SanaNGU/Sentiment-Analysis-for-African-Languages

This Repo describe the code for SemEval 23 Task 12 : AfriSenti-SemEval Shared Task 12 ### To run the code for each task use >python run_train_ma.py --model_name --language --learning_rate --epoch --output_dir -- language can be one of 'am', 'dz' ,'ha', 'ig', 'kr' ,'ma' ,'pcm' ,'pt' ,'sw' ,'ts' ,'twi' ,'yo' example : > python run_train_ma.py --model_name 'alger-ia/dziribert' --language 'ma' --learning_rate 5e-5 --epoch 5.0 --output_dir 'alger-ia/dziribert-ma' ### To run the code for prediction on test set for each task use: >python run_predict_ma.py --model_name --language --output_dir example: > python run_predict_ma.py --model_name 'alger-ia/dziribert' --language 'ma' --output_dir 'alger-ia/dziribert-ma' ## The results for our (Masakhane team) can be found in : TaskA TaskB