<h1 id="wavevapour">WaveVapour</h1>WaveVapour is a vaporwave-focused LoRA designed to bring a strong retro-futuristic aesthetic to your generations. It leans into the classic vaporwave look with neon lighting, nostalgic color palettes, dreamy environments, surreal details, and a distinctly late-80s/90s digital atmosphere.The LoRA works well for everything from portraits and fashion shots to cyberpunk-inspired scenes, retro interiors, cityscapes, and abstract compositions. It can also be mixed with different styles and checkpoints to create more personalized results.<h2 id="recommended-settings">Recommended Settings</h2><ul><li>LoRA strength: 0.7–1.0</li><li>Starting point: 0.8</li><li>Steps: Use your checkpoint's normal recommended range</li><li>CFG: Use your checkpoint's normal recommended range</li><li>Sampler: Any standard sampler should work well</li><li>Scheduler: Basic or Karras</li></ul>For a subtler aesthetic, try 0.5–0.7. For a much stronger vaporwave treatment, push it toward 1.0–1.2, depending on the base model.<h2 id="prompting">Prompting</h2>WaveVapour responds particularly well to prompts describing neon environments, retro technology, glowing lights, sunsets, reflective surfaces, CRT displays, surreal architecture, nostalgic interiors, and atmospheric night scenes.Example:<code>a woman standing in a retro arcade at night, neon lights, CRT monitors, glowing pink and blue lighting, reflective floor, nostalgic 90s atmosphere, cinematic photography</code>The LoRA is intended to be flexible rather than locking you into one specific composition, so experiment with different subjects and environments to find your preferred look.Have fun experimenting with WaveVapour and pushing that retro-futuristic aesthetic in your own direction.<h2 id="krea-2-version-of-this-model-and-more-ai-resources">Krea 2 Version of this Model &amp; more AI resources</h2>KoFI: <a target="_blank" rel="ugc" href="https://www.kofi.com/tensorvizion">https://www.kofi.com/tensorvizion</a>Patreon: <a target="_blank" rel="ugc" href="https://www.patreon.com/tensorvizion">https://www.patreon.com/tensorvizion</a>

WaveVapour

WaveVapour-Qwen-2.1_c1-st4000.safetensors

Qwen 2.1

Z-Image


# 

## Model Introduction

This model is trained using the [Civision/model training](https://modelscope.ai/civision/modelTraining) environment and computing resources provided by ModelScope.

* Model Type: LoRA
* Base Model: [Qwen/Qwen-Image-2.1](https://modelscope.ai/models/Qwen/Qwen-Image-2.1)
* Training Code: [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio)
* Data amount：24
* Total training steps：4000
* License: Apache-2.0

## Inference Code

Install [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio):

```bash
pip install diffsynth
```

Inference:

```python
from diffsynth.pipelines.qwen_image_21 import QwenImage21Pipeline, ModelConfig
import torch

pipe = QwenImage21Pipeline.from_pretrained(
    torch_dtype=torch.bfloat16,
    device="cuda",
    model_configs=[
        ModelConfig(model_id="Qwen/Qwen-Image-2.1", origin_file_pattern="transformer/diffusion_pytorch_model*.safetensors"),
        ModelConfig(model_id="Qwen/Qwen-Image-2.1", origin_file_pattern="text_encoder/model*.safetensors"),
        ModelConfig(model_id="Qwen/Qwen-Image-2.1", origin_file_pattern="vae/diffusion_pytorch_model*.safetensors"),
    ],
    processor_config=ModelConfig(model_id="Qwen/Qwen-Image-2.1", origin_file_pattern="processor/"),
)
pipe.load_lora(pipe.dit, ModelConfig(model_id="tensorvizion/WaveVapour-Qwen-2.1", origin_file_pattern="WaveVapour-Qwen-2.1_c1-st4000.safetensors"))

prompt = "a cat"
image = pipe(prompt, seed=0, num_inference_steps=50)
image.save("image.png")
```


Files with SHA256: cacbda60572655e61e3d7151a9659667da981407cc4377d944b1113976a43857