diff --git a/mindspore/context.py b/mindspore/context.py
index 985270d1fa..361c48c09b 100644
--- a/mindspore/context.py
+++ b/mindspore/context.py
@@ -28,7 +28,7 @@ from mindspore.parallel._auto_parallel_context import _set_auto_parallel_context
     _reset_auto_parallel_context
 
 __all__ = ['GRAPH_MODE', 'PYNATIVE_MODE', 'set_context', 'get_context', 'set_auto_parallel_context',
-           'get_auto_parallel_context', 'reset_auto_parallel_context']
+           'get_auto_parallel_context', 'reset_auto_parallel_context', 'ParallelMode']
 
 GRAPH_MODE = 0
 PYNATIVE_MODE = 1
@@ -647,3 +647,26 @@ def get_context(attr_key):
         raise ValueError(
             "Get context keyword %s is not recognized!" % attr_key)
     return getattr(_context(), attr_key)
+
+class ParallelMode:
+    """
+    Parallel mode options.
+
+    There are five kinds of parallel modes, "STAND_ALONE", "DATA_PARALLEL",
+    "HYBRID_PARALLEL", "SEMI_AUTO_PARALLEL" and "AUTO_PARALLEL". Default: "STAND_ALONE".
+
+        - STAND_ALONE: Only one processor working.
+        - DATA_PARALLEL: Distributing the data across different processors.
+        - HYBRID_PARALLEL: Achieving data parallelism and model parallelism manually.
+        - SEMI_AUTO_PARALLEL: Achieving data parallelism and model parallelism by setting parallel strategies.
+        - AUTO_PARALLEL: Achieving parallelism automatically.
+
+    MODE_LIST: The list for all supported parallel modes.
+    """
+
+    STAND_ALONE = "stand_alone"
+    DATA_PARALLEL = "data_parallel"
+    HYBRID_PARALLEL = "hybrid_parallel"
+    SEMI_AUTO_PARALLEL = "semi_auto_parallel"
+    AUTO_PARALLEL = "auto_parallel"
+    MODE_LIST = [STAND_ALONE, DATA_PARALLEL, HYBRID_PARALLEL, SEMI_AUTO_PARALLEL, AUTO_PARALLEL]
diff --git a/mindspore/nn/layer/embedding.py b/mindspore/nn/layer/embedding.py
index f5be4d9a25..b721df626f 100755
--- a/mindspore/nn/layer/embedding.py
+++ b/mindspore/nn/layer/embedding.py
@@ -20,7 +20,7 @@ from mindspore.common.parameter import Parameter
 from mindspore.common.initializer import initializer
 from mindspore._checkparam import Validator
 from mindspore.communication.management import get_group_size
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.parallel._utils import _get_parallel_mode
 from ..cell import Cell
 from ..._checkparam import Validator as validator, Rel
@@ -129,9 +129,9 @@ class EmbeddingLookup(Cell):
         embedding_size (int): The size of each embedding vector.
         param_init (str): The initialize way of embedding table. Default: 'normal'.
         target (str): Specify the target where the op is executed. The value should in
-        ['DEVICE', 'CPU']. Default: 'CPU'.
-        slice_mode (str): The slicing way in semi auto parallel/auto parallel. The value should get through
-        nn.EmbeddingLookUpSplitMode. Default: 'batch_slice'.
+            ['DEVICE', 'CPU']. Default: 'CPU'.
+        slice_mode (str): The slicing way in semi_auto_parallel/auto_parallel. The value should get through
+            nn.EmbeddingLookUpSplitMode. Default: nn.EmbeddingLookUpSplitMode.BATCH_SLICE.
         manual_shapes (tuple): The accompaniment array in field slice mode.
 
     Inputs:
diff --git a/mindspore/nn/optim/optimizer.py b/mindspore/nn/optim/optimizer.py
index 7cbc9fa097..c60498df21 100755
--- a/mindspore/nn/optim/optimizer.py
+++ b/mindspore/nn/optim/optimizer.py
@@ -29,7 +29,7 @@ from mindspore._checkparam import Validator as validator
 from mindspore._checkparam import Rel
 from mindspore import log as logger
 from mindspore.parallel._utils import _get_global_rank, _get_device_num, _get_parallel_mode
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore import context
 from mindspore.nn.learning_rate_schedule import LearningRateSchedule
 
diff --git a/mindspore/nn/wrap/cell_wrapper.py b/mindspore/nn/wrap/cell_wrapper.py
index c32b7eadd9..d24c166caf 100644
--- a/mindspore/nn/wrap/cell_wrapper.py
+++ b/mindspore/nn/wrap/cell_wrapper.py
@@ -15,7 +15,7 @@
 """Cell_wrapper."""
 from mindspore.parallel._utils import (_get_device_num, _get_mirror_mean,
                                        _get_parallel_mode)
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from ...common import dtype as mstype
 from ...common.parameter import Parameter, ParameterTuple
 from ...ops import composite as C
diff --git a/mindspore/nn/wrap/grad_reducer.py b/mindspore/nn/wrap/grad_reducer.py
index 8e8a9ef756..345854a858 100644
--- a/mindspore/nn/wrap/grad_reducer.py
+++ b/mindspore/nn/wrap/grad_reducer.py
@@ -251,8 +251,9 @@ class DistributedGradReducer(Cell):
         >>> from mindspore.ops import operations as P
         >>> from mindspore.ops import functional as F
         >>> from mindspore import context
+        >>> from mindspore.context import ParallelMode
         >>> from mindspore import nn
-        >>> from mindspore import ParallelMode, ParameterTuple
+        >>> from mindspore import ParameterTuple
         >>>
         >>> device_id = int(os.environ["DEVICE_ID"])
         >>> context.set_context(mode=context.GRAPH_MODE, device_target="Ascend", save_graphs=True,
diff --git a/mindspore/nn/wrap/loss_scale.py b/mindspore/nn/wrap/loss_scale.py
index 88a0f26a34..999873ea6e 100644
--- a/mindspore/nn/wrap/loss_scale.py
+++ b/mindspore/nn/wrap/loss_scale.py
@@ -15,7 +15,7 @@
 """Loss scale cell for loss scale training."""
 import mindspore.context as context
 from mindspore.nn.wrap.grad_reducer import DistributedGradReducer
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.parallel._utils import _get_device_num, _get_parallel_mode, _get_mirror_mean
 from ..cell import Cell
 from ...common import Tensor, RowTensor
diff --git a/mindspore/train/__init__.py b/mindspore/train/__init__.py
index 1895743e20..ea5910be07 100644
--- a/mindspore/train/__init__.py
+++ b/mindspore/train/__init__.py
@@ -18,8 +18,7 @@ High-Level training interfaces.
 Helper functions in train piplines.
 """
 from .model import Model
-from .parallel_utils import ParallelMode
 from .dataset_helper import DatasetHelper
 from . import amp
 
-__all__ = ["Model", "ParallelMode", "DatasetHelper", "amp"]
+__all__ = ["Model", "DatasetHelper", "amp"]
diff --git a/mindspore/train/amp.py b/mindspore/train/amp.py
index e2da1618bf..edbb49894f 100644
--- a/mindspore/train/amp.py
+++ b/mindspore/train/amp.py
@@ -23,7 +23,7 @@ from ..nn.wrap.cell_wrapper import _VirtualDatasetCell
 from ..ops import functional as F
 from ..parallel._utils import _get_parallel_mode
 from .loss_scale_manager import DynamicLossScaleManager, LossScaleManager
-from .parallel_utils import ParallelMode
+from ..context import ParallelMode
 from .. import context
 
