-
Notifications
You must be signed in to change notification settings - Fork 2.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[IE TEST] LRN tests fixed params #743
Conversation
|
||
auto ngPrc = FuncTestUtils::PrecisionUtils::convertIE2nGraphPrc(netPrecision); | ||
auto params = ngraph::builder::makeParams(ngPrc, {inputShapes}); | ||
auto paramIn = | ||
ngraph::helpers::convert2OutputVector(ngraph::helpers::castOps2Nodes<ngraph::op::Parameter>(params)); | ||
|
||
auto lrn = std::make_shared<ngraph::opset1::LRN>(paramIn[0], alpha, beta, bias, size); | ||
std::vector<int64_t> axes; | ||
if (region == "same") { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I propose to parameterize the test directly with the axes
parameter instead of some key-word which defines certain behavior of the test.
There are two advantages to such an approach. First of all, it is to be aligned with opset specification which describes LRN parameters and inputs. And also to make it possible to validate cases different from currently possible axes values {1}
and {2,3,4,...}
d14045b
to
204ec85
Compare
please, take a look @mikhail-treskin |
inference-engine/tests/functional/plugin/shared/src/single_layer_tests/lrn.cpp
Outdated
Show resolved
Hide resolved
e7b54e5
to
7dd0faf
Compare
LGTM |
@mikhail-treskin Please review changes again |
7dd0faf
to
e77d79e
Compare
e77d79e
to
7e9d07d
Compare
7e9d07d
to
91f0437
Compare
Now in LRN tests using
size_t beta, bias
, butngraph::op::LRN
takesdouble beta, bias
. We also need to add type of normalization channel algorithmOutput<Node> axes
.please, take a look @iefode