<h2 id="liquidfire-lora-for-flux-schnell">LiquidFire LoRA for Flux Schnell</h2>Style: Hyper-detailed liquid marble, molten glass, flowing fire, and metallic sheen. Ethereal, dynamic, and surreal textures with a glossy, reflective finish.Schnell Settings:<ul><li>Sampler: DPM++ 2M Karras</li><li>Steps: 20-30</li><li>Strength: 0.4-0.6</li><li>CFG: 3-5</li><li>Resolution: 1024x1024</li><li>Prompt Tips: Use keywords like "liquid marble, molten gold, flowing lava, glassy sheen, intricate swirls, high contrast lighting"</li></ul>Z-Image Settings:<ul><li>Sampler: DPM2</li><li>Steps: 28-32</li><li>Strength: 0.7-0.9</li><li>CFG: 6-7</li><li>Resolution: 1024x1024</li><li>Prompt Tips: Use keywords like "liquid marble, molten gold, flowing lava, glassy sheen, intricate swirls, high contrast lighting"</li></ul><h1 id="generative-ai-resource-tools-and-vault:">Generative AI Resource Tools &amp; Vault:</h1><h3 id="kofi:-https:www.ko-fi.comtensorvizion">KoFI: <a target="_blank" rel="ugc" href="https://www.ko-fi.com/tensorvizion">https://www.ko-fi.com/tensorvizion</a></h3><h3 id="patreon:-https:www.patreon.comtensorvizion">Patreon: <a target="_blank" rel="ugc" href="https://www.patreon.com/tensorvizion">https://www.patreon.com/tensorvizion</a></h3>

LiquidFire

LiquidFire-Z_c1-st3000.safetensors

Z-Image

Flux Schnell


# 

## 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: [Tongyi-MAI/Z-Image](https://modelscope.ai/models/Tongyi-MAI/Z-Image)
* Training Code: [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio)
* Data amount：26
* Total training steps：3000
* License: Apache-2.0

## Inference Code

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

```bash
pip install diffsynth
```

Inference:

```python
from diffsynth.pipelines.z_image import ZImagePipeline, ModelConfig
import torch

pipe = ZImagePipeline.from_pretrained(
    torch_dtype=torch.bfloat16,
    device="cuda",
    model_configs=[
        ModelConfig(model_id="Tongyi-MAI/Z-Image", origin_file_pattern="transformer/*.safetensors"),
        ModelConfig(model_id="Tongyi-MAI/Z-Image", origin_file_pattern="text_encoder/*.safetensors"),
        ModelConfig(model_id="Tongyi-MAI/Z-Image", origin_file_pattern="vae/diffusion_pytorch_model.safetensors"),
    ],
    tokenizer_config=ModelConfig(model_id="Tongyi-MAI/Z-Image", origin_file_pattern="tokenizer/"),
)
pipe.load_lora(pipe.dit, ModelConfig(model_id="tensorvizion/LiquidFire-Z", origin_file_pattern="LiquidFire-Z_c1-st3000.safetensors"))
prompt = "a cat"
image = pipe(prompt=prompt, num_inference_steps=50, cfg_scale=4)
image.save("image.jpg")
```


Files with SHA256: e93dc53ab6f513d2e0f6bff8cbcf20156bf944a285774c204ee94d98e7761445