Hello! I would like to ask you the following question:
1.Is this code from the paper "RouteNet: Leveraging Graph Neural Networks for network modeling and optimization in SDN"?
2.When verifying the generalization ability of the model, use the following line of code：
Is it possible to use other datasets to replace nsfnetbw and synth50bw, such as replacing synth50bw with geant2, and then using gbn to replace geant2 in the testing phase? Will this affect the final performance?
3.
Does the 100000 in this line of code represent the number of epochs, indicating how many training iterations are performed?
Hello! I would like to ask you the following question:
1.Is this code from the paper "RouteNet: Leveraging Graph Neural Networks for network modeling and optimization in SDN"?
2.When verifying the generalization ability of the model, use the following line of code：
Is it possible to use other datasets to replace nsfnetbw and synth50bw, such as replacing synth50bw with geant2, and then using gbn to replace geant2 in the testing phase? Will this affect the final performance?
3.
Does the 100000 in this line of code represent the number of epochs, indicating how many training iterations are performed?