[Auto Parallel] Change the import way of Auto Parallel (#46115)

python/paddle/distributed/auto_parallel/engine.py

@@ -81,7 +81,7 @@ class Engine:
 
             import paddle
             import paddle.vision.transforms as T
-            import paddle.distributed.auto_parallel as auto
+            from paddle.distributed.fleet import auto
             from paddle.vision.datasets import MNIST
 
             transform = T.Compose([
@@ -540,7 +540,7 @@ class Engine:
 
                 import paddle
                 import paddle.vision.transforms as T
-                import paddle.distributed.auto_parallel as auto
+                from paddle.distributed.fleet import auto
                 from paddle.vision.datasets import MNIST
 
                 transform = T.Compose([
@@ -663,7 +663,7 @@ class Engine:
 
                 import paddle
                 import paddle.vision.transforms as T
-                import paddle.distributed.auto_parallel as auto
+                from paddle.distributed.fleet import auto
                 from paddle.vision.datasets import MNIST
 
                 transform = T.Compose([
@@ -771,7 +771,7 @@ class Engine:
 
                 import paddle
                 import paddle.vision.transforms as T
-                import paddle.distributed.auto_parallel as auto
+                from paddle.distributed.fleet import auto
                 from paddle.vision.datasets import MNIST
 
                 transform = T.Compose([
@@ -1041,7 +1041,7 @@ class Engine:
             .. code-block:: python
                 import paddle
                 import paddle.vision.transforms as T
-                import paddle.distributed.auto_parallel as auto
+                from paddle.distributed.fleet import auto
                 from paddle.vision.datasets import MNIST
 
                 transform = T.Compose([
@@ -1107,7 +1107,7 @@ class Engine:
             .. code-block:: python
                 import paddle
                 import paddle.vision.transforms as T
-                import paddle.distributed.auto_parallel as auto
+                from paddle.distributed.fleet import auto
                 from paddle.vision.datasets import MNIST
 
                 transform = T.Compose([

python/paddle/distributed/auto_parallel/interface.py

@@ -55,7 +55,7 @@ def shard_tensor(x, process_mesh=None, shard_spec=None):
         .. code-block:: python
 
             import paddle
-            import paddle.distributed.auto_parallel as auto
+            from paddle.distributed.fleet import auto
 
             mesh = auto.ProcessMesh([[0, 1], [2, 3]], dim_names=["x", "y"])
             x = paddle.ones([4, 6])
@@ -129,7 +129,7 @@ def shard_op(op, process_mesh=None, in_shard_specs=None, out_shard_specs=None):
         .. code-block:: python
 
             import paddle
-            import paddle.distributed.auto_parallel as auto
+            from paddle.distributed.fleet import auto
 
             x = paddle.ones([4, 6])
             y = paddle.zeros([4, 6])

python/paddle/distributed/auto_parallel/planner.py

@@ -22,7 +22,7 @@ from collections import OrderedDict
 import numpy as np
 
 import paddle
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from .cost_model import estimate_cost
 from .dist_op import DistributedOperator
 from .process_group import _g_process_group_map

python/paddle/distributed/auto_parallel/strategy.py

@@ -130,7 +130,7 @@ class Strategy(BaseConfig):
         .. code-block:: python
 
             import paddle
-            import paddle.distributed.auto_parallel as auto
+            from paddle.distributed.fleet import auto
 
             strategy = auto.Strategy()
             sharding = strategy.sharding

python/paddle/distributed/fleet/__init__.py

@@ -90,3 +90,5 @@ distributed_model = distributed_model
 shrink = fleet.shrink
 get_hybrid_communicate_group = fleet.get_hybrid_communicate_group
 distributed_scaler = distributed_scaler
+
+from .. import auto_parallel as auto

python/paddle/fluid/tests/unittests/auto_parallel/amp_pass_unittest.py

@@ -18,7 +18,7 @@ import random
 import numpy as np
 import paddle
 
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.fluid.dygraph.parallel import ParallelEnv
 from get_gpt_model import generate_model, create_data_holder, FakeDataset
 

python/paddle/fluid/tests/unittests/auto_parallel/auto_parallel_relaunch_model.py

@@ -28,7 +28,7 @@ import paddle.utils as utils
 from paddle.fluid import layers
 from paddle.io import IterableDataset, DataLoader
 from paddle.distributed import fleet
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 paddle.enable_static()
 _global_parallel_strategy = None

python/paddle/fluid/tests/unittests/auto_parallel/auto_parallel_relaunch_with_gpt_planner.py

