1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
|
// Copyright (C) 2018 Intel Corporation
//
// SPDX-License-Identifier: Apache-2.0
//
#include "mkldnn_fullyconnected_node.h"
#include "desc_iterator.hpp"
#include <ie_layers.h>
#include <string>
#include <vector>
#include <mkldnn_extension_utils.h>
using namespace mkldnn;
using namespace MKLDNNPlugin;
using namespace InferenceEngine;
MKLDNNFullyConnectedNode::MKLDNNFullyConnectedNode(const InferenceEngine::CNNLayerPtr& layer, const mkldnn::engine& eng) : MKLDNNNode(layer, eng) {
internalBlobDesc.emplace_back([&](primitive_desc_iterator &primitive_desc_it, size_t idx) -> MKLDNNMemoryDesc {
return MKLDNNMemoryDesc(primitive_desc_it.weights_primitive_desc(0).desc());
});
internalBlobDesc.emplace_back([&](primitive_desc_iterator &primitive_desc_it, size_t idx) -> MKLDNNMemoryDesc {
if (internalBlobs.size() <= 1)
return MKLDNNMemoryDesc();
return MKLDNNMemoryDesc(primitive_desc_it.weights_primitive_desc(1).desc());
});
}
void MKLDNNFullyConnectedNode::getSupportedDescriptors() {
if (!descs.empty())
return;
InferenceEngine::Precision precision = getCnnLayer()->insData[0].lock()->getPrecision();
if (precision != InferenceEngine::Precision::FP32)
precision = InferenceEngine::Precision::FP32;
auto inputDataType = MKLDNNExtensionUtils::IEPrecisionToDataType(precision);
precision = getCnnLayer()->outData[0]->getPrecision();
if (precision != InferenceEngine::Precision::FP32)
precision = InferenceEngine::Precision::FP32;
auto outputDataType = MKLDNNExtensionUtils::IEPrecisionToDataType(precision);
auto * fcLayer = dynamic_cast<FullyConnectedLayer*>(getCnnLayer().get());
if (fcLayer == nullptr)
THROW_IE_EXCEPTION << "Cannot convert fully connected layer.";
if (fcLayer->_weights == nullptr) {
THROW_IE_EXCEPTION << "Weights are empty for layer: " << fcLayer->name
<< " used in MKLDNN node: " << getName() << "\n"
<< "Use ReadWeights and SetWeights methods of InferenceEngine::CNNNetReader"
<< " to load them from .bin part of the IR";
}
if (getParentEdges().size() != 1)
THROW_IE_EXCEPTION << "Incorrect number of input edges.";
if (getParentEdges().empty())
THROW_IE_EXCEPTION << "Incorrect number of output edges.";
MKLDNNDims inDims(fcLayer->input()->getDims());
if (inDims.ndims() == 2) {
weightsDims = {fcLayer->_out_num, static_cast<size_t>(inDims.size(1))};
} else if (inDims.ndims() == 4) {
weightsDims = {fcLayer->_out_num, static_cast<size_t>(inDims[1]), static_cast<size_t>(inDims[2]),
static_cast<size_t>(inDims[3])};
} else {
THROW_IE_EXCEPTION << "Unsupported source format for FC layer. Expected 4 or 2, got: "
<< inDims.ndims() << " dims.";
}
internalBlobs.push_back(createInternalBlob(weightsDims, true));
bool withBiases = (fcLayer->_biases != nullptr && fcLayer->_biases->size() != 0);
if (withBiases) {
biasesDims.push_back(static_cast<int>(fcLayer->_out_num));
internalBlobs.push_back(createInternalBlob(biasesDims, false));
}
for (auto format : getAvailableFormatsForDims(getParentEdgeAt(0)->getDims())) {
MKLDNNMemoryDesc in_candidate(inDims, inputDataType, format);
MKLDNNMemoryDesc out_candidate(getChildEdgeAt(0)->getDims(), outputDataType, memory::any);
