README.md

---
license: apache-2.0
base_model:
- Qwen/Qwen-Image-Edit-2509
pipeline_tag: image-to-image
widget:
- text: >-
    apply the image 2 full costume to image 1 singing girl
  output:
    url: workflow-demo2.png
- text: >-
    use image 2 city night view as background for image 1
  output:
    url: workflow-demo.png
- text: >-
    use image 2 as background for image 1 fairy
  output:
    url: workflow-demo3.png
---
## qwen-image-edit-plus-gguf
- run it with `gguf-connector`; simply execute the command below in console/terminal
```
ggc q8
```
>
>GGUF file(s) available. Select which one to use:
>
>1. qwen-image-edit-plus-v2-iq3_s.gguf
>2. qwen-image-edit-plus-v2-iq4_nl.gguf
>4. qwen-image-edit-plus-v2-mxfp4_moe.gguf
>
>Enter your choice (1 to 3): _
>
- opt a `gguf` file in your current directory to interact with; nothing else

![screenshot](https://raw.githubusercontent.com/calcuis/gguf-pack/master/q8.png)

- `ggc q8` accepts multiple image input (see picture above; two images as input)
- as lite lora auto applied, able to generate output with merely 4/8 steps instead of the default 40 steps; save up to 80% loading time

![screenshot](https://raw.githubusercontent.com/calcuis/gguf-pack/master/q8a.png)

- up to 3 pictures plus customize prompt as input (above is 3 images input demo)

![screenshot](https://raw.githubusercontent.com/calcuis/gguf-pack/master/q8b.png)

- though `ggc q8` is accepting single image input (see above), you could opt the legacy `ggc q7` (see below); similar to image-edit model before

```
ggc q7
```

![screenshot](https://raw.githubusercontent.com/calcuis/gguf-pack/master/q7.png)

## **run it with gguf-node via comfyui**
- drag **qwen-image-edit-plus** to > `./ComfyUI/models/diffusion_models`
- *anyone below, drag it to > `./ComfyUI/models/text_encoders`
  - option 1: just **qwen2.5-vl-7b-test** [[5.03GB](https://huggingface.co/calcuis/qwen-image-edit-plus-gguf/blob/main/qwen2.5-vl-7b-test-q4_0.gguf)]
  - option 2: just **qwen2.5-vl-7b-edit** [[7.95GB](https://huggingface.co/calcuis/pig-encoder/blob/main/qwen_2.5_vl_7b_edit-q2_k.gguf)]
  - option 3: both **qwen2.5-vl-7b** [[4.43GB](https://huggingface.co/chatpig/qwen2.5-vl-7b-it-gguf/blob/main/qwen2.5-vl-7b-it-q4_0.gguf)] and **mmproj-clip** [[608MB](https://huggingface.co/chatpig/qwen2.5-vl-7b-it-gguf/blob/main/mmproj-qwen2.5-vl-7b-it-q4_0.gguf)]
- drag **pig** [[254MB](https://huggingface.co/calcuis/pig-vae/blob/main/pig_qwen_image_vae_fp32-f16.gguf)] to > `./ComfyUI/models/vae`

![screenshot](https://raw.githubusercontent.com/calcuis/comfy/master/qi-plus.png)

<Gallery />

## **run it with diffusers**
- might need the most updated git version for `QwenImageEditPlusPipeline`, should after this [pr](https://github.com/huggingface/diffusers/pull/12357/files); for i quant support, should after this [commit](https://github.com/huggingface/diffusers/commit/28106fcac4fd13e7ced5c9eb6803f107e804a08f); install the updated git version diffusers by:
```
pip install git+https://github.com/huggingface/diffusers.git
```
- simply replace `QwenImageEditPipeline` by `QwenImageEditPlusPipeline` from the qwen-image-edit inference example (see [here](https://huggingface.co/calcuis/qwen-image-edit-gguf))
```py
import torch, os
from diffusers import QwenImageTransformer2DModel, GGUFQuantizationConfig, QwenImageEditPlusPipeline
from diffusers.utils import load_image

model_path = "https://huggingface.co/calcuis/qwen-image-edit-plus-gguf/blob/main/qwen-image-edit-plus-v2-iq4_nl.gguf"

transformer = QwenImageTransformer2DModel.from_single_file(
    model_path,
    quantization_config=GGUFQuantizationConfig(compute_dtype=torch.bfloat16),
    torch_dtype=torch.bfloat16,
    config="callgg/image-edit-plus",
    subfolder="transformer"
    )
pipeline = QwenImageEditPipeline.from_pretrained("Qwen/Qwen-Image-Edit-2509", transformer=transformer, torch_dtype=torch.bfloat16)
print("pipeline loaded")
pipeline.enable_model_cpu_offload()
image = load_image("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/cat.png")
prompt = "Add a hat to the cat"
inputs = {
    "image": image,
    "prompt": prompt,
    "generator": torch.manual_seed(0),
    "true_cfg_scale": 2.5,
    "negative_prompt": " ",
    "num_inference_steps": 20,
}
with torch.inference_mode():
    output = pipeline(**inputs)
    output_image = output.images[0]
    output_image.save("output.png")
    print("image saved at", os.path.abspath("output.png"))
```

## **run nunchaku safetensors straight with gguf-connector (experimental feature)**
- run it with the new `q9` connector; simply execute the command below in console/terminal
```
ggc q9
```
>
>Safetensors available. Select which one to use:
>
>1. qwen-image-edit-lite-int4.safetensors
>
>Enter your choice (1 to 1): _
- opt a `safetensors` file in your current directory to interact with; nothing else

![screenshot](https://raw.githubusercontent.com/calcuis/gguf-pack/master/q9.png)
note: able to generate output with 4/8 steps (see above); surprisingly fast even with low end device; compatible with safetensors in nunchaku repo (depends on your machine; opt the right one)

![screenshot](https://raw.githubusercontent.com/calcuis/gguf-pack/master/hats.png)

### **reference**
- gguf-node ([pypi](https://pypi.org/project/gguf-node)|[repo](https://github.com/calcuis/gguf)|[pack](https://github.com/calcuis/gguf/releases))
- gguf-connector ([pypi](https://pypi.org/project/gguf-connector))