You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mxnet.apache.org by GitBox <gi...@apache.org> on 2019/05/13 12:15:51 UTC

[GitHub] [incubator-mxnet] jens-maus commented on issue #5296: KVStore error with certain optimizer settings - R-package

jens-maus commented on issue #5296: KVStore error with certain optimizer settings - R-package
URL: https://github.com/apache/incubator-mxnet/issues/5296#issuecomment-491798252
 
 
   @jeremiedb I just encountered the same issue here with using the R package and multiple GPU training: The training broke with:
   
   ```
   Batch size: 100 
   Auto-select kvstore type = local_update_cpu
   Start training with 2 devices
   Error in kvstore$set.optimizer(optimizer) : 
     kvstore.cc:124: RCheck failed: names.size() == 2 && names[0] == "create.state" && names[1] == "update" Invalid optimizer
   Calls: unetDelineation ... <Anonymous> -> mx.model.train2 -> <Anonymous> -> .External
   Execution halted
   ```
   
   However, forcing the kvstore to `local_allreduce_device` seem to have resolved the issue. Is there anything that I should be concerned of or does the R-package still lack some fixes to work with the default `local_update_cpu`?

----------------------------------------------------------------
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:
users@infra.apache.org


With regards,
Apache Git Services