smbot/config.lua

local preset = rawget(_G, 'preset') or 'ars'

local common_cfg = {
    -- read-only modes:
    playable_mode = false,
    playback_mode = false,

    -- controlling in-game mechanics:
    starting_world = 1, -- set to 0 to randomize the world every epoch.
    starting_level = 1, -- set to 0 to randomize the level every epoch.
    start_big = false,
    starting_lives = 0,
    min_time = 100,
    max_time = 300,
    timer_loser = 1/2, -- decrement the timer when mario is slacking off.

    -- rewards:
    decrement_reward = false, -- bad idea, encourages mario to run into goombas.
    score_multiplier = 1, -- how much the ingame score influences our rewards.

    -- network initialization:
    init_zeros = false, -- instead of random normal noise.

    -- network inputs (connections):
    time_inputs = true, -- binary inputs of global frame count

    -- network layers:
    layernorm = false,
    reduce_tiles = false,
    bias_out = true,

    -- network evaluation (sampling joypad):
    frameskip = 4,
    -- true greedy epsilon has both deterministic and det_epsilon set.
    deterministic = false, -- use argmax on outputs instead of random sampling.
    det_epsilon = false, -- take random actions with probability eps.

    -- evolution strategy and non-rate hyperparemeters:
    es = 'ars',
    ars_lips = false, -- for ARS.
    epoch_top_trials = 9999, -- for ARS.

    -- sampling:
    deviation = 1.0,
    unperturbed_trial = true, -- perform an extra trial without any noise.
    graycode = false, -- for ARS.
    negate_trials = true, -- try pairs of normal and negated noise directions.
    -- AKA antithetic sampling. note that this doubles the number of trials.
    min_refresh = 0.1, -- for SNES.

    -- epoch-related rates:
    learning_rate = 1.0,
    mean_adapt = 1.0, -- for SNES, xNES.
    weight_decay = 0.0,
    sigma_decay = 0.0, -- for SNES.
}

local cfg
if preset == 'snes' then

    cfg = {
        es = 'snes',

        log_fn = 'logs-snes.csv',
        params_fn = 'params-snes.txt',

        starting_lives = 1,
        min_time = 300,
        timer_loser = 1.0,

        epoch_trials = 100,
        negate_trials = false,

        deviation = 1.0,
        min_refresh = 0.2,

        learning_rate = 0.1, -- TODO: rename to learn_primary or something.
        mean_adapt = 0.5, -- TODO: rename to learn_secondary or something.
        weight_decay = 0.02, -- note: multiplied by its std, and mean_adapt.
        sigma_decay = 0.01, -- note: multiplied by learning_rate.
    }

elseif preset == 'xnes' then

    cfg = {
        es = 'xnes',

        log_fn = 'logs-xnes.csv',
        params_fn = 'params-xnes.txt',

        start_big = true,
        min_time = 300,
        timer_loser = 1.0,

        decrement_reward = false,
        score_multiplier = 0,

        init_zeros = true,

        reduce_tiles = true,
        bias_out = false,

        deterministic = false,

        deviation = 0.5,
        negate_trials = false,
        min_refresh = 0.1,

        epoch_trials = 50,

        learning_rate = 0.01,
        mean_adapt = 1.0,
        weight_decay = 0.0,
        sigma_decay = 0.0,
    }

elseif preset == 'ars' then

    cfg = {
        es = 'ars',
        epoch_top_trials = 20,
        ars_lips = false,

        log_fn = 'logs-ars.csv',
        params_fn = 'params-ars.txt',

        min_time = 300,
        timer_loser = 1.0,

        deviation = 0.1,

        epoch_trials = 25,

        learning_rate = 1.0,
        weight_decay = 0.0025,
    }

else

    error("invalid preset: "..tostring(preset))

end

-- TODO: so, uhh..
-- what happens when playback_mode is true but unperturbed_trial is false?

setmetatable(cfg, {
    __index = function(t, n)
        if common_cfg[n] ~= nil then return common_cfg[n] end
        if n == 'log_fn' then return nil end
        if n == 'params_fn' then return nil end
        if n == 'stats_fn' then return nil end
        error("cannot use undeclared config '" .. tostring(n) .. "'", 2)
    end
})

cfg.epoch_top_trials = math.min(cfg.epoch_trials, cfg.epoch_top_trials)

cfg.eps_start = 1.0 * cfg.frameskip / 64
cfg.eps_stop  = 0.1 * cfg.eps_start
cfg.eps_frames = 1000000
cfg.enable_overlay = cfg.playable_mode
cfg.enable_network = not cfg.playable_mode

assert(not cfg.ars_lips or cfg.unperturbed_trial,
       "cfg.unperturbed_trial must be true to use cfg.ars_lips")
assert(not cfg.ars_lips or cfg.negate_trials,
       "cfg.negate_trials must be true to use cfg.ars_lips")
assert(not (cfg.es == 'snes' and cfg.negate_trials),
       "cfg.negate_trials is not yet compatible with SNES")

return cfg