SemEval2021: HaHackathon: Detecting and Rating Humor and Offense
- Developed and trained various machine learning models for Humor classification.
- Used various ML models from classical models to Transformer based models like BERT+BiGRU and achieved F1 of 0.94.
- Also developed a BERT+MLP model to give real number rating to the short text in the range 0-4, achieved a RMSE score of 0.446.