ConfigProto.Experimental.Builder คลาสสุดท้ายแบบคงที่สาธารณะ
Everything inside Experimental is subject to change and is not subject to API stability guarantees in https://www.tensorflow.org/guide/version_compat.Protobuf ประเภท
tensorflow.ConfigProto.Experimental วิธีการสาธารณะ
| ConfigProto.Experimental.Builder | addRepeatedField (com.google.protobuf.Descriptors.FieldDescriptor ช่อง ค่าอ็อบเจ็กต์) |
| ConfigProto ทดลอง | สร้าง () |
| ConfigProto ทดลอง | สร้างบางส่วน () |
| ConfigProto.Experimental.Builder | ชัดเจน () |
| ConfigProto.Experimental.Builder | clearCollectiveDetermisticSequentialExecution () If true, make collective op execution order sequential and deterministic for potentially concurrent collective instances. |
| ConfigProto.Experimental.Builder | clearCollectiveGroupLeader () Task name for group resolution. |
| ConfigProto.Experimental.Builder | clearCollectiveNccl () If true, use NCCL for CollectiveOps. |
| ConfigProto.Experimental.Builder | clearDisableOutputPartitionGraphs () If true, the session will not store an additional copy of the graph for each subgraph. |
| ConfigProto.Experimental.Builder | clearDisableThreadSpinning () If using a direct session, disable spinning while waiting for work in the thread pool. |
| ConfigProto.Experimental.Builder | clearEnableMlirBridge () This field will eventually be deprecated and replaced by mlir_bridge_rollout (b/166038521). |
| ConfigProto.Experimental.Builder | clearEnableMlirGraphOptimization () Whether to enable the MLIR-based Graph optimizations. |
| ConfigProto.Experimental.Builder | clearExecutorType () Which executor to use, the default executor will be used if it is an empty string or "DEFAULT" string executor_type = 3; |
| ConfigProto.Experimental.Builder | clearField (ฟิลด์ com.google.protobuf.Descriptors.FieldDescriptor) |
| ConfigProto.Experimental.Builder | clearMlirBridgeRollout () This field is underdevelopment, for now use enable_mlir_bridge (b/166038521). |
| ConfigProto.Experimental.Builder | clearOneof (com.google.protobuf.Descriptors.OneofDescriptor oneof) |
| ConfigProto.Experimental.Builder | clearOptimizeForStaticGraph () If true, the session may treat the graph as being static for optimization purposes. |
| ConfigProto.Experimental.Builder | clearRecvBufMaxChunk () Guidance to formatting of large RecvBuf fields for transfer. |
| ConfigProto.Experimental.Builder | clearSessionMetadata () Metadata about the session. |
| ConfigProto.Experimental.Builder | clearShareClusterDevicesInSession () This was promoted to a non-experimental API. |
| ConfigProto.Experimental.Builder | clearShareSessionStateInClusterspecPropagation () In the following, session state means the value of a variable, elements in a hash table, or any other resource, accessible by worker sessions held by a TF server. |
| ConfigProto.Experimental.Builder | clearUseNumaAffinity () If true, and supported by the platform, the runtime will attempt to use NUMA affinity where applicable. |
| ConfigProto.Experimental.Builder | clearXlaFusionออโต้จูนเนอร์Thresh () Minimum number of batches run through the XLA graph before XLA fusion autotuner is enabled. |
| ConfigProto.Experimental.Builder | โคลน () |
| บูลีน | getCollectiveDetermisticSequentialExecution () If true, make collective op execution order sequential and deterministic for potentially concurrent collective instances. |
| สตริง | getCollectiveGroupLeader () Task name for group resolution. |
| com.google.protobuf.ByteString | getCollectiveGroupLeaderBytes () Task name for group resolution. |
| บูลีน | getCollectiveNccl () If true, use NCCL for CollectiveOps. |
