• Replies 13 replies
  • Subscribers 434 subscribers
  • Views 6781 views
  • Users 0 members are here
Options
Related

Studio very slow to update main TM

Bruce Campbell

4 minutes to update the main TM with 3 files containing 206 segments, 3124 words!

Studio showing no activity in the Resource Monitor during this time, except for one blip in disk activity I saw when the progress bar lengthened.

Note that this only appears to occur when changes actually have to be made to the TM. If I repeat the update a second time, it completes in a few seconds.

Am I the only one with this problem?

Best regards,
Bruce Campbell
ASAP Language Services
Rate translation
×
Suggest better translation
×
Moderator UI
×
  • Get AI Suggestion
Parents
  • Unfortunately not. I also have a very fast computer with 1 TB SSD disk, and when I update one of my bigger TM's (470 MB) I can wait for about an hour for the task to finish.

    This is a very serious problem, of course, and I wonder why no SDL person has answered here yet?
  • in reply to Søren Vedel
    Hi , et al.

    No special reason... probably just didn't notice the question. Can anyone provide me with some files to look at? A TM and the import files causing the issue.

    Regards

    Paul

    Paul Filkin | RWS Group

    ________________________
    Design your own training!
    You've done the courses and still need to go a little further, or still not clear? 
    Tell us what you need in our Community Solutions Hub

  • in reply to Paul

    Hi Paul,

    Here you can see the log file. The TM is not allowed to be attached here. This is a small TM and it took 4 minutes to update.

     

    --------------------------------------------------------
    Translation memory upgrade started (13-08-2017 10:05:34)
    --------------------------------------------------------

    Working folder: C:\1-TR\Translation Memories\PT-DA\Upgrade 2017-08-13 10-05-07

    ---------------------------------------------------------------------------------
    Export C:\1-TR\Translation Memories\PT-DA\pt-PT_da-DK.sdltm (13-08-2017 10:05:34)
    ---------------------------------------------------------------------------------

    Input TM         : sdltm.file:///C:/1-TR/Translation Memories/PT-DA/pt-PT_da-DK.sdltm
    Target TMX file  : C:\1-TR\Translation Memories\PT-DA\Upgrade 2017-08-13 10-05-07\pt-PT_da-DK.tmx
    Translation units: 20708
    Export completed (13-08-2017 10:05:56)

    ---------------------------------------------------------------
    Create translation memory pt-PT_da-DK (2) (13-08-2017 10:05:56)
    ---------------------------------------------------------------

    Type                          : File
    TM location                   : C:\1-TR\Translation Memories\PT-DA\pt-PT_da-DK (2).sdltm
    Copyright                     :
    Description                   : upload memory
    Expiration Date               : 31-12-9999 23:59:59
    Create character-based indexes: no
    Recognize acronyms            : yes
    Recognize dates               : yes
    Recognize measurements        : yes
    Recognize numbers             : yes
    Recognize times               : yes
    Recognize variables           : yes
    Recognize alphanumeric        : no
    Language pairs                :
                                  : pt-PT -> da-DK
    Language resources            :
    pt-PT                         : (Default Language Resources)
    Fields                        :
                                  : Doc. Type (MultiplePicklist: AA,AB,AC,AD,AG,AL,AM,AN,AS,AT,BR,BU,CB,CC,CM,CN,CO,CP,CQ,CR,CT,CV,DA,DC,DE,DI,DL,DM,DO,DP,DR,DT,DU,DV,ER,EX,FE,FO,FP,FR,FT,GL,HH,IA,IG,IK,IN,LP,LT,LV,MA,MD,MN,MP,NH,NP,NR,NT,OD,OJ,OP,OS,PA,PB,PE,PF,PM,PO,PR,PV,QB,QE,QH,QO,QZ,RC,RE,REF,RF,RH,RM,RR,RT,SP,SR,ST,TA,TC,TR,TT,VI,WD,)
                                  : Year (MultiplePicklist: 1985,1986,1987,1988,1989,1990,1991,1992,1993,1994,1995,1996,1997,1998,1999,2000,2001,2002,2003,2004,2005,2006,2007,2008,2009,2010,2011,2012,2013,2014,2015,2016,2017,2018,2019,2020,2021,2107,)
                                  : Req. Serv. (MultiplePicklist: ---,ACP,ADMI-EXPO,ADMI-IPOL,ADMI-PRES,AFCO,AFET,AGRI,AP,BD,BUDG,BUR,CEC,CLIM,CODE,COMM,CONS,CONT,COR,CPCO,CPDE,CRIM,CRIS,CULT,DELE,DEPO-PRES,DEVE,DROI,ECA,ECON,EESC,EMED,EMPA,EMPL,ENVI,EPRS,EQUI,EUAL,EXPO,FDEP,FEMM,FINP,FINS,FONC,GREF-PRES,GRPO,IMCO,IMPA-IPOL,INLO,INTA,INTE,IPOL,ITEC,ITRE,JTCO,JURI,LEGA-PRES,LIBE,MDEP,MEDI,MEPC,MISC,OPPD,ORGA,ORGA-PRES,PANA,PAPL,PB,PECH,PERS,PETI,QUES-PRES,QUEST,REGI,SAFE,SEANCE,SEDE,SG,SJ,SURE,TDIP,THEM-IPOL,TRAD,TRAD-OLP,TRAN,)
                                  : Translator (MultiplePicklist: corr,flt,inh,revised,)
                                  : Doc. No. (MultipleString)
                                  : Obs. (MultipleString)
                                  : TM Database (MultipleString)
    Complete (13-08-2017 10:05:56)

