See models/ for each respective models' training and inference code. All models are implemented in PyTorch, referencing the original implementations as described in the paper.
The code for the four encoder models is at models/encoder_based/. Encoder-based models are trained using PyTorch Lightning.
Our implementation of the LinkAppend model is available at models/decoder_based/LinkAppend/. In particular, for details on training and inference see the README. The LinkAppend model can be trained using HuggingFace's Trainer.
As mentioned in the paper, we include a full precision recall breakdown for each F1 score. See results/precision_recall.xlsx.
This repo includes the raw model code. Models implemented by @XiyuanZou and @ianporada.
For details, see the paper:
@inproceedings{porada-etal-2024-controlled-reevaluation,
title = "A Controlled Reevaluation of Coreference Resolution Models",
author = "Porada, Ian and
Zou, Xiyuan and
Cheung, Jackie Chi Kit",
booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)",
month = may,
year = "2024",
address = "Torino, Italia",
publisher = "ELRA and ICCL",
url = "https://aclanthology.org/2024.lrec-main.23",
pages = "256--263",
}