Update config.py

Update learning rate and weight decay

config.py

@@ -41,7 +41,7 @@ class Config(object):
     # Validation stats are also calculated at each epoch end and they
     # might take a while, so don't set this too small to avoid spending
     # a lot of time on validation stats.
-    STEPS_PER_EPOCH = 1000
+    STEPS_PER_EPOCH = 670
 
     # Number of validation steps to run at the end of every training epoch.
     # A bigger number improves accuracy of validation stats, but slows
@@ -87,8 +87,8 @@ class Config(object):
     # Images are resized such that the smallest side is >= IMAGE_MIN_DIM and
     # the longest side is <= IMAGE_MAX_DIM. In case both conditions can't
     # be satisfied together the IMAGE_MAX_DIM is enforced.
-    IMAGE_MIN_DIM = 800
-    IMAGE_MAX_DIM = 1024
+    IMAGE_MIN_DIM = 512
+    IMAGE_MAX_DIM = 512
     # If True, pad images with zeros such that they're (max_dim by max_dim)
     IMAGE_PADDING = True  # currently, the False option is not supported
 
@@ -111,7 +111,7 @@ class Config(object):
     MASK_SHAPE = [28, 28]
 
     # Maximum number of ground truth instances to use in one image
-    MAX_GT_INSTANCES = 512
+    MAX_GT_INSTANCES = 256
 
     # Bounding box refinement standard deviation for RPN and final detections.
     RPN_BBOX_STD_DEV = np.array([0.1, 0.1, 0.2, 0.2])
@@ -131,11 +131,11 @@ class Config(object):
     # The Mask RCNN paper uses lr=0.02, but on TensorFlow it causes
     # weights to explode. Likely due to differences in optimzer
     # implementation.
-    LEARNING_RATE = 0.0001
+    LEARNING_RATE = 0.001
     LEARNING_MOMENTUM = 0.9
 
     # Weight decay regularization
-    WEIGHT_DECAY = 0.001
+    WEIGHT_DECAY = 0.0001
 
     # Use RPN ROIs or externally generated ROIs for training
     # Keep this True for most situations. Set to False if you want to train