def rllab_envpolicy_parser(env, args):
if isinstance(args, dict):
args = tonamedtuple(args)
env = RLLabEnv(env, mode=args.control)
if args.algo[:2] == 'tf':
env = TfEnv(env)
# Policy
if args.recurrent:
if args.feature_net:
feature_network = MLP(
name='feature_net',
input_shape=(env.spec.observation_space.flat_dim +
env.spec.action_space.flat_dim, ),
output_dim=args.feature_output,
hidden_sizes=tuple(args.feature_hidden),
hidden_nonlinearity=tf.nn.tanh,
output_nonlinearity=None)
elif args.conv:
strides = tuple(args.conv_strides)
chans = tuple(args.conv_channels)
filts = tuple(args.conv_filters)
assert len(strides) == len(chans) == len(
filts), "strides, chans and filts not equal"
# only discrete actions supported, should be straightforward to extend to continuous
assert isinstance(
env.spec.action_space,
Discrete), "Only discrete action spaces support conv"
feature_network = ConvNetwork(
name='feature_net',
input_shape=env.spec.observation_space.shape,
output_dim=args.feature_output,
conv_filters=chans,
conv_filter_sizes=filts,
conv_strides=strides,
conv_pads=('VALID', ) * len(chans),
hidden_sizes=tuple(args.feature_hidden),
hidden_nonlinearity=tf.nn.relu,
output_nonlinearity=None)
else:
feature_network = None
if args.recurrent == 'gru':
if isinstance(env.spec.action_space, Box):
policy = GaussianGRUPolicy(env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(
args.policy_hidden[0]),
name='policy')
elif isinstance(env.spec.action_space, Discrete):
policy = CategoricalGRUPolicy(
env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(args.policy_hidden[0]),
name='policy',
state_include_action=False if args.conv else True)
else:
raise NotImplementedError(env.spec.observation_space)
elif args.recurrent == 'lstm':
if isinstance(env.spec.action_space, Box):
policy = GaussianLSTMPolicy(
env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(args.policy_hidden),
name='policy')
elif isinstance(env.spec.action_space, Discrete):
policy = CategoricalLSTMPolicy(
env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(args.policy_hidden),
name='policy')
else:
raise NotImplementedError(env.spec.action_space)
else:
raise NotImplementedError(args.recurrent)
elif args.conv:
strides = tuple(args.conv_strides)
chans = tuple(args.conv_channels)
filts = tuple(args.conv_filters)
assert len(strides) == len(chans) == len(
filts), "strides, chans and filts not equal"
# only discrete actions supported, should be straightforward to extend to continuous
assert isinstance(
env.spec.action_space,
Discrete), "Only discrete action spaces support conv"
feature_network = ConvNetwork(
name='feature_net',
input_shape=env.spec.observation_space.shape,
output_dim=env.spec.action_space.n,
conv_filters=chans,
conv_filter_sizes=filts,
conv_strides=strides,
conv_pads=('VALID', ) * len(chans),
hidden_sizes=tuple(args.policy_hidden),
hidden_nonlinearity=tf.nn.relu,
output_nonlinearity=tf.nn.softmax)
policy = CategoricalMLPPolicy(name='policy',
env_spec=env.spec,
prob_network=feature_network)
else:
if isinstance(env.spec.action_space, Box):
policy = GaussianMLPPolicy(env_spec=env.spec,
hidden_sizes=tuple(
args.policy_hidden),
min_std=args.min_std,
name='policy')
elif isinstance(env.spec.action_space, Discrete):
policy = CategoricalMLPPolicy(env_spec=env.spec,
hidden_sizes=tuple(
args.policy_hidden),
name='policy')
else:
raise NotImplementedError(env.spec.action_space)
elif args.algo[:2] == 'th':
# Policy
if args.recurrent:
if args.feature_net:
feature_network = thMLP(
input_shape=(env.spec.observation_space.flat_dim +
env.spec.action_space.flat_dim, ),
output_dim=args.feature_output,
hidden_sizes=tuple(args.feature_hidden),
hidden_nonlinearity=tf.nn.tanh,
output_nonlinearity=None)
else:
feature_network = None
if args.recurrent == 'gru':
if isinstance(env.spec.observation_space, thBox):
policy = thGaussianGRUPolicy(
env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(args.policy_hidden[0]),
)
elif isinstance(env.spec.observation_space, thDiscrete):
