https://github.com/gbaeke/semantic-kernel-demo
https://github.com/gbaeke/semantic-kernel-demo
Last synced: 4 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/gbaeke/semantic-kernel-demo
- Owner: gbaeke
- Created: 2024-03-24T16:31:11.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2024-06-13T13:02:29.000Z (11 months ago)
- Last Synced: 2024-11-09T03:53:50.998Z (6 months ago)
- Language: Python
- Size: 12.7 KB
- Stars: 3
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.MD
Awesome Lists containing this project
README
# Azure OpenAI on your data with Semantic Kernel
Supply an .env file with the following variables:
```
AZURE_OPENAI_DEPLOYMENT_NAME="your_deployment_name"
AZURE_OPENAI_ENDPOINT="https://your_endpoint.openai.azure.com/"
AZURE_OPENAI_API_KEY="Your OpenAI API Key"
AZURE_OPENAI_API_VERSION="2023-12-01-preview"
AZURE_AISEARCH_API_KEY="Your Azure Search API Key"
AZURE_AISEARCH_URL="https://your_endpoint.search.windows.net"
AZURE_AISEARCH_INDEX_NAME="your index name"
MODE="search" # any other value than search will be functions mode
BING_SEARCH_API_KEY="APIKEY"
```
Create a virtual environment with `python -m venv venv`. Activate the virtual environment with `source venv/bin/activate`.
Install the requirements with `pip install -r requirements.txt`.
Run the app with `streamlit run data.py`.
**IMPORTANT**: you need an Azure AI Search index with fields you can map to the fields the code expects. The code expects the following fields:
- titleField
- urlField
- contentFields: list
- vectorFields: list
In the code, you map your fields to these fields like so:
```python
azure_ai_search_settings["fieldsMapping"] = {
"titleField": "Title",
"urlField": "Url",
"contentFields": ["Content"],
"vectorFields": ["contentVector"],
}
```
The vectorField contentVector contains a 1536-dimensional vector that represents the content of the document. You can use the OpenAI API to generate this vector. Use model `text-embedding-ada-002` to generate the vector.