pymupdf / RAG

RAG (Retrieval-Augmented Generation) Chatbot Examples Using PyMuPDF
https://pymupdf.readthedocs.io/en/latest/pymupdf4llm
GNU Affero General Public License v3.0
302 stars 57 forks source link

Changes for v0.0.7 #63

Closed JorjMcKie closed 3 months ago

JorjMcKie commented 3 months ago

Fix #41: Do not use text extraction flag "DEHYPHENATE" because this confuses the span-to-line mapping.

Fix #61: Improve recognition of insignificant vector graphics.