mradermacher

/

NuMarkdown-8B-Thinking-GGUF

like 8

from llama_cpp import Llama

llm = Llama.from_pretrained(
    repo_id="mradermacher/NuMarkdown-8B-Thinking-GGUF",
    filename="NuMarkdown-8B-Thinking.Q4_K_M.gguf",
)

messages = [{
"role": "user",
    "content": [
        {"type": "image", "image": "https://blog.lipsumhub.com/wp-content/uploads/2024/07/lorem-ipsum-meaning-in-english-lipsumhub.jpg"},
        {"type" : "text", "text": "Describe the image"}
    ],
}]

response = llm.create_chat_completion(
    messages = messages
)
response 

Hi, I am initiating the model using llama_cpp, however, there is no reference on how to input the image to transform into markdown.
I am using the following documentation: https://huggingface.co/docs/transformers/en/tasks/image_text_to_text
The following code is not working (it says the image is BLANK). I was wondering if you could reference the correct code.

Thank you