policy = thCategoricalGRUPolicy(
env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(args.policy_hidden[0]),
)
else:
raise NotImplementedError(env.spec.observation_space)
# elif args.recurrent == 'lstm':
# if isinstance(env.spec.action_space, thBox):
# policy = thGaussianLSTMPolicy(env_spec=env.spec,
# feature_network=feature_network,
# hidden_dim=int(args.policy_hidden),
# name='policy')
# elif isinstance(env.spec.action_space, thDiscrete):
# policy = thCategoricalLSTMPolicy(env_spec=env.spec,
# feature_network=feature_network,
# hidden_dim=int(args.policy_hidden),
# name='policy')
# else:
# raise NotImplementedError(env.spec.action_space)
else:
raise NotImplementedError(args.recurrent)
else:
if args.algo == 'thddpg':
assert isinstance(env.spec.action_space, thBox)
policy = thDeterministicMLPPolicy(
env_spec=env.spec,
hidden_sizes=tuple(args.policy_hidden),
)
else:
if isinstance(env.spec.action_space, thBox):
policy = thGaussianMLPPolicy(env_spec=env.spec,
hidden_sizes=tuple(
args.policy_hidden),
min_std=args.min_std)
elif isinstance(env.spec.action_space, thDiscrete):
policy = thCategoricalMLPPolicy(env_spec=env.spec,
hidden_sizes=tuple(
args.policy_hidden),
min_std=args.min_std)
else:
raise NotImplementedError(env.spec.action_space)
if args.control == 'concurrent':
return env, policies
else:
return env, policy
logger.add_text_output(text_log_file)
logger.add_tabular_output(tabular_log_file)
prev_snapshot_dir = logger.get_snapshot_dir()
prev_mode = logger.get_snapshot_mode()
logger.set_snapshot_dir(log_dir)
logger.set_snapshot_mode("gaplast")
logger.set_snapshot_gap(1000)
logger.set_log_tabular_only(False)
logger.push_prefix("[%s] " % "FixMapStartState")
from Algo import parallel_sampler
parallel_sampler.initialize(n_parallel=1)
parallel_sampler.set_seed(0)
policy = CategoricalGRUPolicy(
env_spec=env.spec,
name="gru",
)
baseline = LinearFeatureBaseline(env_spec=env.spec)
with tf.Session() as sess:
writer = tf.summary.FileWriter(logdir=log_dir, )
algo = VPG_t(
env=env,
policy=policy,
baseline=baseline,
batch_size=2048, #2*env._wrapped_env.params['traj_limit'],
max_path_length=env._wrapped_env.params['traj_limit'],
n_itr=10000,
discount=0.95,
def parse_env_args(self, env, args):
if isinstance(args, dict):
args = to_named_tuple(args)
# Multi-agent wrapper
env = RLLabEnv(env, ma_mode=args.control)
env = MATfEnv(env)
# Policy
if args.recurrent:
if args.feature_net:
feature_network = MLP(
name='feature_net',
input_shape=(env.spec.observation_space.flat_dim +
env.spec.action_space.flat_dim, ),
output_dim=args.feature_output,
hidden_sizes=tuple(args.feature_hidden),
hidden_nonlinearity=tf.nn.tanh,
output_nonlinearity=None)
elif args.conv:
strides = tuple(args.conv_strides)
chans = tuple(args.conv_channels)
filts = tuple(args.conv_filters)
assert len(strides) == len(chans) == len(
filts), "strides, chans and filts not equal"
# only discrete actions supported, should be straightforward to extend to continuous
assert isinstance(
env.spec.action_space,
Discrete), "Only discrete action spaces support conv"
feature_network = ConvNetwork(
name='feature_net',
input_shape=env.spec.observation_space.shape,
output_dim=args.feature_output,
conv_filters=chans,
conv_filter_sizes=filts,
conv_strides=strides,
conv_pads=('VALID', ) * len(chans),
hidden_sizes=tuple(args.feature_hidden),
hidden_nonlinearity=tf.nn.relu,
output_nonlinearity=None)
else:
feature_network = None
if args.recurrent == 'gru':
if isinstance(env.spec.action_space, Box):
if args.control == 'concurrent':
policies = [
GaussianGRUPolicy(env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(
args.policy_hidden[0]),
name='policy_{}'.format(agid))
for agid in range(len(env.agents))
]
policy = GaussianGRUPolicy(env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(
args.policy_hidden[0]),
name='policy')
elif isinstance(env.spec.action_space, Discrete):
if args.control == 'concurrent':
policies = [
CategoricalGRUPolicy(
env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(args.policy_hidden[0]),