| ConfigProto ทดลอง | |
| com.google.protobuf.Descriptors.Descriptor แบบคงที่ขั้นสุดท้าย | รับคำอธิบาย () |
| com.google.protobuf.Descriptors.Descriptor | |
| บูลีน | getDisableOutputPartitionGraphs () If true, the session will not store an additional copy of the graph for each subgraph. |
| บูลีน | getDisableThreadSpinning () If using a direct session, disable spinning while waiting for work in the thread pool. |
| บูลีน | getEnableMlirBridge () This field will eventually be deprecated and replaced by mlir_bridge_rollout (b/166038521). |
| บูลีน | getEnableMlirGraphOptimization () Whether to enable the MLIR-based Graph optimizations. |
| สตริง | getExecutorType () Which executor to use, the default executor will be used if it is an empty string or "DEFAULT" string executor_type = 3; |
| com.google.protobuf.ByteString | getExecutorTypeBytes () Which executor to use, the default executor will be used if it is an empty string or "DEFAULT" string executor_type = 3; |
| ConfigProto.Experimental.MlirBridgeRollout | getMlirBridgeRollout () This field is underdevelopment, for now use enable_mlir_bridge (b/166038521). |
| ภายใน | getMlirBridgeRolloutValue () This field is underdevelopment, for now use enable_mlir_bridge (b/166038521). |
| บูลีน | getOptimizeForStaticGraph () If true, the session may treat the graph as being static for optimization purposes. |
| ภายใน | getRecvBufMaxChunk () Guidance to formatting of large RecvBuf fields for transfer. |
| ข้อมูลเมตาของเซสชัน | getSessionMetadata () Metadata about the session. |
| SessionMetadata.Builder | getSessionMetadataBuilder () Metadata about the session. |
| SessionMetadataOrBuilder | getSessionMetadataOrBuilder () Metadata about the session. |
| บูลีน | getShareClusterDevicesInSession () This was promoted to a non-experimental API. |
| บูลีน | getShareSessionStateInClusterspecPropagation () In the following, session state means the value of a variable, elements in a hash table, or any other resource, accessible by worker sessions held by a TF server. |
| บูลีน | getUseNumaAffinity () If true, and supported by the platform, the runtime will attempt to use NUMA affinity where applicable. |
| ยาว | getXlaFusionออโต้จูนเนอร์ Thresh () Minimum number of batches run through the XLA graph before XLA fusion autotuner is enabled. |
| บูลีน | มีSessionMetadata () Metadata about the session. |
| บูลีนสุดท้าย | |
| ConfigProto.Experimental.Builder | mergeFrom (com.google.protobuf.ข้อความ อื่น ๆ ) |
| ConfigProto.Experimental.Builder | mergeFrom (com.google.protobuf.CodedInputStream อินพุต com.google.protobuf.ExtensionRegistryLite extensionRegistry) |
| ConfigProto.Experimental.Builder | |
| ConfigProto.Experimental.Builder สุดท้าย | mergeUnknownFields (com.google.protobuf.UnknownFieldSetknownFields) |
| ConfigProto.Experimental.Builder | setCollectiveDetermisticSequentialExecution (ค่าบูลีน) If true, make collective op execution order sequential and deterministic for potentially concurrent collective instances. |
| ConfigProto.Experimental.Builder | setCollectiveGroupLeader (ค่าสตริง) Task name for group resolution. |
| ConfigProto.Experimental.Builder | setCollectiveGroupLeaderBytes (ค่า com.google.protobuf.ByteString) Task name for group resolution. |
| ConfigProto.Experimental.Builder | setCollectiveNccl (ค่าบูลีน) If true, use NCCL for CollectiveOps. |
| ConfigProto.Experimental.Builder | setDisableOutputPartitionGraphs (ค่าบูลีน) If true, the session will not store an additional copy of the graph for each subgraph. |
