#1129 New Feature - Add option so pick the default model settings: --thinking --instruct

Descripción

It is still not clear which model parameters are being used when launching the server using ./setup.sh --host .

Qwen suggest's sampling parameters; it would be nice to have an option on ./setup.sh  like --thinking  or --instruct, as Qwen's documentation recommends:

> To achieve optimal performance, we recommend the following settings:
> 
> Sampling Parameters: We suggest using the following sets of sampling parameters:
> 
> Thinking Mode: temperature=1.0, top_p=0.95, top_k=20, min_p=0.0, presence_penalty=0.0, repetition_penalty=1.0
> Instruct (or non-thinking) mode: temperature=0.7, top_p=0.80, top_k=20, min_p=0.0, presence_penalty=1.5, repetition_penalty=1.0

Ref: https://huggingface.co/Qwen/Qwen3.8-Flash-Next

EDIT. I will submit a PR for this configuration soon, using Qwen3.8-Flash-Next and Strata :)