---
tags:
- text-to-image
- lora
- diffusers
- template:diffusion-lora
widget:
- text: '3DXLP4, a medium-sized African-American man stands in front of a backdrop of a cityscape. The mans upper torso is adorned with tattoos, including a cross, a gold chain, and a gold bracelet. He is wearing a gold ring, a silver watch, and blue jeans with a black belt. His hair is short and dark, and he is wearing a white bandana tied around his head. The backdrop is a deep blue, with a few buildings visible in the distance.'
  output:
    url: images/1111.png
- text: '3DXLP4, A close-up portrait of a man in a black suit with a black collared button down shirt and a black and white striped tie. The man has dark brown hair and blue eyes. The background is completely dark, creating a stark contrast to the mans face.'
  output:
    url: images/2222.png
- text: '3DXLP4, a blonde woman, wearing a pink bikini, adorned with a gold necklace and matching earrings, stands in front of a dark gray backdrop. She is striking a seductive pose, with her right hand on her hip, adding a pop of color to the scene. The womans lips are pursed, her lips are painted a deep pink, and her hair is a light blonde, her bangs are a darker shade of black, while her left hand is a gold ring.'
  output:
    url: images/3333.png
- text: '3DXLP4, Captured in a close-up, eye-level shot, a woman with long, straight red hair, wearing a red strapless bra, is striking a pose with a lollipop in her right hand. Her lips are painted a vibrant red, adding a pop of color to her face. Her eyes are a piercing blue, while her lips are a darker red. The backdrop is a stark gray, with a statue of liberty in the upper right corner.'
  output:
    url: images/4444.png
base_model: black-forest-labs/FLUX.1-dev
instance_prompt: 3DXLP4
license: creativeml-openrail-m
---
![strangerzonehf/Flux-3DXL-Partfile-0004](images/4x.png)

<Gallery />

## Model description for 3DXL Partfile 0004

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           | 17 & 2550 |
| Epoch                     | 20   | Save Every N Epochs       | 1     |

    Labeling: florence2-en(natural language & English)
    
    Total Images Used for Training : 14

## Best Dimensions & Inference

| **Dimensions** | **Aspect Ratio** | **Recommendation**       |
|-----------------|------------------|---------------------------|
| 1280 x 832      | 3:2              | Best                     |
| 1024 x 1024     | 1:1              | Default                  |

### Inference Range

- **Recommended Inference Steps:** 30–35

## Setting Up
```python
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 = "strangerzonehf/Flux-3DXL-Partfile-0004"
trigger_word = "3DXLP4"  
pipe.load_lora_weights(lora_repo)

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

You should use `3DXLP4` to trigger the image generation.

## Download model

Weights for this model are available in Safetensors format.

[Download](/strangerzonehf/Flux-3DXL-Partfile-0004/tree/main) them in the Files & versions tab.