Hello Sakura Y ,

Please show me the settings from Project settings as in here:

I tested from my side and I have as first suggestion the ones from the TM:

These are my settings in Automation:

Hope this can help you.

Hi Sakura Y 

I believe the setting you are looking for is here:

Hi, thank you for both of your replies.

As you mentioned, the Project Settings appear to be useful.
However, the option does not seem to work as expected.

In the Project Settings, I tried both selecting and clearing the checkbox “Look up segments in MT even if a TM match has been found.”
But the results were the same—machine translation worked regardless of the checkbox setting, as shown below.

When a fazzy match is found

When a 100% match found

Is this the expected behavior, or could it be a bug?
Could you let me know how this option is supposed to work?

Thanks,

Hello Sakura,

we solved this problem on our side by changing the project preparation setting in step 7 Batch-Tasks >>All Language Pairs >> Batch Processing >> Pre-translate Files by changing the standard Translation overwrite mode from Overwrite existing translation if better match found to new: Keep existing translation. Our usual setting is an adapted minimum value of 80 % for the pretranslation (see screenshot below). Before this change we had the NMT matches in target segments after pretranslation although there were high matches in the TM and this problem was solved on our side by the changed setting during project creation. Perhaps this might be an idea for you, too. Have a nice weekend - Best - Renate Dockhorn