| ConfigProto.Experimental.Builder | setDisableThreadSpinning (ค่าบูลีน) If using a direct session, disable spinning while waiting for work in the thread pool. |
| ConfigProto.Experimental.Builder | setEnableMlirBridge (ค่าบูลีน) This field will eventually be deprecated and replaced by mlir_bridge_rollout (b/166038521). |
| ConfigProto.Experimental.Builder | setEnableMlirGraphOptimization (ค่าบูลีน) Whether to enable the MLIR-based Graph optimizations. |
| ConfigProto.Experimental.Builder | setExecutorType (ค่าสตริง) Which executor to use, the default executor will be used if it is an empty string or "DEFAULT" string executor_type = 3; |
| ConfigProto.Experimental.Builder | setExecutorTypeBytes (ค่า com.google.protobuf.ByteString) Which executor to use, the default executor will be used if it is an empty string or "DEFAULT" string executor_type = 3; |
| ConfigProto.Experimental.Builder | setField (ฟิลด์ com.google.protobuf.Descriptors.FieldDescriptor ค่าอ็อบเจ็กต์) |
| ConfigProto.Experimental.Builder | setMlirBridgeRollout (ค่า ConfigProto.Experimental.MlirBridgeRollout ) This field is underdevelopment, for now use enable_mlir_bridge (b/166038521). |
| ConfigProto.Experimental.Builder | setMlirBridgeRolloutValue (ค่า int) This field is underdevelopment, for now use enable_mlir_bridge (b/166038521). |
| ConfigProto.Experimental.Builder | setOptimizeForStaticGraph (ค่าบูลีน) If true, the session may treat the graph as being static for optimization purposes. |
| ConfigProto.Experimental.Builder | setRecvBufMaxChunk (ค่า int) Guidance to formatting of large RecvBuf fields for transfer. |
| ConfigProto.Experimental.Builder | setRepeatedField (ฟิลด์ com.google.protobuf.Descriptors.FieldDescriptor, ดัชนี int, ค่าอ็อบเจ็กต์) |
| ConfigProto.Experimental.Builder | |
| ConfigProto.Experimental.Builder | |
| ConfigProto.Experimental.Builder | setShareClusterDevicesInSession (ค่าบูลีน) This was promoted to a non-experimental API. |
| ConfigProto.Experimental.Builder | setShareSessionStateInClusterspecPropagation (ค่าบูลีน) In the following, session state means the value of a variable, elements in a hash table, or any other resource, accessible by worker sessions held by a TF server. |
| ConfigProto.Experimental.Builder สุดท้าย | setUnknownFields (com.google.protobuf.UnknownFieldSetknownFields) |
| ConfigProto.Experimental.Builder | setUseNumaAffinity (ค่าบูลีน) If true, and supported by the platform, the runtime will attempt to use NUMA affinity where applicable. |
| ConfigProto.Experimental.Builder | setXlaFusionAutotunerThresh (ค่ายาว) Minimum number of batches run through the XLA graph before XLA fusion autotuner is enabled. |
วิธีการสืบทอด
วิธีการสาธารณะ
ConfigProto.Experimental.Builder สาธารณะ addRepeatedField (ฟิลด์ com.google.protobuf.Descriptors.FieldDescriptor ค่าอ็อบเจ็กต์)
ConfigProto.Experimental.Builder สาธารณะ clearCollectiveDetermisticSequentialExecution ()
If true, make collective op execution order sequential and deterministic for potentially concurrent collective instances.
bool collective_deterministic_sequential_execution = 6; ConfigProto.Experimental.Builder สาธารณะ clearCollectiveGroupLeader ()
Task name for group resolution.
string collective_group_leader = 1; ConfigProto.Experimental.Builder สาธารณะ clearCollectiveNccl ()
If true, use NCCL for CollectiveOps. This feature is highly experimental.
bool collective_nccl = 7; ConfigProto.Experimental.Builder สาธารณะ clearDisableOutputPartitionGraphs ()
If true, the session will not store an additional copy of the graph for each subgraph. If this option is set to true when a session is created, the `RunOptions.output_partition_graphs` options must not be set.
