Thank you very much for your outstanding work. Could you please share the detailed hyperparameters used for training and fine-tuning the experiments?