 __all__ = ["build_train_network"]
diff --git a/mindspore/train/model.py b/mindspore/train/model.py
index 7e28eb9704..75f1638a8f 100755
--- a/mindspore/train/model.py
+++ b/mindspore/train/model.py
@@ -30,7 +30,7 @@ from ..parallel._utils import _get_parallel_mode, _get_device_num, _get_global_r
 from ..nn.metrics import Loss
 from .. import nn
 from ..nn.wrap.cell_wrapper import _VirtualDatasetCell
-from .parallel_utils import ParallelMode
+from ..context import ParallelMode
 from ..parallel._utils import _need_to_full, _to_full_tensor
 from ..common import dtype as mstype
 from .dataset_helper import DatasetHelper
diff --git a/mindspore/train/parallel_utils.py b/mindspore/train/parallel_utils.py
deleted file mode 100644
index 4f460a5d98..0000000000
--- a/mindspore/train/parallel_utils.py
+++ /dev/null
@@ -1,41 +0,0 @@
-# Copyright 2020 Huawei Technologies Co., Ltd
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-# http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-# ============================================================================
-"""Parallel utils"""
-
-__all__ = ["ParallelMode"]
-
-
-class ParallelMode:
-    """
-    Parallel mode options.
-
-    There are five kinds of parallel modes, "STAND_ALONE", "DATA_PARALLEL",
-    "HYBRID_PARALLEL", "SEMI_AUTO_PARALLEL" and "AUTO_PARALLEL". Default: "STAND_ALONE".
-
-        - STAND_ALONE: Only one processor working.
-        - DATA_PARALLEL: Distributing the data across different processors.
-        - HYBRID_PARALLEL: Achieving data parallelism and model parallelism manually.
-        - SEMI_AUTO_PARALLEL: Achieving data parallelism and model parallelism by setting parallel strategies.
-        - AUTO_PARALLEL: Achieving parallelism automatically.
-
-    MODE_LIST: The list for all supported parallel modes.
-    """
-
-    STAND_ALONE = "stand_alone"
-    DATA_PARALLEL = "data_parallel"
-    HYBRID_PARALLEL = "hybrid_parallel"
-    SEMI_AUTO_PARALLEL = "semi_auto_parallel"
-    AUTO_PARALLEL = "auto_parallel"
-    MODE_LIST = [STAND_ALONE, DATA_PARALLEL, HYBRID_PARALLEL, SEMI_AUTO_PARALLEL, AUTO_PARALLEL]
diff --git a/model_zoo/official/cv/deeplabv3/train.py b/model_zoo/official/cv/deeplabv3/train.py
index 56ef5b02bb..0269d5f540 100644
--- a/model_zoo/official/cv/deeplabv3/train.py
+++ b/model_zoo/official/cv/deeplabv3/train.py
@@ -17,7 +17,8 @@ import argparse
 from mindspore import context
 from mindspore.communication.management import init
 from mindspore.nn.optim.momentum import Momentum
-from mindspore import Model, ParallelMode
+from mindspore import Model
+from mindspore.context import ParallelMode
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
 from mindspore.train.callback import Callback, CheckpointConfig, ModelCheckpoint, TimeMonitor
 from src.md_dataset import create_dataset
diff --git a/model_zoo/official/cv/faster_rcnn/train.py b/model_zoo/official/cv/faster_rcnn/train.py
index 895ac725cc..a6d09874f9 100644
--- a/model_zoo/official/cv/faster_rcnn/train.py
+++ b/model_zoo/official/cv/faster_rcnn/train.py
@@ -26,7 +26,8 @@ import mindspore.common.dtype as mstype
 from mindspore import context, Tensor
 from mindspore.communication.management import init
 from mindspore.train.callback import CheckpointConfig, ModelCheckpoint, TimeMonitor
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
 from mindspore.nn import SGD
 import mindspore.dataset.engine as de
diff --git a/model_zoo/official/cv/googlenet/train.py b/model_zoo/official/cv/googlenet/train.py
index 5181f9c484..442a4262e5 100644
--- a/model_zoo/official/cv/googlenet/train.py
+++ b/model_zoo/official/cv/googlenet/train.py
@@ -28,7 +28,8 @@ from mindspore import context
 from mindspore.communication.management import init, get_rank
 from mindspore.nn.optim.momentum import Momentum
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor, TimeMonitor
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
 
 from src.config import cifar_cfg as cfg
diff --git a/model_zoo/official/cv/inceptionv3/train.py b/model_zoo/official/cv/inceptionv3/train.py
index f2d2256eef..db7a0d0854 100644
--- a/model_zoo/official/cv/inceptionv3/train.py
+++ b/model_zoo/official/cv/inceptionv3/train.py
@@ -21,7 +21,7 @@ import numpy as np
 import mindspore.nn as nn
 from mindspore import Tensor
 from mindspore import context
-from mindspore import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.communication.management import init, get_rank, get_group_size
 from mindspore.nn.optim.rmsprop import RMSProp
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor, TimeMonitor
diff --git a/model_zoo/official/cv/maskrcnn/train.py b/model_zoo/official/cv/maskrcnn/train.py
index 8df2fab00e..06c5e87b4c 100644
--- a/model_zoo/official/cv/maskrcnn/train.py
+++ b/model_zoo/official/cv/maskrcnn/train.py
@@ -24,7 +24,8 @@ import mindspore.common.dtype as mstype
 from mindspore import context, Tensor
 from mindspore.communication.management import init
 from mindspore.train.callback import CheckpointConfig, ModelCheckpoint, TimeMonitor
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
 from mindspore.nn import SGD
 import mindspore.dataset.engine as de
diff --git a/model_zoo/official/cv/mobilenetv2/train.py b/model_zoo/official/cv/mobilenetv2/train.py
index 4fb800d6dd..02935921e3 100644
--- a/model_zoo/official/cv/mobilenetv2/train.py
+++ b/model_zoo/official/cv/mobilenetv2/train.py
@@ -30,7 +30,8 @@ from mindspore.nn.loss.loss import _Loss
 from mindspore.ops import operations as P
 from mindspore.ops import functional as F
 from mindspore.common import dtype as mstype
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, Callback
 from mindspore.train.loss_scale_manager import FixedLossScaleManager
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
diff --git a/model_zoo/official/cv/mobilenetv2_quant/train.py b/model_zoo/official/cv/mobilenetv2_quant/train.py
index ebe60996cf..0674625382 100644
--- a/model_zoo/official/cv/mobilenetv2_quant/train.py
+++ b/model_zoo/official/cv/mobilenetv2_quant/train.py
@@ -22,7 +22,8 @@ import numpy as np
 from mindspore import context
 from mindspore import Tensor
 from mindspore import nn
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 from mindspore.train.loss_scale_manager import FixedLossScaleManager
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig
 from mindspore.train.serialization import load_checkpoint
diff --git a/model_zoo/official/cv/mobilenetv3/train.py b/model_zoo/official/cv/mobilenetv3/train.py
index 60f3723244..5bd8c974c3 100644
--- a/model_zoo/official/cv/mobilenetv3/train.py
+++ b/model_zoo/official/cv/mobilenetv3/train.py
@@ -28,7 +28,8 @@ from mindspore.nn.loss.loss import _Loss
 from mindspore.ops import operations as P
 from mindspore.ops import functional as F
 from mindspore.common import dtype as mstype
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, Callback
 from mindspore.train.loss_scale_manager import FixedLossScaleManager
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
diff --git a/model_zoo/official/cv/resnet/train.py b/model_zoo/official/cv/resnet/train.py
index 0a891b9163..addb865573 100755
--- a/model_zoo/official/cv/resnet/train.py
+++ b/model_zoo/official/cv/resnet/train.py
@@ -22,7 +22,8 @@ from mindspore import Tensor
 from mindspore import dataset as de
 from mindspore.parallel._auto_parallel_context import auto_parallel_context
 from mindspore.nn.optim.momentum import Momentum
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor, TimeMonitor
 from mindspore.nn.loss import SoftmaxCrossEntropyWithLogits
 from mindspore.train.loss_scale_manager import FixedLossScaleManager
diff --git a/model_zoo/official/cv/resnet50_quant/train.py b/model_zoo/official/cv/resnet50_quant/train.py
index 2e13ec37ff..553597f4c0 100755
--- a/model_zoo/official/cv/resnet50_quant/train.py
+++ b/model_zoo/official/cv/resnet50_quant/train.py
@@ -21,7 +21,8 @@ from mindspore import context
 from mindspore import Tensor
 from mindspore.parallel._auto_parallel_context import auto_parallel_context
 from mindspore.nn.optim.momentum import Momentum
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor, TimeMonitor
 from mindspore.train.loss_scale_manager import FixedLossScaleManager
 from mindspore.train.serialization import load_checkpoint
diff --git a/model_zoo/official/cv/resnet_thor/src/grad_reducer_thor.py b/model_zoo/official/cv/resnet_thor/src/grad_reducer_thor.py
index 86ee3fcc8f..35cbaa1460 100644
--- a/model_zoo/official/cv/resnet_thor/src/grad_reducer_thor.py
+++ b/model_zoo/official/cv/resnet_thor/src/grad_reducer_thor.py
@@ -102,7 +102,8 @@ class DistributedGradReducerThor(Cell):
         >>> from mindspore.ops import functional as F
         >>> from mindspore import context
         >>> from mindspore import nn
-        >>> from mindspore import ParallelMode, ParameterTuple
+        >>> from mindspore import ParameterTuple
+        >>> from mindspore.context import ParallelMode
         >>>
         >>> device_id = int(os.environ["DEVICE_ID"])
         >>> context.set_context(mode=context.GRAPH_MODE, device_target="Ascend", save_graphs=True,
diff --git a/model_zoo/official/cv/resnet_thor/src/model_thor.py b/model_zoo/official/cv/resnet_thor/src/model_thor.py
index 4433dd38ac..1b86acf51f 100644
--- a/model_zoo/official/cv/resnet_thor/src/model_thor.py
+++ b/model_zoo/official/cv/resnet_thor/src/model_thor.py
@@ -18,7 +18,7 @@ import math
 from mindspore.train.callback import RunContext
 from mindspore import context
 from mindspore import nn
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.train.model import Model
 from mindspore.parallel._utils import _need_to_full, _to_full_tensor
 from mindspore.common.dtype import pytype_to_dtype
diff --git a/model_zoo/official/cv/resnet_thor/train.py b/model_zoo/official/cv/resnet_thor/train.py
index d7c667dffa..48dd86d7a0 100644
--- a/model_zoo/official/cv/resnet_thor/train.py
+++ b/model_zoo/official/cv/resnet_thor/train.py
@@ -22,7 +22,7 @@ from mindspore import context
 from mindspore import Tensor
 from mindspore import dataset as de
 from mindspore.parallel._auto_parallel_context import auto_parallel_context
-from mindspore.train.model import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, TimeMonitor, LossMonitor
 from mindspore.train.loss_scale_manager import FixedLossScaleManager
 from mindspore.communication.management import init, get_rank, get_group_size
diff --git a/model_zoo/official/cv/resnext50/train.py b/model_zoo/official/cv/resnext50/train.py
index d2cb72d5d2..c328939e30 100644
--- a/model_zoo/official/cv/resnext50/train.py
+++ b/model_zoo/official/cv/resnext50/train.py
@@ -20,7 +20,7 @@ import datetime
 
