Persistent Issue with Ollama Configuration – LLM Mode Forced to "llamacpp" Despite Correct Configuration

[koffih]

Subject: Persistent Issue with Ollama Configuration – LLM Mode Forced to "llamacpp" Despite Correct Configuration

Hello,

I am currently working on setting up PrivateGPT with Ollama as the LLM provider. Despite having a correct configuration in the `default.yaml` file, I am encountering a persistent issue where the LLM mode is forced to "llamacpp" instead of "ollama". This results in the following error:

ValueError: Unsupported LLM mode: llamacpp

Here is a summary of the problem and the steps I have already taken:

Current Configuration

1. File default.yaml:
   The file /opt/ai-platform/privategpt/private_gpt/settings/profiles/default.yaml contains the following configuration:

   llm:
     provider: ollama
     mode: ollama
     config:
       model: mistral
       temperature: 0.7
       base_url: `http://localhost:11434```