How can I integrate the InternVL-Chat-V1.5 model into a web page without specialized hardware or API?

[JoshuaStudios]

Body

I am developing an entrepreneurship project for my school and I need help to integrate the InternVL-Chat-V1.5 visual language model into our website. I do not have access to an NVIDIA CUDA card for intensive processing nor is there an API available for this model. I am using Windows 11 and VS Code for development. How can I add this model to my website without using advanced hardware or software? I'm looking for a simple way to make the model work so that it can interact with the users of the page. Any advice on how to get started, useful tools or examples would be greatly appreciated.

Open source multimodal AI:
https://huggingface.co/OpenGVLab/InternVL-Chat-V1-5

This is the website where I want to implement open source AI
51673.zip

Guidelines

• I have read and understood this category's guidelines before making this post.

[ItsRaelx]

I don't think you can add this model to your website without specialized graphic card, but you can push this model to something like https://replicate.com/ where they can host it for you and give you api access.

They have some guides on how can you do that for popular programming languages etc.

[JoshuaStudios]

Thank you for your advice. However, since I don't have an NVIDIA graphics card, could you explain to me how to fix the problem with the API? If that's more comfortable for you, we can go over this over Discord. Thanks for your help.

PS: My native language is Spanish.

[NanduWasTaken]

You maybe can run this in an vm(azure, digital cloud) and make an api for it

[CopperEagle]

Hey @JoshuaStudios

There is a community hosted API for this model. The link is also on the official repository of the model you screenshotted above. Here is the link. The basic plan is free but you need to pay if you need many requests per month. Huggingface itself does not offer an inference API becuase the model runs custom code for setup, which is common for Multimodal LLMs on Huggingface AFAIK (same with moondream2).

Hope this helps.

---

Traducción realizada con la versión gratuita del traductor DeepL

Hay una API alojada en la comunidad para este modelo. El enlace también está en el repositorio oficial del modelo que capturaste arriba. Aquí está el enlace. El plan básico es gratuito pero tienes que pagar si necesitas muchas peticiones al mes. Huggingface en sí no ofrece una API de inferencia porque el modelo se ejecuta código personalizado para la configuración, que es común para LLM Multimodal en Huggingface AFAIK (lo mismo con moondream2).

Espero que esto ayude