uploading documents with memory `volatile` and `qdrant` - chatbot cannot answer questions on the document

[ekdnam]

Describe the bug
A clear and concise description of what the bug is.
In ./webapi/appsettings.json, I tried keeping once MemoryStore.Type once asqdrant and then volatile as well.

I uploaded a document via the UI, but the chatbot cannot answer questions on the document.

Something like the following (I uploaded the llama README.md)

To Reproduce
Steps to reproduce the behavior:

1. follow the procedure to turn on the qdrant vector store

Expected behavior
A clear and concise description of what you expected to happen.

Chatbot can answer questions on the document

Screenshots
If applicable, add screenshots to help explain your problem.

Platform

• OS: MacOS
• IDE: Code
• Language: C#
• Source: main branch of repository

Additional context
Add any other context about the problem here.

[crickman]

The "user-intent" included in the prompt info displays this as the intent:

This confirms my suspicion that the quesion is just too ambiguous.

When I ask:

Does LLama support fine-tuning large language models?

The response includes the document snippet.

The experience you've reported is not specific to volitile or qdrant memory.

[AwesomeYuer]

@crickman I found same issue, have this bug been fixed? or in fixing? thanks @AwesomeYuer