This is an LLM chatbot equipped with domain-specific knowledge, leveraging document and vector databases. Users can configure the chatbot's profile to elicit more tailored responses.
One can set up the Anaconda environment from the file environment.yml:
conda env create -f environment.ymlTo insert the notes, run the command-line script insert_notes.py I have built:
python insert_notes.py "<Directory holding the note files>"For more help messages, type the command:
python insert_notes.py --helpThis app is built with Streamlit. Type the following command to run the app in the browser:
streamlit run app.pyImagine wanting to learn how to prepare specific dishes.
I've discovered a fantastic collection of recipes written in Markdown
from this outstanding repository.
(All recipes, formatted as Markdown files, are housed within the recipes directory of this project.)
Although the collection is vast,
I prefer NOT to sift through each recipe manually.
This chatbot offers a solution.
In this demonstration, the chatbot takes on the role of a chef well-versed in a myriad of recipes. I can effortlessly inquire about the preparation of any particular dish.
Click the thumbnail below to watch a video demo:
