Support model-specific max completion token setting

[PeterDaveHello]

Is your feature request related to a problem? Please describe.

It can be inconvenient to frequently adjust the global max completion token setting for different models, as each one supports different token limits. This often requires manual changes to ensure optimal output, which can be time-consuming and error-prone.

Describe the solution you'd like

I propose adding a max completion token setting for each model, allowing users to define specific token limits. The final token count used would be the minimum of the model-specific limit and the global setting, ensuring the model operates within its capabilities.

Additional context

While accurate values are ideal, you might consider using 4096 as a temporary placeholder for built-in models until more precise information is available. For custom models, if no specific value is set, the global setting would apply by default.

BTW, I'm currently trying to implement this feature, but I can't guarantee I'll have enough time or energy to complete it. If this proposal isn't opposed and I find the time, I'll try to work on it when possible.