Dynamically "sources" a provided file

Determine whether we've collected enough data

This function loads the custom algorithms implemented in this repository, and registers them with the tune registry

This function takes a path to a local directory and looks for an `envs` folder, and imports all the available files in there.

This function takes a path to a local directory and looks for a `models` folder, and imports all the available files in there.

Function to register custom preprocessors

Add a step to the current rollout, if we are saving them

Return the action for a batch Returns: action_batch: List of actions for the batch rnn_states: List of RNN states if

Returns model weights. Returns: weights (obj): Serializable copy or view of model weights

Fused compute gradients and apply gradients call. Either this or the combination of compute/apply grads must be implemented by subcla

Runs when an episode is done. Args: worker (RolloutWorker): Reference to the current rollout worker. base_env (BaseEnv

Callback run on the rollout worker before each episode starts. Args: worker (RolloutWorker): Reference to the current rollout work

Runs on each episode step. Args: worker (RolloutWorker): Reference to the current rollout worker. base_env (BaseEnv):

Called immediately after a policy's postprocess_fn is called. You can use this callback to do additional postprocessing for a policy,

Called at the end RolloutWorker.sample(). Args: worker (RolloutWorker): Reference to the current rollout worker. sampl

Called at the end of Trainable.train(). Args: trainer (Trainer): Current trainer instance. result (dict): Dict of resu

Returns the current exploration information of this policy. This information depends on the policy's Exploration object. Ret