train_argument_extraction_bart.sh

python src/train_argument_extraction.py  --lr 2e-5 \
                              --batch_size 8 --gpu 0 \
                              --gradient_accumulation_steps 4 \
                              --opt 'adam' --multi_arg \
                              --transformer_name 'bart-large' \
                              --qg_model_path './model_checkpoint/qg_model_bart' \
                              --models_dir './model_checkpoint/eae_model_bart'