technovangelist.videoprojects/2024-04-04-build-rag-with-python
Matt Williams 8fef00aa94 Update readme
Signed-off-by: Matt Williams <m@technovangelist.com>
2024-04-30 22:40:37 -07:00
..
.gitignore posted the video for 4-4 2024-04-04 20:58:31 -07:00
config.ini posted the video for 4-4 2024-04-04 20:58:31 -07:00
import.py fix: you can only delete something that might exist 2024-04-12 09:55:30 -07:00
readme.md Update readme 2024-04-30 22:40:37 -07:00
requirements.txt ugh, hate the python way of doing things 2024-04-05 07:03:55 -07:00
search.py posted the video for 4-4 2024-04-04 20:58:31 -07:00
sourcedocs.txt posted the video for 4-4 2024-04-04 20:58:31 -07:00
utilities.py posted the video for 4-4 2024-04-04 20:58:31 -07:00

Build RAG with Python

Watch the video

  1. Install the magic tools:
  • Debian/Ubuntu: sudo apt-get install libmagic1`
  • Windows: You'll need DLLs for libmagic. @julian-r maintains a pypi package with the DLLs, you can fetch it with: pip install python-magic-bin
  • Mac: brew install libmagic
  1. Get started by installing the requirements: pip install -r requirements.txt
  2. Make sure you have the models listed in config.ini. so for nomic-embed-text, run ollama pull nomic-embed-text. Update the config to show whatever models you want to use.
  3. Then run ChromaDB in a separate terminal: chroma run --host localhost --port 8000 --path ../vectordb-stores/chromadb
  4. Edit the list of docs in sourcedocs.txt
  5. Import the docs: python3 import.py
  6. Perform a search: python3 search.py <yoursearch>