Possible improvements #121
Description
Following is a list of potential improvements for Lexoid that we can explore:
- VLLM and Ollama support for local models
- OCR using local models
- Structured output with coordinates
- Azure, AWS, Google Vision support
- Layout detection
- Support for TableFormer
- Handle tables separately using Smol VLM
Low priority:
- Docker example
- Hosted endpoint
- Document classification