diff --git a/python/mxnet/kvstore.py b/python/mxnet/kvstore.py
index 190a9c0ad2e3..126c85c7ed9e 100644
--- a/python/mxnet/kvstore.py
+++ b/python/mxnet/kvstore.py
@@ -6,7 +6,7 @@
 import pickle
 from .ndarray import NDArray
 from .base import _LIB
-from .base import check_call, c_array, c_str, string_types, mx_uint, py_str, integer_types
+from .base import check_call, c_array, c_str, string_types, mx_uint, py_str
 from .base import NDArrayHandle, KVStoreHandle
 from . import optimizer as opt
 
diff --git a/python/mxnet/module/module.py b/python/mxnet/module/module.py
index 519499ebcf7f..7e42bfcbb111 100644
--- a/python/mxnet/module/module.py
+++ b/python/mxnet/module/module.py
@@ -8,7 +8,6 @@
 import warnings
 
 from .. import context as ctx
-from .. import ndarray as nd
 from .. import optimizer as opt
 
 from .executor_group import DataParallelExecutorGroup
diff --git a/python/mxnet/ndarray/op.py b/python/mxnet/ndarray/op.py
index 5e4c0855714f..525095b97205 100644
--- a/python/mxnet/ndarray/op.py
+++ b/python/mxnet/ndarray/op.py
@@ -7,7 +7,7 @@
 
 from ..ndarray_doc import _build_doc
 
-# Use different verison of SymbolBase
+# Use different version of SymbolBase
 # When possible, use cython to speedup part of computation.
 # pylint: disable=unused-import
 try:
diff --git a/python/mxnet/optimizer.py b/python/mxnet/optimizer.py
index cc6542b4d61f..d7c0574f962d 100644
--- a/python/mxnet/optimizer.py
+++ b/python/mxnet/optimizer.py
@@ -346,7 +346,8 @@ def create_state(self, index, weight):
                   "multi-precision doesn't supprot non-default weight yet"
             weight_master_copy = array(weight, ctx=weight.context, dtype=numpy.float32)
             if self.momentum != 0.0:
-                momentum = zeros(weight.shape, weight.context, dtype=numpy.float32, stype=weight.stype)
+                momentum = zeros(weight.shape, weight.context, dtype=numpy.float32,
+                                 stype=weight.stype)
             return (momentum, weight_master_copy)
         if weight.dtype == numpy.float16 and not self.multi_precision:
             warnings.warn("Accumulating with float16 in optimizer can lead to "