Cohere Transcribe Arabic Challenges Whisper and OmniASR in Transcription
Le brief IA que les pros lisent chaque soir
Les 7 actus IA du jour, décryptées en 5 min. Gratuit.
Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.
Choisis ton rythme
Gratuit · Pas de spam · Désabonnement en 1 clic
Cohere Transcribe Arabic Challenges Whisper and OmniASR in Transcription
Cohere has launched Cohere Transcribe Arabic, an open-source model designed for Arabic speech recognition. This 2 billion parameter ASR model is, according to Cohere, the most accurate open-source audio-to-text transcription system for Arabic available. It addresses the specific challenges of Arabic speech, including the variety of dialects, Arabic-English bilingual conversations, code-switching, and specialized vocabulary. Cohere claims it outperforms Whisper Large V3, the standard model of Cohere Transcribe, as well as other systems in benchmarks.
Human evaluations of Arabic transcriptions on a scale of 1 to 5 show that Cohere Transcribe Arabic surpasses Whisper Large V3 and the standard model of Cohere Transcribe in terms of overall quality, fidelity to dialect, and code-switching.
The model is distributed under the Apache 2.0 license and is available on Hugging Face as well as via the Cohere API. Additional benchmarks and examples are available on the Cohere blog.
Brief IA — L'actualité IA en français
L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.