aipoly/Molytica_Mix

Molytica: A Universal Vision-Language Foundation Model for the Language of Chemistry

Molytica is an advanced molecular structure recognition vision language model (including the recognition of organic small molecules, organic chemical reactions, and polymers). It is based on visual Large Language Models and uses a carefully customized data synthesis system. Through end-to-end training, the model achieves industry-leading performance. In the era of artificial intelligence, it is expected to greatly accelerate researchers' extraction of relevant data from journal papers, patents, the internet, etc., thereby accelerating the development of Large Language Model technology in the fields of polymer or chemistry science.

Github: https://github.com/lunyang/Molytica