Modeli

Phi-4-multimodal-instruct

Microsoft

Samo v katalogu Odprte uteži

Nazadnje pregledano:

Suverenost in gostovanje

Samo v katalogu

Ta model je v katalogu za referenco. Na LLM EU še ni gostovan in noben lastni endpoint ga tukaj ne streže. Zahteve zanj gredo neposredno k ponudniku, pod lastnimi pogoji tega ponudnika.

Povzetek

Phi-4-multimodal-instruct je 5,6B model, ki sprejema besedilo, slike in govor v eni kontrolni točki, z 128K-žetonskim kontekstom in licenco MIT. Kartica modela poroča 55,1 na MMMU, 93,2 na DocVQA in 62,4 na MathVista ter navaja podprte jezike ločeno za vsako vhodno modalnost. Je kompaktna možnost za dokumentne in govorne cevovode, ki bi sicer potrebovali dva modela.

Specifikacije

Ponudnik
Microsoft (US)
Kontekstno okno
131k
Parametri
5.57B
Modalnosti
text, vision, audio
Licenca
mit · licence
Odprte uteži
yes
Datum izdaje
24. feb. 2025
Naloge
Vid, Povzemanje, Večjezična podpora, Nizki stroški

Jeziki

Močno: en, de, fr, es, it, pt

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

Močno

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

Zadostno

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

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

  • 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

Omejitve

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

Pogosto zastavljena vprašanja

Is Phi-4-multimodal-instruct served on LLM EU?

It has no endpoint in the catalogue, so it is catalogue only — not hosted on LLM EU yet, with no region and no retention to report. A card whose status is "listed" is a catalogue entry, not a running deployment.

Under what licence is Phi-4-multimodal-instruct released?

Phi-4-multimodal-instruct is recorded under mit. The recorded licence text is https://opensource.org/license/mit. The weights are downloadable (open_weights is true). LLM EU republishes the licence id and the link and gives no legal advice; the licence text governs.

How large is Phi-4-multimodal-instruct, and what context window does it have?

Phi-4-multimodal-instruct is recorded at 5.57B (safetensors total 5,574,460,384) (source: https://huggingface.co/api/models/microsoft/Phi-4-multimodal-instruct), with a context window of 131072 tokens (source: https://huggingface.co/microsoft/Phi-4-multimodal-instruct).