MCPcopy Create free account
hub / github.com/DeepRec-AI/DeepRec / run_count

Method run_count

tensorflow/core/profiler/internal/tfprof_node.h:488–504  ·  view source on GitHub ↗

Number of times the graph node is executed. When step < 0, the average number of times executed across all steps.

Source from the content-addressed store, hash-verified

486 // Number of times the graph node is executed. When step < 0, the
487 // average number of times executed across all steps.
488 int64 run_count(int64 step) const {
489 if (execs_.empty()) {
490 return 0;
491 }
492 if (step >= 0) {
493 auto exec = execs_.find(step);
494 if (exec == execs_.end()) {
495 return 0;
496 }
497 return exec->second.run_count();
498 }
499 int64 total_run_count = 0;
500 for (const auto& exec : execs_) {
501 total_run_count += exec.second.run_count();
502 }
503 return total_run_count / execs_.size();
504 }
505 // This is overall computation time, including both cpu and accelerator.
506 // Note, cpu and accelerator might or might not run in parallel.
507 int64 exec_micros(int64 step) const {

Callers

nothing calls this directly

Calls 5

emptyMethod · 0.45
findMethod · 0.45
endMethod · 0.45
run_countMethod · 0.45
sizeMethod · 0.45

Tested by

no test coverage detected