name='policy_{}'.format(agid),
state_include_action=False
if args.conv else True)
for agid in range(len(env.agents))
]
q_network = CategoricalGRUPolicy(
env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(args.policy_hidden[0]),
name='q_network',
state_include_action=False if args.conv else True)
target_q_network = CategoricalGRUPolicy(
env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(args.policy_hidden[0]),
name='target_q_network',
state_include_action=False if args.conv else True)
policy = {
'q_network': q_network,
'target_q_network': target_q_network
}
else:
raise NotImplementedError(env.spec.observation_space)
elif args.recurrent == 'lstm':
if isinstance(env.spec.action_space, Box):
if args.control == 'concurrent':
policies = [
GaussianLSTMPolicy(env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(
args.policy_hidden),
name='policy_{}'.format(agid))
for agid in range(len(env.agents))
]
policy = GaussianLSTMPolicy(
env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(args.policy_hidden),
name='policy')
elif isinstance(env.spec.action_space, Discrete):
if args.control == 'concurrent':
policies = [
CategoricalLSTMPolicy(
env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(args.policy_hidden),
name='policy_{}'.format(agid))
for agid in range(len(env.agents))
]
q_network = CategoricalLSTMPolicy(
env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(args.policy_hidden),
name='q_network')
target_q_network = CategoricalLSTMPolicy(
env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(args.policy_hidden),
name='target_q_network')
policy = {
'q_network': q_network,
'target_q_network': target_q_network
}
else:
raise NotImplementedError(env.spec.action_space)
else:
raise NotImplementedError(args.recurrent)
elif args.conv:
strides = tuple(args.conv_strides)
chans = tuple(args.conv_channels)
filts = tuple(args.conv_filters)
assert len(strides) == len(chans) == len(
filts), "strides, chans and filts not equal"
# only discrete actions supported, should be straightforward to extend to continuous
assert isinstance(
env.spec.action_space,
Discrete), "Only discrete action spaces support conv"
feature_network = ConvNetwork(
name='feature_net',
input_shape=env.spec.observation_space.shape,
output_dim=env.spec.action_space.n,
conv_filters=chans,
conv_filter_sizes=filts,
conv_strides=strides,
conv_pads=(args.conv_pads, ) * len(chans),
hidden_sizes=tuple(args.policy_hidden),
hidden_nonlinearity=tf.nn.relu,
output_nonlinearity=tf.nn.softmax,
batch_normalization=args.batch_normalization)
if args.algo == 'dqn':
q_network = CategoricalMLPPolicy(name='q_network',
env_spec=env.spec,
prob_network=feature_network)
target_q_network = CategoricalMLPPolicy(
name='target_q_network',
env_spec=env.spec,
prob_network=feature_network)
policy = {
'q_network': q_network,
'target_q_network': target_q_network
}
else:
policy = CategoricalMLPPolicy(name='policy',
env_spec=env.spec,
prob_network=feature_network)
else:
if env.spec is None:
networks = [
DQNNetwork(i,
env,
target_network_update_freq=self.args.
target_network_update,
discount_factor=self.args.discount,
batch_size=self.args.batch_size,
learning_rate=self.args.qfunc_lr)
for i in range(env.n)
]
policy = networks
elif isinstance(env.spec.action_space, Box):
policy = GaussianMLPPolicy(env_spec=env.spec,
hidden_sizes=tuple(
args.policy_hidden),
min_std=args.min_std,
name='policy')
elif isinstance(env.spec.action_space, Discrete):
policy = CategoricalMLPPolicy(env_spec=env.spec,
hidden_sizes=tuple(
args.policy_hidden),
name='policy')
else:
raise NotImplementedError(env.spec.action_space)
return env, policy
def main():
now = datetime.datetime.now(dateutil.tz.tzlocal())
rand_id = str(uuid.uuid4())[:5]
timestamp = now.strftime('%Y_%m_%d_%H_%M_%S_%f_%Z')
default_exp_name = 'experiment_%s_%s' % (timestamp, rand_id)
parser = argparse.ArgumentParser()
parser.add_argument('--exp_name',
type=str,
default=default_exp_name,
help='Name of the experiment.')
