An open-source alternative to proprietary molecular descriptor software like Dragon. It utilizes unsupervised pre-training on SMILES strings to generate high-dimensional embeddings for low-data drug discovery tasks.
Tech Stack
Jupyter NotebookChemistry & Materials
Why smiles-transformer?
• State-of-the-art transformer architecture
• Pre-trained weights available
• Handles SMILES syntax natively
Limitations
• Requires GPU for inference
• Complex environment setup
• Limited to SMILES input format
2/26/2026
Last Update
63
Forks
15
Issues
MIT
License
Financial Leak Detected
Stop the "SaaS Tax"
Your team could be burning cash. Switching to smiles-transformer instantly boosts your runway.
