VectorMD enables users to convert markdown documents into a semantic searchable format. By embedding markdown headings into vector space, it can quickly find relevant sections in markdown documents based on natural language queries.
- Semantic Search: Converts markdown documents into a semantically searchable database.
- Efficient Indexing: Employs FAISS for lightning-fast vector searches.
- Easy Integration: Works seamlessly in both CLI environments and directly within Python.
Install VectorMD via pip:
pip install VectorMD-
Initialization:
Convert your markdown document into a semantically searchable format using the following command:
vmd-init --file path_to_code_snippets.md
-
Querying:
After initialization, search for relevant sections in your document using:
vmd docker compose quantized llama2
Use VectorMD directly within your Python scripts:
from vectormd import VectorMD
# Initialize with your markdown file
medicalDB = VectorMD("path_to_medical_markdown.md")
# Query
results = medicalDB.query("HACP empiric tx regimen duration")We welcome contributions! If you'd like to help improve VectorMD, please fork the repository and submit a pull request.
