Initializes a distributed TPU system for use with TensorFlow. Args: embedding_config: If not None, a `TPUEmbeddingConfiguration` proto describing the desired configuration of the hardware embedding lookup tables. If embedding_config is None, no hardware embeddings can be used.
(embedding_config=None, job=None)
| 89 | |
| 90 | @tf_export(v1=["tpu.initialize_system"]) |
| 91 | def initialize_system(embedding_config=None, job=None): |
| 92 | """Initializes a distributed TPU system for use with TensorFlow. |
| 93 | |
| 94 | Args: |
| 95 | embedding_config: If not None, a `TPUEmbeddingConfiguration` proto |
| 96 | describing the desired configuration of the hardware embedding lookup |
| 97 | tables. If embedding_config is None, no hardware embeddings can be used. |
| 98 | job: The job (the XXX in TensorFlow device specification /job:XXX) that |
| 99 | contains the TPU devices that will be initialized. If job=None it is |
| 100 | assumed there is only one job in the TensorFlow flock, and an error will |
| 101 | be returned if this assumption does not hold. |
| 102 | Returns: |
| 103 | A serialized `TopologyProto` that describes the TPU system. Note: |
| 104 | the topology must be evaluated using `Session.run` before it can be used. |
| 105 | """ |
| 106 | config_string = ("" if embedding_config is None else |
| 107 | embedding_config.SerializeToString()) |
| 108 | with ops.device(_tpu_system_device_name(job)): |
| 109 | return tpu_ops.configure_distributed_tpu(embedding_config=config_string) |
| 110 | |
| 111 | |
| 112 | def initialize_system_for_tpu_embedding(embedding_config, job=None): |
nothing calls this directly
no test coverage detected