sampling-create-message-request

record sampling-create-message-request {
  include-context: include-context,
  max-tokens: u64,
  messages: list​<sampling-message>,
  metadata: option​<json>,
  model-preferences: option​<model-preferences>,
  stop-sequences: option​<list​<string>>,
  system-prompt: option​<string>,
  temperature: option​<f64>,
}