The community members are discussing an issue with loading data from a PDF file. The original post shows an error message indicating that the 'str' object has no attribute 'name'. The comments suggest that the community members need to provide a path object instead of a string to load the PDF data correctly. They also discuss potential issues with the PDF file, such as it containing images instead of text, and suggest using OCR tools like OCRmyPDF or HelloRAG to extract the text. The community members also discuss using a service called llamaparse, which offers free usage for up to 1000 pages per day, as a potential solution.
actually if you look at the code provided in the link above, it extracts metadata from the path object of that file. Which is not possible from a string
