Hi,

you should not have to do that...

What happens if you run
  ./experiment.perl -config steps/1/config.1

This will create a new run, but it is the cleanest way to run with a config
file.

-phi

On Fri, Jan 8, 2016 at 3:10 PM, Nicholas Ruiz <nicr...@fbk.eu> wrote:

> Got it. You have to explicitly run the reporting step, after
> removing steps/1/REPORTING_report.1* and evaluation/report.1
>
> ./experiment.perl -continue 1 -final-step REPORTING:report -exec
>
> Although I'm not sure why this is different from the default case, which
> should always run reporting...
>
> zınɹ ʞɔıu
>
> On Fri, Jan 8, 2016 at 2:43 PM, Nicholas Ruiz <nicr...@fbk.eu> wrote:
>
>> Actually, I'm halfway to the solution. I did the steps below:
>>
>> #1: Make sure you get your variable names correct. EMS has a boatload of
>> variables so it's hard to remember what's "output" and what's "reference"
>>
>> #2: Don't have a million config files in a directory. You might be
>> executing the wrong config file. :) [still trying to figure out how not to
>> have one-off configs everywhere]
>>
>> #3: Edit steps/x/config.x instead of editing the original config.
>>
>> #4: Thank the EMS developers for their hard work.
>>
>> However, there's one piece missing. Philipp said:
>> "In you case you also need to remove the steps/1/REPORTING_report.1*
>> step, so it is re-run."
>>
>> I tried deleting the files above, as well as deleting
>> evaluation/report.1, but the reporting module isn't executing. No new 
>> steps/1/REPORTING_report.1*
>> files are being generated, although the dependency list shows
>> 0 REPORTING:report -> run
>>
>> zınɹ ʞɔıu
>>
>> On Fri, Jan 8, 2016 at 1:26 PM, Matthias Huck <mh...@inf.ed.ac.uk> wrote:
>>
>>> So, what has been the proper solution?
>>>
>>> On Fri, 2016-01-08 at 13:20 -0500, Nicholas Ruiz wrote:
>>> > Thanks everyone, it's working now.
>>> >
>>> > zınɹ ʞɔıu
>>>
>>>
>>>
>>>
>>> --
>>> The University of Edinburgh is a charitable body, registered in
>>> Scotland, with registration number SC005336.
>>>
>>>
>>
>
_______________________________________________
Moses-support mailing list
Moses-support@mit.edu
http://mailman.mit.edu/mailman/listinfo/moses-support

Reply via email to