I am planning to use V-REP for Reinforcement Learning and I would, therefore, like to run several v-rep instances in parallel. For that, I use Docker to create v-rep containers. However, my simulation is gradually getting slower as it is being stopped and started again (i.e. reset).
As an example see the following lines of code from a python script that resets the simulation 10 times:
Code: Select all
while trails < 10 vrep.simxStopSimulation(clientID,blocking) vrep.simxGetPingTime(clientID) vrep.simxStartSimulation(clientID,blocking) vrep.simxGetPingTime(clientID) CODE THAT WAITS FOR THE SIM TIME TO REACH 5s trails = trails+1
Is there anything I have forgotten to clean between trails? Or something that I need to include in the python script? Any help and suggestions are well received!