README.md

---
license: mit
tags:
  - stable-diffusion
  - lora
  - safetensors
  - text-to-image
  - real-estate
  - kohya-ss
base_model: runwayml/stable-diffusion-v1-5
---

# Crestline Shreshth

## Model Description
Crestline-Shreshth is a Stable Diffusion 1.5 LoRA fine-tuned for generating photorealistic real estate imagery. It produces high-quality visuals of residential and commercial properties, interiors, and architectural renders — ideal for real estate marketing, virtual staging, and property visualization.

## Model Architecture
- **Base Model**: Stable Diffusion v1.5 (`runwayml/stable-diffusion-v1-5`)
- **Fine-tuning Method**: LoRA (Low-Rank Adaptation) v2
- **Framework**: Diffusers / Kohya-SS
- **Task**: Text-to-Image Generation (Real Estate Domain)

## Training Details
- **Training Tool**: Kohya-SS LoRA trainer
- **Domain**: Real estate photography — interiors, exteriors, architectural renders
- **Version**: v2 (improved over v1 with more training data and better trigger words)
- **Steps**: Fine-tuned to convergence on curated real estate image dataset

## Files
| File | Description |
|------|-------------|
| `Crestline_Shreshth_v2.safetensors` | LoRA weights v2 (safetensors format) |

## Usage

```python
from diffusers import StableDiffusionPipeline
import torch
from huggingface_hub import hf_hub_download

# Download LoRA weights
lora_path = hf_hub_download(repo_id='devanshty/Crestline-Shreshth', filename='Crestline_Shreshth_v2.safetensors')

# Load base pipeline
pipe = StableDiffusionPipeline.from_pretrained(
    "runwayml/stable-diffusion-v1-5",
    torch_dtype=torch.float16
).to("cuda")

# Load LoRA weights
pipe.load_lora_weights(lora_path)

# Generate
image = pipe("luxury modern living room, real estate photography, bright natural light, 4K").images[0]
image.save("output.png")
```

## Download & Use

```python
from huggingface_hub import hf_hub_download
lora_path = hf_hub_download(repo_id='devanshty/Crestline-Shreshth', filename='Crestline_Shreshth_v2.safetensors')
```