Embedding images to Phi-3.5-3.8B-vision-instruct using api server

[kalcao]

Hello!
I'm new to this and having trouble of embedding images to the model.

I'm using python3 -m llama_cpp.server --config config.json to run the server
and it's the config

{
    "host": "0.0.0.0",
    "port": 8080,
    "models": [
        {
            "model": "Phi-3.5-3.8B-vision-instruct-Q8_0.gguf",
            "model_alias": "vision",
            "chat_format": "llama-2",
            "n_gpu_layers": -1,
            "offload_kqv": true,
            "n_threads": 12,
            "n_batch": 512,
            "n_ctx": 2048
        }
    ]
}

this is the json of I'm sending to /v1/chat/completions

{
        "model": "vision",
        "messages": [
        {
            "role": "user",
            "content": [
                {
                    "type": "image_url",
                    "image_url": "https://www.slidecow.com/wp-content/uploads/2018/04/Setting-Up-The-Slide-Text-1000x563.jpg"
                },
                {"type": "text", "text": "What does the image say? Explain what you see"},
            ],
        }
    ],
    }

And it response doesn't make sense at all. What am I doing wrong?

Thanks in advance!