Embedding and Qdrant ?

[jerpic]

Hello,

Any plan to store and retrieve embeddings into the qdrant vector database?

Regards,

[marcominerva]

Saving and retrieving embeddings in a Vector Database is beyond the scope of this library. If you need this feature, I suggest you to take a look at the Kernel Memory library: https://github.com/microsoft/kernel-memory

[marcominerva]

Yes, you can safely use ChatGptNet instead of Semantic Kernel.

[jerpic]

Hello @marcominerva,
Honestly, I'm struggling to find the way for using ChatGptNet instead of Semantic Kernel in your example, method "AskQuestionAsync" from the ApplicationMemoryService class. Any help would be very appreciated ;-)

`public async Task<MemoryResponse?> AskQuestionAsync(Question question, double minimumRelevance = 0.76, string? index = null)
{
// Reformulate the following question taking into account the context of the chat to perform keyword search and embeddings:
var reformulatedQuestion = await chatService.CreateQuestionAsync(question.ConversationId, question.Text);

    // Ask using the embedding search via Kernel Memory and the reformulated question.
    // If tags are provided, use them as filters with OR logic.
    var answer = await memory.AskAsync(reformulatedQuestion, index, filters: question.Tags.ToMemoryFilters(), minRelevance: minimumRelevance);

    if (answer.NoResult == false)
    {
        // If the answer has been found, add the interaction to the chat, so that it will be used for the next reformulation.
        await chatService.AddInteractionAsync(question.ConversationId, question.Text, answer.Result);

        var response = new MemoryResponse(answer.Result, answer.RelevantSources);
        return response;
    }

    return null;
}`

Regards,

[marcominerva]

You need to modify the ChatService.cs file using IChatGptClient instead of IMemoryCache and IChatCompletionService.

[jerpic]

Thank you for your reply.
Sorry to bother you. I thought that the ChatService class was there only to reformulate the question and not to chat about the documents stored into the KernelMemory.
If I change the implementation of ChatService class (for ChatGptNet), how can I use ChatGptNet to stream the answer (coming from var answer = await memory.AskAsync(reformulatedQuestion, index, filters: question.Tags.ToMemoryFilters(), minRelevance: minimumRelevance); )?

[marcominerva]

Unfortunately, at this moment Kernel Memory does not support streaming response.