 import mindspore.nn as nn
 from mindspore import Tensor, context
-from mindspore import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.nn.optim import Momentum
 from mindspore.communication.management import init, get_rank, get_group_size
 from mindspore.train.callback import ModelCheckpoint
diff --git a/model_zoo/official/cv/ssd/src/ssd.py b/model_zoo/official/cv/ssd/src/ssd.py
index fca8a1948d..8fbca66c5c 100644
--- a/model_zoo/official/cv/ssd/src/ssd.py
+++ b/model_zoo/official/cv/ssd/src/ssd.py
@@ -19,6 +19,7 @@ import mindspore.common.dtype as mstype
 import mindspore as ms
 import mindspore.nn as nn
 from mindspore import Parameter, context, Tensor
+from mindspore.context import ParallelMode
 from mindspore.parallel._auto_parallel_context import auto_parallel_context
 from mindspore.communication.management import get_group_size
 from mindspore.ops import operations as P
@@ -388,7 +389,7 @@ class TrainingWrapper(nn.Cell):
         self.reducer_flag = False
         self.grad_reducer = None
         self.parallel_mode = context.get_auto_parallel_context("parallel_mode")
-        if self.parallel_mode in [ms.ParallelMode.DATA_PARALLEL, ms.ParallelMode.HYBRID_PARALLEL]:
+        if self.parallel_mode in [ParallelMode.DATA_PARALLEL, ParallelMode.HYBRID_PARALLEL]:
             self.reducer_flag = True
         if self.reducer_flag:
             mean = context.get_auto_parallel_context("mirror_mean")
diff --git a/model_zoo/official/cv/ssd/train.py b/model_zoo/official/cv/ssd/train.py
index dad5298601..9792892885 100644
--- a/model_zoo/official/cv/ssd/train.py
+++ b/model_zoo/official/cv/ssd/train.py
@@ -21,7 +21,8 @@ import mindspore.nn as nn
 from mindspore import context, Tensor
 from mindspore.communication.management import init
 from mindspore.train.callback import CheckpointConfig, ModelCheckpoint, LossMonitor, TimeMonitor
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
 from src.ssd import SSD300, SSDWithLossCell, TrainingWrapper, ssd_mobilenet_v2
 from src.config import config
diff --git a/model_zoo/official/cv/vgg16/train.py b/model_zoo/official/cv/vgg16/train.py
index ae2f934e1e..d911f3388c 100644
--- a/model_zoo/official/cv/vgg16/train.py
+++ b/model_zoo/official/cv/vgg16/train.py
@@ -29,7 +29,8 @@ from mindspore import context
 from mindspore.communication.management import init, get_rank, get_group_size
 from mindspore.nn.optim.momentum import Momentum
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor, TimeMonitor
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 from mindspore.train.serialization import load_param_into_net, load_checkpoint
 from mindspore.train.loss_scale_manager import FixedLossScaleManager
 from src.dataset import vgg_create_dataset
diff --git a/model_zoo/official/cv/warpctc/src/warpctc_for_train.py b/model_zoo/official/cv/warpctc/src/warpctc_for_train.py
index 8391ffe676..1640342fad 100755
--- a/model_zoo/official/cv/warpctc/src/warpctc_for_train.py
+++ b/model_zoo/official/cv/warpctc/src/warpctc_for_train.py
@@ -16,7 +16,7 @@
 import numpy as np
 from mindspore.parallel._utils import (_get_device_num, _get_mirror_mean,
                                        _get_parallel_mode)
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.common import dtype as mstype
 from mindspore.ops import composite as C
 from mindspore.ops import functional as F
diff --git a/model_zoo/official/cv/warpctc/train.py b/model_zoo/official/cv/warpctc/train.py
index 380308653f..6344368ace 100755
--- a/model_zoo/official/cv/warpctc/train.py
+++ b/model_zoo/official/cv/warpctc/train.py
@@ -21,7 +21,8 @@ import numpy as np
 import mindspore.nn as nn
 from mindspore import context
 from mindspore import dataset as de
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 from mindspore.nn.wrap import WithLossCell
 from mindspore.train.callback import TimeMonitor, LossMonitor, CheckpointConfig, ModelCheckpoint
 from mindspore.communication.management import init, get_group_size, get_rank
diff --git a/model_zoo/official/cv/yolov3_darknet53/eval.py b/model_zoo/official/cv/yolov3_darknet53/eval.py
index cb8a89ec65..7352db7a5d 100644
--- a/model_zoo/official/cv/yolov3_darknet53/eval.py
+++ b/model_zoo/official/cv/yolov3_darknet53/eval.py
@@ -25,7 +25,7 @@ from pycocotools.coco import COCO
 from pycocotools.cocoeval import COCOeval
 
 from mindspore import Tensor
-from mindspore.train import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore import context
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
 import mindspore as ms
diff --git a/model_zoo/official/cv/yolov3_darknet53/src/yolo.py b/model_zoo/official/cv/yolov3_darknet53/src/yolo.py
index eefa1e8bfa..eae0d9f028 100644
--- a/model_zoo/official/cv/yolov3_darknet53/src/yolo.py
+++ b/model_zoo/official/cv/yolov3_darknet53/src/yolo.py
@@ -17,6 +17,7 @@ import mindspore as ms
 import mindspore.nn as nn
 from mindspore.common.tensor import Tensor
 from mindspore import context
+from mindspore.context import ParallelMode
 from mindspore.parallel._auto_parallel_context import auto_parallel_context
 from mindspore.communication.management import get_group_size
 from mindspore.ops import operations as P
@@ -417,7 +418,7 @@ class TrainingWrapper(nn.Cell):
         self.reducer_flag = False
         self.grad_reducer = None
         self.parallel_mode = context.get_auto_parallel_context("parallel_mode")
-        if self.parallel_mode in [ms.ParallelMode.DATA_PARALLEL, ms.ParallelMode.HYBRID_PARALLEL]:
+        if self.parallel_mode in [ParallelMode.DATA_PARALLEL, ParallelMode.HYBRID_PARALLEL]:
             self.reducer_flag = True
         if self.reducer_flag:
             mean = context.get_auto_parallel_context("mirror_mean")
diff --git a/model_zoo/official/cv/yolov3_darknet53/train.py b/model_zoo/official/cv/yolov3_darknet53/train.py
index 63d53f6d2a..877cededeb 100644
--- a/model_zoo/official/cv/yolov3_darknet53/train.py
+++ b/model_zoo/official/cv/yolov3_darknet53/train.py
@@ -18,7 +18,7 @@ import time
 import argparse
 import datetime
 
