forked from google/XNNPACK
-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathconvolution-test-helpers.h
239 lines (227 loc) · 7.18 KB
/
convolution-test-helpers.h
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
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
// Copyright 2022 Google LLC
//
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree.
#pragma once
#include <cstddef>
#include <cstdint>
#include <vector>
#include "xnnpack/quantization.h"
namespace xnnpack {
void compute_convolution_qd8_f32_qc8w_reference_results(
size_t batch_size, size_t output_height, size_t output_width,
size_t input_height, size_t input_width, size_t input_padding_top,
size_t input_padding_right, size_t input_padding_bottom,
size_t input_padding_left, size_t kernel_height, size_t kernel_width,
size_t subsampling_height, size_t subsampling_width, size_t dilation_height,
size_t dilation_width, size_t groups, size_t group_input_channels,
size_t group_output_channels, size_t input_channel_stride,
const std::vector<int8_t>& input, const std::vector<int8_t>& filter,
const std::vector<float>& filter_scale,
const std::vector<xnn_qd8_quantization_params>& quantization_params,
std::vector<float>& output, bool has_bias, const std::vector<float>& bias);
void compute_convolution_qs8_reference_results(
size_t batch_size,
size_t output_height,
size_t output_width,
size_t input_height,
size_t input_width,
size_t input_padding_top,
size_t input_padding_right,
size_t input_padding_bottom,
size_t input_padding_left,
size_t kernel_height,
size_t kernel_width,
size_t subsampling_height,
size_t subsampling_width,
size_t dilation_height,
size_t dilation_width,
size_t groups,
size_t group_input_channels,
size_t group_output_channels,
size_t input_channel_stride,
int8_t input_zero_point,
const std::vector<int8_t>& input,
const std::vector<int8_t>& filter,
std::vector<int32_t>& accumulators,
bool has_bias,
const std::vector<int32_t>& bias);
void compute_convolution_qs8_reference_results(
size_t batch_size,
size_t output_height,
size_t output_width,
size_t input_height,
size_t input_width,
size_t input_padding_top,
size_t input_padding_right,
size_t input_padding_bottom,
size_t input_padding_left,
size_t kernel_height,
size_t kernel_width,
size_t subsampling_height,
size_t subsampling_width,
size_t dilation_height,
size_t dilation_width,
size_t groups,
size_t group_input_channels,
size_t group_output_channels,
int8_t input_zero_point,
const std::vector<int8_t>& input,
const std::vector<int8_t>& filter,
std::vector<int32_t>& accumulators,
bool has_bias,
const std::vector<int32_t>& bias);
void compute_convolution_qu8_reference_results(
size_t batch_size,
size_t output_height,
size_t output_width,
size_t input_height,
size_t input_width,
size_t input_padding_top,
size_t input_padding_right,
size_t input_padding_bottom,
size_t input_padding_left,
size_t kernel_height,
size_t kernel_width,
size_t subsampling_height,
size_t subsampling_width,
size_t dilation_height,
size_t dilation_width,
size_t groups,
size_t group_input_channels,
size_t group_output_channels,
uint8_t input_zero_point,
uint8_t kernel_zero_point,
const std::vector<uint8_t>& input,
const std::vector<uint8_t>& filter,
std::vector<int32_t>& accumulators,
bool has_bias,
const std::vector<int32_t>& bias);
void compute_convolution_qu8_reference_results(
size_t batch_size,
size_t output_height,
size_t output_width,
size_t input_height,
size_t input_width,
size_t input_padding_top,
size_t input_padding_right,
size_t input_padding_bottom,
size_t input_padding_left,
size_t kernel_height,
size_t kernel_width,
size_t subsampling_height,
size_t subsampling_width,
size_t dilation_height,
size_t dilation_width,
size_t groups,
size_t group_input_channels,
size_t group_output_channels,
size_t input_channel_stride,
uint8_t input_zero_point,
uint8_t kernel_zero_point,
const std::vector<uint8_t>& input,
const std::vector<uint8_t>& filter,
std::vector<int32_t>& accumulators,
bool has_bias,
const std::vector<int32_t>& bias);
void compute_depthwise_convolution_qs8_reference_results(
size_t batch_size,
size_t output_height,
size_t output_width,
size_t input_height,
size_t input_width,
size_t input_padding_top,
size_t input_padding_right,
size_t input_padding_bottom,
size_t input_padding_left,
size_t kernel_height,
size_t kernel_width,
size_t subsampling_height,
size_t subsampling_width,
size_t dilation_height,
size_t dilation_width,
size_t input_channels,
size_t depth_multiplier,
size_t input_channel_stride,
int8_t input_zero_point,
const std::vector<int8_t>& input,
const std::vector<int8_t>& filter,
std::vector<int32_t>& accumulators,
bool has_bias,
const std::vector<int32_t>& bias);
void compute_depthwise_convolution_qs8_reference_results(
size_t batch_size,
size_t output_height,
size_t output_width,
size_t input_height,
size_t input_width,
size_t input_padding_top,
size_t input_padding_right,
size_t input_padding_bottom,
size_t input_padding_left,
size_t kernel_height,
size_t kernel_width,
size_t subsampling_height,
size_t subsampling_width,
size_t dilation_height,
size_t dilation_width,
size_t input_channels,
size_t depth_multiplier,
int8_t input_zero_point,
const std::vector<int8_t>& input,
const std::vector<int8_t>& filter,
std::vector<int32_t>& accumulators,
bool has_bias,
const std::vector<int32_t>& bias);
void compute_depthwise_convolution_qu8_reference_results(
size_t batch_size,
size_t output_height,
size_t output_width,
size_t input_height,
size_t input_width,
size_t input_padding_top,
size_t input_padding_right,
size_t input_padding_bottom,
size_t input_padding_left,
size_t kernel_height,
size_t kernel_width,
size_t subsampling_height,
size_t subsampling_width,
size_t dilation_height,
size_t dilation_width,
size_t input_channels,
size_t depth_multiplier,
size_t input_channel_stride,
uint8_t input_zero_point,
uint8_t kernel_zero_point,
const std::vector<uint8_t>& input,
const std::vector<uint8_t>& filter,
std::vector<int32_t>& accumulators,
bool has_bias,
const std::vector<int32_t>& bias);
void compute_depthwise_convolution_qu8_reference_results(
size_t batch_size,
size_t output_height,
size_t output_width,
size_t input_height,
size_t input_width,
size_t input_padding_top,
size_t input_padding_right,
size_t input_padding_bottom,
size_t input_padding_left,
size_t kernel_height,
size_t kernel_width,
size_t subsampling_height,
size_t subsampling_width,
size_t dilation_height,
size_t dilation_width,
size_t input_channels,
size_t depth_multiplier,
uint8_t input_zero_point,
uint8_t kernel_zero_point,
const std::vector<uint8_t>& input,
const std::vector<uint8_t>& filter,
std::vector<int32_t>& accumulators,
bool has_bias,
const std::vector<int32_t>& bias);
} // namespace xnnpack