classe finale statique publique GraphOptions.Builder
tensorflow.GraphOptions
de type Protobuf.GraphOptions
Méthodes publiques
GraphOptions.Builder | addRepeatedField (champ com.google.protobuf.Descriptors.FieldDescriptor, valeur de l'objet) |
Options de graphique | construire () |
Options de graphique | buildPartial () |
GraphOptions.Builder | clair () |
GraphOptions.Builder | clearBuildCostModel () The number of steps to run before returning a cost model detailing the memory usage and performance of each node of the graph. |
GraphOptions.Builder | clearBuildCostModelAfter () The number of steps to skip before collecting statistics for the cost model. |
GraphOptions.Builder | clearEnableBfloat16Sendrecv () If true, transfer float values between processes as bfloat16. |
GraphOptions.Builder | clearEnableRecvScheduling () If true, use control flow to schedule the activation of Recv nodes. |
GraphOptions.Builder | clearField (champ com.google.protobuf.Descriptors.FieldDescriptor) |
GraphOptions.Builder | clearInferShapes () Annotate each Node with Op output shape data, to the extent it can be statically inferred. |
GraphOptions.Builder | clearOneof (com.google.protobuf.Descriptors.OneofDescriptor oneof) |
GraphOptions.Builder | clearOptimizerOptions () Options controlling how graph is optimized. |
GraphOptions.Builder | clearPlacePrunedGraph () Only place the subgraphs that are run, rather than the entire graph. |
GraphOptions.Builder | clearRewriteOptions () Options that control the type and amount of graph rewriting. |
GraphOptions.Builder | clearTimelineStep () If > 0, record a timeline every this many steps. |
GraphOptions.Builder | cloner () |
long | getBuildCostModel () The number of steps to run before returning a cost model detailing the memory usage and performance of each node of the graph. |
long | getBuildCostModelAfter () The number of steps to skip before collecting statistics for the cost model. |
OptionsGraphiques | |
final statique com.google.protobuf.Descriptors.Descriptor | |
com.google.protobuf.Descriptors.Descriptor | |
booléen | getEnableBfloat16Sendrecv () If true, transfer float values between processes as bfloat16. |
booléen | getEnableRecvScheduling () If true, use control flow to schedule the activation of Recv nodes. |
booléen | getInferShapes () Annotate each Node with Op output shape data, to the extent it can be statically inferred. |
Options de l'optimiseur | getOptimizerOptions () Options controlling how graph is optimized. |
OptimizerOptions.Builder | getOptimizerOptionsBuilder () Options controlling how graph is optimized. |
OptimizerOptionsOrBuilder | getOptimizerOptionsOrBuilder () Options controlling how graph is optimized. |
booléen | getPlacePrunedGraph () Only place the subgraphs that are run, rather than the entire graph. |
RéécrivainConfig | getRewriteOptions () Options that control the type and amount of graph rewriting. |
RewriterConfig.Builder | getRewriteOptionsBuilder () Options that control the type and amount of graph rewriting. |
RéécrivainConfigOrBuilder | getRewriteOptionsOrBuilder () Options that control the type and amount of graph rewriting. |
int | getTimelineStep () If > 0, record a timeline every this many steps. |
booléen | hasOptimizerOptions () Options controlling how graph is optimized. |
booléen | hasRewriteOptions () Options that control the type and amount of graph rewriting. |
booléen final | |
GraphOptions.Builder | mergeFrom (com.google.protobuf.Message autre) |
GraphOptions.Builder | mergeFrom (entrée com.google.protobuf.CodedInputStream, com.google.protobuf.ExtensionRegistryLite extensionRegistry) |
GraphOptions.Builder | |
GraphOptions.Builder | mergeRewriteOptions (valeur RewriterConfig ) Options that control the type and amount of graph rewriting. |
GraphOptions.Builder final | mergeUnknownFields (com.google.protobuf.UnknownFieldSet inconnuFields) |
GraphOptions.Builder | setBuildCostModel (valeur longue) The number of steps to run before returning a cost model detailing the memory usage and performance of each node of the graph. |
GraphOptions.Builder | setBuildCostModelAfter (valeur longue) The number of steps to skip before collecting statistics for the cost model. |
GraphOptions.Builder | setEnableBfloat16Sendrecv (valeur booléenne) If true, transfer float values between processes as bfloat16. |
GraphOptions.Builder | setEnableRecvScheduling (valeur booléenne) If true, use control flow to schedule the activation of Recv nodes. |
GraphOptions.Builder | setField (champ com.google.protobuf.Descriptors.FieldDescriptor, valeur de l'objet) |
GraphOptions.Builder | setInferShapes (valeur booléenne) Annotate each Node with Op output shape data, to the extent it can be statically inferred. |
GraphOptions.Builder | |
GraphOptions.Builder | setOptimizerOptions ( OptimizerOptions.Builder builderForValue) Options controlling how graph is optimized. |
GraphOptions.Builder | setPlacePrunedGraph (valeur booléenne) Only place the subgraphs that are run, rather than the entire graph. |
GraphOptions.Builder | setRepeatedField (champ com.google.protobuf.Descriptors.FieldDescriptor, index int, valeur de l'objet) |
GraphOptions.Builder | setRewriteOptions ( RewriterConfig.Builder builderForValue) Options that control the type and amount of graph rewriting. |
GraphOptions.Builder | setRewriteOptions (valeur RewriterConfig ) Options that control the type and amount of graph rewriting. |
GraphOptions.Builder | setTimelineStep (valeur entière) If > 0, record a timeline every this many steps. |
GraphOptions.Builder final | setUnknownFields (com.google.protobuf.UnknownFieldSet inconnuFields) |
Méthodes héritées
Méthodes publiques
public GraphOptions.Builder addRepeatedField (champ com.google.protobuf.Descriptors.FieldDescriptor, valeur de l'objet)
public GraphOptions.Builder clearBuildCostModel ()
The number of steps to run before returning a cost model detailing the memory usage and performance of each node of the graph. 0 means no cost model.
