Cohere launches an open-source voice model specifically for transcription

Read Original
Share: X WhatsApp

Relatively light at just 2 billion parameters, the model is meant for use with consumer-grade GPUs for those who want to self-host it. It currently supports 14 languages.

Continue reading on the source website

To respect copyrights, we only provide a brief summary. Read the full article on TechCrunch.

Read Full Article