-
Notifications
You must be signed in to change notification settings - Fork 2.5k
/
Copy pathtest_convolution_mode.py
138 lines (126 loc) · 8.44 KB
/
test_convolution_mode.py
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
# Copyright (C) 2018-2025 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
import pytest
from pytorch_layer_test_class import PytorchLayerTest
class TestConv2D(PytorchLayerTest):
def _prepare_input(self, ndim=4):
import numpy as np
input_shape = (1, 3, 10, 10, 10)
return (np.random.randn(*input_shape[:ndim]).astype(np.float32),)
def create_model(self, weights_shape, strides, pads, dilations, groups, bias):
import torch
class aten_convolution_mode(torch.nn.Module):
def __init__(self):
super(aten_convolution_mode, self).__init__()
self.weight = torch.randn(weights_shape)
self.bias = None
if bias:
self.bias = torch.randn(weights_shape[0])
self.strides = strides
self.pads = pads
self.dilations = dilations
self.groups = groups
def forward(self, x):
return torch._convolution_mode(x, self.weight, self.bias, self.strides, self.pads, self.dilations,
self.groups)
ref_net = None
return aten_convolution_mode(), ref_net, "aten::_convolution_mode"
@pytest.mark.parametrize("params",
[
{'weights_shape': [1, 3, 3], 'strides': [1], 'pads': 'same', 'dilations': [1],
'groups': 1},
{'weights_shape': [1, 3, 3], 'strides': [1], 'pads': 'valid', 'dilations': [1],
'groups': 1},
{'weights_shape': [1, 3, 3], 'strides': [1], 'pads': 'same', 'dilations': [2],
'groups': 1},
{'weights_shape': [1, 3, 3], 'strides': [1], 'pads': 'valid', 'dilations': [2],
'groups': 1},
{'weights_shape': [3, 1, 1], 'strides': [1], 'pads': 'same', 'dilations': [1],
'groups': 3},
{'weights_shape': [3, 1, 1], 'strides': [1], 'pads': 'valid', 'dilations': [1],
'groups': 3},
{'weights_shape': [1, 3, 3], 'strides': [2], 'pads': 'valid', 'dilations': [1],
'groups': 1},
{'weights_shape': [1, 3, 3], 'strides': [2], 'pads': 'valid', 'dilations': [2],
'groups': 1},
{'weights_shape': [3, 1, 1], 'strides': [1], 'pads': 'same', 'dilations': [2],
'groups': 3},
{'weights_shape': [3, 1, 1], 'strides': [1], 'pads': 'valid', 'dilations': [2],
'groups': 3},
])
@pytest.mark.parametrize("bias", [True, False])
@pytest.mark.nightly
@pytest.mark.precommit
def test_convolution_mode_1d(self, params, bias, ie_device, precision, ir_version):
self._test(*self.create_model(**params, bias=bias),
ie_device, precision, ir_version, dynamic_shapes=params['groups'] == 1,
kwargs_to_prepare_input={'ndim': 3})
@pytest.mark.parametrize("params",
[
{'weights_shape': [1, 3, 3, 3], 'strides': [1, 1], 'pads': 'same', 'dilations': [1, 1],
'groups': 1},
{'weights_shape': [1, 3, 3, 3], 'strides': [1, 1], 'pads': 'valid',
'dilations': [1, 1], 'groups': 1},
{'weights_shape': [1, 3, 3, 3], 'strides': [1, 1], 'pads': 'same', 'dilations': [2, 2],
'groups': 1},
{'weights_shape': [1, 3, 3, 3], 'strides': [1, 1], 'pads': 'valid',
'dilations': [2, 2], 'groups': 1},
