NotesGPT

This repository enables you to interact with your notes and journal entries using a language model. Here's how it works:

1. Embedding Documents:

   The embed.py script reads markdown files from a specified folder and embeds their contents into a Chroma database. This process converts your notes and journal entries into a format that the language model can understand and reference.

2. Querying the LLM:

   The prompt.py script allows you to prompt the language model with queries. When you run this script, it starts an interactive session where you can ask questions or request information. The language model uses the embedded documents to provide detailed and contextually relevant responses, effectively allowing you to "chat" with your notes and journal entries.

Prerequisites

• Docker
• Python

Setup

1. Clone the repository:

   git clone noahskorner/notes-gpt
   cd notes-gpt

2. Set up the virtual environment:

   python -m venv venv
   source venv/Scripts/activate  # On Windows
   source venv/bin/activate    # On Unix or MacOS

3. Install dependencies:

   pip install -r requirements.txt

4. Start the Docker containers:

   ./dev.sh

   This script will:

   • Start the Docker containers defined in docker-compose.yml.
   • Execute the ollama container and pull the llama3.2 model.

Usage

1. Embed Documents:

   To embed documents from a specified folder into the Chroma database, run the following command:

   python embed.py PATH_TO_FOLDER

   Replace PATH_TO_FOLDER with the path to the folder containing your markdown files. This script will read all .md files in the folder and embed their contents into the Chroma database.

2. Prompt the LLM:

   To interact with the LLM using your embedded documents, run the following command:

   python prompt.py

   This will start a prompt where you can enter your queries. The LLM will analyze and reference your notes to provide comprehensive and accurate responses.