About predict results

[zz990816]

Hello, thanks very much for providing the code. The work left a deep impression on me and I try to recurrent the results. However, I can't get the similar results as you show in the paper. And I notice that you recurrent the result of RCR and get much better results than what RCR showed in their paper.
I wonder if you change some hyperparameters or something else to get the results. And can you provide the train,val,test data split of the four dataset if possible.
Thank you very much!

[whitebirds]

Hello, I would also like to try replicating and migrating downstream tasks. I would like to ask you about the dataset you have chosen for replicating the experiment, as well as the processes and issues you need to pay attention to. Thank you very much for your answers.