diff --git a/train.py b/train.py
index c3763fdc..57ce04fd 100644
--- a/train.py
+++ b/train.py
@@ -185,6 +185,7 @@ def train():
                                   hyp=hyp,  # augmentation hyperparameters
                                   rect=opt.rect,  # rectangular training
                                   image_weights=opt.img_weights,
+                                  cache_labels=True if epochs > 10 else False,
                                   cache_images=False if opt.prebias else opt.cache_images)
 
     # Dataloader
diff --git a/utils/datasets.py b/utils/datasets.py
index 6c94b375..0116d38e 100755
--- a/utils/datasets.py
+++ b/utils/datasets.py
@@ -246,7 +246,7 @@ class LoadStreams:  # multiple IP or RTSP cameras
 
 class LoadImagesAndLabels(Dataset):  # for training/testing
     def __init__(self, path, img_size=416, batch_size=16, augment=False, hyp=None, rect=True, image_weights=False,
-                 cache_images=False):
+                 cache_labels=False, cache_images=False):
         path = str(Path(path))  # os-agnostic
         with open(path, 'r') as f:
             self.img_files = [x.replace('/', os.sep) for x in f.read().splitlines()  # os-agnostic
@@ -305,7 +305,7 @@ class LoadImagesAndLabels(Dataset):  # for training/testing
         # Preload labels (required for weighted CE training)
         self.imgs = [None] * n
         self.labels = [None] * n
-        if augment or image_weights:  # cache labels for faster training
+        if cache_labels or image_weights:  # cache labels for faster training
             self.labels = [np.zeros((0, 5))] * n
             extract_bounding_boxes = False
             pbar = tqdm(self.label_files, desc='Reading labels')