How to set simulation parameter values from your training script

You are allowed to set parameter values in your AnyLogic simulation from your python training script. Simply define them in the env_config and env_param:

from alpyperl import AnyLogicEnv
from ray.rllib.algorithms.ppo import PPOConfig

# Set checkpoint directory.
checkpoint_dir = "./resources/trained_policies/cartpole_v0"

# Initialize policy.
policy = (
    PPOConfig()
    .rollouts(
        num_rollout_workers=2,
        num_envs_per_worker=2
    )
    .fault_tolerance(
        recreate_failed_workers=True,
        num_consecutive_worker_failures_tolerance=3
    )
    .environment(
        AnyLogicEnv,
        env_config={
            'run_exported_model': True,
            'exported_model_loc': './resources/exported_models/cartpole_v0',
            'show_terminals': False,
            'verbose': False,
            'checkpoint_dir': checkpoint_dir,
            'env_params': {
                'cartMass': 1.0,
                'poleMass': 0.1,
                'poleLength': 0.5,
            }
        }
    )
    .build()
)

# Perform training.
for _ in range(10):
    result = policy.train()

# Save policy checkpoint.
policy.save(checkpoint_dir)
print(f"Checkpoint saved in directory '{checkpoint_dir}'")

# Close all enviornments.
policy.stop()

Important

You must comply with the following conditions:

• The parameter name must be the same as the parameter name in the AnyLogic model.

• The parameter types suported are only primitive types. (e.g. int, float, bool, etc.)