Recursive Semi-Markov Model

This is the source code for "N-ary Constituent Tree Parsing with Recursive Semi-Markov Model" from ACL 2021.

src_cn and src_en are the codes used in Chinese and English tasks respectively.

Contents

1. Requirements
2. Training
3. Citation
4. Credits

Requirements

• Python : 3.7 or higher.
• Cython : 0.25.2 or any compatible version.
• PyTorch : 1.6.0 or any compatible version.
• transformers : 3.0 or any compatible version.
• EVALB : run make in ./EVALB first to compile evalb for evaluation.

Training

Training requires cloning this repository from GitHub. You can download only one of src_cn and src_en for English or Chinese tasks.

Training Instructions

A new model can be trained using the command python main.py train ... in directory src_cn or src_cn.

The available arguments are listed in the training part at https://github.com/nikitakit/self-attentive-parser.

For each development evaluation, the F-score on the development set is computed and compared to the previous best. If the current model is better, the previous model will be deleted and the current model will be saved. The new filename will be derived from the provided model path base and the development F-score.

As an example, after setting the paths for data and embeddings, you can use the following command to train a model:

python main.py train 
    --use-bert 
    --model-path-base ../model/cn_bert_t1 
    --num-layers 2  
    --learning-rate 0.00001 
    --batch-size 32 
    --eval-batch-size 16 
    --subbatch-max-tokens 500

Evaluation Instructions

A saved model can be evaluated on a test corpus using the command python main.py test ... in directory src_cn or src_cn.

As an example, you can use the following command to evaluate a trained model:

python main.py test --model-path-base ../model/cn_bert_t1_dev=*.pt

Citation

If you are interested in our researches, please cite:

Xin Xin, Jinlong Li and Zeqi Tan. 2021. N-ary Constituent Tree Parsing with Recursive Semi-Markov Model. Accept by ACL-IJCNLP 2021.

Credits

Our code is developed on the base of the open-source Berkekey parser (Kitaev and Klein, 2018; Kitaev et al., 2019)

The code in this repository and portions of this README are based on https://github.com/nikitakit/self-attentive-parser