Tether AI Smashes Language Model Benchmarks with Compact QVAC TranslatePsy-AfriSLM

Tether AI has made significant advancements in natural language processing (NLP) with the release of QVAC TranslatePsy-AfriSLM, a compact model that outperforms larger systems on translation benchmarks for 19 Sub-Saharan African languages.

The model, which has 800 million parameters, was able to run fully locally on laptops or phones and showed impressive results, beating systems up to 152 times larger in size. This achievement is a testament to Tether AI's efforts to improve language models for low-resource languages.

Tether USDT company CEO Paolo Ardoino highlighted the significance of this development, noting that it has extended their company's AI initiatives and sharpened the comparison between small and large language models.