 44 |
+
)
|
| 45 |
+
pipe.load_lora_weights("tonyassi/margiela-ss24-fashion-lora")
|
| 46 |
+
pipe.to("cuda")
|
| 47 |
+
|
| 48 |
+
# Generate image
|
| 49 |
+
prompt = "Margiela SS24 style, megan fox wearing a gold mesh dress with crystals"
|
| 50 |
+
image = pipe(prompt=prompt,
|
| 51 |
+
height=1024,
|
| 52 |
+
width=1024,
|
| 53 |
+
num_inference_steps=50,
|
| 54 |
+
negative_prompt="ugly, deformed face, deformed body").images[0]
|
| 55 |
+
image
|
| 56 |
+
|
| 57 |
+
```
|
| 58 |
+
|
| 59 |
+
## Examples
|
| 60 |
+
|
| 61 |
+
**Margiela SS24 style, megan fox wearing a gold mesh dress with crystals**
|
| 62 |
+
|
| 63 |
+
|
| 64 |
+
**Margiela SS24 style, emma stone wearng a pink sequin leotard, puff sleves, rhinestone stockings**
|
| 65 |
+
|
| 66 |
+
|
| 67 |
+
**Margiela SS24 style, jeff goldblum**
|
| 68 |
+
|
| 69 |
+
|
| 70 |
+
**Margiela SS24 style, megan fox wearing a mesh dress with crystals**
|
| 71 |
+
|
| 72 |
+
|
| 73 |
+
**Margiela SS24 style, jeff goldblum wearing a corset**
|
| 74 |
+
|
| 75 |
+
## Model description
|
| 76 |
+
These are tonyassi/margiela-ss24-fashion-lora LoRA adaption weights for stabilityai/stable-diffusion-xl-base-1.0.
|
| 77 |
+
|
| 78 |
+
The weights were trained using [DreamBooth](https://dreambooth.github.io/).
|
| 79 |
+
|
| 80 |
+
LoRA for the text encoder was enabled: False.
|
| 81 |
+
|
| 82 |
+
Special VAE used for training: madebyollin/sdxl-vae-fp16-fix.
|
| 83 |
+
|
| 84 |
+
## Download model
|
| 85 |
+
Weights for this model are available in Safetensors format.
|
| 86 |
+
|
| 87 |
+
[Download](https://huggingface.co/tonyassi/margiela-ss24-fashion-lora/tree/main) them in the Files & versions tab.
|
| 88 |
+
|
pytorch_lora_weights.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:45486b053b22ab4504b12bab116ae45f27a407a449fd868499506c9d6aa0dd74
|
| 3 |
+
size 170539828
|