Whisper large-v3
Suverenost in gostovanje
V tej uvedbi ni gostovano na LLM EU: na zahteve odgovarja vgrajeno mock zaledje v območju dev-mock. Nobene uteži se ne izvajajo.
Povzetek
Whisper large-v3 is OpenAI’s multilingual speech-recognition and speech-translation model with 1.54B parameters under the Apache-2.0 licence. It has a fixed 30-second receptive field and processes longer audio with a sliding window, and it uses 128 mel frequency bins rather than the 80 of earlier versions. The model card warns that it was evaluated on roughly ten languages and should be evaluated in-domain before deployment.
Specifikacije
- Ponudnik
- OpenAI (US)
- Kontekstno okno
- Ni preverjeno: ponudnik te trditve ni dokumentiral.
- Parametri
- 1.54B
- Modalnosti
- audio
- Licenca
- apache-2.0 · licence
- Odprte uteži
- yes
- Datum izdaje
- 07. nov. 2023
- Naloge
- translation, multilingual-support, summarization
Jeziki
Močno: en, de, fr, es, it, pt, nl, pl
Zadostno: sv, da, fi, cs, ro, el, uk, ru, tr, ar, zh, ja, ko, hi
Jeziki
Močno
- en
- de
- fr
- es
- it
- pt
- nl
- pl
Zadostno
- sv
- da
- fi
- cs
- ro
- el
- uk
- ru
- tr
- ar
- zh
- ja
- ko
- hi
Ni preverjeno: ponudnik te trditve ni dokumentiral. — a language is listed as strong only where a source claims it; the card links the evidence where one exists.
Kdaj ga uporabiti
Prednosti
- one model for transcription and translation to English
- long-standing tooling support across every major inference stack
- Apache-2.0 licence
Omejitve
- fixed 30-second receptive field; longer audio needs chunking logic
- the card only claims strong recognition for about ten languages
- the card advises against using it for classification or speaker attributes