[GitHub] [incubator-mxnet] ptrendx commented on issue #16553: Fix for wrong reqs set after switching from training to inference

2019-10-23 Thread GitBox
ptrendx commented on issue #16553: Fix for wrong reqs set after switching from 
training to inference
URL: https://github.com/apache/incubator-mxnet/pull/16553#issuecomment-545516103
 
 
   As expected, test failed in the new test here: 
http://jenkins.mxnet-ci.amazon-ml.com/job/mxnet-validation/job/centos-gpu/job/PR-15167/50/display/redirect


This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


With regards,
Apache Git Services


[GitHub] [incubator-mxnet] ptrendx commented on issue #16553: Fix for wrong reqs set after switching from training to inference

2019-10-22 Thread GitBox
ptrendx commented on issue #16553: Fix for wrong reqs set after switching from 
training to inference
URL: https://github.com/apache/incubator-mxnet/pull/16553#issuecomment-545206420
 
 
   Ok, I added the test to #15167 that should fail because of this issue. I 
will move the string constants to `const static` members of cached op, merge 
this PR and then merge master to #15167 to show it fixes the problem.


This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


With regards,
Apache Git Services


[GitHub] [incubator-mxnet] ptrendx commented on issue #16553: Fix for wrong reqs set after switching from training to inference

2019-10-21 Thread GitBox
ptrendx commented on issue #16553: Fix for wrong reqs set after switching from 
training to inference
URL: https://github.com/apache/incubator-mxnet/pull/16553#issuecomment-544653387
 
 
   @eric-haibin-lin Yes, it had this bug since the beginning I believe, but I 
don't think there were any ops that would expose it until now.


This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


With regards,
Apache Git Services


[GitHub] [incubator-mxnet] ptrendx commented on issue #16553: Fix for wrong reqs set after switching from training to inference

2019-10-20 Thread GitBox
ptrendx commented on issue #16553: Fix for wrong reqs set after switching from 
training to inference
URL: https://github.com/apache/incubator-mxnet/pull/16553#issuecomment-544336113
 
 
   For 1: this error was found when working on pointwise fusion, which uses 
reqs to compile the right code. I don't know if there is any other operator 
which may not write to an output if there is no backward pass for it (I know 
`mx.sym.contrib.box_nms` would have that behavior, but the current 
implementation writes to the output only used for backward every time, without 
looking at the req value). I can add the test in pointwise fusion PR, show that 
it fails and then merge with master once this fix is in to show that it does 
not fail anymore.
   
   For 2: I agree, probably as `const static` members of `CachedOp`?


This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


With regards,
Apache Git Services