diff --git a/python/paddle/common_ops_import.py b/python/paddle/common_ops_import.py
index 47bce838936d442e4237fc5f04b093520e457646..7169671ad3c246b4560e6e1878e6a33e5f1cdc53 100644
--- a/python/paddle/common_ops_import.py
+++ b/python/paddle/common_ops_import.py
@@ -36,6 +36,6 @@ from paddle.fluid.data_feeder import (  # noqa: F401
     check_variable_and_dtype,
     convert_dtype,
 )
-from paddle.fluid.layers import fill_constant, utils, scale  # noqa: F401
+from paddle.fluid.layers import fill_constant, utils  # noqa: F401
 from paddle.tensor.layer_function_generator import templatedoc  # noqa: F401
 import paddle.fluid as fluid  # noqa: F401
diff --git a/python/paddle/fluid/dygraph/io.py b/python/paddle/fluid/dygraph/io.py
index f84949aa5e014e3d57cbf8bed935396450b2b5f6..ba00deed977a97f39db8c8a6eecdd1903ce2d292 100644
--- a/python/paddle/fluid/dygraph/io.py
+++ b/python/paddle/fluid/dygraph/io.py
@@ -522,7 +522,7 @@ class _ProgramHolder:
         with framework.program_guard(program):
             for i, out in enumerate(self._output_descs):
                 var = program.global_block().var(out.name())
-                var = nn.scale(
+                var = paddle.scale(
                     var, 1.0, name="translated_layer/scale_{}".format(i)
                 )
                 scale_output_vars.append(var)
diff --git a/python/paddle/fluid/layers/io.py b/python/paddle/fluid/layers/io.py
index e92fea5afb1b06f718f153d7f2ded163616f5a9a..2259044e14e79e41edb08b79483d392836f3a835 100644
--- a/python/paddle/fluid/layers/io.py
+++ b/python/paddle/fluid/layers/io.py
@@ -189,6 +189,7 @@ class ListenAndServ:
         .. code-block:: python
 
             import paddle.fluid as fluid
+            import paddle
             with fluid.program_guard(main):
                 serv = layers.ListenAndServ(
                     "127.0.0.1:6170", ["X"], optimizer_mode=False)
@@ -199,7 +200,7 @@ class ListenAndServ:
                         name="X",
                         append_batch_size=False)
                     fluid.initializer.Constant(value=1.0)(x, main.global_block())
-                    layers.scale(x=x, scale=10.0, out=out_var)
+                    paddle.scale(x=x, scale=10.0, out=out_var)
 
             exe = fluid.Executor(place)
             exe.run(main)
diff --git a/python/paddle/fluid/layers/nn.py b/python/paddle/fluid/layers/nn.py
index 18bbc6bc760d4829c0d4198c5d958d4ace993551..d412f344b096f76ac54a8d29c91cafb3644d3d35 100644
--- a/python/paddle/fluid/layers/nn.py
+++ b/python/paddle/fluid/layers/nn.py
@@ -113,7 +113,6 @@ __all__ = [
     'flatten',
     'unique',
     'unique_with_counts',
-    'scale',
     'elementwise_add',
     'elementwise_div',
     'elementwise_sub',
@@ -7924,103 +7923,6 @@ def _elementwise_op(helper):
     return helper.append_activation(out)
 
