Sarcasm Detection using RoBERTa

URL: https://github.com/zabir-nabil/sarcasm-detection-roberta

Description

This repository explores sarcasm detection using various models, including LSTM, GRU, and RoBERTa. The models are trained on datasets such as SARC (reddit), sarcasm_v2, and iSARCASM (twitter).

Methods and Models

• Models Used:
  1. LSTM (Long Short-Term Memory) (Bi-directional)
  2. GRU (Gated Recurrent Unit)
  3. RoBERTa
• Datasets Used:
  1. SARC (Reddit)
  2. sarcasm_v2
  3. iSARCASM (Twitter)

Notebooks

• Sarcasm Detection using GRU and LSTM (Bi-directional) on SARC/sarcasm_v2 dataset (Keras/TensorFlow)
• Sarcasm Detection using RoBERTa on SARC/sarcasm_v2 dataset (PyTorch)

Results

• No specific results or performance metrics provided in the repository.