ML Community Day is November 9! Join us for updates from TensorFlow, JAX, and more Learn more

tensorflow:: ops:: FakeQuantWithMinMaxVarsPerChannelGradient:: Attrs

#include <array_ops.h>

Optional attribute setters for FakeQuantWithMinMaxVarsPerChannelGradient .

Summary

Public attributes

narrow_range_ = false
bool
num_bits_ = 8
int64

Public functions

NarrowRange (bool x)
TF_MUST_USE_RESULT Attrs
Whether to quantize into 2^num_bits - 1 distinct values.
NumBits (int64 x)
TF_MUST_USE_RESULT Attrs
The bitwidth of the quantization; between 2 and 16, inclusive.

Public attributes

narrow_range_

bool tensorflow::ops::FakeQuantWithMinMaxVarsPerChannelGradient::Attrs::narrow_range_ = false

num_bits_

int64 tensorflow::ops::FakeQuantWithMinMaxVarsPerChannelGradient::Attrs::num_bits_ = 8

Public functions

NarrowRange

TF_MUST_USE_RESULT Attrs tensorflow::ops::FakeQuantWithMinMaxVarsPerChannelGradient::Attrs::NarrowRange(
  bool x
)

Whether to quantize into 2^num_bits - 1 distinct values.

Defaults to false

NumBits

TF_MUST_USE_RESULT Attrs tensorflow::ops::FakeQuantWithMinMaxVarsPerChannelGradient::Attrs::NumBits(
  int64 x
)

The bitwidth of the quantization; between 2 and 16, inclusive.

Defaults to 8