IBM / rl-testbed-for-energyplus

Reinforcement Learning Testbed for Power Consumption Optimization using EnergyPlus
MIT License
190 stars 77 forks source link

how to specify number of episodes #72

Open yashviagrawal opened 2 years ago

yashviagrawal commented 2 years ago

Hello,

can you tell me how and where do I specify the number of episodes to run for?

time python3 -m baselines_energyplus.trpo_mpi.run_energyplus --num-timesteps 1000000000 when I run the above command, it keeps on running for innumerable episodes, I ran it for 237 episodes, and I quit the code, but I suspect it would've continued for more episodes had I not stopped.

is there a specific python file, and a proper function in that file, that specifies the no of episodes to run for?

biemann commented 2 years ago

Hi,

An episode is typically 35040 steps long (one step every 15 minutes and there are 35040 quarter hours in a (non-leap) year), you can multiply the number of episodes by this number if you want the number of steps to train the model.