    -----------------------------------------------------------------------------------------------------------------------------
    Import 'C:\1-TR\Translation Memories\PT-DA\pt-PT_da-DK.sdltm' into 'C:\1-TR\Translation Memories\PT-DA' (13-08-2017 10:05:56)
    -----------------------------------------------------------------------------------------------------------------------------

    Import as plain text: False
    Translation units   : 20708
    Import completed (13-08-2017 10:07:01)

    ---------------------------------------------
    Build Translation Model (13-08-2017 10:07:01)
    ---------------------------------------------

    Translation memory  : C:\1-TR\Translation Memories\PT-DA\pt-PT_da-DK (2).sdltm
    Translation units   : 20708
    Build model required: True

    Process completed (13-08-2017 10:07:48)


    ---------------------------------------------
    Align Translation Units (13-08-2017 10:07:48)
    ---------------------------------------------

    Translation memory: C:\1-TR\Translation Memories\PT-DA\pt-PT_da-DK (2).sdltm
    Translation units : 20708
    Alignment required: True

    Process completed (13-08-2017 10:09:40)


    ---------------------------------------------------------
    Translation memory upgrade finished (13-08-2017 10:09:40)
    ---------------------------------------------------------


    Kind regards

    Søren Vedel
  • in reply to Søren Vedel
    Has anyone of you tried to switch off Fuzzy match repair BEFORE updating the TM?
    In my case this is what speeds up that process significantly.

    As for TMX import into a "big mamma" - I wouldn't do that at all. In my case when I prepare projects utilizing my big mamma (1,4 million segments), I create a project TM within the project creation (yes, for the time of project creation "Repair fuzzy matches" is also switched off), then I open such PTM in Studio and import any TMX from the customer there. This is not only faster, but much more secure for me, as I don't know what comes in the TMX. The TM from the customer may be perfect or complete crap (seen all of this in my life), so directly importing it into my TM could spoil that one, which is a quite important asset for me.

    _________________________________________________________

    When asking for help here, please be as accurate as possible. Please always remember to give the exact version of product used and all possible error messages received. The better you describe your problem, the better help you will get.

    Want to learn more about Trados Studio? Visit the Community Hub. Have a good idea to make Trados Studio better? Publish it here.

  • in reply to Jerzy Czopik

    Why is the updated file more than 3 times bigger than my old file?

    Thanks for the tip, Jerzy. Maybe I will try it out next time I don't need my computer for work and have time for doing rather tedious trial-n-error experiments.