int64 build_cost_model = 4;
public GraphOptions.Builder clearBuildCostModelAfter ()
The number of steps to skip before collecting statistics for the cost model.
int64 build_cost_model_after = 9;
public GraphOptions.Builder clearEnableBfloat16Sendrecv ()
If true, transfer float values between processes as bfloat16.
bool enable_bfloat16_sendrecv = 7;
public GraphOptions.Builder clearEnableRecvScheduling ()
If true, use control flow to schedule the activation of Recv nodes. (Currently ignored.)
bool enable_recv_scheduling = 2;
public GraphOptions.Builder clearInferShapes ()
Annotate each Node with Op output shape data, to the extent it can be statically inferred.
bool infer_shapes = 5;
public GraphOptions.Builder clearOptimizerOptions ()
Options controlling how graph is optimized.
.tensorflow.OptimizerOptions optimizer_options = 3;
public GraphOptions.Builder clearPlacePrunedGraph ()
Only place the subgraphs that are run, rather than the entire graph. This is useful for interactive graph building, where one might produce graphs that cannot be placed during the debugging process. In particular, it allows the client to continue work in a session after adding a node to a graph whose placement constraints are unsatisfiable.
bool place_pruned_graph = 6;
public GraphOptions.Builder clearRewriteOptions ()
Options that control the type and amount of graph rewriting. Not currently configurable via the public Python API (i.e. there is no API stability guarantee if you import RewriterConfig explicitly).
.tensorflow.RewriterConfig rewrite_options = 10;
public GraphOptions.Builder clearTimelineStep ()
If > 0, record a timeline every this many steps. EXPERIMENTAL: This currently has no effect in MasterSession.
int32 timeline_step = 8;
public long getBuildCostModel ()
The number of steps to run before returning a cost model detailing the memory usage and performance of each node of the graph. 0 means no cost model.
int64 build_cost_model = 4;
public long getBuildCostModelAfter ()
The number of steps to skip before collecting statistics for the cost model.
int64 build_cost_model_after = 9;
public statique final com.google.protobuf.Descriptors.Descriptor getDescriptor ()
public com.google.protobuf.Descriptors.Descriptor getDescriptorForType ()
public booléen getEnableBfloat16Sendrecv ()
If true, transfer float values between processes as bfloat16.
bool enable_bfloat16_sendrecv = 7;
public booléen getEnableRecvScheduling ()
If true, use control flow to schedule the activation of Recv nodes. (Currently ignored.)
bool enable_recv_scheduling = 2;
public booléen getInferShapes ()
Annotate each Node with Op output shape data, to the extent it can be statically inferred.
bool infer_shapes = 5;
public OptimizerOptions getOptimizerOptions ()
Options controlling how graph is optimized.
.tensorflow.OptimizerOptions optimizer_options = 3;
public OptimizerOptions.Builder getOptimizerOptionsBuilder ()
Options controlling how graph is optimized.
.tensorflow.OptimizerOptions optimizer_options = 3;
public OptimizerOptionsOrBuilder getOptimizerOptionsOrBuilder ()
Options controlling how graph is optimized.
.tensorflow.OptimizerOptions optimizer_options = 3;
public booléen getPlacePrunedGraph ()
Only place the subgraphs that are run, rather than the entire graph. This is useful for interactive graph building, where one might produce graphs that cannot be placed during the debugging process. In particular, it allows the client to continue work in a session after adding a node to a graph whose placement constraints are unsatisfiable.
bool place_pruned_graph = 6;
public RewriterConfig getRewriteOptions ()
Options that control the type and amount of graph rewriting. Not currently configurable via the public Python API (i.e. there is no API stability guarantee if you import RewriterConfig explicitly).
