kelas akhir statis publik ConfigProto.Builder
Session configuration parameters. The system picks appropriate values for fields that are not set.Protobuf tipe
tensorflow.ConfigProto Metode Publik
| ConfigProto.Builder | addAllDeviceFilters (nilai<String> yang dapat diubah) When any filters are present sessions will ignore all devices which do not match the filters. |
| ConfigProto.Builder | addAllSessionInterOpThreadPool (Nilai Iterable<? extends ThreadPoolOptionProto >) This option is experimental - it may be replaced with a different mechanism in the future. |
| ConfigProto.Builder | addDeviceFilters (Nilai string) When any filters are present sessions will ignore all devices which do not match the filters. |
| ConfigProto.Builder | addDeviceFiltersBytes (nilai com.google.protobuf.ByteString) When any filters are present sessions will ignore all devices which do not match the filters. |
| ConfigProto.Builder | addRepeatedField (bidang com.google.protobuf.Descriptors.FieldDescriptor, Nilai objek) |
| ConfigProto.Builder | addSessionInterOpThreadPool (nilai ThreadPoolOptionProto ) This option is experimental - it may be replaced with a different mechanism in the future. |
| ConfigProto.Builder | addSessionInterOpThreadPool (indeks int, nilai ThreadPoolOptionProto ) This option is experimental - it may be replaced with a different mechanism in the future. |
| ConfigProto.Builder | addSessionInterOpThreadPool ( ThreadPoolOptionProto.Pembuat pembangunForValue) This option is experimental - it may be replaced with a different mechanism in the future. |
| ConfigProto.Builder | addSessionInterOpThreadPool (indeks int, ThreadPoolOptionProto.Builder builderForValue) This option is experimental - it may be replaced with a different mechanism in the future. |
| ThreadPoolOptionProto.Builder | addSessionInterOpThreadPoolBuilder (indeks int) This option is experimental - it may be replaced with a different mechanism in the future. |
| ThreadPoolOptionProto.Builder | addSessionInterOpThreadPoolBuilder () This option is experimental - it may be replaced with a different mechanism in the future. |
| KonfigurasiProto | membangun () |
| KonfigurasiProto | |
| ConfigProto.Builder | jernih () |
| ConfigProto.Builder | clearAllowSoftPlacement () Whether soft placement is allowed. |
| ConfigProto.Builder | hapusClusterDef () Optional list of all workers to use in this session. |
| ConfigProto.Builder | |
| ConfigProto.Builder | hapusFilter Perangkat () When any filters are present sessions will ignore all devices which do not match the filters. |
| ConfigProto.Builder | jelasEksperimental () .tensorflow.ConfigProto.Experimental experimental = 16; |
| ConfigProto.Builder | clearField (bidang com.google.protobuf.Descriptors.FieldDescriptor) |
| ConfigProto.Builder | hapusGpuOptions () Options that apply to all GPUs. |
| ConfigProto.Builder | hapusGraphOptions () Options that apply to all graphs. |
| ConfigProto.Builder | hapusInterOpParallelismThreads () Nodes that perform blocking operations are enqueued on a pool of inter_op_parallelism_threads available in each process. |
| ConfigProto.Builder | jelasIntraOpParallelismThreads () The execution of an individual op (for some op types) can be parallelized on a pool of intra_op_parallelism_threads. |
| ConfigProto.Builder | clearIsolateSessionState () If true, any resources such as Variables used in the session will not be shared with other sessions. |
| ConfigProto.Builder | clearLogDevicePlacement () Whether device placements should be logged. |
| ConfigProto.Builder | clearOneof (com.google.protobuf.Descriptors.OneofDescriptor oneof) |
| ConfigProto.Builder | hapusOperationTimeoutInMs () Global timeout for all blocking operations in this session. |
| ConfigProto.Builder | periode penempatan yang jelas () Assignment of Nodes to Devices is recomputed every placement_period steps until the system warms up (at which point the recomputation typically slows down automatically). |
| ConfigProto.Builder | hapusRpcOptions () Options that apply when this session uses the distributed runtime. |
| ConfigProto.Builder | hapusSesiInterOpThreadPool () This option is experimental - it may be replaced with a different mechanism in the future. |
| ConfigProto.Builder | clearShareClusterDevicesInSession () When true, WorkerSessions are created with device attributes from the full cluster. |
| ConfigProto.Builder | hapusUsePerSessionThreads () If true, use a new set of threads for this session rather than the global pool of threads. |
| ConfigProto.Builder | klon () |
| boolean | berisiDeviceCount (kunci string) Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. |
| boolean | dapatkanAllowSoftPlacement () Whether soft placement is allowed. |
| ClusterDef | dapatkanClusterDef () Optional list of all workers to use in this session. |
| ClusterDef.Pembangun | dapatkanClusterDefBuilder () Optional list of all workers to use in this session. |
| ClusterDefOrBuilder | dapatkanClusterDefOrBuilder () Optional list of all workers to use in this session. |
| KonfigurasiProto | |
| com.google.protobuf.Descriptors.Descriptor statis terakhir | |
| com.google.protobuf.Descriptors.Descriptor | |
| Peta<String, Integer> | dapatkan Jumlah Perangkat () Gunakan getDeviceCountMap() sebagai gantinya. |
| ke dalam | dapatkanDeviceCountCount () Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. |
| Peta<String, Integer> | dapatkanDeviceCountMap () Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. |
| ke dalam | getDeviceCountOrDefault (kunci string, int defaultValue) Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. |
| ke dalam | getDeviceCountOrThrow (kunci string) Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. |
| Rangkaian | getDeviceFilters (indeks int) When any filters are present sessions will ignore all devices which do not match the filters. |
| com.google.protobuf.ByteString | getDeviceFiltersBytes (indeks int) When any filters are present sessions will ignore all devices which do not match the filters. |
| ke dalam | dapatkanDeviceFiltersCount () When any filters are present sessions will ignore all devices which do not match the filters. |
| com.google.protobuf.ProtocolStringList | dapatkanDaftarFilterPerangkat () When any filters are present sessions will ignore all devices which do not match the filters. |
| ConfigProto.Eksperimental | dapatkan Eksperimental () .tensorflow.ConfigProto.Experimental experimental = 16; |
| ConfigProto.Eksperimental.Builder | dapatkanExperimentalBuilder () .tensorflow.ConfigProto.Experimental experimental = 16; |
| ConfigProto.ExperimentalOrBuilder | dapatkanExperimentalOrBuilder () .tensorflow.ConfigProto.Experimental experimental = 16; |
| Opsi GPU | dapatkanGpuOptions () Options that apply to all GPUs. |
| Opsi GPU.Pembangun | dapatkanGpuOptionsBuilder () Options that apply to all GPUs. |
| GPUOptionsOrBuilder | dapatkanGpuOptionsOrBuilder () Options that apply to all GPUs. |
| Opsi Grafik | dapatkanGraphOptions () Options that apply to all graphs. |
| GraphOptions.Builder | dapatkanGraphOptionsBuilder () Options that apply to all graphs. |
| GraphOptionsOrBuilder | dapatkanGraphOptionsOrBuilder () Options that apply to all graphs. |
| ke dalam | dapatkanInterOpParallelismThreads () Nodes that perform blocking operations are enqueued on a pool of inter_op_parallelism_threads available in each process. |
| ke dalam | dapatkanIntraOpParallelismThreads () The execution of an individual op (for some op types) can be parallelized on a pool of intra_op_parallelism_threads. |
| boolean | dapatkanIsolateSessionState () If true, any resources such as Variables used in the session will not be shared with other sessions. |
| boolean | dapatkanLogDevicePlacement () Whether device placements should be logged. |
| Peta<String, Integer> | dapatkanMutableDeviceCount () Gunakan pengakses mutasi alternatif sebagai gantinya. |
| panjang | dapatkanOperationTimeoutInMs () Global timeout for all blocking operations in this session. |
| ke dalam | dapatkanPeriode Penempatan () Assignment of Nodes to Devices is recomputed every placement_period steps until the system warms up (at which point the recomputation typically slows down automatically). |
| Opsi RPCO | dapatkanRpcOptions () Options that apply when this session uses the distributed runtime. |
| RPCOptions.Pembangun | dapatkanRpcOptionsBuilder () Options that apply when this session uses the distributed runtime. |
| RPCOptionsOrBuilder | dapatkanRpcOptionsOrBuilder () Options that apply when this session uses the distributed runtime. |
| ThreadPoolOptionProto | getSessionInterOpThreadPool (indeks int) This option is experimental - it may be replaced with a different mechanism in the future. |
| ThreadPoolOptionProto.Builder | getSessionInterOpThreadPoolBuilder (indeks int) This option is experimental - it may be replaced with a different mechanism in the future. |
| Daftar< ThreadPoolOptionProto.Builder > | getSessionInterOpThreadPoolBuilderList () This option is experimental - it may be replaced with a different mechanism in the future. |
| ke dalam | getSessionInterOpThreadPoolCount () This option is experimental - it may be replaced with a different mechanism in the future. |
| Daftar< ThreadPoolOptionProto > | dapatkanSessionInterOpThreadPoolList () This option is experimental - it may be replaced with a different mechanism in the future. |
| ThreadPoolOptionProtoOrBuilder | getSessionInterOpThreadPoolOrBuilder (indeks int) This option is experimental - it may be replaced with a different mechanism in the future. |
| Daftar<? memperluas ThreadPoolOptionProtoOrBuilder > | getSessionInterOpThreadPoolOrBuilderList () This option is experimental - it may be replaced with a different mechanism in the future. |
| boolean | dapatkanShareClusterDevicesInSession () When true, WorkerSessions are created with device attributes from the full cluster. |
| boolean | dapatkanUsePerSessionThreads () If true, use a new set of threads for this session rather than the global pool of threads. |
