feat: set a default temperature in the common local llm settings

[cpacker]

Please describe the purpose of this pull request.

• set a default temperature in the common local llm settings (use 0.8, most user-friendly lm frontends seem to be using 0.7/0.8 atm)
• this shouldn't change LM Studio calls (0.8 default) but it should change vLLM calls (1.0 default)
  • default settings on vLLM make it hard for the LLM to output structured content (eg lists)

How to test

Regression test to check runtime errors on various backends (caused by additional param in payload):

• lmstudio
• ollama
• webui
• llama.cpp
• koboldcpp
• vLLM

Have you tested this PR?

See checklist