"""Python wrappers around TensorFlow ops.
This file is MACHINE GENERATED! Do not edit.
Original C++ source file: math_ops.cc
"""
import collections
from tensorflow.python import pywrap_tfe as pywrap_tfe
from tensorflow.python.eager import context as _context
from tensorflow.python.eager import core as _core
from tensorflow.python.eager import execute as _execute
from tensorflow.python.framework import dtypes as _dtypes
from tensorflow.python.framework import op_def_registry as _op_def_registry
from tensorflow.python.framework import ops as _ops
from tensorflow.python.framework import op_def_library as _op_def_library
from tensorflow.python.util.deprecation import deprecated_endpoints
from tensorflow.python.util import dispatch as _dispatch
from tensorflow.python.util.tf_export import tf_export
from typing import TypeVar
def _abs(x, name=None):
r"""Computes the absolute value of a tensor.
Given a tensor `x`, this operation returns a tensor containing the absolute
value of each element in `x`. For example, if x is an input element and y is
an output element, this operation computes \\(y = |x|\\).
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int8`, `int16`, `int32`, `int64`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Abs", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return _abs_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Abs", x=x, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Abs", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Abs = tf_export("raw_ops.Abs")(_ops.to_raw_op(_abs))
def _abs_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Abs", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Abs", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def accumulate_nv2(inputs, shape, name=None):
r"""Returns the element-wise sum of a list of tensors.
`tf.accumulate_n_v2` performs the same operation as `tf.add_n`, but does not
wait for all of its inputs to be ready before beginning to sum. This can
save memory if inputs are ready at different times, since minimum temporary
storage is proportional to the output size rather than the inputs size.
Unlike the original `accumulate_n`, `accumulate_n_v2` is differentiable.
Returns a `Tensor` of same shape and type as the elements of `inputs`.
Args:
inputs: A list of at least 1 `Tensor` objects with the same type in: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
A list of `Tensor` objects, each with same shape and type.
shape: A `tf.TensorShape` or list of `ints`.
Shape of elements of `inputs`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `inputs`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "AccumulateNV2", name, inputs, "shape", shape)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return accumulate_nv2_eager_fallback(
inputs, shape=shape, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if not isinstance(inputs, (list, tuple)):
raise TypeError(
"Expected list for 'inputs' argument to "
"'accumulate_nv2' Op, not %r." % inputs)
_attr_N = len(inputs)
shape = _execute.make_shape(shape, "shape")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"AccumulateNV2", inputs=inputs, shape=shape, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("N", _op._get_attr_int("N"), "T", _op._get_attr_type("T"),
"shape", _op.get_attr("shape"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"AccumulateNV2", _inputs_flat, _attrs, _result)
_result, = _result
return _result
AccumulateNV2 = tf_export("raw_ops.AccumulateNV2")(_ops.to_raw_op(accumulate_nv2))
def accumulate_nv2_eager_fallback(inputs, shape, name, ctx):
if not isinstance(inputs, (list, tuple)):
raise TypeError(
"Expected list for 'inputs' argument to "
"'accumulate_nv2' Op, not %r." % inputs)
_attr_N = len(inputs)
shape = _execute.make_shape(shape, "shape")
_attr_T, inputs = _execute.args_to_matching_eager(list(inputs), ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_inputs_flat = list(inputs)
_attrs = ("N", _attr_N, "T", _attr_T, "shape", shape)
_result = _execute.execute(b"AccumulateNV2", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"AccumulateNV2", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def acos(x, name=None):
r"""Computes acos of x element-wise.
Provided an input tensor, the `tf.math.acos` operation returns the inverse cosine of each element of the tensor. If `y = tf.math.cos(x)` then, `x = tf.math.acos(y)`.
Input range is `[-1, 1]` and the output has a range of `[0, pi]`.
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int8`, `int16`, `int32`, `int64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Acos", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return acos_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Acos", x=x, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Acos", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Acos = tf_export("raw_ops.Acos")(_ops.to_raw_op(acos))
def acos_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Acos", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Acos", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.acosh', 'acosh')
def acosh(x, name=None):
r"""Computes inverse hyperbolic cosine of x element-wise.
Given an input tensor, the function computes inverse hyperbolic cosine of every element.
Input range is `[1, inf]`. It returns `nan` if the input lies outside the range.
```python
x = tf.constant([-2, -0.5, 1, 1.2, 200, 10000, float("inf")])
tf.math.acosh(x) ==> [nan nan 0. 0.62236255 5.9914584 9.903487 inf]
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Acosh", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_acosh(
(x, name,), None)
if _result is not NotImplemented:
return _result
return acosh_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
acosh, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_acosh(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Acosh", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
acosh, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Acosh", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Acosh = tf_export("raw_ops.Acosh")(_ops.to_raw_op(acosh))
_dispatcher_for_acosh = acosh._tf_type_based_dispatcher.Dispatch
def acosh_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Acosh", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Acosh", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def add(x, y, name=None):
r"""Returns x + y element-wise.
*NOTE*: `Add` supports broadcasting. `AddN` does not. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Given two input tensors, the `tf.add` operation computes the sum for every element in the tensor.
Both input and output have a range `(-inf, inf)`.
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `uint8`, `int8`, `int16`, `int32`, `int64`, `complex64`, `complex128`, `string`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Add", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return add_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Add", x=x, y=y, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Add", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Add = tf_export("raw_ops.Add")(_ops.to_raw_op(add))
def add_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.uint8, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, _dtypes.string, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Add", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Add", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def add_n(inputs, name=None):
r"""Add all input tensors element wise.
Inputs must be of same size and shape.
```python
x = [9, 7, 10]
tf.math.add_n(x) ==> 26
```
Args:
inputs: A list of at least 1 `Tensor` objects with the same type in: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`, `variant`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `inputs`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "AddN", name, inputs)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return add_n_eager_fallback(
inputs, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if not isinstance(inputs, (list, tuple)):
raise TypeError(
"Expected list for 'inputs' argument to "
"'add_n' Op, not %r." % inputs)
_attr_N = len(inputs)
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"AddN", inputs=inputs, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("N", _op._get_attr_int("N"), "T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"AddN", _inputs_flat, _attrs, _result)
_result, = _result
return _result
AddN = tf_export("raw_ops.AddN")(_ops.to_raw_op(add_n))
def add_n_eager_fallback(inputs, name, ctx):
if not isinstance(inputs, (list, tuple)):
raise TypeError(
"Expected list for 'inputs' argument to "
"'add_n' Op, not %r." % inputs)
_attr_N = len(inputs)
_attr_T, inputs = _execute.args_to_matching_eager(list(inputs), ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, _dtypes.variant, ])
_inputs_flat = list(inputs)
_attrs = ("N", _attr_N, "T", _attr_T)
_result = _execute.execute(b"AddN", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"AddN", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def add_v2(x, y, name=None):
r"""Returns x + y element-wise.
*NOTE*: `Add` supports broadcasting. `AddN` does not. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `uint8`, `uint16`, `uint32`, `uint64`, `int8`, `int16`, `int32`, `int64`, `complex64`, `complex128`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "AddV2", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return add_v2_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"AddV2", x=x, y=y, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"AddV2", _inputs_flat, _attrs, _result)
_result, = _result
return _result
AddV2 = tf_export("raw_ops.AddV2")(_ops.to_raw_op(add_v2))
def add_v2_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.uint8, _dtypes.uint16, _dtypes.uint32, _dtypes.uint64, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"AddV2", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"AddV2", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def _all(input, axis, keep_dims=False, name=None):
r"""Computes the "logical and" of elements across dimensions of a tensor.
Reduces `input` along the dimensions given in `axis`. Unless
`keep_dims` is true, the rank of the tensor is reduced by 1 for each entry in
`axis`. If `keep_dims` is true, the reduced dimensions are
retained with length 1.
Args:
input: A `Tensor` of type `bool`. The tensor to reduce.
axis: A `Tensor`. Must be one of the following types: `int32`, `int64`.
The dimensions to reduce. Must be in the range
`[-rank(input), rank(input))`.
keep_dims: An optional `bool`. Defaults to `False`.
If true, retain reduced dimensions with length 1.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "All", name, input, axis, "keep_dims", keep_dims)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return _all_eager_fallback(
input, axis, keep_dims=keep_dims, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"All", input=input, reduction_indices=axis, keep_dims=keep_dims,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("keep_dims", _op._get_attr_bool("keep_dims"), "Tidx",
_op._get_attr_type("Tidx"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"All", _inputs_flat, _attrs, _result)
_result, = _result
return _result
All = tf_export("raw_ops.All")(_ops.to_raw_op(_all))
def _all_eager_fallback(input, axis, keep_dims, name, ctx):
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_attr_Tidx, (axis,) = _execute.args_to_matching_eager([axis], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
input = _ops.convert_to_tensor(input, _dtypes.bool)
_inputs_flat = [input, axis]
_attrs = ("keep_dims", keep_dims, "Tidx", _attr_Tidx)
_result = _execute.execute(b"All", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"All", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def angle(input, Tout=_dtypes.float32, name=None):
r"""Returns the argument of a complex number.
Given a tensor `input` of complex numbers, this operation returns a tensor of
type `float` that is the argument of each element in `input`. All elements in
`input` must be complex numbers of the form \\(a + bj\\), where *a*
is the real part and *b* is the imaginary part.
The argument returned by this operation is of the form \\(atan2(b, a)\\).
For example:
```
# tensor 'input' is [-2.25 + 4.75j, 3.25 + 5.75j]
tf.angle(input) ==> [2.0132, 1.056]
```
@compatibility(numpy)
Equivalent to np.angle.
@end_compatibility
Args:
input: A `Tensor`. Must be one of the following types: `complex64`, `complex128`.
Tout: An optional `tf.DType` from: `tf.float32, tf.float64`. Defaults to `tf.float32`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `Tout`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Angle", name, input, "Tout", Tout)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return angle_eager_fallback(
input, Tout=Tout, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if Tout is None:
Tout = _dtypes.float32
Tout = _execute.make_type(Tout, "Tout")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Angle", input=input, Tout=Tout, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tout",
_op._get_attr_type("Tout"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Angle", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Angle = tf_export("raw_ops.Angle")(_ops.to_raw_op(angle))
def angle_eager_fallback(input, Tout, name, ctx):
if Tout is None:
Tout = _dtypes.float32
Tout = _execute.make_type(Tout, "Tout")
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.complex64, _dtypes.complex128, ], _dtypes.complex64)
_inputs_flat = [input]
_attrs = ("T", _attr_T, "Tout", Tout)
_result = _execute.execute(b"Angle", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Angle", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def _any(input, axis, keep_dims=False, name=None):
r"""Computes the "logical or" of elements across dimensions of a tensor.
Reduces `input` along the dimensions given in `axis`. Unless
`keep_dims` is true, the rank of the tensor is reduced by 1 for each entry in
`axis`. If `keep_dims` is true, the reduced dimensions are
retained with length 1.
Args:
input: A `Tensor` of type `bool`. The tensor to reduce.
axis: A `Tensor`. Must be one of the following types: `int32`, `int64`.
The dimensions to reduce. Must be in the range
`[-rank(input), rank(input))`.
keep_dims: An optional `bool`. Defaults to `False`.
If true, retain reduced dimensions with length 1.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Any", name, input, axis, "keep_dims", keep_dims)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return _any_eager_fallback(
input, axis, keep_dims=keep_dims, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Any", input=input, reduction_indices=axis, keep_dims=keep_dims,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("keep_dims", _op._get_attr_bool("keep_dims"), "Tidx",
_op._get_attr_type("Tidx"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Any", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Any = tf_export("raw_ops.Any")(_ops.to_raw_op(_any))
def _any_eager_fallback(input, axis, keep_dims, name, ctx):
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_attr_Tidx, (axis,) = _execute.args_to_matching_eager([axis], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
input = _ops.convert_to_tensor(input, _dtypes.bool)
_inputs_flat = [input, axis]
_attrs = ("keep_dims", keep_dims, "Tidx", _attr_Tidx)
_result = _execute.execute(b"Any", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Any", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def approximate_equal(x, y, tolerance=1e-05, name=None):
r"""Returns the truth value of abs(x-y) < tolerance element-wise.
Args:
x: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
y: A `Tensor`. Must have the same type as `x`.
tolerance: An optional `float`. Defaults to `1e-05`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "ApproximateEqual", name, x, y, "tolerance", tolerance)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return approximate_equal_eager_fallback(
x, y, tolerance=tolerance, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if tolerance is None:
tolerance = 1e-05
tolerance = _execute.make_float(tolerance, "tolerance")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"ApproximateEqual", x=x, y=y, tolerance=tolerance, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "tolerance",
_op.get_attr("tolerance"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"ApproximateEqual", _inputs_flat, _attrs, _result)
_result, = _result
return _result
ApproximateEqual = tf_export("raw_ops.ApproximateEqual")(_ops.to_raw_op(approximate_equal))
def approximate_equal_eager_fallback(x, y, tolerance, name, ctx):
if tolerance is None:
tolerance = 1e-05
tolerance = _execute.make_float(tolerance, "tolerance")
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T, "tolerance", tolerance)
_result = _execute.execute(b"ApproximateEqual", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"ApproximateEqual", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def arg_max(input, dimension, output_type=_dtypes.int64, name=None):
r"""Returns the index with the largest value across dimensions of a tensor.
Note that in case of ties the identity of the return value is not guaranteed.
Usage:
```python
import tensorflow as tf
a = [1, 10, 26.9, 2.8, 166.32, 62.3]
b = tf.math.argmax(input = a)
c = tf.keras.backend.eval(b)
# c = 4
# here a[4] = 166.32 which is the largest element of a across axis 0
```
Args:
input: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`, `bool`.
dimension: A `Tensor`. Must be one of the following types: `int16`, `int32`, `int64`.
int16, int32 or int64, must be in the range `[-rank(input), rank(input))`.
Describes which dimension of the input Tensor to reduce across. For vectors,
use dimension = 0.
output_type: An optional `tf.DType` from: `tf.int16, tf.uint16, tf.int32, tf.int64`. Defaults to `tf.int64`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `output_type`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "ArgMax", name, input, dimension, "output_type", output_type)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return arg_max_eager_fallback(
input, dimension, output_type=output_type, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if output_type is None:
output_type = _dtypes.int64
output_type = _execute.make_type(output_type, "output_type")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"ArgMax", input=input, dimension=dimension, output_type=output_type,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tidx",
_op._get_attr_type("Tidx"), "output_type",
_op._get_attr_type("output_type"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"ArgMax", _inputs_flat, _attrs, _result)
_result, = _result
return _result
ArgMax = tf_export("raw_ops.ArgMax")(_ops.to_raw_op(arg_max))
def arg_max_eager_fallback(input, dimension, output_type, name, ctx):
if output_type is None:
output_type = _dtypes.int64
output_type = _execute.make_type(output_type, "output_type")
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, _dtypes.bool, ])
_attr_Tidx, (dimension,) = _execute.args_to_matching_eager([dimension], ctx, [_dtypes.int16, _dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [input, dimension]
_attrs = ("T", _attr_T, "Tidx", _attr_Tidx, "output_type", output_type)
_result = _execute.execute(b"ArgMax", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"ArgMax", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def arg_min(input, dimension, output_type=_dtypes.int64, name=None):
r"""Returns the index with the smallest value across dimensions of a tensor.
Note that in case of ties the identity of the return value is not guaranteed.
Usage:
```python
import tensorflow as tf
a = [1, 10, 26.9, 2.8, 166.32, 62.3]
b = tf.math.argmin(input = a)
c = tf.keras.backend.eval(b)
# c = 0
# here a[0] = 1 which is the smallest element of a across axis 0
```
Args:
input: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`, `bool`.
dimension: A `Tensor`. Must be one of the following types: `int32`, `int64`.
int32 or int64, must be in the range `[-rank(input), rank(input))`.
Describes which dimension of the input Tensor to reduce across. For vectors,
use dimension = 0.
output_type: An optional `tf.DType` from: `tf.int32, tf.int64`. Defaults to `tf.int64`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `output_type`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "ArgMin", name, input, dimension, "output_type", output_type)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return arg_min_eager_fallback(
input, dimension, output_type=output_type, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if output_type is None:
output_type = _dtypes.int64
output_type = _execute.make_type(output_type, "output_type")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"ArgMin", input=input, dimension=dimension, output_type=output_type,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tidx",
_op._get_attr_type("Tidx"), "output_type",
_op._get_attr_type("output_type"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"ArgMin", _inputs_flat, _attrs, _result)
_result, = _result
return _result
ArgMin = tf_export("raw_ops.ArgMin")(_ops.to_raw_op(arg_min))
def arg_min_eager_fallback(input, dimension, output_type, name, ctx):
if output_type is None:
output_type = _dtypes.int64
output_type = _execute.make_type(output_type, "output_type")
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, _dtypes.bool, ])
_attr_Tidx, (dimension,) = _execute.args_to_matching_eager([dimension], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [input, dimension]
_attrs = ("T", _attr_T, "Tidx", _attr_Tidx, "output_type", output_type)
_result = _execute.execute(b"ArgMin", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"ArgMin", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.asin', 'asin')
def asin(x, name=None):
r"""Computes the trignometric inverse sine of x element-wise.
The `tf.math.asin` operation returns the inverse of `tf.math.sin`, such that
if `y = tf.math.sin(x)` then, `x = tf.math.asin(y)`.
**Note**: The output of `tf.math.asin` will lie within the invertible range
of sine, i.e [-pi/2, pi/2].
For example:
```python
# Note: [1.047, 0.785] ~= [(pi/3), (pi/4)]
x = tf.constant([1.047, 0.785])
y = tf.math.sin(x) # [0.8659266, 0.7068252]
tf.math.asin(y) # [1.047, 0.785] = x
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int8`, `int16`, `int32`, `int64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Asin", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_asin(
(x, name,), None)
if _result is not NotImplemented:
return _result
return asin_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
asin, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_asin(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Asin", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
asin, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Asin", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Asin = tf_export("raw_ops.Asin")(_ops.to_raw_op(asin))
_dispatcher_for_asin = asin._tf_type_based_dispatcher.Dispatch
def asin_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Asin", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Asin", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.asinh', 'asinh')
def asinh(x, name=None):
r"""Computes inverse hyperbolic sine of x element-wise.
Given an input tensor, this function computes inverse hyperbolic sine
for every element in the tensor. Both input and output has a range of
`[-inf, inf]`.
```python
x = tf.constant([-float("inf"), -2, -0.5, 1, 1.2, 200, 10000, float("inf")])
tf.math.asinh(x) ==> [-inf -1.4436355 -0.4812118 0.8813736 1.0159732 5.991471 9.903487 inf]
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Asinh", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_asinh(
(x, name,), None)
if _result is not NotImplemented:
return _result
return asinh_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
asinh, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_asinh(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Asinh", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
asinh, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Asinh", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Asinh = tf_export("raw_ops.Asinh")(_ops.to_raw_op(asinh))
_dispatcher_for_asinh = asinh._tf_type_based_dispatcher.Dispatch
def asinh_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Asinh", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Asinh", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.atan', 'atan')
def atan(x, name=None):
r"""Computes the trignometric inverse tangent of x element-wise.
The `tf.math.atan` operation returns the inverse of `tf.math.tan`, such that
if `y = tf.math.tan(x)` then, `x = tf.math.atan(y)`.
**Note**: The output of `tf.math.atan` will lie within the invertible range
of tan, i.e (-pi/2, pi/2).
For example:
```python
# Note: [1.047, 0.785] ~= [(pi/3), (pi/4)]
x = tf.constant([1.047, 0.785])
y = tf.math.tan(x) # [1.731261, 0.99920404]
tf.math.atan(y) # [1.047, 0.785] = x
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int8`, `int16`, `int32`, `int64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Atan", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_atan(
(x, name,), None)
if _result is not NotImplemented:
return _result
return atan_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
atan, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_atan(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Atan", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
atan, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Atan", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Atan = tf_export("raw_ops.Atan")(_ops.to_raw_op(atan))
_dispatcher_for_atan = atan._tf_type_based_dispatcher.Dispatch
def atan_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Atan", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Atan", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.atan2', 'atan2')
def atan2(y, x, name=None):
r"""Computes arctangent of `y/x` element-wise, respecting signs of the arguments.
This is the angle \\( \theta \in [-\pi, \pi] \\) such that
\\[ x = r \cos(\theta) \\]
and
\\[ y = r \sin(\theta) \\]
where \\(r = \sqrt{x^2 + y^2} \\).
For example:
>>> x = [1., 1.]
>>> y = [1., -1.]
>>> print((tf.math.atan2(y,x) * (180 / np.pi)).numpy())
[ 45. -45.]
Args:
y: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
x: A `Tensor`. Must have the same type as `y`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `y`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Atan2", name, y, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_atan2(
(y, x, name,), None)
if _result is not NotImplemented:
return _result
return atan2_eager_fallback(
y, x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
atan2, (), dict(y=y, x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_atan2(
(y, x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Atan2", y=y, x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
atan2, (), dict(y=y, x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Atan2", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Atan2 = tf_export("raw_ops.Atan2")(_ops.to_raw_op(atan2))
_dispatcher_for_atan2 = atan2._tf_type_based_dispatcher.Dispatch
def atan2_eager_fallback(y, x, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([y, x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
(y, x) = _inputs_T
_inputs_flat = [y, x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Atan2", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Atan2", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.atanh', 'atanh')
def atanh(x, name=None):
r"""Computes inverse hyperbolic tangent of x element-wise.
Given an input tensor, this function computes inverse hyperbolic tangent
for every element in the tensor. Input range is `[-1,1]` and output range is
`[-inf, inf]`. If input is `-1`, output will be `-inf` and if the
input is `1`, output will be `inf`. Values outside the range will have
`nan` as output.
```python
x = tf.constant([-float("inf"), -1, -0.5, 1, 0, 0.5, 10, float("inf")])
tf.math.atanh(x) ==> [nan -inf -0.54930615 inf 0. 0.54930615 nan nan]
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Atanh", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_atanh(
(x, name,), None)
if _result is not NotImplemented:
return _result
return atanh_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
atanh, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_atanh(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Atanh", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
atanh, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Atanh", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Atanh = tf_export("raw_ops.Atanh")(_ops.to_raw_op(atanh))
_dispatcher_for_atanh = atanh._tf_type_based_dispatcher.Dispatch
def atanh_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Atanh", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Atanh", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def batch_mat_mul(x, y, adj_x=False, adj_y=False, name=None):
r"""Multiplies slices of two tensors in batches.
Multiplies all slices of `Tensor` `x` and `y` (each slice can be
viewed as an element of a batch), and arranges the individual results
in a single output tensor of the same batch size. Each of the
individual slices can optionally be adjointed (to adjoint a matrix
means to transpose and conjugate it) before multiplication by setting
the `adj_x` or `adj_y` flag to `True`, which are by default `False`.
The input tensors `x` and `y` are 2-D or higher with shape `[..., r_x, c_x]`
and `[..., r_y, c_y]`.
The output tensor is 2-D or higher with shape `[..., r_o, c_o]`, where:
r_o = c_x if adj_x else r_x
c_o = r_y if adj_y else c_y
It is computed as:
output[..., :, :] = matrix(x[..., :, :]) * matrix(y[..., :, :])
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int32`, `int64`, `complex64`, `complex128`.
2-D or higher with shape `[..., r_x, c_x]`.
y: A `Tensor`. Must have the same type as `x`.
2-D or higher with shape `[..., r_y, c_y]`.
adj_x: An optional `bool`. Defaults to `False`.
If `True`, adjoint the slices of `x`. Defaults to `False`.
adj_y: An optional `bool`. Defaults to `False`.
If `True`, adjoint the slices of `y`. Defaults to `False`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "BatchMatMul", name, x, y, "adj_x", adj_x, "adj_y", adj_y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return batch_mat_mul_eager_fallback(
x, y, adj_x=adj_x, adj_y=adj_y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if adj_x is None:
adj_x = False
adj_x = _execute.make_bool(adj_x, "adj_x")
if adj_y is None:
adj_y = False
adj_y = _execute.make_bool(adj_y, "adj_y")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"BatchMatMul", x=x, y=y, adj_x=adj_x, adj_y=adj_y, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "adj_x",
_op._get_attr_bool("adj_x"), "adj_y",
_op._get_attr_bool("adj_y"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"BatchMatMul", _inputs_flat, _attrs, _result)
_result, = _result
return _result
BatchMatMul = tf_export("raw_ops.BatchMatMul")(_ops.to_raw_op(batch_mat_mul))
def batch_mat_mul_eager_fallback(x, y, adj_x, adj_y, name, ctx):
if adj_x is None:
adj_x = False
adj_x = _execute.make_bool(adj_x, "adj_x")
if adj_y is None:
adj_y = False
adj_y = _execute.make_bool(adj_y, "adj_y")
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T, "adj_x", adj_x, "adj_y", adj_y)
_result = _execute.execute(b"BatchMatMul", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"BatchMatMul", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def batch_mat_mul_v2(x, y, adj_x=False, adj_y=False, name=None):
r"""Multiplies slices of two tensors in batches.
Multiplies all slices of `Tensor` `x` and `y` (each slice can be
viewed as an element of a batch), and arranges the individual results
in a single output tensor of the same batch size. Each of the
individual slices can optionally be adjointed (to adjoint a matrix
means to transpose and conjugate it) before multiplication by setting
the `adj_x` or `adj_y` flag to `True`, which are by default `False`.
The input tensors `x` and `y` are 2-D or higher with shape `[..., r_x, c_x]`
and `[..., r_y, c_y]`.
The output tensor is 2-D or higher with shape `[..., r_o, c_o]`, where:
r_o = c_x if adj_x else r_x
c_o = r_y if adj_y else c_y
It is computed as:
output[..., :, :] = matrix(x[..., :, :]) * matrix(y[..., :, :])
*NOTE*: `BatchMatMulV2` supports broadcasting in the batch dimensions. More
about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html).
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int16`, `int32`, `int64`, `complex64`, `complex128`.
2-D or higher with shape `[..., r_x, c_x]`.
y: A `Tensor`. Must have the same type as `x`.
2-D or higher with shape `[..., r_y, c_y]`.
adj_x: An optional `bool`. Defaults to `False`.
If `True`, adjoint the slices of `x`. Defaults to `False`.
adj_y: An optional `bool`. Defaults to `False`.
If `True`, adjoint the slices of `y`. Defaults to `False`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "BatchMatMulV2", name, x, y, "adj_x", adj_x, "adj_y", adj_y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return batch_mat_mul_v2_eager_fallback(
x, y, adj_x=adj_x, adj_y=adj_y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if adj_x is None:
adj_x = False
adj_x = _execute.make_bool(adj_x, "adj_x")
if adj_y is None:
adj_y = False
adj_y = _execute.make_bool(adj_y, "adj_y")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"BatchMatMulV2", x=x, y=y, adj_x=adj_x, adj_y=adj_y, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "adj_x",
_op._get_attr_bool("adj_x"), "adj_y",
_op._get_attr_bool("adj_y"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"BatchMatMulV2", _inputs_flat, _attrs, _result)
_result, = _result
return _result
BatchMatMulV2 = tf_export("raw_ops.BatchMatMulV2")(_ops.to_raw_op(batch_mat_mul_v2))
def batch_mat_mul_v2_eager_fallback(x, y, adj_x, adj_y, name, ctx):
if adj_x is None:
adj_x = False
adj_x = _execute.make_bool(adj_x, "adj_x")
if adj_y is None:
adj_y = False
adj_y = _execute.make_bool(adj_y, "adj_y")
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T, "adj_x", adj_x, "adj_y", adj_y)
_result = _execute.execute(b"BatchMatMulV2", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"BatchMatMulV2", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def batch_mat_mul_v3(x, y, Tout, adj_x=False, adj_y=False, name=None):
r"""Multiplies slices of two tensors in batches.
Multiplies all slices of `Tensor` `x` and `y` (each slice can be
viewed as an element of a batch), and arranges the individual results
in a single output tensor of the same batch size. Each of the
individual slices can optionally be adjointed (to adjoint a matrix
means to transpose and conjugate it) before multiplication by setting
the `adj_x` or `adj_y` flag to `True`, which are by default `False`.
The input tensors `x` and `y` are 2-D or higher with shape `[..., r_x, c_x]`
and `[..., r_y, c_y]`.
The output tensor is 2-D or higher with shape `[..., r_o, c_o]`, where:
r_o = c_x if adj_x else r_x
c_o = r_y if adj_y else c_y
It is computed as:
output[..., :, :] = matrix(x[..., :, :]) * matrix(y[..., :, :])
*NOTE*: `BatchMatMulV3` supports broadcasting in the batch dimensions. More
about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html).
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `uint8`, `int8`, `int16`, `int32`, `int64`, `complex64`, `complex128`.
2-D or higher with shape `[..., r_x, c_x]`.
y: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `uint8`, `int8`, `int16`, `int32`, `int64`, `complex64`, `complex128`.
2-D or higher with shape `[..., r_y, c_y]`.