-from mindspore import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.nn.optim.momentum import Momentum
 from mindspore import Tensor
 import mindspore.nn as nn
diff --git a/model_zoo/official/cv/yolov3_darknet53_quant/eval.py b/model_zoo/official/cv/yolov3_darknet53_quant/eval.py
index 24260f6ee9..99f50fe370 100644
--- a/model_zoo/official/cv/yolov3_darknet53_quant/eval.py
+++ b/model_zoo/official/cv/yolov3_darknet53_quant/eval.py
@@ -25,7 +25,7 @@ from pycocotools.coco import COCO
 from pycocotools.cocoeval import COCOeval
 
 from mindspore import Tensor
-from mindspore.train import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore import context
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
 import mindspore as ms
diff --git a/model_zoo/official/cv/yolov3_darknet53_quant/src/yolo.py b/model_zoo/official/cv/yolov3_darknet53_quant/src/yolo.py
index e794218636..81b00303df 100644
--- a/model_zoo/official/cv/yolov3_darknet53_quant/src/yolo.py
+++ b/model_zoo/official/cv/yolov3_darknet53_quant/src/yolo.py
@@ -17,6 +17,7 @@ import mindspore as ms
 import mindspore.nn as nn
 from mindspore.common.tensor import Tensor
 from mindspore import context
+from mindspore.context import ParallelMode
 from mindspore.parallel._auto_parallel_context import auto_parallel_context
 from mindspore.communication.management import get_group_size
 from mindspore.ops import operations as P
@@ -417,7 +418,7 @@ class TrainingWrapper(nn.Cell):
         self.reducer_flag = False
         self.grad_reducer = None
         self.parallel_mode = context.get_auto_parallel_context("parallel_mode")
-        if self.parallel_mode in [ms.ParallelMode.DATA_PARALLEL, ms.ParallelMode.HYBRID_PARALLEL]:
+        if self.parallel_mode in [ParallelMode.DATA_PARALLEL, ParallelMode.HYBRID_PARALLEL]:
             self.reducer_flag = True
         if self.reducer_flag:
             mean = context.get_auto_parallel_context("mirror_mean")
diff --git a/model_zoo/official/cv/yolov3_darknet53_quant/train.py b/model_zoo/official/cv/yolov3_darknet53_quant/train.py
index 75d1eb0902..047c43f84c 100644
--- a/model_zoo/official/cv/yolov3_darknet53_quant/train.py
+++ b/model_zoo/official/cv/yolov3_darknet53_quant/train.py
@@ -19,7 +19,7 @@ import time
 import argparse
 import datetime
 
-from mindspore import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.nn.optim.momentum import Momentum
 from mindspore import Tensor
 from mindspore import context
diff --git a/model_zoo/official/cv/yolov3_resnet18/src/yolov3.py b/model_zoo/official/cv/yolov3_resnet18/src/yolov3.py
index c33ed1a0d3..2fe4c8f07c 100644
--- a/model_zoo/official/cv/yolov3_resnet18/src/yolov3.py
+++ b/model_zoo/official/cv/yolov3_resnet18/src/yolov3.py
@@ -19,6 +19,7 @@ import numpy as np
 import mindspore as ms
 import mindspore.nn as nn
 from mindspore import context, Tensor
+from mindspore.context import ParallelMode
 from mindspore.parallel._auto_parallel_context import auto_parallel_context
 from mindspore.communication.management import get_group_size
 from mindspore.common.initializer import TruncatedNormal
@@ -652,7 +653,7 @@ class TrainingWrapper(nn.Cell):
         self.reducer_flag = False
         self.grad_reducer = None
         self.parallel_mode = context.get_auto_parallel_context("parallel_mode")
-        if self.parallel_mode in [ms.ParallelMode.DATA_PARALLEL, ms.ParallelMode.HYBRID_PARALLEL]:
+        if self.parallel_mode in [ParallelMode.DATA_PARALLEL, ParallelMode.HYBRID_PARALLEL]:
             self.reducer_flag = True
         if self.reducer_flag:
             mean = context.get_auto_parallel_context("mirror_mean")
diff --git a/model_zoo/official/cv/yolov3_resnet18/train.py b/model_zoo/official/cv/yolov3_resnet18/train.py
index e0d309be9c..6aadb95327 100644
--- a/model_zoo/official/cv/yolov3_resnet18/train.py
+++ b/model_zoo/official/cv/yolov3_resnet18/train.py
@@ -29,7 +29,8 @@ import mindspore.nn as nn
 from mindspore import context, Tensor
 from mindspore.communication.management import init
 from mindspore.train.callback import CheckpointConfig, ModelCheckpoint, LossMonitor, TimeMonitor
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
 from mindspore.common.initializer import initializer
 
diff --git a/model_zoo/official/nlp/bert/run_pretrain.py b/model_zoo/official/nlp/bert/run_pretrain.py
index 73b1021003..d644a1166c 100644
--- a/model_zoo/official/nlp/bert/run_pretrain.py
+++ b/model_zoo/official/nlp/bert/run_pretrain.py
@@ -24,7 +24,7 @@ import mindspore.communication.management as D
 import mindspore.common.dtype as mstype
 from mindspore import context
 from mindspore.train.model import Model
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.nn.wrap.loss_scale import DynamicLossScaleUpdateCell
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, TimeMonitor
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
diff --git a/model_zoo/official/nlp/bert/src/bert_for_finetune.py b/model_zoo/official/nlp/bert/src/bert_for_finetune.py
index 1b147982f0..886cc15398 100644
--- a/model_zoo/official/nlp/bert/src/bert_for_finetune.py
+++ b/model_zoo/official/nlp/bert/src/bert_for_finetune.py
@@ -25,7 +25,7 @@ from mindspore.common.tensor import Tensor
 from mindspore.common.parameter import Parameter
 from mindspore.common import dtype as mstype
 from mindspore.nn.wrap.grad_reducer import DistributedGradReducer
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.communication.management import get_group_size
 from mindspore import context
 from .bert_for_pre_training import clip_grad
diff --git a/model_zoo/official/nlp/bert/src/bert_for_pre_training.py b/model_zoo/official/nlp/bert/src/bert_for_pre_training.py
index 84f442c22c..b57f93143a 100644
--- a/model_zoo/official/nlp/bert/src/bert_for_pre_training.py
+++ b/model_zoo/official/nlp/bert/src/bert_for_pre_training.py
@@ -24,7 +24,7 @@ from mindspore.common.tensor import Tensor
 from mindspore.common.parameter import Parameter
 from mindspore.common import dtype as mstype
 from mindspore.nn.wrap.grad_reducer import DistributedGradReducer
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.communication.management import get_group_size
 from mindspore import context
 from mindspore.ops import _selected_ops
diff --git a/model_zoo/official/nlp/bert_thor/run_pretrain.py b/model_zoo/official/nlp/bert_thor/run_pretrain.py
index 5c5e1c282f..222412e44e 100644
--- a/model_zoo/official/nlp/bert_thor/run_pretrain.py
+++ b/model_zoo/official/nlp/bert_thor/run_pretrain.py
@@ -35,7 +35,7 @@ from mindspore import log as logger
 from mindspore.nn.optim import Lamb, Momentum, AdamWeightDecay
 from mindspore.nn.wrap.loss_scale import DynamicLossScaleUpdateCell
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, TimeMonitor
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
 
 _current_dir = os.path.dirname(os.path.realpath(__file__))
diff --git a/model_zoo/official/nlp/bert_thor/src/bert_for_pre_training.py b/model_zoo/official/nlp/bert_thor/src/bert_for_pre_training.py
index 807d5a5d31..81271a6992 100644
--- a/model_zoo/official/nlp/bert_thor/src/bert_for_pre_training.py
+++ b/model_zoo/official/nlp/bert_thor/src/bert_for_pre_training.py
@@ -27,7 +27,7 @@ from mindspore.ops import _selected_ops
 from mindspore.ops import composite as C
 from mindspore.ops import functional as F
 from mindspore.ops import operations as P
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from .bert_model import BertModel
 from .config import cfg
 from .lr_generator import get_bert_damping
diff --git a/model_zoo/official/nlp/bert_thor/src/grad_reducer_thor.py b/model_zoo/official/nlp/bert_thor/src/grad_reducer_thor.py
index 47b86a4e65..cd0cc34819 100644
--- a/model_zoo/official/nlp/bert_thor/src/grad_reducer_thor.py
+++ b/model_zoo/official/nlp/bert_thor/src/grad_reducer_thor.py
@@ -102,7 +102,8 @@ class DistributedGradReducerThor(Cell):
         >>> from mindspore.ops import functional as F
         >>> from mindspore import context
         >>> from mindspore import nn
-        >>> from mindspore import ParallelMode, ParameterTuple
+        >>> from mindspore import ParameterTuple
+        >>> from mindspore.context import ParallelMode
         >>>
         >>> device_id = int(os.environ["DEVICE_ID"])
         >>> context.set_context(mode=context.GRAPH_MODE, device_target="Ascend", save_graphs=True,
diff --git a/model_zoo/official/nlp/bert_thor/src/model_thor.py b/model_zoo/official/nlp/bert_thor/src/model_thor.py
index 2ccb8b6ec6..01697f65a7 100644
--- a/model_zoo/official/nlp/bert_thor/src/model_thor.py
+++ b/model_zoo/official/nlp/bert_thor/src/model_thor.py
@@ -36,7 +36,7 @@ from mindspore.parallel._utils import _need_to_full
 from mindspore.train import amp
 from mindspore.parallel._utils import _to_full_tensor
 from mindspore.train.callback import _InternalCallbackParam, RunContext, _CallbackManager
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from .dataset_helper import DatasetHelper
 
