fix codestyle && some doc error

python/paddle/dataset/uci_housing.py

@@ -51,7 +51,7 @@
 UCI_TRAIN_DATA = None
 UCI_TEST_DATA = None
 
-FLUID_URL_MODEL = 'https://github.com/PaddlePaddle/book/raw/develop/01.fit_a_line/base/fit_a_line.base.tar'
+FLUID_URL_MODEL = 'https://github.com/PaddlePaddle/book/raw/develop/01.fit_a_line/base/fit_a_line.fluid.tar'
 FLUID_MD5_MODEL = '6e6dd637ccd5993961f68bfbde46090b'
 
 
@@ -152,7 +152,7 @@ def reader():
 
 def base_model():
     parameter_tar = paddle.dataset.common.download(
-        FLUID_URL_MODEL, 'uci_housing', FLUID_MD5_MODEL, 'fit_a_line.base.tar'
+        FLUID_URL_MODEL, 'uci_housing', FLUID_MD5_MODEL, 'fit_a_line.fluid.tar'
     )
 
     tar = tarfile.TarFile(parameter_tar, mode='r')

python/paddle/distributed/auto_parallel/static/auto_align_tool.py

@@ -21,6 +21,8 @@
 
 import paddle
 import paddle.distributed as dist
+from paddle.base import core
+from paddle.base.framework import Program
 from paddle.distributed.auto_parallel.static.converter import Converter
 from paddle.distributed.auto_parallel.static.dist_context import (
     get_default_distributed_context,
@@ -30,8 +32,6 @@
     is_forward_op,
     is_loss_op,
 )
-from paddle.base import core
-from paddle.base.framework import Program
 from paddle.static.io import deserialize_program
 
 _valid_types = [

python/paddle/distributed/auto_parallel/static/completion.py

@@ -15,8 +15,8 @@
 import copy
 import logging
 
-from paddle.distributed.fleet.meta_optimizers.common import OpRole
 from paddle.base.core import get_spmd_rule  # noqa: F401
+from paddle.distributed.fleet.meta_optimizers.common import OpRole
 from paddle.framework import core
 
 from ..process_mesh import ProcessMesh, compute_compatible_process_mesh

python/paddle/distributed/auto_parallel/static/engine.py

@@ -24,8 +24,8 @@
 import paddle
 import paddle.distributed.auto_parallel.static.utils as auto_utils
 from paddle import static, utils
-from paddle.distributed import fleet
 from paddle.base.executor import _to_name_str
+from paddle.distributed import fleet
 from paddle.framework import IrGraph
 from paddle.framework import _current_expected_place as _get_device
 from paddle.framework import core, in_dynamic_mode

python/paddle/distributed/auto_parallel/static/tuner/rule_based_tuner.py

@@ -26,6 +26,9 @@
 import numpy as np
 
 import paddle
+from paddle.base import program_guard
+from paddle.base.backward import append_backward
+from paddle.base.framework import Parameter, unique_name
 from paddle.distributed.auto_parallel.process_mesh import ProcessMesh
 from paddle.distributed.auto_parallel.static.cluster_v2 import DeviceMesh
 from paddle.distributed.auto_parallel.static.completion import Completer
@@ -48,9 +51,6 @@
     print_program_with_dist_attr,
 )
 from paddle.distributed.fleet.meta_optimizers.common import OpRole
-from paddle.base import program_guard
-from paddle.base.backward import append_backward
-from paddle.base.framework import Parameter, unique_name
 
 from ....utils.log_utils import get_logger
 from ..graph import Graph

python/paddle/distributed/communication/stream/all_gather.py

@@ -15,8 +15,8 @@
 import paddle
 import paddle.distributed as dist
 from paddle import framework
-from paddle.distributed.communication.group import _get_global_group
 from paddle.base import data_feeder
+from paddle.distributed.communication.group import _get_global_group
 
 
 def _all_gather_into_tensor_in_dygraph(

python/paddle/distributed/communication/stream/all_reduce.py

@@ -13,12 +13,12 @@
 # limitations under the License.
 
 from paddle import framework
+from paddle.base import data_feeder
 from paddle.distributed.communication.group import (
     _get_global_group,
     _warn_cur_rank_not_in_group,
 )
 from paddle.distributed.communication.reduce import ReduceOp, _get_reduce_op
-from paddle.base import data_feeder
 
 
 def _all_reduce_in_dygraph(tensor, op, group, sync_op, use_calc_stream):

python/paddle/distributed/communication/stream/all_to_all.py

@@ -15,11 +15,11 @@
 import paddle
 import paddle.distributed as dist
 from paddle import framework
+from paddle.base import data_feeder
 from paddle.distributed.communication.group import (
     _get_global_group,
     _warn_cur_rank_not_in_group,
 )
-from paddle.base import data_feeder
 
 
 def _all_to_all_tensor_in_dygraph(

python/paddle/distributed/communication/stream/broadcast.py

@@ -13,12 +13,12 @@
 # limitations under the License.
 
