English | 简体中文

python == 3.8

• torch == 1.5.0
• transformers == 3.4.0
• opt-einsum == 3.3.0
• ujson
• deepke

git clone https://github.com/zjunlp/DeepKE.git
cd DeepKE/example/re/document

• Create and enter the python virtual environment.
• Install dependencies: pip install -r requirements.txt.

• Dataset

  • Download the dataset to this directory.

    wget 121.41.117.246:8080/Data/re/document/data.tar.gz
    tar -xzvf data.tar.gz

  • The dataset DocRED is stored in data:

    • dev.json：Validation set

    • rel_info.json：Relation set

    • rel2id.json：Relation labels - ID

    • test.json：Test set

    • train_annotated.json：Training set annotated manually

    • train_distant.json: Training set generated by distant supervision

• Training

  • Parameters, model paths and configuration for training are in the conf folder and users can modify them before training.

  • Training on DocRED

    python run.py

  • The trained model is stored in the current directory by default.

  • Start to train from last-trained model
    

    modify train_from_saved_model in .yaml as the path of the last-trained model

  • Logs for training are stored in the current directory by default and the path can be configured by modifying log_dir in .yaml

• Prediction

  python predict.py

  • After prediction, generated result.json is stored in the current directory

DocuNet