fashn-vton-1.5

GitHub Repository: fashn-vton-1.5

Hugging Face Model: fashn-vton-1.5

Official Website Introduction: fashn-vton-1-5

Downloading Model Files

Model Directory Structure

/fashnvton1.5/weights

├── dwpose

│ ├── dw-ll_ucoco_384.onnx

│ └── yolox_l.onnx

├── fashn-human-parser

│ ├── README.md

│ ├── config.json

│ ├── model.safetensors

│ └── preprocessor_config.json

└── model.safetensors

Instructions Click the blue text links above to download the files. Create a dwpose folder and place dw-ll_ucoco_384.onnx and yolox_l.onnx inside it. Create a fashn-human-parser folder and place README.md, config.json, model.safetensors, and preprocessor_config.json inside it. Create a weights folder, place the model.safetensors file inside it, and then move both the dwpose and fashn-human-parser folders into this weights folder.

# Download the Archive

This archive, fashnvton1.5, comes pre-configured with the necessary Python environment and corresponding dependencies.

Download Link: fashnvton1.5

If you wish to mirror the files to your own storage or increase your download speed, you can register via my referral link: pcloud

After extracting the archive, move the relevant model files to the designated directory—for example, the weights folder.

Next, double-click start.bat to launch the application with a single click.

You will see the terminal window open as the application starts up.

Once the application has successfully launched, open the provided URL in your web browser. http://127.0.0.1:7860/

You should see the initial interface.

Select an image from the examples folder to serve as the sample photo, then click “Start.”

After waiting a short while, you will observe that the outfit worn by the girl in the top-left corner has been replaced by the outfit worn by the girl in the bottom-left corner.

You can choose to perform a virtual try-on for the upper body, lower body, or the entire outfit.

The input image can be a standalone picture of clothing or a photo of another person; the system will automatically process and interpret it.

Note

My system configuration features an RTX 3060 graphics card with 12GB of VRAM.