Hi,

just a comment: EMS (experiment.perl) is re-running steps for which the
script/binary has changed. So if you check out an update out of SVN, it will
notice that update. The timestamp is stored in *INFO files, so you could
(if you knew when you updated your versions) backtrack to the corresponding
SVN release.

-phi

On Wed, Sep 28, 2011 at 2:09 PM, Lane Schwartz <[email protected]> wrote:
> On Wed, Sep 28, 2011 at 5:55 AM, Barry Haddow
> <[email protected]> wrote:
>> Hi
>>
>> So it looks like a date-stamped release folder is not popular. It would be
>> possible to have a version stamp inserted into the binaries and scripts and
>> have this read by ems - would this be useful?
>
> Yes, it would be nice to be able to run the moses binary with some
> sort of --version flag to see what revision it was compiled from.
>
>
>> And would a 'make install' be of any use? It would mean that anyone who 
>> wanted
>> a date-stamped release could make one with --prefix, and would help users to
>> separate files needed at runtime from those needed at build time.
>
> Agreed.
>
>
>> If we do have a 'make install', should the scripts be squished into the bin
>> directory, or have their directory structure preserved? If we go for the
>> latter, then there would be the option of running them from a source 
>> directory
>> too.
>
>> On Tuesday 27 September 2011 23:53:29 Raphael Payen wrote:
>>> I like your idea of a VERSION.txt file. Maybe containing also the
>>> configure line.
>>>
>>> By the way, there is already a make install which is configurable with
>>> "configure --prefix=..." ; I dont understand if you want more than
>>> this.
>>>
>>> Raphael
>>>
>>> 2011/9/27 Lane Schwartz <[email protected]>:
>>> > I would like to see a true make install option. I don't care what the
>>> > default location is, as long as it's configurable.
>>> >
>>> > At the least, I would prefer to see the binaries (and preferably
>>> > scripts, too) installed to a bin directory within the moses dir
>>> > instead of being spread throughout the source directories.
>>> >
>>> > I agree that it can be useful to know what version you're running on.
>>> > But I don't think time stamped directories are the way to do it. I
>>> > would propose that make install creates a VERSION.txt file that lists
>>> > the svn version number, or some other version number.
>>> >
>>> > Lane
>>> >
>>> > On Tue, Sep 27, 2011 at 6:07 AM, Hieu Hoang <[email protected]> wrote:
>>> >> I'd go with Ondrej and avoid any release procedure and run everything
>>> >> in-situ. We can debug and commit scripts easier.
>>> >>
>>> >> Or have a very simple release that moves all executables and scripts
>>> >> into a bin directory. If the user wants to move/rename, it's their
>>> >> choice
>>> >>
>>> >> On 27 September 2011 16:30, Barry Haddow <[email protected]> wrote:
>>> >>> Hi
>>> >>>
>>> >>> I think it's useful to have a date-stamped release directory, so that
>>> >>> you can
>>> >>> work on the code and have experiments running, but keep tracking of
>>> >>> which version was used in which experiment. Many people at Edinburgh
>>> >>> make a copy of
>>> >>> the moses binary with the svn version number as suffix on the name.
>>> >>>
>>> >>> How about having a 'make install, which by default installs to a
>>> >>> date-stamped
>>> >>> directory, with the standard bin, lib etc. pattern, with all the
>>> >>> scripts under
>>> >>> bin? How big a change would that be?
>>> >>>
>>> >>> cheers - Barry
>>> >>>
>>> >>> On Friday 23 September 2011 08:46:06 Ondrej Bojar wrote:
>>> >>> > Hi,
>>> >>> >
>>> >>> > mert-moses-new.pl is outdated, isn't it? There should be only
>>> >>> > mert-moses.pl these days.
>>> >>> >
>>> >>> > The 'scripts releasing' is of mine and dates back to JHU workshop in
>>> >>> > 2006. Back then, we were hacking the scripts and running experiments
>>> >>> > at the same time, so we needed some track of what version of scripts
>>> >>> > was used for that particular failing experiment.
