[jira] [Updated] (SYSTEMML-1774) Improve Parfor parallelism for deep learning

["Fei Hu (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/SYSTEMML-1774?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Fei Hu updated SYSTEMML-1774:
-----------------------------
    External issue URL:   (was: https://issues.apache.org/jira/browse/SYSTEMML-1760)

> Improve Parfor parallelism for deep learning
> --------------------------------------------
>
>                 Key: SYSTEMML-1774
>                 URL: https://issues.apache.org/jira/browse/SYSTEMML-1774
>             Project: SystemML
>          Issue Type: Improvement
>          Components: Algorithms
>    Affects Versions: SystemML 1.0
>            Reporter: Fei Hu
>              Labels: deeplearning
>
> When running the  [distributed MNIST LeNet example | https://github.com/apache/systemml/blob/master/scripts/nn/examples/mnist_lenet_distrib_sgd.dml], each mini-batch could ideally run in parallel without interaction. We try to force {{parfor (j in 1:parallel_batches)}} at line 137 of {{nn/examples/mnist_lenet_distrib_sgd.dml}} to use {{REMOTE_SPARK}} mode, but got some errors about {{org.apache.sysml.runtime.DMLRuntimeException: Not supported: Instructions of type other than CP instructions}}. More log information can be found at the following comments. One example of the errors is that at the convolution layer, we need to randomly generate a matrix, but SystemML choose {{RandSPInstruction}} instead of {{DataGenCPInstruction}}, which may be because SystemML could not determine the row number of the matrix. For this distributed MNIST LeNet  example, using CPInstruction may achieve better performance. 



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)