bool disable_output_partition_graphs = 14; ConfigProto.Experimental.Builder สาธารณะ clearDisableThreadSpinning ()
If using a direct session, disable spinning while waiting for work in the thread pool. This may result in higher latency for completing ops, but in the case where there is a lot of spinning may result in lower CPU usage.
bool disable_thread_spinning = 9; ConfigProto.Experimental.Builder สาธารณะ clearEnableMlirBridge ()
This field will eventually be deprecated and replaced by mlir_bridge_rollout (b/166038521). Whether to enable the MLIR-based TF->XLA bridge. This is a replacement to the existing bridge, and not ready for production usage yet. If this option is set to true when a session is created, MLIR is used to perform the set of graph transformations to put the graph in a form that can be executed with delegation of some computations to an accelerator. This builds on the model of XLA where a subset of the graph is encapsulated and attached to a "compile" operation, whose result is fed to an "execute" operation. The kernel for these operations is responsible to lower the encapsulated graph to a particular device.
bool enable_mlir_bridge = 13; ConfigProto.Experimental.Builder สาธารณะ clearEnableMlirGraphOptimization ()
Whether to enable the MLIR-based Graph optimizations. This will become a part of standard Tensorflow graph optimization pipeline, currently this is only used for gradual migration and testing new passes that are replacing existing optimizations in Grappler.
bool enable_mlir_graph_optimization = 16; ConfigProto.Experimental.Builder สาธารณะ clearExecutorType ()
Which executor to use, the default executor will be used if it is an empty string or "DEFAULT"
string executor_type = 3; ConfigProto.Experimental.Builder clearField สาธารณะ (ฟิลด์ com.google.protobuf.Descriptors.FieldDescriptor)
ConfigProto.Experimental.Builder สาธารณะ clearMlirBridgeRollout ()
This field is underdevelopment, for now use enable_mlir_bridge (b/166038521). Whether to enable the MLIR-based TF->XLA bridge.
.tensorflow.ConfigProto.Experimental.MlirBridgeRollout mlir_bridge_rollout = 17; ConfigProto.Experimental.Builder สาธารณะ clearOneof (com.google.protobuf.Descriptors.OneofDescriptor oneof)
ConfigProto.Experimental.Builder สาธารณะ clearOptimizeForStaticGraph ()
If true, the session may treat the graph as being static for optimization purposes. If this option is set to true when a session is created, the full GraphDef must be passed in a single call to Session::Create(), and Session::Extend() may not be supported.
bool optimize_for_static_graph = 12; ConfigProto.Experimental.Builder สาธารณะ clearRecvBufMaxChunk ()
Guidance to formatting of large RecvBuf fields for transfer. Any positive value sets the max chunk size. 0 defaults to 4096. Any negative value indicates no max, i.e. one chunk only.
int32 recv_buf_max_chunk = 4; ConfigProto.Experimental.Builder สาธารณะ clearSessionMetadata ()
Metadata about the session. If set, this can be used by the runtime and the Ops for debugging, monitoring, etc. NOTE: This is currently used and propagated only by the direct session.
.tensorflow.SessionMetadata session_metadata = 11; ConfigProto.Experimental.Builder สาธารณะ clearShareClusterDevicesInSession ()
This was promoted to a non-experimental API. Please use ConfigProto.share_cluster_devices_in_session instead.
bool share_cluster_devices_in_session = 10; ConfigProto.Experimental.Builder สาธารณะ clearShareSessionStateInClusterspecPropagation ()
In the following, session state means the value of a variable, elements in a hash table, or any other resource, accessible by worker sessions held by a TF server. When ClusterSpec propagation is enabled, the value of isolate_session_state is ignored when deciding whether to share session states in a TF server (for backwards compatibility reasons). - If share_session_state_in_clusterspec_propagation is true, the session states are shared. - If share_session_state_in_clusterspec_propagation is false, session states are isolated. When clusterspec propagation is not used, the value of share_session_state_in_clusterspec_propagation is ignored when deciding whether to share session states in a TF server. - If isolate_session_state is true, session states are isolated. - If isolate_session_state is false, session states are shared. TODO(b/129330037): Add a single API that consistently treats isolate_session_state and ClusterSpec propagation.
