To control the length of reasoning I want to be able to define the number of tokens the model can use in the final answer
You won't be notified about changes to this idea.
Guest