 
-def scale(x, scale=1.0, bias=0.0, bias_after_scale=True, act=None, name=None):
-    """
-
-    Putting scale and bias to the input Tensor as following:
-
-    ``bias_after_scale`` is True:
-
-    .. math::
-                            Out=scale*X+bias
-
-    ``bias_after_scale`` is False:
-
-    .. math::
-                            Out=scale*(X+bias)
-
-    Args:
-        x(Tensor): Input N-D Tensor of scale operator. Data type can be float32, float64, int8, int16, int32, int64, uint8.
-        scale(float|Tensor): The scale factor of the input, it should be a float number or a Tensor with shape [1] and data type as float32.
-        bias(float): The bias to be put on the input.
-        bias_after_scale(bool): Apply bias addition after or before scaling. It is useful for numeric stability in some circumstances.
-        act(str, optional): Activation applied to the output such as tanh, softmax, sigmoid, relu.
-        name(str, optional): The default value is None. Normally there is no need for user to set this property.  For more information, please refer to :ref:`api_guide_Name`
-
-    Returns:
-        Tensor: Output tensor of scale operator, with shape and data type same as input.
-
-    Examples:
-
-        .. code-block:: python
-
-            # scale as a float32 number
-            import paddle
-
-            data = paddle.randn(shape=[2,3], dtype='float32')
-            res = paddle.scale(data, scale=2.0, bias=1.0)
-
-        .. code-block:: python
-
-            # scale with parameter scale as a Tensor
-            import paddle
-
-            data = paddle.randn(shape=[2, 3], dtype='float32')
-            factor = paddle.to_tensor([2], dtype='float32')
-            res = paddle.scale(data, scale=factor, bias=1.0)
-
-    """
-
-    if in_dygraph_mode():
-        out = _C_ops.scale(x, scale, float(bias), bias_after_scale)
-        return dygraph_utils._append_activation_in_dygraph(out)
-    if _non_static_mode():
-        _scale = scale.numpy().item(0) if isinstance(scale, Variable) else scale
-        out = _legacy_C_ops.scale(
-            x,
-            'scale',
-            float(_scale),
-            'bias',
-            float(bias),
-            'bias_after_scale',
-            bias_after_scale,
-        )
-        return dygraph_utils._append_activation_in_dygraph(out)
-
-    check_variable_and_dtype(
-        x,
-        "x",
-        [
-            'float16',
-            'uint16',
-            'float32',
-            'float64',
-            'int8',
-            'int16',
-            'int32',
-            'int64',
-            'uint8',
-        ],
-        "scale",
-    )
-    inputs = {'X': [x]}
-    attrs = {
-        'bias': float(bias),
-        'bias_after_scale': bias_after_scale,
-    }
-    if isinstance(scale, Variable):
-        inputs['ScaleTensor'] = [scale]
-    else:
-        attrs['scale'] = float(scale)
-    helper = LayerHelper('scale', **locals())
-    out = helper.create_variable_for_type_inference(dtype=x.dtype)
-
-    helper.append_op(
-        type='scale', inputs=inputs, outputs={'Out': out}, attrs=attrs
-    )
-    return helper.append_activation(out)
-
-
 def elementwise_add(x, y, axis=-1, act=None, name=None):
     """
 
diff --git a/python/paddle/fluid/nets.py b/python/paddle/fluid/nets.py
index c4bd6cf81f5f425711b01590a101b6354d40bf58..5cd8380eba58651c4f2a7e896a73f4f82f547ed1 100644
--- a/python/paddle/fluid/nets.py
+++ b/python/paddle/fluid/nets.py
@@ -620,7 +620,7 @@ def scaled_dot_product_attention(
     v = __split_heads(v, num_heads)
 
     key_dim_per_head = keys.shape[-1] // num_heads
-    scaled_q = layers.scale(x=q, scale=key_dim_per_head**-0.5)
+    scaled_q = paddle.scale(x=q, scale=key_dim_per_head**-0.5)
     product = layers.matmul(x=scaled_q, y=k, transpose_y=True)
 
     x = paddle.reshape(x=product, shape=[-1, product.shape[-1]])
diff --git a/python/paddle/fluid/tests/book/test_recommender_system.py b/python/paddle/fluid/tests/book/test_recommender_system.py
index b947102d23a30902260c7c0ba205539290cb6c5a..c270c87b3c2acc44dce6d24b9722890d40ac6ee6 100644
--- a/python/paddle/fluid/tests/book/test_recommender_system.py
+++ b/python/paddle/fluid/tests/book/test_recommender_system.py
@@ -162,7 +162,7 @@ def model():
 
     # need cos sim
     inference = layers.cos_sim(X=usr_combined_features, Y=mov_combined_features)
-    scale_infer = layers.scale(x=inference, scale=5.0)
+    scale_infer = paddle.scale(x=inference, scale=5.0)
 
     label = layers.data(name='score', shape=[1], dtype='float32')
     square_cost = layers.square_error_cost(input=scale_infer, label=label)
diff --git a/python/paddle/fluid/tests/unittests/collective/fleet/parallel_dygraph_transformer.py b/python/paddle/fluid/tests/unittests/collective/fleet/parallel_dygraph_transformer.py
index c052de953dec193a9f230290dcf23c7a451e2bd1..bcdd7329bb18e73d6c2d284a3b192a069caf353e 100644
--- a/python/paddle/fluid/tests/unittests/collective/fleet/parallel_dygraph_transformer.py
+++ b/python/paddle/fluid/tests/unittests/collective/fleet/parallel_dygraph_transformer.py
@@ -537,7 +537,7 @@ class PrepareEncoderDecoderLayer(Layer):
 
