Helsinki-NLP / OPUS-CAT

OPUS-CAT is a collection of software which make it possible to OPUS-MT neural machine translation models in professional translation. OPUS-CAT includes a local offline MT engine and a collection of CAT tool plugins.
MIT License
69 stars 11 forks source link

Slow fine-tuning / instant MT adaptation #59

Open Khalid-kamal opened 1 year ago

Khalid-kamal commented 1 year ago

Is there a way to accelerate fine-tuning with large data? I think the fastest way is to fine-tune instantly after each segment either in trados or in memoq after confirming a sentence

TommiNieminen commented 1 year ago

Sorry for the late reply, I missed this. It would be nice to have instant fine-tuning after confirming new translations, but it's not possible to implement it with the current MT engine setup. The MT engine will be eventually changed, though, so I'll keep this in mind.

martinengelke commented 1 month ago

At this point I would suggest a Tool for creating a new rule in an efficient and fast way during translation jobs:

Often it comes to misleading translations I would like to clarify and save that translation-pair as a 'rule'. For me it seems unpractical to use the UI for that, so that's why I am trying to find out how to do it at least via command line for now. There should be a file with the stored rules somewhere ...