Spaces:
Running
Running
| vectordbText = """ | |
| We will use the [Qdrant](https://qdrant.tech/documentation/) server deployment as microservice. \ | |
| You can either deploy it as individually or you can use it as one server to serve multiple \ | |
| chatbots (like in image below) by having multiple collections (multiple collection can also serve one chatbot) | |
| # Qdrant Vector Database Server on Hugging Face Spaces | |
| [ChaBo_QdrantServer](https://huggingface.co/spaces/GIZ/chatfed_QdrantServer/blob/main/README.md) Space hosts \ | |
| a Qdrant vector database instance. This is just a Infrastructural component and doesnt\ | |
| not serve any user application through its User Interface. However the admin task can be performed by\ | |
| accessing "<embedded space url>/dashboard" Ex:https://giz-chatfed-qdrantserver.hf.space/dashboard \ | |
| which is passsword protected. | |
| **Persistence:** Data is stored persistently in the `/data/qdrant_data` directory due to enabled persistent storage. | |
| **How to connect:** | |
| From your client application (e.g., your retrieval microservice), use the `qdrant-client` \ | |
| with the host set to your Space's direct URL and the appropriate port: | |
| ```python | |
| from qdrant_client import QdrantClient | |
| # Replace with your actual Space URL (e.g., https://your-username-qdrant-server.hf.space) | |
| QDRANT_HOST = "giz-chatfed-qdrantserver.hf.space" | |
| client = QdrantClient( | |
| host = QDRANT_HOST, | |
| port=443, # very important that port to be used for python client | |
| https=True, | |
| api_key = <QDRANT_API_KEY>,) | |
| ``` | |
| **API Documentation**: https://api.qdrant.tech/api-reference | |
| """ |