ast_toolbox.policies.go_explore_policy module

class ast_toolbox.policies.go_explore_policy.GoExplorePolicy(env_spec, name='GoExplorePolicy')

Bases: garage.tf.policies.base.StochasticPolicy

A stochastic policy for Go-Explore that takes actions uniformally at random.

Parameters:

• env_spec (garage.envs.EnvSpec) – Environment specification.
• name (str) – Name for the tensors.

dist_info(obs, state_infos)

Distribution info.

Return the distribution information about the actions.

Parameters:

• obs (array_like) – Observation values.
• state_infos (dict) – A dictionary whose values should contain information about the state of the policy at the time it received the observation.

dist_info_sym(obs_var, state_info_vars, name='dist_info_sym')

Symbolic graph of the distribution.

Return the symbolic distribution information about the actions.

Parameters:

• obs_var (tf.Tensor) – Symbolic variable for observations.
• state_infos (dict) – A dictionary whose values should contain information about the state of the policy at the time it received the observation.
• name (str) – Name of the symbolic graph.

get_action(observation)

Get action sampled from the policy.

Parameters:observation (array_like) – Observation from the environment. Returns:array_like – Action sampled from the policy.

get_actions(observations)

Get actions sampled from the policy.

Parameters:observations (list[array_like]) – Observations from the environment. Returns:array_like – Actions sampled from the policy.

log_diagnostics(paths)

Log extra information per iteration based on the collected paths.

reset(dones=None)

Reset the policy.

If dones is None, it will be by default np.array([True]) which implies the policy will not be “vectorized”, i.e. number of parallel environments for training data sampling = 1.

Parameters:dones (array_like) – Bool that indicates terminal state(s).

terminate()

Clean up operation.

distribution

Distribution.

Returns:Distribution.

vectorized

Indicates whether the policy is vectorized. If True, it should implement get_actions(), and support resetting with multiple simultaneous states.