ollama support - read for review

[jonathanalexander229]

Here's a summary of the changes we made to add Ollama support to your AI hedge fund application:

1. Environment Configuration:

   • Added OLLAMA_BASE_URL and OLLAMA_MODEL to .env.example to allow configuration of the Ollama instance and model
   • These variables are loaded by the existing dotenv configuration in main.py

2. Dependencies:

   • Added langchain-community package to pyproject.toml for Ollama integration
   • This package provides the ChatOllama implementation

3. Model Provider Updates in models.py:

   • Added import for ChatOllama: from langchain_community.chat_models import ChatOllama
   • Added OLLAMA to the ModelProvider enum
   • Added a single Ollama model entry to the AVAILABLE_MODELS list
   • Updated the get_model function to handle the Ollama provider type

4. Simplified Implementation:

   • Added error handling to check for environment variables
   • Kept the implementation simple and aligned with other model providers
   • Pulls the actual model name from environment variables when needed

These changes allow your application to use locally hosted Ollama models alongside the cloud-based LLM options, all configured through your .env file.