This subfolder contains all the scripts that were used to train the universal FlashMD models.
The number in train_<number>.sh refers to the number of MD steps that are to be predicted
by the FlashMD models. The "data/" subfolder should contain the processed universal datasets.