@@ -19,7 +19,7 @@ import sys
 
 import numpy as np
 
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from auto_parallel_relaunch_model import mlp_pretrain_forward
 from auto_parallel_relaunch_model import batch_generator_creator
 

python/paddle/fluid/tests/unittests/auto_parallel/clip_grad_by_global_norm.py

@@ -18,7 +18,7 @@ import random
 import numpy as np
 import paddle
 
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.fluid.dygraph.parallel import ParallelEnv
 from get_gpt_model import generate_model, create_data_holder, FakeDataset
 

python/paddle/fluid/tests/unittests/auto_parallel/engine_api.py

@@ -28,7 +28,7 @@ import paddle.utils as utils
 from paddle.fluid import layers
 from paddle.io import Dataset, IterableDataset, DataLoader
 
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.optimizer.lr import CosineAnnealingDecay
 from paddle.fluid.dataloader.collate import default_collate_fn
 

python/paddle/fluid/tests/unittests/auto_parallel/engine_api_dp.py

@@ -28,7 +28,7 @@ import paddle.utils as utils
 from paddle.fluid import layers
 from paddle.io import Dataset, DataLoader
 
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 paddle.enable_static()
 batch_size = 2

python/paddle/fluid/tests/unittests/auto_parallel/get_gpt_model.py

@@ -17,7 +17,7 @@ import numpy as np
 import random
 
 import paddle
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 sys.path.append("..")
 import auto_parallel_gpt_model as modeling

python/paddle/fluid/tests/unittests/auto_parallel/gradient_merge_pass_unittest.py

@@ -18,7 +18,7 @@ import random
 import numpy as np
 import paddle
 
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.fluid.dygraph.parallel import ParallelEnv
 from get_gpt_model import generate_model, create_data_holder, FakeDataset
 

python/paddle/fluid/tests/unittests/auto_parallel/high_order_grad.py

@@ -16,7 +16,7 @@ import random
 import paddle
 import unittest
 import numpy as np
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.incubate.autograd import Hessian
 
 np.random.seed(1234)

python/paddle/fluid/tests/unittests/auto_parallel/iterable_dataset.py

@@ -29,7 +29,7 @@ from paddle.fluid import layers
 from paddle.io import Dataset, IterableDataset, DataLoader
 from paddle.static import InputSpec
 
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.optimizer.lr import CosineAnnealingDecay
 from paddle.fluid.dataloader.collate import default_collate_fn
 

python/paddle/fluid/tests/unittests/auto_parallel/optimization_tuner_api.py

@@ -28,7 +28,7 @@ import paddle.utils as utils
 from paddle.fluid import layers
 from paddle.io import Dataset, IterableDataset, DataLoader
 
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from engine_api_dp import MyDataset
 
 paddle.enable_static()

python/paddle/fluid/tests/unittests/auto_parallel/recompute_pass_unittest.py

@@ -18,7 +18,7 @@ import random
 import numpy as np
 import paddle
 
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.fluid.dygraph.parallel import ParallelEnv
 from get_gpt_model import generate_model, create_data_holder, FakeDataset
 

python/paddle/fluid/tests/unittests/auto_parallel/sharding_pass_unittest.py

@@ -18,7 +18,7 @@ import random
 import numpy as np
 import paddle
 
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.fluid.dygraph.parallel import ParallelEnv
 from get_gpt_model import generate_model, create_data_holder, FakeDataset
 

python/paddle/fluid/tests/unittests/auto_parallel/test_base_cost.py

@@ -24,7 +24,7 @@ import paddle.nn as nn
 import paddle.static as static
 import paddle.nn.functional as F
 import paddle.utils as utils
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.completion import Completer
 from paddle.distributed.auto_parallel.dist_context import DistributedContext
 from paddle.distributed import fleet

python/paddle/fluid/tests/unittests/auto_parallel/test_dist_context.py

@@ -25,7 +25,7 @@ import paddle.static as static
 import paddle.nn.functional as F
 
 from paddle.distributed import fleet
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.dist_context import DistributedContext
 from paddle.distributed.auto_parallel.utils import print_program_with_dist_attr
 

python/paddle/fluid/tests/unittests/auto_parallel/test_dist_embedding.py

@@ -14,7 +14,7 @@
 
 import unittest
 import paddle
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 from paddle.fluid import program_guard
 from paddle.fluid.backward import append_backward

