Skip to content

Comments

Add Docling Support, Pydantic JSON Validation, and Configurable LLM API URL#101

Open
mohamed-em2m wants to merge 4 commits intoVectifyAI:mainfrom
mohamed-em2m:main
Open

Add Docling Support, Pydantic JSON Validation, and Configurable LLM API URL#101
mohamed-em2m wants to merge 4 commits intoVectifyAI:mainfrom
mohamed-em2m:main

Conversation

@mohamed-em2m
Copy link

@mohamed-em2m mohamed-em2m commented Feb 5, 2026

This pull request introduces several enhancements to improve flexibility, structure, and extensibility:

✅ Added Docling support for improved document processing and parsing.

✅ Integrated Pydantic models to ensure structured and validated JSON outputs.

✅ Added support for custom LLM API base URLs, allowing easier integration with self-hosted or alternative LLM providers.

♻️ Refactored parts of the codebase to improve clarity and maintainability.

@banghasan
Copy link

👍

@mohamed-em2m mohamed-em2m changed the title adding configure to support set base url of llms api Add Docling Support, Pydantic JSON Validation, and Configurable LLM API URL Feb 11, 2026
@adityasasidhar
Copy link

@mohamed-em2m

Hey, I saw your implementation and it looks fantastic !!

Thomasyyj added a commit to Thomasyyj/PageIndex that referenced this pull request Feb 18, 2026
…tructured output

Integrate new features from PR VectifyAI#101 into LiteLLM architecture:
- Add PDFReader class supporting docling, PyMuPDF, and PyPDF2 parsers
- Add response_format (Pydantic BaseModel) support to all LLM API functions
- Add read_pdf() utility function with page/full output formats
- Update extract_text_from_pdf, get_text_of_pages, get_page_tokens to use PDFReader
- Add docling and pydantic to requirements.txt
- Add PDF parser config options to config.yaml

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants