/CofiPara

Primary LanguagePython

CofiPara

Official PyTorch implementation for the paper - CofiPara: A Coarse-to-fine Paradigm for Multimodal Sarcasm Target Identification with Large Multimodal Models.

(ACL 2024: The 62nd Annual Meeting of the Association for Computational Linguistics, August 2024, Bangkok.) [paper]

Environment

python==3.10
torch==2.0.1
(pip install -r requirements.txt)

Please also install groudingdino in GroundingDINO repository.

Dataset

Please refer to MSD_Data and MSTI_Data

Training and Test

Edit model_config.py to change training and test settings.

python main.py

Citation

@inproceedings{cofipara2024msti,
  title={CofiPara: A Coarse-to-fine Paradigm for Multimodal Sarcasm Target Identification with Large Multimodal Models},
  author={Lin, Hongzhan and Chen, Zixin and
    Luo, Ziyang and
    Cheng, Mingfei and
    Ma, Jing and
    Chen, Guang},
  booktitle={Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year={2024}
}