 from paddle import framework
+from paddle.base import data_feeder
 from paddle.distributed.communication.group import (
     _get_global_group,
     _get_or_throw_group_rank,
     _warn_cur_rank_not_in_group,
 )
-from paddle.base import data_feeder
 
 
 def _broadcast_in_dygraph(

python/paddle/distributed/communication/stream/recv.py

@@ -13,12 +13,12 @@
 # limitations under the License.
 
 from paddle import framework
+from paddle.base import data_feeder
 from paddle.distributed.communication.group import (
     _get_global_group,
     _get_or_throw_group_rank,
     _warn_cur_rank_not_in_group,
 )
-from paddle.base import data_feeder
 
 
 def _recv_in_dygraph(

python/paddle/distributed/communication/stream/reduce.py

@@ -13,13 +13,13 @@
 # limitations under the License.
 
 from paddle import framework
+from paddle.base import data_feeder
 from paddle.distributed.communication.group import (
     _get_global_group,
     _get_or_throw_group_rank,
     _warn_cur_rank_not_in_group,
 )
 from paddle.distributed.communication.reduce import ReduceOp, _get_reduce_op
-from paddle.base import data_feeder
 
 
 def _reduce_in_dygraph(

python/paddle/distributed/communication/stream/reduce_scatter.py

@@ -15,12 +15,12 @@
 import paddle
 import paddle.distributed as dist
 from paddle import framework
+from paddle.base import data_feeder
 from paddle.distributed.communication.group import (
     _get_global_group,
     _warn_cur_rank_not_in_group,
 )
 from paddle.distributed.communication.reduce import ReduceOp, _get_reduce_op
-from paddle.base import data_feeder
 
 
 def _reduce_scatter_tensor_in_dygraph(

python/paddle/distributed/communication/stream/scatter.py

@@ -17,12 +17,12 @@
 import paddle
 import paddle.distributed as dist
 from paddle import framework
+from paddle.base import data_feeder
 from paddle.distributed.communication.group import (
     _get_global_group,
     _get_or_throw_group_rank,
     _warn_cur_rank_not_in_group,
 )
-from paddle.base import data_feeder
 
 
 def _scatter_tensor_in_dygraph(

python/paddle/distributed/communication/stream/send.py

@@ -13,12 +13,12 @@
 # limitations under the License.
 
 from paddle import framework
+from paddle.base import data_feeder
 from paddle.distributed.communication.group import (
     _get_global_group,
     _get_or_throw_group_rank,
     _warn_cur_rank_not_in_group,
 )
-from paddle.base import data_feeder
 
 
 def _send_in_dygraph(

python/paddle/distributed/fleet/base/distributed_strategy.py

@@ -19,10 +19,10 @@
 import google.protobuf.text_format
 
 import paddle
-from paddle.distributed.fleet.proto import distributed_strategy_pb2
-from paddle.distributed.fleet.utils.log_util import logger
 from paddle.base.framework import _global_flags
 from paddle.base.wrapped_decorator import wrap_decorator
+from paddle.distributed.fleet.proto import distributed_strategy_pb2
+from paddle.distributed.fleet.utils.log_util import logger
 
 __all__ = []
 

python/paddle/distributed/fleet/base/role_maker.py

@@ -20,10 +20,10 @@
 import numpy as np
 
 import paddle
+from paddle.base import core
 from paddle.distributed.fleet.base.private_helper_function import (
     wait_server_ready,
 )
-from paddle.base import core
 
 from ...backup_env import getenv_or_backup
 

python/paddle/distributed/fleet/layers/mpu/mp_layers.py

@@ -14,8 +14,8 @@
 
 import paddle
 from paddle.autograd import PyLayer
-from paddle.distributed import fleet
 from paddle.base import core
+from paddle.distributed import fleet
 from paddle.nn import functional as F
 
 from ....communication.reduce import ReduceOp, _get_reduce_op

python/paddle/distributed/fleet/layers/mpu/mp_ops.py

@@ -15,8 +15,8 @@
 import paddle
 from paddle import _legacy_C_ops
 from paddle.autograd import PyLayer
-from paddle.distributed import collective
 from paddle.base.data_feeder import check_dtype, check_variable_and_dtype
+from paddle.distributed import collective
 from paddle.framework import LayerHelper, _create_tensor, in_dynamic_mode
 from paddle.nn import Layer
 from paddle.nn.utils import dygraph_utils

python/paddle/distributed/fleet/layers/mpu/random.py

@@ -18,9 +18,9 @@
 
 import paddle
 from paddle import _legacy_C_ops
-from paddle.common_ops_import import Variable
 from paddle.base import core
 from paddle.base.data_feeder import check_variable_and_dtype
+from paddle.common_ops_import import Variable
 from paddle.framework import LayerHelper, in_dynamic_mode
 