>>> >>> >
>>> >>> > As soon as anyone has some spare time I'd suggest:
>>> >>> >
>>> >>> > - delete the 'releasing' code in the Makefile
>>> >>> > - make sure the main make compiles everything: all moseses, all
>>> >>> > auxiliary binaries in scripts etc.
>>> >>> > - I'd avoid implementing 'make install', because all the tools in
>>> >>> > scripts know where their frieds are sitting in terms of relative
>>> >>> > paths. The install would need to preserve the complicated structure
>>> >>> > anyway so there's not much point in having the install as 'cp -r'
>>> >>> > does the same job.
>>> >>> >
>>> >>> > Cheers, O.
>>> >>> >
>>> >>> > On 09/23/2011 09:20 AM, Hieu Hoang wrote:
>>> >>> > > agreed, we were having an offlist moan about the same thing. The
>>> >>> > > separation between decoding&  training, and release date-stamp
>>> >>> > > thingy is
>>> >>> > > historical and quite silly.
>>> >>> > >
>>> >>> > > The directories&  release procedure has to be rationalised.
>>> >>> > >
>>> >>> > > Someone will do it eventually...
>>> >>> > >
>>> >>> > > On 23 September 2011 14:04, Joerg
>>> >>>
>>> >>> Tiedemann<[email protected]>wrote:
>>> >>> > >> By the way, what is the use of a date-stamped directory anyway?
>>> >>> > >> I find if rather disturbing that all the binaries and scripts are
>>> >>> > >> distributed all over the place.
>>> >>> > >> moses-cmd/src
>>> >>> > >> moses-chart-cmd/src
>>> >>> > >> misc
>>> >>> > >> scripts
>>> >>> > >> released scripts in a date-stamped directory of your choice
>>> >>> > >> ...
>>> >>> > >>
>>> >>> > >> J�rg
>>> >>> > >>
>>> >>> > >>
>>> >>> > >> On Thu, Sep 22, 2011 at 4:34 PM, Barry Haddow
>>> >>> > >>
>>> >>> > >> <[email protected]>  wrote:
>>> >>> > >>> Hi Hieu et al
>>> >>> > >>>
>>> >>> > >>> I replied to the OP on this, but forgot to CC it to the list.
>>> >>> > >>>
>>> >>> > >>> There was a change in mert some time in the summer (from Prague)
>>> >>> > >>> meaning
>>> >>> > >>
>>> >>> > >> that
>>> >>> > >>
>>> >>> > >>> if you run an old mert-moses.perl with a new mert binary, you get
>>> >>> > >>> an
>>> >>> > >>
>>> >>> > >> error
>>> >>> > >>
>>> >>> > >>> with exit code 3. I suspect this is the problem here since the
>>> >>> > >>
>>> >>> > >> scripts-rootdir
>>> >>> > >>
>>> >>> > >>> suggests the scripts are from March. Checking mert.log would
>>> >>> > >>> confirm the diagnosis.
>>> >>> > >>>
>>> >>> > >>> The solution is to rerun 'make release' in the scripts directory,
>>> >>> > >>> and
>>> >>> > >>> use
>>> >>> > >>
>>> >>> > >> the
>>> >>> > >>
>>> >>> > >>> new scripts-rootdir.
>>> >>> > >>>
>>> >>> > >>> As an aside, I should say that I'm not keen on our two-level make
>>> >>> > >>> setup,
>>> >>> > >>
>>> >>> > >> and
>>> >>> > >>
>>> >>> > >>> would like to find something simpler. Probably a one level make,
>>> >>> > >>> with a
>>> >>> > >>> conventional 'make install', but with the default being to
>>> >>> > >>> install in a
>>> >>> > >>
>>> >>> > >> date
>>> >>> > >>
>>> >>> > >>> stamped directory.
>>> >>> > >>>
>>> >>> > >>> cheers - Barry
>>> >>> > >>>
>>> >>> > >>> On Thursday 22 Sep 2011 14:59:26 Hieu Hoang wrote:
>>> >>> > >>>> i'm afraid i don't know what the problem is. The n-best-list and
>>> >>> > >>>> ini
>>> >>> > >>>> files look ok. I'm running mert and moses-chart that was checked
>>> >>> > >>>> out
>>> >>> > >>>> on the 13th September and ran ok.