 
diff --git a/model_zoo/official/nlp/mass/src/transformer/transformer_for_train.py b/model_zoo/official/nlp/mass/src/transformer/transformer_for_train.py
index 3cb1b3739a..fd5974e162 100644
--- a/model_zoo/official/nlp/mass/src/transformer/transformer_for_train.py
+++ b/model_zoo/official/nlp/mass/src/transformer/transformer_for_train.py
@@ -22,7 +22,7 @@ from mindspore.common.tensor import Tensor
 from mindspore.common.parameter import Parameter
 from mindspore.common import dtype as mstype
 from mindspore.nn.wrap.grad_reducer import DistributedGradReducer
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.parallel._utils import _get_device_num, _get_parallel_mode, _get_mirror_mean
 
 from .transformer import Transformer
diff --git a/model_zoo/official/nlp/mass/train.py b/model_zoo/official/nlp/mass/train.py
index 14ab3c1b82..0562d75b3d 100644
--- a/model_zoo/official/nlp/mass/train.py
+++ b/model_zoo/official/nlp/mass/train.py
@@ -26,7 +26,8 @@ from mindspore.nn.optim import Adam, Lamb
 from mindspore.train.model import Model
 from mindspore.train.loss_scale_manager import DynamicLossScaleManager, FixedLossScaleManager
 from mindspore.train.callback import CheckpointConfig, ModelCheckpoint
-from mindspore import context, ParallelMode, Parameter
+from mindspore import context, Parameter
+from mindspore.context import ParallelMode
 from mindspore.communication import management as MultiAscend
 from mindspore.train.serialization import load_checkpoint
 
diff --git a/model_zoo/official/nlp/tinybert/run_general_distill.py b/model_zoo/official/nlp/tinybert/run_general_distill.py
index ff771428db..1cbab6b33f 100644
--- a/model_zoo/official/nlp/tinybert/run_general_distill.py
+++ b/model_zoo/official/nlp/tinybert/run_general_distill.py
@@ -24,7 +24,7 @@ import mindspore.common.dtype as mstype
 from mindspore import context
 from mindspore.train.model import Model
 from mindspore.train.callback import TimeMonitor
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.nn.optim import AdamWeightDecay
 from mindspore.nn.wrap.loss_scale import DynamicLossScaleUpdateCell
 from mindspore import log as logger
diff --git a/model_zoo/official/nlp/tinybert/src/tinybert_for_gd_td.py b/model_zoo/official/nlp/tinybert/src/tinybert_for_gd_td.py
index f003ec26e7..4e595ec4e6 100644
--- a/model_zoo/official/nlp/tinybert/src/tinybert_for_gd_td.py
+++ b/model_zoo/official/nlp/tinybert/src/tinybert_for_gd_td.py
@@ -26,7 +26,7 @@ from mindspore.common import dtype as mstype
 from mindspore.common.parameter import Parameter
 from mindspore.communication.management import get_group_size
 from mindspore.nn.wrap.grad_reducer import DistributedGradReducer
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
 from .tinybert_model import BertModel, TinyBertModel, BertModelCLS
 
diff --git a/model_zoo/official/nlp/transformer/src/transformer_for_train.py b/model_zoo/official/nlp/transformer/src/transformer_for_train.py
index 32d5ad7e20..f26396d1f1 100644
--- a/model_zoo/official/nlp/transformer/src/transformer_for_train.py
+++ b/model_zoo/official/nlp/transformer/src/transformer_for_train.py
@@ -22,7 +22,7 @@ from mindspore.common.tensor import Tensor
 from mindspore.common.parameter import Parameter, ParameterTuple
 from mindspore.common import dtype as mstype
 from mindspore.nn.wrap.grad_reducer import DistributedGradReducer
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.parallel._utils import _get_device_num, _get_parallel_mode, _get_mirror_mean
 from mindspore.communication.management import get_group_size
 from mindspore import context
diff --git a/model_zoo/official/nlp/transformer/train.py b/model_zoo/official/nlp/transformer/train.py
index 2ac2c09369..36edcb0854 100644
--- a/model_zoo/official/nlp/transformer/train.py
+++ b/model_zoo/official/nlp/transformer/train.py
@@ -29,7 +29,7 @@ from mindspore.train.callback import Callback, TimeMonitor
 from mindspore.train.serialization import load_checkpoint, load_param_into_net
 import mindspore.dataset.engine as de
 import mindspore.communication.management as D
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore import context
 
 from src.transformer_for_train import TransformerTrainOneStepCell, TransformerNetworkWithLoss, \
diff --git a/model_zoo/official/recommend/deepfm/train.py b/model_zoo/official/recommend/deepfm/train.py
index 39d96a510f..c9f2549fac 100644
--- a/model_zoo/official/recommend/deepfm/train.py
+++ b/model_zoo/official/recommend/deepfm/train.py
@@ -19,7 +19,8 @@ import argparse
 import random
 import numpy as np
 
-from mindspore import context, ParallelMode
+from mindspore import context
+from mindspore.context import ParallelMode
 from mindspore.communication.management import init, get_rank, get_group_size
 from mindspore.train.model import Model
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, TimeMonitor
diff --git a/model_zoo/official/recommend/wide_and_deep/src/callbacks.py b/model_zoo/official/recommend/wide_and_deep/src/callbacks.py
index 40dba54578..c48ac8a699 100644
--- a/model_zoo/official/recommend/wide_and_deep/src/callbacks.py
+++ b/model_zoo/official/recommend/wide_and_deep/src/callbacks.py
@@ -17,7 +17,7 @@ callbacks
 import time
 from mindspore.train.callback import Callback
 from mindspore import context
-from mindspore.train import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.communication.management import get_rank
 
 def add_write(file_path, out_str):
diff --git a/model_zoo/official/recommend/wide_and_deep/src/wide_and_deep.py b/model_zoo/official/recommend/wide_and_deep/src/wide_and_deep.py
index e873c15ad0..6a03d6feff 100644
--- a/model_zoo/official/recommend/wide_and_deep/src/wide_and_deep.py
+++ b/model_zoo/official/recommend/wide_and_deep/src/wide_and_deep.py
@@ -23,7 +23,7 @@ from mindspore.ops import operations as P
 from mindspore.nn import Dropout
 from mindspore.nn.optim import Adam, FTRL, LazyAdam
 from mindspore.common.initializer import Uniform, initializer
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.nn.wrap.grad_reducer import DistributedGradReducer
 from mindspore.communication.management import get_group_size
 
diff --git a/model_zoo/official/recommend/wide_and_deep/train_and_eval_auto_parallel.py b/model_zoo/official/recommend/wide_and_deep/train_and_eval_auto_parallel.py
index 5440873d2c..5fd6e1afad 100644
--- a/model_zoo/official/recommend/wide_and_deep/train_and_eval_auto_parallel.py
+++ b/model_zoo/official/recommend/wide_and_deep/train_and_eval_auto_parallel.py
@@ -20,7 +20,7 @@ import sys
 import mindspore.dataset.engine as de
 from mindspore import Model, context
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, TimeMonitor
-from mindspore.train import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.communication.management import get_rank, get_group_size, init
 from mindspore.parallel import set_multi_subgraphs
 from mindspore.nn.wrap.cell_wrapper import VirtualDatasetCellTriple
diff --git a/model_zoo/official/recommend/wide_and_deep/train_and_eval_distribute.py b/model_zoo/official/recommend/wide_and_deep/train_and_eval_distribute.py
index 7e99aa72bb..9bd312a621 100644
--- a/model_zoo/official/recommend/wide_and_deep/train_and_eval_distribute.py
+++ b/model_zoo/official/recommend/wide_and_deep/train_and_eval_distribute.py
@@ -20,7 +20,7 @@ import sys
 import numpy as np
 from mindspore import Model, context
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, TimeMonitor
-from mindspore.train import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.communication.management import get_rank, get_group_size, init
 
 from src.wide_and_deep import PredictWithSigmoid, TrainStepWrap, NetWithLossClass, WideDeepModel
diff --git a/model_zoo/official/recommend/wide_and_deep/train_and_eval_parameter_server.py b/model_zoo/official/recommend/wide_and_deep/train_and_eval_parameter_server.py
index 5f93e202da..f52e673b74 100644
--- a/model_zoo/official/recommend/wide_and_deep/train_and_eval_parameter_server.py
+++ b/model_zoo/official/recommend/wide_and_deep/train_and_eval_parameter_server.py
@@ -20,7 +20,7 @@ import sys
 import numpy as np
 from mindspore import Model, context
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, TimeMonitor
-from mindspore.train import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.communication.management import get_rank, get_group_size, init
 