bool share_session_state_in_clusterspec_propagation = 8; ConfigProto.Experimental.Builder สาธารณะ clearUseNumaAffinity ()
If true, and supported by the platform, the runtime will attempt to use NUMA affinity where applicable. One consequence will be the existence of as many CPU devices as there are available NUMA nodes.
bool use_numa_affinity = 5; ConfigProto.Experimental.Builder สาธารณะ clearXlaFusionAutotunerThresh ()
Minimum number of batches run through the XLA graph before XLA fusion autotuner is enabled. Default value of zero disables the autotuner. The XLA fusion autotuner can improve performance by executing a heuristic search on the compiler parameters.
int64 xla_fusion_autotuner_thresh = 15; บูลีนสาธารณะ getCollectiveDetermisticSequentialExecution ()
If true, make collective op execution order sequential and deterministic for potentially concurrent collective instances.
bool collective_deterministic_sequential_execution = 6; สตริงสาธารณะ getCollectiveGroupLeader ()
Task name for group resolution.
string collective_group_leader = 1; สาธารณะ com.google.protobuf.ByteString getCollectiveGroupLeaderBytes ()
Task name for group resolution.
string collective_group_leader = 1; บูลีนสาธารณะ getCollectiveNccl ()
If true, use NCCL for CollectiveOps. This feature is highly experimental.
bool collective_nccl = 7; สาธารณะคงที่สุดท้าย com.google.protobuf.Descriptors.Descriptor getDescriptor ()
สาธารณะ com.google.protobuf.Descriptors.Descriptor getDescriptorForType ()
บูลีนสาธารณะ getDisableOutputPartitionGraphs ()
If true, the session will not store an additional copy of the graph for each subgraph. If this option is set to true when a session is created, the `RunOptions.output_partition_graphs` options must not be set.
bool disable_output_partition_graphs = 14; บูลีนสาธารณะ getDisableThreadSpinning ()
If using a direct session, disable spinning while waiting for work in the thread pool. This may result in higher latency for completing ops, but in the case where there is a lot of spinning may result in lower CPU usage.
bool disable_thread_spinning = 9; บูลีนสาธารณะ getEnableMlirBridge ()
This field will eventually be deprecated and replaced by mlir_bridge_rollout (b/166038521). Whether to enable the MLIR-based TF->XLA bridge. This is a replacement to the existing bridge, and not ready for production usage yet. If this option is set to true when a session is created, MLIR is used to perform the set of graph transformations to put the graph in a form that can be executed with delegation of some computations to an accelerator. This builds on the model of XLA where a subset of the graph is encapsulated and attached to a "compile" operation, whose result is fed to an "execute" operation. The kernel for these operations is responsible to lower the encapsulated graph to a particular device.
bool enable_mlir_bridge = 13; บูลีนสาธารณะ getEnableMlirGraphOptimization ()
Whether to enable the MLIR-based Graph optimizations. This will become a part of standard Tensorflow graph optimization pipeline, currently this is only used for gradual migration and testing new passes that are replacing existing optimizations in Grappler.
bool enable_mlir_graph_optimization = 16; สตริงสาธารณะ getExecutorType ()
Which executor to use, the default executor will be used if it is an empty string or "DEFAULT"
string executor_type = 3; สาธารณะ com.google.protobuf.ByteString getExecutorTypeBytes ()
Which executor to use, the default executor will be used if it is an empty string or "DEFAULT"
string executor_type = 3; ConfigProto.Experimental.MlirBridgeRollout สาธารณะ รับ MlirBridgeRollout ()
This field is underdevelopment, for now use enable_mlir_bridge (b/166038521). Whether to enable the MLIR-based TF->XLA bridge.
.tensorflow.ConfigProto.Experimental.MlirBridgeRollout mlir_bridge_rollout = 17; int สาธารณะ getMlirBridgeRolloutValue ()
This field is underdevelopment, for now use enable_mlir_bridge (b/166038521). Whether to enable the MLIR-based TF->XLA bridge.
