This contains code for the NLP-letters project.
Check ben-yu and TomOrth folders have individual scripts. Check their respective readmes for information
For shallow models:
- python final_experiments.py -model svm -dataset full_text -preprocessing replace
- python final_experiments.py -model svm -dataset s1_s2 -preprocessing replace
- python final_experiments.py -model rf -dataset full_text -preprocessing replace
- python final_experiments.py -model rf -dataset s1_s2 -preprocessing replace
For BERT training, use final_bert_training.py or Final_BERT_Training notebook in Colab and comment and uncomment the sections of the script/notebook you need to run for the specific model.
- Install Anaconda https://docs.anaconda.com/free/anaconda/install/
- Create conda env
conda env create -f environment.yml - Activate environment
conda activate nlp-letters