 from src.wide_and_deep import PredictWithSigmoid, TrainStepWrap, NetWithLossClass, WideDeepModel
diff --git a/model_zoo/official/recommend/wide_and_deep_multitable/src/wide_and_deep.py b/model_zoo/official/recommend/wide_and_deep_multitable/src/wide_and_deep.py
index c36a7ea66f..7d9b566a88 100644
--- a/model_zoo/official/recommend/wide_and_deep_multitable/src/wide_and_deep.py
+++ b/model_zoo/official/recommend/wide_and_deep_multitable/src/wide_and_deep.py
@@ -24,7 +24,7 @@ from mindspore.ops import operations as P
 from mindspore.nn import Dropout, Flatten
 from mindspore.nn.optim import Adam, FTRL
 from mindspore.common.initializer import Uniform, initializer
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.nn.wrap.grad_reducer import DistributedGradReducer
 
 
diff --git a/model_zoo/official/recommend/wide_and_deep_multitable/train_and_eval_distribute.py b/model_zoo/official/recommend/wide_and_deep_multitable/train_and_eval_distribute.py
index 756a227fa3..99fe089fe6 100644
--- a/model_zoo/official/recommend/wide_and_deep_multitable/train_and_eval_distribute.py
+++ b/model_zoo/official/recommend/wide_and_deep_multitable/train_and_eval_distribute.py
@@ -20,7 +20,7 @@ import numpy as np
 from mindspore import Model, context
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig
 from mindspore.train.callback import TimeMonitor
-from mindspore.train import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.communication.management import get_rank, get_group_size, init
 
 from src.wide_and_deep import PredictWithSigmoid, TrainStepWrap, NetWithLossClass, WideDeepModel
diff --git a/tests/st/auto_parallel/resnet50_expand_loss.py b/tests/st/auto_parallel/resnet50_expand_loss.py
index f46bb587fb..fc6d00ff61 100644
--- a/tests/st/auto_parallel/resnet50_expand_loss.py
+++ b/tests/st/auto_parallel/resnet50_expand_loss.py
@@ -28,7 +28,8 @@ from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import operations as P
 from mindspore.parallel import set_algo_parameters
 from mindspore.train.callback import Callback
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 
 context.set_context(mode=context.GRAPH_MODE, device_target="Ascend")
 context.set_context(device_id=int(os.getenv('DEVICE_ID')))
diff --git a/tests/st/mem_reuse/resnet_cifar_memreuse.py b/tests/st/mem_reuse/resnet_cifar_memreuse.py
index 3a63dbbbc2..fc4cf42b51 100644
--- a/tests/st/mem_reuse/resnet_cifar_memreuse.py
+++ b/tests/st/mem_reuse/resnet_cifar_memreuse.py
@@ -30,7 +30,8 @@ from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import functional as F
 from mindspore.ops import operations as P
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 
 random.seed(1)
 np.random.seed(1)
diff --git a/tests/st/mem_reuse/resnet_cifar_normal.py b/tests/st/mem_reuse/resnet_cifar_normal.py
index d96d016e66..83241220cd 100644
--- a/tests/st/mem_reuse/resnet_cifar_normal.py
+++ b/tests/st/mem_reuse/resnet_cifar_normal.py
@@ -30,7 +30,8 @@ from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import functional as F
 from mindspore.ops import operations as P
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 
 random.seed(1)
 np.random.seed(1)
diff --git a/tests/st/model_zoo_tests/wide_and_deep/python_file_for_ci/train_and_test_multinpu_ci.py b/tests/st/model_zoo_tests/wide_and_deep/python_file_for_ci/train_and_test_multinpu_ci.py
index 073c23423e..22cb8bb40e 100644
--- a/tests/st/model_zoo_tests/wide_and_deep/python_file_for_ci/train_and_test_multinpu_ci.py
+++ b/tests/st/model_zoo_tests/wide_and_deep/python_file_for_ci/train_and_test_multinpu_ci.py
@@ -19,7 +19,7 @@ import os
 import sys
 from mindspore import Model, context
 from mindspore.train.callback import TimeMonitor
-from mindspore.train import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.communication.management import get_rank, get_group_size, init
 from mindspore.parallel import set_multi_subgraphs
 from mindspore.nn.wrap.cell_wrapper import VirtualDatasetCellTriple
diff --git a/tests/st/model_zoo_tests/wide_and_deep/python_file_for_ci/wide_and_deep.py b/tests/st/model_zoo_tests/wide_and_deep/python_file_for_ci/wide_and_deep.py
index c31b3b5b0e..6944b46b2d 100644
--- a/tests/st/model_zoo_tests/wide_and_deep/python_file_for_ci/wide_and_deep.py
+++ b/tests/st/model_zoo_tests/wide_and_deep/python_file_for_ci/wide_and_deep.py
@@ -25,7 +25,7 @@ from mindspore.nn.optim import Adam, FTRL
 from mindspore.common.initializer import Uniform, initializer
 # from mindspore.train.callback import ModelCheckpoint, CheckpointConfig
 from mindspore.parallel._utils import _get_device_num, _get_parallel_mode, _get_mirror_mean
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.nn.wrap.grad_reducer import DistributedGradReducer
 from mindspore.communication.management import get_group_size
 import numpy as np
diff --git a/tests/st/model_zoo_tests/wide_and_deep/train_and_test_multinpu_ci_data_parallel.py b/tests/st/model_zoo_tests/wide_and_deep/train_and_test_multinpu_ci_data_parallel.py
index 0f909b0236..eaae9dfff3 100644
--- a/tests/st/model_zoo_tests/wide_and_deep/train_and_test_multinpu_ci_data_parallel.py
+++ b/tests/st/model_zoo_tests/wide_and_deep/train_and_test_multinpu_ci_data_parallel.py
@@ -20,7 +20,7 @@ import sys
 import numpy as np
 from mindspore import Model, context
 from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, TimeMonitor
-from mindspore.train import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.communication.management import get_rank, get_group_size, init
 
 from src.wide_and_deep import PredictWithSigmoid, TrainStepWrap, NetWithLossClass, WideDeepModel
diff --git a/tests/st/model_zoo_tests/yolov3/src/yolov3.py b/tests/st/model_zoo_tests/yolov3/src/yolov3.py
index c33ed1a0d3..2fe4c8f07c 100644
--- a/tests/st/model_zoo_tests/yolov3/src/yolov3.py
+++ b/tests/st/model_zoo_tests/yolov3/src/yolov3.py
@@ -19,6 +19,7 @@ import numpy as np
 import mindspore as ms
 import mindspore.nn as nn
 from mindspore import context, Tensor
+from mindspore.context import ParallelMode
 from mindspore.parallel._auto_parallel_context import auto_parallel_context
 from mindspore.communication.management import get_group_size
 from mindspore.common.initializer import TruncatedNormal
@@ -652,7 +653,7 @@ class TrainingWrapper(nn.Cell):
         self.reducer_flag = False
         self.grad_reducer = None
         self.parallel_mode = context.get_auto_parallel_context("parallel_mode")
-        if self.parallel_mode in [ms.ParallelMode.DATA_PARALLEL, ms.ParallelMode.HYBRID_PARALLEL]:
+        if self.parallel_mode in [ParallelMode.DATA_PARALLEL, ParallelMode.HYBRID_PARALLEL]:
             self.reducer_flag = True
         if self.reducer_flag:
             mean = context.get_auto_parallel_context("mirror_mean")
diff --git a/tests/st/networks/models/bert/src/bert_for_pre_training.py b/tests/st/networks/models/bert/src/bert_for_pre_training.py
index 2577cf617a..0a0675a805 100644
--- a/tests/st/networks/models/bert/src/bert_for_pre_training.py
+++ b/tests/st/networks/models/bert/src/bert_for_pre_training.py
@@ -24,7 +24,7 @@ from mindspore.common.tensor import Tensor
 from mindspore.common.parameter import Parameter, ParameterTuple
 from mindspore.common import dtype as mstype
 from mindspore.nn.wrap.grad_reducer import DistributedGradReducer
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.communication.management import get_group_size
 from mindspore import context
 from .bert_model import BertModel
diff --git a/tests/st/networks/models/bert/src/utils.py b/tests/st/networks/models/bert/src/utils.py
index 2b19d3d291..bcea50dc3b 100644
--- a/tests/st/networks/models/bert/src/utils.py
+++ b/tests/st/networks/models/bert/src/utils.py
@@ -26,7 +26,7 @@ from mindspore.common.tensor import Tensor
 from mindspore.common.parameter import Parameter, ParameterTuple
 from mindspore.common import dtype as mstype
 from mindspore.nn.wrap.grad_reducer import DistributedGradReducer
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from mindspore.communication.management import get_group_size
 from mindspore import context
 from mindspore.model_zoo.Bert_NEZHA.bert_model import BertModel
diff --git a/tests/st/networks/models/resnet50/src_thor/dataset_helper.py b/tests/st/networks/models/resnet50/src_thor/dataset_helper.py
index 4e900fd38a..448589b6f7 100644
--- a/tests/st/networks/models/resnet50/src_thor/dataset_helper.py
+++ b/tests/st/networks/models/resnet50/src_thor/dataset_helper.py
@@ -16,7 +16,7 @@
 from mindspore._checkparam import check_bool
 from mindspore.parallel._utils import _get_device_num, _get_parallel_mode, _to_full_shapes
 from mindspore.train._utils import _exec_datagraph, _get_types_and_shapes
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 
 def _send_data(dataset):
     """Engine dataset to write data to tdt queue."""
diff --git a/tests/st/networks/models/resnet50/src_thor/grad_reducer_thor.py b/tests/st/networks/models/resnet50/src_thor/grad_reducer_thor.py
index e84c941249..97d2cb2270 100644
--- a/tests/st/networks/models/resnet50/src_thor/grad_reducer_thor.py
+++ b/tests/st/networks/models/resnet50/src_thor/grad_reducer_thor.py
@@ -103,7 +103,8 @@ class DistributedGradReducerThor(Cell):
         >>> from mindspore.ops import functional as F
         >>> from mindspore import context
         >>> from mindspore import nn
-        >>> from mindspore import ParallelMode, ParameterTuple
+        >>> from mindspore import ParameterTuple
+        >>> from mindspore.context import ParallelMode
         >>>
         >>> device_id = int(os.environ["DEVICE_ID"])
         >>> context.set_context(mode=context.GRAPH_MODE, device_target="Ascend", save_graphs=True,
diff --git a/tests/st/networks/models/resnet50/src_thor/model_thor.py b/tests/st/networks/models/resnet50/src_thor/model_thor.py
index 2399c4a01b..8144b1cd80 100644
--- a/tests/st/networks/models/resnet50/src_thor/model_thor.py
+++ b/tests/st/networks/models/resnet50/src_thor/model_thor.py
@@ -30,7 +30,7 @@ from mindspore.parallel._utils import _get_parallel_mode, _get_device_num, _get_
     _get_parameter_broadcast, _device_number_check, _parameter_broadcast_check
 from mindspore.train import amp
 from mindspore.train.callback import _InternalCallbackParam, RunContext, _CallbackManager
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 
 from .dataset_helper import DatasetHelper
 
