Paper Summarization with RAG: Single Document Q&A Demo

A demo of paper2summary - a RAG-powered research paper Q&A system.

🔗 GitHub: https://github.com/gabe-zhang/paper2summary
🔗 Fine-tuned model: https://huggingface.co/gabe-zhang/Llama-PaperSummarization-LoRA
🔗 Demo GUI: Kotaemon v0.9.3 - https://github.com/Cinnamon/kotaemon

This video demonstrates single document question-answering with the "Attention Is All You Need" paper (Transformer architecture). The system uses:
- GPT-4o-mini for chat
- text-embedding-3-small for embeddings
- Retrieval-Augmented Generation for accurate citations

Features shown:
- PDF upload and indexing
- Natural language queries about paper content
- Relevance-scored evidence snippets with source citations

#RAG #LLM #MachineLearning #ResearchPapers #AI

Видео Paper Summarization with RAG: Single Document Q&A Demo канала Gabe Z.