diff --git a/dataset/dataset.py b/dataset/dataset.py
index 5119a732aa05708a4dbec17f85de3a05eec1dee0..e06b3f1febd7768d3a447cedd93c99aae1b22734 100644
--- a/dataset/dataset.py
+++ b/dataset/dataset.py
@@ -219,9 +219,9 @@ class ImageFolderDuo(data.Dataset):
 
 def load_data_duo(base_dir, batch_size, args, shuffle=True):
     train_transform = transforms.Compose(
-        [Random_erasing(args.augment_args[0], args.augment_args[3]),
-         Random_int_noise(args.augment_args[1], args.augment_args[4]),
-         Random_shift_rt(args.augment_args[2], args.augment_args[5], args.augment_args[6]),
+        [#Random_erasing(args.augment_args[0], args.augment_args[3]),
+         #Random_int_noise(args.augment_args[1], args.augment_args[4]),
+         #Random_shift_rt(args.augment_args[2], args.augment_args[5], args.augment_args[6]),
          transforms.Resize((224, 224)),
          Threshold_noise(args.noise_threshold),
          Log_normalisation(),
diff --git a/main.py b/main.py
index 6aa0b89809a21b52e9f2e7c924533ef89aae8052..2cd28d3050f8ab7ae463989b269f1b2bfc129394 100644
--- a/main.py
+++ b/main.py
@@ -254,10 +254,10 @@ def run_duo(args):
     plt.plot(train_loss)
     plt.plot(train_loss)
 
-    plt.savefig(f'output/training_plot_model_{args.model}_noise_{args.noise_threshold}_lr_{args.lr}_optim_{args.optim + ("_momentum_"+str(args.momentum) if args.optim=="SGD" else "_betas_" + str(args.beta1)+ "_" +str(args.beta2))}.png')
+    plt.savefig(f'output/Atraining_plot_model_{args.model}_noise_{args.noise_threshold}_lr_{args.lr}_optim_{args.optim + ("_momentum_"+str(args.momentum) if args.optim=="SGD" else "_betas_" + str(args.beta1)+ "_" +str(args.beta2))}.png')
     #load and evaluate best model
     load_model(model, args.save_path)
-    make_prediction_duo(model,data_test, f'output/model_{args.model}_noise_{args.noise_threshold}_lr_{args.lr}_optim_{args.optim + ("_momentum_"+str(args.momentum) if args.optim=="SGD" else "_betas_" + str(args.beta1)+ "_" +str(args.beta2))}.png')
+    make_prediction_duo(model,data_test, f'output/Amodel_{args.model}_noise_{args.noise_threshold}_lr_{args.lr}_optim_{args.optim + ("_momentum_"+str(args.momentum) if args.optim=="SGD" else "_betas_" + str(args.beta1)+ "_" +str(args.beta2))}.png')
     return best_loss,best_acc
 
 def make_prediction_duo(model, data, f_name):