piyushghai commented on issue #11626: [MXNET-651] MXNet Model Backwards
Compatibility Checker
URL: https://github.com/apache/incubator-mxnet/pull/11626#issuecomment-408271045
@marcoabreu The Jenkins CI [1] build is giving an error on : the import
statement for MXNet under the Inference
piyushghai commented on issue #11626: [MXNET-651] MXNet Model Backwards
Compatibility Checker
URL: https://github.com/apache/incubator-mxnet/pull/11626#issuecomment-405675518
@marcoabreu ping.
This is an automated message
piyushghai commented on issue #11626: [MXNET-651] MXNet Model Backwards
Compatibility Checker
URL: https://github.com/apache/incubator-mxnet/pull/11626#issuecomment-404623349
@marcoabreu I have removed real world datasets, and instead I am now using
randomly generated datasets. I have
piyushghai commented on issue #11626: [MXNET-651] MXNet Model Backwards
Compatibility Checker
URL: https://github.com/apache/incubator-mxnet/pull/11626#issuecomment-404271838
@ThomasDelteil What are your thoughts on using real world datasets v/s
random ones for this ?
piyushghai commented on issue #11626: [MXNET-651] MXNet Model Backwards
Compatibility Checker
URL: https://github.com/apache/incubator-mxnet/pull/11626#issuecomment-404268363
@marcoabreu I totally agree with you. I'm currently working adding toy
examples which use random weights on simple
piyushghai commented on issue #11626: [MXNET-651] MXNet Model Backwards
Compatibility Checker
URL: https://github.com/apache/incubator-mxnet/pull/11626#issuecomment-404054059
We cannot unittest it as we are not testing the various model save/load APIs
on the same version. We are trying to
piyushghai commented on issue #11626: [MXNET-651] MXNet Model Backwards
Compatibility Checker
URL: https://github.com/apache/incubator-mxnet/pull/11626#issuecomment-404053135
I am checking the following scenarios to save/load models :
1. Hybridized models import/export API from [Gluon
piyushghai commented on issue #11626: [MXNET-651] MXNet Model Backwards
Compatibility Checker
URL: https://github.com/apache/incubator-mxnet/pull/11626#issuecomment-404051396
@szha Here's what this PR does :
"Backward compatibility aims to check whether models trained on earlier