Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/embedelite/sk-hackathon
The EmbedElite Semantic Kernel plugin submission for the SK Hackathon
https://github.com/embedelite/sk-hackathon
Last synced: about 1 month ago
JSON representation
The EmbedElite Semantic Kernel plugin submission for the SK Hackathon
- Host: GitHub
- URL: https://github.com/embedelite/sk-hackathon
- Owner: embedelite
- Created: 2023-07-23T17:00:58.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2023-07-26T22:19:12.000Z (over 1 year ago)
- Last Synced: 2024-08-01T13:18:05.717Z (4 months ago)
- Language: Python
- Homepage: https://www.embedelite.com
- Size: 174 KB
- Stars: 5
- Watchers: 2
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
- awesome-semantickernel - EmbedElite marketplace - made premium context based on embeddings via an API (Plugins)
README
Relevant LLM-ready assets for professionals.
> 🥳 We are among the winners of the SK Hackathon! Thanks to the SK Team and community for your interest and votes. Since the hackathon has ended, we had to deactivate the API key used in this repo. Also we are going to refine our plugin/provide more integration examples soon. If you would like to get access to our API or discover more embeddings/RAG solutions, please check our website [www.embedelite.com]([linkurl](https://www.embedelite.com))
# SK Hackathon EmbedElite Plugin
This project is creating a plugin for the EmbedElite marketplace for the SK Hackathon. The Sementic Kernel plugin facilitates fetching ready-made premium context based on embeddings via an API.
# About EmbedElite
EmbedElite is a marketplace for LLM assets, i.e., data chunks, priced by demand and queried via Retrieval-Augmented Generation (“RAG”)
On one side, data vendors feed information onto the platform and are compensated per usage of their data and retrieval algorithm
On the other side, data consumers perform queries enriched by the relevant assets and algorithms, paying per query and tokens retrieved
# How to Use EmbedElite
### Endpoints for Data Consumers
#### Request Type: POST
URL:
```http request
https://api.embedelite.com/query
```
Headers:
```http request
API-Key:
Content-Type: application/json
```
Body (JSON):
```json
{
"query": "",
"product_id": "",
"rag_id": "",
"price_floor": 1.0,
"price_cap": 5.0,
"currency": "EUR"
}
```
cUrl example:
```bash
curl -X POST https://api.embedelite.com/query \
-H "API-Key: sk-ee-9m839d3n98nh39fh9f3mhe98h3" \
-H "Content-Type: application/json" \
-d '{
"query": "What are the VAT rules in Germany if I sell services?",
"product_id": "vat-rules-eu",
"rag_id": "RAG_aks298msd9nj34hncs",
"price_floor": 1.0,
"price_cap": 5.0,
"currency": "EUR"
}'
```
Response Example:
```json
{
"response": "If you sell services in Germany, you usually charge the German VAT rate of 19%. However, there are exceptions for broadcasting, telecommunication, and electronically-supplied services, which must be charged at the VAT rate of the customer's country. If your customer is in another EU country, your invoice must contain specific information, such as the customer's name and address, the date of the invoice, the VAT rate, and the total amount including VAT. If your customer is outside the European Union, you must not charge VAT.",
"updated_at": "2023-07-25"
"currency": "EUR",
"paid": 2.3,
"originalQuery": "What are the VAT rules in Germany if I sell services?",
"price_cap": 2.3,
"price_floor": 2.3,
}
```
## For Vendors
Data vendors can simple push content via a POST endpoint to the platform. EmbedElite will keep the data confidential and intellectual ownership stays with the vendor. The ownership is defined in the metadata field. If you are a data vendor, please contact us for the data vendor API access: [email protected].
## Requirements to use the plugin
- [Python](https://www.python.org/downloads/) 3.8 or higher
- [Poetry](https://python-poetry.org/) for package handling and dependency management
- [Semantic Kernel Tools](https://marketplace.visualstudio.com/items?itemName=ms-semantic-kernel.semantic-kernel)
## Sample Configuration
A `.env` file housed within the project is used for configuring the sample. It contains API keys and other confidential settings.
Ensure you possess an
[Open AI API Key](https://openai.com/api/) or an
[Azure Open AI service key](https://learn.microsoft.com/azure/cognitive-services/openai/quickstart?pivots=rest-api)
Create a new file titled `.env` by duplicating the `.env.example` file. Then transfer your API keys into the new `.env` file:
```
OPENAI_API_KEY=""
OPENAI_ORG_ID=""
AZURE_OPENAI_DEPLOYMENT_NAME=""
AZURE_OPENAI_ENDPOINT=""
AZURE_OPENAI_API_KEY=""
EMBEDELITE_API_KEY=""
```
## Sample Execution
Within Visual Studio Code, press `F5` to run the console application. As defined in `launch.json` and `tasks.json`, Visual Studio Code will implement `poetry install` and then `python hello_world/main.py`
To build and execute the console application from the terminal, apply the following commands:
```powershell
poetry install
poetry run python main.py
```