 __all__ = []

python/paddle/distributed/fleet/meta_optimizers/dgc_optimizer.py

@@ -19,9 +19,9 @@
 __all__ = []
 
 import paddle
-from paddle.common_ops_import import LayerHelper
 from paddle.base import framework
 from paddle.base.dygraph import base as imperative_base
+from paddle.common_ops_import import LayerHelper
 from paddle.framework import core, in_dynamic_mode
 from paddle.nn.clip import ClipGradByNorm, append_gradient_clip_ops
 from paddle.optimizer import Momentum, Optimizer

python/paddle/distributed/fleet/meta_parallel/sharding/group_sharded_stage3.py

@@ -22,8 +22,8 @@
 import paddle.distributed as dist
 from paddle import framework, nn
 from paddle.autograd import PyLayer
-from paddle.distributed import collective
 from paddle.base.framework import EagerParamBase
+from paddle.distributed import collective
 from paddle.framework import core
 from paddle.nn import ClipGradByGlobalNorm
 

python/paddle/distributed/fleet/meta_parallel/sharding/group_sharded_utils.py

@@ -20,9 +20,9 @@
 
 import paddle
 from paddle import _C_ops, _legacy_C_ops
-from paddle.common_ops_import import dygraph_only
 from paddle.base import core
 from paddle.base.dygraph import to_variable
+from paddle.common_ops_import import dygraph_only
 from paddle.nn import clip
 
 

python/paddle/distributed/fleet/recompute/recompute_hybrid.py

@@ -158,7 +158,7 @@ def forward(
                 # Note:
                 #  If not marked non_differentiable, all output tensors' attr `stop gradient`
                 #  will be reset to `False` in c++ backend.
-                #  See https://github.com/PaddlePaddle/Paddle/blob/9d62efb0e6e5373823039d9eda96cd5905426c0a/paddle/base/pybind/eager_py_layer.cc#L388
+                #  See https://github.com/PaddlePaddle/Paddle/blob/9d62efb0e6e5373823039d9eda96cd5905426c0a/paddle/fluid/pybind/eager_py_layer.cc#L388
                 if framework.in_dynamic_mode() and state:
                     ctx.mark_non_differentiable(arg)
             else:

python/paddle/distributed/fleet/scaler.py

@@ -18,8 +18,8 @@
 
 import paddle
 from paddle import _C_ops, _legacy_C_ops
-from paddle.distributed import fleet
 from paddle.base.dygraph import to_variable
+from paddle.distributed import fleet
 from paddle.framework import core
 
 from .base.topology import ParallelMode

python/paddle/distributed/fleet/utils/hybrid_parallel_inference.py

@@ -16,10 +16,9 @@
 
 import numpy as np
 
-from paddle.distributed import fleet
-
 # (TODO: GhostScreaming) It will be removed later.
 from paddle.base import core
+from paddle.distributed import fleet
 from paddle.framework import Block, Program, in_dynamic_mode
 
 

python/paddle/distributed/fleet/utils/hybrid_parallel_util.py

@@ -14,16 +14,16 @@
 
 import paddle
 from paddle import framework
+
+# (TODO: GhostScreaming) It will be removed later.
+from paddle.base import core
 from paddle.distributed.parallel import (
     _split_tensors,
     build_groups,
     in_dynamic_mode,
     sync_params_buffers,
 )
 
-# (TODO: GhostScreaming) It will be removed later.
-from paddle.base import core
-
 from .log_util import logger
 
 __all__ = []

python/paddle/distributed/fleet/utils/mix_precision_utils.py

@@ -20,13 +20,13 @@
 
 import paddle
 from paddle import _legacy_C_ops, nn
+from paddle.base import framework
+from paddle.base.dygraph import base as imperative_base
+from paddle.base.dygraph import to_variable
 from paddle.distributed import fleet
 from paddle.distributed.fleet.utils.hybrid_parallel_util import (
     obtain_optimizer_parameters_list,
 )
-from paddle.base import framework
-from paddle.base.dygraph import base as imperative_base
-from paddle.base.dygraph import to_variable
 from paddle.framework import core
 
 

python/paddle/distributed/fleet/utils/sequence_parallel_utils.py

@@ -17,12 +17,12 @@
 import paddle
 from paddle import distributed as dist
 from paddle.autograd import PyLayer
+from paddle.base import core
 from paddle.distributed import fleet
 from paddle.distributed.fleet.meta_parallel import get_rng_state_tracker
 from paddle.distributed.fleet.utils.hybrid_parallel_util import (
     fused_allreduce_gradients_with_group,
 )
-from paddle.base import core
 from paddle.nn import Layer
 from paddle.nn import functional as F
 

python/paddle/distributed/fleet/utils/tensor_parallel_utils.py

@@ -22,8 +22,8 @@
 ch.setFormatter(formatter)
 logger.addHandler(ch)
 
-from paddle.distributed.fleet.meta_optimizers.common import OP_ROLE_KEY
 from paddle.base import core
+from paddle.distributed.fleet.meta_optimizers.common import OP_ROLE_KEY
 from paddle.static import Parameter
 
 _supported_optimizer_type = [

python/paddle/distributed/launch/context/device.py

@@ -14,10 +14,9 @@
 
 import os
 
-from paddle.device import get_available_custom_device
-
 # (TODO: GhostScreaming) It will be removed later.
 from paddle.base import core
+from paddle.device import get_available_custom_device
 
 
 class DeviceType: