Quantizes then dequantizes a tensor.
This is almost identical to QuantizeAndDequantizeV2, except that it returns a gradient of 1 for inputs that are within the quantization range, or 0 otherwise.
Nested Classes
class | QuantizeAndDequantizeV4.Options | Optional attributes for QuantizeAndDequantizeV4
|
Public Methods
Output<T> |
asOutput()
Returns the symbolic handle of a tensor.
|
static QuantizeAndDequantizeV4.Options |
axis(Long axis)
|
static <T extends Number> QuantizeAndDequantizeV4<T> | |
static QuantizeAndDequantizeV4.Options |
narrowRange(Boolean narrowRange)
|
static QuantizeAndDequantizeV4.Options |
numBits(Long numBits)
|
Output<T> |
output()
|
static QuantizeAndDequantizeV4.Options |
rangeGiven(Boolean rangeGiven)
|
static QuantizeAndDequantizeV4.Options |
roundMode(String roundMode)
|
static QuantizeAndDequantizeV4.Options |
signedInput(Boolean signedInput)
|
Inherited Methods
Public Methods
public Output<T> asOutput ()
Returns the symbolic handle of a tensor.
Inputs to TensorFlow operations are outputs of another TensorFlow operation. This method is used to obtain a symbolic handle that represents the computation of the input.
public static QuantizeAndDequantizeV4.Options axis (Long axis)
Parameters
axis | If specified, this axis is treated as a channel or slice axis, and a separate quantization range is used for each channel or slice along this axis. |
---|
public static QuantizeAndDequantizeV4<T> create (Scope scope, Operand<T> input, Operand<T> inputMin, Operand<T> inputMax, Options... options)
Factory method to create a class wrapping a new QuantizeAndDequantizeV4 operation.
Parameters
scope | current scope |
---|---|
input | Tensor to quantize and then dequantize. |
inputMin | If `range_given == True`, this specifies the minimum input value that needs to be represented, otherwise it is determined from the min value of the `input` tensor. |
inputMax | If `range_given == True`, this specifies the maximum input value that needs to be represented, otherwise it is determined from the max value of the `input` tensor. |
options | carries optional attributes values |
Returns
- a new instance of QuantizeAndDequantizeV4
public static QuantizeAndDequantizeV4.Options narrowRange (Boolean narrowRange)
Parameters
narrowRange | If True, then the absolute value of the quantized minimum value is the same as the quantized maximum value, instead of 1 greater. i.e. for 8 bit quantization, the minimum value is -127 instead of -128. |
---|
public static QuantizeAndDequantizeV4.Options numBits (Long numBits)
Parameters
numBits | The bitwidth of the quantization. |
---|
public static QuantizeAndDequantizeV4.Options rangeGiven (Boolean rangeGiven)
Parameters
rangeGiven | Whether the range is given or should be determined from the `input` tensor. |
---|
public static QuantizeAndDequantizeV4.Options roundMode (String roundMode)
Parameters
roundMode | The 'round_mode' attribute controls which rounding tie-breaking algorithm is
used when rounding float values to their quantized equivalents. The following
rounding modes are currently supported:
|
---|
public static QuantizeAndDequantizeV4.Options signedInput (Boolean signedInput)
Parameters
signedInput | Whether the quantization is signed or unsigned. (actually this parameter should have been called `signed_output`) |
---|