diff --git a/flair/trainers/trainer.py b/flair/trainers/trainer.py
index 257ac54606..4c5e13df2b 100644
--- a/flair/trainers/trainer.py
+++ b/flair/trainers/trainer.py
@@ -88,16 +88,16 @@ def find_learning_rate(self,
                 f.write(f'{itr}\t{datetime.datetime.now():%H:%M:%S}\t{learning_rate}\t{loss_item}\n')
 
             if stop_early and loss_item > 4 * best_loss:
-                log_line()
+                log_line(log)
                 log.info('loss diverged - stopping early!')
                 break
 
         self.model.load_state_dict(model_state)
         self.model.to(model_device)
 
-        log_line()
+        log_line(log)
         log.info(f'learning rate finder finished - plot {learning_rate_tsv}')
-        log_line()
+        log_line(log)
 
         return Path(learning_rate_tsv)
 
diff --git a/flair/training_utils.py b/flair/training_utils.py
index be071570b5..3be4d8e43e 100644
--- a/flair/training_utils.py
+++ b/flair/training_utils.py
@@ -9,9 +9,6 @@
 from functools import reduce
 
 
-log = logging.getLogger(__name__)
-
-
 class Metric(object):
 
     def __init__(self, name):
@@ -232,8 +229,5 @@ def convert_labels_to_one_hot(label_list: List[List[str]], label_dict: Dictionar
     return [[1 if l in labels else 0 for l in label_dict.get_items()] for labels in label_list]
 
 
-def log_line(logger = None):
-    if logger is not None:
-        logger.info('-' * 100)
-    else:
-        log.info('-' * 100)
+def log_line(logger):
+    logger.info('-' * 100)