Add Support for Multiple Models

[JustinHsu1019]

Description

The project currently supports only a single model for generating Selenium scripts. Users may want to choose between different models (e.g., GPT-4, GPT-4-turbo, Gemini, Llama, or other APIs) for flexibility and potentially faster responses.

Proposed Solution

• Modularize the model invocation logic by moving it into a separate file or module.
• Add support for multiple models by creating a configuration file or menu where users can select the desired model.
• Implement a simple extension menu for model selection with options to configure model-specific parameters.

[guoriyue]

Support GPT4 GPT3.5 and DeepSeek now

[JustinHsu1019]

Great! Thank you~