.tensorflow.ConfigProto.Experimental.MlirBridgeRollout mlir_bridge_rollout = 17; บูลีนสาธารณะ getOptimizeForStaticGraph ()
If true, the session may treat the graph as being static for optimization purposes. If this option is set to true when a session is created, the full GraphDef must be passed in a single call to Session::Create(), and Session::Extend() may not be supported.
bool optimize_for_static_graph = 12; int สาธารณะ getRecvBufMaxChunk ()
Guidance to formatting of large RecvBuf fields for transfer. Any positive value sets the max chunk size. 0 defaults to 4096. Any negative value indicates no max, i.e. one chunk only.
int32 recv_buf_max_chunk = 4; SessionMetadata สาธารณะ getSessionMetadata ()
Metadata about the session. If set, this can be used by the runtime and the Ops for debugging, monitoring, etc. NOTE: This is currently used and propagated only by the direct session.
.tensorflow.SessionMetadata session_metadata = 11; SessionMetadata.Builder สาธารณะ getSessionMetadataBuilder ()
Metadata about the session. If set, this can be used by the runtime and the Ops for debugging, monitoring, etc. NOTE: This is currently used and propagated only by the direct session.
.tensorflow.SessionMetadata session_metadata = 11; SessionMetadataOrBuilder สาธารณะ getSessionMetadataOrBuilder ()
Metadata about the session. If set, this can be used by the runtime and the Ops for debugging, monitoring, etc. NOTE: This is currently used and propagated only by the direct session.
.tensorflow.SessionMetadata session_metadata = 11; บูลีนสาธารณะ getShareClusterDevicesInSession ()
This was promoted to a non-experimental API. Please use ConfigProto.share_cluster_devices_in_session instead.
bool share_cluster_devices_in_session = 10; บูลีนสาธารณะ getShareSessionStateInClusterspecPropagation ()
In the following, session state means the value of a variable, elements in a hash table, or any other resource, accessible by worker sessions held by a TF server. When ClusterSpec propagation is enabled, the value of isolate_session_state is ignored when deciding whether to share session states in a TF server (for backwards compatibility reasons). - If share_session_state_in_clusterspec_propagation is true, the session states are shared. - If share_session_state_in_clusterspec_propagation is false, session states are isolated. When clusterspec propagation is not used, the value of share_session_state_in_clusterspec_propagation is ignored when deciding whether to share session states in a TF server. - If isolate_session_state is true, session states are isolated. - If isolate_session_state is false, session states are shared. TODO(b/129330037): Add a single API that consistently treats isolate_session_state and ClusterSpec propagation.
bool share_session_state_in_clusterspec_propagation = 8; บูลีนสาธารณะ getUseNumaAffinity ()
If true, and supported by the platform, the runtime will attempt to use NUMA affinity where applicable. One consequence will be the existence of as many CPU devices as there are available NUMA nodes.
bool use_numa_affinity = 5; สาธารณะยาว getXlaFusionAutotunerThresh ()
Minimum number of batches run through the XLA graph before XLA fusion autotuner is enabled. Default value of zero disables the autotuner. The XLA fusion autotuner can improve performance by executing a heuristic search on the compiler parameters.
int64 xla_fusion_autotuner_thresh = 15; บูลีนสาธารณะ hasSessionMetadata ()
Metadata about the session. If set, this can be used by the runtime and the Ops for debugging, monitoring, etc. NOTE: This is currently used and propagated only by the direct session.
.tensorflow.SessionMetadata session_metadata = 11; บูลีนสุดท้ายสาธารณะ isInitialized ()
ConfigProto.Experimental.Builder สาธารณะ ผสานจาก (com.google.protobuf.CodedInputStream อินพุต com.google.protobuf.ExtensionRegistryLite extensionRegistry)
ขว้าง
| IOข้อยกเว้น |
|---|
ConfigProto.Experimental.Builder สาธารณะ ผสานSessionMetadata (ค่า SessionMetadata )
Metadata about the session. If set, this can be used by the runtime and the Ops for debugging, monitoring, etc. NOTE: This is currently used and propagated only by the direct session.