| boolean | hasClusterDef () Optional list of all workers to use in this session. |
| boolean | memiliki Eksperimental () .tensorflow.ConfigProto.Experimental experimental = 16; |
| boolean | memilikiGpuOptions () Options that apply to all GPUs. |
| boolean | hasGraphOptions () Options that apply to all graphs. |
| boolean | memilikiRpcOptions () Options that apply when this session uses the distributed runtime. |
| boolean terakhir | |
| ConfigProto.Builder | |
| ConfigProto.Builder | mergeExperimental ( ConfigProto.Nilai eksperimental ) .tensorflow.ConfigProto.Experimental experimental = 16; |
| ConfigProto.Builder | mergeFrom (com.google.protobuf.Pesan lainnya) |
| ConfigProto.Builder | mergeFrom (com.google.protobuf.CodedInputStream masukan, com.google.protobuf.ExtensionRegistryLite extensionRegistry) |
| ConfigProto.Builder | |
| ConfigProto.Builder | |
| ConfigProto.Builder | mergeRpcOptions (nilai RPCOptions ) Options that apply when this session uses the distributed runtime. |
| ConfigProto.Builder terakhir | mergeUnknownFields (com.google.protobuf.UnknownFieldSet unknownFields) |
| ConfigProto.Builder | putAllDeviceCount (nilai Peta<String, Integer>) Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. |
| ConfigProto.Builder | putDeviceCount (kunci string, nilai int) Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. |
| ConfigProto.Builder | hapusDeviceCount (kunci string) Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. |
| ConfigProto.Builder | hapusSessionInterOpThreadPool (int indeks) This option is experimental - it may be replaced with a different mechanism in the future. |
| ConfigProto.Builder | setAllowSoftPlacement (nilai boolean) Whether soft placement is allowed. |
| ConfigProto.Builder | setClusterDef ( ClusterDef.Pembuat pembangunForValue) Optional list of all workers to use in this session. |
| ConfigProto.Builder | |
| ConfigProto.Builder | setDeviceFilters (indeks int, nilai String) When any filters are present sessions will ignore all devices which do not match the filters. |
| ConfigProto.Builder | setEksperimental ( ConfigProto.Nilai eksperimental ) .tensorflow.ConfigProto.Experimental experimental = 16; |
| ConfigProto.Builder | setEksperimental ( ConfigProto.Experimental.Pembuat pembangunForValue) .tensorflow.ConfigProto.Experimental experimental = 16; |
| ConfigProto.Builder | setField (bidang com.google.protobuf.Descriptors.FieldDescriptor, Nilai objek) |
| ConfigProto.Builder | |
| ConfigProto.Builder | |
| ConfigProto.Builder | |
| ConfigProto.Builder | |
| ConfigProto.Builder | setInterOpParallelismThreads (nilai int) Nodes that perform blocking operations are enqueued on a pool of inter_op_parallelism_threads available in each process. |
| ConfigProto.Builder | setIntraOpParallelismThreads (nilai int) The execution of an individual op (for some op types) can be parallelized on a pool of intra_op_parallelism_threads. |
| ConfigProto.Builder | setIsolateSessionState (nilai boolean) If true, any resources such as Variables used in the session will not be shared with other sessions. |
| ConfigProto.Builder | setLogDevicePlacement (nilai boolean) Whether device placements should be logged. |
| ConfigProto.Builder | setOperationTimeoutInMs (nilai panjang) Global timeout for all blocking operations in this session. |
| ConfigProto.Builder | setPlacementPeriod (nilai int) Assignment of Nodes to Devices is recomputed every placement_period steps until the system warms up (at which point the recomputation typically slows down automatically). |
| ConfigProto.Builder | setRepeatedField (bidang com.google.protobuf.Descriptors.FieldDescriptor, indeks int, Nilai objek) |
| ConfigProto.Builder | setRpcOptions (nilai RPCOptions ) Options that apply when this session uses the distributed runtime. |
| ConfigProto.Builder | setRpcOptions ( RPCOptions.Pembuat pembangunForValue) Options that apply when this session uses the distributed runtime. |
| ConfigProto.Builder | setSessionInterOpThreadPool (indeks int, nilai ThreadPoolOptionProto ) This option is experimental - it may be replaced with a different mechanism in the future. |
| ConfigProto.Builder | setSessionInterOpThreadPool (indeks int, ThreadPoolOptionProto.Builder builderForValue) This option is experimental - it may be replaced with a different mechanism in the future. |
| ConfigProto.Builder | setShareClusterDevicesInSession (nilai boolean) When true, WorkerSessions are created with device attributes from the full cluster. |
| ConfigProto.Builder terakhir | setUnknownFields (com.google.protobuf.UnknownFieldSet unknownFields) |
| ConfigProto.Builder | setUsePerSessionThreads (nilai boolean) If true, use a new set of threads for this session rather than the global pool of threads. |
Metode Warisan
Metode Publik
public ConfigProto.Builder addAllDeviceFilters (nilai<String> yang dapat diubah)
When any filters are present sessions will ignore all devices which do not match the filters. Each filter can be partially specified, e.g. "/job:ps" "/job:worker/replica:3", etc.