Tout: A `tf.DType` from: `tf.bfloat16, tf.half, tf.float32, tf.float64, tf.int16, tf.int32, tf.int64, tf.complex64, tf.complex128`.
If not spcified, Tout is the same type to input type.
adj_x: An optional `bool`. Defaults to `False`.
If `True`, adjoint the slices of `x`. Defaults to `False`.
adj_y: An optional `bool`. Defaults to `False`.
If `True`, adjoint the slices of `y`. Defaults to `False`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `Tout`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "BatchMatMulV3", name, x, y, "Tout", Tout, "adj_x", adj_x,
"adj_y", adj_y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return batch_mat_mul_v3_eager_fallback(
x, y, Tout=Tout, adj_x=adj_x, adj_y=adj_y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
Tout = _execute.make_type(Tout, "Tout")
if adj_x is None:
adj_x = False
adj_x = _execute.make_bool(adj_x, "adj_x")
if adj_y is None:
adj_y = False
adj_y = _execute.make_bool(adj_y, "adj_y")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"BatchMatMulV3", x=x, y=y, Tout=Tout, adj_x=adj_x, adj_y=adj_y,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("Ta", _op._get_attr_type("Ta"), "Tb", _op._get_attr_type("Tb"),
"Tout", _op._get_attr_type("Tout"), "adj_x",
_op._get_attr_bool("adj_x"), "adj_y",
_op._get_attr_bool("adj_y"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"BatchMatMulV3", _inputs_flat, _attrs, _result)
_result, = _result
return _result
BatchMatMulV3 = tf_export("raw_ops.BatchMatMulV3")(_ops.to_raw_op(batch_mat_mul_v3))
def batch_mat_mul_v3_eager_fallback(x, y, Tout, adj_x, adj_y, name, ctx):
Tout = _execute.make_type(Tout, "Tout")
if adj_x is None:
adj_x = False
adj_x = _execute.make_bool(adj_x, "adj_x")
if adj_y is None:
adj_y = False
adj_y = _execute.make_bool(adj_y, "adj_y")
_attr_Ta, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.uint8, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
_attr_Tb, (y,) = _execute.args_to_matching_eager([y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.uint8, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x, y]
_attrs = ("Ta", _attr_Ta, "Tb", _attr_Tb, "Tout", Tout, "adj_x", adj_x,
"adj_y", adj_y)
_result = _execute.execute(b"BatchMatMulV3", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"BatchMatMulV3", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.betainc', v1=['math.betainc', 'betainc'])
@deprecated_endpoints('betainc')
def betainc(a, b, x, name=None):
r"""Compute the regularized incomplete beta integral \\(I_x(a, b)\\).
The regularized incomplete beta integral is defined as:
\\(I_x(a, b) = \frac{B(x; a, b)}{B(a, b)}\\)
where
\\(B(x; a, b) = \int_0^x t^{a-1} (1 - t)^{b-1} dt\\)
is the incomplete beta function and \\(B(a, b)\\) is the *complete*
beta function.
Args:
a: A `Tensor`. Must be one of the following types: `float32`, `float64`.
b: A `Tensor`. Must have the same type as `a`.
x: A `Tensor`. Must have the same type as `a`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `a`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Betainc", name, a, b, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_betainc(
(a, b, x, name,), None)
if _result is not NotImplemented:
return _result
return betainc_eager_fallback(
a, b, x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
betainc, (), dict(a=a, b=b, x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_betainc(
(a, b, x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Betainc", a=a, b=b, x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
betainc, (), dict(a=a, b=b, x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Betainc", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Betainc = tf_export("raw_ops.Betainc")(_ops.to_raw_op(betainc))
_dispatcher_for_betainc = betainc._tf_type_based_dispatcher.Dispatch
def betainc_eager_fallback(a, b, x, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([a, b, x], ctx, [_dtypes.float32, _dtypes.float64, ])
(a, b, x) = _inputs_T
_inputs_flat = [a, b, x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Betainc", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Betainc", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def bincount(arr, size, weights, name=None):
r"""Counts the number of occurrences of each value in an integer array.
Outputs a vector with length `size` and the same dtype as `weights`. If
`weights` are empty, then index `i` stores the number of times the value `i` is
counted in `arr`. If `weights` are non-empty, then index `i` stores the sum of
the value in `weights` at each index where the corresponding value in `arr` is
`i`.
Values in `arr` outside of the range [0, size) are ignored.
Args:
arr: A `Tensor` of type `int32`. int32 `Tensor`.
size: A `Tensor` of type `int32`. non-negative int32 scalar `Tensor`.
weights: A `Tensor`. Must be one of the following types: `int32`, `int64`, `float32`, `float64`.
is an int32, int64, float32, or float64 `Tensor` with the same
shape as `arr`, or a length-0 `Tensor`, in which case it acts as all weights
equal to 1.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `weights`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Bincount", name, arr, size, weights)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return bincount_eager_fallback(
arr, size, weights, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Bincount", arr=arr, size=size, weights=weights, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Bincount", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Bincount = tf_export("raw_ops.Bincount")(_ops.to_raw_op(bincount))
def bincount_eager_fallback(arr, size, weights, name, ctx):
_attr_T, (weights,) = _execute.args_to_matching_eager([weights], ctx, [_dtypes.int32, _dtypes.int64, _dtypes.float32, _dtypes.float64, ])
arr = _ops.convert_to_tensor(arr, _dtypes.int32)
size = _ops.convert_to_tensor(size, _dtypes.int32)
_inputs_flat = [arr, size, weights]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Bincount", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Bincount", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def bucketize(input, boundaries, name=None):
r"""Bucketizes 'input' based on 'boundaries'.
For example, if the inputs are
boundaries = [0, 10, 100]
input = [[-5, 10000]
[150, 10]
[5, 100]]
then the output will be
output = [[0, 3]
[3, 2]
[1, 3]]
Args:
input: A `Tensor`. Must be one of the following types: `int32`, `int64`, `float32`, `float64`.
Any shape of Tensor contains with int or float type.
boundaries: A list of `floats`.
A sorted list of floats gives the boundary of the buckets.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `int32`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Bucketize", name, input, "boundaries", boundaries)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return bucketize_eager_fallback(
input, boundaries=boundaries, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if not isinstance(boundaries, (list, tuple)):
raise TypeError(
"Expected list for 'boundaries' argument to "
"'bucketize' Op, not %r." % boundaries)
boundaries = [_execute.make_float(_f, "boundaries") for _f in boundaries]
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Bucketize", input=input, boundaries=boundaries, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "boundaries",
_op.get_attr("boundaries"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Bucketize", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Bucketize = tf_export("raw_ops.Bucketize")(_ops.to_raw_op(bucketize))
def bucketize_eager_fallback(input, boundaries, name, ctx):
if not isinstance(boundaries, (list, tuple)):
raise TypeError(
"Expected list for 'boundaries' argument to "
"'bucketize' Op, not %r." % boundaries)
boundaries = [_execute.make_float(_f, "boundaries") for _f in boundaries]
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.int32, _dtypes.int64, _dtypes.float32, _dtypes.float64, ])
_inputs_flat = [input]
_attrs = ("T", _attr_T, "boundaries", boundaries)
_result = _execute.execute(b"Bucketize", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Bucketize", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def cast(x, DstT, Truncate=False, name=None):
r"""Cast x of type SrcT to y of DstT.
Args:
x: A `Tensor`.
DstT: A `tf.DType`.
Truncate: An optional `bool`. Defaults to `False`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `DstT`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Cast", name, x, "DstT", DstT, "Truncate", Truncate)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return cast_eager_fallback(
x, DstT=DstT, Truncate=Truncate, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
DstT = _execute.make_type(DstT, "DstT")
if Truncate is None:
Truncate = False
Truncate = _execute.make_bool(Truncate, "Truncate")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Cast", x=x, DstT=DstT, Truncate=Truncate, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("SrcT", _op._get_attr_type("SrcT"), "DstT",
_op._get_attr_type("DstT"), "Truncate",
_op._get_attr_bool("Truncate"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Cast", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Cast = tf_export("raw_ops.Cast")(_ops.to_raw_op(cast))
def cast_eager_fallback(x, DstT, Truncate, name, ctx):
DstT = _execute.make_type(DstT, "DstT")
if Truncate is None:
Truncate = False
Truncate = _execute.make_bool(Truncate, "Truncate")
_attr_SrcT, (x,) = _execute.args_to_matching_eager([x], ctx, [])
_inputs_flat = [x]
_attrs = ("SrcT", _attr_SrcT, "DstT", DstT, "Truncate", Truncate)
_result = _execute.execute(b"Cast", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Cast", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def ceil(x, name=None):
r"""Returns element-wise smallest integer not less than x.
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Ceil", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return ceil_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Ceil", x=x, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Ceil", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Ceil = tf_export("raw_ops.Ceil")(_ops.to_raw_op(ceil))
def ceil_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Ceil", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Ceil", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def _clip_by_value(t, clip_value_min, clip_value_max, name=None):
r"""Clips tensor values to a specified min and max.
Given a tensor `t`, this operation returns a tensor of the same type and
shape as `t` with its values clipped to `clip_value_min` and `clip_value_max`.
Any values less than `clip_value_min` are set to `clip_value_min`. Any values
greater than `clip_value_max` are set to `clip_value_max`.
Args:
t: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
A `Tensor`.
clip_value_min: A `Tensor`. Must have the same type as `t`.
A 0-D (scalar) `Tensor`, or a `Tensor` with the same shape
as `t`. The minimum value to clip by.
clip_value_max: A `Tensor`. Must have the same type as `t`.
A 0-D (scalar) `Tensor`, or a `Tensor` with the same shape
as `t`. The maximum value to clip by.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `t`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "ClipByValue", name, t, clip_value_min, clip_value_max)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return _clip_by_value_eager_fallback(
t, clip_value_min, clip_value_max, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"ClipByValue", t=t, clip_value_min=clip_value_min,
clip_value_max=clip_value_max, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"ClipByValue", _inputs_flat, _attrs, _result)
_result, = _result
return _result
ClipByValue = tf_export("raw_ops.ClipByValue")(_ops.to_raw_op(_clip_by_value))
def _clip_by_value_eager_fallback(t, clip_value_min, clip_value_max, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([t, clip_value_min, clip_value_max], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
(t, clip_value_min, clip_value_max) = _inputs_T
_inputs_flat = [t, clip_value_min, clip_value_max]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"ClipByValue", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"ClipByValue", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def _complex(real, imag, Tout=_dtypes.complex64, name=None):
r"""Converts two real numbers to a complex number.
Given a tensor `real` representing the real part of a complex number, and a
tensor `imag` representing the imaginary part of a complex number, this
operation returns complex numbers elementwise of the form \\(a + bj\\), where
*a* represents the `real` part and *b* represents the `imag` part.
The input tensors `real` and `imag` must have the same shape.
For example:
```
# tensor 'real' is [2.25, 3.25]
# tensor `imag` is [4.75, 5.75]
tf.complex(real, imag) ==> [[2.25 + 4.75j], [3.25 + 5.75j]]
```
Args:
real: A `Tensor`. Must be one of the following types: `float32`, `float64`.
imag: A `Tensor`. Must have the same type as `real`.
Tout: An optional `tf.DType` from: `tf.complex64, tf.complex128`. Defaults to `tf.complex64`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `Tout`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Complex", name, real, imag, "Tout", Tout)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return _complex_eager_fallback(
real, imag, Tout=Tout, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if Tout is None:
Tout = _dtypes.complex64
Tout = _execute.make_type(Tout, "Tout")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Complex", real=real, imag=imag, Tout=Tout, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tout",
_op._get_attr_type("Tout"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Complex", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Complex = tf_export("raw_ops.Complex")(_ops.to_raw_op(_complex))
def _complex_eager_fallback(real, imag, Tout, name, ctx):
if Tout is None:
Tout = _dtypes.complex64
Tout = _execute.make_type(Tout, "Tout")
_attr_T, _inputs_T = _execute.args_to_matching_eager([real, imag], ctx, [_dtypes.float32, _dtypes.float64, ], _dtypes.float32)
(real, imag) = _inputs_T
_inputs_flat = [real, imag]
_attrs = ("T", _attr_T, "Tout", Tout)
_result = _execute.execute(b"Complex", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Complex", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def complex_abs(x, Tout=_dtypes.float32, name=None):
r"""Computes the complex absolute value of a tensor.
Given a tensor `x` of complex numbers, this operation returns a tensor of type
`float` or `double` that is the absolute value of each element in `x`. All
elements in `x` must be complex numbers of the form \\(a + bj\\). The absolute
value is computed as \\( \sqrt{a^2 + b^2}\\).
For example:
>>> x = tf.complex(3.0, 4.0)
>>> print((tf.raw_ops.ComplexAbs(x=x, Tout=tf.dtypes.float32, name=None)).numpy())
5.0
Args:
x: A `Tensor`. Must be one of the following types: `complex64`, `complex128`.
Tout: An optional `tf.DType` from: `tf.float32, tf.float64`. Defaults to `tf.float32`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `Tout`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "ComplexAbs", name, x, "Tout", Tout)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return complex_abs_eager_fallback(
x, Tout=Tout, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if Tout is None:
Tout = _dtypes.float32
Tout = _execute.make_type(Tout, "Tout")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"ComplexAbs", x=x, Tout=Tout, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tout",
_op._get_attr_type("Tout"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"ComplexAbs", _inputs_flat, _attrs, _result)
_result, = _result
return _result
ComplexAbs = tf_export("raw_ops.ComplexAbs")(_ops.to_raw_op(complex_abs))
def complex_abs_eager_fallback(x, Tout, name, ctx):
if Tout is None:
Tout = _dtypes.float32
Tout = _execute.make_type(Tout, "Tout")
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.complex64, _dtypes.complex128, ], _dtypes.complex64)
_inputs_flat = [x]
_attrs = ("T", _attr_T, "Tout", Tout)
_result = _execute.execute(b"ComplexAbs", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"ComplexAbs", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def conj(input, name=None):
r"""Returns the complex conjugate of a complex number.
Given a tensor `input` of complex numbers, this operation returns a tensor of
complex numbers that are the complex conjugate of each element in `input`. The
complex numbers in `input` must be of the form \\(a + bj\\), where *a* is the
real part and *b* is the imaginary part.
The complex conjugate returned by this operation is of the form \\(a - bj\\).
For example:
```
# tensor 'input' is [-2.25 + 4.75j, 3.25 + 5.75j]
tf.conj(input) ==> [-2.25 - 4.75j, 3.25 - 5.75j]
```
Args:
input: A `Tensor`. Must be one of the following types: `complex64`, `complex128`, `variant`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `input`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Conj", name, input)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return conj_eager_fallback(
input, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Conj", input=input, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Conj", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Conj = tf_export("raw_ops.Conj")(_ops.to_raw_op(conj))
def conj_eager_fallback(input, name, ctx):
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.complex64, _dtypes.complex128, _dtypes.variant, ], _dtypes.complex64)
_inputs_flat = [input]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Conj", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Conj", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.cos', 'cos')
def cos(x, name=None):
r"""Computes cos of x element-wise.
Given an input tensor, this function computes cosine of every
element in the tensor. Input range is `(-inf, inf)` and
output range is `[-1,1]`. If input lies outside the boundary, `nan`
is returned.
```python
x = tf.constant([-float("inf"), -9, -0.5, 1, 1.2, 200, 10000, float("inf")])
tf.math.cos(x) ==> [nan -0.91113025 0.87758255 0.5403023 0.36235774 0.48718765 -0.95215535 nan]
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Cos", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_cos(
(x, name,), None)
if _result is not NotImplemented:
return _result
return cos_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
cos, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_cos(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Cos", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
cos, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Cos", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Cos = tf_export("raw_ops.Cos")(_ops.to_raw_op(cos))
_dispatcher_for_cos = cos._tf_type_based_dispatcher.Dispatch
def cos_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Cos", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Cos", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.cosh', 'cosh')
def cosh(x, name=None):
r"""Computes hyperbolic cosine of x element-wise.
Given an input tensor, this function computes hyperbolic cosine of every
element in the tensor. Input range is `[-inf, inf]` and output range
is `[1, inf]`.
```python
x = tf.constant([-float("inf"), -9, -0.5, 1, 1.2, 2, 10, float("inf")])
tf.math.cosh(x) ==> [inf 4.0515420e+03 1.1276259e+00 1.5430807e+00 1.8106556e+00 3.7621956e+00 1.1013233e+04 inf]
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Cosh", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_cosh(
(x, name,), None)
if _result is not NotImplemented:
return _result
return cosh_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
cosh, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_cosh(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Cosh", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
cosh, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Cosh", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Cosh = tf_export("raw_ops.Cosh")(_ops.to_raw_op(cosh))
_dispatcher_for_cosh = cosh._tf_type_based_dispatcher.Dispatch
def cosh_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Cosh", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Cosh", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('linalg.cross', v1=['linalg.cross', 'cross'])
@deprecated_endpoints('cross')
def cross(a, b, name=None):
r"""Compute the pairwise cross product.
`a` and `b` must be the same shape; they can either be simple 3-element vectors,
or any shape where the innermost dimension is 3. In the latter case, each pair
of corresponding 3-element vectors is cross-multiplied independently.
Args:
a: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `int64`, `bfloat16`, `uint16`, `half`, `uint32`, `uint64`.
A tensor containing 3-element vectors.
b: A `Tensor`. Must have the same type as `a`.
Another tensor, of same type and shape as `a`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `a`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Cross", name, a, b)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_cross(
(a, b, name,), None)
if _result is not NotImplemented:
return _result
return cross_eager_fallback(
a, b, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
cross, (), dict(a=a, b=b, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_cross(
(a, b, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Cross", a=a, b=b, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
cross, (), dict(a=a, b=b, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Cross", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Cross = tf_export("raw_ops.Cross")(_ops.to_raw_op(cross))
_dispatcher_for_cross = cross._tf_type_based_dispatcher.Dispatch
def cross_eager_fallback(a, b, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([a, b], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.int64, _dtypes.bfloat16, _dtypes.uint16, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
(a, b) = _inputs_T
_inputs_flat = [a, b]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Cross", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Cross", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def cumprod(x, axis, exclusive=False, reverse=False, name=None):
r"""Compute the cumulative product of the tensor `x` along `axis`.
By default, this op performs an inclusive cumprod, which means that the first
element of the input is identical to the first element of the output:
```python
tf.cumprod([a, b, c]) # => [a, a * b, a * b * c]
```
By setting the `exclusive` kwarg to `True`, an exclusive cumprod is
performed instead:
```python
tf.cumprod([a, b, c], exclusive=True) # => [1, a, a * b]
```
By setting the `reverse` kwarg to `True`, the cumprod is performed in the
opposite direction:
```python
tf.cumprod([a, b, c], reverse=True) # => [a * b * c, b * c, c]
```
This is more efficient than using separate `tf.reverse` ops.
The `reverse` and `exclusive` kwargs can also be combined:
```python
tf.cumprod([a, b, c], exclusive=True, reverse=True) # => [b * c, c, 1]
```
Args:
x: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
A `Tensor`. Must be one of the following types: `float32`, `float64`,
`int64`, `int32`, `uint8`, `uint16`, `int16`, `int8`, `complex64`,
`complex128`, `qint8`, `quint8`, `qint32`, `half`.
axis: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A `Tensor` of type `int32` (default: 0). Must be in the range
`[-rank(x), rank(x))`.
exclusive: An optional `bool`. Defaults to `False`.
If `True`, perform exclusive cumprod.
reverse: An optional `bool`. Defaults to `False`.
A `bool` (default: False).
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Cumprod", name, x, axis, "exclusive", exclusive, "reverse",
reverse)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return cumprod_eager_fallback(
x, axis, exclusive=exclusive, reverse=reverse, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if exclusive is None:
exclusive = False
exclusive = _execute.make_bool(exclusive, "exclusive")
if reverse is None:
reverse = False
reverse = _execute.make_bool(reverse, "reverse")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Cumprod", x=x, axis=axis, exclusive=exclusive, reverse=reverse,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("exclusive", _op._get_attr_bool("exclusive"), "reverse",
_op._get_attr_bool("reverse"), "T", _op._get_attr_type("T"),
"Tidx", _op._get_attr_type("Tidx"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Cumprod", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Cumprod = tf_export("raw_ops.Cumprod")(_ops.to_raw_op(cumprod))
def cumprod_eager_fallback(x, axis, exclusive, reverse, name, ctx):
if exclusive is None:
exclusive = False
exclusive = _execute.make_bool(exclusive, "exclusive")
if reverse is None:
reverse = False
reverse = _execute.make_bool(reverse, "reverse")
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tidx, (axis,) = _execute.args_to_matching_eager([axis], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [x, axis]
_attrs = ("exclusive", exclusive, "reverse", reverse, "T", _attr_T, "Tidx",
_attr_Tidx)
_result = _execute.execute(b"Cumprod", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Cumprod", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def cumsum(x, axis, exclusive=False, reverse=False, name=None):
r"""Compute the cumulative sum of the tensor `x` along `axis`.
By default, this op performs an inclusive cumsum, which means that the first
element of the input is identical to the first element of the output:
```python
tf.cumsum([a, b, c]) # => [a, a + b, a + b + c]
```
By setting the `exclusive` kwarg to `True`, an exclusive cumsum is
performed instead:
```python
tf.cumsum([a, b, c], exclusive=True) # => [0, a, a + b]
```
By setting the `reverse` kwarg to `True`, the cumsum is performed in the
opposite direction:
```python
tf.cumsum([a, b, c], reverse=True) # => [a + b + c, b + c, c]
```
This is more efficient than using separate `tf.reverse` ops.
The `reverse` and `exclusive` kwargs can also be combined:
```python
tf.cumsum([a, b, c], exclusive=True, reverse=True) # => [b + c, c, 0]
```
Args:
x: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
A `Tensor`. Must be one of the following types: `float32`, `float64`,
`int64`, `int32`, `uint8`, `uint16`, `int16`, `int8`, `complex64`,
`complex128`, `qint8`, `quint8`, `qint32`, `half`.
axis: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A `Tensor` of type `int32` (default: 0). Must be in the range
`[-rank(x), rank(x))`.
exclusive: An optional `bool`. Defaults to `False`.
If `True`, perform exclusive cumsum.
reverse: An optional `bool`. Defaults to `False`.
A `bool` (default: False).
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Cumsum", name, x, axis, "exclusive", exclusive, "reverse",
reverse)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return cumsum_eager_fallback(
x, axis, exclusive=exclusive, reverse=reverse, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if exclusive is None:
exclusive = False
exclusive = _execute.make_bool(exclusive, "exclusive")
if reverse is None:
reverse = False
reverse = _execute.make_bool(reverse, "reverse")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Cumsum", x=x, axis=axis, exclusive=exclusive, reverse=reverse,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("exclusive", _op._get_attr_bool("exclusive"), "reverse",
_op._get_attr_bool("reverse"), "T", _op._get_attr_type("T"),
"Tidx", _op._get_attr_type("Tidx"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Cumsum", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Cumsum = tf_export("raw_ops.Cumsum")(_ops.to_raw_op(cumsum))
def cumsum_eager_fallback(x, axis, exclusive, reverse, name, ctx):
if exclusive is None:
exclusive = False
exclusive = _execute.make_bool(exclusive, "exclusive")
if reverse is None:
reverse = False
reverse = _execute.make_bool(reverse, "reverse")
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tidx, (axis,) = _execute.args_to_matching_eager([axis], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [x, axis]
_attrs = ("exclusive", exclusive, "reverse", reverse, "T", _attr_T, "Tidx",
_attr_Tidx)
_result = _execute.execute(b"Cumsum", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Cumsum", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def cumulative_logsumexp(x, axis, exclusive=False, reverse=False, name=None):
r"""Compute the cumulative product of the tensor `x` along `axis`.
By default, this op performs an inclusive cumulative log-sum-exp,
which means that the first
element of the input is identical to the first element of the output:
```python
tf.math.cumulative_logsumexp([a, b, c]) # => [a, log(exp(a) + exp(b)), log(exp(a) + exp(b) + exp(c))]
```
By setting the `exclusive` kwarg to `True`, an exclusive cumulative log-sum-exp is
performed instead:
```python
tf.cumulative_logsumexp([a, b, c], exclusive=True) # => [-inf, a, log(exp(a) * exp(b))]
```
Note that the neutral element of the log-sum-exp operation is `-inf`,
however, for performance reasons, the minimal value representable by the
floating point type is used instead.
By setting the `reverse` kwarg to `True`, the cumulative log-sum-exp is performed in the
opposite direction.
Args:
x: A `Tensor`. Must be one of the following types: `half`, `float32`, `float64`.
A `Tensor`. Must be one of the following types: `float16`, `float32`, `float64`.
axis: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A `Tensor` of type `int32` (default: 0). Must be in the range
`[-rank(x), rank(x))`.
exclusive: An optional `bool`. Defaults to `False`.
If `True`, perform exclusive cumulative log-sum-exp.
reverse: An optional `bool`. Defaults to `False`.
A `bool` (default: False).
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "CumulativeLogsumexp", name, x, axis, "exclusive", exclusive,
"reverse", reverse)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return cumulative_logsumexp_eager_fallback(
x, axis, exclusive=exclusive, reverse=reverse, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if exclusive is None:
exclusive = False
exclusive = _execute.make_bool(exclusive, "exclusive")
if reverse is None:
reverse = False
reverse = _execute.make_bool(reverse, "reverse")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"CumulativeLogsumexp", x=x, axis=axis, exclusive=exclusive,
reverse=reverse, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("exclusive", _op._get_attr_bool("exclusive"), "reverse",
_op._get_attr_bool("reverse"), "T", _op._get_attr_type("T"),
"Tidx", _op._get_attr_type("Tidx"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"CumulativeLogsumexp", _inputs_flat, _attrs, _result)
_result, = _result
return _result
CumulativeLogsumexp = tf_export("raw_ops.CumulativeLogsumexp")(_ops.to_raw_op(cumulative_logsumexp))
def cumulative_logsumexp_eager_fallback(x, axis, exclusive, reverse, name, ctx):
if exclusive is None:
exclusive = False
exclusive = _execute.make_bool(exclusive, "exclusive")
if reverse is None:
reverse = False
reverse = _execute.make_bool(reverse, "reverse")
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.half, _dtypes.float32, _dtypes.float64, ])
_attr_Tidx, (axis,) = _execute.args_to_matching_eager([axis], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [x, axis]
_attrs = ("exclusive", exclusive, "reverse", reverse, "T", _attr_T, "Tidx",
_attr_Tidx)
_result = _execute.execute(b"CumulativeLogsumexp", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"CumulativeLogsumexp", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def dense_bincount(input, size, weights, binary_output=False, name=None):
r"""Counts the number of occurrences of each value in an integer array.
Outputs a vector with length `size` and the same dtype as `weights`. If
`weights` are empty, then index `i` stores the number of times the value `i` is
counted in `arr`. If `weights` are non-empty, then index `i` stores the sum of
the value in `weights` at each index where the corresponding value in `arr` is
`i`.
Values in `arr` outside of the range [0, size) are ignored.
Args:
input: A `Tensor`. Must be one of the following types: `int32`, `int64`.
1D or 2D int `Tensor`.
size: A `Tensor`. Must have the same type as `input`.
non-negative int scalar `Tensor`.
weights: A `Tensor`. Must be one of the following types: `int32`, `int64`, `float32`, `float64`.
is an int32, int64, float32, or float64 `Tensor` with the same
shape as `arr`, or a length-0 `Tensor`, in which case it acts as all weights
equal to 1.
binary_output: An optional `bool`. Defaults to `False`.
bool; Whether the kernel should count the appearance or number of occurrences.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `weights`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "DenseBincount", name, input, size, weights, "binary_output",
binary_output)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return dense_bincount_eager_fallback(
input, size, weights, binary_output=binary_output, name=name,
ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if binary_output is None:
binary_output = False
binary_output = _execute.make_bool(binary_output, "binary_output")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"DenseBincount", input=input, size=size, weights=weights,
binary_output=binary_output, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("Tidx", _op._get_attr_type("Tidx"), "T",
_op._get_attr_type("T"), "binary_output",
_op._get_attr_bool("binary_output"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"DenseBincount", _inputs_flat, _attrs, _result)
_result, = _result
return _result
DenseBincount = tf_export("raw_ops.DenseBincount")(_ops.to_raw_op(dense_bincount))
def dense_bincount_eager_fallback(input, size, weights, binary_output, name, ctx):
if binary_output is None:
binary_output = False
binary_output = _execute.make_bool(binary_output, "binary_output")
_attr_Tidx, _inputs_Tidx = _execute.args_to_matching_eager([input, size], ctx, [_dtypes.int32, _dtypes.int64, ])
(input, size) = _inputs_Tidx
_attr_T, (weights,) = _execute.args_to_matching_eager([weights], ctx, [_dtypes.int32, _dtypes.int64, _dtypes.float32, _dtypes.float64, ])
_inputs_flat = [input, size, weights]
_attrs = ("Tidx", _attr_Tidx, "T", _attr_T, "binary_output", binary_output)
_result = _execute.execute(b"DenseBincount", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"DenseBincount", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.digamma', v1=['math.digamma', 'digamma'])
@deprecated_endpoints('digamma')
def digamma(x, name=None):
r"""Computes Psi, the derivative of Lgamma (the log of the absolute value of
`Gamma(x)`), element-wise.