.tensorflow.SessionMetadata session_metadata = 11; ConfigProto.Experimental.Builder สาธารณะขั้นสุดท้าย ผสาน UnknownFields (com.google.protobuf.UnknownFieldSetknownFields)
ConfigProto.Experimental.Builder สาธารณะ setCollectiveDetermisticSequentialExecution (ค่าบูลีน)
If true, make collective op execution order sequential and deterministic for potentially concurrent collective instances.
bool collective_deterministic_sequential_execution = 6; ConfigProto.Experimental.Builder สาธารณะ setCollectiveGroupLeader (ค่าสตริง)
Task name for group resolution.
string collective_group_leader = 1; ConfigProto.Experimental.Builder สาธารณะ setCollectiveGroupLeaderBytes (ค่า com.google.protobuf.ByteString)
Task name for group resolution.
string collective_group_leader = 1; ConfigProto.Experimental.Builder สาธารณะ setCollectiveNccl (ค่าบูลีน)
If true, use NCCL for CollectiveOps. This feature is highly experimental.
bool collective_nccl = 7; ConfigProto.Experimental.Builder สาธารณะ setDisableOutputPartitionGraphs (ค่าบูลีน)
If true, the session will not store an additional copy of the graph for each subgraph. If this option is set to true when a session is created, the `RunOptions.output_partition_graphs` options must not be set.
bool disable_output_partition_graphs = 14; ConfigProto.Experimental.Builder สาธารณะ setDisableThreadSpinning (ค่าบูลีน)
If using a direct session, disable spinning while waiting for work in the thread pool. This may result in higher latency for completing ops, but in the case where there is a lot of spinning may result in lower CPU usage.
bool disable_thread_spinning = 9; ConfigProto.Experimental.Builder สาธารณะ setEnableMlirBridge (ค่าบูลีน)
This field will eventually be deprecated and replaced by mlir_bridge_rollout (b/166038521). Whether to enable the MLIR-based TF->XLA bridge. This is a replacement to the existing bridge, and not ready for production usage yet. If this option is set to true when a session is created, MLIR is used to perform the set of graph transformations to put the graph in a form that can be executed with delegation of some computations to an accelerator. This builds on the model of XLA where a subset of the graph is encapsulated and attached to a "compile" operation, whose result is fed to an "execute" operation. The kernel for these operations is responsible to lower the encapsulated graph to a particular device.
bool enable_mlir_bridge = 13;ConfigProto.Experimental.Builder สาธารณะ setEnableMlirGraphOptimization (ค่าบูลีน)
Whether to enable the MLIR-based Graph optimizations. This will become a part of standard Tensorflow graph optimization pipeline, currently this is only used for gradual migration and testing new passes that are replacing existing optimizations in Grappler.
bool enable_mlir_graph_optimization = 16; ConfigProto.Experimental.Builder setExecutorType สาธารณะ (ค่าสตริง)
Which executor to use, the default executor will be used if it is an empty string or "DEFAULT"
string executor_type = 3; ConfigProto.Experimental.Builder setExecutorTypeBytes สาธารณะ (ค่า com.google.protobuf.ByteString)
Which executor to use, the default executor will be used if it is an empty string or "DEFAULT"
string executor_type = 3; ConfigProto.Experimental.Builder setField สาธารณะ (com.google.protobuf.Descriptors.FieldDescriptor ฟิลด์ ค่าอ็อบเจ็กต์)
ConfigProto.Experimental.Builder setMlirBridgeRollout สาธารณะ (ค่า ConfigProto.Experimental.MlirBridgeRollout )
This field is underdevelopment, for now use enable_mlir_bridge (b/166038521). Whether to enable the MLIR-based TF->XLA bridge.
