Time-line for code releasre

[moeinheidari7829]

Dear authors,

Thank you so much for open-sourcing this great work! Is there an approximate timeline of the code release?

Bests

[mrokuss]

Hi Moein, thanks so much for your interest in our work!

We’re currently cleaning things up and running the necessary internal checks. We’re aiming to have the code released around the end of the year. We’ll keep you posted.

Thanks again for your patience!

[luvargas2]

Hi @mrokuss , congratulations on the Voxtell work! I’ve been diving into the code, and it’s very elegant.
I 'm currently exploring an extension to your architecture and would like to benchmark against your reported results. Could you provide an estimated timeline for the release of the pretrained model weights and the fine-tuning scripts?

[mrokuss]

Hi @luvargas2 and @moeinheidari7829, thanks a lot for the kind words and for your interest in VoxTell!

We’ve just released the pretrained model weights here (we recommend v1.1):
https://huggingface.co/mrokuss/VoxTell

We also released the Napari plugin for inference and interactive use:
https://github.com/MIC-DKFZ/napari-voxtell

The code for fine-tuning and additional training-related components will need a bit more time. We’ll share updates as soon as it’s ready.

In the meantime, feel free to reach out if anything is unclear or if you have questions about benchmarking or the current release.

Best,
Max

[Treeboy2762]

Hi @mrokuss, massive congratulations on pulling off this project!

I would like to extend upon VoxTell and tweak a bit of the training procedures, especially on RexGroundingCT. Is this model designed to be compatible with the nnUNet training scheme only? I was wondering if I should write a training script myself rather than waiting for its integration to the nnUNet pipeline.

Any advice would be appreciated :) Happy new year!
Daesung