def policy_fn(name, ob_space, ac_space):
return mlp_policy.MlpPolicy(
name=name,
ob_space=ob_space,
ac_space=ac_space,
hid_size=args.size,
num_hid_layers=args.layers,
activation=args.activation,
keep=args.keep,
)
def policy_fn(name, ob_space, ac_space):
return mlp_policy.MlpPolicy(
name=name,
ob_space=ob_space,
ac_space=ac_space,
hid_size=args.hid_size,
num_hid_layers=args.num_hid_layers,
bound_by_sigmoid=args.bound_by_sigmoid,
sigmoid_coef=args.sigmoid_coef,
activation=args.activation,
normalize_obs=not args.nonormalize_obs,
gaussian_fixed_var=not args.nogaussian_fixed_var,
avg_norm_symmetry=args.avg_norm_symmetry,
symmetric_interpretation=args.symmetric_interpretation,
stdclip=args.stdclip,
actions=args.actions,
gaussian_bias=args.gaussian_bias,
gaussian_from_binary=args.gaussian_from_binary,
parallel_value=args.parallel_value,
pv_layers=args.pv_layers,
pv_hid_size=args.pv_hid_size,
three=args.three)
def policy_fn(name, ob_space, ac_space):
return mlp_policy.MlpPolicy(name=name, ob_space=ob_space, ac_space=ac_space,
hid_size=64, num_hid_layers=2)