garage.tf.samplers.batch_sampler module

Collects samples in parallel using a stateful pool of workers.

class BatchSampler(algo, env, n_envs)

Bases: garage.sampler.sampler_deprecated.BaseSampler

Collects samples in parallel using a stateful pool of workers.

Parameters:

• algo (garage.np.algos.RLAlgorithm) – The algorithm.
• env (gym.Env) – The environment.
• n_envs (int) – Number of environments.

obtain_samples(itr, batch_size=None, whole_paths=True)

Collect samples for the given iteration number.

Parameters:

• itr (int) – Number of iteration.
• batch_size (int) – Number of environment steps in one batch.
• whole_paths (bool) – Whether to use whole path or truncated.

Returns:

A list of paths.

Return type:

list[dict]

shutdown_worker()

Terminate workers if necessary.

start_worker()

Initialize the sampler.

worker_init_tf(g)

Initialize the tf.Session on a worker.

Parameters:g (object) – Global state object.

worker_init_tf_vars(g)

Initialize the policy parameters on a worker.

Parameters:g (object) – Global state object.