https://github.com/langurmonkey/gaiasky-ai-assistant
AI assistant for Gaia Sky using local Ollama models and RAG with the Gaia Sky documentation site and the home page
https://github.com/langurmonkey/gaiasky-ai-assistant
Last synced: about 1 year ago
JSON representation
AI assistant for Gaia Sky using local Ollama models and RAG with the Gaia Sky documentation site and the home page
- Host: GitHub
- URL: https://github.com/langurmonkey/gaiasky-ai-assistant
- Owner: langurmonkey
- License: gpl-3.0
- Created: 2025-03-27T07:39:09.000Z (over 1 year ago)
- Default Branch: master
- Last Pushed: 2025-03-27T07:42:40.000Z (over 1 year ago)
- Last Synced: 2025-04-23T04:59:53.737Z (about 1 year ago)
- Language: Python
- Size: 115 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Gaia Sky AI assistant
This project implements an AI assistant for Gaia Sky. It consists of a RAG setup that scrapes the Gaia Sky documentation and home page, transforms it to text, does the embedding to vectors, and persists them to Chroma DB. Then, a chatbot is provided to query any Ollama model with relevant context.
This project is described in the following blog post:
https://tonisagrista.com/blog/2025/gaiasky-ai-assistant
## Running
In order to run it, clone the repository and set up the virtual environment with `pipenv`:
```bash
# Install dependencies
pipenv install
# Enter the virtual environment
pipenv shell
```
Then, run the `gsai.py` script. Use the `--scrape` flag the first time so that the websites are scraped and the database is populated and saved to disk.
```bash
# Scrape websites, create embeddings, store to disk
gsai.py --scrape
```
Once the websites have been scraped, you can run the chatbot by launching the program with no arguments.
```bash
# Run the chatbot
gsai.py
```
Set the temperature with `-t`, and the number of results with `-k`. When in doubt, use the `-h` argument to show the help.