Try Docling

Import

Code

from docling.document_converter import DocumentConverter
from IPython.display import Markdown, display
from warnings import filterwarnings
from os import environ

Disable Warnings

Code

environ["TF_CPP_MIN_LOG_LEVEL"] = "3"
environ["CUDA_VISIBLE_DEVICES"] = ""
filterwarnings(
    "ignore", message="'pin_memory' argument is set as true but no accelerator is found"
)

Document Converter

Code

converter = DocumentConverter()
result = converter.convert("https://arxiv.org/pdf/2408.09869")

Results

Code

display(Markdown(result.document.export_to_markdown()))