I\u2019m excited to share Content Extractor with Vision LLM<\/strong>, an open-source Python tool that extracts content from documents (PDF, DOCX, PPTX), describes embedded images using Vision Language Models<\/strong>I\u2019m excited to share Content Extractor with Vision LLM<\/strong>, an open-source Python tool that extracts content from documents (PDF, DOCX, PPTX), describes embedded images using Vision Language Models<\/strong>, and saves the results in clean Markdown files.<\/p>\n This is an evolving project, and I\u2019d love your feedback, suggestions, and contributions to make it even better!<\/p>\n Multi-format support<\/strong>: Extract text and images from PDF, DOCX, PPTX<\/strong>. Advanced image description<\/strong>: Choose from local models<\/strong> (Ollama’s llama3.2-vision) or cloud models<\/strong> (OpenAI GPT-4 Vision). Two PDF processing modes<\/strong>: Text + Images<\/strong>: Extract text and embedded images. Page as Image<\/strong>: Preserve complex layouts with high-resolution page images. Markdown outputs<\/strong>: Text and image descriptions are neatly formatted. CLI interface<\/strong>: Simple command-line interface for specifying input\/output folders and file types. Modular & extensible<\/strong>: Built with SOLID principles for easy customization. Detailed logging<\/strong>: Logs all operations with timestamps. <\/p>\n Programming<\/strong>: Python 3.12 Document processing<\/strong>: PyMuPDF, python-docx, python-pptx Vision Language Models<\/strong>: Ollama llama3.2-vision, OpenAI GPT-4 Vision <\/p>\n Clone the repo and install dependencies using Poetry<\/strong>. System dependencies like LibreOffice<\/strong> and poppler<\/strong> are required for processing specific file types.<\/p>\n Detailed setup instructions: GitHub Repo<\/a><\/p>\n Clone the repo and install dependencies. Start the Ollama server: ollama serve. Pull the llama3.2-vision model: ollama pull llama3.2-vision. Run the tool:bashCopy codepoetry run python main.py<\/a> –source \/path\/to\/source –output \/path\/to\/output –type pdf Review results in clean Markdown format, including extracted text and image descriptions. <\/p>\n This is a work in progress, and I\u2019d love your input to:<\/p>\n Improve features and functionality Test with different use cases Compare image descriptions from models Suggest new ideas or report bugs <\/p>\n GitHub: Content Extractor with Vision LLM<\/a><\/p>\n Feel free to open issues, create pull requests, or fork the repo for your own projects.<\/p>\n This tool has a lot of potential, and with your help, it can become a robust library for document content extraction and image analysis. Let me know your thoughts, ideas, or any issues you encounter!<\/p>\n Looking forward to your feedback, contributions, and testing results.<\/p>\n , and saves the results in clean Markdown files.<\/p>\n This is an evolving project, and I\u2019d love your feedback, suggestions, and contributions to make it even better!<\/p>\n Multi-format support<\/strong>: Extract text and images from PDF, DOCX, PPTX<\/strong>. Advanced image description<\/strong>: Choose from local models<\/strong> (Ollama’s llama3.2-vision) or cloud models<\/strong> (OpenAI GPT-4 Vision). Two PDF processing modes<\/strong>: Text + Images<\/strong>: Extract text and embedded images. Page as Image<\/strong>: Preserve complex layouts with high-resolution page images. Markdown outputs<\/strong>: Text and image descriptions are neatly formatted. CLI interface<\/strong>: Simple command-line interface for specifying input\/output folders and file types. Modular & extensible<\/strong>: Built with SOLID principles for easy customization. Detailed logging<\/strong>: Logs all operations with timestamps. <\/p>\n Programming<\/strong>: Python 3.12 Document processing<\/strong>: PyMuPDF, python-docx, python-pptx Vision Language Models<\/strong>: Ollama llama3.2-vision, OpenAI GPT-4 Vision <\/p>\n Clone the repo and install dependencies using Poetry<\/strong>. System dependencies like LibreOffice<\/strong> and poppler<\/strong> are required for processing specific file types.<\/p>\n Detailed setup instructions: GitHub Repo<\/a><\/p>\n Clone the repo and install dependencies. Start the Ollama server: ollama serve. Pull the llama3.2-vision model: ollama pull llama3.2-vision. Run the tool:bashCopy codepoetry run python main.py<\/a> –source \/path\/to\/source –output \/path\/to\/output –type pdf Review results in clean Markdown format, including extracted text and image descriptions. <\/p>\n This is a work in progress, and I\u2019d love your input to:<\/p>\n Improve features and functionality Test with different use cases Compare image descriptions from models Suggest new ideas or report bugs <\/p>\n GitHub: Content Extractor with Vision LLM<\/a><\/p>\n Feel free to open issues, create pull requests, or fork the repo for your own projects.<\/p>\n This tool has a lot of potential, and with your help, it can become a robust library for document content extraction and image analysis. Let me know your thoughts, ideas, or any issues you encounter!<\/p>\n Looking forward to your feedback, contributions, and testing results.<\/p>\n<\/div>\n submitted by \/u\/Electrical-Two9833 <\/a> \u2728 Key Features<\/h1>\n
\ud83d\udee0\ufe0f Tech Stack<\/h1>\n
\ud83d\udce6 Installation<\/h1>\n
\ud83d\ude80 How to Use<\/h1>\n
\ud83d\udca1 Why Share?<\/h1>\n
\ud83d\udcc2 Repo & Contribution<\/h1>\n
\ud83e\udd1d Let\u2019s Collaborate!<\/h1>\n
\u2728 Key Features<\/h1>\n
\ud83d\udee0\ufe0f Tech Stack<\/h1>\n
\ud83d\udce6 Installation<\/h1>\n
\ud83d\ude80 How to Use<\/h1>\n
\ud83d\udca1 Why Share?<\/h1>\n
\ud83d\udcc2 Repo & Contribution<\/h1>\n
\ud83e\udd1d Let\u2019s Collaborate!<\/h1>\n
[link]<\/a><\/span> [comments]<\/a><\/span><\/p>