parser.add_argument('--discount', type=float, default=0.99)
parser.add_argument('--gae_lambda', type=float, default=1.0)
parser.add_argument('--reward_scale', type=float, default=1.0)
parser.add_argument('--n_iter', type=int, default=250)
parser.add_argument('--sampler_workers', type=int, default=1)
parser.add_argument('--max_traj_len', type=int, default=250)
parser.add_argument('--update_curriculum',
action='store_true',
default=False)
parser.add_argument('--n_timesteps', type=int, default=8000)
parser.add_argument('--control', type=str, default='centralized')
parser.add_argument('--rectangle', type=str, default='10,10')
parser.add_argument('--map_type', type=str, default='rectangle')
parser.add_argument('--n_evaders', type=int, default=5)
parser.add_argument('--n_pursuers', type=int, default=2)
parser.add_argument('--obs_range', type=int, default=3)
parser.add_argument('--n_catch', type=int, default=2)
parser.add_argument('--urgency', type=float, default=0.0)
parser.add_argument('--pursuit', dest='train_pursuit', action='store_true')
parser.add_argument('--evade', dest='train_pursuit', action='store_false')
parser.set_defaults(train_pursuit=True)
parser.add_argument('--surround', action='store_true', default=False)
parser.add_argument('--constraint_window', type=float, default=1.0)
parser.add_argument('--sample_maps', action='store_true', default=False)
parser.add_argument('--map_file', type=str, default='../maps/map_pool.npy')
parser.add_argument('--flatten', action='store_true', default=False)
parser.add_argument('--reward_mech', type=str, default='global')
parser.add_argument('--catchr', type=float, default=0.1)
parser.add_argument('--term_pursuit', type=float, default=5.0)
parser.add_argument('--recurrent', type=str, default=None)
parser.add_argument('--policy_hidden_sizes', type=str, default='128,128')
parser.add_argument('--baselin_hidden_sizes', type=str, default='128,128')
parser.add_argument('--baseline_type', type=str, default='linear')
parser.add_argument('--conv', action='store_true', default=False)
parser.add_argument('--max_kl', type=float, default=0.01)
parser.add_argument('--checkpoint', type=str, default=None)
parser.add_argument('--log_dir', type=str, required=False)
parser.add_argument('--tabular_log_file',
type=str,
default='progress.csv',
help='Name of the tabular log file (in csv).')
parser.add_argument('--text_log_file',
type=str,
default='debug.log',
help='Name of the text log file (in pure text).')
parser.add_argument('--params_log_file',
type=str,
default='params.json',
help='Name of the parameter log file (in json).')
