This is a BERT NER model using PyTorch adapted from Abhishek (https://github.com/abhishekkrthakur/bert-entity-extraction). MLFlow is wrapped around the training.py to log training results locally. Script is set with CUDA enabled. Model training parameters can be set in config.py, eg. Batch_size, and Epoch.
Understand the script by watching the Youtube video at https://www.youtube.com/watch?v=MqQ7rqRllIc&t=238s
Download the data from https://www.kaggle.com/datasets/abhinavwalia95/entity-annotated-corpus and unzip to input\ folder.
Clone BERT based model: https://huggingface.co/bert-base-uncased to input\bert_base_uncased folder.