This projects tries to determine the sentiment of the arabic tweets that written in colloquial Arabic.
Here I used a pre-trained models from hugging face and then fine tune them and chose a champion model.
This project is coded using pytorch framework and achieved the best score on kaggle private competition
This repo is good for starting dealing with arabic language
pip install gdown
pip install pyarabic
pip install pandas
pip install farasapy
pip install numpy
pip install emoji
pip install transformers
pip install nltk
# Install torch with cuda 11.3
pip3 install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu113
- Data cleaning
- Text embedding
- Use machine learning algorithms
- Use pretrained models and finetune them