bash run.sh
fragment_tokenizer/tokenizer.py
Defines Fragment Augmentation, Encoder and Decoder.moses_train_test_split.py
Takes inmoses_dataset.csv
and produces train, test moses data.encode_moses.py
Produces failed and augmented encoded moses smiles files.model/tokenizer.py
Creates FragmentBERT Tokenizer.model.py
Train the model and log to WandB.
conda env create -f environment.yml
conda activate fbdd