Facebook's MMS-300m Multilingual Speech Model Trends on Hugging Face

facebook · hf · 2026-09-02

Facebook's (Meta) MMS-300m model is trending on Hugging Face. Built with Transformers and PyTorch using the wav2vec2 architecture, this pre-trained model supports a wide array of language tags (ab, af, ak, am, ar, etc.), focusing on advancing multilingual speech processing capabilities.

Original post →

More from Models

Models channel →