About MolBERT
MolBERT is a BERT-like language model tailored for molecules represented as SMILES strings. It leverages unsupervised pretraining to learn contextualized molecular embeddings that improve downstream tasks such as property prediction and molecule generation.