     def forward(self, src_word, src_pos):
         src_word_emb = self._input_emb(src_word)
-        src_word_emb = fluid.layers.scale(
+        src_word_emb = paddle.scale(
             x=src_word_emb, scale=self._src_emb_dim**0.5
         )
         # # TODO change this to fit dynamic length input
diff --git a/python/paddle/fluid/tests/unittests/dist_transformer.py b/python/paddle/fluid/tests/unittests/dist_transformer.py
index 5a02ddf7005954a998dc5f553dfc782f1acf9150..4778a942dd5f79acc94169d1ef76b562ddbd0c91 100644
--- a/python/paddle/fluid/tests/unittests/dist_transformer.py
+++ b/python/paddle/fluid/tests/unittests/dist_transformer.py
@@ -1173,7 +1173,7 @@ def multi_head_attention(
         """
         Scaled Dot-Product Attention
         """
-        scaled_q = layers.scale(x=q, scale=d_model**-0.5)
+        scaled_q = paddle.scale(x=q, scale=d_model**-0.5)
         product = layers.matmul(x=scaled_q, y=k, transpose_y=True)
         if attn_bias:
             product += attn_bias
@@ -1305,7 +1305,7 @@ def prepare_encoder(
             ),
         )
 
-    src_word_emb = layers.scale(x=src_word_emb, scale=src_emb_dim**0.5)
+    src_word_emb = paddle.scale(x=src_word_emb, scale=src_emb_dim**0.5)
     src_pos_enc = layers.embedding(
         src_pos,
         size=[src_max_len, src_emb_dim],
diff --git a/python/paddle/fluid/tests/unittests/dygraph_to_static/bert_dygraph_model.py b/python/paddle/fluid/tests/unittests/dygraph_to_static/bert_dygraph_model.py
index 0bccfea79d1781791d0722dc6c3eeb6eb474102b..f35e7a973d3fa3aa48f93c7531cca331a4b7912c 100644
--- a/python/paddle/fluid/tests/unittests/dygraph_to_static/bert_dygraph_model.py
+++ b/python/paddle/fluid/tests/unittests/dygraph_to_static/bert_dygraph_model.py
@@ -276,7 +276,7 @@ class BertModelLayer(Layer):
         self_attn_mask = fluid.layers.matmul(
             x=input_mask, y=input_mask, transpose_y=True
         )
-        self_attn_mask = fluid.layers.scale(
+        self_attn_mask = paddle.scale(
             x=self_attn_mask, scale=10000.0, bias=-1.0, bias_after_scale=False
         )
         n_head_self_attn_mask = paddle.stack(
diff --git a/python/paddle/fluid/tests/unittests/dygraph_to_static/transformer_dygraph_model.py b/python/paddle/fluid/tests/unittests/dygraph_to_static/transformer_dygraph_model.py
index dc2f00584cc039d6ac2aab5231d299e4c2c3c94f..9819b79ae6f768a90fd273e2e7a6a69179dfb93f 100644
--- a/python/paddle/fluid/tests/unittests/dygraph_to_static/transformer_dygraph_model.py
+++ b/python/paddle/fluid/tests/unittests/dygraph_to_static/transformer_dygraph_model.py
@@ -342,7 +342,7 @@ class WrapEncoder(Layer):
 
     def forward(self, src_word, src_pos, src_slf_attn_bias):
         word_emb = self.word_embedder(src_word)
-        word_emb = layers.scale(x=word_emb, scale=self.emb_dim**0.5)
+        word_emb = paddle.scale(x=word_emb, scale=self.emb_dim**0.5)
         pos_enc = self.pos_encoder(src_pos)
         pos_enc.stop_gradient = True
         emb = word_emb + pos_enc
@@ -546,7 +546,7 @@ class WrapDecoder(Layer):
         caches=None,
     ):
         word_emb = self.word_embedder(trg_word)
-        word_emb = layers.scale(x=word_emb, scale=self.emb_dim**0.5)
+        word_emb = paddle.scale(x=word_emb, scale=self.emb_dim**0.5)
         pos_enc = self.pos_encoder(trg_pos)
         pos_enc.stop_gradient = True
         emb = word_emb + pos_enc
diff --git a/python/paddle/fluid/tests/unittests/ipu/test_scale_op_ipu.py b/python/paddle/fluid/tests/unittests/ipu/test_scale_op_ipu.py
index 17d5b8207c1211a07fdc2b6b8d1afba8bb29d729..e95a86799e32a5b4abc7515fac6dc1eb60c18ae5 100644
--- a/python/paddle/fluid/tests/unittests/ipu/test_scale_op_ipu.py
+++ b/python/paddle/fluid/tests/unittests/ipu/test_scale_op_ipu.py
@@ -55,7 +55,7 @@ class TestBase(IPUOpTest):
         x = paddle.static.data(
             name=self.feed_list[0], shape=self.feed_shape[0], dtype='float32'
         )
-        out = paddle.fluid.layers.scale(x, **self.attrs)
+        out = paddle.scale(x, **self.attrs)
         self.fetch_list = [out.name]
 