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Digamma", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_digamma(
(x, name,), None)
if _result is not NotImplemented:
return _result
return digamma_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
digamma, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_digamma(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Digamma", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
digamma, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Digamma", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Digamma = tf_export("raw_ops.Digamma")(_ops.to_raw_op(digamma))
_dispatcher_for_digamma = digamma._tf_type_based_dispatcher.Dispatch
def digamma_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Digamma", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Digamma", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def div(x, y, name=None):
r"""Returns x / y element-wise.
*NOTE*: `Div` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `uint8`, `int8`, `uint16`, `int16`, `int32`, `uint32`, `uint64`, `int64`, `complex64`, `complex128`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Div", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return div_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Div", x=x, y=y, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Div", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Div = tf_export("raw_ops.Div")(_ops.to_raw_op(div))
def div_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.uint8, _dtypes.int8, _dtypes.uint16, _dtypes.int16, _dtypes.int32, _dtypes.uint32, _dtypes.uint64, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Div", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Div", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def div_no_nan(x, y, name=None):
r"""Returns 0 if the denominator is zero.
*NOTE*: `DivNoNan` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Args:
x: A `Tensor`. Must be one of the following types: `half`, `float32`, `bfloat16`, `float64`, `complex64`, `complex128`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "DivNoNan", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return div_no_nan_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"DivNoNan", x=x, y=y, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"DivNoNan", _inputs_flat, _attrs, _result)
_result, = _result
return _result
DivNoNan = tf_export("raw_ops.DivNoNan")(_ops.to_raw_op(div_no_nan))
def div_no_nan_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.half, _dtypes.float32, _dtypes.bfloat16, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"DivNoNan", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"DivNoNan", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def equal(x, y, incompatible_shape_error=True, name=None):
r"""Returns the truth value of (x == y) element-wise.
*NOTE*: `Equal` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
```python
x = tf.constant([2, 4])
y = tf.constant(2)
tf.math.equal(x, y) ==> array([True, False])
x = tf.constant([2, 4])
y = tf.constant([2, 4])
tf.math.equal(x, y) ==> array([True, True])
```
Args:
x: A `Tensor`.
y: A `Tensor`. Must have the same type as `x`.
incompatible_shape_error: An optional `bool`. Defaults to `True`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Equal", name, x, y, "incompatible_shape_error",
incompatible_shape_error)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return equal_eager_fallback(
x, y, incompatible_shape_error=incompatible_shape_error, name=name,
ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if incompatible_shape_error is None:
incompatible_shape_error = True
incompatible_shape_error = _execute.make_bool(incompatible_shape_error, "incompatible_shape_error")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Equal", x=x, y=y, incompatible_shape_error=incompatible_shape_error,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "incompatible_shape_error",
_op._get_attr_bool("incompatible_shape_error"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Equal", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Equal = tf_export("raw_ops.Equal")(_ops.to_raw_op(equal))
def equal_eager_fallback(x, y, incompatible_shape_error, name, ctx):
if incompatible_shape_error is None:
incompatible_shape_error = True
incompatible_shape_error = _execute.make_bool(incompatible_shape_error, "incompatible_shape_error")
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T, "incompatible_shape_error",
incompatible_shape_error)
_result = _execute.execute(b"Equal", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Equal", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.erf', v1=['math.erf', 'erf'])
@deprecated_endpoints('erf')
def erf(x, name=None):
r"""Computes the [Gauss error function](https://en.wikipedia.org/wiki/Error_function) of `x` element-wise. In statistics, for non-negative values of $x$, the error function has the following interpretation: for a random variable $Y$ that is normally distributed with mean 0 and variance $1/\sqrt{2}$, $erf(x)$ is the probability that $Y$ falls in the range $[−x, x]$.
For example:
>>> tf.math.erf([[1.0, 2.0, 3.0], [0.0, -1.0, -2.0]])
<tf.Tensor: shape=(2, 3), dtype=float32, numpy=
array([[ 0.8427007, 0.9953223, 0.999978 ],
[ 0. , -0.8427007, -0.9953223]], dtype=float32)>
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Erf", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_erf(
(x, name,), None)
if _result is not NotImplemented:
return _result
return erf_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
erf, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_erf(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Erf", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
erf, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Erf", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Erf = tf_export("raw_ops.Erf")(_ops.to_raw_op(erf))
_dispatcher_for_erf = erf._tf_type_based_dispatcher.Dispatch
def erf_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Erf", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Erf", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.erfc', v1=['math.erfc', 'erfc'])
@deprecated_endpoints('erfc')
def erfc(x, name=None):
r"""Computes the complementary error function of `x` element-wise.
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Erfc", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_erfc(
(x, name,), None)
if _result is not NotImplemented:
return _result
return erfc_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
erfc, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_erfc(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Erfc", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
erfc, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Erfc", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Erfc = tf_export("raw_ops.Erfc")(_ops.to_raw_op(erfc))
_dispatcher_for_erfc = erfc._tf_type_based_dispatcher.Dispatch
def erfc_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Erfc", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Erfc", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def erfinv(x, name=None):
r"""TODO: add doc.
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Erfinv", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return erfinv_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Erfinv", x=x, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Erfinv", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Erfinv = tf_export("raw_ops.Erfinv")(_ops.to_raw_op(erfinv))
def erfinv_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Erfinv", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Erfinv", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def euclidean_norm(input, axis, keep_dims=False, name=None):
r"""Computes the euclidean norm of elements across dimensions of a tensor.
Reduces `input` along the dimensions given in `axis`. Unless
`keep_dims` is true, the rank of the tensor is reduced by 1 for each entry in
`axis`. If `keep_dims` is true, the reduced dimensions are
retained with length 1.
Args:
input: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
The tensor to reduce.
axis: A `Tensor`. Must be one of the following types: `int32`, `int64`.
The dimensions to reduce. Must be in the range
`[-rank(input), rank(input))`.
keep_dims: An optional `bool`. Defaults to `False`.
If true, retain reduced dimensions with length 1.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `input`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "EuclideanNorm", name, input, axis, "keep_dims", keep_dims)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return euclidean_norm_eager_fallback(
input, axis, keep_dims=keep_dims, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"EuclideanNorm", input=input, reduction_indices=axis,
keep_dims=keep_dims, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("keep_dims", _op._get_attr_bool("keep_dims"), "T",
_op._get_attr_type("T"), "Tidx", _op._get_attr_type("Tidx"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"EuclideanNorm", _inputs_flat, _attrs, _result)
_result, = _result
return _result
EuclideanNorm = tf_export("raw_ops.EuclideanNorm")(_ops.to_raw_op(euclidean_norm))
def euclidean_norm_eager_fallback(input, axis, keep_dims, name, ctx):
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tidx, (axis,) = _execute.args_to_matching_eager([axis], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [input, axis]
_attrs = ("keep_dims", keep_dims, "T", _attr_T, "Tidx", _attr_Tidx)
_result = _execute.execute(b"EuclideanNorm", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"EuclideanNorm", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def exp(x, name=None):
r"""Computes exponential of x element-wise. \\(y = e^x\\).
This function computes the exponential of every element in the input tensor.
i.e. `exp(x)` or `e^(x)`, where `x` is the input tensor.
`e` denotes Euler's number and is approximately equal to 2.718281.
Output is positive for any real input.
```python
x = tf.constant(2.0)
tf.math.exp(x) ==> 7.389056
x = tf.constant([2.0, 8.0])
tf.math.exp(x) ==> array([7.389056, 2980.958], dtype=float32)
```
For complex numbers, the exponential value is calculated as follows:
```
e^(x+iy) = e^x * e^iy = e^x * (cos y + i sin y)
```
Let's consider complex number 1+1j as an example.
e^1 * (cos 1 + i sin 1) = 2.7182818284590 * (0.54030230586+0.8414709848j)
```python
x = tf.constant(1 + 1j)
tf.math.exp(x) ==> 1.4686939399158851+2.2873552871788423j
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Exp", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return exp_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Exp", x=x, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Exp", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Exp = tf_export("raw_ops.Exp")(_ops.to_raw_op(exp))
def exp_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Exp", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Exp", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.expm1', v1=['math.expm1', 'expm1'])
@deprecated_endpoints('expm1')
def expm1(x, name=None):
r"""Computes `exp(x) - 1` element-wise.
i.e. `exp(x) - 1` or `e^(x) - 1`, where `x` is the input tensor.
`e` denotes Euler's number and is approximately equal to 2.718281.
```python
x = tf.constant(2.0)
tf.math.expm1(x) ==> 6.389056
x = tf.constant([2.0, 8.0])
tf.math.expm1(x) ==> array([6.389056, 2979.958], dtype=float32)
x = tf.constant(1 + 1j)
tf.math.expm1(x) ==> (0.46869393991588515+2.2873552871788423j)
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Expm1", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_expm1(
(x, name,), None)
if _result is not NotImplemented:
return _result
return expm1_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
expm1, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_expm1(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Expm1", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
expm1, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Expm1", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Expm1 = tf_export("raw_ops.Expm1")(_ops.to_raw_op(expm1))
_dispatcher_for_expm1 = expm1._tf_type_based_dispatcher.Dispatch
def expm1_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Expm1", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Expm1", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def floor(x, name=None):
r"""Returns element-wise largest integer not greater than x.
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Floor", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return floor_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Floor", x=x, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Floor", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Floor = tf_export("raw_ops.Floor")(_ops.to_raw_op(floor))
def floor_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Floor", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Floor", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export(v1=['floor_div'])
@deprecated_endpoints('floor_div')
def floor_div(x, y, name=None):
r"""Returns x // y element-wise.
*NOTE*: `floor_div` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `uint8`, `int8`, `uint16`, `int16`, `int32`, `uint32`, `uint64`, `int64`, `complex64`, `complex128`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "FloorDiv", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_floor_div(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return floor_div_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
floor_div, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_floor_div(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"FloorDiv", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
floor_div, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"FloorDiv", _inputs_flat, _attrs, _result)
_result, = _result
return _result
FloorDiv = tf_export("raw_ops.FloorDiv")(_ops.to_raw_op(floor_div))
_dispatcher_for_floor_div = floor_div._tf_type_based_dispatcher.Dispatch
def floor_div_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.uint8, _dtypes.int8, _dtypes.uint16, _dtypes.int16, _dtypes.int32, _dtypes.uint32, _dtypes.uint64, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"FloorDiv", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"FloorDiv", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.floormod', 'math.mod', v1=['math.floormod', 'floormod', 'math.mod', 'mod'])
@deprecated_endpoints('floormod', 'mod')
def floor_mod(x, y, name=None):
r"""Returns element-wise remainder of division. When `x < 0` xor `y < 0` is
true, this follows Python semantics in that the result here is consistent
with a flooring divide. E.g. `floor(x / y) * y + mod(x, y) = x`.
*NOTE*: `math.floormod` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Args:
x: A `Tensor`. Must be one of the following types: `int8`, `int16`, `int32`, `int64`, `uint8`, `uint16`, `uint32`, `uint64`, `bfloat16`, `half`, `float32`, `float64`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "FloorMod", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_floor_mod(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return floor_mod_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
floor_mod, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_floor_mod(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"FloorMod", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
floor_mod, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"FloorMod", _inputs_flat, _attrs, _result)
_result, = _result
return _result
FloorMod = tf_export("raw_ops.FloorMod")(_ops.to_raw_op(floor_mod))
_dispatcher_for_floor_mod = floor_mod._tf_type_based_dispatcher.Dispatch
def floor_mod_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.uint8, _dtypes.uint16, _dtypes.uint32, _dtypes.uint64, _dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"FloorMod", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"FloorMod", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.greater', 'greater')
def greater(x, y, name=None):
r"""Returns the truth value of (x > y) element-wise.
*NOTE*: `math.greater` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Example:
```python
x = tf.constant([5, 4, 6])
y = tf.constant([5, 2, 5])
tf.math.greater(x, y) ==> [False, True, True]
x = tf.constant([5, 4, 6])
y = tf.constant([5])
tf.math.greater(x, y) ==> [False, False, True]
```
Args:
x: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `int64`, `bfloat16`, `uint16`, `half`, `uint32`, `uint64`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Greater", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_greater(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return greater_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
greater, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_greater(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Greater", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
greater, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Greater", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Greater = tf_export("raw_ops.Greater")(_ops.to_raw_op(greater))
_dispatcher_for_greater = greater._tf_type_based_dispatcher.Dispatch
def greater_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.int64, _dtypes.bfloat16, _dtypes.uint16, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Greater", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Greater", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.greater_equal', 'greater_equal')
def greater_equal(x, y, name=None):
r"""Returns the truth value of (x >= y) element-wise.
*NOTE*: `math.greater_equal` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Example:
```python
x = tf.constant([5, 4, 6, 7])
y = tf.constant([5, 2, 5, 10])
tf.math.greater_equal(x, y) ==> [True, True, True, False]
x = tf.constant([5, 4, 6, 7])
y = tf.constant([5])
tf.math.greater_equal(x, y) ==> [True, False, True, True]
```
Args:
x: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `int64`, `bfloat16`, `uint16`, `half`, `uint32`, `uint64`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "GreaterEqual", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_greater_equal(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return greater_equal_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
greater_equal, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_greater_equal(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"GreaterEqual", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
greater_equal, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"GreaterEqual", _inputs_flat, _attrs, _result)
_result, = _result
return _result
GreaterEqual = tf_export("raw_ops.GreaterEqual")(_ops.to_raw_op(greater_equal))
_dispatcher_for_greater_equal = greater_equal._tf_type_based_dispatcher.Dispatch
def greater_equal_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.int64, _dtypes.bfloat16, _dtypes.uint16, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"GreaterEqual", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"GreaterEqual", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def _histogram_fixed_width(values, value_range, nbins, dtype=_dtypes.int32, name=None):
r"""Return histogram of values.
Given the tensor `values`, this operation returns a rank 1 histogram counting
the number of entries in `values` that fall into every bin. The bins are
equal width and determined by the arguments `value_range` and `nbins`.
```python
# Bins will be: (-inf, 1), [1, 2), [2, 3), [3, 4), [4, inf)
nbins = 5
value_range = [0.0, 5.0]
new_values = [-1.0, 0.0, 1.5, 2.0, 5.0, 15]
with tf.get_default_session() as sess:
hist = tf.histogram_fixed_width(new_values, value_range, nbins=5)
variables.global_variables_initializer().run()
sess.run(hist) => [2, 1, 1, 0, 2]
```
Args:
values: A `Tensor`. Must be one of the following types: `int32`, `int64`, `float32`, `float64`.
Numeric `Tensor`.
value_range: A `Tensor`. Must have the same type as `values`.
Shape [2] `Tensor` of same `dtype` as `values`.
values <= value_range[0] will be mapped to hist[0],
values >= value_range[1] will be mapped to hist[-1].
nbins: A `Tensor` of type `int32`.
Scalar `int32 Tensor`. Number of histogram bins.
dtype: An optional `tf.DType` from: `tf.int32, tf.int64`. Defaults to `tf.int32`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `dtype`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "HistogramFixedWidth", name, values, value_range, nbins,
"dtype", dtype)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return _histogram_fixed_width_eager_fallback(
values, value_range, nbins, dtype=dtype, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if dtype is None:
dtype = _dtypes.int32
dtype = _execute.make_type(dtype, "dtype")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"HistogramFixedWidth", values=values, value_range=value_range,
nbins=nbins, dtype=dtype, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "dtype",
_op._get_attr_type("dtype"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"HistogramFixedWidth", _inputs_flat, _attrs, _result)
_result, = _result
return _result
HistogramFixedWidth = tf_export("raw_ops.HistogramFixedWidth")(_ops.to_raw_op(_histogram_fixed_width))
def _histogram_fixed_width_eager_fallback(values, value_range, nbins, dtype, name, ctx):
if dtype is None:
dtype = _dtypes.int32
dtype = _execute.make_type(dtype, "dtype")
_attr_T, _inputs_T = _execute.args_to_matching_eager([values, value_range], ctx, [_dtypes.int32, _dtypes.int64, _dtypes.float32, _dtypes.float64, ])
(values, value_range) = _inputs_T
nbins = _ops.convert_to_tensor(nbins, _dtypes.int32)
_inputs_flat = [values, value_range, nbins]
_attrs = ("T", _attr_T, "dtype", dtype)
_result = _execute.execute(b"HistogramFixedWidth", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"HistogramFixedWidth", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.igamma', v1=['math.igamma', 'igamma'])
@deprecated_endpoints('igamma')
def igamma(a, x, name=None):
r"""Compute the lower regularized incomplete Gamma function `P(a, x)`.
The lower regularized incomplete Gamma function is defined as:
\\(P(a, x) = gamma(a, x) / Gamma(a) = 1 - Q(a, x)\\)
where
\\(gamma(a, x) = \\int_{0}^{x} t^{a-1} exp(-t) dt\\)
is the lower incomplete Gamma function.
Note, above `Q(a, x)` (`Igammac`) is the upper regularized complete
Gamma function.
Args:
a: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
x: A `Tensor`. Must have the same type as `a`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `a`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Igamma", name, a, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_igamma(
(a, x, name,), None)
if _result is not NotImplemented:
return _result
return igamma_eager_fallback(
a, x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
igamma, (), dict(a=a, x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_igamma(
(a, x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Igamma", a=a, x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
igamma, (), dict(a=a, x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Igamma", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Igamma = tf_export("raw_ops.Igamma")(_ops.to_raw_op(igamma))
_dispatcher_for_igamma = igamma._tf_type_based_dispatcher.Dispatch
def igamma_eager_fallback(a, x, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([a, x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
(a, x) = _inputs_T
_inputs_flat = [a, x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Igamma", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Igamma", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def igamma_grad_a(a, x, name=None):
r"""Computes the gradient of `igamma(a, x)` wrt `a`.
Args:
a: A `Tensor`. Must be one of the following types: `float32`, `float64`.
x: A `Tensor`. Must have the same type as `a`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `a`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "IgammaGradA", name, a, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return igamma_grad_a_eager_fallback(
a, x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"IgammaGradA", a=a, x=x, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"IgammaGradA", _inputs_flat, _attrs, _result)
_result, = _result
return _result
IgammaGradA = tf_export("raw_ops.IgammaGradA")(_ops.to_raw_op(igamma_grad_a))
def igamma_grad_a_eager_fallback(a, x, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([a, x], ctx, [_dtypes.float32, _dtypes.float64, ])
(a, x) = _inputs_T
_inputs_flat = [a, x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"IgammaGradA", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"IgammaGradA", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.igammac', v1=['math.igammac', 'igammac'])
@deprecated_endpoints('igammac')
def igammac(a, x, name=None):
r"""Compute the upper regularized incomplete Gamma function `Q(a, x)`.
The upper regularized incomplete Gamma function is defined as:
\\(Q(a, x) = Gamma(a, x) / Gamma(a) = 1 - P(a, x)\\)
where
\\(Gamma(a, x) = \int_{x}^{\infty} t^{a-1} exp(-t) dt\\)
is the upper incomplete Gamma function.
Note, above `P(a, x)` (`Igamma`) is the lower regularized complete
Gamma function.
Args:
a: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
x: A `Tensor`. Must have the same type as `a`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `a`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Igammac", name, a, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_igammac(
(a, x, name,), None)
if _result is not NotImplemented:
return _result
return igammac_eager_fallback(
a, x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
igammac, (), dict(a=a, x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_igammac(
(a, x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Igammac", a=a, x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
igammac, (), dict(a=a, x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Igammac", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Igammac = tf_export("raw_ops.Igammac")(_ops.to_raw_op(igammac))
_dispatcher_for_igammac = igammac._tf_type_based_dispatcher.Dispatch
def igammac_eager_fallback(a, x, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([a, x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
(a, x) = _inputs_T
_inputs_flat = [a, x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Igammac", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Igammac", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def imag(input, Tout=_dtypes.float32, name=None):
r"""Returns the imaginary part of a complex number.
Given a tensor `input` of complex numbers, this operation returns a tensor of
type `float` that is the imaginary part of each element in `input`. All
elements in `input` must be complex numbers of the form \\(a + bj\\), where *a*
is the real part and *b* is the imaginary part returned by this operation.
For example:
```
# tensor 'input' is [-2.25 + 4.75j, 3.25 + 5.75j]
tf.imag(input) ==> [4.75, 5.75]
```
Args:
input: A `Tensor`. Must be one of the following types: `complex64`, `complex128`.
Tout: An optional `tf.DType` from: `tf.float32, tf.float64`. Defaults to `tf.float32`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `Tout`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Imag", name, input, "Tout", Tout)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return imag_eager_fallback(
input, Tout=Tout, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if Tout is None:
Tout = _dtypes.float32
Tout = _execute.make_type(Tout, "Tout")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Imag", input=input, Tout=Tout, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tout",
_op._get_attr_type("Tout"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Imag", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Imag = tf_export("raw_ops.Imag")(_ops.to_raw_op(imag))
def imag_eager_fallback(input, Tout, name, ctx):
if Tout is None:
Tout = _dtypes.float32
Tout = _execute.make_type(Tout, "Tout")
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.complex64, _dtypes.complex128, ], _dtypes.complex64)
_inputs_flat = [input]
_attrs = ("T", _attr_T, "Tout", Tout)
_result = _execute.execute(b"Imag", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Imag", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def inv(x, name=None):
r"""Computes the reciprocal of x element-wise.
I.e., \\(y = 1 / x\\).
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int8`, `int16`, `int32`, `int64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Inv", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return inv_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Inv", x=x, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Inv", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Inv = tf_export("raw_ops.Inv")(_ops.to_raw_op(inv))
def inv_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Inv", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Inv", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def inv_grad(y, dy, name=None):
r"""Computes the gradient for the inverse of `x` wrt its input.
Specifically, `grad = -dy * y*y`, where `y = 1/x`, and `dy`
is the corresponding input gradient.
Args:
y: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
dy: A `Tensor`. Must have the same type as `y`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `y`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "InvGrad", name, y, dy)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return inv_grad_eager_fallback(
y, dy, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"InvGrad", y=y, dy=dy, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"InvGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
InvGrad = tf_export("raw_ops.InvGrad")(_ops.to_raw_op(inv_grad))
def inv_grad_eager_fallback(y, dy, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([y, dy], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
(y, dy) = _inputs_T
_inputs_flat = [y, dy]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"InvGrad", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"InvGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.is_finite', v1=['math.is_finite', 'debugging.is_finite', 'is_finite'])
@deprecated_endpoints('debugging.is_finite', 'is_finite')
def is_finite(x, name=None):
r"""Returns which elements of x are finite.
@compatibility(numpy)
Equivalent to np.isfinite
@end_compatibility
Example:
```python
x = tf.constant([5.0, 4.8, 6.8, np.inf, np.nan])
tf.math.is_finite(x) ==> [True, True, True, False, False]
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "IsFinite", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_is_finite(
(x, name,), None)
if _result is not NotImplemented:
return _result
return is_finite_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
is_finite, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_is_finite(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"IsFinite", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
is_finite, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"IsFinite", _inputs_flat, _attrs, _result)
_result, = _result
return _result
IsFinite = tf_export("raw_ops.IsFinite")(_ops.to_raw_op(is_finite))
_dispatcher_for_is_finite = is_finite._tf_type_based_dispatcher.Dispatch
def is_finite_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"IsFinite", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"IsFinite", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.is_inf', v1=['math.is_inf', 'debugging.is_inf', 'is_inf'])
@deprecated_endpoints('debugging.is_inf', 'is_inf')
def is_inf(x, name=None):
r"""Returns which elements of x are Inf.
@compatibility(numpy)
Equivalent to np.isinf
@end_compatibility
Example:
```python
x = tf.constant([5.0, np.inf, 6.8, np.inf])
tf.math.is_inf(x) ==> [False, True, False, True]
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "IsInf", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_is_inf(
(x, name,), None)
if _result is not NotImplemented:
return _result
return is_inf_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
is_inf, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_is_inf(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"IsInf", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
is_inf, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"IsInf", _inputs_flat, _attrs, _result)
_result, = _result
return _result
IsInf = tf_export("raw_ops.IsInf")(_ops.to_raw_op(is_inf))
_dispatcher_for_is_inf = is_inf._tf_type_based_dispatcher.Dispatch
def is_inf_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"IsInf", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"IsInf", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.is_nan', v1=['math.is_nan', 'debugging.is_nan', 'is_nan'])
@deprecated_endpoints('debugging.is_nan', 'is_nan')
def is_nan(x, name=None):
r"""Returns which elements of x are NaN.
@compatibility(numpy)
Equivalent to np.isnan
@end_compatibility
Example:
```python
x = tf.constant([5.0, np.nan, 6.8, np.nan, np.inf])
tf.math.is_nan(x) ==> [False, True, False, True, False]
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "IsNan", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_is_nan(
(x, name,), None)
if _result is not NotImplemented:
return _result
return is_nan_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
is_nan, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_is_nan(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"IsNan", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
is_nan, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"IsNan", _inputs_flat, _attrs, _result)
_result, = _result
return _result
IsNan = tf_export("raw_ops.IsNan")(_ops.to_raw_op(is_nan))
_dispatcher_for_is_nan = is_nan._tf_type_based_dispatcher.Dispatch
def is_nan_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"IsNan", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"IsNan", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.less', 'less')
def less(x, y, name=None):
r"""Returns the truth value of (x < y) element-wise.
*NOTE*: `math.less` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Example:
```python
x = tf.constant([5, 4, 6])
y = tf.constant([5])
tf.math.less(x, y) ==> [False, True, False]
x = tf.constant([5, 4, 6])
y = tf.constant([5, 6, 7])
tf.math.less(x, y) ==> [False, True, True]
```
Args:
x: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `int64`, `bfloat16`, `uint16`, `half`, `uint32`, `uint64`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Less", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_less(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return less_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
less, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_less(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Less", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
less, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Less", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Less = tf_export("raw_ops.Less")(_ops.to_raw_op(less))
_dispatcher_for_less = less._tf_type_based_dispatcher.Dispatch
def less_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.int64, _dtypes.bfloat16, _dtypes.uint16, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Less", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Less", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.less_equal', 'less_equal')
def less_equal(x, y, name=None):
r"""Returns the truth value of (x <= y) element-wise.
*NOTE*: `math.less_equal` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Example:
```python
x = tf.constant([5, 4, 6])
y = tf.constant([5])
tf.math.less_equal(x, y) ==> [True, True, False]
x = tf.constant([5, 4, 6])
y = tf.constant([5, 6, 6])
tf.math.less_equal(x, y) ==> [True, True, True]
```
Args:
x: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `int64`, `bfloat16`, `uint16`, `half`, `uint32`, `uint64`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "LessEqual", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_less_equal(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return less_equal_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
less_equal, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_less_equal(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"LessEqual", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
less_equal, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"LessEqual", _inputs_flat, _attrs, _result)
_result, = _result
return _result
LessEqual = tf_export("raw_ops.LessEqual")(_ops.to_raw_op(less_equal))
_dispatcher_for_less_equal = less_equal._tf_type_based_dispatcher.Dispatch
def less_equal_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.int64, _dtypes.bfloat16, _dtypes.uint16, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"LessEqual", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"LessEqual", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.lgamma', v1=['math.lgamma', 'lgamma'])
@deprecated_endpoints('lgamma')
def lgamma(x, name=None):
r"""Computes the log of the absolute value of `Gamma(x)` element-wise.
For positive numbers, this function computes log((input - 1)!) for every element in the tensor.
`lgamma(5) = log((5-1)!) = log(4!) = log(24) = 3.1780539`
Example:
```python
x = tf.constant([0, 0.5, 1, 4.5, -4, -5.6])
tf.math.lgamma(x) ==> [inf, 0.5723649, 0., 2.4537368, inf, -4.6477685]
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Lgamma", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_lgamma(
(x, name,), None)
if _result is not NotImplemented:
return _result
return lgamma_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
lgamma, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_lgamma(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Lgamma", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
lgamma, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Lgamma", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Lgamma = tf_export("raw_ops.Lgamma")(_ops.to_raw_op(lgamma))
_dispatcher_for_lgamma = lgamma._tf_type_based_dispatcher.Dispatch
def lgamma_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Lgamma", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Lgamma", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def lin_space(start, stop, num, name=None):
r"""Generates values in an interval.
A sequence of `num` evenly-spaced values are generated beginning at `start`.
If `num > 1`, the values in the sequence increase by `stop - start / num - 1`,
so that the last one is exactly `stop`.
For example:
```
tf.linspace(10.0, 12.0, 3, name="linspace") => [ 10.0 11.0 12.0]
```
Args:
start: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
0-D tensor. First entry in the range.
stop: A `Tensor`. Must have the same type as `start`.
0-D tensor. Last entry in the range.
num: A `Tensor`. Must be one of the following types: `int32`, `int64`.
