Setting up AI services
There are two types of AI services that Phaeton provides, namely, initial AI code security check for detection of malicious code before submission and AI code generation.
It is possbile to use AI services for coding help and for other purposes. For this you can use Jupyter AI extension.
To setup Jupyter AI extension, click on the icon with chat bubbles on left menu bar inside Jupyterlab.
Afterwards, click on “⚙️Start Here” button.
- As an example we provide access to Azure OpenAI API. Run the code in the following cell (
Shift + EnterorCtrl + Enter) to see the configuration for Azure OpenAI API for Jupyter AI.
import os
print(f"Deployment name: {os.environ.get('AZURE_AI_DEPLOYMENT')}")
print(f"Base API URL: {os.environ.get('AZURE_AI_ENDPOINT')}")
print(f"API version: {os.environ.get('AZURE_AI_API_VERSION')}")
print(f"Azure OpenAI API KEY: {os.environ.get('AZURE_AI_API_KEY')}")- Fill in the configuration form with the access information you obtained above.
- Afterwards you can use chat functionality of Jupyter AI extension.
- Embedding models will help you provide documents and ask questions related to contents of the documents.