parser.add_argument('--seed', type=int, help='Random seed for numpy')
parser.add_argument('--args_data',
type=str,
help='Pickled data for stub objects')
parser.add_argument('--snapshot_mode',
type=str,
default='all',
help='Mode to save the snapshot. Can be either "all" '
'(all iterations will be saved), "last" (only '
'the last iteration will be saved), or "none" '
'(do not save snapshots)')
parser.add_argument(
'--log_tabular_only',
type=ast.literal_eval,
default=False,
help=
'Whether to only print the tabular log information (in a horizontal format)'
)
args = parser.parse_args()
parallel_sampler.initialize(n_parallel=args.sampler_workers)
if args.seed is not None:
set_seed(args.seed)
parallel_sampler.set_seed(args.seed)
args.hidden_sizes = tuple(map(int, args.policy_hidden_sizes.split(',')))
if args.checkpoint:
with tf.Session() as sess:
data = joblib.load(args.checkpoint)
policy = data['policy']
env = data['env']
else:
if args.sample_maps:
map_pool = np.load(args.map_file)
else:
if args.map_type == 'rectangle':
env_map = TwoDMaps.rectangle_map(
*map(int, args.rectangle.split(',')))
elif args.map_type == 'complex':
env_map = TwoDMaps.complex_map(
*map(int, args.rectangle.split(',')))
else:
raise NotImplementedError()
map_pool = [env_map]
env = PursuitEvade(map_pool,
n_evaders=args.n_evaders,
n_pursuers=args.n_pursuers,
obs_range=args.obs_range,
n_catch=args.n_catch,
train_pursuit=args.train_pursuit,
urgency_reward=args.urgency,
surround=args.surround,
sample_maps=args.sample_maps,
constraint_window=args.constraint_window,
flatten=args.flatten,
reward_mech=args.reward_mech,
catchr=args.catchr,
term_pursuit=args.term_pursuit)
env = TfEnv(
RLLabEnv(StandardizedEnv(env,
scale_reward=args.reward_scale,
enable_obsnorm=False),
mode=args.control))
if args.recurrent:
if args.conv:
feature_network = ConvNetwork(
name='feature_net',
input_shape=emv.spec.observation_space.shape,
output_dim=5,
conv_filters=(16, 32, 32),
conv_filter_sizes=(3, 3, 3),
conv_strides=(1, 1, 1),
conv_pads=('VALID', 'VALID', 'VALID'),
hidden_sizes=(64, ),
hidden_nonlinearity=tf.nn.relu,
output_nonlinearity=tf.nn.softmax)
else:
feature_network = MLP(
name='feature_net',
input_shape=(env.spec.observation_space.flat_dim +
env.spec.action_space.flat_dim, ),
output_dim=5,
hidden_sizes=(256, 128, 64),
hidden_nonlinearity=tf.nn.tanh,
output_nonlinearity=None)
if args.recurrent == 'gru':
policy = CategoricalGRUPolicy(env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(
args.policy_hidden_sizes),
name='policy')
elif args.recurrent == 'lstm':
policy = CategoricalLSTMPolicy(env_spec=env.spec,
feature_network=feature_network,
hidden_dim=int(
args.policy_hidden_sizes),
name='policy')
elif args.conv:
feature_network = ConvNetwork(
name='feature_net',
input_shape=env.spec.observation_space.shape,
output_dim=5,
conv_filters=(8, 16),
conv_filter_sizes=(3, 3),
conv_strides=(2, 1),
conv_pads=('VALID', 'VALID'),
hidden_sizes=(32, ),
hidden_nonlinearity=tf.nn.relu,
output_nonlinearity=tf.nn.softmax)
policy = CategoricalMLPPolicy(name='policy',
env_spec=env.spec,
prob_network=feature_network)
else:
policy = CategoricalMLPPolicy(name='policy',
env_spec=env.spec,
hidden_sizes=args.hidden_sizes)
if args.baseline_type == 'linear':
baseline = LinearFeatureBaseline(env_spec=env.spec)
else:
baseline = ZeroBaseline(env_spec=env.spec)
# logger
default_log_dir = config.LOG_DIR
if args.log_dir is None:
log_dir = osp.join(default_log_dir, args.exp_name)
else:
log_dir = args.log_dir
tabular_log_file = osp.join(log_dir, args.tabular_log_file)
text_log_file = osp.join(log_dir, args.text_log_file)
params_log_file = osp.join(log_dir, args.params_log_file)
logger.log_parameters_lite(params_log_file, args)
logger.add_text_output(text_log_file)
logger.add_tabular_output(tabular_log_file)
prev_snapshot_dir = logger.get_snapshot_dir()
prev_mode = logger.get_snapshot_mode()
logger.set_snapshot_dir(log_dir)
logger.set_snapshot_mode(args.snapshot_mode)
logger.set_log_tabular_only(args.log_tabular_only)
