garage.tf.samplers
¶
Samplers which run agents that use Tensorflow in environments.
-
class
TFWorkerClassWrapper
(wrapped_class)¶ Acts like a Worker class, but is actually an object.
When called, constructs the wrapped class and wraps it in a
TFWorkerWrapper
.- Parameters
wrapped_class (type) – The class to wrap. Should be a subclass of
Worker
.
-
class
TFWorkerWrapper
¶ Bases:
garage.sampler.Worker
Wrapper around another workers that initializes a TensorFlow Session.
-
worker_init
(self)¶ Initialize a worker.
-
shutdown
(self)¶ Perform shutdown processes for TF.
-
property
agent
(self)¶ Policy: the worker’s agent.
-
property
env
(self)¶ Environment: Worker’s environment.
-
update_agent
(self, agent_update)¶ Update the worker’s agent, using agent_update.
- Parameters
agent_update (object) – An agent update. The exact type of this argument depends on the
Worker
implementation.
-
update_env
(self, env_update)¶ Update the worker’s env, using env_update.
- Parameters
env_update (object) – An environment update. The exact type of this argument depends on the
Worker
implementation.
-
rollout
(self)¶ Sample a single episode of the agent in the environment.
- Returns
- Batch of sampled episodes. May be truncated if
max_episode_length is set.
- Return type
-
start_episode
(self)¶ Begin a new episode.
-
step_episode
(self)¶ Take a single time-step in the current episode.
- Returns
- True iff the episode is done, either due to the environment
indicating termination of due to reaching max_episode_length.
- Return type
-
collect_episode
(self)¶ Collect the current episode, clearing the internal buffer.
- Returns
- Batch of sampled episodes. May be truncated if the
episodes haven’t completed yet.
- Return type
-