## For training (train a model from scratch)
call main.py
--lr: learning rate (ADAM optimiser)
--b1: beta1 (ADAM optimiser)
--b2: beta2 (ADAM optimiser)
--n_epoches: number of epoch for the training
--decay_epoch: decay for learning rate scheduler
--batch_size: batch size
--data_dir: path to data, the dir must contain two dir "photo_jpg" and "monet_jpg"
--tracking: flag, enable mlflow tracking
--track_name: name of the run for mlflow tracking
--profiling: flag, enable profiling with pytorch profiler
--amp: flag, enable mixed precision

example: 
main.py --lr 0.0005 --b1 0.5 --b2 0.996 --n_epoches 120 --decay_epoch 20 --batch_size 8 /
--data_dir ./datasets/monet_kaggle/ --tracking --track_name run_test --profiling  --amp


## For evaluation (transform every photo of "photo_jpg" directory to Money style)
--evaluation: enable evaluation
--model_path: path to trained model "its weights"
--data_dir: path to data, the dir must contain "photo_jpg" dir

example: 
main.py --evaluation --model_path ./model.pt --data_dir ./datasets/monet_kaggle/