Skip to content

Instantly share code, notes, and snippets.

View maglore9900's full-sized avatar

maglore9900

18 followers · 16 following
View GitHub Profile
@maglore9900
maglore9900 / pros_cons.md
Created December 21, 2024 16:11
Metadata Strategy Pros Cons
Single-File Metadata Easy to set up Less context for queries
Per-Document Metadata Precise filtering possible Requires preprocessing
@maglore9900
maglore9900 / lessons_learned.md
Created December 21, 2024 16:05
Issue Solution Note
Large memory usage Incremental writes Reduced RAM footprint
Metadata ambiguity Use structured metadata keys Improves query relevance
Lack of GPU support (Win) Use Anaconda or Linux Significantly faster batch processing
@maglore9900
maglore9900 / README.md
Created April 4, 2024 14:44 — forked from disler/README.md
Use these Prompt Chains to build HIGH QUALITY AI Agents (Agentic Building Blocks)

Setup

  1. Create a new directory with these three files (requirements.txt, main.py, README.md)
  2. python -m venv venv
  3. source venv/bin/activate
  4. pip install -r requirements.txt
  5. python main.py
  6. Update main() to run the example prompt chains