Finetuning for Sarcasm Detection using a Pruned Dataset

Models

All the models are available in Models directory. In addition to models reported in the paper, we have implented XLNet and Electra models, which are available in in the models directory.

To run the models, first make sure to install the requirements using below command:

pip install -r requirements.txt

Then run each models by using calling python for the model name like:

python ./Models/model-name/model-name.py

Download SARC (train-balanced-sarcasm.csv) from: https://www.kaggle.com/datasets/sherinclaudia/sarcastic-comments-on-reddit

To Run Inference on SARC

python ./Models/RoBERTa/RoBERTa_Reddit_Ranking.py

To Reproduce best results in the paper

python ./Models/RoBERTa/RoBERTa.py

Codebase is fork of: https://github.com/AmirAbaskohi/SemEval2022-Task6-Sarcasm-Detection ; Read their paper: https://arxiv.org/pdf/2204.08198.pdf