0-D tensor. Number of values to generate.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `start`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "LinSpace", name, start, stop, num)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return lin_space_eager_fallback(
start, stop, num, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"LinSpace", start=start, stop=stop, num=num, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tidx",
_op._get_attr_type("Tidx"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"LinSpace", _inputs_flat, _attrs, _result)
_result, = _result
return _result
LinSpace = tf_export("raw_ops.LinSpace")(_ops.to_raw_op(lin_space))
def lin_space_eager_fallback(start, stop, num, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([start, stop], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
(start, stop) = _inputs_T
_attr_Tidx, (num,) = _execute.args_to_matching_eager([num], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [start, stop, num]
_attrs = ("T", _attr_T, "Tidx", _attr_Tidx)
_result = _execute.execute(b"LinSpace", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"LinSpace", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.log', v1=['math.log', 'log'])
@deprecated_endpoints('log')
def log(x, name=None):
r"""Computes natural logarithm of x element-wise.
I.e., \\(y = \log_e x\\).
Example:
>>> x = tf.constant([0, 0.5, 1, 5])
>>> tf.math.log(x)
<tf.Tensor: shape=(4,), dtype=float32, numpy=array([ -inf, -0.6931472, 0. , 1.609438 ], dtype=float32)>
See: https://en.wikipedia.org/wiki/Logarithm
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Log", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_log(
(x, name,), None)
if _result is not NotImplemented:
return _result
return log_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
log, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_log(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Log", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
log, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Log", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Log = tf_export("raw_ops.Log")(_ops.to_raw_op(log))
_dispatcher_for_log = log._tf_type_based_dispatcher.Dispatch
def log_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Log", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Log", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.log1p', v1=['math.log1p', 'log1p'])
@deprecated_endpoints('log1p')
def log1p(x, name=None):
r"""Computes natural logarithm of (1 + x) element-wise.
I.e., \\(y = \log_e (1 + x)\\).
Example:
>>> x = tf.constant([0, 0.5, 1, 5])
>>> tf.math.log1p(x)
<tf.Tensor: shape=(4,), dtype=float32, numpy=array([0. , 0.4054651, 0.6931472, 1.7917595], dtype=float32)>
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Log1p", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_log1p(
(x, name,), None)
if _result is not NotImplemented:
return _result
return log1p_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
log1p, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_log1p(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Log1p", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
log1p, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Log1p", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Log1p = tf_export("raw_ops.Log1p")(_ops.to_raw_op(log1p))
_dispatcher_for_log1p = log1p._tf_type_based_dispatcher.Dispatch
def log1p_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Log1p", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Log1p", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.logical_and', 'logical_and')
def logical_and(x, y, name=None):
r"""Returns the truth value of x AND y element-wise.
Logical AND function.
Requires that `x` and `y` have the same shape or have
[broadcast-compatible](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
shapes. For example, `x` and `y` can be:
- Two single elements of type `bool`.
- One `tf.Tensor` of type `bool` and one single `bool`, where the result will
be calculated by applying logical AND with the single element to each
element in the larger Tensor.
- Two `tf.Tensor` objects of type `bool` of the same shape. In this case,
the result will be the element-wise logical AND of the two input tensors.
You can also use the `&` operator instead.
Usage:
>>> a = tf.constant([True])
>>> b = tf.constant([False])
>>> tf.math.logical_and(a, b)
<tf.Tensor: shape=(1,), dtype=bool, numpy=array([False])>
>>> a & b
<tf.Tensor: shape=(1,), dtype=bool, numpy=array([False])>
>>> c = tf.constant([True])
>>> x = tf.constant([False, True, True, False])
>>> tf.math.logical_and(c, x)
<tf.Tensor: shape=(4,), dtype=bool, numpy=array([False, True, True, False])>
>>> c & x
<tf.Tensor: shape=(4,), dtype=bool, numpy=array([False, True, True, False])>
>>> y = tf.constant([False, False, True, True])
>>> z = tf.constant([False, True, False, True])
>>> tf.math.logical_and(y, z)
<tf.Tensor: shape=(4,), dtype=bool, numpy=array([False, False, False, True])>
>>> y & z
<tf.Tensor: shape=(4,), dtype=bool, numpy=array([False, False, False, True])>
This op also supports broadcasting
>>> tf.logical_and([[True, False]], [[True], [False]])
<tf.Tensor: shape=(2, 2), dtype=bool, numpy=
array([[ True, False],
[False, False]])>
The reduction version of this elementwise operation is `tf.math.reduce_all`.
Args:
x: A `tf.Tensor` of type bool.
y: A `tf.Tensor` of type bool.
name: A name for the operation (optional).
Returns:
A `tf.Tensor` of type bool with the shape that `x` and `y` broadcast to.
Args:
x: A `Tensor` of type `bool`.
y: A `Tensor` of type `bool`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "LogicalAnd", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_logical_and(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return logical_and_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
logical_and, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_logical_and(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"LogicalAnd", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
logical_and, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ()
_inputs_flat = _op.inputs
_execute.record_gradient(
"LogicalAnd", _inputs_flat, _attrs, _result)
_result, = _result
return _result
LogicalAnd = tf_export("raw_ops.LogicalAnd")(_ops.to_raw_op(logical_and))
_dispatcher_for_logical_and = logical_and._tf_type_based_dispatcher.Dispatch
def logical_and_eager_fallback(x, y, name, ctx):
x = _ops.convert_to_tensor(x, _dtypes.bool)
y = _ops.convert_to_tensor(y, _dtypes.bool)
_inputs_flat = [x, y]
_attrs = None
_result = _execute.execute(b"LogicalAnd", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"LogicalAnd", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.logical_not', 'logical_not')
def logical_not(x, name=None):
r"""Returns the truth value of `NOT x` element-wise.
Example:
>>> tf.math.logical_not(tf.constant([True, False]))
<tf.Tensor: shape=(2,), dtype=bool, numpy=array([False, True])>
Args:
x: A `Tensor` of type `bool`. A `Tensor` of type `bool`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "LogicalNot", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_logical_not(
(x, name,), None)
if _result is not NotImplemented:
return _result
return logical_not_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
logical_not, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_logical_not(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"LogicalNot", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
logical_not, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ()
_inputs_flat = _op.inputs
_execute.record_gradient(
"LogicalNot", _inputs_flat, _attrs, _result)
_result, = _result
return _result
LogicalNot = tf_export("raw_ops.LogicalNot")(_ops.to_raw_op(logical_not))
_dispatcher_for_logical_not = logical_not._tf_type_based_dispatcher.Dispatch
def logical_not_eager_fallback(x, name, ctx):
x = _ops.convert_to_tensor(x, _dtypes.bool)
_inputs_flat = [x]
_attrs = None
_result = _execute.execute(b"LogicalNot", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"LogicalNot", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.logical_or', 'logical_or')
def logical_or(x, y, name=None):
r"""Returns the truth value of x OR y element-wise.
Logical OR function.
Requires that `x` and `y` have the same shape or have
[broadcast-compatible](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
shapes. For example, `x` and `y` can be:
- Two single elements of type `bool`.
- One `tf.Tensor` of type `bool` and one single `bool`, where the result will
be calculated by applying logical OR with the single element to each
element in the larger Tensor.
- Two `tf.Tensor` objects of type `bool` of the same shape. In this case,
the result will be the element-wise logical OR of the two input tensors.
You can also use the `|` operator instead.
Usage:
>>> a = tf.constant([True])
>>> b = tf.constant([False])
>>> tf.math.logical_or(a, b)
<tf.Tensor: shape=(1,), dtype=bool, numpy=array([ True])>
>>> a | b
<tf.Tensor: shape=(1,), dtype=bool, numpy=array([ True])>
>>> c = tf.constant([False])
>>> x = tf.constant([False, True, True, False])
>>> tf.math.logical_or(c, x)
<tf.Tensor: shape=(4,), dtype=bool, numpy=array([False, True, True, False])>
>>> c | x
<tf.Tensor: shape=(4,), dtype=bool, numpy=array([False, True, True, False])>
>>> y = tf.constant([False, False, True, True])
>>> z = tf.constant([False, True, False, True])
>>> tf.math.logical_or(y, z)
<tf.Tensor: shape=(4,), dtype=bool, numpy=array([False, True, True, True])>
>>> y | z
<tf.Tensor: shape=(4,), dtype=bool, numpy=array([False, True, True, True])>
This op also supports broadcasting
>>> tf.logical_or([[True, False]], [[True], [False]])
<tf.Tensor: shape=(2, 2), dtype=bool, numpy=
array([[ True, True],
[ True, False]])>
The reduction version of this elementwise operation is `tf.math.reduce_any`.
Args:
x: A `tf.Tensor` of type bool.
y: A `tf.Tensor` of type bool.
name: A name for the operation (optional).
Returns:
A `tf.Tensor` of type bool with the shape that `x` and `y` broadcast to.
Args:
x: A `Tensor` of type `bool`.
y: A `Tensor` of type `bool`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "LogicalOr", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_logical_or(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return logical_or_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
logical_or, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_logical_or(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"LogicalOr", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
logical_or, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ()
_inputs_flat = _op.inputs
_execute.record_gradient(
"LogicalOr", _inputs_flat, _attrs, _result)
_result, = _result
return _result
LogicalOr = tf_export("raw_ops.LogicalOr")(_ops.to_raw_op(logical_or))
_dispatcher_for_logical_or = logical_or._tf_type_based_dispatcher.Dispatch
def logical_or_eager_fallback(x, y, name, ctx):
x = _ops.convert_to_tensor(x, _dtypes.bool)
y = _ops.convert_to_tensor(y, _dtypes.bool)
_inputs_flat = [x, y]
_attrs = None
_result = _execute.execute(b"LogicalOr", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"LogicalOr", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def mat_mul(a, b, transpose_a=False, transpose_b=False, name=None):
r"""Multiply the matrix "a" by the matrix "b".
The inputs must be two-dimensional matrices and the inner dimension of
"a" (after being transposed if transpose_a is true) must match the
outer dimension of "b" (after being transposed if transposed_b is
true).
*Note*: The default kernel implementation for MatMul on GPUs uses
cublas.
Args:
a: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int32`, `int64`, `complex64`, `complex128`.
b: A `Tensor`. Must have the same type as `a`.
transpose_a: An optional `bool`. Defaults to `False`.
If true, "a" is transposed before multiplication.
transpose_b: An optional `bool`. Defaults to `False`.
If true, "b" is transposed before multiplication.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `a`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "MatMul", name, a, b, "transpose_a", transpose_a, "transpose_b",
transpose_b)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return mat_mul_eager_fallback(
a, b, transpose_a=transpose_a, transpose_b=transpose_b, name=name,
ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if transpose_a is None:
transpose_a = False
transpose_a = _execute.make_bool(transpose_a, "transpose_a")
if transpose_b is None:
transpose_b = False
transpose_b = _execute.make_bool(transpose_b, "transpose_b")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"MatMul", a=a, b=b, transpose_a=transpose_a, transpose_b=transpose_b,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("transpose_a", _op._get_attr_bool("transpose_a"), "transpose_b",
_op._get_attr_bool("transpose_b"), "T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"MatMul", _inputs_flat, _attrs, _result)
_result, = _result
return _result
MatMul = tf_export("raw_ops.MatMul")(_ops.to_raw_op(mat_mul))
def mat_mul_eager_fallback(a, b, transpose_a, transpose_b, name, ctx):
if transpose_a is None:
transpose_a = False
transpose_a = _execute.make_bool(transpose_a, "transpose_a")
if transpose_b is None:
transpose_b = False
transpose_b = _execute.make_bool(transpose_b, "transpose_b")
_attr_T, _inputs_T = _execute.args_to_matching_eager([a, b], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
(a, b) = _inputs_T
_inputs_flat = [a, b]
_attrs = ("transpose_a", transpose_a, "transpose_b", transpose_b, "T",
_attr_T)
_result = _execute.execute(b"MatMul", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"MatMul", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def _max(input, axis, keep_dims=False, name=None):
r"""Computes the maximum of elements across dimensions of a tensor.
Reduces `input` along the dimensions given in `axis`. Unless
`keep_dims` is true, the rank of the tensor is reduced by 1 for each entry in
`axis`. If `keep_dims` is true, the reduced dimensions are
retained with length 1.
Args:
input: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `int64`, `bfloat16`, `uint16`, `half`, `uint32`, `uint64`, `qint8`, `quint8`, `qint32`, `qint16`, `quint16`.
The tensor to reduce.
axis: A `Tensor`. Must be one of the following types: `int32`, `int64`.
The dimensions to reduce. Must be in the range
`[-rank(input), rank(input))`.
keep_dims: An optional `bool`. Defaults to `False`.
If true, retain reduced dimensions with length 1.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `input`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Max", name, input, axis, "keep_dims", keep_dims)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return _max_eager_fallback(
input, axis, keep_dims=keep_dims, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Max", input=input, reduction_indices=axis, keep_dims=keep_dims,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("keep_dims", _op._get_attr_bool("keep_dims"), "T",
_op._get_attr_type("T"), "Tidx", _op._get_attr_type("Tidx"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Max", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Max = tf_export("raw_ops.Max")(_ops.to_raw_op(_max))
def _max_eager_fallback(input, axis, keep_dims, name, ctx):
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.int64, _dtypes.bfloat16, _dtypes.uint16, _dtypes.half, _dtypes.uint32, _dtypes.uint64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.qint16, _dtypes.quint16, ])
_attr_Tidx, (axis,) = _execute.args_to_matching_eager([axis], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [input, axis]
_attrs = ("keep_dims", keep_dims, "T", _attr_T, "Tidx", _attr_Tidx)
_result = _execute.execute(b"Max", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Max", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.maximum', 'maximum')
def maximum(x, y, name=None):
r"""Returns the max of x and y (i.e. x > y ? x : y) element-wise.
Example:
>>> x = tf.constant([0., 0., 0., 0.])
>>> y = tf.constant([-2., 0., 2., 5.])
>>> tf.math.maximum(x, y)
<tf.Tensor: shape=(4,), dtype=float32, numpy=array([0., 0., 2., 5.], dtype=float32)>
Note that `maximum` supports [broadcast semantics](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html) for `x` and `y`.
>>> x = tf.constant([-5., 0., 0., 0.])
>>> y = tf.constant([-3.])
>>> tf.math.maximum(x, y)
<tf.Tensor: shape=(4,), dtype=float32, numpy=array([-3., 0., 0., 0.], dtype=float32)>
The reduction version of this elementwise operation is `tf.math.reduce_max`
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int8`, `uint8`, `int16`, `uint16`, `int32`, `uint32`, `int64`, `uint64`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Maximum", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_maximum(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return maximum_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
maximum, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_maximum(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Maximum", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
maximum, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Maximum", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Maximum = tf_export("raw_ops.Maximum")(_ops.to_raw_op(maximum))
_dispatcher_for_maximum = maximum._tf_type_based_dispatcher.Dispatch
def maximum_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int8, _dtypes.uint8, _dtypes.int16, _dtypes.uint16, _dtypes.int32, _dtypes.uint32, _dtypes.int64, _dtypes.uint64, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Maximum", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Maximum", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def mean(input, axis, keep_dims=False, name=None):
r"""Computes the mean of elements across dimensions of a tensor.
Reduces `input` along the dimensions given in `axis`. Unless
`keep_dims` is true, the rank of the tensor is reduced by 1 for each entry in
`axis`. If `keep_dims` is true, the reduced dimensions are
retained with length 1.
Args:
input: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
The tensor to reduce.
axis: A `Tensor`. Must be one of the following types: `int32`, `int64`.
The dimensions to reduce. Must be in the range
`[-rank(input), rank(input))`.
keep_dims: An optional `bool`. Defaults to `False`.
If true, retain reduced dimensions with length 1.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `input`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Mean", name, input, axis, "keep_dims", keep_dims)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return mean_eager_fallback(
input, axis, keep_dims=keep_dims, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Mean", input=input, reduction_indices=axis, keep_dims=keep_dims,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("keep_dims", _op._get_attr_bool("keep_dims"), "T",
_op._get_attr_type("T"), "Tidx", _op._get_attr_type("Tidx"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Mean", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Mean = tf_export("raw_ops.Mean")(_ops.to_raw_op(mean))
def mean_eager_fallback(input, axis, keep_dims, name, ctx):
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tidx, (axis,) = _execute.args_to_matching_eager([axis], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [input, axis]
_attrs = ("keep_dims", keep_dims, "T", _attr_T, "Tidx", _attr_Tidx)
_result = _execute.execute(b"Mean", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Mean", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def _min(input, axis, keep_dims=False, name=None):
r"""Computes the minimum of elements across dimensions of a tensor.
Reduces `input` along the dimensions given in `axis`. Unless
`keep_dims` is true, the rank of the tensor is reduced by 1 for each entry in
`axis`. If `keep_dims` is true, the reduced dimensions are
retained with length 1.
Args:
input: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `int64`, `bfloat16`, `uint16`, `half`, `uint32`, `uint64`, `qint8`, `quint8`, `qint32`, `qint16`, `quint16`.
The tensor to reduce.
axis: A `Tensor`. Must be one of the following types: `int32`, `int64`.
The dimensions to reduce. Must be in the range
`[-rank(input), rank(input))`.
keep_dims: An optional `bool`. Defaults to `False`.
If true, retain reduced dimensions with length 1.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `input`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Min", name, input, axis, "keep_dims", keep_dims)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return _min_eager_fallback(
input, axis, keep_dims=keep_dims, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Min", input=input, reduction_indices=axis, keep_dims=keep_dims,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("keep_dims", _op._get_attr_bool("keep_dims"), "T",
_op._get_attr_type("T"), "Tidx", _op._get_attr_type("Tidx"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Min", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Min = tf_export("raw_ops.Min")(_ops.to_raw_op(_min))
def _min_eager_fallback(input, axis, keep_dims, name, ctx):
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.int64, _dtypes.bfloat16, _dtypes.uint16, _dtypes.half, _dtypes.uint32, _dtypes.uint64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.qint16, _dtypes.quint16, ])
_attr_Tidx, (axis,) = _execute.args_to_matching_eager([axis], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [input, axis]
_attrs = ("keep_dims", keep_dims, "T", _attr_T, "Tidx", _attr_Tidx)
_result = _execute.execute(b"Min", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Min", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.minimum', 'minimum')
def minimum(x, y, name=None):
r"""Returns the min of x and y (i.e. x < y ? x : y) element-wise.
Both inputs are number-type tensors (except complex). `minimum` expects that
both tensors have the same `dtype`.
Examples:
>>> x = tf.constant([0., 0., 0., 0.])
>>> y = tf.constant([-5., -2., 0., 3.])
>>> tf.math.minimum(x, y)
<tf.Tensor: shape=(4,), dtype=float32, numpy=array([-5., -2., 0., 0.], dtype=float32)>
Note that `minimum` supports [broadcast semantics](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html) for `x` and `y`.
>>> x = tf.constant([-5., 0., 0., 0.])
>>> y = tf.constant([-3.])
>>> tf.math.minimum(x, y)
<tf.Tensor: shape=(4,), dtype=float32, numpy=array([-5., -3., -3., -3.], dtype=float32)>
The reduction version of this elementwise operation is `tf.math.reduce_min`
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int8`, `uint8`, `int16`, `uint16`, `int32`, `uint32`, `int64`, `uint64`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Minimum", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_minimum(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return minimum_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
minimum, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_minimum(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Minimum", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
minimum, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Minimum", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Minimum = tf_export("raw_ops.Minimum")(_ops.to_raw_op(minimum))
_dispatcher_for_minimum = minimum._tf_type_based_dispatcher.Dispatch
def minimum_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int8, _dtypes.uint8, _dtypes.int16, _dtypes.uint16, _dtypes.int32, _dtypes.uint32, _dtypes.int64, _dtypes.uint64, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Minimum", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Minimum", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def mod(x, y, name=None):
r"""Returns element-wise remainder of division. This emulates C semantics in that
the result here is consistent with a truncating divide. E.g.
`tf.truncatediv(x, y) * y + truncate_mod(x, y) = x`.
*NOTE*: `Mod` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Args:
x: A `Tensor`. Must be one of the following types: `int32`, `int64`, `half`, `half`, `bfloat16`, `float32`, `float64`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Mod", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return mod_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Mod", x=x, y=y, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Mod", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Mod = tf_export("raw_ops.Mod")(_ops.to_raw_op(mod))
def mod_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.int32, _dtypes.int64, _dtypes.half, _dtypes.half, _dtypes.bfloat16, _dtypes.float32, _dtypes.float64, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Mod", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Mod", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def mul(x, y, name=None):
r"""Returns x * y element-wise.
*NOTE*: `Multiply` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `uint8`, `int8`, `uint16`, `int16`, `int32`, `uint32`, `uint64`, `int64`, `complex64`, `complex128`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Mul", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return mul_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Mul", x=x, y=y, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Mul", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Mul = tf_export("raw_ops.Mul")(_ops.to_raw_op(mul))
def mul_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.uint8, _dtypes.int8, _dtypes.uint16, _dtypes.int16, _dtypes.int32, _dtypes.uint32, _dtypes.uint64, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Mul", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Mul", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def mul_no_nan(x, y, name=None):
r"""Returns x * y element-wise. Returns zero if y is zero, even if x if infinite or NaN.
*NOTE*: `MulNoNan` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "MulNoNan", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return mul_no_nan_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"MulNoNan", x=x, y=y, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"MulNoNan", _inputs_flat, _attrs, _result)
_result, = _result
return _result
MulNoNan = tf_export("raw_ops.MulNoNan")(_ops.to_raw_op(mul_no_nan))
def mul_no_nan_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"MulNoNan", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"MulNoNan", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def ndtri(x, name=None):
r"""TODO: add doc.
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Ndtri", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return ndtri_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Ndtri", x=x, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Ndtri", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Ndtri = tf_export("raw_ops.Ndtri")(_ops.to_raw_op(ndtri))
def ndtri_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Ndtri", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Ndtri", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.negative', 'negative')
def neg(x, name=None):
r"""Computes numerical negative value element-wise.
I.e., \\(y = -x\\).
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int8`, `int16`, `int32`, `int64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Neg", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_neg(
(x, name,), None)
if _result is not NotImplemented:
return _result
return neg_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
neg, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_neg(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Neg", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
neg, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Neg", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Neg = tf_export("raw_ops.Neg")(_ops.to_raw_op(neg))
_dispatcher_for_neg = neg._tf_type_based_dispatcher.Dispatch
def neg_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Neg", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Neg", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.nextafter')
def next_after(x1, x2, name=None):
r"""Returns the next representable value of `x1` in the direction of `x2`, element-wise.
This operation returns the same result as the C++ std::nextafter function.
It can also return a subnormal number.
@compatibility(cpp)
Equivalent to C++ std::nextafter function.
@end_compatibility
Args:
x1: A `Tensor`. Must be one of the following types: `float64`, `float32`.
x2: A `Tensor`. Must have the same type as `x1`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x1`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "NextAfter", name, x1, x2)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_next_after(
(x1, x2, name,), None)
if _result is not NotImplemented:
return _result
return next_after_eager_fallback(
x1, x2, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
next_after, (), dict(x1=x1, x2=x2, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_next_after(
(x1, x2, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"NextAfter", x1=x1, x2=x2, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
next_after, (), dict(x1=x1, x2=x2, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"NextAfter", _inputs_flat, _attrs, _result)
_result, = _result
return _result
NextAfter = tf_export("raw_ops.NextAfter")(_ops.to_raw_op(next_after))
_dispatcher_for_next_after = next_after._tf_type_based_dispatcher.Dispatch
def next_after_eager_fallback(x1, x2, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x1, x2], ctx, [_dtypes.float64, _dtypes.float32, ], _dtypes.float32)
(x1, x2) = _inputs_T
_inputs_flat = [x1, x2]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"NextAfter", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"NextAfter", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def not_equal(x, y, incompatible_shape_error=True, name=None):
r"""Returns the truth value of (x != y) element-wise.
*NOTE*: `NotEqual` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Args:
x: A `Tensor`.
y: A `Tensor`. Must have the same type as `x`.
incompatible_shape_error: An optional `bool`. Defaults to `True`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `bool`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "NotEqual", name, x, y, "incompatible_shape_error",
incompatible_shape_error)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return not_equal_eager_fallback(
x, y, incompatible_shape_error=incompatible_shape_error, name=name,
ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if incompatible_shape_error is None:
incompatible_shape_error = True
incompatible_shape_error = _execute.make_bool(incompatible_shape_error, "incompatible_shape_error")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"NotEqual", x=x, y=y,
incompatible_shape_error=incompatible_shape_error,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "incompatible_shape_error",
_op._get_attr_bool("incompatible_shape_error"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"NotEqual", _inputs_flat, _attrs, _result)
_result, = _result
return _result
NotEqual = tf_export("raw_ops.NotEqual")(_ops.to_raw_op(not_equal))
def not_equal_eager_fallback(x, y, incompatible_shape_error, name, ctx):
if incompatible_shape_error is None:
incompatible_shape_error = True
incompatible_shape_error = _execute.make_bool(incompatible_shape_error, "incompatible_shape_error")
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T, "incompatible_shape_error",
incompatible_shape_error)
_result = _execute.execute(b"NotEqual", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"NotEqual", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.polygamma', v1=['math.polygamma', 'polygamma'])
@deprecated_endpoints('polygamma')
def polygamma(a, x, name=None):
r"""Compute the polygamma function \\(\psi^{(n)}(x)\\).
The polygamma function is defined as:
\\(\psi^{(a)}(x) = \frac{d^a}{dx^a} \psi(x)\\)
where \\(\psi(x)\\) is the digamma function.
The polygamma function is defined only for non-negative integer orders \\a\\.
Args:
a: A `Tensor`. Must be one of the following types: `float32`, `float64`.
x: A `Tensor`. Must have the same type as `a`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `a`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Polygamma", name, a, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_polygamma(
(a, x, name,), None)
if _result is not NotImplemented:
return _result
return polygamma_eager_fallback(
a, x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
polygamma, (), dict(a=a, x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_polygamma(
(a, x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Polygamma", a=a, x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
polygamma, (), dict(a=a, x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Polygamma", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Polygamma = tf_export("raw_ops.Polygamma")(_ops.to_raw_op(polygamma))
_dispatcher_for_polygamma = polygamma._tf_type_based_dispatcher.Dispatch
def polygamma_eager_fallback(a, x, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([a, x], ctx, [_dtypes.float32, _dtypes.float64, ])
(a, x) = _inputs_T
_inputs_flat = [a, x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Polygamma", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Polygamma", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def _pow(x, y, name=None):
r"""Computes the power of one value to another.
Given a tensor `x` and a tensor `y`, this operation computes \\(x^y\\) for
corresponding elements in `x` and `y`. For example:
```
# tensor 'x' is [[2, 2]], [3, 3]]
# tensor 'y' is [[8, 16], [2, 3]]
tf.pow(x, y) ==> [[256, 65536], [9, 27]]
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `float32`, `half`, `float64`, `int8`, `int16`, `int32`, `int64`, `complex64`, `complex128`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Pow", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return _pow_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Pow", x=x, y=y, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Pow", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Pow = tf_export("raw_ops.Pow")(_ops.to_raw_op(_pow))
def _pow_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.float32, _dtypes.half, _dtypes.float64, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Pow", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Pow", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def prod(input, axis, keep_dims=False, name=None):
r"""Computes the product of elements across dimensions of a tensor.
Reduces `input` along the dimensions given in `axis`. Unless
`keep_dims` is true, the rank of the tensor is reduced by 1 for each entry in
`axis`. If `keep_dims` is true, the reduced dimensions are
retained with length 1.
Args:
input: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
The tensor to reduce.
axis: A `Tensor`. Must be one of the following types: `int32`, `int64`.
The dimensions to reduce. Must be in the range
`[-rank(input), rank(input))`.
keep_dims: An optional `bool`. Defaults to `False`.
If true, retain reduced dimensions with length 1.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `input`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Prod", name, input, axis, "keep_dims", keep_dims)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return prod_eager_fallback(
input, axis, keep_dims=keep_dims, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Prod", input=input, reduction_indices=axis, keep_dims=keep_dims,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("keep_dims", _op._get_attr_bool("keep_dims"), "T",
_op._get_attr_type("T"), "Tidx", _op._get_attr_type("Tidx"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Prod", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Prod = tf_export("raw_ops.Prod")(_ops.to_raw_op(prod))
def prod_eager_fallback(input, axis, keep_dims, name, ctx):
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tidx, (axis,) = _execute.args_to_matching_eager([axis], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [input, axis]
_attrs = ("keep_dims", keep_dims, "T", _attr_T, "Tidx", _attr_Tidx)
_result = _execute.execute(b"Prod", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Prod", _inputs_flat, _attrs, _result)
_result, = _result
return _result
_QuantizeDownAndShrinkRangeOutput = collections.namedtuple(
"QuantizeDownAndShrinkRange",
["output", "output_min", "output_max"])
def quantize_down_and_shrink_range(input, input_min, input_max, out_type, name=None):
r"""Convert the quantized 'input' tensor into a lower-precision 'output', using the
actual distribution of the values to maximize the usage of the lower bit depth
and adjusting the output min and max ranges accordingly.
