delete mean api (#48764)

python/paddle/fluid/clip.py

@@ -119,6 +119,8 @@ class ErrorClipByValue(BaseErrorClipAttr):
         .. code-block:: python
 
             import paddle.fluid as fluid
+            import paddle
+            paddle.enable_static()
             BATCH_SIZE = 128
             CLIP_MAX = 2e-6
             CLIP_MIN = -1e-6
@@ -132,11 +134,12 @@ class ErrorClipByValue(BaseErrorClipAttr):
                     input=hidden2, size=10, act='softmax')
                 label = fluid.layers.data(name='y', shape=[1], dtype='int64')
                 cost = fluid.layers.cross_entropy(input=predict, label=label)
-                avg_cost = fluid.layers.mean(cost)
+                avg_cost = paddle.mean(cost)
             prog_clip = prog.clone()
             prog_clip.block(0).var(hidden1.name)._set_error_clip(
                 fluid.clip.ErrorClipByValue(
                     max=CLIP_MAX, min=CLIP_MIN)
+                    )
     """
 
     def __init__(self, max, min=None):

python/paddle/fluid/contrib/layers/nn.py

@@ -1965,6 +1965,8 @@ def fused_bn_add_act(
 
             import paddle
             import paddle.fluid as fluid
+            import paddle
+            paddle.enable_static()
 
             paddle.enable_static()
             # required: gpu
@@ -1997,7 +1999,7 @@ def fused_bn_add_act(
                     fused_bn_add_act = fluid.contrib.layers.fused_bn_add_act(conv1_2, bn)
                     prediction = fluid.layers.fc(input=fused_bn_add_act, size=10, act='softmax')
                     loss = fluid.layers.cross_entropy(input=prediction, label=y)
-                    loss = fluid.layers.mean(loss)
+                    loss = paddle.mean(loss)
                     sgd = fluid.optimizer.SGD(learning_rate=0.001)
                     sgd = fluid.contrib.mixed_precision.decorate(
                         sgd, use_dynamic_loss_scaling=True, init_loss_scaling=128.0)

python/paddle/fluid/contrib/slim/quantization/quantization_pass.py

@@ -27,10 +27,10 @@ from .... import unique_name
 
 from ....framework import Program, program_guard, default_startup_program
 from ....data import data
-from ....layers import mean
 from ....executor import scope_guard
 from ....framework import _get_paddle_place
 from . import utils
+import paddle
 
 __all__ = [
     'QuantizationTransformPass',
@@ -927,7 +927,7 @@ class QuantizationTransformPass:
                 out_node = func(in_node)
                 graph.out_node_mapping_table[out_node.name] = var_node.name()
                 # loss shape must be 1 when minimize
-                loss = mean(out_node)
+                loss = paddle.mean(out_node)
                 if not graph._for_test:
                     assert (
                         self._optimizer

python/paddle/fluid/layers/nn.py

@@ -86,7 +86,6 @@ __all__ = [
     'elementwise_mul',
     'clip',
     'clip_by_norm',
-    'mean',
     'mul',
     'merge_selected_rows',
     'get_tensor_from_selected_rows',
@@ -3368,47 +3367,6 @@ def clip_by_norm(x, max_norm, name=None):
     return out
 
 
-@deprecated(since="2.0.0", update_to="paddle.mean")
-@templatedoc()
-def mean(x, name=None):
-    """
-    ${comment}
-
-    Args:
-        x(${x_type}): ${x_comment}
-        name(basestring|None): Name of the output.
-
-    Returns:
-        out(${out_type}): ${out_comment}
-
-    Examples:
-        .. code-block:: python
-
-            import paddle
-            import paddle.fluid as fluid
-            paddle.enable_static()
-
-            input = fluid.layers.data(
-                name='data', shape=[2, 3], dtype='float32')
-            mean = paddle.mean(input)
-    """
-
-    if _in_legacy_dygraph():
-        return _legacy_C_ops.mean(x)
-    if in_dygraph_mode():
-        return _C_ops.mean_all(x)
-
-    helper = LayerHelper("mean", **locals())
-    check_variable_and_dtype(x, 'x', ['float16', 'float32', 'float64'], 'mean')
-    out = helper.create_variable_for_type_inference(dtype=x.dtype)
-
-    helper.append_op(
-        type="mean", inputs={"X": x}, attrs={}, outputs={"Out": out}
-    )
-
-    return out
-
-
 @templatedoc()
 def merge_selected_rows(x, name=None):
     """