     def run_model(self, exec_mode):
@@ -126,7 +126,7 @@ class TestCase5(TestBase):
         y = paddle.static.data(
             name=self.feed_list[1], shape=self.feed_shape[1], dtype='float32'
         )
-        out = paddle.fluid.layers.scale(x, scale=y, **self.attrs)
+        out = paddle.scale(x, scale=y, **self.attrs)
         self.fetch_list = [out.name]
 
 
diff --git a/python/paddle/fluid/tests/unittests/ipu/test_varname_inplace_ipu.py b/python/paddle/fluid/tests/unittests/ipu/test_varname_inplace_ipu.py
index bc5464941a574f061da4e03c199e2a3dfaeed059..61b38e80e7873ac9a7e675f303e9e714696ea619 100644
--- a/python/paddle/fluid/tests/unittests/ipu/test_varname_inplace_ipu.py
+++ b/python/paddle/fluid/tests/unittests/ipu/test_varname_inplace_ipu.py
@@ -62,9 +62,9 @@ class TestBase(IPUOpTest):
                 add1 = paddle.fluid.layers.elementwise_add(x, x)
                 reshape = paddle.reshape(add1, **self.attrs)
                 add2 = paddle.fluid.layers.elementwise_add(reshape, reshape)
-                scale1 = paddle.fluid.layers.scale(add2)
-                scale2 = paddle.fluid.layers.scale(scale1, scale=1.3, bias=0.5)
-                scale3 = paddle.fluid.layers.scale(scale2, scale=2, bias=0.7)
+                scale1 = paddle.scale(add2)
+                scale2 = paddle.scale(scale1, scale=1.3, bias=0.5)
+                scale3 = paddle.scale(scale2, scale=2, bias=0.7)
 
             fetch_list = [scale3.name]
 
diff --git a/python/paddle/fluid/tests/unittests/ir/inference/test_trt_scale_op.py b/python/paddle/fluid/tests/unittests/ir/inference/test_trt_scale_op.py
index ffa83eb875595e408a0535156480793fc5ccb5b5..0ffabd017814131ecee30ad7b852a424bc9d24cd 100644
--- a/python/paddle/fluid/tests/unittests/ir/inference/test_trt_scale_op.py
+++ b/python/paddle/fluid/tests/unittests/ir/inference/test_trt_scale_op.py
@@ -17,6 +17,7 @@ import unittest
 import numpy as np
 from inference_pass_test import InferencePassTest
 
+import paddle
 import paddle.fluid as fluid
 import paddle.fluid.core as core
 from paddle.fluid.core import AnalysisConfig, PassVersionChecker
@@ -39,7 +40,7 @@ class TRTScaleTest(InferencePassTest):
         self.fetch_list = [out]
 
     def append_scale(self, data):
-        return fluid.layers.scale(
+        return paddle.scale(
             x=data, scale=2.0, bias=-1.0, bias_after_scale=False
         )
 
