tf.raw_ops.TPUReplicateMetadata

Metadata indicating how the TPU computation should be replicated.

tf.raw_ops.TPUReplicateMetadata(
    num_replicas,
    num_cores_per_replica=1,
    topology='',
    use_tpu=True,
    device_assignment=[],
    computation_shape=[],
    host_compute_core=[],
    padding_map=[],
    step_marker_location='STEP_MARK_AT_ENTRY',
    allow_soft_placement=False,
    use_spmd_for_xla_partitioning=False,
    tpu_compile_options_proto='',
    name=None
)

This operation holds the metadata common to operations of a tpu.replicate() computation subgraph.

num_replicasAn int that is >= 0. Number of replicas of the computation
num_cores_per_replicaAn optional int. Defaults to 1. Number of cores per replica. Used for model parallelism.
topologyAn optional string. Defaults to "". TopologyProto indicating the topology of the TPU pod slice.
use_tpuAn optional bool. Defaults to True. Whether to place the computation on the TPU.
device_assignmentAn optional list of ints. Defaults to []. The assignment of devices for the computation.
computation_shapeAn optional list of ints. Defaults to []. DEPRECATED. Use num_cores_per_replica instead.
host_compute_coreAn optional list of strings. Defaults to [].
padding_mapAn optional list of strings. Defaults to [].
step_marker_locationAn optional string. Defaults to "STEP_MARK_AT_ENTRY".
allow_soft_placementAn optional bool. Defaults to False.
use_spmd_for_xla_partitioningAn optional bool. Defaults to False.
tpu_compile_options_protoAn optional string. Defaults to "".
nameA name for the operation (optional).

The created Operation.