I have several large documents that I needed to pre-translate with TM and MT. To my surprise, about 5% of segments are not pre-translated.
These are my settings:
Anything known about this?
Daniel
What version of Studio and what MT engine?
This problem has been seen before and is possibly related to a timeout. We did resolve this by extending the time quite some time ago, so it would be useful to know which version of Studio and which machine translation engines you're using.
Paul Filkin | RWS
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
This is still an issue with larger files. About 8% of text remained untranslated after running the pre-translate task. Running the task again does not help. THis is very time-consuming - is there a known workaround?
Daniel
Change the minimum match value to 100% and try again.
_________________________________________________________
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.
It is set to 100%.
Can you check the content of the remaining segments? Maybe this way we can find out why they are untranslated.
_________________________________________________________
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.
Seemingly random. "50 x 5 x 120 cm." or "This product has these features." Or: "With pinch-free hinges."
Daniel
The segment with measures can be explained, because it is all variables. But I have no idea why the others remained, sorry.
_________________________________________________________
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.
The segment with measures can be explained, because it is all variables. But I have no idea why the others remained, sorry.
_________________________________________________________
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.
I am convinced it has something to do with segment numbers, not segment content. Segments are being sent to the MT provider and inserted, and at some point this iteration does not work properly. That's my guess at least. If I find the time I might run this over various big files and see whether the gaps coincide with segment numbering, character count etc.
Daniel
I experience a similar behavior with OPUSCAT MT Engine. It skips all segments that have any fuzzy match. As was mentioned above, the minimum match value is set to 100%, so that it neither inserts fuzzy matches, nor the MT result.
Wow, Stepan Konev, thank you for this answer!
The solution is to disable all TMs before starting the the pre-translation task. Studio finds no fuzzy matches and translates all segments with MT. This is a good workaround. tried it with DeepL and running out of time to try with other MT providers, but I would be surprised if they behave differently.
Still, this is a bug. Maybe Paul Filkin could add it to the list?
Daniel
Translate
