|
a |
|
b/ViTPose/tests/test_onnx.py |
|
|
1 |
# Copyright (c) OpenMMLab. All rights reserved. |
|
|
2 |
import os.path as osp |
|
|
3 |
import tempfile |
|
|
4 |
|
|
|
5 |
import torch.nn as nn |
|
|
6 |
|
|
|
7 |
from tools.deployment.pytorch2onnx import _convert_batchnorm, pytorch2onnx |
|
|
8 |
|
|
|
9 |
|
|
|
10 |
class DummyModel(nn.Module): |
|
|
11 |
|
|
|
12 |
def __init__(self): |
|
|
13 |
super().__init__() |
|
|
14 |
self.conv = nn.Conv3d(1, 2, 1) |
|
|
15 |
self.bn = nn.SyncBatchNorm(2) |
|
|
16 |
|
|
|
17 |
def forward(self, x): |
|
|
18 |
return self.bn(self.conv(x)) |
|
|
19 |
|
|
|
20 |
def forward_dummy(self, x): |
|
|
21 |
return (self.forward(x), ) |
|
|
22 |
|
|
|
23 |
|
|
|
24 |
def test_onnx_exporting(): |
|
|
25 |
with tempfile.TemporaryDirectory() as tmpdir: |
|
|
26 |
out_file = osp.join(tmpdir, 'tmp.onnx') |
|
|
27 |
model = DummyModel() |
|
|
28 |
model = _convert_batchnorm(model) |
|
|
29 |
# test exporting |
|
|
30 |
pytorch2onnx(model, (1, 1, 1, 1, 1), output_file=out_file) |