Extract Triples from text

Step1:

Predownload chinese bert model and extract it to pretrained_bert_model.

wget https://s3.amazonaws.com/models.huggingface.co/bert/bert-base-chinese.tar.gz
wget https://s3.amazonaws.com/models.huggingface.co/bert/bert-base-chinese-vocab.txt
tar xvf bert-base-chinese.tar.gz && rm bert-base-chinese.tar.gz
mv pytorch_model.bin pretrained_bert_model/
mv bert_config.json pretrained_bert_model/config.json
mv bert-base-chinese-vocab.txt pretrained_bert_model/vocab.txt

Step2:

Preprocess raw data into Pytorch Dataset.

python run.py --do_preprocess

Step3:

Begin training our model. Note: log file is in log and saved model is in save.

python run.py --do_train

Step4:

Predict model and parse the output of model to csv file.

python run.py --do_predict