repeated string device_filters = 4; public ConfigProto.Builder addAllSessionInterOpThreadPool (Nilai Iterable<? extends ThreadPoolOptionProto >)
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; public ConfigProto.Builder addDeviceFilters (Nilai string)
When any filters are present sessions will ignore all devices which do not match the filters. Each filter can be partially specified, e.g. "/job:ps" "/job:worker/replica:3", etc.
repeated string device_filters = 4; ConfigProto.Builder publik addDeviceFiltersBytes (nilai com.google.protobuf.ByteString)
When any filters are present sessions will ignore all devices which do not match the filters. Each filter can be partially specified, e.g. "/job:ps" "/job:worker/replica:3", etc.
repeated string device_filters = 4; public ConfigProto.Builder addRepeatedField (bidang com.google.protobuf.Descriptors.FieldDescriptor, Nilai objek)
ConfigProto.Builder publik addSessionInterOpThreadPool (nilai ThreadPoolOptionProto )
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; ConfigProto.Builder publik addSessionInterOpThreadPool (indeks int, nilai ThreadPoolOptionProto )
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; ConfigProto.Builder addSessionInterOpThreadPool publik ( ThreadPoolOptionProto.Builder builderForValue)
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; ConfigProto.Builder publik addSessionInterOpThreadPool (int indeks, ThreadPoolOptionProto.Builder builderForValue)
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12;ThreadPoolOptionProto.Builder publik addSessionInterOpThreadPoolBuilder (int indeks)
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; ThreadPoolOptionProto.Builder publik addSessionInterOpThreadPoolBuilder ()
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; ConfigProto.Builder publik clearAllowSoftPlacement ()
Whether soft placement is allowed. If allow_soft_placement is true, an op will be placed on CPU if 1. there's no GPU implementation for the OP or 2. no GPU devices are known or registered or 3. need to co-locate with reftype input(s) which are from CPU.
bool allow_soft_placement = 7; ConfigProto.Builder publik clearClusterDef ()
Optional list of all workers to use in this session.
.tensorflow.ClusterDef cluster_def = 14; ConfigProto.Builder publik clearDeviceFilters ()
When any filters are present sessions will ignore all devices which do not match the filters. Each filter can be partially specified, e.g. "/job:ps" "/job:worker/replica:3", etc.
repeated string device_filters = 4; ConfigProto.Builder publik clearExperimental ()
.tensorflow.ConfigProto.Experimental experimental = 16;
ConfigProto.Builder publik clearGpuOptions ()
Options that apply to all GPUs.
.tensorflow.GPUOptions gpu_options = 6; ConfigProto.Builder publik clearGraphOptions ()
Options that apply to all graphs.
.tensorflow.GraphOptions graph_options = 10; ConfigProto.Builder publik clearInterOpParallelismThreads ()
Nodes that perform blocking operations are enqueued on a pool of inter_op_parallelism_threads available in each process. 0 means the system picks an appropriate number. Negative means all operations are performed in caller's thread. Note that the first Session created in the process sets the number of threads for all future sessions unless use_per_session_threads is true or session_inter_op_thread_pool is configured.
int32 inter_op_parallelism_threads = 5; ConfigProto.Builder publik clearIntraOpParallelismThreads ()
The execution of an individual op (for some op types) can be
parallelized on a pool of intra_op_parallelism_threads.
0 means the system picks an appropriate number.
If you create an ordinary session, e.g., from Python or C++,
then there is exactly one intra op thread pool per process.
The first session created determines the number of threads in this pool.
All subsequent sessions reuse/share this one global pool.
There are notable exceptions to the default behavior describe above:
1. There is an environment variable for overriding this thread pool,
named TF_OVERRIDE_GLOBAL_THREADPOOL.
