A simple chat bot for IBRS data
- Install the following Microsoft VScode extentions
- Jupyter - this enables you to run the notbooks directly in VScode rather than in Colab.
- Python - this enable you to run python in VScode
-
Clone this repo to your local machine. The easy way is to use VScode clone command.
CTL-SHIFT-P Git: Clone Clone from GitHub https://github.com/Kevin-McIsaac/IBRS.git
-
Check that python and pip are installed. In a terminal
python --version pip --version
-
Consider consider using a python virtual environment rather than a global installation. You can use VSCode using
CTRL-SHIFT P Venv CTRL-SHIFT P Python: Select Interpreter
or do this manually
python -m venv .venv
-
If you open the terminal using VScode this will automatically activated the the local virtual venv, other wise use 'source', then install the python modules listed in requirements.txt.
source .venv/bin/activate pip install -r requirements.txt
-
Create a .env file with the necessary API information or upload the version I sent you.
# IBRS Pinecone credentials INDEX='ibrs' PINECONE_API_KEY='' PINECONE_ENV='' #IBRS OPEN API credentials OPENAI_API_KEY= '' #IBRS WordPress credentials IBRS_API_BEARER = ''
The Q&A bot is called gideon.py and written in python using streamlit. You can run this using
streamlit run gideon.py
This will print a URL to the local web page to access the running application.