def run_task(snapshot_config, *_):
"""Run task."""
with LocalTFRunner(snapshot_config=snapshot_config) as runner:
# env = TfEnv(normalize(MassSpringEnv_OptK_HwAsAction(params), normalize_action=False, normalize_obs=False, normalize_reward=True, reward_alpha=0.1))
env = TfEnv(MassSpringEnv_OptK_HwAsAction(params))
zip_project(log_dir=runner._snapshotter._snapshot_dir)
comp_policy_model = MLPModel(
output_dim=1,
hidden_sizes=params.comp_policy_network_size,
hidden_nonlinearity=tf.nn.tanh,
output_nonlinearity=tf.nn.tanh,
)
mech_policy_model = MechPolicyModel_OptK_HwAsAction(params)
policy = CompMechPolicy_OptK_HwAsAction(
name='comp_mech_policy',
env_spec=env.spec,
comp_policy_model=comp_policy_model,
mech_policy_model=mech_policy_model)
# baseline = GaussianMLPBaseline(
# env_spec=env.spec,
# regressor_args=dict(
# hidden_sizes=params.baseline_network_size,
# hidden_nonlinearity=tf.nn.tanh,
# use_trust_region=True,
# ),
# )
baseline = LinearFeatureBaseline(env_spec=env.spec)
algo = CMAES(env_spec=env.spec,
policy=policy,
baseline=baseline,
**params.cmaes_algo_kwargs)
runner.setup(algo, env)
runner.train(**params.cmaes_train_kwargs)
parser = argparse.ArgumentParser()
parser.add_argument('--seed',
default=int(now.timestamp()),
type=int,
help='seed')
parser.add_argument('--exp_id',
default=now.strftime("%Y_%m_%d_%H_%M_%S"),
help='experiment id (suffix to data directory name)')
args = parser.parse_args()
exp_prefix = 'cmaes_ppo_opt_k_{0}_{1}_params/seed_{2}'.format(
args.exp_id, params.n_springs, args.seed)
# CMA-ES global optimization
options = params.cmaes_options
options['seed'] = args.seed
options['verb_filenameprefix'] = os.path.join(os.environ['PROJECTDIR'],
'data/local',
exp_prefix.replace('_',
'-'), '-')
x0 = params.cmaes_x0
sigma0 = params.cmaes_sigma0
es = cma.CMAEvolutionStrategy(x0, sigma0, options)
es.optimize(cmaes_obj_fcn, args=[exp_prefix])
es.result_pretty()
zip_project(log_dir=os.path.join(os.environ['PROJECTDIR'], 'data/local',
exp_prefix.replace('_', '-')))