You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mxnet.apache.org by gi...@git.apache.org on 2017/08/23 11:15:40 UTC

[GitHub] cjolivier01 commented on issue #7559: nightly build stochastically choose optimizer

cjolivier01 commented on issue #7559: nightly build stochastically choose optimizer
URL: https://github.com/apache/incubator-mxnet/pull/7559#issuecomment-324107424
 
 
   Yes, it would be caught by any optimizer other than sgd.  When he test is
   run, the full set of options is displayed, so copy-pasting that will
   reproduce.
   
   On Tue, Aug 22, 2017 at 10:57 AM, Madan Jampani <no...@github.com>
   wrote:
   
   > *@madjam* commented on this pull request.
   > ------------------------------
   >
   > In tests/nightly/test_all.sh
   > <https://github.com/apache/incubator-mxnet/pull/7559#discussion_r134556457>
   > :
   >
   > > @@ -72,9 +80,15 @@ check_val() {
   >  example_dir=../../example/image-classification
   >  # python: lenet + mnist
   >  test_lenet() {
   > -    python $example_dir/train_mnist.py \
   > -        --data-dir `pwd`/data/mnist/ --network lenet --gpus $gpus --num-epochs 10 \
   > -        2>&1 | tee log
   > +    optimizer=$(get_random_optimizer)
   >
   > My concern was that it might be hard to reproduce failures. Granted that
   > someone who knows the test and looks at the specific failure should be able
   > to figure it out. For example, the bug you fixed yesterday, would that have
   > been immediately caught by this test?
   >
   > I do see your concern about longer run time.
   >
   > ?
   > You are receiving this because you authored the thread.
   > Reply to this email directly, view it on GitHub
   > <https://github.com/apache/incubator-mxnet/pull/7559#discussion_r134556457>,
   > or mute the thread
   > <https://github.com/notifications/unsubscribe-auth/AKts_e1_aeXq_y2VvUJIbnlPNm4Z1Ru0ks5saxaWgaJpZM4O-32h>
   > .
   >
   
 
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services