2. When connecting to a server, such as a remote `tf.train.Server`
instance, then this option will be ignored altogether.
int32 intra_op_parallelism_threads = 2; ConfigProto.Builder publik clearIsolateSessionState ()
If true, any resources such as Variables used in the session will not be shared with other sessions. However, when clusterspec propagation is enabled, this field is ignored and sessions are always isolated.
bool isolate_session_state = 15; ConfigProto.Builder publik clearLogDevicePlacement ()
Whether device placements should be logged.
bool log_device_placement = 8; ConfigProto.Builder publik clearOperationTimeoutInMs ()
Global timeout for all blocking operations in this session. If non-zero, and not overridden on a per-operation basis, this value will be used as the deadline for all blocking operations.
int64 operation_timeout_in_ms = 11; ConfigProto.Builder publik clearPlacementPeriod ()
Assignment of Nodes to Devices is recomputed every placement_period steps until the system warms up (at which point the recomputation typically slows down automatically).
int32 placement_period = 3; ConfigProto.Builder publik clearRpcOptions ()
Options that apply when this session uses the distributed runtime.
.tensorflow.RPCOptions rpc_options = 13; ConfigProto.Builder publik clearSessionInterOpThreadPool ()
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; ConfigProto.Builder publik clearShareClusterDevicesInSession ()
When true, WorkerSessions are created with device attributes from the full cluster. This is helpful when a worker wants to partition a graph (for example during a PartitionedCallOp).
bool share_cluster_devices_in_session = 17; ConfigProto.Builder publik clearUsePerSessionThreads ()
If true, use a new set of threads for this session rather than the global pool of threads. Only supported by direct sessions. If false, use the global threads created by the first session, or the per-session thread pools configured by session_inter_op_thread_pool. This option is deprecated. The same effect can be achieved by setting session_inter_op_thread_pool to have one element, whose num_threads equals inter_op_parallelism_threads.
bool use_per_session_threads = 9; boolean publik berisiDeviceCount (kunci string)
Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. If a particular device type is not found in the map, the system picks an appropriate number.
map<string, int32> device_count = 1; getAllowSoftPlacement boolean publik ()
Whether soft placement is allowed. If allow_soft_placement is true, an op will be placed on CPU if 1. there's no GPU implementation for the OP or 2. no GPU devices are known or registered or 3. need to co-locate with reftype input(s) which are from CPU.
bool allow_soft_placement = 7; ClusterDef publik getClusterDef ()
Optional list of all workers to use in this session.
.tensorflow.ClusterDef cluster_def = 14; ClusterDef.Builder publik getClusterDefBuilder ()
Optional list of all workers to use in this session.
.tensorflow.ClusterDef cluster_def = 14; ClusterDefOrBuilder publik getClusterDefOrBuilder ()
Optional list of all workers to use in this session.
.tensorflow.ClusterDef cluster_def = 14; public static final com.google.protobuf.Descriptors.Descriptor getDescriptor ()
com.google.protobuf.Descriptors.Descriptor publik getDescriptorForType ()
int publik getDeviceCountCount ()
Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. If a particular device type is not found in the map, the system picks an appropriate number.
map<string, int32> device_count = 1; Peta publik<String, Integer> getDeviceCountMap ()
Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. If a particular device type is not found in the map, the system picks an appropriate number.
map<string, int32> device_count = 1; public int getDeviceCountOrDefault (kunci string, int defaultValue)
Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. If a particular device type is not found in the map, the system picks an appropriate number.
map<string, int32> device_count = 1;public int getDeviceCountOrThrow (kunci string)
Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. If a particular device type is not found in the map, the system picks an appropriate number.
map<string, int32> device_count = 1; String publik getDeviceFilters (indeks int)
When any filters are present sessions will ignore all devices which do not match the filters. Each filter can be partially specified, e.g. "/job:ps" "/job:worker/replica:3", etc.
repeated string device_filters = 4; publik com.google.protobuf.ByteString getDeviceFiltersBytes (indeks int)
When any filters are present sessions will ignore all devices which do not match the filters. Each filter can be partially specified, e.g. "/job:ps" "/job:worker/replica:3", etc.
repeated string device_filters = 4; int publik getDeviceFiltersCount ()
When any filters are present sessions will ignore all devices which do not match the filters. Each filter can be partially specified, e.g. "/job:ps" "/job:worker/replica:3", etc.
repeated string device_filters = 4; publik com.google.protobuf.ProtocolStringList getDeviceFiltersList ()
When any filters are present sessions will ignore all devices which do not match the filters. Each filter can be partially specified, e.g. "/job:ps" "/job:worker/replica:3", etc.
repeated string device_filters = 4; ConfigProto publik.Eksperimental getEksperimental ()
.tensorflow.ConfigProto.Experimental experimental = 16;
ConfigProto.Experimental.Builder publik getExperimentalBuilder ()
.tensorflow.ConfigProto.Experimental experimental = 16;
ConfigProto.ExperimentalOrBuilder publik getExperimentalOrBuilder ()
.tensorflow.ConfigProto.Experimental experimental = 16;
Opsi GPU publik getGpuOptions ()
Options that apply to all GPUs.
