If necessary, flatten output.

If `persistent` is true, GradientTape will not eagerly delete backward functions (and hence the tensors they keep alive). Instead, everything is delet

Returns true iff two operands are compatible according to the broadcasting rule.

Returns the encapculated llvm::Function.

If not nullptr, the kernel invoke functions defined in the library. E.g., CHECK_NOTNULL(function_library())->Run("Foo", ...).

Returns a tensor that evaluates to `True` if this optional has a value. Args: name: (Optional.) A name for the created operation. Retu

Return the number of ops in the model.

Returns a builder to create a new {@link Operation}. @param type of the Operation (i.e., identifies the computation to be performed) @param name to r

Register this dispatcher as a handler for `op`. Args: op: Python function: the TensorFlow operation that should be handled. Must ha

Execute the graph fragments necessary to compute all requested fetches. <p><b>WARNING:</b> The caller assumes ownership of all returned {@link Tensor

Test the model on a single batch of samples. Arguments: x: Input data. It could be: - A Numpy array (or array-like), or a list

Fill entries with keys [start,limit).

Return true iff the rank and all of the dimensions are well defined TODO(irving): Rename to is_fully_defined now that it's fast.

Pushes one entry while maintaining approximation error invariants.

Sets tensor to dynamic.

Verifies that HloEvaluator evaluates a HLO instruction that performs element-wise addition with 2 operands.

Creates a `Sum` instance. Args: name: (Optional) string name of the metric instance. dtype: (Optional) data type of the metric result

Get the containing `tf.distribute.Strategy`. This should not generally be needed except when creating a new `ReplicaContext` and to validate

Adds metric tensor to the layer. Args: value: Metric tensor. aggregation: Sample-wise metric reduction function. If `aggregation=None

Returns a context manager that makes this object the default session. Use with the `with` keyword to specify that calls to `tf.Operation.run`

Returns this shape as a `TensorShapeProto`.

Clients can manipulate the underlying NodeDef via this accessor.

Reduces the input dataset to a single element. The transformation calls `reduce_func` successively on every element of the input dataset unti

Reverse a tensor along the specified axes. Arguments: x: Tensor to reverse. axes: Integer or iterable of integers. Axes to re

Iterates over the time dimension of a tensor. Arguments: step_function: RNN step function. Args; input; Tensor with s

Element-wise tanh. Arguments: x: A tensor or variable. Returns: A tensor.

If `must_be_true()` is true this SymbolPredicate represents the proposition "tensor_id() is live and evaluates to true". If `must_be_true()` is false

static */

Returns a list of `Operation`s that consume this tensor. Returns: A list of `Operation`s.

Covariance. Covariance is (possibly) defined only for non-scalar-event distributions. For example, for a length-`k`, vector-valued distribut

Reshape the input to the original shape. This is the inverse of encode. Encode must have been called at least once prior to this method being

Filters this dataset according to `predicate`. ```python d = tf.data.Dataset.from_tensor_slices([1, 2, 3]) d = d.filter(lambda x: x < 3)