README.md

Effective Cascade Dual-Decoder Model for Joint Entity and RelationExtraction

Effective Cascade Dual-Decoder Model for Joint Entity and RelationExtraction. [submitted]

Requirements

All experiments are conducted with an NVDIA GeForce RTX 2080 Ti.

The main requirements are:

• python = 3.6
• torch = 1.1.0
• transformers = 3.5.1 (Online)

Usage

Training

1. Partial Match:

   python train.py --data_dir=dataset/WebNLG-P/data --id=WebNLG-P --classemb_num=214 --entityclass_num=2 --relationclass_num=171

2. Exact Match:

   python othertrain.py --data_dir=dataset/WebNLG-E/data --id=WebNLG-E --classemb_num=255 --entityclass_num=2 --relationclass_num=211

Testing

1. Partial Match:

   python othereval.py --model_dir=./saved_models/WebNLG-P --data_dir=dataset/WebNLG-P/data

2. Exact Match:

   python eval.py --model_dir=./saved_models/WebNLG-E --data_dir=dataset/WebNLG-E/data

Related Repo

Codes are adapted from the repositories of Joint Extraction of Entities and Relations Based on a Novel Decomposition and A Novel Cascade Binary Tagging Framework for Relational Triple Extraction.