.tensorflow.GPUOptions gpu_options = 6; GPUOptions.Builder publik getGpuOptionsBuilder ()
Options that apply to all GPUs.
.tensorflow.GPUOptions gpu_options = 6; GPUOptionsOrBuilder publik getGpuOptionsOrBuilder ()
Options that apply to all GPUs.
.tensorflow.GPUOptions gpu_options = 6; GraphOptions publik getGraphOptions ()
Options that apply to all graphs.
.tensorflow.GraphOptions graph_options = 10; GraphOptions.Builder publik getGraphOptionsBuilder ()
Options that apply to all graphs.
.tensorflow.GraphOptions graph_options = 10; publik GraphOptionsOrBuilder getGraphOptionsOrBuilder ()
Options that apply to all graphs.
.tensorflow.GraphOptions graph_options = 10; publik int getInterOpParallelismThreads ()
Nodes that perform blocking operations are enqueued on a pool of inter_op_parallelism_threads available in each process. 0 means the system picks an appropriate number. Negative means all operations are performed in caller's thread. Note that the first Session created in the process sets the number of threads for all future sessions unless use_per_session_threads is true or session_inter_op_thread_pool is configured.
int32 inter_op_parallelism_threads = 5; publik int getIntraOpParallelismThreads ()
The execution of an individual op (for some op types) can be
parallelized on a pool of intra_op_parallelism_threads.
0 means the system picks an appropriate number.
If you create an ordinary session, e.g., from Python or C++,
then there is exactly one intra op thread pool per process.
The first session created determines the number of threads in this pool.
All subsequent sessions reuse/share this one global pool.
There are notable exceptions to the default behavior describe above:
1. There is an environment variable for overriding this thread pool,
named TF_OVERRIDE_GLOBAL_THREADPOOL.
2. When connecting to a server, such as a remote `tf.train.Server`
instance, then this option will be ignored altogether.
int32 intra_op_parallelism_threads = 2; boolean publik getIsolateSessionState ()
If true, any resources such as Variables used in the session will not be shared with other sessions. However, when clusterspec propagation is enabled, this field is ignored and sessions are always isolated.
bool isolate_session_state = 15; getLogDevicePlacement boolean publik ()
Whether device placements should be logged.
bool log_device_placement = 8; Peta publik<String, Integer> getMutableDeviceCount ()
Gunakan pengakses mutasi alternatif sebagai gantinya.
getOperationTimeoutInMs publik yang panjang ()
Global timeout for all blocking operations in this session. If non-zero, and not overridden on a per-operation basis, this value will be used as the deadline for all blocking operations.
int64 operation_timeout_in_ms = 11; int publik getPlacementPeriod ()
Assignment of Nodes to Devices is recomputed every placement_period steps until the system warms up (at which point the recomputation typically slows down automatically).
int32 placement_period = 3; RPCOptions publik getRpcOptions ()
Options that apply when this session uses the distributed runtime.
.tensorflow.RPCOptions rpc_options = 13; RPCOptions.Builder publik getRpcOptionsBuilder ()
Options that apply when this session uses the distributed runtime.
.tensorflow.RPCOptions rpc_options = 13; RPCOptionsOrBuilder publik getRpcOptionsOrBuilder ()
Options that apply when this session uses the distributed runtime.
.tensorflow.RPCOptions rpc_options = 13; ThreadPoolOptionProto publik getSessionInterOpThreadPool (int indeks)
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12;ThreadPoolOptionProto.Builder publik getSessionInterOpThreadPoolBuilder (int indeks)
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; Daftar publik< ThreadPoolOptionProto.Builder > getSessionInterOpThreadPoolBuilderList ()
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; int publik getSessionInterOpThreadPoolCount ()
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; Daftar publik< ThreadPoolOptionProto > getSessionInterOpThreadPoolList ()
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; publik ThreadPoolOptionProtoOrBuilder getSessionInterOpThreadPoolOrBuilder (int indeks)
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; Daftar Publik<? memperluas ThreadPoolOptionProtoOrBuilder > getSessionInterOpThreadPoolOrBuilderList ()
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; boolean publik getShareClusterDevicesInSession ()
When true, WorkerSessions are created with device attributes from the full cluster. This is helpful when a worker wants to partition a graph (for example during a PartitionedCallOp).
bool share_cluster_devices_in_session = 17; boolean publik getUsePerSessionThreads ()
If true, use a new set of threads for this session rather than the global pool of threads. Only supported by direct sessions. If false, use the global threads created by the first session, or the per-session thread pools configured by session_inter_op_thread_pool. This option is deprecated. The same effect can be achieved by setting session_inter_op_thread_pool to have one element, whose num_threads equals inter_op_parallelism_threads.
bool use_per_session_threads = 9; boolean publik hasClusterDef ()
Optional list of all workers to use in this session.