python/paddle/fluid/optimizer.py

@@ -1452,7 +1452,7 @@ class SGDOptimizer(Optimizer):
                 y = fluid.layers.data(name='y', shape=[1], dtype='float32')
                 y_predict = fluid.layers.fc(input=x, size=1, act=None)
                 cost = paddle.nn.functional.square_error_cost(input=y_predict, label=y)
-                avg_cost = fluid.layers.mean(cost)
+                avg_cost = paddle.mean(cost)
 
                 sgd_optimizer = fluid.optimizer.SGD(learning_rate=0.001)
                 sgd_optimizer.minimize(avg_cost)
@@ -1654,7 +1654,7 @@ class MomentumOptimizer(Optimizer):
                 y = fluid.layers.data(name='y', shape=[1], dtype='float32')
                 y_predict = fluid.layers.fc(input=x, size=1, act=None)
                 cost = paddle.nn.functional.square_error_cost(input=y_predict, label=y)
-                avg_cost = fluid.layers.mean(cost)
+                avg_cost = paddle.mean(cost)
 
                 moment_optimizer = fluid.optimizer.MomentumOptimizer(learning_rate=0.001, momentum=0.9)
                 moment_optimizer.minimize(avg_cost)
@@ -2232,7 +2232,7 @@ class AdamOptimizer(Optimizer):
                 y = fluid.data(name='y', shape=[None, 1], dtype='float32')
                 y_predict = fluid.layers.fc(input=x, size=1, act=None)
                 cost = paddle.nn.functional.square_error_cost(input=y_predict, label=y)
-                avg_cost = fluid.layers.mean(cost)
+                avg_cost = paddle.mean(cost)
 
                 adam_optimizer = fluid.optimizer.AdamOptimizer(0.01)
                 adam_optimizer.minimize(avg_cost)
@@ -2261,7 +2261,7 @@ class AdamOptimizer(Optimizer):
                 y = fluid.data(name='y', shape=[None, 1], dtype='float32')
                 y_predict = fluid.layers.fc(input=x, size=1, act=None)
                 cost = paddle.nn.functional.square_error_cost(input=y_predict, label=y)
-                avg_cost = fluid.layers.mean(cost)
+                avg_cost = paddle.mean(cost)
 
                 # define beta decay variable
                 def get_decayed_betas(beta1_init, beta2_init, decay_steps, decay_rate, epsilon_init):
@@ -2641,6 +2641,8 @@ class AdamaxOptimizer(Optimizer):
 
           import paddle.fluid as fluid
           import numpy
+          import paddle
+          paddle.enable_static()
 
           # First create the Executor.
           place = fluid.CPUPlace() # fluid.CUDAPlace(0)
@@ -2651,7 +2653,7 @@ class AdamaxOptimizer(Optimizer):
           with fluid.program_guard(train_program, startup_program):
               data = fluid.data(name='X', shape=[None, 1], dtype='float32')
               hidden = fluid.layers.fc(input=data, size=10)
-              loss = fluid.layers.mean(hidden)
+              loss = paddle.mean(hidden)
               adam = fluid.optimizer.AdamaxOptimizer(learning_rate=0.2)
               adam.minimize(loss)
 
@@ -2816,6 +2818,8 @@ class DpsgdOptimizer(Optimizer):
 
           import paddle.fluid as fluid
           import numpy
+          import paddle
+          paddle.enable_static()
 
           # First create the Executor.
           place = fluid.CPUPlace() # fluid.CUDAPlace(0)
@@ -2826,7 +2830,7 @@ class DpsgdOptimizer(Optimizer):
           with fluid.program_guard(train_program, startup_program):
               data = fluid.layers.data(name='X', shape=[1], dtype='float32')
               hidden = fluid.layers.fc(input=data, size=10)
-              loss = fluid.layers.mean(hidden)
+              loss = paddle.mean(hidden)
               optimizer = fluid.optimizer.Dpsgd(learning_rate=0.01, clip=10.0, batch_size=16.0, sigma=1.0)
               optimizer.minimize(loss)
 
@@ -3291,7 +3295,7 @@ class RMSPropOptimizer(Optimizer):
                 y = fluid.layers.data(name='y', shape=[1], dtype='float32')
                 y_predict = fluid.layers.fc(input=x, size=1, act=None)
                 cost = paddle.nn.functional.square_error_cost(input=y_predict, label=y)
-                avg_cost = fluid.layers.mean(cost)
+                avg_cost = paddle.mean(cost)
 
                 rms_optimizer = fluid.optimizer.RMSProp(learning_rate=0.1)
                 rms_optimizer.minimize(avg_cost)
@@ -3510,7 +3514,7 @@ class FtrlOptimizer(Optimizer):
                 y = fluid.layers.data(name='y', shape=[1], dtype='float32')
                 y_predict = fluid.layers.fc(input=x, size=1, act=None)
                 cost = paddle.nn.functional.square_error_cost(input=y_predict, label=y)
-                avg_cost = fluid.layers.mean(cost)
+                avg_cost = paddle.mean(cost)
 
                 ftrl_optimizer = fluid.optimizer.Ftrl(learning_rate=0.1)
                 ftrl_optimizer.minimize(avg_cost)
@@ -3679,11 +3683,13 @@ class LambOptimizer(AdamOptimizer):
     Examples:
         .. code-block:: python
 
