Ahoj,

you don't need to include extra headers for SALM, you just need to use
g++ 4.1 or older to build it.

Parallelization would be nice, it's really really slow now.

If you merge it it will be better than me pushing my changes, since I
believe you know more what's going on (my version will almost surely
break with parallel corpora, factors etc)

Karel

On Wed, Aug 22, 2012 at 9:29 AM, Aleš Tamchyna <[email protected]> wrote:
> Hi,
>
> sigtest filter was added to EMS in branch damt_phrase, you can check it
> out. There's also a script for its parallelization.
>
> I didn't merge this into master because the sigtest filter still depends
> on SALM (which in turn doesn't compile on Linux unless extra header
> files are included in the source code) -- but maybe it's worth merging
> anyway?
>
> Best,
> Ales
>
> Dne 22.8.2012 01:14, Karel Bílek napsal(a):
>> OK.
>>
>> I will add support of sigtest-filter (
>> http://www.statmt.org/moses/?n=Moses.AdvancedFeatures#ntoc19 ) which
>> ships with moses to my EMS experiment.meta ; I will add this to moses
>> on git as another pull request as I properly test it
>>
>> It will use the first case I mentioned because I don't use factors and
>> I don't understand them as well :)
>>
>> BTW
>> there is a slight error in here
>> http://www.statmt.org/moses/?n=Moses.AdvancedFeatures#ntoc19
>> Build Instructions
>>
>> you can't just build it with "make", since it will work only with g++
>> 4.1 and not newer, you have to edit the makefile
>> also, it's not in "directory sigtest-filter", but in 
>> "contrib/sigtest-filter".
>>
>> Karel
>>
>> On Tue, Aug 21, 2012 at 10:31 PM, Philipp Koehn<[email protected]>  wrote:
>>> Hi,
>>>
>>> "ignore" implies that not doing this steps also means that all
>>> preceding steps do not need to be carried out (unless required
>>> by other steps), while "pass" implies only that this step requires
>>> no processing and hence the output of the preceding step
>>> can be used.
>>>
>>> -phi
>>>
>>> On Tue, Aug 21, 2012 at 1:23 PM, Karel Bílek<[email protected]>  wrote:
>>>> By the way, one more question. (I am new to EMS.)
>>>>
>>>> What is the difference between "ignore-if" and "pass-unless"?
>>>>
>>>> K
>>>>
>>>> On Tue, Aug 21, 2012 at 10:16 PM, Karel Bílek<[email protected]>  wrote:
>>>>> Thanks for tips!
>>>>>
>>>>> Yes, I do want to use the same pruned tables for tuning and testing.
>>>>> Currently, I do have only one translation table.
>>>>>
>>>>> KB
>>>>>
>>>>> On Tue, Aug 21, 2012 at 10:13 PM, Philipp Koehn<[email protected]>  
>>>>> wrote:
>>>>>> Hi,
>>>>>>
>>>>>> there are several places where you could do this, depending
>>>>>> on what your goal is.
>>>>>>
>>>>>> If you want use the same pruned phrase table for tuning
>>>>>> and testing, it makes sense to add such a step into the
>>>>>> model training part.
>>>>>>
>>>>>> Currently, there is a phrase translation table step:
>>>>>>
>>>>>> build-ttable
>>>>>>    in: extracted-phrases lexical-translation-table domains
>>>>>>    out: phrase-translation-table
>>>>>>
>>>>>> which feeds into the configuration file creation step:
>>>>>>
>>>>>> create-config
>>>>>>    in: reordering-table phrase-translation-table generation-table
>>>>>> sparse-lexical domains INTERPOLATED-LM:binlm LM:binlm
>>>>>>    out: config
>>>>>>
>>>>>> So, you could put something inbetween, but changing the
>>>>>> input of the create-config step into:
>>>>>>
>>>>>> create-config
>>>>>>    in: reordering-table pruned-phrase-translation-table
>>>>>> generation-table sparse-lexical domains INTERPOLATED-LM:binlm LM:binlm
>>>>>>    out: config
>>>>>>
>>>>>> and creating a step
>>>>>>
>>>>>> prune-phrase-table
>>>>>>    in: phrase-translation-table
>>>>>>    out:  pruned-phrase-translation-table
>>>>>>
>>>>>> However, this may get a bit messy when you have multiple phrase 
>>>>>> translation
>>>>>> tables (due to factored models, for instance), and lexicalized 
>>>>>> reordering models
>>>>>> which have to have the same phrase entries as the phrase translation 
>>>>>> table.
>>>>>> In that case, it would better to place the pruning step after the
>>>>>> "create-config"
>>>>>>
>>>>>> prune-tables
>>>>>>    in: config
>>>>>>    out: pruned-config
>>>>>>
>>>>>> and then have the subsequent steps take pruned-config as input.
>>>>>>
>>>>>> -phi
>>>>>>
>>>>>>
>>>>>> On Tue, Aug 21, 2012 at 12:55 PM, Karel Bílek<[email protected]>  
>>>>>> wrote:
>>>>>>> Hello everyone,
>>>>>>>
>>>>>>> I have a question about EMS.
>>>>>>>
>>>>>>> I want to prune the phrase table right before TUNING step. I already
>>>>>>> do have a script for that; but I don't know where to "plug" it into
>>>>>>> config EMS file.
>>>>>>>
>>>>>>> I can do the pruning "by hand" and then write the address directly
>>>>>>> into "phrase-translation-table" but that seems to be against the
>>>>>>> spirit of EMS.
>>>>>>>
>>>>>>> Thanks,
>>>>>>> Karel Bílek
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> 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
> _______________________________________________
> 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