action = act(np.array(obs)[None], update_eps=update_eps, **kwargs)[0]
reset = False
new_obs, rew, done, info = env.step(action)
replay_buffer.add(obs, action, rew, new_obs, float(done))
obs = new_obs
if done:
num_iters_since_reset = 0
obs = env.reset()
reset = True
if (num_iters > max(5 * args.batch_size, args.replay_buffer_size // 200) and
num_iters % args.learning_freq == 0):
# Sample a bunch of transitions from replay buffer
if args.prioritized:
experience = replay_buffer.sample(args.batch_size, beta=beta_schedule.value(num_iters))
(obses_t, actions, rewards, obses_tp1, dones, weights, batch_idxes) = experience
else:
obses_t, actions, rewards, obses_tp1, dones = replay_buffer.sample(args.batch_size)
weights = np.ones_like(rewards)
# Minimize the error in Bellman's equation and compute TD-error
td_errors = train(obses_t, actions, rewards, obses_tp1, dones, weights)
# Update the priorities in the replay buffer
if args.prioritized:
new_priorities = np.abs(td_errors) + args.prioritized_eps
replay_buffer.update_priorities(batch_idxes, new_priorities)
# Update target network.
if num_iters % args.target_update_freq == 0:
update_target()
if start_time is not None:
def learn(
env,
p_dist_func,
lr=5e-4,
eps=0.0003125,
max_timesteps=100000,
buffer_size=50000,
exp_t1=1e6,
exp_p1=0.1,
exp_t2=25e6,
exp_p2=0.01,
# exploration_fraction=0.1,
# exploration_final_eps=0.02,
train_freq=1,
batch_size=32,
print_freq=1,
checkpoint_freq=10000,
learning_starts=1000,
gamma=0.95,
target_network_update_freq=500,
prioritized_replay=False,
prioritized_replay_alpha=0.6,
prioritized_replay_beta0=0.4,
prioritized_replay_beta_iters=None,
prioritized_replay_eps=1e-6,
num_cpu=16,
param_noise=False,
callback=None,
dist_params=None,
n_action=None,
action_map=None):
"""Train a distdeepq model.
Parameters
-------
env: gym.Env
environment to train on
p_dist_func: (tf.Variable, int, str, bool) -> tf.Variable
the model that takes the following inputs:
observation_in: object
the output of observation placeholder
num_actions: int
number of actions
scope: str
reuse: bool
should be passed to outer variable scope
and returns a tensor of shape (batch_size, num_actions) with values of every action.
lr: float
learning rate for adam optimizer
max_timesteps: int
number of env steps to optimizer for
buffer_size: int
size of the replay buffer
exploration_fraction: float
fraction of entire training period over which the exploration rate is annealed
exploration_final_eps: float
final value of random action probability
train_freq: int
update the model every `train_freq` steps.
set to None to disable printing
batch_size: int
size of a batched sampled from replay buffer for training
print_freq: int
how often to print out training progress
set to None to disable printing
checkpoint_freq: int
how often to save the model. This is so that the best version is restored
at the end of the training. If you do not wish to restore the best version at
the end of the training set this variable to None.
learning_starts: int
how many steps of the model to collect transitions for before learning starts
gamma: float
discount factor
target_network_update_freq: int
update the target network every `target_network_update_freq` steps.
prioritized_replay: True
if True prioritized replay buffer will be used.
prioritized_replay_alpha: float
alpha parameter for prioritized replay buffer
prioritized_replay_beta0: float
initial value of beta for prioritized replay buffer
prioritized_replay_beta_iters: int
number of iterations over which beta will be annealed from initial value
to 1.0. If set to None equals to max_timesteps.
prioritized_replay_eps: float
epsilon to add to the TD errors when updating priorities.
num_cpu: int
number of cpus to use for training
callback: (locals, globals) -> None
function called at every steps with state of the algorithm.
If callback returns true training stops.
Returns
-------
act: ActWrapper
Wrapper over act function. Adds ability to save it and load it.
See header of baselines/distdeepq/categorical.py for details on the act function.
"""
# Create all the functions necessary to train the model
sess = make_session(num_cpu=num_cpu)
sess.__enter__()
# logger.configure()
def make_obs_ph(name):
return U.BatchInput(env.observation_space.shape, name=name)
if dist_params is None:
raise ValueError('dist_params is required')
# z, dz = build_z(**dist_params)
act, train, update_target, debug = distdeepq.build_train(
make_obs_ph=make_obs_ph,
p_dist_func=p_dist_func,
# num_actions=env.action_space.n,
n_action=n_action,
optimizer=tf.train.AdamOptimizer(learning_rate=lr, epsilon=eps),
gamma=gamma,
grad_norm_clipping=10,
param_noise=param_noise,
dist_params=dist_params)
act_params = {
'make_obs_ph': make_obs_ph,
'p_dist_func': p_dist_func,
'num_actions': n_action,
'dist_params': dist_params
}
# Create the replay buffer
if prioritized_replay:
replay_buffer = PrioritizedReplayBuffer(buffer_size,
alpha=prioritized_replay_alpha)
if prioritized_replay_beta_iters is None:
prioritized_replay_beta_iters = max_timesteps
beta_schedule = LinearSchedule(prioritized_replay_beta_iters,
initial_p=prioritized_replay_beta0,
final_p=1.0)
else:
replay_buffer = ReplayBuffer(buffer_size)
