---
tags:
- text-to-image
- lora
- diffusers
- template:diffusion-lora
widget:
- text: '3D Realism, a medium-angle shot captures a womans face, adorned with a brown fedora hat, a purple scarf, and a blue denim jean jacket. The womans hair is braided, adding a touch of texture to her outfit. The backdrop is a stark white wall, creating a stark contrast to the womans outfit.'
  output:
    url: images/4.png
- text: '3D Realism, A medium-built man is sitting on a park bench during sunset, dressed in a red checkered flannel shirt and blue jeans. He is holding a steaming cup of coffee in both hands. His curly black hair is slightly tousled, and a small scar is visible on his left cheek. The background features a glowing orange sky and silhouettes of trees.'
  output:
    url: images/5.png
- text: '3D Realism, A medium-sized woman stands on a sandy beach at sunset. She is wearing a long white summer dress and a wide-brimmed straw hat. Her long, wavy blond hair flows freely in the wind. She holds a bouquet of wildflowers in her left hand. The background shows a vibrant sunset over the ocean, with waves gently lapping the shore.'
  output:
    url: images/6.png
- text: '3D Realism, a medium-sized man stands against a dark brown backdrop. He is dressed in a black t-shirt, adorned with a black beard and mustache. His hair is short and neatly combed. His eyes are slightly open, as if he is looking into the camera. His arms are adorned with tattoos, adding a pop of color to the scene. The mans expression is neutral, with a slight smile on his face.'
  output:
    url: images/1.png
- text: '3D Realism, A close-up shot of a womans face, taken from a low-angle perspective. The womans eyes are a piercing blue, and her hair is a vibrant shade of brown. Her eyebrows are a darker brown, and she is wearing a pair of silver hoop earrings. Her lips are a lighter shade of pink, and the earrings are a dull gold. The background is a stark white, and there is a slight shadow on her face, adding depth to the composition.'
  output:
    url: images/2.png
- text: '3D Realism, a medium-sized Asian man, dressed in a blue and black striped sweater, is adorned with a black watch with a silver band. His left hand is raised in the air, adding a pop of color to the scene. His hair is styled in a short, spiky cut, and he is smiling, with a slight smile on his face. The backdrop is a muted yellow, creating a stark contrast to the mans attire.'
  output:
    url: images/3.png
base_model: black-forest-labs/FLUX.1-dev
instance_prompt: 3D Realism
license: creativeml-openrail-m
---
![adas.png](https://cdn-uploads.huggingface.co/production/uploads/65bb837dbfb878f46c77de4c/RIOA43KjJTdM7xS_Pn0Q_.png)

<Gallery />

# Model description for Flux-3D-Realism-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           | 13 & 2050 |
| Epoch                     | 10   | Save Every N Epochs       | 1     |

    Labeling: florence2-en(natural language & English)
    
    Total Images Used for Training : 10

## 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-3D-Realism-LoRA"
trigger_word = "3D Realism"  
pipe.load_lora_weights(lora_repo)

device = torch.device("cuda")
pipe.to(device)
```

## Trigger words

You should use `3D Realism` to trigger the image generation.

## Download model

Weights for this model are available in Safetensors format.

[Download](/strangerzonehf/Flux-3D-Realism-LoRA/tree/main) them in the Files & versions tab.