memotion2020-code

Memotion Analysis 2020 Paper link: https://www.aclweb.org/anthology/2020.semeval-1.152.pdf This repo contains the base code used Team KAFK for the Memotion Analysis Task. Since the data is publicly posted on Kaggle, we have released two notebooks containing the code for our classifier and exploratory data analysis at Kaggle itself. Please find their links below.

Notebooks:

1. EDA Notebook: https://www.kaggle.com/coseck/memotion-eda
2. Classifier Notebook: https://www.kaggle.com/coseck/team-kafk-classification-models-for-task-a-and-b

Credits:

• RAdam : https://github.com/LiyuanLucasLiu/RAdam
• LookAhead: https://github.com/lonePatient/lookahead_pytorch
• Transformers: https://github.com/huggingface/transformers
• EffecientNet: https://github.com/lukemelas/EfficientNet-PyTorch

If you use our scripts please cite:

@inproceedings{das2020kafk,
  title={KAFK at SemEval-2020 Task 8: Extracting Features from Pre-trained Neural Networks to Classify Internet Memes},
  author={Das, Kaushik Amar and Baruah, Arup and Barbhuiya, Ferdous Ahmed and Dey, Kuntal},
  booktitle={Proceedings of the Fourteenth Workshop on Semantic Evaluation},
  pages={1148--1154},
  year={2020}
}