Error: Unhandled exception. System.AggregateException: One or more errors occurred. with HTTP status code: BadRequest. Content: {"object":"error","message":"Invalid model: None","type":"invalid_model","param":null,"code":"1500"}).

public class Model
{
    private List<ChatMessage> _chatHistory = [];
    private const string DocumentPath = "posts.txt";
    private const string QdrantCollectionName = "HotDog_Project";
    private ISemanticTextMemory _memory;
    private MistralClient _client;
    
    public Model(string mistralApiKey,string qdrantApiKey ,string qdrantEndPoint)
    {
        var mistralApiAuth = new APIAuthentication(mistralApiKey);
        _client = new MistralClient(mistralApiAuth);
        
        HttpClient qdrantHttpClient = new HttpClient();
        qdrantHttpClient.BaseAddress = new Uri(qdrantEndPoint);
        qdrantHttpClient.DefaultRequestHeaders.Add("api-key", qdrantApiKey);
        
        var qdrantClient = new QdrantVectorDbClient(qdrantHttpClient,512);

        var embeddingService = _client.Embeddings.AsTextEmbeddingGenerationService();
        
        _memory = new SemanticTextMemory(new QdrantMemoryStore(qdrantClient),
            embeddingService);
        
        Task.Run(async () => await LoadAndEmbedDocuments()).Wait();
    }
    private async Task LoadAndEmbedDocuments()
    {
        // reading and splitting the text 
        var documentText = await File.ReadAllTextAsync(DocumentPath);
        var chunks = SplitTextIntoChunks(documentText, 300);
        // embedding text
        foreach (var (chunk, index) in chunks.Select((chunk, index) => (chunk,  index)))
        {
            var chunkId = $"chunk_{index}"; // Example: "chunk_0", "chunk_1"
            **await _memory.SaveInformationAsync(QdrantCollectionName, chunk, chunkId);**
        }

    }

The error originates from the _memory.SaveInformationAsync function that tries to automatically use the registered mistral embedding text embedding service. I get that the reason for the bad request is that the "model" is missing in the request headers but I cannot find a way to customize the request headers to add the model parameter.

I tried creating a separate client for embedding and doing the following:

var embeddingHttpClient = new HttpClient();
        embeddingHttpClient.DefaultRequestHeaders.Add("model",ModelDefinitions.MistralEmbed);
        var embeddingClient = new MistralClient(mistralApiAuth, embeddingHttpClient);
        var embeddingService = embeddingClient.Embeddings.AsTextEmbeddingGenerationService();
_memory = new SemanticTextMemory(new QdrantMemoryStore(qdrantClient),
            embeddingService);

This doesn't work either.

I also checked the Mistral.SDK documentation but it does not show an example for integrating with the SemanticTextMemory object.