Open wallian opened 5 years ago
Just realized I forgot to reply to this before -- sorry about that @wallian!
You'd have to make/include the JSON file yourself, but tbh these models are somewhat out of date. You'll find much better models at https://github.com/olivettigroup/materials-synthesis-generative-models which is a release that corresponds to a newer paper.
There is a far better token classifier here https://github.com/olivettigroup/materials-synthesis-generative-models/blob/master/models/token_classifier.py and the repo includes the annotated data required to train said model.
Thank you so much for sharing these source codes.
I tried them, and got following error.
Could you share this "extraction_lexicon.json"? Is this already exists in somewhere? Or should I create it by myself?
Thank you!