That's nice. Translation is one of those fields that, besides maybe LibreTranslate, free software was always quite far behind on.
Tangent: Why does free software generally lack machine learning features? Is the expertise just not there or is it rather the cultural stigma of "evil black-boxes"?
Trying to figure out the most expedient way to set it up and test it. Anyone have any ideas?
Is there a place where I can test this online?
I can not help but wonder if it can translate any non previously seen language like protoindoeuropean.
(a) it does not appear to be "open source" for any such definition (b) the paragraph that seems to be actionable has no hyperlinks for ether word "here"
> You can explore a demo of NLLB-200 here, showing how the model can translate stories from around the world, and read the research paper here.