>>> >>> > >>>>
>>> >>> > >>>> i'm not familiar with the mert code so i can't tell you if the
>>> >>> > >>>> changes
>>> >>> > >>>> you made are good. However, you're welcome to post the changes
>>> >>> > >>>> to the
>>> >>> > >>>> mailing list and someone might be know better than i do.
>>> >>> > >>>>
>>> >>> > >>>> On 22/09/2011 15:34, Prasanth K wrote:
>>> >>> > >>>>> Hi Hieu,
>>> >>> > >>>>>
>>> >>> > >>>>> I am attaching the following files:
>>> >>> > >>>>>
>>> >>> > >>>>> multi-threads_tune-run1.100best.out - first 1000 lines from the
>>> >>> > >>>>> multi-threaded decoder
>>> >>> > >>>>> single-threads_tune-run1.100best.out - the same from the
>>> >>> > >>>>> single-threaded decoder
>>> >>> > >>>>>
>>> >>> > >>>>> multi-threads_tune-run1.moses.ini - the ini file used at the
>>> >>> > >>>>> beginning of the tuning when using threads
>>> >>> > >>>>> single-threads_tune-run1.moses.ini - the ini file used at the
>>> >>> > >>>>> beginning of the tuning when threads were not used
>>> >>> > >>>>> single-threads_tune-run2.moses.ini - the ini file used at the
>>> >>> > >>>>> beginning of the second iteration when threads were not used
>>> >>> > >>>>>
>>> >>> > >>>>> - Prasanth
>>> >>> > >>>>>
>>> >>> > >>>>> On Thu, Sep 22, 2011 at 10:19 AM, Hieu
>>> >>> > >>>>> Hoang<[email protected] <mailto:[email protected]>>  wrote:
>>> >>> > >>>>>
>>> >>> > >>>>>      can you send me your ini file, and a few lines of the
>>> >>> > >>>>> n-best list. For accuracy, send them as attachements, not cut&
>>> >>> > >>>>>  paste into
>>> >>> > >>>>> the email.
>>> >>> > >>>>>
>>> >>> > >>>>>      I'll try&  see what the problem is
>>> >>> > >>>>>
>>> >>> > >>>>>
>>> >>> > >>>>>      On 22 September 2011 14:57, Prasanth
>>> >>> > >>>>> K<[email protected]
>>> >>> > >>>>>      <mailto:[email protected]>>  wrote:
>>> >>> > >>>>>
>>> >>> > >>>>>          Hi all,
>>> >>> > >>>>>
>>> >>> > >>>>>          I am facing the same error that Cyrine Nasri mentioned
>>> >>> > >>>>> in this thread.
>>> >>> > >>>>>
>>> >>> > >>>>>          I will try and give more information that what has
>>> >>> > >>>>> already
>>> >>> > >>>>>          been mentioned.
>>> >>> > >>>>>          1.  I was using a single-threaded version of moses
>>> >>> > >>>>> until earlier and was having no problem with the experiments
>>> >>> > >>>>> using
>>> >>> > >>
>>> >>> > >> EMS.
>>> >>> > >>
>>> >>> > >>>>>          2.  I recently shifted to a multi-threaded version,
>>> >>> > >>>>> and tried the same experiment again with EMS.
>>> >>> > >>>>>          This time, the tuning process crashes after a single
>>> >>> > >>>>> iteration with exactly the same error as mentioned below.
>>> >>> > >>>>>          3.  I have used 10 threads for decoding in the tuning
>>> >>> > >>>>> process, and am using a server with 32Gb ram for running the
>>> >>> > >>>>> experiments. (Might not be related, but just thought I should
>>> >>> > >>>>> mention!)
>>> >>> > >>>>>
>>> >>> > >>>>>          I am not sure if Cyrine's problem was with the
>>> >>> > >>>>> multi-threaded version as well, but could some one point out as
>>> >>> > >>>>> to what might be wrong in this picture ?
