diff --git a/TTS/bin/train_glow_tts.py b/TTS/bin/train_glow_tts.py
index 14a20149..072ad41b 100644
--- a/TTS/bin/train_glow_tts.py
+++ b/TTS/bin/train_glow_tts.py
@@ -497,6 +497,7 @@ def main(args):  # pylint: disable=redefined-outer-name
     criterion = GlowTTSLoss()
 
     if args.restore_path:
+        print(f" > Restoring from {os.path.basename(args.restore_path)} ...")
         checkpoint = torch.load(args.restore_path, map_location='cpu')
         try:
             # TODO: fix optimizer init, model.cuda() needs to be called before
@@ -514,7 +515,7 @@ def main(args):  # pylint: disable=redefined-outer-name
 
         for group in optimizer.param_groups:
             group['initial_lr'] = c.lr
-        print(" > Model restored from step %d" % checkpoint['step'],
+        print(f" > Model restored from step {checkpoint['step']:d}",
               flush=True)
         args.restore_step = checkpoint['step']
     else:
@@ -542,7 +543,8 @@ def main(args):  # pylint: disable=redefined-outer-name
         best_loss = float('inf')
         print(" > Starting with inf best loss.")
     else:
-        print(args.best_path)
+        print(" > Restoring best loss from "
+              f"{os.path.basename(args.best_path)} ...")
         best_loss = torch.load(args.best_path,
                                map_location='cpu')['model_loss']
         print(f" > Starting with loaded last best loss {best_loss}.")
diff --git a/TTS/bin/train_speedy_speech.py b/TTS/bin/train_speedy_speech.py
index 4e521451..8e9dbc2e 100644
--- a/TTS/bin/train_speedy_speech.py
+++ b/TTS/bin/train_speedy_speech.py
@@ -461,6 +461,7 @@ def main(args):  # pylint: disable=redefined-outer-name
     criterion = SpeedySpeechLoss(c)
 
     if args.restore_path:
+        print(f" > Restoring from {os.path.basename(args.restore_path)} ...")
         checkpoint = torch.load(args.restore_path, map_location='cpu')
         try:
             # TODO: fix optimizer init, model.cuda() needs to be called before
@@ -506,7 +507,8 @@ def main(args):  # pylint: disable=redefined-outer-name
         best_loss = float('inf')
         print(" > Starting with inf best loss.")
     else:
-        print(args.best_path)
+        print(" > Restoring best loss from "
+              f"{os.path.basename(args.best_path)} ...")
         best_loss = torch.load(args.best_path,
                                map_location='cpu')['model_loss']
         print(f" > Starting with loaded last best loss {best_loss}.")
diff --git a/TTS/bin/train_tacotron.py b/TTS/bin/train_tacotron.py
index cdc68c94..2382c0be 100644
--- a/TTS/bin/train_tacotron.py
+++ b/TTS/bin/train_tacotron.py
@@ -534,12 +534,13 @@ def main(args):  # pylint: disable=redefined-outer-name
     # setup criterion
     criterion = TacotronLoss(c, stopnet_pos_weight=c.stopnet_pos_weight, ga_sigma=0.4)
     if args.restore_path:
+        print(f" > Restoring from {os.path.basename(args.restore_path)}...")
         checkpoint = torch.load(args.restore_path, map_location='cpu')
         try:
-            print(" > Restoring Model.")
+            print(" > Restoring Model...")
             model.load_state_dict(checkpoint['model'])
             # optimizer restore
-            print(" > Restoring Optimizer.")
+            print(" > Restoring Optimizer...")
             optimizer.load_state_dict(checkpoint['optimizer'])
             if "scaler" in checkpoint and c.mixed_precision:
                 print(" > Restoring AMP Scaler...")
@@ -547,7 +548,7 @@ def main(args):  # pylint: disable=redefined-outer-name
             if c.reinit_layers:
                 raise RuntimeError
         except (KeyError, RuntimeError):
-            print(" > Partial model initialization.")
+            print(" > Partial model initialization...")
             model_dict = model.state_dict()
             model_dict = set_init_dict(model_dict, checkpoint['model'], c)
             # torch.save(model_dict, os.path.join(OUT_PATH, 'state_dict.pt'))
@@ -585,7 +586,8 @@ def main(args):  # pylint: disable=redefined-outer-name
         best_loss = float('inf')
         print(" > Starting with inf best loss.")
     else:
-        print(args.best_path)
+        print(" > Restoring best loss from "
+              f"{os.path.basename(args.best_path)} ...")
         best_loss = torch.load(args.best_path,
                                map_location='cpu')['model_loss']
         print(f" > Starting with loaded last best loss {best_loss}.")
diff --git a/TTS/bin/train_vocoder_gan.py b/TTS/bin/train_vocoder_gan.py
index ecc33288..9043a560 100644
--- a/TTS/bin/train_vocoder_gan.py
+++ b/TTS/bin/train_vocoder_gan.py
@@ -485,6 +485,7 @@ def main(args):  # pylint: disable=redefined-outer-name
     criterion_disc = DiscriminatorLoss(c)
 