.tensorflow.ConfigProto.Experimental.MlirBridgeRollout mlir_bridge_rollout = 17; ConfigProto.Experimental.Builder สาธารณะ setMlirBridgeRolloutValue (ค่า int)
This field is underdevelopment, for now use enable_mlir_bridge (b/166038521). Whether to enable the MLIR-based TF->XLA bridge.
.tensorflow.ConfigProto.Experimental.MlirBridgeRollout mlir_bridge_rollout = 17; ConfigProto.Experimental.Builder สาธารณะ setOptimizeForStaticGraph (ค่าบูลีน)
If true, the session may treat the graph as being static for optimization purposes. If this option is set to true when a session is created, the full GraphDef must be passed in a single call to Session::Create(), and Session::Extend() may not be supported.
bool optimize_for_static_graph = 12; ConfigProto.Experimental.Builder สาธารณะ setRecvBufMaxChunk (ค่า int)
Guidance to formatting of large RecvBuf fields for transfer. Any positive value sets the max chunk size. 0 defaults to 4096. Any negative value indicates no max, i.e. one chunk only.
int32 recv_buf_max_chunk = 4; ConfigProto.Experimental.Builder setRepeatedField สาธารณะ (ฟิลด์ com.google.protobuf.Descriptors.FieldDescriptor, ดัชนี int, ค่าอ็อบเจ็กต์)
ConfigProto.Experimental.Builder สาธารณะ setSessionMetadata (ค่า SessionMetadata )
Metadata about the session. If set, this can be used by the runtime and the Ops for debugging, monitoring, etc. NOTE: This is currently used and propagated only by the direct session.
.tensorflow.SessionMetadata session_metadata = 11; ConfigProto.Experimental.Builder สาธารณะ setSessionMetadata ( SessionMetadata.Builder builderForValue)
Metadata about the session. If set, this can be used by the runtime and the Ops for debugging, monitoring, etc. NOTE: This is currently used and propagated only by the direct session.
.tensorflow.SessionMetadata session_metadata = 11; ConfigProto.Experimental.Builder สาธารณะ setShareClusterDevicesInSession (ค่าบูลีน)
This was promoted to a non-experimental API. Please use ConfigProto.share_cluster_devices_in_session instead.
bool share_cluster_devices_in_session = 10; ConfigProto.Experimental.Builder สาธารณะ setShareSessionStateInClusterspecPropagation (ค่าบูลีน)
In the following, session state means the value of a variable, elements in a hash table, or any other resource, accessible by worker sessions held by a TF server. When ClusterSpec propagation is enabled, the value of isolate_session_state is ignored when deciding whether to share session states in a TF server (for backwards compatibility reasons). - If share_session_state_in_clusterspec_propagation is true, the session states are shared. - If share_session_state_in_clusterspec_propagation is false, session states are isolated. When clusterspec propagation is not used, the value of share_session_state_in_clusterspec_propagation is ignored when deciding whether to share session states in a TF server. - If isolate_session_state is true, session states are isolated. - If isolate_session_state is false, session states are shared. TODO(b/129330037): Add a single API that consistently treats isolate_session_state and ClusterSpec propagation.
bool share_session_state_in_clusterspec_propagation = 8; ConfigProto.Experimental.Builder สุดท้ายสาธารณะ setUnknownFields (com.google.protobuf.UnknownFieldSetknownFields)
ConfigProto.Experimental.Builder สาธารณะ setUseNumaAffinity (ค่าบูลีน)
If true, and supported by the platform, the runtime will attempt to use NUMA affinity where applicable. One consequence will be the existence of as many CPU devices as there are available NUMA nodes.
bool use_numa_affinity = 5; ConfigProto.Experimental.Builder สาธารณะ setXlaFusionAutotunerThresh (ค่ายาว)
Minimum number of batches run through the XLA graph before XLA fusion autotuner is enabled. Default value of zero disables the autotuner. The XLA fusion autotuner can improve performance by executing a heuristic search on the compiler parameters.
int64 xla_fusion_autotuner_thresh = 15;