>>> >>> > >>>>>
>>> >>> > >>>>>          - Prasanth
>>> >>> > >>>>>
>>> >>> > >>>>>          On Wed, Jul 27, 2011 at 5:59 PM, Cyrine NASRI
>>> >>> > >>>>>
>>> >>> > >>>>>  <[email protected]<mailto:[email protected]>>
>>> >>> > >>
>>> >>> > >> wrote:
>>> >>> > >>>>>              Hello,
>>> >>> > >>>>>              I 'm trying to launch the mert using this command:
>>> >>> > >>>>>
>>> >>> > >>>>>              ./mert-moses-new.pl<http://mert-moses-new.pl>
>>> >>> > >>>>>              /users/parole/cnasri/moses_work/corpus/source2TOK
>>> >>> > >>>>>              /users/parole/cnasri/moses_work/corpus/ref2TOK
>>> >>> > >>>>>
>>> >>> > >>>>> /users/parole/cnasri/moses_work/moses/moses-cmd/src/moses
>>> >>> > >>
>>> >>> > >> /users/parole/cnasri/moses_work/moses-scripts/scripts-20110727-154
>>> >>> > >>3/trai n
>>> >>> > >>
>>> >>> > >>>>> ing/moses.ini --working-dir
>>> >>> > >>>>> /users/parole/cnasri/moses_work/tuning/
>>> >>> > >>>>> --mertdir /users/parole/cnasri/moses_work/moses/mert
>>> >>> > >>>>>
>>> >>> > >>>>>              But there is only one iteration, after it stops
>>> >>> > >>>>> and
>>> >>> > >>
>>> >>> > >> itshow:
>>> >>> > >>>>>              Peeking at the beginning of nbestlist to get order
>>> >>> > >>>>> of scores: run1.best100.out
>>> >>> > >>>>>              The decoder returns the scores in this order: d lm
>>> >>> > >>>>> w tm
>>> >>> > >>>>> tm
>>> >>> > >>
>>> >>> > >> tm
>>> >>> > >>
>>> >>> > >>>>>              Executing: gzip -f run1.best100.out
>>> >>> > >>>>>              Scoring the nbestlist.
>>> >>> > >>>>>              Executing:
>>> >>> > >>>>>
>>> >>> > >>>>>  /users/parole/cnasri/moses_work/moses/mert/extractor
>>> >>> > >>>>> --scconfig case:true --scfile run1.scores.dat --ffile
>>> >>> > >>>>> run1.features.dat -r
>>> >>> > >>>>>              /users/parole/cnasri/moses_work/corpus/ref2TOK -n
>>> >>> > >>>>>              run1.best100.out.gz>  extract.out 2>  extract.err
>>> >>> > >>>>>              Executing: \cp -f init.opt run1.init.opt
>>> >>> > >>>>>              Executing:
>>> >>> > >>>>> /users/parole/cnasri/moses_work/moses/mert/mert -d 6
>>> >>> > >>>>>  --scconfig case:true -n 20 --ffile run1.features.dat --scfile
>>> >>> > >>>>> run1.scores.dat --ifile run1.init.opt 2>  mert.log Exit code: 3
>>> >>> > >>>>>              Failed to run mert at ./mert-moses-new.pl
>>> >>> > >>>>>              <http://mert-moses-new.pl>  line 752.
>>> >>> > >>>>>               and in the mert log :
>>> >>> > >>>>>
>>> >>> > >>>>>              Seeding random numbers with system clock
>>> >>> > >>>>>
>>> >>> > >>>>>              :Too few minimum weights.
>>> >>> > >>>>>
>>> >>> > >>>>>              error could not initialize start point with
>>> >>> > >>>>>
>>> >>> > >>>>>              I have not idea how i resolve this problem.
>>> >>> > >>>>>              Any idea please?
