Trouble replicating model performance #2
Description
Hi, I'm having trouble replicating model performance as described in the paper. Are the hyperparameters described here the ones used for the final model? Could you provide more details on the cross-validation and train:test split strategy?