logger.push_prefix("[%s] " % args.exp_name)
algo = TRPO(
env=env,
policy=policy,
baseline=baseline,
batch_size=args.n_timesteps,
max_path_length=args.max_traj_len,
n_itr=args.n_iter,
discount=args.discount,
gae_lambda=args.gae_lambda,
step_size=args.max_kl,
optimizer=ConjugateGradientOptimizer(hvp_approach=FiniteDifferenceHvp(
base_eps=1e-5)) if args.recurrent else None,
mode=args.control,
)
algo.train()
def run_task(v, num_cpu=8, log_dir="./data", ename=None, **kwargs):
from scipy.stats import bernoulli, uniform, beta
import tensorflow as tf
from assistive_bandits.experiments.l2_rnn_baseline import L2RNNBaseline
from assistive_bandits.experiments.tbptt_optimizer import TBPTTOptimizer
from sandbox.rocky.tf.policies.categorical_gru_policy import CategoricalGRUPolicy
from assistive_bandits.experiments.pposgd_clip_ratio import PPOSGD
if not local_test and force_remote:
import rl_algs.logger as rl_algs_logger
log_dir = rl_algs_logger.get_dir()
if log_dir is not None:
log_dir = osp.join(log_dir, str(v["n_episodes"]))
log_dir = osp.join(log_dir, v["human_policy"])
text_output_file = None if log_dir is None else osp.join(log_dir, "text")
tabular_output_file = None if log_dir is None else osp.join(
log_dir, "train_table.csv")
info_theory_tabular_output = None if log_dir is None else osp.join(
log_dir, "info_table.csv")
rag = uniform_bernoulli_iterator()
bandit = BanditEnv(n_arms=v["n_arms"],
reward_dist=bernoulli,
reward_args_generator=rag,
horizon=v["n_episodes"])
pi_H = human_policy_dict[v["human_policy"]](bandit)
h_wrapper = human_wrapper_dict[v["human_wrapper"]]
env = h_wrapper(bandit, pi_H, penalty=v["intervention_penalty"])
if text_output_file is not None:
logger.add_text_output(text_output_file)
logger.add_tabular_output(tabular_output_file)
logger.log("Training against {}".format(v["human_policy"]))
logger.log("Setting seed to {}".format(v["seed"]))
env.seed(v["seed"])
baseline = L2RNNBaseline(
name="vf",
env_spec=env.spec,
log_loss_before=False,
log_loss_after=False,
hidden_nonlinearity=getattr(tf.nn, v["nonlinearity"]),
weight_normalization=v["weight_normalization"],
layer_normalization=v["layer_normalization"],
state_include_action=False,
hidden_dim=v["hidden_dim"],
optimizer=TBPTTOptimizer(
batch_size=v["opt_batch_size"],
n_steps=v["opt_n_steps"],
n_epochs=v["min_epochs"],
),
batch_size=v["opt_batch_size"],
n_steps=v["opt_n_steps"],
)
policy = CategoricalGRUPolicy(env_spec=env.spec,
hidden_nonlinearity=getattr(
tf.nn, v["nonlinearity"]),
hidden_dim=v["hidden_dim"],
state_include_action=True,
name="policy")
n_itr = 3 if local_test else 100
# logger.log('sampler_args {}'.format(dict(n_envs=max(1, min(int(np.ceil(v["batch_size"] / v["n_episodes"])), 100)))))
# parallel_sampler.initialize(6)
# parallel_sampler.set_seed(v["seed"])
algo = PPOSGD(
env=env,
policy=policy,
baseline=baseline,
batch_size=v["batch_size"],
max_path_length=v["n_episodes"],
# 43.65 env time
sampler_args=dict(n_envs=max(
1, min(int(np.ceil(v["batch_size"] / v["n_episodes"])), 100))),
# 100 threads -> 1:36 to sample 187.275
# 6 threads -> 1:31
# force_batch_sampler=True,
n_itr=n_itr,
step_size=v["mean_kl"],
clip_lr=v["clip_lr"],
log_loss_kl_before=False,
log_loss_kl_after=False,
use_kl_penalty=v["use_kl_penalty"],
min_n_epochs=v["min_epochs"],
entropy_bonus_coeff=v["entropy_bonus_coeff"],
optimizer=TBPTTOptimizer(
batch_size=v["opt_batch_size"],
n_steps=v["opt_n_steps"],
n_epochs=v["min_epochs"],
),
discount=v["discount"],
gae_lambda=v["gae_lambda"],
use_line_search=True
# scope=ename
)
sess = tf.Session()
with sess.as_default():
algo.train(sess)
if text_output_file is not None:
logger.remove_tabular_output(tabular_output_file)
logger.add_tabular_output(info_theory_tabular_output)
# Now gather statistics for t-tests and such!
for human_policy_name, human_policy in human_policy_dict.items():
logger.log("-------------------")
logger.log("Evaluating against {}".format(human_policy.__name__))
logger.log("-------------------")
logger.log("Obtaining Samples...")