.tensorflow.ClusterDef cluster_def = 14; boolean publik hasEksperimental ()
.tensorflow.ConfigProto.Experimental experimental = 16;
boolean publik hasGpuOptions ()
Options that apply to all GPUs.
.tensorflow.GPUOptions gpu_options = 6; boolean publik hasGraphOptions ()
Options that apply to all graphs.
.tensorflow.GraphOptions graph_options = 10; boolean publik hasRpcOptions ()
Options that apply when this session uses the distributed runtime.
.tensorflow.RPCOptions rpc_options = 13; boolean akhir publik diinisialisasi ()
ConfigProto.Builder mergeClusterDef publik (nilai ClusterDef )
Optional list of all workers to use in this session.
.tensorflow.ClusterDef cluster_def = 14; ConfigProto.Builder mergeExperimental publik ( ConfigProto.Nilai eksperimental)
.tensorflow.ConfigProto.Experimental experimental = 16;
ConfigProto.Builder mergeFrom (com.google.protobuf.CodedInputStream masukan, com.google.protobuf.ExtensionRegistryLite extensionRegistry)
Melempar
| Pengecualian IO |
|---|
ConfigProto.Builder mergeGpuOptions publik (nilai GPUOptions )
Options that apply to all GPUs.
.tensorflow.GPUOptions gpu_options = 6; ConfigProto.Builder mergeGraphOptions publik (nilai GraphOptions )
Options that apply to all graphs.
.tensorflow.GraphOptions graph_options = 10; ConfigProto.Builder mergeRpcOptions publik (nilai RPCOptions )
Options that apply when this session uses the distributed runtime.
.tensorflow.RPCOptions rpc_options = 13; ConfigProto.Builder mergeUnknownFields final publik (com.google.protobuf.UnknownFieldSet unknownFields)
public ConfigProto.Builder putAllDeviceCount (nilai Map<String, Integer>)
Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. If a particular device type is not found in the map, the system picks an appropriate number.
map<string, int32> device_count = 1; public ConfigProto.Builder putDeviceCount (kunci string, nilai int)
Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. If a particular device type is not found in the map, the system picks an appropriate number.
map<string, int32> device_count = 1; ConfigProto.Builder publik deleteDeviceCount (kunci string)
Map from device type name (e.g., "CPU" or "GPU" ) to maximum number of devices of that type to use. If a particular device type is not found in the map, the system picks an appropriate number.
map<string, int32> device_count = 1; ConfigProto.Builder publik hapusSessionInterOpThreadPool (int indeks)
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; ConfigProto.Builder setAllowSoftPlacement publik (nilai boolean)
Whether soft placement is allowed. If allow_soft_placement is true, an op will be placed on CPU if 1. there's no GPU implementation for the OP or 2. no GPU devices are known or registered or 3. need to co-locate with reftype input(s) which are from CPU.
bool allow_soft_placement = 7; ConfigProto.Builder setClusterDef publik ( ClusterDef.Builder builderForValue)
Optional list of all workers to use in this session.
.tensorflow.ClusterDef cluster_def = 14; ConfigProto.Builder setClusterDef publik (nilai ClusterDef )
Optional list of all workers to use in this session.
.tensorflow.ClusterDef cluster_def = 14;ConfigProto.Builder setDeviceFilters publik (indeks int, nilai String)
When any filters are present sessions will ignore all devices which do not match the filters. Each filter can be partially specified, e.g. "/job:ps" "/job:worker/replica:3", etc.
repeated string device_filters = 4; ConfigProto.Builder setExperimental publik ( ConfigProto.Nilai eksperimental)
.tensorflow.ConfigProto.Experimental experimental = 16;
ConfigProto.Builder setEksperimental publik ( ConfigProto.Experimental.Builder builderForValue)
.tensorflow.ConfigProto.Experimental experimental = 16;
public ConfigProto.Builder setField (bidang com.google.protobuf.Descriptors.FieldDescriptor, Nilai objek)
ConfigProto.Builder setGpuOptions publik ( GPUOptions.Builder builderForValue)
Options that apply to all GPUs.
.tensorflow.GPUOptions gpu_options = 6; ConfigProto.Builder setGpuOptions publik (nilai GPUOptions )
Options that apply to all GPUs.
.tensorflow.GPUOptions gpu_options = 6; ConfigProto.Builder setGraphOptions publik ( GraphOptions.Builder builderForValue)
Options that apply to all graphs.
