Phi-4-multimodal-instruct

Microsoft

Samo katalog Otvoreni utezi

Suverenost i hosting

Samo katalog

Ovaj je model u katalogu za referencu. Još nije hostan na LLM EU i nijedna vlastita krajnja točka ga ovdje ne poslužuje. Zahtjevi za njega idu izravno pružatelju, prema uvjetima tog pružatelja.

Sažetak

Phi-4-multimodal-instruct is a 5.6B model that accepts text, images and speech in one checkpoint, with a 128K-token context and the MIT licence. The model card reports 55.1 on MMMU, 93.2 on DocVQA and 62.4 on MathVista, and lists the supported languages separately for each input modality. It is a compact option for document and speech pipelines that would otherwise need two models.

Specifikacije

Pružatelj
Microsoft (US)
Kontekstni prozor
131k
Parametri
5.57B
Modaliteti
text, vision, audio
Licenca
mit · licence
Otvoreni utezi
yes
Datum izlaska
24. velj 2025.
Zadaci
vision, summarization, multilingual-support, cheap

Jezici

Jako: en, de, fr, es, it, pt

Zadovoljavajuće: nl, pl, sv, da, fi, cs, ru, uk, ar, zh, ja, ko, hi, tr

Jezici

Jako

  • en
  • de
  • fr
  • es
  • it
  • pt

Zadovoljavajuće

  • nl
  • pl
  • sv
  • da
  • fi
  • cs
  • ru
  • uk
  • ar
  • zh
  • ja
  • ko
  • hi
  • tr

Nije provjereno: pružatelj nije dokumentirao ovu tvrdnju. — a language is listed as strong only where a source claims it; the card links the evidence where one exists.

Kada ga koristiti

Prednosti

  • text, image and speech input at 5.6B parameters
  • 128K-token context for document batches
  • MIT licence
  • 23 declared languages including most large EU languages

Ograničenja

  • modality-specific quality is below dedicated single-modality models
  • English-centric training data
  • no evaluation has been run by LLM EU