test_pi_H = human_policy(bandit)
test_env = h_wrapper(bandit, test_pi_H, penalty=0.)
eval_sampler = VectorizedSampler(algo, n_envs=100)
algo.batch_size = v["num_eval_traj"] * v["n_episodes"]
algo.env = test_env
logger.log("algo.env.pi_H has class: {}".format(
algo.env.pi_H.__class__))
eval_sampler.start_worker()
paths = eval_sampler.obtain_samples(-1)
eval_sampler.shutdown_worker()
rewards = []
H_act_seqs = []
R_act_seqs = []
best_arms = []
optimal_a_seqs = []
for p in paths:
a_Rs = env.action_space.unflatten_n(p['actions'])
obs_R = env.observation_space.unflatten_n(p['observations'])
best_arm = np.argmax(p['env_infos']['arm_means'][0])
H_act_seqs.append(obs_R[:, 1])
R_act_seqs.append(a_Rs)
best_arms.append(best_arm)
optimal_a_seqs.append(
[best_arm for _ in range(v["n_episodes"])])
rewards.append(np.sum(p['rewards']))
#feel free to add more data
logger.log("NumTrajs {}".format(v["num_eval_traj"]))
logger.log("AverageReturn {}".format(np.mean(rewards)))
logger.log("StdReturn {}".format(np.std(rewards)))
logger.log("MaxReturn {}".format(np.max(rewards)))
logger.log("MinReturn {}".format(np.min(rewards)))
optimal_a_H_freqs = _frequency_agreement(H_act_seqs,
optimal_a_seqs)
optimal_a_R_freqs = _frequency_agreement(R_act_seqs,
optimal_a_seqs)
for t in range(v["n_episodes"]):
logger.record_tabular("PolicyExecTime", 0)
logger.record_tabular("EnvExecTime", 0)
logger.record_tabular("ProcessExecTime", 0)
logger.record_tabular("Tested Against", human_policy_name)
logger.record_tabular("t", t)
logger.record_tabular("a_H_agreement", optimal_a_H_freqs[t])
logger.record_tabular("a_R_agreement", optimal_a_R_freqs[t])
H_act_seqs_truncated = [a_Hs[0:t] for a_Hs in H_act_seqs]
R_act_seqs_truncated = [a_Rs[0:t] for a_Rs in R_act_seqs]
h_mutual_info = _mutual_info_seqs(H_act_seqs_truncated,
best_arms, v["n_arms"] + 1)
r_mutual_info = _mutual_info_seqs(R_act_seqs_truncated,
best_arms, v["n_arms"] + 1)
logger.record_tabular("h_mutual_info", h_mutual_info)
logger.record_tabular("r_mutual_info", r_mutual_info)
logger.record_tabular("a_H_opt_freq", optimal_a_H_freqs[t])
logger.record_tabular("a_R_opt_freq", optimal_a_R_freqs[t])
logger.dump_tabular()
test_env = h_wrapper(bandit, human_policy(bandit), penalty=0.)
logger.log("Printing Example Trajectories")
for i in range(v["num_display_traj"]):
observation = test_env.reset()
policy.reset()
logger.log("-- Trajectory {} of {}".format(
i + 1, v["num_display_traj"]))
logger.log("t \t obs \t act \t reward \t act_probs")
for t in range(v["n_episodes"]):
action, act_info = policy.get_action(observation)
new_obs, reward, done, info = test_env.step(action)
logger.log("{} \t {} \t {} \t {} \t {}".format(
t, observation, action, reward, act_info['prob']))
observation = new_obs
if done:
logger.log("Total reward: {}".format(
info["accumulated rewards"]))
break
if text_output_file is not None:
logger.remove_text_output(text_output_file)
logger.remove_tabular_output(info_theory_tabular_output)