README.md

---
tags:
- text-to-image
- lora
- diffusers
- template:diffusion-lora
- stable-diffusion-3.5-large-turbo
- hyper-realistic
- realism
- epic-realism
widget:
- text: >-
    Hyper Realistic, dark haired woman is standing in front of orange light, in
    glossy transparent plastic futuristic raincoat, short hair, volumetric
    lighting, cinematic angle, glamour vibrant.
  output:
    url: images/222.webp
- text: >-
    Photograph of two young men, one with short blonde hair and the other with
    brown hair, standing in front of mountains and wearing old West-style and
    tank tops, abs, fit, The image is in a vintage black-and-white photography
    style, in the style of Richard Avedon. --v 6. 1 --ar 113:128
  output:
    url: images/000.webp
- text: >-
    A photorealistic portrait of a 22-year-old Indian woman smiles confidently
    at the camera, with long, straight black hair and warm brown eyes. She wears
    a yet casual dress that combines comfort and style, posed in a full-body
    shot with her hands casually resting at her sides, showcasing a stylish
    watch on her left wrist. The image features high detail and quality,
    captured in soft, natural lighting that accentuates her approachable
    expression as she stands in a modern urban environment.
  output:
    url: images/444.webp
base_model: stabilityai/stable-diffusion-3.5-large-turbo
instance_prompt: hyper realistic
license: creativeml-openrail-m
---
# SD3.5-4/8Step-Large-Turbo-HyperRealistic-LoRA

Demo space + Grid Gen Upto 3X2 : https://huggingface.co/spaces/prithivMLmods/SD3.5-TURBO-REALISM (Not In Use)

<Gallery />

**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/SD3.5-Large-Turbo-HyperRealistic-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           | 20 & 2.2k |
| Epoch                     | 15     | Save Every N Epochs       | 1      |

    Labeling: florence2-en(natural language & English)
    
    Total Images Used for Training : 30

## Sample Image

![prithivMLmods/SD3.5-Large-Turbo-HyperRealistic-LoRA](images/1000.webp)

![prithivMLmods/SD3.5-Large-Turbo-HyperRealistic-LoRA](images/1001.webp)

## App File Structure

/project-root/

├── .gitattributes         
├── README.md      
├── app.py             
├── pythonproject.py     


## Setting Up

```
from diffusers import DiffusionPipeline
import torch

device = "cuda" if torch.cuda.is_available() else "cpu"
model_repo_id = "stabilityai/stable-diffusion-3.5-large-turbo"

torch_dtype = torch.bfloat16 if torch.cuda.is_available() else torch.float32

pipe = DiffusionPipeline.from_pretrained(model_repo_id, torch_dtype=torch_dtype)
pipe = pipe.to(device)

pipe.load_lora_weights("prithivMLmods/SD3.5-Large-Turbo-HyperRealistic-LoRA", weight_name="SD3.5-4Step-Large-Turbo-HyperRealistic-LoRA.safetensors")
trigger_word = "hyper realistic"  # Specify trigger word for LoRA
pipe.fuse_lora(lora_scale=1.0))
```
## Trigger words 🧨

> [!WARNING]
> **Trigger words:**  You should use `hyper realistic` to trigger the image generation.

## Download model

Weights for this model are available in Safetensors format.

[Download](/prithivMLmods/SD3.5-4Step-Large-Turbo-HyperRealistic-LoRA/tree/main) them in the Files & versions tab.