While testing a bit more with the plugin i found 2 use cases that could benefit from a better control over temperature (amount of randomness in output) and frequency penalty.
When asking for summary or other factual information, it is good to set the temperature low, but when i ask to come up with new ideas based on my notes, it might be beneficial to have a way higher randomness in the answer.
Both options are useable within the openAI API (see also text-generator plugin wich provides this access also).
Therefor i would request to make these changeable in the settings.
While testing a bit more with the plugin i found 2 use cases that could benefit from a better control over temperature (amount of randomness in output) and frequency penalty.
When asking for summary or other factual information, it is good to set the temperature low, but when i ask to come up with new ideas based on my notes, it might be beneficial to have a way higher randomness in the answer.
Both options are useable within the openAI API (see also text-generator plugin wich provides this access also). Therefor i would request to make these changeable in the settings.