DeepGS
DeepGS copied to clipboard
Issue with the code
Hello!
Line 393 in the training script (run_training.sh) MSE_dev = tester.test(dataset_dev)[0] raises an error ValueError: Found array with 0 sample(s) (shape=(0,)) while a minimum of 1 is required.
This is probably because dataset_dev, dataset_test = split_dataset(dataset_, 0.5) produces dataset_dev as []
Is it possible for you to look into this please?
Thank you and best wishes, Kadi
Hi! I got the same problems. Have you solved this problem?
He did the wrong the pre-progress. The length of the 5 components are not the same length -> can not use the zip function to create the dataset -> dataset is 1 then it's not satisfied that batch size.
I think what you said is right. And do you have any solutions?
He did the wrong the pre-progress. The length of the 5 components are not the same length -> can not use the zip function to create the dataset -> dataset is 1 then it's not satisfied that batch size.
Have you solved this problem?
Hi, I'm facing the same error. Was anyone able to resolve it ?
The process should be done by the author. This repository uses the same method with TransformerCPI. you can find it here: https://github.com/lifanchen-simm/transformerCPI
So sad that the author has abandoned us.