@@ -71,7 +72,7 @@ class TRTScaleShape2Test(InferencePassTest):
         self.fetch_list = [out]
 
     def append_scale(self, data):
-        return fluid.layers.scale(
+        return paddle.scale(
             x=data, scale=2.0, bias=-1.0, bias_after_scale=False
         )
 
diff --git a/python/paddle/fluid/tests/unittests/ir/test_ir_fusion_group_pass.py b/python/paddle/fluid/tests/unittests/ir/test_ir_fusion_group_pass.py
index 4d4fdc3e279af4d1b5fb1ed6a14b919d497900d6..19754fd6f2da0540f1f70fbb4cb240feb5e6cd8e 100644
--- a/python/paddle/fluid/tests/unittests/ir/test_ir_fusion_group_pass.py
+++ b/python/paddle/fluid/tests/unittests/ir/test_ir_fusion_group_pass.py
@@ -207,7 +207,7 @@ class FusionGroupPassFillConstantTest(FusionGroupPassTest):
 
             tmp_0 = layers.elementwise_add(self.feed_vars[0], self.feed_vars[1])
             tmp_1 = layers.fill_constant(shape=[2, 2], dtype=dtype, value=2.0)
-            tmp_2 = layers.scale(
+            tmp_2 = paddle.scale(
                 tmp_1, scale=3.0, bias=1.0, bias_after_scale=True
             )
             tmp_3 = layers.elementwise_mul(tmp_2, tmp_0)
diff --git a/python/paddle/fluid/tests/unittests/mlu/test_scale_op_mlu.py b/python/paddle/fluid/tests/unittests/mlu/test_scale_op_mlu.py
index c7858a482295d534c1ed472007cd76ce2a951645..df164e26950619409a9fa93fa2fc4358064b52e2 100644
--- a/python/paddle/fluid/tests/unittests/mlu/test_scale_op_mlu.py
+++ b/python/paddle/fluid/tests/unittests/mlu/test_scale_op_mlu.py
@@ -131,7 +131,7 @@ class TestScaleOpSelectedRows(unittest.TestCase):
 class TestScaleRaiseError(unittest.TestCase):
     def test_errors(self):
         def test_type():
-            fluid.layers.scale([10])
+            paddle.scale([10])
 
         self.assertRaises(TypeError, test_type)
 
diff --git a/python/paddle/fluid/tests/unittests/test_array_read_write_op.py b/python/paddle/fluid/tests/unittests/test_array_read_write_op.py
index f5f54017d9983a9c0d420bceaeed3a002e9e6cbf..e8d2579b3f2914506529199252eaf1ba0b5e32a4 100644
--- a/python/paddle/fluid/tests/unittests/test_array_read_write_op.py
+++ b/python/paddle/fluid/tests/unittests/test_array_read_write_op.py
@@ -80,7 +80,7 @@ class TestArrayReadWrite(unittest.TestCase):
         self.assertEqual(outs[0], outs[1])
 
         total_sum = layers.sums(input=[a_sum, x_sum])
-        total_sum_scaled = layers.scale(x=total_sum, scale=1 / 6.0)
+        total_sum_scaled = paddle.scale(x=total_sum, scale=1 / 6.0)
 
         append_backward(total_sum_scaled)
 
@@ -117,7 +117,7 @@ class TestArrayReadWrite(unittest.TestCase):
             total_sum_dygraph = layers.sums(
                 input=[a_sum_dygraph, x_sum_dygraph]
             )
-            total_sum_scaled_dygraph = layers.scale(
+            total_sum_scaled_dygraph = paddle.scale(
                 x=total_sum_dygraph, scale=1 / 6.0
             )
             total_sum_scaled_dygraph.backward()
diff --git a/python/paddle/fluid/tests/unittests/test_dist_train.py b/python/paddle/fluid/tests/unittests/test_dist_train.py
index 65753803282fa977bbc410629070ef55d29d7026..3067321289b9cf1eae8d8972745a951a989a322c 100644
--- a/python/paddle/fluid/tests/unittests/test_dist_train.py
+++ b/python/paddle/fluid/tests/unittests/test_dist_train.py
@@ -29,6 +29,7 @@ import paddle.fluid.layers.ops as ops
 from dist_test_utils import remove_ps_flag
 
 from paddle.fluid import core
+import paddle
 
 RPC_OP_ROLE_ATTR_NAME = (
     op_role_attr_name
@@ -150,7 +151,7 @@ class TestSendOp(unittest.TestCase):
                 append_batch_size=False,
             )
             fluid.initializer.Constant(value=2.3)(x, main.global_block())
-            o = layers.scale(x=x, scale=10.0)
+            o = paddle.scale(x=x, scale=10.0)
         exe = fluid.Executor(place)
         self.local_out = exe.run(main, fetch_list=[o])
 
diff --git a/python/paddle/fluid/tests/unittests/test_eager_deletion_recurrent_op.py b/python/paddle/fluid/tests/unittests/test_eager_deletion_recurrent_op.py
index 21ebf05038e8ce44ca713add39bca8738e311c55..7d53e2aac86233a3847c731834a399fb63b0a961 100644
--- a/python/paddle/fluid/tests/unittests/test_eager_deletion_recurrent_op.py
+++ b/python/paddle/fluid/tests/unittests/test_eager_deletion_recurrent_op.py
@@ -155,7 +155,7 @@ class EagerDeletionRecurrentOpTest1(unittest.TestCase):
             h_pre = rnn.memory(init=h_boot)
             x_t = rnn.step_input(x)
 
