Video-LLaVA with transformers library

[Kamakshi8104]

Model description

Video-LLaVA is a multimodal model that is trained on both images and videos simultaneously. I feel that it would be highly beneficial if it is added to the transformers library as it is a pretty good choice when it comes to video-question answering

Open source status

• The model implementation is available
• The model weights are available

Provide useful links for the implementation

https://huggingface.co/LanguageBind/Video-LLaVA-7B

[zucchini-nlp]

If the authors will not be willing to contribute, I want to work on this

cc @gante

[amyeroberts]

Great @zucchini-nlp! I'd reach out on the model page on the hub, asking the authors if they'd like to add it to the library. If they're happy to just leave as-is it's all yours!