add log_f1 argument

Author: ahmedbesbes

train.py

@@ -78,13 +78,6 @@ def train(model, training_generator, optimizer, criterion, epoch, writer, log_fi
         lr = optimizer.state_dict()["param_groups"][0]["lr"]
 
         if (iter % print_every == 0) and (iter > 0):
-            intermediate_report = classification_report(
-                y_true, y_pred, output_dict=True)
-
-            f1_by_class = 'F1 Scores by class: '
-            for class_name in class_names:
-                f1_by_class += f"{class_name} : {np.round(intermediate_report[class_name]['f1-score'], 4)} |"
-
             print("[Training - Epoch: {}], LR: {} , Iteration: {}/{} , Loss: {}, Accuracy: {}".format(
                 epoch + 1,
                 lr,
@@ -93,7 +86,16 @@ def train(model, training_generator, optimizer, criterion, epoch, writer, log_fi
                 losses.avg,
                 accuracies.avg
             ))
-            print(f1_by_class)
+
+            if bool(args.log_f1):
+                intermediate_report = classification_report(
+                    y_true, y_pred, output_dict=True)
+
+                f1_by_class = 'F1 Scores by class: '
+                for class_name in class_names:
+                    f1_by_class += f"{class_name} : {np.round(intermediate_report[class_name]['f1-score'], 4)} |"
+
+                print(f1_by_class)
 
     f1_train = f1_score(y_true, y_pred, average='weighted')
 
@@ -403,6 +405,7 @@ def run(args, both_cases=False):
     parser.add_argument('--workers', type=int, default=1)
     parser.add_argument('--log_path', type=str, default='./logs/')
     parser.add_argument('--log_every', type=int, default=100)
+    parser.add_argument('--log_f1', type=int, default=1, choices=[0, 1])
     parser.add_argument('--flush_history', type=int,
                         default=1, choices=[0, 1])
     parser.add_argument('--output', type=str, default='./models/')