Bitcasts a tensor from one type to another without copying data.
Given a tensor `input`, this operation returns a tensor that has the same buffer data as `input` with datatype `type`.
If the input datatype `T` is larger than the output datatype `type` then the shape changes from [...] to [..., sizeof(`T`)/sizeof(`type`)].
If `T` is smaller than `type`, the operator requires that the rightmost dimension be equal to sizeof(`type`)/sizeof(`T`). The shape then goes from [..., sizeof(`type`)/sizeof(`T`)] to [...].
tf.bitcast() and tf.cast() work differently when real dtype is casted as a complex dtype (e.g. tf.complex64 or tf.complex128) as tf.cast() make imaginary part 0 while tf.bitcast() gives module error. For example,
Example 1:
>>> a = [1., 2., 3.] >>> equality_bitcast = tf.bitcast(a, tf.complex128) Traceback (most recent call last): ... InvalidArgumentError: Cannot bitcast from 1 to 18 [Op:Bitcast] >>> equality_cast = tf.cast(a, tf.complex128) >>> print(equality_cast) tf.Tensor([1.+0.j 2.+0.j 3.+0.j], shape=(3,), dtype=complex128)
Example 2:
>>> tf.bitcast(tf.constant(0xffffffff, dtype=tf.uint32), tf.uint8)
Example 3:
>>> x = [1., 2., 3.]
>>> y = [0., 2., 3.]
>>> equality= tf.equal(x,y)
>>> equality_cast = tf.cast(equality,tf.float32)
>>> equality_bitcast = tf.bitcast(equality_cast,tf.uint8)
>>> print(equality)
tf.Tensor([False True True], shape=(3,), dtype=bool)
>>> print(equality_cast)
tf.Tensor([0. 1. 1.], shape=(3,), dtype=float32)
>>> print(equality_bitcast)
tf.Tensor(
[[ 0 0 0 0]
[ 0 0 128 63]
[ 0 0 128 63]], shape=(3, 4), dtype=uint8)
NOTE: Bitcast is implemented as a low-level cast, so machines with different
endian orderings will give different results. A copy from input buffer to output
buffer is made on BE machines when types are of different sizes in order to get
the same casting results as on LE machines.
Public Methods
Output<U> |
asOutput()
Returns the symbolic handle of a tensor.
|
static <U, T> Bitcast<U> | |
Output<U> |
output()
|
Inherited Methods
Public Methods
public Output<U> asOutput ()
Returns the symbolic handle of a tensor.
Inputs to TensorFlow operations are outputs of another TensorFlow operation. This method is used to obtain a symbolic handle that represents the computation of the input.
public static Bitcast<U> create (Scope scope, Operand<T> input, Class<U> type)
Factory method to create a class wrapping a new Bitcast operation.
Parameters
scope | current scope |
---|
Returns
- a new instance of Bitcast