ray.rllib.policy.sample_batch.MultiAgentBatch.env_steps#

MultiAgentBatch.env_steps() → int[source]#

The number of env steps (there are >= 1 agent steps per env step).

Returns:: The number of environment steps contained in this batch.