Tether, the issuer of the USDT stablecoin, released an open-source neural machine translation model family called TranslatePsy-AfriSLM, covering 19 Sub-Saharan African languages. These models, developed by Tether’s QVAC division, weigh only 21 to 35 MB per language pair and can run offline directly on a phone without server connectivity. The associated dataset, TranslatePsy-AfriSLM-Synthetic-Mix, contains approximately 215 million bidirectional training examples paired with English and is available on Hugging Face under a CC BY-NC 4.0 license. The models are compatible with Bergamot, the open-source translation engine used in Firefox, and process a sentence in approximately 46 milliseconds, up to 78 times faster than larger models like Salamandra-2B. This initiative aims to reinforce Tether’s presence in regions where its stablecoin is widely adopted, particularly in Africa, Southeast Asia, and Latin America.
Source: Read the original article
