About release of the pre-trained Part-to-Code model

Thanks for your impressive work and open-source!

According to the paper, the framework relies on a foundational Part-to-Code inference model pretrained on a large-scale Synthetic Part Dataset

Could you please release the pre-trained checkpoint for this specific Part-to-Code model? 
It would be helpful to try to reconstruct other types of objects.

Thank you for your contribution again!