search

haotian-liu / LLaVA

[NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.
https://llava.hliu.cc
Apache License 2.0
20.15k stars 2.22k forks source link

use model as an API [Feature request] #656

Open stablegpt5279 opened 1 year ago

stablegpt5279 commented 1 year ago

feature

how to use the model as an API for production applications??

xts-bit commented 1 year ago

@haotian-liu Same any solutions?

hszhoushen commented 1 year ago

feature

how to use the model as an API for production applications??

I have a solution, add the code with flask (a web application package), so that the LLaVa model can be a server that can receive the post from any client, each time the model receives the post, it will process the image, and send back the language description to the client.

hszhoushen commented 1 year ago

@haotian-liu Same any solutions?

I have a solution, add the code with flask (a web application package), so that the LLaVa model can be a server that can receive the post from any client, each time the model receives the post, it will process the image, and send back the language description to the client.