.tensorflow.RewriterConfig rewrite_options = 10;
public RewriterConfig.Builder getRewriteOptionsBuilder ()
Options that control the type and amount of graph rewriting. Not currently configurable via the public Python API (i.e. there is no API stability guarantee if you import RewriterConfig explicitly).
.tensorflow.RewriterConfig rewrite_options = 10;
public RewriterConfigOrBuilder getRewriteOptionsOrBuilder ()
Options that control the type and amount of graph rewriting. Not currently configurable via the public Python API (i.e. there is no API stability guarantee if you import RewriterConfig explicitly).
.tensorflow.RewriterConfig rewrite_options = 10;
public int getTimelineStep ()
If > 0, record a timeline every this many steps. EXPERIMENTAL: This currently has no effect in MasterSession.
int32 timeline_step = 8;
public booléen hasOptimizerOptions ()
Options controlling how graph is optimized.
.tensorflow.OptimizerOptions optimizer_options = 3;
public booléen hasRewriteOptions ()
Options that control the type and amount of graph rewriting. Not currently configurable via the public Python API (i.e. there is no API stability guarantee if you import RewriterConfig explicitly).
.tensorflow.RewriterConfig rewrite_options = 10;
public final booléen isInitialized ()
public GraphOptions.Builder mergeFrom (entrée com.google.protobuf.CodedInputStream, com.google.protobuf.ExtensionRegistryLite extensionRegistry)
Jetés
IOException |
---|
public GraphOptions.Builder mergeOptimizerOptions (valeur OptimizerOptions )
Options controlling how graph is optimized.
.tensorflow.OptimizerOptions optimizer_options = 3;
public GraphOptions.Builder mergeRewriteOptions (valeur RewriterConfig )
Options that control the type and amount of graph rewriting. Not currently configurable via the public Python API (i.e. there is no API stability guarantee if you import RewriterConfig explicitly).
.tensorflow.RewriterConfig rewrite_options = 10;
public final GraphOptions.Builder mergeUnknownFields (com.google.protobuf.UnknownFieldSet unknownFields)
public GraphOptions.Builder setBuildCostModel (valeur longue)
The number of steps to run before returning a cost model detailing the memory usage and performance of each node of the graph. 0 means no cost model.
int64 build_cost_model = 4;
public GraphOptions.Builder setBuildCostModelAfter (valeur longue)
The number of steps to skip before collecting statistics for the cost model.
int64 build_cost_model_after = 9;
public GraphOptions.Builder setEnableBfloat16Sendrecv (valeur booléenne)
If true, transfer float values between processes as bfloat16.
bool enable_bfloat16_sendrecv = 7;
public GraphOptions.Builder setEnableRecvScheduling (valeur booléenne)
If true, use control flow to schedule the activation of Recv nodes. (Currently ignored.)
bool enable_recv_scheduling = 2;
public GraphOptions.Builder setField (champ com.google.protobuf.Descriptors.FieldDescriptor, valeur de l'objet)
public GraphOptions.Builder setInferShapes (valeur booléenne)
Annotate each Node with Op output shape data, to the extent it can be statically inferred.
bool infer_shapes = 5;
public GraphOptions.Builder setOptimizerOptions (valeur OptimizerOptions )
Options controlling how graph is optimized.
.tensorflow.OptimizerOptions optimizer_options = 3;
public GraphOptions.Builder setOptimizerOptions ( OptimizerOptions.Builder builderForValue)
Options controlling how graph is optimized.
.tensorflow.OptimizerOptions optimizer_options = 3;
public GraphOptions.Builder setPlacePrunedGraph (valeur booléenne)
Only place the subgraphs that are run, rather than the entire graph. This is useful for interactive graph building, where one might produce graphs that cannot be placed during the debugging process. In particular, it allows the client to continue work in a session after adding a node to a graph whose placement constraints are unsatisfiable.
bool place_pruned_graph = 6;
public GraphOptions.Builder setRepeatedField (champ com.google.protobuf.Descriptors.FieldDescriptor, index int, valeur de l'objet)
public GraphOptions.Builder setRewriteOptions ( RewriterConfig.Builder builderForValue)
Options that control the type and amount of graph rewriting. Not currently configurable via the public Python API (i.e. there is no API stability guarantee if you import RewriterConfig explicitly).
.tensorflow.RewriterConfig rewrite_options = 10;
public GraphOptions.Builder setRewriteOptions (valeur RewriterConfig )
Options that control the type and amount of graph rewriting. Not currently configurable via the public Python API (i.e. there is no API stability guarantee if you import RewriterConfig explicitly).
.tensorflow.RewriterConfig rewrite_options = 10;
public GraphOptions.Builder setTimelineStep (valeur int)
If > 0, record a timeline every this many steps. EXPERIMENTAL: This currently has no effect in MasterSession.
int32 timeline_step = 8;