Skip to content

Latest commit

 

History

History
69 lines (51 loc) · 1.53 KB

README.md

File metadata and controls

69 lines (51 loc) · 1.53 KB
Raw

RAP for DEGREE

The code is basically borrowed from here, with slight modification.

Requirements

General

  • Python (verified on 3.8)

Python Packages

  • see requirements.txt
conda create -n degree python=3.8
conda activate degree
pip install -r requirements.txt

Dataset

Preprocessing

ace05e

  1. Prepare data processed from DyGIE++
  2. Put the processed data into the folder processed_data/ace05e_dygieppformat
  3. Run ./scripts/process_ace05e.sh

The data folder should be like following after preprocessing

processed_data/ace05e_bart
├── dev.w1.oneie.json
├── test.w1.oneie.json
├── train.001.w1.oneie.json
├── train.003.w1.oneie.json
├── train.005.w1.oneie.json
├── ...
└── train.w1.oneie.json

Retrieving

Run the following command to generate retrieved reference for dataset.

cd ../../retrieval
python retrieve.py --base_model degree

And the retrieved data will be in the folder processed_data/retrieved

Training

Generate data

python degree/generate_data_degree_e2e.py -c config/config_degree_e2e_ace05e.json

train

python degree/train_degree_e2e.py -c config/config_degree_e2e_ace05e.json

The model will be stored at ./output/degree_e2e_ace05e/[timestamp]/best_model.mdl in default.

Evaluation

Evaluate on the model

python degree/eval_end2endEE.py -c config/config_degree_e2e_ace05e.json -e [e2e_model]