[
https://issues.apache.org/jira/browse/MADLIB-1228?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16435890#comment-16435890
]
ASF GitHub Bot commented on MADLIB-1228:
----------------------------------------
GitHub user njayaram2 opened a pull request:
https://github.com/apache/madlib/pull/262
MLP: Print loss of all iterations in verbose mode
JIRA: MADLIB-1228
- When verbose mode was set in mlp classification and regression training,
the loss was printed only for 2 -> n-1 iterations. This commit fixes it
and now prints loss for 1->n iterations (or whatever iteration
convergence happens in).
- This commit also fixes another bug that was discovered in
__utils_normalize_data_grouping in utilis_regularization.py_in. The
issue was that if grouping_col had more than one column specified in
it, the query used for normalization used to error out due to
ambiguity in column names referenced. This commit prefixes the table
name while accessing the grouping column names now.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/madlib/madlib bugfix/mlp/verbose-iteration-num
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/madlib/pull/262.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #262
----
commit 6c0d19c027c5e5a750258d22936c894797e79f4f
Author: Nandish Jayaram <njayaram@...>
Date: 2018-04-12T00:22:18Z
MLP: Print loss of all iterations in verbose mode
JIRA: MADLIB-1228
- When verbose mode was set in mlp classification and regression training,
the loss was printed only for 2 -> n-1 iterations. This commit fixes it
and now prints loss for 1->n iterations (or whatever iteration
convergence happens in).
- This commit also fixes another bug that was discovered in
__utils_normalize_data_grouping in utilis_regularization.py_in. The
issue was that if grouping_col had more than one column specified in
it, the query used for normalization used to error out due to
ambiguity in column names referenced. This commit prefixes the table
name while accessing the grouping column names now.
----
> MLP verbose - print loss for all iterations in verbose mode
> -----------------------------------------------------------
>
> Key: MADLIB-1228
> URL: https://issues.apache.org/jira/browse/MADLIB-1228
> Project: Apache MADlib
> Issue Type: Bug
> Components: Module: Neural Networks
> Reporter: Nandish Jayaram
> Priority: Major
> Fix For: v1.14
>
>
> For MLP running in verbose mode, for n iterations, currently we print loss
> for 2->n-1. We should instead print out loss for 1->n when:
> * running for max iterations &
> * when converges (i.e., does not run to max iterations, so print loss for
> 1->converging_iteration)
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