-            h = layers.scale(
+            h = paddle.scale(
                 x=layers.elementwise_add(x=h_pre, y=x_t),
                 scale=self.py_rnn.scale,
             )
@@ -431,8 +431,8 @@ class EagerDeletionRecurrentOpMultipleMemoryTest(EagerDeletionRecurrentOpTest1):
             h_pre2 = rnn.memory(init=h_boot2)
             x_t = rnn.step_input(x)
 
-            mem1 = layers.scale(x=h_pre1, scale=1.0)
-            mem2 = layers.scale(x=h_pre2, scale=1.0)
+            mem1 = paddle.scale(x=h_pre1, scale=1.0)
+            mem2 = paddle.scale(x=h_pre2, scale=1.0)
             out = layers.sums(input=[mem1, x_t, mem2])
 
             rnn.update_memory(h_pre1, mem1)
@@ -691,7 +691,7 @@ class EagerDeletionFarwardOnlyRnnAndBackwardRnnTest(
                 h_pre = forward_only_rnn.memory(init=h_boot)
                 x_t = forward_only_rnn.step_input(x)
 
-                h = layers.scale(
+                h = paddle.scale(
                     x=layers.elementwise_add(x=h_pre, y=x_t),
                     scale=self.py_rnn.scale,
                 )
@@ -707,7 +707,7 @@ class EagerDeletionFarwardOnlyRnnAndBackwardRnnTest(
                 h_pre = rnn.memory(init=h_boot)
                 x_t = rnn.step_input(x)
 
-                h = layers.scale(
+                h = paddle.scale(
                     x=layers.elementwise_add(x=h_pre, y=x_t),
                     scale=self.py_rnn.scale,
                 )
diff --git a/python/paddle/fluid/tests/unittests/test_imperative_transformer_sorted_gradient.py b/python/paddle/fluid/tests/unittests/test_imperative_transformer_sorted_gradient.py
index f8671a76e2cd64e09486efb3e102703325f57336..b5c398b36c26a626afb38bbb29044e833c774851 100644
--- a/python/paddle/fluid/tests/unittests/test_imperative_transformer_sorted_gradient.py
+++ b/python/paddle/fluid/tests/unittests/test_imperative_transformer_sorted_gradient.py
@@ -692,7 +692,7 @@ class PrepareEncoderDecoderLayer(Layer):
 
     def forward(self, src_word, src_pos):
         src_word_emb = self._input_emb(src_word)
-        src_word_emb = fluid.layers.scale(
+        src_word_emb = paddle.scale(
             x=src_word_emb, scale=self._src_emb_dim**0.5
         )
         # # TODO change this to fit dynamic length input
diff --git a/python/paddle/fluid/tests/unittests/test_layers.py b/python/paddle/fluid/tests/unittests/test_layers.py
index f62adf08ecf952dc32dd5971720299ac705f4dbd..752a089dadf834f9cfdc15202401ecabe8a9dbb1 100644
--- a/python/paddle/fluid/tests/unittests/test_layers.py
+++ b/python/paddle/fluid/tests/unittests/test_layers.py
@@ -3632,7 +3632,7 @@ class TestBook(LayerTest):
                 dtype='float32',
                 append_batch_size=False,
             )
-            out = layers.scale(input, scale=scale_var)
+            out = paddle.scale(input, scale=scale_var)
             return out
 
     def make_iou_similarity(self):
diff --git a/python/paddle/fluid/tests/unittests/test_recurrent_op.py b/python/paddle/fluid/tests/unittests/test_recurrent_op.py
index e8069c5f06363dfb25b428015caa841ff50f36b0..e0abda02b37f3ba4ef9e8fdbb7e165b3cce95ab1 100644
--- a/python/paddle/fluid/tests/unittests/test_recurrent_op.py
+++ b/python/paddle/fluid/tests/unittests/test_recurrent_op.py
@@ -151,7 +151,7 @@ class RecurrentOpTest1(unittest.TestCase):
             h_pre = rnn.memory(init=h_boot)
             x_t = rnn.step_input(x)
 