[input_min, input_max] are scalar floats that specify the range for the float
interpretation of the 'input' data. For example, if input_min is -1.0f and
input_max is 1.0f, and we are dealing with quint16 quantized data, then a 0
value in the 16-bit data should be interpreted as -1.0f, and a 65535 means 1.0f.
This operator tries to squeeze as much precision as possible into an output with
a lower bit depth by calculating the actual min and max values found in the
data. For example, maybe that quint16 input has no values lower than 16,384 and
none higher than 49,152. That means only half the range is actually needed, all
the float interpretations are between -0.5f and 0.5f, so if we want to compress
the data into a quint8 output, we can use that range rather than the theoretical
-1.0f to 1.0f that is suggested by the input min and max.
In practice, this is most useful for taking output from operations like
QuantizedMatMul that can produce higher bit-depth outputs than their inputs and
may have large potential output ranges, but in practice have a distribution of
input values that only uses a small fraction of the possible range. By feeding
that output into this operator, we can reduce it from 32 bits down to 8 with
minimal loss of accuracy.
Args:
input: A `Tensor`. Must be one of the following types: `qint8`, `quint8`, `qint32`, `qint16`, `quint16`.
input_min: A `Tensor` of type `float32`.
The float value that the minimum quantized input value represents.
input_max: A `Tensor` of type `float32`.
The float value that the maximum quantized input value represents.
out_type: A `tf.DType` from: `tf.qint8, tf.quint8, tf.qint32, tf.qint16, tf.quint16`.
The type of the output. Should be a lower bit depth than Tinput.
name: A name for the operation (optional).
Returns:
A tuple of `Tensor` objects (output, output_min, output_max).
output: A `Tensor` of type `out_type`.
output_min: A `Tensor` of type `float32`.
output_max: A `Tensor` of type `float32`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "QuantizeDownAndShrinkRange", name, input, input_min, input_max,
"out_type", out_type)
_result = _QuantizeDownAndShrinkRangeOutput._make(_result)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return quantize_down_and_shrink_range_eager_fallback(
input, input_min, input_max, out_type=out_type, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
out_type = _execute.make_type(out_type, "out_type")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"QuantizeDownAndShrinkRange", input=input, input_min=input_min,
input_max=input_max, out_type=out_type,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("Tinput", _op._get_attr_type("Tinput"), "out_type",
_op._get_attr_type("out_type"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"QuantizeDownAndShrinkRange", _inputs_flat, _attrs, _result)
_result = _QuantizeDownAndShrinkRangeOutput._make(_result)
return _result
QuantizeDownAndShrinkRange = tf_export("raw_ops.QuantizeDownAndShrinkRange")(_ops.to_raw_op(quantize_down_and_shrink_range))
def quantize_down_and_shrink_range_eager_fallback(input, input_min, input_max, out_type, name, ctx):
out_type = _execute.make_type(out_type, "out_type")
_attr_Tinput, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.qint16, _dtypes.quint16, ])
input_min = _ops.convert_to_tensor(input_min, _dtypes.float32)
input_max = _ops.convert_to_tensor(input_max, _dtypes.float32)
_inputs_flat = [input, input_min, input_max]
_attrs = ("Tinput", _attr_Tinput, "out_type", out_type)
_result = _execute.execute(b"QuantizeDownAndShrinkRange", 3,
inputs=_inputs_flat, attrs=_attrs, ctx=ctx,
name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"QuantizeDownAndShrinkRange", _inputs_flat, _attrs, _result)
_result = _QuantizeDownAndShrinkRangeOutput._make(_result)
return _result
_QuantizedAddOutput = collections.namedtuple(
"QuantizedAdd",
["z", "min_z", "max_z"])
def quantized_add(x, y, min_x, max_x, min_y, max_y, Toutput=_dtypes.qint32, name=None):
r"""Returns x + y element-wise, working on quantized buffers.
Args:
x: A `Tensor`. Must be one of the following types: `qint8`, `quint8`, `qint32`, `qint16`, `quint16`.
y: A `Tensor`. Must be one of the following types: `qint8`, `quint8`, `qint32`, `qint16`, `quint16`.
min_x: A `Tensor` of type `float32`.
The float value that the lowest quantized `x` value represents.
max_x: A `Tensor` of type `float32`.
The float value that the highest quantized `x` value represents.
min_y: A `Tensor` of type `float32`.
The float value that the lowest quantized `y` value represents.
max_y: A `Tensor` of type `float32`.
The float value that the highest quantized `y` value represents.
Toutput: An optional `tf.DType` from: `tf.qint8, tf.quint8, tf.qint32, tf.qint16, tf.quint16`. Defaults to `tf.qint32`.
name: A name for the operation (optional).
Returns:
A tuple of `Tensor` objects (z, min_z, max_z).
z: A `Tensor` of type `Toutput`.
min_z: A `Tensor` of type `float32`.
max_z: A `Tensor` of type `float32`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "QuantizedAdd", name, x, y, min_x, max_x, min_y, max_y,
"Toutput", Toutput)
_result = _QuantizedAddOutput._make(_result)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return quantized_add_eager_fallback(
x, y, min_x, max_x, min_y, max_y, Toutput=Toutput, name=name,
ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if Toutput is None:
Toutput = _dtypes.qint32
Toutput = _execute.make_type(Toutput, "Toutput")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"QuantizedAdd", x=x, y=y, min_x=min_x, max_x=max_x, min_y=min_y,
max_y=max_y, Toutput=Toutput, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T1", _op._get_attr_type("T1"), "T2", _op._get_attr_type("T2"),
"Toutput", _op._get_attr_type("Toutput"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"QuantizedAdd", _inputs_flat, _attrs, _result)
_result = _QuantizedAddOutput._make(_result)
return _result
QuantizedAdd = tf_export("raw_ops.QuantizedAdd")(_ops.to_raw_op(quantized_add))
def quantized_add_eager_fallback(x, y, min_x, max_x, min_y, max_y, Toutput, name, ctx):
if Toutput is None:
Toutput = _dtypes.qint32
Toutput = _execute.make_type(Toutput, "Toutput")
_attr_T1, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.qint16, _dtypes.quint16, ])
_attr_T2, (y,) = _execute.args_to_matching_eager([y], ctx, [_dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.qint16, _dtypes.quint16, ])
min_x = _ops.convert_to_tensor(min_x, _dtypes.float32)
max_x = _ops.convert_to_tensor(max_x, _dtypes.float32)
min_y = _ops.convert_to_tensor(min_y, _dtypes.float32)
max_y = _ops.convert_to_tensor(max_y, _dtypes.float32)
_inputs_flat = [x, y, min_x, max_x, min_y, max_y]
_attrs = ("T1", _attr_T1, "T2", _attr_T2, "Toutput", Toutput)
_result = _execute.execute(b"QuantizedAdd", 3, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"QuantizedAdd", _inputs_flat, _attrs, _result)
_result = _QuantizedAddOutput._make(_result)
return _result
_QuantizedMatMulOutput = collections.namedtuple(
"QuantizedMatMul",
["out", "min_out", "max_out"])
def quantized_mat_mul(a, b, min_a, max_a, min_b, max_b, Toutput=_dtypes.qint32, transpose_a=False, transpose_b=False, Tactivation=_dtypes.quint8, name=None):
r"""Perform a quantized matrix multiplication of `a` by the matrix `b`.
The inputs must be two-dimensional matrices and the inner dimension of
`a` (after being transposed if `transpose_a` is non-zero) must match the
outer dimension of `b` (after being transposed if `transposed_b` is
non-zero).
Args:
a: A `Tensor`. Must be one of the following types: `qint8`, `quint8`, `qint32`, `qint16`, `quint16`.
Must be a two-dimensional tensor.
b: A `Tensor`. Must be one of the following types: `qint8`, `quint8`, `qint32`, `qint16`, `quint16`.
Must be a two-dimensional tensor.
min_a: A `Tensor` of type `float32`.
The float value that the lowest quantized `a` value represents.
max_a: A `Tensor` of type `float32`.
The float value that the highest quantized `a` value represents.
min_b: A `Tensor` of type `float32`.
The float value that the lowest quantized `b` value represents.
max_b: A `Tensor` of type `float32`.
The float value that the highest quantized `b` value represents.
Toutput: An optional `tf.DType` from: `tf.qint8, tf.quint8, tf.qint32, tf.qint16, tf.quint16`. Defaults to `tf.qint32`.
transpose_a: An optional `bool`. Defaults to `False`.
If true, `a` is transposed before multiplication.
transpose_b: An optional `bool`. Defaults to `False`.
If true, `b` is transposed before multiplication.
Tactivation: An optional `tf.DType` from: `tf.qint8, tf.quint8, tf.qint32, tf.qint16, tf.quint16`. Defaults to `tf.quint8`.
The type of output produced by activation function
following this operation.
name: A name for the operation (optional).
Returns:
A tuple of `Tensor` objects (out, min_out, max_out).
out: A `Tensor` of type `Toutput`.
min_out: A `Tensor` of type `float32`.
max_out: A `Tensor` of type `float32`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "QuantizedMatMul", name, a, b, min_a, max_a, min_b, max_b,
"Toutput", Toutput, "transpose_a", transpose_a, "transpose_b",
transpose_b, "Tactivation", Tactivation)
_result = _QuantizedMatMulOutput._make(_result)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return quantized_mat_mul_eager_fallback(
a, b, min_a, max_a, min_b, max_b, Toutput=Toutput,
transpose_a=transpose_a, transpose_b=transpose_b,
Tactivation=Tactivation, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if Toutput is None:
Toutput = _dtypes.qint32
Toutput = _execute.make_type(Toutput, "Toutput")
if transpose_a is None:
transpose_a = False
transpose_a = _execute.make_bool(transpose_a, "transpose_a")
if transpose_b is None:
transpose_b = False
transpose_b = _execute.make_bool(transpose_b, "transpose_b")
if Tactivation is None:
Tactivation = _dtypes.quint8
Tactivation = _execute.make_type(Tactivation, "Tactivation")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"QuantizedMatMul", a=a, b=b, min_a=min_a, max_a=max_a, min_b=min_b,
max_b=max_b, Toutput=Toutput,
transpose_a=transpose_a, transpose_b=transpose_b,
Tactivation=Tactivation, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T1", _op._get_attr_type("T1"), "T2", _op._get_attr_type("T2"),
"Toutput", _op._get_attr_type("Toutput"), "transpose_a",
_op._get_attr_bool("transpose_a"), "transpose_b",
_op._get_attr_bool("transpose_b"), "Tactivation",
_op._get_attr_type("Tactivation"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"QuantizedMatMul", _inputs_flat, _attrs, _result)
_result = _QuantizedMatMulOutput._make(_result)
return _result
QuantizedMatMul = tf_export("raw_ops.QuantizedMatMul")(_ops.to_raw_op(quantized_mat_mul))
def quantized_mat_mul_eager_fallback(a, b, min_a, max_a, min_b, max_b, Toutput, transpose_a, transpose_b, Tactivation, name, ctx):
if Toutput is None:
Toutput = _dtypes.qint32
Toutput = _execute.make_type(Toutput, "Toutput")
if transpose_a is None:
transpose_a = False
transpose_a = _execute.make_bool(transpose_a, "transpose_a")
if transpose_b is None:
transpose_b = False
transpose_b = _execute.make_bool(transpose_b, "transpose_b")
if Tactivation is None:
Tactivation = _dtypes.quint8
Tactivation = _execute.make_type(Tactivation, "Tactivation")
_attr_T1, (a,) = _execute.args_to_matching_eager([a], ctx, [_dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.qint16, _dtypes.quint16, ])
_attr_T2, (b,) = _execute.args_to_matching_eager([b], ctx, [_dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.qint16, _dtypes.quint16, ])
min_a = _ops.convert_to_tensor(min_a, _dtypes.float32)
max_a = _ops.convert_to_tensor(max_a, _dtypes.float32)
min_b = _ops.convert_to_tensor(min_b, _dtypes.float32)
max_b = _ops.convert_to_tensor(max_b, _dtypes.float32)
_inputs_flat = [a, b, min_a, max_a, min_b, max_b]
_attrs = ("T1", _attr_T1, "T2", _attr_T2, "Toutput", Toutput, "transpose_a",
transpose_a, "transpose_b", transpose_b, "Tactivation", Tactivation)
_result = _execute.execute(b"QuantizedMatMul", 3, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"QuantizedMatMul", _inputs_flat, _attrs, _result)
_result = _QuantizedMatMulOutput._make(_result)
return _result
_QuantizedMulOutput = collections.namedtuple(
"QuantizedMul",
["z", "min_z", "max_z"])
def quantized_mul(x, y, min_x, max_x, min_y, max_y, Toutput=_dtypes.qint32, name=None):
r"""Returns x * y element-wise, working on quantized buffers.
Args:
x: A `Tensor`. Must be one of the following types: `qint8`, `quint8`, `qint32`, `qint16`, `quint16`.
y: A `Tensor`. Must be one of the following types: `qint8`, `quint8`, `qint32`, `qint16`, `quint16`.
min_x: A `Tensor` of type `float32`.
The float value that the lowest quantized `x` value represents.
max_x: A `Tensor` of type `float32`.
The float value that the highest quantized `x` value represents.
min_y: A `Tensor` of type `float32`.
The float value that the lowest quantized `y` value represents.
max_y: A `Tensor` of type `float32`.
The float value that the highest quantized `y` value represents.
Toutput: An optional `tf.DType` from: `tf.qint8, tf.quint8, tf.qint32, tf.qint16, tf.quint16`. Defaults to `tf.qint32`.
name: A name for the operation (optional).
Returns:
A tuple of `Tensor` objects (z, min_z, max_z).
z: A `Tensor` of type `Toutput`.
min_z: A `Tensor` of type `float32`.
max_z: A `Tensor` of type `float32`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "QuantizedMul", name, x, y, min_x, max_x, min_y, max_y,
"Toutput", Toutput)
_result = _QuantizedMulOutput._make(_result)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return quantized_mul_eager_fallback(
x, y, min_x, max_x, min_y, max_y, Toutput=Toutput, name=name,
ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if Toutput is None:
Toutput = _dtypes.qint32
Toutput = _execute.make_type(Toutput, "Toutput")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"QuantizedMul", x=x, y=y, min_x=min_x, max_x=max_x, min_y=min_y,
max_y=max_y, Toutput=Toutput, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T1", _op._get_attr_type("T1"), "T2", _op._get_attr_type("T2"),
"Toutput", _op._get_attr_type("Toutput"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"QuantizedMul", _inputs_flat, _attrs, _result)
_result = _QuantizedMulOutput._make(_result)
return _result
QuantizedMul = tf_export("raw_ops.QuantizedMul")(_ops.to_raw_op(quantized_mul))
def quantized_mul_eager_fallback(x, y, min_x, max_x, min_y, max_y, Toutput, name, ctx):
if Toutput is None:
Toutput = _dtypes.qint32
Toutput = _execute.make_type(Toutput, "Toutput")
_attr_T1, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.qint16, _dtypes.quint16, ])
_attr_T2, (y,) = _execute.args_to_matching_eager([y], ctx, [_dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.qint16, _dtypes.quint16, ])
min_x = _ops.convert_to_tensor(min_x, _dtypes.float32)
max_x = _ops.convert_to_tensor(max_x, _dtypes.float32)
min_y = _ops.convert_to_tensor(min_y, _dtypes.float32)
max_y = _ops.convert_to_tensor(max_y, _dtypes.float32)
_inputs_flat = [x, y, min_x, max_x, min_y, max_y]
_attrs = ("T1", _attr_T1, "T2", _attr_T2, "Toutput", Toutput)
_result = _execute.execute(b"QuantizedMul", 3, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"QuantizedMul", _inputs_flat, _attrs, _result)
_result = _QuantizedMulOutput._make(_result)
return _result
def ragged_bincount(splits, values, size, weights, binary_output=False, name=None):
r"""Counts the number of occurrences of each value in an integer array.
Outputs a vector with length `size` and the same dtype as `weights`. If
`weights` are empty, then index `i` stores the number of times the value `i` is
counted in `arr`. If `weights` are non-empty, then index `i` stores the sum of
the value in `weights` at each index where the corresponding value in `arr` is
`i`.
Values in `arr` outside of the range [0, size) are ignored.
Args:
splits: A `Tensor` of type `int64`. 1D int64 `Tensor`.
values: A `Tensor`. Must be one of the following types: `int32`, `int64`.
2D int `Tensor`.
size: A `Tensor`. Must have the same type as `values`.
non-negative int scalar `Tensor`.
weights: A `Tensor`. Must be one of the following types: `int32`, `int64`, `float32`, `float64`.
is an int32, int64, float32, or float64 `Tensor` with the same
shape as `input`, or a length-0 `Tensor`, in which case it acts as all weights
equal to 1.
binary_output: An optional `bool`. Defaults to `False`.
bool; Whether the kernel should count the appearance or number of occurrences.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `weights`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "RaggedBincount", name, splits, values, size, weights,
"binary_output", binary_output)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return ragged_bincount_eager_fallback(
splits, values, size, weights, binary_output=binary_output,
name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if binary_output is None:
binary_output = False
binary_output = _execute.make_bool(binary_output, "binary_output")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"RaggedBincount", splits=splits, values=values, size=size,
weights=weights, binary_output=binary_output,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("Tidx", _op._get_attr_type("Tidx"), "T",
_op._get_attr_type("T"), "binary_output",
_op._get_attr_bool("binary_output"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"RaggedBincount", _inputs_flat, _attrs, _result)
_result, = _result
return _result
RaggedBincount = tf_export("raw_ops.RaggedBincount")(_ops.to_raw_op(ragged_bincount))
def ragged_bincount_eager_fallback(splits, values, size, weights, binary_output, name, ctx):
if binary_output is None:
binary_output = False
binary_output = _execute.make_bool(binary_output, "binary_output")
_attr_Tidx, _inputs_Tidx = _execute.args_to_matching_eager([values, size], ctx, [_dtypes.int32, _dtypes.int64, ])
(values, size) = _inputs_Tidx
_attr_T, (weights,) = _execute.args_to_matching_eager([weights], ctx, [_dtypes.int32, _dtypes.int64, _dtypes.float32, _dtypes.float64, ])
splits = _ops.convert_to_tensor(splits, _dtypes.int64)
_inputs_flat = [splits, values, size, weights]
_attrs = ("Tidx", _attr_Tidx, "T", _attr_T, "binary_output", binary_output)
_result = _execute.execute(b"RaggedBincount", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"RaggedBincount", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def _range(start, limit, delta, name=None):
r"""Creates a sequence of numbers.
This operation creates a sequence of numbers that begins at `start` and
extends by increments of `delta` up to but not including `limit`.
For example:
```
# 'start' is 3
# 'limit' is 18
# 'delta' is 3
tf.range(start, limit, delta) ==> [3, 6, 9, 12, 15]
```
Args:
start: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int8`, `int16`, `int32`, `int64`, `uint16`, `uint32`.
0-D (scalar). First entry in the sequence.
limit: A `Tensor`. Must have the same type as `start`.
0-D (scalar). Upper limit of sequence, exclusive.
delta: A `Tensor`. Must have the same type as `start`.
0-D (scalar). Optional. Default is 1. Number that increments `start`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `start`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Range", name, start, limit, delta)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return _range_eager_fallback(
start, limit, delta, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Range", start=start, limit=limit, delta=delta, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("Tidx", _op._get_attr_type("Tidx"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Range", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Range = tf_export("raw_ops.Range")(_ops.to_raw_op(_range))
def _range_eager_fallback(start, limit, delta, name, ctx):
_attr_Tidx, _inputs_Tidx = _execute.args_to_matching_eager([start, limit, delta], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.uint16, _dtypes.uint32, ], _dtypes.int32)
(start, limit, delta) = _inputs_Tidx
_inputs_flat = [start, limit, delta]
_attrs = ("Tidx", _attr_Tidx)
_result = _execute.execute(b"Range", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Range", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def real(input, Tout=_dtypes.float32, name=None):
r"""Returns the real part of a complex number.
Given a tensor `input` of complex numbers, this operation returns a tensor of
type `float` that is the real part of each element in `input`. All elements in
`input` must be complex numbers of the form \\(a + bj\\), where *a* is the real
part returned by this operation and *b* is the imaginary part.
For example:
```
# tensor 'input' is [-2.25 + 4.75j, 3.25 + 5.75j]
tf.real(input) ==> [-2.25, 3.25]
```
Args:
input: A `Tensor`. Must be one of the following types: `complex64`, `complex128`.
Tout: An optional `tf.DType` from: `tf.float32, tf.float64`. Defaults to `tf.float32`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `Tout`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Real", name, input, "Tout", Tout)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return real_eager_fallback(
input, Tout=Tout, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if Tout is None:
Tout = _dtypes.float32
Tout = _execute.make_type(Tout, "Tout")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Real", input=input, Tout=Tout, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tout",
_op._get_attr_type("Tout"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Real", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Real = tf_export("raw_ops.Real")(_ops.to_raw_op(real))
def real_eager_fallback(input, Tout, name, ctx):
if Tout is None:
Tout = _dtypes.float32
Tout = _execute.make_type(Tout, "Tout")
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.complex64, _dtypes.complex128, ], _dtypes.complex64)
_inputs_flat = [input]
_attrs = ("T", _attr_T, "Tout", Tout)
_result = _execute.execute(b"Real", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Real", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('realdiv')
def real_div(x, y, name=None):
r"""Returns x / y element-wise for real types.
If `x` and `y` are reals, this will return the floating-point division.
*NOTE*: `Div` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `uint8`, `int8`, `uint16`, `int16`, `int32`, `uint32`, `uint64`, `int64`, `complex64`, `complex128`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "RealDiv", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_real_div(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return real_div_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
real_div, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_real_div(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"RealDiv", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
real_div, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"RealDiv", _inputs_flat, _attrs, _result)
_result, = _result
return _result
RealDiv = tf_export("raw_ops.RealDiv")(_ops.to_raw_op(real_div))
_dispatcher_for_real_div = real_div._tf_type_based_dispatcher.Dispatch
def real_div_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.uint8, _dtypes.int8, _dtypes.uint16, _dtypes.int16, _dtypes.int32, _dtypes.uint32, _dtypes.uint64, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"RealDiv", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"RealDiv", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.reciprocal', v1=['math.reciprocal', 'reciprocal'])
@deprecated_endpoints('reciprocal')
def reciprocal(x, name=None):
r"""Computes the reciprocal of x element-wise.
I.e., \\(y = 1 / x\\).
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int8`, `int16`, `int32`, `int64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Reciprocal", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_reciprocal(
(x, name,), None)
if _result is not NotImplemented:
return _result
return reciprocal_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
reciprocal, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_reciprocal(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Reciprocal", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
reciprocal, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Reciprocal", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Reciprocal = tf_export("raw_ops.Reciprocal")(_ops.to_raw_op(reciprocal))
_dispatcher_for_reciprocal = reciprocal._tf_type_based_dispatcher.Dispatch
def reciprocal_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Reciprocal", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Reciprocal", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def reciprocal_grad(y, dy, name=None):
r"""Computes the gradient for the inverse of `x` wrt its input.
Specifically, `grad = -dy * y*y`, where `y = 1/x`, and `dy`
is the corresponding input gradient.
Args:
y: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
dy: A `Tensor`. Must have the same type as `y`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `y`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "ReciprocalGrad", name, y, dy)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return reciprocal_grad_eager_fallback(
y, dy, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"ReciprocalGrad", y=y, dy=dy, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"ReciprocalGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
ReciprocalGrad = tf_export("raw_ops.ReciprocalGrad")(_ops.to_raw_op(reciprocal_grad))
def reciprocal_grad_eager_fallback(y, dy, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([y, dy], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
(y, dy) = _inputs_T
_inputs_flat = [y, dy]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"ReciprocalGrad", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"ReciprocalGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
_RequantizationRangeOutput = collections.namedtuple(
"RequantizationRange",
["output_min", "output_max"])
def requantization_range(input, input_min, input_max, name=None):
r"""Computes a range that covers the actual values present in a quantized tensor.
Given a quantized tensor described by `(input, input_min, input_max)`, outputs a
range that covers the actual values present in that tensor. This op is typically
used to produce the `requested_output_min` and `requested_output_max` for
`Requantize`.
Args:
input: A `Tensor`. Must be one of the following types: `qint8`, `quint8`, `qint32`, `qint16`, `quint16`.
input_min: A `Tensor` of type `float32`.
The float value that the minimum quantized input value represents.
input_max: A `Tensor` of type `float32`.
The float value that the maximum quantized input value represents.
name: A name for the operation (optional).
Returns:
A tuple of `Tensor` objects (output_min, output_max).
output_min: A `Tensor` of type `float32`.
output_max: A `Tensor` of type `float32`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "RequantizationRange", name, input, input_min, input_max)
_result = _RequantizationRangeOutput._make(_result)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return requantization_range_eager_fallback(
input, input_min, input_max, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"RequantizationRange", input=input, input_min=input_min,
input_max=input_max, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("Tinput", _op._get_attr_type("Tinput"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"RequantizationRange", _inputs_flat, _attrs, _result)
_result = _RequantizationRangeOutput._make(_result)
return _result
RequantizationRange = tf_export("raw_ops.RequantizationRange")(_ops.to_raw_op(requantization_range))
def requantization_range_eager_fallback(input, input_min, input_max, name, ctx):
_attr_Tinput, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.qint16, _dtypes.quint16, ])
input_min = _ops.convert_to_tensor(input_min, _dtypes.float32)
input_max = _ops.convert_to_tensor(input_max, _dtypes.float32)
_inputs_flat = [input, input_min, input_max]
_attrs = ("Tinput", _attr_Tinput)
_result = _execute.execute(b"RequantizationRange", 2, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"RequantizationRange", _inputs_flat, _attrs, _result)
_result = _RequantizationRangeOutput._make(_result)
return _result
_RequantizationRangePerChannelOutput = collections.namedtuple(
"RequantizationRangePerChannel",
["output_min", "output_max"])
def requantization_range_per_channel(input, input_min, input_max, clip_value_max, name=None):
r"""Computes requantization range per channel.
Args:
input: A `Tensor`. Must be one of the following types: `qint8`, `quint8`, `qint32`, `qint16`, `quint16`.
The original input tensor.
input_min: A `Tensor` of type `float32`.
The minimum value of the input tensor
input_max: A `Tensor` of type `float32`.
The maximum value of the input tensor.
clip_value_max: A `float`.
The maximum value of the output that needs to be clipped.
Example: set this to 6 for Relu6.
name: A name for the operation (optional).
Returns:
A tuple of `Tensor` objects (output_min, output_max).
output_min: A `Tensor` of type `float32`.
output_max: A `Tensor` of type `float32`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "RequantizationRangePerChannel", name, input, input_min,
input_max, "clip_value_max", clip_value_max)
_result = _RequantizationRangePerChannelOutput._make(_result)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return requantization_range_per_channel_eager_fallback(
input, input_min, input_max, clip_value_max=clip_value_max,
name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
clip_value_max = _execute.make_float(clip_value_max, "clip_value_max")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"RequantizationRangePerChannel", input=input, input_min=input_min,
input_max=input_max,
clip_value_max=clip_value_max,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "clip_value_max",
_op.get_attr("clip_value_max"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"RequantizationRangePerChannel", _inputs_flat, _attrs, _result)
_result = _RequantizationRangePerChannelOutput._make(_result)
return _result
RequantizationRangePerChannel = tf_export("raw_ops.RequantizationRangePerChannel")(_ops.to_raw_op(requantization_range_per_channel))
def requantization_range_per_channel_eager_fallback(input, input_min, input_max, clip_value_max, name, ctx):
clip_value_max = _execute.make_float(clip_value_max, "clip_value_max")
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.qint16, _dtypes.quint16, ], _dtypes.qint32)
input_min = _ops.convert_to_tensor(input_min, _dtypes.float32)
input_max = _ops.convert_to_tensor(input_max, _dtypes.float32)
_inputs_flat = [input, input_min, input_max]
_attrs = ("T", _attr_T, "clip_value_max", clip_value_max)
_result = _execute.execute(b"RequantizationRangePerChannel", 2,
inputs=_inputs_flat, attrs=_attrs, ctx=ctx,
name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"RequantizationRangePerChannel", _inputs_flat, _attrs, _result)
_result = _RequantizationRangePerChannelOutput._make(_result)
return _result
_RequantizeOutput = collections.namedtuple(
"Requantize",
["output", "output_min", "output_max"])
def requantize(input, input_min, input_max, requested_output_min, requested_output_max, out_type, name=None):
r"""Converts the quantized `input` tensor into a lower-precision `output`.
Converts the quantized `input` tensor into a lower-precision `output`, using the
output range specified with `requested_output_min` and `requested_output_max`.
`[input_min, input_max]` are scalar floats that specify the range for the float
interpretation of the `input` data. For example, if `input_min` is -1.0f and
`input_max` is 1.0f, and we are dealing with `quint16` quantized data, then a 0
value in the 16-bit data should be interpreted as -1.0f, and a 65535 means 1.0f.
