ONE - On-device Neural Engine
Loading...
Searching...
No Matches
exo::FuseRsqrtPass Struct Referencefinal

Class to fuse TFLSqrt that is divided(TFLDiv) by 1, into TFLRsqrt. More...

#include <FuseRsqrtPass.h>

Collaboration diagram for exo::FuseRsqrtPass:

Public Member Functions

const char * name (void) const final
 
bool run (loco::Graph *g) final
 Run the pass.
 
- Public Member Functions inherited from logo::Pass
virtual ~Pass ()=default
 

Detailed Description

Class to fuse TFLSqrt that is divided(TFLDiv) by 1, into TFLRsqrt.

<BEFORE>

TFLConst(1) ---— \ A — TFLSqrt — TFLDiv — B

<AFTER>

A — TFLRsqrt — B

Definition at line 38 of file FuseRsqrtPass.h.

Member Function Documentation

◆ name()

const char * exo::FuseRsqrtPass::name ( void  ) const
inlinefinalvirtual

Reimplemented from logo::Pass.

Definition at line 40 of file FuseRsqrtPass.h.

40{ return "exo::FuseRsqrtPass"; }

◆ run()

bool exo::FuseRsqrtPass::run ( loco::Graph graph)
finalvirtual

Run the pass.

Returns
false if there was nothing changed

Implements logo::Pass.

Definition at line 80 of file FuseRsqrtPass.cpp.

81{
82 bool changed = false;
83 for (auto node : loco::active_nodes(loco::output_nodes(g)))
84 {
85 if (auto div = as_candidate(node))
86 {
87 fuse_rsqrt(div);
88 changed = true;
89 }
90 }
91
92 return changed;
93}
std::set< loco::Node * > active_nodes(const std::vector< loco::Node * > &roots)
Enumerate all the nodes required to compute "roots".
std::vector< Node * > output_nodes(Graph *)
Definition Graph.cpp:101

References loco::active_nodes(), and loco::output_nodes().

Referenced by package.infer.session::inference().


The documentation for this struct was generated from the following files: