Downloads · 30 days
0
QuantaBricks/riemanngen
riemanngen is a machine learning model from QuantaBricks. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
Trained checkpoints for the RiemannGen project: two from-scratch MolMIM-style latent sequence models.
Downloads · 30 days
0
Access
Public
Updated Jul 5, 2026
Repo size
211 MB
Likes
0
Public
Click a slice to open those files.
.pt211 MB · 100%
From the Hugging Face model README
Trained checkpoints for the RiemannGen project: two from-scratch MolMIM-style latent sequence models.
atom/final-atom.pt + atom/vocab_new.txt -- plain-SMILES model.fragment/final.pt + fragment/vocab_new.txt -- SAFE-fragment-encoded
model, for scaffold growth / local fragment editing.Licensed under Apache License 2.0 -- free to use, modify, and redistribute.
If you use these checkpoints or the RiemannGen codebase, please cite:
@misc{riemanngen2026,
title = {RiemannGen: latent-space molecule generation with atom and fragment backends},
author = {XeonChem},
year = {2026},
url = {https://huggingface.co/XeonChem/riemanngen}
}