Allow GPT model to be configured via config file

[t4nature]

Summary:
This update introduces the capability for users to select the desired GPT model version via a configuration file. It moves away from the hardcoded GPT-3.5 model, enhancing flexibility and user control over model selection.

Changes:

• Removed the hardcoded reference to GPT-3.5 model.
• Implemented functionality to read the GPT model version from the configuration file (config.example.json), allowing users to specify their preferred model.
• Updated config.example.json to include a placeholder for the GPT model version.
• Modified documentation to guide users on how to specify the model version in the configuration file.

Motivation:
This change was motivated by the need to empower users with the flexibility to easily switch between different GPT models based on their requirements or the availability of newer models from OpenAI, without needing to alter the code.

Testing:
The feature was tested by specifying different GPT model versions in the config.example.json file. The application successfully utilized the specified models at runtime. Tests were conducted with GPT-3.5, GPT-4 (hypothetical future model), ensuring that the application behaves as expected across different model configurations.

The test code has not been uploaded.