seyonec's picture
Create README.md
4922f62

RoBERTa model trained on 1M SMILES from PubChem 77M set in MoleculeNet. Uses Smiles-Tokenizer