createDescriptor({in_candidate}, {out_candidate});
}
}
void MKLDNNFullyConnectedNode::createPrimitive() {
if (prim)
return;
auto prim_desc = createPrimitiveDescriptor<inner_product_forward::primitive_desc, inner_product_forward::desc>();
if (internalBlobs.size() > 1) {
prim.reset(new inner_product_forward(prim_desc,
getParentEdgeAt(0)->getMemory().GetPrimitive(),
internalBlobMemory[0]->GetPrimitive(),
internalBlobMemory[1]->GetPrimitive(),
getChildEdgeAt(0)->getMemory().GetPrimitive()));
} else {
prim.reset(new inner_product_forward(prim_desc,
getParentEdgeAt(0)->getMemory().GetPrimitive(),
internalBlobMemory[0]->GetPrimitive(),
getChildEdgeAt(0)->getMemory().GetPrimitive()));
}
}
bool MKLDNNFullyConnectedNode::created() const {
return getType() == FullyConnected;
}
memory::format MKLDNNFullyConnectedNode::weightsFormatForSrcFormat(memory::format sourceFormat) {
switch (sourceFormat) {
case memory::format::x:
return memory::format::x;
case memory::format::nc:
return memory::format::oi;
case memory::format::nchw:
return memory::format::oihw;
case memory::format::nChw8c:
return memory::format::oIhw8i;
case memory::format::nChw16c:
return memory::format::oIhw16i;
default:
THROW_IE_EXCEPTION << "Unsupported source format for node " << getName();
}
}
const std::vector<impl_desc_type>& MKLDNNFullyConnectedNode::getPrimitivesPriority() {
std::vector<impl_desc_type> priorities = {
impl_desc_type::unknown,
impl_desc_type::gemm_blas,
impl_desc_type::gemm_avx512,
impl_desc_type::gemm_avx2,
impl_desc_type::gemm_sse42,
impl_desc_type::gemm_any,
impl_desc_type::gemm,
impl_desc_type::jit_uni_dw,
impl_desc_type::jit_uni_1x1,
impl_desc_type::jit_uni,
impl_desc_type::jit_avx512_dw,
impl_desc_type::jit_avx512_1x1,
impl_desc_type::jit_avx512,
impl_desc_type::jit_avx2_dw,
impl_desc_type::jit_avx2_1x1,
impl_desc_type::jit_avx2,
impl_desc_type::jit_sse42_dw,
impl_desc_type::jit_sse42_1x1,
impl_desc_type::jit_sse42,
impl_desc_type::ref,
};
for (const auto& impl : priorities) {
if (std::find(implPriorities.begin(), implPriorities.end(), impl) == implPriorities.end())
implPriorities.push_back(impl);
}
return implPriorities;
}
void MKLDNNFullyConnectedNode::createDescriptor(const std::vector<InferenceEngine::TensorDesc> &inputDesc,
const std::vector<InferenceEngine::TensorDesc> &outputDesc) {
MKLDNNMemoryDesc in_candidate(inputDesc[0]);
MKLDNNMemoryDesc out_candidate(outputDesc[0]);
memory::format weights_fmt = weightsFormatForSrcFormat(in_candidate.getFormat());
MKLDNNMemoryDesc wgh_candidate(MKLDNNDims(weightsDims), in_candidate.getDataType(), weights_fmt);
MKLDNNMemoryDesc bias_candidate(MKLDNNDims(biasesDims), in_candidate.getDataType(), memory::any);
if (internalBlobs.size() > 1) {
MKLDNNDescriptor desc(std::shared_ptr<inner_product_forward::desc>(
new inner_product_forward::desc(prop_kind::forward_scoring, in_candidate, wgh_candidate,
bias_candidate, out_candidate)));
descs.push_back(desc);
} else {
MKLDNNDescriptor desc(std::shared_ptr<inner_product_forward::desc>(
new inner_product_forward::desc(prop_kind::forward_scoring, in_candidate, wgh_candidate,
out_candidate)));
descs.push_back(desc);
}
}
|