python/paddle/fluid/tests/unittests/auto_parallel/test_dist_matmul.py

@@ -14,7 +14,7 @@
 
 import unittest
 import paddle
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 from paddle.fluid import program_guard
 from paddle.fluid.backward import append_backward

python/paddle/fluid/tests/unittests/auto_parallel/test_dist_op_cost.py

@@ -16,7 +16,7 @@ import unittest
 import copy
 
 import paddle
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.cluster import Cluster
 from paddle.distributed.auto_parallel.operators.common import get_distributed_operator_impl_container, is_elementwise_op
 

python/paddle/fluid/tests/unittests/auto_parallel/test_dist_pnorm.py

@@ -14,7 +14,7 @@
 
 import unittest
 import paddle
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 from paddle.fluid import program_guard
 from paddle.fluid.backward import append_backward

python/paddle/fluid/tests/unittests/auto_parallel/test_dist_reshape.py

@@ -14,7 +14,7 @@
 
 import unittest
 import paddle
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 from paddle.fluid import program_guard
 from paddle.fluid.backward import append_backward

python/paddle/fluid/tests/unittests/auto_parallel/test_dist_slice.py

@@ -14,7 +14,7 @@
 
 import unittest
 import paddle
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.utils import print_program_with_dist_attr
 
 paddle.enable_static()

python/paddle/fluid/tests/unittests/auto_parallel/test_interface.py

@@ -21,7 +21,7 @@ import paddle.nn as nn
 import paddle.nn.functional as F
 import paddle.static as static
 import paddle.distributed as dist
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.dist_context import get_default_distributed_context
 from paddle.distributed.auto_parallel.process_mesh import ProcessMesh
 from paddle.distributed.auto_parallel.utils import print_program_with_dist_attr

python/paddle/fluid/tests/unittests/auto_parallel/test_lr_grad_clip.py

@@ -20,7 +20,7 @@ import numpy as np
 import paddle
 import paddle.nn as nn
 import paddle.nn.functional as F
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 import paddle.distributed.fleet as fleet
 
 from paddle.io import Dataset

python/paddle/fluid/tests/unittests/auto_parallel/test_pass_quantization.py

@@ -18,7 +18,7 @@ import random
 import numpy as np
 import paddle
 
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from get_gpt_model import generate_model, create_data_holder, FakeDataset
 
 paddle.enable_static()

python/paddle/fluid/tests/unittests/auto_parallel/test_prim_dist_op.py

@@ -14,13 +14,13 @@
 
 import unittest
 import paddle
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 from paddle.fluid import program_guard
 from paddle.incubate.autograd import prim2orig, enable_prim, prim_enabled
 from paddle.fluid.layer_helper import LayerHelper
 from paddle.distributed.auto_parallel.utils import print_program_with_dist_attr
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.completion import Completer
 from paddle.distributed.auto_parallel.partitioner import Partitioner
 from paddle.distributed.auto_parallel.utils import set_var_dist_attr

python/paddle/fluid/tests/unittests/auto_parallel/test_process_mesh.py

@@ -19,7 +19,7 @@ import paddle.fluid as fluid
 import paddle.nn as nn
 import paddle.nn.functional as F
 import paddle.static as static
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.process_mesh import ProcessMesh
 from paddle.distributed.auto_parallel.dist_context import get_default_distributed_context
 from paddle.distributed.auto_parallel.utils import print_program_with_dist_attr

python/paddle/fluid/tests/unittests/auto_parallel/test_strategy.py

@@ -15,7 +15,7 @@
 import os
 # import yaml
 import unittest
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 
 class TestStrategy(unittest.TestCase):

python/paddle/fluid/tests/unittests/auto_parallel/test_to_static.py

@@ -20,7 +20,7 @@ import numpy as np
 import paddle
 import paddle.nn as nn
 import paddle.nn.functional as F
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 import paddle.distributed.fleet as fleet
 
 from paddle import LazyGuard

python/paddle/fluid/tests/unittests/auto_parallel/test_while_op_completion.py

@@ -19,7 +19,7 @@ import paddle.nn as nn
 import paddle.utils as utils
 import paddle.static as static
 import paddle.nn.functional as F
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 from paddle.distributed import fleet
 from paddle.distributed.auto_parallel.completion import Completer

python/paddle/fluid/tests/unittests/auto_parallel/test_while_op_partition.py

@@ -20,7 +20,7 @@ import paddle.utils as utils
 import paddle.fluid as fluid
 import paddle.static as static
 import paddle.nn.functional as F
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 from paddle.distributed import fleet
 from paddle.distributed.auto_parallel.completion import Completer

python/paddle/fluid/tests/unittests/auto_parallel_autoconvert.py

@@ -25,7 +25,7 @@ import paddle.nn as nn
 import paddle.utils as utils
 import paddle.static as static
 import paddle.nn.functional as F
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 from paddle.distributed import fleet
 from paddle.fluid.initializer import NumpyArrayInitializer

python/paddle/fluid/tests/unittests/auto_parallel_data_unshard.py

@@ -23,7 +23,7 @@ import random
 import paddle
 import paddle.nn as nn
 import paddle.fluid.core as core
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 import paddle.nn.functional as F
 from paddle.distributed import fleet
 

python/paddle/fluid/tests/unittests/auto_parallel_gpt_model.py

@@ -22,7 +22,7 @@ import paddle
 import paddle.nn as nn
 import paddle.nn.functional as F
 import paddle.tensor as tensor
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle import fluid
 from paddle.fluid import layers
 from paddle.distributed import fleet

python/paddle/fluid/tests/unittests/auto_parallel_save_load.py

@@ -25,7 +25,7 @@ import paddle.nn as nn
 import paddle.utils as utils
 import paddle.static as static
 import paddle.nn.functional as F
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 from paddle.distributed import fleet
 from paddle.fluid.initializer import NumpyArrayInitializer

python/paddle/fluid/tests/unittests/collective/fleet/auto_parallel_parallelizer.py

@@ -23,7 +23,7 @@ import paddle.nn.functional as F
 import paddle.utils as utils
 from paddle.fluid import layers
 from paddle.distributed import fleet
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.utils import print_program_with_dist_attr
 import paddle.fluid.core as core
 

python/paddle/fluid/tests/unittests/distributed_passes/auto_parallel_pass_test_base.py

@@ -25,7 +25,7 @@ from collections import OrderedDict
 from dist_pass_test_base import DistPassTestBase
 
 import paddle.distributed.fleet as fleet
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 sys.path.append("..")
 import auto_parallel_gpt_model as modeling

python/paddle/fluid/tests/unittests/distributed_passes/test_auto_parallel_data_parallel_optimization_pass.py

@@ -20,7 +20,7 @@ import unittest
 import paddle
 import paddle.nn as nn
 import paddle.distributed.fleet as fleet
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.dist_context import get_default_distributed_context
 from paddle.distributed.passes import new_pass, PassManager, PassContext
 from auto_parallel_pass_test_base import AutoPallelPassTestBase

python/paddle/fluid/tests/unittests/distributed_passes/test_auto_parallel_gradient_merge_pass.py

@@ -26,7 +26,7 @@ import paddle.utils as utils
 import paddle.static as static
 import paddle.nn.functional as F
 import paddle.distributed.fleet as fleet
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 
 from paddle.fluid.initializer import NumpyArrayInitializer
 from auto_parallel_pass_test_base import AutoPallelPassTestBase

python/paddle/fluid/tests/unittests/distributed_passes/test_auto_parallel_recompute_pass.py

@@ -20,7 +20,7 @@ import unittest
 import paddle
 import paddle.nn as nn
 import paddle.distributed.fleet as fleet
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.passes import new_pass, PassManager
 from auto_parallel_pass_test_base import AutoPallelPassTestBase
 

python/paddle/fluid/tests/unittests/distributed_passes/test_auto_parallel_sharding_pass.py

@@ -20,7 +20,7 @@ import unittest
 import paddle
 import paddle.nn as nn
 import paddle.distributed.fleet as fleet
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.passes import new_pass, PassManager
 from auto_parallel_pass_test_base import AutoPallelPassTestBase
 

python/paddle/fluid/tests/unittests/test_auto_parallel_completion.py

@@ -26,7 +26,7 @@ import paddle.utils as utils
 import paddle.tensor as tensor
 from paddle.fluid import layers
 from paddle.nn.layer.transformer import _convert_param_attr_to_list
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.completion import Completer
 from paddle.distributed.auto_parallel.utils import check_distributed_attr_for_program
 from paddle.distributed.auto_parallel.utils import print_program_with_dist_attr

python/paddle/fluid/tests/unittests/test_auto_parallel_completion_gpt.py

@@ -30,7 +30,7 @@ from paddle.nn.layer.transformer import _convert_param_attr_to_list
 from paddle.fluid.initializer import Normal, Constant, NumpyArrayInitializer
 from paddle.distributed.fleet import fleet
 import paddle.static as static
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.completion import Completer
 from paddle.distributed.auto_parallel.utils import check_distributed_attr_for_program
 from paddle.distributed.auto_parallel.utils import print_program_with_dist_attr

