unreal.LearningAgentsImitationTrainerTrainingSettings

class unreal.LearningAgentsImitationTrainerTrainingSettings

Bases: StructBase

The configurable settings for the training process.

C++ Source:

  • Plugin: LearningAgents

  • Module: LearningAgentsTraining

  • File: LearningAgentsImitationTrainer.h

Editor Properties: (see get_editor_property/set_editor_property)

  • action_entropy_weight (float): [Read-Write] Weighting used for the entropy bonus. Larger values encourage larger action noise and therefore greater exploration but can make actions very noisy.

  • action_regularization_weight (float): [Read-Write] Weight used to regularize actions. Larger values will encourage smaller actions but too large will cause actions to become always zero.

  • batch_count_per_evaluation (int32): [Read-Only] How many batches to perform evaluation on? Randomly chosen each evaluation. Currently not used in Python

  • batch_size (uint32): [Read-Write] Batch size to use for training. Smaller values tend to produce better results at the cost of slowing down training. Large batch sizes are much more computationally efficient when training on the GPU.

  • device (LearningAgentsTrainingDevice): [Read-Write] The device to train on.

  • evaluation_frequency (int32): [Read-Only] How many training iteration loops between an evaluation run. Currently not used in Python

  • iterations_per_snapshot (int32): [Read-Write] If bSaveSnapshots is true, the snapshots will be saved at an interval defined by the specified number of iterations.

  • learning_rate (float): [Read-Write] Learning rate of the policy network. Typical values are between 0.001 and 0.0001.

  • learning_rate_decay (float): [Read-Write] Amount by which to multiply the learning rate every time it decays.

  • learning_rate_decay_step_size (int32): [Read-Write] The number of iterations to train before updating the learning rate.

  • m_lflow_tracking_uri (str): [Read-Write] The URI of the MLflow Tracking Server to log to.

  • number_of_iterations (int32): [Read-Write] The number of iterations to run before ending training.

  • observation_noise_scale (float): [Read-Write] A multiplicative scaling factor that controls the observation noise that increases the perturbations added to observations.

  • random_seed (int32): [Read-Write] The seed used for any random sampling the trainer will perform, e.g. for weight initialization.

  • run_evaluation (bool): [Read-Only] Should evaluation run during the training process. Currently not used in Python

  • save_snapshots (bool): [Read-Write] If true, snapshots of the trained networks will be emitted to the intermediate directory.

  • train_eval_dataset_split (float): [Read-Only] How much data should be used for evaluation. Currently not used in Python

  • use_m_lflow (bool): [Read-Write] If true, MLflow will be used for experiment tracking.

  • use_tensorboard (bool): [Read-Write] If true, TensorBoard logs will be emitted to the intermediate directory.

    TensorBoard will only work if it is installed in Unreal Engine’s python environment. This can be done by enabling the “Tensorboard” plugin in your project.

  • weight_decay (float): [Read-Write] Amount of weight decay to apply to the network. Larger values encourage network weights to be smaller but too large a value can cause the network weights to collapse to all zeros.

  • window (uint32): [Read-Write] The number of consecutive steps of observations and actions over which to train the policy. Increasing this value will encourage the policy to use its memory effectively. Too large and training can become unstable. Given we don’t know the memory state during imitation learning it is better this is slightly larger than when we are doing reinforcement learning.