     if args.restore_path:
+        print(f" > Restoring from {os.path.basename(args.restore_path)}...")
         checkpoint = torch.load(args.restore_path, map_location='cpu')
         try:
             print(" > Restoring Generator Model...")
@@ -523,7 +524,7 @@ def main(args):  # pylint: disable=redefined-outer-name
         for group in optimizer_disc.param_groups:
             group['lr'] = c.lr_disc
 
-        print(" > Model restored from step %d" % checkpoint['step'],
+        print(f" > Model restored from step {checkpoint['step']:d}",
               flush=True)
         args.restore_step = checkpoint['step']
     else:
@@ -549,10 +550,11 @@ def main(args):  # pylint: disable=redefined-outer-name
         best_loss = float('inf')
         print(" > Starting with inf best loss.")
     else:
-        print(args.best_path)
+        print(" > Restoring best loss from "
+              f"{os.path.basename(args.best_path)} ...")
         best_loss = torch.load(args.best_path,
                                map_location='cpu')['model_loss']
-        print(f" > Starting with loaded last best loss {best_loss}.")
+        print(f" > Starting with best loss of {best_loss}.")
     keep_best = c.get('keep_best', False)
     keep_after = c.get('keep_after', 10000)  # void if keep_best False
 
diff --git a/TTS/bin/train_vocoder_wavegrad.py b/TTS/bin/train_vocoder_wavegrad.py
index 7846aae5..271e8d4c 100644
--- a/TTS/bin/train_vocoder_wavegrad.py
+++ b/TTS/bin/train_vocoder_wavegrad.py
@@ -354,6 +354,7 @@ def main(args):  # pylint: disable=redefined-outer-name
         criterion.cuda()
 
     if args.restore_path:
+        print(f" > Restoring from {os.path.basename(args.restore_path)}...")
         checkpoint = torch.load(args.restore_path, map_location='cpu')
         try:
             print(" > Restoring Model...")
@@ -397,7 +398,8 @@ def main(args):  # pylint: disable=redefined-outer-name
         best_loss = float('inf')
         print(" > Starting with inf best loss.")
     else:
-        print(args.best_path)
+        print(" > Restoring best loss from "
+              f"{os.path.basename(args.best_path)} ...")
         best_loss = torch.load(args.best_path,
                                map_location='cpu')['model_loss']
         print(f" > Starting with loaded last best loss {best_loss}.")
diff --git a/TTS/bin/train_vocoder_wavernn.py b/TTS/bin/train_vocoder_wavernn.py
index 44ffef14..5fde5025 100644
--- a/TTS/bin/train_vocoder_wavernn.py
+++ b/TTS/bin/train_vocoder_wavernn.py
@@ -383,6 +383,7 @@ def main(args):  # pylint: disable=redefined-outer-name
 
     # restore any checkpoint
     if args.restore_path:
+        print(f" > Restoring from {os.path.basename(args.restore_path)}...")
         checkpoint = torch.load(args.restore_path, map_location="cpu")
         try:
             print(" > Restoring Model...")
@@ -420,7 +421,8 @@ def main(args):  # pylint: disable=redefined-outer-name
         best_loss = float('inf')
         print(" > Starting with inf best loss.")
     else:
-        print(args.best_path)
+        print(" > Restoring best loss from "
+              f"{os.path.basename(args.best_path)} ...")
         best_loss = torch.load(args.best_path,
                                map_location='cpu')['model_loss']
         print(f" > Starting with loaded last best loss {best_loss}.")
diff --git a/TTS/utils/arguments.py b/TTS/utils/arguments.py
index d05936dc..44345dd5 100644
--- a/TTS/utils/arguments.py
+++ b/TTS/utils/arguments.py
@@ -169,7 +169,6 @@ def process_args(args, model_type):
         args.restore_path, best_model = get_last_models(args.continue_path)
         if not args.best_path:
             args.best_path = best_model
-        print(f" > Training continues for {args.restore_path}")
 
     # setup output paths and read configs
     c = load_config(args.config_path)
@@ -184,8 +183,7 @@ def process_args(args, model_type):
     if model_class == "TTS":
         check_config_tts(c)
     elif model_class == "VOCODER":
-        print("Vocoder config checker not implemented, "
-              "skipping ...")
+        print("Vocoder config checker not implemented, skipping ...")
     else:
         raise ValueError(f"model type {model_type} not recognized!")