SPRINGLab/shiksha-MT-nllb-3.3B
Translation • Updated • 3 • 1
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
A curated subset of Bharat Parallel Corpus Collection (BPCC) for 8 Indian languages. Translation pairs are filtered with LABSE score(>0.9) and further preprocessed. Useful for training high-quality translation models.