README.md

---
tags:
  - stable-diffusion
  - lora
  - fluxart
  - image-generation
license: creativeml-openrail-m
base_model: FluxArt/stable-flux-v1.5
library_name: diffusers
pipeline_tag: text-to-image
---

# Flux LoRA v0.004

LoRA-модификация для модели [FluxArt/stable-flux-v1.5](https://huggingface.co/FluxArt/stable-flux-v1.5), предназначенная для генерации кинематографичных портретов в стиле киберпанк, неон-нуар и sci-fi.

## 🧠 Использование

```python
from diffusers import StableDiffusionPipeline
pipe = StableDiffusionPipeline.from_pretrained("FluxArt/stable-flux-v1.5").to("cuda")
pipe.load_lora_weights("sergSt/flux-lora", weight_name="flux-lora-000004.safetensors")
pipe.fuse_lora()
image = pipe("cinematic portrait of a cyberpunk samurai").images[0]
image.save("output.png")