    I only updated my TM because of the message: "take benefit from the UpLift blabla...." message, hoping everything went faster and better than previously.
  • in reply to Søren Vedel
    This is simply because of the fragment matching. Extra information is being added to TM. And the UpLift blabla is one of the most interesting new features, providing you new matches composed of fragments from previous matches or matches repaired with your terminology entries. Sometimes, in my GE-PL jobs I get fuzzy matches of 70%+ quality, where I just need to confirm without any changes or the only necessary change is adapt word endings. I would say this is additional leverage of maybe 2% from the TM and more active usage of TB. After having used it for a while, when you are then forced to work with a software, which "nails you on a cross", you notice, how good it is in facht.

    _________________________________________________________

    When asking for help here, please be as accurate as possible. Please always remember to give the exact version of product used and all possible error messages received. The better you describe your problem, the better help you will get.

    Want to learn more about Trados Studio? Visit the Community Hub. Have a good idea to make Trados Studio better? Publish it here.

  • in reply to Jerzy Czopik
    Ok, makes sense. Thanks for clarifying!

    But I have actually switched the fragment matching off, because it apparantly gave bad results in my case. I find Google Translate API with neural switched on terribly good, by the way. Most of the results I have to read twice to find an error.

    But I must try to find a good tutorial to guide me how to use UpLift, and of course, what it actually "is".

    By the way, now we're talking about new features, I saw in the latest MemoQ update a wonderful feature - amongst many other - where I am able to filter my fuzzy results totally the way I wish. E.g. "Show me results between 87% and 92% or whatever range I'd like to filter. Of course combined with the rest of the filter options at the same time.

    Is this possible in Studio? And I guess I should ask this question elsewhere...?
  • in reply to Søren Vedel
    I am not aware to have the possibility to filter the matches that way here.

    Speaking about UpLift: it really depends very strongly on the language direction.
    In my case, GE-PL is delivering really high quality results, while when translating PL-GE the results are more or less not present...
    In a nutshell: the fuzzy match repair feature replaces parts of the current fuzzy match by parts found in other matches (100%, CM, fuzzy, auto concordance) or in termbase.
    So if you have had translated: "Temperature is measured with a thermometer" and have temperature, thermometer, pressure and manometer in your TB, the new sentence "Pressure is measured with a manometer" could be derived from the previous match and TB. In my case this delivers very useful results for GE-PL, which is my main language direction.
    However, all this slows Studio significantly down - so it is simply necessary to experiment, for which jobs/customers it works well and where it is better to switch it off. However, I would never ever resign of this feature. Simply because I'm certain it will be getting better in the next releases.

    _________________________________________________________

    When asking for help here, please be as accurate as possible. Please always remember to give the exact version of product used and all possible error messages received. The better you describe your problem, the better help you will get.

    Want to learn more about Trados Studio? Visit the Community Hub. Have a good idea to make Trados Studio better? Publish it here.

Reply
  • in reply to Søren Vedel
    I am not aware to have the possibility to filter the matches that way here.

    Speaking about UpLift: it really depends very strongly on the language direction.
    In my case, GE-PL is delivering really high quality results, while when translating PL-GE the results are more or less not present...
    In a nutshell: the fuzzy match repair feature replaces parts of the current fuzzy match by parts found in other matches (100%, CM, fuzzy, auto concordance) or in termbase.
    So if you have had translated: "Temperature is measured with a thermometer" and have temperature, thermometer, pressure and manometer in your TB, the new sentence "Pressure is measured with a manometer" could be derived from the previous match and TB. In my case this delivers very useful results for GE-PL, which is my main language direction.
    However, all this slows Studio significantly down - so it is simply necessary to experiment, for which jobs/customers it works well and where it is better to switch it off. However, I would never ever resign of this feature. Simply because I'm certain it will be getting better in the next releases.

    _________________________________________________________

    When asking for help here, please be as accurate as possible. Please always remember to give the exact version of product used and all possible error messages received. The better you describe your problem, the better help you will get.

    Want to learn more about Trados Studio? Visit the Community Hub. Have a good idea to make Trados Studio better? Publish it here.

Children
No Data