I did not know about this resource that allows you to test various LLM models and even adjust parameters. One of the examples: build.nvidia.com/google/gemma-7b Adjust parameters like temperature, top_p, etc.