Pawel Raasz
Pawel Raasz
> @praasz, please comment on value outside Kaldi. Kaldi support was stopped a year ago. I see setting shapes for variables reasonable in the context that we started to produce...
@itikhono , @jane-intel could you review the PR?
@jane-intel could check if anything from ticket [CVS-94507](https://jira.devtools.intel.com/browse/CVS-94507) is missing. I think current shape inference works correctly and recalculate concatenation axis on each use. Added test to verify it.
> @praasz the problem was about this: https://github.com/openvinotoolkit/openvino/blob/master/src/core/src/op/concat.cpp#L45 https://github.com/openvinotoolkit/openvino/blob/master/src/core/include/openvino/op/concat.hpp#L43 https://github.com/openvinotoolkit/openvino/blob/master/src/core/include/openvino/op/concat.hpp#L60:L63 > ... Sure we can review if setter/getter could be marked as deprecated, I assume deprecation message should be like...
> General questions: > > * Why did we pick to use strings to identify variables? We could've used indices or instances instead. It looks like name is user friendly...
> @praasz should we leave it as part of C++ API? It looks like is not required to make it in C++ API, the current API can be used in...
@akuporos could you check bindings changes?
@mryzhov Could you check if ConvertPrecision should insert Clamp operator (or have special mode) in cases where simple value conversion may produce incorrect value?
build_jenkins