Follow-up of Benchmark Confirmation on coGN and MODNet

[chao1224]

Hi there,

Thank you again for providing this nice benchmark tool!

Previously, I posted a confirmation issue (#254) on the data splits for SchNet training. Now I want to double-check if the coGN and MODNet have the same issue: that 80%-20% for training and test, and no validation is used. Can anyone help confirm this?

Any help is appreciated!

Regards,
Shengchao