dag_defaults.json

{
    "tag": "default",
    "data": {
      "dataset_type": "dummy",
      "time_periods": 2,
      "days": 10,
      "dummy_data_param": 4,
      "orders_density": 0.1,
      "order_distr": "star",
      "number_of_cars": 10,
      "time_periods_per_hour": 1,
      "order_sampling_multiplier": 1,
      "driver_sampling_multiplier": 1
    },
    "solver": {
      "cA2C": {
        "iterations": 20,
        "gamma": 0.9,
        "learning_rate": 1e-3,
        "weight_poorest": 0,
        "normalize_rewards": 1,
        "minimum_reward": 1,
        "batch_size": 2000,
        "include_income_to_observation": 0,
        "poorest_first": 1,
        "idle_reward": 0,
        "hold_observation": 1,
        "draw_freq": 2,
        "draw": 1,
        "entropy_coef": 0.01,
        "action_mask": 0,
        "discrete": 0,
        "eval_freq": 0,
        "robust": 0
      },
      "Gym": {
        "num_cpu": 4,
        "training_iterations": 100000,
        "gamma": 0.99,
        "ent_coef": 0.01,
        "learning_rate": 0.00025,
        "vf_coef": 0.5,
        "max_grad_norm": 0.5,
        "lam": 0.95,
        "noptepochs": 4,
        "cliprange": 0.2,
        "weight_poorest": 0,
        "normalize_rewards": 0,
        "minimum_reward": 1,
        "include_income_to_observation": 1,
        "poorest_first": 1,
        "callback": 1,
        "save_freq": 100000,
        "eval_freq": 1,
        "idle_reward": 0,
        "hold_observation": 1,
        "draw": 1,
        "draw_freq": 2,
        "robust": 0,
        "robust_nu": 0.1,
        "robust_epsilon": 1,
        "robust_cmin": 0,
        "robust_cmax": 500,
        "robust_gamma": 0.3,
        "action_mask": 1,
        "discrete": 0,
        "batch_env": 0,
        "runs": 1
      },
      "No": {},
      "Diff": {},
      "Oriented": {},
      "RobustA2C": {
        "robust_gamma": 0.3,
        "cmin": 0,
        "cmax": 10000,
        "robust_epsilon": 0.1,
        "robust_nu": 0.1,
        "iterations": 20,
        "hold_observation": 1,
        "gamma": 0.9,
        "learning_rate": 1e-3,
        "weight_poorest": 0,
        "normalize_rewards": [1, 0],
        "minimum_reward": 1,
        "idle_reward": 0,
        "batch_size": 2000,
        "include_income_to_observation": 0,
        "poorest_first": 1,
        "draw_freq": 2,
        "draw": 1
      },
      "Split": {
        "shrinking_fraction": 0.8,
        "iterations": 10,
        "subsolver": "Diff",
        "weight_poorest": 0,
        "hold_observation": 1,
        "normalize_rewards": 1,
        "minimum_reward": 0,
        "idle_reward": 0,
        "poorest_first": 1,
        "include_income_to_observation": 0
      }
    },
    "problem": {
      "wc": 0.1,
      "testing_epochs": 1,
      "count_neighbors": 1,
      "penalty_for_invalid_action": 1000,
      "randomize_drivers": 1
    },
    "full_rerun": 1,
    "DEBUG": 0,
    "seed": 123,
    "runs": 1
}