---
tags:
- text-to-image
- lora
- diffusers
- template:diffusion-lora
- Pastel
widget:
- text: 'Pastel BG, a young woman with brown hair and blue eyes stands in front of a colorful backdrop. The womans face is adorned with freckles, adding a pop of color to her outfit. The backdrop is a vibrant shade of purple, with yellow stars and stripes on it.'
  output:
    url: images/PB1.png
- text: 'Pastel BG, An eye-level view of a gray tabby cat with long white whiskers and a pink nose. The cats head is tilted slightly to the right, and its eyes are wide open. Its ears are pointed up, and the cats fur is a mix of gray and black. The background is a combination of pink, purple, and yellow, with white dots dotting the background. To the left of the cat, there is a purple star with a white butterfly on it.'
  output:
    url: images/PB2.png
- text: 'Pastel BG, a man stands in front of a colorful backdrop. He is dressed in a light pink suit jacket, a yellow collared shirt, and a pair of sunglasses. His hair is styled in a short bob, and his eyes are slightly open. His lips are slightly parted, as if he is looking to the right. The backdrop is a combination of pink, yellow, and green, with small white stars on the right side of the wall.'
  output:
    url: images/PB3.png
base_model: black-forest-labs/FLUX.1-dev
instance_prompt: Pastel BG
license: creativeml-openrail-m
---
# Pastel-BG-Flux-LoRA

<Gallery />

- Hosted Here🧨: https://huggingface.co/spaces/prithivMLmods/FLUX-LoRA-DLC

**The model is still in the training phase. This is not the final version and may contain artifacts and perform poorly in some cases.**

## Model description 

**prithivMLmods/Pastel-BG-Flux-LoRA**

Image Processing Parameters 

| Parameter                 | Value  | Parameter                 | Value  |
|---------------------------|--------|---------------------------|--------|
| LR Scheduler              | constant | Noise Offset              | 0.03   |
| Optimizer                 | AdamW  | Multires Noise Discount   | 0.1    |
| Network Dim               | 64     | Multires Noise Iterations | 10     |
| Network Alpha             | 32     | Repeat & Steps           | 28 & 3340|
| Epoch                     | 15  | Save Every N Epochs       | 1      |

    Labeling: florence2-en(natural language & English)
    
    Total Images Used for Training : 18 [ Hi-RES ]
    
## Best Dimensions

- 1024 x 1024 (Default)
  
## Setting Up
```
import torch
from pipelines import DiffusionPipeline

base_model = "black-forest-labs/FLUX.1-dev"
pipe = DiffusionPipeline.from_pretrained(base_model, torch_dtype=torch.bfloat16)

lora_repo = "prithivMLmods/Pastel-BG-Flux-LoRA"
trigger_word = "Pastel BG"  
pipe.load_lora_weights(lora_repo)

device = torch.device("cuda")
pipe.to(device)
```
## Trigger words

You should use `Pastel BG` to trigger the image generation.

## Download model

Weights for this model are available in Safetensors format.

[Download](/prithivMLmods/Pastel-BG-Flux-LoRA/tree/main) them in the Files & versions tab.