-Made it so that the user, when running Remy, must supply a protobuf that specifies the specific configuration range of networks to train the remy on
-Made a separate script that creates this protobuf for you when specifying the range via command line
-Made the configrange object ranges over all of the parameters - it was only ranges over some of them (just rtt and link speed) - now has ranges for mean on duration, off duration, senders
-Modified these ranges to have a low, high, and increment for sampling
-Added buffer size as a parameter
-Write a vector of trained networks as a protobuf to the outputted dna files: this is useful later if we ever wanted to know what networks a certain remy was trained on (if, in a certain branch, the networks are chosen randomly, like they are currently)
-Modified the evaluator, instead of randomly selecting some of the networks specified by the config range, to do a nested for loop to consider all options within the cube of configuration options
-Made it so that the user, when running Remy, must supply a protobuf that specifies the specific configuration range of networks to train the remy on -Made a separate script that creates this protobuf for you when specifying the range via command line
-Made the configrange object ranges over all of the parameters - it was only ranges over some of them (just rtt and link speed) - now has ranges for mean on duration, off duration, senders -Modified these ranges to have a low, high, and increment for sampling
-Added buffer size as a parameter
-Write a vector of trained networks as a protobuf to the outputted dna files: this is useful later if we ever wanted to know what networks a certain remy was trained on (if, in a certain branch, the networks are chosen randomly, like they are currently)
-Modified the evaluator, instead of randomly selecting some of the networks specified by the config range, to do a nested for loop to consider all options within the cube of configuration options