Explore my workflow creation space! Feel free to click the link below and experience the future of AIGC generation directly.<a target="_blank" rel="ugc" href="https://www.runninghub.ai?inviteCode=rh-v1559">https://www.runninghub.ai?inviteCode=rh-v1559</a>Fan benefits,register to get 1000 points,daily login 100 points,play 4090!Experience the super power of 48G.## Model OverviewWelcome to Krea2_MangaTint_FilmBloom!An aesthetic enhancement LoRA designed to merge Retro Film Halation, dreamy light leaks, warm color grading, and vibrant anime artistry, newly added to the Krea2_MangaTint family.This model brings a nostalgic, cinematic atmosphere to your illustrations. It softly diffuses bright highlights with iconic red/golden film halation, subtle lens flares, organic film grain, and rich color tone mapping without losing line sharpness. Perfect for generating youth anime portraits, golden hour cityscapes, retro aesthetic posters, and emotion-rich storybook scenes.## ⚙️ Recommended Parameters (Optimized for Krea2 Turbo)To capture the delicate light leaks and velvety halation glow, please apply the following configurations:* Base Architecture: Krea2* **Inference Pipeline:** Krea2 Turbo Version* Sampling Steps: 8 steps (Strictly optimized for Turbo speed)* CFG Scale: 1.0 (Crucial for Turbo models)* LoRA Weight: 0.8 - 1.0 (0.9 is highly recommended for optimal film warmth)* Core Trigger Tags: <code>mangatint style</code>, <code>filmbloom scenery</code>, <code>film halation</code>, <code>analog film aesthetic</code>, <code>warm light leak</code>, <code>dreamy bloom effect</code>, <code>retro color grade</code>, <code>masterpiece illustration</code>## 🎨 Prompting Guide* To boost signature film halation and light leaks: <code>red film halation around highlights</code>, <code>vintage light leak</code>, <code>soft lens flare</code>, <code>golden hour lighting</code>, <code>subtle film grain</code>* To enhance cinematic anime atmosphere: <code>nostalgic vibe</code>, <code>cinematic lighting</code>, <code>rich warm color palette</code>, <code>retro anime aesthetic</code>

Krea2_MangaTint_FilmBloom

Krea2_MangaTint_FilmBloom.safetensors

v1.0


# Krea2_漫彩_胶片晕光彩

## 模型介绍

本模型依托魔搭社区（ModelScope）AIGC专区[模型训练](https://modelscope.cn/aigc/modelTraining)环境与算力完成训练。

* 模型类型：LoRA
* 基础模型：[krea/Krea-2-Turbo](https://modelscope.cn/models/krea/Krea-2-Turbo)
* 训练代码：[DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio)
* 训练数据量：97
* 总训练步数：7000
* 开源协议：Apache-2.0

## 推理代码

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

```bash
pip install diffsynth
```

开始推理：

```python
from diffsynth.pipelines.krea2 import Krea2Pipeline, ModelConfig
import torch

pipe = Krea2Pipeline.from_pretrained(
    torch_dtype=torch.bfloat16,
    device="cuda",
    model_configs=[
        ModelConfig(model_id="krea/Krea-2-Turbo", origin_file_pattern="turbo.safetensors"),
        ModelConfig(model_id="Qwen/Qwen3-VL-4B-Instruct", origin_file_pattern="*.safetensors"),
        ModelConfig(model_id="Qwen/Qwen-Image", origin_file_pattern="vae/diffusion_pytorch_model.safetensors"),
    ],
    tokenizer_config=ModelConfig(model_id="Qwen/Qwen3-VL-4B-Instruct", origin_file_pattern=""),
)
pipe.load_lora(pipe.dit, ModelConfig(model_id="zhouwenbin1994/Krea2_MangaTint_FilmBloom", origin_file_pattern="Krea2_MangaTint_FilmBloom_c2-st7000.safetensors"))
prompt = "a cat"
image = pipe(
    prompt, seed=0,
    num_inference_steps=8, cfg_scale=1, mu=1.15,
)
image.save("image.jpg")
```


Files with SHA256: 1a4f6101b05f761ff5199a88e6b258692def3edf11974ce7d0be72c7c0ccd7bf