QuantizedConv2d

public final class QuantizedConv2d

Computes a 2D convolution given quantized 4D input and filter tensors.

The inputs are quantized tensors where the lowest value represents the real number of the associated minimum, and the highest represents the maximum. This means that you can only interpret the quantized output in the same way, by taking the returned minimum and maximum values into account.

Nested Classes

class QuantizedConv2d.Options Optional attributes for QuantizedConv2d  

Constants

String OP_NAME The name of this op, as known by TensorFlow core engine

Public Methods

static <V extends TType> QuantizedConv2d<V>
create(Scope scope, Operand<? extends TType> input, Operand<? extends TType> filter, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, Options... options)
Factory method to create a class wrapping a new QuantizedConv2d operation.
static QuantizedConv2d.Options
dilations(List<Long> dilations)
Output<TFloat32>
maxOutput()
The float value that the highest quantized output value represents.
Output<TFloat32>
minOutput()
The float value that the lowest quantized output value represents.
Output<V>
output()

Inherited Methods

Constants

public static final String OP_NAME

The name of this op, as known by TensorFlow core engine

Constant Value: "QuantizedConv2D"

Public Methods

public static QuantizedConv2d<V> create (Scope scope, Operand<? extends TType> input, Operand<? extends TType> filter, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, Options... options)

Factory method to create a class wrapping a new QuantizedConv2d operation.

Parameters
scope current scope
filter filter's input_depth dimension must match input's depth dimensions.
minInput The float value that the lowest quantized input value represents.
maxInput The float value that the highest quantized input value represents.
minFilter The float value that the lowest quantized filter value represents.
maxFilter The float value that the highest quantized filter value represents.
strides The stride of the sliding window for each dimension of the input tensor.
padding The type of padding algorithm to use.
options carries optional attributes values
Returns
  • a new instance of QuantizedConv2d

public static QuantizedConv2d.Options dilations (List<Long> dilations)

Parameters
dilations 1-D tensor of length 4. The dilation factor for each dimension of `input`. 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 of `data_format`, see above for details. Dilations in the batch and depth dimensions must be 1.

public Output<TFloat32> maxOutput ()

The float value that the highest quantized output value represents.

public Output<TFloat32> minOutput ()

The float value that the lowest quantized output value represents.

public Output<V> output ()