config.toml

[run]
save_dir = ".\\training_data\\"
run_name = "test-run"
n_episodes = 50
use_cuda = true
seed = 0

[agent]
n_hidden_layers = 2
n_hidden_units = 64
lr = 4.4e-4
gamma = 0.99
batch_size = 256
buffer_size = 500_000
polyak_step_size = 0.995
update_every = 1
lambda_t = 400.0
lambda_s = 400.0
n_taus = 8
n_cos = 64
n_taus_exp = 8   # number of taus used for calculating the expectation
risk_distortion = "wang"  # Options: "wang", "", "", ""
risk_measure = -0.5

[training]
t_max = 30.0
c = [4, 1, 1]
loc_penalty = -500

[training.pitch_ref]
block_width = 10.0
smooth_width = 6.0
max_ampl = 20.0
nr_levels = 15
vary_timings = 0.1
start_with_zero = false

[training.roll_ref]
block_width = 8.0
smooth_width = 3.0
max_ampl = 25.0
nr_levels = 15
vary_timings = 0.1
start_with_zero = false