OptionalfrequencyPenalize new tokens based on their frequency in the text so far.
Model identifier used for the chat completion.
OptionalpresencePenalize new tokens based on whether they appear in the text so far.
LLM provider to route the request through.
Model identifier used for thread summarization.
OptionaltemperatureSampling temperature between 0 and 2.
OptionaltopNucleus sampling probability mass.
Parameters for chat completion requests sent through the launcher API.