[fluid_ops] clean data_norm in python (#70358)

co63oc · web-flow · commit da3257aff668 · 2025-04-11T14:35:12.000+08:00
diff --git a/python/paddle/static/nn/__init__.py b/python/paddle/static/nn/__init__.py
@@ -21,7 +21,6 @@
     conv2d_transpose,
     conv3d,
     conv3d_transpose,
-    data_norm,
     deform_conv2d,
     embedding,
     fc,
@@ -58,7 +57,6 @@
     'conv2d_transpose',
     'conv3d',
     'conv3d_transpose',
-    'data_norm',
     'deform_conv2d',
     'group_norm',
     'instance_norm',
diff --git a/python/paddle/static/nn/common.py b/python/paddle/static/nn/common.py
@@ -457,214 +457,6 @@ def continuous_value_model(input, cvm, use_cvm=True):
     return out
 
 
-@static_only
-def data_norm(
-    input,
-    act=None,
-    epsilon=1e-05,
-    param_attr=None,
-    data_layout='NCHW',
-    in_place=False,
-    name=None,
-    moving_mean_name=None,
-    moving_variance_name=None,
-    do_model_average_for_mean_and_var=True,
-    slot_dim=-1,
-    sync_stats=False,
-    summary_decay_rate=0.9999999,
-    enable_scale_and_shift=False,
-):
-    r"""
-
-    **Data Normalization Layer**
-
-    This op can be used as a normalizer function for conv2d and fully_connected operations.
-    The required data format for this layer is one of the following:
-
-    1. NHWC `[batch, in_height, in_width, in_channels]`
-
-    2. NCHW `[batch, in_channels, in_height, in_width]`
-
-    :math:`input` is the input features over a mini-batch.
-
-    ..  math::
-
-        \mu_{\beta} &\gets \frac{1}{m} \sum_{i=1}^{m} x_i \qquad &//
-        \ mini-batch\ mean \\
-        \sigma_{\beta}^{2} &\gets \frac{1}{m} \sum_{i=1}^{m}(x_i -
-        \mu_{\beta})^2 \qquad &//\ mini-batch\ variance \\
-        \hat{x_i} &\gets \frac{x_i - \mu_\beta} {\sqrt{
-        \sigma_{\beta}^{2} + \epsilon}} \qquad &//\ normalize \\
-        y_i &\gets \gamma \hat{x_i} + \beta \qquad &//\ scale\ and\ shift
-
-    Args:
-        input (Tensor): The input Tensor.
-        act (str, optional): Activation type, linear|relu|prelu|... Default: None.
-        epsilon(float, optional): Whether to add small values into the variance during calculations
-            to prevent division by zero. Default: 1e-05.
-        param_attr (ParamAttr, optional): The parameter attribute for Parameter `scale`. Default: None.
-        data_layout (str, optional): Specify the data format of the input, and the data format of the output
-            will be consistent with that of the input. An optional string from: `"NCHW"`, `"NHWC"`.
-            The default is `"NCHW"`. When it is `"NCHW"`, the data is stored in the order of:
-            `[batch_size, input_channels, input_height, input_width]`. Default: `"NCHW"`.
-        in_place (bool, optional): Make the input and output of batch norm reuse memory. Default: False.
-        name (str, optional): A name for this layer (optional). If set None, the layer
-            will be named automatically. Default: None.
-        moving_mean_name (str, optional): The name of moving_mean which store the global Mean. Default: None.
-        moving_variance_name (str, optional): The name of the moving_variance which store the global Variance. Default: None.
-        do_model_average_for_mean_and_var (bool, optional): Whether parameter mean and variance
-            should do model average when model average is enabled. Default: True.
-        slot_dim (int, optional): The embedding dimension of one slot. Slot is a set of one specific feature. In pslib mode,
-            we distinguish feature ids by slot and pull their embeddings from parameter server (pslib). The first
-            place of the embedding is the historical show number (occurrence time of this feature id with a label 0).
-            If the input of this op is concated by slot-wise embeddings, and the show number is zero when this slot
-            is new or empty, the normalization result may be impractical. To avoid this, we add slot_dim to locate
-            the show number and judge if the show number is zero. If so, we choose to skip normalization on this
-            embedding. Default: -1.
-        sync_stats (bool, optional): When running with multiple GPU cards, using allreduce to sync the
-            summary messages. Default: False.
-        summary_decay_rate (float, optional): The decay rate when updating summary. Default: 0.9999999.
-        enable_scale_and_shift (bool, optional): do scale&shift after normalization. Default: False.
-
-    Returns:
-        Tensor: A tensor which is the result after applying data normalization on the input.
