Title are not generated for Phi-3

[newptcai]

What happened?

I run phi-3 with Ollama. I can have conversation. But the titles are not created.

Steps to Reproduce

I run ollama through docker and LibreChat via docker compose. Here's the section in librechat.yaml

    # Ollama Example
    - name: "Ollama"
      apiKey: "ollama"
      # use 'host.docker.internal' instead of localhost if running LibreChat in a docker container
      baseURL: "http://localhost:11434/v1/chat/completions" 
      models:
        default: [
          "phi3",
          "llama3",
          "mistral",
          "codellama",
          "gemma:2b",
          "gemma:7b"
          ]
        fetch: false # fetching list of models is not supported
      titleConvo: true
      titleModel: "llama3"
      summarize: false
      summaryModel: "llama3"
      forcePrompt: false
      modelDisplayLabel: "Ollama"
      addParams:
            "stop": [
              "<|start_header_id|>",
              "<|end_header_id|>",
              "<|eot_id|>",
              "<|reserved_special_token|>",
              "<|end|>",
              "<|assistant|>"
            ]

What browsers are you seeing the problem on?

No response

Relevant log output

No response

Screenshots

Code of Conduct

• I agree to follow this project's Code of Conduct

[danny-avila]

This is a new feature introduced today, but for both titleModel and summaryModel, you can now set it to "current_model" which will use the currently selected model for title generation.

Maybe you are having an issue with ollama attempting to load llama3 and it times out since it was not loaded already.

I use phi3 without issues this way.

Also, try commenting or removing addParams for the token delimiters, as they may no longer be necessary as a default set of values are no longer passed.

My config:

    - name: "Ollama"
      apiKey: "ollama"
      baseURL: "http://localhost:11434/v1/"
      models:
        default: [
          "llama3:latest",
          "command-r",
          "mixtral",
          "phi3",
          ]
        fetch: false # fetching list of models is not supported
      titleConvo: true
      titleModel: "current_model"
      modelDisplayLabel: "Ollama"

You will have to update to today's latest commit, let me know if you need help updating

[newptcai]

Hi, thanks. Since I am using docker, I will have to wait.

[hbprotoss]

Not working in v0.7.5

    - name: "DeepSeek"
      apiKey: "${DEEPSEEK_API_KEY}"
      baseURL: "https://api.deepseek.com"
      models:
        default: 
          [
            'deepseek-chat'
          ]
        fetch: true
        titleConvo: true
        titleModel: "deepseek-chat"
        summarize: false
        summaryModel: "deepseek-chat"
        forcePrompt: false
        modelDisplayLabel: "DeepSeek"

the log shows it is still using gpt-3.5-turbo for title generating

2024-11-28T05:48:10.833Z debug: [OpenAIClient] chatCompletion
{
  baseURL: "https://api.deepseek.com",
    modelOptions.model: "gpt-3.5-turbo",
    modelOptions.user: "67359dd97be51631f75859aa",
    modelOptions.temperature: 0.2,
    modelOptions.presence_penalty: 0,
    modelOptions.frequency_penalty: 0,
    modelOptions.max_tokens: 16,
    // 1 message(s)
    modelOptions.messages: [{"role":"system","content":"Please generate a concise, 5-word-or-less title for the conversation, us... [truncated]],
}
2024-11-28T05:48:10.838Z debug: Making request to https://api.deepseek.com/chat/completions
2024-11-28T05:48:10.926Z warn: [OpenAIClient.chatCompletion][create] API error
2024-11-28T05:48:10.927Z error: [OpenAIClient.chatCompletion] Unhandled error type Error: 400 Model Not Exist
2024-11-28T05:48:10.927Z error: [OpenAIClient] There was an issue generating the title with the completion method Error: 400 Model Not Exist
2024-11-28T05:48:10.927Z debug: [OpenAIClient] Convo Title: New Chat
2024-11-28T05:48:10.927Z debug: [saveConvo] api/server/services/Endpoints/openAI/addTitle.js