+            import paddle
             import paddle.fluid as fluid
+            paddle.enable_static()
 
             data = fluid.data(name='x', shape=[-1, 5], dtype='float32')
             hidden = fluid.layers.fc(input=data, size=10)
-            cost = fluid.layers.mean(hidden)
+            cost = paddle.mean(hidden)
 
             def exclude_fn(param):
                 return param.name.endswith('.b_0')
@@ -3885,8 +3891,10 @@ class ModelAverage(Optimizer):
 
       .. code-block:: python
 
+        import paddle
         import paddle.fluid as fluid
         import numpy
+        paddle.enable_static()
 
         # First create the Executor.
         place = fluid.CPUPlace()  # fluid.CUDAPlace(0)
@@ -3898,7 +3906,7 @@ class ModelAverage(Optimizer):
             # build net
             data = fluid.data(name='X', shape=[None, 1], dtype='float32')
             hidden = fluid.layers.fc(input=data, size=10)
-            loss = fluid.layers.mean(hidden)
+            loss = paddle.mean(hidden)
             optimizer = fluid.optimizer.Momentum(learning_rate=0.2, momentum=0.1)
             optimizer.minimize(loss)
 
@@ -4064,6 +4072,8 @@ class ModelAverage(Optimizer):
 
             import paddle.fluid as fluid
             import numpy
+            import paddle
+            paddle.enable_static()
 
             # First create the Executor.
             place = fluid.CPUPlace()  # fluid.CUDAPlace(0)
@@ -4075,7 +4085,7 @@ class ModelAverage(Optimizer):
                 # build net
                 data = fluid.data(name='X', shape=[None, 1], dtype='float32')
                 hidden = fluid.layers.fc(input=data, size=10)
-                loss = fluid.layers.mean(hidden)
+                loss = paddle.mean(hidden)
                 optimizer = fluid.optimizer.Momentum(learning_rate=0.2, momentum=0.1)
                 optimizer.minimize(loss)
 
@@ -4118,6 +4128,8 @@ class ModelAverage(Optimizer):
 
             import paddle.fluid as fluid
             import numpy
+            import paddle
+            paddle.enable_static()
 
             # First create the Executor.
             place = fluid.CPUPlace()  # fluid.CUDAPlace(0)
@@ -4129,7 +4141,7 @@ class ModelAverage(Optimizer):
                 # build net
                 data = fluid.data(name='X', shape=[None, 1], dtype='float32')
                 hidden = fluid.layers.fc(input=data, size=10)
-                loss = fluid.layers.mean(hidden)
+                loss = paddle.mean(hidden)
                 optimizer = fluid.optimizer.Momentum(learning_rate=0.2, momentum=0.1)
                 optimizer.minimize(loss)
 

python/paddle/fluid/regularizer.py

@@ -68,6 +68,8 @@ class L2DecayRegularizer(WeightDecayRegularizer):
 
             # Example1: set Regularizer in optimizer
             import paddle.fluid as fluid
+            import paddle
+            paddle.enable_static()
 
             main_prog = fluid.Program()
             startup_prog = fluid.Program()
@@ -77,7 +79,7 @@ class L2DecayRegularizer(WeightDecayRegularizer):
                 hidden = fluid.layers.fc(input=data, size=128, act='relu')
                 prediction = fluid.layers.fc(input=hidden, size=10, act='softmax')
                 loss = fluid.layers.cross_entropy(input=prediction, label=label)
-                avg_loss = fluid.layers.mean(loss)
+                avg_loss = paddle.mean(loss)
             optimizer = fluid.optimizer.Adagrad(
                 learning_rate=1e-4,
                 regularization=fluid.regularizer.L2Decay(
@@ -87,6 +89,8 @@ class L2DecayRegularizer(WeightDecayRegularizer):
 