diff --git a/tests/st/networks/models/resnet50/test_resnet50_imagenet.py b/tests/st/networks/models/resnet50/test_resnet50_imagenet.py
index 220b986208..42b209f25e 100644
--- a/tests/st/networks/models/resnet50/test_resnet50_imagenet.py
+++ b/tests/st/networks/models/resnet50/test_resnet50_imagenet.py
@@ -24,7 +24,8 @@ import numpy as np
 from mindspore import context, Tensor
 from mindspore.communication.management import init
 from mindspore.parallel._auto_parallel_context import auto_parallel_context
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 from mindspore.train.callback import Callback
 from mindspore.train.loss_scale_manager import FixedLossScaleManager
 import mindspore.nn as nn
diff --git a/tests/st/tbe_networks/resnet_cifar.py b/tests/st/tbe_networks/resnet_cifar.py
index c6b1ee0a78..c40b4809be 100644
--- a/tests/st/tbe_networks/resnet_cifar.py
+++ b/tests/st/tbe_networks/resnet_cifar.py
@@ -32,7 +32,8 @@ from mindspore.communication.management import init
 from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import operations as P
 from mindspore.parallel._auto_parallel_context import auto_parallel_context
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 
 random.seed(1)
 np.random.seed(1)
diff --git a/tests/st/tbe_networks/test_resnet_cifar_8p.py b/tests/st/tbe_networks/test_resnet_cifar_8p.py
index 56d6a91d64..7eefdcc7a9 100644
--- a/tests/st/tbe_networks/test_resnet_cifar_8p.py
+++ b/tests/st/tbe_networks/test_resnet_cifar_8p.py
@@ -32,7 +32,8 @@ from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import operations as P
 from mindspore.parallel._auto_parallel_context import auto_parallel_context
 from mindspore.train.callback import Callback
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 
 random.seed(1)
 np.random.seed(1)
diff --git a/tests/ut/python/communication/test_data_parallel_dense.py b/tests/ut/python/communication/test_data_parallel_dense.py
index ef042e3f39..d2fdf2d158 100644
--- a/tests/ut/python/communication/test_data_parallel_dense.py
+++ b/tests/ut/python/communication/test_data_parallel_dense.py
@@ -25,7 +25,7 @@ from mindspore.common.api import _executor
 from mindspore.nn import Momentum
 from mindspore.nn import TrainOneStepCell, WithLossCell
 from mindspore.ops import operations as P
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 
 
 class DenseMMNet(nn.Cell):
diff --git a/tests/ut/python/communication/test_data_parallel_lenet.py b/tests/ut/python/communication/test_data_parallel_lenet.py
index 7a5062b941..0897023a69 100755
--- a/tests/ut/python/communication/test_data_parallel_lenet.py
+++ b/tests/ut/python/communication/test_data_parallel_lenet.py
@@ -21,7 +21,8 @@ import numpy as np
 
 import mindspore.context as context
 import mindspore.nn as nn
-from mindspore import Tensor, Model, ParallelMode
+from mindspore import Tensor, Model
+from mindspore.context import ParallelMode
 from mindspore.nn.optim import Momentum
 from mindspore.ops import operations as P
 
diff --git a/tests/ut/python/communication/test_data_parallel_resnet.py b/tests/ut/python/communication/test_data_parallel_resnet.py
index fa798c0616..ad60c3c957 100644
--- a/tests/ut/python/communication/test_data_parallel_resnet.py
+++ b/tests/ut/python/communication/test_data_parallel_resnet.py
@@ -19,7 +19,8 @@ import numpy as np
 
 import mindspore.context as context
 import mindspore.nn as nn
-from mindspore import Tensor, Model, ParallelMode
+from mindspore import Tensor, Model
+from mindspore.context import ParallelMode
 from mindspore.nn.optim import Momentum
 from mindspore.ops.operations import TensorAdd
 from ....dataset_mock import MindData
diff --git a/tests/ut/python/model/test_mix_precision.py b/tests/ut/python/model/test_mix_precision.py
index 4570bd243a..9faf7341b6 100644
--- a/tests/ut/python/model/test_mix_precision.py
+++ b/tests/ut/python/model/test_mix_precision.py
@@ -26,7 +26,7 @@ from mindspore.nn import TrainOneStepCell, WithLossCell
 from mindspore.ops import composite as C
 from mindspore.ops import operations as P
 from mindspore.ops import functional as F
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from tests.ops_common import convert
 from ....train_step_wrap import train_step_with_loss_warp
 
diff --git a/tests/ut/python/parallel/test_allreduce_fusion.py b/tests/ut/python/parallel/test_allreduce_fusion.py
index c93df7ffb1..d1d9c74e5d 100644
--- a/tests/ut/python/parallel/test_allreduce_fusion.py
+++ b/tests/ut/python/parallel/test_allreduce_fusion.py
@@ -22,7 +22,8 @@ from mindspore.nn.loss import SoftmaxCrossEntropyWithLogits
 from mindspore.nn.optim.momentum import Momentum
 from mindspore.parallel import _cost_model_context as cost_model_context
 from mindspore.parallel._auto_parallel_context import auto_parallel_context
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from tests.dataset_mock import MindData
 
 
diff --git a/tests/ut/python/parallel/test_alltoall.py b/tests/ut/python/parallel/test_alltoall.py
index 96ff843504..53a77a0bc6 100644
--- a/tests/ut/python/parallel/test_alltoall.py
+++ b/tests/ut/python/parallel/test_alltoall.py
@@ -24,7 +24,8 @@ from mindspore.nn.loss import SoftmaxCrossEntropyWithLogits
 from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import operations as P
 from mindspore.parallel._utils import _reset_op_id
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from tests.dataset_mock import MindData
 
 
diff --git a/tests/ut/python/parallel/test_auto_parallel_onehot.py b/tests/ut/python/parallel/test_auto_parallel_onehot.py
index 302de23a50..32ee1c99ed 100644
--- a/tests/ut/python/parallel/test_auto_parallel_onehot.py
+++ b/tests/ut/python/parallel/test_auto_parallel_onehot.py
@@ -23,7 +23,8 @@ from mindspore.common.parameter import Parameter
 from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import composite as C
 from mindspore.ops import operations as P
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from tests.dataset_mock import MindData
 from tests.ut.python.ops.test_math_ops import VirtualLoss
 
diff --git a/tests/ut/python/parallel/test_auto_parallel_resnet.py b/tests/ut/python/parallel/test_auto_parallel_resnet.py
index 5a117bbb1a..9d4d2f2755 100644
--- a/tests/ut/python/parallel/test_auto_parallel_resnet.py
+++ b/tests/ut/python/parallel/test_auto_parallel_resnet.py
@@ -29,7 +29,8 @@ from mindspore.ops import operations as P
 from mindspore.parallel import _cost_model_context as cost_model_context
 from mindspore.parallel import set_algo_parameters
 from mindspore.parallel._utils import _reset_op_id as resset_op_id
-from mindspore.train.model import Model, ParallelMode
+from mindspore.train.model import Model
+from mindspore.context import ParallelMode
 