>>> >>> > >>>>>
>>> >>> > >>>>>              Thank you
>>> >>> > >>>>>
>>> >>> > >>>>>
>>> >>> > >>>>>
>>> >>> > >>>>>              _______________________________________________
>>> >>> > >>>>>              Moses-support mailing list
>>> >>> > >>>>>
>>> >>> > >>>>>  [email protected]<mailto:[email protected]>
>>> >>> > >>>>> http://mailman.mit.edu/mailman/listinfo/moses-support
>>> >>> > >>>>>
>>> >>> > >>>>>
>>> >>> > >>>>>
>>> >>> > >>>>>
>>> >>> > >>>>>
>>> >>> > >>>>>            --- J.B.S. Haldane
>>> >>> > >>>>>
>>> >>> > >>>>>          _______________________________________________
>>> >>> > >>>>>          Moses-support mailing list
>>> >>> > >>>>>          [email protected]<mailto:[email protected]>
>>> >>> > >>>>>          http://mailman.mit.edu/mailman/listinfo/moses-support
>>> >>> > >>>>>
>>> >>> > >>>>>
>>> >>> > >>>>>
>>> >>> > >>>>>
>>> >>> > >>>>>
>>> >>> > >>>>>
>>> >>> > >>>>>    --- J.B.S. Haldane
>>> >>> > >>>
>>> >>> > >>> _______________________________________________
>>> >>> > >>> Moses-support mailing list
>>> >>> > >>> [email protected]
>>> >>> > >>> http://mailman.mit.edu/mailman/listinfo/moses-support
>>> >>> > >>
>>> >>> > >> --
>>> >>> > >>
>>> >>> > >>
>>> >>> > >> ******************************************************************
>>> >>> > >>****** ********** J�rg Tiedemann
>>> >>> > >> [email protected]
>>> >>> > >>   Dep. of Linguistics and Philology
>>> >>> > >> http://stp.lingfil.uu.se/~joerg/
>>> >>> > >>   Uppsala University                                  tel:  +46
>>> >>> > >> (0)18 -
>>> >>> > >> 471 1412
>>> >>> > >>   Box 635, SE-751 26 Uppsala/SWEDEN   fax: +46 (0)18 - 471 1094
>>> >>> > >
>>> >>> > > _______________________________________________
>>> >>> > > Moses-support mailing list
>>> >>> > > [email protected]
>>> >>> > > http://mailman.mit.edu/mailman/listinfo/moses-support
>>> >>>
>>> >>> --
>>> >>> The University of Edinburgh is a charitable body, registered in
>>> >>> Scotland, with registration number SC005336.
>>> >>>
>>> >>>
>>> >>> _______________________________________________
>>> >>> Moses-support mailing list
>>> >>> [email protected]
>>> >>> http://mailman.mit.edu/mailman/listinfo/moses-support
>>> >>
>>> >> _______________________________________________
>>> >> Moses-support mailing list
>>> >> [email protected]
>>> >> http://mailman.mit.edu/mailman/listinfo/moses-support
>>> >
>>> > --
>>> > When a place gets crowded enough to require ID's, social collapse is not
>>> > far away.  It is time to go elsewhere.  The best thing about space travel
>>> > is that it made it possible to go elsewhere.
>>> >                 -- R.A. Heinlein, "Time Enough For Love"
>>> >
>>> > _______________________________________________
>>> > Moses-support mailing list
>>> > [email protected]
>>> > http://mailman.mit.edu/mailman/listinfo/moses-support
>>>
>>> _______________________________________________
>>> Moses-support mailing list
>>> [email protected]
>>> http://mailman.mit.edu/mailman/listinfo/moses-support
>>>
>>
>> --
>> Barry Haddow
>> University of Edinburgh
>> +44 (0) 131 651 3173
>>
>> --
>> The University of Edinburgh is a charitable body, registered in
>> Scotland, with registration number SC005336.
>>
>>
>> _______________________________________________
>> Moses-support mailing list
>> [email protected]
>> http://mailman.mit.edu/mailman/listinfo/moses-support
>>
>
>
>
> --
> When a place gets crowded enough to require ID's, social collapse is not
> far away.  It is time to go elsewhere.  The best thing about space travel
> is that it made it possible to go elsewhere.
>                 -- R.A. Heinlein, "Time Enough For Love"
>
> _______________________________________________
> Moses-support mailing list
> [email protected]
> http://mailman.mit.edu/mailman/listinfo/moses-support
>

_______________________________________________
Moses-support mailing list
[email protected]
http://mailman.mit.edu/mailman/listinfo/moses-support

Reply via email to