williamgjn/herbkg

Herbal entity extraction based on BioBERT

Additional DataSet：

Link：https://pan.baidu.com/s/1pK7edbDEcQ5AjwWczaJxHg

PassWord：herb

Pretrained model:

Link：https://pan.baidu.com/s/1hO__P44XxzgUsqKQvrzcFg

PassWord：herb

Demo

https://adventurous-production-58c.notion.site/HerbKG-Documentation-b0e82bbcc29f4754950d701f2397e832

Trainning demo:

->pip install -r requirements.txt

->python run_classifier.py --task_name=mrpc --do_train=True --do_eval=True --do_predict=True do_train_and_eval=True --save_checkpoint_max=10000 --num_train_epochs=1 --max_seq_length=512 --train_batch_size=16 --data_dir=DiseaseData --vocab_file=pretrained/vocab.txt --bert_config_file=pretrained/config.json --init_checkpoint=pretrained/model.ckpt --output_dir=outputs