nn_example_cfg.json

{
    "experiment": {
        "gamma": 0.98,
        "horizon": 5,
        "base_policy": 0.8,
        "eval_policy": 0.2,
        "stochastic_env": true,
        "stochastic_rewards": false,
        "sparse_rewards": false,
        "num_traj": 8,
        "is_pomdp": false,
        "pomdp_horizon": 2,
        "seed": 1000,
        "experiment_number": 0,
        "access": 0,
        "secret": 0,
        "to_regress_pi_b": {
            "to_regress": false,
            "model": "defaultCNN",
            "max_epochs": 100,
            "batch_size": 32,
            "clipnorm": 1.0
        },
        "frameskip": 1,
        "frameheight": 1
    },
    "models": {
        "FQE": {
            "model": "defaultCNN",
            "convergence_epsilon": 1e-4,
            "max_epochs": 100,
            "batch_size": 32,
            "clipnorm": 1.0
        },
        "Retrace": {
            "model": "defaultCNN",
            "convergence_epsilon": 1e-4,
            "max_epochs": 3,
            "batch_size": 32,
            "clipnorm": 1.0,
            "lamb": 0.9
        },
        "Tree-Backup": {
            "model": "defaultCNN",
            "convergence_epsilon": 1e-4,
            "max_epochs": 3,
            "batch_size": 32,
            "clipnorm": 1.0,
            "lamb": 0.9
        },
        "Q^pi(lambda)": {
            "model": "defaultCNN",
            "convergence_epsilon": 1e-4,
            "max_epochs": 3,
            "batch_size": 32,
            "clipnorm": 1.0,
            "lamb": 0.9
        },
        "Q-Reg": {
            "model": "defaultCNN",
            "convergence_epsilon": 1e-4,
            "max_epochs": 100,
            "batch_size": 32,
            "clipnorm": 1.0
        },
        "MRDR": {
            "model": "defaultCNN",
            "convergence_epsilon": 1e-4,
            "max_epochs": 100,
            "batch_size": 32,
            "clipnorm": 1.0
        },
        "IH": {
            "model": "defaultCNN",
            "convergence_epsilon": 1e-4,
            "max_epochs": 100,
            "batch_size": 32,
            "clipnorm": 1.0
        },
        "MBased": {
            "model": "defaultModelBasedCNN",
            "convergence_epsilon": 1e-4,
            "max_epochs": 100,
            "batch_size": 32,
            "clipnorm": 1.0,
            "max_traj_length": 25
        },
        "IS": {"model": "tabular"}
    }
}