diff --git a/scripts/detection/faster_rcnn/train_faster_rcnn.py b/scripts/detection/faster_rcnn/train_faster_rcnn.py
index 52d2449e1e..d9d17464f0 100644
--- a/scripts/detection/faster_rcnn/train_faster_rcnn.py
+++ b/scripts/detection/faster_rcnn/train_faster_rcnn.py
@@ -531,12 +531,13 @@ def train(net, train_data, val_data, eval_metric, batch_size, ctx, args):
                         metric_losses[k].append(result[k])
                     for k in range(len(add_losses)):
                         add_losses[k].append(result[len(metric_losses) + k])
+            trainer.step(batch_size)
+
             for metric, record in zip(metrics, metric_losses):
                 metric.update(0, record)
             for metric, records in zip(metrics2, add_losses):
                 for pred in records:
                     metric.update(pred[0], pred[1])
-            trainer.step(batch_size)
 
             # update metrics
             if (not args.horovod or hvd.rank() == 0) and args.log_interval \
diff --git a/scripts/instance/mask_rcnn/train_mask_rcnn.py b/scripts/instance/mask_rcnn/train_mask_rcnn.py
index 36afa78ea0..e8100e21d3 100644
--- a/scripts/instance/mask_rcnn/train_mask_rcnn.py
+++ b/scripts/instance/mask_rcnn/train_mask_rcnn.py
@@ -10,7 +10,7 @@
 os.environ['MXNET_EXEC_BULK_EXEC_MAX_NODE_TRAIN_BWD'] = '25'
 os.environ['MXNET_GPU_COPY_NTHREADS'] = '1'
 os.environ['MXNET_OPTIMIZER_AGGREGATION_SIZE'] = '54'
-os.environ['MXNET_USE_FUSION'] = '0
+os.environ['MXNET_USE_FUSION'] = '0'
 
 import logging
 import time
@@ -559,6 +559,7 @@ def train(net, train_data, val_data, eval_metric, batch_size, ctx, logger, args)
             metric.reset()
         tic = time.time()
         btic = time.time()
+        speed = []
         train_data_iter = iter(train_data)
         next_data_batch = next(train_data_iter)
         next_data_batch = split_and_load(next_data_batch, ctx_list=ctx)
@@ -595,12 +596,14 @@ def train(net, train_data, val_data, eval_metric, batch_size, ctx, logger, args)
             except StopIteration:
                 pass
 
+            trainer.step(batch_size)
+
             for metric, record in zip(metrics, metric_losses):
                 metric.update(0, record)
             for metric, records in zip(metrics2, add_losses):
                 for pred in records:
                     metric.update(pred[0], pred[1])
-            trainer.step(batch_size)
+
             if (not args.horovod or hvd.rank() == 0) and args.log_interval \
                     and not (i + 1) % args.log_interval:
                 msg = ','.join(['{}={:.3f}'.format(*metric.get()) for metric in metrics + metrics2])