-            h = layers.scale(
+            h = paddle.scale(
                 x=layers.elementwise_add(x=h_pre, y=x_t),
                 scale=self.py_rnn.scale,
             )
@@ -419,8 +419,8 @@ class RecurrentOpMultipleMemoryTest(RecurrentOpTest1):
             h_pre2 = rnn.memory(init=h_boot2)
             x_t = rnn.step_input(x)
 
-            mem1 = layers.scale(x=h_pre1, scale=1.0)
-            mem2 = layers.scale(x=h_pre2, scale=1.0)
+            mem1 = paddle.scale(x=h_pre1, scale=1.0)
+            mem2 = paddle.scale(x=h_pre2, scale=1.0)
             out = layers.sums(input=[mem1, x_t, mem2])
 
             rnn.update_memory(h_pre1, mem1)
diff --git a/python/paddle/fluid/tests/unittests/test_scale_op.py b/python/paddle/fluid/tests/unittests/test_scale_op.py
index 07be7620a93a1174dc8cd8a63dd6708bcc730341..8082128a02fce46156869a44075a0fa6a293d4f7 100644
--- a/python/paddle/fluid/tests/unittests/test_scale_op.py
+++ b/python/paddle/fluid/tests/unittests/test_scale_op.py
@@ -133,7 +133,7 @@ class TestScaleOpSelectedRows(unittest.TestCase):
 class TestScaleRaiseError(unittest.TestCase):
     def test_errors(self):
         def test_type():
-            fluid.layers.scale([10])
+            paddle.scale([10])
 
         self.assertRaises(TypeError, test_type)
 
diff --git a/python/paddle/fluid/tests/unittests/transformer_model.py b/python/paddle/fluid/tests/unittests/transformer_model.py
index cf564e771e26f0e34633c9758c8d6700a8572986..dfb29fdeb40358cb8243f6894f436e41488ede37 100644
--- a/python/paddle/fluid/tests/unittests/transformer_model.py
+++ b/python/paddle/fluid/tests/unittests/transformer_model.py
@@ -161,7 +161,7 @@ def multi_head_attention(
             sum_out = layers.reduce_sum(exp_out, dim=-1, keep_dim=False)
             return layers.elementwise_div(x=exp_out, y=sum_out, axis=0)
 
-        scaled_q = layers.scale(x=q, scale=d_model**-0.5)
+        scaled_q = paddle.scale(x=q, scale=d_model**-0.5)
         product = layers.matmul(x=scaled_q, y=k, transpose_y=True)
         weights = __softmax(layers.elementwise_add(x=product, y=attn_bias))
         if dropout_rate:
diff --git a/python/paddle/static/io.py b/python/paddle/static/io.py
index 47e3dddbbd893970418f11a164483f5052b038ad..fdcfc1a657069d9fa77c4826cfaeb214c06ccdb5 100644
--- a/python/paddle/static/io.py
+++ b/python/paddle/static/io.py
@@ -26,7 +26,6 @@ from paddle.fluid import (
     CompiledProgram,
     default_main_program,
     Program,
-    layers,
     unique_name,
     program_guard,
 )
@@ -201,7 +200,7 @@ def normalize_program(program, feed_vars, fetch_vars):
         uniq_fetch_vars = []
         for i, var in enumerate(fetch_vars):
             if var.dtype != paddle.bool:
-                var = layers.scale(
+                var = paddle.scale(
                     var, 1.0, name="save_infer_model/scale_{}".format(i)
                 )
             uniq_fetch_vars.append(var)
diff --git a/python/paddle/tensor/ops.py b/python/paddle/tensor/ops.py
index c4d8b01762a2af9e854874945b330ee0519f91e4..2897f4fd85a1b21075b6ce682f0c0becf62d2692 100644
--- a/python/paddle/tensor/ops.py
+++ b/python/paddle/tensor/ops.py
@@ -52,7 +52,7 @@ __inplace_unary_func__ = [
 __all__ = []
 
 # It is a hot fix in some unittest using:
-#   fluid.layers.scale(x=x, scale=10.0, out=out_var)
+#   paddle.scale(x=x, scale=10.0, out=out_var)
 # e.g.: test_program_code.py, test_dist_train.py
 globals()['_scale'] = generate_layer_fn('scale')