Creates a Dataset that returns pseudorandom numbers.
tf.raw_ops.RandomDatasetV2(
seed,
seed2,
seed_generator,
output_types,
output_shapes,
rerandomize_each_iteration=False,
metadata='',
name=None
)
Creates a Dataset that returns a stream of uniformly distributed pseudorandom 64-bit signed integers. It accepts a boolean attribute that determines if the random number generators are re-applied at each epoch. The default value is True which means that the seeds are applied and the same sequence of random numbers are generated at each epoch. If set to False, the seeds are not re-applied and a different sequence of random numbers are generated at each epoch.
In the TensorFlow Python API, you can instantiate this dataset via the class tf.data.experimental.RandomDatasetV2.
Args | |
|---|---|
seed | A Tensor of type int64. A scalar seed for the random number generator. If either seed or seed2 is set to be non-zero, the random number generator is seeded by the given seed. Otherwise, a random seed is used. |
seed2 | A Tensor of type int64. A second scalar seed to avoid seed collision. |
seed_generator | A Tensor of type resource. A resource for the random number seed generator. |
output_types | A list of tf.DTypes that has length >= 1. |
output_shapes | A list of shapes (each a tf.TensorShape or list of ints) that has length >= 1. |
rerandomize_each_iteration | An optional bool. Defaults to False. A boolean attribute to rerandomize the sequence of random numbers generated at each epoch. |
metadata | An optional string. Defaults to "". |
name | A name for the operation (optional). |
Returns | |
|---|---|
A Tensor of type variant. |