{ }
returns f(inputs)
, where f
's body is placed and partitioned.
tf.raw_ops.PartitionedCall(
args,
Tout,
f,
config='',
config_proto='',
executor_type='',
name=None
)
Asynchronously executes a function, potentially across multiple devices but within a single process. The kernel places and partitions a given function's underlying graph, and executes each of the partitioned subgraphs as a function.
Returns | |
---|---|
A list of Tensor objects of type Tout .
|