beta_schedule = None
# Create the schedule for exploration starting from 1.
#exploration = LinearSchedule(schedule_timesteps=int(exploration_fraction * max_timesteps),
# initial_p=1.0,
# final_p=exploration_final_eps)
# exploration = PiecewiseSchedule([(0, 1.0),(max_timesteps/25, 0.1),
# (max_timesteps, 0.01)], outside_value=0.01)
exploration = PiecewiseSchedule([(0, 1.0), (exp_t1, exp_p1),
(exp_t2, exp_p2)],
outside_value=exp_p2)
# Initialize the parameters and copy them to the target network.
U.initialize()
update_target()
episode_rewards = [0.0]
saved_mean_reward = None
obs = env.reset()
reset = True
with tempfile.TemporaryDirectory() as td:
model_saved = False
model_file = os.path.join(td, "model")
for t in range(max_timesteps):
if callback is not None:
if callback(locals(), globals()):
break
# Take action and update exploration to the newest value
kwargs = {}
if not param_noise:
update_eps = exploration.value(t)
update_param_noise_threshold = 0.
else:
update_eps = 0.
# Compute the threshold such that the KL divergence between perturbed and non-perturbed
# policy is comparable to eps-greedy exploration with eps = exploration.value(t).
# See Appendix C.1 in Parameter Space Noise for Exploration, Plappert et al., 2017
# for detailed explanation.
update_param_noise_threshold = -np.log(1. - exploration.value(
t) + exploration.value(t) / float(env.action_space.n))
kwargs['reset'] = reset
kwargs[
'update_param_noise_threshold'] = update_param_noise_threshold
kwargs['update_param_noise_scale'] = True
action = act(np.array(obs)[None], update_eps=update_eps,
**kwargs)[0]
reset = False
action_val = action_map[action]
new_obs, rew, done, _ = env.step(action_val)
# env.render()
# rew = rew-1 for proposed loss with new metric
# rew = rew-1
# Store transition in the replay buffer.
replay_buffer.add(obs, action, rew, new_obs, float(done))
obs = new_obs
episode_rewards[-1] += rew
if done:
obs = env.reset()
episode_rewards.append(0.0)
reset = True
if t > learning_starts and t % train_freq == 0:
# Minimize the error in Bellman's equation on a batch sampled from replay buffer.
if prioritized_replay:
experience = replay_buffer.sample(
batch_size, beta=beta_schedule.value(t))
(obses_t, actions, rewards, obses_tp1, dones, weights,
batch_idxes) = experience
else:
obses_t, actions, rewards, obses_tp1, dones = replay_buffer.sample(
batch_size)
weights, batch_idxes = np.ones_like(rewards), None
errors = train(obses_t, actions, rewards, obses_tp1, dones,
weights)
if prioritized_replay:
new_priorities = np.abs(errors) + prioritized_replay_eps
replay_buffer.update_priorities(batch_idxes,
new_priorities)
if t > learning_starts and t % target_network_update_freq == 0:
# Update target network periodically.
update_target()
mean_100ep_reward = round(np.mean(episode_rewards[-101:-1]), 1)
num_episodes = len(episode_rewards)
if done and print_freq is not None and len(
episode_rewards) % print_freq == 0:
logger.record_tabular("steps", t)
logger.record_tabular("episodes", num_episodes)
logger.record_tabular("mean 100 episode reward",
mean_100ep_reward)
logger.record_tabular("% time spent exploring",
int(100 * exploration.value(t)))
# debug['pi'] = tf.Print(debug['pi'], [debug['pi'], "target pi"])
# tf.Print(debug['mu'], [debug['mu'], "target mu"])
# tf.Print(debug['sigma'], [debug['sigma'], "target sigma"])
logger.dump_tabular()
if (checkpoint_freq is not None and t > learning_starts
and num_episodes > 100 and t % checkpoint_freq == 0):
if saved_mean_reward is None or mean_100ep_reward > saved_mean_reward:
if print_freq is not None:
logger.log(
"Saving model due to mean reward increase: {} -> {}"
.format(saved_mean_reward, mean_100ep_reward))
U.save_state(model_file)
model_saved = True
saved_mean_reward = mean_100ep_reward
if model_saved:
if print_freq is not None:
logger.log("Restored model with mean reward: {}".format(
saved_mean_reward))
U.load_state(model_file)
return ActWrapper(act, act_params)