Args:
input: A `Tensor`. Must be one of the following types: `qint8`, `quint8`, `qint32`, `qint16`, `quint16`.
input_min: A `Tensor` of type `float32`.
The float value that the minimum quantized input value represents.
input_max: A `Tensor` of type `float32`.
The float value that the maximum quantized input value represents.
requested_output_min: A `Tensor` of type `float32`.
The float value that the minimum quantized output value represents.
requested_output_max: A `Tensor` of type `float32`.
The float value that the maximum quantized output value represents.
out_type: A `tf.DType` from: `tf.qint8, tf.quint8, tf.qint32, tf.qint16, tf.quint16`.
The type of the output. Should be a lower bit depth than Tinput.
name: A name for the operation (optional).
Returns:
A tuple of `Tensor` objects (output, output_min, output_max).
output: A `Tensor` of type `out_type`.
output_min: A `Tensor` of type `float32`.
output_max: A `Tensor` of type `float32`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Requantize", name, input, input_min, input_max,
requested_output_min, requested_output_max, "out_type", out_type)
_result = _RequantizeOutput._make(_result)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return requantize_eager_fallback(
input, input_min, input_max, requested_output_min,
requested_output_max, out_type=out_type, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
out_type = _execute.make_type(out_type, "out_type")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Requantize", input=input, input_min=input_min, input_max=input_max,
requested_output_min=requested_output_min,
requested_output_max=requested_output_max,
out_type=out_type, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("Tinput", _op._get_attr_type("Tinput"), "out_type",
_op._get_attr_type("out_type"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Requantize", _inputs_flat, _attrs, _result)
_result = _RequantizeOutput._make(_result)
return _result
Requantize = tf_export("raw_ops.Requantize")(_ops.to_raw_op(requantize))
def requantize_eager_fallback(input, input_min, input_max, requested_output_min, requested_output_max, out_type, name, ctx):
out_type = _execute.make_type(out_type, "out_type")
_attr_Tinput, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.qint16, _dtypes.quint16, ])
input_min = _ops.convert_to_tensor(input_min, _dtypes.float32)
input_max = _ops.convert_to_tensor(input_max, _dtypes.float32)
requested_output_min = _ops.convert_to_tensor(requested_output_min, _dtypes.float32)
requested_output_max = _ops.convert_to_tensor(requested_output_max, _dtypes.float32)
_inputs_flat = [input, input_min, input_max, requested_output_min, requested_output_max]
_attrs = ("Tinput", _attr_Tinput, "out_type", out_type)
_result = _execute.execute(b"Requantize", 3, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Requantize", _inputs_flat, _attrs, _result)
_result = _RequantizeOutput._make(_result)
return _result
_RequantizePerChannelOutput = collections.namedtuple(
"RequantizePerChannel",
["output", "output_min", "output_max"])
def requantize_per_channel(input, input_min, input_max, requested_output_min, requested_output_max, out_type=_dtypes.quint8, name=None):
r"""Requantizes input with min and max values known per channel.
Args:
input: A `Tensor`. Must be one of the following types: `qint8`, `quint8`, `qint32`, `qint16`, `quint16`.
The original input tensor.
input_min: A `Tensor` of type `float32`.
The minimum value of the input tensor
input_max: A `Tensor` of type `float32`.
The maximum value of the input tensor.
requested_output_min: A `Tensor` of type `float32`.
The minimum value of the output tensor requested.
requested_output_max: A `Tensor` of type `float32`.
The maximum value of the output tensor requested.
out_type: An optional `tf.DType` from: `tf.qint8, tf.quint8, tf.qint32, tf.qint16, tf.quint16`. Defaults to `tf.quint8`.
The quantized type of output tensor that needs to be converted.
name: A name for the operation (optional).
Returns:
A tuple of `Tensor` objects (output, output_min, output_max).
output: A `Tensor` of type `out_type`.
output_min: A `Tensor` of type `float32`.
output_max: A `Tensor` of type `float32`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "RequantizePerChannel", name, input, input_min, input_max,
requested_output_min, requested_output_max, "out_type", out_type)
_result = _RequantizePerChannelOutput._make(_result)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return requantize_per_channel_eager_fallback(
input, input_min, input_max, requested_output_min,
requested_output_max, out_type=out_type, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if out_type is None:
out_type = _dtypes.quint8
out_type = _execute.make_type(out_type, "out_type")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"RequantizePerChannel", input=input, input_min=input_min,
input_max=input_max,
requested_output_min=requested_output_min,
requested_output_max=requested_output_max,
out_type=out_type, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "out_type",
_op._get_attr_type("out_type"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"RequantizePerChannel", _inputs_flat, _attrs, _result)
_result = _RequantizePerChannelOutput._make(_result)
return _result
RequantizePerChannel = tf_export("raw_ops.RequantizePerChannel")(_ops.to_raw_op(requantize_per_channel))
def requantize_per_channel_eager_fallback(input, input_min, input_max, requested_output_min, requested_output_max, out_type, name, ctx):
if out_type is None:
out_type = _dtypes.quint8
out_type = _execute.make_type(out_type, "out_type")
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.qint16, _dtypes.quint16, ], _dtypes.qint32)
input_min = _ops.convert_to_tensor(input_min, _dtypes.float32)
input_max = _ops.convert_to_tensor(input_max, _dtypes.float32)
requested_output_min = _ops.convert_to_tensor(requested_output_min, _dtypes.float32)
requested_output_max = _ops.convert_to_tensor(requested_output_max, _dtypes.float32)
_inputs_flat = [input, input_min, input_max, requested_output_min, requested_output_max]
_attrs = ("T", _attr_T, "out_type", out_type)
_result = _execute.execute(b"RequantizePerChannel", 3, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"RequantizePerChannel", _inputs_flat, _attrs, _result)
_result = _RequantizePerChannelOutput._make(_result)
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.rint', v1=['math.rint', 'rint'])
@deprecated_endpoints('rint')
def rint(x, name=None):
r"""Returns element-wise integer closest to x.
If the result is midway between two representable values,
the even representable is chosen.
For example:
```
rint(-1.5) ==> -2.0
rint(0.5000001) ==> 1.0
rint([-1.7, -1.5, -0.2, 0.2, 1.5, 1.7, 2.0]) ==> [-2., -2., -0., 0., 2., 2., 2.]
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Rint", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_rint(
(x, name,), None)
if _result is not NotImplemented:
return _result
return rint_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
rint, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_rint(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Rint", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
rint, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Rint", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Rint = tf_export("raw_ops.Rint")(_ops.to_raw_op(rint))
_dispatcher_for_rint = rint._tf_type_based_dispatcher.Dispatch
def rint_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Rint", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Rint", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def round(x, name=None):
r"""Rounds the values of a tensor to the nearest integer, element-wise.
Rounds half to even. Also known as bankers rounding. If you want to round
according to the current system rounding mode use std::cint.
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int8`, `int16`, `int32`, `int64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Round", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return round_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Round", x=x, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Round", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Round = tf_export("raw_ops.Round")(_ops.to_raw_op(round))
def round_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Round", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Round", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def rsqrt(x, name=None):
r"""Computes reciprocal of square root of x element-wise.
I.e., \\(y = 1 / \sqrt{x}\\).
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Rsqrt", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return rsqrt_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Rsqrt", x=x, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Rsqrt", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Rsqrt = tf_export("raw_ops.Rsqrt")(_ops.to_raw_op(rsqrt))
def rsqrt_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Rsqrt", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Rsqrt", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def rsqrt_grad(y, dy, name=None):
r"""Computes the gradient for the rsqrt of `x` wrt its input.
Specifically, `grad = dy * -0.5 * y^3`, where `y = rsqrt(x)`, and `dy`
is the corresponding input gradient.
Args:
y: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
dy: A `Tensor`. Must have the same type as `y`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `y`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "RsqrtGrad", name, y, dy)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return rsqrt_grad_eager_fallback(
y, dy, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"RsqrtGrad", y=y, dy=dy, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"RsqrtGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
RsqrtGrad = tf_export("raw_ops.RsqrtGrad")(_ops.to_raw_op(rsqrt_grad))
def rsqrt_grad_eager_fallback(y, dy, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([y, dy], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
(y, dy) = _inputs_T
_inputs_flat = [y, dy]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"RsqrtGrad", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"RsqrtGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.segment_max', v1=['math.segment_max', 'segment_max'])
@deprecated_endpoints('segment_max')
def segment_max(data, segment_ids, name=None):
r"""Computes the maximum along segments of a tensor.
Read
[the section on segmentation](https://tensorflow.org/api_docs/python/tf/math#Segmentation)
for an explanation of segments.
Computes a tensor such that
\\(output_i = \max_j(data_j)\\) where `max` is over `j` such
that `segment_ids[j] == i`.
If the max is empty for a given segment ID `i`, `output[i] = 0`.
Caution: On CPU, values in `segment_ids` are always validated to be sorted,
and an error is thrown for indices that are not increasing. On GPU, this
does not throw an error for unsorted indices. On GPU, out-of-order indices
result in safe but unspecified behavior, which may include treating
out-of-order indices as the same as a smaller following index.
<div style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;">
<img style="width:100%" src="https://www.tensorflow.org/images/SegmentMax.png" alt>
</div>
For example:
>>> c = tf.constant([[1,2,3,4], [4, 3, 2, 1], [5,6,7,8]])
>>> tf.math.segment_max(c, tf.constant([0, 0, 1])).numpy()
array([[4, 3, 3, 4],
[5, 6, 7, 8]], dtype=int32)
Args:
data: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `int64`, `bfloat16`, `uint16`, `half`, `uint32`, `uint64`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor whose size is equal to the size of `data`'s
first dimension. Values should be sorted and can be repeated.
Caution: The values are always validated to be sorted on CPU, never validated
on GPU.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SegmentMax", name, data, segment_ids)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_segment_max(
(data, segment_ids, name,), None)
if _result is not NotImplemented:
return _result
return segment_max_eager_fallback(
data, segment_ids, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
segment_max, (), dict(data=data, segment_ids=segment_ids,
name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_segment_max(
(data, segment_ids, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SegmentMax", data=data, segment_ids=segment_ids, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
segment_max, (), dict(data=data, segment_ids=segment_ids, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tindices",
_op._get_attr_type("Tindices"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SegmentMax", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SegmentMax = tf_export("raw_ops.SegmentMax")(_ops.to_raw_op(segment_max))
_dispatcher_for_segment_max = segment_max._tf_type_based_dispatcher.Dispatch
def segment_max_eager_fallback(data, segment_ids, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.int64, _dtypes.bfloat16, _dtypes.uint16, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tindices, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ])
_inputs_flat = [data, segment_ids]
_attrs = ("T", _attr_T, "Tindices", _attr_Tindices)
_result = _execute.execute(b"SegmentMax", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SegmentMax", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.segment_mean', v1=['math.segment_mean', 'segment_mean'])
@deprecated_endpoints('segment_mean')
def segment_mean(data, segment_ids, name=None):
r"""Computes the mean along segments of a tensor.
Read
[the section on segmentation](https://tensorflow.org/api_docs/python/tf/math#Segmentation)
for an explanation of segments.
Computes a tensor such that
\\(output_i = \frac{\sum_j data_j}{N}\\) where `mean` is
over `j` such that `segment_ids[j] == i` and `N` is the total number of
values summed.
If the mean is empty for a given segment ID `i`, `output[i] = 0`.
Caution: On CPU, values in `segment_ids` are always validated to be sorted,
and an error is thrown for indices that are not increasing. On GPU, this
does not throw an error for unsorted indices. On GPU, out-of-order indices
result in safe but unspecified behavior, which may include treating
out-of-order indices as a smaller following index when computing the numerator
of the mean.
<div style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;">
<img style="width:100%" src="https://www.tensorflow.org/images/SegmentMean.png" alt>
</div>
For example:
>>> c = tf.constant([[1.0,2,3,4], [4, 3, 2, 1], [5,6,7,8]])
>>> tf.math.segment_mean(c, tf.constant([0, 0, 1])).numpy()
array([[2.5, 2.5, 2.5, 2.5],
[5., 6., 7., 8.]], dtype=float32)
Args:
data: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor whose size is equal to the size of `data`'s
first dimension. Values should be sorted and can be repeated.
Caution: The values are always validated to be sorted on CPU, never validated
on GPU.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SegmentMean", name, data, segment_ids)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_segment_mean(
(data, segment_ids, name,), None)
if _result is not NotImplemented:
return _result
return segment_mean_eager_fallback(
data, segment_ids, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
segment_mean, (), dict(data=data, segment_ids=segment_ids,
name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_segment_mean(
(data, segment_ids, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SegmentMean", data=data, segment_ids=segment_ids, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
segment_mean, (), dict(data=data, segment_ids=segment_ids,
name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tindices",
_op._get_attr_type("Tindices"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SegmentMean", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SegmentMean = tf_export("raw_ops.SegmentMean")(_ops.to_raw_op(segment_mean))
_dispatcher_for_segment_mean = segment_mean._tf_type_based_dispatcher.Dispatch
def segment_mean_eager_fallback(data, segment_ids, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tindices, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ])
_inputs_flat = [data, segment_ids]
_attrs = ("T", _attr_T, "Tindices", _attr_Tindices)
_result = _execute.execute(b"SegmentMean", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SegmentMean", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.segment_min', v1=['math.segment_min', 'segment_min'])
@deprecated_endpoints('segment_min')
def segment_min(data, segment_ids, name=None):
r"""Computes the minimum along segments of a tensor.
Read
[the section on segmentation](https://tensorflow.org/api_docs/python/tf/math#Segmentation)
for an explanation of segments.
Computes a tensor such that
\\(output_i = \min_j(data_j)\\) where `min` is over `j` such
that `segment_ids[j] == i`.
If the min is empty for a given segment ID `i`, `output[i] = 0`.
Caution: On CPU, values in `segment_ids` are always validated to be sorted,
and an error is thrown for indices that are not increasing. On GPU, this
does not throw an error for unsorted indices. On GPU, out-of-order indices
result in safe but unspecified behavior, which may include treating
out-of-order indices as the same as a smaller following index.
<div style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;">
<img style="width:100%" src="https://www.tensorflow.org/images/SegmentMin.png" alt>
</div>
For example:
>>> c = tf.constant([[1,2,3,4], [4, 3, 2, 1], [5,6,7,8]])
>>> tf.math.segment_min(c, tf.constant([0, 0, 1])).numpy()
array([[1, 2, 2, 1],
[5, 6, 7, 8]], dtype=int32)
Args:
data: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `int64`, `bfloat16`, `uint16`, `half`, `uint32`, `uint64`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor whose size is equal to the size of `data`'s
first dimension. Values should be sorted and can be repeated.
Caution: The values are always validated to be sorted on CPU, never validated
on GPU.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SegmentMin", name, data, segment_ids)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_segment_min(
(data, segment_ids, name,), None)
if _result is not NotImplemented:
return _result
return segment_min_eager_fallback(
data, segment_ids, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
segment_min, (), dict(data=data, segment_ids=segment_ids,
name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_segment_min(
(data, segment_ids, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SegmentMin", data=data, segment_ids=segment_ids, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
segment_min, (), dict(data=data, segment_ids=segment_ids, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tindices",
_op._get_attr_type("Tindices"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SegmentMin", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SegmentMin = tf_export("raw_ops.SegmentMin")(_ops.to_raw_op(segment_min))
_dispatcher_for_segment_min = segment_min._tf_type_based_dispatcher.Dispatch
def segment_min_eager_fallback(data, segment_ids, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.int64, _dtypes.bfloat16, _dtypes.uint16, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tindices, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ])
_inputs_flat = [data, segment_ids]
_attrs = ("T", _attr_T, "Tindices", _attr_Tindices)
_result = _execute.execute(b"SegmentMin", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SegmentMin", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.segment_prod', v1=['math.segment_prod', 'segment_prod'])
@deprecated_endpoints('segment_prod')
def segment_prod(data, segment_ids, name=None):
r"""Computes the product along segments of a tensor.
Read
[the section on segmentation](https://tensorflow.org/api_docs/python/tf/math#Segmentation)
for an explanation of segments.
Computes a tensor such that
\\(output_i = \prod_j data_j\\) where the product is over `j` such
that `segment_ids[j] == i`.
If the product is empty for a given segment ID `i`, `output[i] = 1`.
Caution: On CPU, values in `segment_ids` are always validated to be sorted,
and an error is thrown for indices that are not increasing. On GPU, this
does not throw an error for unsorted indices. On GPU, out-of-order indices
result in safe but unspecified behavior, which may include treating
out-of-order indices as the same as a smaller following index.
<div style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;">
<img style="width:100%" src="https://www.tensorflow.org/images/SegmentProd.png" alt>
</div>
For example:
>>> c = tf.constant([[1,2,3,4], [4, 3, 2, 1], [5,6,7,8]])
>>> tf.math.segment_prod(c, tf.constant([0, 0, 1])).numpy()
array([[4, 6, 6, 4],
[5, 6, 7, 8]], dtype=int32)
Args:
data: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor whose size is equal to the size of `data`'s
first dimension. Values should be sorted and can be repeated.
Caution: The values are always validated to be sorted on CPU, never validated
on GPU.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SegmentProd", name, data, segment_ids)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_segment_prod(
(data, segment_ids, name,), None)
if _result is not NotImplemented:
return _result
return segment_prod_eager_fallback(
data, segment_ids, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
segment_prod, (), dict(data=data, segment_ids=segment_ids,
name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_segment_prod(
(data, segment_ids, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SegmentProd", data=data, segment_ids=segment_ids, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
segment_prod, (), dict(data=data, segment_ids=segment_ids,
name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tindices",
_op._get_attr_type("Tindices"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SegmentProd", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SegmentProd = tf_export("raw_ops.SegmentProd")(_ops.to_raw_op(segment_prod))
_dispatcher_for_segment_prod = segment_prod._tf_type_based_dispatcher.Dispatch
def segment_prod_eager_fallback(data, segment_ids, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tindices, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ])
_inputs_flat = [data, segment_ids]
_attrs = ("T", _attr_T, "Tindices", _attr_Tindices)
_result = _execute.execute(b"SegmentProd", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SegmentProd", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.segment_sum', v1=['math.segment_sum', 'segment_sum'])
@deprecated_endpoints('segment_sum')
def segment_sum(data, segment_ids, name=None):
r"""Computes the sum along segments of a tensor.
Read
[the section on segmentation](https://tensorflow.org/api_docs/python/tf/math#Segmentation)
for an explanation of segments.
Computes a tensor such that
\\(output_i = \sum_j data_j\\) where sum is over `j` such
that `segment_ids[j] == i`.
If the sum is empty for a given segment ID `i`, `output[i] = 0`.
Caution: On CPU, values in `segment_ids` are always validated to be sorted,
and an error is thrown for indices that are not increasing. On GPU, this
does not throw an error for unsorted indices. On GPU, out-of-order indices
result in safe but unspecified behavior, which may include treating
out-of-order indices as the same as a smaller following index.
<div style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;">
<img style="width:100%" src="https://www.tensorflow.org/images/SegmentSum.png" alt>
</div>
For example:
>>> c = tf.constant([[1,2,3,4], [4, 3, 2, 1], [5,6,7,8]])
>>> tf.math.segment_sum(c, tf.constant([0, 0, 1])).numpy()
array([[5, 5, 5, 5],
[5, 6, 7, 8]], dtype=int32)
Args:
data: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor whose size is equal to the size of `data`'s
first dimension. Values should be sorted and can be repeated.
Caution: The values are always validated to be sorted on CPU, never validated
on GPU.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SegmentSum", name, data, segment_ids)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_segment_sum(
(data, segment_ids, name,), None)
if _result is not NotImplemented:
return _result
return segment_sum_eager_fallback(
data, segment_ids, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
segment_sum, (), dict(data=data, segment_ids=segment_ids,
name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_segment_sum(
(data, segment_ids, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SegmentSum", data=data, segment_ids=segment_ids, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
segment_sum, (), dict(data=data, segment_ids=segment_ids, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tindices",
_op._get_attr_type("Tindices"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SegmentSum", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SegmentSum = tf_export("raw_ops.SegmentSum")(_ops.to_raw_op(segment_sum))
_dispatcher_for_segment_sum = segment_sum._tf_type_based_dispatcher.Dispatch
def segment_sum_eager_fallback(data, segment_ids, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tindices, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ])
_inputs_flat = [data, segment_ids]
_attrs = ("T", _attr_T, "Tindices", _attr_Tindices)
_result = _execute.execute(b"SegmentSum", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SegmentSum", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def select(condition, x, y, name=None):
r"""Selects elements from `x` or `y`, depending on `condition`.
The `x`, and `y` tensors must all have the same shape, and the
output will also have that shape.
The `condition` tensor must be a scalar if `x` and `y` are scalars.
If `x` and `y` are vectors or higher rank, then `condition` must be either a
scalar, a vector with size matching the first dimension of `x`, or must have
the same shape as `x`.
The `condition` tensor acts as a mask that chooses, based on the value at each
element, whether the corresponding element / row in the output should be
taken from `x` (if true) or `y` (if false).
If `condition` is a vector and `x` and `y` are higher rank matrices, then
it chooses which row (outer dimension) to copy from `x` and `y`.
If `condition` has the same shape as `x` and `y`, then it chooses which
element to copy from `x` and `y`.
For example:
```python
# 'condition' tensor is [[True, False]
# [False, True]]
# 't' is [[1, 2],
# [3, 4]]
# 'e' is [[5, 6],
# [7, 8]]
select(condition, t, e) # => [[1, 6], [7, 4]]
# 'condition' tensor is [True, False]
# 't' is [[1, 2],
# [3, 4]]
# 'e' is [[5, 6],
# [7, 8]]
select(condition, t, e) ==> [[1, 2],
[7, 8]]
```
Args:
condition: A `Tensor` of type `bool`.
x: A `Tensor` which may have the same shape as `condition`.
If `condition` is rank 1, `x` may have higher rank,
but its first dimension must match the size of `condition`.
y: A `Tensor` with the same type and shape as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `t`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Select", name, condition, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return select_eager_fallback(
condition, x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Select", condition=condition, t=x, e=y, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Select", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Select = tf_export("raw_ops.Select")(_ops.to_raw_op(select))
def select_eager_fallback(condition, x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [])
(x, y) = _inputs_T
condition = _ops.convert_to_tensor(condition, _dtypes.bool)
_inputs_flat = [condition, x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Select", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Select", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def select_v2(condition, t, e, name=None):
r"""TODO: add doc.
Args:
condition: A `Tensor` of type `bool`.
t: A `Tensor`.
e: A `Tensor`. Must have the same type as `t`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `t`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SelectV2", name, condition, t, e)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return select_v2_eager_fallback(
condition, t, e, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SelectV2", condition=condition, t=t, e=e, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SelectV2", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SelectV2 = tf_export("raw_ops.SelectV2")(_ops.to_raw_op(select_v2))
def select_v2_eager_fallback(condition, t, e, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([t, e], ctx, [])
(t, e) = _inputs_T
condition = _ops.convert_to_tensor(condition, _dtypes.bool)
_inputs_flat = [condition, t, e]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"SelectV2", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SelectV2", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sigmoid(x, name=None):
r"""Computes sigmoid of `x` element-wise.
Specifically, `y = 1 / (1 + exp(-x))`.
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Sigmoid", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sigmoid_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Sigmoid", x=x, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Sigmoid", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Sigmoid = tf_export("raw_ops.Sigmoid")(_ops.to_raw_op(sigmoid))
def sigmoid_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Sigmoid", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Sigmoid", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sigmoid_grad(y, dy, name=None):
r"""Computes the gradient of the sigmoid of `x` wrt its input.
Specifically, `grad = dy * y * (1 - y)`, where `y = sigmoid(x)`, and
`dy` is the corresponding input gradient.
Args:
y: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
dy: A `Tensor`. Must have the same type as `y`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `y`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SigmoidGrad", name, y, dy)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sigmoid_grad_eager_fallback(
y, dy, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SigmoidGrad", y=y, dy=dy, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SigmoidGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SigmoidGrad = tf_export("raw_ops.SigmoidGrad")(_ops.to_raw_op(sigmoid_grad))
def sigmoid_grad_eager_fallback(y, dy, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([y, dy], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
(y, dy) = _inputs_T
_inputs_flat = [y, dy]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"SigmoidGrad", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SigmoidGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sign(x, name=None):
r"""Returns an element-wise indication of the sign of a number.
`y = sign(x) = -1` if `x < 0`; 0 if `x == 0`; 1 if `x > 0`.
For complex numbers, `y = sign(x) = x / |x|` if `x != 0`, otherwise `y = 0`.
Example usage:
>>> tf.math.sign([0., 2., -3.])
<tf.Tensor: shape=(3,), dtype=float32, numpy=array([ 0., 1., -1.], dtype=float32)>
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int8`, `int16`, `int32`, `int64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Sign", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sign_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Sign", x=x, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Sign", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Sign = tf_export("raw_ops.Sign")(_ops.to_raw_op(sign))
def sign_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Sign", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Sign", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.sin', 'sin')
def sin(x, name=None):
r"""Computes sine of x element-wise.
Given an input tensor, this function computes sine of every
element in the tensor. Input range is `(-inf, inf)` and
output range is `[-1,1]`.
```python
x = tf.constant([-float("inf"), -9, -0.5, 1, 1.2, 200, 10, float("inf")])
tf.math.sin(x) ==> [nan -0.4121185 -0.47942555 0.84147096 0.9320391 -0.87329733 -0.54402107 nan]
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Sin", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_sin(
(x, name,), None)
if _result is not NotImplemented:
return _result
return sin_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
sin, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_sin(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Sin", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
sin, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Sin", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Sin = tf_export("raw_ops.Sin")(_ops.to_raw_op(sin))
_dispatcher_for_sin = sin._tf_type_based_dispatcher.Dispatch
def sin_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Sin", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Sin", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.sinh', 'sinh')
def sinh(x, name=None):
r"""Computes hyperbolic sine of x element-wise.
Given an input tensor, this function computes hyperbolic sine of every
element in the tensor. Input range is `[-inf,inf]` and output range
is `[-inf,inf]`.
```python
x = tf.constant([-float("inf"), -9, -0.5, 1, 1.2, 2, 10, float("inf")])
tf.math.sinh(x) ==> [-inf -4.0515420e+03 -5.2109528e-01 1.1752012e+00 1.5094614e+00 3.6268604e+00 1.1013232e+04 inf]
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Sinh", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_sinh(
(x, name,), None)
if _result is not NotImplemented:
return _result
return sinh_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
sinh, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_sinh(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Sinh", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
sinh, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Sinh", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Sinh = tf_export("raw_ops.Sinh")(_ops.to_raw_op(sinh))
_dispatcher_for_sinh = sinh._tf_type_based_dispatcher.Dispatch
def sinh_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Sinh", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Sinh", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sobol_sample(dim, num_results, skip, dtype=_dtypes.float32, name=None):
r"""Generates points from the Sobol sequence.
Creates a Sobol sequence with `num_results` samples. Each sample has dimension
`dim`. Skips the first `skip` samples.
Args:
dim: A `Tensor` of type `int32`.
Positive scalar `Tensor` representing each sample's dimension.
num_results: A `Tensor` of type `int32`.
Positive scalar `Tensor` of dtype int32. The number of Sobol points to return
in the output.
skip: A `Tensor` of type `int32`.
Positive scalar `Tensor` of dtype int32. The number of initial points of the
Sobol sequence to skip.
dtype: An optional `tf.DType` from: `tf.float32, tf.float64`. Defaults to `tf.float32`.
The type of the sample. One of: `float32` or `float64`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `dtype`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SobolSample", name, dim, num_results, skip, "dtype", dtype)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sobol_sample_eager_fallback(
dim, num_results, skip, dtype=dtype, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if dtype is None:
dtype = _dtypes.float32
dtype = _execute.make_type(dtype, "dtype")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SobolSample", dim=dim, num_results=num_results, skip=skip,
dtype=dtype, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("dtype", _op._get_attr_type("dtype"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SobolSample", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SobolSample = tf_export("raw_ops.SobolSample")(_ops.to_raw_op(sobol_sample))
def sobol_sample_eager_fallback(dim, num_results, skip, dtype, name, ctx):
if dtype is None:
dtype = _dtypes.float32
dtype = _execute.make_type(dtype, "dtype")
dim = _ops.convert_to_tensor(dim, _dtypes.int32)
num_results = _ops.convert_to_tensor(num_results, _dtypes.int32)
skip = _ops.convert_to_tensor(skip, _dtypes.int32)
_inputs_flat = [dim, num_results, skip]
_attrs = ("dtype", dtype)
_result = _execute.execute(b"SobolSample", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SobolSample", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sparse_bincount(indices, values, dense_shape, size, weights, binary_output=False, name=None):
r"""Counts the number of occurrences of each value in an integer array.
Outputs a vector with length `size` and the same dtype as `weights`. If
`weights` are empty, then index `i` stores the number of times the value `i` is
counted in `arr`. If `weights` are non-empty, then index `i` stores the sum of
the value in `weights` at each index where the corresponding value in `arr` is
`i`.
