https://github.com/eavanvalkenburg/sk-demo-cloudbrew
https://github.com/eavanvalkenburg/sk-demo-cloudbrew
Last synced: 3 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/eavanvalkenburg/sk-demo-cloudbrew
- Owner: eavanvalkenburg
- Created: 2024-12-12T10:54:40.000Z (6 months ago)
- Default Branch: main
- Last Pushed: 2024-12-13T12:59:09.000Z (6 months ago)
- Last Synced: 2024-12-29T15:39:03.610Z (5 months ago)
- Language: Python
- Size: 90.8 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Semantic Kernel Demo for CloudBrew 2024 - A two-day Microsoft Azure event
This demo shows how to use [Semantic Kernel](https://github.com/microsoft/semantic-kernel) to run a web app using mesop that uses both online models with OpenAI and offline models with Ollama.
## Prerequisites
Copy the .env.example file and rename it to .env. Fill in the values for the following variables:
- OPENAI_API_KEY
- OPENAI_CHAT_MODEL_ID (recommended is GPT-4o)
- OPENAI_EMBEDDING_MODEL_ID (recommended is text-embedding-3-small)
- AZURE_AI_SEARCH_ENDPOINT
- AZURE_AI_SEARCH_API_KEY
- AZURE_AI_SEARCH_INDEX_NAME
You can also adjust the Qdrant settings.
The data loading requires this environment variable:
- GITHUB_TOKEN
### qdrant
Setup qdrant, this can be done in a container locally, or on a remote server. The following command will start a qdrant container on your local machine:
```bash
docker run -d -p 6333:6333 -p 6334:6334 qdrant/qdrant:latest
```
### venv
Create a virtual environment and activate it:
```bash
uv venv --python=3.12
source .venv/bin/activate (or .venv/Scripts/activate on Windows)
uv pip install -r requirements.txt --prerelease=allow
```
## Loading data
To load the data, run the following command, depending on the repo used, this might take a while:
```bash
source .venv/bin/activate (or .venv/Scripts/activate on Windows)
python data_ingestion/main.py
```
You can control which service you want to have indexed, by using the following flags:
```bash
python data/main.py --no-azure # this will only index Qdrant
```
or
```bash
python data/main.py --no-qdrant # this will only index Azure AI
```
By default, both will be indexed.
To review the data when you have Qdrant running locally you can open: `http://localhost:6333/dashboard` in your browser.
## Running the app
To run the app, simply run the following command:
```bash
source .venv/bin/activate
mesop main.py
```
The app will then be available in your browser at `http://localhost:32123/chat`.
When you want to go offline, either set the `MODE` environment variable to `offline` (this won't be picked up until you return mesop) or turn off the network.
If the `MODE` environment is anything other then `offline` or `online`, the app try to actually get a connection (it does a test to the `1.1.1.1` DNS Server by default).