Paralled sentences
Hey, someone figured out how to have parallel sentences in the dataset, the author says they have used a ngram method to build multiway data (that I suppose is parallel data https://arxiv.org/pdf/2010.10239.pdf). There is a multiway dir but I don't understand the organization of the file, Is someone know how to manage this?
I'm interested in this too, but don't no how to open this files(((
What are "canary" files?
I mean you should use this one: https://github.com/UKPLab/sentence-transformers/blob/master/examples/applications/parallel-sentence-mining/bitext_mining.py
Read the code and you'll understand what it does
Hello, the paper obviously explain about the parallel bilingual dataset which covers 156 languages. I think that we should easily reach it without any pair mining process or something else? Could you please share any repository to let us download sentence pairs ?