Using smollm2 to interact with your Bear notes.

Search through and ask questions about your Bear notes using local AI models served by Ollama.

Run the following:

make init # pulls model down, starts ollama, syncs your Bear notes to a folder, and then indexes them as embeddings into ChromaDB sqlite db
make search # search through your notes using semantic search
make ask # generate AI responses to your prompts

• Semantic Search: Use make search to find relevant notes based on meaning, not just keywords
• AI Generation: Use make ask to generate AI responses to your prompts
• Bear Integration: Automatically syncs with your Bear notes
• Local Processing: All processing happens on your machine using Ollama

• Bear notes installed on your Mac
• Ollama installed on your Mac
• Python installed on your Mac
  • Python 3.11.1 is recommended

• smollm2 - a small language model for AI responses
• nomic-embed-text - a small text embedding model for semantic search
• ChromaDB - an open-source vector database to store and query embeddings
• Ollama - an open-source language model runner

Install this tool using pip:

pip install notes-analyzer

For help, run:

notes-analyzer --help

Basic commands:

make sync   # Sync Bear notes to local directory
make index  # Create searchable index of your notes
make search # Search through your notes
make ask    # Generate AI responses to prompts

You can also use the Python module directly:

python -m notes_analyzer --help

To contribute to this tool, first checkout the code. Then create a new virtual environment:

cd notes-analyzer
python -m venv venv
source venv/bin/activate

Now install the dependencies and test dependencies:

pip install -e '.[test]'

To run the tests:

python -m pytest