interface publique GraphOptionsOrBuilder
| Sous-classes indirectes connues |
Méthodes publiques
| abstrait 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. |
| abstrait long | getBuildCostModelAfter () The number of steps to skip before collecting statistics for the cost model. |
| booléen abstrait | getEnableBfloat16Sendrecv () If true, transfer float values between processes as bfloat16. |
| booléen abstrait | getEnableRecvScheduling () If true, use control flow to schedule the activation of Recv nodes. |
| booléen abstrait | getInferShapes () Annotate each Node with Op output shape data, to the extent it can be statically inferred. |
| abstrait OptimizerOptions | getOptimizerOptions () Options controlling how graph is optimized. |
| abstrait OptimizerOptionsOrBuilder | getOptimizerOptionsOrBuilder () Options controlling how graph is optimized. |
| booléen abstrait | getPlacePrunedGraph () Only place the subgraphs that are run, rather than the entire graph. |
| résumé RewriterConfig | getRewriteOptions () Options that control the type and amount of graph rewriting. |
| résumé RewriterConfigOrBuilder | getRewriteOptionsOrBuilder () Options that control the type and amount of graph rewriting. |
| abstrait entier | getTimelineStep () If > 0, record a timeline every this many steps. |
| booléen abstrait | hasOptimizerOptions () Options controlling how graph is optimized. |
| booléen abstrait | hasRewriteOptions () Options that control the type and amount of graph rewriting. |
Méthodes publiques
public abstrait 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 abstrait long getBuildCostModelAfter ()
The number of steps to skip before collecting statistics for the cost model.
int64 build_cost_model_after = 9; public abstrait booléen getEnableBfloat16Sendrecv ()
If true, transfer float values between processes as bfloat16.
bool enable_bfloat16_sendrecv = 7; public abstrait booléen getEnableRecvScheduling ()
If true, use control flow to schedule the activation of Recv nodes. (Currently ignored.)
bool enable_recv_scheduling = 2; public abstrait booléen getInferShapes ()
Annotate each Node with Op output shape data, to the extent it can be statically inferred.
bool infer_shapes = 5; résumé public OptimizerOptions getOptimizerOptions ()
Options controlling how graph is optimized.
.tensorflow.OptimizerOptions optimizer_options = 3; public abstrait OptimizerOptionsOrBuilder getOptimizerOptionsOrBuilder ()
Options controlling how graph is optimized.
.tensorflow.OptimizerOptions optimizer_options = 3; public abstrait 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 abstrait 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 abstrait 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 abstrait int getTimelineStep ()
If > 0, record a timeline every this many steps. EXPERIMENTAL: This currently has no effect in MasterSession.
int32 timeline_step = 8; public abstrait booléen hasOptimizerOptions ()
Options controlling how graph is optimized.
.tensorflow.OptimizerOptions optimizer_options = 3; public abstrait 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;