Fix: Ensure correct usage of gpt-4o instead of defaulting to gpt-4

[felipe-de-godoy]

Overview

This pull request addresses an issue where the configuration intended to use gpt-4o was defaulting to gpt-4. The fix ensures that the system correctly uses gpt-4o when specified.

Related Issue

• Issue Misconfiguration when using gpt-4o leads to gpt-4 being used #20: Misconfiguration when using gpt-4o leads to gpt-4 being used

Changes Made

• Modified the model initialization logic to correctly handle the gpt-4o parameter.

Testing

• Verified that specifying gpt-4o in the configuration now correctly initializes the gpt-4o model.

Impacts

• gpt-4o is cheaper than 'gpt-4', so using the wrong is costly