[GitHub] [incubator-mxnet] MrChengmo commented on issue #19949: DistributeTraining throw "dmlc::Error" when using nn.Embedding(sparse_grad=True)

[GitBox]

MrChengmo commented on issue #19949:
URL: 
https://github.com/apache/incubator-mxnet/issues/19949#issuecomment-785559171


   > Do you observe the Storage fallback also on single machine? Or does it 
only occur in the distributed setting?
   
   - Set `kv_store="local"` & `nn.Embedding(sparse_gard)`,  works well on 
single machine (distributed_traing=False), use command `python -u train.py `.
   - Set `kv_store="dist_async"` & `nn.Embedding(sparse_gard)`,  set 
`distributed_traing=True`, use command `bash local_cluster` to simulate 
distributed training， throw error


----------------------------------------------------------------
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



---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@mxnet.apache.org
For additional commands, e-mail: issues-h...@mxnet.apache.org