swarmrl.trainers.episodic_trainer Module API Reference¶
Module for the EpisodicTrainer
EpisodicTrainer
¶
Bases: Trainer
Class for the simple MLP RL implementation.
Attributes¶
rl_protocols : list(protocol) A list of RL protocols to use in the simulation.
Source code in swarmrl/trainers/episodic_trainer.py
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 |
|
perform_rl_training(get_engine, system, n_episodes, episode_length, reset_frequency=1, load_bar=True)
¶
Perform the RL training.
Parameters¶
get_engine : callable Function to get the engine for the simulation. system_runner : espressomd.System Engine used to perform steps for each agent. n_episodes : int Number of episodes to use in the training. episode_length : int Number of time steps in one episode. reset_frequency : int (default=1) After how many episodes is the simulation reset. load_bar : bool (default=True) If true, show a progress bar.
Notes¶
If you are using semi-episodic training but your task kills the simulation, the system will be reset.
Source code in swarmrl/trainers/episodic_trainer.py
26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 |
|