feat: add support for gpt5.

[aliarabat]

✨ Feature: Support for New GPT-5 Models

This Pull Request introduces full compatibility with OpenAI's new model series: gpt-5 and gpt-5-mini.

This upgrade is essential as the latest models use a revised API schema, making them incompatible with the current version of TopicGPT.

⚙️ Technical Changes
The primary changes address necessary parameter updates to correctly invoke the new gpt-5 models:

1. Renamed Max Tokens Parameter:

• The max_tokens argument is deprecated and no longer supported when invoking the gpt-5 series.

• Action: I have globally replaced instances of max_tokens with the new, required parameter: max_completion_tokens.

2. Adjusted Temperature Default:

• The gpt-5 model series is designed to accept a default temperature value of 1.0.

• Action: I have updated all relevant model invocation code to explicitly set the temperature to 1.0 to ensure optimal and consistent model behavior.

💡 Future Improvement (Follow-up Task)
While the current implementation hardcodes the temperature to 1.0, it would significantly improve the user experience to make this configurable.

Proposal: I suggest creating a follow-up issue (which I am happy to open) to refactor the temperature parameter. This would involve moving it to a configurable setting (e.g., via environment variables or a configuration file) to allow users to easily adjust the model's temperature without modifying the codebase.

✅ Testing Notes
I have verified successful completion calls using the following models: gpt-4o, gpt-4o-mini, gpt-5, gpt-5-mini, as well as Ollama models, including gemma3:latest and llama3.1:latest.