{'weights_shape': [3, 1, 1, 1], 'strides': [1, 1], 'pads': 'same', 'dilations': [1, 1],
'groups': 3},
{'weights_shape': [3, 1, 1, 1], 'strides': [1, 1], 'pads': 'valid',
'dilations': [1, 1], 'groups': 3},
{'weights_shape': [1, 3, 3, 3], 'strides': [2, 2], 'pads': 'valid',
'dilations': [1, 1], 'groups': 1},
{'weights_shape': [1, 3, 3, 3], 'strides': [2, 2], 'pads': 'valid',
'dilations': [2, 2], 'groups': 1},
{'weights_shape': [1, 3, 3, 3], 'strides': [2, 1], 'pads': 'valid',
'dilations': [1, 1], 'groups': 1},
{'weights_shape': [3, 1, 1, 1], 'strides': [2, 2], 'pads': 'valid',
'dilations': [1, 1], 'groups': 3},
{'weights_shape': [3, 1, 1, 1], 'strides': [2, 2], 'pads': 'valid',
'dilations': [2, 2], 'groups': 3},
{'weights_shape': [3, 1, 1, 1], 'strides': [2, 1], 'pads': 'valid',
'dilations': [1, 1], 'groups': 3},
{'weights_shape': [3, 1, 1, 1], 'strides': [1, 1], 'pads': 'same', 'dilations': [2, 1],
'groups': 3},
{'weights_shape': [3, 1, 1, 1], 'strides': [1, 1], 'pads': 'valid',
'dilations': [2, 1], 'groups': 3},
{'weights_shape': [3, 1, 1, 1], 'strides': [1, 1], 'pads': 'same', 'dilations': [2, 2],
'groups': 3},
{'weights_shape': [3, 1, 1, 1], 'strides': [1, 1], 'pads': 'valid',
'dilations': [2, 2], 'groups': 3},
])
@pytest.mark.parametrize("bias", [True, False])
@pytest.mark.nightly
@pytest.mark.precommit
def test_convolution_mode_2d(self, params, bias, ie_device, precision, ir_version):
self._test(*self.create_model(**params, bias=bias),
ie_device, precision, ir_version, dynamic_shapes=params['groups'] == 1)
@pytest.mark.parametrize("params",
[
{'weights_shape': [1, 3, 3, 3, 3], 'strides': [1, 1, 1], 'pads': 'same',
'dilations': [1, 1, 1], 'groups': 1},
{'weights_shape': [1, 3, 3, 3, 3], 'strides': [1, 1, 1], 'pads': 'valid',
'dilations': [1, 1, 1], 'groups': 1},
{'weights_shape': [3, 1, 1, 1, 1], 'strides': [1, 1, 1], 'pads': 'same',
'dilations': [1, 1, 1], 'groups': 3},
{'weights_shape': [3, 1, 1, 1, 1], 'strides': [1, 1, 1], 'pads': 'valid',
'dilations': [1, 1, 1], 'groups': 3},
{'weights_shape': [1, 3, 3, 3, 3], 'strides': [2, 2, 1], 'pads': 'valid',
'dilations': [1, 1, 1], 'groups': 1},
{'weights_shape': [1, 3, 3, 3, 3], 'strides': [2, 2, 2], 'pads': 'valid',
'dilations': [1, 1, 1], 'groups': 1},
{'weights_shape': [1, 3, 3, 3, 3], 'strides': [2, 2, 2], 'pads': 'valid',
'dilations': [2, 2, 2], 'groups': 1},
{'weights_shape': [3, 1, 1, 1, 1], 'strides': [1, 1, 1], 'pads': 'same',
'dilations': [2, 1, 2], 'groups': 3},
{'weights_shape': [3, 1, 1, 1, 1], 'strides': [1, 1, 1], 'pads': 'valid',
'dilations': [2, 1, 2], 'groups': 3},
])
@pytest.mark.parametrize("bias", [True, False])
@pytest.mark.nightly
@pytest.mark.precommit
def test_convolution_mode_3d(self, params, bias, ie_device, precision, ir_version):
self._test(*self.create_model(**params, bias=bias),
ie_device, precision, ir_version, dynamic_shapes=params['groups'] == 1,
kwargs_to_prepare_input={'ndim': 5})