I am looking to fine tune a GNN and my supervisor suggested exploring different learning rates. I came across this tutorial video where he mentions that a randomised log space s