ray.rllib.evaluation.sampler.SamplerInput#

class ray.rllib.evaluation.sampler.SamplerInput[source]#

Bases: ray.rllib.offline.input_reader.InputReader

Reads input experiences from an existing sampler.

Methods

__init__()

get_data()

Called by self.next() to return the next batch of data.

get_extra_batches()

Returns list of extra batches since the last call to this method.

get_metrics()

Returns list of episode metrics since the last call to this method.

tf_input_ops([queue_size])

Returns TensorFlow queue ops for reading inputs from this reader.