ita-mdt_weights

<p align="center"> <h1 style="font-size: 1.5em;">ITA-MDT: Image-Timestep-Adaptive Masked Diffusion Transformer Framework for Image-Based Virtual Try-On</h1> </p> <p align="center"> <a href="https://jiwoohong93.github.io/ita-mdt/"><b>Project Page</b></a>&nbsp&nbsp | &nbsp&nbsp <a href="https://github.com/jiwoohong93/ita-mdt_code">Repository</a>&nbsp&nbsp | &nbsp&nbsp <a href="https://arxiv.org/pdf/2503.20418">Paper</a>&nbsp&nbsp | &nbsp&nbsp <b>CVPR 2025</b> </p>

Details about the model, implementation, and results can be found through the sources above.

License: CC BY-NC-SA 4.0

Contact: [email protected]

---

How to Download Weights

The repository contains three weight files:

• ema_0.9999_2000000.pt → Exponential Moving Average (EMA) weights.
• model2000000.pt → Raw model checkpoint.
• opt2000000.pt → Optimizer state for resuming training.

[2025-10-08] Reuploaded with the correct model weights.

If you only plan to run generation, downloading ema_0.9999_2000000.pt is sufficient.

---

Option 1: Download using 'huggingface_hub'

from huggingface_hub import hf_hub_download

# Download only the EMA weights for generation
hf_hub_download(
    repo_id="jiwoohong93/ita-mdt_weights",
    filename="ema_0.9999_2000000.pt"
)

# (Optional) Download the raw model checkpoint
hf_hub_download(
    repo_id="jiwoohong93/ita-mdt_weights",
    filename="model2000000.pt"
)

# (Optional) Download the optimizer state
hf_hub_download(
    repo_id="jiwoohong93/ita-mdt_weights",
    filename="opt2000000.pt"
)

Option 2: Download with wget

# Download only the EMA weights for generation
wget https://huggingface.co/jiwoohong93/ita-mdt_weights/resolve/main/ema_0.9999_2000000.pt

# (Optional) Download the raw model checkpoint
wget https://huggingface.co/jiwoohong93/ita-mdt_weights/resolve/main/model2000000.pt

# (Optional) Download the optimizer state
wget https://huggingface.co/jiwoohong93/ita-mdt_weights/resolve/main/opt2000000.pt

Citation

We kindly encourage citation of our work if you find it useful.

@article{hong2025ita,
    title={ITA-MDT: Image-Timestep-Adaptive Masked Diffusion Transformer Framework for Image-Based Virtual Try-On},
    author={Hong, Ji Woo and Ton, Tri and Pham, Trung X and Koo, Gwanhyeong and Yoon, Sunjae and Yoo, Chang D},
    journal={arXiv preprint arXiv:2503.20418},
    year={2025}
}