Values in `arr` outside of the range [0, size) are ignored.
Args:
indices: A `Tensor` of type `int64`. 2D int64 `Tensor`.
values: A `Tensor`. Must be one of the following types: `int32`, `int64`.
1D int `Tensor`.
dense_shape: A `Tensor` of type `int64`. 1D int64 `Tensor`.
size: A `Tensor`. Must have the same type as `values`.
non-negative int scalar `Tensor`.
weights: A `Tensor`. Must be one of the following types: `int32`, `int64`, `float32`, `float64`.
is an int32, int64, float32, or float64 `Tensor` with the same
shape as `input`, or a length-0 `Tensor`, in which case it acts as all weights
equal to 1.
binary_output: An optional `bool`. Defaults to `False`.
bool; Whether the kernel should count the appearance or number of occurrences.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `weights`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SparseBincount", name, indices, values, dense_shape, size,
weights, "binary_output", binary_output)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sparse_bincount_eager_fallback(
indices, values, dense_shape, size, weights,
binary_output=binary_output, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if binary_output is None:
binary_output = False
binary_output = _execute.make_bool(binary_output, "binary_output")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SparseBincount", indices=indices, values=values,
dense_shape=dense_shape, size=size, weights=weights,
binary_output=binary_output, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("Tidx", _op._get_attr_type("Tidx"), "T",
_op._get_attr_type("T"), "binary_output",
_op._get_attr_bool("binary_output"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SparseBincount", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SparseBincount = tf_export("raw_ops.SparseBincount")(_ops.to_raw_op(sparse_bincount))
def sparse_bincount_eager_fallback(indices, values, dense_shape, size, weights, binary_output, name, ctx):
if binary_output is None:
binary_output = False
binary_output = _execute.make_bool(binary_output, "binary_output")
_attr_Tidx, _inputs_Tidx = _execute.args_to_matching_eager([values, size], ctx, [_dtypes.int32, _dtypes.int64, ])
(values, size) = _inputs_Tidx
_attr_T, (weights,) = _execute.args_to_matching_eager([weights], ctx, [_dtypes.int32, _dtypes.int64, _dtypes.float32, _dtypes.float64, ])
indices = _ops.convert_to_tensor(indices, _dtypes.int64)
dense_shape = _ops.convert_to_tensor(dense_shape, _dtypes.int64)
_inputs_flat = [indices, values, dense_shape, size, weights]
_attrs = ("Tidx", _attr_Tidx, "T", _attr_T, "binary_output", binary_output)
_result = _execute.execute(b"SparseBincount", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SparseBincount", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sparse_mat_mul(a, b, transpose_a=False, transpose_b=False, a_is_sparse=False, b_is_sparse=False, name=None):
r"""Multiply matrix "a" by matrix "b".
The inputs must be two-dimensional matrices and the inner dimension of "a" must
match the outer dimension of "b". Both "a" and "b" must be `Tensor`s not
`SparseTensor`s. This op is optimized for the case where at least one of "a" or
"b" is sparse, in the sense that they have a large proportion of zero values.
The breakeven for using this versus a dense matrix multiply on one platform was
30% zero values in the sparse matrix.
The gradient computation of this operation will only take advantage of sparsity
in the input gradient when that gradient comes from a Relu.
Args:
a: A `Tensor`. Must be one of the following types: `float32`, `bfloat16`.
b: A `Tensor`. Must be one of the following types: `float32`, `bfloat16`.
transpose_a: An optional `bool`. Defaults to `False`.
transpose_b: An optional `bool`. Defaults to `False`.
a_is_sparse: An optional `bool`. Defaults to `False`.
b_is_sparse: An optional `bool`. Defaults to `False`.
name: A name for the operation (optional).
Returns:
A `Tensor` of type `float32`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SparseMatMul", name, a, b, "transpose_a", transpose_a,
"transpose_b", transpose_b, "a_is_sparse", a_is_sparse, "b_is_sparse",
b_is_sparse)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sparse_mat_mul_eager_fallback(
a, b, transpose_a=transpose_a, transpose_b=transpose_b,
a_is_sparse=a_is_sparse, b_is_sparse=b_is_sparse, name=name,
ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if transpose_a is None:
transpose_a = False
transpose_a = _execute.make_bool(transpose_a, "transpose_a")
if transpose_b is None:
transpose_b = False
transpose_b = _execute.make_bool(transpose_b, "transpose_b")
if a_is_sparse is None:
a_is_sparse = False
a_is_sparse = _execute.make_bool(a_is_sparse, "a_is_sparse")
if b_is_sparse is None:
b_is_sparse = False
b_is_sparse = _execute.make_bool(b_is_sparse, "b_is_sparse")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SparseMatMul", a=a, b=b, transpose_a=transpose_a,
transpose_b=transpose_b, a_is_sparse=a_is_sparse,
b_is_sparse=b_is_sparse, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("transpose_a", _op._get_attr_bool("transpose_a"), "transpose_b",
_op._get_attr_bool("transpose_b"), "a_is_sparse",
_op._get_attr_bool("a_is_sparse"), "b_is_sparse",
_op._get_attr_bool("b_is_sparse"), "Ta",
_op._get_attr_type("Ta"), "Tb", _op._get_attr_type("Tb"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SparseMatMul", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SparseMatMul = tf_export("raw_ops.SparseMatMul")(_ops.to_raw_op(sparse_mat_mul))
def sparse_mat_mul_eager_fallback(a, b, transpose_a, transpose_b, a_is_sparse, b_is_sparse, name, ctx):
if transpose_a is None:
transpose_a = False
transpose_a = _execute.make_bool(transpose_a, "transpose_a")
if transpose_b is None:
transpose_b = False
transpose_b = _execute.make_bool(transpose_b, "transpose_b")
if a_is_sparse is None:
a_is_sparse = False
a_is_sparse = _execute.make_bool(a_is_sparse, "a_is_sparse")
if b_is_sparse is None:
b_is_sparse = False
b_is_sparse = _execute.make_bool(b_is_sparse, "b_is_sparse")
_attr_Ta, (a,) = _execute.args_to_matching_eager([a], ctx, [_dtypes.float32, _dtypes.bfloat16, ], _dtypes.float32)
_attr_Tb, (b,) = _execute.args_to_matching_eager([b], ctx, [_dtypes.float32, _dtypes.bfloat16, ], _dtypes.float32)
_inputs_flat = [a, b]
_attrs = ("transpose_a", transpose_a, "transpose_b", transpose_b,
"a_is_sparse", a_is_sparse, "b_is_sparse", b_is_sparse, "Ta", _attr_Ta,
"Tb", _attr_Tb)
_result = _execute.execute(b"SparseMatMul", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SparseMatMul", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sparse_segment_mean(data, indices, segment_ids, name=None):
r"""Computes the mean along sparse segments of a tensor.
See `tf.sparse.segment_sum` for usage examples.
Like `SegmentMean`, but `segment_ids` can have rank less than `data`'s first
dimension, selecting a subset of dimension 0, specified by `indices`.
Args:
data: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor. Has same rank as `segment_ids`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor. Values should be sorted and can be repeated.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SparseSegmentMean", name, data, indices, segment_ids)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sparse_segment_mean_eager_fallback(
data, indices, segment_ids, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SparseSegmentMean", data=data, indices=indices,
segment_ids=segment_ids, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tidx",
_op._get_attr_type("Tidx"), "Tsegmentids",
_op._get_attr_type("Tsegmentids"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SparseSegmentMean", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SparseSegmentMean = tf_export("raw_ops.SparseSegmentMean")(_ops.to_raw_op(sparse_segment_mean))
def sparse_segment_mean_eager_fallback(data, indices, segment_ids, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_attr_Tidx, (indices,) = _execute.args_to_matching_eager([indices], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_attr_Tsegmentids, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [data, indices, segment_ids]
_attrs = ("T", _attr_T, "Tidx", _attr_Tidx, "Tsegmentids",
_attr_Tsegmentids)
_result = _execute.execute(b"SparseSegmentMean", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SparseSegmentMean", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sparse_segment_mean_grad(grad, indices, segment_ids, output_dim0, name=None):
r"""Computes gradients for SparseSegmentMean.
Returns tensor "output" with same shape as grad, except for dimension 0 whose
value is output_dim0.
Args:
grad: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
gradient propagated to the SparseSegmentMean op.
indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
indices passed to the corresponding SparseSegmentMean op.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
segment_ids passed to the corresponding SparseSegmentMean op.
output_dim0: A `Tensor` of type `int32`.
dimension 0 of "data" passed to SparseSegmentMean op.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `grad`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SparseSegmentMeanGrad", name, grad, indices, segment_ids,
output_dim0)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sparse_segment_mean_grad_eager_fallback(
grad, indices, segment_ids, output_dim0, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SparseSegmentMeanGrad", grad=grad, indices=indices,
segment_ids=segment_ids,
output_dim0=output_dim0, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tidx",
_op._get_attr_type("Tidx"), "Tsegmentids",
_op._get_attr_type("Tsegmentids"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SparseSegmentMeanGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SparseSegmentMeanGrad = tf_export("raw_ops.SparseSegmentMeanGrad")(_ops.to_raw_op(sparse_segment_mean_grad))
def sparse_segment_mean_grad_eager_fallback(grad, indices, segment_ids, output_dim0, name, ctx):
_attr_T, (grad,) = _execute.args_to_matching_eager([grad], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_attr_Tidx, (indices,) = _execute.args_to_matching_eager([indices], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_attr_Tsegmentids, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
output_dim0 = _ops.convert_to_tensor(output_dim0, _dtypes.int32)
_inputs_flat = [grad, indices, segment_ids, output_dim0]
_attrs = ("T", _attr_T, "Tidx", _attr_Tidx, "Tsegmentids",
_attr_Tsegmentids)
_result = _execute.execute(b"SparseSegmentMeanGrad", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SparseSegmentMeanGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sparse_segment_mean_with_num_segments(data, indices, segment_ids, num_segments, name=None):
r"""Computes the mean along sparse segments of a tensor.
Like `SparseSegmentMean`, but allows missing ids in `segment_ids`. If an id is
missing, the `output` tensor at that position will be zeroed.
Read
[the section on segmentation](https://tensorflow.org/api_docs/python/tf/math#Segmentation)
for an explanation of segments.
Args:
data: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor. Has same rank as `segment_ids`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor. Values should be sorted and can be repeated.
num_segments: A `Tensor`. Must be one of the following types: `int32`, `int64`.
Should equal the number of distinct segment IDs.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SparseSegmentMeanWithNumSegments", name, data, indices,
segment_ids, num_segments)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sparse_segment_mean_with_num_segments_eager_fallback(
data, indices, segment_ids, num_segments, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SparseSegmentMeanWithNumSegments", data=data, indices=indices,
segment_ids=segment_ids,
num_segments=num_segments,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tidx",
_op._get_attr_type("Tidx"), "Tnumsegments",
_op._get_attr_type("Tnumsegments"), "Tsegmentids",
_op._get_attr_type("Tsegmentids"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SparseSegmentMeanWithNumSegments", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SparseSegmentMeanWithNumSegments = tf_export("raw_ops.SparseSegmentMeanWithNumSegments")(_ops.to_raw_op(sparse_segment_mean_with_num_segments))
def sparse_segment_mean_with_num_segments_eager_fallback(data, indices, segment_ids, num_segments, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_attr_Tidx, (indices,) = _execute.args_to_matching_eager([indices], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_attr_Tnumsegments, (num_segments,) = _execute.args_to_matching_eager([num_segments], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_attr_Tsegmentids, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [data, indices, segment_ids, num_segments]
_attrs = ("T", _attr_T, "Tidx", _attr_Tidx, "Tnumsegments",
_attr_Tnumsegments, "Tsegmentids", _attr_Tsegmentids)
_result = _execute.execute(b"SparseSegmentMeanWithNumSegments", 1,
inputs=_inputs_flat, attrs=_attrs, ctx=ctx,
name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SparseSegmentMeanWithNumSegments", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sparse_segment_sqrt_n(data, indices, segment_ids, name=None):
r"""Computes the sum along sparse segments of a tensor divided by the sqrt of N.
N is the size of the segment being reduced.
See `tf.sparse.segment_sum` for usage examples.
Args:
data: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor. Has same rank as `segment_ids`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor. Values should be sorted and can be repeated.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SparseSegmentSqrtN", name, data, indices, segment_ids)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sparse_segment_sqrt_n_eager_fallback(
data, indices, segment_ids, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SparseSegmentSqrtN", data=data, indices=indices,
segment_ids=segment_ids, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tidx",
_op._get_attr_type("Tidx"), "Tsegmentids",
_op._get_attr_type("Tsegmentids"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SparseSegmentSqrtN", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SparseSegmentSqrtN = tf_export("raw_ops.SparseSegmentSqrtN")(_ops.to_raw_op(sparse_segment_sqrt_n))
def sparse_segment_sqrt_n_eager_fallback(data, indices, segment_ids, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_attr_Tidx, (indices,) = _execute.args_to_matching_eager([indices], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_attr_Tsegmentids, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [data, indices, segment_ids]
_attrs = ("T", _attr_T, "Tidx", _attr_Tidx, "Tsegmentids",
_attr_Tsegmentids)
_result = _execute.execute(b"SparseSegmentSqrtN", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SparseSegmentSqrtN", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sparse_segment_sqrt_n_grad(grad, indices, segment_ids, output_dim0, name=None):
r"""Computes gradients for SparseSegmentSqrtN.
Returns tensor "output" with same shape as grad, except for dimension 0 whose
value is output_dim0.
Args:
grad: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
gradient propagated to the SparseSegmentSqrtN op.
indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
indices passed to the corresponding SparseSegmentSqrtN op.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
segment_ids passed to the corresponding SparseSegmentSqrtN op.
output_dim0: A `Tensor` of type `int32`.
dimension 0 of "data" passed to SparseSegmentSqrtN op.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `grad`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SparseSegmentSqrtNGrad", name, grad, indices, segment_ids,
output_dim0)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sparse_segment_sqrt_n_grad_eager_fallback(
grad, indices, segment_ids, output_dim0, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SparseSegmentSqrtNGrad", grad=grad, indices=indices,
segment_ids=segment_ids,
output_dim0=output_dim0, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tidx",
_op._get_attr_type("Tidx"), "Tsegmentids",
_op._get_attr_type("Tsegmentids"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SparseSegmentSqrtNGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SparseSegmentSqrtNGrad = tf_export("raw_ops.SparseSegmentSqrtNGrad")(_ops.to_raw_op(sparse_segment_sqrt_n_grad))
def sparse_segment_sqrt_n_grad_eager_fallback(grad, indices, segment_ids, output_dim0, name, ctx):
_attr_T, (grad,) = _execute.args_to_matching_eager([grad], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_attr_Tidx, (indices,) = _execute.args_to_matching_eager([indices], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_attr_Tsegmentids, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
output_dim0 = _ops.convert_to_tensor(output_dim0, _dtypes.int32)
_inputs_flat = [grad, indices, segment_ids, output_dim0]
_attrs = ("T", _attr_T, "Tidx", _attr_Tidx, "Tsegmentids",
_attr_Tsegmentids)
_result = _execute.execute(b"SparseSegmentSqrtNGrad", 1,
inputs=_inputs_flat, attrs=_attrs, ctx=ctx,
name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SparseSegmentSqrtNGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sparse_segment_sqrt_n_with_num_segments(data, indices, segment_ids, num_segments, name=None):
r"""Computes the sum along sparse segments of a tensor divided by the sqrt of N.
N is the size of the segment being reduced.
Like `SparseSegmentSqrtN`, but allows missing ids in `segment_ids`. If an id is
missing, the `output` tensor at that position will be zeroed.
Read
[the section on segmentation](https://tensorflow.org/api_docs/python/tf/math#Segmentation)
for an explanation of segments.
Args:
data: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor. Has same rank as `segment_ids`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor. Values should be sorted and can be repeated.
num_segments: A `Tensor`. Must be one of the following types: `int32`, `int64`.
Should equal the number of distinct segment IDs.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SparseSegmentSqrtNWithNumSegments", name, data, indices,
segment_ids, num_segments)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sparse_segment_sqrt_n_with_num_segments_eager_fallback(
data, indices, segment_ids, num_segments, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SparseSegmentSqrtNWithNumSegments", data=data, indices=indices,
segment_ids=segment_ids,
num_segments=num_segments,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tidx",
_op._get_attr_type("Tidx"), "Tnumsegments",
_op._get_attr_type("Tnumsegments"), "Tsegmentids",
_op._get_attr_type("Tsegmentids"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SparseSegmentSqrtNWithNumSegments", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SparseSegmentSqrtNWithNumSegments = tf_export("raw_ops.SparseSegmentSqrtNWithNumSegments")(_ops.to_raw_op(sparse_segment_sqrt_n_with_num_segments))
def sparse_segment_sqrt_n_with_num_segments_eager_fallback(data, indices, segment_ids, num_segments, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_attr_Tidx, (indices,) = _execute.args_to_matching_eager([indices], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_attr_Tnumsegments, (num_segments,) = _execute.args_to_matching_eager([num_segments], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_attr_Tsegmentids, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [data, indices, segment_ids, num_segments]
_attrs = ("T", _attr_T, "Tidx", _attr_Tidx, "Tnumsegments",
_attr_Tnumsegments, "Tsegmentids", _attr_Tsegmentids)
_result = _execute.execute(b"SparseSegmentSqrtNWithNumSegments", 1,
inputs=_inputs_flat, attrs=_attrs, ctx=ctx,
name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SparseSegmentSqrtNWithNumSegments", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sparse_segment_sum(data, indices, segment_ids, name=None):
r"""Computes the sum along sparse segments of a tensor.
Read
[the section on segmentation](https://tensorflow.org/api_docs/python/tf/math#Segmentation)
for an explanation of segments.
Like `SegmentSum`, but `segment_ids` can have rank less than `data`'s first
dimension, selecting a subset of dimension 0, specified by `indices`.
For example:
```python
c = tf.constant([[1,2,3,4], [-1,-2,-3,-4], [5,6,7,8]])
# Select two rows, one segment.
tf.sparse_segment_sum(c, tf.constant([0, 1]), tf.constant([0, 0]))
# => [[0 0 0 0]]
# Select two rows, two segment.
tf.sparse_segment_sum(c, tf.constant([0, 1]), tf.constant([0, 1]))
# => [[ 1 2 3 4]
# [-1 -2 -3 -4]]
# Select all rows, two segments.
tf.sparse_segment_sum(c, tf.constant([0, 1, 2]), tf.constant([0, 0, 1]))
# => [[0 0 0 0]
# [5 6 7 8]]
# Which is equivalent to:
tf.segment_sum(c, tf.constant([0, 0, 1]))
```
Args:
data: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `int64`, `bfloat16`, `uint16`, `half`, `uint32`, `uint64`.
indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor. Has same rank as `segment_ids`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor. Values should be sorted and can be repeated.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SparseSegmentSum", name, data, indices, segment_ids)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sparse_segment_sum_eager_fallback(
data, indices, segment_ids, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SparseSegmentSum", data=data, indices=indices,
segment_ids=segment_ids, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tidx",
_op._get_attr_type("Tidx"), "Tsegmentids",
_op._get_attr_type("Tsegmentids"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SparseSegmentSum", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SparseSegmentSum = tf_export("raw_ops.SparseSegmentSum")(_ops.to_raw_op(sparse_segment_sum))
def sparse_segment_sum_eager_fallback(data, indices, segment_ids, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.int64, _dtypes.bfloat16, _dtypes.uint16, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tidx, (indices,) = _execute.args_to_matching_eager([indices], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_attr_Tsegmentids, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [data, indices, segment_ids]
_attrs = ("T", _attr_T, "Tidx", _attr_Tidx, "Tsegmentids",
_attr_Tsegmentids)
_result = _execute.execute(b"SparseSegmentSum", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SparseSegmentSum", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sparse_segment_sum_grad(grad, indices, segment_ids, output_dim0, name=None):
r"""Computes gradients for SparseSegmentSum.
Returns tensor "output" with same shape as grad, except for dimension 0 whose
value is output_dim0.
Args:
grad: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`.
gradient propagated to the SparseSegmentSum op.
indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
indices passed to the corresponding SparseSegmentSum op.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
segment_ids passed to the corresponding SparseSegmentSum op.
output_dim0: A `Tensor` of type `int32`.
dimension 0 of "data" passed to SparseSegmentSum op.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `grad`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SparseSegmentSumGrad", name, grad, indices, segment_ids,
output_dim0)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sparse_segment_sum_grad_eager_fallback(
grad, indices, segment_ids, output_dim0, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SparseSegmentSumGrad", grad=grad, indices=indices,
segment_ids=segment_ids,
output_dim0=output_dim0, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tidx",
_op._get_attr_type("Tidx"), "Tsegmentids",
_op._get_attr_type("Tsegmentids"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SparseSegmentSumGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SparseSegmentSumGrad = tf_export("raw_ops.SparseSegmentSumGrad")(_ops.to_raw_op(sparse_segment_sum_grad))
def sparse_segment_sum_grad_eager_fallback(grad, indices, segment_ids, output_dim0, name, ctx):
_attr_T, (grad,) = _execute.args_to_matching_eager([grad], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
_attr_Tidx, (indices,) = _execute.args_to_matching_eager([indices], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_attr_Tsegmentids, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
output_dim0 = _ops.convert_to_tensor(output_dim0, _dtypes.int32)
_inputs_flat = [grad, indices, segment_ids, output_dim0]
_attrs = ("T", _attr_T, "Tidx", _attr_Tidx, "Tsegmentids",
_attr_Tsegmentids)
_result = _execute.execute(b"SparseSegmentSumGrad", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SparseSegmentSumGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sparse_segment_sum_with_num_segments(data, indices, segment_ids, num_segments, name=None):
r"""Computes the sum along sparse segments of a tensor.
Like `SparseSegmentSum`, but allows missing ids in `segment_ids`. If an id is
missing, the `output` tensor at that position will be zeroed.
Read
[the section on segmentation](https://tensorflow.org/api_docs/python/tf/sparse#Segmentation)
for an explanation of segments.
For example:
```python
c = tf.constant([[1,2,3,4], [-1,-2,-3,-4], [5,6,7,8]])
tf.sparse_segment_sum_with_num_segments(
c, tf.constant([0, 1]), tf.constant([0, 0]), num_segments=3)
# => [[0 0 0 0]
# [0 0 0 0]
# [0 0 0 0]]
tf.sparse_segment_sum_with_num_segments(c,
tf.constant([0, 1]),
tf.constant([0, 2],
num_segments=4))
# => [[ 1 2 3 4]
# [ 0 0 0 0]
# [-1 -2 -3 -4]
# [ 0 0 0 0]]
```
Args:
data: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `int64`, `bfloat16`, `uint16`, `half`, `uint32`, `uint64`.
indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor. Has same rank as `segment_ids`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A 1-D tensor. Values should be sorted and can be repeated.
num_segments: A `Tensor`. Must be one of the following types: `int32`, `int64`.
Should equal the number of distinct segment IDs.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SparseSegmentSumWithNumSegments", name, data, indices,
segment_ids, num_segments)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sparse_segment_sum_with_num_segments_eager_fallback(
data, indices, segment_ids, num_segments, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SparseSegmentSumWithNumSegments", data=data, indices=indices,
segment_ids=segment_ids,
num_segments=num_segments,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tidx",
_op._get_attr_type("Tidx"), "Tnumsegments",
_op._get_attr_type("Tnumsegments"), "Tsegmentids",
_op._get_attr_type("Tsegmentids"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SparseSegmentSumWithNumSegments", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SparseSegmentSumWithNumSegments = tf_export("raw_ops.SparseSegmentSumWithNumSegments")(_ops.to_raw_op(sparse_segment_sum_with_num_segments))
def sparse_segment_sum_with_num_segments_eager_fallback(data, indices, segment_ids, num_segments, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.int64, _dtypes.bfloat16, _dtypes.uint16, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tidx, (indices,) = _execute.args_to_matching_eager([indices], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_attr_Tnumsegments, (num_segments,) = _execute.args_to_matching_eager([num_segments], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_attr_Tsegmentids, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [data, indices, segment_ids, num_segments]
_attrs = ("T", _attr_T, "Tidx", _attr_Tidx, "Tnumsegments",
_attr_Tnumsegments, "Tsegmentids", _attr_Tsegmentids)
_result = _execute.execute(b"SparseSegmentSumWithNumSegments", 1,
inputs=_inputs_flat, attrs=_attrs, ctx=ctx,
name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SparseSegmentSumWithNumSegments", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sqrt(x, name=None):
r"""Computes square root of x element-wise.
I.e., \\(y = \sqrt{x} = x^{1/2}\\).
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Sqrt", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sqrt_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Sqrt", x=x, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Sqrt", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Sqrt = tf_export("raw_ops.Sqrt")(_ops.to_raw_op(sqrt))
def sqrt_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Sqrt", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Sqrt", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sqrt_grad(y, dy, name=None):
r"""Computes the gradient for the sqrt of `x` wrt its input.
Specifically, `grad = dy * 0.5 / y`, where `y = sqrt(x)`, and `dy`
is the corresponding input gradient.
Args:
y: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
dy: A `Tensor`. Must have the same type as `y`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `y`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SqrtGrad", name, y, dy)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sqrt_grad_eager_fallback(
y, dy, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SqrtGrad", y=y, dy=dy, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SqrtGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SqrtGrad = tf_export("raw_ops.SqrtGrad")(_ops.to_raw_op(sqrt_grad))
def sqrt_grad_eager_fallback(y, dy, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([y, dy], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
(y, dy) = _inputs_T
_inputs_flat = [y, dy]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"SqrtGrad", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SqrtGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.square', 'square')
def square(x, name=None):
r"""Computes square of x element-wise.
I.e., \\(y = x * x = x^2\\).
>>> tf.math.square([-2., 0., 3.])
<tf.Tensor: shape=(3,), dtype=float32, numpy=array([4., 0., 9.], dtype=float32)>
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int8`, `int16`, `int32`, `int64`, `uint8`, `uint16`, `uint32`, `uint64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Square", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_square(
(x, name,), None)
if _result is not NotImplemented:
return _result
return square_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
square, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_square(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Square", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
square, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Square", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Square = tf_export("raw_ops.Square")(_ops.to_raw_op(square))
_dispatcher_for_square = square._tf_type_based_dispatcher.Dispatch
def square_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.uint8, _dtypes.uint16, _dtypes.uint32, _dtypes.uint64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Square", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Square", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.squared_difference', v1=['math.squared_difference', 'squared_difference'])
@deprecated_endpoints('squared_difference')
def squared_difference(x, y, name=None):
r"""Returns conj(x - y)(x - y) element-wise.
*NOTE*: `math.squared_difference` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int32`, `int64`, `complex64`, `complex128`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "SquaredDifference", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_squared_difference(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return squared_difference_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
squared_difference, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_squared_difference(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"SquaredDifference", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
squared_difference, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"SquaredDifference", _inputs_flat, _attrs, _result)
_result, = _result
return _result
SquaredDifference = tf_export("raw_ops.SquaredDifference")(_ops.to_raw_op(squared_difference))
_dispatcher_for_squared_difference = squared_difference._tf_type_based_dispatcher.Dispatch
def squared_difference_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"SquaredDifference", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"SquaredDifference", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def sub(x, y, name=None):
r"""Returns x - y element-wise.
*NOTE*: `tf.subtract` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Both input and output have a range `(-inf, inf)`.
Example usages below.
Subtract operation between an array and a scalar:
>>> x = [1, 2, 3, 4, 5]
>>> y = 1
>>> tf.subtract(x, y)
<tf.Tensor: shape=(5,), dtype=int32, numpy=array([0, 1, 2, 3, 4], dtype=int32)>
>>> tf.subtract(y, x)
<tf.Tensor: shape=(5,), dtype=int32,
numpy=array([ 0, -1, -2, -3, -4], dtype=int32)>
Note that binary `-` operator can be used instead:
>>> x = tf.convert_to_tensor([1, 2, 3, 4, 5])
>>> y = tf.convert_to_tensor(1)
>>> x - y
<tf.Tensor: shape=(5,), dtype=int32, numpy=array([0, 1, 2, 3, 4], dtype=int32)>
Subtract operation between an array and a tensor of same shape:
>>> x = [1, 2, 3, 4, 5]
>>> y = tf.constant([5, 4, 3, 2, 1])
>>> tf.subtract(y, x)
<tf.Tensor: shape=(5,), dtype=int32,
numpy=array([ 4, 2, 0, -2, -4], dtype=int32)>
**Warning**: If one of the inputs (`x` or `y`) is a tensor and the other is a
non-tensor, the non-tensor input will adopt (or get casted to) the data type
of the tensor input. This can potentially cause unwanted overflow or underflow
conversion.
