bnn-upc / papers Goto Github PK

RouteNet-Erlang-main/Scheduling/Delay/check_predictions.py:

problem:
1、model.load_weights('./ckpt_dir/{}'.format(best))
Exception has occurred: ValueError
Unable to load weights saved in HDF5 format into a subclassed Model which has not created its variables yet. Call the Model first, then load the weights.
2、I can't find the following dataset
ds_test = input_fn('../../data/gnnet_data_set_evaluation_delays', label='PktsDrop', shuffle=False)
3、Why under this directory（data)
sys.path.insert(1, '../../data/')

Hi~I'm a student interested in your research, and I'm tring to reproduce the user case of network optimization with a packet-level simulator, I want to ask something about TM:
1.does TM mean the average rate of packets?
2.I think the higher TI means smaller interval in sending packets, but as you described in the paper , "In each source-destination pair, inter-packet arrival times are modeled with an exponential distribution whose mean is derived from the traffic defined in TM.",maybe is there anything different with arrival times and sending interval? And how can we modify the simulation settings in sending packets to get the similar results with RouteNet predictions?
Hope to get your help, thank you very much.