-
-    Examples:
-
-        .. code-block:: python
-
-            >>> import paddle
-            >>> paddle.enable_static()
-
-            >>> x = paddle.randn(shape=[32, 100])
-            >>> hidden2 = paddle.static.nn.data_norm(input=x)
-    """
-    helper = LayerHelper('data_norm', **locals())
-    dtype = helper.input_dtype()
-
-    input_shape = input.shape
-    if len(input_shape) < 2:
-        raise ValueError(
-            f"The shape pf Input < 2 (got {len(input_shape)}D input, input shape is: {input_shape})"
-        )
-    if data_layout == 'NCHW':
-        channel_num = input_shape[1]
-    else:
-        if data_layout == 'NHWC':
-            channel_num = input_shape[-1]
-        else:
-            raise ValueError("unsupported data layout:" + data_layout)
-
-    param_shape = [channel_num]
-
-    batch_size_default = 1e4
-    batch_sum_default = 0.0
-    batch_square_sum_default = 1e4
-    scale_w_default = 1.0
-    bias_default = 0.0
-
-    if param_attr and isinstance(param_attr, dict):
-        batch_size_default = param_attr.get("batch_size", 1e4)
-        batch_sum_default = param_attr.get("batch_sum", 0.0)
-        batch_square_sum_default = param_attr.get("batch_square", 1e4)
-    if enable_scale_and_shift:
-        scale_w_default = param_attr.get("scale_w", 1.0)
-        bias_default = param_attr.get("bias", 0.0)
-
-    # create scale and shift(bias) when enable_scale_and_shift is True
-    if name is None:
-        name = "dn"
-    if enable_scale_and_shift:
-        scale_w = helper.create_parameter(
-            attr=ParamAttr(
-                name=name + '.scale_w',
-                initializer=Constant(value=float(scale_w_default)),
-                trainable=True,
-            ),
-            shape=param_shape,
-            dtype=input.dtype,
-        )
-        bias = helper.create_parameter(
-            attr=ParamAttr(
-                name=name + '.bias',
-                initializer=Constant(value=float(bias_default)),
-                trainable=True,
-            ),
-            shape=param_shape,
-            dtype=input.dtype,
-        )
-    # create parameter
-    batch_size = helper.create_parameter(
-        attr=ParamAttr(
-            name=name + '.batch_size',
-            initializer=Constant(value=float(batch_size_default)),
-            trainable=True,
-        ),
-        shape=param_shape,
-        dtype=input.dtype,
-    )
-
-    batch_sum = helper.create_parameter(
-        attr=ParamAttr(
-            name=name + '.batch_sum',
-            initializer=Constant(value=float(batch_sum_default)),
-            trainable=True,
-        ),
-        shape=param_shape,
-        dtype=input.dtype,
-    )
-
-    batch_square_sum = helper.create_parameter(
-        attr=ParamAttr(
-            name=name + '.batch_square_sum',
-            initializer=Constant(value=float(batch_square_sum_default)),
-            trainable=True,
-        ),
-        shape=param_shape,
-        dtype=input.dtype,
-    )
-
-    means = helper.create_variable(dtype=dtype, stop_gradient=True)
-    scales = helper.create_variable(dtype=dtype, stop_gradient=True)
-
-    data_norm_out = input if in_place else helper.create_variable(dtype=dtype)
-
-    inputs = {
-        "X": input,
-        "BatchSize": batch_size,
-        "BatchSum": batch_sum,
-        "BatchSquareSum": batch_square_sum,
-    }
-    attrs = {
-        "epsilon": epsilon,
-        "data_layout": data_layout,
-        "sync_stats": sync_stats,
-        "summary_decay_rate": summary_decay_rate,
-    }
-    if slot_dim > 0:
-        attrs["slot_dim"] = slot_dim
-    if enable_scale_and_shift:
-        attrs["enable_scale_and_shift"] = enable_scale_and_shift
-    if enable_scale_and_shift:
-        inputs["scale_w"] = scale_w
-        inputs["bias"] = bias
-    helper.append_op(
-        type="data_norm",
-        inputs=inputs,
-        outputs={
-            "Y": data_norm_out,
-            "Means": means,
-            "Scales": scales,
-            "BatchSize": batch_size,
-            "BatchSum": batch_sum,
-            "BatchSquareSum": batch_square_sum,
-        },
-        attrs=attrs,
-    )
-
-    return helper.append_activation(data_norm_out)
-
-
 def group_norm(
     input,
     groups,
diff --git a/test/legacy_test/test_directory_migration.py b/test/legacy_test/test_directory_migration.py
@@ -88,7 +88,6 @@ def test_new_directory(self):
             'paddle.static.nn.conv3d',
             'paddle.static.nn.conv3d_transpose',
             'paddle.static.nn.create_parameter',
-            'paddle.static.nn.data_norm',
             'paddle.static.nn.deform_conv2d',
             'paddle.static.nn.group_norm',
             'paddle.static.nn.instance_norm',
@@ -173,7 +172,6 @@ def test_old_directory(self):
             'paddle.declarative.conv3d_transpose',
             'paddle.declarative.create_parameter',
             'paddle.declarative.crf_decoding',
-            'paddle.declarative.data_norm',
             'paddle.declarative.deformable_conv',
             'paddle.declarative.group_norm',
             'paddle.declarative.hsigmoid',