For example,
>>> x = tf.constant([1, 2], dtype=tf.int8)
>>> y = [2**8 + 1, 2**8 + 2]
>>> tf.subtract(x, y)
<tf.Tensor: shape=(2,), dtype=int8, numpy=array([0, 0], dtype=int8)>
When subtracting two input values of different shapes, `tf.subtract` follows the
[general broadcasting rules](https://numpy.org/doc/stable/user/basics.broadcasting.html#general-broadcasting-rules)
. The two input array shapes are compared element-wise. Starting with the
trailing dimensions, the two dimensions either have to be equal or one of them
needs to be `1`.
For example,
>>> x = np.ones(6).reshape(2, 3, 1)
>>> y = np.ones(6).reshape(2, 1, 3)
>>> tf.subtract(x, y)
<tf.Tensor: shape=(2, 3, 3), dtype=float64, numpy=
array([[[0., 0., 0.],
[0., 0., 0.],
[0., 0., 0.]],
[[0., 0., 0.],
[0., 0., 0.],
[0., 0., 0.]]])>
Example with inputs of different dimensions:
>>> x = np.ones(6).reshape(2, 3, 1)
>>> y = np.ones(6).reshape(1, 6)
>>> tf.subtract(x, y)
<tf.Tensor: shape=(2, 3, 6), dtype=float64, numpy=
array([[[0., 0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0., 0.]],
[[0., 0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0., 0.]]])>
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `uint8`, `int8`, `uint16`, `int16`, `int32`, `int64`, `complex64`, `complex128`, `uint32`, `uint64`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Sub", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return sub_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Sub", x=x, y=y, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Sub", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Sub = tf_export("raw_ops.Sub")(_ops.to_raw_op(sub))
def sub_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.uint8, _dtypes.int8, _dtypes.uint16, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, _dtypes.uint32, _dtypes.uint64, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Sub", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Sub", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def _sum(input, axis, keep_dims=False, name=None):
r"""Computes the sum of elements across dimensions of a tensor.
Reduces `input` along the dimensions given in `axis`. Unless
`keep_dims` is true, the rank of the tensor is reduced by 1 for each entry in
`axis`. If `keep_dims` is true, the reduced dimensions are
retained with length 1.
Args:
input: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
The tensor to reduce.
axis: A `Tensor`. Must be one of the following types: `int32`, `int64`.
The dimensions to reduce. Must be in the range
`[-rank(input), rank(input))`.
keep_dims: An optional `bool`. Defaults to `False`.
If true, retain reduced dimensions with length 1.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `input`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Sum", name, input, axis, "keep_dims", keep_dims)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return _sum_eager_fallback(
input, axis, keep_dims=keep_dims, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Sum", input=input, reduction_indices=axis, keep_dims=keep_dims,
name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("keep_dims", _op._get_attr_bool("keep_dims"), "T",
_op._get_attr_type("T"), "Tidx", _op._get_attr_type("Tidx"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Sum", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Sum = tf_export("raw_ops.Sum")(_ops.to_raw_op(_sum))
def _sum_eager_fallback(input, axis, keep_dims, name, ctx):
if keep_dims is None:
keep_dims = False
keep_dims = _execute.make_bool(keep_dims, "keep_dims")
_attr_T, (input,) = _execute.args_to_matching_eager([input], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tidx, (axis,) = _execute.args_to_matching_eager([axis], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [input, axis]
_attrs = ("keep_dims", keep_dims, "T", _attr_T, "Tidx", _attr_Tidx)
_result = _execute.execute(b"Sum", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Sum", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.tan', 'tan')
def tan(x, name=None):
r"""Computes tan of x element-wise.
Given an input tensor, this function computes tangent of every
element in the tensor. Input range is `(-inf, inf)` and
output range is `(-inf, inf)`. If input lies outside the boundary, `nan`
is returned.
```python
x = tf.constant([-float("inf"), -9, -0.5, 1, 1.2, 200, 10000, float("inf")])
tf.math.tan(x) ==> [nan 0.45231566 -0.5463025 1.5574077 2.572152 -1.7925274 0.32097113 nan]
```
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `int8`, `int16`, `int32`, `int64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Tan", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_tan(
(x, name,), None)
if _result is not NotImplemented:
return _result
return tan_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
tan, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_tan(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Tan", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
tan, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Tan", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Tan = tf_export("raw_ops.Tan")(_ops.to_raw_op(tan))
_dispatcher_for_tan = tan._tf_type_based_dispatcher.Dispatch
def tan_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.int8, _dtypes.int16, _dtypes.int32, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Tan", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Tan", _inputs_flat, _attrs, _result)
_result, = _result
return _result
[文档]@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.tanh', 'nn.tanh', 'tanh')
def tanh(x, name=None):
r"""Computes hyperbolic tangent of `x` element-wise.
Given an input tensor, this function computes hyperbolic tangent of every
element in the tensor. Input range is `[-inf, inf]` and
output range is `[-1,1]`.
>>> x = tf.constant([-float("inf"), -5, -0.5, 1, 1.2, 2, 3, float("inf")])
>>> tf.math.tanh(x)
<tf.Tensor: shape=(8,), dtype=float32, numpy=
array([-1.0, -0.99990916, -0.46211717, 0.7615942 , 0.8336547 ,
0.9640276 , 0.9950547 , 1.0], dtype=float32)>
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Tanh", name, x)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_tanh(
(x, name,), None)
if _result is not NotImplemented:
return _result
return tanh_eager_fallback(
x, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
tanh, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_tanh(
(x, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Tanh", x=x, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
tanh, (), dict(x=x, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Tanh", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Tanh = tf_export("raw_ops.Tanh")(_ops.to_raw_op(tanh))
_dispatcher_for_tanh = tanh._tf_type_based_dispatcher.Dispatch
def tanh_eager_fallback(x, name, ctx):
_attr_T, (x,) = _execute.args_to_matching_eager([x], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
_inputs_flat = [x]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Tanh", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Tanh", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def tanh_grad(y, dy, name=None):
r"""Computes the gradient for the tanh of `x` wrt its input.
Specifically, `grad = dy * (1 - y*y)`, where `y = tanh(x)`, and `dy`
is the corresponding input gradient.
Args:
y: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `complex64`, `complex128`.
dy: A `Tensor`. Must have the same type as `y`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `y`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "TanhGrad", name, y, dy)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return tanh_grad_eager_fallback(
y, dy, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"TanhGrad", y=y, dy=dy, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"TanhGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
TanhGrad = tf_export("raw_ops.TanhGrad")(_ops.to_raw_op(tanh_grad))
def tanh_grad_eager_fallback(y, dy, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([y, dy], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
(y, dy) = _inputs_T
_inputs_flat = [y, dy]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"TanhGrad", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"TanhGrad", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('truncatediv')
def truncate_div(x, y, name=None):
r"""Returns x / y element-wise for integer types.
Truncation designates that negative numbers will round fractional quantities
toward zero. I.e. -7 / 5 = -1. This matches C semantics but it is different
than Python semantics. See `FloorDiv` for a division function that matches
Python Semantics.
*NOTE*: `truncatediv` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Args:
x: A `Tensor`. Must be one of the following types: `bfloat16`, `half`, `float32`, `float64`, `uint8`, `int8`, `uint16`, `int16`, `int32`, `uint32`, `uint64`, `int64`, `complex64`, `complex128`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "TruncateDiv", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_truncate_div(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return truncate_div_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
truncate_div, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_truncate_div(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"TruncateDiv", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
truncate_div, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"TruncateDiv", _inputs_flat, _attrs, _result)
_result, = _result
return _result
TruncateDiv = tf_export("raw_ops.TruncateDiv")(_ops.to_raw_op(truncate_div))
_dispatcher_for_truncate_div = truncate_div._tf_type_based_dispatcher.Dispatch
def truncate_div_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.uint8, _dtypes.int8, _dtypes.uint16, _dtypes.int16, _dtypes.int32, _dtypes.uint32, _dtypes.uint64, _dtypes.int64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"TruncateDiv", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"TruncateDiv", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('truncatemod')
def truncate_mod(x, y, name=None):
r"""Returns element-wise remainder of division. This emulates C semantics in that
the result here is consistent with a truncating divide. E.g. `truncate(x / y) *
y + truncate_mod(x, y) = x`.
*NOTE*: `truncatemod` supports broadcasting. More about broadcasting
[here](http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html)
Args:
x: A `Tensor`. Must be one of the following types: `int32`, `int64`, `bfloat16`, `half`, `float32`, `float64`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "TruncateMod", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_truncate_mod(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return truncate_mod_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
truncate_mod, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_truncate_mod(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"TruncateMod", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
truncate_mod, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"TruncateMod", _inputs_flat, _attrs, _result)
_result, = _result
return _result
TruncateMod = tf_export("raw_ops.TruncateMod")(_ops.to_raw_op(truncate_mod))
_dispatcher_for_truncate_mod = truncate_mod._tf_type_based_dispatcher.Dispatch
def truncate_mod_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.int32, _dtypes.int64, _dtypes.bfloat16, _dtypes.half, _dtypes.float32, _dtypes.float64, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"TruncateMod", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"TruncateMod", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.unsorted_segment_max', v1=['math.unsorted_segment_max', 'unsorted_segment_max'])
@deprecated_endpoints('unsorted_segment_max')
def unsorted_segment_max(data, segment_ids, num_segments, name=None):
r"""Computes the maximum along segments of a tensor.
Read
[the section on segmentation](https://tensorflow.org/api_docs/python/tf/math#Segmentation)
for an explanation of segments.
This operator is similar to `tf.math.unsorted_segment_sum`,
Instead of computing the sum over segments, it computes the maximum such that:
\\(output_i = \max_{j...} data[j...]\\) where max is over tuples `j...` such
that `segment_ids[j...] == i`.
If the maximum is empty for a given segment ID `i`, it outputs the smallest
possible value for the specific numeric type,
`output[i] = numeric_limits<T>::lowest()`.
If the given segment ID `i` is negative, then the corresponding value is
dropped, and will not be included in the result.
Caution: On CPU, values in `segment_ids` are always validated to be less than
`num_segments`, and an error is thrown for out-of-bound indices. On GPU, this
does not throw an error for out-of-bound indices. On Gpu, out-of-bound indices
result in safe but unspecified behavior, which may include ignoring
out-of-bound indices or outputting a tensor with a 0 stored in the first
dimension of its shape if `num_segments` is 0.
<div style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;">
<img style="width:100%" src="https://www.tensorflow.org/images/UnsortedSegmentMax.png" alt>
</div>
For example:
>>> c = tf.constant([[1,2,3,4], [5,6,7,8], [4,3,2,1]])
>>> tf.math.unsorted_segment_max(c, tf.constant([0, 1, 0]), num_segments=2).numpy()
array([[4, 3, 3, 4],
[5, 6, 7, 8]], dtype=int32)
Args:
data: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `int64`, `bfloat16`, `uint16`, `half`, `uint32`, `uint64`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A tensor whose shape is a prefix of `data.shape`.
The values must be less than `num_segments`.
Caution: The values are always validated to be in range on CPU, never validated
on GPU.
num_segments: A `Tensor`. Must be one of the following types: `int32`, `int64`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "UnsortedSegmentMax", name, data, segment_ids, num_segments)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_unsorted_segment_max(
(data, segment_ids, num_segments, name,), None)
if _result is not NotImplemented:
return _result
return unsorted_segment_max_eager_fallback(
data, segment_ids, num_segments, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
unsorted_segment_max, (), dict(data=data, segment_ids=segment_ids,
num_segments=num_segments,
name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_unsorted_segment_max(
(data, segment_ids, num_segments, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"UnsortedSegmentMax", data=data, segment_ids=segment_ids,
num_segments=num_segments, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
unsorted_segment_max, (), dict(data=data, segment_ids=segment_ids,
num_segments=num_segments, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tindices",
_op._get_attr_type("Tindices"), "Tnumsegments",
_op._get_attr_type("Tnumsegments"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"UnsortedSegmentMax", _inputs_flat, _attrs, _result)
_result, = _result
return _result
UnsortedSegmentMax = tf_export("raw_ops.UnsortedSegmentMax")(_ops.to_raw_op(unsorted_segment_max))
_dispatcher_for_unsorted_segment_max = unsorted_segment_max._tf_type_based_dispatcher.Dispatch
def unsorted_segment_max_eager_fallback(data, segment_ids, num_segments, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.int64, _dtypes.bfloat16, _dtypes.uint16, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tindices, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ])
_attr_Tnumsegments, (num_segments,) = _execute.args_to_matching_eager([num_segments], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [data, segment_ids, num_segments]
_attrs = ("T", _attr_T, "Tindices", _attr_Tindices, "Tnumsegments",
_attr_Tnumsegments)
_result = _execute.execute(b"UnsortedSegmentMax", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"UnsortedSegmentMax", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.unsorted_segment_min', v1=['math.unsorted_segment_min', 'unsorted_segment_min'])
@deprecated_endpoints('unsorted_segment_min')
def unsorted_segment_min(data, segment_ids, num_segments, name=None):
r"""Computes the minimum along segments of a tensor.
Read
[the section on segmentation](https://tensorflow.org/api_docs/python/tf/math#Segmentation)
for an explanation of segments.
This operator is similar to `tf.math.unsorted_segment_sum`,
Instead of computing the sum over segments, it computes the minimum such that:
\\(output_i = \min_{j...} data_[j...]\\) where min is over tuples `j...` such
that `segment_ids[j...] == i`.
If the minimum is empty for a given segment ID `i`, it outputs the largest
possible value for the specific numeric type,
`output[i] = numeric_limits<T>::max()`.
For example:
>>> c = tf.constant([[1,2,3,4], [5,6,7,8], [4,3,2,1]])
>>> tf.math.unsorted_segment_min(c, tf.constant([0, 1, 0]), num_segments=2).numpy()
array([[1, 2, 2, 1],
[5, 6, 7, 8]], dtype=int32)
If the given segment ID `i` is negative, then the corresponding value is
dropped, and will not be included in the result.
Caution: On CPU, values in `segment_ids` are always validated to be less than
`num_segments`, and an error is thrown for out-of-bound indices. On GPU, this
does not throw an error for out-of-bound indices. On Gpu, out-of-bound indices
result in safe but unspecified behavior, which may include ignoring
out-of-bound indices or outputting a tensor with a 0 stored in the first
dimension of its shape if `num_segments` is 0.
Args:
data: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `int64`, `bfloat16`, `uint16`, `half`, `uint32`, `uint64`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A tensor whose shape is a prefix of `data.shape`.
The values must be less than `num_segments`.
Caution: The values are always validated to be in range on CPU, never validated
on GPU.
num_segments: A `Tensor`. Must be one of the following types: `int32`, `int64`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "UnsortedSegmentMin", name, data, segment_ids, num_segments)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_unsorted_segment_min(
(data, segment_ids, num_segments, name,), None)
if _result is not NotImplemented:
return _result
return unsorted_segment_min_eager_fallback(
data, segment_ids, num_segments, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
unsorted_segment_min, (), dict(data=data, segment_ids=segment_ids,
num_segments=num_segments,
name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_unsorted_segment_min(
(data, segment_ids, num_segments, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"UnsortedSegmentMin", data=data, segment_ids=segment_ids,
num_segments=num_segments, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
unsorted_segment_min, (), dict(data=data, segment_ids=segment_ids,
num_segments=num_segments, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tindices",
_op._get_attr_type("Tindices"), "Tnumsegments",
_op._get_attr_type("Tnumsegments"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"UnsortedSegmentMin", _inputs_flat, _attrs, _result)
_result, = _result
return _result
UnsortedSegmentMin = tf_export("raw_ops.UnsortedSegmentMin")(_ops.to_raw_op(unsorted_segment_min))
_dispatcher_for_unsorted_segment_min = unsorted_segment_min._tf_type_based_dispatcher.Dispatch
def unsorted_segment_min_eager_fallback(data, segment_ids, num_segments, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.int64, _dtypes.bfloat16, _dtypes.uint16, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tindices, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ])
_attr_Tnumsegments, (num_segments,) = _execute.args_to_matching_eager([num_segments], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [data, segment_ids, num_segments]
_attrs = ("T", _attr_T, "Tindices", _attr_Tindices, "Tnumsegments",
_attr_Tnumsegments)
_result = _execute.execute(b"UnsortedSegmentMin", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"UnsortedSegmentMin", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.unsorted_segment_prod', v1=['math.unsorted_segment_prod', 'unsorted_segment_prod'])
@deprecated_endpoints('unsorted_segment_prod')
def unsorted_segment_prod(data, segment_ids, num_segments, name=None):
r"""Computes the product along segments of a tensor.
Read
[the section on segmentation](https://tensorflow.org/api_docs/python/tf/math#Segmentation)
for an explanation of segments.
This operator is similar to `tf.math.unsorted_segment_sum`,
Instead of computing the sum over segments, it computes the product of all
entries belonging to a segment such that:
\\(output_i = \prod_{j...} data[j...]\\) where the product is over tuples
`j...` such that `segment_ids[j...] == i`.
For example:
>>> c = tf.constant([[1,2,3,4], [5,6,7,8], [4,3,2,1]])
>>> tf.math.unsorted_segment_prod(c, tf.constant([0, 1, 0]), num_segments=2).numpy()
array([[4, 6, 6, 4],
[5, 6, 7, 8]], dtype=int32)
If there is no entry for a given segment ID `i`, it outputs 1.
If the given segment ID `i` is negative, then the corresponding value is
dropped, and will not be included in the result.
Caution: On CPU, values in `segment_ids` are always validated to be less than
`num_segments`, and an error is thrown for out-of-bound indices. On GPU, this
does not throw an error for out-of-bound indices. On Gpu, out-of-bound indices
result in safe but unspecified behavior, which may include ignoring
out-of-bound indices or outputting a tensor with a 0 stored in the first
dimension of its shape if `num_segments` is 0.
Args:
data: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A tensor whose shape is a prefix of `data.shape`.
The values must be less than `num_segments`.
Caution: The values are always validated to be in range on CPU, never validated
on GPU.
num_segments: A `Tensor`. Must be one of the following types: `int32`, `int64`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "UnsortedSegmentProd", name, data, segment_ids, num_segments)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_unsorted_segment_prod(
(data, segment_ids, num_segments, name,), None)
if _result is not NotImplemented:
return _result
return unsorted_segment_prod_eager_fallback(
data, segment_ids, num_segments, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
unsorted_segment_prod, (), dict(data=data,
segment_ids=segment_ids,
num_segments=num_segments,
name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_unsorted_segment_prod(
(data, segment_ids, num_segments, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"UnsortedSegmentProd", data=data, segment_ids=segment_ids,
num_segments=num_segments, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
unsorted_segment_prod, (), dict(data=data, segment_ids=segment_ids,
num_segments=num_segments,
name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tindices",
_op._get_attr_type("Tindices"), "Tnumsegments",
_op._get_attr_type("Tnumsegments"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"UnsortedSegmentProd", _inputs_flat, _attrs, _result)
_result, = _result
return _result
UnsortedSegmentProd = tf_export("raw_ops.UnsortedSegmentProd")(_ops.to_raw_op(unsorted_segment_prod))
_dispatcher_for_unsorted_segment_prod = unsorted_segment_prod._tf_type_based_dispatcher.Dispatch
def unsorted_segment_prod_eager_fallback(data, segment_ids, num_segments, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tindices, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ])
_attr_Tnumsegments, (num_segments,) = _execute.args_to_matching_eager([num_segments], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [data, segment_ids, num_segments]
_attrs = ("T", _attr_T, "Tindices", _attr_Tindices, "Tnumsegments",
_attr_Tnumsegments)
_result = _execute.execute(b"UnsortedSegmentProd", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"UnsortedSegmentProd", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.unsorted_segment_sum', v1=['math.unsorted_segment_sum', 'unsorted_segment_sum'])
@deprecated_endpoints('unsorted_segment_sum')
def unsorted_segment_sum(data, segment_ids, num_segments, name=None):
r"""Computes the sum along segments of a tensor.
Read
[the section on segmentation](https://tensorflow.org/api_docs/python/tf/math#Segmentation)
for an explanation of segments.
Computes a tensor such that
\\(output[i] = \sum_{j...} data[j...]\\) where the sum is over tuples `j...` such
that `segment_ids[j...] == i`. Unlike `SegmentSum`, `segment_ids`
need not be sorted and need not cover all values in the full
range of valid values.
If the sum is empty for a given segment ID `i`, `output[i] = 0`.
If the given segment ID `i` is negative, the value is dropped and will not be
added to the sum of the segment.
`num_segments` should equal the number of distinct segment IDs.
Caution: On CPU, values in `segment_ids` are always validated to be less than
`num_segments`, and an error is thrown for out-of-bound indices. On GPU, this
does not throw an error for out-of-bound indices. On Gpu, out-of-bound indices
result in safe but unspecified behavior, which may include ignoring
out-of-bound indices or outputting a tensor with a 0 stored in the first
dimension of its shape if `num_segments` is 0.
<div style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;">
<img style="width:100%" src="https://www.tensorflow.org/images/UnsortedSegmentSum.png" alt>
</div>
>>> c = [[1,2,3,4], [5,6,7,8], [4,3,2,1]]
>>> tf.math.unsorted_segment_sum(c, [0, 1, 0], num_segments=2).numpy()
array([[5, 5, 5, 5],
[5, 6, 7, 8]], dtype=int32)
Args:
data: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
segment_ids: A `Tensor`. Must be one of the following types: `int32`, `int64`.
A tensor whose shape is a prefix of `data.shape`.
The values must be less than `num_segments`.
Caution: The values are always validated to be in range on CPU, never validated
on GPU.
num_segments: A `Tensor`. Must be one of the following types: `int32`, `int64`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `data`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "UnsortedSegmentSum", name, data, segment_ids, num_segments)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_unsorted_segment_sum(
(data, segment_ids, num_segments, name,), None)
if _result is not NotImplemented:
return _result
return unsorted_segment_sum_eager_fallback(
data, segment_ids, num_segments, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
unsorted_segment_sum, (), dict(data=data, segment_ids=segment_ids,
num_segments=num_segments,
name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_unsorted_segment_sum(
(data, segment_ids, num_segments, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"UnsortedSegmentSum", data=data, segment_ids=segment_ids,
num_segments=num_segments, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
unsorted_segment_sum, (), dict(data=data, segment_ids=segment_ids,
num_segments=num_segments, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"), "Tindices",
_op._get_attr_type("Tindices"), "Tnumsegments",
_op._get_attr_type("Tnumsegments"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"UnsortedSegmentSum", _inputs_flat, _attrs, _result)
_result, = _result
return _result
UnsortedSegmentSum = tf_export("raw_ops.UnsortedSegmentSum")(_ops.to_raw_op(unsorted_segment_sum))
_dispatcher_for_unsorted_segment_sum = unsorted_segment_sum._tf_type_based_dispatcher.Dispatch
def unsorted_segment_sum_eager_fallback(data, segment_ids, num_segments, name, ctx):
_attr_T, (data,) = _execute.args_to_matching_eager([data], ctx, [_dtypes.float32, _dtypes.float64, _dtypes.int32, _dtypes.uint8, _dtypes.int16, _dtypes.int8, _dtypes.complex64, _dtypes.int64, _dtypes.qint8, _dtypes.quint8, _dtypes.qint32, _dtypes.bfloat16, _dtypes.uint16, _dtypes.complex128, _dtypes.half, _dtypes.uint32, _dtypes.uint64, ])
_attr_Tindices, (segment_ids,) = _execute.args_to_matching_eager([segment_ids], ctx, [_dtypes.int32, _dtypes.int64, ])
_attr_Tnumsegments, (num_segments,) = _execute.args_to_matching_eager([num_segments], ctx, [_dtypes.int32, _dtypes.int64, ], _dtypes.int32)
_inputs_flat = [data, segment_ids, num_segments]
_attrs = ("T", _attr_T, "Tindices", _attr_Tindices, "Tnumsegments",
_attr_Tnumsegments)
_result = _execute.execute(b"UnsortedSegmentSum", 1, inputs=_inputs_flat,
attrs=_attrs, ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"UnsortedSegmentSum", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.xdivy')
def xdivy(x, y, name=None):
r"""Returns 0 if x == 0, and x / y otherwise, elementwise.
Args:
x: A `Tensor`. Must be one of the following types: `half`, `float32`, `float64`, `complex64`, `complex128`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Xdivy", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_xdivy(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return xdivy_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
xdivy, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_xdivy(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Xdivy", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
xdivy, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Xdivy", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Xdivy = tf_export("raw_ops.Xdivy")(_ops.to_raw_op(xdivy))
_dispatcher_for_xdivy = xdivy._tf_type_based_dispatcher.Dispatch
def xdivy_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Xdivy", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Xdivy", _inputs_flat, _attrs, _result)
_result, = _result
return _result
def xlog1py(x, y, name=None):
r"""Returns 0 if x == 0, and x * log1p(y) otherwise, elementwise.
Args:
x: A `Tensor`. Must be one of the following types: `half`, `float32`, `float64`, `complex64`, `complex128`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Xlog1py", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
return xlog1py_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
# Add nodes to the TensorFlow graph.
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Xlog1py", x=x, y=y, name=name)
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Xlog1py", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Xlog1py = tf_export("raw_ops.Xlog1py")(_ops.to_raw_op(xlog1py))
def xlog1py_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Xlog1py", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Xlog1py", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.xlogy')
def xlogy(x, y, name=None):
r"""Returns 0 if x == 0, and x * log(y) otherwise, elementwise.
Args:
x: A `Tensor`. Must be one of the following types: `half`, `float32`, `float64`, `complex64`, `complex128`.
y: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Xlogy", name, x, y)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_xlogy(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
return xlogy_eager_fallback(
x, y, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
xlogy, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_xlogy(
(x, y, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Xlogy", x=x, y=y, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
xlogy, (), dict(x=x, y=y, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Xlogy", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Xlogy = tf_export("raw_ops.Xlogy")(_ops.to_raw_op(xlogy))
_dispatcher_for_xlogy = xlogy._tf_type_based_dispatcher.Dispatch
def xlogy_eager_fallback(x, y, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, y], ctx, [_dtypes.half, _dtypes.float32, _dtypes.float64, _dtypes.complex64, _dtypes.complex128, ])
(x, y) = _inputs_T
_inputs_flat = [x, y]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Xlogy", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Xlogy", _inputs_flat, _attrs, _result)
_result, = _result
return _result
@_dispatch.add_fallback_dispatch_list
@_dispatch.add_type_based_api_dispatcher
@tf_export('math.zeta', v1=['math.zeta', 'zeta'])
@deprecated_endpoints('zeta')
def zeta(x, q, name=None):
r"""Compute the Hurwitz zeta function \\(\zeta(x, q)\\).
The Hurwitz zeta function is defined as:
\\(\zeta(x, q) = \sum_{n=0}^{\infty} (q + n)^{-x}\\)
Args:
x: A `Tensor`. Must be one of the following types: `float32`, `float64`.
q: A `Tensor`. Must have the same type as `x`.
name: A name for the operation (optional).
Returns:
A `Tensor`. Has the same type as `x`.
"""
_ctx = _context._context or _context.context()
tld = _ctx._thread_local_data
if tld.is_eager:
try:
_result = pywrap_tfe.TFE_Py_FastPathExecute(
_ctx, "Zeta", name, x, q)
return _result
except _core._NotOkStatusException as e:
_ops.raise_from_not_ok_status(e, name)
except _core._FallbackException:
pass
try:
_result = _dispatcher_for_zeta(
(x, q, name,), None)
if _result is not NotImplemented:
return _result
return zeta_eager_fallback(
x, q, name=name, ctx=_ctx)
except _core._SymbolicException:
pass # Add nodes to the TensorFlow graph.
except (TypeError, ValueError):
_result = _dispatch.dispatch(
zeta, (), dict(x=x, q=q, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
else:
_result = _dispatcher_for_zeta(
(x, q, name,), None)
if _result is not NotImplemented:
return _result
# Add nodes to the TensorFlow graph.
try:
_, _, _op, _outputs = _op_def_library._apply_op_helper(
"Zeta", x=x, q=q, name=name)
except (TypeError, ValueError):
_result = _dispatch.dispatch(
zeta, (), dict(x=x, q=q, name=name)
)
if _result is not _dispatch.OpDispatcher.NOT_SUPPORTED:
return _result
raise
_result = _outputs[:]
if _execute.must_record_gradient():
_attrs = ("T", _op._get_attr_type("T"))
_inputs_flat = _op.inputs
_execute.record_gradient(
"Zeta", _inputs_flat, _attrs, _result)
_result, = _result
return _result
Zeta = tf_export("raw_ops.Zeta")(_ops.to_raw_op(zeta))
_dispatcher_for_zeta = zeta._tf_type_based_dispatcher.Dispatch
def zeta_eager_fallback(x, q, name, ctx):
_attr_T, _inputs_T = _execute.args_to_matching_eager([x, q], ctx, [_dtypes.float32, _dtypes.float64, ])
(x, q) = _inputs_T
_inputs_flat = [x, q]
_attrs = ("T", _attr_T)
_result = _execute.execute(b"Zeta", 1, inputs=_inputs_flat, attrs=_attrs,
ctx=ctx, name=name)
if _execute.must_record_gradient():
_execute.record_gradient(
"Zeta", _inputs_flat, _attrs, _result)
_result, = _result
return _result