Cohere has released the open-source speech transcription model Transcribe.


It is a 2 billion parameter Conformer ASR supporting 14 languages, capable of local deployment, and also accessible via API and Model Vault, with the download link directly on Hugging Face.
People managing inference environments, doing meeting transcriptions, customer service voice, and real-time speech agents should check it out. The average WER on the HF Open ASR Leaderboard is 5.42%, which is quite impressive.
Project page:
View Original
post-image
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pin