python/paddle/fluid/tests/unittests/test_auto_parallel_cost_model.py

@@ -22,7 +22,7 @@ import paddle.nn as nn
 import paddle.static as static
 import paddle.nn.functional as F
 import paddle.utils as utils
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.completion import Completer
 from paddle.distributed.auto_parallel.dist_context import DistributedContext
 from paddle.distributed import fleet

python/paddle/fluid/tests/unittests/test_auto_parallel_dist_tensor.py

@@ -17,7 +17,7 @@ import unittest
 
 import paddle
 from paddle.fluid import core
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.completion import Completer
 from paddle.distributed import fleet
 from paddle.distributed.auto_parallel.parallelizer import AutoParallelizer

python/paddle/fluid/tests/unittests/test_auto_parallel_mapper.py

@@ -36,7 +36,7 @@ from paddle.nn.layer.transformer import _convert_param_attr_to_list
 from paddle.fluid.initializer import Normal, Constant, NumpyArrayInitializer
 from paddle.distributed import fleet
 
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.completion import Completer
 from paddle.distributed.auto_parallel.parallelizer import AutoParallelizer
 from paddle.distributed.auto_parallel.dist_context import DistributedContext

python/paddle/fluid/tests/unittests/test_auto_parallel_partitioner.py

@@ -27,7 +27,7 @@ import paddle.utils as utils
 import paddle.tensor as tensor
 from paddle.fluid import layers
 from paddle.nn.layer.transformer import _convert_param_attr_to_list
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.completion import Completer
 from paddle.distributed.auto_parallel.utils import check_distributed_attr_for_program
 from paddle.distributed.auto_parallel.utils import print_program_with_dist_attr

python/paddle/fluid/tests/unittests/test_auto_parallel_partitioner_gpt.py

@@ -30,7 +30,7 @@ from paddle.nn.layer.transformer import _convert_param_attr_to_list
 from paddle.fluid.initializer import Normal, Constant, NumpyArrayInitializer
 from paddle.distributed import fleet
 import paddle.static as static
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.completion import Completer
 from paddle.distributed.auto_parallel.utils import check_distributed_attr_for_program
 from paddle.distributed.auto_parallel.utils import print_program_with_dist_attr

python/paddle/fluid/tests/unittests/test_auto_parallel_reshard.py

@@ -21,7 +21,7 @@ import paddle.nn as nn
 import paddle.static as static
 import paddle.nn.functional as F
 import paddle.utils as utils
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.completion import Completer
 from paddle.distributed.auto_parallel.dist_context import DistributedContext
 from paddle.distributed import fleet

python/paddle/fluid/tests/unittests/test_auto_parallel_reshard_dpmppp.py

@@ -21,7 +21,7 @@ import paddle.nn as nn
 import paddle.static as static
 import paddle.nn.functional as F
 import paddle.utils as utils
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.completion import Completer
 from paddle.distributed.auto_parallel.dist_context import DistributedContext
 from paddle.distributed import fleet

python/paddle/fluid/tests/unittests/test_auto_parallel_reshard_mppp.py

@@ -21,7 +21,7 @@ import paddle.nn as nn
 import paddle.static as static
 import paddle.nn.functional as F
 import paddle.utils as utils
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.completion import Completer
 from paddle.distributed.auto_parallel.dist_context import DistributedContext
 from paddle.distributed import fleet

python/paddle/fluid/tests/unittests/test_auto_parallel_reshard_serial.py

@@ -25,7 +25,7 @@ import paddle.nn as nn
 import paddle.static as static
 import paddle.nn.functional as F
 import paddle.utils as utils
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.dist_context import get_default_distributed_context
 from paddle.distributed import fleet
 from paddle.distributed.auto_parallel.partitioner import Partitioner

python/paddle/fluid/tests/unittests/test_auto_parallel_searcher.py

@@ -25,7 +25,7 @@ import paddle.static as static
 import paddle.nn.functional as F
 import paddle.utils as utils
 from paddle.distributed import fleet
-import paddle.distributed.auto_parallel as auto
+from paddle.distributed.fleet import auto
 from paddle.distributed.auto_parallel.cluster import Cluster
 from paddle.distributed.auto_parallel.utils import SerialProgramInfo
 from paddle.distributed.auto_parallel.planner import PlanSpace, PlanFilter