             # Example2: set Regularizer both in ParamAttr and optimizer
             import paddle.fluid as fluid
+            import paddle
+            paddle.enable_static()
 
             l1 = fluid.regularizer.L1Decay(regularization_coeff=0.1)
             l2 = fluid.regularizer.L2Decay(regularization_coeff=0.1)
@@ -97,7 +101,7 @@ class L2DecayRegularizer(WeightDecayRegularizer):
             hidden1 = fluid.layers.fc(x, 8, param_attr=w_param)  # fc_0.w_0(L1), fc_0.b_0
             hidden2 = fluid.layers.fc(hidden1, 16, param_attr=w_param)   # fc_1.w_0(L1), fc_1.b_0
             predict = fluid.layers.fc(hidden2, 32)    # fc_3.w_0, fc_3.b_0
-            avg_loss = fluid.layers.mean(predict)
+            avg_loss = paddle.mean(predict)
 
             # set L2 regularization in optimizer
             optimizer = fluid.optimizer.SGD(learning_rate=1e-4, regularization=l2)
@@ -181,7 +185,8 @@ class L1DecayRegularizer(WeightDecayRegularizer):
 
             # Example1: set Regularizer in optimizer
             import paddle.fluid as fluid
-
+            import paddle
+            paddle.enable_static()
             main_prog = fluid.Program()
             startup_prog = fluid.Program()
             with fluid.program_guard(main_prog, startup_prog):
@@ -190,7 +195,7 @@ class L1DecayRegularizer(WeightDecayRegularizer):
                 hidden = fluid.layers.fc(input=data, size=128, act='relu')
                 prediction = fluid.layers.fc(input=hidden, size=10, act='softmax')
                 loss = fluid.layers.cross_entropy(input=prediction, label=label)
-                avg_loss = fluid.layers.mean(loss)
+                avg_loss = paddle.mean(loss)
             optimizer = fluid.optimizer.Adagrad(
                 learning_rate=1e-4,
                 regularization=fluid.regularizer.L1DecayRegularizer(
@@ -200,7 +205,8 @@ class L1DecayRegularizer(WeightDecayRegularizer):
 
             # Example2: set Regularizer both in ParamAttr and optimizer
             import paddle.fluid as fluid
-
+            import paddle
+            paddle.enable_static()
             l1 = fluid.regularizer.L1Decay(regularization_coeff=0.1)
             l2 = fluid.regularizer.L2Decay(regularization_coeff=0.1)
             x = fluid.layers.uniform_random([3,4])
@@ -210,7 +216,7 @@ class L1DecayRegularizer(WeightDecayRegularizer):
             hidden1 = fluid.layers.fc(x, 8, param_attr=w_param)  # fc_0.w_0(L1), fc_0.b_0
             hidden2 = fluid.layers.fc(hidden1, 16, param_attr=w_param)  # fc_1.w_0(L1), fc_1.b_0
             predict = fluid.layers.fc(hidden2, 32)   # fc_3.w_0, fc_3.b_0
-            avg_loss = fluid.layers.mean(predict)
+            avg_loss = paddle.mean(predict)
 
             # set L2 regularization in optimizer
             optimizer = fluid.optimizer.SGD(learning_rate=1e-4, regularization=l2)

python/paddle/fluid/tests/unittests/test_dist_fleet_ps13.py

@@ -66,7 +66,7 @@ class TestPSPassWithBow(unittest.TestCase):
                 ),
                 loss_op2,
             )
-            avg_cost = fluid.layers.mean(loss_op3)
+            avg_cost = paddle.mean(loss_op3)
             return avg_cost
 
         is_distributed = False

python/paddle/fluid/transpiler/distribute_transpiler.py

@@ -288,12 +288,12 @@ class DistributeTranspiler:
 
             paddle.enable_static()
 
-            x = fluid.data(name='x', shape=[13], dtype='float32')
+            x = fluid.data(name='x', shape=[1,13], dtype='float32')
             y = fluid.data(name='y', shape=[1], dtype='float32')
             y_predict = fluid.layers.fc(input=x, size=1, act=None)
 
             cost =paddle.nn.functional.square_error_cost(input=y_predict, label=y)
-            avg_loss = fluid.layers.mean(cost)
+            avg_loss = paddle.mean(cost)
 
             sgd_optimizer = fluid.optimizer.SGD(learning_rate=0.001)
             sgd_optimizer.minimize(avg_loss)

tools/infrt/fake_models/multi_fc.py

@@ -42,7 +42,7 @@ for i in range(num_layers - 1):
     )
 
 cost = fluid.layers.square_error_cost(fc_out, label)
-avg_cost = fluid.layers.mean(cost)
+avg_cost = paddle.mean(cost)
 
 optimizer = fluid.optimizer.SGD(learning_rate=0.001)
 optimizer.minimize(avg_cost)