Molecule Transformers
Molecule Transformers is a collection of recipes for pre-training and fine-tuning molecular transformer language models, including BART, BERT, etc.
Pinned Repositories
moleculenet-bert-ssl
Semi-supervised learning techniques (pseudo-label, mixmatch, and co-training) for pre-trained BERT language model amidst low-data regime based on molecular SMILES from the Molecule Net benchmark.
moleculenet-smiles-bert-mixup
Training pre-trained BERT language model on molecular SMILES from the Molecule Net benchmark by leveraging mixup and enumeration augmentations.
moleculetransformers.github.io
Documentation for the Molecule Transformers.
rdkit-benchmarking-platform-transformers
Port of RDKit Benchmarking platform for pre-trained transformers-based language models for virtual screening drug discovery task.
smiles-augment
Augment molecular SMILES with methods including enumeration, and mixup, for low-data regime settings for downstream supervised drug discovery tasks.
smiles-featurizers
Extract Molecular SMILES embeddings from language models pre-trained with various objectives architectures.
Molecule Transformers's Repositories
MoleculeTransformers/smiles-featurizers
Extract Molecular SMILES embeddings from language models pre-trained with various objectives architectures.
MoleculeTransformers/moleculenet-smiles-bert-mixup
Training pre-trained BERT language model on molecular SMILES from the Molecule Net benchmark by leveraging mixup and enumeration augmentations.
MoleculeTransformers/moleculetransformers.github.io
Documentation for the Molecule Transformers.
MoleculeTransformers/moleculenet-bert-ssl
Semi-supervised learning techniques (pseudo-label, mixmatch, and co-training) for pre-trained BERT language model amidst low-data regime based on molecular SMILES from the Molecule Net benchmark.
MoleculeTransformers/smiles-augment
Augment molecular SMILES with methods including enumeration, and mixup, for low-data regime settings for downstream supervised drug discovery tasks.
MoleculeTransformers/rdkit-benchmarking-platform-transformers
Port of RDKit Benchmarking platform for pre-trained transformers-based language models for virtual screening drug discovery task.