Please read before posting a question:

For questions about H2O software features, "where can I find an example of [x]?, "how can I use GPUs with H2O?", etc (questions that do not involve any user code), or to communicate with H2O developers directly about a particular issue, please post those questions here on h2ostream.
For code questions, we encourage you to post a Minimal, Complete and Verifiable Example (MCVE) at StackOverflow.com. If you are new to Stack Overflow, please visit the MCVE link to ensure you ask your question properly and that it contains all the necessary information (or if may be flagged & closed). Please use the h2o and/or sparkling-water tags.
For algorithm or data science questions ("what causes overfitting in GBMs?"), please use Cross Validated (on Stack Exchange). If the question is specific to the H2O implementations of algorithms, you can use the h2o tag.
For bug reports or feature requests, you can file those directly in our issue tracking system located here: http://jira.h2o.ai When filing a bug report, please make sure to include a reproducible example if possible and H2O version & client (R, Py, etc) information. You must create a free JIRA account using your email address to file a report.
IF IN DOUBT about which stack exchange site to post to, choose stack overflow.

Before posting a question on either site, please do the following:

Search h2ostream or Stack Overflow to see if your question has already been answered.
Search the H2O User Guide or any of the other many documentation resources at docs.h2o.ai.
Do not double-post your question: choose h2ostream --or-- Stack Overflow, but not both.

Notice to h2ostream users:
For every post please provide the following at the top of your post:

1) What version of H2O are you using.
2) Specify the type of machine your using (i.e. OS X 10.11.4, Windows 10, etc).
3) Specify what language you are working in and what version (i.e. Python 2.7, Spark 1.6.1, etc)
4) The code you were executing when you received an error message (please provide a reproducible example if possible).
5) Copy and paste in your error message.
6) Type of data you are using (if applicable).

Deep learning issue

roimi omosewo

Tom Kraljevic