Merge pull request #48 from iamarunbrahma/refactor_v13

Refactor whitespace in parser.py for improved readability; streamline image processing code and async calls.

Update .gitignore to include .venv and uv.lock; remove uv.lock file as it is no longer needed.

Add Pillow dependency for image processing; refactor whitespace in various files for consistency and readability.

Refactor LLM class to improve Gemini client integration; streamline template accessors and update test cases for markdown generation.

update uv.lock

Add support for dynamic vision model discovery and refactor constants; introduce ModelDetector class for improved model capability detection.

Update dependencies and documentation for Gemini model; remove DeepSeek references and examples.

Refactor LLM class to streamline Ollama client initialization and enhance concurrency support.

Replace PyMuPDF with pypdfium2 for PDF processing and update related code in vision_parse module.

Merge pull request #32 from iamarunbrahma/add_streamlit_app

Remove unnecessary blank line in Streamlit app

Add Streamlit demo application for Vision Parse PDF extraction

Merge pull request #31 from iamarunbrahma/updated_markdown_prompt

Refine markdown prompt template text extraction instructions