TF2 tensor generators

Generators

The following APIs generate new tensors.

Generator	Main alias of	Level	Description	Returns Tensor?
tf.Tensor		1	A tf.Tensor represents a multidimensional array of elements.	TRUE
tf.SparseTensor	tf.sparse.SparseTensor	1	Represents a sparse tensor.	TRUE
tf.ones		1	Creates a tensor with all elements set to one (1).	TRUE
tf.fill		1	Creates a tensor filled with a scalar value.	TRUE
tf.zeros		1	Creates a tensor with all elements set to zero.	TRUE
tf.zeros_like		1	Creates a tensor with all elements set to zero.	TRUE
tf.one_hot		1	Returns a one-hot tensor.	TRUE
tf.eye		1	Construct an identity matrix, or a batch of matrices.	TRUE
tf.Variable		1	A variable maintains shared, persistent state manipulated by a program.	FALSE
tf.constant		1	Creates a constant tensor from a tensor-like object.	TRUE
tf.convert_to_tensor		1	Converts the given value to a Tensor.	TRUE
tf.range		1	Creates a sequence of numbers.	TRUE
tf.sparse.SparseTensor		2	Represents a sparse tensor.	TRUE
tf.linalg.eye	tf.eye	2	Construct an identity matrix, or a batch of matrices.	TRUE
tf.keras.Input		2	A Keras tensor is a symbolic tensor-like object, which we augment with certain attributes that allow us to build a Keras model just by knowing the inputs and outputs of the model.	FALSE
tf.ragged.constant		2	Constructs a constant RaggedTensor from a nested Python list.	TRUE
tf.ragged.range		2	Returns a RaggedTensor containing the specified sequences of numbers.	TRUE
tf.RaggedTensor.from_nested_row_lengths		2	Creates a RaggedTensor from a nested list of row_lengths tensors.	TRUE
tf.RaggedTensor.from_nested_row_splits		2	Creates a RaggedTensor from a nested list of row_splits tensors.	TRUE
tf.RaggedTensor.from_nested_value_rowids		2	Creates a RaggedTensor from a nested list of value_rowids tensors.	TRUE
tf.RaggedTensor.from_row_lengths		2	Creates a RaggedTensor with rows partitioned by row_lengths.	TRUE
tf.RaggedTensor.from_row_limits		2	Creates a RaggedTensor with rows partitioned by row_limits.	TRUE
tf.RaggedTensor.from_row_splits		2	Creates a RaggedTensor with rows partitioned by row_splits.	TRUE
tf.RaggedTensor.from_row_starts		2	Creates a RaggedTensor with rows partitioned by row_starts.	TRUE
tf.RaggedTensor.from_value_rowids		2	Creates a RaggedTensor with rows partitioned by value_rowids.	TRUE
tf.random.gamma		2	Draws shape samples from each of the given Gamma distribution(s).	TRUE
tf.random.normal		2	Outputs random values from a normal distribution.	TRUE
tf.random.poisson		2	Draws shape samples from each of the given Poisson distribution(s).	TRUE
tf.random.uniform		2	Outputs random values from a uniform distribution.	TRUE
tf.random.truncated_normal		2	Outputs random values from a truncated normal distribution.	TRUE
tf.sparse.eye		2	Creates a two-dimensional sparse tensor with ones along the diagonal.	TRUE
tf.experimental.numpy.ndarray	tf.Tensor	3	A tf.experimental.numpy.ndarray represents a multidimensional array of elements.	TRUE
tf.keras.layers.Input	tf.keras.Input	3	A Keras tensor is a symbolic tensor-like object, which we augment with certain attributes that allow us to build a Keras model just by knowing the inputs and outputs of the model.	FALSE

Non-scalar Generators

The following APIs generate new tensors consisting of other new tensors.

Generator	Main alias of	Level	Description	Returns Tensor?
tf.range		1	Creates a sequence of numbers.	TRUE