I've been trying to train "Microsoft custom translator" on my dictionary several times, but all are failed. When I'm choosing only "Phrase dictionary" - model creation finishes successfully. But when I'm trying to create a training model - every time I've got the error "DATA PROCESSING FAILED"
My dictionary data is a set of 40.000 russian words and their english translation. Words are already mapped, so I'm using .align files.
What can be the reason of this error?
One of the common reasons for failed training with training document type is the number of parallel sentences would be less than the minimum required 10,000. For more information about the different document types and min/max, plz refer to