Aliases:
tf.nn.depthwise_conv2d_backprop_filter
tf.nn.depthwise_conv2d_native_backprop_filter
tf.nn.depthwise_conv2d_backprop_filter(
input,
filter_sizes,
out_backprop,
strides,
padding,
data_format='NHWC',
dilations=[1, 1, 1, 1],
name=None
)
Defined in generated file: tensorflow/python/ops/gen_nn_ops.py
.
Computes the gradients of depthwise convolution with respect to the filter.
Args:
input
: ATensor
. Must be one of the following types:half
,bfloat16
,float32
,float64
. 4-D with shape based ondata_format
. For example, ifdata_format
is 'NHWC' theninput
is a 4-D[batch, in_height, in_width, in_channels]
tensor.filter_sizes
: ATensor
of typeint32
. An integer vector representing the tensor shape offilter
, wherefilter
is a 4-D[filter_height, filter_width, in_channels, depthwise_multiplier]
tensor.out_backprop
: ATensor
. Must have the same type asinput
. 4-D with shape based ondata_format
. For example, ifdata_format
is 'NHWC' then out_backprop shape is[batch, out_height, out_width, out_channels]
. Gradients w.r.t. the output of the convolution.strides
: A list ofints
. The stride of the sliding window for each dimension of the input of the convolution.padding
: Astring
from:"SAME", "VALID"
. The type of padding algorithm to use.data_format
: An optionalstring
from:"NHWC", "NCHW"
. Defaults to"NHWC"
. Specify the data format of the input and output data. With the default format "NHWC", the data is stored in the order of: [batch, height, width, channels]. Alternatively, the format could be "NCHW", the data storage order of: [batch, channels, height, width].dilations
: An optional list ofints
. Defaults to[1, 1, 1, 1]
. 1-D tensor of length 4. The dilation factor for each dimension ofinput
. If set to k > 1, there will be k-1 skipped cells between each filter element on that dimension. The dimension order is determined by the value ofdata_format
, see above for details. Dilations in the batch and depth dimensions must be 1.name
: A name for the operation (optional).
Returns:
A Tensor
. Has the same type as input
.