This is a casenote script file containing model training and testing using LEGAL BERT model published in huggingface
- The project is developed using the latest version of tensorflow (wheel is for Mac M1)
- The most critical dependency is HuggingFace, python3
During development, the only input file needed is a csv file containing:
- col1: Sentence
- col2-n: the six label categories of the sentence
First add a data.csv file under data/ folder, then a test run is as simple as
python3 tfBertClf.py