.tensorflow.GraphOptions graph_options = 10; ConfigProto.Builder setGraphOptions publik (nilai GraphOptions )
Options that apply to all graphs.
.tensorflow.GraphOptions graph_options = 10; ConfigProto.Builder setInterOpParallelismThreads publik (nilai int)
Nodes that perform blocking operations are enqueued on a pool of inter_op_parallelism_threads available in each process. 0 means the system picks an appropriate number. Negative means all operations are performed in caller's thread. Note that the first Session created in the process sets the number of threads for all future sessions unless use_per_session_threads is true or session_inter_op_thread_pool is configured.
int32 inter_op_parallelism_threads = 5; public ConfigProto.Builder setIntraOpParallelismThreads (nilai int)
The execution of an individual op (for some op types) can be
parallelized on a pool of intra_op_parallelism_threads.
0 means the system picks an appropriate number.
If you create an ordinary session, e.g., from Python or C++,
then there is exactly one intra op thread pool per process.
The first session created determines the number of threads in this pool.
All subsequent sessions reuse/share this one global pool.
There are notable exceptions to the default behavior describe above:
1. There is an environment variable for overriding this thread pool,
named TF_OVERRIDE_GLOBAL_THREADPOOL.
2. When connecting to a server, such as a remote `tf.train.Server`
instance, then this option will be ignored altogether.
int32 intra_op_parallelism_threads = 2; public ConfigProto.Builder setIsolateSessionState (nilai boolean)
If true, any resources such as Variables used in the session will not be shared with other sessions. However, when clusterspec propagation is enabled, this field is ignored and sessions are always isolated.
bool isolate_session_state = 15; ConfigProto.Builder setLogDevicePlacement publik (nilai boolean)
Whether device placements should be logged.
bool log_device_placement = 8; public ConfigProto.Builder setOperationTimeoutInMs (nilai panjang)
Global timeout for all blocking operations in this session. If non-zero, and not overridden on a per-operation basis, this value will be used as the deadline for all blocking operations.
int64 operation_timeout_in_ms = 11; ConfigProto.Builder setPlacementPeriod publik (nilai int)
Assignment of Nodes to Devices is recomputed every placement_period steps until the system warms up (at which point the recomputation typically slows down automatically).
int32 placement_period = 3; public ConfigProto.Builder setRepeatedField (bidang com.google.protobuf.Descriptors.FieldDescriptor, indeks int, Nilai objek)
ConfigProto.Builder setRpcOptions publik (nilai RPCOptions )
Options that apply when this session uses the distributed runtime.
.tensorflow.RPCOptions rpc_options = 13; ConfigProto.Builder setRpcOptions publik ( RPCOptions.Builder builderForValue)
Options that apply when this session uses the distributed runtime.
.tensorflow.RPCOptions rpc_options = 13; ConfigProto.Builder publik setSessionInterOpThreadPool (indeks int, nilai ThreadPoolOptionProto )
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; ConfigProto.Builder setSessionInterOpThreadPool publik (indeks int, ThreadPoolOptionProto.Builder builderForValue)
This option is experimental - it may be replaced with a different mechanism in the future. Configures session thread pools. If this is configured, then RunOptions for a Run call can select the thread pool to use. The intended use is for when some session invocations need to run in a background pool limited to a small number of threads: - For example, a session may be configured to have one large pool (for regular compute) and one small pool (for periodic, low priority work); using the small pool is currently the mechanism for limiting the inter-op parallelism of the low priority work. Note that it does not limit the parallelism of work spawned by a single op kernel implementation. - Using this setting is normally not needed in training, but may help some serving use cases. - It is also generally recommended to set the global_name field of this proto, to avoid creating multiple large pools. It is typically better to run the non-low-priority work, even across sessions, in a single large pool.
repeated .tensorflow.ThreadPoolOptionProto session_inter_op_thread_pool = 12; ConfigProto.Builder publik setShareClusterDevicesInSession (nilai boolean)
When true, WorkerSessions are created with device attributes from the full cluster. This is helpful when a worker wants to partition a graph (for example during a PartitionedCallOp).
bool share_cluster_devices_in_session = 17; ConfigProto.Builder setUnknownFields final publik (com.google.protobuf.UnknownFieldSet unknownFields)
public ConfigProto.Builder setUsePerSessionThreads (nilai boolean)
If true, use a new set of threads for this session rather than the global pool of threads. Only supported by direct sessions. If false, use the global threads created by the first session, or the per-session thread pools configured by session_inter_op_thread_pool. This option is deprecated. The same effect can be achieved by setting session_inter_op_thread_pool to have one element, whose num_threads equals inter_op_parallelism_threads.
bool use_per_session_threads = 9;