[Usage]: The example of using microsoft/Phi-4-multimodal-instruct audio

[engchina]

Your current environment

How to use microsoft/Phi-4-multimodal-instruct audio by using vllm?

Here,
I can see an example of using vision, but how to use audio? Please help！

How would you like to use vllm

I want to run inference of a [specific model](put link here). I don't know how to integrate it with vllm.

Before submitting a new issue...

• Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.

[DarkLight1337]

See the audio-language example: https://github.com/vllm-project/vllm/blob/main/examples%2Foffline_inference%2Faudio_language.py

[engchina]

@DarkLight1337
Thanks for your reply.

Is there any example by using api similar as below?

#14119

[DarkLight1337]

See https://github.com/vllm-project/vllm/blob/main/examples/offline_inference/audio_language.py

[engchina]

Thanks，I found this link which is exactly I'm looking for.
https://docs.vllm.ai/en/latest/getting_started/examples/openai_chat_completion_client_for_multimodal.html

[DarkLight1337]

Oops, sorry I posted the wrong link. The link you found is the correct one.