Is there a design problem with the reward function of large language models?
@chenbin8253 Жыл бұрын
Many problems. Such as how to design the function form, component, hyperparameters etc. Design Reward function for a specific task or general tasks ...