Hello, I am currently working on my graduation thesis and using the dataset pcqm4m-v2. My task is actually similar to the KDDCUP challenge. I have encountered a problem now, which is that if I train the model based on the dataset you divided, I can only calculate the MAE of my model on the validation set. However, I need to know how the MAE is on the test set, which can be found on the dataset you provided, The HOMO-LUMO energy gaps corresponding to the test set are all nan. May I ask if it is convenient for you to provide me with the actual values of the test set?
Hello, I am currently working on my graduation thesis and using the dataset pcqm4m-v2. My task is actually similar to the KDDCUP challenge. I have encountered a problem now, which is that if I train the model based on the dataset you divided, I can only calculate the MAE of my model on the validation set. However, I need to know how the MAE is on the test set, which can be found on the dataset you provided, The HOMO-LUMO energy gaps corresponding to the test set are all nan. May I ask if it is convenient for you to provide me with the actual values of the test set?