 context.set_context(mode=context.GRAPH_MODE, device_target="Ascend")
 context.set_context(device_id=0)
diff --git a/tests/ut/python/parallel/test_batchnorm_batch_parallel.py b/tests/ut/python/parallel/test_batchnorm_batch_parallel.py
index 21d5003b4c..236cc263df 100644
--- a/tests/ut/python/parallel/test_batchnorm_batch_parallel.py
+++ b/tests/ut/python/parallel/test_batchnorm_batch_parallel.py
@@ -26,7 +26,8 @@ from mindspore.nn.layer.pooling import MaxPool2d
 from mindspore.nn.loss import SoftmaxCrossEntropyWithLogits
 from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import operations as P
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from tests.dataset_mock import MindData
 
 dev_num = 8
diff --git a/tests/ut/python/parallel/test_bn_prelu_cell.py b/tests/ut/python/parallel/test_bn_prelu_cell.py
index 07f5d3906b..0694c331b1 100644
--- a/tests/ut/python/parallel/test_bn_prelu_cell.py
+++ b/tests/ut/python/parallel/test_bn_prelu_cell.py
@@ -27,7 +27,7 @@ from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import functional as F
 from mindspore.ops import operations as P
 from mindspore.train.model import Model
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 from tests.dataset_mock import MindData
 
 
diff --git a/tests/ut/python/parallel/test_dataset_interface.py b/tests/ut/python/parallel/test_dataset_interface.py
index 0e70b2513c..02153cadf1 100644
--- a/tests/ut/python/parallel/test_dataset_interface.py
+++ b/tests/ut/python/parallel/test_dataset_interface.py
@@ -22,7 +22,8 @@ from mindspore.common.parameter import Parameter, ParameterTuple
 from mindspore.nn.loss import SoftmaxCrossEntropyWithLogits
 from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import composite as C, functional as F, operations as P
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from mindspore.train.loss_scale_manager import DynamicLossScaleManager
 from tests.dataset_mock import MindData
 
diff --git a/tests/ut/python/parallel/test_full_batch.py b/tests/ut/python/parallel/test_full_batch.py
index 70a68a5b00..ddb0b057e2 100644
--- a/tests/ut/python/parallel/test_full_batch.py
+++ b/tests/ut/python/parallel/test_full_batch.py
@@ -23,7 +23,8 @@ from mindspore.nn.loss import SoftmaxCrossEntropyWithLogits
 from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import operations as P
 from mindspore.parallel._utils import _reset_op_id
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from tests.dataset_mock import MindData
 
 class Dataset(MindData):
diff --git a/tests/ut/python/parallel/test_gather_v2_primitive.py b/tests/ut/python/parallel/test_gather_v2_primitive.py
index 4c0534aad0..8ad626e184 100644
--- a/tests/ut/python/parallel/test_gather_v2_primitive.py
+++ b/tests/ut/python/parallel/test_gather_v2_primitive.py
@@ -27,7 +27,8 @@ from mindspore.nn.optim import Momentum
 from mindspore.ops import composite as C
 from mindspore.ops import functional as F
 from mindspore.ops import operations as P
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 
 context.set_context(mode=context.GRAPH_MODE)
 device_number = 32
diff --git a/tests/ut/python/parallel/test_loss_scale.py b/tests/ut/python/parallel/test_loss_scale.py
index 3469400d2a..88997160f8 100644
--- a/tests/ut/python/parallel/test_loss_scale.py
+++ b/tests/ut/python/parallel/test_loss_scale.py
@@ -25,7 +25,8 @@ from mindspore.ops import functional as F
 from mindspore.nn.optim.momentum import Momentum
 from mindspore.nn.wrap.loss_scale import DynamicLossScaleUpdateCell
 import mindspore.nn as nn
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from tests.dataset_mock import MindData
 
 
diff --git a/tests/ut/python/parallel/test_one_dev.py b/tests/ut/python/parallel/test_one_dev.py
index 056f4a15c7..ec0990761d 100644
--- a/tests/ut/python/parallel/test_one_dev.py
+++ b/tests/ut/python/parallel/test_one_dev.py
@@ -25,7 +25,8 @@ from mindspore.nn.loss import SoftmaxCrossEntropyWithLogits
 from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import operations as P
 from mindspore.parallel._utils import _reset_op_id
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from tests.dataset_mock import MindData
 
 context.set_context(mode=context.GRAPH_MODE)
diff --git a/tests/ut/python/parallel/test_one_hot_net.py b/tests/ut/python/parallel/test_one_hot_net.py
index 33c8fcbc82..32f0d2a85c 100644
--- a/tests/ut/python/parallel/test_one_hot_net.py
+++ b/tests/ut/python/parallel/test_one_hot_net.py
@@ -25,7 +25,8 @@ from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import composite as C
 from mindspore.ops import functional as F
 from mindspore.ops import operations as P
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from tests.dataset_mock import MindData
 from tests.ut.python.ops.test_math_ops import VirtualLoss
 
diff --git a/tests/ut/python/parallel/test_operator_model_parallel.py b/tests/ut/python/parallel/test_operator_model_parallel.py
index 788521c525..96e3c86f51 100644
--- a/tests/ut/python/parallel/test_operator_model_parallel.py
+++ b/tests/ut/python/parallel/test_operator_model_parallel.py
@@ -29,7 +29,8 @@ from mindspore.nn.loss import SoftmaxCrossEntropyWithLogits
 from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import operations as P
 from mindspore.ops.operations import TensorAdd
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from tests.dataset_mock import MindData
 
 dev_num = 8
diff --git a/tests/ut/python/parallel/test_optimizer.py b/tests/ut/python/parallel/test_optimizer.py
index e848f4ed9b..0d6de9526a 100644
--- a/tests/ut/python/parallel/test_optimizer.py
+++ b/tests/ut/python/parallel/test_optimizer.py
@@ -23,7 +23,7 @@ from mindspore.nn import Dense
 from mindspore.nn import Momentum
 from mindspore.nn import TrainOneStepCell, WithLossCell
 from mindspore.ops import operations as P
-from mindspore.train.parallel_utils import ParallelMode
+from mindspore.context import ParallelMode
 
 
 class Net(nn.Cell):
diff --git a/tests/ut/python/parallel/test_prelu_cell.py b/tests/ut/python/parallel/test_prelu_cell.py
index dca467ef8d..074e585f53 100644
--- a/tests/ut/python/parallel/test_prelu_cell.py
+++ b/tests/ut/python/parallel/test_prelu_cell.py
@@ -24,7 +24,8 @@ from mindspore.nn.loss import SoftmaxCrossEntropyWithLogits
 from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import functional as F
 from mindspore.ops import operations as P
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from tests.dataset_mock import MindData
 
 context.set_context(mode=context.GRAPH_MODE)
diff --git a/tests/ut/python/parallel/test_reshape.py b/tests/ut/python/parallel/test_reshape.py
index 9cfb376e1b..d2e08230be 100644
--- a/tests/ut/python/parallel/test_reshape.py
+++ b/tests/ut/python/parallel/test_reshape.py
@@ -28,7 +28,8 @@ from mindspore.ops import functional as F
 from mindspore.ops import operations as P
 from mindspore.ops.operations.comm_ops import _VirtualDataset
 from mindspore.parallel import set_algo_parameters
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from tests.dataset_mock import MindData
 from tests.ut.python.ops.test_math_ops import VirtualLoss
 
diff --git a/tests/ut/python/parallel/test_transpose.py b/tests/ut/python/parallel/test_transpose.py
index b0b917bf19..2669a57433 100644
--- a/tests/ut/python/parallel/test_transpose.py
+++ b/tests/ut/python/parallel/test_transpose.py
@@ -21,7 +21,8 @@ from mindspore.common.parameter import Parameter
 from mindspore.nn.loss import SoftmaxCrossEntropyWithLogits
 from mindspore.nn.optim.momentum import Momentum
 from mindspore.ops import operations as P
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from tests.dataset_mock import MindData
 
 
diff --git a/tests/ut/python/train/test_amp.py b/tests/ut/python/train/test_amp.py
index 6406a4b80d..0cba6bfa08 100644
--- a/tests/ut/python/train/test_amp.py
+++ b/tests/ut/python/train/test_amp.py
@@ -20,7 +20,8 @@ import mindspore.context as context
 from mindspore import Tensor
 from mindspore import amp
 from mindspore import nn
-from mindspore.train import Model, ParallelMode
+from mindspore.train import Model
+from mindspore.context import ParallelMode
 from mindspore.common import dtype as mstype
 from ....dataset_mock import MindData
 from mindspore.parallel._auto_parallel_context import auto_parallel_context