Update README.md

README.md

@@ -93,7 +93,7 @@ We believe these benchmarks will be generally useful and plan to open-source the
 
 All Shisa V2 models inherit the [chat templates](https://huggingface.co/docs/transformers/v4.37.1/chat_templating) of their respective base models and have been tested and validated for proper inference with both [vLLM](https://github.com/vllm-project/vllm) and [SGLang](https://github.com/sgl-project/sglang).
 
-We recommend running at lower temperatures: 0.0-0.2 for factual answers and 0.5-0.7 for creative tasks.
+Running sampler sweeps, we found the models operate well across a variety of temperatures in most settings. For translation tasks specifically, we recommend a lower temperatures (0.2) to increase accuracy. For role-play and creative tasks, a higher temp (eg 1.0) seems to give good results. To prevent cross-lingual token leakage we recommend a top_p of 0.9 or min_p of 0.1.
 
 No additional safety alignment has been done on these models, so they will largely inherit the base models' biases and safety profiles.