From 6940dd720ebb6b3d1df8ca203ab696daefe58189 Mon Sep 17 00:00:00 2001 From: Jim Flynn Date: Fri, 20 Mar 2020 12:25:56 +0000 Subject: renamed Documentation folder 20.02 and added .nojekyll file Signed-off-by: Jim Flynn --- 20.02/namespacearmnn_caffe_parser.xhtml | 208 ++++++++++++++++++++++++++++++++ 1 file changed, 208 insertions(+) create mode 100644 20.02/namespacearmnn_caffe_parser.xhtml (limited to '20.02/namespacearmnn_caffe_parser.xhtml') diff --git a/20.02/namespacearmnn_caffe_parser.xhtml b/20.02/namespacearmnn_caffe_parser.xhtml new file mode 100644 index 0000000000..a6d721b8e3 --- /dev/null +++ b/20.02/namespacearmnn_caffe_parser.xhtml @@ -0,0 +1,208 @@ + + + + + + + + + + + + + +ArmNN: armnnCaffeParser Namespace Reference + + + + + + + + + + + + + + + + +
+
+ + + + ArmNN + + + +
+
+  20.02 +
+
+
+ + + + + + + +
+
+ +
+
+
+ +
+ +
+
+ + +
+ +
+ +
+ +
+
armnnCaffeParser Namespace Reference
+
+
+ +

Caffe networks are loaded from protobuf files (binary or text) using the protobuf library and the generated code from caffe.pb.h. +More...

+ + + + + + + + + + +

+Classes

class  CaffeParser
 
class  CaffeParserBase
 
class  ICaffeParser
 
class  RecordByRecordCaffeParser
 
+ + + + + +

+Typedefs

using BindingPointInfo = armnn::BindingPointInfo
 
using ICaffeParserPtr = std::unique_ptr< ICaffeParser, void(*)(ICaffeParser *parser)>
 
+ + + +

+Functions

BlobShape TensorDescToBlobShape (const TensorInfo &desc)
 
+

Detailed Description

+

Caffe networks are loaded from protobuf files (binary or text) using the protobuf library and the generated code from caffe.pb.h.

+

This gives us a caffe::NetParameter which is an in-memory version of the file. This contains a flat list of Caffe 'layers' (e.g. convolution, pooling etc.). Each layer has inputs (called "bottoms") and outputs (called "tops"). Data flows from bottom to top. The bottoms of a layer refer to the tops of other layers, not their names. The names of layers seem to be arbitrary (you could rename a layer and the network wouldn't need any other changes).

+

Some layers (e.g. Relu) can be configured so that their top and bottom are both the same. This is called an "in-place" layer and is a Caffe runtime feature used to reduce memory usage by modifying tensors in-place. This isn't relevant to the parser and so we preprocess these layers to convert them to regular layers, to result in a consistent graph structure.

+

Typedef Documentation

+ +

◆ BindingPointInfo

+ +
+
+ +

Definition at line 19 of file ICaffeParser.hpp.

+ +
+
+ +

◆ ICaffeParserPtr

+ +
+
+ + + + +
using ICaffeParserPtr = std::unique_ptr<ICaffeParser, void(*)(ICaffeParser* parser)>
+
+ +

Definition at line 22 of file ICaffeParser.hpp.

+ +
+
+

Function Documentation

+ +

◆ TensorDescToBlobShape()

+ +
+
+ + + + + + + + +
BlobShape armnnCaffeParser::TensorDescToBlobShape (const TensorInfodesc)
+
+ +

Definition at line 325 of file CaffeParser.cpp.

+ +

References TensorInfo::GetNumDimensions(), and TensorInfo::GetShape().

+ +

Referenced by CaffeParserBase::AddConvLayerWithDepthwiseConv(), CaffeParserBase::AddConvLayerWithSplits(), and CaffeParserBase::ParseConvLayer().

+
326 {
327  BlobShape ret;
328  for (unsigned int i = 0; i < desc.GetNumDimensions(); ++i)
329  {
330  ret.add_dim(i);
331  ret.set_dim(boost::numeric_cast<int>(i), desc.GetShape()[i]);
332  }
333 
334  return ret;
335 }
const TensorShape & GetShape() const
Definition: Tensor.hpp:88
+
unsigned int GetNumDimensions() const
Definition: Tensor.hpp:92
+
+
+
+
+
+ + + + -- cgit v1.2.1