README_from_modelscope.md

---
base_model: MusePublic/FLUX.1-Kontext-Dev@v1
cover_images:
- _cover_images_/26_blingbling.jpg
- _cover_images_/9_blingbling.jpg
- _cover_images_/6_blingbling.jpg
- _cover_images_/12_blingbling.jpg
- _cover_images_/15_blingbling.jpg
- _cover_images_/5_blingbling.jpg
- _cover_images_/11_blingbling.jpg
frameworks:
- Pytorch
license: Apache License 2.0
tags:
- LoRA
- text-to-image
tasks:
- text-to-image-synthesis
vision_foundation: FLUX_1

#model-type:
##如 gpt、phi、llama、chatglm、baichuan 等
#- gpt

#domain:
##如 nlp、cv、audio、multi-modal
#- nlp

#language:
##语言代码列表 https://help.aliyun.com/document_detail/215387.html?spm=a2c4g.11186623.0.0.9f8d7467kni6Aa
#- cn 

#metrics:
##如 CIDEr、Blue、ROUGE 等
#- CIDEr

#tags:
##各种自定义，包括 pretrained、fine-tuned、instruction-tuned、RL-tuned 等训练方法和其他
#- pretrained

#tools:
##如 vllm、fastchat、llamacpp、AdaSeq 等
#- vllm
---

# 玻璃雕像 - Kontext 图像编辑 LoRA

## 模型介绍

本 LoRA 模型是基于 [Kontext](https://www.modelscope.cn/models/black-forest-labs/FLUX.1-Kontext-dev) 模型和 [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio) 训练的 LoRA 模型，使用本模型后，可输入指令 `Transform into a glass sculpture.` 将画面中的主体内容转换为晶莹的玻璃雕像。

## 模型效果

||样例 1|样例 2|样例 3|
|-|-|-|-|
|原图|![](./assets/9.jpg)|![](./assets/6.jpg)|![](./assets/12.jpg)|
|生成图|![](./assets/9_blingbling.jpg)|![](./assets/6_blingbling.jpg)|![](./assets/12_blingbling.jpg)|

||样例 4|样例 5|样例 6|
|-|-|-|-|
|原图|![](./assets/15.jpg)|![](./assets/5.jpg)|![](./assets/11.jpg)|
|生成图|![](./assets/15_blingbling.jpg)|![](./assets/5_blingbling.jpg)|![](./assets/11_blingbling.jpg)|

## 使用说明

本模型基于框架 [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio/tree/main/examples/flux) 训练，请先安装

```
git clone https://github.com/modelscope/DiffSynth-Studio.git  
cd DiffSynth-Studio
pip install -e .
```

```python
import torch
from diffsynth.pipelines.flux_image_new import FluxImagePipeline, ModelConfig
from PIL import Image
from modelscope import snapshot_download

snapshot_download("DiffSynth-Studio/FLUX.1-Kontext-dev-lora-blingbling", cache_dir="./models")
pipe = FluxImagePipeline.from_pretrained(
    torch_dtype=torch.bfloat16,
    device="cuda",
    model_configs=[
        ModelConfig(model_id="black-forest-labs/FLUX.1-Kontext-dev", origin_file_pattern="flux1-kontext-dev.safetensors"),
        ModelConfig(model_id="black-forest-labs/FLUX.1-dev", origin_file_pattern="text_encoder/model.safetensors"),
        ModelConfig(model_id="black-forest-labs/FLUX.1-dev", origin_file_pattern="text_encoder_2/"),
        ModelConfig(model_id="black-forest-labs/FLUX.1-dev", origin_file_pattern="ae.safetensors"),
    ],
)
pipe.load_lora(pipe.dit, "models/DiffSynth-Studio/FLUX.1-Kontext-dev-lora-blingbling/model.safetensors", alpha=1)

image = Image.open("your_image.jpg")
image = pipe(
    prompt="Transform into a glass sculpture.",
    kontext_images=image,
    embedded_guidance=2.5,
    seed=0,
)
image.save("output.jpg")
```